Combined Liver and Small Bowel Transplantation in a Child with Epithelial Dysplasia

Epithelial dysplasia (ED) is a recently recognized disorder of the digestive mucosa, probably genetically transmitted as autosomal recessive. It appears in the neonatal period, causing refractory watery diarrhea. Histologic study of the intestinal mucosa shows moderate to severe villous atrophy and a characteristic focal crowding of epithelial cells (called tufts) (1,2). The only available treatment is total parenteral nutrition (TPN), which may cause well-known complications, especially liver disease, in these very young children. We present here first case of a child with ED and TPN-related cirrhosis who underwent successful small bowel and liver transplantation. The indication and timing of this procedure are discussed and its difficulties highlighted. CASE REPORT The child was the second daughter of healthy, unrelated French parents. Her older sister was healthy, whereas a younger one, born 2 years later, also had ED. Our patient was born full term, weighing 3450 g, after a normal pregnancy. She was admitted to the hospital on postnatal day 4 for diarrhea and dehydration. Diarrhea persisted (80-100 ml/kg/day) despite bowel rest. Total parenteral nutrition was instituted. Treatment with methylprednisolone was tentatively administered without effect. The child was first seen at our institution at 4 months of age when she already had cholestasis. Histologic study of duodenal biopsy specimens showed findings typical of epithelial dysplasia (Fig. 1). During the next 2 years, liver disease worsened, and liver biopsy performed when she was 3 years old revealed cirrhosis. At time of transplantation, her bilirubin was 143µM/l and prothrombin time was 43% of control.FIG. 1: Intestinal mucosa before transplantation. Severe villous atrophy, focal crowding of enterocytes without inflammation(Periodic acid-Schiff, magnification, × 25).At 3 years 10 months of age, she received a combined liver and small bowel transplantation from a 7-year-old donor. Her native small bowel was resected from the first jejunal loop to 10 cm before the ileocecal valve. The intestinal graft was anastomosed end-to-end to the proximal loop and put in terminal ileostomy. A gastrostomy tube was placed. Transplantation of the whole liver was performed as usual, with an end-to-end choledochal anastomosis. Initial immunosuppression was achieved with tacrolimus, administered intravenously for the first 5 days, then through the gastrostomy to maintain blood concentrations between 15 and 20 ng/ml the first 2 weeks, then 10 to 15 ng/ml methylprednisolone (bolus of 20 mg/kg, then 2 mg/kg day) and 1 mg/kg per day azathioprine were administered concomitantly. Protocol intestinal biopsies were performed twice weekly the first 2 weeks, then weekly (Fig. 2). Follow-up is summarized inTable 1.TABLE 1: Summary of follow-up after transplantationFIG. 2: Biopsy of the transplanted small bowel 1 month after transplant. Normal villous architecture, normal epithelium, no inflammation. Hematoxylin-eosin; magnification, × 10.One month after transplantation, a bile leak was diagnosed on percutaneous cholangiography performed because of cytolysis, cholestasis, and a collection of fluid on ultrasonography. The child underwent percutaneous drainage, then Roux-en-Y anastomosis, without further biliary problems. Enteral nutrition was initiated with a protein hydrolysate formula(Pepti-Junior, Nutricia, Rueil-Malmaison, France) at 3 weeks, when oral nutrition was allowed. The child showed no interest in eating and had to be fed through the gastrotomy. Intestinal transit time was short and unreactive to loperamide or acetorphan. Steatorrhea was noted (15 g/day 5 months after transplant, and 10 g/day at 1 year). Parenteral nutrition was discontinued 6 months after transplant, and ileostomy closure was performed 9 months after transplant. A severe episode of diarrhea caused by rotavirus occurred at 1 year, and was successfully treated with bowel rest and intravenous rehydration. Liver biopsy performed 6 months after transplant because of cytolysis showed centrilobular inflammatory lesions (Fig. 3). Azathioprine was discontinued without biologic improvement. These lesions were subsequently seen but were less prominent on all successive biopsy specimens. No episode of ischemia or vascular problems had occurred that could explain them. They were thus interpreted as mild vascular rejection(3). It was decided to keep the blood concentration level no lower than 7 to 8 ng/ml. Seroconversion to Epstein-Barr virus (EBV) occurred 2 months after transplant (donor positive, recipient negative at time of transplantation). Biologic lymphoproliferative disease was diagnosed 1 month afterward on the presence of circulating activated B cells and oligoclonal immunoglobulins on protein electrophoresis. It was controlled by lessening immunosuppression. Twenty-one months after transplant, herpes simplex virus (HSV) meningoencephalitis was diagnosed according to clinical, electroencephalographic (temporal localization), and biologic criteria and was successfully treated with intravenous acyclovir. At the same time, the EBV genome was found in the cerebrospinal fluid by polymerase chain reaction. A liver biopsy, performed because of cytolysis, showed portal spaces infiltrated with activated B cells expressing the latent membrane protein of EBV (4) and positively stained with the EBER-1 (small RNA of EBV) probe (5) (Fig. 4). Immunosuppression was lowered (tacrolimus concentration between 5 and 7 ng/ml; prednisone 0.15 mg/kg). A control biopsy 1 month later showed mild acute rejection, without any B cells. Tacrolimus was increased for a few weeks then decreased. Liver biopsy 6 months after encephalitis while transaminases were at still abnormal levels (alanine aminotransferase, 100 IU/ml) showed improved centrilobular lesions, with normal portal spaces and no B cells.FIG. 3: Liver biopsy specimen 6 months after transplant: Centrilobular inflammatory lesions with stenosis of the centrolobular vein. Chronic (vascular) rejection. Hematoxylin-eosin; magnification × 25.FIG. 4: Liver biopsy 19 months after transplant: In situ hybridization with EBER-1 probe revealing nuclear staining of lymphocytes in portal space and sinusoids. Magnification, ×40).Three years after transplantation, the