Hepatic steatosis due to total parenteral nutrition: The influence of short-gut syndrome, refeeding, and small bowel transplantation

JOURNAL

OF SURGICAL

RESEARCH

50,

335-343

(1991)

Hepatic Steatosis due to Total Parenteral Nutrition: The Influence Short-Gut Syndrome, Refeeding, and Small Bowel Transplantation

of

J. M. LANGREHR, M.D.,’ M. J. REILLY, B. S., B. BANNER, M.D.,* V. J. WARTY, PH.D.,? K. K. W. LEE, M.D., AND W. H. SCHRAUT, M.D.’ Department of Surgery, *Department School of Medicine, University Presented

at the Annual

Meeting

of the Association

of Pathology, of Pittsburgh, for Academic

This study was undertaken to determine whether refeeding through the native small intestine or through a small bowel transplant would reverse hepatic steatosis induced by total parenteral nutrition (TPN), and of what influence a coexisting short-gut syndrome is. Three short-gut syndromes of different severity were established in Lewis rats (short-gut I, mild; short-gut II, moderate; short-gut III, severe). TPN was administered for 10 days and the animals were refed for 20 days. A liver biopsy after the TPN period confirmed a mild to moderate fatty infiltration of the liver in all groups. After the refeeding period a second liver biopsy was obtained and no evidence of hepatic steatosis was observed in Groups 1, 2, 3, and 4 (normal Lewis rat, short-gut I, II, and III). The animals in group 5 (shortgut I) received a syngeneic small bowel transplant after discontinuation of TPN. After the refeeding period the liver biopsies showed no evidence of fatty infiltration. The intestinal graft also reversed the nutritional deficiencies which were observed in the animals with short-gut and showed normal body weight gain and nitrogen and fat uptake in comparison to the normal animals (Group 1). These data show that a small bowel graft is capable of reversing the deleterious sequelae of short-gut syndrome as well as the TPN-related hepatic steatosis. 0 1991 Academic Press, Inc.

INTRODUCTION Since its establishment as an effective therapeutic modality in the late 196Os,total parenteral nutrition (TPN) has been used widely to support patients who suffer from a variety of malabsorption/malnutrition syndromes.

1 Supported by the German Research Council (La-621/2-l). Present address: Department of Surgery, UKRV, Free University of Berlin, Spandauer Damm 130, 1000 Berlin 19, Germany. ’ To whom reprint requests should be addressed at Department of Surgery, University of Pittsburgh, 497 Scaife Hall, Pittsburgh, PA 15261. 335

and TDepartment of Clinical Chemistry, Pittsburgh, Pennsylvania 15261 Surgery,

Houston,

Texas,

November

14-17,

1990

Some clinical situations require only short periods of TPN, whereas the massive loss of intestine, a clinical catastrophe described under the term short-gut syndrome, requires the patient to rely on TPN entirely and indefinitely. Through the years, TPN has evolved into a standardized form of therapy, and its potential risks and side effects such as hepatobiliary abnormalities, altered bone metabolism, catheter-related complications, and psychiatric disorders have become known [l, 21. TPNrelated or -induced liver impairment can become progressive and life threatening, having no other recourse than discontinuation of TPN. This would only be possible if a new additional intestinal absorptive surface-i.e., a small bowel transplant-could be provided, in the hopes that liver impairment would be reversed by resuming enteral alimentation [3,4]. Since Peden [5] first reported TPN-related morphologic alterations and dysfunction of the liver, many clinical and experimental studies have addressed this problem. The most frequently observed alteration is steatosis, which seems to occur relatively early after institution of TPN, whereas cholestasis, which is less commonly seen, appears to have a later onset [l, 21. Other reported indicators of hepatic dysfunction are elevation of serum transaminases [6], hyperammonuria [7], and elevation of serum cholesterol levels [8]. The pathogenesis of TPN-related hepatic steatosis is not clearly defined. Several mechanisms have been proposed such as decreased hepatic triglyceride secretion [9], intestinal overgrowth with anaerobic flora inducing toxic steatosis [lo], excess carbohydrate calories [8,11], and insulin/glucagon imbalance [12, 131. Recent overviews by Klein and Sax suggest a multifactorial etiology and pathogenesis [ 1, 141. The clinical significance of TPN-related hepatic steatosis is disputed [l, 141.Allardyce reported that elevated liver enzymes in patients on TPN returned to normal values after discontinuation of TPN, suggesting reversal of hepatic cellular impairment [ 151. In some patients on long-term TPN, however, progressive severe chronic liver disease evolves [16]. The lack of definitive conclu0022-4804/91$1.50 Copyright 0 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.

