Morbidity and Mortality of Short-Bowel Syndrome Acquired in Infancy: An Update

By Arthur Cooper, Thomas F. Floyd, Ar thur J. Ross, III, Harry C. Bishop,

John M. Templeton, Jr, and Mor i tz M. Ziegler

�9 The advent of to ta l parentera l nutr i t ion (TPN) has made survival beyond infancy possible for large numbers of pat ients who have sustained massive small intest inal loss due to a var ie ty of int raabdomi- nal catastrophes. However , the quant i ty and qual i ty of life have been l imited by the deve lopment of late sequalae due both to the pro t rac ted use of TPN and the long-term compl icat ions of foreshor ten ing of the gut. To determine to wha t ex ten t the morb id i ty and morta l i ty of shor t -bowel syndrome (SBS) may have improved over the last 10 years, we rev iewed our exper ience since 1973 w i th pat ients losing more than 50% of to ta l small intest inal mass in infancy. The et iologies of SBS in the 16 s tudy pat ients were necrot iz ing enterocol i t is (6), midgut volvulus (5), mult ip le atresias (3), gastroschisis (1), and congeni- tal SBS (1). Overall survival was 81%; tota l small intest inal length (SIL) at the t ime of diagnosis was 44.2 • 7.9 cm in survivors and 30.3 • 7.8 cm in nonsurvivors, probabi l i ty values not signif icant. Al though no pat ient survived w i thou t an i leocecal valve whose to ta l SlL was greater than 20 cm, the three deaths in this series were not related di rect ly to the SlL, but to end-stage l iver disease result ing f rom TPN-associated cholestasis. Among the survi- vors, adaptat ion to enteral feedings required 13.8 • 2,5 me, during which t ime weaning f rom TPN occurred; we igh t at adaptat ion was 6.87 • 1.32 kg. A mean fo l low up of 3.3 • 0.8 years permi t ted assessment of qual i ty of life; the mean score was 3,0 • 0.3 on an arb i t rary scale (excel lent = 4, good = 3, fair = 2, poor = 1), whi le the mean number of stools per diem was 3.1 • 0.7. Finally, near-normal somatic g rowth was achieved in survivors weaned f rom TPN in that mean we igh t (W), height (HI, and weight - for -he ight (W/H) all approached or exceeded the 25th percent i le (W = 26th percent i le, H = 24th percent i le, W / H = 41st percenti le). We conclude that the long- term out look for most pat ients w i th SBS result ing f rom massive small intest inal loss in infancy is good, al though TPN-associated cholestasis may lead to end stage l iver disease and death in a minor i ty of pat ients.

INDEX WORDS: Short bowel syndrome; tota l paren- teral nutr i t ion.

M ORE THAN A DECADE AGO, Wil- more concluded that survival was

unlikely for those infants who required extensive intestinal resections if the remaining small intes- tinal segment measured less than 40 cm. 1 Since that time, the advent of total parenteral nutrition

(TPN) has made survival beyond infancy possi- ble for large numbers of patients who have sustained massive small intestinal loss due to a variety of intraabdominal catastrophes. How- ever, the quantity and quality of life for such patients have been limited by the development of late sequelae, due both to the protracted use of TPN, and the long term complications of fore- shortening of the gut. Therefore, to determine to what extent the morbidity and mortality of the short-bowel syndrome (SBS) may have improved over the last 10 years, during which time the use of TPN has become routine, we reviewed our experience since 1973 with patients who under- went massive intestinal resection in infancy.

MATERIALS AND METHODS

Review of the hospital charts and office records of all infants who survived the loss of more than 50% of their total intestinal mass, exclusive of the duodenum and colon, between 1973 and 1983 at The Children's Hospital of Phila- delphia provided the basis for the study; this was supple- mented by personal interview with the patient and family in every case. Age, sex, diagnosis, gestational age, age at operations, weights at birth and operations, length of time required for complete adaptation to enteral feedings, and weight at the time of adaptation were obtained from the hospital chart; the type of operation and measured length of remaining small bowel, exclusive of the duodenum, which was normal in all cases, were determined from the operative note. TPN-associated cholestasis was defined by the presence of direct and total hyperbilirubinemia without another cause that resolved with cessation of TPN, or was documented by liver biopsy. The length of the remaining intestinal segment was measured by placing a single strand of 3-0 surgical silk at the time of the initial operation along the unstretched antime-

From the Department of Surgery, Children's Hospital of Philadelphia, and the Harrison Department of Surgical Research, University of Pennsylvania School of Medicine, Philadelphia.

