Surgical aspects in treatment of inflammatory bowel disease · Chapter 7 The Management of Anastomotic Pouch Sinus After ileal-pouch-anal anastomosis. Diseases of the Colon & Rectum,

Surgical Aspects in the Treatment of Inflammatory Bowel Diseaseusama ahmed ali

Surgical aspects in the treatment of

inflammatory bowel disease

Usama Ahmed Ali

Surgical aspects in the treatment of inflammatory bowel disease

PhD thesis, Utrecht University, The Netherlands

© U. Ahmed Ali, Utrecht, The Netherlands, 2019

Copyright of the published articles has been transferred to the respective publishers. No

part of this thesis may be reproduced, stored or transmitted, in any form or by any means,

without prior permission of the author.

The research in this thesis was financially supported by a personal grant from the Utrecht

University, The Netherlands and the Research Fellowship Program of the Cleveland Clinic,

Cleveland, United States of America.

Publication of this thesis was financially supported by Chirurgische Fonds UMC Utrecht,

Applied Medical and Erbe Nederland B.V.

Cover illustration: Rosalia Simunovic, RosalisArts, webshop: www.etsy.com/shop/RosalisArt

Lay-out: Optima Grafische Communicatie, Rotterdam, The Netherlands

Printed by: Optima Grafische Communicatie, Rotterdam, The Netherlands

Surgical aspects in treatment of

inflammatory bowel disease

Chirurgische aspecten in de behandeling van inflammatoire darmziekte

(met een samenvatting in het Nederlands)

Proefschrift

ter verkrijging van de graad van doctor aan de Universiteit Utrecht

op gezag van de rector magnificus, prof.dr. H.R.B.M. Kummeling,

ingevolge het besluit van het college voor promoties in het openbaar

te verdedigen op vrijdag 29 maart 2019 des middags te 2.30 uur

door

Usama Ahmed Ali

geboren op 19 april 1983 te Tripoli, Libië

Promotoren: Prof. dr. R. van Hillegersberg

Prof. dr. C.J. van Laarhoven

Prof. dr. R.P. Kiran

For my lovely wife – Amasi – and adored son – Shahien

Table of conTenTs

Chapter 1 General introduction and outline of the thesis 11

PART I CURRENT OUTCOMES IN COLORECTAL SURGERY

Chapter 2 Update of complications and functional outcome of the ileal

pouch-anal anastomosis: overview of evidence and meta-analysis

of 96 observational studies.

International Journal of Colorectal Disease, 2012.

21

Chapter 3 Actual versus estimated length of stay after colorectal surgery -

which factors influence a deviation?

The American Journal of Surgery, 2013.

43

Chapter 4 Risk and Location of Cancer in Patients with Preoperative Colitis-

Associated Dysplasia Undergoing Proctocolectomy.

Annals of Surgery, 2014.

59

PART II OPTIMIZING SURGICAL TREATMENT FOR INFLAMMATORY

BOWEL DISEASE

Chapter 5 Open versus laparoscopic (assisted) ileal pouch-anal anastomosis

for ulcerative colitis and familial adenomatous polyposis.

The Cochrane Database of Systematic Reviews, 2009.

81

Chapter 6 Surgical treatment of Ulcerative colitis: Ileorectal vs. Ileal-pouch

anal anastomosis.

World Journal of Gastroenterology, 2014.

113

Chapter 7 The Management of Anastomotic Pouch Sinus After ileal-pouch-

anal anastomosis.

Diseases of the Colon & Rectum, 2012.

131

Chapter 8 Impact of preoperative immunosuppressive agents on

postoperative outcomes in Crohn’s disease.


147

Chapter 9 Ileal pouch-anal anastomosis with close rectal dissection using

automated vessel sealers for ulcerative colitis: a promising

alternative.

Digestive Surgery, 2011.

169

PART III SUMMARY AND CONCLUSIONS

Chapter 10 Summary 185

Chapter 11 Gained insights and future perspectives 195

Chapter 12 Nederlandse samenvatting (Dutch summary) 205

APPENDICES

Review committee 217

Acknowledgement 218

List of publications 221

Curriculum Vitae 227

1 General introduction

and outline of the thesis

Chapter 1 13

1General inTroducTion

Inflammatory bowel disease (IBD) includes a group of chronic, relapsing, inflammatory

disorders of the gastrointestinal tract. It mainly includes ulcerative colitis (UC) and Crohn’s

disease (CD), with indeterminate colitis as an intermediate form in which there is difficulty

in distinguishing between these two conditions. Although the etiology and pathogenesis

of IBD are still poorly understood, it is widely accepted that several genetic and environ-

mental factors are involved. Several dozen genes have been identified by genome-wide

association studies to increase the susceptibility to IBD.1 Similarly, several environmental

risk factors are associated with IBD such as smoking, composition of the intestinal flora,

past history of appendectomy, certain medication, diet and stress.2

UC and CD differ with regards to various characteristics.3 UC usually starts in the rectum

and can extend in a continuous pattern to involve the complete colon. It is limited to the

mucosa and sometimes the submucosa. On the other hand, CD often starts at the terminal

ileum and extends in an irregular fashion with so called ‘skip’ lesions. The inflammation is

often transmural and can involve any part of the gastrointestinal tract. IBD commonly af-

fects people at an early age, with around 25% of patients diagnosed in the first 2 decades

(mostly between 13 to 18 years).4,5 The incidence and prevalence of IBD are increasing

worldwide, with the highest reported incidence in Europe of 24.3 per 100,000 person-

years for UC and 12.7 per 100,000 person-years for CD, and a respective prevalence of

505 per 100,000 and 322 per 100,000.6 Both UC and CD are lifelong diseases with a

significant impact on quality of life, personal burden and costs for patients and society.7,8

Therapy in IBD is directed at resolving symptoms as effectively as possible and subsequently

maintaining symptom-free periods. Lifelong medical treatment is required and currently

forms the corner-stone of management, especially with the development of several bio-

logic therapies that have proven to be effective in the majority of patients. The use of

these agents has been shown to facilitate remission, improve quality of life, reduce steroid

dependency and delay surgery.9-11

For UC, indications for surgical treatment include acute pancolitis, failure of medical treat-

ment, patient’s preference and the presence of dysplasia or neoplasia.12 The gold standard

for surgical treatment is a total proctocolectomy in which the complete colon and rectum

are removed.13 This surgery provides relief of symptoms and eliminates the risk of develop-

ing colorectal cancer. To re-establish continuity of the intestinal tract usually a restorative

ileal pouch-anal anastomosis (IPAA) procedure is performed. In this procedure, the terminal

ileum is used to form a reservoir to replace the removed colon, and is then anastomosed

14 Chapter 1

to the anus.13 The IPAA is currently the surgical treatment of choice for UC due to its good

functional results and high patient satisfaction.13,14

In CD, however, surgery is not curative, since the disease can (and does) recur throughout

the intestinal tract. This combined with the good clinical outcome of newer immuno-

modulatory drugs, has limited the indications of surgery to those patients with symptoms

or complications that are intractability to medical management.15 The surgical goal in CD

is, therefore, to limit the resection of bowel as much as possible. Limited small-bowel

resections, especially ileocecal resections, are the most common procedures in CD.

Surgery for IBD has certain challenges, including a relatively young patient population,

diseases that are chronic and recurring in nature and a continuously changing balance

between medical and surgical interventions resulting in a selection of sicker patients to be

operated. In this thesis we will examine certain aspects of the surgical management of IBD

as outlined below.

ouTline of The Thesis

In the first part of the thesis we review the current state of important outcomes in IBD sur-

gery. First, in Chapter 2 we provide a comprehensive update of the postoperative morbidity

and functional outcome after restorative proctocolectomy by means of an IPAA procedure.

This procedure has been performed for several decades now; a period that encompasses

many medical and technical advances. We examine whether these developments were able

to improve the outcome of this procedure by performing a systematic review summarizing

outcomes after the year 2000 and compare these to previous outcomes. Subsequently, in

Chapter 3 we focus on the length of hospital stay in colorectal surgery and its impact on

cost of care. In this era of ever-increasing costs it is prudent to examine how costs can be

reduced most effectively. Therefore, we have studied factors affecting the difference be-

tween the actual length of hospital stay, which represents the costs made by the hospital,

versus the estimated length of stay, representing the reimbursed amount received from

insurance companies. By selectively targeting modifiable factors identified by such analysis,

an optimized effort can be undertaken to reduce the financial burden for hospitals. In

Chapter 4, we examine the topic of oncological outcome after proctocolecotmy in colitis

patients with dysplasia. We evaluate the correlation between the preoperative findings

at colonoscopic surveillance (in terms of grade, location and appearance of dysplasia)

and those of the pathohistological findings after proctocolectomy. By doing so we aim at

improving our ability to predict the presence and location of cancer in colitis patients with

dysplasia, and facilitate recommendations regarding the necessity and type of surgery.

Chapter 1 15

1In the second part, we examine several aspects of the surgical treatment of inflammatory

bowel disease. First, in Chapter 5 we compare the open versus laparoscopic approach for

performing a restorative proctocolectomy with IPAA. With the advancements of minimal-

invasive surgery the laparoscopic approach for this demanding operation has been increas-

ingly popular. In this systematic Cochrane review we evaluate the presumed benefits of the

laparoscopic approach in a direct comparison. Next, we examine a potential alternative to

the IPAA procedure. While IPAA allows for removal of all diseases tissue and offers good

long-term results, it remains a large and complex procedure. Certain patients might benefit

from a less invasive treatment such as a total colectomy with ileorectal anastomosis, with

less postoperative complications as a major presumed advantage. In Chapter 6, we review

the literature and examine the benefits and harms of both approaches. Subsequently, we

focus on the optimal management of an anastomotic pouch sinus, an uncommon but

serious postoperative complication after IPAA. In Chapter 7 we retrospectively examine

a large prospectively collected database, identify all patients with this complication, and

map the received treatment and its outcome. Based on this, we provide recommenda-

tions on how best to manage this challenging complication. In Chapter 8, we evaluate

the impact of preoperative immunosuppressive agents on outcomes of surgery in CD.

While immunosuppressive agents are essential in the management of this disease, their

safety in the preoperative period is controversial. In this extensive systematic review and

meta-analysis, we examine the effect of various common immunosuppressive agents on

postoperative complications, especially infectious complications. Finally, in Chapter 9,

we examine whether a change in the way the IPAA procedure is performed might affect

postoperative outcome. In a pilot study, the close rectal dissection technique is attempted

instead of the more common total mesorectal excision approach. Safety, feasibility and

potential benefits are evaluated.

16 Chapter 1

references

1. Barrett JC, Hansoul S, Nicolae DL, Cho

JH, Duerr RH, Rioux JD, Brant SR, et al.

Genome-wide association defines more

than 30 distinct susceptibility loci for

crohn’s disease. Nature genetics. 2008; 40:

955-62.

2. Ye Y, Pang Z, Chen W, Ju S, Zhou C. The

epidemiology and risk factors of inflamma-

tory bowel disease. International journal of

clinical and experimental medicine. 2015; 8:

22529-42.

3. Price AB, Morson BC. Inflammatory bowel

disease: The surgical pathology of crohn’s

disease and ulcerative colitis. Human pa-

thology. 1975; 6: 7-29.

4. Benchimol EI, Fortinsky KJ, Gozdyra P, Van

den Heuvel M, Van Limbergen J, Griffiths

AM. Epidemiology of pediatric inflamma-

tory bowel disease: A systematic review of

international trends. Inflammatory bowel

diseases. 2011; 17: 423-39.

5. Mir-Madjlessi SH, Michener WM, Farmer

RG. Course and prognosis of idiopathic

ulcerative proctosigmoiditis in young pa-

tients. Journal of pediatric gastroenterology

and nutrition. 1986; 5: 571-5.

6. Molodecky NA, Soon IS, Rabi DM, Ghali

WA, Ferris M, Chernoff G, Benchimol EI, et

al. Increasing incidence and prevalence of

the inflammatory bowel diseases with time,

based on systematic review. Gastroenterol-

ogy. 2012; 142: 46-54.e42; quiz e30.

7. Kawalec P. Indirect costs of inflammatory

bowel diseases: Crohn’s disease and ulcer-

ative colitis. A systematic review. Archives of

medical science: AMS. 2016; 12: 295-302.

8. Floyd DN, Langham S, Severac HC,

Levesque BG. The economic and quality-

of-life burden of crohn’s disease in europe

and the united states, 2000 to 2013: A

systematic review. Digestive diseases and

sciences. 2015; 60: 299-312.

9. D’Haens GR. Top-down therapy for ibd:

Rationale and requisite evidence. Nature

reviews. Gastroenterology & hepatology.

2010; 7: 86-92.

10. Ford AC, Sandborn WJ, Khan KJ, Hanauer

SB, Talley NJ, Moayyedi P. Efficacy of bio-

logical therapies in inflammatory bowel dis-

ease: Systematic review and meta-analysis.

The American journal of gastroenterology.

2011; 106: 644-59, quiz 60.

11. Archer R, Tappenden P, Ren S, Martyn-St

James M, Harvey R, Basarir H, Stevens J,

et al. Infliximab, adalimumab and golim-

umab for treating moderately to severely

active ulcerative colitis after the failure of

conventional therapy (including a review

of ta140 and ta262): Clinical effectiveness

systematic review and economic model.

Health technology assessment (Winchester,

England). 2016; 20: 1-326.

12. Cohen JL, Strong SA, Hyman NH, Buie WD,

Dunn GD, Ko CY, Fleshner PR, et al. Practice

parameters for the surgical treatment of

ulcerative colitis. Diseases of the colon and

rectum. 2005; 48: 1997-2009.

13. Fazio VW, Ziv Y, Church JM, Oakley JR,

Lavery IC, Milsom JW, Schroeder TK. Ileal

pouch-anal anastomoses complications and

function in 1005 patients. Annals of sur-

gery. 1995; 222: 120-7.

14. Hueting WE, Buskens E, van der Tweel I,

Gooszen HG, van Laarhoven CJ. Results

and complications after ileal pouch anal

anastomosis: A meta-analysis of 43 obser-

vational studies comprising 9,317 patients.

Digestive surgery. 2005; 22: 69-79.

15. Seifarth C, Kreis ME, Grone J. Indications

and specific surgical techniques in crohn’s

disease. Viszeralmedizin. 2015; 31: 273-9.

Part I

Current outComes In ColoreCtal surgery

S. de Zeeuw1

U. Ahmed Ali1

R.A.R.T. Donders2

W.E. Hueting3

F. Keus1

C.J.H.M. van Laarhoven1

1 Department of Surgery, Radboud University Medical Centre, Nijmegen2 Department of Epidemiology, Biostatistics and HTA, Radboud University Medical Centre,

Nijmegen3 Department of Surgery, Diakonessenhuis, Leiden

2 update of complications and

functional outcome of the

ileal pouch-anal anastomosis:

overview of evidence and meta-

analysis of 96 observational

studies

International Journal of Colorectal Disease, 2012.

22 Chapter 2

absTracT

backgroundThe ileal pouch-anal anastomosis (IPAA) is the procedure of choice for the surgical treat-

ment of ulcerative colitis and familial adenomatous polyposis, since its introduction over

30 years ago. We aim at providing a comprehensive update of the outcome of the IPAA

over the past decades.

MethodsA systematic review with an extensive search in PubMed, EMBASE and The Cochrane

Library was conducted. All studies published after 2000 reporting on complications or

functional outcome after a primary open IPAA procedure for UC or FAP were selected.

Study characteristics, functional outcome and complications were extracted. A review

with similar methodology conducted 10-years earlier was used to evaluate developments

in outcome over time. Pooled estimates were compared using a random-effects logistic

meta-analyzing technique. Analyses focusing on the effect of time of study conductance,

centralization and variation in surgical techniques were performed.

resultsFifty three studies including 14,966 patients were included. Pooled rate of pouch failure

and pelvic sepsis were 4.3% (95% CI 3.5 – 6.3) and 7.5% (95% CI 6.1 – 9.1), respectively.

Compared to studies published before 2000, a reduction of 2.5% was observed in the

pouch failure rate (p = 0,0038). Analysis on the effect of the time of study conductance

confirmed a decline in pouch failure. Functional outcome remained stable over time, with

a 24-hour defecation frequency of 5.9 (95% CI 5.0 to 6.9). Technical surgery aspects did

not have an important effect on outcome.

conclusionThis review provides up to date outcome estimates of the IPAA procedure that can be

useful as reference values for practice and research. It is also shows a reduction in pouch

failure over time.

Chapter 2 23

2

inTroducTion

Restorative proctocolectomy by means of an ileal pouch-anal anastomosis (IPAA) is the

procedure of choice for the surgical treatment of ulcerative colitis (UC) and familial ad-

enomatous polyposis (FAP). The IPAA procedure generally results in acceptable long-term

functional outcomes and improvement of quality of life. It is, however, associated with

substantial morbidity with complications up to 50% of patients1. Since its introduction in

1978, the IPAA procedure has continuously been subjected to attempts of improvements

in technique.2 Additionally, much has been achieved in improving the peri-operative care.

Centralization of complex surgery has been a recent development in many countries. All

these developments may have resulted in improved outcomes.

A large number of observational studies have reported outcomes of the IPAA procedure,

most commonly from a selected group of patients in one hospital. This review aims to

provide an overview of the available evidence and to evaluate the effect of the continuous

developments on outcomes. Previously, a systematic review summarizing complications

and functional results after an IPAA procedure in studies published until 2000 was con-

ducted. This current systematic review provides an update regarding the outcomes of the

IPAA procedure and uses the combined data set of both reviews to analyze changes in the

outcomes of the IPAA procedure.

MeThods

This systematic review aims to provide an overview regarding the outcomes of the open

IPAA procedure. Additionally, it evaluates the effect of development of practice on the

most important outcomes over time.

search strategyA systematic literature search with predefined search terms was carried out in 4 electronic

databases: MEDLINE (1-1-2000 to 1-1-2010), EMBASE (1-1-2000 to 1-1-2010), and the

Cochrane Library (Issue 1, 2010)) (Figure 1). Two authors (SZ and UAA) independently

performed the selection of the publications according to inclusion criteria. Disagreements

were discussed with a third reviewer (FK). Additional relevant studies were looked for by

cross-reference checking of all included studies. Language restrictions were not applied.

24 Chapter 2

selection of studies

Inclusion criteria

Title and abstract of all identified publications (prospective and retrospective and observa-

tional and comparison studies) were screened according to the following inclusion criteria:

the study population consists of adult patients with established UC or FAP undergoing a

primary IPAA procedure; the intervention is a clearly documented open IPAA procedure in

an elective setting with or without a diverting ileostomy and irrespective of the number of

stages of the operation; the study reports at least one of the primary outcomes reported

below; the study includes a consecutive series with a minimal sample size of 50 patients.

10301 studies from search

terms

236 studies extracted by title

94 studies read by two

authors

53 studies included for meta-

analysis

58 studies < 50 patients

4 studies with morbus Crohn as population

21 studies with other outcome measures

12 studies with patients < 18 years

47 studies with laparoscopic surgery

Figure 1. Flow-chart of study selection resulting in the 54 included studies

Chapter 2 25

2

Exclusion criteria

Studies were excluded from this systematic review if they included children (<18 years) or

primarily elderly patients (> 65 years). Studies were excluded as well when the IPAA was

performed for other indications than UC or FAP (e.g. Morbus Crohn) or when a second-

ary IPAA was performed (e.g. after ileo-rectal anastomosis). Studies were also excluded

when there was selective outcome reporting, i.e. studies focusing on one parameter (e.g.

pouchitits) without reporting of any additional outcomes.

outcomes of interest and definitionsPrimary outcomes were pouch failure, pelvic sepsis, and severe day incontinence. Second-

ary outcomes included other complications (stricture, fistula, small bowel obstruction,

pouchitis, and sexual dysfunction) and parameters of functional outcome (defecation

frequency and incontinence). Table 1 shows the definitions of all outcomes.

Table 1. Definition of outcomes

Pouch failure Pouch excision or a nonfunctioning pouch at 12 months after IPAA

Pelvic sepsis Pelvic abscess, anastomotic leakage or dehiscence, pelvic/perineal wound infection

Fistula Any pouch-related fistula

Stricture Anastomotic fibrosis necessitating dilatation

Mild fecal incontinence Soiling, spotting in underwear

Severe fecal incontinence Regularly severe leakage or fecal loss passive fecal incontinence

data extractionAfter assessment for eligibility, two authors (SZ and UAA) independently extracted the fol-

lowing data if available: numbers of patients, patient characteristics, dates of start and end

of the inclusion period, duration of follow-up, variations in surgical technique, numbers

and type of complications, and parameters of pouch functional outcome. Authors were

contacted when data were missing. One author provided updated unpublished informa-

tion.3 Double publications describing identical populations were considered as one study.

data of studies published until 2000Previously, a systematic review was performed summarizing complications and functional

results after an IPAA procedure in studies published until 20004. For time-frame analyses

regarding the development of clinical practice and its effect on outcomes, data from both

reviews were combined. Both reviews were conducted with similar methodology, and the

full data set from the previous review was available.

26 Chapter 2

statistical analysisRelevant study characteristics and outcomes were extracted and presented for each study

individually. Outcomes were subsequently pooled and cumulative probabilities were calcu-

lated with 95% confidence intervals (95% CI). Data was pooled using a random-effects

model, based on the Restricted Maximum Likelihood (REML) estimator in order to incorpo-

rate the heterogeneity between studies. Pooled results of this review (all studies published

since 2000) were compared with pooled results of the previously published review (all stud-

ies published before 2000) by analyzing the complete set in one random-effects model and

incorporating a dummy variable that coded for period. The Knapp and Hartung adjustment

was used to obtain estimates and confidence intervals.5 Additionally, analysis focusing on

the effect of time of study conductance, centralization, and variation in techniques were

performed using linear regression and one-way ANOVA as appropriate. Data management

and statistical analyses were conducted using SPSS (version 15) and R Statistical Software

(R Development Core Team, version 2.11.1) with the metafor package.6,7

resulTs

selection processThe search resulted in a total of 10,301 hits. Initial selections based on titles identified 236

potentially relevant articles. Further selections based on abstracts excluded 142 studies.

The full-text of the remaining 94 studies was evaluated. Finally, a total of 53 studies with

14,966 patients were included (Figure 1).

description of identified studiesThe most important characteristics of the included studies are presented in Table 2. Most

studies were retrospective cohort studies (70%). The median sample size was 127 (range

50 to 2490) over a median inclusion periods of 12 years (range 4 to 30). A diverting

ileostomy was used in 79% (range 5 – 100%) of all the IPAA procedures. A handsewn

anastomosis was used in 40% of the procedures (range 0 – 100%). Median duration of

follow-up was 75 months (range 6 – 180). 31 studies mentioned postoperative mortality

with a median of 0% (range 0 – 2.9%).

Chapter 2 27

2

Table 2. Characteristics of studies and selected outcomes following IPAA in the 54 included studies published since 2000

Author Year of

pub.

Inclusion period

No. of patients

FU(mnt)

Age Gender(%

female)

Typeof

disease

Handsewn anastomo-

sis (%)

Withdivertingileostoma

(%)

Fonkalsrud8 2000 1993-1997 77 34.5 50 UC 57.1 100

Karlbom9 2000 1983-1996 168 29 32 39.3 UC 58.9 70.8

Seidel10 2000 1985-1996 55 31.2 45.5 Both

Simchuk11 2000 1987-1996 114 38 43.9 Both 100

Gullberg12 2001 1990-1995 86 72 34.1 45.3 Both 2.3 10.5

Barton13 2001 1983-2000 110 30 36.8 Both 100 100

Heuschen14 2001 1991-2000 171 30 36.7 45.6 UC 100 66.7

Regimbeau15 2001 1984-1998 172 60 36 47.1 Both 100 100

Blumberg16 2001 1982-1995 145 34 49 Both 15.2

Dayton17 2002 1982-2001 644 79 37.2 39.4 UC

Rossi18 2002 1989-2000 75 40 33.3 Both 54.7 100

Bullard19 2002 1980-1992 154 151 34 44.2 Both 64

Gignoux20 2002 1993-1998 84 22 33 42.9 Both 100

Mukhtar21 2002 1990-1997 50 35 44 Both 2 98

Rudolph22 2002 1991-1999 71 47 38 56 UC 84.5

Carmon23 2003 1990-2001 77 51 38 55.8 UC 72.7

Teixeira24 2003 1986-2000 80 108 50 UC 11.3 77.5

Michelassi25 2003 1987-2002 391 37 33.7 44.2 Both 29.9 65

Holubar26 2003 51 85 42.1 51 UC 70.6

Fowler27 2003 1984-2001 106 40 50 Both 39.4 41.3

Parc28 2004 1983-1990 110 175 26.5 43.6 FAP 100

Lin29 2004 1985-2002 61 60 37 54 Both 67.2

Hueting30 2004 1989-2000 111 42 35.4 68.5 UC

Gosselink31 2004 1989-2001 127 56 36 42.5 UC 21.6

Krausz32 2005 1984-2004 174 64 34.1 47.7 Both 54 87.4

Hallberg33 2005 1990-1997 100 53 32 39 UC 17

Swenson34 2005 1995-2003 54 22 35 35.2 57.4

Chapman35 2005 1981-2000 2002 121 33.5 46.6 Both 5

Kok Sun Ho36 2005 1989-2001 330 32 59.1 UC

Ikeuchi37 2005 1999-2003 242 32.1 50.8 UC 100 38

Arai38 2005 1993-2003 296 53 33.8 43.2 UC 96.3 55.4

Araki39 2006 1998-2003 220 34.5 46.8 Both 47.3

Rickard40 2006 1982-2002 516 33 45.2 Both 32.4 88

Bednarz41 2007 1994-2005 108 60 50 Both 18.5 100

Berndtsson42 2007 1982-1995 370 180 34 45.7 UC

Gualdrini43 2007 1987-2005 391 45.6 35.4 51.7 Both

28 Chapter 2

outcomes

Complications

Data on complications were available from all 53 studies. We found a pooled incidence of

pouch failure of 4.3% (95% CI 3.5 – 6.3) (Figure 2). A sensitivity analysis including only

studies with minimal follow-up of 5 years showed a pouch failure rate of 4.7%. The pooled

incidence of pelvic sepsis was 7.5% (95% CI 6.1 – 9.1) (Figure 3). Pooled incidence rates

for other important complications are presented in Table 3.

Functional results

Data on functional results after IPAA were available from 26 studies including 5,321

patients. The pooled incidence of mild and severe fecal incontinence during the day was

14.3% (CI 7.3 – 25.9) and 6.1% (2.9 – 12.3), respectively (Table 3). Mean defecation

frequency was 5.9 (4.9 – 6.7) per 24 hours with a mean nighttime frequency of 1.5 (1.0

– 2.1).

Table 2. (continued)

Author Year of

pub.

Inclusion period

No. of patients

FU(mnt)

Age Gender(%

female)

Typeof

disease

Handsewn anastomo-

sis (%)

Withdivertingileostoma

(%)

Kiran44 2008 2016 180 37.7 43.9 Both 15.4 85.9

Tulchinsky3 2008 1986-2005 211 124.6 35.4 58.3 UC 16.5 79

Walker45 2008 1987-2004 178 84 31 44.9 Both 6.7 100

Ferrante46 2008 1990-2004 173 78 39 38.7 UC 2.3 85

Leal47 2008 1984-2006 69 82 29.6 59.4 FAP 100 100

Zarate48 2008 1984-2006 107 34.4 57 Both 12.1 100

Pinto49 2009 2001-2008 66 21.1 52.8 66.7 Both 18.2 100

Leowardi50 2009 1988-1996 197 147.6 33 45.2 UC 100 100

Norwood51 2009 1997-2007 107 77 38 45.8 Both 99

Loftus52 2009 2001-2005 237 5.9 38.9 45 UC

Tekkis53 2009 1976-2006 2491 74.4 35.9 47.9 Both 33.3 67.3

Scarpa54 2009 1984-2004 58 136 38 34.5 UC 0

Wasmuth55 2009 1984-2006 304 120 34.4 38.8 Both 40.1 84

Andersson56 2009 1984-2003 110 144 35 45.5 Both 47.3

Fichera57 2009 2002-2007 106 24.8 36.9 42.2 UC 71.6

Rink2 2009 1990-2002 131 85 33 46.6 UC 84

Pricolo58 2009 1991-2008 312 36 46.8 Both 8

Chapter 2 29

2

Comparison with studies published before 2000

The previous review of studies published before 2000 identified 43 studies with a total of

9,317 patients.4 Median duration of follow-up was 36.7 months (range 12-99). Detailed

discussion of the characteristics of these studies is published elsewhere.4

We compared incidences of the most important complications reported in studies pub-

lished prior to 2000 versus the incidences reported in studies published since 2000 (Table

3). Pouch failure decreased between the two time periods from 6.8% to 4.3% (p=0.0038).

Pelvic sepsis showed a decrease from 9.5% to 7.4 (p=0.1284). Other complications and the

functional results showed no significant differences between the two groups of studies.

0 10 20 30 40

Percentage

PricoloRinkFichera WasmuthScarpaTekkisNorwoodLeowardiZarateLealFerranteWalkerTulchinsky KiranGualdriniBerndtssonBednarzAraiKok Sun HoChapmanHallbergKrauszGosselinkParcFowlerHolubarMichelassiTeixeiraCarmonRudolphMukhtarGignouxBullardRossiDaytonBlumbergRegimbeauHeuschenBartonGullbergSimchukSeidelKarlbom

2009200920092009200920092009200920082008200820082008200820072007200720052005200520052005200420042003200320032003200320022002200220022002200220012001200120012001200020002000

312118106304

582491107197107

69173178211

2016391370108296330

200290

174127110106

51391

8077715084

15475

644145172171110

86114

55168

3.8] 14.0] 9.6]

11.1] 12.8] 8.8] 9.5]

21.5] 10.7] 16.3] 9.7] 5.1] 4.9] 6.8] 7.8]

14.7] 6.3] 3.1] 4.8] 7.0] 7.5]

10.4] 12.1] 6.2] 9.6]

13.6] 5.6] 8.3] 9.4]

17.6] 14.6] 8.0]

12.4] 8.9] 2.7] 7.1] 6.8] 9.8]

12.7] 8.8] 7.8]

15.6] 12.2]

,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,

0.7 4.0 1.4 5.1 0.9 6.7 1.411.3 2.0 3.0 2.7 0.5 0.7 4.8 3.3 8.3 0.1 0.3 1.2 4.9 0.2 3.1 3.2 0.1 1.4 0.0 1.9 0.2 0.0 3.8 1.0 0.2 4.0 0.2 0.7 1.0 1.2 2.8 3.1 0.6 0.9 1.8 4.1

[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[

1.6 7.6 3.8 7.6 3.4 7.7 3.715.7 4.7 7.2 5.2 1.7 1.9 5.7 5.111.1 0.9 1.0 2.4 5.9 1.1 5.7 6.3 0.9 3.8 0.0 3.3 1.2 0.0 8.5 4.0 1.2 7.1 1.3 1.4 2.8 2.9 5.3 6.4 2.3 2.6 5.5 7.1

5.3], 3.5[ 4.3RE Model

Author Year Study Size

Figure 2. Proportions of pouch failure following an IPAA procedure in studies published since 2000The size of the dot correlates with the population size of the study. The lines represent the 95% CI.The pooled incidence of all studies is 4.3%.

30 Chapter 2

The effect of time of study conductance on iPaa outcomesResults of both reviews were combined to analyze the effect of time of study conductance

on the outcomes of the IPAA procedure. For this, the relationship between the median

year of inclusion of each study and the outcomes observed in that study was evaluated.

Regression analysis showed that studies with an earlier median year of inclusion have a

higher rate of pouch failure than studies with a more recent median year of inclusion

(p<0,001). Figure 4 is a scatter-plot depicting this relationship. In a sensitivity analysis

excluding the 5 studies with a pouch failure rate of >12.5%, the relationship remained

statistically significant (p=0.009). Additionally, we categorized the median year of inclusion

of all studies (published before and after 2000) into four interquartile ranges (IQR) in order

to evaluate the median rate of pouch failure over these four time periods. Figure 5 shows

0 10 20 30 40

Percentage

PricoloRinkFichera AnderssonWasmuthScarpaTekkisLoftusNorwoodPintoZarateLealWalkerKiranGualdriniBednarzRickardArakiAraiIkeuchiChapmanSwensonHallbergKrauszGosselinkHuetingParcLinFowlerHolubarMichelassiTeixeiraCarmonMukhtarGignouxRossiDaytonBlumbergRegimbeauHeuschenBartonGullbergSimchukSeidelKarlbomFonkalsrud

2009200920092009200920092009200920092009200820082008200820072007200620062005200520052005200520052004200420042003200320032003200320032002200220022002200120012001200120012000200020002000

312131106110304

582491

237107

66107

69178

2016391108516220296242

20025490

174127111110

61106

51391

8077508475

644145172171110

86114

55168

77

2.5] 5.9]

20.0] 11.6] 24.9] 12.8] 13.1] 32.1] 22.0] 11.3] 10.7] 19.8] 9.4] 6.6]

27.5] 6.3] 7.1]

12.6] 7.8]

14.4] 8.2]

15.9] 22.0] 14.5] 13.1] 23.3] 6.2]

14.2] 12.0] 12.6] 10.8] 17.2] 9.8]

19.5] 15.0] 16.7] 6.0]

14.8] 13.3] 25.9] 6.2]

18.9] 11.2] 24.4] 12.9] 8.6]

,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,

0.2 0.4 7.3 2.5

15.9 0.9

10.621.0 8.6 0.8 2.0 4.9 2.7 4.6

19.1 0.1 3.3 5.2 2.8 6.7 6.0 1.8 7.7 5.7 3.7 9.7 0.1 1.6 2.6 0.3 5.4 4.2 0.7 3.0 3.2 3.6 2.9 5.3 4.9

14.1 0.1 5.5 2.4 6.2 4.5 0.2

[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[

0.6 1.5

12.3 5.5

20.1 3.4

11.826.214.0 3.0 4.7

10.1 5.1 5.5

23.0 0.9 4.8 8.2 4.7 9.9 7.0 5.6

13.3 9.2 7.1

15.3 0.9 4.9 5.7 2.0 7.7 8.8 2.6 8.0 7.1 8.0 4.2 9.0 8.1

19.3 0.9

10.5 5.3

12.7 7.7 1.3

9.1], 6.1[ 7.5RE Model

Author Year Study Size

Figure 3. Proportions of pelvic sepsis following an IPAA procedure in studies published since 2000The size of the dot correlates with the population size of the study. The lines represent the 95% CI.The pooled incidence of all studies is 7.5%.

Chapter 2 31

2

a consecutive decline in the pouch failure rate over these time periods (p=0.005, one-way

ANOVA analysis).

centralization: trends over timeThe median inclusion rate for all studies was 14 patients per year (range from 2.9 to 105).

We evaluated the effect of time of study conductance (median year of inclusion) on the

inclusion rate (Figure 6). The figure shows that between 1991 and 2000 an increase was

seen in studies from centers including less than 1 patient per month, from 8 studies (30%)

before 1991 to 23 studies (46%) between 1991 and 2000. It also shows that there are not

yet enough studies available that have been conducted after 2000 to draw conclusions

Table 3. Pooled incidences of complications and functional outcome following an IPAA procedures in studies published since 2000 compared to studies published before 2000

Since 2000 Before 2000P-

valueNo. of studies

No. of patients

Pooled incidences

No. of studies

No. of patients

Pooled incidences

Complications % (95% CI) % (95%CI)

Pouch failure 43 13249 4.3 (3.5–5.3) 39 8877 6.8 (5.8–8.4) 0.004

Pouch failure: FU > 5 yrs

22 9873 4.7 (3.4–6.4) 11 3198 8.5 (5.4–13.2)

Pelvic sepsis 46 13450 7.5 (6.1–9.1) 41 9082 9.5 (8.2–10.9) NS

Fistula 38 12398 4.5 (3.5–5.7) 30 5120 5.5 (4.3–7.0) NS

Stricture 35 12219 10.7 (8.2–13.8) 28 5185 9.2 (6.8–12.4) NS

Pouchitis 39 12685 26.8 (21.0–33.5) 33 7289 18.8 (15.7–22.4) NS

Sexual dysfunction

13 6131 3.0 (1.7–5.2) 21 5112 3.6 (2.7–4.7) NS

Small bowel obstr.

34 11895 11.4 (9.1–14.1) 27 5853 13.1 (11.0–15.7) NS

Fecal incontinence

%(95%CI) % (95%CI)

Mild day incont. 21 6988 14.3 (7.3–25.9) 31 4313 17.0 (12.8–22.2) NS

Severe day incont.

13 3718 6.1 (2.9–12.3) 27 3914 3.7 (2.8–4.8) NS

Mild night incont.

9 5423 17.3 (4.7–46.8) 17 2582 13.1 (9.5–17.9) NS

Severe night incont.

10 3614 7.6 (2.5–21.3) 9 1271 4.5 (3.0–6.7) NS

Frequency mean(95%CI) mean (95%CI)

Frequency day 26 5321 5.7 (4.9–6.7) 13 2277 5.2 (4.0–6.7) NS

Frequency night 22 7117 1.5 (1.0–2.1) 20 2950 1.0 (0.6–1.6) NS

Frequency 24 h 26 5132 5.9 (5.0–6.9) 20 3547 5.2 (4.4–6.1) NS

CI: confidence interval; FU: Follow-up period; Yrs: years

32 Chapter 2

Figure 4. Scatter plot depicting the median year of inclusion and the rate of pouch failure of studies included in both reviews

Figure 5. Error-bar chart depicting the median rate of pouch failure compared to the median year of inclusion (divided into interquartile ranges)

Chapter 2 33

2

about the current situation. Analysis of the effect of the inclusion rate on the outcomes of

the IPAA did not show any benefit on the outcomes of the IPAA.

Technical aspects of the iPaaOverall, 43 out of 78 studies (55%) used a protective ileostomy as a routine part of the IPAA

procedure (routine use defined as ileostomy performed in more than 80% of patients). The

percentage has decreased from almost 70% (16 out of 23) in studies conducted between

1980 and 1990, to 47% (21 out of 44) in studies conducted after 1990. The number of

studies performing an ileostomy in less than 50% of patients increased from 11% to 32%

over the same periods. The rate of ileostomy was not associated with any of the outcomes

of the IPAA. Type of anastomosis (handsewn vs. stapled) was only available from 31 studies

published since 2000 (the previous review did not register this characteristic). Fifteen stud-

ies (48%) still performed a handsewn anastomosis in more than 50% of patients. There

were no differences in functional outcome or complications between studies performing

different types of anastomosis.

Figure 6. Relationship between moment of study conductance (median year of inclusion) and inclu-sion rate of patients, divided in the past three decades

34 Chapter 2

discussion

The open IPAA operation is the most common surgical reconstruction procedure in patients

with UC and FAP. Randomized trials on IPAA are rare1 due to low incidences as well as the

complexity of patient selection and the intervention. Therefore, most studies are retro-

spective case-series. Estimates of the outcomes of the IPAA procedure vary considerably.

Acceptable complication and functional outcome rates are, therefore, difficult to set. This

systematic review provides up to date and reliable incidences of important outcomes of the

IPAA procedure, that could provide a reference values for these outcomes. Additionally, this

review may provide insights in changes of practice over time and their effects on outcomes.

Several important finding are provided. First, it shows that the rate of pouch failure after

an IPAA procedure has declined continuously over time. This decline was larger in the

earlier period of the IPAA, but seems to continue over time. Secondly, a large proportion

of studies with this complex type of surgery are conducted in centers that include less than

one patient per month. Thirdly, the functional outcome of the IPAA seems not to improve

over time, despite several developments in surgical technique and peri-operative care. This

may represent an intrinsic limitation of the IPAA technique regarding functional outcome.

This review has some limitations. First, many of the included studies are retrospective

cohort series, which may limit the reliability of the results. However, the overwhelming

majority of the evidence in this field consists of this type of studies. This review provides

an overview of this evidence, thereby increases its utility for clinicians and researchers.

Moreover, the statistical techniques used are rather conservative to allow for incorporation

of the heterogeneity of the studies.

Second, publication bias is a factor that should be taken into consideration. For this purpose,

we excluded series with less than 50 patients. Studies with a small sample size do not only

increase heterogeneity of results, but are also more subject to publication bias; i.e. small

studies with ‘unfavorable’ results are less likely to get published. By introducing a threshold

we aimed to prevent including small studies with overestimated intervention effects.

Finally, the lack of individual patient data limits the analyses that could be performed.

For example, many studies report a cohort in which several types of operative techniques

have been used (e.g. several types of anastomses). This prohibits analyzing the effect of

variations in technique on a patient level, since only aggregate outcomes of the complete

cohorts are available. Collecting individual patient data of these large numbers of studies,

especially considering the dated studies, is not feasible. We, therefore, limited our analyses

to factors that could be analyzed reliably on an aggregate level.

Chapter 2 35

2

Many factors may have contributed to the reduction of the rate of pouch failure observed.

The additional analyses in this review were unable to pinpoint factors responsible for the

observed reduction in pouch failure. This is probably caused by the lack of individual pa-

tient data (as discussed above) and we believe that general improvements in care over the

past decades, and the increased experience of surgeons with this approach are the most

important factors contributing to this finding.

Another explanation that could be considered is the variation in the surgical techniques

that have been introduced and sometimes largely employed over the last decades. Such

variations include the various types of pouches used and the use of the double-stapled

anastomosis. The routinely constructed deviating ileostoma does not seem to improve

the frequency of postoperative complications, but only their severity. Despite comparative

studies and several large meta-analyses published previously, none of these variations in

technique could be identified as factors associated with the reduction of postoperative

complications, including pouch failure.59,60 This is in line with the findings in this review.

Fewer studies reported on other outcomes, which decreased the power of the analyses

concerning these other outcomes. Secondly, it could be that the observed decrease in the

rate of pouch failure may be, at least in part, a result of smaller decreases in various other

postoperative complications. Complications like pelvic sepsis, fistulae and strictures could

all result in pouch failure, provided they were severe enough. Smaller decreases in these

outcomes can be difficult to show statistically. Overall, however, they may have caused the

observed substantial decrease in pouch failure rate.

Recently, discussions on the relation between volume of complex surgical interventions

and outcomes have gained wide attention.61-64 Certain complex surgical interventions have

been centralized in high volume hospitals with improved outcomes.65-68 Centralization still

has to be reflected in the data in the years to come. Therefore, this review was unable to

evaluate the potential influence of this development on current practice. However, in the

period from 1991 to 2000 nearly 50% of studies were performed in ‘low inclusion rate’

centers (including less than 1 patient per month). This is a substantial increase from the

30% observed in the period from1980 to 1990. Taking into account that we only included

series of 50 patients or more (Figure 1), may mean that the true proportion of studies

conducted in such ‘low inclusion rate’ centers may be much higher. With the repeatedly

established relationship between volume and outcome in several fields of surgery, this is a

troublesome observation. Centralization of the complex IPAA procedure may offer a way

to further improve outcomes in future.

36 Chapter 2

Finally, we found no statistically significant differences in pooled estimates of the func-

tional outcome over time. There were also no differences in functional outcomes between

various surgical factors studied in this review. This is in agreement with findings in literature

showing no benefits in terms of functional outcome of technical developments of the

IPAA procedure, like type of anastomosis69 or laparoscopic approach.1 In most patients,

however, functional outcome of the IPAA is considered to be highly satisfactorily, allowing

for an acceptable quality of life and social functioning.70,71 Therefore, reducing complica-

tions, especially severe complications like pelvic sepsis and pouch failure, seems to be a

more important goal for future surgical developments.

conclusions

This systematic review and meta-analysis summarizes the available evidence and provides

up to date estimates for complications and functional outcome after an IPAA procedure

that can be useful as reference values for practice and research. It also shows a reduction

in pouch failure over time in patients operated on using the IPAA procedure. The functional

results remain unchanged over time and between various operative factors.

Chapter 2 37

2

references

1. Ahmed Ali U, Keus F, Heikens JT, Bemel-

man WA, Berdah SV, Gooszen HG, van

Laarhoven CJ. Open versus laparoscopic

(assisted) ileo pouch anal anastomosis for

ulcerative colitis and familial adenomatous

polyposis. The Cochrane database of sys-

tematic reviews. 2009: Cd006267.

2. Rink AD, Radinski I, Vestweber KH. Does

mesorectal preservation protect the ileoanal

anastomosis after restorative proctocolec-

tomy? Journal of gastrointestinal surgery:

official journal of the Society for Surgery of

the Alimentary Tract. 2009; 13: 120-8.

3. Tulchinsky H, Dotan I, Alper A, Brazowski

E, Klausner JM, Halpern Z, Rabau M.

Comprehensive pouch clinic concept for

follow-up of patients after ileal pouch anal

anastomosis: Report of 3 years’ experience

in a tertiary referral center. Inflammatory

bowel diseases. 2008; 14: 1125-32.







5. Knapp G, Hartung J. Improved tests for

a random effects meta-regression with

a single covariate. Statistics in medicine.

2003; 22: 2693-710.

6. Team RDC. A language and environment

for statistical computing. Vienna, Austria;

2010.

7. Viechtbauer W. Conducting meta-analyses

in r with the metafor package. Journal of

Statistical Software. 2010; 36: 1-48.

8. Fonkalsrud EW, Thakur A, Roof L. Com-

parison of loop versus end ileostomy for

fecal diversion after restorative procto-

colectomy for ulcerative colitis. Journal of

the American College of Surgeons. 2000;

190: 418-22.

9. Karlbom U, Raab Y, Ejerblad S, Graf W,

Thorn M, Pahlman L. Factors influencing

the functional outcome of restorative proc-

tocolectomy in ulcerative colitis. The British

journal of surgery. 2000; 87: 1401-8.

10. Seidel SA, Newman M, Sharp KW. Ileoanal

pouch versus ileostomy: Is there a dif-

ference in quality of life? The American

surgeon. 2000; 66: 540-6; discussion 46-7.

11. Simchuk EJ, Thirlby RC. Risk factors and

true incidence of pouchitis in patients

after ileal pouch-anal anastomoses. World


12. Gullberg K, Liljeqvist L. Stapled ileoanal

pouches without loop ileostomy: A pro-

spective study in 86 patients. International

journal of colorectal disease. 2001; 16:

221-7.

13. Barton JG, Paden MA, Lane M, Postier RG.

Comparison of postoperative outcomes

in ulcerative colitis and familial polyposis

patients after ileoanal pouch operations.

American journal of surgery. 2001; 182:

616-20.

14. Heuschen UA, Hinz U, Allemeyer EH, Lucas

M, Heuschen G, Herfarth C. One- or two-

stage procedure for restorative proctocolec-

tomy: Rationale for a surgical strategy in

ulcerative colitis. Annals of surgery. 2001;

234: 788-94.

15. Regimbeau JM, Panis Y, Pocard M, Haute-

feuille P, Valleur P. Handsewn ileal pouch-

anal anastomosis on the dentate line

after total proctectomy: Technique to avoid

incomplete mucosectomy and the need for

long-term follow-up of the anal transition

zone. Diseases of the colon and rectum.

2001; 44: 43-50; discussion 50-1.

16. Blumberg D, Opelka FG, Hicks TC, Timmcke

AE, Beck DE. Restorative proctocolectomy:

Ochsner clinic experience. Southern medi-

cal journal. 2001; 94: 467-71.

38 Chapter 2

17. Dayton MT, Larsen KR, Christiansen DD.

Similar functional results and complica-

tions after ileal pouch-anal anastomosis in

patients with indeterminate vs ulcerative

colitis. Archives of surgery (Chicago, Ill.:

1960). 2002; 137: 690-4; discussion 94-5.

18. Rossi HL, Brand MI, Saclarides TJ. Anal com-

plications after restorative proctocolectomy

(j-pouch). The American surgeon. 2002; 68:

628-30.

19. Bullard KM, Madoff RD, Gemlo BT. Is

ileoanal pouch function stable with time?

Results of a prospective audit. Diseases of

the colon and rectum. 2002; 45: 299-304.

20. Gignoux BM, Dehni N, Parc R, Tiret E. [ileal

pouch anal-anastomosis without protective

ileostomy]. Gastroenterologie clinique et

biologique. 2002; 26: 671-4.

21. Mukhtar H, Zak A, Berry AR. Restorative

proctocolectomy: One surgeon’s experience

of a now routine procedure in a district

general hospital. Journal of the Royal Col-

lege of Surgeons of Edinburgh. 2002; 47:

400-6.

22. Rudolph WG, Uthoff SM, McAuliffe TL,

Goode ET, Petras RE, Galandiuk S. Indeter-

minate colitis: The real story. Diseases of


23. Carmon E, Keidar A, Ravid A, Goldman G,

Rabau M. The correlation between quality

of life and functional outcome in ulcerative

colitis patients after proctocolectomy

ileal pouch anal anastomosis. Colorectal

disease: the official journal of the Associa-

tion of Coloproctology of Great Britain and

Ireland. 2003; 5: 228-32.

24. Teixeira MG, Ponte AC, Sousa M, Almeida

MG, Silva Filho E, Calache JE, Habr-Gama

A, et al. Short- and long-term outcomes of

ileal pouch-anal anastomosis for ulcerative

colitis. Revista do Hospital das Clinicas.

2003; 58: 193-8.

25. Michelassi F, Lee J, Rubin M, Fichera A,

Kasza K, Karrison T, Hurst RD. Long-term

functional results after ileal pouch anal

restorative proctocolectomy for ulcerative

colitis: A prospective observational study.

Annals of surgery. 2003; 238: 433-41; dis-

cussion 42-5.

26. Holubar S, Hyman N. Continence altera-

tions after ileal pouch-anal anastomosis do

not diminish quality of life. Diseases of the

colon and rectum. 2003; 46: 1489-91.

27. Fowler AL, Turner BN, Thomson WH. A

study of the complications and pelvic

visceral function after restorative proc-

tocolectomy and w pouch construction.

Colorectal disease: the official journal of

the Association of Coloproctology of Great

Britain and Ireland. 2003; 5: 342-6.

28. Parc Y, Piquard A, Dozois RR, Parc R, Tiret

E. Long-term outcome of familial adeno-

matous polyposis patients after restorative

coloproctectomy. Annals of surgery. 2004;

239: 378-82.

29. Lin JJ, Song ZF, Xu JH. [the prognosis of

the total proctocolectomy and ileal-pouch

anal anastomosis]. Zhonghua wai ke za

zhi [Chinese journal of surgery]. 2004; 42:

861-3.

30. Hueting WE, Gooszen HG, van Laarhoven

CJ. Sexual function and continence after

ileo pouch anal anastomosis: A comparison

between a meta-analysis and a question-

naire survey. International journal of

colorectal disease. 2004; 19: 215-8.

31. Gosselink MP, Schouten WR, van Lieshout

LM, Hop WC, Laman JD, Ruseler-van

Embden JG. Delay of the first onset of pou-

chitis by oral intake of the probiotic strain

lactobacillus rhamnosus gg. Diseases of the

colon and rectum. 2004; 47: 876-84.

32. Krausz MM, Duek SD. Restorative proc-

tocolectomy with ileal pouch-anal anas-

tomosis for ulcerative colitis and familial

adenomatous polyposis: Twenty years

follow-up in 174 patients. The Israel Medi-

cal Association journal: IMAJ. 2005; 7: 23-7.

33. Hallberg H, Stahlberg D, Akerlund JE. Ileal

pouch-anal anastomosis (ipaa): Functional

outcome after postoperative pelvic sepsis.

Chapter 2 39

2

A prospective study of 100 patients. Inter-

national journal of colorectal disease. 2005;

20: 529-33.

34. Swenson BR, Hollenbeak CS, Poritz LS, Kol-

tun WA. Modified two-stage ileal pouch-

anal anastomosis: Equivalent outcomes

with less resource utilization. Diseases of


35. Chapman JR, Larson DW, Wolff BG, Dozois

EJ, Cima RR, Pemberton JH, Crownhart BS,

et al. Ileal pouch-anal anastomosis: Does

age at the time of surgery affect outcome?

Archives of surgery (Chicago, Ill.: 1960).

2005; 140: 534-9; discussion 39-40.

36. Ho KS, Chang CC, Baig MK, Borjesson L,

Nogueras JJ, Efron J, Weiss EG, et al. Ileal

pouch anal anastomosis for ulcerative colitis

is feasible for septuagenarians. Colorectal



Ireland. 2006; 8: 235-8.

