Gut, 1983, 24, 392-398

Functional and structural studies of ileal reservoirs usedfor continent urostomy and ileostomyB M PHILIPSON, N G KOCK, R JAGENBURG, c AHREN, L NORLEN,J W L ROBINSON, AND H MENGE

From the Departments ofSurgery II, Clinical Chemistry, Pathology and Urology, Sahlgrenska sjukhuset,GBteborg, Sweden; Chirurgie Experimentale, CHUV, Lausanne, Switzerland, and Klinikum Steglitz, FreieUniversitat, Berlin, Federal Republic ofGermany

SUMMARY The structure and function of the mucosa has been studied in continent ileostomy andurostomy reservoirs, the latter being a receptacle for diverted urine constructed out of ilealtissue. Morphometric evaluation was performed by a microdissection technique and functionalstudies involved the assessment of L-phenylafanine absorption by the whole pouch in vivo andthe uptake of the same amino acid by biopsy samples in vitro. Endoscopic examination revealedfairly homogeneous villous structure in the ileostomy reservoirs. In the urostomies, there was a

gradual appearance, as a function of the postoperative time interval, of areas of flat mucosa

intermingled with villous regions. When possible, biopsies of the two types of mucosa were

studied separately. Even in the villous regions of the urostomies, the size of the villi was smallerin all dimensions than the villi of ileostomy samples. The uptake of phenylalanine in vitro,however, was the same in villous samples from urostomies and in ileostomy biopsies. Theabsorption of phenylalanine in vivo by the whole pouch was lower in the urostomies. There was a

good correlation between amino acid absorption in vivo and in vitro in the ileostomies, but not inthe urostomies. This is attributed principally to the variable proportions of avillar and villousmucosa in the latter material, as there is a clear reduction in absorption in vivo in the longestestablished urostomy pouches.

In 1967 the continent ileostomy was introduced inclinical practice as an alternative to the conventionalileostomy.1 The intra-abdominal ileal reservoirconstructed with this technique has also been usedrecently to achieve continence in patients requiringurinary diversion.2 3 The mucosa of ileostomyreservoirs underwent structural changes soon afterconstruction, but in long term follow-up, a tendencytowards normalisation was noted.4 The localabsorption of L-phenylalanine in ileostomyreservoirs has been investigated both under clinicaland experimental conditions, the results indicatingfinite, albeit somewhat decreased, absorptive

4 6capacity.Morphological studies in 'urostomy' reservoirs

have revealed signs of progressive villous atrophyand flattening of the epithelial cells in the mucosa

Address for correspondence: B M Philipson. Kirurgiska kliniken II,Sahlgrenska sjukhuset, S-413 45 Goteborg, Sweden.

Received for publication 6 August 1982

after prolonged exposure to urine.7 8 As urinarydiversion via an ileal conduit may lead to metabolicdisturbances in patients with reduced renalfunction,9 10 it seemed of interest to study theabsorptive capacity of the reservoir mucosa whichhad been constantly exposed to urine. For thisreason we have examined the local absorption ofL-phenylalanine in the reservoirs used for collectionof urine and compared the results with thoseobtained in reservoirs accumulating ileal effluent. Inaddition, biopsy samples were taken from bothseries of reservoirs for morphological and functionalinvestigations in vitro.

Methods

PATIENTSTen patients, seven women and three men, withurinary diversion via a continent ileal reservoir werestudied. The mean age was 408 years (range 21-65years) and the mean postoperative observation time

392

on 4 April 2019 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.24.5.392 on 1 M

ay 1983. Dow

nloaded from

Continent ileostomies and urostomies

25 months (range five months to six years). Theindications for the construction of these continenturostomies are shown in Table I. In seven patients,this was a secondary operation because ofmalfunction of an earlier procedure, whereas inthree patients, the continent urostomy was aprimary diversion.

Eight patients with continent ileostomies werealso studied after proctocolectomy for ulcerativecolitis. In this group, there were four women andfour men: the mean age was 41*0 years (range 27-63years), and mean postoperative observation time 42months (range six months to six years). All patientsgave their informed consent to the experiments.The investigations were carried out on two

consecutive days. On both days the patients hadcontrolled food and fluid intake before the investi-gation, with avoidance of coffee and tea.