336

JOURNAL

SHORT-GUT

3

OF

SURGICAL

SHORT-GUT

RESEARCH:

2

FIG. 1. Operative procedures. Three different short-gut syndromes were established by resecting parts of the intestine, as indicated by the dashed lines. All bowel and anastomoses were performed with 7-O Novafil.

sions led us to investigate systematically whether related hepatic steatosis is reversible when TPN continued and enteral alimentation is resumed, through the native intestine or by an intestinal plant. MATERIALS

AND

TPNis diseither trans-

METHODS

Animals Male Lewis rats, weighing 180-220 g, were purchased from Harlan Sprague-Dawley, Inc. (Indianapolis, IN). The animals were housed in single metabolic cages (Nalgene Corp., Rochester, NY) in a conventional facility and fed Purina rodent chow (Purina 5001, Purina Mills, St. Louis, MO) and tap water ad Zibitum. Environmental acclimatization was allowed for 1 week before the animals entered the experiment.

VOL.

50, NO.

4, APRIL

1991

proximately 2 cm of small bowel in place. For short-gut II, the first jejunal loop (ca. 5 cm) was left in place and the remaining small bowel was resected to the ileocecal valve. Then the ileal stump was closed and the jejunal loop was primarily anastomosed to the ascending colon in an end-to-side fashion. In this model, the cecum functions as a self-emptying reservoir. The third short-gut model (short-gut III) was obtained by leaving the first jejunal loop (5 cm) in place with resection of the remaining small bowel and cecum. Then, a primary end-to-side jejunocolonic anastomosis was performed, as in shortgut II. A central venous catheter was placed, using the method described by Popp and Brennan [17]. Briefly, the animal’s ventral and dorsal neck was shaved and thoroughly rinsed with 70% propylalcohol. Then the rats were placed in the supine position, covered with sterile drapes, and the incisional site was rinsed again. After isolating the right internal jugular vein, a venotomy was made and a silicone catheter (Silastic, o.d. = 0.95 mm, Dow Corning, Inc., Midland, MI) was advanced into the vena cava superior under aseptic conditions. The catheter was tunneled subcutaneously to the back of the animal, exiting between the scapulae and connected to a swivel/spring coil apparatus (Harvard Bioscience, South Natwick, MA). This device allows the animal to move freely in its individual cage. For syngeneic small bowel transplantation, the small bowel of an age-matched Lewis rat was harvested and transplanted as previously described [ 181. The graft was isolated from the ligament of Treitz to the ileocecal valve on a superior mesenteric artery (SMA)-portal vein vascular pedicle. The bowel lumen was irrigated with cold saline containing neomycin (0.05%) and the graft was flushed intravascularly with cold heparin-saline (10 units/ml). The graft was then wrapped in saline-soaked gauze and stored at 4°C while the recipient was prepared for implantation. Revascularization consisted of end-toside venous and arterial anastomosis to the vena cava inferior and the infrarenal aorta. All grafts were placed in gastrointestinal continuity following the vascular anastomosis. Grafts not surviving beyond 5 days were considered a technical failure and were excluded from further analysis (technical failure rate was approximately 15%). Liver specimens for histology were obtained by performing a wedge biopsy. All biopsies were taken from the right lateral lobe.

Surgical Procedures Three different short-gut models were produced, as shown in Fig. 1. The rats were anesthetized with an ip injection of pentobarbital at a dosage of 50 mg/kg body weight (BW). Short-gut I was established by resecting the entire small bowel of the animal from the ligament of Treitz to the ileocecal valve. Then a primary end-to-end small bowel anastomosis was performed, which left ap-

TPN Solution All animals received a balanced TPN solution (Table l), containing 20% dextrose, 5% amino acids, 4% fat, adequate vitamins, electrolytes, and trace elements. Choline chloride (Sigma, St. Louis, MO) was added to a final concentration of 0.1%. All other parts of the solution were purchased from Abbott Laboratories (North

LANGREHR

TABLE Composition 500 ml 10% amino acids” 286 ml 70% dextrose 200 ml 20% fat 2 ml 50% choline chlorideb 10 ml multivitamins’ 3 ml trace elementsd 22 ml standard electrolytes’ 1024 ml stock

solution

ET

AL.:

