Short bowel syndrome : Surgical therapy

Many surgical solutions to short bowel syndrome have been proposed; however, none has proven robe uniformly successful. Some of these solutions, combined with optimal medical management, may represent the patient's only hope for survival without parenteral nutrition. Most forms of surgical therapy are supportive and aim at controll ing three basic pathophysiological defects: decreased intestinal transit time, gastric hypersecretion, and reduced functional mucosa! surface area. Conservative resection and, thus, prevention of short bowel syndrome remains the best form of treatment at present. In the fu ture, small bowel transplantation may prove to be an important advance in therapy; however, this remains largely experimental due to continued problems with rejection. Can J Gastroenterol 1990;4(4): 167-173

tilm nf a few extra centimetre~ nf bowel can have a profound influence on the intestine's a bi li t y to adapt, and thus on the eventual prognosis.Many aurhors have advocated construct ion of multiple ostomies with a second look procedure for all margina l bowel which may l:ie vial:ile (l-3).
Surgical treatment can l,e c lassified into three ca tegories based on pathophysio logic processes: slowi ng transit time; decreasing gasrric secretions; nnd increasing the effective mucosa!surface area (4).

SLOWING INTESTINAL
TRANSIT Surgical therapy to decrease intestinal transit time is based on the construc tion of a partial small bowel obst ruction .Many different methods have been devised; however, a ll are unpredictable in their eventual effe ct.Most of the complications which arise from each merhod are due co bowel obstruction.Antiperistaltic segments: Reversal of intestinal segments to create anciperistaltic regions was originally described by Mall in 1896 (5).He reversed variable segments of small bowel in th ree dogs, two of which d ied of bowel obstruction, while the third lived fo r th ree months.At laparotomy two months later, the segment showed rete nti o n o f reversed peris t alsis.T h rough the use of glass beajs as markers, Mall a lso demonstrated a delay 111 intesti na l transit time.The procedure was viewed as impossible until the 1950s when Hammer (6) demon traced that the d uodenum could he reversed successfull y resulting in an increase in intestinal transit t ime (6).Later in 1955, Hammer demonstrated t hat an 80% small howel resection in the dog could be tolerated with the addition of a reversed ilea! segment as short as 2 inches (7).In a controll ed fo llow-up exreriment, he demonstrated th.at dogs could survive a 90% resection for up to two years with rhe reversal of a small segment of intestine.
The first human report of egmental reversal was by Gibson (8) in a female with short bowel syndrome from a mesenteric infarction.She recovered fo llowing the reversal of a 7.5 cm segment of jejunum.Subsequently, Srahlgren (9) demonstrated that the fecal output of fat and, to a lesser degree, nitrogen, decreased with the insertion of an antiperistaltic segment in the jeju num of dogs.Us ing 3 co 4 cm paired proximal and distal reversed segments, Keller ( 10) proposed that the mechanism by which intestinal reversal worked was a prolongation of intestina l transit with an increase 111 the contact time of mucosaand luminalcontents.Venables m 1966 ( 11) in the first human study of absorpt ion showed a 50% decrease in fecal fat absorption with the insertion of a reversed segment.
T here are two c rit ica 1 factors in c reating antiperistaltic segments: the length of the segment and its location.If the segment is too short there will be no benefit, whereas a long segment will cause a bowel obstruction.Fink and O lson ( 12) favoured a 10 cm segment located proximally to retard gastric emptying; however, the weight of evidence is now fo r distal placement.Wilmo re ( 13) fai led in his attempt to treat ileostomy diarrhea over the lo ng term with a distal 10 cm segment; however, there was a small short term decrease in ileosromy vo lume.Pertsemlidis ( I 4) a lso felt that the segment should be placed as distally as possible, and used a 14 cm segment 12 cm proximal to the ostomy to prevent stagnant loop syndrome and bacterial overgrowth.Warden in 1978 ( 15) reported five babies with a 3 cm reversal placed as distally as possible, with survival in fo ur.