DAY I ABSORPTION OF L-PHENYLALANINE INVIVO

Immediately before the study, the reservoir wasrinsed with 0-9% isotonic saline and emptied. Twohundred and fifty millilitres of a test solutioncontaining L-phenylalanine 20 mmol/l and 51Cr-EDTA 5 uCi/l (non-absorbable marker) dissolved ina Ringer solution pH7.4 containing 30 mmol/lglucose was instilled into the reservoir. Afterinstillation, aliquots were taken at two, 10, 20, 30,40, 50, and 60 minutes, whereafter the reservoir wasemptied. Before the samples were taken, thecontents were mixed by repeated aspiration andinjection of the test solution. Phenylalanine wasdetermined by ion-exchange chromatography(Beckman Amino Acid Analyzer model 120 C) and1Cr activity in a conventional y-counter. Thephenylalanine concentration in the reservoir wascorrected for the corresponding 51Cr-EDTA activityand the rate of absorption was calculated by linearregression analysis of the relationship betweenconcentration and time. Net water movementsacross the pouch mucosa cannot be establisheddirectly from the changes in 51Cr-EDTA activity inview of the fact that fluid is continually entering the

Table 1 Indications for urinary diversion in 1Opatients

Number ofDiagnosis patients

Urinary incontinence 3Bladder extrophy 2Neurogenic bladder 2Interstitial cystitis 1Carcinoma of urinary bladder 1Carcinoma of uterine cervix 1

pouch throughout the experimental period, from theileum or from the upper urinary tract.

DAY 2 EXCISION OF BIOPSY SAMPLES FOR STUDYIN VITRO

After the emptying of the reservoirs, endoscopy wasperformed with a cystoscope. The macroscopicappearance of the mucosa was registered anddirected biopsies were taken for morphometricstudies and uptake in vitro.

MORPHOLOGICAL STUDIESBiopsy samples were fixed in Bouins solution. Afterdehydration and orientation, they were embeddedin plastic, sectioned at 1.5 ,um and stained withhaematoxylin-eosin. The number of mitoses in agiven sample magnified 40x was counted and theaverage of 10 such measurements was taken as anestimate of the mitotic activity. Other samples werefixed in ethanol/glacial acetic acid (3:1, v/v) for 24hours, then rinsed with and stored in 75% ethanol.These samples were evaluated morphometrically bymicrodissection.11 12 They were first placed underthe dissection microscope with the villous surfacefacing downwards, flattened and photographed toobtain the serosal area. Next, they were stained withthe Feulgen reagent, reorientated so that the villiwere facing upwards and again photographed so thatthe number of villi per unit serosal surface could beestimated. Microdissection was then performedunder the stereomicroscope to liberate individualvilli and crypts. From each sample, 10 random villiwere chosen for the determination of their height,width, and breadth at base and apex; from thesevalues the surface area of the individual villus can beassessed. 12 Finally, the lengths of 10 random cryptswere recorded.

DETERMINATION OF L-PHENYLALANINEACCUMULATION IN VITROThree parallel biopsy samples from each patientwere incubated for one hour at 37°C in Krebsbicarbonate buffer containing 1 mmol/l L-(3H)-phenylalanine and a tracer amount of 14COOH-inulin for the determination of the extracellularspace of the specimen. After the incubation, thesamples were rinsed, blotted, weighed and dissolvedindividually in 1 ml SolueneO-350 by warming to55°C. After the dissolution, the samples wereacidified with 100 ,u1 33% acetic acid and added to 10ml Instagel® for counting in a liquid scintillationspectrometer under conditions chosen to discrimi-nate between the two isotopes. Aliquots of theincubation medium were counted under identicalconditions. Under the assumption that the tissuecontained 80% water, the results for each sample

393

on 4 April 2019 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.24.5.392 on 1 M

ay 1983. Dow

nloaded from

Philipson, Kock, Jagenburg, Ahren, Norlen, Robinson, and Menge

were expressed as a distribution ratio between intra-and extracellular water, by correcting the totaluptake of phenylalanine for the extracellularcompartment, as determined with inulin.13