Presented before the 15th Annual Meeting of the Ameri- can Pediatric Surgical Association, Marco Beach, Florida, May 9-12, 1984.

Address reprint requests to Arthur Cooper, MD, Depart- ment of Surgery, The Children's Hospital of Philadelphia, One Children's Center, 34th Street and Civic Center Boule- vard, Philadelphia, PA 19104.

�9 1984 by Grune & Stratton, Inc. 0022-3468/84/1906~)016503.00/0

Journal of Pediatric Surgery, Vol 19, No 6 (December), 1984 711

712 COOPER ET AL

senteric border of the intestine, then measuring the length of the silk strand against a ruler. Current weight (W), length or height (H), weight-for-length or height (W/H), and total serum protein and albumin levels at the time of follow up were obtained from the office record; number of stools per day, type of diet and medications, and quality of life were determined at the time of personal interview. Weight was measured with a beam balance scale, length with a length board, and height with a stadiometer; quality of life was determined by the parents on an arbitrary scale (excellent = 4, good = 3, fair = 2, poor = 1). Finally, length of follow up was determined for each child, and somatic growth parame- ters were plotted on the standard charts developed by the National Center for Health Statistics. Data were then tabu- lated and, where appropriate, analyzed statistically using the Student's t test.

Nutritional management during the period of adaptation relied heavily upon the use of TPN. During the perioperative period, balanced TPN (10 to 12.5 g/kg/d dextrose, 1.5 to 2.0 g/kg/d crystalline amino acids, 1.0 to 3.0 g/kg/d fat emul- sion) sufficient to provide at least 60 kcal/kg/d was adminis- tered by peripheral vein. However, once the bloodstream had been sterilized, as proven by serial blood cultures, long term central venous access was obtained, usually within 7 to 10 days of operation, by means of a tunnelled, cuffed Silastic catheter; balanced TPN (20 to 25 g/kg/d dextrose, 2.0 to 2.5 g/kg/d crystalline amino acids, 1.0 to 3.0 g/kg/d fat emul- sion) sufficient to provide at least 100 kcal/kg/d was then administered by central vein. Once gastrointestinal motility was restored, enteral refeeding was slowly begun using dilute elemental formula (Pregestimil; Mead-Johnson, Evansville, IN) and advanced as tolerated. Continuous feedings, either by nasogastric tube or gastrostomy tube, were given if bolus feedings were poorly absorbed. Cholestyramine, loperamide, and rarely, oral antibiotics, were utilized as appropriate if diarrhea halted the progression of feedings; sodium bicarbon- ate was added to the formula if hyperchloremic metabolic acidosis persisted despite, or because of, the addition of cholestyramine. In selected cases, most often those with massive ileal resection, a formula rich in medium-chain triglycerides (Portagen; Mead-Johnson, Evansville, IN) was

substituted for the more elemental formula. Weight, fluid intake and output, and stool pH, sugar, and water content were strictly monitored throughout the entire course of hospitalization; blood counts, blood chemistries, and liver function tests were also obtained at appropriate intervals, especially during administration of TPN. Steady weight gain on bolus feedings was the major criterion for discharge, although selected patients were managed at home either with continuous enteral or cycled parenteral feedings, or both.

R E S U L T S

S i x t e e n pa t ien ts r e q u i r i n g mass ive resec t ion of

m o r e t h a n 50% of to ta l in tes t ina l mass in i n fancy

were i nc luded in the s tudy (Tab le s 1, 2, and 3).