TPN-RELATED

HEPATIC

STEATOSIS

AND

TABLE

1

of TPN

337

REFEEDING

2

Experimental

Solution 206 680 376 4

kcal kcal kcal kcal

1266 kcal

’ Aminosyn 5%, 500 ml contains 74 meq of acetate from acetic acid used in processing and from lysine acetate (essential amino acid). b Cystalline choline chloride was diluted in saline. c Ascorbic acid (vitamin C), 100 mg; vitamin A (retinal), 1 mg (3300 units); ergocalciferol (vitamin D), 5 fig (200 units); thiamine HCl (vitamin Bl), 3 mg; riboflavin (vitamin B2), 3.6 mg; pyridoxine HCl, 4 mg; niacinamide, 40 mg; dexpanthenol, 15 mg; vitamin E, 10 mg (10 units); biotin, 60 fig; folic acid, 400 pg; cyanocobalamin (vitamin B12), 5 Pg. d Zinc sulfate, 3 mg; copper (cupric sulfate), 1.2 mg; manganese sulfate, 0.3 mg; chromium chloride, 12 pg; selenium, 60 yg. ’ Sodium chloride, 35 meq (14 ml); potassium phosphate, 20.4 mM (6.8 ml); magnesium sulfate, 5 meq (1.25 ml).

Chicago, IL) and mixed to appropriate final concentrations by the hospital pharmacy using standard procedures. The solution was administered with a Gilson Minipuls 2 peristaltic pump (Gilson, Inc., Middleton, WI) and the flow was adjusted according to standard curves. The amount infused was measured daily and adjusted to 22 to 25 ml/100 g BW/day. This equals approximately 26 to 30 kcal/lOO g BW/day, which meets the nutritional requirements for laboratory rats [191. Experimental Design The experimental plan, as shown in Table 2, was the same for all groups. On Day 0, nitrogen and fat balances were performed. On Day 2, insertion of the catheter and abdominal surgery were performed in one procedure. Then the animals were allowed to acclimate to the swivel/spring coil device for 2 days. During these 2 days, the infused amount of TPN solution was slowly increased. Starting on Day 4, the full amount of TPN was given for 10 consecutive days. No food was given between Day 2 and Day 14, but the animals had free access to water. On Day 14, the catheter was removed and a liver biopsy was performed. In Group 5, syngeneic small bowel transplantation was performed at the same time. After discontinuation of TPN, the animals were refed with normal rat chow for 20 days. Only the transplanted group was fasted for 1 more day after TPN to await abeyance of an ileus. On Days 33 and 34 of the experiment, nitrogen and fat balances and a second liver biopsy were performed. Group 1 consisted of normal rats receiving TPN and refeeding as described above (Table 3). The animals in

Setup Procedure

Day 0

First

nitrogen

balacne,

2

Operation-Catheter into preparation of short-gut, infusion of TPN solution,

4

Full amount

7

Second

of TPN

nitrogen

Third

nitrogen

13

Fourth

14

Catheter removal, chemistry, small start refeeding Fifth

nitrogen

Liver

biopsy,

33

Fifth

nitrogen

34

Liver

biopsy,

internal jugular vein, start of continuous serum for blood chemistry

(approx

28 kcal/lOO

g BW/day)

balance

nitrogen

24

fat balance

balance

10

23

first

balance liver biopsy, serum bowel transplantation

balance, serum

balance, serum

second

for blood

fat balance chemistry

second

for blood

for blood (Group

5),

for Group for Group

4 4

fat balance chemistry

Group 2 received short-gut I on Day 2 and were refed for 20 days after discontinuation of TPN. In Group 3, all animals received short-gut II and were refed for 20 days after the period of TPN. In a pilot study, animals with the short-gut III preparation were noted to develop accelerated emaciation in the refeeding phase. We therefore refed the animals in this group (Group 4) only for 10 days after TPN. The animals in Group 5 received a short-gut I on Day 2 and were then isografted on Day 14 and refed for 20 days. Nutritional

Parameters

All animals were evaluated and weighed daily for the duration of the experiment. Total nitrogen contents of collected stool and urine samples were determined using a modified Kjeldahl method [20]. During TPN the stool production vanished quickly (1 to 2 days); so only urine nitrogen content was measured for this period. Fecal fat

TABLE Experimental Group 1 2 3 4 5

(n)

(6)

Short-gut

(6) (6) (6)

Groups Therapy

None I

(7)

3

II III I

after

TPN

Refeeding for 20 days Refeeding for 20 days Refeeding for 20 days Refeeding for 10 days Small bowel transplant on Day 14 and refeeding for 20 days

338

JOURNAL

OF

SURGICAL

RESEARCH:

300 Normal Lewis

275 G E

250 225

p

200

z

175

P

d

150 125

.!