Results

Endoscopic examination of ileostomy reservoirsrevealed mucosa covering the entire surface of thepouch. The shape of the villi varied from finger-liketo club-like, with occasional leaves. In the urostomyreservoirs, there was considerable variation in theendoscopic appearance, dependent on the post-operative interval. In those of recent construction,almost unchanged ileal mucosa persisted. At laterstages, greater variation was noted, with the appear-ance of villous ridges. In three patients with themost long established pouches frankly avillarregions were observed, where underlying vesselswere easily discernible. Amongst these areas, therewere islands of mucosa with regular intestinalappearance. Figure 1 illustrates these two extremes.Microscopic examination of biopsy samples of

ileostomy reservoirs confirmed earlier observa-tions.4 The villi were stunted and the crypts werelengthened. In one patient, a pronounced inflamma-tion was noted in the lamina propria, accompaniedby considerable shortening of the villi. The hetero-

geneity of the mucosa from samples of urostomyreservoirs previously observed8 was confirmed onmicroscopic examination. Biopsies from intestinalareas showed shortened villi and lengthened cryptswith a normal surface epithelium. The laminapropria and the submucosa disclosed a chronicoedema. The number of inflammatory cells waswithin normal limits. Biopsies from avillar mucosashowed almost total loss of crypts and a proprianlayer covered by a single row of low cubic epi-thelium. No signs of fibrosis, metaplasia or dysplasiawere ever encountered. Two samples from the samelong established urostomy pouch are contrasted inFig. 2: one portion of the mucosa reveals islands ofintestinal appearance with villi and lengthenedcrypts, whereas the other portion shows a smooth,avillar surface covered with a single row of low cubicepithelium.Morphometric evaluation of samples from

ileostomy reservoirs and of biopsies specificallytaken from areas of intestinal appearance inurostomy reservoirs is presented in Table 2. In twopatients, microdissection was not possible becauseof the absence of villi; this does not mean that theendoscopic samples were taken from avillar regions,as long crypts were encountered in this material.The morphometric data clearly show that the villousheight and breadth, as well as the villous surface

Fig. 1 Endoscopic appearance ofurostomy reservoirs. Picture on left shows surface of recently constructed reservoir withfinger-like villi. That on the right representsfive-year-old reservoir with islands of 'intestinal-like' mucosa surrounded byavillar regions.

394

on 4 April 2019 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.24.5.392 on 1 M

ay 1983. Dow

nloaded from

Continent ileostomies and urostomies

Fig. 2 Microphoto oftwo areas ofa urostomy reservoirfive years after construction. On the left, section from'intestinal-like' island with cylindrical epithelium, whereas on right avillar surface covered with flat epithelial layer is shown.

area and the ratio of the mucosal to serosal surface,were significantly smaller in the urostomyspecimens. The number of villi per unit area wassignificantly larger and the crypt length was smallerin the urostomies.

In addition, the villous height and crypt lengthwere assessed on haematoxylin-eosin stainedspecimens. There was generally close agreementbetween the values obtained by the two techniques.Where such agreement did not occur, we tend toascribe the differences to the heterogeneity of theoriginal samples. We also determined the mitoticactivity in the crypts in this material, and found1*8±0.32 mitoses per section in the ileostomyreservoirs and 3O0±094 mitosis in the urostomies(means ± SEM), difference not significant.For the evaluation of the uptake of L-

phenylalanine in vitro, the biopsy samples fromurostomy reservoirs having intestinal or avillar

appearances were examined separately. The resultsin Table 3 show that the samples of intestinalappearance transport this amino acid to the sameextent as biopsy samples from ileostomy reservoirs.On the other hand, the three avillar specimens werehardly able to accumulate phenylalanine above adistribution ratio of unity. In addition, the extra-cellular space of the avillar samples was muchgreater than that of the other specimens.The absorption of phenylalanine in vivo was

significantly larger in the ileostomy reservoirs thanin the urostomies (43-8±9-17 vs 16.6±4.72 ,umol/min; t=2-63, p<0-02). In the ileostomy patients, agood correlation was observed between the uptakeof phenylalanine in vitro, expressed as a distributionratio, and the absorption of the same amino acid invivo (Fig. 3). A similar correlation was not obtainedfor the urostomy patients.