37. Ikeuchi H, Nakano H, Uchino M, Nakamura

M, Noda M, Yanagi H, Yamamura T. Safety

of one-stage restorative proctocolectomy

for ulcerative colitis. Diseases of the colon

and rectum. 2005; 48: 1550-5.

38. Arai K, Koganei K, Kimura H, Akatani M,

Kitoh F, Sugita A, Fukushima T. Incidence

and outcome of complications following

restorative proctocolectomy. American


39. Araki T, Parc Y, Lefevre J, Dehni N, Tiret E,

Parc R. The effect on morbidity of mesen-

tery lengthening techniques and the use

of a covering stoma after ileoanal pouch

surgery. Diseases of the colon and rectum.

2006; 49: 621-8.

40. Rickard MJ, Young CJ, Bissett IP, Stitz R,

Solomon MJ. Ileal pouch-anal anastomosis:

The australasian experience. Colorectal



Ireland. 2007; 9: 139-45.

41. Bednarz W, Olewinski R, Woldan J. Ileal-

pouch-anal anastomosis after restorative

proctocolectomy in patients with ulcerative

colitis or familial adenomatous polypo-

sis—11 years of experience. Polski Przeglad

Chirurgiczny. 2007; 79: 652–62.

42. Berndtsson I, Lindholm E, Oresland T,

Borjesson L. Long-term outcome after

ileal pouch-anal anastomosis: Function and

health-related quality of life. Diseases of


43. Gualdrini U, Coraglio M, Filippa L, Mas-

ciangioli G, Gutierrez A, Lumi C, Graziano

A. [long term functional outcomes after

ileoanal pouch in patients with and without

postoperative complications]. Acta gastro-

enterologica Latinoamericana. 2007; 37:

143-9.

44. Kiran RP, Remzi FH, Fazio VW, Lavery IC,

Church JM, Strong SA, Hull TL. Complica-

tions and functional results after ileoanal

pouch formation in obese patients. Journal

of gastrointestinal surgery: official journal

of the Society for Surgery of the Alimentary

Tract. 2008; 12: 668-74.

45. Walker LR, Bulow S. [restorative procto-

colectomy with an ileoanal pouch. Post-

operative course and long-term functional

results]. Ugeskrift for laeger. 2008; 170:

1721-5.

46. Ferrante M, Declerck S, De Hertogh G, Van

Assche G, Geboes K, Rutgeerts P, Penninckx

F, et al. Outcome after proctocolectomy

with ileal pouch-anal anastomosis for ulcer-

ative colitis. Inflammatory bowel diseases.

2008; 14: 20-8.

47. Leal RF, Ayrizono Mde L, Coy CS, Fagundes

JJ, Goes JR. [short-term and long-term post-

operative complications after ileal pouch-

anal anastomosis in familial adenomatous

polyposis]. Arquivos de gastroenterologia.

2008; 45: 106-10.

48. Zarate CA, Zuniga DA, Pinedo MG, Lopez

KF, Molina PM, Viviani GP. [ileal pouch-anal

anastomosis for ulcerative colitis: Compli-

cations and long term functional results].

Revista medica de Chile. 2008; 136: 467-74.

40 Chapter 2

49. Pinto RA, Canedo J, Murad-Regadas S,

Regadas SF, Weiss EG, Wexner SD. Ileal

pouch-anal anastomosis in elderly patients:

Is there a difference in morbidity compared

with younger patients? Colorectal disease:

the official journal of the Association of Co-

loproctology of Great Britain and Ireland.

2011; 13: 177-83.

50. Leowardi C, Hinz U, Tariverdian M, Kienle P,

Herfarth C, Ulrich A, Kadmon M. Long-term

outcome 10 years or more after restorative

proctocolectomy and ileal pouch-anal

anastomosis in patients with ulcerative

colitis. Langenbeck’s archives of surgery /

Deutsche Gesellschaft fur Chirurgie. 2010;

395: 49-56.

51. Norwood MG, Mann CD, West K, Miller

AS, Hemingway D. Restorative procto-

colectomy. Does ethnicity affect outcome?




52. Loftus EV, Jr., Friedman HS, Delgado

DJ, Sandborn WJ. Colectomy subtypes,

follow-up surgical procedures, postsurgical

complications, and medical charges among

ulcerative colitis patients with private health

insurance in the united states. Inflamma-

tory bowel diseases. 2009; 15: 566-75.

53. Tekkis PP, Lovegrove RE, Tilney HS, Smith JJ,

Sagar PM, Shorthouse AJ, Mortensen NJ, et

al. Long-term failure and function after re-

storative proctocolectomy - a multi-centre

study of patients from the uk national

ileal pouch registry. Colorectal disease: the

official journal of the Association of Co-


2010; 12: 433-41.

54. Scarpa M, Mescoli C, Rugge M, D’Inca R,

Ruffolo C, Polese L, D’Amico DF, et al. Re-

storative proctocolectomy for inflammatory

bowel disease: The padova prognostic score

for colitis in predicting long-term outcome

and quality of life. International journal of


55. Wasmuth HH, Trano G, Endreseth B, Ryd-

ning A, Wibe A, Myrvold HE. Long-term

surgical load in patients with ileal pouch-

anal anastomosis. Colorectal disease: the



2009; 11: 711-8.

56. Andersson T, Lunde OC, Johnson E, Moum

T, Nesbakken A. Long-term functional

outcome and quality of life after restorative

proctocolectomy with ileo-anal anastomo-

sis for colitis. Colorectal disease: the official

journal of the Association of Coloproctol-

ogy of Great Britain and Ireland. 2011; 13:

431-7.

57. Fichera A, Silvestri MT, Hurst RD, Rubin

MA, Michelassi F. Laparoscopic restorative

proctocolectomy with ileal pouch anal

anastomosis: A comparative observational

study on long-term functional results.

Journal of gastrointestinal surgery: official

journal of the Society for Surgery of the

Alimentary Tract. 2009; 13: 526-32.

58. Pricolo VE. Ileal pouch-anal anastomosis for

ulcerative colitis--the rhode island experi-

ence. Medicine and health, Rhode Island.

2009; 92: 100-2.

59. Lovegrove RE, Constantinides VA, Heriot

AG, Athanasiou T, Darzi A, Remzi FH, Nich-

olls RJ, et al. A comparison of hand-sewn

versus stapled ileal pouch anal anastomosis

(ipaa) following proctocolectomy: A

meta-analysis of 4183 patients. Annals of

surgery. 2006; 244: 18-26.

60. Weston-Petrides GK, Lovegrove RE, Tilney

HS, Heriot AG, Nicholls RJ, Mortensen NJ,

Fazio VW, et al. Comparison of outcomes

after restorative proctocolectomy with or

without defunctioning ileostomy. Archives

of surgery (Chicago, Ill.: 1960). 2008; 143:

406-12.

61. Begg CB, Cramer LD, Hoskins WJ, Brennan

MF. Impact of hospital volume on operative

mortality for major cancer surgery. Jama.

1998; 280: 1747-51.

Chapter 2 41

2

62. Killeen SD, O’Sullivan MJ, Coffey JC,

Kirwan WO, Redmond HP. Provider volume

and outcomes for oncological procedures.

The British journal of surgery. 2005; 92:

389-402.

63. Halm EA, Lee C, Chassin MR. Is volume re-

lated to outcome in health care? A system-

atic review and methodologic critique of

the literature. Annals of internal medicine.

2002; 137: 511-20.

64. Dudley RA, Johansen KL, Brand R, Rennie

DJ, Milstein A. Selective referral to high-vol-

ume hospitals: Estimating potentially avoid-

able deaths. Jama. 2000; 283: 1159-66.

65. Hollenbeck BK, Wei Y, Birkmeyer JD.

Volume, process of care, and operative

mortality for cystectomy for bladder cancer.

Urology. 2007; 69: 871-5.

66. Birkmeyer NJ, Goodney PP, Stukel TA,

Hillner BE, Birkmeyer JD. Do cancer centers

designated by the national cancer institute

have better surgical outcomes? Cancer.

2005; 103: 435-41.

67. Birkmeyer JD, Siewers AE, Finlayson EV,

Stukel TA, Lucas FL, Batista I, Welch HG, et

al. Hospital volume and surgical mortality in

the united states. The New England journal

of medicine. 2002; 346: 1128-37.

68. Birkmeyer JD, Sun Y, Wong SL, Stukel TA.

Hospital volume and late survival after

cancer surgery. Annals of surgery. 2007;

245: 777-83.

69. Schluender SJ, Mei L, Yang H, Fleshner PR.

Can a meta-analysis answer the question:

Is mucosectomy and handsewn or double-

stapled anastomosis better in ileal pouch-

anal anastomosis? The American surgeon.

2006; 72: 912-6.

70. Heikens JT, de Vries J, van Laarhoven CJ.

Quality of life, health-related quality of life

and health status in patients having restor-

ative proctocolectomy with ileal pouch-anal

anastomosis for ulcerative colitis: A system-

atic review. Colorectal disease: the official



536-44.

71. Thirlby RC, Sobrino MA, Randall JB. The

long-term benefit of surgery on health-

related quality of life in patients with

inflammatory bowel disease. Archives of

surgery (Chicago, Ill.: 1960). 2001; 136:

521-7.

U. Ahmed Ali

T. Dunne

B. Gurland

J.D. Vogel

R.P. Kiran

Department of Colorectal Surgery, Digestive Disease Institute, Cleveland Clinic, Cleveland, Ohio, USA

3 actual versus estimated lenGth

of stay after colorectal

surGery – which factors

influence a deviation?

The American Journal of Surgery, 2013.

44 Chapter 3

absTracT

backgroundDue to increasing health costs, reimbursement has shifted from a fee-for-service to fixed

reimbursement schemes, such as the Diagnosis Related Groups (DRG). Studying the devia-

tions of actual outcome compared to those expected based on such scheme provides more

insights in costs and identifies potential target for saving. This study determines factors

associated with deviation in length of hospital stay (LOS) from that determined by DRG.

MethodsCohort study from a prospectively collected database including consecutive patients oper-

ated in a high-volume colorectal surgery department in 2009.

resultsFor 1,461 included patients, average expected and actual LOS were 8.17 (IQR 4.7-11.9)

and 8.31 (IQR 4-10) days, respectively. The most prominent factors associated with an

increase of LOS from expected were parenteral nutrition (5.11 days), emergency room

admittance (3.67) and ileus (3.45) (all p≤0.001). Other independently associated factors

included blood transfusion, anastomotic leak, sepsis, pulmonary embolism and surgeon.

Patients with a higher severity-illness-index and postoperative intensive care stay had a

lower than expected LOS.

conclusionsIdentifying predictors of deviation of LOS offers an opportunity to identify maximally cost-

effective targets for hospital cost savings. We identified multiple factors that may serve as

targets for specific cost savings measures. DRG related outcomes also offer a tool to more

accurately compare results from hospitals with different case-mix.

Chapter 3 45

3

inTroducTion

Colorectal surgical procedures carry an increased risk of certain complications, like post-

operative ileus and anastomotic leakage. These complications are known risk factors for

prolonged hospital stay.1-4 Studies have looked at the influence of these factors on the

absolute length of hospital stay5-7 (LOS), but correlating these differences with deviation

from what is expected by risk estimation models, like the Diagnosis Related Groups (DRG),

is yet to be done. DRG is a classification system intended to relate the type of patients a

hospital treats to the costs incurred by the hospital.8 In the United States DRG is used to

determine hospital payment for Medicare patients.9 Other similar systems exist around the

world.10-12

The DRG offers an estimation of the expected LOS based on preoperative patient charac-

teristics, type of surgery and postoperative complications. Trying to predict deviation from

this estimate has certain advantages. Such analysis offers insights to physicians and depart-

ments to which factors result in the greatest increase in cost. This allows for prioritization

of research and management goals.13-16 As an example, a previous study from our group

showed that mean direct hospital costs were significantly lower for laparoscopic compared

to open colectomy, although DRG classifications (and thus payment to hospitals) were

similar.17 Additionally, the DRG classification offers a reference value that is more reliable

than absolute LOS alone. Since the DRG classification corrects for the complexity of patient

population (i.e. case-mix), results can be better interpreted across different sources and are

thus more generalizable. Therefore, the aim of this study was to evaluate factors associated

with deviation of actual LOS from expected based on the DRG classification.

MaTerials and MeThods

study populationAfter study approval by the institutional review board (IRB), all consecutive patients oper-

ated on from January 12th to December 17th 2009 in the Department of Colorectal Surgery

at the Cleveland Clinic were included. Patients were selected based on procedural codes

for colorectal surgical procedures. Patients had to be 18 years or older. No exclusion criteria

were used.

study variablesClinical and administrative data respectively were obtained from the prospectively col-

lected clinical “Outcomes database” and departmental financial records. Clinical data col-

lected included patient characteristics, medical history, co-morbidities, smoking, laboratory

46 Chapter 3

profiles, operative details and postoperative outcome (actual LOS and 30 day-postoperative

complications). Patients were accrued into the database on the day of discharge. A prede-

termined dataset was collected, including diagnosis, laboratory tests, surgical details and

in-hospital complications. Anastomotic leak was defined as any clinically evident leak or

extravasation of contrast from the anastomosis detected by radiology studies. Prolonged

postoperative ileus was defined as absence of gastrointestinal function for longer than 5

days or the need for placement of a nasogastric tube. Other complications were also based

on specific definitions. All patients were contacted after discharge to obtain a complete

30-day follow-up. Administrative data included DRG classification, expected LOS, severity

of illness index (from 1 to 4 indicating respectively, minor, moderate, major, or extreme

severity of illness )8,18, and source of admittance (i.e. clinic referral, emergency room or

transfer).

data and statistical analysisDifference between actual and expected LOS was calculated for all patients. Variables

univariatly associated with deviation of LOS were included in a multivariate model. The

Wilcoxon rank sum test, Kruskal-Wallis test and the Spearman correlation test were used

for univariate analysis as appropriate. Linear regression analysis with forward stepwise

inclusion was used for the multivariate model. A p-value of 0.05 was used as cut-off point

for statistical significance.

resulTs

A total of 1,461 patients with a mean age of 50 (±18) years, and with 755 (52%) males

were included (Table 1). Stoma takedown was the indication for surgery in 361 (25%)

patients, cancer in 257 (18%), Crohn’s disease in 204 (14%) and ulcerative colitis in 200

(14%) patients. Patients were coded to 76 different surgical procedures. Most frequently

performed operations were enterostomy closure, laparoscopic and open partial colectomy,

exploratory laparotomy and laparoscopic total colectomy. About 1196 (82%) patients were

assigned to five distinct DRG codes (Table 1). The average expected LOS based on the DRG

classification was 8.17 (interquartile range 4.7 to 11.9) days, while the average actual

LOS was 8.31 (interquartile range 4 to 10) days. The average actual LOS was higher than

expected by 0.14 (interquartile range -2.2 to 1.8) days per patient. Actual hospital stay was

shorter than or equal to expected in 804 (55%) patients, and longer than expected in the

remaining 657 (45%). The proportion of patients that stayed at least five, seven or 10 days

longer than expected was 136 (9.3%), 91 (6.2%) and 41 (2.8%), respectively.

Chapter 3 47

3

Table 1. Patient characteristics

Characteristics Result

Age (years) [mean (SD)] 50 (18)

Sex (males) [n (%)] 755 (52%)

BMI [mean (SD)] 26.3 (6.2)

Expected LOS (days) [mean (SD)] 8.17 (4.97)

Actual LOS (days) [mean (SD)] 8.31 (6.58)

Actual - Expected LOS (days) [mean (SD)] 0.14 (4.69)

Diagnosis / Indication of surgery [n (%)]

Stoma removal 361 (25%)

Cancer 257 (18%)

Crohn’s disease 204 (14%)

Ulcerative colitis 200 (14%)

Diverticulitis 93 (6%)

Polyp 58 (4%)

Other 288 (20%)

ASA classification [n (%)]

1. No disturbance 31 (2%)

2. Mild disturbance 828 (57%)

3. Moderate disturbance 541 (37%)

4. Severe disturbance 58 (4%)

Surgical procedure [n (%)]

Closure enterostomy 339 (23%)

Laparoscopic partial colectomy 253 (17%)

Partial colectomy 152 (10%)

Exploratory laparotomy 88 (6%)

Laparoscopic total colectomy 79 (5%)

Adominoperineal resection 66 (5%)

Ileal pouch anal anastomosis 59 (4%)

Other 425 (29%)

DRG codes [n (%)]

Major bowel procedure w CC (330) 476 (33%)

Major bowel procedure w/o CC (331) 245 (17%)

Major bowel procedure w MCC (329) 238 (16%)

Minor bowel procedure w/o CC (346) 131 (9%)

Minor bowel procedure w CC (345) 106 (7%)

Rectal resection w CC (333) 42 (3%)

Other 223 (15%)

ASA: American Society of Anesthesiologists. BMI: body mass index. (M)CC: (major) complications and comor-bidites. LOS: length of hospital stay.

48 Chapter 3

Table 2. Univariate analysis of preoperative and intra-operative factors potentially associated with differences between actual and expected length of hospital stay

N n (%)* P-value

General characteristics

Sex (males) 1461 755 (52%) 0.54

Age at surgery 1455 - 0.74

BMI 1450 - 0.61

Diagnosis** 1458 - <0.001

ASA classification 1460 - 0.59

Preoperative characteristics

Diabetes 1460 120 (8%) 0.10

COPD 1460 24 (2%) 0.35

Dialysis 1460 9 (1%) 0.77

Smoking (current smokers) 1460 280 (19%) 0.046

Creatinine levels 1458 - 0.54

Albumin levels 1037 - 0.96

White blood count 1458 - 0.01

Hematocrit 1457 - 0.76

Platelet count 1457 - 0.41

History of chemotherapy 1460 61 (4%) 0.22

History of pelvic radiation 1460 43 (3%) 0.007

Surgical details

Surgical procedure 1460 - 0.01

Surgical approach‡ 1460 - <0.001

Operative time 1461 - 0.08

Type of intestinal anastomosis§ 1442 - 0.09

Creation of stoma 1459 597 (41%) 0.07

Blood transfusions (intra-operative) 1457 163 (11%) 0.74

Intra-operative complications

Any complication 1461 197 (13%) 0.56

Small bowel puncture or laceration 1460 158 (11%) 0.05

ASA: American Society of Anesthesiologists. BMI: body mass index. COPD: chronic obstructive pulmonary disease.* Number of patients with a certain characteristic (for dichotomous factors only).** Diagnosis of cancer was associated with an increase in LOS from expected; while diverticulitis and ulcerative colitis were associated with a decrease in LOS from expected.‡ Surgical approach is classified as open, laparoscopic, hand-assisted, or converted. Open surgery and conver-sion were associated with increase LOS from expected, laparoscopic and hand-assisted were associated with a decrease from expected.§ Type of intestinal anastomosis is classified as stapled anastomosis, hand sewn anastomosis, other techniques or no anastomosis.

Chapter 3 49

3

risk factors for deviation of hospital stayPreoperative and postoperative factors that were univariably associated with deviation

of LOS from expected are shown in Table 2 and Table 3, respectively. Diagnosis, type of

surgical approach (laparoscopic, open or hand-assisted), postoperative parenteral nutri-

tion, multiple postoperative complications, DRG classification and severity of illness were

all strongly associated with deviation from expected LOS (p<0.001). Additionally, other

factors showed sufficient association to be included in multivariate analysis. On multivari-

ate analysis, diagnosis, surgeon, postoperative recovery unit, use of parenteral nutrition,

several postoperative complications, DRG classification, source of admittance and severity

of illness were independently associated with deviation of LOS from expected (Table 4).

Postoperative factors were generally more strongly associated with a deviation of LOS than

preoperative factors. Need for parenteral nutrition, prolonged ileus, elevated creatinine

(serum creatinine >1 mg/dL), need for blood transfusions and anastomotic leaks were each

Table 3. Univariate analysis of administrative and postoperative factors potentially associated with difference between actual and expected length of hospital stay

N n (%)* P-value

Administrative characteristics

DRG Code 1461 - <0.001

Admitting Source** 1461 - 0.001

Severity of illness index‡ 1461 - <0.001

Postoperative recovery

Initial postoperative facility§ 1459 - 0.01

Need for parenteral nutrition 1460 181 (12%) <0.001

Postoperative complications

Urinary tract infection 1459 42 (3%) 0.002

Pulmonary embolism / DVT 1460 43 (3%) 0.001

Prolonged ileus 1460 223 (15%) <0.001

Blood transfusion 1460 268 (18%) 0.003

New arrhythmias 1460 38 (3%) 0.85

Deep SSI (organ space) 1460 103 (7%) <0.001

Wound infection 1459 139 (10%) 0.001

Anastomotic leak 1461 43 (3%) <0.001

Elevated creatinine (>1 mg/dL) 1460 53 (4%) 0.002

Sepsis 1460 113 (8%) <0.001

DRG: Diagnosis related groups. DVT: deep venous thrombosis. SSI: surgical site infection.* Number of patients with a certain characteristic (for dichotomous factors only).** Admitting source scored as outpatient clinic referral, emergency room, transfers from other hospital, trans-fers from other health care facility or unknown.‡ Severity of illness index classified into four categories minor, moderate, major and extreme severity.§ Initial postoperative facility classified as regular ward, post-aneasthesia care unit (overnight stay), intensive care unit.

50 Chapter 3

Table 4. Independent factors associated with differences between actual and expected length of hospital stay

N / n (%)* Estimate (SE)** P-value

General characteristics

Diagnosis

- Ulcerative colitis 200 (14%) -0.77 (0.30) 0.010

Surgical details

Surgeon‡ 1461 up to ± 1.25 0.017

Postoperative recovery

Initial postoperative facility 0.04

- Regular nursing ward 1386 (95%) Ref.

- ICU 59 (4%) -1.25 (0.60) 0.04

- PACU overnight admission 14 (1%) 1.17 (1.04) 0.27

Use of parenteral nutrition 181 (12%) 5.11 (0.37) <0.001

Postoperative complications

DVT/PE 43 (3%) 1.90 (0.62) 0.002

Prolonged ileus 223 (15%) 3.45 (0.30) <0.001

Blood transfusion (postoperatively) 268 (18%) 2.26 (0.31) <0.001

Wound infection 139 (10%) 0.79 (0.37) 0.03

Elevated creatinine (>1 mg/dL) 53 (4%) 2.49 (0.58) <0.001

Anastomotic leak 43 (3%) 2.15 (0.65) 0.001

Sepsis 113 (8%) 1.97 (0.44) <0.01

Administrative characteristics

DRG

- Rectal resection with CC (code 333) 42 (3%) 1.38 (0.60) 0.02

Admitting Source <0.001

- Outpatient clinic 1305 (89%) Ref.

- Emergency room 58 (4%) 3.67 (0.53) <0.001

- Other hospital 86 (6%) 1.32 (0.48) 0.006

- Other health care facility 10 (1%) -1.10 (1.22) 0.36

Severity <0.001

- Minor 417 (29%) Ref.

- Moderate 593 (41%) -1.06 (0.25) <0.001

- Major 282 (19%) -4.57 (0.34) <0.001

- Extreme 169 (12%) -7.56 (0.48) <0.001

CC: complications and co-morbidities. SE: standard error. SSI: surgical site infection.* For continuous outcomes, the number of patients valid for analysis. For dichotomous factors, number and percentage of patients with a certain characteristic.** The estimate indicates the covariate-adjusted deviation in days of LOS from expected for the respective factor compared to the deviation in the group not having the factor (or reference group in case of categorical variables).‡ Different surgeons were responsible for a deviation of up to ±1.25 days from the mean deviation of LOS observed in this cohort.

Chapter 3 51

3

associated with an increase of LOS of over 2 days from expected, compared to not having

these complications (Table 4). Additionally, sepsis, venous thromboembolism and wound

infections were independently associated with increases of LOS from expected.

Patients admitted from the emergency room and from other hospitals had an LOS that was

longer than expected by 3.67 (0.52) and 1.32 (0.48) days, respectively, compared to pa-

tients admitted through clinic referral. Of the 58 patients admitted through the emergency

room, only 3 (5%) had previous surgery in our department. Different surgeons were also

independently associated with a deviation of LOS from expected of up to ±1.25 days from

the mean deviation of LOS observed for all surgeons. Patients with increasing severity of ill-

ness had a shorter than expected LOS. Actual LOS for moderate, major and extreme sever-

ity was -1.06 (0.53), -4.57 (0.34) and -7.65 (0.48) days shorter than expected, respectively,

compared to minor severity. Patients admitted to the ICU as the initial postoperative facility

also stayed shorter than expected by -1.25 (1.04) days, compared to those discharged to

the regular nursing floor.

In order to evaluate whether the use of laparoscopic surgery was responsible for LOS

differences amongst surgeons, data were also specifically examined for patients who

underwent laparoscopic surgery (n=315). For these patients, the influence of surgeon was

not statistically significant, both in univariate and multivariate analysis. When data was

further scrutinized, there was however a large variation between different surgeons. Most

of this variation was between surgeons who regularly performed laparoscopic surgery and

those who did not. Of a total of 12 surgeons, 5 had performed less than 10 laparoscopic

surgeries each, while 7 performed over 20 laparoscopic surgeries during the study period.

When the 7 surgeons with over 20 laparoscopic surgeons were analyzed (patients n=292),

only a variation of about ±0.5 LOS days was observed.

discussion

This study examined the difference between actual and expected LOS based on the DRG

classification in a large colorectal surgical department. Overall, the DRG accurately indi-

cated the actual LOS with an average difference of 0.14 days (interquartile range -2.2

to 1.8). We did identify several important factors that were associated with significant

deviation of LOS from expected. In this regard, postoperative complications were found to

be more influential than pre- or intra-operative risk factors. Thorough examination of the

identified factors is necessary in order to evaluate potential ways of reducing their burden

on patients’ health and hospital costs.

52 Chapter 3

The need for parenteral nutrition and admittance from the emergency room resulted in the

largest increases from expected with an average covariate-adjusted increase of 5.1 and 3.5

days, respectively. Parenteral nutrition has previously been associated with postoperative

morbidity, due to hyperglycemia and infectious complications.19-21 A consequent reduction

in parenteral nutrition use in some disciplines demonstrated decreased complications and

hospital resource utilization.22 At our institution, parenteral nutrition is only administered

when necessitated by the patient’s condition. In the postoperative phase, our policy is

to initiate parenteral nutrition in patients with delayed recovery of gastrointestinal func-

tion or to continue this until the re-establishment of an adequate and consistent oral

intake that can maintain nutrition can be ascertained when patients on this preoperatively

undergo surgery. It is, therefore, likely a surrogate of protracted recovery. Consequently,

directly influencing this factor is probably difficult. A potential limitation of this study

is that we were unable to separate out if the timing of TPN, whether preoperative or

postoperative, influenced its associated with outcomes in terms of LOS. Although, it may

be expected that certain diagnoses, such as inflammatory bowel disease may be expected

to have an increased use of preoperative parenteral nutrition, in a post-hoc analysis, no

associations were seen between these diagnoses and longer than expected LOS. Further,

postoperative parenteral nutrition remained an important factor on multivariate analysis

even when accounting for both diagnosis and ileus. Thus, these results indirectly suggest

that preoperative parenteral nutrition may have had a limited influence, although this

may also have been due to the relatively low number of patients receiving the therapy

preoperatively in our study.

In a similar way, emergency room admittance is not an easy target for care providers to

reduce LOS. In tertiary centers serving a large urban area, emergency admittance is known

to contribute to increased LOS and expenses.23,24 The increased severity of patients present-

ing in the emergency room and time needed to perform tests that are routinely performed

outside the hospital are factors that may lead to increased LOS.24

Prolonged ileus was the postoperative complication associated with the largest increase in

LOS, averaging more than 3 days from expected. Due to the nature of colorectal surgery,

prolonged ileus is a fairly common complication.1,5 It also is particularly difficult to prevent.

We previously showed that two of the three independent predictors of ileus were unmodi-

fiable (i.e. age and previous abdominal surgery).1 The third factor (preoperative narcotics)

is a potential target for LOS reduction. A recent double-blind randomized trial showed that

a short course of systemic lidocaine postoperatively was successful in improving gastroin-

testinal motility and shortening of the LOS compared to placebo.25 Another double-blind

randomized trial showed that intravenous Ketorolac was able to reduce the rate of post-

operative ileus in patients undergoing laparoscopic colon surgery.16 The study also showed

Chapter 3 53

3

“promise in reducing hospital stay”, although the outcome was not statistically significant,

probably due to relatively small sample size (n=44).16

Other complications, like elevated creatinine, need for blood transfusion, anastomotic

leak, sepsis and deep-venous thrombosis were also associated with a substantial increase

in LOS. These postoperative complications are known for their deleterious influence on

clinical outcome, and several studies have investigated factors contributing to their occur-

rence.26-32 Our study highlights the importance of addressing these specific factors from a

cost perspective. Measures aimed at reducing these complications has the largest potential

impact for hospitals reimbursed through a DRG based system. Therefore, departments and

hospitals can prioritize and select their goals in a more informed manner.

Another potential factor that emerged was the role of the individual surgeon. Surgeon had

an important influence on deviation of LOS of up to ±1.25 days compared to the mean of all

surgeons. This suggests that modifiable factors likely exist and further supports a strategy of

identifying and targeting potential factors that influence outcomes and LOS based on best

evidence. Enhanced recovery (fast-track) programs may offer a way of reducing differences

in LOS between care providers. Such programs help decrease variations in care by different

physicians and have shown to reduce LOS.33-37 Although our patients are managed using

standardized care pathways, individual variations exist amongst surgeons for preoperative

and postoperative care and decisions in addition to any potential intra-operative differ-

ences. While the inclusion of a large group of patients undergoing different operations by

several surgeons serves the current aim of the study well, these very factors limit our ability

to study the specific role of surgeon on outcome. In particular, such factors as experience,

certain variations in pre and postoperative care decisions amongst surgeons, and other

specific factors such as surgical technique were not recorded. While the use of laparo-

scopic surgery seemed to minimize the influence of surgeon on LOS, differences amongst

surgeons persisted, with surgeons who regularly performed laparoscopic surgery tending

to have a lower variation in LOS than those who performed laparoscopy less commonly.

Future studies that prospectively collect such differences in pre-, intra-, and post-operative

factors between surgeons and their influence on LOS may help elucidate the reasons for

these differences in LOS.

Patients with severe illness and those initially admitted to ICU after surgery, i.e. patients

identified as potential high risk for complications, stayed at our hospital shorter than ex-

pected. The DRG classification allocates a larger margin of fluctuation in the expected LOS for

severe patients, which may in part be responsible for these results. In a post-hoc analysis, we

observed a strong relation between the presence of preoperative co-morbidities and direct

postoperative admittance to ICU. First, we found a highly significant association between

54 Chapter 3

ASA classification and rate of ICU admittance, with 0%, 1.2%, 5.0% and 36% of patients

being directly admitted to ICU for ASA 1, 2, 3 and 4, respectively (p<0.001). We also found

a significant association between direct postoperative ICU admittance and preoperative

work-up for co-morbidities, including having had a cardiac consult (p<0.001) or stress test

(p<0.001). These results suggest that admittance directly to ICU was indeed highly associated

with presence of known preoperative co-morbid states, and thus supports the explanation

that DRG classification currently may overestimate the length of stay in patients with such co-

morbidities. These results also suggests that our department performed better than expected

in this complex category of patients. This is in line with previous evidence showing that

tertiary high volume centers generally perform better than lower volume centers.5,38

In addition to identifying factors that are important for LOS and costs, this study also pro-

vides a potential way for comparing results from various institutions more reliably. Rhodes

et al. have shown that patient mix is probably the most important factor in determining

LOS.7 Schoetz et al. warned against accepting rigid criteria for length of stay for patients

undergoing colorectal surgery, since differences in clinical characteristics can require dif-

ferences in approach.39 Although the DRG can not explain individual patient variation of

LOS and thus is not suitable for detailed subgroup analysis7, it does provide a more reliable

way for comparing the overall LOS between different hospital. For example, in the study by

Faiz et al. comparing LOS after colorectal surgery in various English NHS hospitals, it was

observed that wide variation existed in the rates of prolonged stay at all levels of hospital

volume.5 Correlating such findings to a case-mix adjusted expectation can provide insight

whether these variations were mainly driven by differences in the patients case-mix or

more so by different practices between hospitals.

A potential limitation of this study is the retrospective nature of the analysis. Thus, some

potentially relevant factors identified in previous studies, such as delays time of presenta-

tion and socioeconomic status of patients, were not available for this analysis.5,6 Despite

this drawbacks, this study measures the relative effect of co-morbidities and complications

on LOS, providing another dimension to the importance of reduction of these events on

health costs not previously provided by estimates of absolute LOS alone.

conclusion

Identifying predictors of deviation of LOS offers an opportunity to identify maximally cost-

effective targets for hospital cost savings. We identified multiple factors that may serve as

targets for specific cost savings measures. DRG related outcomes also offer a tool to more

accurately compare results from hospitals with different case-mix.

Chapter 3 55

3

references

1. Kronberg U, Kiran RP, Soliman MS, Hammel

JP, Galway U, Coffey JC, Fazio VW. A char-

acterization of factors determining postop-

erative ileus after laparoscopic colectomy

enables the generation of a novel predictive

score. Ann Surg. 2011; 253: 78-81.

2. Retchin SM, Penberthy L, Desch C, Brown

R, Jerome-D’Emilia B, Clement D. Periop-

erative management of colon cancer under

medicare risk programs. Arch.Intern.Med.

1997; 157: 1878-84.

3. Bokey EL, Chapuis PH, Fung C, Hughes

WJ, Koorey SG, Brewer D, Newland RC.

Postoperative morbidity and mortality fol-

lowing resection of the colon and rectum

for cancer. Dis Colon Rectum. 1995; 38:

480-86.

4. Colorectal Cancer Collaborative G. Surgery

for colorectal cancer in elderly patients:

A systematic review. Lancet. 2000; 356:

968-74.

5. Faiz O, Haji A, Burns E, Bottle A, Kennedy R,

Aylin P. Hospital stay amongst patients un-

dergoing major elective colorectal surgery:

Predicting prolonged stay and readmissions

in nhs hospitals. Colorectal Dis. 2011; 13:

816-22.

6. Reddy KM, Meyer CE, Palazzo FF, Conaghan

P, Blunt MC, Stebbings WS, Leicester RJ, et

al. Postoperative stay following colorectal

surgery: A study of factors associated with

prolonged hospital stay. Ann R Coll.Surg

Engl. 2003; 85: 111-14.

7. Rhodes RS, Sharkey PD, Horn SD. Effect of

patient factors on hospital costs for major

bowel surgery: Implications for managed

health care. Surgery. 1995; 117: 443-50.

8. Averill RF, Goldfield N, Hughes JS, Bonazelli

J, McCullough EC, Steinbeck BA, Mullin R,

et al. All patient refined diagnosis related

groups (apr-drgs). Methodology overview.

Wallingford, CT, USA: 3M Health Informa-

tion System; 2003.

9. Averill RF, Kalison MJ, Vertrees JC, Goldfield

NI. Achieving short-term medicare savings

through the expansion of the prospective

payment system. Health Care Manage.Rev.

1996; 21: 18-25.

10. Pirson M, Schenker L, Martins D, Dung D,

Chale JJ, Leclercq P. What can we learn

from international comparisons of costs by

drg? Eur J Health Econ. 2012.

11. Trocchi P, Kluttig A, Dralle H, Sekulla C, Bier-

mann M, Stang A. Thyroid cancer surgery in

germany: An analysis of the nationwide drg

statistics 2005-2006. Langenbecks Arch

Surg. 2012; 397: 421-28.

12. Oostenbrink JB, Rutten FF. Cost assessment

and price setting of inpatient care in the

netherlands. The dbc case-mix system.

Health Care Manag.Sci. 2006; 9: 287-94.

13. Christensen HK, Thaysen HV, Rodt SA,

Carlsson P, Laurberg S. Short hospital stay

and low complication rate are possible with

a fully implemented fast-track model after

elective colonic surgery. Eur.Surg Res. 2011;

46: 156-61.

14. Kehlet H, Mogensen T. Hospital stay of 2

days after open sigmoidectomy with a

multimodal rehabilitation programme. Br.J

Surg. 1999; 86: 227-30.

15. Murphy ME, Noetscher CM. Reducing hos-

pital inpatient lengths of stay. J Nurs.Care

Qual. 1999; Spec No: 40-54.

16. Schlachta CM, Burpee SE, Fernandez C,

Chan B, Mamazza J, Poulin EC. Optimizing

recovery after laparoscopic colon surgery

(oral-cs): Effect of intravenous ketorolac on

length of hospital stay. Surg Endosc. 2007;

21: 2212-19.

17. Senagore AJ, Brannigan A, Kiran RP, Brady

K, Delaney CP. Diagnosis-related group

assignment in laparoscopic and open

colectomy: Financial implications for payer

and provider. Dis Colon Rectum. 2005; 48:

1016-20.

56 Chapter 3

18. Horn SD, Sharkey PD, Chambers AF, Horn

RA. Severity of illness within drgs: Impact

on prospective payment. Am J Public

Health. 1985; 75: 1195-99.

19. Pasquel FJ, Smiley D, Spiegelman R, Lin E,

Peng L, Umpierrez GE. Hyperglycemia is

associated with increased hospital com-

plications and mortality during parenteral

nutrition. Hosp.Pract (Minneap.). 2011; 39:

81-88.

20. Sena MJ, Utter GH, Cuschieri J, Maier RV,

Tompkins RG, Harbrecht BG, Moore EE, et

al. Early supplemental parenteral nutrition

is associated with increased infectious com-

plications in critically ill trauma patients. J

Am Coll.Surg. 2008; 207: 459-67.

21. Ishizuka M, Nagata H, Takagi K, Kubota

K. Total parenteral nutrition is a major risk

factor for central venous catheter-related

bloodstream infection in colorectal cancer

patients receiving postoperative chemo-

therapy. Eur.Surg Res. 2008; 41: 341-45.

22. Rhee P, Hadjizacharia P, Trankiem C, Chan

L, Salim A, Brown C, Green D, et al. What

happened to total parenteral nutrition? The

disappearance of its use in a trauma inten-

sive care unit. J Trauma. 2007; 63: 1215-22.

23. Grannemann TW, Brown RS, Pauly MV. Es-

timating hospital costs. A multiple-output

analysis. J Health Econ. 1986; 5: 107-27.

24. Thorpe KE. Why are urban hospital costs

so high? The relative importance of patient

source of admission, teaching, competi-

tion, and case mix. Health Serv.Res. 1988;

22: 821-36.

25. Herroeder S, Pecher S, Schonherr ME,

Kaulitz G, Hahnenkamp K, Friess H, Bottiger

BW, et al. Systemic lidocaine shortens length

of hospital stay after colorectal surgery: A

double-blinded, randomized, placebo-con-

trolled trial. Ann Surg. 2007; 246: 192-200.

26. Causey MW, Maykel JA, Hatch Q, Miller S,

Steele SR. Identifying risk factors for renal

failure and myocardial infarction following

colorectal surgery. J Surg Res. 2011; 170:

32-37.

27. El-Gazzaz G, Geisler D, Hull T. Risk of

clinical leak after laparoscopic versus open

bowel anastomosis. Surg Endosc. 2010; 24:

1898-903.

28. Kiran RP, da Luz MA, Remzi FH, Church

JM, Lavery I, Hammel J, Fazio VW. Factors

associated with septic complications after

restorative proctocolectomy. Ann Surg.

2010; 251: 436-40.

29. Kooby DA, Stockman J, Ben-Porat L, Gonen

M, Jarnagin WR, DeMatteo RP, Tuorto S,

et al. Influence of transfusions on periop-

erative and long-term outcome in patients

following hepatic resection for colorectal

metastases. Ann Surg. 2003; 237: 860-69.

30. Ondrula DP, Nelson RL, Prasad ML, Coyle

BW, Abcarian H. Multifactorial index of

preoperative risk factors in colon resec-

tions. Dis Colon Rectum. 1992; 35: 117-22.

31. Ortega-Deballon P, Radais F, Facy O, d’Athis

P, Masson D, Charles PE, Cheynel N, et al.

C-reactive protein is an early predictor of

septic complications after elective colorec-

tal surgery. World J Surg. 2010; 34: 808-14.

32. Stender MT, Nielsen TS, Frokjaer JB, Larsen

TB, Lundbye-Christensen S, Thorlacius-

Ussing O. High preoperative prevalence

of deep venous thrombosis in patients

with colorectal cancer. Br.J Surg. 2007; 94:

1100-03.

33. Lovely JK, Maxson PM, Jacob AK, Cima RR,

Horlocker TT, Hebl JR, Harmsen WS, et al.

Case-matched series of enhanced versus

standard recovery pathway in minimally

invasive colorectal surgery. Br.J Surg. 2011.

34. Lee Y, Fleming FJ, Deeb AP, Gunzler D,

Messing S, Monson JR. A laparoscopic

approach reduces short-term complications

and length of stay following ileocolic resec-

tion in crohn’s disease: An analysis of out-

comes from the nsqip database. Colorectal

Dis. 2011.

35. Feroci F, Kroning KC, Lenzi E, Moraldi

L, Cantafio S, Scatizzi M. Laparoscopy

within a fast-track program enhances the

Chapter 3 57

3

short-term results after elective surgery for

resectable colorectal cancer. Surg Endosc.

2011; 25: 2919-25.

36. Muller S, Zalunardo MP, Hubner M, Clavien

PA, Demartines N. A fast-track program

reduces complications and length of

hospital stay after open colonic surgery.

Gastroenterology. 2009; 136: 842-47.

37. Kehlet H. Fast-track colorectal surgery.

Lancet. 2008; 371: 791-93.

38. Mehta RH, Liang L, Karve AM, Hernandez

AF, Rumsfeld JS, Fonarow GC, Peterson ED.

Association of patient case-mix adjustment,

hospital process performance rankings, and

eligibility for financial incentives. JAMA.

2008; 300: 1897-903.

39. Schoetz DJ, Jr., Bockler M, Rosenblatt MS,

Malhotra S, Roberts PL, Murray JJ, Coller

JA, et al. “Ideal” length of stay after col-

ectomy: Whose ideal? Dis Colon Rectum.

1997; 40: 806-10.

R.P. Kiran

U. Ahmed Ali

P.J. Nisar

W. Khoury

J. Gu

B. Shen*

F.H. Remzi

J.P. Hammel

I.C. Lavery

V.W. Fazio

J.R. Goldblum**

Departments of Colorectal Surgery, *Gastroenterology and **Anatomic Pathology, Digestive Disease Institute, Cleveland Clinic, Cleveland, Ohio, USA

4 risk and location of cancer

in patients with preoperative

colitis-associated dysplasia

underGoinG proctocolectomy

Annals of Surgery, 2014.

60 Chapter 4

absTracT

backgroundThe ability to predict the presence and location of cancer in colitis patients with dysplasia

is essential to facilitate recommendations regarding the necessity and type of surgery. The

aim of this study is to evaluate the influence of preoperative dysplasia grade, appearance

and site on risk and location of cancer in patients with colitis.

MethodsUlcerative and indeterminate colitis patients who underwent proctocolectomy for dysplasia

were retrospectively selected. Patient characteristics and findings at colonoscopic surveil-

lance were associated with findings on the surgical specimen by regression analysis.

resultsFrom 1984-2007, 348 proctocolectomy specimens with preoperative dysplasia showed

cancer in 51 (15%) and dysplasia in 172 (49%) cases. Patients with preoperative high-

grade dysplasia (HGD) had cancer in 29% compared to 3% in LGD (p<0.001). Patients with

preoperative dysplasia-associated lesion/mass (DALM) had cancer in 25% compared to 8%

in flat dysplasia (p<0.001). In LGD with DALM, the risk of cancer was not significantly

higher that in flat LGD (7% vs. 2%, p=0.3), but risk of cancer or HGD showed a a three-

fold increase (29% vs. 9%, p=0.015). On multivariate analysis, HGD, DALM and disease

duration were independent risk factors for postoperative cancer. In patients with isolated

colonic dysplasia above the sigmoid level, postoperative rectal involvement was limited.

conclusionsRisk of cancer for patients with HGD or DALM is substantial. Despite low risk of cancer

in patients with flat LGD, threshold for surgery should be low given the high prevalence

of postoperative pathologic findings. Only in selected cases, colonoscopic surveillance

after discussion of associated risks may be acceptable, provided high patient compliance.

Surgery remains the only modality that can eliminate the risk of cancer. The location of

preoperative dysplasia may allow for the clarification of the need for proctectomy especially

in the poor risk surgical patient.

Chapter 4 61

4

inTroducTion

Cancer in inflammatory bowel disease (IBD), is thought to develop through a sequence of

chronic inflammation, dysplasia and cancer.1-3 There is, however, significant variation in

the reported rates of progression.4-9 Dysplasia, especially high-grade dysplasia (HGD) and

dysplasia-associated lesions or masses (DALM) on colonoscopy, are associated with a high

risk of concomitant cancer of up to 58%.10-12 The correlation between low-grade dysplasia

(LGD) and cancer, on the other hand, is less clear. The reported risk of synchronous cancer

varies between 0 and 19%, and progression rates to HGD or cancer range between 0.5

and 54%.6-8,13-15 Considering these contrasting results, treatment strategies for patients

with LGD remain controversial, including either a low threshold for surgery or medical

management with close colonoscopic surveillance.6,9,12,13,16

Correlation of the type and location of dysplasia on colonoscopy with the type and site of

pathological findings in the resected specimen is not well studied. Such knowledge can

help guide decision-making regarding the optimal surgical procedure.17 For patients with

HGD or colorectal cancer, a total proctocolectomy has been widely adopted as the preferred

surgical treatment.18,19 However, whether less extensive resections are a safe alternative in

selected cases and whether removal of the anorectal mucosa should be routinely added in

patients with dysplasia are not clear.20

The aims of this study are to assess the frequency and location of dysplasia and cancer in proc-

tocolectomy specimens in patients diagnosed with preoperative dysplasia, and to identify risk

factors associated with the presence of colorectal cancer and HGD, in an attempt to provide

physicians with better estimates of risk and to aid decision making in this complex disease.

PaTienTs and MeThods

study populationPatients with ulcerative or indeterminate colitis undergoing proctocolectomy for preopera-

tive dysplasia at our institution from 1984-2007 were identified. Patients undergoing an

ileal pouch anal anastomosis were retrospectively identified from a prospective institutional

review board (IRB) approved pouch database. Patients undergoing total proctocolectomy

with end ileostomy were identified by searching the pathology database, hospital bill-

ing records and operating lists in our department. Additional data was collected from

endoscopy, pathology reports and patient charts. Presence of dysplasia was determined

from preoperative endoscopic and histopathology reports. Patients with lesions classified

as indefinite for dysplasia (IND) were included, while non-dysplastic lesions were excluded.

62 Chapter 4

study variables and definitionsPatient and disease characteristics including age, gender, diagnoses, duration and extent of

disease, primary sclerosing cholangitis (PSC) and surgical details were evaluated. In order to

assess whether control of colitis was associated with cancer risk, the presence or absence

of symptoms as a risk factor was also assessed. Symptomatic colitis was defined as the

presence of active disease with symptoms of diarrhea or rectal bleeding in the preoperative

period. PSC was defined as the presence of intra- or extrahepatic bile duct abnormalities

in the form of beading, duct ectasia, and stricturing of the intra- or extrahepatic bile ducts

documented in the medical record from endoscopic retrograde cholangiopancreatography,

magnetic resonance cholangiopancreatography, and/or liver biopsy. Preoperative dyspla-

sia was classified according to grade (LGD, HGD and IND), appearance (flat, DALM or

adenoma-like lesions (ALM)) and focality (unifocal or multifocal).21 DALM was defined as

any mass or plaque-like lesion within inflamed areas of the colorectal mucosa, and ALM

as any mass that has the appearance of sporadic adenomas in non-inflamed areas of the

colon or rectum.22 As part of the standard preoperative workflow at our hospital, before

the confirmation or rejection of a diagnosis of dysplasia, all histopathology slides includ-

ing biopsies from colonoscopies performed elsewhere, are examined by two experienced

gastrointestinal pathologists and the findings discussed until an agreement is reached.

Location of dysplasia was classified as rectal, colonic or both (colonic and rectal). The

last 15 cm above the anal verge was considered to be the rectum. Colonic location was

categorized into cecum, ascending, transverse, descending and sigmoid colon. Postopera-

tive lesions located in colonic segments adjacent to preoperative dysplasia location were

considered to be in the same segment due to the potential limitation of endoscopy in

accurately identifying colonic segments. Dysplasia surveillance in our IBD Clinic has fol-

lowed the initial American College of Gastroenterology (ACG) Practicing Guidelines since

1997.23 After 8-10 years of colitis, patients typically underwent surveillance colonoscopy

every 1- 3 years with multiple biopsies from each segment of the colon and rectum with a

total minimum of 32 tissue samples. For polypoid or mass lesions, additional biopsies were

taken. The practice pattern before the publication of the ACG guideline varied amongst

gastroenterologists. Chromoendoscopy is occasionally used, but is not considered routine

practice in our clinic. As expected, equipment has improved gradually, with newer scopes

that allowed the use of narrow band imaging acquired in 2008. However, there was no

major change in how the surveillance was performed until 2008.

study outcomesThe main study outcome was histopathology of the colectomy specimen, which was classi-

fied as dysplasia (HGD, LGD or IND), cancer or neither. Colectomy specimens were handled

in a standardized manner. Sections taken from the proximal and distal surgical margin,

Chapter 4 63

4

every 10 cm throughout the specimen and from any gross lesions. Additional sections were

taken from the area of preoperative dysplasia.

data and statistical analysisCorrelations between grade, appearance and site of dysplasia and postoperative findings

were examined. Factors associated with the finding of cancer and cancer or HGD in the

surgical specimen were assessed by univariate and multivariate analysis. Risk factors evalu-

ated included age, gender, diagnoses, grade, appearance and multifocality of preoperative

dysplasia, duration of disease, symptomatic disease, extent of colitis and PSC. Categorical

variables were summarized as numbers and percentages, and quantitative variables as

means and standard deviations (SD) or medians with inter-quartile ranges (IQR), according

to normality. Univariate associations were analyzed by Chi-square, Wilcoxon rank sum,

Kruskal-Wallis tests and logistic regression, as appropriate. Logistic regression with step-

wise backward selection of variables was used for the multivariate models.

resulTs

Patient population and overall findingsWe included 348 (235 male) colitis patients diagnosed with preoperative dysplasia. Dyspla-

sia was the primary indication for surgery in 291 (83%) patients, and a secondary indication

in 57 (17%). Median age at surgery was 46 (IQR 38 to 57) years and median duration of

disease was 14.6 (IQR 8.7 to 20.9) years. Overall, 282 (81%) patients had ulcerative colitis

and 66 (19%) indeterminate colitis on final pathology. Three hundred thirty two (95%)

patients underwent restorative proctocolectomy, while 16 (5%) patients underwent total

proctocolectomy (TPC) with end ileostomy. The most common reasons for TPC with end

ileostomy were sphincter impairment, incontinence, older age and patient choice. Disease

was symptomatic in 282 (81%) patients and PSC diagnosed in 60 (17%) patients. Divided

into three equal periods (1984 – 1992, 1992 – 2000 and 2000 – 2007), the numbers of

operated patients were 71 (20%), 143 (41%) and 134 (39%) patients, respectively.

Postoperative pathology showed colorectal cancer in 51 (15%) patients, dysplasia in 172

(49%) and no cancer or dysplasia in the remaining 125 (36%). Of the 51 patients with

cancer, 11 (22%) had intramucosal adenocarcinoma, 17 (35%) had stage I disease, 9

(18%) had stage II, 11 (22%) had stage III, with the stage not well documented in 3

patient. Dysplasia on final pathology was graded as HGD in 100 (58%) patients, LGD in 59

(34%) and IND in 13 (8%).

64 Chapter 4

Preoperative dysplasia and postoperative findingsPreoperative dysplasia was HGD in 159 (46%), LGD in 136 (39%) and IND in 53 (15%)

patients (Table 1). Symptomatic disease, DALM appearance and multifocal dysplasia were

more frequently seen in patients with HGD (Table 1). Patients with HGD had a 29% risk of

malignancy in the surgical specimen, compared to 3% in patients with LGD (p<0.001). The

risk of HGD in the proctocolectomy specimen was 49% in patients with preoperative HGD,

compared to 12% for patients with preoperative LGD or IND (p<0.001).

In 316 (91%) patients, the appearance of dysplasia during endoscopy was documented.