168
In summary, intestinal reversal to create an antiperiscal tic region anJ delay transit bone of the more common surgical interventions used in the treatment of ,hort howcl syndrome.The ideal length in the adu lt seems to be 10 to l 5 cm with placement as distally as possihle.Nevertheless, inconsistent results and difficulty in predicting which patients are likely to respond have consistently dampened enthusiasm for this arproach.Recirculating loops: The recirculating loop is an extension of intestinal reversal and was first suggesteJ by Stafford et a l in 1959 (16).Mackby in 1965 (I 7) showed an improved surv ival in dogs when a recirculating loop was combined with a separate a ntiperistaltic segment.Aleman ( 18) in a controlled experimen t o n 20 dogs found that animals with recirculating loops had better fat absorption hut lost more weight, and often died due to anorexia compared with controls.In a separate study comparing recirculating loops co segmental reversals, the latter were more often technically successful and had fewe r complications (19).Although a further report of successful clinical application of the recirculating loop was submitted in 1975, the procedure has fa llen into general disfavour due to the strong evidence against its being superior to a simple reversal and to the fact that the procedure is complicated, uses long lengths of bowel and is fraught with complicatio ns causing a high mortality race (20).Colon interposition: Colon interposition has also been used in the treatment of short bowel syndrome.Use of the colon has three main advantages: a n intrinsically slow peristalsis; decreased likelihood of causing obstruction due to the fact that colon can be placed isoperistaltically; and no requirement for the use of small bowel which is a lready compromised in residual length .Initia lly, antipe ristaltic colonic segments were used and shown to be of benefit in a dog model (21).In 1967, Trinkle anJ Bryant (22) placed a 5 cm ant iperistaltic segmen t of t ransverse colo n in a three-week-old infant after massive resection for a mid gut vo lvulus.Although the patient died, she initially gained weight and haJ an 111crease m transit time with a Jecrease in stool frequency.Later, isoperistalric colon interposition was investigated hy H.utcher (23,24 ).In a series of experi• ments with beagle puppies, he showed increased survival with the interposition of a 15 cm segment of isoperistaluc colon both pre-ilea!and pre-jejuna! after a 90% small bowel resection.The animals in both cases attained approximately 70% of their expected growth with a decrease in mortality and morbid ity compared with controls.Garcia et al (25) interposed a 24 cm segment of isopcristaltic colon to a L 5 cm length of small bowel with an imacc ileocecal va lve in a huma n patient.After an aJaptive period, intesti nal transit time increased from 10 to 105 mins am! the patient was able Lo maintain adequate nutri t ion by oral feedi ng.In this patient, the development of a D-lactic acidcmia was postulated to be Jue to bacterial overgrowth.Comparison of iso-and antiperistal tic colonic interposition was examined by LloyJ (26,27) using a 90% small howel resection in the rat.In a series of experiments, he found t hat antiperisraltic colonic interposi tion effectively prolonged transit time but had other un predictable effects.Some animals acqu ired a small bowel obstruction and had, on average, lower body weights with on ly a slight increase in rhe absorption of albumin and no change in t he absorption of fat compared to controls.H e concluded that there was no advantage to ant1peristaltic interposition over isoperista lcic a nd , indeed, there may be disadvantages.Further studies in dogs by Carner (28) showed that the insertion of a 20 cm length of anti peristaltic colon made no d iffere nce in xylose absorption, 24 h fat excretion , bowel transit time or ave rage weight loss.Although his nu mbers were small, he concluded t hat antiperistalt ic colonic interposition was of no benefit.
G lick et al in 1984 (29) published a series on six infants in which a proximal isoperistaltic segmen t of colon was used to t reat short bowel syndro me.The segments were 11 to 15 cm in length and three of the infants survived whi le the ochers d ied of sepsis and total parenteral nutrition-inducet.lliver failure.His conclusions were that surviva l was associated with greater residual small bowel length, colon interposition at a younger age and a shorter duration of medical management.Unfortunately, the average residual bowel length in the wrvivors was 86 cm and these patients may well have adapted on their own with nme.Brolin (30) puhlished a case report of a 34-year-old female with 12 cm of residual le ngth of sma ll bowel in whom an isoperistaltic colonic interposition was done.Postoperatively, she did well in the shon te rm; however, no nutritional studies were done.