In the case of the ileostomy reservoirs, there was

Table 2 Microdissection data concerning ileostomy and urostomy mucosae

Villousbreadth (pm) Ratio Crypt

mucosal:serosal lengthheight (pm) width (,um) apex base permm2 surface (mm2) surface (pm)

Ileostomy 419±12-1 145±5.0 286±12.6 325±21-3 11-9±0-69 0-307±0.0221 3-66±0-361 391±28-3Urostomy 287±39-4 148±5-3 157±7-4 184±7-8 15-6±1-46 0-115±0-0186 1.90±0.377 258±34-7P <0-01 NS <0-001 <0-001 <0-05 <0-001 <0-005 <0-01

Results are means ± SEM of eight samples in each group. Two additional samples have been excluded from the urostomy group, as theypossessed no villi for dissection; they did, however, possess crypts and so they did not represent samples of the smooth mucosa discussed inthe text and illustrated in Fig. 2. Probability of difference between populations evaluated by t test (NS = not significant at p<O0.5).

395

on 4 April 2019 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.24.5.392 on 1 M

ay 1983. Dow

nloaded from

396 Philipson, Kock, Jagenburg, Ahre'n, Norlen, Robinson, and Menge

Table 3 L-phenylalanine accumulation by biopsies in vitro

Phenylalanine uptakeExtracellular space (%) p.mol/lO0 mgfresh tissue Distribution ratio

1. Ileostomies(n=8) 23±4.2 0-57±0-098 8-2±1-372. Urostomies-villous samples (N=10) 36±2.7 . 0.59±0.104 11.1±2.163. Urostomies- avillar samples (N=3) 63±1-9 0.11±0.013 2-2±0-49tl_2 2.61* 0-12 1-11t2-3 8.17* 4.50t 4.00t

Results are means ± SEM of number of samples stated. Statistical evaluation of difference by ttest. *p<0.05; tp<001; *p<0001.

no obvious change in the absorption of phenyl-alanine in vivo with the length of the postoperativeinterval (Fig. 4). The one particularly low value inthis group (marked with an asterisk in the figure)concerns the patient with an inflamed mucosa towhom we have referred above. In contrast, therewas a marked trend in the urostomy patientstowards reduced phenylalanine absorption as afunction of time. In our experience, regions ofavillar mucosa first appeared after 20 months'observation time. For that reason, we have con-sidered patients with reservoirs established for morethan 20 months separately. In all five urostomyreservoirs of that age, the absorption of phenyl-alanine was less than 16 ,mol/min, whereas in all

ileostomies with the exception of the patient withinflammation of the mucosa, the correspondingabsorption was greater than 19 ,umol/min.

Discussion

In ileostomy reservoirs, even on a long term basis,the villous structure of the mucosa is maintainedwith relatively discrete modifications, despitealterations in the intraluminal environment. Indeed,according to our previous studies, a reduction invillous height and an increase in crypt length,accompanied by an increase in cell turnover,develop during the early postoperative period,whereas these parameters tend to normalise at later

80.A

A8A

A

r = 0 81

10 20 30 40 50 60 70 80Absorption in vivo(,umol min )

Fig. 3 Correlation between uptake ofL-phenylalanine invitro, expressed as distribution ratio between intra- andextracellular concentrations, and absorption ofsame aminoacid in vivo in continent ileostomy reservoirs.

60-

40 -

20

.EC:

c 80-

20

40-

20 -

A

AA

A

A

in,A

AA~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

*

0

0o

0~~~~~~~~~~n .

0 20 40

Months after operution60 80

Fig. 4 Absorption ofL-phenylalanine in vivo in ileostomy(upperpanel) and urostomy (lowerpanel) reservoirs as

function ofpostoperative time interval. Asterisk refers topatient with pronounced inflammation oflamina propriaand reduced villous height. Vertical dotted line refers topostoperative time interval after which heterogeneity ofmucosal structure ofurostomy was encountered atendoscopy.