Dysplasia was flat in 118 (37%) patients, appeared as DALM in 149 (47%) and as ALM

in the remaining 49 (16%) patients. Preoperative HGD and multifocal dysplasia were

significantly more common in DALM. Cancer on final pathology was significantly more

Table 1. Patient characteristics and pathological outcome by preoperative dysplasia grade

High GradeDysplasia(N=159)

Low Grade Dysplasia(N=136)

Indefinite for Dysplasia

(N=53)

P-value

Age in years (median-IQR) 46 (38 - 58) 45 (37 - 56) 47 (37 - 58) 0.5

Gender (male) 112 (70%) 89 (65%) 34 (64%) 0.6

Duration of disease in years (median-IQR)

15.5 (10.0 – 21.3) 13.8 (8.4 – 21.9) 11.7 (7.0 – 20.0) 0.1

Symptomatic disease 134 (84%) 102 (75%) 46 (87)% 0.05

Primary sclerosing cholangitis 35 (22%) 18 (13%) 7 (13%) 0.1

Extent of Disease

0.4

Pancolitis 120 (76%) 88 (65%) 38 (72%)

Left sided colitis 16 (10%) 22 (16%) 9 (16%)

Proctitis / Proctosigmoiditis 6 (4%) 9 (7%) 3 (6%)

Other / inactive 17 (11%) 17 (13%) 3 (6%)

Appearance of dysplasia

<0.001Flat 36 (23%) 56 (41%) 26 (49%)

DALM 95 (60%) 42 (31%) 12 (23%)

ALM 18 (11%) 23 (17%) 8 (15%)

Multifocal dysplasia 94 (59%) 46 (34%) 12 (23%) <0.001

Final pathology

Cancer 46 (29%) 4 (3%) 1 (2%) <0.001

HGD 78 (49%) 16 (12%) 6 (12%) <0.001

LGD 9 (6%) 46 (34%) 4 (8%) <0.001

Indefinite for dysplasia 3 (2%) 5 (4%) 5 (9%) 0.04

No dysplasia or cancer 23 (15%) 65 (48%) 36 (70%) <0.001

ALM: adenoma like mass. DALM: dysplasia associated lesion or mass. IQR: Interquartile range. HGD: high grade dysplasia. LGD: low grade dysplasia.

Chapter 4 65

4

frequent in DALM compared to flat dysplasia and ALM (25%, 8%, 0.4%, respectively,

p<0.001). Postoperative HGD was also more common in DALM compared to other types

of dysplasia appearance (41%, 19% and 29%, respectively, p<0.001). In flat dysplasia,

multifocality was not associated with a higher risk of cancer (12% vs. 4%, p=0.2), but it

was significantly associated with the risk of cancer or HGD (37% vs. 19%, p=0.03).

For patients with LGD and HGD, characteristics of patients and pathological outcome for

flat dysplasia, DALM and ALM are presented in Table 2. Risk of cancer in both flat and

DALM HGD (22% and 35%, respectively) was significantly higher than either flat or DALM

LGD (2% and 7%, respectively) (p<0.001). Patients with preoperative LGD with DALM

had a much higher risk of having pathological findings in the postoperative specimen

(p<0.001) compared to patients with flat LGD or ALM. The risk for cancer and/or HGD

was significantly higher in patients with LGD with DALM compared with flat LGD (29% vs.

9% patients, p=0.015). In flat LGD, multifocal dysplasia was not associated with cancer

risk (5% vs. 0%, p=0.4). For patients with HGD, extent of disease was greater in flat and

DALM dysplasia when compared to ALM. Risk of cancer was highest in DALM HGD (35%),

followed by flat HGD (22%) and ALM (6%) (p=0.08).

site of preoperative dysplasiaFor 336 patients with available data, 201 (58%) patients had dysplasia limited to the

colon, 82 (24%) had rectal dysplasia and 53 (15%) had colonic and rectal dysplasia (Table

3). Preoperative HGD and DALM were more frequent with combined colonic and rectal

dysplasia. Risk of cancer on final pathology was highest in the combined colonic and rectal

dysplasia group. For 215 patients, the exact location of both preoperative and postopera-

tive dysplasia within the colon was available (Table 3). Despite a high rate of concordance

between the location of preoperative and postoperative pathological findings, postopera-

tive lesions (dysplasia or cancer) were present in remote segments that were not similar or

adjacent to the segment of preoperative dysplasia in 11% (24 / 215) of patients. In patients

with preoperative dysplasia limited to the colon (n=201), rectal dysplasia on final pathology

was present in 13 (7%) patients and rectal cancer was seen in three patients (1.5%). When

only patients with colonic involvement above the level of the sigmoid were assessed, the

rates of rectal dysplasia and cancer were 3% (3/98) and 0% (0/98), respectively.

In patients with preoperative dysplasia limited to the colon (n=201), 41 (20%) had a muco-

sectomy with a handsewn anastomosis. None of these patients had evidence of dysplasia

in the mucosectomy specimen. On the other hand, of the 135 patients with rectal dysplasia

preoperatively, 56 (41%) patients underwent mucosectomy and handsewn anastomosis.

Nine (16%) of these patients had evidence of dysplasia in the mucosectomy specimen;

HGD was seen in 4 (7%) patients, LGD in 3 (5%) and indefinite dysplasia in 2 (4%). The

66 Chapter 4

Tab

le 2

. Pat

ien

t ch

arac

teri

stic

s an

d p

ath

olo

gic

al o

utc

om

e b

y p

reo

per

ativ

e d

ysp

lasi

a ap

pea

ran

ce

Low

gra

de

dys

pla

sia

Hig

h g

rad

e d

ysp

lasi

a

Flat

(N=

56)

DA

LM(N

=42

)A

LM(N

=23

)P-

valu

eFl

at(N

=36

)D

ALM

(N=

95)

ALM

(N=

18)

P-va

lue

Ag

e in

yea

rs (

med

ian

-IQ

R)

45 (3

7-54

)47

(40-

57)

42 (3

5–57

)0.

746

(34–

54)

45 (3

8–57

)62

(43–

73)

0.01

Gen

der

(m

ale)

32 (5

7%)

35 (8

3%)

13 (5

7%)

0.01

26 (7

1%)

68 (7

2%)

14 (7

8%)

0.9

Du

rati

on

of

dis

ease

in y

ears

(m

edia

n-I

QR

)13

.0(8

.9–1

9.7)

16.2

(7.8

-29.

8)10

.8(5

.1–2

0.7)

0.3

15.5

(10.

0–22

.6)

15.8

(10.

2–21

.9)

14.9

(8.9

–17.

1)0.

5

Sym

pto

mat

ic d

isea

se42

(75%

)29

(69%

)20

(87%

)0.

331

(86%

)83

(88%

)11

(61%

)0.

6

Prim

ary

scle

rosi

ng

ch

ola

ng

itis

9 (1

6%)

5 (1

2%)

2 (9

%)

0.6

4 (1

1%)

24 (2

5%)

3 (1

7%)

0.2

Exte

nt

of

Dis

ease

0.03

0.2

Panc

oliti

s45

(81%

)20

(48%

)14

(61%

)29

(81%

)73

(77%

)11

(61%

)

Left

sid

ed c

oliti

s4

(7%

)11

(26%

)4

(17%

)2

(6%

)8

(8%

)5

(28%

)

Proc

titis

/ Pr

octo

sigm

oidi

tis2

(4%

)3

(7%

)3

(13%

)3

(8%

)1

(1%

)1

(6%

)

Oth

er /

inac

tive

5 (9

%)

8 (1

9%)

2 (9

%)

2 (6

%)

13 (1

4%)

1 (6

%)

Mu

ltif

oca

l dys

pla

sia

21 (3

8%)

19 (4

5%)

6 (2

6%)

0.3

21 (5

8%)

60 (6

3%)

9 (5

0%)

0.6

Fin

al p

ath

olo

gy

Can

cer

1 (2

%)

3 (7

%)

0 (0

%)

0.2

8 (2

2%)

34 (3

5%)

1 (6

%)

0.08

HG

D4

(7%

)9

(21%

)3

(13%

)0.

116

(44%

)50

(52%

)11

(61%

)0.

5

LGD

18 (3

2%)

22 (5

2%)

4 (1

7%)

0.01

3 (8

%)

3 (3

%)

1 (6

%)

0.5

Inde

finite

for

dys

plas

ia3

(5%

)1

(2%

)0

(0%

)0.

41

(3%

)1

(1%

)1

(6%

)0.

4

No

dysp

lasi

a or

can

cer

30 (5

4%)

7 (1

7%)

16 (7

0%)

<0.

001

8 (2

2%)

9 (1

0%)

4 (2

2%)

0.03

ALM

: Ade

nom

a lik

e le

sion

. DA

LM: d

yspl

asia

ass

ocia

ted

lesi

on o

r m

ass.

IQR:

Inte

rqua

rtile

ran

ge.H

GD

: hig

h gr

ade

dysp

lasi

a. L

GD

: low

gra

de d

yspl

asia

.

Chapter 4 67

4

Table 3. Patient characteristics and pathological outcome by location of preoperative dysplasia

Colonicdysplasia(n=201)

Rectaldysplasia

(n=82)

Colonic & Rectal dysplasia

(n=53)P-value

Age at Surgery- years[median (IQR)]

46 (38 - 56) 49 (40 - 59) 45 (34 - 54) 0.2

Gender (male) 134 (67%) 55 (67%) 40 (76%) 0.5

Duration of Disease- years [median (IQR)]

15.3 (9.4 - 22.0) 15.4 (9.0 - 22.3) 13.3 (8.0 - 18.4) 0.2

Symptomatic disease 172 (86%) 55 (67%) 45 (85%) 0.001

Primary sclerosing cholangitis 36 (18%) 7 (9%) 14 (26%) 0.02

Extent of Disease

0.01

Pancolitis 145 (72%) 50 (61%) 42 (79%)

Left sided colitis 24 (12%) 15 (18%) 6 (11%)

Proctitis / Proctosigmoiditis 7 (4%) 10 (12%) 0 (0%)

Other / inactive 25 (12%) 7 (9%) 5 (10%)

Preoperative dysplasia Grade

0.008HGD 85 (42%) 37 (45%) 37 (70%)

LGD 86 (43%) 31 (38%) 13 (25%)

Indefinite 30 (15%) 14 (17%) 3 (6%)

Appearance of dysplasia

<0.001Flat 82 (41%) 26 (32%) 9 (17%)

DALM 68 (34%) 45 (55%) 36 (68%)

ALM 36 (18%) 9 (11%) 4 (8%)

Multifocal dysplasia 88 (44%) 11 (13%) 53 (100%) <0.001

Pouch Anastomosis Type <0.001

Handsewn 41 (20%) 31 (38%) 25 (47%)

Final pathology

Cancer 30 (15%) 7 (9%) 14 (26%) 0.02

HGD 55 (27%) 24 (29%) 20 (38%) 0.3

LGD 35 (17%) 15 (18%) 9 (17%) 1.0

Indefinite for dysplasia 8 (4%) 3 (4%) 1 (2%) 0.8

No dysplastic changes 73 (36%) 33 (40%) 9 (17%) 0.04

Postoperative site related to preoperative site*

117 (97%) 50 (100%) 43 (100%) 0.2

Presence of remote postoperative dysplasia sites*

19 (16%) 1 (2%) 4 (9%) 0.03

68 Chapter 4

four patients with HGD in the mucosectomy specimen also had HGD in the preoperative

colonoscopy biopsies. Details of the location of preoperative dysplasia within the rectum

were available in 45 patients (lower rectum:13, mid rectum: 12, upper rectum: 20). In

patients with dysplasia in the lower rectum, eight patients (7 with HGD and 1 with LGD)

underwent mucosectomy and handsewn anastomosis. Two (25%) patients had evidence

of dysplasia (1 HGD and 1 LGD) in the mucosectomy specimen. In patients with dysplasia

of the mid rectum, 6 patients (3 with HGD, 2 with LGD and 1 indefinite for dysplasia)

underwent mucosectomy and handsewn anastomosis. The mucosectomy specimen of

one (16%) patient showed lesions indefinite for dyplasia. In patients with dysplasia in the

upper rectum, 4 patients (2 with HGD, 2 with LGD) underwernt mucosectomy. One (25%)

patient with LGD preoperatively had LGD in the mucosectomy specimen.

risk factors for postoperative cancer and dysplasiaUnivariate analysis of factors associated with the presence of cancer in the colectomy

specimen in patients with preoperative dysplasia is shown in Figure 1. Presence of can-

cer in the colectomy was significantly associated with duration of disease, symptomatic

disease, presence of preoperative HGD, DALM, multifocal dysplasia and combined colon

and rectum site of dysplasia. These factors were included in a multivariate analysis (Table

4). Disease duration, HGD and DALM were found to be independent predictive factors for

cancer in this model.

Factors associated with the presence of cancer or HGD in the colectomy specimen were also

analyzed (Figure 2). There was a significant association for this outcome with symptomatic

Table 3. (continued)

Colonicdysplasia(n=201)

Rectaldysplasia

(n=82)

Colonic & Rectal dysplasia

(n=53)P-value

Site of postoperative pathological findings**

<0.001

Cecum 45 (22%) 0 (0%) 12 (23%)

Ascending 35 (17%) 0 (0%) 11 (21%)

Transverse 47 (23%) 1 (1%) 15 (28%)

Descending 41 (20%) 1 (1%) 19 (36%)

Sigmoid 54 (27%) 3 (4%) 22 (43%)

Rectum 13 (7%) 47 (57%) 35 (66%)

ALM: adenoma like mass. DALM: dysplasia associated lesion or mass. IQR: Interquartile range.HGD: high grade dysplasia. LGD: low grade dysplasia.* Percentages are based on patients in which the exact location of dysplasia is documented both pre- and postoperatively (n =215).** Some patients had multiple lesions in different locations, while others did not have any pathological findings. Percentages are of total number of patients.

Chapter 4 69

4

Figure 1. Factors associated with cancer in patients with colitis undergoing proctocolectomy for dysplasia

Figure 2. Factors associated with cancer and/or dysplasia in patients with colitis undergoing procto-colectomy for dysplasia

70 Chapter 4

disease, extensive inflammation (i.e. pancolitis), preoperative HGD, DALM, multifocality

and combined colon and rectum site dysplasia. There was also a trend towards an as-

sociation with duration of disease, but it did not reach statistical significance (p=0.1).

Multivariate analysis identified HGD and DALM as independent predictive factors for this

outcome (Table 4).

Table 4. Multivariate analysis for risk factors of cancer and high grade dysplasia in patients with preoperative dysplasia

Multivariate analysisOdds ratio

(95% Confidence interval)P-value

Risk factors for cancer

Dysplasia grade

<0.001- LGD 1.32 (0.14 - 12.54)

- HGD 12.09 (1.57 - 93.19)

DALM 3.08 (1.39 – 6.80) 0.005

Duration of disease (10 years) 1.53 (1.05 - 2.25) 0.03

Symptomatic disease 2.71 (0.85 - 8.58) 0.09

Risk factors for cancer or high grade dysplasia

Dysplasia grade

<0.001- LGD 1.13 (0.41 – 3.13)

- HGD 20.80 (7.77 – 55.67)

DALM 3.87 (2.10 – 7.14) <0.001

DALM: dysplasia associated lesion or mass. HGD: high grade dysplasia. LGD: low grade dysplasia

discussion

Optimal management of IBD patients with dysplasia on colonoscopy is important. In this

study, synchronous colorectal cancer was present in 15% of patients with colitis who

underwent restorative proctocolectomy for dysplasia. Patients with HGD and DALM had

a higher risk of cancer than patients with LGD and flat dysplasia, respectively. Multivari-

ate analysis additionally identified duration of disease as an independent risk factor for

cancer. This study also identified certain relationships between the location of preoperative

dysplasia and the postoperative location of pathological findings, which could be used to

clarify the need for proctectomy especially in the poor risk surgical patient.

Given the high risk of cancer, our findings support surgery in all patients with HGD. This

approach is supported by previous results and is in accordance with the ACG practice

guidelines.16,18,24,25 We found that the risk of cancer is further increased depending upon

additional factors such as duration of disease and the presence of DALM, consistent with

previous reports.1-3,12,26,27 For patients with preoperative LGD, cancer was detected in only

Chapter 4 71

4

3% of proctocolectomy specimens. The risk for either cancer or HGD was 15%. Patients

with LGD appearing as DALM had a higher rate of cancer (7%) compared to patients

with flat LGD (2%), but this difference was not statistically significant. When the risk for

postoperative cancer and HGD was considered, patients with LGD appearing as DALM

compared to flat LGD had a three-fold increase in risk from 9% to 29% (p=0.015). These

results suggest that the presence of DALM, even in the setting of LGD, is a sufficient

argument to support prophylactic surgical treatment. This position is also supported by

other current evidence and guidelines.16,18,24,28

Whether surgery should be undertaken in flat LGD is a matter of continued controversy,

considering the variability in the reported cancer risk.6,12,13 Current ACG guidelines prompt

consideration of surgery in patients with flat LGD to prevent progression to a higher grade

of neoplasia.16 However, the Medical Position Statement by the American Gastroentero-

logical Association (AGA) finds that current evidence is insufficient to either advice for

or against colectomy for flat LGD.25,29 These unclear recommendations leave much room

for debate. In this study we have shown that the risk of cancer in flat LGD is 2%, which

is at the lower end of the range (0 to 20%) previously reported.6,13-15 Without further

specification, however, these results make it difficult to decide for or against surgery. It is,

therefore, important to take other factors into account that might aid in this decision. This

study shows that about 1 in 10 patients with flat LGD had either cancer or HGD at the time

of diagnosis. In addition, 53% of patients with LGD had pathological findings of dysplasia

or cancer in the surgical specimen. These findings support a low threshold for surgery in

all patients with LGD, since this is the only treatment modality that can eliminate the risk

of cancer. Presence of other factors associated with cancer, like long duration of disease,

should also further tip the balance towards surgery. Other factors that could be taken into

account include multifocal dysplasia and extensive disease (i.e. pancolitis), which were uni-

variately associated with malignant change in our study. PSC and family history have also

been shown to be associated with increased risk of cancer in other reports.1,2,28 When none

of these factors are present, both surgery and a conservative approach may be considered,

the choice being individualized based on an informed discussion with the patient. The

high 5-year progression rates of up to 54% from LGD to HGD or cancer reported in several

studies is a further ongoing important consideration in favour of surgery.7,8,15,28

Considering the presence of dysplasia and cancer in segments adjacent to and remote

from the preoperative location, total proctocolectomy (+/- restoration of intestinal continu-

ity) is the ideal procedure for most patients undergoing surgery. Patients with isolated

colonic dysplasia without sigmoid involvement had low rates of rectal dysplasia (3%) and

rectal cancer (0%). Thus, colectomy with rectal preservation may obviate some of the risks

of a pelvic dissection when such patients have significant co-morbidity but require surgery.

72 Chapter 4

Select patients unwilling to have a permanent stoma and unsuitable for restorative surgery

may also be candidates for this approach, especially if the cancer risk or other conditions

preclude a strategy of surveillance alone.

Additionally, in patients with isolated colonic dysplasia who underwent mucosectomy

(n=41), none of the mucosectomy specimens showed dysplastic changes. Given the poor

functional outcomes for patients after mucosectomy and handsewn anastomosis30,31 and

also the small chance of finding cancer or dysplasia in the rectum or mucosectomy speci-

men, patients with isolated colonic dysplasia may safely be offered a stapled anastomosis.

In patients with preoperative rectal dysplasia, the risk of cancer or dysplasia in the rectum

is considerable (about 60%). In these patients, a proctocolectomy is virtually always the

management of choice. The type of anastomosis, however, is still a matter of controversy.

In series with long-term follow-up, dysplasia and cancer in the anal transitional zone after

stapled pouch surgery has been found to be infrequent.32-34 Dysplasia mostly developed

in the first two to three years, if at all, and often disappeared on repeated biopsies. None

of the series found cancer in the anal transition zone with a follow-up of 10 years or

more. These findings, combined with several reports of worse functional outcomes after

mucosectomy with handsewn anastomosis, support the use of a stapled approach in most

patients, especially when concerns regarding sphincter function exist. In selected cases,

like the presence of dysplasia near the anal transitional zone or if patients are considered

high risk for cancer for other reasons, mucosectomy may be considered.

When considering trends over time in our series, there was an increase in the number of

patients operated on for dysplasia in the later parts compared to the earlier parts of the

study. Reasons for this could be several, such as an increased awareness of the risk of

cancer, more standardized surveillance, increased incidence and prevalence of ulcerative

colitis and lower threshold for surgery due to advances in the field.

This study has some limitations. The study is retrospective, which limits the availability of

certain data. This was especially the case for some variables on preoperative colonoscopy,

which were not always well documented (e.g. exact location and size of lesions). Another

limitation is that this cohort only included patients who had proctocolectomy. This provides

a reliable reference test for the presence of cancer, in the form of complete pathological

examination of colon and rectum. On the other hand, it might introduce a certain degree

of selection bias. First, some patients with dysplasia may have been managed medically by

their physicians and not referred for surgery. Since HGD is widely considered an absolute

indication for surgery,16,23,25 we expect that this selection bias is limited in patients with

HGD. With LGD there is the possibility that physicians may have advised some patients not

to undergo surgery, since controversy still exists about this decision. Such referral bias is

Chapter 4 73

4

expected to occur more often in patients with fewer risk factors for cancer, and less often

in patients with factors that might increase this risk (e.g. prolonged severe disease, family

history, etc). In these latter patients, the decision to refer to surgery will probably have

been made more promptly. Therefore, if this referral bias has influenced our study, it would

probably have skewed our population to include patients with LGD judged as having a

greater risk of cancer. Although this should be taken into consideration when evaluating

results, we should recognize that the observed risk of cancer in LGD patients in this study

is actually in the lower range of that reported in the literature (3% vs. 0% to 20%).

The strength of this study is the inclusion of patients with a complete pathological evalu-

ation of the colon and rectum. This clearly provides the most reliable way of assessing the

presence of any synchronous pathological findings. The detailed review of all data and the

classification of endoscopic and pathologic findings based on current standard definitions

is also a strength. Some controversy in the literature on the topic relates to the constantly

changing definitions of these findings. Thus, our ability to classify and correlate the preop-

erative and final pathology findings based on uniform, standard definitions, is a particular

strength of this study. Finally, this study has a large sample size, with the total numbers of

patients with LGD (n=136) and LGD and DALM (n=42) being greater than even a previous

meta-analysis7, where the corresponding numbers of patients used to calculate the risk of

synchronous cancer were 98 and 31 respectively. Thus, the large numbers of patients in

the current study provide data that are less susceptible for over- and underestimation. The

study hence provides reliable estimates of cancer risk after proctocolectomy for colitis pa-

tients with dysplasia based on a consideration of patient, disease and endoscopic factors.

This information aids decision-making for clinicians and patients alike facing the dilemma

of choosing the optimal management strategy in area of controversy but great clinical

importance.

In conclusion, the presence of HGD or DALM should favour surgical intervention as the

treatment of choice. Despite a low risk of cancer in patients with flat LGD, the threshold

for surgery should be low given the high likelihood of finding dysplasia or cancer in the

postoperative pathology and hence the potential further ongoing risk of cancer. A detailed

discussion of associated risks and consideration of factors potentially increasing the risk

of cancer especially in the context of the beneficial effect of surgery in control of colitis

and thus reducing symptoms are important steps in decision-making. The specific medical

condition of the patient, especially duration of disease, severity and extent of disease,

multifocal dysplasia, family history and perioperative risk along with the risk of future

cancer based on current findings are additional factors that should be taken into con-

sideration in this decision. In most patients, especially if other factors that are associated

with risk of cancer are present, surgery should be considered. In selected cases without

74 Chapter 4

such factors, colonoscopic surveillance after discussion of associated risks may be accept-

able, provided high patient compliance can be assured. The location of dysplasia, whether

colonic or rectal, may further clarify the optimal surgical approach in individual patients.

Total proctocolectomy and restorative pouch surgery is currently the gold standard for

surgical management. Isolated colonic dysplasia without sigmoid involvement may allow

for subtotal colectomy in selected poor risk surgical patients. A stapled anastomosis may

be safely undertaken in patients with isolated colonic dysplasia and the majority of patients

with rectal dysplasia. Mucosectomy and hand-sewn anastomosis should be reserved for

specific circumstances as determined on a case by case basis.

Chapter 4 75

4

references

1. Gupta RB, Harpaz N, Itzkowitz S, Hossain

S, Matula S, Kornbluth A, Bodian C, et

al. Histologic inflammation is a risk factor

for progression to colorectal neoplasia in

ulcerative colitis: A cohort study. Gastroen-

terology. 2007; 133: 1099-105.

2. Lichtenstein GR, Rutgeerts P. Importance

of mucosal healing in ulcerative colitis.

Inflamm.Bowel Dis. 2010; 16: 338-46.

3. Rubin DT, Lashner BA. Will a 5-asa a day

keep the cancer (and dysplasia) away? Am J

Gastroenterol. 2005; 100: 1354-56.

4. Eaden JA, Abrams KR, Mayberry JF. The risk

of colorectal cancer in ulcerative colitis: A

meta-analysis. Gut. 2001; 48: 526-35.

5. Ullman T, Odze R, Farraye FA. Diagnosis

and management of dysplasia in patients

with ulcerative colitis and crohn’s disease

of the colon. Inflamm.Bowel Dis. 2009; 15:

630-38.

6. Befrits R, Ljung T, Jaramillo E, Rubio C. Low-

grade dysplasia in extensive, long-standing

inflammatory bowel disease: A follow-up

study. Dis Colon Rectum. 2002; 45: 615-20.

7. Thomas T, Abrams KA, Robinson RJ,

Mayberry JF. Meta-analysis: Cancer risk of

low-grade dysplasia in chronic ulcerative

colitis. Aliment.Pharmacol.Ther. 2007; 25:

657-68.

8. Ullman T, Croog V, Harpaz N, Sachar D,

Itzkowitz S. Progression of flat low-grade

dysplasia to advanced neoplasia in patients

with ulcerative colitis. Gastroenterology.

2003; 125: 1311-19.

9. Itzkowitz SH, Present DH. Consensus con-

ference: Colorectal cancer screening and

surveillance in inflammatory bowel disease.


10. Rutter MD, Saunders BP, Wilkinson KH,

Rumbles S, Schofield G, Kamm MA, Wil-

liams CB, et al. Thirty-year analysis of a

colonoscopic surveillance program for neo-

plasia in ulcerative colitis. Gastroenterology.

2006; 130: 1030-38.

11. Eaden J, Abrams K, McKay H, Denley

H, Mayberry J. Inter-observer variation

between general and specialist gastrointes-

tinal pathologists when grading dysplasia

in ulcerative colitis. J Pathol. 2001; 194:

152-57.

12. Blackstone MO, Riddell RH, Rogers BH,

Levin B. Dysplasia-associated lesion or mass

(dalm) detected by colonoscopy in long-

standing ulcerative colitis: An indication

for colectomy. Gastroenterology. 1981; 80:

366-74.

13. Lim CH, Axon AT. Low-grade dysplasia:

Nonsurgical treatment. Inflamm.Bowel Dis.

2003; 9: 270-72.

14. Lindberg B, Persson B, Veress B, Ingelman-

Sundberg H, Granqvist S. Twenty years’

colonoscopic surveillance of patients with

ulcerative colitis. Detection of dysplastic

and malignant transformation. Scand.J

Gastroenterol. 1996; 31: 1195-204.

15. Bernstein CN, Shanahan F, Weinstein WM.

Are we telling patients the truth about sur-

veillance colonoscopy in ulcerative colitis?

Lancet. 1994; 343: 71-74.

16. Kornbluth A, Sachar DB. Ulcerative colitis

practice guidelines in adults: American

college of gastroenterology, practice pa-

rameters committee. Am J Gastroenterol.

2010; 105: 501-23.

17. Lindberg J, Stenling R, Palmqvist R, Rute-

gard J. Surgery for neoplastic changes in

ulcerative colitis--can limited resection be

justified? Outcome for patients who under-

went limited surgery. Colorectal Dis. 2006;

8: 551-56.

18. Thomas T, Nair P, Dronfield MW, Mayberry

JF. Management of low and high-grade

dysplasia in inflammatory bowel disease:

The gastroenterologists’ perspective and

current practice in the united kingdom.

Eur.J Gastroenterol Hepatol. 2005; 17:

1317-24.

76 Chapter 4

19. Itzkowitz SH, Harpaz N. Diagnosis and

management of dysplasia in patients with

inflammatory bowel diseases. Gastroenter-

ology. 2004; 126: 1634-48.

20. Chambers WM, McC Mortensen NJ. Should

ileal pouch-anal anastomosis include muco-

sectomy? Colorectal Dis. 2007; 9: 384-92.

21. Riddell RH, Goldman H, Ransohoff DF,

Appelman HD, Fenoglio CM, Haggitt RC,

Ahren C, et al. Dysplasia in inflammatory

bowel disease: Standardized classification

with provisional clinical applications. Hum.

Pathol. 1983; 14: 931-68.

22. van Schaik FD, Offerhaus GJ, Schipper

ME, Siersema PD, Vleggaar FP, Oldenburg

B. Endoscopic and pathological aspects of

colitis-associated dysplasia. Nat Rev Gastro-

enterol Hepatol. 2009; 6: 671-78.

23. Kornbluth A, Sachar DB. Ulcerative colitis

practice guidelines in adults. American

college of gastroenterology, practice pa-

rameters committee. Am J Gastroenterol.

1997; 92: 204-11.

24. Sjoqvist U. Dysplasia in ulcerative colitis--

clinical consequences? Langenbecks Arch

Surg. 2004; 389: 354-60.

25. Farraye FA, Odze RD, Eaden J, Itzkowitz SH,

McCabe RP, Dassopoulos T, Lewis JD, et

al. Aga medical position statement on the

diagnosis and management of colorectal

neoplasia in inflammatory bowel disease.


26. Triantafillidis JK, Nasioulas G, Kosmidis PA.

Colorectal cancer and inflammatory bowel

disease: Epidemiology, risk factors, mecha-

nisms of carcinogenesis and prevention

strategies. Anticancer Res. 2009; 29:

2727-37.

27. Rubin DT. An updated approach to dys-

plasia in ibd. J Gastrointest.Surg. 2008; 12:

2153-56.

28. Ullman TA. Patients with low-grade dyspla-

sia should be advised to undergo colectomy.


29. Farraye FA, Odze RD, Eaden J, Itzkowitz

SH. Aga technical review on the diagnosis

and management of colorectal neoplasia in

inflammatory bowel disease. Gastroenter-

ology. 2010; 138: 746-74, 74.

30. Kirat HT, Remzi FH, Kiran RP, Fazio VW.

Comparison of outcomes after hand-sewn

versus stapled ileal pouch-anal anastomosis

in 3,109 patients. Surgery. 2009; 146:

723-29.





(ipaa) following proctocolectomy: A meta-

analysis of 4183 patients. Ann Surg. 2006;

244: 18-26.

32. Remzi FH, Fazio VW, Delaney CP, Preen

M, Ormsby A, Bast J, O’Riordain MG, et

al. Dysplasia of the anal transitional zone

after ileal pouch-anal anastomosis: Results

of prospective evaluation after a minimum

of ten years. Dis Colon Rectum. 2003; 46:

6-13.

33. O’Riordain MG, Fazio VW, Lavery IC, Remzi

F, Fabbri N, Meneu J, Goldblum J, et al.

Incidence and natural history of dysplasia of

the anal transitional zone after ileal pouch-

anal anastomosis: Results of a five-year

to ten-year follow-up. Dis Colon Rectum.

2000; 43: 1660-65.

34. Coull DB, Lee FD, Henderson AP, Anderson

JH, McKee RF, Finlay IG. Risk of dysplasia in

the columnar cuff after stapled restorative

proctocolectomy. Br.J Surg. 2003; 90: 72-75.

Part II

oPtImIzIng surgICal treatment for Inflammatory bowel dIsease

U. Ahmed Ali1

F. Keus2

J.T. Heikens3

W.A. Bemelman4

S.V. Berdah5

H.G. Gooszen1

C.J.H.M. van Laarhoven6

1Department of Surgery, University Medical Center Utrecht, Utrecht, Netherlands2Department of Surgery, Diakonessenhuis, Utrecht, Netherlands3Department of Surgery, St Elisabeth Hospital, Tilburg, Netherlands4Department of Surgery, Academic Medical Centre, Amsterdam, Amsterdam, Netherlands5Department of Digestive Surgery, Hôpital Nord - Marseille, Marseille, France6Department of Surgery, Radboud University Medical Center, Nijmegen, Netherlands

5 open versus laparoscopic

(assisted) ileal pouch-anal

anastomosis for ulcerative

colitis and familial adenomatous

polyposis

Cochrane Database of Systematic Reviews, 2009.

82 Chapter 5

absTracT

backgroundRestorative proctocolectomy with ileal pouch-anal anastomosis (IPAA) is the main surgi-

cal treatment for patients with ulcerative colitis (UC) and familial adenomatous polyposis

(FAP). With the advancements of minimal-invasive surgery this demanding operation is

increasingly being performed laparoscopically. Our objective is to perform a systematic

review to compare the beneficial and harmful effects of laparoscopic versus open IPAA for

patients with UC and FAP.

MethodsWe searched The Cochrane IBD/FBD Group Specialized Trial Register (April 2007), The

Cochrane Library (Issue 1, 2007), MEDLINE (1990 to April 2007), EMBASE (1990 to April

2007), ISI Web of Knowledge (1990 to April 2007) and the web casts of the American

Society of Colon and Rectal Surgeons (ASCRS) (up to 2006) for all trials comparing open

versus laparoscopic IPAA. All trials in patients with UC or FAP comparing any kind of lapa-

roscopic IPAA versus open IPAA. The methodological quality of included trials was evalu-

ated. Analysis of RCTs and non-RCTs was performed separately. Sensitivity and subgroup

analyses were performed if appropriate.

resultsEleven trials included 607 patients of whom 253 (41%) in the laparoscopic IPAA group.

Only one of the included trials was a randomized controlled trial. There were no signifi-

cant differences in mortality or complications between the two groups. Reoperation and

readmission rates were not significantly different. Operative time was significantly longer

in the laparoscopic group both in the RCT and meta-analysis of non-RCTs (weighted mean

difference (WMD) 91 minutes; 95% Confidence Interval (CI) 53 to 130). There were no sig-

nificant differences between the two groups regarding postoperative recovery parameters.

Sensitivity analysis using imputed data showed slightly improved postoperative recovery in

the laparoscopic group. Total incision length was significantly shorter in the laparoscopic

group, while two trials evaluating cosmesis found significantly higher cosmesis scores in

the laparoscopic group. Other long-term outcomes were poorly reported.

conclusionsThe laparoscopic IPAA is a feasible and safe procedure. Short-term advantages of the

laparoscopic approach seem to be limited and their clinical significance is arguable. Large

high-quality trials focusing on differences regarding specific postoperative complications,

cosmesis, quality of life and costs are needed.

Chapter 5 83

5

inTroducTion

Ulcerative colitis (UC) is an inflammatory disease of the large intestine of uncertain etiology

characterized by recurring episodes of inflammation primarily involving the mucosal layer

and occasionally the submucosa of the colon. The estimated incidence of UC is 8.7/100.000

in Europe.1 Whereas medical treatment is the mainstream for the management of UC, sev-

eral evidence-based indications for surgical treatment have been identified, including acute

pancolitis, intractability to medical treatment and the presence of dysplasia or neoplasia.2

Familial adenomatous polyposis (FAP) is a rare autosomal dominant disease, character-

ized by the development of hundreds to thousands of adenomatous polyps in the colon

and rectum of affected individuals leading to cancer at young age, if left untreated. The

incidence is around 2 per million and the prevalence is around 40 per million.3,4 Because of

the inevitable development of cancer in this disease, colectomy is warranted for all patients

with FAP at a certain point of their disease.

Both UC and FAP are limited to the colon and rectum. Therefore, proctocolectomy pro-

vides relief of symptoms and eliminates the risk of developing colorectal cancer. Since its

introduction in 1978 the restorative ileal pouch-anal anastomosis (IPAA) has gained wide

acceptance in the surgical treatment of patients with UC and FAP.5 The open IPAA has

become the standard surgical approach for UC and FAP, due to its good functional results

and high patient satisfaction.6-9

Over the past decades laparoscopic techniques have evolved rapidly and large bowel resec-

tions were increasingly being performed laparoscopically. The first laparoscopic-assisted

restorative proctocolectomy with IPAA (LA-IPAA) was described in 1992,10 and technical

feasibility of this procedure has been shown repeatedly.11-14 Reduced post-operative pain

quicker recovery and superior cosmetic results are presumed advantages of the LA-IPAA

over the open IPAA. However, reports have shown inconsistent results. Early reports tend

to be skeptical with respect to the benefits of the LA-IPAA,15-17 while recent studies tend to

show more favorable results of the laparoscopic technique.18-20

As patients with UC and FAP requiring surgery generally are young, active and highly

motivated individuals, minimal-invasive surgery may especially be appealing in this group

of patients.12 Cosmesis and patients’ satisfaction could therefore play important role in the

choice between these two approaches. However, possible benefits regarding these items

are of secondary importance compared to primary outcomes of surgery, like postoperative

morbidity and mortality. To be able to compare both techniques in a satisfactory manner a

systematic evaluation of the benefits and harms of open versus LA-IPAA is needed.

84 Chapter 5

MeThods

objectivesTo study whether laparoscopic and open IPAA for UC and FAP are different in terms of

primary (mortality and complications) and secondary outcomes (operative time, hospital

stay, convalescence, cosmesis, functional outcome and quality of life). If data were present,

differences in other secondary outcomes were compared as well.

criteria for considering studies for this review

Types of studies

Due to the paucity of randomized clinical trials, non- randomized controlled clinical trials

comparing open IPAA versus LA-IPAA were also included in this review. Trials were included

if they performed a direct comparison of open IPAA versus LA-IPAA, irrespectively of ran-

domization, prospective data collection, number of patients or language of the article.

Types of participants

Studies including patients with UC or FAP who underwent an IPAA procedure were in-

cluded. If studies sporadically included patients with other diseases, they were included

as long as the main population consisted of UC or FAP patients. Studies including mainly

patients with other diseases were excluded, unless they presented the data for the UC and

FAP patients separately. When multiple studies have overlapping patient populations, only

the most recent publication was included in the review.

Types of interventions

Studies comparing any type of open IPAA to any type of LA-IPAA were included. The

following classification of the surgical procedures (based on intention-to-treat) was used:

• 'LaparoscopicIPAA'includedthoseproceduresstartedasalaparoscopicprocedure,with

creation of any kind of pneumoperitoneum (by Veress needle or open introduction) or

mechanicalabdominalwalllift,irrespectiveofthenumberoftrocarsused.'Laparoscopic-

assistedIPAA'includedthoseproceduresinwhichanadditionalsmallincisionlaparotomy

was used (e.g. Pfannenstiel or subumbilical midline incision) to facilitate the laparoscopic

IPAA procedure.

• Inallothercasesthesurgicalinterventionwasclassifiedas'openIPAA'.

Types of outcome measures

Primary outcome measures were mortality and complications (except minor complications).

Secondary outcome measures were all other outcomes assessed in the comparison of the

two operative techniques. These included minor complications, operative time, operative

Chapter 5 85

5

blood loss, time to bowel movement, time to regular diet, hospital stay, readmission rate,

reoperation rate, incision length, cosmesis, functional outcome (fecal and sexual function)

and costs.

Complications were classified into the following categories:

• Intraoperativecomplications:allcomplicationsoccurringanddetected intraoperatively,

like small bowel perforation and severe intraoperative bleeding.

• Procedurespecificcomplications:pouchfailure,pelvicsepsis,pouchfistula,anastomotic

leakage and strictures.

• Severe postoperative complications: e.g. intra-abdominal abscesses, bleeding, sepsis,

burst abdomen (Platzbauch) and myocardial infarction.

• Mildpostoperativecomplications:e.g.includingprolongedileus,woundinfections,uri-

nary tract infections, urinary retention, pleural effusion, late incisional hernia, and deep

venous thrombosis. Other postoperative complications were categorized appropriately at

first encounter.

• Totalcomplications:thetotalnumberofallcomplicationsperstudy.

The following definitions were used for the procedure specific complications:

• Pouchfailure:pouchexcisionoranon-functioningpouchat12monthsafterIPAApro-

cedure.

• Pelvicsepsis:pelvicabscess,anastomoticleakageordehiscenceorpelvic/perinealwound

infection.

• Pouchfistula:anypouchrelatedfistula.

• Stricture:anastomoticfibrosisnecessitatingdilatation.

Functional outcome was assessed using the following items:

• Defecationfrequency:timesofdefecationperday,nightorper24hours.

• Mildfecalincontinence:soilingorspottinginunderwear.

• Severefecalincontinence:regularlysevereleakageorfecallossorpassivefecalinconti-

nence.

• Urgefecal incontinence: inability todeferdefecationmorethan15minutesafterfirst

urge.

• Sexualdysfunction:retrogradeejaculation,erectiondisorderordyspareunia.

search methods for identification of studiesWe have searched the following databases:

• TheCochraneIBD/FBDGroupSpecializedTrialRegister(NonMEDLINERecords)

• TheCochraneLibrary(includingTheCochraneDatabaseofSystematicReviews,Database

of Abstracts

86 Chapter 5

• ofReviewsofEffects(DARE),TheCochraneCentralRegisterofControlledTrials(CEN-

TRAL), Health

• TechnologyAssessment (HTA)Database,NHSEconomicEvaluationDatabase) (Issue1,

2007)

• MEDLINE(1990toApril2007)

• EMBASE(1990toApril2007)

• ISIWebofKnowledge(WebofScience)(1990toApril2007)

• WebcastsoftheannualmeetingsoftheAmericanSocietyofColonandRectalSurgeons

(ASCRS)

Since the LA-IPAA was not described before 1992 the search was started from the year

1990. As a fall-safe strategy we pre-specified that if any eligible article published before

1992 was to be found, the search would have been expanded to start from 1985. During

the search no eligible articles prior to 1992 were found, thus the search remained un-

changed. References of included trials and relevant reviews encountered during the search

were searched manually. Finally, all authors of included trials were requested for additional

information on any published, unpublished or ongoing trials.

data collection and analysisThe review was conducted according to the pre-specified protocol and the recommenda-

tions from the Cochrane Handbook for Systematic Reviews of Interventions.21,22

Selection of studies

Titles of all retrieved articles were first screened by UAA or FK and all obviously irrelevant

reports were excluded. Subsequently, abstracts of selected articles were reviewed by both

UAA and FK independently and differences were resolved by discussion with CVL, if neces-

sary. In case of any uncertainty, articles were always selected for the subsequent step.

Finally, the full-text of all selected articles was reviewed by UAA and FK independently to

determine the eligibility of the article for this review.

Extraction of data

Two reviewers (UAA and FK or JH) independently extracted all relevant data. For each study

patient characteristics, study characteristics, data needed for the methodological quality

assessment of the study and the primary and secondary outcomes were extracted according

to availability. Data regarding patient characteristics included number of patients in each

group, age, gender, BMI and diagnoses of included patients. Data regarding study charac-

teristics included study design, sample size information, inclusion and exclusion criteria of

the study, follow-up period, loss to follow-up, surgical experience and information regarding

surgical techniques. Individual authors were contacted if any essential data were missing.

Chapter 5 87

5

assessment of methodological quality of included studiesIn this review both randomized controlled trials (RCTs) and non- randomized controlled

trials (non-RCTs) were included. For both types a different assessment method was chosen.

Assessment of methodological quality of randomized clinical trials

Based on the available empirical evidence we assessed the methodological quality of RCTs

using the following items.22-25

– Generation of the allocation sequence

• Adequate, if the allocation sequence was generated by a computer or random

number table. Drawing of lots, tossing of a coin, shuffling of cards, or throwing

dice was considered as adequate if a person who was not otherwise involved in the

recruitment of participants performed the procedure.

• Unclear,ifthetrialwasdescribedasrandomized,butthemethodusedforgeneration

of the allocation sequence was not described.

• Inadequate,ifasysteminvolvingdates,names,oradmittancenumberswasusedfor

the allocation of patients.

– Allocation concealment

• Adequate, iftheallocationofpatients involvedacentral independentunit,on-site

locked computer, or sealed envelopes.

• Unclear,ifthetrialwasdescribedasrandomized,butthemethodusedtoconcealthe

allocation was not described.

• Inadequate,iftheallocationsequencewasknowntotheinvestigatorswhoassigned

participants or if the study was quasi- randomized.

– Blinding

• Adequate,ifthetrialwasdescribed(atleast)asblindtoparticipantsorassessorsand

the method of blinding was described. We are well aware that it is very difficult to

properly blind trials comparing surgical treatments, therefore one level of blinding

was considered adequate.

• Unclear,ifthetrialwasdescribedas(double)blind,butthemethodofblindingwas

not described.

• Notperformed,ifthetrialwasnotblinded.

– Follow-up

• Adequate,ifthenumbersandreasonsfordropoutsandwithdrawalsinallinterven-

tion groups were described or if it was specified that there were no dropouts or

withdrawals.

• Unclear,ifthereportgavetheimpressionthattherehadbeennodropoutsorwith-

drawals, but this was not specifically stated.

• Inadequate, if thenumberor reasons fordropoutsandwithdrawalswerenotde-

scribed.

88 Chapter 5

Assessment of methodological quality of non- randomized clinical trials

Quality assessment for non- randomized clinical trials is a complex topic and is generally

considered to be an area of ongoing research (Cochrane Handbook of Systematic Reviews

of Interventions, section 6.8).22 In an extensive review of this topic investigators reviewed

193 tools for quality assessment in literature,26 and concluded that based on quality and

design of these tools only 6 tools were suitable for use in systematic reviews. They also

concluded that all 6 tools needed some type of modification before being fully suitable for

that purpose.

In this review we used a modification of the Methodological Index for Non-Randomized

Studies (MINORS).27 This tool is one of the few validated and tested methods specifically

developed for the assessment of quality of non- randomized trials. To adhere to the guide-

lines of the Cochrane Handbook the following modifications were applied:

• Fourofthe12itemsoftheMINORSweredisregardedinthequalityassessment,because

these items related to the applicability, reporting quality, and precision of the results,

rather than the validity of the assessed trials.

• Everyitemwasusedindependentlytodistinguishbetweenhighandlowqualitytrials,

rather than using the sum score of this list. The draw-backs of using summary scores

are the decreased transparency to readers of the review and the evidence showing that

different scales could results in contradicting results when applied in the same review.28

The modified MINORS list is outlined in Table 1. Every study was assessed using this method

by UAA and JH independently. Discrepancies were solved by consensus discussion with a

third reviewer, CVL, if necessary.

statistical analysisData from RCTs and non-RCT were analyzed separately. With adequate data available

statistical analysis of binary data was conducted using relative risks (RR) as the summary

statistic. Trials with zero events in both arms were excluded from meta-analyses. However,

a sensitivity analysis using risk differences (RD) was performed with inclusion of these trials,

and in case of inconsistency the results of this sensitivity analysis were reported.

For continuous outcomes weighted mean differences (WMD) were used as the summary

statistic. Authors, however, often presented their results in medians with ranges due to

suspicion of skewed data, while means with their standard deviations (SD) are needed

for meta-analysis. Authors were contacted for additional data if necessary. Additionally,

sensitivity analyses imputing data for missing means and standard deviations (calculated

from available medians and ranges) were performed.29

Chapter 5 89

5

Heterogeneity was calculated using Higgins chi-square (χ²) test and quantified by mea-

suring I².30 A χ²-test with a P-value of <0.10 was considered to indicate the presence

of heterogeneity, while an I²>50% was considered to suggest a marked inconsistency

in effect between studies. The fixed-effect model was only used if no heterogeneity was

present. In all other cases the random-effects model was used. If excessive heterogeneity

was present, data were re-checked first. If heterogeneity persisted, subgroup or sensitivity

analyses were used to explore its causes. When adequate reasons were present extreme

outliers were excluded in sensitivity analyses. In situations of excessive heterogeneity that

could not be explained, we refrained from reporting a pooled estimate.

Bias detection

Funnel plots were used to provide a visual assessment of whether treatment estimates

were associated with study size. This may help identify the presence of publication or other

type of biases.31-33

Table 1. Modified Methodological Index for Non-Randomized Studies (MINORS)

Item*

1. Inclusion of consecutive patients: all patients potentially fit for inclusion (satisfying the criteria for inclusion) have been included in the study during the study period (no exclusion or details about the reasons for exclusion).

2. Prospective collection of data: data were collected according to a protocol established before the beginning of the study.

3. Unbiased assessment of the study endpoint: blind evaluation of objective endpoints and double-blind evaluation of subjective endpoints. Otherwise the reasons for not blinding should be stated.

4. Follow-up period appropriate to the aim of the study: the follow-up should be sufficiently long to allow the assessment of the main endpoint and possible adverse events.

5. Loss to follow up less than 5%^: all patients should be included in the follow up. Otherwise, the proportion lost to follow up should not exceed the proportion experiencing the major endpoint; or if the numbers and reasons for dropouts and withdrawals in all intervention groups were described.

6. An adequate control group: having a gold standard diagnostic test or therapeutic intervention recognized as the optimal intervention according to the available published data

7. Contemporary groups: control and studied group should be managed during the same time period (no historical comparison).

8. Baseline equivalence of groups **: the groups should be similar regarding the criteria other than the studied endpoints, i.e. absence of confounding factors that could bias the interpretation of the results.

* Items were scored ‘adequate’ if condition was satisfied, ‘inadequate’ if condition was not satisfied and ‘un-clear’ if information regarding item was not reported.^ Measured at time of discharge, since most trials only followed patients until discharge.** In this review baselines of the two groups should be equivalent regarding age, gender, BMI and distribution of diagnoses (UC/FAP).

90 Chapter 5

Subgroup and sensitivity analyses

Subgroup analyses were performed according to the methodological quality of the in-

cluded trials:

• ForRCTs:adequatecomparedtounclear/inadequateregardingthefourqualitycriteria

used to assess the methodological quality of RCTs.

• For non-RCTs: adequate compared to unclear/inadequate regarding each item of the

modified MINORS list on which the studies differed.

Furthermore, causes of heterogeneity were explored by performing sensitivity analysis

based on surgical technique and other factors that may explain heterogeneity.

Statistical analysis was conducted using the statistical package (RevMan Analyses v. 5.0.16)

provided by The Cochrane Collaboration and was performed by UAA, FK and CVL.

resulTs

description of studies

Search and trial identification

The systematic search was conducted in The Cochrane IBD/FBD Group Specialized Trial

Register (Non-MEDLINE Records) (931 records, 0 selected), The Cochrane Library, Issue

1, 2007 (649 hits, 5 selected), The National Library of Medicine (MEDLINE) via PubMed

(4455 hits, 117 selected), Exerpta Medica via EMBASE (6830 hits, 503 selected), ISI Web

of Knowledge (1573 hits, 54 selected), and web casts of the annual meetings of the

American Society of Colon and Rectal Surgeons (ASCRS) (all published web cast until 2006,

2 selected).

Altogether, the search resulted in 14,438 hits. The first selection was performed based on

the titles of publications and all clearly irrelevant publications were excluded. A total of

679 hits were considered possibly relevant based on their titles. After correction for dupli-

cates, 590 publications remained. The abstracts of these 590 publications were reviewed

independently by two reviewers (UAA and FK). Differences between UAA and FK were

discussed with CVL. A total of 552 publications could be rejected based on their abstracts.

Eventually, 37 publications were selected for full-text review.

From the selected 37, eleven publications were included, and one protocol of an ongoing

RCT was identified.34 A cross-reference search was performed of all included publications as

well as five reviews considered relevant to our review.35-39 This resulted in the identification

Chapter 5 91

5

of one additional publication.40 Consequently, 12 publications were finally included in this

review. A flow-chart of the selection process is provided in Figure 1.

Figure 1. Flow-chart of the selection process

Of these 12 publications, two described the short- and long-term outcomes of a partially

overlapping patient population (Larson 2005; Larson 2006).19,41 These publications were

considered to be two distinct trials, but data of only one of these two publications was

used in any single meta-analysis to prevent double reporting of patients. Two other pub-

lications (Maartense 2004; Polle 2007) described the short- and long-term outcomes of

the same trial.42,43 These two publications were considered one trial and data was handled

accordingly. In summary, 12 publications, describing 11 trials were included in this review.