TPN Period I

100

Refeeding

Period

I

,'.'I""I""1.".I.."I"'.I'.'.I( 0 5 10

15

I

20

25

30

35

Day of Experiment FIG. 2. Body weights were obtained daily and the data are expressed as means + SEM. The SEM was less than 3% of the mean at all time points and therefore was not depicted. The weight gain on TPN was statistically significant (P c 0.01) for all groups compared to the starting weight. Also statistically significantly different was the weight gain (Groups 1 [Normal Lewis] and 5 [Short-Gut I plus SBTX) and the weight loss (Groups 3 [Short-Gut II] and 4 [Short-Gut III]) (P < 0.002) during refeeding, whereas the small weight gain in Group 2 [Short-Gut I] was not significant.

excretion was measured before administration of TPN and after refeeding, according to the method described by Amenta [Zl]. Blood Chemistry On Days 2,14, and 34 (Day 24 for Group 4) serum was obtained from each animal and a blood chemistry profile, including SGOT, SGPT, total protein, albumin, direct and indirect bilirubin, glucose, alkaline phosphatase, creatinine, blood urea nitrogen (BUN), amylase, glucose, lactate dehydrogenase, cholesterol, and triglycerides, was performed using standard laboratory methods. Histology All liver specimens obtained were fixed in 10% buffered Formalin and stained with hematoxylin/eosin and oil red 0 stain for microscopic evaluation. Statistics All results were expressed as means + standard error. The Student t test was used to calculate P values. P values < 0.05 were considered to be statistically significant. RESULTS

Clinical Course and Nutritional

Parameters

All animals of all experimental groups had comparable body weight (Fig. 2) and comparable total food in-

VOL.

50, NO.

4, APRIL

1991

take (Table 4) and were in positive nitrogen and positive fat balance (Table 5) at the beginning of the experiment. After an initial weight loss due to the fasting prior to abdominal surgery and the recovery from the operation, all animals appeared clinically healthy and steadily gained weight on TPN. Stool production vanished in all groups before the full amount of TPN was instituted. As depicted in Fig. 2, the body weight on Day 14 (completion of TPN) showed a statistically significant increase (P < 0.01) in all groups compared to the starting weight. Since the amount of TPN solution administered was adjusted daily according to the body weight, the average total nitrogen intake during TPN was similar in all groups. The nitrogen retained during the period of TPN averaged O-14-0.16 g/day (Groups 1,2, and 5). The animals in Groups 3 and 4 retained only 0.076 to 0.077 g nitrogen. This was statistically significantly less than compared to Groups 1, 2, and 5 (Table 5). During the refeeding period, the normal animals in Group 1 appeared clinically healthy and constantly gained weight. The mean body weight on Day 34 for this group was 263.8 f 2.39 g (P < 0.001 compared to Day 14). The total nitrogen intake on Day 33 was 0.533 + 0.025 g and the total fat intake was 0.9 + 0.01 g (Table 4). All animals had positive nitrogen and fat balances (net nitrogen intake 0.143 + 0.021 g, net fat intake 0.64 f 0.01 g) (Table 5). The animals with short-gut I (Group 2) initially lost weight during refeeding, but then slowly regained their weight by Day 34 (not significantly different compared to Day 14) (Fig. 2). The total nitrogen intake (0.633 + 0.038 g) and nitrogen balance (0.151 f 0.103 g) on Day 33 were the same as on Day 0. The total fat intake on Day 33 (0.92 f 0.6 g) was unchanged from preoperative values. Net fat intake (0.35 + 0.8 g) was significantly reduced (P < 0.05) compared to Group 1 (normal animals) . In Group 3 (short-gut II), diarrhea and weight loss began with refeeding. The body weight stabilized after approximately 5 days, but was significantly less (P < 0.002) compared to Day 14 (213.3 f 1.1 g) and to Day 34 (188.3 f 2.47 g) (Fig. 2). Except for the diarrhea, which persisted during the 34-day observation period, all animals appeared healthy. The total nitrogen intake on Day 33 for this group was 0.952 f 0.088 g, a statistically significant increase (P c 0.006) compared to Groups 1,2, and 5 on Day 33. A dramatic decrease in nitrogen retention (-0.12 f 0.11 g) on Day 33 was noted. This was statistically significant (P < 0.001) compared to Groups 1, 2, and 5. The fat balance on Day 33 for this group showed comparable results. While total fat intake (1.27 f 0.13 g) was increased (not statistically significant), the net fat intake (-0.26 + 0.08 g) was statistically significantly decreased (P < 0.001 compared to Groups 1, 2, and 5 on Day 33). The animals of Group 4 (short-gut III) had a rapid weight loss after discontinuation of TPN. All animals

LANGREHR

ET

AL.:

TPN-RELATED

Total

HEPATIC

Nitrogen Total

Group

(n)

1 (Normal) 2 (Short-gut I) 3 (Short-gut II) 4 (Short-gut III) 5 (Short-gut I + small bowel TX)

