ANZ J. Surg.

2003;

73

: 929–931

ORIGINAL ARTICLE

Original Article

PROXIMAL GASTRIC EXCLUSION FOR UNRESECTABLE GASTRIC CANCER

K

EVIN

D

OLAN

AND

H

ENRY

S

UE

-L

ING

Department of Surgery, The General Infirmary at Leeds, Leeds, United Kingdom

Background:

Patients with unresectable distal gastric cancer causing obstruction have classically undergone palliative gastro-jejunostomy, but high mortality rates and delayed return of gastric emptying have been reported. The aim of the present study was tocompare gastrojejunostomy and proximal gastric exclusion in patients with unresectable distal gastric cancer.

Methods:

Until 1996, patients with unresectable obstructing distal gastric cancer underwent antecolic gastrojejunostomy, butsince 1997 we have performed proximal gastric exclusion for these patients. Mortality, morbidity, time taken to resume oral fluidsand normal diet, length of palliation and survival were compared.

Results:

There was no mortality in either the gastrojejunostomy group (

n

= 4) or the exclusion group (

n

= 6). A single patient inthe gastrojejunostomy group developed a sacral sore and another patient had recurrent vomiting following gastrojejunostomy. Exclu-sion resulted in a quicker return to diet and a slightly longer survival, although these were not statistically significant.

Conclusion:

Proximal gastric exclusion offers a safe, quick and life-enduring palliation for unresectable malignant gastric outletobstruction.

Key words: cancer, exclusion, gastrojejunostomy, stomach, unresectable.

INTRODUCTION

The last quarter of a century has witnessed a significant decreasein the incidence of gastric cancer, particularly of the distal stom-ach.

1

Unfortunately, however, the proportion of patients in thedeveloped world with early gastric cancer has remained small,

2

and up to 75% of all gastric cancers are not resectable.

3

Less thantwo-thirds of those patients who are referred for a surgicalopinion undergo curative resection, with the majority of theremainder undergoing palliative resection or gastrojejunostomy,and a smaller number having laparotomy only or no treatment atall.

4,5

Patients undergoing palliative resection have a significantlylonger survival than patients undergoing gastrojejunostomy,although this may reflect more locally advanced unresectabledisease in patients undergoing gastrojejunostomy.

4–6

Survival ofpatients undergoing gastrojejunostomy for unresectable disease issimilar to survival of patients undergoing laparotomy only.

4,5

Thequality of palliation afforded by gastrojejunostomy was notdescribed in either of these series,

4,5

and the role of gastrojejunos-tomy as palliation for unresectable distal gastric cancer has beenquestioned.

7

A mortality rate of 22% and delayed gastric empty-ing in 21% has been described in patients undergoing gastrojeju-nostomy for advanced cancer of the stomach.

8

A comprehensiveanalysis of gastroenterostomies performed for a variety of pathol-ogies revealed that preoperative malignant obstruction of thestomach was the most significant predictor of postoperativedelayed return of gastric emptying.

9

As a result of these data, our policy for unresectable obstruct-ing cancer of the distal stomach changed from gastrojejunostomyto a modified version of the Devine exclusion procedure. Thispaper compares the results of gastrojejunostomy and gastricexclusion performed at our institution over the last 8 years.

METHODS

Between 1993 and 1996, inclusive patients with unresectableobstructing cancer of the distal stomach underwent antecolic, iso-peristaltic gastrojejunostomy.

Antral exclusion was initially described in 1925 for difficultduodenal ulcers,

10

and performed for unresectable antral cancer in1936.

11

At our institution, we perform a more proximal exclusionthan that previously described by dividing the stomach at thejunction of the proximal one-third and distal two-thirds, over-sewing the distal staple line and performing an antecolic gastro-jejunostomy to the proximal stomach in a Polya reconstruction(Fig. 1). This has been our procedure of choice for unresectableobstructing cancer of the distal stomach since 1997.

Data regarding mortality and morbidity, time to resumption offree fluids and normal diet, hospital stay, clinical outcome andsurvival were retrospectively collected from the hospital casenotes.

RESULTS

Four patients underwent gastrojejunostomy between 1993 and1996, and proximal gastric exclusion has been performed in sixpatients since 1997. They were of similar ages and a singlepatient in each group was female (Table 1). In all patients, cancerwas unresectable due to invasion of the pancreas, and the major-ity of patients in each group also had distant metastases at thetime of surgery (Table 1).

There was no operative mortality. A single patient in thegastrojejunostomy group developed a sacral sore. Patients under-

K. Dolan

MD, FRCS;

H. Sue-Ling

MD, FRCS

Correspondence: Kevin Dolan, Suite 109, Joondalup Health Campus,Shenton Avenue, Joondalup, WA 6027, Australia.Email: medkd@iprimus.com.au

Accepted for publication 14 July 2003.