14 -

12 -

10-

2 8-8

6-

4-

2-1

1 1 5

on 4 April 2019 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.24.5.392 on 1 M

ay 1983. Dow

nloaded from

Continent ileostomies and urostomies 397

stages.5 In contrast, there are progressive alterationsin the mucosal structure of the urostomy reservoirs.Although systematic long term studies are not yetavailable, owing to the novelty of the technique, itappears that the normal ileal mucosa is graduallytransformed into a structure lacking villi and crypts,but resembling a lamina propria covered by a cubicor flat epithelium.8 Nevertheless, in all cases studiedup to the present time, some areas of mucosa withintestinal features persist. The structure of these'intestinal islands', with their lengthened crypts andshortened villi, and occasional flat mucosa withcoeliac-like characteristics,8 provides some indica-tion as to how this mucosal transformation hasdeveloped. Changes of this nature only occur whenthe noxious influence is present in the lumen.14 Wetherefore consider the changes that we haveobserved to be an adaptive phenomenon associatedwith the long term exposure to urine.The fact that even the smooth areas of the

urostomy mucosa which no longer exhibit intestinalcharacteristics are covered with a monocellularepithelial layer requires some comment. In earlierstudies we have noted the occasional occurrence ofareas of denuded mucosa7 8 which has beeninterpreted in terms of intermittent erosion of thesurface epithelium. It is known that the healing ofchemically produced ulcers occurs by theprogression of a unicellular layer from survivingcrypts;15 we believe that in the present instance, theflat or cubic epithelium that covers the laminapropria of the smooth areas of the urostomypouches originates in a similar manner.The biopsy samples from urostomy reservoirs

with villous structure exhibited an unexpectedlygood capacity for phenylalanine uptake in vitro.Indeed, the distribution ratio established by thesespecimens was as high as that of the ileostomybiopsies. Hitherto no data are available, to ourknowledge, on the uptake of amino acids by normalhuman ileal biopsy samples, so it is not knownwhether these distribution ratios correspond to thenorm. On the other hand, when avillar samples wereincubated in the same fashion, the uptake ofphenylalanine was found to be much smaller, withthe establishment of distribution ratios that are onlyslightly greater than unity.Whereas there was a good correlation between

the absorption of phenylalanine in vivo and in vitroin the ileostomy reservoirs, a similar correlation didnot exist in the urostomy patients. Two factors maybe responsible for this observation: first, the propor-tions of smooth and 'intestinal-like' mucosa in theurostomy pouches was variable. Whereas the uptakeof phenylalanine in vitro was assessed separately onthe two types of tissue - and differed considerably in

these - the absorption in vivo was necessarilydetermined for the whole pouch. This incidentallyexplains why the absorption of phenylalanine in vivotends to fall with an increasing postoperativeinterval, as the proportion of smooth mucosaappears to increase with time. No such functionalimpairment was observed in vitro in the 'intestinal-like' biopsies. Secondly, the absorption of relativelyhigh concentrations of amino acids in vivo isbelieved to occur simultaneously by transcellularand paracellular pathways.16 In the urostomyreservoirs into which hypertonic urine is continuallyflowing, there is likely to exist a permanent para-cellular bulk flow of water from the blood to thelumen via the paracellular shunts. This would tendto hinder any paracellular absorption of phenyl-alanine, and could contribute to the low values forthis process registered in the urostomy pouches.

In the previous publication we reported kidneyfunction data for patients with long term urostomyreservoirs and found no negative influence of thepouch per se on any parameter.3 The present studyshows that urine has no acute harmful effect on theintestinal mucosa used for the construction of thepouch. But in the long term, we have observed agradual transformation of the ileal mucosa into aless differentiated structure which absorbs lessamino acids. It is possible that this structure mayalso absorb less of the harmful components of urineand actually protect the patient from metabolicdisturbances.