6

Day

STEATOSIS

TABLE

4

Intake

from

Food

nitrogen

intake

(g)

0

TPN

period”

Total Day

0.023 0.031 0.048 0.047

0.365 0.41 0.377 0.376

+ f f f

0.003 0.007 0.006 0.007

0.533 0.633

6 6

+ f f +

6

0.537

* 0.034

0.401

+ 0.004

0.577

Group

(n)

Day

1 (Normal) 2 (Short-gut I) II) 3 (Short-gut 4 (Short-gut III) I 5 (Short-gut + small bowel TX)

6 7 6

0.143

6

0.147

k +f 2

6

0.165

k 0.027

0.189 0.135

0 0.017 0.017 0.023 0.015

+ 0.048

cages. Chow

nitrogen

33

0.9 f 0.01 0.92 f 0.06 1.27 t 0.13d 0.98 + O.lld

0.04 0.06 0.07 0.07

+ 0.05

0.76

and TPN

Day

0.87

+ 0.03

and fat contents

were

Blood Chemistry Serum for blood chemistry profiles was obtained at the beginning of the experiment, after the TPN period

TABLE

Net

0.029’

(g)

0

0.93 + 0.83 t 0.73 + 0.73 +

0.025 0.038 0.088'

fat intake

The animals in Group 5 received a small bowel isograft on Day 14 and lost weight during the recovery phase after the operation. They then steadily gained weight, appeared clinically healthy and showed body weights on Day 34 (257.5 f 7.04 g) similar to those of Group 1 (normal animals) (Fig. 2). The total nitrogen intake (0.577 + 0.048 g) and the net nitrogen intake (0.114 + 0.064 g) as well as the total fat intake (0.87 + 0.03 g) and the net fat intake (0.57 + 0.02 g) were almost equal to those seen in Group 1 animals.

had severe diarrhea and showed clinical signs of emaciation after 8 to 9 days of refeeding. We therefore sacrificed the animals after 10 days of refeeding (Day 24). The mean body weight at that time was 149.2 f 5.89 g (P < 0.001 compared to Day 14) (Fig. 2). The total nitrogen intake (0.782 f 0.029 g) was statistically significantly increased (P < 0.01 compared to Groups 1, 2, and 5 on Day 33) and the net nitrogen retention (-0.149 & 0.065 g) was statistically significantly decreased (P f 0.003 compared to Groups 1,2, and 5 on Day 33). The total fat intake (0.98 k 0.11 g) was similar to that of the other groups, but the net fat intake (-0.13 + 0.07 g) was statistically significantly decreased (P < 0.002 compared to Groups 1, 2, and 5 on Day 33).

Net Nitrogen

Day

33

f + 0.952 + 0.782 -+

Note. Twenty-four-hour metabolic balances were performed using single metabolic calculated. Data are expressed as means f SEM. ’ An average value for the total nitrogen intake per day during TPN was calculated. * P < 0.006 compared with Groups 1, 2, and 5. ’ P < 0.01 compared with Groups 1, 2, and 5. d Not significant compared with Groups 1,2, and 5.

339

REFEEDING

or TPN

0.468 0.588 0.522 0.521

7

AND

5

and Net Fat Intake nitrogen TPN

intake

(g)

Net fat intake

period”

0.142 0.158 0.076 0.077

-t f f k

0.005 0.01

0.139

+ 0.017

0.008' 0.012'

Day 0.143

0.151 -0.12

-0.149 0.114

33

Day 0

(g)

0.021

0.67

0.04 0.03'

0.57 0.5 0.44

-If f +

0.04 0.06 0.06 0.04

0.64 0.35 -0.26 -0.13

* 0.03

0.53

f 0.05

0.57

* * * -c

0.05d

Note. Twenty-four-hour metabolic balances were performed using single metabolic cages. Urine content were determined using standard methods. Chow nitrogen and fat contents were calculated. a An average value for the net nitrogen intake per day during TPN was calculated. b P < 0.001 compared with Groups 1 and 2, P < 0.007 compared with Group 5. ’ P d 0.001 compared with Groups 1 and 2, P < 0.005 compared with Group 5. d P < 0.003 compared with Groups 1,2, and 5. e P C 0.003 compared with Groups 1, 2, and 5. /P < 0.05 compared with Group 1. g P 4 0.001 compared with Groups 1, 2, and 5. h P < 0.002 compared with Groups 1,2, and 5.

33

Day _t f -c +

0.01 0.08’ O.OSg 0.07h

It 0.02

and stool nitrogen content and stool fat All data are expressed as means ~fr SEM.

340

JOURNAL

TPN Period

OF

SURGICAL

5

= !‘,j * t

10

,

Refeeding

6

.