T h e e t io logies of S B S were nec ro t i z ing en te roco-

litis (6) , m i d g u t vo lvulus wi th g a n g r e n e (5),

m u l t i p l e a t res ias (3), gas t roschis i s (1), and con-

gen i t a l S B S due to vo lvulus (1). T h e overa l l

surv iva l r a t e was 81% ( 1 3 / 1 6 ) . S u r v i v a l was 86%

( 6 / 7 ) in pa t ien ts hav ing m o r e t h a n 40 c m of

sma l l bowel r ema in ing , 100% ( 4 / 4 ) in pa t i en t s

hav ing 15 to 40 cm of sma l l bowel r e m a i n i n g and

an i n t ac t i leocecal va lve , and 60% ( 3 / 5 ) in

pa t i en t s hav ing 15 to 40 c m of sma l l bowel

r e m a i n i n g bu t no i l eoceca l valve; indeed , 50%

( 1 / 2 ) o f pa t ien ts whose to ta l S I L was less t han

20 c m surv ived wi thou t an i leoceca l valve. S t r ik-

ingly, t he th ree dea ths in this series d id no t seem

to be r e l a t ed d i rec t ly to m a l n u t r i t i o n induced

e i the r by the dec reased S I L or the absence o f the

i l eoceca l va lve (Fig 1), bu t r a t h e r to end s tage

l iver d i sease resul t ing f r o m T P N - a s s o c i a t e d cho-

lestasis . I t is in te res t ing to note tha t al l dea ths

o c c u r r e d in infants w h o had r e q u i r e d bowel

resec t ions for nec ro t i z ing enterocol i t i s , a condi-

Table 1. Pat ient History and Ex tent of Resect ion Following Mass ive Small Intestinal Resect ion in Infancy

Residual Small Bowel Case Gestational Birth Weight at Ileocecal No. Sex Diagnosis Age Weight Operation Total/Jejunum/Ileum Valve

1 F NEC 31 wk 1.37 kg 1.36 kg 105 cm 85 cm 20 cm 0

2 F NEC 30 wk 0.92 kg 0.90 kg 100 cm "intact" 0 0

3 F gastrochisis 38 wk 2.39 kg 2.39 kg 56 cm 46 crn 10 crn 0

4 F multiple atresias 35 wk 2.35 kg 2.35 kg 50 cm 40 cm 10 cm 0

5 M volvulus term ? 4.10 kg 45 cm 30 cm 15 cm +

6 F diaphragmatic hernia,

NEC 36 wk 2.30 kg 2.30 kg 44 cm 40 cm 4 cm +

7 F congenital SBS due to in

utero volvulus 35 wk 1.70 kg 1.89 kg 42 cm ? ? +

8 F volvulus term ? 4.30 kg 37 cm 21 crn 14 cm +

9 F volvulus term (adopted) 4.00 kg 31 cm 37 cm 4 cm +

10 F NEC 37 wk 2.10 kg 2.10 kg 30cm 3Ocm 0 0

11 F ileal atresia 35 wk 3.01 kg 3.01 kg 29 cm 29 cm 0 0

12 M NEC 31 wk 1.44 kg ? 24 cm 24 cm 0 0

13 F volvulus 37 wk 2.60 kg 2.60 kg 23 cm 16 cm 8 cm +

14 F ileal atresia with

perforation 31 wk 1.30 kg 1.30 kg 19 cm 15 cm 4 cm 0

15 F gastrosehisis term ? ? 17 cm 17 cm 0 0

16 M volvulus term 3.27 kg 5.60 kg 15 cm 10 cm 5 crn +

S H O R T - B O W E L S Y N D R O M E 7 1 3

T a b l e 2 . E x t e n t o f R e s e c t i o n a n d D e t a i l s o f A d a p t a t i o n F o l l o w i n g M a s s i v e S m a l l I n t e s t i n a l R e s e c t i o n i n I n f a n c y

Residual Colon Case Length of Time Required Weight at No. Ascending/Transverse/Descending Survival Follow Up for Adaptation Adaptation Diet Medications

1 + + + alive 3.75 yr 8 mo 2.9 kg table food vitamins

2 0 " l h " + alive 3.67 yr 18 mo 7.5 kg table food (Pregestimil cholestyramine, Polycitra,

supplement) vitamins

3 + + + alive 4,08 yr 26 mo 9.8 kg table food none

4 + "1/2 . . . . 1/z" alive 1.16 yr not yet adapted ~4 strength Presgestimil cholestyramine, Ioperamide.

NaHCO 3 5 + + + alive 6.00 yr ? table food none

6 § + + died 1 .O6 yr did not adapt - - Pregestimit

7 + + + alive 2 .00 yr 15 mo 6.2 kg table food none

8 + + + alive 9.55 yr 71~ mo 4.9 kg table food none

9 + + + alive 0 .50 yr not yet adapted - - 90% strength Pregestimil cholestyramine, Ioperamide,

vitamins

10 O 0 0 diecl 1.75 yr did not adapt - - NPO

11 + + + alive 5.42 yr 10 mo 3.9 kg table food (lactose free) cholestyramine, vitamins

12 0 "1t2"' + alive 0 .50 yr not yet adapted - - ~4 strength Pregestimi] vitamins

13 Jr Jr Jr alive 1.33 yr not yet adapted - - full strength Portagen cholestyramlee, Ioperamide

14 Jr + + alive 0.75 yr not yet adapted - - NPO none

15 O 0 Jr died 1.17 yr did not adapt Pregestimil cimetidine, leperamide,

vitamins

16 Jr + Jr alive 5.17 yr 12 mo 12.9 kg table food (Osmolite cholestyramine, Polycitra,

supplement) NaHCO3, vitamins, ZnSO 4

tion known to predispose to the development of TPN-associated cholestasis. However, it should also be noted that among the survivors, three of whom also had necrotizing enterocolitis, the inci- dence of TPN-associated cholestasis was 92% (12/13) . Fortunately, in two thirds of these patients (8 /12) , cholestasis was mild, and resolved following cessation of TPN, but in one

third (4 /12) , it persisted to the time of follow-up examination.