930 DOLAN AND SUE-LING

going proximal gastric exclusion commenced free fluids and dieton average 2 days earlier than those with a gastrojejunostomy,although this was not statistically significant (Table 1). There wasno difference in hospital stay, although discharge relied on socialas well as clinical factors. A single patient in the gastrojejunos-tomy group developed recurrent vomiting 4 weeks followingsurgery, presumably due to cancer progression, and was treatedconservatively. Patients in the exclusion group had a slightlylonger survival than those with a gastrojejunostomy (Table 1).

DISCUSSION

Although the numbers in our study are small, there is an encour-aging trend for patients undergoing proximal gastric exclusionto commence diet quicker and to live longer than those patientsundergoing gastrojejunostomy. Similar results have beenreported following comparison of gastrojejunostomy and modi-fied Devine exclusion for patients with duodenal obstructionsecondary to pancreatic cancer.

12

It is also reassuring that there

was no mortality or morbidity associated with proximal gastricexclusion in our study, and that none of the patients developedrecurrent vomiting. This would suggest that proximal gastricexclusion has afforded our patients excellent palliation of theirprimary disease before succumbing to the general widespreadeffects of cancer.

The time to recommence diet, the length of hospital stay andsurvival in our study are similar to that of 20 patients withobstructing antral cancer undergoing a more distal gastric exclu-sion in Hong Kong.

10

In the Hong Kong study, a single patienthad persistent vomiting and two further patients developed recur-rent vomiting, perhaps due to the more distal exclusion. Webelieve that proximal exclusion results in a more direct andgravity-assisted route out of the stomach. As in our study, distalgastric exclusion has been shown to prolong survival when com-pared with gastrojejunostomy, and a decreased risk of bleedingfollowing gastric exclusion has also been reported.

11

An alternative to gastric exclusion has been described, namelystomach-partitioning gastrojejunostomy, in which the stomach isnot transected but is partially partitioned by a transverse stapleline from the greater curve that stops short of the lesser curve.

13

The gastrojejunostomy is fashioned to the proximal part ofthe stomach. The authors state that eight patients undergoingstomach-partitioning gastrojejunostomy had better palliation andlonger survival than 13 patients with gastrojejunostomy alone.However, the results of that study need to be viewed with cautionas only one of eight patients in the stomach-partitioning grouphad an antecolic gastrojejunostomy compared with six patients inthe gastrojejunostomy alone group, and patients in the stomach-partitioning group also had chemotherapy. The reason given forperforming stomach-partitioning rather than exclusion is a beliefthat the communication between the proximal and distal parti-tions of the stomach at the lesser curve prevents distension andpossible perforation of the distal stomach as performed in exclu-sion. Perforation of the distal stomach has not been reported withdistal gastric exclusion, and did not occur in our series of proxi-mal gastric exclusion. We believe that stomach-partitioning mayallow filling of the distal partition of the stomach, leading to full-ness, nausea and possible bleeding from the cancer.

There are a number of series of Wallstent insertion for pal-liation of malignant gastric outlet obstruction, but technical fail-ures have been described in up to 25% and failure to improvesymptoms in 5–25%.

14–21

Problems with stent migration andtumour ingrowth have been reported in approximately 15% ofpatients,

19–21

and in an attempt to overcome these problems,coaxial insertion of uncovered and covered expandable metalstents was performed.

22

However, even with two stents 8% ofpatients had no change in their symptoms, 8% suffered stentmigration and 25% had tumour ingrowth.

22

Endoscopic stentingresulted in a shorter hospital stay than open gastrojejunostomy inpatients with pancreatic cancer,

17

although this advantage may belost when compared with laparoscopic gastric exclusion.

23

Pallia-tion afforded by stenting has not yet been compared with gastricexclusion.

Proximal gastric exclusion offers a safe and life-enduring pal-liation of unresectable malignant gastric outlet obstruction, andappears to be an acceptable alternative to standard gastrojejunos-tomy. A larger, randomized trial of proximal gastric exclusionand standard gastrojejunostomy would be required to confirm thesignificance of the encouraging trends in our study, but we do notfeel that this would be ethical in these unfortunate patients withsevere symptoms and a short lifespan.

Fig. 1.

Proximal gastric exclusion for unresectable distal gastriccancer. C, cancer, DS, distal stomach; GJ, gastrojejunostomy; J, jeju-num; PS, proximal stomach.

Table 1.

Comparison of gastrojejunostomy and proximal gastricexclusion for unresectable distal gastric cancer

Gastrojejunostomy (

n

= 4)Proximal gastric exclusion (

n

= 6)

Age (years) 73 (61–78) 73 (66–82)Male : female 3 : 1 5 : 1Distant metastases 3 4Morbidity 1 (sacral sore) 0Mortality 0 0Free fluids (day) 7 (4–8) 5 (4–6)Normal diet (day) 9 (8–11) 7 (6–8)Hospital stay (days) 11 (9–13) 11 (9–17)Vomiting 1 0Survival (months) 3 (1–8) 4 (1–8)

PROXIMAL GASTRIC EXCLUSION 931

REFERENCES

1. Dolan K, Sutton R, Walker SJ, Morris AI, Campbell F,Williams EMI. New classification of oesophageal and gastriccarcinomas derived from changing patterns in epidemiology.