This study was supported by grants from theSwedish Medical Research Council (project No.B80-17X-00577-16C) and the Swedish ILCOAssociation. We are grateful to Harriet Andersson,Kristina Bostrom, Barbro Ekmark and ElisabetLindholm (Goteborg), Sylvianne Henriot(Lausanne), and Ursula Feldman (Berlin) for skilfultechnical assistance.

References

1 Kock NG. Intra-abdominal 'reservoir' in patients withpermanent ileostomy. Arch Surg 1969; 99: 223-31.

2 Kock NG, Nilson AE, Norl6n L, Sundin T, Trasti H.Urinary diversion via a continent ileum reservoir.Clinical experience. Scand J Urol Nephrol 1978; suppl49: 23-31.

3 Kock NG, Nilson AE, Nilsson LO, Norlen UJ,Philipson BM. Urinary diversion via a continent ilealreservoir. Clinical results in 12 patients. J Urol 1982;28: 469-75.

on 4 April 2019 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.24.5.392 on 1 M

ay 1983. Dow

nloaded from

398 Philipson, Kock, Jagenburg, Ahren, Norlen, Robinson, and Menge

4 Philipson B, Brandberg A, Jagenburg R, Kock NG,Lager I, Ahr6n C. Mucosal morphology, bacteriologyand absorption in intra-abdominal ileostomy reservoir.Scand J Gastroenterol 1975; 10: 145-53.

5 Nilsson LO, Kock NG, Lindgren I, Myrvold HE,Philipson BM, Ahr6n C. Morphological and histo-chemical changes in the mucosa of the continentileostomy reservoir 6-10 years after its construction.Scand J Gastroenterol 1980; 15: 737-47.

6 Philipson B, Kock NG, Robinson JWL, Menge H,Mirkovitch V. Function and structure of continentileostomy reservoirs in dogs. Gut 1975; 16: 132-6.

7 Hansson H-A, Kock NG, Norlen L, Philipson B, TrastiH, Ahr6n C. Morphological observations in pedicledileal grafts used for construction of continent reservoirsfor urine. Scand J Urol Nephrol 1978; suppl 49: 49-61.

8 Philipson BM, Nilsson LO, Norlen L, Kock NG,Ahr6n C. Mucosal adaption in ileum after long timeexposure to urine: a study in patients with continenturostomy. In: Robinson JWL, Dowling RH, RieckenEO, eds. Mechanisms of intestinal adaption. Lancaster:MTP Press, 1982: 613-20.

9 Creevy CD. Renal complications after ileal diversion ofthe urine in non-neoplastic disorders. J Urol 1960; 83:394-7.

10 Giesy JD, Hodges CV. Flaccid paralysis associated

with hyperchloremic acidosis and hypokalemiafollowing ileal loop urinary diversion. J Urol 1965; 94:243-6.

11 Clarke RM. Mucosal architecture and epithelial cellproduction rate in the small intestine of the albino rat. JAnat 1970; 107: 519-29.

12 Lorenz-Meyer H, Kohn R, Riecken EO. Vergleichverschiedener morphometrischer Methoden zurErfassung der Schleimhautoberflache des Ratten-dunndarms und deren Beziehung zur Funktion. Histo-chemistry 1976; 49: 123-9.

13 Robinson JWL, Mirkovitch V. The recovery offunction and microcirculation in small intestinal loopsfollowing ischaemia. Gut 1972; 13: 784-9.

14 Menge H, Robinson JWL, Riecken EO. Anpassungs-moglichkeiten der Dunndarmschleimhaut anverschiedene intraluminale Milieuveranderungen. ZGastroenterol 1976; 14: 420-33.

15 Stamm B, Mirkovitch V, Winistorfer B, RobinsonJWL, Ozzello L. Regeneration and functional recoveryof canine intestinal mucosa following injury caused byformalin. Virchows Arch [Cell Patholl 1974; 17: 137-48.

16 Robinson JWL, Antonioli JA. Is paracellularmovement of importance in the intestinal absorption oforganic solutes? Gastroenterol Clin Biol 1980; 4: 78-86.

on 4 April 2019 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.24.5.392 on 1 M

ay 1983. Dow

nloaded from