Total small intestinal length (SIL) at the time of diagnosis (mean _+ SEM) was 44.2 _+ 7.9 cm (range, 15 to 105 cm) in survivors and 30.3 _+ 7.8 cm (range, 17 to 44 cm) in nonsurvivors, proba- bility values not significant. Among the survi- vors, complete adaptation to enteral feedings

T a b l e 3 . E n d R e s u l t F o l l o w i n g M a s s i v e S m a l l I n t e s t i n a l R e s e c t i o n i n I n f a n c y

Weight for Case Height Total No. of Stools Quality of TPN Associated No, Weight Height (percentile) Protein Albumin Per Diem Life* Cholestasis

1 t6 .6 kg 103 cm 5 0 - 7 0 6.3 g/dL 4.0 g/dL 2, formed 4 mild and resolved

(50 -75 percentile) (50 -75 percentile)

2 13.2 kg 98 cm 10-25 7.4 g/dL 4.6 g/dL 2 -3 , loose 4 chronic

( 10 -25 percentile) ( 25 -50 percentile)

9 14.8 kg 95 cm 5 0 - 7 5 6.9 g/dL 4,0 g/dL 1-2, formed 4 chronic

( 10 -25 percentile) ( 5 -10 percentile)

4 5,83 kg 66 cm <5 5.8 g/dL 3.6 g/dL (ileostomy) 2 mild and resolved (<5 percentile) (<5 percentile)

5 19.1 kg 102 cm 9 0 - 9 5 - - 1, formed 4 mild and resolved

( 2 5 - 5 0 percentile) (<5 percentile)

6 10.2 kg 74 cm 5 6.6 g/dL 3.4 g/dL - - - - end stage liver disease

7 7.60 kg 72 cm 5 0 - 7 4 6.2 g/dL 3.7 g/dL 3 -4 , formed 4 none

(<5 percentile) (<5 percentile)

8 28.6 kg 132 cm <5 6.5 g/dL - - 1 -2 , formed 4 mild and resolved ( 2 5 - 5 0 percentile) ( 2 5 - 5 0 percentile~

9 5.49 kg 65 cm 5 5.9 g/dL 3.9 g/dL (continuous) 2 chronic

(<5 p~centile) ( 2 5 - 5 0 percentile)

10 7_65 72.5 cm 72.5 6.1 g/dL 2.2 g/~L - - - - end stage liver disease

11 16.3 kg 99 cm 7 5 - 9 0 6.4 g/dL 3.8 g/dL 2, formed 4 mild and resolved ( 10 -25 peccentile} ( <5 percentile)

12 4.43 kg 55 cm 50 - - 4.9 g/dL 6, formed 2 chronic

(<5 percentile) (<5 percentile)

13 11.8 kg 76 cm >95 6,2 g/dL 4.1 g/dL 3 -4 , loose 1 mild and resolved ( 7 5 - 9 0 percentile) ( 10 -25 percentile)

14 7.O1 kg 63 cm 7 5 - 9 0 5.1 g/dL 3.3 g/dL (fistula) 1 chronic (<5 percentile) (<5 percentile)

15 5.6 kg 63 cm 10 -25 7.0 g/dL 3.0 g/dL - - - - end stage liver disease

15 20.3 kg "60 th percentile" 5 0 - 7 5 7.1 g/dL 4.5 g/dL 3, formed 3. mild and resolved

*Excellent = 4, good = 3, fair = 2, poor = 1.