92 Chapter 5

Finally, two publications were translated from French and Japanese (Berdah 2004; Otani

2001).44,45 Additional information on methodology was obtained regarding seven out of

11 trials and additional data (i.e. individual patient data) was obtained from two trials

(Berdah 2004; Maartense 2004).42,44

Patient characteristics

The 11 trials included 607 patients, of whom 253 (41%) in the laparoscopic and 354 (59%)

in the open IPAA group. A total of 516 (85%) patients suffered from UC and 89 (15%)

from FAP. Two studies exclusively included patients with UC,40,45 whereas all other studies

included both UC and FAP patients. None of the studies presented results of UC and FAP

patients separately. Baseline characteristics of included patients are shown in Table 2.

Trial designs

Of the 11 included trials, only one was a randomized controlled trial.42 Four of the 10

non- randomized trials collected their data prospectively. The other 6 had retrospectively

data-collection or did not clearly specify this aspect. Ten trials had a mono-center design,

and one trial was conducted in two centers.42 Characteristics of the included trials are

shown in Table 3.

Surgical interventions

The types of performed surgical interventions varied between the included trials. Eight

trials compared the laparoscopic-assisted IPAA (LA-IPAA) with the conventional open IPAA.

The other three trials compared slightly different types of procedures: one trial (Brown

2001) compared the LA-IPAA with a mini-open IPAA46, another trial (Maartense 2004)

compared the laparoscopic hand-assisted IPAA (HA-IPAA) with the conventional open

IPAA,42 and yet another trial (Larson 2006) compared a combined group of LA-IPAA and

HA-IPAA with the conventional open IPAA. We considered all trials comparing procedures

using a laparoscopic technique versus open IPAA.

The HA-IPAA differs from the LA-IPAA in that the accessory incision is performed at the

start of the operation and covered with an air sealed hand-port. Through this hand-port

manual assistance could be provided during the different stages of the operation. In gen-

eral, HA-IPAA and LA-IPAA were performed using a small Pfannenstiel incision, facilitating

open rectum resection and open creation of a J-pouch by means of a double-stapling

technique to construct the ileal pouch-anal anastomosis.

Outcome measures

Primary outcome measures, mortality and complications, were well reported. Due to

the brief follow-up period in most trials mainly short-term complications were reported.

Chapter 5 93

5

Tab

le 2

. Pat

ien

ts’ c

har

acte

rist

ics

of

incl

ud

ed s

tud

ies

Tria

l

NA

ge

Gen

der

(M

:F)

BM

ID

iag

no

sis

(UC

:FA

P)

LA-IP

AA

/O

pen

IPA

ALA

-IPA

AO

pen

IPA

ALA

-IPA

AO

pen

IPA

ALA

-IPA

AO

pen

IPA

ALA

-IPA

AO

pen

IPA

A

Ara

ki 2

00140

21/1

127

.2(8

.1)^

31.1

(11.

2)^

8:13

2:09

31.2

(4.5

)^33

.1(4

.8)^

21:0

011

:00

Berd

ah 2

00444

12/1

232

(16-

60)

31 (2

1-58

)6:

066:

0624

(17-

29)

22 (2

0-26

)11

:01

11:0

1

Brow

n 20

0146

13/1

332

(16-

69)

29 (1

5-59

)4:

085:

08-

-2:

9 +

1 o

ther

6:07

Dun

ker

2001

4715

/17

30.6

(7.1

) ^*

39.2

(8.4

) ^*

4:11

9:08

23.6

(4.6

) ^24

.5 (3

.2)

14:0

114

:03

Has

him

oto

2001

4811

/13

30.0

(19-

47)

30.0

(18-

49)

4:07

7:06

--

6:05

6:07

Lars

on 2

00519

33/3

328

(18-

56)

27 (1

7-56

)6:

276:

2721

.7 (1

7-31

)22

.3 (1

8-33

)31

:02:

0031

:02:

00

Lars

on 2

00641

100/

200

32 (1

7-66

)32

(17-

64)

40:6

080

:120

22.4

(17-

34)

23 (1

6-32

)98

:02:

0019

1:09

:00

Maa

rten

se 2

00442

30/3

029

(16-

57)*

35 (1

6-57

)*9:

2115

:15

22.6

(18.

1-34

.7)

23.3

(17.

2-34

.2)

20:1

020

:10

Mar

cello

200

02020

/20

25 (1

2-61

)26

(9-6

1)15

:05

15:0

524

(18-

32)

24 (1

8-30

)13

:07

13:0

7

Ota

ni 2

00145

10/1

830

.2(1

1.8)

^39

.6(1

7.7)

^-

--

-10

:00

18:0

0

Schm

itt 1

99417

22/2

031

(12-

59)

34 (1

7-64

)11

:11

11:0

9-

-16

:5 +

1 o

ther

15:0

5

* si

gnifi

cant

diff

eren

ce; ^

mea

n (S

D).

IPA

A: i

leal

pou

ch-a

nal a

nast

omos

is; L

A-IP

AA

: lap

aros

copi

c (a

ssis

ted)

IPA

A; M

: mal

e; F

: fem

ale;

UC

: ulc

erat

ive

colit

is; F

AP:

fam

ilial

ade

nom

atou

s po

lypo

sis.

94 Chapter 5

Tab

le 3

. Ch

arac

teri

stic

s o

f in

clu

ded

stu

die

s

Tria

lN

Cas

es/C

ontr

ols

Des

ign

Mat

chin

g^

Follo

w-u

p

per

iod

Loss

to

Fo

llow

-up

LA-IP

AA

Loss

to

Fo

llow

-up

Ope

n IP

AA

Co

nve

rsio

n r

ate

(%)

Ara

ki 2

00140

21/1

1R,

NR

-D

/C0

0-

Berd

ah 2

00444

12/1

2P,

NR

1,2,

3,4

> 3

yea

rs0

03/

12 (2

5%)

Brow

n 20

0146

12/1

3R,

NR

-D

/C0

00/

12 (0

%)

Dun

ker

2001

4715

/17

R, N

R4,

5,8

Mea

n 16

mon

th0

00/

15 (0

%)

Has

him

oto

2001

4811

/13

R, N

R-

D/C

00

0/11

(0%

)

Lars

on 2

00519

33/3

3P,

NR

1,2,

3,4,

6>

12

mon

ths

4N

/A-

Lars

on 2

00641

100/

200

P+R,

NR

1,2,

3,5,

690

day

s0

06/

100

(6%

)

Maa

rten

se 2

00442

30/3

0RC

T-

30 d

ays

2/30

(6,6

%)

3/30

(10%

)0/

30 (0

%)

Mar

cello

200

02020

/20

P, N

R1,

2,3,

4,7

D/C

00

0/20

(%)

Ota

ni 2

00145

10/1

8N

S, N

R-

D/C

00

-

Schm

itt 1

99417

22/2

0P,

NR

1,2,

4D

/C0

0-

Tota

ls25

3/35

4*9/

200

(4.5

%)

* La

rson

200

5 w

as n

ot a

dded

up

to t

he t

otal

of

patie

nts

to a

void

dup

licat

ion

patie

nt d

uplic

atio

n.^

1:

age;

2:

gend

er;

3: B

MI;

4: d

iagn

osis

; 5:

ope

rativ

e te

chni

que;

6:

date

of

oper

atio

n; 7

: se

verit

y of

dis

ease

; 8:

dur

atio

n of

fol

low

-up.

Eac

h nu

mbe

r in

dica

tes

that

the

stu

dy

had

activ

ely

mat

ched

bet

wee

n th

e tw

o gr

oups

reg

ardi

ng t

hat

item

.IP

AA

: ile

al p

ouch

-ana

l ana

stom

osis

; LA

-IPA

A:

lapa

rosc

opic

(as

sist

ed)

IPA

A;

R: r

etro

spec

tive;

P:

pros

pect

ive;

P+

R: p

rosp

ectiv

e fo

r th

e LA

-IPA

A g

roup

, an

d re

tros

pect

ive

for

the

open

-IPA

A g

roup

; NS:

not

spe

cifie

d; N

R: n

on-r

ando

miz

ed t

rial;

RCT:

ran

dom

ized

con

trol

led

tria

l; N

/A: n

ot a

pplic

able

; D/C

: unt

il di

scha

rge.

Chapter 5 95

5

Long-term outcomes were reported by 4 trials only (Berdah 2004; Dunker 2001; Larson

2005; Maartense 2004).19,42,44,47

Secondary outcomes were reported variably. Based on availability of data, meta-analysis

was performed for complications, operative time, blood loss, time to bowel movement,

time to regular diet, hospital stay, re-operation rate and incision length. Cosmesis, func-

tional outcome and costs were reviewed without meta-analysis.

risk of bias in included studiesSince only one RCT was identified all trials, including the RCT, were assessed using the

modified MINORS. Differences were identified regarding three items only:

• Prospective collectionofdata:five trials (45%) scored 'adequate'and six trials (55%)

scored'unclear/inadequate'.

• Contemporarygroupsofcasesandcontrols:seventrials(64%)scored'adequate'and

fourtrials(36%)scored'unclear/inadequate'.

• Baselineequivalenceofgroups:seventrials(64%)scored'adequate'andthefourtrials

(36%)scored'inadequate'.

Only three trials, including the RCT, scored adequate for these three items. Detailed meth-

odological assessment of trials is listed in Table 4.

Table 4. Methodological assessment of included studies

Trial

Item 1Con-

secutive patients

Item 2Prospec-

tive design

Item 3Assess-ment of

out-comes

Item 4FU-

period

Item 5Loss to

FU< 5%

Item 6Ad-

equate control group

Item 7Contem-porary control group

Item 8Baseline equiva-lence

Araki 200140 A U I A A A I A

Berdah 200444 A A I A A A A A

Brown 200146 A I I A A A I A

Dunker 200147 A I I A A A A I

Hashimoto 200148 A I I A A A I I

Larson 200519 A A I A A A A A

Larson 200641 A I I A A A A A

Maartense 200442 A A I A A A A A

Marcello 200020 A A I A A A A A

Otani 200145 A U I A A A U I

Schmitt 199417 A A I A A A A I

A: Adequate; I: Inadequate; U: Unclear

96 Chapter 5

effects of interventionsMeta-analysis of RCTs was not feasible, since only one RCT was identified. Results of this

RCT are presented in Table 5. When adequate data were available, a meta-analysis of

the non- randomized trials was performed. In this meta-analysis five comparisons were

conducted. In three comparisons trials were subdivided into subgroups of high- and low-

quality trials based on the three methodological quality criteria. In the fourth comparison

trials meeting all three methodological criteria (highest-quality trials) were set against all

other trials. The fifth comparison contained additional sensitivity analyses. The findings of

the RCT as well as the results of the meta-analysis of non-RCTs are discussed below.

No significant differences were observed regarding mortality, intra-operative complications,

procedure specific complications, severe complications, minor complications, readmission

and reoperation rate in any of the performed comparisons.

Table 5. Reported outcomes by Maartense 2004 (RCT)

OutcomeLA-IPAA(N=30)

Open IPAA(N=30)

P-value*

Mortality 0 (0%) 0 (0%) NS

Intraoperative complications 1 (3%) 1 (3%) NS

Procedure specific complications 4 (13%) 6 (20%) NS

Severe postoperative complications 0 (0%) 2 (7%) NS

Mild postoperative complications 9 (30%) 7 (23%) NS

Total complications 14 (47%) 16 (53%) NS

Operative time 214 (149 - 400) 133 ( 97 - 260) p<0.001

Blood loss 263 (75–1200) 300 (50–800) p=0.98

Time to regulat diet 6 (4 - 19) 7 (4 - 15) 0.6

Hospital stay 10 (5 - 13) 11 (6 - 28) 0.767

Readmission 5/23* (22%) 3/23* (13%) NS

Reoperation 5 (17%) 5 (17%) NS

N (%) / Median (range)* Only reported in a sub-group of patients (N=23).IPAA: ileal pouch-anal anastomosis; LA-IPAA: laparoscopic (assisted) IPAA. NS = not statistically significant.

Mortality

Mortality was not reported in two trials (Hashimoto 2001; Otani 2001).45,48 In the nine

trials reporting mortality a total of 232 patients and 323 patients were included in the

laparoscopic and the open group, respectively. Eight trials reported zero mortality in both

groups. With one death in the open group by one trial (Araki 2001), there was no statisti-

cally significant difference.40

Chapter 5 97

5

Intraoperative complications

Intraoperative complications were reported by 5 trials including 130 and 230 patients in

the laparoscopic and open groups, respectively. The RCT and one additional trial reported

one intraoperative complication in each group, while three trials reported zero intraopera-

tive complications. No significant differences were observed between the two groups.

Procedure specific complications

Procedure specific complications were reported in 8 trials, with 2 trials reporting zero

complications. The RCT reported 4/30 (13%) and 6/30 (20%) complications in the

laparoscopic and open group, respectively (not statistically significant). Five non-RCTs were

pooled showing 6/132 (4.5%) procedure specific complications in the laparoscopic group

and 6/224 (2.7%) in the open group. Differences were not statistically significant (Relative

risk (RR) 0.81; 95% confidence interval (CI) 0.32 to 2.02). Heterogeneity was not present.

Severe complications

Severe complications were reported in 9 trials, with 3 trials reporting zero complications.

The RCT reported 0/30 (0%) and 2/30 (7%) severe complications in the laparoscopic and

open group, respectively (not statistically significant). Five non-RCTs were pooled showing

8/157 (5.1%) severe complications in the laparoscopic group and 20/258 (7.8%) in the

open group. Differences were not statistically significant (RR 0.65; 95% CI 0.29 to 1.48).

Heterogeneity was not present.

Minor complications

Minor complications were reported in 9 trials. The RCT reported 9/30 (30%) and 7/30

(23%) minor complications in the laparoscopic and open group, respectively (not statisti-

cally significant). Eight non-RCTs were pooled. There were 59/213 (27.7%) minor compli-

cations in the laparoscopic group and 83/306 (27.1%) in the open group. Differences were

not statistically significant (RR 1.05; 95% CI 0.78 to 1.41). No significant heterogeneity

was present.

Total complications

Total complications consisted of the sum of all complications in the aforementioned cat-

egories. The RCT showed 14/30 (47%) and 16/30 (53%) total complications in the laparo-

scopic and open group, respectively (not statistically significant). Eight non-RCTs with 213

patients in the laparoscopic group and 306 patients in the open group were pooled. A total

of 80 (37.6%) complications in the laparoscopic group and 127 (41.5%) in the open group

were observed. There were no significant differences between the two groups (RR 0.91;

95% CI 0.73 to 1.14) (Figure 2). No heterogeneity was observed between the studies. In a

funnel plot on total complications we did not find arguments for bias (Figure 3).

98 Chapter 5

Operative time

Operative time was reported in 10 trials, but only five provided data in means with their

standard-deviations. Both the RCT and the pooled non-RCTs showed a significantly longer

operative time for the laparoscopic procedure. The RCT showed an increase in the median

operative time from 133 (range 97 to 260) for the open IPAA to 214 (range 149 to 400)

Figure 2. Meta-analysis total complications

Figure 3. Funnel plot of trials reporting total complications, including 95% confidence interval lines. There are no indications for bias.

Chapter 5 99

5

for the laparoscopic approach (p=<0.001). The meta-analysis of four non-RCTs included 58

patients in each group and showed a significantly longer operative time in the laparoscopic

group (weighted mean difference (WMD) 92 minutes; 95% CI 53 to 130). The random-

effects model was used, since substantial heterogeneity was present (χ² 11.02 (p=0.01),

I² 73%).

Sensitivity analysis imputing data for missing values included 223 and 324 patients in the

laparoscopic and open groups, respectively, and confirmed the findings of the primary

meta-analysis (WMD 82 minutes; 95% CI 60 to 105). Heterogeneity remained significant

(χ² 36.30 (p<0.0001), I² 75.2%). Further exploring of data in subgroup analyses did not

identify clear causes for the observed heterogeneity.

Blood loss

Operative blood loss was reported by five trials. The RCT reported no difference between

the two groups with a median blood loss of 263 mL (range 75 to 1200) in the laparoscopic

and 300 mL (range 50 to 800) in the open group (p=0.98). Two non-RCTs provided data

in means and SDs and were pooled. The meta-analysis included 31 and 29 patients in the

laparoscopic and open groups, respectively. There was no significant difference in blood

loss between the two techniques (WMD -99 mL; 95% CI -261 to 64). As heterogeneity was

present, the random-effects model was applied (χ² 2.70 (p<0.10), I² 63.0%). A sensitivity

analysis imputing data for two more trials included 62 patients in each group and did show

a significant reduction in blood loss using the laparoscopic technique (WMD -138 mL; 95%

CI -235 to -41). Adding these two extra trials reduced heterogeneity significantly (χ² 4.57

(p<0.21), I² 34.4%).

Time to bowel movement

Time to bowel movement was reported by three non-RCTs. All trials found a significant

shorter time to bowel movement in the laparoscopic group. Data of only one trial was

provided in means and standard-deviations (Araki 2001).40 Therefore, only pooled data

from a sensitivity analysis imputing data for means and SD from the two other trials was

available. This analysis included 141 and 231 patients in the laparoscopic and open group,

respectively, and showed a significant shorter time to bowel movement in the laparoscopic

Figure 4. Sensitivity analysis time to bowel movement (days) with imputed data

100 Chapter 5

group (WMD -1.96 days; 95% CI -3.45 to -0.46) (Figure 4). The random-effects model was

used, since heterogeneity was present (χ² 6.74 (p=0.03), I² =70%).

Time to regular diet

Time to regular diet was reported by six trials. The RCT showed a median of 6 days (range

4 to 19) in the laparoscopic and 7 days (range 4-15) in the open group (p=0.6). Only two

non-RCTs including 25 and 35 patients in the laparoscopic and open groups, respectively,

provided data in means and standard-deviations. No significant difference between both

groups was found when pooling these trials (WMD -2.72 days, 95% CI -8.33 to 2.88).

A sensitivity analysis with imputed data including three additional trials, with a total of 148

patients in the laparoscopic and 261 patients in the open group, did result in a significant

difference in favor of the laparoscopic group (WMD -1.47 days; 95% CI -2.25 to -0.69)

(Figure 5). No significant heterogeneity was present.

Figure 5. Sensitivity analysis time to regular diet (days) with imputed data

Hospital stay

Hospital stay was reported by 9 trials. The RCT showed no significant difference between

the two groups, with a median stay of 10 days (range 5 to 13) in the laparoscopic and 11

days (range 6-28) in the open group. Four non-RCTs, including 48 patients in the laparo-

scopic and 40 in the open group, presented data in means with SD. Pooling these trials

showed a significantly shorter hospital stay for the laparoscopic procedure compared to the

open technique (WMD -2.66 days; 95% CI -4.28, -1.04). No heterogeneity was present.

A sensitivity analysis using imputed data included five additional trials, with a total of 213

patients in the laparoscopic and 306 in the open group. This sensitivity analysis, showed a

significant shorter hospital stay for the laparoscopic group as well (WMD -2.12 days; 95%

CI -3.12 to -1.12) (Figure 6). The random-effects model was used, since heterogeneity was

present (χ² 12.33 (p=0.09), I² 43%).

Chapter 5 101

5

Figure 6. Sensitivity analysis duration of hospital stay (days) with imputed data

Readmission rate

Readmission rate was reported by two trials. The RCT reported 5/23 (22%) and 3/23 (13%)

readmissions in the laparoscopic and open group, respectively. Larson 2006 found 21/100

(21%) readmissions in the laparoscopic and 44/200 (22%) in the open group.41 Both stud-

ies showed no significant difference between the two groups.

Reoperation rate

Re-operation rate was reported by seven trials. The RCT reported 5/30 (17%) reoperations

in both groups (not statistically significant). Six non-RCTs were pooled. There were 7/172

(4.0%) reoperations in the laparoscopic and 16/275 (5.8%) in the open group. Differences

were not statistically significant (RR 0.74; 95% CI 0.32 to 1.71). No significant heterogene-

ity was present.

Incision length

Incision length was reported by two studies (Brown 2001, Dunker 2001).46,47 Both studies

showed a significant smaller incision length for the laparoscopic group. Since only one

trial provided means and SDs, no meta-analysis was performed. Sensitivity analysis with

imputed data, including 27 patients in the laparoscopic and 30 patients in the open group,

showed a significantly shorter incision in the laparoscopic group as well (WMD -7.79 cm;

95% CI -9.68 to -5.9). There was no heterogeneity present.

Cosmesis

Cosmesis scores were reported by two trials (Dunker 2001; Maartense 2004).42,47 Both

studies used the same cosmesis scale and both showed significantly higher cosmesis scores

in the laparoscopic group. The RCT showed a significant increase from a mean of 14.7

points in the open group to 18.5 in the laparoscopic groups (SD not reported, p=0.01).

Dunker 2001 reported an increase from a mean of 16 (4.6) points in the open to 19.8 (4.6)

in the laparoscopic group (p=0.03).47

102 Chapter 5

Functional outcome

Defecation frequency

Defecation frequency was reported by four trials. However, pooling data was not possible

due to inconsistencies in reporting of results. All available data of these four trials are

presented in Table 6. The defecation frequency was reported by three trials (1 RCT and 2

non-RCTs) per day and night separately, all with a follow-up period of at least 12 months.

One trial (Otani 2001) reported the defecation frequency per 24 hours at discharge.45 No

significant differences were reported between the two groups by any of the four trials. Two

sensitivity analyses imputing missing data for non-RCTs reporting the defecation frequency

per day and night separately, showed no significant differences as well.

Table 6. Available data for defecation frequency and costs

OUTCOME N LA-IPAA Open IPAA

DefecationFrequency

TrialLA-IPAA /

Open-IPAAMean/Median

(SD/Range)Mean/Median

(SD/Range)P-value*

- during the day Dunker 200147 15/17 5.70 (1.30) 6.30 (2.00) NS

Larson 200519 33/33 7 (3-14) 6 (3-12) 0.23*

Maartense 200442 22/23 6.09 (2.29) 5.35 (1.82) 0.161

- during the night Dunker 200147 15/17 1.0 (0.7) 1.3 (0.7) NS

Larson 200519 33/33 1 (1-5) 2 (1-4) 0.86*

Maartense 200442 22/23 2.14 (1.91) 1.78 (1.41) 0.371

- per 24 hours Otani 200145 10/18 8.00 (2.3) 11.00 (1.0) NS

Costs (Euros) TrialLA-IPAA /

Open-IPAAMean (SD) Mean (SD) P-value*

operative costs Maartense 200442 30/30 3467 (291) 1756 (171) 0.001

overall costs Maartense 200442 30/30 18,733 (8667) 16,830 (8626) 0.095

* As provided by author.IPAA: ileal pouch-anal anastomosis; LA-IPAA: laparoscopic (assisted) IPAA

Fecal incontinence

Four trials reported on this outcome. Every study used its own classification for fecal incon-

tinence and pooling results was not possible. None of the four trials found any significant

difference between the two groups regarding daytime continence, overnight continence,

soiling, urge incontinence, or any of the used measure of fecal continence.

Sexual function

Sexual function was reported by three trials. All trials measured and reported sexual

function differently. None of these trials identified a significant difference between the

laparoscopic and open group regarding this outcome as well.

Chapter 5 103

5

Costs

Only one trial (Maartense 2004) reported on differences in costs between the laparoscopic

and open IPAA (Table 6).42 Operative costs were significantly higher in the laparoscopic

group. However, when the overall total costs (including costs for hospital stay, relaparoto-

mies and readmission, etc.) were analyzed no significant difference was found.

discussion

This systematic review evaluating differences between laparoscopic and open IPAA

procedures shows several important findings. First, only one randomized trial has been

conducted so far on LA-IPAA versus open IPAA. Therefore, lower level evidence had to be

included in this review with consequently an increased risk for introducing bias. Second,

no significant differences were found in the primary outcome measures: mortality and

complications. Also readmission and reoperation rates were not significantly different.

Third, laparoscopic IPAA is associated with a significantly longer operative time. Fourth,

presumed short-term benefits of the LA-IPAA regarding convalescence could not be con-

firmed reliably. Fifth, follow-up periods of most studies were inappropriate for evaluation

of long-term outcomes.

This review has encountered several methodological problems. First, by including non-

randomized studies the risk for bias and its influence on false conclusions is considerable.26

Non-randomized trials have a higher risk of bias, due to lack of standardized protocols,

unclear methodology, high risk of selection bias and inability to match for all confounders.

Moreover, the methodological quality of most included trials was moderate or poor. For

example, only 3 out of the 10 non- randomized trials collected their data prospectively and

had a sufficient level of base-line equivalence between the two groups. However, these

lower level evidence studies are the best we have so far. While future research is obviously

needed, we hope that by reviewing this evidence systematically we can provide clinicians

with a balanced understanding of the present evidence and its possible implications.

Methodological quality assessment of non- randomized trials has proved to be problematic

as well. Contrary to RCTs, little evidence exists regarding important items in the method-

ological assessment of non- randomized trials. Moreover, few validated scales and lists

exist for this purpose and there is no consensus regarding the tool of choice. The MINORS

index, chosen for its validation and simplicity, proved to have its own shortcomings as well.

It was liable to confuse the validity of trials with other items like quality of reporting and

precision of results, and it used a summary score to indicate the overall quality of trials.

Both properties have important drawback and their use is discouraged by the Cochrane

104 Chapter 5

guidelines.22 The applied modifications did solve these problems, but they also may have

undermined the validity of the scale. Empirical evidence regarding methodological criteria

influencing reliability of the results of non- randomized trials is needed for better assess-

ment of this type of trials.

Another problem that faced this review was the paucity of data regarding several out-

comes. This was caused by the relatively small number of available trials, but also by the

small numbers of patients in every trial and the limited amount of data suitable for meta-

analysis. Especially the latter point was of considerable importance. For several continuous

outcomes less than half of the available trials could be included in the meta-analysis. This

could undermine the reliability of results, as illustrated by the inconsistent results of the

meta-analysis and the sensitivity analysis with imputed data, seen in several outcomes.

Additionally, most trials had a short follow-up period, usually until discharge, which made

it difficult to evaluate important long-term outcomes, like functional outcome and quality

of life.

During the conduct of this review another meta-analysis comparing open versus laparo-

scopic IPAA was published.49 General conclusions of this meta-analysis were quite similar to

ours, although there were some important differences. First, the search strategy performed

was less extensive and limits were applied for ‘study type’ in PubMed. This resulted in

failure to identify two relevant trials.44,45 Secondly, this meta-analysis included just over

half the numbers of patients included in our meta-analysis (329 compared to 607), mainly

due to a large trial which was published recently.41 Also, two trials were included in this

meta-analysis that were excluded by us for having partial overlap in patients with other

included trials.14,50 The last important difference is that the authors imputed means and

standard deviations using medians and ranges in their meta-analyses without specification

in the methods section. Some of their findings were therefore partly based on imputed

data and may very well be biased.

The absence in this review of differences between the open and laparoscopic IPAA con-

sidering the primary outcomes, mortality and complications, suggests that the LA-IPAA is

a feasible and safe procedure. This is in accordance to the results of several relative large

series of LA-IPAA published recently.12,51 However we have to emphasize that the number

of included patients is rather low to be able to detect all clinically relevant differences.

Especially detailed evaluation of individual complications, like wound infections or late

incisional hernia that may benefit from a minimal invasive approach, is not possible based

on the available data.

Chapter 5 105

5

Regarding postoperative recovery, this review could not reliable identify clinically significant

benefits of the laparoscopic approach. Hospital stay was shorter for the laparoscopic ap-

proach in both the meta-analysis of non-RCTs (WMD -2.7 days) and the sensitivity analysis

with imputed data (WMD -2.1 days). However, this was not supported by the result of

the RCT. One explanation could be that the RCT used a hand-assisted approach for the

laparoscopic IPAA, while most other trials used a laparoscopic-assisted approach. Sensitiv-

ity analyses imputing missing data also suggested that the laparoscopic technique was

associated with less blood loss, time to bowel movement and time to regular diet. As

proved by these inconsistent findings, more data is needed before a final word could be

said regarding this matter.

A point worth considering is that the clinical relevance of such relatively small benefits

regarding the postoperative recovery remains questionable. With a complex operation like

the IPAA a one or a two day reduction of hospital stay is not likely to be decisive in the

choice of operative technique. Neither will be a slightly faster normalization of bowel func-

tion or return to normal diet. Especially with the increased implementation of ‘fast-track’

perioperative care programs, the differences between open and laparoscopic techniques

could diminish even further. ‘Fast-track’ programs have already been applied in a wide

variety of colonic operations and results seem to confirm their ability to accelerate recovery,

reduce morbidity and shorten hospital stay.52 In time, implementation of such programs

even in complex procedures, like the IPAA, seems likely. Therefore, future studies compar-

ing open and laparoscopic IPAA should probably focus on other clinically more relevant

outcomes, like specific complications, costs and long-term outcomes.

One of the most relevant long-term outcomes is maybe the cosmetic result of surgery and

its impact in this predominately young group of patients. In this review, two trials showed

a significantly shorter incision in the laparoscopic IPAA. Also, two trials reported results

of cosmesis and body image. Both trials observed significantly higher scores regarding

cosmesis in the laparoscopic group. Additionally, in the long-term follow-up of the only

RCT body image scores in female patients were found to be significantly higher in the

laparoscopic group.42 These results suggest that the shorter incision of the laparoscopic

approach is associated with long-lasting benefits in the perception of patients of their

own body, which could be an interesting argument in favor of the laparoscopic technique.

While these results need conformation from other trials, the social and psychological

impact could be significant. With an ever increasing patient’s awareness a smaller scar and

improved cosmesis could be important factor in guiding the choice of operative technique

in clinical practice.

106 Chapter 5

Costs also play an important role when evaluating new operative techniques, since fear of

increased costs could be a strong motive against their implementation. In this review only

one trial reported on costs and found no significant differences when the overall costs,

including admission, complications and readmission, were evaluated.42 The costs of the

laparoscopic procedure was, however, significantly higher than the open technique. Trials

comparing costs of other types of colonic surgery give contradicting results: some trials

show no differences in costs between laparoscopic and open approaches,53 while others

show results favoring the laparoscopic approach.54 It is therefore not possible to draw

reliable conclusion from these studies and future research is required to resolve this issue

as well.

In this context the ongoing LapCon-Pouch Trial is an important step in the right direction.34

This single-center, patient blinded trial will randomize 160 patients into two groups: a

totally laparoscopic IPAA group and a conventional open IPAA group. This trial will be the

largest randomized clinical trial and will have the largest number of total laparoscopic IPAA

patients included in any report so far. It will have the power to answer several interesting

question and with a follow-up period of 12 month more light could be shed on important

long-term outcomes. While the sample-size is not large enough to resolve all issues, it

should be noted that for a relatively infrequent operation as the IPAA a sample of this size

is substantial. It is therefore to be hoped that other high-volume centers will follow the

example of this trial and that in the near future a meta-analysis based solely on RCTs would

become feasible.

conclusions

implications for practiceThe laparoscopic IPAA is a safe procedure that could be performed successfully in centers

experienced in laparoscopic and restorative pouch surgery. The laparoscopic approach

seems to be associated with some short-term advantages regarding postoperative recov-

ery, but these advantages seem to be limited and their clinical significance is arguable. For

a complex operation like the IPAA other outcomes, like specific complications, long-term

functional outcome, cosmesis and costs are more likely to influence the choice of the

operative technique. This review have shown that for cosmesis there are some data favor-

ing the laparoscopic approach, but that the evidence is still inconclusive and more research

is needed before a general recommendation can be made. There is also some evidence that

costs, a crucial item in today’s health care, may not become a decisive item in the decision

between open and laparoscopic IPAA.

Chapter 5 107

5

implications for researchHigh volume colorectal surgical centers should include patients scheduled for laparoscopic

IPAA in high-quality RCTs with sufficiently long follow-up to be able to reliable assess

differences in relevant long-term outcomes. The focus of these trials should be specific

postoperative complications, cosmesis, quality of life and costs. Trials should provide data

suitable for meta-analysis in there published publications, or at least make it readily available

upon request. There is need for empirical evidence to guide the process of methodological

quality assessment of non- randomized controlled trials.

108 Chapter 5

references

1. Shivananda S, Lennard-Jones J, Logan R,

Fear N, Price A, Carpenter L, van Blanken-

stein M. Incidence of inflammatory bowel

disease across europe: Is there a difference

between north and south? Results of the

european collaborative study on inflamma-

tory bowel disease (ec-ibd). Gut. 1996; 39:

690-97.

2. Cohen JL, Strong SA, Hyman NH, Buie

WD, Dunn GD, Ko CY, Fleshner PR, et al.

Standards practice task force american so-

ciety of colon and rectal surgeons. Practice


ulcerative colitis. Diseases of the Colon &

Rectum. 2005; 48: 1997-2009.

3. Bulow S. Results of national registration of

familial adenomatous polyposis. Gut. 2003;

52: 742-46.

4. Jarvinen HJ. Epidemiology of familial

adenomatous polyposis in finland: Impact

of family screening on the colorectal cancer

rate and survival. Gut. 1992; 33: 357-60.

5. Parks AG, Nicholls RJ. Proctocolectomy

without ileostomy for ulcerative colitis.

British Medical Journal. 1978; 2: 85-88.

6. Cohen Z, McLeod RS, Stephen W, Stern HS,

O’Connor B, Reznick R. Continuing evolu-

tion of the pelvic pouch procedure. Annals

of Surgery. 1992; 216: 506-11.

7. Fazio VW, Ziv Y, Church JM, Oakley JR,

Lavery IC, Milsom JW, Schroeder TK. Ileal

pouch-anal anastomoses complications and

function in 1005 patients. Annals of Sur-

gery. 1995; 222: 120-27.






Digestive Surgery. 2005; 22: 69-79.

9. Marcello PW, Roberts PL, Schoetz Dj Jr,

Coller JA, Murray JJ, Veidenheimer MC.

Long-term results of the ileoanal pouch

procedure. Archives of Surgery. 1993; 128:

500-03.

10. Peters WR. Laparoscopic total proctocolec-

tomy with creation of ileostomy for ulcer-

ative colitis: Report of two cases. Journal of

Laparoendoscopic Surgery. 1992; 2: 175-78.

11. Casillas S, Delaney CP. Laparoscopic surgery

for inflammatory bowel disease. Digestive

Surgery. 2005; 22: 135-42.

12. Ky AJ, Sonoda T, Milsom JW. One-stage

laparoscopic restorative proctocolectomy:

An alternative to the conventional ap-

proach? Diseases of the Colon & Rectum.

2002; 45: 207-10.

13. Santoro E, Carlini M, Carboni F, Feroce A.

Laparoscopic total proctocolectomy with

ileal j pouch-anal anastomosis. Hepatogas-

troenterology. 1999; 46: 894-99.

14. Wexner SD, Johansen OB, Nogueras JJ, Jag-

elman DG. Laparoscopic total abdominal

colectomy - a prospective trial. Diseases of

the Colon & Rectum. 1992; 35: 651-55.

15. Reissman P, Salky BA, Pfeifer J, Edye M,

Jagelman DG, Wexner SD. Laparoscopic

surgery in the management of inflamma-

tory bowel disease. American Journal of

Surgery. 1996; 171: 47-50.

16. Sardinha TC, Wexner SD. Laparoscopy for

inflammatory bowel disease: Pros and cons.

World Journal of Surgery. 1998; 22: 370-74.

17. Schmitt SL, Cohen SM, Wexner SD, Nogu-

eras JJ, Jagelman DG. Does laparoscopic-

assisted ileal pouch anal anastomosis

reduce the length of hospitalization?

International Journal of Colorectal Disease.

1994; 9: 134-37.

18. Gill TS, Karantana A, Rees J, Pandey S,

Dixon AR. Laparoscopic proctocolectomy

with restorative ileal-anal pouch. Colorectal

Disease. 2004; 6: 458-61.

Chapter 5 109

5

19. Larson DW, Dozois EJ, Piotrowicz K,

Cima RR, Wolff BG, Young-Fadok TM.

Laparoscopic-assisted vs. Open ileal pouch-

anal anastomosis: Functional outcome in a

casematched series. Diseases of the Colon

& Rectum. 2005; 48: 1845-50.

20. Marcello PW, Milsom JW, Wong SK, Ham-

merhofer KA, Goormastic M, Church JM,

Fazio VW. Laparoscopic restorative proc-

tocolectomy: Case-matched comparative

study with open restorative proctocolec-

tomy. Diseases of the Colon & Rectum.

2000; 43: 604-08.

21. Ahmed Ali U, Keus E, Heikens JT, Goo-

szen HG, van Laarhoven CJHM. Open

versus laparoscopic (assisted) ileo pouch

anal anastomosis for ulcerative colitis or

familial adenomatous polyposis (protocol).

Cochrane Database of Systematic Reviews.

2006.

22. Higgins JPT, Green S. Cochrane handbook

for systematic reviews of interventions

version 5.1.0 [updated march 2011]. The

Cochrane Collaboration 2011 available

from www.cochrane-handbook.org.

23. Kjaergard LL, Villumsen J, Gluud C. Re-

ported methodological quality and discrep-

ancies between large and small randomised

trials in meta-analyses. Annals of Internal

Medicine. 2001; 135: 982-89.

24. Moher D, Pham B, Jones A, Cook DJ, Jadad

AR, Moher M, et al. Does quality of reports

of randomised trials affect estimates of

intervention efficacy reported in meta-

analyses? Lancet. 1998; 352: 609-13.

25. Schulz KF, Chalmers I, Hayer R, Altman D.

Empirical evidence of bias. JAMA. 1995;

273: 408-12.

26. Deeks JJ, Dinnes J, D’Amico R, Sakarovitch

C, Song F, Petticrew M, Altman DG, et al.

Evaluating non-randomised intervention

studies. Health Technology Assessment.

2003; 7: iii-x, 1-173.

27. Slim K, Nini E, Forestier D, Kwiatkowski

F, Panis Y, Chipponi J. Methodological

index for non-randomized studies (minors):

Development and validation of a new

instrument. ANZ Journal of Surgery. 2003;

73: 712-16.

28. Juni P, Witschi A, Block R, Egger M. The

hazards of scoring the quality of clinical

trials for meta-analysis. JAMA. 1999; 282:

1054-60.

29. Hozo SP, Djulbegovic B, Hozo I. Estimating

the mean and variance from the median,

range, and the size of a sample. BMC Medi-

cal Research Methodology. 2005; 5: 13-13.

30. Higgins JPT, Thompson SG. Quantifying

heterogeneity in a meta-analysis. Statistics

in Medicine. 2002; 21: 1539-58.

31. Begg CB, Mazumdar M. Operating charac-

teristics of a rank correlation test for publi-

cation bias. Biometrics. 1994; 50: 1088-101.

32. Egger M, Davey Smith G, Schneider M,

Minder C. Bias in meta-analysis detected

by a simple graphical test. BMJ. 1997; 315:

629-34.

33. Macaskill P, Walter SD, Irwig L. A compari-

son of methods to detect publication bias

in meta-analysis. Statistics in Medicine.

2001; 20: 641-54.

34. Antolovic D, Kienle P, Knaebel HP, Schmidt

J, Gutt CN, Weitz J, Koch M, et al. Totally

laparoscopic versus conventional ileoanal

pouch procedure--design of a single-centre,

expertise based randomised controlled

trial to compare the laparoscopic and

conventional surgical approach in patients

undergoing primary elective restorative

proctocolectomy--lapconpouch-trial. BMC

Surgery. 2006; 6: 13-13.

35. Ballantyne GH, Leahy PF. Hand-assisted

laparoscopic colectomy: Evolution to a

clinically useful technique. Diseases of the

Colon & Rectum. 2004; 47: 753-65

36. Bemelman WA, Hogezand van RA, Meijer-

ink W, Griffioen G, Ringers J. Laparoscopic-

assisted bowel resections in inflammatory

bowel disease: State of the art. Netherlands

Journal of Medicine. 1998; 53: S39–s46

110 Chapter 5

37. Chung CC, Tsang WWC, Kwok SY, Li

MKW. Laparoscopy and its current role

in the management of colorectal disease.

Colorectal Disease. 2003; 5: 528–43

38. Schwenk W, Haase O, Neudecker J, Müller

JM. Short term benefits for laparoscopic

colorectal resection. Cochrane Database of

Systematic Reviews. 2005

39. Tan JJY, Tjandra JJ. Laparoscopic surgery for

ulcerative colitis – a meta-analysis. Colorec-

tal Disease. 2006; 8: 626–36

40. Araki Y, Ishibashi N, Ogata Y, Shirouzo K,

Isomoto H. The usefulness of restorative

laparoscopic-assisted total colectomy for

ulcerative colitis. Kurume Medical Journal.

2001; 48: 99-103.

41. Larson DW, Cima RR, Dozois EJ, Davies M,

Piotrowicz K, Barnes SA, Wolff B, et al.

Safety, feasibility, and short-term outcomes

of laparoscopic illeal-pouch-anal anasto-

mosis - a single institutional case-matched

experience. Annals of Surgery. 2006; 243:

667-72.

42. Maartense S, Dunker MS, Slors JF, Cuesta

MA, Gouma DJ, van Deventer SJ, van Bo-

degraven AA, et al. Hand-assisted laparo-

scopic versus open restorative proctocolec-

tomy with ileal pouch anal anastomosis: A

randomized trial. Annals of Surgery. 2004;

240: 984-91.

43. Polle SW, Dunker MS, Slors JFM, Sprangers

MA, Cuesta MA, Gouma DJ, Bemelman

WA. Body image, cosmesis, quality of life,

and functional outcome of hand-assisted

laparoscopic versus open restorative

proctocolectomy: Long-term results of a

randomized trial. Surgical Endoscopy and

Other Interventional Techniques. 2007; 21:

1301-07.

44. Berdah SV, Barthet M, Emungania O, Orsoni

P, Alliot P, Grimaud JC, Brunet C. Two stage

videoassisted restorative proctocolectomy.

Early experience of 12 cases. Annales de

Chirurgie. 2004; 129: 332-36.

45. Otani Y, Nakamura T, Kanazawa H, Aihara

N, Kokuba Y, Ihara A, Kakita A. Article in

japanese. Japanese Journal of Gastroen-

terological Surgery. 2001; 34: 351-56.

46. Brown SR, Eu KW, Seow-Choen F. Consecu-

tive series of laparoscopic-assisted vs. Mini-

laparotomy restorative proctocolectomies.

Diseases of the Colon & Rectum. 2001; 44:

397-400.

47. Dunker MS, Bemelman WA, Slors JFM,

van Duijvendijk P, Gouma DJ. Functional

outcome, quality of life, body image, and

cosmesis in patients after laparoscopic as-

sisted and conventional restorative procto-

colectomy - a comparative study. Diseases

of the Colon & Rectum. 2001; 44: 1800-07.

48. Hashimoto A, Funayama Y, Naito H, Fu-

kushima K, Shibata C, Naitoh T, Shibuya

K, et al. Laparascope-assisted versus con-

ventional restorative proctocolectomy with

rectal mucosectomy. Surgery Today. 2001;

31: 210-14.

49. Tilney HS, Lovegrove RE, Heriot AG, Purkay-

astha S, Constantinides V, Nicholls RJ, Tek-

kis PP. Comparison of short-term outcomes

of laparoscopic vs open approaches to ileal

pouch surgery. Int J Colorectal Dis. 2007;

22: 531-42.

50. Young-Fadok TM, Dozois E, Sandborn

WJ, Tremaine WJ. A case matched study

of laparoscopic proctocolectomy and ileal

pouch-anal anastomosis (pc-ipaa) versus

open pc-ipaa for ulcerative colitis. Gastro-

enterology. 2001; Suppl: 2302-02.

51. Kienle P, Z’Graggen K, Schmidt J, Benner

A, Weitz J, Buchler MW. Laparoscopic

restorative proctocolectomy. British Journal

of Surgery. 2005; 92: 88-93.

52. Wind J, Polle SW, Fung Kon Jin PH, Dejong

CH, von Meyenfeldt MF, Ubbink DT, Gouma

DJ, et al. Systematic review of enhanced

recovery programmes in colonic surgery.

British Journal of Surgery. 2006; 93: 800-09.

Chapter 5 111

5

53. Pokala N, Delaney CP, Senagore AJ, Brady

KM, Fazio VW. Laparoscopic vs open total

colectomy: A case-matched comparative

study. Surgical Endoscopy and Other Inter-

ventional Techniques. 2005; 19: 531-35.

54. Senagore AJ, Luchtefeld MA, Mackeigan

JM, Mazier WP. Open colectomy versus lap-

aroscopic colectomy: Are there differences?

American Surgeon. 1993; 59: 549-53.

D. Scoglio1

U. Ahmed Ali2

A. Fichera1

1Department of Surgery, University of Washington Medical Center, Seattle, Washington, USA2Department of Surgery, University Medical Center Utrecht, Utrecht, The Netherlands

6 surGical treatment of ulcerative

colitis: ileorectal vs. ileal

pouch-anal anastomosis

World Journal of Gastroenterology, 2014.

114 Chapter 6

absTracT

backgroundTotal proctocolectomy with ileal pouch-anal anastomosis (IPAA) is the current gold stan-

dard in the surgical treatment of ulcerative colitis (UC) refractory to medical management.

However, it remains a procedure of significant magnitude and complications. colectomy

with ileorectal anastomosis (IRA) might offer an alternative with good long-term functional

results in a selected group of patients.

MethodsA review of literature is performed to compare the recent evidence on IRA vs. IPAA. Studies

including adult patients with UC and describing postoperative outcomes of IRA in terms of

morbidity, mortality, functional outcome and risk of cancer were identified and discussed.

resultsThe overall mortality was comparable after IRA (0 and 4.0%) versus IPAA (0 and 2.9%).

Overall morbidity for IRA was between 8% and 28%, which was lower than for IPAA

(15% to 63%). Failure of the reservoir was higher after IRA with a rate of 7% to 16% after

5-years, compared to 4.7% for IPAA. The risk of cancer was higher after IRA, with a rate

of 0 to 8% compared to <0.5% for IPAA, and a cumulative risk of cancer after 20 years

of 6% to 14% for IRA compared to 4.2% for the IPAA. Functional results seemed to be

better after IRA with lower frequency of bowel movements and less night-time seepage.

conclusionIPAA is the preferred approach for most patients with UC given its curative nature and

lower risk of failure and cancer. Nonetheless, patients with higher surgical risk with mild

rectal involvement and no evidence of dysplasia or cancer at the time of intervention might

benefit from the good functional outcomes of an IRA procedure. Patients should be aware

of the risk of recurrent disease and cancer, and be compliant with surveillance.

Chapter 6 115

6

inTroducTion

The main goals of surgical treatment for ulcerative colitis (UC) are not only to alleviate

symptoms and minimize cancer risk but also to obtain good functional outcomes and

improve quality of life. Until the 1950s total proctocolectomy with end-ileostomy (TPC)

was the only available procedure for UC patients failing medical management.

In the 1940s reports of subtotal colectomy with ileorectal anastomosis (IRA) as an alterna-

tive to TPC in selected patients were first published.1 During the 1950s and 1960s, Aylett

became the leading proponent of this procedure describing it as a way to avoid a permanent

stoma.2 At that time IRA represented a valid alternative to TPC in high selected patients

with minimal rectal inflammation. It was a less invasive operation, performed in one stage

and not requiring pelvic dissection with the associated risk of sexual dysfunction.3-5

In 1978, Parks et al. described an ileal pouch-anal anastomosis (IPAA).6 Since then, IPAA

has become the procedure of choice for patients affected by UC with excellent long-term

functional results, low risk of persistent cuff inflammation or neoplastic degeneration in

the retained rectum.7,8 Consequently, many surgeons have abandoned IRA in favor of IPAA

and TPC has remained an option for patients not candidates for IPAA. The counterargu-

ment is that IPAA, as a major procedure, carries its own risks including anastomotic failure

and pelvic sepsis that could result in poor pouch function, pouchitis and infertility in young

women, as well as pelvic nerve damage and portal vein thrombosis.3-5,9,10 In addition few

cases of cancer have been reported arising not only in the anal transitional zone but also

in the pouch itself.7,8

Interestingly, recent series of selected UC patients undergoing IRA showed long-term func-

tional results similar to IPAA.11-14 The aim of the current study was to review and compare

the most recent literature on IRA and IPAA as it pertains to postoperative morbidity and

mortality, failure rates, functional outcomes and cancer risk. It will help surgeons to provide

a tailored treatment for UC patients.

MeThods

literature search and study selectionTo identify relevant studies on the IRA procedure, we searched PubMed with predefined

search terms ‘ileorectal anastomosis ulcerative colitis’ on December, 2013. A total of 218

hits were identified and were screened by DS. For postoperative outcomes after IPAA in

terms of morbidity and mortality, a recent systematic review performed by the authors

116 Chapter 6

was used. For risk of cancer an additional search of PubMed with the terms pouch (pouch

ulcerative colitis AND cancer [tiab]) was performed, yielding 46 hits that were screened by

UAA.

Inclusion criteria were cohort studies including adult patients with UC and describing

postoperative outcomes of IRA in terms of morbidity, mortality, functional outcome and

risk of cancer. Exclusion criteria were case-reports, studies including a majority of patients

with other diseases than UC, mixed series with other interventions in which IRA results

were not presented separately and studies not reporting on relevant outcomes, written in

a language other than English, or presented in abstract form only.

study outcomesStudy outcomes were morbidity, especially serious procedure specific postoperative com-

plications such as pouch failure and anastomotic leak, morbidity, functional outcome and

long-term risk of cancer. Secondary outcomes included individual complications depending

on available data.

data extraction and analysisTitles and abstracts of all retrieved records, and subsequently full-text articles, were

examined by a single author (DS or UAA). The following data, were appropriate, were

extracted: inclusion period, sample size, duration of follow-up, reported outcomes and

results, including estimated cumulative risk of cancer if available.

Study characteristics and reported outcomes are discussed in a narrative review. Relevant

figure on study outcomes are presented in tables for ease of comparison. Due to difference

in indication between the two procedures, the study populations are quite different. This

combined with paucity of data on IRA A, did not allow for a meaningful (comparative)

meta-analysis of the procedures.

resulTs

ileorectal anastomosisUC begins in the rectum and extends proximally in a continuous fashion. The severity of

the disease also seems to be higher distally with the exception of fulminant pancolitis.

However, distal disease is sometimes alleviated by topical treatment and patients with mini-

mal rectal involvement and no dysplastic changes in the rectum could be considered for

IRA. Furthermore, an adequate rectal compliance and a normal anal sphincters function are

critical for good long-term results. These functions can easily be assessed by digital rectal

Chapter 6 117

6

examination but more accurately by rigid/flexible proctoscopy and manometry. Patients

with poor sphincter function, severe rectal disease, and non-distensible rectum should not

be offered an IRA. On the contrary, patients with colitis associated colorectal cancer and

advanced metastatic disease may benefit from an IRA because of their short life expectancy

and the palliative nature of their treatment.

Several studies have shown IRA for UC to be safe, with low postoperative morbidity and

mortality.11,13,15-17 During the years, overall morbidity has been reported between 8% and

28% and mortality between 0 and 4% (Table 1). Most studies focused their attention

on postoperative complications such as small bowel obstruction, anastomotic leak and

abdominal abscess. The fatal events were due to anastomotic leak (with subsequent sepsis)

and a pulmonary embolism.

Table 1. Morbidity and mortality after ileorectal anastomosis

Study Period N° Anastomotic leak (%)

Proctitis (%)

Need for proctectomy

(%)

Overall Morbidity

(%)

Mortality (%)

da Luz Moreira11 1971-2006 86 2.3 28 53 8 0

Leijonmarck13 1955-1984 51 3.9 45 57 16 4

Pastore15 1974-1990 48 4.4 10 17 23 0

Börjesson16 1997-2003 32 3.1 9.3 12 28 0

Grundfest17 1957-1977 89 9.1 11 21 16 0

Elton18 1990-1999 18 5.6 11 17 22 0

Andersson19 1992-2006 105 2.8 8.6 13 12 0

Lepistö20 1978-2000 20 - 45 35 - -

Oakley21 1960-1982 288 - 41 55 - -

The majority of published data has included mainly primary anastomosis with leak rates

ranging from 2% to 9%.11,13,15-19,22 Diverting ileostomies have been utilized in selective

cases at the surgeon discretion. Turnbull et al. suggested that preservation of grossly

involved recto-sigmoid colon was the main cause of IRA failure.23 In his opinion an anasto-

mosis at 6cm or less above peritoneal reflection improved rectal inflammation during the

first months and reduced the likelihood of IRA failure.