I-(

15

20

25

30

35

Day of Experiment

TPNPeriod

Refeeding

Period

1

t

ol.‘..u..‘.r..-‘m’.

0

10

20

30

Day of Experiment

1991

Histology Two liver biopsies were obtained from each animal. As exemplified in Figs. 4 and 5, all animals showed a mildto-moderate fatty infiltration after the lo-day course of TPN. A zone I (periportal lobulus) distribution of the fatty infiltration was seen in most of the sections (Fig. 4). Micro- and macrovesicular fat mostly in the hepatocytes, but also in nonparechymal cells, was observed (Fig. 5). This finding correlated closely with consecutive sections stained with oil red 0 stain. No differences in the severity or the distribution of the steatosis between the experimental groups were observed. The histologic sections from the biopsies taken after the refeeding period were uniform in all animals. Neither fatty infiltration nor other pathologic features, such as cholestasis or fibrosis, were detected.

DISCUSSION

60: s E :

4, APRIL

,

40: 20:

50, NO.

Period

‘20

5 =

VOL.

However, the serum triglyceride levels increased significantly (P < 0.05 in all groups) during the TPN period and remained elevated after the refeeding period (statistically not significant) (Fig. 3). All other tested parameters, including creatinine, BUN, amylase, glucose, and lactate dehydrogenase, remained normal during the experiment. L

0

RESEARCH:

407

0

10

20

30

Day of Experiment

FIG. 3. Serum SGOT, cholesterol, and triglyceride levels are depicted before and after the TPN period and after the refeeding period. The data are expressed as means + SEM. Group 1 (normal Lewis), open squares; Group 2 (short-gut I), open circles; Group 3 (short-gut II), solid triangles; Group 4 (short-gut III), open triangles; Group 5 (short-gut I plus small bowel isograft), solid circles.

and after the refeeding period. Using this design, each rat served as its own control. As indicators of hepatobiliary abnormalities, serum levels of SGOT, SGPT, total protein, albumin, direct and indirect bilirubin, and alkaline phosphatase were measured. SGOT levels showed a uniform increase in all groups at the time TPN was discontinued, by the end of the refeeding period, the levels had normalized (Fig. 3). All other parameters remained in the normal range during the entire experiment. As parameters of fat metabolism, serum cholesterol and triglycerides were determined. The serum cholesterol levels increased slightly during the TPN period and decreased after refeeding. Most of the changes were marginally or not at all statistically significant (Fig. 3).

The improved surgical and medical treatment of conditions such as bowel infarction, Crohn’s disease, congenital short bowel syndrome, and a variety of other diseases has resulted in an increased patient population with a short-gut syndrome. The introduction and standardization of total parenteral nutrition (TPN) has provided a life-saving therapy in the presence of organ failure, a situation comparable to renal dialysis [2]. Small bowel transplantation is thought to be the physiologic treatment for most of these short-gut syndromes. The recent advances in immunosuppressive therapy and the better understanding of transplantation immunology in general have made the first successful attempts in clinical intestinal transplantation possible [22-241. The present study was undertaken against this background to determine the effect of a coexisting short-gut syndrome on the development of TPN-related hepatic steatosis and the capability of an intestinal graft to reverse TPN and short-gut syndrome-related pathologic conditions. Using the well-described model of TPN in the rat, we found that short-gut syndromes of different severity, compared with normal animals, have no differential effect on the development of TPN-related hepatic steatosis or increases in liver enzyme levels. This confirms under standardized experimental conditions the clinical finding that hepatic steatosis and hepatic enzyme alterations associated with TPN are not dependent on the underlying disease [l, 21.

LANGREHR

FIG. 4. TPN period hematoxylin magnification,

ET

AL.:

TPN-RELATED

HEPATIC

STEATOSIS

AND

REFEEDING

341

(A) Histologic section (stained with hematoxylin and eosin) from an animal with short-gut I plus small bowel transplant after the (Day 14). Note the zone 1 distribution of the lipid vacuoles (original magnification, 10X). (B) Histologic section (stained with and eosin) of the same animal as in A after the refeeding period (Day 34). Note the absence of lipid vacuoles seen at Day 14 (original 10X).