lsoperisca I tic colonic interposition is a viable a lternat ive in the treatment of short bowel syndrome.Proximal placement seems co be best; however, the optimal length of the segment has yet to be detennined.The range in the literature is from 8 co 24 cm.A 11 of these lengths are reported to prolong intestinal transit t ime.Whether or not the segment has the abi li ty to absorb nutrients has nor heen established.The actual mechanism of improved absorption is presumed to be through prolonged intestinal transit time a nd greater mucosa!contact time with luminal contents.There appear to be no short term difficulties with the procedure.In the long term, bacterial overgrowth due to stasis with resultant n-lactic acidcmia and possible encephalopathy may prove to be a problem, especiall y in the infant population.Experimental valves: The recognition that the presence of an ileocecal valve has a positive effect on the residual intestine's ability to adapt to resection has led co attempts to create an 'artificial' ileocecal va lve.Experimental va lves have been looked at as another means of slowing intestinal transit t ime.Intestinal valves of many different types have been used, including ablation of the outer lo ngit udinal muscle layer (3 l ,32), the reverse incussusception valve (33 ), and several ocher va lves of similar construction (34)(35)(36)(37)(38).
The valves have all increased transit time, but the effects have not always been predictable.Schille r (31) attempted to show that the o uter muscle ablating sphincter was better than intestinal reversal in promoting survival after a 90% resection; however, his numbers were very small.Waddell (33) presented three patien ts with an incussusception valve: one patient having marked success, one satisfactory, and the third eventually acqu iring a bowel ohstruction such that the va lve had co be raken down.Vinograd (36) published a study of the submucosally tunnelled valve constructed much like the re-implantation of a ureter into the hlaJder.He found the optimal length to be 4 to 6 cm in a dog model in prolonging intestinal transit time wirh no demonstrable reflux.Chardavoyne et a l (39) examined t he efficacy of a surgically created nipple valve in a dog hy introducing labelled hacteria below the valve.The occurrence ofhacteriaahove the valve was no different than with an intact ileocecal valve.
le is nm surprising that clinical series examining imescinal valves are very few in number.The unpredictability, the occurrence of bowel obstruction and th e loss of further in testinal length to construct the valves arc all major deterrents co their routine use.Retrograde luminal pacing: It has heen known for.ome time that the propagation of peristalsis distally in the in iestine is ch rough electrical impulses from a pacemaker in the duodenum.Phillips et a l (40), in a series of experiments in dogs, demonstrated that retrograde electrical pacing of intact jejunum enhances absorpt ion of glucose, water and sodium.When the bowel was transsected co e liminate t he proximal pacemaker, retrograde pacing had a profound effect on absorption such that it was greater than the absorption from in tact jejunum with or without retrograde pacing.It appears that retrograde pacing slowed transit time and in some cases reversed the flow of luminal contents (41,42).Layzell (43) later demonstrated that dogs with retrograde luminal pacing showed an increase in body weight and a decrease in fecal fat and nitrogen excretion.
Investigation in to the mechanism of action has shown chat the effect of pacing can be inhibited by an a lphaadrenergic blockade, and thus the effect Short bowel syndrome: Surgical therapy must be mediated, at least in part, hy ;m adrenergic mechanism ( 44 ).
This therapy has remained cxpenmenta 1, since entrainment requires transsection of the duodenum to eliminate the intrinsic pacemaker, prolonged function of the electrodes has been difficult, and the electrodes must be implanted su rgically and removed su rgically when they foil.