IRA does not involve extensive pelvic dissection, unlike IPAA, minimizing the risk of sexual

and urinary dysfunction. Hence, higher fertility rates may be expected in IRA patients com-

pared to IPAA although definitive studies providing evidence for better fertility rates in UC

patients are lacking. Thus, colectomy with IRA could be considered when treating women

in their reproductive age.22 Some authors have shown acceptable long-term success rate

after IRA.2,11,13,15,18,20 Aylett et al. reported on a total of 300 cases operated on over a

118 Chapter 6

ten-year period with only 7% failure rates.2 Lepistö et al. and Pastore et al. reported a

cumulative probability of having a functioning IRA at five years of 84%.15,20 Elton et al. had

88% success in their 18 patients, but the follow-up in that study was shorter.18 Ten year cu-

mulative success (69%) in Lepistö’s series was higher than reported by Leijonmarck (51%)

in 1990.13,20 At 20-years the current probability of having a functioning IRA has ranged

between 46 and 69 percent.22 A recent study compared 22 IRA with 66 IPAA patients

matched for age, gender, and follow-up time, including IRAs performed in the past 25

years showed a cumulative probability of having a functioning IRA at 5, 10, 15 and 20 years

of 81, 74, 56 and 46 percent respectively in accordance with previously published work.11

In terms of functional results, the da Luz Moreira’s series reported six bowel movements

per day (range 2-11), 1/22 (5%) night-time seepage and 15/22 (68%) reporting frequent

urgency.11 Leijonmarck et al. showed four bowel movements per day and none during

the night, with 100% of continence (25% of patients are on antidiarrheal medication),

after a mean follow-up of 13 years.13 Elton et al. had no significant difference between

preoperative and 1-year postoperative stool frequency, 11/12 patients had no problems

with continence, and three were using antidiarrheal medication.18 Pastore et al. described

a median number of six bowel movements per day (range 2-20).15 The median number of

nocturnal bowel movements was 1 (range 0-10) and three patients had more than eight

daily stools with frequent soiling and urgency. At the time of follow-up, anti-diarrheal

medications were taken by 53.3% of patients, whereas 31.3% required low doses of

systemic or topical steroids. More than 90% of patients considered that their health status

had improved after the operation. Quality of life was improved in 84%.

All the studies above showed that IRA is a safe procedure with an acceptable function

and quality of life, but unfortunately it is not necessarily a definitive operation, especially

for young patients. Specifically, Andersson et al. reported an estimated cumulative failure

rate of 10.1% and 24.1% respectively at 5 and 10 years.19 In his series Leijonmarck et al.

had 57% of failure after 13-year follow-up.13 Pastore et al. suggested that time between

IRA and additional surgery in his series was 3.9 ± 4.7 years.15 In da Luz Moreira’s series, 38

patients (44%) continued to have a functioned IRA after a median follow-up of 11 years

(range 1-30 years).11 The rectum was resected in 46 (53%) of 86 patients and the median

follow-up between IRA and completion proctectomy was 10 years (range 1-33 years).

The main indication for proctectomy is recurrent proctitis refractory to medical manage-

ment, followed by dysplasia or cancer, and the development of Crohn’s disease.11,13,15-21

The options for these patients include IPAA, Brooke ileostomy, or a continent ileostomy

(Kock pouch). Very Often IPAA can be safely performed in the majority of these patients

thus preserving bowel continuity and avoiding permanent fecal diversion.11

Chapter 6 119

6

Cancer risk after IRA

Mucosal dysplasia is a premalignant pathological state associated with long standing UC.24

Dysplasia in general is considered an indication for surgery in UC, even though this para-

digm is rapidly changing. Epithelial dysplasia of the colon and rectum was graded as mild,

moderate, or severe depending on whether the upper one-third, upper two-third, or entire

glands displayed nuclear anisocytosis and hyperchromatism, as well as loss of nuclear po-

larity and the normal goblet cell configuration of colonic mucosa. Since dysplastic changes

were often patchy, only the highest degree of dysplasia was considered. Johnson et al. has

shown in 1983 that the probability of developing rectal adenocarcinoma after a diagnosis

of mild or severe dysplasia in IRA patients reached 42% at nine years from diagnosis.25

The rate of dysplasia and cancer, in patients with UC, increases with time and leaving the

rectum in place contributes to the increased risk. The overall cumulative probability of

rectal dysplasia in the retained rectum increases from 9% at 10 years to 25% at 20 years.11

The overall incidence of rectal cancer after an IRA varies in the literature based on length

of follow-up, ranging from 0 to 18%. Grundfest et al. reported on four patients who

developed carcinoma of the rectum during their study period (4.8% at 8-year follow-up),

although he estimated the risk of rectal cancer to be 13% at more than 25 years of follow-

up.17 Oakley et al. found nine patients with rectal cancer in the stump (3.1%) at an 8-year

follow-up while Andersson et al. showed an overall risk of cancer of 1.9% at a 5.4-year

follow-up.19,21 However, some series reported higher rates of degeneration as Baker et al.

who described, in 1978, a cumulative cancer risk of 6% after 20 years rising to 18% after

35 years in a series of 374 unselected patients.26 In da Luz Moreira’s series, the cumulative

probability of developing dysplasia and cancer was 7, 9, 20 and 25% and 0, 2, 5 and 14%

at 5, 10, 15 and 20 years respectively.11 On the other hand, Leijonmarck et al. and Lepistö

et al. have reported no case of cancer in more recent series at 13 and 18-year follow-up,

respectively.13,20 Pastore et al. showed a cumulative probability of remaining free of cancer

around 85.5% at 12 years (95%CI: 57.7-100).15

Most patients who develop rectal cancer in the retained rectum presented at an advanced

stage (stage III-IV) suggesting the possibility of a more aggressive biology and making

close surveillance imperative.11,27 For instance, in Baker’s study 62% of patients who had

developed rectal cancer died within three years of diagnosis. Johnson et al. reported a total

of 10 rectal cancers, 8 of which had either nodal or distant metastases.27 The patients in

the series reported by Oakley et al. fared better, with just 2 of 9 patients with rectal cancer

dying over a 22-year time period.21 Rectal biopsies taken from multiple sites every 6 to 12

months are advised following IRA in UC patients. If dysplasia is found, completion proctec-

tomy is indicated. Patients with long standing UC who are not able or willing to undergo

surveillance should not be offered an IRA. It is also important to emphasize that colectomy

with IRA should not be offered to patients with preexisting dysplasia or cancer due to the

120 Chapter 6

increased risk of further neoplastic degeneration.28 In addition, the presence of dysplasia or

cancer in the resected colon should cause particular concern about the fate of the remain-

ing rectum suggesting that a completion proctectomy would be indicated in these cases.

In fact, Oakley et al. reported on five surviving patients who had cancer in their colonic

specimens; three of the five were found on follow-up to have cancer or severe dysplasia in

the rectal remnant.12 Grundfest et al. described nine patients with a colitis-associated colon

cancer or severe dysplasia who underwent subtotal colectomy, eight of whom survived; of

the eight, five developed severe dysplasia or cancer in the retained rectum.17

ileal Pouch-anal anastomosisRestorative proctocolectomy with IPAA is currently the procedure of choice for the surgical

treatment of UC. The main reason for its popularity is its avoidance of a permanent stoma

with stable functional results and good quality of life. In over 30 years of its existence,

the IPAA has undergone several refinements in the quest of achieving optimal results.

Examples include different shapes of the pouch, different anastomotic techniques, use of

defunctioning ileostomy and various dissection methods.29-31 Surgeons have also obtained

greater experience and familiarity with the technique, which has also benefited outcomes.

A large body of literature exists on the outcomes of IPAA. Most studies, however, are

retrospective cohorts reporting outcomes from a single institution. Due to large variations

between studies, an overview is needed for reliably assessment of the IPAA outcomes. A

meta-analysis of 43 observational studies, all published before 2000, has provided pooled

estimates of complications and functional outcomes after IPAA.32 This meta-analysis showed

a pouch failure risk of 6.8% (95%CI: 5.4-8.4), increasing to 8.5% (95%CI: 5.4-13.2) when

only patients with a minimal follow-up of 5 years were considered.32 Other pouch related

complications were also studied. Pelvic sepsis and pouch fistulas, both major postoperative

complications, were observed in 9.5% (95%CI: 8.2-10.9) and 5.5% (95%CI: 4.3-7.0),

respectively. Sexual dysfunction was present in 3.4% (95%CI: 2.7-4.7), while pouchitis was

reported in 18.8% (95%CI: 15.7-22.4).

A recent meta-analysis, including 53 studies published after 2000, showed significant

improvements in some but not all of these complications.33 The overall rate of pouch

failure was significantly reduced to 4.3% (95%CI: 3.5-5.3), and pouch failure after at least

5 years of follow-up was 4.7% (95%CI: 3.4-6.4). An trends towards improvement was also

seen in other complications, although differences were not statistically significant. Pelvic

sepsis, pouch fistula and sexual dysfunction were reported in 7.5% (95%CI: 6.1-9.1),

4.5% (95%CI: 3.5-5.7) and 3.0% (95%CI: 1.7-5.2) of patients. The only complication

showing a substantial increase was pouchitis, with a rate of 26.8% (95%CI: 21.0-33.5).

Chapter 6 121

6

Thus it seems that the rate of complications after IPAA has declined over time (Table 2).

The authors of the meta-analyses have noticed that the decline was largest in the earlier

period of the IPAA, but seems to have continued over time. Nonetheless, IPAA remains a

complex surgery with substantial risk of morbidity. Unfortunately, these reviews did not

provide estimates of the overall morbidity since large variations exists in definition between

studies. In a recent review comparing laparoscopic to open IPAA, we observed an overall

morbidity rate after open IPAA ranging from 15% to 63%, with a total of 127 events in

306 patients (42%).34 Severe postoperative complications were reported by 6 studies with

rates ranging from 5% to 27%, with a total 28 events in 269 patients (14%).34 The high

rate of pouchitis is also worrisome, since this complication can affect functional outcomes,

quality of life and might also increase risk of dysplasia in the pouch.35,36 It should be noted

that the meta-analyses discussed above did not distinguish between acute and chronic

pouchitis, which is an important distinction in terms of course and health implications.35-37

Table 2. Morbidity and moratlity after ileal pouch-anal anastomsosis

Study(Period)

N ofstudies

N ofpatients

Pelvic sepsis(%, 95%CI)

Pouch failure(%, 95%CI)

Pouchitis(%, 95%CI)

Mortality(%, range)

Heuting 2005(Studies<2000)32

43 9,317 9.5 (8.2–10.9) 6.8 (5.8–8.4) 18.8 (15.7–22.4) -

De Zeeuw 2012(Studies ≥2000)33

53 14,966 7.5 (6.1–9.1) 4.3 (3.5–5.3) 26.8 (21.0–33.5) 0 (0 – 2.9)

CI: confidence interval.

Functional outcomes after IPAA were similar in studies published before and after 2000.33

Average frequency of bowel movements per 24 hours was 5.9 (95%CI: 5.0-6.9), of which

1.5 (95%CI: 1.0-2.1) overnight. Mild and severe faecal incontinence were reported in

14.3% (7.3-25.9) and 6.1% (2.9-12.3) of patients, respectively. The authors conclude that

functional outcomes of IPAA may be determined by an intrinsic limitation of the IPAA

procedure, rather than growing expertise or technical refinement. This is in line with other

studies showing no improvement in functional outcomes based on technical developments,

such as type of anastomosis or laparoscopic approach.34,37 However, most patients consider

the functional outcome after IPAA to be highly satisfactory, with good quality of life and

social functionality that are comparable to those in a healthy reference population.38,39

As expected, achieving these adequate quality of life scores was highly correlated with

achieving of good functional outcomes.40

Cancer risk after IPAA

The IPAA has as an important advantage the removal of the whole colon and virtually the

entire rectum as part of the procedure. This minimizes chances of colon and rectal cancer

in this high-risk population. A proctocolectomy should be considered almost mandatory

122 Chapter 6

when dysplasia is present. Even when only low-grade dysplasia has been identified by colo-

noscopy, the risk remains substantial. In such patients, studies show a risk of concomitant

cancer or high-grade dysplasia of 15% and a 5-year progression rate of up to 54% if not

operated on.28,41

When a double-stapled approach for IPAA is used, a mucosal remnant at the anal transition

zone (ATZ) is left in place. The risk of cancer in this area is a matter of controversy. In three

series with long-term follow-up focused on this outcome, dysplasia and cancer in the anal

transitional zone after stapled pouch surgery was found to be infrequent.7,42,43 Dysplasia

was observed in 8/178 (4.4%), 7/210 (3.3%) and 0/135 (0%) after at least 10 years of

follow-up. In most of these cases, dysplasia developed in the first 2 to 3 years and often

disappeared on repeated biopsies. None of the series found cancer in the ATZ after such

prolonged follow-up. These data strongly emphasize the extent to which IPAA minimizes

the risk of cancer.

The best evidence regarding the development of dysplasia and adenocarcinoma after IPAA

can be obtained from a recent study from the Cleveland Clinic, in which 3,203 patients

undergoing an IPAA from 1984 to 2009 were analysed.44 Cumulative incidences for pouch

neoplasia at 5, 10, 15, 20, and 25 years were 0.9%, 1.3%, 1.9%, 4.2%, and 5.1%, re-

spectively. Overall, 23 patients (0.72%) developed dysplasia, while 11 (0.36%) developed

adenocarcinoma of the pouch and/or the ATZ. Risk factors for pouch neoplasia were also

evaluated. Preoperative established cancer (hazard ratios (HR) 13.43, 95%CI: 3.96-45.53,

p<0.001) or dysplasia (HR 3.62, 95%CI: 1.59-8.23, p=0.002) were the only independent

factors associated with increased risk of pouch neoplasia. Mucosectomy did not protect

against this risk, and the rate of pouch cancer was actually higher after mucosectomy with

a rate of 1.3% (6/451) compared to 0.3% (9/2734) after the double-stapled approach. The

authors concluded that the risk for neoplasia in patients with UC and IPAA is small, and

that it is mainly determined by the presence of preoperative dysplasia or cancer.

Additionally, in a review of literature, 26 published case reports were identified between

1984 and 2008.45 Certain observations from this review are noteworthy. First, of the 26

carcinomas, 14 (52%) arose from rectal mucosa or from the anal transition zone, while

6 (23%) were from ileal pouch mucosa. Second, adenocarcinomas developed after

mucosectomy in 17 patients, and after a double-stapled approach in 8 patients (1 case

not reported). Also worth noting, the indication for the IPAA was due to neoplasia in 19

patients (9 cancers and 10 dysplasia) and non-neoplasia in 6 patients. The median time for

development of pouch lesions was the shortest in patients operated on for cancer (median

3 years), compared to a median of 6.5 in the other patients. This review is in line with

results from the above mentioned study, and further establishes the following conclusions:

Chapter 6 123

6

1) the low number of reported cases, 2) cancer can develop both after mucosectomy

or double-stapled approach, 3) the close relationship between surgery for neoplasia and

development of cancer.

The review was not able to estimate the incidence of cancer after IPAA, since the total

number of IPAA cases was not stated in most case reports. Branco et al. did publish their

own case as part of this review, which was the first case they observed in a cohort of 520

patients (0.2%) from 1978 to 2008.45 This percentage is also in line with the Cleveland

study.44

Despite this seemingly small risk, surveillance of selected patients has been recommended

by some authors.46,47 This approach might especially be important in UC patients with

dysplasia or cancer present at time of surgery, or patients with retained rectal mucosa and

active inflammation (i.e. cuffitis). Also the presence of chronic pouchitis might be a valid

indication for surveillance, since this has been associated with increased risk of low-grade

dysplasia (odds ratio 13.48, P < 0.02), as well as high-grade dysplasia (3/66 vs 0/210,

p=0.01).36 An overview of the current evidence on the risk of cancer after IRA vs. IPAA is

presented in Table 3.

Table 3. Risk of cancer after ileorectal vs. ileal pouch-anal anastomosis in ulcerative colitis

Ileorectalanastomosis

Period N Follow-up average (years)

Overall cancerrate (%)

Estimated cumulative risk after 20 years (%)

da Luz Moreira11 1971-2006 86 9 8 14

Leijonmarck13 1955-1984 51 13 0 -

Pastore15 1974-1990 48 6.3 2 14.3*

Börjesson16 1997-2003 32 3.5 0 -

Grundfest17 1957-1977 89 8 4.8 5 ± 3.5

Elton18 1990-1999 18 2.6 - -

Andersson19 1992-2006 105 5.4 - 2.1

Lepistö20 1978-2000 20 18 0 -

Oakley21 1960-1982 288 8.2 3.1 -

Baker26 1952-1976 374 >10 5.9 6 ± 2

ileal pouch-analanastomosis

Period N Follow-up average (years)

Overall cancerrate (%)

Estimated cumulative risk after 20 years (%)

Kariv44 1984-2009 3,203 ±12 0.4 4%

Branco45 1978-2008 520 ±15 0.2 -

* Cumulative risk at 12 years (rather than 20).

124 Chapter 6

conclusion

In the current era IPAA is the preferred approach for patients with UC requiring surgical

treatment. The removal of all diseased mucosa and the lower risk of cancer after IPAA

compared to IRA are the main advantages of this technique (Table 4). Therefore, IPAA

should certainly be performed when the rectum is actively involved in the disease or when

dysplasia or cancer are present in any part of the colon or rectum. Nonetheless, there is

still a role for IRA and TPC for selected patients and for patients not candidates for IPAA.

Total abdominal colectomy with IRA is justified in UC patients with normal anal sphincters

tone without severe perineal disease, and spared and distensible rectum with no evidence

of dysplasia or cancer at the time of intervention. It can be also proposed to young women

as a possible interim procedure based on concerns for infertility after IPAA.

The risk of cancer is of particular concern in the comparison between these two tech-

niques. Current evidence shows a large variation in the reported rates of cancer after IRA

from 0% to 8%. For IPAA, this risk is much smaller, and two large series have shown a rate

of cancer of about 0.3%. Few studies have calculated the cumulative risk of cancer as well.

Similarly, estimated cumulative risk of cancer after 20 years was higher after IRA (6% to

14%) compared to IPAA (4.2%) (Table 4).

Therefore, every patient undergoing IRA should be informed about the risk of recurrent

proctitis and cancer in long standing disease. They have to fully understand the need for

meticulous surveillance and agree to comply with at least yearly endoscopy with rectal

Table 4. Main advantages and disadvantages of ileorectal anastomosis (IRA) and ileal pouch-anal anastomosis (IPAA)

IRA IPAA

Advantages Easier operation Lower risk of cancer

Lower overall morbidity No need for medical therapy

Lower risk of urinary and sexual dysfunction Less urgency

Fewer bowel movements per day

Better continence

Disadvantages Need for maintenance therapy Major operation

Risk of recurrent/persistent disease Risk of postoperative complications

Higher risk of neoplastic degeneration Pouchitis

Need for strict surveillance

More dietary and work restrictions

Chapter 6 125

6

biopsies. Unless these conditions are met, patients should not be offered an IRA. Also,

patients with widely metastatic colorectal cancer may benefit from an IRA as a palliative

procedure.

Functional results seem to be better after IRA with lower frequency of bowel movements

and less night-time seepage but with more urgency compared to patients with an IPAA.

The overall quality of life is similar, although the IRA group has significantly more dietary

and work restrictions.11 Finally, TPC still remains the procedure of choice in patients with

impaired anal sphincter function and high-risk of pouch failure.

126 Chapter 6

references

1. Baker WN. The results of ileorectal anas-

tomosis at st mark’s hospital from 1953 to

1968. Gut. 1970; 11: 235-9.

2. Aylett SO. Three hundred cases of diffuse

ulcerative colitis treated by total colectomy

and ileo-rectal anastomosis. British medical

journal. 1966; 1: 1001-5.

3. Hueting WE, Gooszen HG, van Laarhoven

CJ. Sexual function and continence after

ileo pouch anal anastomosis: A comparison

between a meta-analysis and a question-

naire survey. International journal of


4. Gorgun E, Remzi FH, Goldberg JM, Thorn-

ton J, Bast J, Hull TL, Loparo B, et al. Fertility

is reduced after restorative proctocolec-

tomy with ileal pouch anal anastomosis: A

study of 300 patients. Surgery. 2004; 136:

795-803.

5. Waljee A, Waljee J, Morris AM, Higgins

PD. Threefold increased risk of infertility: A

meta-analysis of infertility after ileal pouch

anal anastomosis in ulcerative colitis. Gut.

2006; 55: 1575-80.



British medical journal. 1978; 2: 85-8.

7. Remzi FH, Fazio VW, Delaney CP, Preen

M, Ormsby A, Bast J, O’Riordain MG, et

al. Dysplasia of the anal transitional zone

after ileal pouch-anal anastomosis: Results

of prospective evaluation after a minimum

of ten years. Diseases of the colon and

rectum. 2003; 46: 6-13.

8. Das P, Johnson MW, Tekkis PP, Nicholls

RJ. Risk of dysplasia and adenocarcinoma

following restorative proctocolectomy for

ulcerative colitis. Colorectal disease: the



2007; 9: 15-27.

9. Remzi FH, Fazio VW, Oncel M, Baker ME,

Church JM, Ooi BS, Connor JT, et al. Portal

vein thrombi after restorative proctocolec-

tomy. Surgery. 2002; 132: 655-61; discus-

sion 61-2.

10. Sagap I, Remzi FH, Hammel JP, Fazio VW.

Factors associated with failure in manag-

ing pelvic sepsis after ileal pouch-anal

anastomosis (ipaa)--a multivariate analysis.

Surgery. 2006; 140: 691-703; discussion

03-4.

11. da Luz Moreira A, Kiran RP, Lavery I. Clini-

cal outcomes of ileorectal anastomosis for

ulcerative colitis. The British journal of

surgery. 2010; 97: 65-9.

12. Oakley JR, Jagelman DG, Fazio VW, Lavery

IC, Weakley FL, Easley K, Farmer RG. Com-

plications and quality of life after ileorectal

anastomosis for ulcerative colitis. American


13. Leijonmarck CE, Lofberg R, Ost A, Hellers G.

Long-term results of ileorectal anastomosis

in ulcerative colitis in stockholm county.

Diseases of the colon and rectum. 1990; 33:

195-200.

14. Gruner OP, Flatmark A, Naas R, Fretheim B,

Gjone E. Ileorectal anastomosis in ulcerative

colitis. Results in 57 patients. Scandinavian

journal of gastroenterology. 1975; 10:

641-6.

15. Pastore RL, Wolff BG, Hodge D. Total

abdominal colectomy and ileorectal anas-

tomosis for inflammatory bowel disease.


1455-64.

16. Borjesson L, Lundstam U, Oresland T, Brev-

inge H, Hulten L. The place for colectomy

and ileorectal anastomosis: A valid surgical

option for ulcerative colitis? Techniques in

coloproctology. 2006; 10: 237-41; discus-

sion 41.

17. Grundfest SF, Fazio V, Weiss RA, Jagelman

D, Lavery I, Weakley FL, Turnbull RB, Jr. The

risk of cancer following colectomy and

Chapter 6 127

6

ileorectal anastomosis for extensive mu-

cosal ulcerative colitis. Annals of surgery.

1981; 193: 9-14.

18. Elton C, Makin G, Hitos K, Cohen CR. Mor-

tality, morbidity and functional outcome

after ileorectal anastomosis. The British


19. Andersson P, Norblad R, Soderholm JD,

Myrelid P. Ileorectal anastomosis in com-

parison with ileal pouch anal anastomosis

in reconstructive surgery for ulcerative

colitis--a single institution experience.

Journal of Crohn’s & colitis. 2014; 8: 582-9.

20. Lepisto A, Jarvinen HJ. Fate of the rectum

after colectomy with ileorectal anastomosis

in ulcerative colitis. Scandinavian journal of

surgery: SJS: official organ for the Finnish

Surgical Society and the Scandinavian

Surgical Society. 2005; 94: 40-2.

21. Oakley JR, Lavery IC, Fazio VW, Jagelman

DG, Weakley FL, Easley K. The fate of the

rectal stump after subtotal colectomy for


rectum. 1985; 28: 394-6.

22. da Luz Moreira A, Lavery IC. Ileorectal

anastomosis and proctocolectomy with end

ileostomy for ulcerative colitis. Clinics in

colon and rectal surgery. 2010; 23: 269-73.

23. Turnbull RB, Jr. Surgical treatment of ulcer-

ative colitis: Early results after colectomy

and low ileorectal anastomosis. Diseases of


24. Morson BC, Pang LS. Rectal biopsy as an

aid to cancer control in ulcerative colitis.

Gut. 1967; 8: 423-34.

25. Johnson WR, McDermott FT, Pihl E, Hughes

ES. Mucosal dysplasia. A major predictor

of cancer following ileorectal anastomosis.


697-700.

26. Baker WN, Glass RE, Ritchie JK, Aylett SO.

Cancer of the rectum following colectomy

and ileorectal anastomosis for ulcerative

colitis. The British journal of surgery. 1978;

65: 862-8.

27. Johnson WR, McDermott FT, Hughes ES,

Pihl EA, Milne BJ, Price AB. The risk of

rectal carcinoma following colectomy in


rectum. 1983; 26: 44-6.

28. Kiran RP, Ahmed Ali U, Nisar PJ, Khoury W,

Gu J, Shen B, Remzi FH, et al. Risk and loca-

tion of cancer in patients with preoperative

colitis-associated dysplasia undergoing

proctocolectomy. Annals of surgery. 2014;

259: 302-9.







surgery. 2006; 244: 18-26.







406-12.













33. de Zeeuw S, Ahmed Ali U, Donders RA,

Hueting WE, Keus F, van Laarhoven CJ.

Update of complications and functional

outcome of the ileo-pouch anal anastomo-

sis: Overview of evidence and meta-analysis

of 96 observational studies. International


843-53.

128 Chapter 6








35. Abdelrazeq AS, Lund JN, Leveson SH.

Implications of pouchitis on the functional

results following stapled restorative proc-

tocolectomy. Diseases of the colon and

rectum. 2005; 48: 1700-7.

36. Banasiewicz T, Marciniak R, Paszkowski J,

Krokowicz P, Kaczmarek E, Walkowiak J,

Szmeja J, et al. Pouchitis may increase the

risk of dysplasia after restorative procto-

colectomy in patients with ulcerative colitis.




37. Schluender SJ, Mei L, Yang H, Fleshner PR.

Can a meta-analysis answer the question:

Is mucosectomy and handsewn or double-

stapled anastomosis better in ileal pouch-

anal anastomosis? The American surgeon.

2006; 72: 912-6.

38. Heikens JT, de Vries J, van Laarhoven CJ.

Quality of life, health-related quality of life

and health status in patients having restor-

ative proctocolectomy with ileal pouch-anal

anastomosis for ulcerative colitis: A system-

atic review. Colorectal disease: the official



536-44.

39. Thirlby RC, Sobrino MA, Randall JB. The

long-term benefit of surgery on health-

related quality of life in patients with

inflammatory bowel disease. Archives of

surgery (Chicago, Ill.: 1960). 2001; 136:

521-7.

40. Carmon E, Keidar A, Ravid A, Goldman G,

Rabau M. The correlation between quality

of life and functional outcome in ulcerative

colitis patients after proctocolectomy

ileal pouch anal anastomosis. Colorectal



Ireland. 2003; 5: 228-32.

41. Thomas T, Abrams KA, Robinson RJ, May-

berry JF. Meta-analysis: Cancer risk of low-

grade dysplasia in chronic ulcerative colitis.

Alimentary pharmacology & therapeutics.

2007; 25: 657-68.

42. O’Riordain MG, Fazio VW, Lavery IC, Remzi

F, Fabbri N, Meneu J, Goldblum J, et al.

Incidence and natural history of dysplasia of

the anal transitional zone after ileal pouch-

anal anastomosis: Results of a five-year to

ten-year follow-up. Diseases of the colon

and rectum. 2000; 43: 1660-5.

43. Coull DB, Lee FD, Henderson AP, Anderson

JH, McKee RF, Finlay IG. Risk of dysplasia in

the columnar cuff after stapled restorative

proctocolectomy. The British journal of

surgery. 2003; 90: 72-5.

44. Kariv R, Remzi FH, Lian L, Bennett AE, Kiran

RP, Kariv Y, Fazio VW, et al. Preoperative

colorectal neoplasia increases risk for

pouch neoplasia in patients with restorative

proctocolectomy. Gastroenterology. 2010;

139: 806-12, 12.e1-2.

45. Branco BC, Sachar DB, Heimann TM, Sarpel

U, Harpaz N, Greenstein AJ. Adenocarcino-

ma following ileal pouch-anal anastomosis

for ulcerative colitis: Review of 26 cases.

Inflammatory bowel diseases. 2009; 15:

295-9.

46. Duff SE, O’Dwyer ST, Hulten L, Willen

R, Haboubi NY. Dysplasia in the ileoanal

pouch. Colorectal disease: the official jour-

nal of the Association of Coloproctology of

Great Britain and Ireland. 2002; 4: 420-9.

47. Hurlstone DP, Shorthouse AJ, Cross SS,

Brown S, Sanders DS, Lobo AJ. High-

magnification chromoscopic pouchoscopy:

A novel in vivo technique for surveillance of

the anal transition zone and columnar cuff

following ileal pouch-anal anastomosis.

Techniques in coloproctology. 2004; 8: 173-

8; discussion 78.

U. Ahmed Ali

B. Shen*

F.H. Remzi

R.P. Kiran

Departments of Colorectal Surgery and *Gastroenterology, Digestive Disease Institute, Cleveland Clinic, Cleveland, Ohio, USA

7 the manaGement of anastomotic

pouch sinus after ileal pouch-

anal anastomosis (ipaa)


132 Chapter 7

absTracT

backgroundAnastomotic sinus is a relatively uncommon complication after an ileal pouch anal anasto-

mosis (IPAA). Disease course is poorly defined and management can be challenging. The

aim of this study is to evaluate frequency, management and outcome of this complication.

MethodsWe performed a retrospective designed cohort study from a prospectively collected

database of a high volume specialized Colorectal Surgery department. Patients with an

anastomotic sinus after pouch surgery from 1997 to 2009 were included. Management

and outcomes of these patients were evaluated. Main endpoints were the rates of sinus

healing and pouch failure.

resultsFrom 2286 patients who underwent an IPAA procedure, 45 (2.0%) patients were identified

with an anastomotic pouch sinus. There were 32 (71%) males, and the mean age was 40

(±13) years. The pouch sinus was initially managed by observation in 23 (51%) patients,

drainage of the sinus in 9 (20%), unroofing of the sinus tract in 8 (18%), sinus closure in 3

(7%) and a diverting ileostomy in 2 (4%). In 28 patients (62%), subsequent treatment was

necessary. Sinus healing was achieved in 27 (60%) patients, whereas 15 (33%) eventually

developed pouch failure. Of the treatment modalities applied, a strategy with observation

as initial treatment was the most successful with a healing rate of 65%. The healing rate

was significantly lower in symptomatic patients compared to asymptomatic patients (30%

vs. 84%, p=0.001). Pouch failure was also higher (45% vs. 24%, p=0.14). No other factors

associated with healing rate or pouch failure were identified.

conclusionAnastomotic pouch sinuses after pouch surgery are associated with a high rate of pouch

failure. Symptomatic presentation is a significant predictor for low healing rates and is

associated with a high risk of pouch failure. Observation and watchful monitoring is the

initial treatment of choice when permitted by the patient’s condition.

Chapter 7 133

7

inTroducTion

In an ileal pouch anal anastomosis (IPAA) procedure, the integrity of the anastomosis is an

important determinant of the long-term viability of the constructed pouch. Anastomotic

leak and fistulae are well known complications that could threaten the structure of the

anastomosis and are important predictors of pouch failure.1,2 An anastomotic sinus of

the pouch is a less known condition that occurs in 2.8 to 8% of patients undergoing an

IPAA procedure.3-5 It is thought to be a special form of anastomotic leak in which the leak

is confined to a blind-ending track.3-5 The clinical impact of an anastomotic sinus of the

pouch can be significant. It can delay ileostomy closure when shown in the precedent

Gastrografin enema and require prolonged evaluation and follow-up. It also can require

intervention aimed at facilitating the closure of the sinus, with added costs and burden to

the patient and health system. Furthermore, it has the potential to eventually cause pouch

failure, resulting in major surgical interventions such as pouch revision, redo pouches or

pouch excision with hospital admission and significant potential morbidity.3,5 Addition-

ally, there have been previous reports of malignant changes occurring in various types of

fistulae, draining sinuses and chronic inflammatory tracts.6-9

The optimal management of anastomotic sinuses in an ileoanal pouch is not well known.

Many surgeons choose to delay ileostomy reversal and repeat the pouchogram at a later

time, hoping for eventual spontaneous resolution, which has been reported to occur in

approximately 40 to 50% of patients.3,10 Several surgical therapies have been described

for the management of patients with persistent sinuses including debridement, unroofing,

occlusive treatment with fibrin glue, pouch revision and redo pouch.3-5,10,11 Results of these

interventions vary greatly between reports, and large series are very scarce.5 In this study,

we set out to evaluate the frequency, management and outcome of anastomotic pouch

sinuses after IPAA in a specialized colorectal surgical center.

PaTienTs and MeThods

PatientsAll patients who underwent restorative proctocolectomy in the study period were included

in a prospective ileal pouch database. Patients who developed an anastomotic pouch sinus

after IPAA between 1997 and 2009 were identified from this prospectively maintained

institutional review board approved database.

134 Chapter 7

inclusion and exclusion criteriaPatients who had a primary or redo IPAA for ulcerative colitis or indeterminate colitis and

subsequently developed a pouch sinus were included.

surgical treatmentAt our institution, most patients undergo a double-stapled ileal J-pouch anastomosis

with a defunctioning loop ileostomy. The choice of operation is based on surgeon prefer-

ence, with the selective performance of a one-stage proctocolectomy and IPAA without

a defunctioning ostomy based on an individual patient’s circumstances.12 In the absence

of any specific concerns after surgery, we routinely perform a gastrografin enema just

before closure of the ileostomy to ensure integrity of the pouch anastomosis. In patients

with complications (e.g. pelvic sepsis or anastomotic leaks) other approaches, like exam

under anesthesia (EUA), pouchoscopy and concomitant closure of ileostomy if satisfactory

healing is deemed to have occurred, may be chosen by the surgeon. All ileostomy closures

were performed at our institution.

definition of a pouch sinus and diagnostic modalitiesA pouch sinus was defined as a blind-ending tract arising from the area of the pouch

anal anastomosis. All pouch sinuses were identified by imaging modalities (pouchogram)

or endoscopy or by examination under anesthesia (EUA) performed by an experienced

colorectal surgeon. Pouchograms were performed by means of Gastrografin enema.

study variablesData regarding patient demographics (age, gender and preoperative diagnosis), medical

and surgical history, diagnostic modality and management of the sinus were collected. Fac-

tors potentially influencing outcome of treatment were collected and analyzed. These were

pouch construction (primary pouch vs. redo pouch), history of postoperative anastomotic

leak or pelvic sepsis after IPAA, symptomatic presentation of sinus and time to ileostomy

closure (before vs. after ileostomy closure). The primary outcomes evaluated were sinus

healing and pouch failure. Healing was defined as complete healing of the sinus confirmed

by imaging (pouchogram), endoscopy or EUA. Pouch failure was defined as any condition

necessitating permanent diversion through an ileostomy or excision of the pouch. Thus,

patients who had successful redo of the pouch as part of the treatment with an eventually

healed sinus were considered to have developed pouch failure, although such a sinus was

deemed to have eventually healed from the standpoint of healing rates.

data and statistical analysisCategorical variables were summarized as frequencies and percentages. Quantitative vari-

ables were summarized by medians with ranges and mean with standard deviations (SD).

Chapter 7 135

7

The different treatment modalities and associated healing rate are presented and discussed.

The course of pouch sinuses diagnosed before ileostomy closure and those diagnosed after

ileostomy closure were compared. Statistical testing of categorical variables was performed

with a Chi-square test, with a p-value of 0.05 as cut-off point for statistical significance.

resulTs

Patients and characteristicsBetween 1997 and 2009, a total of 2286 patients underwent an IPAA procedure at the

Colorectal Surgical Department of the Cleveland Clinic. A total of 45 (2.0%) patients with a

chronic anastomotic pouch sinus were identified. The mean age of these patients at pouch

surgery was 40 (±13) years old, and 32 (71%) patients were male. Preoperative diagnosis

was ulcerative colitis in 41 (91%) patients and indeterminate colitis in 4 (9%). Final pathol-

ogy showed ulcerative colitis in 30 (67%), indeterminate colitis favoring ulcerative colitis in

7 (16%), indeterminate colitis in 7 (16%) and Crohn’s disease in 1 (2%) patient. At subse-

quent follow-up, mucosal apperance and histology acquired during pouch examination did

raise the suspicion of Crohn’s disease in 9 (20%) patients. Preoperatively, 26 (58%) patients

were on 1 or more immunomodulatory drugs. Thirty-five (78%) patients developed an

anastomotic pouch sinus after a primary IPAA procedure. The procedure was performed as

a 1-stage procedure in 1 patient, 2-stage procedure in 19 patients and a 3-stage procedure

in 15 patients. The remaining 10 (22%) developed the pouch sinus after a redo IPAA. A

double-stapled approach was used in 34 (76%) patients and handsewn anastomosis was

used in 11 (24%) patients. All but 2 patients (96%) had a defunctioning loop ileostomy.

Twenty-four (53%) patients had a previous anastomotic leak or pelvic abscess.

Twenty (44%) patients were asymptomatic at the time of diagnosis, while 25 (56%) pre-

sented with one or more symptoms. Symptoms in patients who were diverted included

fever, perianal / back pain and rectal drainage. Symptoms in patients who were not diverted

were fever, perianal / abdominal pain, pouchitis like symptoms (pain, urgency), severe

rectal burning, diarrhea, irritation and itching. A pouch sinus was noticed before closure of

the diverting ileostomy in 23 (51%) patients, and after closure of the diverting ileostomy

in the remaining 22 (49%) patients. All patients who developed a sinus after ileostomy

closure had normal pre-takedown evaluation. The median duration between IPAA surgery

and ileostomy closure was 5 months (IQR 3 to 8).

Anastomotic pouch sinus was initially diagnosed by endoscopy in 23 (51%) patients, Gas-

trografin enema in 15 (33%) patients and by exam under anesthesia (EUA) in the remaining

7 (16%) patients (Figure 1).13 Diagnosis was confirmed in all patients either by additional

136 Chapter 7

diagnostic modalities or by imaging modalities used for the treatment and follow-up of

patients. Of the 23 (51%) patients who were initially diagnosed with a pouch sinus on

pouchoscopy, all but 5 patients had this diagnosis confirmed by atleast one other form of

imaging. In 12 patients, the diagnosis was confirmed by a gastrografin enema. In these 12

patients, no discrepancies were seen between gastrografin enema and pouchoscopy. The

median duration between surgery and the initial diagnosis of the sinus was 6 months (IQR

2 to 22.5). In patients diagnosed with a sinus before ileostomy closure, the median dura-

tion from IPAA to diagnosis of the sinus was 2 months (IQR 2 to 3). For patients diagnosed

with a sinus after ileostomy closure this median duration was 22.5 (IQR 8 to 64) months.

ManagementThe initial management of a pouch sinus included watchful observation in 23 (51%) pa-

tients, drain placement in the sinus in 9 (20%) patients, unroofing of the sinus in 8 (18%)

patients, sinus closure in 3 (7%), construction of a diverting ileostomy in 2 (4%) patients

(Table 1). For patients undergoing observation as initial treatment, the first reassessment of

the sinus after its initial detection was performed after a median of 2 months (IQR 1 to 3.5

months). After initial treatment, 28 (62%) patients required additional treatments (Figure

2). These included unroofing of the sinus in 14 (50%) patients, drainage in 3 (11%), fibrin

glue in 2 (7%), sinus closure in 2 (7%), diversion in 5 (18%), pouch redo in 1 (4%) and

pouch excision in 1 (4%) patients. Nine (20%) patients required one more subsequent

treatment, including unroofing (x2), drainage (x1), fibrin glue (1x), diversion (x1), redo

pouch (x2) and pouch excision (x2).

Figure 1. Distal pouch sinus at the anastomosis on pouchoscopy (A) and pouchogram by means of a Gastrografin enema (B) (arrows)

Chapter 7 137

7

Table 1. Treatment patterns of anastomotic pouch sinus and associated rates of healing and pouch failure

Patterns of treatment NSinus Healed

n (%)Pouch failure

n (%)

Number of required treatments

Only 1 treatment 17 12 (71%) 5 (29%)

Two or more treatments 28 15 (54%) 10 (36%)

Initial treatment*

Observation 23 15 (65%) 5 (22%)

Drainage 9 5 (56%) 3 (33%)

Unroofing 8 4 (50%) 4 (50%)

Sinus closure 3 1 (33%) 2 (67%)

Diversion 2 2 (100%) 1 (50%)

* Healing and pouch failure rates represent the final outcome of patients after all interventions.

Sinus (n=45)

Observation (n=23)

Drainage (n=9)

- Unroofing (n=4) - Drainage (n=1) - Diversion (n=1) - Drainage &

unroofing (n=1) - Diversion & redo

pouch (n=1)

- Unroofing (n=4) - Diversion (n=2) - Improved, but not yet healed (n=2)

- Unroofing (n=1) - Sinus closure (n=1) - Fibrin glue (n=1)

- Unroofing (n=1) - Diversion (n=1) - Unroofing & redo pouch (n=1) - Fibrin glue & drainage (n=1) - Not healed (n=1)

- Unroofing (n=1) - Sinus closure & Diversion (n=1)

Diversion (n=2)

- Unroofing (n=1) - Drainage & Fibrin glue (n=1)

Unroofing (n=8)

- Redo pouch (n=1)

- Pouch excision (n=1) - Not helaed (n=3)

None

Sinus not healed or pouch failure.

Sinus healed after additional treatment.

Sinus healed after initial treatment.

Healed (n=7)

Sinus closure (n=3)

Healed (n=1) Healed (n=3) Healed (n=1) -

None

Figure 2. A diagram showing the treatment modalities applied for pouch sinus and the outcome of these treatments

138 Chapter 7

outcomesProven complete healing of the sinus was eventually achieved in 27 (60%) of patients

(Table 2). The median time from the diagnosis of the sinus to healing was 7.5 months

(IQR 2 to 28). In two patients substantial reduction of the size of the sinus was seen on

imaging, but the sinus was not yet completely healed. Two patients lacked recent imaging

to confirm healing of the sinus, although these patients had normally functioning pouches.

A total of 15 (33%) patients eventually had pouch failure.

The influence of various factors on the healing and pouch failure rates is presented (Table

2). The healing rate was slightly higher (63% vs. 50%, p=0.76) and pouch failure rate lower

(29% vs. 50%, p=0.21) in patients who underwent a primary IPAA procedure compared

to patients who developed a sinus after a redo IPAA procedure, respectively. Patients with

a history of an overt postoperative anastomotic leak or pelvic sepsis had similar healing

rate of the sinus (63% vs. 57%, p=0.49) and pouch failure rate (38% vs. 29%, p=0.53)

as patients without a history of these complications. Symptomatic sinus presentation was

significantly associated with a lower healing rate (30% vs. 84%, p=0.001), pouch failure

was higher but did not reach statistical significance (45% vs. 24%, p=0.14). Finally, pouch

sinuses noticed before closure of the ileostomy had a higher rate of healing (70% vs. 50%,

p=0.33) compared to sinuses noticed after closure of ileostomy, but pouch failure rates

were quite similar (30% vs. 36%, p=0.67). Eighteen patients had an incidental finding of a

sinus tract on preoperative investigation performed to ensure anastomotic integrity before

loop ileostomy closure. These were managed at the discretion of the operating surgeon by

Table 2. Evaluation of potential factors associated with healing and pouch failure in patients with an anastomotic pouch sinus

Potential risk factors NHealing rate

n (%)P-value

Pouch failuren (%)

P-value

All patients 45 27 (60%) - 15 (33%) -

Pouch construction

- Primary pouch 35 22 (63%)0.756

10 (29%)0.205

- Redo pouch 10 5 (50%) 5 (50%)

History of anastomotic leak or pelvic sepsis

- Yes 24 15 (63%)0.490

9 (38%)0.526

- No 21 12 (57%) 6 (29%)

Symptomatic sinus presentation

- Yes 20 6 (30%)0.001

9 (45%)0.138

- No 25 21 (84%) 6 (24%)

Time to ileostomy closure

- Before closure 23 16 (70%)0.334

7 (30%)0.673

- After closure 22 11 (50%) 8 (36%)

Chapter 7 139

7

means of observation in 10 (55%), drainage in 4 (22%), unroofing in 2 (11%), and sinus

closure in 2 (11%) patients. Healing rate was 83%, while pouch failure rate was 22% for

these patients.

Of the treatment modalities applied, a strategy with observation as initial treatment was

the most used treatment strategy (23 (51%) of patients), with a healing rate of 65%. In

treating asymptomatic patients all treatment modalities seemed more successful than in

symptomatic patients (Table 3). Observation resulted in higher healing rates for asymptom-

atic compared to symptomatic patients (79% vs. 44%, p=0.19, respectively). The use of

invasive treatment modalities was fairly similar in asymptomatic and symptomatic groups

of patients, but healing rates were generally higher in asymptomatic patients. Initial treat-

ment with drainage (100% vs. 20%, p=0.016) and unroofing (100% vs. 0%, p=0.005)

were both significantly associated with higher healing rates in asymptomatic patients as

when compared to symptomatic patients.

Table 3. Treatments and healing rates of anastomotic pouch sinus in patients with asymptomatic vs. symptomatic presentation

Patterns of treatment

N

Asymptomatic Symptomatic

P-valuen Healed n (%) n Healed n (%)

Number of treatments

Only 1 treatment 17 11 9 (82%) 6 3 (50%) 0.0793

Two or more treatments 28 14 12 (85%) 14 3 (21%) 0.012

Initial treatment*

Observation 23 14 11 (79%) 9 4 (44%) 0.188

Drainage 9 4 4 (100%) 5 1 (20%) 0.016

Unroofing 8 4 4 (100%) 4 0 (0%) 0.005

Sinus closure 3 2 1 (50%) 1 0 (0%) 0.386

Diversion 2 1 1 (100%) 1 1 (100%) -

* Healing and pouch failure rates represent the final outcome of patients after all interventions.

In treating patients diagnosed with pouch sinuses before vs. after ileostomy closure a

slight difference in the initial treatment modalities preferred was noted (Table 4). More

patients with sinuses diagnosed before ileostomy closure were treated with drainage (7

vs. 2 patients) and sinus closure (3 vs. 0 patients) compared to patients diagnosed after

ileostomy closure. On the other hand, unroofing was a more preferred initial treatment in

patients diagnosed with sinuses after ileostomy closure (6 out of 8 patients).

Observation was performed initially in an equal proportion of patients in both groups with

similar rates of success (Table 4). In patients where the pouch sinus was noticed before

diverting ileostomy closure, the ileostomy was closed after complete healing of the sinus

140 Chapter 7

in most patients. In 5 (22%) patients, however, the sinus was deemed acceptable enough

to allow closure of the ileostomy before complete healing. In all cases, the size of the sinus

was noted to be getting progressively smaller and the surgeons deemed the sinus small

enough to not pose a threat to the integrity of the anastomosis. Two of these patients had

an initial asymptomatic presentation, and both of them achieved complete healing. The 3

other patients had a symptomatic presentation initially; none of them achieved complete

healing after ileostomy closure, and all experienced pouch failure eventually. Of the 10

patients with a suspicion of Crohn’s disease, the sinus healed in 5 (50%) patients and 4

(40%) patients developed pouch failure.

discussion

An anastomotic pouch sinus is a relatively uncommon complication after IPAA. Its outcome

and consequences for treatment are not well studied. Only few reports have presented

outcomes of this condition in the last 15 years, most of which are relatively small (Table

5). Moreover, previous studies have typically presented the outcome of pouch sinuses in

highly selected populations based on patient presentation or type of treatment.3-5,11 There

is, therefore, a paucity of data regarding the treatment and outcome of pouch sinuses,

especially from large series. The present study has shown that pouch sinuses are associated

with pouch failure in about 33% of cases, making it an important risk factor for pouch

failure. This is comparable to the pouch failure rate of patients suffering from pouch-

related fistulae.1 We also showed that the risk for pouch failure was not equal across all

patients. Patients with an asymptomatic presentation tend to have a much better outcome

Table 4. Treatments and healing rates in patients diagnosed with an anastomotic pouch sinus before and after ileostomy closure

Patterns of treatment

N

Before IL closure After IL closure

P-valuen Healed n (%) n Healed n (%)

Number of treatments

Only 1 treatment 17 9 8 (89%) 8 4 (50%) 0.117

Two or more treatments 28 14 8 (57%) 14 5 (50%) 0.340

Initial treatment*

Observation 23 11 8 (73%) 12 7 (85%) 0.599

Drainage 9 7 5 (71%) 2 0 (0%) 0.073

Unroofing 8 2 2 (100%) 6 2 (33%) 0.102

Sinus closure 3 3 1 (33%) 0 - -

Diversion 2 0 - 2 2 (100%) -

IL: diverting ileostomy.* Healing and pouch failure rates represent the final outcome of patients after all interventions.

Chapter 7 141

7

in terms of healing of the sinus and preservation of a functional pouch. These findings

are important in that they can raise the alertness of physicians to this condition, provide

guidance in the treatment of its different presentations and help counsel patients.

The overall healing rate in this study was about 60%, which is lower than healing rates

reported in previous studies (Table 5). However, when only asymptomatic patients are

observed, the healing rate of 84% was fairly similar to previous reports. The improved out-

come of asymptomatic patients can potentially be explained by differences in the severity

of the condition. Smaller non-inflamed sinuses are more likely to present asymptomatically

than larger or inflamed sinuses. This could also explain the tendency of asymptomatic

sinuses to heal more often under conservative treatment. Another explanation could be

the effect of clinical symptoms on the choice of treatment. Severe clinical symptoms may

necessitate the abandonment of a conservative approach with the use of surgical interven-

tions earlier, with a potentially less favorable outcome.

The optimal treatment for anastomotic pouch sinuses is not well defined. Most surgeons

adapt a conservative wait-and-see strategy in which they observe the patient and repeat

imaging modalities at regular intervals to monitor the healing of the sinus.3,5 Results from

this study support this conservative approach, especially for asymptomatic patients. When

a conservative watchful waiting approach is not possible or insufficient to achieve complete

healing, other treatments may be applied. Several treatments have been described, includ-

ing drainage of the sinus, surgical closing of the sinus and permanent fecal diversion.3,11

There are no studies comparing these treatments, and all available evidence is descriptive.

In 1997, Whitlow et al. described a new method for treating pouch sinuses by unroof-

ing of the anastomosis, whereby the common wall between the sinus and the adjacent

bowel lumen (i.e. pouch) is divided.4 In 6 patients treated by this approach, all achieved

complete sinus healing by 12 months after surgery. In the current study, 8 patients were

Table 5. Studies describing the outcome of anastomotic pouch sinuses after ileal pouch anal anas-tomosis

Study N Selection criteria Initial treatments Healing rate

Akbari 20093 22Asymptomatic, before ileostomy

closureObservation

(± debridement)20/21 (95%)

Nyam 19975 41Asymptomatic, before ileostomy

closureObservation 40 / 41 (95%)

Swain 200411 7 Fibrin glue treatment Fibrin glue 7 / 7 (100%

Whitlow 19974 6 Unroofing as treatment Unroofing 6 / 6 (100%)

Current study 45Symptomatic and asymptomatic,

irrespective of ileostomy closure or type of treatment

Depending on patient condition and

presentation27 / 45 (60%)

EUA: exam under anesthesia

142 Chapter 7

treated with unroofing of the sinus as the initial treatment with 4 (50%) patients achieving

complete healing. All 4 patients had an asymptomatic presentation of the sinus. Addition-

ally, 14 patients underwent unroofing of the sinus as an additional treatment (9 after

observation, 3 after drainage, 1 after sinus closure and 1 after diversion). Seven patients

achieved complete healing of the sinus, 5 of which were asymptomatic. In all, 50% (11

of 22 patients) had healing. However, unroofing seems to be a promising treatment in

asymptomatic sinuses (81% (9 of 11 patients) healed), but results seem to be less favorable

in symptomatic patients (18% (2 of 11 patients) healed).