We further observed that the consistently induced hepatic steatosis of mild or moderate degree was clearly associated with the administration of TPN, since fasted and nonfasted animals with different short-gut syndromes consistently failed to show any hepatic alterations (data not shown). This result is in contrast with the experience in abdominal bypass procedures, in which hepatic steatosis occurs [27]. A possible explanation for this phenomenon is the reported bacterial overgrowth in the defunctionalized parts of the bowel, since

bacterial overgrowth has been shown to be associated with hepatic steatosis [lo, 271. When the animals were refed, the steatosis vanished quickly in all experimental groups. This finding indicates that restoration of enteral alimentation, even if nutritionally inadequate (as in Groups 3 and 4), quickly catalyses the clearing of the fatty infiltration. It also coincides with the clinical finding that cessation of TPN and restoration of enteral feeding ameliorate hepatic impairments [15, 251. We therefore conclude that the

342

JOURNAL

OF

SURGICAL

RESEARCH:

VOL.

50, NO.

4, APRIL

1991

FIG. 5. (A) Histologic section (stained with hematoxylin and eosin) from an animal with short-gut III after TPN period (Day 14). Note the microand macrovesicular lipid vacuoles in parenchymal and nonparenchymal cells (original magnification, 40X). (B) Histologic section (stained with hematoxylin and eosin) of the same animal as in A after the refeeding period (Day 24). Note the absence of lipid vacuoles seen on Day 14 (original magnification, 40X).

TPN-related hepatic steatosis in our model is of a storage type. Since the experimental data reported by various groups using different time spans of TPN all coincide in showing hepatic steatosis, we further hypothesize that the constant parenteral inflow of nutrients overloads the liver for a certain time period. This hypothesis is supported by the clinical finding that hepatic steatosis and enzyme alterations often resolve even under continuation of TPN [ 1,261. However, it needs to be shown-once the chronic hepatic alterations seen in

clinical practice can be simulated in the animal modelwhether chronic changes are as readily reversible. The questions evolving from these findings, whether addition of enteral feeding for TPN or fasting the animals after discontinuation of TPN will prevent or reverse hepatic steatosis, are currently under investigation in our laboratory. A recent study by Stank0 et al. shows that patients with massive loss of intestine are at increased risk of developing severe hepatic complications after 4 to 10

LANGREHR

ET

AL.:

TPN-RELATED

HEPATIC

months of TPN [28]. Further, it has been shown that a certain percentage of patients on long-term TPN eventually will develop steatohepatitis and liver failure [ 161. In the second part of our study, we demonstrated that a small bowel isograft not only reverses the hepatic steatosis and the serum enzyme alterations caused by 10 days of TPN, but also successfully reverses the nutritional and metabolic consequences of short-gut syndrome (normal liver histology, serum enzymes, body weight gain, and nitrogen and fat balances in Group 5 after refeeding). In contrast, animals with different short-gut syndromes showed clear signs of malnutrition (negative nitrogen and fat balances despite increased nutrient intake, Tables 4 and 5) and even lethal emaciation (Group 4). Assuming that adequate immunosuppressive therapy and postoperative management of an intestinal transplant recipient will be possible in the near future, we propose that a small bowel transplant should be performed early enough to assure that the side effects of TPN and the pathologic conditions caused by the underlying short-gut syndrome can be reversed. Since Grant has shown that a combined liver/small bowel transplantation can be successfully performed [24], there might be a viable treatment modality available for the patient who develops chronic liver failure under long-term TPN and therefore would be a candidate for a combined liver/small bowel transplant.

9.

10.

11.

12.

2.

3.

4. 5. 6.

7.

8.

Hardy, M. A., and Chabot, J. Transplantation of the small intestine-Its time has come. Literature Scan: Transplantation 6: 1, 1990. Schraut, W. H. Current status of small-bowel transplantation. Gastroenterology 94: 525, 1988. Peden, V. H., Witzleben, C. L., and Skelton, M. A. teral nutrition. J. Pediatr. 78: 180, 1971. Grant, J. P., Cox, C. E., Kleinman, L. M., Maher, man, M. A., Tangrea, J. A., Brown, J. H., Gross, R. M., and Jones, R. S. Serum hepatic enzyme and vations during parenteral nutrition. Surg. Gynecol. 573,1977.

Total

Keim, N. L., and Mares-Perlman, J. A. Development of hepatic steatosis and essential fatty acid deficiency with hypercaloric, fat-free parenteral nutrition. J. Nutr. 114: 1807, 1984. Li, S., Nussbaum, M. S., Teague, D., Gapen, C. L., Dayal, R., and Fisher, J. E. Increasing dextrose concentrations in total parenteral nutrition (TPN) causes alterations in hepatic morphology and plasma levels of insulin and glucagon in rats. J. Surg. Res. 44: 639,1988.

Sax, H. C., and Bower, R. H. Hepatic complications enteral nutrition. JPEN 12: 615, 1988.

15.

Allardyce, D. B., Salvian, A. J., and Quenville, N. F. Cholestatic jaundice during total parenteral nutrition. Can. J. Surg. 21: 332, 1978.