714 COOPER ET AL

I 1 0

I00

u

z 9O

"r 80

(.9 Z W -- 70

.J

o ii1

5O _J . J

~ 4 0

~ 3o

~ 2o

IO

�9 ALIVE * DEAD

ILEOCECAL ILEOCECAL VALVE VALVE I N TACT RESECTED

Fig 1. Long-term survival following massive intestinal resection in infants is related neither to residual small bowel length nor to the presence of the ileocecal valve.

required 13.8 _+ 2.5 mo (range, 8 to 26 mo), during which time weaning from parenteral feed- ing occurred; the mean weight at adaptation was 6.87 +_ 1.32 kg (range, 2.91 to 12.9 kg). The majority (7/9) of patients weaned from TPN were tolerating a regular diet without supple- mental medications at the time of follow-up examination; the two patients not yet weaned both continued to require dilute elemental for- mula (3/4 strength Pregestimil), but one of them also required cholestryramine, loperamide, and sodium bicarbonate to avoid excessive output from a proximal colostomy in the absence of the ileocecal valve. The mean number of stools per diem was 3.1 _+ 0.7 (range, 1 to 8), and total serum protein and albumin concentrations were normal in all but two of the survivors, as reflected by mean values of 6.3 +_ 0.2 g /dL (range, 5.2 to 7.4 g /dL) and 4.0 _ 0.1 g /dL (range, 3.3 to 4.9 g /dL) for these two parameters, respectively.

A mean follow up of 3.3 ___ 0.8 years (range, 0.5 to 9.6 years) permitted a meaningful assess- ment by the parents of quality of life; the mean score was 3.0 _+ 0.3, based on an arbitrary scale (excellent = 4, good = 3, fair = 2, poor = 1).

Impossible to convey in any numerical sense, however, was the overwhelming sense on the part of the parents that the outcome was worth their many trials and tribulations, despite the extraor- dinary demands placed on the families, and especially the mothers, of these patients, both during the initial period of hospitalization and for several months thereafter.

Finally, near-normal somatic growth was achieved in that mean weight (W), length or height (H), and weight-for-length or height (W/ H) all approached or exceeded the 25th percen- tile (W = 26th percentile, H = 24th percentile, W / H = 41st percentile). However, better somatic growth was not consistently observed in patients with longer lengths of residual small intestine (Fig 2).

Although not included in this analysis, there is a small group of patients who deserve special mention. These are patients who underwent less than massive small intestinal resection, but in whom subtotal large intestinal resections were also performed, together with removal of the ileocecal valve (Tables 4, 5, and 6). For all intents and purposes, these patients were indis- tinguishable from those with short-bowel syn- drome, and their management was no less prob- lematic. Fortunately, in both of these cases, survival was achieved, but in each case, certain dietary restrictions, such as the avoidance of

I

~40 Z

~- 30 "1"

h i

2O >- a o m

I 0 ~r . . . . . SMALL,, BOWEL, IN CM

2 4 6 B I0 AGE IN YR

Fig 2. Body weights of long-term survivors following massive intestinal resection in infants demonstrate near- normal somatic growth, although better somatic growth was not consistently observed in patients with longer lengths of residual small bowel. The three lines represent the lOth, 50th and 90th percentiles of body weight, as adapted from the anthropometric charts of the National Center for Health Statistics.

SHORT-BOWEl, SYNDROME

Table 4. Patient History and Extent of Resection Following Massive Large Intestinal Resection in Infancy

715

Residual Small Bowel Case Gestational Birth Age at Weight at Ileocecal No. Sex Diagnosis Age Weight Operation Operation Total/Jejunum/Ileum Valve

1 F NEC 34 wk 2.27 kg 10 d 2 .40 kg 220 cm " in tac t " - 2 0 cm 0

2 F NEC 32 wk 1.47 kg 6 d 1.51 kg 180 cm " in tac t " - 3 0 cm 0

Table 5. Extent of Resection and Details of Adaptation Following Massive Large Intestinal Resection in Infancy

Residual Colon Case Length of Time Required Weight at No. Ascending/Transverse/Descending Survival Follow Up for Adaptation Adaptation Diet Medications

1 0 0 " ' ~ " alive 8.17 yr 20 mo 9 .70 kg table food Ioperamide,

(lactose vitamins, ZnSO 4

free)

2 0 0 + alive 7.67 yr 8 mo 3.65 kg table food - -

Table 6. End Result Following Massive Large Intestinal Resection in Infancy

Case Weight for Total No. of Stools Quality of TPN Associated No. Weight Height Height Protein Albumin Per Diem Life Cholestasis

1 27.5 kg 127 cm 5 0 - 7 5 percentile - - - - 4 - 5 , loose 4 none

( 5 0 - 7 5 ( 2 5 - 5 0

percentile) percentile)

2 23 .0 kg 120 cm 5 0 - 7 5 percentile 7.8 g/d[, 4 .5 g/dL 2 - 3 , loose 4 none

*Excellent = 4, good = 3, fair = 2, poor = 1.

fresh fruit, have remained necessary, due pre- sumably to the loss of the water-absorptive capacity of the proximal large intestine.