The use of fibrin glue for closure of pouch sinuses was also recently reported by Swain et

al.11 Only 3 patients were treated by this approach in our series, with 2 patients achieving

complete healing (Figure 2). With regards to transanal drainage of the sinuses, in total 12

patients had this treatment at any time and 8 (66%) patients eventually healed. Again,

healing rate in asymptomatic patients was much higher (100% (7 out of 7 patients))

compared to symptomatic patients (20% (1 out of 5 patients)). Treatments utilized before

or after ileostomy closure did not differ substantially (Table 4). In both groups, conservative

treatment was chosen first in the majority of patients with comparable success rates (73%

vs. 85% for before and after ileostomy closure, respectively). Some small differences did

exist (slightly greater use of drainage and sinus closure in the group that developed a

sinus before ileostomy closure, and greater use of unroofing of the sinus in the group

that developed a sinus after ileostomy closure), but these differences probably reflect an

individual physician’s choice in a particular patient. Diversion was performed in 8 out of

the 22 patients with a sinus diagnosed after ileostomy closure (Figure 2). In the major-

ity of cases, therefore, other treatment modalities were considered more appropriate in

this group of patients. This is understandable, since an anastomotic sinus, especially if

asymptomatic, may be a relatively small defect, which may not always justify a re-operation

for diversion. When other measures fail, a redo pouch is the ultimate surgical option in

the treatment of pouch sinuses. Three (1%) patients were treated with a redo pouch in

this series. Two (66%) of these patients achieved successful healing and had a functional

pouch. The third patient, unfortunately, did not benefit from this procedure and the new

pouch was eventually excised.

The current study includes prospectively collected information for a large cohort of patients

undergoing IPAA who developed pouch sinuses. Although it does have the disadvantages

of a retrospective design, the inclusion of a relatively large number of patients who have

varying presentations of this rare and complex condition may provide better information

that helps guide physicians faced with its management. The small numbers of patients

in the subgroups may have also led to a type II error, resulting in an insignificant p-value.

Nevertheless, the details of the outcomes with the various treatment strategies will provide

Chapter 7 143

7

information previously not clearly available. This may help patients and clinicians make

decisions when faced with the quandary of how best to manage this difficult condition.

Based on the results of this study, we propose the following algorithm for the management

of a sinus after IPAA. Sinuses detected in patients who have intestinal continuity restored

and are otherwise asymptomatic, are best left alone without intervention. When a pouch

sinus is incidentally encountered (e.g. on a Gastrografin enema) prior to ileostomy clo-

sure, we suggest delaying ileostomy closure for a few months until healing of the sinus is

demonstrated. Symptomatic sinuses and those that are defunctioned, but non-healing on

watchful waiting alone, may be managed by trans-anal drain placement, unroofing of the

sinus or injection of fibrin glue. Consideration may be given to simply closing the ileostomy

in selected asymptomatic patients with a small persistent sinus after due discussion with

the patient about the pros and cons of this approach. On the other hand, the creation of a

defunctioning ostomy may be considered in circumstances where eventual healing of the

sinus may be expected in patients who are not defunctioned. When attempts at healing

have failed or may not reasonably be anticipated due to the size or persistence of the sinus,

a redo IPAA may be considered in symptomatic patients and those with large sinuses.

conclusion

Anastomotic pouch sinuses after IPAA are associated with a high rate of pouch failure.

Symptomatic presentation is a significant predictor for low healing rates and is associated

with a high risk of pouch failure. Observation and watchful monitoring is the treatment of

choice when permitted by the patient’s condition. Optimal choice of subsequent treatment

is not well studied and comparative studies are needed.

144 Chapter 7

references

1. Nisar PJ, Kiran RP, Shen B, Remzi FH, Fazio

VW. Factors associated with ileoanal pouch

failure in patients developing early or late

pouch-related fistula. Dis.Colon Rectum.

2011; 54: 446-53.

2. Forbes SS, O’Connor BI, Victor JC, Cohen

Z, McLeod RS. Sepsis is a major predictor of

failure after ileal pouch-anal anastomosis.

Dis.Colon Rectum. 2009; 52: 1975-81.

3. Akbari RP, Madoff RD, Parker SC, Hager-

man G, Minami S, Bullard Dunn KM,

Mellgren AF. Anastomotic sinuses after

ileoanal pouch construction: Incidence,

management, and outcome. Dis.Colon

Rectum. 2009; 52: 452-55.

4. Whitlow CB, Opelka FG, Gathright JB,

Jr., Beck DE. Treatment of colorectal and

ileoanal anastomotic sinuses. Dis.Colon

Rectum. 1997; 40: 760-63.

5. Nyam DC, Wolff BG, Dozois RR, Pemberton

JH, Mathison SM. Does the presence of

a pre-ileostomy closure asymptomatic

pouch-anastomotic sinus tract affect the

success of ileal pouch-anal anastomosis?

J.Gastrointest.Surg. 1997; 1: 274-77.

6. Cruickshank AH, McConnel EM, Miller DG.

Malignancy in scars, chronic ulcers, and

sinuses. J.Clin.Pathol. 1963; 16: 573-80.

7. Gifford J, Saltzstein SL, Barone RM.

Adenocarcinoma occurring in association

with a chronic sinus tract and biliary fistula.

Cancer. 1981; 47: 2093-97.

8. McCune WS, Thistlethwaite JR. Fistula

cancer. Ann.Surg. 1959; 149: 815-20.

9. Skir I. Mucinous carcinoma associated with

fistulas of long-standing. Am.J.Surg. 1948;

75: 285-89.

10. Arumainayagam N, Chadwick M, Roe

A. The fate of anastomotic sinuses after

total mesorectal excision for rectal cancer.

Colorectal Dis. 2009; 11: 288-90.

11. Swain BT, Ellis CN. Fibrin glue treatment of

low rectal and pouch-anal anastomotic si-

nuses. Dis.Colon Rectum. 2004; 47: 253-55.

12. Remzi FH, Fazio VW, Gorgun E, Ooi BS,

Hammel J, Preen M, Church JM, et al. The

outcome after restorative proctocolectomy

with or without defunctioning ileostomy.

Dis Colon Rectum. 2006; 49: 470-77.

13. Tang L, Cai H, Moore L, Shen B. Evaluation

of endoscopic and imaging modalities in

the diagnosis of structural disorders of the

ileal pouch. Inflamm.Bowel Dis. 2010; 16:

1526-31.

U. Ahmed Ali1

S.T. Martin2

A.D. Rao3

R.P. Kiran3

1 Department of Surgery, University Medical Center Utrecht, Utrecht, The Netherlands2 Department of Colorectal Surgery, St. Vincent’s University Hospital, Dublin, Ireland3 Division of Colorectal Surgery, Department of Surgery, Columbia University Medical Center, New York, USA

8 impact of preoperative

immunosuppressive aGents on

postoperative outcomes in

crohn’s disease


148 Chapter 8

absTracT

backgroundImmunosuppressive agents are essential in the management of Crohn’s disease. Their

safety before surgery is, however, still controversial. Therefore, this review aims to evalu-

ate whether preoperative use of immunosuppressive agents is associated with increased

postoperative complications.

MethodsA literature search of PubMed, EMBASE, and The Cochrane Library was undertaken. All

studies describing postoperative outcomes of patients undergoing bowel resections for

Crohn’s disease were included if they reported data comparing patients on preoperative

immunosuppressive agents to an appropriate control group. Main outcomes were overall

and infectious complications.

resultsTwenty-one eligible studies (20 retrospective and 1 prospective) with 6899 patients were

included. When individual studies were examined, only 2/14 (14%), 4/13 (31%) and 1/8

(13%) studies found an association between postoperative complications and preoperative

anti-tumor necrosis (TNF) agents, corticosteroids and thiopurines, respectively. In meta-

analysis, patients on anti-TNF agents (Relative Risk (RR) 1.29, 95% Confidence Interval

(CI) 1.07 to 1.55), and corticosteroids (RR 1.55, 95% CI 1.23 to 1.95) were found to have

a higher risk of postoperative infectious complications. The use of anti-TNF agents was

also significantly associated with wound infection (RR 1.62, 95% CI 1.12 to 2.34) and

septic shock (RR 1.81, 95% CI 1.03 to 3.17). There was no association between use of

thiopurines or combined immunomodulator drugs and postoperative complications.

conclusionsPatients with Crohn’s disease on preoperative immunosuppressive agents are at higher risk

for complications. Both corticosteroids and anti-TNFα agents may increase risk of infections

and septic shock. A preoperative drug free interval, when feasible, might be considered to

reduce the risk of infections. The adoption of any operative strategies that modify these

outcomes may additionally counter these risks.

Chapter 8 149

8

inTroducTion

Crohn’s disease (CD) is a chronic, inflammatory disease of the gastrointestinal tract, char-

acterized by alternating periods of relapse and remission. Traditionally, medical treatment

consisted of aminosalicyclates, antibiotics, immunosuppressive and modulating agents

such as corticosteroids, thiopurines (6-mercaptopurine, azathioprine) and methotrexate.1,2

Recently, anti-tumor necrosis alpha (TNF) drugs have been used to induce and maintain

remission in moderate to severe CD.3,4 Increasingly, there is a vogue for “top-down” ap-

proach to the treatment of patients with moderate and severe colitis, with early treatment

with anti-TNF drugs followed by other immunomodulator drugs, if needed, with reported

superior clinical outcomes.5 The use of these agents has been shown to facilitate remission,

improve quality of life, reduce steroid dependency and delay surgery.4,6 Controversy regard-

ing the potential adverse effects of these agents on outcomes of surgery, however, exists.7,8

Data suggest that these agents may be independently associated with septic complications

in the post-operative period, when used as monotherapy or in combination.9-11 Previous

data from our institution implicated infliximab use in the 3 months preceding surgery in

the increased rates of post-operative sepsis, abscess and readmission,9 other reports do

not show any association between adverse septic complications after surgery and the use

of infliximab and/or steroids.12-16 In fact, one study has suggested improved outcomes for

patients receiving immunomodulating agents.17 The aim of this systematic review is to

determine whether pre-operative immunosuppression and/or the use of immunodulator

agents are associated with adverse outcomes following surgery in Corhn’s disease patients.


study selectionA systematic literature search with predefined search terms was carried out in MEDLINE,

EMBASE and The Cochrane Library for articles published until February 20th, 2013 (Figure 1).

inclusion criteriaTitles and abstracts of all identified records were screened and selected according to the

following inclusion criteria: study population - adult patients with CD; intervention - bowel

resection procedures for CD; study period - published in 1990 or later; research question -

comparison of the effect of any preoperative immunosuppressive agent on the postopera-

tive outcomes of surgery; study outcomes – the study reported at least one of the outcome

measures listed below; publication - published as a full article in a peer-reviewed journal.

If studies reported a mix of patients that included CD, the study was only included if CD

was the most common disease reported. Studies were excluded if written in a language

150 Chapter 8

other English (only 1 Korean study was excluded due to language), or were presented in

abstract form only.

study outcomesPrimary outcomes were total overall complications and total infectious complications. In-

fectious complications included wound infections, wound dehiscence, anastomotic leaks,

(intra-abdominal) abscess, sepsis, pneumonia, peritonitis, bacteremia, and other compli-

cations thought to have an infectious etiology. Secondary outcomes included individual

complications depending on available data.

data extractionTitles and abstracts of all retrieved records, and subsequently full-text articles, were exam-

ined independently by 2 authors (UAA and SM). The following data were also extracted

independently by these two authors: inclusion period, population characteristics, study

design, reported outcomes and results. The time period in which the medications were

administered prior to surgery (drug window) for patients in the treatment group was also

extracted. Discrepancies were resolved by discussion and agreement. Accuracy of search

results and data extraction was independently assessed by a third investigator (ADR).

risk of confounding biasConsidering the high risk of confounding in non-randomized studies, we adapted a simple

system for evaluating the risk of confounding in studies included in our meta-analysis.

• Lowriskofconfoundingbias: studies thatpresenteddatabasedonmatchedanalysis

(matched for important confounders).

• Mediumriskofconfoundingbias:studiesthatdidnotcorrectforpotentialconfound-

ers, but presented complete baselines, with equal characteristics for important potential

confounders.

• Highriskofconfoundingbias:studiesthatdidnotcorrectforpotentialconfounders,and

did not present baseline characteristics or presented unequal baseline characteristics.

statistical analysisStudy characteristics and reported outcomes were presented. Outcomes reported by 2 or

more studies were pooled in meta-analyses. Pooled results were presented for all studies

(total group) and stratified for risk of confounding (low, medium and high groups). In

addition, meta-analysis of only the low and medium groups was presented. A sensitivity

analysis excluding studies with mixed populations of patients (i.e. studies that addition-

ally included patients with diagnoses other than Crohn’s disease) was also performed to

determine the influence of these data on the overall results.

Chapter 8 151

8

Meta-analysis results were presented as risk ratios (RRs) with 95% confidence intervals (CI).

Empirical evidence suggests that there is no difference in consistency across trials between

RR and OR when variations in control group event rates are present.18 This fact combined

with the possibility of overestimating a benefit when OR are used, have made us choose

RR.19 Statistical heterogeneity was defined as presence of variability in the intervention ef-

fects between studies, and is considered reflective of clinical or methodological differences.

Heterogeneity was calculated using Higgins χ²-test and quantified by the I² measure.20

I² values of 25%, 50%, and 75% are considered to represent low, moderate, and high

heterogeneity, respectively.20 For this review, a χ²-test with a p-value of < 0.05 or I² > 50%

were considered to indicate significant heterogeneity. The fixed-effect model using the

Mantel–Haenszel method was used if no significant heterogeneity was present. In cases of

significant heterogeneity, the random-effects model using DerSimonian and Laird method

was used, and this is indicated in the results. Non-randomized studies with zero events in

one arm were included in the analysis by adding a continuity correction of 0.5 to the cells

in the 2 × 2 table of that study. Review Manager (version 5) provided by The Cochrane

Collaboration was used for data management and statistical analyses.

resulTs

description of studiesA total of 2787 potentially eligible reports were identified by our search. Through a two

stage selection process (Figure 1), a total of 21 eligible studies were identified.9,12-17,21-34

These studies included a total of 6899 patients. The inclusion period in these studies

ranged from 1981 to 2010. All studies included CD patients, although 5 studies addi-

tionally included patients with other diagnoses. All but one study were retrospectively

designed. Fourteen studies reported data on the influence of anti-TNF alpha drugs, 13 on

the influence of corticosteroids, and 8 on the influence of thiopurines on postoperative

outcome. Additionally, 6 studies reported the outcome after the use of a combination of

immunosuppressive agents. One study (Page 2002), reported outcomes for patients <60

years and patients ≥60 years seperately.16 Data of both age groups were included (Page

2002a for patients <60 and Page 2002b for patients ≥60 years). Table 1 describes the most

important datapoints of the included studies.

Table 1 also presents the main conclusions of the individual studies. Individually, most

studies failed to find any significant association between preoperative immunosuppressive

agents and postoperative complications. Only 2 (14%) out of 14, 3 (31%) out of 14 and

152 Chapter 8

1 (13%) out of 8 studies found an association between postoperative complications and

preoperative anti-TNF agents, corticosteroid and thiopurines, respectively.

anti-Tumor necrosis factor (Tnf) agentsFourteen studies reported on the effect of preoperative anti-TNF agents on postoperative

outcomes, with a total of 5634 patients. Table 2 shows characteristics and results of these

studies. Most studies used a period of 3 months preoperatively as cut-off for including

patients into the anti-TNF group. Only 3 studies were matched (low risk of confounding),

while 4 others were not matched but had equivalent baseline characteristics (medium risk

of confounding). Meta-analysis showed no difference in total postoperative complication

rates between anti-TNF and no-TNF groups (Relative risk (RR) 1.16, 95% CI 0.97 to 1.40).

When only low and medium risk of confounding studies were included a trend towards

higher postoperative complications in the anti-TNF group was observed, but it did not

reach statistical significance (RR 1.35, 95% CI 1.00 to 1.83). Sensitivity analysis excluding

studies with other disease than Crohn’s disease also showed no difference between groups

Papers excluded by detailed review n=74 - Abstract only n=15 - Reviews n=14 - Letters & comments n=11 - Relevant outcomes not reported n=7 - Not focused on immunosuppressives n=7 - Duplicates n=6 - Not about Crohn’s Disease n=4 - Postoperative immunosuppressives n=3 - Published before 1990 n=2 - Focus on recurrence of disease n=1 - Case report n=1 - Other languages (Korean) n=1 - Not a surgical cohort n=1 - Animal study n=1

General Search Syntax: (immunosuppression OR immunosuppressive OR azathioprine OR Imuran OR cyclosporine OR 6-MP OR 6-mercaptopurine OR Puri-Nethol OR infliximab OR Remicade OR Humira OR adalimumab OR (anti AND TNF AND alpha) OR corticosteroids) AND (crohn's disease OR inflammatory bowel disease OR IBD) AND (Preoperative or “pre-operative” or postoperative or “post-operative” or perioperative or “peri-operative” or surgical complications or surgical outcome or surgical outcomes or postoperative complications)

Total (n=2787)

n=1077

n=36

Papers screened after search (n=2504)

Potentially relevant studies (n=95)

Papers excluded n=283 - Filtering double papers n=283 Papers excluded n=2409 - Clearly not relevant based on title and abstract n=2409

Studies included in review (n=21, 20 retrospective and 1

prospective)

n=1674 Medline (via PubMed)

EMBASE

The Cochrane Library

Figure 1. Flow chart of the selection process

Chapter 8 153

8

Tab

le 1

. Ch

arac

teri

stic

s an

d r

epo

rted

au

tho

r’s

con

clu

sio

ns

of

incl

ud

ed s

tud

ies

Stu

dy

IDC

ou

ntr

yPe

rio

dN

Dia

gn

ose

sSu

rgic

alin

terv

enti

on

Typ

eD

esig

n &

Dat

a co

llect

ion

Stu

die

d d

rug

sD

oes

Dru

g in

crea

se

com

plic

atio

ns?

Ab

erra

200

321U

SA19

92-2

000

159

CD

(55%

), U

CRs

.C

ohor

tR,

RC

S, T

hio

CS:

Yes

. Thi

o: N

o.

Ap

pau

200

89U

SA19

98-2

007

404

CD

Rs.

Coh

ort

R, R

a-TN

FIF

X: Y

es.

Baf

ford

201

212U

SA19

99-2

010

196

CD

Rs.,

Stom

.C

ohor

tR,

Pa-

TNF,

CS,

Thi

o, M

ixN

o.

Bru

ewer

200

313G

erm

any

1982

-200

039

7C

DRs

.C

ohor

tR,

RC

SN

o.

Can

edo

201

014U

SA20

00-2

008

225

CD

Rs.

Coh

ort

R, P

a-TN

F, T

hio

No.

Co

lom

bel

200

420U

SA19

98-2

001

270

CD

Rs.,

Stric

.C

ohor

tR,

Ra-

TNF,

CS,

Thi

oN

o.

El-H

uss

un

a 20

1223

Den

mar

k20

00-2

007

417

CD

Rs.

Coh

ort

R, R

a-TN

F, C

S, T

hio

CS:

Yes

. a-T

NF

& T

hio:

No.

Ind

ar 2

00924

USA

1999

-200

711

2C

DRs

., St

ric.,

Stom

.C

ohor

tR,

Ra-

TNF,

CS,

Any

No.

Kas

par

ek 2

01225

Ger

man

y20

01-2

008

168

CD

Rs.

Mat

ched

R, P

a-TN

FN

o.

Ku

nit

ake

2008

15U

SA19

93-2

007

413

CD

(45%

), U

C, I

CRs

., St

ric.

Coh

ort

R, R

a-TN

FN

o.

Mar

chal

200

426Be

lgiu

m19

98-2

002

79C

DRs

.M

atch

edR,

Ra-

TNF

No.

Mas

care

nh

as 2

01227

USA

2003

-201

093

CD

Rs.

Coh

ort

R, R

Any

No.

Myr

elid

200

928Sw

eden

1989

-200

234

3C

DRs

., St

ric.

Coh

ort

R, P

CS,

Thi

oC

S: N

o. T

hio:

Yes

.

Nas

ir 2

01029

USA

2005

-200

937

0C

DRs

.C

ohor

tR,

Pa-

TNF,

CS,

Thi

o, M

ixN

o.

No

rgar

d 2

01230

Den

mar

k20

00-2

010

2293

CD

Rs.

Coh

ort

R, R

a-TN

FN

o.

Pag

e 20

02a16

USA

1989

-199

910

5C

D (6

7%),

UC

Rs.

Coh

ort

R, R

CS,

Thi

oN

o.

Pag

e 20

02b

16U

SA19

89-1

999

105

CD

(67%

), U

CRs

.C

ohor

tR,

RC

S, T

hio

No.

Post

199

131G

erm

any

1981

-198

942

9C

DRs

., St

ric.,

Stom

.C

ohor

tR,

PC

SC

S: Y

es.

Riz

zo 2

01132

Italy

2004

-201

011

4C

D (7

6), U

CRs

.C

ohor

tR,

Ra-

TNF,

CS,

Thi

oC

S: Y

es. a

-TN

F: N

o.

Sam

pie

tro

200

933Ita

lly19

93-2

007

393

CD

Rs.

Coh

ort

P, P

CS,

Thi

o, M

ixN

o.

Tay

2003

17U

SA19

98-2

002

100

CD

Rs.,

Stric

.C

ohor

tR,

Ra-

TNF,

CS,

Mix

CS:

No.

Any

: Les

s co

mpl

icat

ions

.

Wat

erm

an 2

01334

Can

ada

2000

-201

047

3C

D (6

0%),

UC

, IC

Rs.

Mat

ched

R, P

a-TN

FN

o.

a-TN

F: a

nti-t

umor

nec

rosi

s fa

ctor

. C

D:

Cro

hn’s

dise

ase.

CS:

cor

ticos

tero

ids.

P:

pros

pect

ive.

R:

retr

ospe

ctiv

e. R

s.:

bow

el R

esec

tions

. St

ric.:

Stric

turo

plas

ty.

Stom

.: St

oma

plac

e-m

ent/

clos

ure.

Thi

o: T

hiop

urin

es. U

C: U

lcer

ativ

e co

litis

.

154 Chapter 8

Tab

le 2

. Po

sto

per

ativ

e o

utc

om

e in

pat

ien

ts w

ith

an

d w

ith

ou

t p

reo

per

ativ

e an

ti-t

um

or

nec

rosi

s fa

cto

r ag

ents

Stu

dy

IDD

rug

win

do

w

Ris

k o

f co

nfo

un

din

g

bia

s in

MA

a

N C

ases

vs. C

on

tro

ls

Tota

l co

m-

plic

atio

ns

(%)

Tota

l in

fec-

tio

us

com

-p

licat

ion

s(%

)

An

asto

mo

tic

leak

(%)

Intr

a-ab

do

min

al

Ab

sces

s(%

)

Wo

un

din

fect

ion

(%)

Sep

tic

sho

ck(%

)R

eop

erat

ion

(%)

Mo

rtal

ity

(%)

Kas

par

ek 2

01225

<3m

oLo

w (M

)48

vs.

48

56 v

s. 4

235

vs.

38

4 vs

. 13

6 vs

. 10

19 v

s. 1

5-

23 v

s. 2

1-

Mar

chal

200

426A

nyLo

w (M

)40

vs.

39

48 v

s. 3

835

vs.

18

0 vs

. 55

vs. 3

--

--

Wat

erm

an

2013

34<

6mo

Low

(M)

195

vs. 2

78-

20 v

s. 1

33

vs. 5

4 vs

. 510

vs.

42

vs. 1

4 vs

. 81

vs 0

.03

Ap

pau

200

89<

3mo

Med

ium

60 v

s. 3

29-

42 v

s. 1

910

vs.

410

vs.

4-

20 v

s. 1

08

vs. 3

2 vs

. 0

El-H

uss

un

a 20

1223

<3m

oM

ediu

m32

vs.

385

-9

vs. 1

3-

--

--

-

Riz

zo 2

01132

<3m

oM

ediu

m54

vs.

60

26 v

s. 1

717

vs.

13

7 vs

. 5-

--

-

Tay

2003

17<

8wks

Med

ium

22 v

s. 7

8-

14 v

s. 1

0-

--

--

-

Baf

ford

201

212<

3mo

Hig

h (B

NE)

35 v

s. 1

61-

20 v

s. 2

4-

--

--

-

Can

edo

201

014<

3mo

Hig

h (B

NE)

65 v

s. 7

5-

25 v

s. 1

73

vs. 1

3 vs

. 314

vs.

11

-3

vs. 3

0 vs

. 0

Co

lom

bel

20

0422

<2m

oH

igh

(BN

R)52

vs.

218

23 v

s. 2

317

vs.

20

--

--

--

Ind

ar 2

00924

<1m

oH

igh

(BN

R)17

vs.

95

35 v

s. 2

8-

--

--

--

Ku

nit

ake

2008

15<

3mo

Hig

h (B

NE)

101

vs. 3

1217

vs.

16

6 vs

. 10

3 vs

. 3-

--

-2

vs. 3

Nas

ir 2

01029

<2m

oH

igh

(BN

E)11

9 vs

. 251

30 v

s. 2

83

vs. 2

b-

--

--

-

No

rgar

d 2

01230

<3m

oH

igh

(BN

E)21

4 vs

. 207

9-

-4

vs. 3

--

-8

vs. 9

0.5

vs. 3

Out

com

e da

ta p

rese

nted

as

perc

enta

ges

of c

ases

(%) v

s. p

erce

ntag

es o

f co

ntro

ls (%

). M

A: m

eta-

anal

ysis

.a L

ow(M

): lo

w ri

sk b

ecau

se o

f app

ropr

iate

mat

chin

g. M

ediu

m: n

ot m

atch

ed, b

ut b

asel

ine

equi

vale

nt b

etw

een

grou

ps. H

igh(

BNE)

: Hig

h ris

k, b

asel

ine

not e

qual

bet

wee

n gr

oups

. H

igh(

BNR)

: hig

h ris

k, b

asel

ine

not

repo

rted

.b R

epor

ted

infe

ctio

us c

ompl

icat

ions

lim

ited

to m

ajor

abd

omin

al c

ompl

icat

ions

(abs

cess

es, l

eaks

).

Chapter 8 155

8

(RR 1.15, 95% 0.93 to 1.41). Infectious complications were significantly higher in the anti-

TNF groups (RR 1.29, 95% CI 1.07 to 1.55) (Figure 2). When studies were categorized ac-

cording to risk of confounding, the difference was significant in both the low and medium

risk of confounding groups (RR for low and medium risk of confounding 1.52, 95% CI

1.22 to 1.89). Results were consistent with sensitivity analysis excluding studies with other

patients than CD (RR 1.31, 96% 1.06 to 1.64). The rates of wound infection (RR 1.62,

95% CI 1.12 to 2.34) and septic shock (RR 1.81, 95% CI 1.03 to 3.17) were higher in the

a-TNF group compared to control (reported by three and two studies, respectively). Results

were consistent in the subgroup of only low and medium risk of confounding. In sensitivity

Study or Subgroup1.2.1 Low risk of confoundingKasparek 2012Marchal 2004Waterman 2013Subtotal (95% CI)Total eventsHeterogeneity: Chi² = 3.18, df = 2 (P = 0.20); I² = 37%Test for overall effect: Z = 2.28 (P = 0.02)

1.2.2 Medium risk of confoundingAppau 2008El-Hussuna 2012Rizzo 2011Tay 2003Subtotal (95% CI)Total eventsHeterogeneity: Chi² = 5.11, df = 3 (P = 0.16); I² = 41%Test for overall effect: Z = 3.05 (P = 0.002)

1.2.3 High risk of confoundingBaford 2012Canedo 2010Colombel 2004Kunitake 2008Nasir 2010Subtotal (95% CI)Total eventsHeterogeneity: Chi² = 3.70, df = 4 (P = 0.45); I² = 0%Test for overall effect: Z = 0.33 (P = 0.74)

Total (95% CI)Total eventsHeterogeneity: Chi² = 19.12, df = 11 (P = 0.06); I² = 42%Test for overall effect: Z = 2.68 (P = 0.007)

Events

171439

70

25393

40

716

964

42

152

Total

4840

195283

60325422

168

356552

101119372

823

Events

187

35

60

6149

88

126

38134332

5

131

317

Total

4839

278365

3293856078

852

16175

218312251

1017

2234

Weight

11.8%4.6%

18.9%35.4%

12.3%4.9%5.0%2.3%

24.6%

8.9%7.9%

10.9%10.3%

2.1%40.1%

100.0%

M-H, Fixed, 95% CI

0.94 [0.56, 1.60]1.95 [0.88, 4.31]1.59 [1.05, 2.41]1.42 [1.05, 1.92]

2.25 [1.54, 3.27]0.74 [0.24, 2.23]1.25 [0.52, 3.01]1.33 [0.38, 4.59]1.66 [1.20, 2.29]

0.85 [0.41, 1.74]1.42 [0.74, 2.73]0.88 [0.46, 1.68]0.58 [0.25, 1.35]1.69 [0.46, 6.17]0.94 [0.67, 1.32]

1.29 [1.07, 1.55]

Anti-TNF agents Control Risk Ratio Risk RatioM-H, Fixed, 95% CI

0.1 0.2 0.5 1 2 5 10Favours anti-TNF agents Favours control

Figure 2. Meta-analysis of postoperative infectious complication with and without preoperative anti-tumor necrosis factor (TNF) agents

156 Chapter 8

Tab

le 3

. Po

sto

per

ativ

e o

utc

om

e in

pat

ien

ts w

ith

an

d w

ith

ou

t p

reo

per

ativ

e st

ero

id u

se

Stu

dy

IDD

rug

w

ind

ow

Ris

k o

f co

nfo

un

din

g

bia

s in

MA

b

N C

ases

vs.

C

on

tro

ls

Tota

l co

m-

plic

atio

ns

(%)

Tota

l in

fec-

tio

us

com

-p

licat

ion

s(%

)

An

asto

mo

tic

leak

(%)

Intr

a-ab

do

min

al

Ab

sces

s(%

)

Wo

un

d

infe

ctio

n(%

)

Sep

tic

sho

ck(%

)R

eop

erat

ion

(%)

Mo

rtal

ity

(%)

Bru

ewer

20

0313

a<

1mo

Med

ium

219

vs. 1

77-

11 v

s. 8

3 vs

. 25

vs. 3

--

3 vs

. 32

vs. 2

El-H

uss

un

a 20

1223

<1w

kM

ediu

m66

vs.

351

-20

vs.

11

--

--

--

Riz

zo 2

01132

NR

Med

ium

48 v

s. 6

625

vs.

18

17 v

s. 1

4-

--

--

-

Tay

2003

17N

RM

ediu

m50

vs.

50

-12

vs.

10

--

--

--

Ab

erra

200

321N

RH

igh

(BN

E)56

vs.

51

-29

vs.

65

vs. 0

5 vs

. 05

vs. 0

5 vs

. 4-

-

Baf

ford

201

212<

6wks

Hig

h (B

NE)

72 v

s. 1

24-

29 v

s. 1

9-

--

--

-

Co

lom

bel

20

0422

aN

RH

igh

(BN

R)77

vs.

193

27 v

s. 2

223

vs.

18

--

--

-

Ind

ar 2

00924

<1m

oH

igh

(BN

R)47

vs.

65

28 v

s. 3

1-

--

--

--

Myr

elid

200

928<

2wk

Hig

h (B

NR)

87 v

s. 2

569

vs. 7

--

--

--

Nas

ir 2

01029

NR

Hig

h (B

NE)

151

vs. 2

1928

vs.

29

1 vs

. 3-

--

--

Pag

e 20

02a16

NR

Hig

h (B

NR)

62 v

s. 1

321

vs.

15

--

--

-

Pag

e 20

02b

16N

RH

igh

(BN

R)26

vs.

442

vs.

75

--

--

--

-

Post

199

131N

RH

igh

(BN

R)26

5 vs

. 89

12 v

s. 7

6 vs

. 0-

--

--

-

Sam

pie

tro

20

0933

NR

Hig

h (B

NR)

165

vs. 2

733

vs. 7

--

--

--

-

Out

com

e da

ta p

rese

nted

as

perc

enta

ges

of c

ases

(%) v

s. p

erce

ntag

es o

f co

ntro

ls (%

). N

R: n

ot r

epor

ted.

MA

: met

a-an

alys

is.

a Th

ese

two

stud

ies

also

rep

orte

d a

sub-

grou

p an

alys

is b

etw

een

high

dos

e of

ste

roid

s vs

. no

ster

oids

.b

Low

(M):

low

risk

bec

ause

of a

ppro

pria

te m

atch

ing.

Med

ium

: not

mat

ched

, but

bas

elin

e eq

uiva

lent

bet

wee

n gr

oups

. Hig

h(BN

E): H

igh

risk,

bas

elin

e no

t equ

al b

etw

een

grou

ps.

Hig

h(BN

R): h

igh

risk,

bas

elin

e no

t re

port

ed.

Chapter 8 157

8

analysis excluding studies with diseases other than CD, the risk of wound infection was no

longer statistically significant (RR 1.29, 95% CI 0.68 to 2.44), but septic shock (reported

only by 1 study) remained statistically significant (Table 2). Other individual complications

showed no differences.

corticosteroidsWe identified 13 studies that reported on the effect of preoperative corticosteroids on

postoperative outcomes, with a total of 3510 patients (Table 3). Unfortunately, most

studies only reported data on total and infectious complications, but not on individual

complications. Studies often did not report the cut-off point in terms of the dose and

preoperative time period over which steroids were used. None of the studies were

matched, but 4 studies had equivalent baseline characteristics. Two studies reported a

subgroup analysis comparing patients on high-dose of preoperative corticosteroids vs. no

steroids.13,22 Breuwer 2003 reported on a subgroup of 73 patients receiving 20 mg or more

/ day of corticosteroids.13 Colombel 2004 reported on a subgroup of 43 patients with 40

mg or more /day.22 Although data of these subgroups are not presented in the table, we

performed an additional meta-analysis (i.e. sensitivity analysis) in which we substituted

the overall data of these studies with the data from these subgroups only. This allowed

us to evaluate whether an increased dose of steroids would affect our results (testing

consistency of results).

Meta-analysis showed no difference in total postoperative complication rates for patients

on corticosteroid vs. those not on steroids (RR 1.03, 95% CI 0.83 to 1.27). This was

consistent in our subgroup analysis of only low and medium risk of confounding and

sensitivity analysis excluding studies with other diseases than CD. The difference became

more pronounced when only patients on high-dose corticosteroids were included for the

two studies above, but did not reach significance (RR 1.12, 95% CI 0.74 to 1.68, I²=67%,

Random-effects model (REM)). The risk of infectious complications was significantly higher

in the corticosteroid group (RR 1.55, RR 95% CI 1.23 to 1.95) (Figure 3). These results were

consistent in the subgroup analysis of only low and medium risk of confounding studies

(RR 1.45, 95% CI 1.01 to 2.08), and in the sensitivity analysis excluding studies with

diseases other than CD (RR 1.45, 95% CI 1.13 to 1.87). The difference was greater when

only patients on high-dose corticosteroids were included (RR 1.67, 95% CI 1.31 to 2.13).

Data were insufficient to show differences in individual complications.

ThiopurinesEight studies reporting on the impact of preoperative thiopurines on postoperative out-

comes were included, with a total of 1934 patients (Table 4). Most studies only reported

on total and infectious complications, without reporting on individual complications. The

158 Chapter 8

timeframe for preoperative thiopurine use in these studies varied from 2 weeks to 3 months.

None of the studies were matched. None of the pooled outcomes showed a significant

difference between the two groups. The pooled RR estimates for total complications and

infectious postoperative complications were 0.97 (95% CI 0.69 to 1.36) and 1.23 (95% CI

0.66 to 2.29, I²=81%, REM), respectively.

Mixed immunosuppressive agentsSix studies compared postoperative outcomes of patients who were on a combination of

immunosuppressive agents preoperatively to those not taking these medications. They

included a total of 1264 patients (Table 5). Three studies included patients on any immu-

nosuppressive agents, including steroids and anti-TNF agents. One study (Nasir 2010) did

not include patients who were on anti-TNF agents and corticosteroids (i.e. only included

patients on other immunosuppressants).23 Two studies (Sampietero 2009 and Tay 2003)

Study or Subgroup4.3.1 Medium risk of confoundingBruewer 2003El-Hussuna 2012Rizzo 2011Tay 2003Subtotal (95% CI)Total eventsHeterogeneity: Chi² = 0.75, df = 3 (P = 0.86); I² = 0%Test for overall effect: Z = 2.01 (P = 0.04)

4.3.3 High risk of confoundingAberra 2003Baford 2012Colombel 2004Myrelid 2009Nasir 2010Post 1991Subtotal (95% CI)Total eventsHeterogeneity: Chi² = 9.07, df = 5 (P = 0.11); I² = 45%Test for overall effect: Z = 3.10 (P = 0.002)

Total (95% CI)Total eventsHeterogeneity: Chi² = 9.66, df = 9 (P = 0.38); I² = 7%Test for overall effect: Z = 3.67 (P = 0.0002)

Events

2413

86

51

162118

82

16

81

132

Total

219664850

383

56727787

151265708

1091

Events

1439

95

67

3243418

70

86

153

Total

177351

6650

644

51124193256219

89932

1576

Weight

16.1%12.8%

7.9%5.2%

42.0%

3.3%18.3%20.2%

9.5%5.9%0.8%

58.0%

100.0%

M-H, Fixed, 95% CI

1.39 [0.74, 2.60]1.77 [1.00, 3.13]1.22 [0.51, 2.94]1.20 [0.39, 3.68]1.45 [1.01, 2.08]

4.86 [1.50, 15.70]1.51 [0.91, 2.51]1.33 [0.80, 2.20]1.31 [0.59, 2.90]0.41 [0.09, 1.97]

11.17 [0.68, 184.22]1.62 [1.19, 2.19]

1.55 [1.23, 1.95]

Steroids Control Risk Ratio Risk RatioM-H, Fixed, 95% CI

0.05 0.2 1 5 20Favours steroids Favours control

Figure 3. Meta-analysis of postoperative infectious complication with and without preoperative cor-ticosteroid

Chapter 8 159

8

Tab

le 4

. Po

sto

per

ativ

e o

utc

om

e in

pat

ien

ts w

ith

an

d w

ith

ou

t p

reo

per

ativ

e th

iop

uri

nes

Stu

dy

IDD

rug

w

ind

ow

Ris

k o

f co

nfo

un

din

g

bia

s in

MA

a

N C

ases

vs.

C

on

tro

ls

Tota

l co

m-

plic

atio

ns

(%)

Tota

l in

fec-

tio

us

com

-p

licat

ion

s(%

)

An

asto

mo

tic

leak

(%)

Intr

a-ab

do

min

al

Ab

sces

s(%

)

Wo

un

d

infe

ctio

n(%

)

Sep

tic

sho

ck(%

)R

eop

erat

ion

(%)

Mo

rtal

ity

(%)

El-H

uss

un

a 20

1223

<1m

oM

ediu

m16

6 vs

. 251

-10

vs.

14

--

--

--

Riz

zo 2

01132

NR

Med

ium

34 v

s. 8

015

vs.

24

9 vs

. 18

--

--

--

Ab

erra

200

321<

2wk

Hig

h (B

NE)

18 v

s. 5

1-

72 v

s. 6

6 vs

. 00

vs. 0

17 v

s. 0

6 vs

. 4-

-

Baf

ford

201

212<

3mo

Hig

h (B

NE)

69 v

s. 1

27-

13 v

s. 2

8-

--

--

-

Can

edo

201

014<

1mo

Hig

h (B

NE)

85 v

s. 7

5-

33 v

s. 1

75

vs. 1

7 vs

. 37

vs. 1

1-

8 vs

. 30

vs. 0

Co

lom

bel

20

0422

<1m

oH

igh

(BN

R)10

5 vs

. 165

25 v

. 22

19 v

s. 1

9-

--

--

-

Myr

elid

200

928<

6wk

Hig

h (B

NR)

51 v

s. 2

9210

vs.

6-

--

--

-

Pag

e 20

02a16

NR

Hig

h (B

NR)

20 v

s. 5

515

vs.

22

--

--

--

-

Pag

e 20

02b

16N

RH

igh

(BN

R)13

vs.

17

54 v

s. 4

1-

--

--

--

Out

com

e da

ta p

rese

nted

as

perc

enta

ges

of c

ases

(%) v

s. p

erce

ntag

es o

f co

ntro

ls (%

). N

R: n

ot r

epor

ted.

MA

: met

a-an

alys

is.

a M

ediu

m: n

ot m

atch

ed, b

ut b

asel

ine

equi

vale

nt b

etw

een

grou

ps. H

igh(

BNE)

: Hig

h ris

k, b

asel

ine

not

equa

l bet

wee

n gr

oups

. Hig

h(BN

R): h

igh

risk,

bas

elin

e no

t re

port

ed.

160 Chapter 8

Tab

le 5

. Po

sto

per

ativ

e o

utc

om

e in

pat

ien

ts w

ith

an

d w

ith

ou

t p

reo

per

ativ

e im

mu

no

sup

pre

ssiv

e d

rug

s in

co

mb

inat

ion

Stu

dy

IDD

rug

w

ind

ow

Ris

k o

f co

nfo

un

din

g

bia

s in

MA

a

N C

ases

vs.

C

on

tro

ls

Tota

l co

mp

licat

ion

s(%

)

Tota

l in

fect

iou

s co

mp

licat

ion

s(%

)

An

asto

mo

tic

leak

(%)

Intr

a-ab

do

min

al

Ab

sces

s(%

)

Wo

un

d

infe

ctio

n(%

)M

ort

alit

y(%

)

Baf

ford

201

212<

3mo

Hig

h (B

NE)

127

vs. 6

9-

24 v

s. 2

0-

--

-

Ind

ar 2

00924

<1m

oH

igh

(BN

R)69

vs.

43

32 v

s. 2

619

vs.

73

vs. 2

1 vs

03

vs 0

0 vs

. 0

Mas

care

nh

as 2

01227

<3m

oM

ediu

m19

vs.

74

21 v

s. 8

--

--

0 vs

. 0

Nas

ir 2

01029

bN

RH

igh

(BN

E)11

5 vs

. 255

25 v

s. 3

02

vs. 3

--

--

Sam

pie

tro

200

933 c

NR

Hig

h (B

NR)

60 v

s. 3

338

vs. 1

7-

--

--

Tay

2003

17 c

<2m

oM

ediu

m72

vs.

28

-6

vs 2

51

vs 1

13

vs 1

4-

-

Out

com

e da

ta p

rese

nted

as

perc

enta

ges

of c

ases

(%) v

s. p

erce

ntag

es o

f co

ntro

ls (%

). N

R: n

ot r

epor

ted.

MA

: met

a-an

alys

is.

a Med

ium

: not

mat

ched

, but

bas

elin

e eq

uiva

lent

bet

wee

n gr

oups

. Hig

h(BN

E): H

igh

risk,

bas

elin

e no

t eq

ual b

etw

een

grou

ps. H

igh(

BNR)

: hig

h ris

k, b

asel

ine

not

repo

rted

.b In

clud

ed a

ll im

mun

osup

pres

sive

age

nts

othe

r th

an a

nti-t

umur

nec

rosi

s fa

ctor

(TN

F) a

gent

s an

d st

eroi

dc In

clud

ed d

rugs

cla

ssifi

ed a

s im

mun

olog

ical

and

bio

logi

cal d

rugs

(exc

lude

d st

eroi

ds)

Chapter 8 161

8

only included drugs that they classified as immunologic and/or biologic drugs (effectively

excluding corticosteroids).17,33 None of the studies was matched, and only few reported

data on individual complications. None of the pooled outcomes showed a significant dif-

ference. The pooled RR estimates for total and infectious postoperative complications were

0.89 (95% CI 0.67 to 1.18) and 1.01 (95% CI 0.67 to 1.54), respectively.

discussion

This study provides a comprehensive review of the impact of immunosuppressive agents on

postoperative outcomes in CD, and identified several important findings. Firstly, patients

on preoperative anti-TNF agents or corticosteroids were at higher risk for developing post-

operative infectious complications. The use of anti-TNF agents was also significantly associ-

ated with individual infectious complications (i.e. wound infection and septic shock). For

corticosteroids, the available data also suggest a dose-related response to its deleterious

effect on postoperative complications. This review also provides a comprehensive review

on the state of current evidence on this topic, and identified several shortcomings of the

current studies relating to the study question.

Anti-TNF medication has become one of the most important and widely used medica-

tions in the treatment of CD.3,35 Its use in the preoperative period, however, has triggered

concerns among surgeons and gastroenterologists alike about a potential increased risk of

associated postoperative (infectious) complications.36 In earlier reviews that did not include

a meta-analysis, the conclusions tended to suggest the lack of any such association.36,37

One reason for this could be that the small sample size and methodological limitations of

these studies precluded their ability to detect potential differences in outcomes. As shown

in this review, only 2 (14%) out of 14 studies found an association between postoperative

complications and preoperative anti-TNF agents (Table 1). However, when meta-analysis

was performed, a clear increase in overall infectious complications as well as in several in-

dividual infectious complications was observed. As a significant proportion of patients with

CD undergo at least one surgical procedure in the course of the disease,38,39 these findings

are likely to be of great importance in clinical practice. Thus, when patients on anti-TNF

agents require surgery, a preoperative drug free interval, when feasible, is advisable and

might reduce the risk of septic complications. The adoption of specific operative strategies

that may reduce these risks or their effects, such as the use of a temporary ostomy, may

additionally be considered. The findings also support patient education and counseling of

potential adverse outcomes of these agents when patients undergo surgery.

162 Chapter 8

A recent review has similarly reported a significant increase in infectious complications

after anti-TNF treatment in CD patients.10 However, that review had several limitations and

based its results on only 6 out of the 12 studies that were included in our meta-analysis.

Four of these studies were published after their search, while two that were published

prior to the search were omitted. Most importantly, the review did not stratify studies

based on the risk of confounding. Given the heterogeneity and the high risk of confound-

ing in several studies, the findings of that review were inconclusive. The inclusion of a

greater number of studies and the consideration of the confounding risk allowed us to

identify potentially important differences in individual infectious complications, such as

wound infection. The association between immunosuppressive agents and delayed wound

healing and infection has been the topic of several in vitro and in-vivo studies, which

have suggested a strong pathophysiological basis for this phenomenon.8 In addition, septic

shock was found to be significantly associated with anti-TNF treatment in meta-analysis.

Although only two studies reported on this outcome, the severity of this complication war-

rants serious consideration of these results, and future research should focus on providing

more data on this outcome.

We also found an increased risk of postoperative infectious complications for CD patients

on corticosteroid. These deleterious effects of steroids on postoperative infectious compli-

cations are not specific to CD since previous studies have described such an association

in both general medical patients and patients with inflammatory bowel disease in gen-

eral.11,40 A further interesting finding of our study was that subgroup analyses suggested

a dose-effect relationship between preoperative corticosteroids and rates of infectious

complications. This suggests that lowering the dose of corticosteroids in anticipation of

surgery might decrease risk of complications in patients, when discontinuing the drug

is not possible. Finally, our findings suggest that based on current evidence, there is no

association between the use of thiopurines or other immunomodulatory drugs in combina-

tion on postoperative complications. This is in line with other reviews in the literature,7,37

although this review does include the largest number of studies and patients, and is cur-

rently the only review to combine current data on these medications into a meta-analysis.

Our review also highlights current limitations of the published data on this topic. Firstly,

most studies are underpowered, which does not allow the identification of potential im-

portant clinical differences. Similarly, most studies were non-randomized and unmatched,

thus leading to significant differences between baseline characteristics of the study groups.

Although some studies used regression to correct for baseline discrepancies, the inclusion

of many covariates in regression models could further decrease the power of these studies.

This might explain the failure of many of the studies to identify the differences that were

observed in this meta-analysis. Also, poor reporting of comparative data has proven to be a

Chapter 8 163

8

limitation to meta-analyses of secondary outcomes of interest; several important individual

complications, such as anastomotic leaks and pelvic and intra-abdominal abscesses that

might influence the decision for surgery. It may have also been interesting to study the

relation between the time of last drug administration and post-operative complications.

Most studies, however, did not aim to study this effect, and unfortunately the data were

therefore not suitable for this analysis.

Despite the continuing controversy on this clinically relevant topic, the quality of evidence

has essentially remained suboptimal over the last 20 years. It is likely that randomized studies

or large prospective cohort studies will remain a significant challenge due to the complexity

of the question at hand. The accumulation of data on all important confounding factors

and the use of a matched study design may allow for more appropriate interpretation of

results and allow meta-analyses. Performing meta-analysis with aggregate published data

is subject to certain limitations. For example, patients may have more than 1 complication,

and it is often difficult to determine the exact number of patients that had complica-

tions when different subgroups (e.g. mild and severe complications) are combined into

a single (total complications) group. In this regard, performing an individual patient data

meta-analysis confers certain advantages facilitating more accurate and powerful analyses,

adjustment for confounders and the opportunity for subgroup analyses. Performing an

individual patient data meta-analysis, however, is often limited by the resources, time and

cooperation required.41

The results of this review, which currently provides the best evidence on the topic, suggests

that patients with Crohn’s disease on immunosuppressive therapy in the peri-operative

period may have adverse postoperative outcomes.. We acknowledge that this review does

have some limitations. Firstly, five studies (24%) included patients with diagnoses other

than CD, e.g. ulcerative colitis and indeterminate colitis. We sought to overcome this by

including such studies if CD was the largest group of patients (in most cases >60% of

patients). In addition, the results of our sensitivity analysis excluding these five studies was

consistent with our main meta-analysis, adding to the robustness of our findings. Secondly,

there were differences between studies in regard to the definition of their endpoints.

Therefore, we focused on two major outcomes (i.e. total complications, and infectious

complications), and did not include other classifications that may have been introduced by

individual studies (e.g. major vs. minor, early vs. late). Under infectious complications we

included the outcomes reported by each study that were consistent with our definition.

Also, the retrospective nature of studies and the varying methods of analyses used alters

the robustness of the conclusions that can be derived. In this regard, the reliability of the

findings in this review are strengthened by the fact that we stratified results of studies

based on the risk of confounding, and showed consistent results in the low and medium

164 Chapter 8

risk of bias subgroup. In the subgroup of studies that had a high risk of confounding no

significant difference was observed for most comparisons. It seems, therefore, that studies

that correct well for potential confounding are more likely to find an association between

anti-TNF agents and postoperative complications. A similar pattern was recognized in an

earlier review of studies of anti-TNF agents and postoperative complications in patients

with ulcerative colitis.36 Authors of that review noticed that studies that specifically ad-

justed for disease severity were more likely to find a positive association between anti-TNF

drugs and postoperative complications.

In conclusion, this review suggests that CD patients on preoperative immunosuppressive

agents are at higher risk for complications. Both corticosteroids and anti-TNF agents may

increase risk of infections and septic shock. A preoperative drug free interval, when fea-

sible, is advisable and might reduce the risk of infections.

Chapter 8 165

8

references

1. Rasmussen SN, Binder V, Maier K, Bond-

esen S, Fischer C, Klotz U, Hansen SH, et al.

Treatment of crohn’s disease with peroral

5-aminosalicylic acid. Gastroenterology.

1983; 85: 1350-53.

2. Schultz MG, Rieder HL, Hersh T, Riepe S.

Remission of crohn’s disease with antimy-

cobacterial chemotherapy. Lancet. 1987; 2:

1391-92.

3. D’Haens GR, Panaccione R, Higgins PD,

Vermeire S, Gassull M, Chowers Y, Hanauer

SB, et al. The london position statement

of the world congress of gastroenterol-

ogy on biological therapy for ibd with the

european crohn’s and colitis organization:

When to start, when to stop, which drug to

choose, and how to predict response? Am J

Gastroenterol. 2011; 106: 199-212.

4. Ford AC, Sandborn WJ, Khan KJ, Hanauer

SB, Talley NJ, Moayyedi P. Efficacy of bio-

logical therapies in inflammatory bowel dis-


Am J Gastroenterol. 2011; 106: 644-59,

quiz.

5. D’Haens GR. Top-down therapy for ibd:

Rationale and requisite evidence. Nat Rev

Gastroenterol Hepatol. 2010; 7: 86-92.

6. Matsumoto T, Iida M, Motoya S, Haruma K,

Suzuki Y, Kobayashi K, Ito H, et al. Thera-

peutic efficacy of infliximab on patients

with short duration of crohn’s disease: A

japanese multicenter survey. Dis Colon

Rectum. 2008; 51: 916-23.