child is living a normal life at home, attends school, has normal growth (-1 SD for height and weight), and normal psychomotor development, except for eating. She still refuses to eat more than half her daily intake. The remaining nourishment is given through the gastrostomy. No intestinal biopsy has been performed since the stoma closure. Findings in the last liver tests were nearly normal. There is no sign of lymphoproliferative disease. DISCUSSION Epithelial dysplasia, also known as tufting enteropathy, is a genetic disease responsible for intractable diarrhea of infancy. The main histologic features are villous atrophy, focal disorganization and crowding of surface enterocytes, pseudocystic formations in the glands, and abnormal regenerative cryptae (1,2). Abnormalities in the laminin and heparan sulfate proteoglycan in the basal membrane of intestinal epithelium and in adhesion molecules on enterocytes were described, and suggest a possible defect in differentiation or proliferation of enterocytes(2,6). In our experience, some patients seem especially prone to the development of early severe liver disease while receiving TPN, as our patient did, and later on, her younger sister. It is notable that both sisters were first seen at the same age (2 years) with clinically severe liver disease (hepatosplenomegaly, portal hypertension, and jaundice). In contrast, other patients reach adolescence without hepatopathy(2). The specific and unknown genetic defect might be a main determinant of liver impairment in this disease. With the introduction of tacrolimus, intestinal transplantation has become a therapeutic option for TPN-dependent patients. It must be combined with liver transplantation in the presence of end-stage liver disease, as was our patient's case. The survival rate of patients from large centers is at least 65% to 70% at 1 year, somewhat better for isolated intestinal transplantation(7,8). In the case of our patient, most problems came from the liver rather than the bowel. Biliary complications, either bile leak or stenosis, are encountered in 10% to 25% of liver transplantations (9). The second and more worrying problem for the future of this transplanted liver was its histologic appearance, with inflammatory centrilobular lesions suggestive of chronic rejection (3). Hepatic ischemia could have induced such lesions, but no episode of ischemia was recorded. Furthermore, the lesions of chronic rejection are interpreted as ischemic and resulting from foamy arteriopathy (3). Despite the absence of ductopenia and the treatment with tacrolimus, an effective drug for chronic rejection, we had no other explanation for these lesions. It is difficult to imagine any correlation between the pre- and posttransplantation liver disease. The EBV-induced lymphoproliferative syndrome is a major complication of immunosuppression, especially of tacrolimus therapy, and remains one of the more worrying problems in small bowel transplantation, for which heavy immunosuppression is needed (10). It is our practice after transplantation to control regularly EBV serology and virus in blood(by polymerase chain reaction), and protein immunoelectrophoresis. When there is suspicion of lymphoproliferative syndrome, study of B cells and their spontaneous proliferation is performed, allowing us to to reduce the immunosuppression quickly. However, the tacrolimus level at the time of herpes simplex encephalitis and lymphoproliferative syndrome in the cerebrospinal fluid and liver was not high, but it was adequate, because acute rejection occurred when the dosage was lowered. The cumulative dose of tacrolimus may also be important in determining the disease. It seems that primary infection occurring around the transplantation, when the recipient is EBV-negative and the donor EBV-positive and when immunosuppression is at its highest level, is the more deleterious (10,11). Epstein-Barr virus-negative grafts should be used in EBV-negative children, but that is impractical because of the availability of organs. All efforts should be made to lower immunosuppression safely but rapidly in the first months, especially in these cases of a mismatch in EBV serology. It is remarkable that no episode of bowel rejection occurred, neither when the histologic liver lesions were at their worst, nor when acute hepatic rejection occurred after encephalitis. The transplanted bowel had near normal function, allowing normal food intake despite a mild degree of malabsorption. Nevertheless, its reabsorption capacities were unable to defend against acute aggression, such as rotavirus diarrhea. This has also been observed by others(7,8,12,13) and has several possible explanations: denervation, absence of lymphatic drainage, ischemia-related injury, and immunosuppressive treatment(7). Intravenous rehydration should be instituted early in these patients when diarrhea occurs. The main digestive problem was psychological and was related to the neonatal onset of ED. This child, who had not been educated to oral feeding and the feeling of hunger and satiety, was not interested in food-a not uncommon problem in this type of transplantation (13). Children receiving TPN should be encouraged very early to eat even small amounts of food, to educate their orality and taste and to prevent this posttransplantation problem that will limit their social integration. In conclusion, ED is a disease that causes definitive intestinal insufficiency, for which life-long TPN or intestinal transplantation are the only therapeutic options. The indication and timing of intestinal transplantation are therefore difficult to decide because TPN is generally well tolerated for long periods of time (14), and the complications of transplantation may be life-threatening (7,8,10,11,13). Nevertheless, it appears that early liver disease develops in some patients with intestinal dysplasia. Combined liver and small bowel transplantation is the only solution when cirrhosis has developed, and brings with it all the difficulties of a combined transplantation. Therefore, liver function should be monitored carefully, and isolated intestinal transplantation should be proposed for these patients before irreversibly severe liver disease is established. ADDENDUM More than 3 years after transplantation, the child began to take food orally. She no longer uses the gastrostomy.