DECREASING GASTRIC HYPERSECRETION
As mentioned previously, gastric hypersecretion has been recognized as a component of the short bowel syndrome for some time.This was felt to be detrimemal for a number of reasons: there would be an increase in the volume of intestinal sec retions presented to an already compromised absorptive system; t he acidification of the lumen of the intestine would hinder the action of the digestive enzymes; and the resultant osmotic and volume load would further compound the fluid, electrolyte and nutrient losses al ready occurring in the compromised bowel.
The realization chat patients with prev inus vagotom ies and gastreccom ies were more tolerant of massive resection led to the use of these procedures fo r the control of gastric hypersecret ion in short bowel syndrome.Today, with the emergence of the H2 receptor antagon ists and the realization that gastric hypersecretion is probably tempora ry, there is no place fo r the surgical treatment of gastric hypersecretion.

INCREASING THE EFFECTIVE MUCOSAL SURFACE AREA
Attempts at increasing the effective mucosa!surface area has led to three completely different approaches: t he growth of neomucosa; bowel lengthening and capering procedures; and small bowel transplantation.Neomucosa: Neomucosa was originally investigated by Cywes in 1968 ( 45) and Binningcon in 1974 (46).T hey observed chat the ~erosa of neighbouring bowel could be used as a bed co stimulate the growth of a mucosa!covering.In a series of experiments on dogs and pigs, they found chat if the small bowel was opened along its an timesenteric border anJ sutured lu the ,em~,t ol n neighhnurmg r1ece nf hnwd ma longt• 1ud111al intestinal paiLh, the 111terven mg sero"t would be co,TreJ with a mucosa wh1<.:h was rrnven to c<1nra 1n emyme le,•eb ~,mdar to nl)tmal mucosa ( 46) 56) puhl1shed a report of another infant with short bowel syndrome who had an mtest mal lengthen111g procedure done.Post• operatively, ~he remained nn total parenteral nutrition for four weeks hut was then ,tble t0 progress to a regular J1et.He moJ1fied the procedure to some extent l:,y re-anascomosing the lengthened bowel 111 a helical forma• t1on to rrevent traction on the mesentcnc vessels.The problems en• countereJ postoperatively 111cluded bacterial overgrowth rmximal ro the lengthened loop, nutrit10nal 111• tolerance and gastmesophagcal reflux.Bacterial overgrowth was felt ro be due to roor pcriscalsb 111 the proximal dilated duodenum which was nm tapered.Ora l tolerance of feed mg was complete by eight months, although continuous tube feed suprlemencat1on still main• tained a large rroportlon of the caloric intake.U rcc ho lme was used to combat the reflux with complete success.
Thompson ct al 111 1985 ( 5 7) were unsucLessful 111 their attemr1 to appl\ the 81anch 1 prtJLedurl' 10 a child with the short bowel syndrome.After tb 1J ing the bowel they rL••anastomn,eJ 1t, on ly to have one segment of the lengthened bowel become nnnv1ahle lntercmngly, the pallent ,1pparcntly 1mrmveJ clinically and the ,1uthor~ at• trihuteJ this to the result.inttapering of the rcmammg segment nf l:iowel.
A direct comparison pf 1he efficacy of mtestinal lcngthenmg relative to other surgical treatments of ,hon howd syndrome has been rerformed in a pig model by Stgalct et .11(unruhli,hed data).They compan:d isorerntaltic colon 1nterrosn 1on and 1nteslln~l lengthening.While hoch group, showed superior weight ga111 to control arnmals rece1v111g no treatment, mtestinal lengthening was found co he superior to 1sopern,taluc colon inter• prn,1tioning.
Intestinal lengthen m g arpears 10 he a v iable alternative m the surgical ireat• mcnt of short bowel sy ndrome.Ap pl1camm of this procedure seems to h~ tailored to the subset of the population in which the residual bowel has dilated so markedly that peristalsis 1s relative!}ineffective (58).The advantages m elude the face that nn mucosa!surface area need be sacrificed; a slowing of transit ume is obtained wilhout cau,mg a funct 1onal ol:istruction; and normal peristalsis is restored to ;1 previou:,ly Ji. lated segment.Small bowel transplantation: Succe~ful small bowel transplantarnm may he the ultimate treatment for short bowel syndrome.The tech111cal fcas1bil1ty was first escal:ilished expernnentally in the late 1960s by Ull1he1 111 a Jog m()(.lel ( 59).Hts procedure cons isted of transplaming the entire small intestine with vascular anastom<):,is between the respective mesenteric vessels of the graft and hnst.I !is autografts survived indefinitel y, but the a llografo all rejected in a maner of days.At present, the maior obstacle is ~ti II rejection, as the large amount of lymrho1J us,ut' present 111 the transplanted organ make, it extremely tmmunoge111c (60,61) Control of rejection can be approache<l by either general 1mmuno:,uppress1on of the host or by alreramm of the 1m munogen1c1ty of the donor organ.Numerous methods of temporary control llf reiection have hccn accomplished 111 the dog moJcl using co1wcnc1nn,1l hmt 1mmunosuppress1vc I hcrapy.Unfortunately, b()th a:athiupnnc and predn1 wne interfere w11 h muc.:w,al cell replication and thus the function ofthl• graft (60).Anulymphocytl' scrum alone has hcen shown to be 1m11fficien1 111 averting rCJCCtl(m tn Jng~ (62).66) who Jemomcrated that dogs with incont1nuny grafts given oral cyclosponne survived fnr less than one month (66).This emphasized the fact 1hac absorpuve func.:tinn of an nllografr i\ mitially suhoptimal, and 1mmunosuppressi ve therapy should he av compl1sheJ parcnterally.Furtht.'r,tudies by Ricour (67) and Crane (68) m pigs confirm these findings .The aJ-Jmon of predmsone Lo cyclnspnnne therapy has been shown in dogs to improve survival over c.:yclospnnne alone (69); however, these results have nnt heen comistendy reproduced.High Jose cyclrn,porme therapy may not he w1thouc its own problem ; a rcvers1hlc impairment of 1ncescinal absorption with a protein-losing cmeropathy has been described 1n dogs (70).Experimental altcmpts at passive immune enhancement (antiJonor antibody) have been entirely un~uccessful 111 prolonging survival (61).

CONCLUSIONS
Management of the short bowel syndrome contmues co be a difficult clmical problem.Over the long term, the clm1uan depends on mtrin~ic adaptation of the residual in1estinc and this, 1n turn, is dependent on length, type ,ind functional state of the residual hnwel alnng wtth the presence or ah-,encc of an ileocecal va lve and colon.The mecha111sms of adapt,lt1on arc n11t ent1rcly under~too<l, thus prevcnung active 1ntcrvenllonal therapy to accelerate adapcamm.As the moM cnmmon cause of .,honhnll'el syndrome ts su rgical resection, the best therapy is prevention using conservative resections anJ 'seconJ look' proceJures in the case of howcl of quc~ti(mable viab ility.Pre~ent therapy, whe ther medical or surgical, is purely supportive ACKNOWLEDGEMENTS: Dr Fcdorak 1s the rccipientd a C linical lnvestigarorsh1p from the Alberrn Heritage Foundation for Medical Research.Th is work 1, supported by a grant from the Alberta Heritage Foundarion for Medical Research and the Departmcnt of Surgery, University of Alhena.The ,1urhon, express appreciatinn ro Mrs Michel PollMJ for her expert scc reranal a,sim111ce.