DISCUSSION

Since 1925, numerous reports have docu- mented survival for large numbers of infants who have undergone massive small bowel resection for a variety of intra-abdominal catastrophes. 2-8 By and large, these reports have confirmed that patients who have undergone resection of no more than 80% to 85% of mean intestinal length have thrived, following a more or less lengthy initial period during which time the remaining bowel "adapted" to its increased level of func- tion. They have also reported the survival of many patients who have undergone even more extensive resections. Indeed, several anecdotal reports have appeared recently which have docu- mented survival in patients with as little as 13, 12, 11, and 8 cm of small bowel remaining. 9-~2 However, it should be noted that in each of these cases, the ileocecal valve was intact, and that lengthy periods were required before intestinal adaptation was complete. Moreover, each of these infants continued to require some form of

specialized nutritional support following dis- charge, including supplemental enteral feeding, and in one case, parenteral nutrition.

Wilmore, in 1972, attempted to determine which factors appeared best to correlate with a successful outcome following massive intestinal resection in a newborn infant. I On the basis of an extensive review of the English literature, he concluded that among infants less than 2 months of age at the time of operation, survival was highly likely if the remaining segment of intes- tine exceeded 40 cm in length, thus confirming the early experiments of Clatworthy and asso- ciates, ~3 but highly unlikely if the remaining intestinal segment was less than 15 cm in length. However, perhaps even more important was his suggestion that infants whose remaining intesti- nal segments measured less than 40 cm in length required the presence of an intact ileocecal valve for survival, thus confirming the clinical observa- tions of Benson and his colleagues, 5 and the experimental studies of both Hammer ~4 and Keller ~5 and their colleagues.

The issue of whether the absolute length of intestine remaining, or the total percentage of bowel remaining, is better predictor of survival

716 COOPER ET AL

was also addressed by Wilmore, although he reached no firm conclusion. However, several authors have stressed the wide variability and length of the small intestine in the infant, 16-~9 serving to highlight the concept that making careful measurement of the length remaining of the intestinal segments at operation is preferred over measuring the length of the resected seg- ment itself. Although there is no data to confirm that isometric segments of intestine removed from different individuals have the same absorp- tive capacities, and while substantial controversy continues to exist regarding the optimal method for determining intestinal length in situ, mea- surement of the length of the unstretched anti- mesenteric border of the remaining intestinal segment still seems to be the most reliable way to assess the quantity of intestinal mass remaining, hence the likelihood of survival. However, we take exception to the notion that SBS can be defined in strictly anatomic terms, since many infants whose total SIL may exceed 75 cm, the outer limits proposed by Rickham 7 for inclusion in the SBS category, may have clinical courses identical to those whose total SIL may be under 75 cm. Thus, we have included in this series all patients whose total SIL was less than 50% of that expected either for gestational age, in the case of the premature infant, or of crown-heel length, in the case of the term infant. It should, of course, be obvious that since no one knows the minimum length of the small bowel necessary for survival, the use of either definition to predict survival in an individual case cannot be justified. Furthermore, since the treatment would be the same in any event, the issue does not seem to be a critical one.

The present data are interesting from a num- ber of different aspects. First, the long-term outlook for most patients with SBS resulting from massive small intestinal loss is surprisingly good. The overall survival in this series compares favorably with that reported by Wilmore, 1 while the distribution of patients in both of these studies with regard to the length of residual intestine and the presence or absence of the ileocecal valve seems quite similar. Although our survival is less than that reported both by Ben- son 5 and by Riekham, v the average length of residual intestine in both of their series was considerably longer; indeed, the survival of our patients with similar lengths of residual intestine is comparable. However, while the distribution of

patients with regard to anatomy may be similar among all these groups, it is quite dissimilar with regard to primary diagnosis. In particular, the largest number of patients in our series underwent massive intestinal resection for necrotizing entero- colitis. This is important, because the functional and/or adaptive capability of the residual intes- tine may also have been adversely affected by the primary disease, thus rendering the anatomically short bowel functionally even shorter. In this regard, it is interesting to note that both of the patients described in Table 2 had severe necrotiz- ing enterocolitis, with multiple skip lesions evident at the time of their operations.