7. Afif W, Loftus EV, Jr. Safety profile of ibd

therapeutics: Infectious risks. Gastroenterol

Clin North Am. 2009; 38: 691-709.

8. Busti AJ, Hooper JS, Amaya CJ, Kazi S. Ef-

fects of perioperative antiinflammatory and

immunomodulating therapy on surgical

wound healing. Pharmacotherapy. 2005;

25: 1566-91.

9. Appau KA, Fazio VW, Shen B, Church JM,

Lashner B, Remzi F, Brzezinski A, et al. Use

of infliximab within 3 months of ileocolonic

resection is associated with adverse post-

operative outcomes in crohn’s patients. J

Gastrointest.Surg. 2008; 12: 1738-44.

10. Kopylov U, Ben-Horin S, Zmora O, Eliakim

R, Katz LH. Anti-tumor necrosis factor and

postoperative complications in crohn’s dis-



11. Subramanian V, Saxena S, Kang JY, Pollok

RC. Preoperative steroid use and risk of

postoperative complications in patients

with inflammatory bowel disease undergo-

ing abdominal surgery. Am J Gastroenterol.

2008; 103: 2373-81.

12. Bafford AC, Powers S, Ha C, Kruse D,

Gorfine SR, Chessin DB, Bauer JJ. Immu-

nosuppressive therapy does not increase

operative morbidity in patients with crohn’s

disease. J Clin Gastroenterol. 2012.

13. Bruewer M, Utech M, Rijcken EJ, Anthoni

C, Laukoetter MG, Kersting S, Senninger N,

et al. Preoperative steroid administration:

Effect on morbidity among patients under-

going intestinal bowel resection for crohns

disease. World J Surg. 2003; 27: 1306-10.

14. Canedo J, Lee SH, Pinto R, Murad-Regadas

S, Rosen L, Wexner SD. Surgical resection

in crohn’s disease: Is immunosuppressive

medication associated with higher post-

operative infection rates? Colorectal Dis.

2011; 13: 1294-98.

15. Kunitake H, Hodin R, Shellito PC, Sands

BE, Korzenik J, Bordeianou L. Perioperative

treatment with infliximab in patients with

crohn’s disease and ulcerative colitis is

not associated with an increased rate of

postoperative complications. J Gastrointest.

Surg. 2008; 12: 1730-36.

16. Page MJ, Poritz LS, Kunselman SJ, Koltun

WA. Factors affecting surgical risk in elderly

patients with inflammatory bowel disease. J

Gastrointest.Surg. 2002; 6: 606-13.

166 Chapter 8

17. Tay GS, Binion DG, Eastwood D, Otterson

MF. Multivariate analysis suggests improved

perioperative outcome in crohn’s disease

patients receiving immunomodulator

therapy after segmental resection and/or

strictureplasty. Surgery. 2003; 134: 565-72.

18. Deeks JJ. Issues in the selection of a sum-

mary statistic for meta-analysis of clinical

trials with binary outcomes. Stat.Med.

2002; 21: 1575-600.

19. Altman DG, Deeks JJ, Sackett DL. Odds

ratios should be avoided when events are

common. BMJ. 1998; 317: 1318.

20. Higgins JP, Thompson SG, Deeks JJ, Altman

DG. Measuring inconsistency in meta-

analyses. BMJ. 2003; 327: 557-60.

21. Aberra FN, Lewis JD, Hass D, Rombeau JL,

Osborne B, Lichtenstein GR. Corticosteroids

and immunomodulators: Postoperative

infectious complication risk in inflammatory

bowel disease patients. Gastroenterology.

2003; 125: 320-27.

22. Colombel JF, Loftus EV, Jr., Tremaine WJ,

Pemberton JH, Wolff BG, Young-Fadok

T, Harmsen WS, et al. Early postoperative

complications are not increased in patients

with crohn’s disease treated perioperatively

with infliximab or immunosuppressive ther-

apy. Am J Gastroenterol. 2004; 99: 878-83.

23. El-Hussuna A, Andersen J, Bisgaard T, Jess P,

Henriksen M, Oehlenschlager J, Thorlacius-

Ussing O, et al. Biologic treatment or

immunomodulation is not associated with

postoperative anastomotic complications

in abdominal surgery for crohn’s disease.

Scand J Gastroenterol. 2012; 47: 662-68.

24. Indar AA, Young-Fadok TM, Heppell J,

Efron JE. Effect of perioperative immuno-

suppressive medication on early outcome

in crohn’s disease patients. World J Surg.

2009; 33: 1049-52.

25. Kasparek MS, Bruckmeier A, Beigel F, Muller

MH, Brand S, Mansmann U, Jauch KW, et

al. Infliximab does not affect postopera-

tive complication rates in crohn’s patients

undergoing abdominal surgery. Inflamm.

Bowel Dis. 2012; 18: 1207-13.

26. Marchal L, D’Haens G, Van AG, Vermeire

S, Noman M, Ferrante M, Hiele M, et al.

The risk of post-operative complications as-

sociated with infliximab therapy for crohn’s

disease: A controlled cohort study. Aliment.

Pharmacol.Ther. 2004; 19: 749-54.

27. Mascarenhas C, Nunoo R, Asgeirsson T,

Rivera R, Kim D, Hoedema R, Dujovny N,

et al. Outcomes of ileocolic resection and

right hemicolectomies for crohn’s patients

in comparison with non-crohn’s patients

and the impact of perioperative immu-

nosuppressive therapy with biologics and

steroids on inpatient complications. Am J

Surg. 2012; 203: 375-78.

28. Myrelid P, Olaison G, Sjodahl R, Nystrom

PO, Almer S, Andersson P. Thiopurine

therapy is associated with postoperative

intra-abdominal septic complications in

abdominal surgery for crohn’s disease. Dis

Colon Rectum. 2009; 52: 1387-94.

29. Nasir BS, Dozois EJ, Cima RR, Pemberton

JH, Wolff BG, Sandborn WJ, Loftus EV, et

al. Perioperative anti-tumor necrosis fac-

tor therapy does not increase the rate of

early postoperative complications in crohn’s

disease. J Gastrointest.Surg. 2010; 14:

1859-65.

30. Norgard BM, Nielsen J, Qvist N, Gradel KO,

de Muckadell OB, Kjeldsen J. Pre-operative

use of anti-tnf-alpha agents and the risk

of post-operative complications in patients

with ulcerative colitis - a nationwide cohort

study. Aliment.Pharmacol.Ther. 2012; 35:

1301-09.

31. Post S, Betzler M, von DB, Schurmann G,

Kuppers P, Herfarth C. Risks of intestinal

anastomoses in crohn’s disease. Ann Surg.

1991; 213: 37-42.

32. Rizzo G, Armuzzi A, Pugliese D, Verbo A,

Papa A, Mattana C, Rapaccini GL, et al.

Anti-tnf-alpha therapies do not increase

early postoperative complications in

Chapter 8 167

8

patients with inflammatory bowel disease.

An italian single-center experience. Int J

Colorectal Dis. 2011; 26: 1435-44.

33. Sampietro GM, Corsi F, Maconi G, Ardiz-

zone S, Frontali A, Corona A, Porro GB, et

al. Prospective study of long-term results

and prognostic factors after conservative

surgery for small bowel crohn’s disease. Clin

Gastroenterol Hepatol. 2009; 7: 183-91.

34. Waterman M, Xu W, Dinani A, Steinhart

AH, Croitoru K, Nguyen GC, McLeod RS,

et al. Preoperative biological therapy and

short-term outcomes of abdominal surgery

in patients with inflammatory bowel dis-

ease. Gut. 2013; 62: 387-94.

35. Swaminath A, Lebwohl B, Capiak KM,

Present DH. Practice patterns in the use of

anti-tumor necrosis factor alpha agents in

the management of crohn’s disease: A us

national practice survey comparing experts

and non-experts. Dig Dis Sci. 2011; 56:

1160-64.

36. Holubar SD, Cima RR, Pemberton JH. Does

infliximab increase complications after

surgery for inflammatory bowel disease?

F1000.Med Rep. 2009; 1.

37. Subramanian V, Pollok RC, Kang JY, Kumar

D. Systematic review of postoperative com-

plications in patients with inflammatory

bowel disease treated with immunomodu-

lators. Br.J Surg. 2006; 93: 793-99.

38. Louis E, Collard A, Oger AF, Degroote E,

Aboul Nasr El Yafi FA, Belaiche J. Behaviour

of crohn’s disease according to the vienna

classification: Changing pattern over the

course of the disease. Gut. 2001; 49:

777-82.

39. Mekhjian HS, Switz DM, Watts HD,

Deren JJ, Katon RM, Beman FM. National

cooperative crohn’s disease study: Factors

determining recurrence of crohn’s disease

after surgery. Gastroenterology. 1979; 77:

907-13.

40. Stuck AE, Minder CE, Frey FJ. Risk of

infectious complications in patients taking

glucocorticosteroids. Rev Infect.Dis. 1989;

11: 954-63.

41. Lyman GH, Kuderer NM. The strengths

and limitations of meta-analyses based on

aggregate data. BMC Med Res Methodol.

2005; 5: 14.

S. de Zeeuw

U. Ahmed Ali

M.B.M. van der Kolk

C.J.H.M. van Laarhoven

Department of Surgery, Radboud University Medical Centre, Nijmegen

9 ileo-pouch anal anastomosis

with close-rectal dissection

usinG automated vessel

sealers for ulcerative colitis:

a promisinG alternative

Digestive Surgery, 2011.

170 Chapter 9

absTracT

backgroundDespite decennia of experience, the ileo pouch anal anastomosis for ulcerative colitis is

still associated with high complication rates. The development of automatic vessel sealers

has resulted in the revive of a promising surgical alternative to the conventional procedure:

the close rectal dissection. By preserving the mesorectal layer it is hypothesized that nerve-

related and other postoperative complications can be reduced.

MethodsAll patients with ulcerative colitis with indication for restorative proctocolectomy at our

institution during the pilot study underwent the close-rectal pouch procedure with tempo-

rary diverting ileostomy. Standardized clinical history, anorectal physiology measurements,

endoscopic and histological examination were carried out before and after surgery.

resultsThe procedure was technically successful in all 10 patients, with a median age of 41 years

and a median postoperative follow-up period of 16 months. There were no cases of pelvic

sepsis and bladder- or sexual dysfunction. The median daytime defecation frequency was

6.0. Endoscopic and histological examination showed no abnormalities. Anorectal physiol-

ogy supported the good functional results.

conclusionThe preliminary results of the close-rectal pouch procedure are promising, with good

functional results and a low complication rate after one year.

Chapter 9 171

9

inTroducTion

Surgery is an essential modality in the treatment of ulcerative colitis (UC). In UC the most

common surgical indication is severe refractory colitis. The side effects associated with

chronic medical therapy, in particular corticosteroids, are another reason for patients and

doctors to seek a more definitive solution. Eventually, up to 30% of all UC patients will

undergo surgery after 10 years of disease onset.1,2

The ability to restore the oral-anal continuity has been an important mile stone in the

surgical management of UC.3 Two surgical restorative options gained wide acceptance.

The ileorectal anastomosis (IRA) proved useful in certain patient subgroups, mainly in UC

patients with mild proctitis. The downside is the residual rectal mucosa, and the associated

risk of malignant degeneration. The second option, the restorative proctocolectomy by

means of an ileo pouch anal anastomosis (IPAA) progressed to be the gold standard for

restorative surgery for the majority of UC patients.3

Despite decennia of experience, the IPAA is still associated with high complication rates.

The adoption of the double-stapled approach for the anastomosis by most surgeons

have arguable resulted in an improvement in functional results.4 Studies regarding the

routine use of a defunctioning ileostomy did not show any reliable differences in terms

of complication rate5, especially pelvic sepsis. The use of an ileostomy therefore remains a

matter of surgeon’s preference and experience.6 Currently, major complications still pose

an important challenge for restorative surgery. Pelvic sepsis occurs in up to 12% of all

cases7-10, sexual dysfunction in 4%11, and the rate of pouch failure varies up to 8%.10,11

The close rectal or perimuscular dissection (CRD)12 has been first described in 1972. Despite

promising preliminary results, the technique was found to be laborious and few surgeons

adopted it. With the development of the electrothermal bipolar vessel sealer (EBVS) and

ultrasound dissection13 the technique has been substantially simplified. Several series have

since shown that this technique is safe and promising.14-18 The main distinction of the CRD

technique is in the way the dissection of the rectum is performed. In this technique the

rectum is dissected through the nonanatomic perimuscular plane in an effort to avoid dam-

age to the pelvic autonomic nerves, which lie immediately outside the mesorectal plane.

This is hypothesized to reduce nerve-related postoperative complications like sexual dys-

function. In addition, with the preservation of the mesorectal fat it is hypothesized that

there will be less room in the pelvis for postoperative hematoma or fluid collections. This

can reduce the risk of subsequent infection and abscess formation in the deep pelvis.

Finally, in case of a potential leakage of the anastomosis, the mesorectal fat can help

172 Chapter 9

contain such leakage, protecting the remainder of the pelvis and potentially avoiding pelvic

sepsis. The goal of this pilot study is to evaluate whether the CR-IPAA is a safe procedure

that could potentially reduce the complication rate, especially in terms of pelvic sepsis and

autonomic nerve damage.


ethics and patient selectionThe medical ethics committee of the Radboud University Nijmegen Medical Centre ap-

proved this study. Ten patients, all with UC, with an indication for surgical treatment by

means of an IPAA procedure during this pilot study, were suitable candidates for the study.

Written informed consent was obtained before inclusion of any patient.

surgical techniquePatients were admitted to the hospital one day before surgery to undergo preoperative

assessment and preparations. Antibiotic prophylaxis were administrated to all patients im-

mediately before surgery. The patient was placed in the lithotomy position. The procedure

was performed through a pfannenstiel or a low midline incision. (Laparoscopic) subtotal

colectomy with end ileostomy was in most cases performed as a staged procedure prior

to proctectomy and IPAA. This was the case for both elective and acute (e.g. toxic mega

colon) procedures. During the pouch formation, the rectum was dissected according to

the CRD technique.12 The dissection was conducted circulair to the mesorectum, in the

nonanatomic perimuscular plane, thereby preserving the mesorectum fat layer just outside

the rectum. The dissection was aided by electrothermal bipolar vessel sealer (Ligasure®,

Covidien, USA). Subsequently, a J-pouch was constructed from the distal ileum in a con-

ventional manner. The anastomosis was performed as a stapled pouch-anal anastomosis.

Because the mesorectal thickness is minimized to almost zero at the level of the pelvic floor,

the pouch-anal anastomosis is at the same level and identical to the anastomosis of the

conventional IPAA procedure. A diverting ileostomy was performed routinely in all patients.

Oro-anal continuity was restored approximately three months after clinical and radiological

assessment of a healed pouch and pouch-anal anastomosis.

follow-upPatients were seen in the outpatient clinic at 3 weeks and 3, 6 and 12 months after surgery.

At every visit a standardized history was taken, with emphasis on defecation frequency and

level of continence. Any complications were recorded. Pelvic sepsis was defined as a pelvic

abscess, anastomotic leakage or dehiscence or a pelvic or perineal wound infection.

Chapter 9 173

9

functional resultsDefecation frequency and questions about incontinence were scored. Physical examination

was performed, including inspection of the anus and digital rectal examination. There were

four question addressing sexual and bladder dysfunction (Table 1).

Table 1. Questions regarding sexual and bladder dysfunction

Bladder dysfunction Do you experience any problems with urination?

Sexual dysfunction for women Do you experience pain during sexual intercourse?

Sexual dysfunction for menDo you experience a problem with getting an erection?

Do you suffer from retrograde ejaculation?

anorectal physiologyEvaluation of anorectal function was performed preoperatively, and one year postopera-

tive. Anal sphincter pressure was measured by circular 4-channel low compliance water

fusion manometry (Mui scientific, Mississauga, ON, Canada). Volume capacity was mea-

sured by means of balloon distention. All measurements were performed in the left supine

position.19-21

endoscopy and pathologyEndoscopy was performed at three months and one year to evaluate the pouch mucosa,

inflammatory activity and to take multiple biopsies. The proctocolectomy specimens were

processed for histopathology to examine the dissection planes and confirm the diagnosis

colitis ulcerosa. Biopsies of the mucosa, taken at endoscopy, were processed for detailed

routine histopathological evaluation.

Mesorectal evaluationTo support the close-rectal dissection principle, we performed MRI scans pre- and post-

operatively. To our knowledge, there is no objective measurement for the volume of the

mesorectum. Mesorectum is mobile and adapts its shape to every change in rectum or

pouch volume. Therefore, we determined the presence of the mesorectum at three levels

on the MRI scans.

statistical analysisDescriptive statistics were used to summarize patient demographics, clinical data, surgical

details, morbidity and fecal incontinence. Continuous data were tested using the paired-

t-test and otherwise the Wilcoxon signed-rank test was used. Statistical significance was

considered when p<0.05. SPSS® version 16.0 (SPSS, Chicago, Illinois, USA) was used for

all statistical analysis.

174 Chapter 9

resulTs

All patients referred to our institution for an IPAA procedure during the study period con-

sented to participate in this study and were thus included. In total 10 consecutive patients

underwent an IPAA procedure with CRD dissection. The group consisted of six men and four

women, all diagnosed with UC. Six of ten suffered severe colitis (Mayo score of 3 points22),

the other patients suffered medical refractory mild to moderate colitis. Half of the patients

were operated on in an acute stage (e.g. toxic colitis). None of the subtotal colectomy

specimens showed evidence of dysplasia. The median age at surgery was 41 years (range

21 to 53). Surgical data and pre-operative medication specifications are listed in Table 2.

Table 2. Surgical data of the patient population (N=10)

Characteristic N (%)

Immunmodulating medication before colectomy 5 (50%)

Immunomodulating medication before pouch construction 0 (0%)

Referred after first-stage surgery in other hospital 5 (50%)

Stages:- Two-stage procedure- Three-stage procedure

9 (90%)1 (10%)

surgical parametersThe procedure was technically successful in all patients. The median surgical time was

247 minutes (range 133 to 332). The median blood loss was 425 ml (range 200 to 900).

Hospital stay was median 9 days (range 6 to 16 days). The median interval to closure of the

ileostomy was 4.5 months.

follow-upAll patients completed the direct postoperative follow-up. The median follow-up period

was 16 months (range 7 to 37).

Postoperative complicationsOne patient who developed a 5 cm presacral abscess one month postoperatively, which

was drained radiologically. This patient suffered minimal systemic symptoms and recovery

was uneventful. Five patients had minor complications. Two patients had a gastroparesis,

for which conservative treatment was sufficient. Three infectious complications occurred

in three patients: two urinary tract infections and one pulmonary infection, all could be

treated with antibiotics. None of the patients suffered bladder- or sexual dysfunction

preoperatively and postoperatively there were no cases of bladder dysfunction or sexual

dysfunction.

Chapter 9 175

9

functional resultsThe recorded daily defecation frequency was a median of 6.0 in all patients (range 3 to

8.5). The median nocturnal frequency after one year was 1.5 (range 0.5 to 3.5). Three

patients recorded occasionally nocturnal soiling. Passive and urge incontinence was not

reported.

anorectal physiologyAnorectal manometry was performed in all patients to objectively assess the anorectal

function. For the seven patient with at least one year follow-up the preoperative and

one year follow-up manometry results were compared. The maximal tolerated volume

increased from 68 ml (preoperative rectum measurements) to a median of 193 ml after

one year (postoperative pouch measurement) (p=0,03). First sense and first urge volumes

also increased significantly after one year (p=0.01 and p=0.02). All the results are shown

in Table 3.

Table 3. Pre- and postoperative anal sphincter manometry and reservoir function

Referencevalues

Medianpreoperative

values (range)

Medianpostoperative value

at 1 year (range) p-value

Anal sphincter

Maximal resting pressure (mm Hg) 43–89 60 (30–70) 60 (23–90) 0.63

Maximal squeeze pressure (mm Hg) 58–122 123 (38–225) 129 (38–360) 0.89

Reservoir function

First sense volume (ml) 35 30 (9–40) (rectum) 50 (15–200) (pouch) 0.01

First urge volume (ml) 80–120 43 (24–60) (rectum) 73 (60–200) (pouch) 0.02

Maximal tolerated volume (ml) 200 68 (27–80) (rectum) 193 (82–203) (pouch) 0.03

endoscopy and pathologyEndoscopy results after 1 year revealed signs of inflammation in the two patients. One

patient had symptoms of pouchitis at the moment of endoscopy. Microscopic examination

of the proctocolectomy specimens confirmed the diagnosis of UC in all patients. Histo-

pathology confirmed the diagnosis pouchitis on the basis of the biopsies taken in two

patients. No dysplasia or carcinoma was present in any specimen or at follow-up.

Mesorectal evaluationAll postoperative MRI scans showed the presence of vital mesorectum tissue in situ. This

supports the principle of close-rectal dissection. Figure 1 presents the MRI scans of a pa-

tient, showing the mesorectum at three levels in the pelvis, preoperatively and 6 months

after surgery. This indicates the presence of viable mesorectum postoperative comparable

to preoperative without any objective differences in volume.

176 Chapter 9

discussion

The surgical principles of the IPAA technique, originate from the total mesorectal excision

(TME). The TME technique is based on oncological principles for rectal carcinomas. The

TME removes the entire rectum including the mesorectum.23 In patients with oncological

diseased rectums the mesorectum has to be removed because of possible lymphatic in-

volvement. This is where the discrepancy comes in: the proctectomy for UC is no oncologic

surgery. There is no reason to go along with the TME principle. The CRD is a well known

“old” technique, but used to be too laborious to gain support.12 It leaves the mesorectum

in situ. With the advent of the automated vessel sealers, the CRD has gained renewed

interest.

This pilot study aimed to assess to the feasibility and evaluate the potential of the CR-IPAA

with automated vessel sealers to offer an alternative for the conventional IPAA. Several

important findings have been made. First, the new CR-IPAA technique is a safe proce-

dure, which could easily be learned and performed with the aid of the new technical

Figure 1. Comparison of the mesorectum preoperatively surrounding the rectum versus postopera-tively surrounding the ileo-anal pouchArrows indicate the mesorectum. Very low in the pelvis, just above the anal sphincter, there is physically no mesorectum in situ. At middle-high and high levels in the pelvis the mesorectum is shown. There are no objec-tive differences in mesorectum volume.

Chapter 9 177

9

developments like the electrothermal bipolar vessel sealer or ultrasound dissection. This

straight forward technique is reflected in a short operative time and with low blood loss,

comparable to the open IPAA procedure.24

In addition, the theoretical advantage of more space after mesorectal excision for the

pouch doesn’t seem to last: the defecation frequency after CR-IPAA remains the same

compared to the IPAA after TME principles. The mesorectum seems to function as a kind

of “communicating vessel” together with the pouch in the pelvis. The MRI scans confirm

this theory.

This pilot study showed a promising tendency regarding pelvic-related postoperative

complications. Only one of the patients developed a retropouch-localised abscess, without

systemic sign of infection and therefore no pelvic sepsis This is in line with the hypothesis

that the remnant mesorectum can protect against pelvic sepsis if there is an anastomotic

leakage: symptoms will be minimized. Pre-operative high dose steroid use has a negative

influences on the morbidity.25 Therefore we performed so many three-stage procedures.

The immunomodulating medication could be reduced after the first stage.

In addition, there were no signs of autonomic nerve damage in the ten patients: no sexual

dysfunction and bladder dysfunction was reported. This also contributes to the hypothesis

that the close-rectal dissection protects against nerve damage.

The functional outcome at one year follow-up after the CR-IPAA was comparable to

that of the conventional IPAA observed in literature. In an extensive review about the

long-term outcome of the conventional IPAA, Hueting et al.11 observed that an average

daytime frequency of 5.2 and nocturnal frequency of 1.0 was achieved. Tekkis et al.10

reported a daytime frequency of 5 and nocturnal frequency of 1.0. Whether in the long

term the retained mesorectum will hamper the reservoir function of the pouch needs to

be researched.

First sense, first urge and maximum tolerated volume of the pouch are closely related to

functional outcome.26 Therefore, the significantly increased volumes of these parameters

in our population support a good functional outcome. Our results of maximal anal rest-

ing pressure are comparable to a study of Selvaggi et al, while the maximum tolerated

volume of Selvaggi seems to be higher.27 But the functional results in terms of defecation

frequency are comparable. Anorectal manometry measurements should only be used as

a contribution to the overall results. Defecation frequency and continence are the most

important parameters in terms of functional results.

178 Chapter 9

A limitation of this pilot study is the small sample size. A larger study with long-term

follow-up will be needed to reliably assess the hypothesized benefits of the CR-IPAA. There

are twelve studies reporting to have used the close-rectal dissection technique.14-18,28-34

One study34 has a population with both close-rectal and total mesorectal excision, so

the results of this study cannot be analyzed. But the five largest studies, with over 100

patients, show contradictory results: Rink et al.18 and Berry et al.14 show a very low leakage

rate and no autonomic nerve damage respectively, due to the close-rectal dissection. But

Jarvinen et al.15, Araki et al.28 and Régimbeau et al.33 present a normal leakage rate and

sexual dysfunction percentage, comparable to the conventional technique. The reason for

these varying results can be different use of technique: partial versus complete and circular

close-rectal dissection. None of the studies specify the technique. Another reason for the

difference in results is the use of conventional versus new diathermic instruments.

However, with the recent developed automated vessel sealers, the traditional laborious

close-rectal IPAA technique has been greatly simplified. The circular close rectal dissection

with automated vessel sealers is technically easy, comes along with minimal blood loss and

a vital, dry mesorectal bedding, which shows in these 10 patient to function as a shield in

case of pelvic sepsis and protects the autonomic nerves.

conclusion

Given the ease of the CRD technique, the absence of oncological arguments and potential

advantages on complications, the question should be raised whether the CR-IPAA should

replace the TME technique in case of an IPAA for UC patients. Results so far of the CR-IPAA

need further investigation to substantiate potential benefits of this technique in a large

well conducted trial.

Chapter 9 179

9

references

1. Cohen JL, Strong SA, Hyman NH, Buie WD,

Dunn GD, Ko CY, Fleshner PR, et al. Practice



rectum. 2005; 48: 1997-2009.

2. Cottone M, Scimeca D, Mocciaro F, Civita-

vecchia G, Perricone G, Orlando A. Clinical

course of ulcerative colitis. Digestive and

liver disease: official journal of the Italian

Society of Gastroenterology and the Italian

Association for the Study of the Liver. 2008;

40 Suppl 2: S247-52.



British medical journal. 1978; 2: 85-8.







surgery. 2006; 244: 18-26.







406-12.

6. de Montbrun SL, Johnson PM. Proximal

diversion at the time of ileal pouch-anal

anastomosis for ulcerative colitis: Current

practices of north american colorectal sur-

geons. Diseases of the colon and rectum.

2009; 52: 1178-83.

7. Chapman JR, Larson DW, Wolff BG, Dozois

EJ, Cima RR, Pemberton JH, Crownhart BS,

et al. Ileal pouch-anal anastomosis: Does

age at the time of surgery affect outcome?

Archives of surgery (Chicago, Ill.: 1960).

2005; 140: 534-9; discussion 39-40.

8. Dayton MT, Larsen KR, Christiansen DD.

Similar functional results and complica-

tions after ileal pouch-anal anastomosis in

patients with indeterminate vs ulcerative

colitis. Archives of surgery (Chicago, Ill.:

1960). 2002; 137: 690-4; discussion 94-5.

9. Kiran RP, Remzi FH, Fazio VW, Lavery IC,

Church JM, Strong SA, Hull TL. Complica-

tions and functional results after ileoanal

pouch formation in obese patients. Journal

of gastrointestinal surgery: official journal

of the Society for Surgery of the Alimentary

Tract. 2008; 12: 668-74.

10. Tekkis PP, Lovegrove RE, Tilney HS, Smith JJ,

Sagar PM, Shorthouse AJ, Mortensen NJ, et

al. Long-term failure and function after re-

storative proctocolectomy - a multi-centre

study of patients from the uk national

ileal pouch registry. Colorectal disease: the



2010; 12: 433-41.







12. Lee EC, Dowling BL. Perimuscular excision

of the rectum for crohn’s disease and ulcer-

ative colitis. A conservation technique. The

British journal of surgery. 1972; 59: 29-32.

13. Kossi J, Luostarinen M, Kontula I, Laato

M. Laparoscopic sigmoid and rectal resec-

tion using an electrothermal bipolar vessel

sealing device. Journal of laparoendoscopic

& advanced surgical techniques. Part A.

2007; 17: 719-22.

14. Berry AR, de Campos R, Lee EC. Perineal

and pelvic morbidity following perimuscu-

lar excision of the rectum for inflammatory

bowel disease. The British journal of sur-

gery. 1986; 73: 675-7.

15. Jarvinen HJ, Luukkonen P. Experience with

restorative proctocolectomy in 201 pa-

tients. Annales chirurgiae et gynaecologiae.

1993; 82: 159-64.

180 Chapter 9

16. Lindsey I, George BD, Kettlewell MG,

Mortensen NJ. Impotence after mesorectal

and close rectal dissection for inflammatory

bowel disease. Diseases of the colon and

rectum. 2001; 44: 831-5.

17. Pescatori M, Mattana C, Castagneto M.

Clinical and functional results after restor-

ative proctocolectomy. The British journal

of surgery. 1988; 75: 321-4.







19. Andriesse GI, Gooszen HG, Schipper

ME, Akkermans LM, van Vroonhoven TJ,

van Laarhoven CJ. Functional results and

visceral perception after ileo neo-rectal

anastomosis in patients: A pilot study. Gut.

2001; 48: 683-9.

20. Strijbos SA, Hueting WE, Schipper ME,

Oostvogel HJ, van Vroonhoven TJ, Gooszen

HG, van Laarhoven CJ. The ileo neo rectal

anastomosis (inra) in patients with familial

adenomatous polyposis: Clinical results at

two years. Colorectal disease: the official



354-9.

21. van Laarhoven CJ, Hueting WE, Schipper

ME, Oostvogel HJ, Akkermans LM, van

Vroonhoven TJ, Gooszen HG. Ileo-neorectal

anastomosis: Medium- and long-term

follow-up of 37 patients. Digestive surgery.

2004; 21: 371-8; discussion 79.

22. Schroeder KW, Tremaine WJ, Ilstrup DM.

Coated oral 5-aminosalicylic acid therapy

for mildly to moderately active ulcerative

colitis. A randomized study. The New

England journal of medicine. 1987; 317:

1625-9.

23. Maurer CA, Renzulli P, Kull C, Kaser SA,

Mazzucchelli L, Ulrich A, Buchler MW.

The impact of the introduction of total

mesorectal excision on local recurrence rate

and survival in rectal cancer: Long-term

results. Annals of surgical oncology. 2011;

18: 1899-906.








25. Ismael H, Horst M, Farooq M, Jordon J,

Patton JH, Rubinfeld IS. Adverse effects of

preoperative steroid use on surgical out-

comes. American journal of surgery. 2011;

201: 305-8; discussion 08-9.

26. Rink AD, Nagelschmidt M, Radinski I, Vest-

weber KH. Evaluation of vector manometry

for characterization of functional outcome

after restorative proctocolectomy. Interna-

tional journal of colorectal disease. 2008;

23: 807-15.

27. Selvaggi F, Giuliani A, Gallo C, Signori-

ello G, Riegler G, Canonico S. Randomized,

controlled trial to compare the j-pouch

and w-pouch configurations for ulcerative

colitis in the maturation period. Diseases of


28. Araki T, Parc Y, Lefevre J, Dehni N, Tiret E,

Parc R. The effect on morbidity of mesen-

tery lengthening techniques and the use

of a covering stoma after ileoanal pouch

surgery. Diseases of the colon and rectum.

2006; 49: 621-8.

29. Gignoux BM, Dehni N, Parc R, Tiret E. [ileal

pouch anal-anastomosis without protective

ileostomy]. Gastroenterologie clinique et

biologique. 2002; 26: 671-4.

30. Gullberg K, Liljeqvist L. Stapled ileoanal

pouches without loop ileostomy: A pro-

spective study in 86 patients. International


221-7.

31. Hallberg H, Stahlberg D, Akerlund JE. Ileal

pouch-anal anastomosis (ipaa): Functional

outcome after postoperative pelvic sepsis.

Chapter 9 181

9

A prospective study of 100 patients. Inter-

national journal of colorectal disease. 2005;

20: 529-33.

32. Mukhtar H, Zak A, Berry AR. Restorative

proctocolectomy: One surgeon’s experience

of a now routine procedure in a district

general hospital. Journal of the Royal Col-

lege of Surgeons of Edinburgh. 2002; 47:

400-6.

33. Regimbeau JM, Panis Y, Pocard M, Haute-

feuille P, Valleur P. Handsewn ileal pouch-

anal anastomosis on the dentate line

after total proctectomy: Technique to avoid

incomplete mucosectomy and the need for

long-term follow-up of the anal transition

zone. Diseases of the colon and rectum.

2001; 44: 43-50; discussion 50-1.

34. Wasmuth HH, Trano G, Endreseth B, Ryd-

ning A, Wibe A, Myrvold HE. Long-term

surgical load in patients with ileal pouch-

anal anastomosis. Colorectal disease: the



2009; 11: 711-8.

Part III

summary and ConClusIons

10 summary

Chapter 10 187

10

suMMary

The first part of this thesis focusses on evaluating the current state of important outcomes

in inflammatory bowel disease (IBD) surgery. First, we evaluated the contemporary out-

comes of the ileal pouch-anal anastomosis (IPAA) procure in terms of complications and

functional outcome (Chapter 2). This treatment of choice in patient with ulcerative colitis

(UC) has been performed for over 30 years, a period encompassing various improvements

in surgical equipment and perioperative care, as well as fine-tuning of surgical technique.

We performed an extensive review of literature and identified 53 observational studies

from the year 2000 onwards. Results were compared to a previously performed systematic

review with similar methodology describing the results of 43 studies published before

2000. This allowed for comparison of pooled estimates of a large and representative

sample from both eras with a total of 24,283 included patients.

We found a pooled rate of pouch failure of 4.3% (95% CI 3.5 to 6.3) for studies published

after 2000, which was 2.5% less than the rate before 2000 (p=0.004). An additional

time-effect analysis confirmed a decline in pouch failure over time. The rate of pelvic

sepsis was 7.5% (95% CI 6.1 – 9.1) and 9.5 (8.2–10.9) in the periods after and before

2000, respectively, but the difference was not statistically significant. Functional outcome

remained stable over time, with a 24-hour defecation frequency of 5.9 (95% CI 5.0 to

6.9). Variations in the technical aspects of the operations, such as routine use of ileostomy,

did not have an important effect on outcome. Thus, after decennia of experience with the

IPAA, the rate of certain severe complications such as pouch failure and pelvic sepsis seem

to have declined, but functional outcomes did not improve. The estimates provided by this

review can be useful as reference values for practice and research.

Secondly, we examined the length of hospital stay after colorectal surgery from a hospital

cost-perspective. In countries all over the world a shift has been observed in reimburse-

ment schemes for health care. Instead of the traditional fee-for-service scheme, countries

are adapting the fixed reimbursement scheme, such as the Diagnosis Related Groups (DRG)

in the USA and the ‘Diagnose-Behandelcombinatie’ (DBC) in the Netherlands. In these

schemes, patients are assigned to predetermined groups based on diagnosis, received

treatments and several other characteristics (e.g. co-morbidities or occurrence of certain

complications). Hospitals are reimbursed a certain fixed amount per patient based on the

group he or she is assigned to. For each group, an estimate of the expected hospital stay

is used to calculate this fixed reimbursement.

In contrary to the usual approach in which the absolute hospital stay is studied, we exam-

ined which factors are responsible for the largest deviation of the actual versus estimated

188 Chapter 10

hospital stay, and thus for the largest financial burden for hospitals (Chapter 3). A total of

1,461 patients were included in this single-department consecutive cohort. The average

expected and actual hospital stay were virtually equal, being 8.17 (IQR 4.7-11.9) and 8.31

(IQR 4.0-10.0) days, respectively. As hypothesized, certain factors were associated with

a large difference between the actual and expected length of stay. Parenteral nutrition

showed the largest deviation with an increase of about 5.1 days. Other factors showing

significant increases were emergency room admittance (3.67), ileus (3.45), blood transfu-

sion (2.3), anastomotic leak (2.2), sepsis (2.0) and venous thrombosis (1.9 days). While not

all factors are modifiable, some of them form a potential target for cost-savings measures.

The next step will be to examine hospital policies regarding factors showing large increases

and attempt to modify them were possible to effectuate the largest possible savings.

In the final chapter of part I, we examine the risk of cancer in colitis patients diagnosed

with dysplasia at colonoscopic surveillance. The aim of the study is to utilize the informa-

tion available preoperatively, such as patient characteristics and findings at colonoscopic

surveillance, to predict the presence and location of cancer in such patients. For this, we

identified 348 patients with colitis that underwent a proctocolectomy due to dysplasia in a

period of over 20 years (Chapter 4). Pathohistological examination showed the presence

of cancer in 51 (15%) patients, while some degree of dysplasia was observed in about

50% of patients. In patients with high-grade dysplasia preoperatively, the risk of cancer

was 29%. The presence of dysplasia-associated lesions or masses (DALM) preoperatively

was also associated with high risk of cancer of about 25%. These results were confirmed

by multivariate analysis in which high-grade dysplasia, DALM and disease duration were

independent risk factors for postoperative cancer. These results support surgical manage-

ment by a proctocolecotmy in virtually all patients with either high-grade dysplasia or

DALM during preoperative surveillance.

Patients with low-grade dysplasia (LGD) without DALM (i.e. flat LGD) had a relatively low

risk of cancer (2%) and a moderate risk of high-grade dysplasia (7%). Management of

these patients is challenging since it entails a balance between the small but present risk

of cancer and the risk of morbidity and mortality associated with surgery. This paper helps

physicians inform their patients more accurately and thus allows for better shared-decision

making. In regard to the location of histopathological findings, we observed that dysplasia

and cancer can be present in segments adjacent to and remote from the preoperative

location. This also supports the choice for a total proctocolectomy (+/- restoration of intes-

tinal continuity) for most patients. Patients with isolated colonic dysplasia without sigmoid

involvement had low rates of rectal dysplasia (3%) and rectal cancer (0%). Therefore, in

selected patients with a high-risk of morbidity a colectomy with rectal preservation may

obviate some of the risks of a pelvic dissection and provide an acceptable alternative.

Chapter 10 189

10

In part II, we evaluate various aspects of the surgical treatment of IBD aiming at optimizing

these treatments and findings ways to improve patients’ outcomes. First, we compare the

upcoming laparoscopic approach for the IPAA procedure with the established open ap-

proach. With the advancements of minimal-invasive surgery this demanding operation has

been increasingly performed laparoscopically. We evaluated the presumed benefits of the

laparoscopic approach in a systematic Cochrane review (Chapter 5). Eleven comparative

trials (including 607 patients) were included, of which only 1 was randomized. Analysis

showed no significant differences in complications or functional outcomes. Operative time

was significantly longer in the laparoscopic approach by about 1.5 hours. In regard to post-

operative recovery parameters, no significant differences were seen when only published

data suitable for meta-analysis (i.e. means and standard deviations) were used. However,

when all available published data were used to impute the missing measures, a significant

reduction in time-to-bowel movement (-1.96 days), time-to-regular diet (-1.30 days) and

hospital stay (-1.88 days) were observed. Total incision length was significantly shorter in

the laparoscopic group, while two trials evaluating cosmesis found significantly improved

cosmesis scores in the laparoscopic group. While more research is required, this review did

show that the laparoscopic IPAA is a feasible and safe procedure that has the potential for

quicker postoperative recovery. Moreover, cosmesis and body image were superior in the

laparoscopic group, which can be important in this young, active and highly motivated

patient group.

While IPAA is the treatment of choice for most patients, certain patients might benefit

from a less invasive procedure to reduce the operative morbidity. In such patients, a total

proctocolectomy with end-ileostomy is possible, but it does not allow for restoring the

continuity of the intestinal tract. In certain cases, a subtotal colectomy with ileorectal

anastomosis (IRA) might offer a valid alternative. IRA offers a one stage procedure that

does not require pelvic dissection, potentially reducing surgical morbidity and avoiding

the associated risk of sexual dysfunction. In order to examine whether these proposed

benefits are supported by evidence, we performed a narrative review comparing the IRA

to the IPAA procedure (Chapter 6). Overall morbidity was less after IRA, ranging from 8%

to 28%, compared to 15% to 63% after IPAA. Severe postoperative complications, except

for reservoir failure, were lower after IRA. Anastomotic leak, for example, was reported

between the 2.3 and 9.1% after IRA (weighted average of 4.5%), compared to an average

pelvic sepsis rate of 7.5% (95% CI 6.1 to 9.1) after IPAA. Rate of sexual dysfunction was

low after IPAA (around 3%), but this complication was virtually non-existent after IRA.

The overall mortality was comparable after IRA (0 and 4.0%) versus IPAA (0 and 2.9%).

Failure of the reservoir was higher after IRA with a reported frequency after 5-years of 7%

to 16%, compared to an average of 4.7% for IPAA. The risk of cancer was also higher

after IRA with reported frequency between 0 to 8% compared to <0.5% for IPAA, and

190 Chapter 10

an estimated cumulative risk of cancer after 20 years of 6% to 14% for IRA compared

to 4.2% for the IPAA approach. Functional results seemed to be better after IRA with

lower frequency of bowel movements and less night-time seepage, but with more urgency

compared to patients with an IPAA.

Thus, current evidence supports IPAA as the preferred approach for patients with UC

requiring surgical treatment. The definitive nature of the procedure with low risk of failure

and removal of all diseased mucosa with lower risk of cancer are the main advantages of

this technique. Nonetheless, there might be a role for IRA in selected patients. Patients

with higher surgical risk with mild rectal involvement and no evidence of dysplasia or

cancer at the time of intervention might benefit from the good functional outcomes of

an IRA procedure. These patients should, however, be aware about the risk of recurrent

proctitis and cancer in long standing disease, and have to fully understand the need for

meticulous surveillance. Unless these conditions are met, patients should not be offered an

IRA. Finally, patients not suitable for IPAA or with widely metastatic colorectal cancer may

also benefit from an IRA procedure.

Subsequently, we focused on the optimal management of an anastomotic pouch sinus.

This serious complication after IPAA is challenging to treat and carries a high risk of pouch

failure. Due to its relative infrequency, its course and optimal management are not well

defined. Therefore, we set out to evaluate the frequency, management and outcome of

anastomotic pouch sinus in a large prospectively collected database encompassing 2,286

patients that have underwent an IPAA procedure (Chapter 7). A total of 45 (2.0%) patients

were diagnosed with an anastomotic pouch sinus confirmed by imaging or examination

under anaesthesia. The pouch sinus was initially managed by observation in 23 (51%)

patients. Other treatment modalities included transrectal drainage, unroofing of the sinus

tract, sinus closure or diverting of the fecal stream. Healing rate after initial treatment

alone was low (26%). In 28 patients (62%), subsequent treatment was necessary. Eventu-

ally, sinus healing was achieved in 27 (60%) patients, whereas 15 (33%) developed pouch

failure. When examining factors influencing the rate of healing only symptomatic presenta-

tion was associated with worsened outcomes. The healing rate was significantly lower in

those patients compared to asymptomatic patients (30% vs. 84%, p=0.001). Pouch failure

tended to higher as well (45% vs. 24%, p=0.14). Of the treatments applied, a strategy

with observation as initial treatment was the most successful with a healing rate of 65%.

Based on findings in this study as well as a review of literature, we proposed an algo-

rithm for management of a pouch sinus after IPAA. Observation and watchful waiting

are recommended in asymptomatic patients or patients with incidentally encountered

sinuses (e.g. on a Gastrografin enema) prior to ileostomy closure (with delay of ileostomy

Chapter 10 191

10

close). Symptomatic sinuses and those that are defunctioned but non-healing on watchful

waiting alone, may be managed by trans-anal drain placement, unroofing of the sinus or

injection of fibrin glue. No clear evidence exists to favour one approach over the other,

and experience of the treating physician should be guiding. A defunctioning ostomy may

be considered in circumstances where eventual healing of the sinus may be expected in

patients who are not defunctioned. When attempts at healing have failed or may not

reasonably be anticipated due to the size or persistence of the sinus, a redo IPAA may be

considered.

Next we evaluated the impact of immunosuppressive agents in patients with CD that are

undergoing surgery. While such agents have become a corner-stone in the management

of CD, their impact on postoperative complication, especially infectious complications,

has been a topic of concern. To gain more insight into this potentially harmful effect, we

performed a systematic review including comparative studies on this topic (Chapter 8).

A total of 21 comparative studies including 6,899 patients were identified. We analysed

data separately for anti-tumor necrosis factor (TNF) agents (14 studies), corticosteroids (14

studies), thiopurines (9 studies) and mixed immunosuppressive agents (6 studies).

In meta-analysis, patients receiving anti-TNF agents and corticosteroids before surgery were

found to have a higher risk of postoperative infectious complications with a RR of 1.29

(95% CI 1.07 to 1.55) and 1.55 (95% CI 1.23 to 1.95), respectively. These results were

consistent in sensitivity analysis of studies with the highest methodological quality. Suf-

ficient data to compare individual complications were only present in the anti-TNF group,

which should a significant increase in rates of wound infection (RR 1.62, 95% CI 1.12 to

2.34) and septic shock (RR 1.81, 95% CI 1.03 to 3.17) when these drugs are administered

within 3 months of surgery. For corticosteroids, the risk of infectious complications was

greater when only patients on high-dose corticosteroids were included (RR 1.67, 95%

CI 1.31 to 2.13). Based on the limited data available, no association was seen between

use of thiopurines (RR 1.23 (95% CI 0.66 to 2.29) or combined immunomodulator drugs

(RR 1.01 (95% CI 0.67 to 1.54) and infectious postoperative complications. Overall, this

review shows that certain commonly used immunosuppressive agents do increase risk of

infectious postoperative complications, and that this should be considered when informing

patients and making decisions regarding the optimal timing of surgery.

Finally, we examined a variation in the way the pelvic dissection is performed during the IPAA

procedure. Pelvic dissection is most commonly performed according to the total mesorectal

excision (TME) principles, in which both the rectum and mesorectum are excised en-bloc.

This technique is the golden-standard for oncological resection, but is not strictly necessary

in benign disease. An alternative is the close-rectal dissection, which has been proposed

192 Chapter 10

in the 1970’s but was too laborious given the technical equipment at that time. In this

technique the rectum is dissected through the nonanatomic perimuscular plane, thereby

preserving the mesorectal fat layer that is just outside the rectum. Hypothesized benefits

include preservation of the pelvic autonomic nerves (responsible for sexual dysfunction)

and allowing for less room in the pelvis for postoperative hematoma or fluid collections,

potentially reducing the risk pelvic infectious complication. Due to developments in surgi-

cal equipment, especially electrothermal bipolar vessel sealer and ultrasound dissection,

this technique has been substantially simplified.

To examine the feasibility and evaluate potential benefits, we performed a pilot study

including 10 consecutive patients with UC undergoing an IPAA procedure (Chapter 9).

The procedure was technically successful in all patients. After a median follow-up of 16

months, no cases of pelvic sepsis or bladder- or sexual dysfunction were observed. The me-

dian daytime defecation frequency was 6.0, and anorectal physiological testing supported

the good functional results. Thus, results of the close-rectal pouch procedure in the current

era are promising, and support further comparative research to validate and quantify the

potential benefits of this technique.

11 Gained insiGhts and

future perspectives

Chapter 11 197

11

inTroducTion

In the preceding chapters, findings from several studies on the surgical treatment of in-

flammatory bowel disease (IBD) have been presented. In this final chapter, we present a

topic based discussion of the most important gained insights and future perspective based

on the entirety of the thesis.

role and alTernaTives of The ileal Pouch-anal anasToMosis (iPaa)

Based on current evidence, the IPAA is considered the current gold-standard for the treat-

ment of patients with ulcerative colitis (UC). An important finding from our thesis was

the stagnation (or stability) of functional outcomes after the IPAA procedure even after

3 decades of experience (Chapter 2). This probably represents an intrinsic limitation of

the IPAA procedure, which could be used to counsel patients and also to set accurate

expectations of outcomes in scientific research. We also showed that the laparoscopic ap-

proach has shown short-term advantages when compared to the open approach in terms

of postoperative recovery and cosmesis (Chapter 5). This combined with the increased

experience in laparoscopic colorectal surgery, has resulted in the laparoscopic IPAA to be-

come the preferred treatment modality by both physicians and patients. We also showed

promising results of a technical modification of the IPAA procedure (Chapter 9). This so

called close-rectal dissection (CRD) has been revived due to recent technical advances in

surgical instrumentation making it easier and safer. While only 10 patients were included

in this pilot study, results show that this alternative is feasible, safe and was associated with

promising results in regard to postoperative complications. A recent RCT confirmed the

hypothesized benefits of the CRD approach.1 In this RCT including 59 patients, the CRD

approach led to a lower severe complication rate (2 of 28 (7%) vs. 10 of 31 (32%)) and

better short-term quality of life than the traditional approach.

While the (laparoscopic) IPAA is suitable for most patients with UC, certain subgroups can

be identified in which an alternative approach such as the ileorectal anastomosis (IRA) is

beneficial. Due to unequal target populations, direct comparison between these two pro-

cedures was not feasible. A narrative approach in which all relevant evidence is presented

was performed, with focus on risk of postoperative complications and cancer as important

determinates in the choice between the two interventions (Chapter 6). This has resulted in

concrete recommendations to guide selection of patients for each procedure. Patients with

higher surgical risk with mild rectal involvement and no evidence of dysplasia or cancer

at the time of intervention might benefit from the good functional outcomes of an IRA

198 Chapter 11

procedure. It can be also proposed to young women as a possible interim procedure based

on concerns for infertility after IPAA. Patients should be aware of the risk of recurrent

disease and cancer, and be compliant with surveillance.

In regard to the increased risk of cancer in patients with UC, this thesis have shown that

patients with high-grade dysplasia or dysplasia-associated lesion/mass on colonoscopic

surveillance have a high risk of synchronize colon carcinoma, and should undergo a proc-

tocolectomy (Chapter 4). Management of patients with flat low-grade dysplasia (LGD),

on the other hand, is particularly challenging. In our study we have shown that the risk

of cancer in flat LGD is 2%, which is at the lower end of the range reported by previous

studies (i.e. 0 to 20%).2-4 Triggered by the large variation reported in literature, the Position

Statement by the American Gastroenterological Association (AGA) states that current evi-

dence is insufficient to either advice for or against colectomy for flat LGD.5,6 We propose an

individualized approach in which surgery should be considered in patients with additional

risk factors for cancer as identified in literature. Such risk factors included young age, long

duration of disease, multifocal dysplasia especially if located distally, extensive disease (e.g.

pancolitis) and positive family history.7,8 In other cases, colonoscopic surveillance after dis-

cussion of associated risks is acceptable, provided high patient compliance can be assured.

Future research, especially longitudinal cohort studies aiming at reliably identifying the

rate and predictors of progression of flat LGD to advanced neoplasia is needed to further

fine-tune these recommendations.

PrevenTion and ManaGeMenT of coMPlicaTions afTer ibd surGery

Given the increased importance of immunosuppressive agents in the treatment of Crohn’s

disease (CD), we evaluated their effect on postoperative complications (Chapter 8). We

were able to show an increase in overall infectious complications for certain commonly

used immunosuppressive agents. While not a substitute for well-designed and performed

studies, the synthesis of all available evidence into meta-analyses with consideration of

the risk of bias allowed us to identify these potentially important differences. We feel that

this review provides sufficient evidence to advice a preoperative drug free interval when

feasible to reduce the risk of infectious complications. The adoption of specific operative

strategies that may reduce these risks or their effects, such as the use of a temporary

ostomy, may additionally be considered in high risk patients.