16.

Bowyer, B. A., Fleming, C. R., Petz, L. J., and McGill, D. B. Does long-term home parenteral nutrition in adult patients cause chronic liver disease? JPEN 9: 11, 1985. Popp, M. B., and Brennan, M. F. Long-term vascular access in the rat: Importance of asepsis. Am. J. Physiol. 241: H606,1981. Lee, K. K. W., and Schraut, W. H. Structure and function of orthotopic small bowel allografts in rats treated with cyclosporine. Am. J. Surg. 151: 55, 1986. Rogers, A. E. Nutrition. In H. J. Baker, J. R. Lindsey, and S. H. Weisbroth (Eds.), The Laboratory Rat. Orlando: Academic Press, 1979. Vol. 1, pp. 123-152. Tietz, N. Textbook of Clinical Chemistry. Philadelphia: Saunders, 1986. P. 911. Amenta, J. S. A rapid extraction and quantification of total lipids and lipid fractions in blood and feces. Clin. Chem. 16: 339,197O. Deltz, E., Schroeder, P., Gundlack, M., Gebhardt, H., Hansmann, M. L., and Leimenstoll, G. Successful clinical small bowel transplantation. Transplant. Proc. 22(6): 2501, 1990.

17. 18.

20. 21. 22.

of total

par-

23.

Goulet, O., Revillon, Y., Jan, D., Nezelof, C., Brousse, N., CerfBensussan, N., Pellerin, D., and Ricour, C. Small intestinal transplantation (SIT) in children. Transplant. Proc. 22(6): 2499,199o.

24.

Grant, D., Wall, W., Mimeault, R., Zhong, R., Ghent, C., B., Stilles, C., and Duff, J. Successful small-bowel/liver plantation. Lancet 335: 181, 1990. Postuma, R., and Trevenen, C. L. Liver disease in infants ing total parenteral nutrition. Pediatrics 63: 110, 1979. Bengoa, J. M., Hanauer, S. B., Sitrin, M. D., Baker, A. Rosenberg, I. H. Pattern and prognosis of liver function normalities during parenteral nutrition in inflammatory disease. Hepatology 5: 79, 1985.

25.

Touloukian, R. G., and Seashore, J. H. Hepatic secretory obstruction with total parenteral nutrition in the infant. J. Pediatr. Surg. 10: 353,198O. Boelhouwer, R. U., King, W. W-K., Kingsworth, A. N., Weening, J. J., Young, V. R., and Malt, R. A. Fat-based (intralipid 20%) versus carbohydrate-based total parenteral nutrition: Effects on hepatic structure and function in rats. JPEN 7: 530, 1983.

Hall, R. I., Grant, J. P., Ross, L. H., Coleman, R. A., Bozovic, M. G., and Quarfordt, S. H. Pathogenesis of hepatic steatosis in the parenterally fed rat. J. Clin. Invest. 74: 1658, 1984. Freund, H. R., Muggia-Sullam, M., LaFrance, R., Enrione, E. B., Popp, M. B., and Bjornson, H. S. A possible beneficial effect of metronidazole in reducing TPN-associated liver function derangements. J. Surg. Res. 38: 356, 1985.

14.

paren-

M. M., PittE., Beazley, bilirubin eleObstet. 145:

343

REFEEDING

Li, S., Nussbaum, M. S., McFadden, D. W., Gapen, C. L., Dayal, R., and Fisher, J. E. Addition of glucagon to total parenteral nutrition (TPN) prevents hepatic steatosis in rats. Surgery 104: 350,198s.

19.

Klein, S., and Nealon, W. H. Hepatobiliary abnormalities associated with total parenteral nutrition. Semin. Liver Dis. 8: 237, 1988. Wolman, S., Jeejeebhoy, K. N., Stewart, S., and Greig, P. D. Experience in home parenteral nutrition and indications for small-bowel transplantation. In E. Deltz, A. Thiede, and H. Hamelmann (Eds.), Small Bowel Transplantation. Berlin: SpringerVerlag, 1985. Pp. 214-221.

AND

13.

REFERENCES 1.

STEATOSIS

26.

27.

28.

Garcia, transreceivL., and test abbowel

Hollenbeck, J. I., Oleary, J. P., Maher, J. W., and Woodward, E. R. An etiologic basis for fatty liver after jejunoileal bypass. J. Surg. Res. 18: 83, 1975. Stanko, R. T., Nathan, G., Mendelow, H., and Adibi, S. A. Development of hepatic cholestasis and fibrosis in patients with massive loss of intestine supported by prolonged parenteral nutrition. Gastroenterology 92: 197, 1987.