Second, the length of time required before adaptation to enteral feedings corresponded roughly to the time when caloric requirements per unit of body mass are decreasing. This sug- gests that the final phase of adaptation may represent simply a waiting period, during which time the size of the organism and its total absorp- tive capacity may increase proportionately, but during which time its total nutrient requirements per unit mass must decrease sufficiently that the already maximally adapted intestine, which pre- viously had been unable, simply on the basis of its limited surface area, to absorb a sufficient vol- ume of nutrients to sustain life independent of artificial nutritional support, may now do so, due not to any intrinsic change in function, but rather to the relative decrease in energy needs on the part of the organism as a whole. This is a hypothesis that, of course, is yet unproven.

Third, the leading cause of death from SBS is no longer protein calorie malnutrition resulting from the inability of the foreshortened gut to absorb enough food. It is end-stage liver disease associated with TPN. This was, in fact, the only cause of death observed in our series of patients.

Finally, due most probably to the routine use of TPN, neither the length of the residual seg- ment of intestine, nor the presence or absence of the ileocecal valve, could be shown to impaet significantly on survival. This fact, perhaps more than any other, underlies our willingness to try to gain survival for all victims of severe SBS, and we are encouraged that a solution to the vexing problem of TPN-associated cholestasis is closer at hand, and that progress in the field of small bowel transplantation may soon provide hope for those infants whose total small intestinal mass, even after adaptation, will not support life.

SHORT-BOWEL SYNDROME 717

REFERENCES

1. Wilmore DW: Factors correlating with a successful outcome following extensive intestinal resection in newborn infants. J Pediatr 80:88-95, 1972

2. Jackson RH: Extensive resection of the small intestine in an infant. Surg Gynecol Obstet 40:55-61, 1925

3. Potts W J: Pediatric surgery. JAMA 157:627-630, 1955

4. Pilling GP, Cresson SL: Massive resection of the small intestine in the neonatal period: Report of two successful cases and review of the literature. Pediatrics 19:940-948, 1957

5. Benson CE, Lloyd JR, Crabbenhoft KL: The surgical and metabolic aspects of massive small bowel resection in the newborn. J Pediatr Surg 2:227-240, 1967

6. Valman HB: Growth and fat absorption after resection of ileum in childhood. J Pediatrics 88:41-45, 1976

7. Rickham PP, Irving I, Shmerling DH: Long-term results following extensive small intestinal resection in the neonatal period. Prog Pediatr Surg 10:65-75, 1977

8. Bohane TD, Haka-Ikse K, Biggar WD, et al: A clinical study of young infants after small intestinal resection. J Pediatrics 94:552-558, 1979

9. Postuma R, Moroz S, Friesen F: Extreme short-bowel syndrome in an infant. J Pediatr Surg 18:264-268, 1983

10. Holt D, Easa D, Shim W, et al: Survival after massive

small intestinal resection in a neonate. Am J Dis Child 136:79-80, 1982

11. Kurz R, Sauer H: Treatment and metabolic findings in extreme short-bowel syndrome with 11 cm jejunal rem- nant. J Pediatr Surg 18:257-263, 1983

12. Bell M J, Martin LW, Schubert WK, et al: Massive small-bowel resection in an infant: Long-term management and intestinal adaptation. J Pediatr Surg 8:197-204, 1973

13. Clatworthy HW, Saleeby R, Lovingood C: Extensive small bowel resection in young dogs: Its effect on growth and development. Surgery 32:341-351, 1952

14. Hammer JM, Seay PH, Johnston RL, et al: The effect of antiperistaltic bowel segments on intestinal emptying time. Arch Surg 79:537-541, 1959

15. Keller JW, Stewart WRC, Westerheide R, et al: Prolonged survival with paired reversed segment after mas- sive intestinal resection. Arch Surg 91:174-179, 1965

16. Robbin L: The length of the large and the small intestine in young children. Am J Dis Child 19:370-374, 1920

17. Bryant J: Observations upon the growth and length of the human intestine. Am J Med Sci 167:499-520, 1924

18. Siebert JR: Small-intestine length in infants and chil- dren. Am J Dis Child 134:593 595, 1980

19. Touloukian R J, Walker Smith G J: Normal intestinal length in preterm infants. J Pediatr Surg 18:720-723, 1983

D i s c u s s i o n

Judah Folkman (Boston): I think this is an important study by Cooper and his colleagues because it reminds us that given time, the intesti- nal mucosa has enormous potential for regrowth. However, while one waits for this regrowth that may take up to a year or more, there is an accumulating risk of all of the sequelae of total parenteral nutrition.