Br. J. Cancer

1999;

80

: 834–42.2. Sue-Ling HM, Johnston D, Martin IG

et al.

Gastric cancer:a curable disease in Britain.

BMJ

1993;

307

: 591–6.3. Fielding JW, Roginski C, Ellis DJ

et al.

Clinicopathologicalstaging of gastric cancer.

Br. J. Surg.

1984;

71

: 677–80.4. Bozzeti F, Bonfanti G, Audisio RA

et al.

Prognosis of patientsafter palliative surgical procedures for carcinoma of the stomach.

Surg. Gynecol. Obstet.

1987;

164

: 151–4.5. Doglietti GB, Pacelli F, Caprino P, Sgadari A, Crucitti F. Sur-

gery: independent prognostic factor in curable and far advancedgastric cancer.

World J. Surg.

2000;

24

: 459–63.6. Hanazaki K, Sodeyama H, Mochizuli Y

et al.

Palliative gas-trectomy for advanced cancer.

Hepatogastroenterology

2001;

48

: 285–90.7. Kikuchi S, Tsutsumi O, Kobayashi N

et al.

Does gastrojejunos-tomy for unresectable cancer of the gastric antrum offer satisfac-tory palliation?

Hepatogastroenterology

1999;

46

: 584–7.8. Lo NN, Kee SG, Nambiar R. Palliative gastrojejunostomy for

advanced cancer of the stomach.

Ann. Acad. Med. Singapore

1991;

20

: 356–8.9. Woods SDS, Mitchell GJ. Delayed return of gastric emptying

after gastroenterostomies.

Br. J. Surg.

1989;

76

: 145–8.10. Kwok SPY, Chung SCS, Griffin SM, Li AKC. Devine exclusion

for unresectable carcinoma of the stomach.

Br. J. Surg.

1991;

78

:684–5.

11. Nakata Y, Kimura K, Tomioka N, Sato M, Watanabe Y,Kawachi K. Gastric exclusion for unresectable gastric cancer.

Hepatogastroenterology

1999;

46

: 2654–7.12. Kato K, Morita T, Miyasaka Y, Kondo S, Katoh H. Modified

Devine exclusion for unresectable pancreatic head carcinoma.

Hepatogastroenterology

2001;

48

: 569–71.13. Kaminishi M, Yamaguchi H, Shimizu N

et al.

Stomach-partitioning

gastrojejunostomy for unresectable gastric carcinoma.

Arch. Surg.

1997;

132

: 184–7.14. Soetikno RM, Lichtenstein DR, Vandervoort J

et al.

Palliation ofmalignant gastric outlet obstruction using an endoscopicallyplaced Wallstent.

Gastrointest. Endosc.

1998;

47

: 267–70.15. Park KB, Do YS, Kang WK

et al.

Malignant obstruction ofgastric outlet and duodenum: palliation with flexible coveredmetallic stents.

Radiology

2001;

219

: 679–83.16. Wai CT, Ho KY, Yeoh KG, Lim SG. Palliation of malignant

gastric outlet obstruction caused by gastric cancer with self-expandable metal stents.

Surg. Laparosc. Percutan. Tech.

2001;

11

: 161–4.17. Wong YT, Brams DM, Munson L

et al.

Gastric outlet obstruc-tion secondary to pancreatic cancer: surgical vs endoscopic palli-ation.

Surg. Endosc.

2002;

16

: 310–12.18. Espinel J, Vivas S, Munoz F, Jorquera F, Olcoz JL. Palliative

treatment of malignant obstruction of the gastric outlet using anendoscopically placed enteral Wallstent.

Dig. Dis. Sci.

2001;

46

:2322–4.

19. Aviv RI, Shyamalan G, Khan FH

et al.

Use of stents in the treat-ment of malignant gastric outlet obstruction.

Clin. Radiol.

2002;

57

: 587–92.20. Kaw M, Singh S, Gagneja H, Azad P. Role of self-expandable

metal stents in the palliation of malignant duodenal obstruction.

Surg. Endosc.

2003;

17

: 646–50.21. Kim JH, Yoo BM, Lee KJ

et al.

Self-expanding coil stent with along delivery system for palliation of unresectable malignantgastric outlet obstruction: a prospective study.

Endoscopy

2001;

33

: 838–42.22. Jung GS, Song HY, Seo TS

et al.

Malignant gastric outletobstruction: treatment by means of coaxial placement of uncov-ered and covered expandable nitinol stents.

J. Vasc. Interv.Radiol.

2002;

13

: 275–83.23. Ammori BJ, Boreham B. Laparoscopic Devine exclusion gastro-

enterostomy for the palliation of unresectable and obstructinggastric carcinoma.

Surg. Laparosc. Percutan. Tech.

2002;

12

:353–5.