For studying uncommon postoperative complications, such as anastomotic pouch sinuses

after IPAA, large surgical registries are indispensable. Using such database we were able

Chapter 11 199

11

for the first time to present the full spectrum of incidence, presentation, treatment and

outcomes of this complication (Chapter 7). These findings are important in that they can

raise the alertness of physicians to this condition, provide guidance in the treatment of its

different presentations and help counsel patients. This study shows that observation was

the preferred initial management for pouch sinuses with reasonable healing rates. When

this treatment strategy fails, several other options as well as their success rates have been

presented. No superior method was identified, and choice of treatment should rely on

availability and experience of the treating physician.

challenGes in surGical research MeThodoloGy

Finding the ideal timing to perform randomized trials for innovative surgical treatments

is challenging.9 RCTs should be done as early as possible after introduction of a new

intervention to reduce the possibility that the new innovation becomes widely adopted

and strong preferences of physicians or patients hinders delivery of the allocated interven-

tion. However, randomization needs to be sufficiently delayed until the surgical team is

adequately prepared, the new procedure is considered safe and feasible, and the learn-

ing curve has been passed.10 This combined with the cost, ethical and organisational

restraints of performing RCTs makes surgeons often prefer non-randomized studies or

(even worse) historic comparison of cohorts over time. This challenge is clearly illustrated

in our systematic review of open versus laparoscopic IPAA (Chapter 5). Only one RCT was

identified comparing laparoscopic vs. open IPAA, compared to 10 non-randomized studies.

Moreover, it turned-out the second on-going RCT that was identified by our review had

to be stopped prematurely due to insufficient recruitment because of a marked patients’

preference in favour of the laparoscopic approach.11 The researchers were eventually only

able to recruit 42 out of the 160 intended patients.

This consequently results in two important challenges for the surgical field. First, given the

relatively short window in which RCTs can be conducted, performing RCTs as soon as pos-

sible should receive the highest priority of all innovating centres. Surgeons have to take the

initiative in discussing this limitation with ethical boards as well as governmental regulatory

and funding organisation in order to facilitate the conductance and funding of RCTs for

promising surgical innovations. Secondly, we need to find ways to deal with the suboptimal

body of evidence that will inevitably be produced. This has to be done in a way that

ensures maximal benefit of the available good quality evidence, without getting misled by

the potential biases of lower quality evidence. The Cochrane review performed in Chapter

5 of this thesis is an attempt in this direction, since it was one of the few Cochrane reviews

incorporating a meta-analysis of non-randomized trials at the time. It was also one of the

200 Chapter 11

first Cochrane reviews to incorporate imputed estimates of mean and standard deviation

when these were not published, but other data such as medians and ranges were available.

Currently, similar approaches are becoming more familiar. For example, the GRADE clas-

sification which is now part of all Cochrane reviews allows for non-randomized evidence to

be classified as moderate or even high quality, when studies are well performed and results

show certain signs that reinforces their reliability (e.g. a strong magnitude of effects or if

all plausible biases would reduce an apparent treatment effect).12

Finally, in relation to health costs research, by looking at clinical outcomes from the

perspective of disproportionate costs compared to the reimbursement received by the

hospital, we add a dimension that is often overlooked (Chapter 3). Results from such

approach can help guide policy making and prioritize research funding and efforts towards

targeting factors with the largest cost impact. A limitation of this approach is the fact that

regional and local differences between hospitals might make results of a study not directly

generalizable to other hospitals. Thus, each hospital has to perform their own analysis of

factors causing the most excessive costs. This same limitation, however, does come with

an unexpected advantage. Using this methodology of reporting outcomes offers a tool

to more accurately compare results of hospitals with different patient populations, since

outcomes are normalized based on the expected outcomes of that case-mix.

conclusions

The IPAA is the gold-standard treatment of patients with UC. Functional results after this

procedure has been stable for the last 30 years, suggesting no further improvement is to

be expected. The laparoscopic IPAA has been associated with improved convalescence and

cosmetics and have become the preferred method. Performing an IPAA procedure using

a close-rectal dissection, rather than the usual total mesorectal excision (TME) method,

might reduce the risk of certain postoperative complications and increase the quality of

life. For specific patients an IRA might be a suitable alternative, although the risk of cancer

and proctitis should be taken into account. In regard to the risk of cancer, colitis patients

with high-grade dysplasia or dysplasia-associated lesion/mass on colonoscopic surveillance

have a high risk of synchronize colon carcinoma, and should undergo a proctocolectomy.

Patients with low-grade dysplasia need to undergo a personalized assessment in which

surgery should be considered in patients with additional risk factors for cancer.

Use of certain immunosuppressive agents prior to surgery in patients with CD, is associated

with an increase in overall infectious complications. Physicians should be aware of this

increased risk and a preoperative drug free interval when feasible is advised to reduce this

Chapter 11 201

11

risk. For the management of anastomotic pouch sinuses after IPAA, observation has been

shown to be the optimal initial management. If this fails, several options are available

with current evidence being insufficient to provide recommendations on their efficacy.

Therefore, the choice of treatment should rely on availability and experience of the treating

physician.

Performing randomized controlled trials in the field of surgery remains challenging. A

genuine effort to perform RCTs in the brief window of opportunity is important. Similarly

a collective initiative to reduce the financial and organization burden of surgical RCTs on

physicians and institutions is warranted. Finally, identification of clinical outcomes that

result in the largest disparity between costs incurred and the reimbursement received by

hospitals can help reduce health costs. Targeting these outcomes, if possible, to reduce

their occurrence can provide an efficient way to maximise health cost savings.

202 Chapter 11

references

1. Bartels SA, Gardenbroek TJ, Aarts M, Pon-

sioen CY, Tanis PJ, Buskens CJ, Bemelman

WA. Short-term morbidity and quality of

life from a randomized clinical trial of close

rectal dissection and total mesorectal exci-

sion in ileal pouch-anal anastomosis. The

British journal of surgery. 2015; 102: 281-7.

2. Befrits R, Ljung T, Jaramillo E, Rubio C. Low-

grade dysplasia in extensive, long-standing

inflammatory bowel disease: A follow-up

study. Diseases of the colon and rectum.

2002; 45: 615-20.

3. Lim CH, Axon AT. Low-grade dysplasia:

Nonsurgical treatment. Inflammatory

bowel diseases. 2003; 9: 270-2; discussion

73-5.

4. Lindberg B, Persson B, Veress B, Ingelman-

Sundberg H, Granqvist S. Twenty years’

colonoscopic surveillance of patients with

ulcerative colitis. Detection of dysplastic

and malignant transformation. Scandina-

vian journal of gastroenterology. 1996; 31:

1195-204.

5. Farraye FA, Odze RD, Eaden J, Itzkowitz

SH. Aga technical review on the diagnosis

and management of colorectal neoplasia

in inflammatory bowel disease. Gastroen-

terology. 2010; 138: 746-74, 74.e1-4; quiz

e12-3.

6. Farraye FA, Odze RD, Eaden J, Itzkowitz SH,

McCabe RP, Dassopoulos T, Lewis JD, et

al. Aga medical position statement on the

diagnosis and management of colorectal

neoplasia in inflammatory bowel disease.


7. Jess T, Simonsen J, Jorgensen KT, Pedersen

BV, Nielsen NM, Frisch M. Decreasing risk

of colorectal cancer in patients with

inflammatory bowel disease over 30 years.

Gastroenterology. 2012; 143: 375-81.e1;

quiz e13-4.

8. Navaneethan U, Jegadeesan R, Gutierrez

NG, Venkatesh PG, Hammel JP, Shen B, Ki-

ran RP. Progression of low-grade dysplasia

to advanced neoplasia based on the loca-

tion and morphology of dysplasia in ulcer-

ative colitis patients with extensive colitis

under colonoscopic surveillance. Journal of

Crohn’s & colitis. 2013; 7: e684-91.

9. Ergina PL, Cook JA, Blazeby JM, Boutron

I, Clavien PA, Reeves BC, Seiler CM, et al.

Challenges in evaluating surgical innova-

tion. Lancet (London, England). 2009; 374:

1097-104.

10. Barkun JS, Aronson JK, Feldman LS, Mad-

dern GJ, Strasberg SM, Altman DG, Barkun

JS, et al. Evaluation and stages of surgical

innovations. Lancet (London, England).

2009; 374: 1089-96.

11. Schiessling S, Leowardi C, Kienle P, Antolovic

D, Knebel P, Bruckner T, Kadmon M, et al.

Laparoscopic versus conventional ileoanal

pouch procedure in patients undergoing

elective restorative proctocolectomy (lap-

conpouch trial)-a randomized controlled

trial. Langenbeck’s archives of surgery /

Deutsche Gesellschaft fur Chirurgie. 2013;

398: 807-16.

12. Guyatt GH, Oxman AD, Kunz R, Vist GE,

Falck-Ytter Y, Schunemann HJ. What is

“quality of evidence” and why is it impor-

tant to clinicians? BMJ (Clinical research

ed.). 2008; 336: 995-8.

12 nederlandse samenvattinG

(dutch summary)

Chapter 12 207

12

saMenvaTTinG

In het eerste deel van dit proefschrift wordt de huidige status van chirurgie van inflam-

matoire darmziekten (IBD) onderzocht. De ileo-anale anastomose met ileumreservoir (IPAA)

wordt ruim 30 jaar verricht en wordt gezien als de ‘gouden standaard’ voor de chirurgische

behandeling van patiënten met colitis ulcerosa. De invloeden op deze operatie van ontwik-

kelingen in de perioperatieve zorg en verfijning van chirurgische technieken worden in een

systematische review van de literatuur onderzocht (Hoofdstuk 2). Behandeluitkomsten in

53 observationele studies na het jaar 2000 werden vergeleken met de resultaten van 43

studies die gepubliceerd waren voor het jaar 2000.

Het falen van het ileumreservoir was gedaald van 6,8% (95% betrouwbaarheidsinterval

(BI) 5,8 tot 8,4) voor 2000 naar 4,3% (BI 3,5 tot 5,3) na 2000 (p=0,004). Er werd ook een

daling geconstateerd in het percentage van septische complicaties van 9,5% (BI 8,2 tot

10,9) naar 7,5% (BI 6,1 tot 9,1), maar dit verschil was niet statistisch significant. Functi-

onele uitkomsten waren gelijk in de tijd. Variaties in de operatieve procedure, zoals het

wel of niet aanleggen van een ontlastend stoma, hadden geen effect op deze uitkomsten.

Concluderend heeft in de afgelopen jaren een verfijning van de IPAA techniek gezorgd

voor een daling van de meest ernstige complicaties, zonder dat de functionele uitkomsten

verder verbeterden. Deze getallen kunnen als een graadmeter gebruikt worden voor de

huidige praktijk en wetenschappelijk onderzoek. De stabiele functionele uitkomsten repre-

senteren waarschijnlijk de biologische limiet dat het ileumreservoir kan realiseren.

Een ander actueel onderwerp is het bestuderen van de opnameduur na colorectale

chirurgie in het kader van kosteneffectiviteit. Wereldwijd is er een verschuiving gaande

van directe vergoedingen (‘fee-for-service’) naar geïntegreerde vergoedingen zoals het

DBC (Diagnose-Behandelcombinatie) systeem in Nederland. Een vergelijkbaar systeem in

de VS is de ‘Diagnosis Related Groups’ (DRG). Ziekenhuizen krijgen een vast bedrag per

patiënt afhankelijk van de diagnose, behandeling en een aantal andere karakteristieken

(bijvoorbeeld co-morbiditeiten en complicaties). Patiënten worden op basis van deze gege-

vens in groepen verdeeld. Bij iedere groep hoort een verwachte opnameduur waarop de

vergoeding is berekend.

In de meeste wetenschappelijke studies wordt de absolute opnameduur bestudeerd. In

onze studie werd gekeken naar de discrepantie tussen de verwachte opnameduur (o.b.v.

de DRG groep) en de daadwerkelijke opnameduur (Hoofdstuk 3). In totaal werden er

1.461 opeenvolgende patiënten die colorectale operaties hebben ondergaan geïncludeerd.

Overall was de gemiddelde verwachte en daadwerkelijke opnameduur voor de gehele

groep vrijwel gelijk, namelijk 8,17 (interkwartielbereik (IQR) 4,7 tot 11,9) en 8,31 (IQR 4,0

208 Chapter 12

tot 10) dagen, respectievelijk. Via multivariate analyse werden verschillende individuele

factoren nader onderzocht. De opnameduur werd vooral beïnvloed door het gebruik van

parenterale voeding (5,1 dagen langer dan verwacht). Andere factoren die grote verschil-

len lieten zien waren opname vanaf de spoedeisende hulp (3,7 dagen), ileus (3,5 dagen),

bloedtransfusie (2,3 dagen), naadlekkage (2,2 dagen), sepsis (2,0 dagen) en veneuze

trombose (1,9 dagen). Deze resultaten kunnen gebruikt worden om een prioritering aan

te brengen in de factoren waarnaar gekeken dient te worden om een maximale kosten-

besparing te realiseren. Een gedegen analyse op ziekenhuis- en afdelingsniveau zijn hierbij

essentieel om te kijken naar verbeterpunten in relevante protocollen en procedures (bijv.

omtrent parenterale voeding).

In Hoofdstuk 4, werd het risico op kanker bij patiënten met colitis ulcerosa waarbij

dysplasie is vastgesteld bij coloscopie onderzocht. Het doel was om te beoordelen of de

aanwezigheid en locatie van kanker voorspeld kan worden op basis van de coloscopie. Er

werden 348 patiënten met colitis ulcerosa geïncludeerd die een proctocolectomie had-

den ondergaan in verband met dysplasie. Pathologisch onderzoek liet kanker zien in 51

(15%) van de patiënten. In patiënten met hooggradig dysplasie was de kans op kanker

29%. De aanwezigheid van de zogenaamde met dysplasie samenhangende afwijking

of massa (‘dysplasia associated lesion or mass’ (DALM)) ging gepaard met een kans op

kanker van 25%. In multivariate analyse, waren hooggradig dysplasie, DALM en duur van

de ziekte onafhankelijke voorspellers voor de aanwezigheid van kanker. Deze resultaten

ondersteunen de keuze voor een resectie als standaardbehandeling voor patiënten met

deze afwijkingen.

Patiënten met laaggradige dysplasia zonder DALM hadden een relatief laag risico van

kanker (2%) en middelmatig risico op hooggradige dysplasie (7%). Behandeling van deze

patiënten is uitdagend, aangezien het risico van morbiditeit en mortaliteit van de operatie

gewogen dient te worden tegen het lage maar reële risico op kanker. Ten aanzien van de

locatie van de afwijkingen, werd aangetoond dat in het postoperatieve preparaat kanker

of dysplasie regelmatig gevonden wordt in andere delen van het colon dan waar de pre-

operatieve afwijking werd gezien. Daardoor blijft een proctocolectomie de operatie van

keuze voor de meeste patiënten. Bij patiënten met uitsluitend afwijkingen proximaal van

het sigmoïd bij de coloscopie, was de kans op rectale afwijkingen laag (i.e. 3% dysplasie

en 0% kanker). In geselecteerde patiënten met hoge operatierisico’s kan in deze groep

patiënten een subtotale colectomie overwogen worden om het complicatierisico’s bij een

rectumresectie te vermijden.

In het tweede deel van dit proefschrift, wordt nader in gegaan op het optimaliseren van

de chirurgische behandelingen bij IBD. Door de ontwikkeling van de minimaal-invasieve

Chapter 12 209

12

chirurgie wordt de IPAA steeds vaker laparoscopisch uitgevoerd. In Hoofdstuk 5, werd

in een systematische Cochrane review deze techniek onderzocht. Analyse van 11 verge-

lijkende studies met 607 patiënten liet geen verschil zien in complicaties en functionele

uitkomst. Operatietijd van de laparoscopische operatie was significant langer (gemiddeld

1,5 uur langer). Ten aanzien van het postoperatief herstel, werden er geen significante

verschillen gezien bij het analyseren van alleen de gepubliceerde resultaten die geschikt

waren voor meta-analyse. Echter, bij het includeren van alle gepubliceerde data en impu-

teren van ontbrekende parameters, werd een significante reductie gezien in tijd tot eerste

defecatie (−1,96 dagen), tijd tot normaal dieet (−1,3 dagen) en ziekenhuisverblijf (−1,88)

in de laparoscopische groep. Totale incisie lengte was significant korter en patiëntbeoorde-

ling van het cosmetisch aspect was significant beter in de laparoscopie groep. Deze review

laat zien dat de laparoscopische IPAA een veilig procedure is met potentie tot verbetering

van de postoperatieve uitkomsten. Het cosmetische voordeel kan bovendien zwaarwegend

zijn, gezien de voornamelijk jonge patiënten die voor deze operatie in aanmerking komen.

Bij hoog operatierisico patiënten kan een subtotale colectomie met ileorectale anastomose

(IRA) een alternatief bieden voor de IPAA. IRA is een operatie die geen dissectie vereist

in het kleine bekken, met potentieel een verminderd risico op chirurgische complicaties

en seksuele dysfunctie. Om te beoordelen of deze beoogde voordelen ook worden on-

dersteund door de literatuur, werd een review uitgevoerd waarbij de IRA en de IPAA met

elkaar werd vergeleken (Hoofdstuk 6). De totale morbiditeit was lager bij de IRA (8% tot

28%) vergeleken met de IPAA (15% tot 63%). Ernstige complicaties, zonder falen van

het reservoir, waren lager na de IRA. De kans op naadlekkage was tussen de 2,3% en

9,1% na de IRA (gewogen gemiddelde 4,5%), vergeleken met een gemiddelde kans van

7,5% (95% betrouwbaarheidsinterval van 6,1% tot 9,1%) voor de IPAA. Het percentage

seksuele dysfunctie was laag na de IPAA (rond 3%), maar deze complicatie werd niet ge-

rapporteerd na de IRA. De totale mortaliteit was vergelijkbaar tussen de IRA (0 tot 4%) en

de IPAA (0 tot 2,9%). Falen van het reservoir was hoger in de IRA groep, met 5 jaarsfalen

van 7% tot 16%, vergeleken met gemiddeld 4,7% na IPAA. De kans op maligniteit na

een IRA was ook hoger, met een incidentie van 0% tot 8% vergeleken met incidentie van

<0,5% voor de IPAA. Het cumulatieve risico op kanker na de IRA wordt geschat op 6% tot

14% vergeleken met 4,2% voor de IPAA. Functionele uitkomsten lijken beter te zijn na de

IRA, met lagere defecatiefrequentie en minder kans op lichte incontinentie. Kans op ‘urge

incontinentie’ is echter licht hoger bij de IRA vergeleken met de IPAA.

Deze resultaten ondersteunen de keuze voor de IPAA als voorkeursoperatie voor patiënten

met colitis ulcerosa. Het feit dat de IPAA een definitieve behandeling is met laag risico

op kanker is de voornaamste reden hiervoor. Desondanks kan de IRA soms nuttig zijn

in geselecteerde patiënten. Patiënten met een hoog operatierisico en weinig tot geen

210 Chapter 12

rectale betrokkenheid, kunnen voordeel hebben van de minder grote operatie en de goede

functionele resultaten na de IRA. Wel moeten ze goed worden voorgelicht over de kans

op het ontwikkelen van proctitis en kanker, en bereid zijn onder langdurige endoscopische

controles te blijven. Tenslotte, kan de IRA ook geschikt zijn voor patiënten die geen kandi-

daat zijn voor de IPAA of die reeds gemetastaseerde ziekte hebben.

Een ileumreservoir sinus na een IPAA operatie is een ernstige complicatie, die moeilijk te

behandelen is en een hoog risico heeft op pouchfalen. Om de incidentie, behandeling en

uitkomst beter in kaart te brengen, hebben we gekeken naar een grote cohort van 2.286

patiënten die een IPAA hebben ondergaan (Hoofdstuk 7). In totaal waren er 45 (2,0%)

patiënten die een ileumreservoir sinus ontwikkelden. In 23 (51%) patiënten werd de sinus

aanvankelijk conservatief behandeld. Bij de overige patiënten werd de sinus op verschil-

lende manieren behandelend, waaronder transrectale drainage, ‘deroofing’ van de sinus,

primair sluiten van de sinus en aanleggen van een deviërend stoma. Het genezingspercen-

tage na de initiële invasieve behandeling was laag (26%). Een vervolgbehandeling was bij

28 (62%) patiënten nodig. Dit resulteerde uiteindelijke in genezing bij 27 (60%) patiënten

en in falen van het ileumreservoir bij 15 (33%) patiënten. Het bestaan van klachten ten

tijde van presentatie was een belangrijke voorspellende factor voor het falen van behande-

ling met een genezingspercentage van slechts 30%, vergeleken met 84% bij patiënten

zonder klachten (p=0,001). Van de toegepaste behandelingen was observatie de meest

succesvolle initiële behandeling met genezing bij 65%.

Op basis van bevindingen in deze studie, gecombineerd met een review van de bestaande

literatuur, werd een behandelalgoritme opgesteld voor de behandeling van een ileumre-

servoir sinus na een IPAA procedure. Expectatief beleid wordt geadviseerd bij asymptomati-

sche patiënten als initiële behandeling. Bij patiënten met symptomen en patiënten die een

deviërend stoma hebben, kan een invasieve behandeling ingezet worden (i.e. drainage,

‘deroofing’ of primair sluiten). Er is onvoldoende bewijs om te bepalen welke van deze

behandelingen het beste is. De ervaring van de behandelende arts hoort hierbij leidend te

zijn. Het aanleggen van een deviërend stoma kan overwogen worden om de genezingskans

van de sinus te maximaliseren. Bij sinussen waarbij de verwachting op genezing klein is,

op basis van hun grootte en duur van bestaan, dient een redo-IPAA overwogen te worden.

In Hoofdstuk 8 werd het effect van immunosuppressiva op de operatierisico’s bij patiën-

ten met de ziekte van Crohn geëvalueerd. Immunosuppressiva vormen in de hedendaagse

geneeskunde de belangrijkste behandeling voor de ziekte van Crohn. Hun invloed op het

ontstaan van infectieuze complicaties bij een noodzaak tot een operatie blijft echter een

reden van zorg. Om meer inzicht hierin te krijgen, werd een systematische review ver-

richt. Er werden 21 vergelijkende studies met in totaal 6.899 patiënten geïdentificeerd.

Chapter 12 211

12

Postoperatieve resultaten bij gebruik van ‘anti-tumor necrosis factor’ (anti-TNF) (14 studies),

corticosteroïden (14 studies), thiopurines (9 studies) en gemengde immunosuppressiva (6

studies) werden afzonderlijk geanalyseerd.

Een meta-analyse liet een hogere kans zien op postoperatieve infectieuze complicaties bij

patiënten die voor de operatie anti-TNF (RR 1,29, 95% BI 1,07 tot 1,55) of corticosteroïden

(RR 1,55, 95% BI 1,23 tot 1,95) gebruikten. Sensitiviteitsanalyse van studies met hoge

kwaliteit bevestigde deze resultaten. Voor anti-TNF gebruik waren er voldoende data om

afzonderlijke complicaties te analyseren. Er werd een significante toename gezien in het

risico op wondinfecties (RR 1,62, 95% BI 1,12 tot 2,34) en septische shock (RR 1,81, 95%

BI 1,03 tot 3,17) bij gebruik van anti-TNF middelen binnen 3 maanden voor de operatie.

Bij corticosteroïden werd geconstateerd dat patiënten die een hogere dosering innamen,

een extra hoog risico hadden op infectieuze complicaties (RR 1,67, 95% BI 1,31 tot 2,13).

Er werd geen verband gezien tussen het optreden van infectieuze complicaties en gebruik

van thiopurines (RR 1,23, 95% BI 0,66 tot 2,29) of gemengde immunosuppressiva (RR

1,01, 95% BI 0,67 tot 1,54). Concluderend, laat deze review zien dat sommige veel ge-

bruikte immunosuppressiva een verhoogde kans geven op infectieuze complicaties, en dat

dit meegenomen moet worden bij het informeren van patiënten en bij het bepalen van het

optimale tijdstip voor een eventuele operatie.

In Hoofdstuk 9 worden de resultaten van een nieuwe technische variatie voor de rectale

dissectie bij een IPAA procedure beschreven. De dissectie in het kleine bekken wordt door-

gaans uitgevoerd volgens de ‘totale mesorectale excisie’ (TME), waarbij zowel het rectum

als mesorectum en-bloc worden gereseceerd. TME wordt gezien als de gouden standaard

voor oncologische resecties, maar is niet strikt noodzakelijk voor benigne aandoeningen.

Een alternatief hiervoor is het ‘close-rectal’ dissectie, waarbij de dissectie plaatsvindt dicht

op de wand van het rectum. Deze techniek werd voor het eerst in de jaren 70 voorgesteld,

maar werd toen als ‘te veeleisend’ ervaren. Het voordeel van deze techniek is het intact

houden van de mesorectale vetlaag rondom het rectum. Dit is de laag waarin de autonome

zenuwen lopen die o.a. verantwoordelijke zijn voor de blaas- en seksuele functie. De mo-

gelijke voordelen van het ‘close-rectal’ dissectie is het behoud van deze functies en het

minder radicaal uitruimen van het kleine bekken, waardoor er minder ruimte is voor stase

van bloed en vocht en hierdoor minder kans op infectieuze complicaties. Met de huidige

chirurgische instrumenten zoals elektrische en ultrasonische dissectoren is deze techniek

eenvoudiger geworden.

Om de haalbaarheid van deze dissectie te evalueren en om een inschatting te krijgen van

de potentiële voordelen, hebben we een pilot studie uitgevoerd bij 10 patiënten. Deze

patiënten ondergingen allen een IPAA in verband met colitis ulcerosa. De procedure was

212 Chapter 12

technisch gelukt in alle patiënten. Niemand ontwikkelde septische ontstekingen van het

kleine bekken. Na een mediane follow-up van 16 maanden, werden ook geen blaas- of

seksuele dysfunctie klachten waargenomen. De mediane defecatiefrequentie was 6,0 keer

per dag, en anorectale manometrie liet goede functionele resultaten zien. De resultaten

van deze dissectietechniek met de huidige beschikbare chirurgische instrumenten zijn zeer

bemoedigend. Grotere studies zijn nodig om de mogelijke voordelen beter te bestuderen.

aPPendICes

217

A

Appendices

review coMMiTTee

• Prof.dr.R.L.A.W.Bleys,DepartmentofAnatomy,UtrechtUniversity

• Prof.dr.G.J.Groen,DepartmentofAnesthesia,UniversityofGroningen

• Prof.dr.I.Q.Molenaar,DepartmentofSurgery,UtrechtUniversity

• Prof.dr.H.M.Verkooijen,DepartmentofRadiology,UtrechtUniversity

• Prof.dr.F.P.Vleggaar,DepartmentofGastroenterology,UtrechtUniversity

218 Appendices

acknowledGMenT

I would like to express my greatest appreciation to all the patients, their families, co-

authors, local investigators and colleagues that were involved in the studies included in this

thesis. Without you, this would not have been possible.

Prof. dr. van Laarhoven, dear Kees, it is with you that I took my first steps in research and had

the chance to build a solid and rich foundation. Our first article was a Cochrane review; a

challenging endeavor that proved to be highly rewarding. Several Cochrane reviews followed

and those form an important part of both PhD dissertations. You also supported me to fol-

low a course on Evidence-Based Medicine at the University of Oxford, a unique experience

that I cherish. This eventually evolved into completing this prestigious Master’s program, as

you yourself have done before. Thank you for your early recognition and encouragement.

Prof. dr. Kiran, dear Ravi, we only met shortly in Oxford, but there was an unmistakable

connection and appreciation. Your invitation to come to the Cleveland Clinic was an honor,

and working with you on a daily basis was a true reward. I have learned so much from

you about research, medicine and life. The greatest realization of all is that who you work

with is probably the greatest factor for success. Thank you for a wonderful experience, and

thank you for your trust, advice and wholehearted support.

Prof. dr. van Hillegersberg, dear Richard, our work together was brief, but your professional

career is an example for every researcher and physician. The devotion with which you devel-

oped the minimal invasive approach in upper gastrointestinal surgery has transformed the field

regionally and nationally. Thank you for your involvement and facilitation of this endeavor.

Dear review committee, professor Ronald Bleys, professor Gerbrand Groen, professor

Quitus Molenaar, professor Lenny Verkooijen, professor Frank Vleggaar, thank you for

reviewing and assessment of this thesis.

Dr. Keus, dear Erik, you helped me with my first steps in research and showed me how to turn

an idea into a publication. More importantly, you taught me the value of high quality research

and that it does not come easy: for our first Cochrane review we screened over 14,000 hits…

twice!! I really enjoyed our visit to the Cochrane group in Denmark and the discussions we

had there with Gluud and Wetterslev. Thank you for your company and support.

Drs. de Zeeuw, dear Sharonne, partner in the pouch research in Nijmegen. It was a pleasure

to work with you on all kinds of projects. Your contribution to this thesis is substantial. It is

unfortunate that our paths have split somewhere along the road. I can’t believe time went

219

A

Appendices

so fast and that you are now in your final year of residency. I wish you all the luck in the

world with your surgical career and lovely family.

Dr. Heikens, dear Joost, my companion to my first trip to Oxford! Thanks for the unforget-

table moments and many tips about research, surgery and life!

Dr. Fichera, dear Alessandro, it was an honor to work with you. Always upbeat and charm-

ing, but also always present for colleagues and patients. I learned a lot from you about

colorectal surgery and Italy (especially Juventus)! Thanks for having me in the team in

Seattle and for supervising one of the chapters of this thesis.

Dr. Scoglio, dear Danielle, thanks for your work on this thesis and the good (albeit short)

time we had together in Seattle.

Dear Dilara Khoshknabi, team mate in the cubicles of the Cleveland Clinic! I still remember

when we had to re-extract lost data of over 800 patients in less than a week! We were

unbeatable. Thank you for your unparalleled effort and great work ethic. It was a pleasure.

Dear Kathrina Allen, the busy bee of the Cleveland Clinic. Thank you for your help and

assistance in all matters during my stay in Cleveland.

Dear prof. dr. Groen, dear Gerbrand, it was always great to walk into your room at the

UMCU and just be surprised by the unexpected turns our conversations took. Thank you

for all the great stories and life lessons. Also, thank you for your passionate and amusing

way of teaching. I still remember your not so subtle hint for the C7 dermatome.

Dear prof. dr. Elkhammas, dear uncle Elmahdi, few people can even begin to compare

to your kindness, cheerful personality, especially in combination with your skills and ac-

complishments. I really enjoyed my time working with you and the transplant team at Ohio

State. Thank you for you hospitality, boundless support and especially the great BBQs we

had together in Columbus.

To my fellow researchers in the Cleveland Clinic, Awad Jarrar, Galal El Gazzaz and Faisal

Elagili. We had a blissful and memorable time together. Thank you immensely for being a

second family to me during my time there.

Dear Dr. Moayeri, dear Nizar, colleague, friend and paranimf. Thanks for your companion-

ship in computer science, medical school, student activities, research and trading cars. It is

always a pleasure to have you nearby.

220 Appendices

Friends for life, Khalid Annejar, Fatih Arslan, Mohamed Attrach, Mohamed Fathi and

Nicolas Schroten, thank you for your friendship and all the nice moments we had together.

My dear supervisors and fellow residents at the University Medical Center Utrecht and the

St. Antonius Hospital, thank you for being my extended family in the past four years.

Nobody has been more important to me in the pursuit of this thesis than my family:

First and foremost, I would like to thank my parents, whose love, support and guidance

have made me the person I am today. I would like to express my endless gratitude to my

father, Hadi, the person to whom I owe so much, not in the least my drive for knowledge,

perseverance and sense of responsibility. Thank you for putting our wellbeing and educa-

tion first, knowing that it has not always been easy. Circumstances might have increased

the distance between us, but I will never forget the blessed time we had together.

To my mother, the dearest person to my heart and greatest love of my life. You raised us

with unconditional love, boundless caring and infinite giving that are truly unmatched.

There is nothing we can say or do that can express the level of gratitude we owe you. May

God enable us to be the sons and daughters you hope for, and allow us to serve you in

every way that you desire.

My siblings, Iman, Inas, Isra, Inam and Anas, five unique personalities that have had a

profound impact on my life. Life has not always been easy, but we have stood the test time

and time again. Thank you all for being there, and thank you for keeping me honest when

needed! I wish you all the peace and happiness in the world.

My friend, companion and love, Amasi: thank you for your friendship, support and pa-

tience. I am grateful for all the joyful and blissful times together. I am even more grateful

for your limitless support through the demanding tasks of life, and the direly needed

consolation when (on occasion) things really go south. Thank you for bearing with my

flaws and imperfections all those years. Most importantly, thank you for making me feel at

home every day of my life.

My son, Shahien, the newest addition to our family. You might be too young to realize this,

but ‘oh God!’ you are one busy little fella! According to your grandmother the apple did

not fall far from the tree (and as you can see your father did eventually alright, so we have

good hope for you too!). We are truly blessed to have you. You bring an incredible amount

of joy to us: me, your mom, grandmother, aunts and uncle. We love you immensely.

221

A

Appendices

lisT of PublicaTions

1. Van Dalen AHM, Ahmed Ali U, Murray AC, Kiran RP. Optimizing Patient Selection

for Laparoscopic and Open Colorectal Cancer Resections: A National Surgical Quality

Improvement Program–Matched Analysis.

The American Surgeon, February 2019.

2. Schepers NJ, Bakker OJ, Besselink MG, Ahmed Ali U, Bollen TL, Gooszen HG, van Sant-

voort HC, Bruno MJ, for the Dutch Pancreatitis Study Group. Impact of Characteristics of

Organ Failure and Infected Necrosis on Mortality in Necrotizing Pancreatitis.

Gut, June 2018.

3. Ahmed Ali U, ten Hove JR, Reiber BM, van der Sluis PC, Besselink MG. Sample Size of

Surgical Randomized Controlled Trials: a Lack of Improvement over Time.

Journal of Surgical Research. 2018 Aug;228:1-7.

4. Issa Y, van Santvoort HC, van Dieren S, Besselink MG, Boermeester MA, Ahmed Ali

U. Diagnosing Chronic Pancreatitis: Comparison and Evaluation of Different Diagnostic

Tools.

Pancreas. 2017 Oct;46(9):1158-1164.

5. Rijkers AP, Bakker OJ, Ahmed Ali U, Hagenaars JCJP, van Santvoort HC, Besselink MG,

Bollen TL, van Eijck CH; Dutch Pancreatitis Study Group. Risk of Pancreatic Cancer After

a Primary Episode of Acute Pancreatitis.

Pancreas. 2017 Sep;46(8):1018-1022.

6. Ahmed Ali U, Beata RM, ten Hove JR, MD; van der Sluis PC, Gooszen HG, Boermeester

MA, Besselink GH, Journal impact factor and methodological quality of surgical random-

ized controlled trials: a systematic reviewan empirical study.

Langenbecks Arch Surg. 2017 Nov;402(7):1015-1022.

7. Frima H, Hulsmans MH, Houwert RM, Ahmed Ali U, Verleisdonk EJ, Sommer C, van

Heijl M. End cap versus no end cap in intramedullary nailing for displaced midshaft

clavicle fractures: influence on implant-related irritation.

Eur J Trauma Emerg Surg. 2018 Feb;44(1):119-124.

8. van Rosmalen BV, Alldinger I, Cieslak KP, Wennink R, Clarke M, Ahmed Ali U, Besselink

MG. Worldwide trends in volume and quality of published protocols of randomized

controlled trials.

PLoS One. 2017 Mar 15;12(3).

9. Broeders JA, Ahmed Ali U, Fischer G. Systematic review and meta-analysis of random-

ized clinical trials (RCTs) comparing topical calcineurin inhibitors with topical corticoste-

roids for atopic dermatitis: A 15-year experience.

J Am Acad Dermatol. 2016 Aug;75(2):410-419.

222 Appendices

10. Houwert RM, Smeeing DP, Ahmed Ali U, Hietbrink F, Kruyt MC, van der Meijden OA.

Plate fixation or intramedullary fixation for midshaft clavicle fractures: a systematic

review and meta-analysis of randomized controlled trials and observational studies.

J Shoulder Elbow Surg. 2016 Jul;25(7):1195-203.

11. Ahmed Ali U, Issa Y, Hagenaars JC, Bakker OJ, van Goor H, Nieuwenhuijs VB, Bollen

TL, van Ramshorst B, Witteman BJ, Brink MA, Schaapherder AF, Dejong CH, Spanier BW,

Heisterkamp J, van der Harst E, van Eijck CH, Besselink MG, Gooszen HG, van Santvoort

HC, Boermeester MA; Dutch Pancreatitis Study Group. Risk of Recurrent Pancreatitis and

Progression to Chronic Pancreatitis After a First Episode of Acute Pancreatitis.

Clin Gastroenterol Hepatol. 2016 May;14(5):738-46.

12. Broeders JA, Ahmed Ali U, Molyneux AJ, Poncyljusz W, Raymond J, White PM, Steinfort

B. Bioactive versus bare platinum coils for the endovascular treatment of intracranial

aneurysms: systematic review and meta-analysis of randomized clinical trials.

J Neurointerv Surg. 2016 Sep;8(9):898-908.

13. Ahmed Ali U, Pahlplatz JM, Nealon WH, van Goor H, Gooszen HG, Boermeester MA.

Endoscopic or surgical intervention for painful obstructive chronic pancreatitis.

Cochrane Database Syst Rev. 2015 Mar 19;(3):CD007884.

14. Ahmed Ali U, Issa Y, van Goor H, van Eijck CH, Nieuwenhuijs VB, Keulemans Y, Fock-

ens P, Busch OR, Drenth JP, Dejong CH, van Dullemen HM, van Hooft JE, Siersema PD,

Spanier BW, Poley JW, Poen AC, Timmer R, Seerden T, Tan AC, Thijs WJ, Witteman BJ,

Romkens TE, Roeterdink AJ, Gooszen HG, van Santvoort HC, Bruno MJ, Boermeester

MA; Dutch Pancreatitis Study Group. Dutch Chronic Pancreatitis Registry (CARE): design

and rationale of a nationwide prospective evaluation and follow-up.

Pancreatology. 2015 Jan-Feb;15(1):46-52.

15. Hollemans RA, Bollen TL, van Brunschot S, Bakker OJ, Ahmed Ali U, van Goor H, Boer-

meester MA, Gooszen HG, Besselink MG, van Santvoort HC; Dutch Pancreatitis Study

Group. Predicting Success of Catheter Drainage in Infected Necrotizing Pancreatitis.

Ann Surg. 2016 Apr;263(4):787-92.

16. Bakker OJ, van Brunschot S, van Santvoort HC, Besselink MG, Bollen TL, Boermeester

MA, Dejong CH, van Goor H, Bosscha K, Ahmed Ali U, Bouwense S, et al.; Dutch

Pancreatitis Study Group. Early versus on-demand nasoenteric tube feeding in acute

pancreatitis.

N Engl J Med. 2014 Nov 20;371(21):1983-93.

17. Scoglio D, Ahmed Ali U, Fichera A. Surgical treatment of ulcerative colitis: ileorectal vs

ileal pouch-anal anastomosis.

World J Gastroenterol. 2014 Oct 7;20(37):13211-8.

18. Ahmed Ali U, Dunne T, Gurland B, Vogel JD, Kiran RP. Actual versus estimated length

of stay after colorectal surgery - which factors influence a deviation?

Am J Surg. 2014 Oct;208(4):663-9.

223

A

Appendices

19. Ahmed Ali U, Boermeester MA. [Appraising the evidence for new treatments].

Ned Tijdschr Geneeskd. 2014;158:A7281.

20. Ahmed Ali U, Martin ST, Rao AD, Kiran RP. Impact of preoperative immunosuppressive

agents on postoperative outcomes in Crohn’s disease.

Dis Colon Rectum. 2014 May;57(5):663-74.

21. Bakker OJ, van Santvoort H, Besselink MG, Boermeester MA, van Eijck C, Dejong K,

van Goor H, Hofker S, Ahmed Ali U, Gooszen HG, Bollen TL; Dutch Pancreatitis Study

Group. Extrapancreatic necrosis without pancreatic parenchymal necrosis: a separate

entity in necrotising pancreatitis?

Gut. 2013 Oct;62(10):1475-80.

22. Bouwense SA, Ahmed Ali U, ten Broek R, Issa Y, van Eijck CH, Wilder-Smith OH, van

Goor H. Pain outcome after pancreatic surgery for pain of chronic pancreatitis: Relation

to altered central pain processing.

Br J Surg. 2013 Dec;100(13):1797-804

23. Issa Y, Ahmed Ali U, Bouwense SA, van Santvoort HC, van Goor H. Preoperative Opioid

Use And The Outcome Of Thoracoscopic Splanchnicectomy In Chronic Pancreatitis: A

Systematic Review.

Surg Endosc. 2014 Feb;28(2):405-12.

24. Ahmed Ali U, van der Sluis PC, Issa Y, AbouHabaga B, Gooszen HG, Flum DR, Algra

A, Besselink MG. Trends in worldwide volume and methodological quality of surgical

randomized controlled trials.

Ann Surg. 2013 Aug;258(2):199-207.

25. Besselink MG, Ahmed Ali U. Nederland wereldwijd koploper in kwaliteit van chirur-

gische RCTs.

Nederlands Tijdschrift voor Heelkunde. 2013 Jul;22(4):10-1.

26. Ahmed Ali U, Vogel JD. Safety of Surgical Resident Training.

Adv Surg. 2013;47:45-57. (Book chapter)

27. Kiran RP, Ahmed Ali U, Nisar PJ, Khoury W, Gu J, Shen B, Remzi FH, Hammel JP, Lavery

IC, Fazio VW, Goldblum JR. Risk and Location of Cancer in Patients with Preoperative

Colitis-Associated Dysplasia Undergoing Proctocolectomy.

Ann Surg. 2014 Feb;259(2):302-9.

28. Ahmed Ali U, Issa Y, Bruno MJ, van Goor H, van Santvoort H, Busch ORC, Dejong CHC,

Nieuwenhuijs VB, van Eijck CH, van Dullemen HM, Fockens P, Siersema PD, Gouma

DJ, MA Boermeester, et al. Early Surgery versus Optimal Current Step-up Practice for

Chronic Pancreatitis (ESCAPE): design and rationale of a randomized trial.

BMC Gastroenterol. 2013 Mar 18;13(1):49.

224 Appendices

29. Broeders JA, Roks DJ, Ahmed Ali U, Watson DI, Baigrie RJ, Cao Z, Hartmann J, Maddern

GJ. Laparoscopic Anterior 180-Degree Versus Nissen Fundoplication for Gastroesopha-

geal Reflux Disease Systematic Review and Meta-Analysis of Randomized Clinical Trials.

Ann Surg. 2013 May;257(5):850-9.

30. Shapiro R, Ahmed Ali U, Lavery IC, Kiran RP. Endorectal ultrasound does not reliably

identify patients with uT3 rectal cancer who can avoid neoadjuvant chemoradiotherapy.

Int J Colorectal Dis. 2013 Feb; 28(7):993-1000.

31. Ahmed Ali U, Nieuwenhuijs VB, van Eijck CH, Gooszen HG, van Dam RM, Busch OR, Di-

jkgraaf MG, Mauritz FA, Jens S, Mast J, van Goor H, Boermeester MA; Dutch Pancreatitis

Study Group. Clinical outcome in relation to timing of surgery in chronic pancreatitis: a

nomogram to predict pain relief.

Arch Surg. 2012 Oct;147(10):925-32.

32. Kiran RP, Ahmed Ali U, Coffey JC, Vogel JD, Pokala N, Fazio VW. Impact of Resident

Participation in Surgical Operations on Postoperative Outcomes: National Surgical Qual-

ity Improvement Program.

Ann Surg. 2012 Sep;256(3):469-75.

33. Ahmed Ali U, Shen B, Remzi FH, Kiran RP. The Management of Anastomotic Pouch

Sinus After IPAA.

Dis Colon Rectum. 2012 May;55(5):541-8.

34. van Esch AA, Ahmed Ali U, van Goor H, Bruno MJ, Drenth JPH. Wide variation in

diagnostic and therapeutic strategies in chronic pancreatitis: A Dutch national survey.

Journal of the Pancreas. 2012 Jul 10;13(4):394-401.

35. de Zeeuw S, Ahmed Ali U, Donders RA, Hueting WE, Keus F, van Laarhoven CJ. Update

of complications and functional outcome of the ileo-pouch anal anastomosis: overview

of evidence and meta-analysis of 96 observational studies.

Int J Colorectal Dis. 2012 Apr;27(4):553.

36. Ahmed Ali U, Bruno MJ, Issa Y, Gooszen HG, Fockens P, Boermeester MA; Pancreatitis

Werkgroep Nederland. [Better pain management in chronic pancreatitis through early

surgery?].

Ned Tijdschr Geneeskd. 2012;156(5):A4469. Dutch.

37. Ahmed Ali U, Pahlplatz JM, Nealon WH, van Goor H, Gooszen HG, Boermeester MA.

Endoscopic or surgical intervention for painful obstructive chronic pancreatitis.

Cochrane Database Syst Rev. 2012 Jan 18;1:CD007884.

38. de Zeeuw S, Ahmed Ali U, van der Kolk MB, van Laarhoven KC. Ileal pouch anal anas-

tomosis with close rectal dissection using automated vessel sealers forulcerative colitis: a

promising alternative.

Dig Surg. 2011;28(5-6):345-51.

225

A

Appendices

39. van Santvoort HC, Bakker OJ, Bollen TL, Besselink MG, Ahmed Ali U, SchrijverAM,

Boermeester MA, et al; Dutch Pancreatitis Study Group. A conservative and minimally

invasive approach to necrotizing pancreatitis improves outcome.

Gastroenterology. 2011 Oct;141(4):1254-63.

40. Broeders JA, Roks DJ, Ahmed Ali U, Draaisma WA, Smout AJ, Hazebroek EJ. Lapa-

roscopic anterior versus posterior fundoplication for gastroesophageal reflux disease:

systematic review and meta-analysis of randomized clinical trials.

Ann Surg. 2011 Jul;254(1):39-47.

41. Ahmed Ali U, Jens S, Busch OR, Keus F, van Goor H, Gooszen HG, Boermeester MA.

Antioxidants for pain in chronic pancreatitis.

Cochrane Database Syst Rev. 2014 Aug 21;(8):CD008945.

42. Bakker OJ, van Santvoort HC, van Brunschot S, Ahmed Ali U, Besselink MG,Boermeester

MA, et al; Dutch Pancreatitis Study Group. Pancreatitis, very early compared with normal

start of enteral feeding (PYTHON trial): design and rationale of a randomised controlled

multicenter trial.

Trials. 2011 Mar 10;12:73.

43. Broeders JA, Mauritz FA, Ahmed Ali U, Draaisma WA, Ruurda JP, Gooszen HG,Smout AJ,

Broeders IA, Hazebroek EJ. Systematic review and meta-analysis oflaparoscopic Nissen

(posterior total) versus Toupet (posterior partial) fundoplication for gastro-oesophageal

reflux disease.

Br J Surg. 2010 Sep;97(9):1318-30.

44. Besselink MG, van Santvoort HC, Renooij W, de Smet MB, Boermeester MA, FischerK,

Timmerman HM, Ahmed Ali U, Cirkel GA, et al; Dutch Acute Pancreatitis Study Group.

Intestinal barrier dysfunction in a randomized trial of a specific probiotic composition in

acute pancreatitis.

Ann Surg. 2009 Nov;250(5):712-9.

45. Ahmed Ali U, Keus F, Heikens JT, Bemelman WA, Berdah SV, Gooszen HG, vanLaar-

hoven CJ. Open versus laparoscopic (assisted) ileo pouch anal anastomosis for ulcerative

colitis and familial adenomatous polyposis.

Cochrane Database Syst Rev. 2009 Jan 21;(1): CD006267.

46. Keus F, Ahmed Ali U, Noordergraaf GJ, Roukema JA, Gooszen HG, van LaarhovenCJ.

Laparoscopic vs. small incision cholecystectomy: Implications for pulmonary function

and pain. A randomized clinical trial.

Acta Anaesthesiol Scand. 2008 Mar;52(3):363-73.

226 Appendices

collaborator47. van Santvoort HC, Besselink MG, Bakker OJ, Hofker HS, Boermeester MA, Dejong CH,

van Goor H, Schaapherder AF, van Eijck CH, Bollen TL, van Ramshorst B, Nieuwenhuijs

VB, Timmer R, Laméris JS, Kruyt PM, Manusama ER, van der Harst E, van der Schelling

GP, Karsten T, Hesselink EJ, van Laarhoven CJ, Rosman C, Bosscha K, de Wit RJ, Houdijk

AP, van Leeuwen MS, Buskens E, Gooszen HG; Dutch Pancreatitis Study Group. A step-

up approach or open necrosectomy for necrotizing pancreatitis.

N Engl J Med. 2010 Apr 22;362(16):1491-502.

48. van Santvoort HC, Besselink MG, de Vries AC, Boermeester MA, Fischer K, Bollen TL,

Cirkel GA, Schaapherder AF, Nieuwenhuijs VB, van Goor H, Dejong CH, van Eijck CH,

Witteman BJ, Weusten BL, van Laarhoven CJ, Wahab PJ, Tan AC, Schwartz MP, van der

Harst E, Cuesta MA, Siersema PD, Gooszen HG, van Erpecum KJ; Dutch Acute Pancre-

atitis Study Group. Early endoscopic retrograde cholangiopancreatography in predicted

severe acute biliary pancreatitis: a prospective multicenter study.

Ann Surg. 2009 Jul;250(1):68-75.

49. Besselink MG, van Santvoort HC, Buskens E, Boermeester MA, van Goor H, Timmerman

HM, Nieuwenhuijs VB, Bollen TL, van Ramshorst B, Witteman BJ, Rosman C, Ploeg RJ,

Brink MA, Schaapherder AF, Dejong CH, Wahab PJ, van Laarhoven CJ, van der Harst E,

van Eijck CH, Cuesta MA, Akkermans LM, Gooszen HG; Acute Pancreatitis Werkgroep

Nederland. [Probiotic prophylaxis in patients with predicted severe acute pancreatitis: a

randomised, double-blind, placebo-controlled trial].

Ned Tijdschr Geneeskd. 2008 Mar 22;152(12):685-96.

50. Besselink MG, van Santvoort HC, Buskens E, Boermeester MA, van Goor H, Timmerman

HM, Nieuwenhuijs VB, Bollen TL, van Ramshorst B, Witteman BJ, Rosman C, Ploeg RJ,

Brink MA, Schaapherder AF, Dejong CH, Wahab PJ, van Laarhoven CJ, van der Harst

E, van Eijck CH, Cuesta MA, Akkermans LM, Gooszen HG; Dutch Acute Pancreatitis

Study Group. Probiotic prophylaxis in predicted severe acute pancreatitis: a randomised,

double-blind, placebo-controlled trial.

Lancet. 2008 Feb 23;371(9613):651-659.

227

A

Appendices

curriculuM viTae

Usama Ahmed Ali was born on 19 April 1983 in Tripoli,

Libya. Both his parents have their roots in Gherian (غريان) a

small mountain town 80 kilometers to the south of the Libyan

coast. He came with his family to the Netherlands at 9 years

of age to call Utrecht home for most of his subsequent life.

He graduated high-school from College De Klop in 2001, and

enrolled into a double-major in Medicine and Computer Sci-

ence at Utrecht University. He obtained a Bachelor›s degree in

Computer Science in 2006, followed by his Medical Degree in

2007. During his studies he participated in several scientific

projects, most notably on colorectal surgery with dr. Eric Keus and prof. dr. van Laarhoven.

After graduation, he commenced with his PhD program within the Dutch Pancreatitis Study

Group with the goal of initiating research on chronic pancreatitis under the supervision of

prof. dr. Boermeester and prof. dr. Gooszen. This resulted in setting-up and obtaining

grants for two long-term longitudinal projects, namely the CARE-national database and

the ESCAPE-randomized trial. While recruitment for both projects was ongoing, he crossed

the Atlantic to Columbus, Ohio in 2011 in chase of his love and by that time his spouse

Amasi Elbakush. There, he worked as a research fellow at the Department of Colorectal

Surgery of the Cleveland Clinic under supervision of prof. dr. Ravi Kiran. He then completed

his US Medical Licensing Exams and took a brief clinical training program at the University

of Washington in Seattle. In the meantime he read part-time for a Master’s degree on

Evidence-Based Medicine at the University of Oxford, which he obtained in 2012 with

distinction. Eventually he returned to his hometown Utrecht to start his General Surgery

residency, which he commenced in 2014 at the University Medical Center Utrecht. The two

research tracks he has engaged in, i.e. chronic pancreatitis and colorectal surgery, evolved

into the current two PhD dissertations.

Documents

Surgical aspects in treatment of inflammatory bowel disease · Chapter 7 The Management of Anastomotic Pouch Sinus After ileal-pouch-anal anastomosis. Diseases of the Colon & Rectum,