Because of these accumulating risks we and others have begun to reexamine the whole ques- tion of whether one could provide these babies with an accessory gut by transplanting fetal gut subcutaneously. We have previously demon- strated that in rats this works well and that an accessory subcutaneous gut can support nutrition (J Surg Res 15:176, 1973; Ann Surg 179:109, 1974). We had put the problem aside for a number of years. However, the increasing num- bers of children with short-gut syndrome and the cholestasis that results from TPN, suggests a need for this approach.

Robert Touloukian (New Haven, Conn): Last year we reported to this association the near doubling of the length of the small intestine during the last 12 weeks of gestation. Wilmore's report on short-gut patients receiving TPN 12

years ago reflects survival of babies who were undoubtedly closer to full term than many oper- ated babies in our nurseries today who are likely to weigh a kilogram or less. I believe that we should report our survival data on short-gut babies by the actual percentage of normal bowel remaining at the time of initial surgery rather than by the specific length of the gut remnant. Would the authors choose to comment and also indicate the number of extreme low-birth-weight preterm infants in their study group? An impor- tant factor predisposing these babies to cholesta- sis is partial small-bowel obstruction, particu- larly with relatively high anastomoses. This causes an increased enterohepatic recirculation of conjugated bilirubin and taxes the excretory function of the liver. Would the authors clarify that GI series were done to rule out anastomotic obstruction in their patients.

Arnold Coran (Ann Arbor, Mich): I would like to ask the authors a couple of questions. Did they look at the liver histology in any of these patients and did they see any pattern of the cholestasis? Second, since the length of intestine does not correlate directly with the ability of the intestine to absorb, do they have any absorption

718 COOPER ET AL

data showing what was the ability of these vari- ous babies to absorb nutrients such as fat, carbo- hydrate, and proteins?

John Seashore (New Haven, Conn): I think we all recognize the association of T P N and choles- tasis, but the incidence of liver problems seems extraordinarily high in this group. I wonder if the authors could comment on whether some of this might not be related to the short gut syndrome itself, as is seen in jejunoileal bypass patients, and not necessarily to the TPN.

Denis R. King (Columbus, Ohio): Since the average investment in each of these patients is between a half and perhaps three quarters of a million dollars, based on conservative figures for intensive care hospitalization on the East Coast, I would like to ask the authors, "Are you getting your patients out of the hospital into a hospice environment? Are you sending them home on TPN? What are you doing to decrease the amount of money invested in these unfortunate individuals?"

Arthur Cooper (closing): I would like to thank all of the discussants for their comments. I think Dr Folkman's remarks are particularly interest- ing. He has underscored the fact that adaptation to enteral feedings in these patients takes about 1 year, which corresponds to the time when caloric requirements per unit mass are decreasing. Thus, rather than actual physiologic adaptation, which we know occurs much sooner than 1 year, what we may be observing is that the total caloric needs of the organism as a whole are decreasing relative to its size, such that by the time 1 year of age is reached, the bowel is finally capable of absorbing enough nutrients. Dr Touloukian, the majority of our patients were not preterm, but term infants, so I am not sure that our overall statistics would be altered significantly by con-

sidering the bowel lengths in terms of percent- ages rather than absolute lengths. Moreover, looking at your own data on bowel length in preterm infants, which we have referenced in this paper, the wide standard deviations observed would suggest that use of percentages rather than absolute lengths might not be the optimal way to look at this question, as has been under- scored by several authors in the literature throughout the years. Finally, so far as the question of residual obstruction is concerned, virtually every one of these patients had an upper gastrointestinal examination for transit time at some point or other during his hospital course, and there was no evidence of residual obstruc- tion. Dr Coran, the question of liver histology is a good one. All of the infants who died had biopsy- proven cholestasis. The features were considered entirely consistent with TPN-associated choles- tasis by our pathologist, Dr Witzleben, who was the first to report this condition, and there were no other causes identified. Unfortunately, we have not yet studied the absorptive function of these patients in sufficient detail to warrant our drawing any conclusions about this. Dr Seashore, the question of whether TPN-associated choles- tasis may actually be due to the short gut itself rather than the TPN solution is still open, and certainly there is data to support both concepts. However, I am not sure at this point how we could find the answer in this group of patients. Finally, Dr King, indeed very few of our patients have gone home on continuous enteral and/or cycled total parenteral nutrition. We do make every at tempt to get them out of the hospital as early as we possibly can. I must say that our success rate in achieving this goal has not been as good as we would have liked, but we are hopeful that we will be able to do better in the future.