Treatment of Microvillus Inclusion Disease by Intestinal Transplantation

Abstract

Microvillus inclusion disease (microvillus) is an uncommon form of congenital protracted diarrhea usually starting in the early neonatal period. It was first reported in 1978 (1) and is probably of autosomal recessive inheritance. Clinical presentation with secretory diarrhea in the first week of life is typical, with massive stools and electrolyte loss even when no enteral nutrition is provided. The diagnosis is based on the ultrastructural finding of intracytoplasmic inclusions that are lined by intact microvilli (2), which are present in the absorptive surface epithelial cells of the small and large intestine, associated with poorly developed surface brush border microvilli. It is thought to be due to a defect in brush border assembly and cellular differentiation, which may be secondary to defective meiosis (3). Recently, defects in small intestine apical but not basolateral membrane transport systems have been demonstrated in this condition (4). There is no effective nutritional or drug therapy, and these children are dependent on parenteral nutrition and intravenous replacement of fluid and electrolytes. Intestinal transplantation has been described in three children (5–7), two of whom were alive at follow-up. In all three cases the native ileocecal valve and at least part of the native colon were removed at the time of transplantation because of concern that, because the colonic enterocytes also expressed microvillus, the native colon would be a source of postoperative secretory diarrhea. The patient described in this report is a 2-year-old girl with MVID who underwent liver and small bowel transplantation with the preservation of 15 cm of the native ileum, the ileocecal valve, and the entire colon, with no deleterious effects. CASE REPORT The patient was born at 34 weeks' gestation. Colonic dilation had been noted on antenatal ultrasound examination at 30 weeks. Her birth weight was 2.1 kg. She did not need resuscitation and was admitted to the special care unit for only 3 days. She was the third child of healthy consanguineous parents. They had one healthy child, their first child having died at 5 days of age of unknown cause. A maternal aunt and uncle had died at 14 and 28 days of age of unknown cause. The patient became acutely unwell on day 6 of life with inability to feed adequately, hypoglycemia, dehydration, shock, and acidosis. Parenteral nutrition (PN) was initiated on day 12. She had a complicated course over the next 4 months with episodes of severe acidosis, hyponatremia, hypokalemia, line sepsis, and necrotizing enterocolitis. She had stool losses of 300 to 400 ml/kg·day when feeding by mouth, which reduced to 150 to 200 ml/kg·day when nothing was consumed orally. When clinically more stable, she underwent duodenal biopsy, which confirmed the diagnosis of MVID. Subsequently, a colonic specimen also showed classic features of MVID with microvillus inclusions visible on electron microscopy (Fig. 1). Liver function remained abnormal but worsened at the age of 4 months. The microvillus had marked failure to thrive. At 9 months of age, she was referred for consideration for small bowel transplantation.FIG. 1.: Transmission electron micrograph of colonic biopsy specimen shows microvillous inclusions (thin arrows) and secretory granules (thick arrows). The inclusions were widespread, involving almost every enterocyte (original magnification, ×33,000).Irreversible intestinal failure, complicated by PN liver disease was confirmed. A liver specimen showed micronodular cirrhosis, Scheuer score of 3 (8), and severe fatty change. The patient had no varices or ascites, but her spleen measured 8.3 cm. A medical and developmental assessment indicated no contraindications to transplantation. She was listed for a combined liver and small bowel transplantation at the age of 10 months. At the age of 2 years she received a combined en bloc liver and small bowel graft including the donor duodenum and pancreatic head (thus preserving the whole common bile duct). The combined graft was transplanted in one step by performing only a caval anastomosis and an arterial reconstruction from an aortic conduit. Although the donor weighed 8 kg and recipient 7.8 kg, the donor small bowel was too large and was reduced by removing 30 cm of donor ileum. Fifteen centimeters of the native ileum, the native ileocecal valve, and total colon were preserved. The native jejunum was anastomosed end to side to the donor jejunum, and donor and native ileum were anastomosed end to end. An ileostomy was formed 20 cm proximal to the ileal anastomosis (Fig. 2). It is routine practice to form an ileostomy, to allow access to perform a biopsy in the transplanted small bowel.FIG. 2.: Operative technique for en bloc liver and small bowel transplantation in the patient, with the preservation of 15 cm of native ileum, native ileocecal valve, and the entire colon. 1 = liver graft, 2 = small bowel graft, 3 = stoma, 4 = native ileum and colon, 5 = jejuno-jejunal anastomosis, 6 = ileo-ileal anastomosis, 7 = donor portal vein, 8 = native hepatic vein, 9 = hepatic artery, 10 = superior mesenteric artery, 11 = arterial conduit, 12 = native duodenum, 13 = donor duodenum and head of donor pancreas.The postoperative course was complicated by hematemesis due to gastritis on day 3, which was managed conservatively, and an episode of intravenous catheter sepsis on day 7 due to Staphylococcus aureus, which was treated by changing the catheter and intravenous flucloxacillin. Enteral feeding was initiated on postoperative day 4 with oral rehydration solution (Dioralyte; Rhône–Poulenc Rorer, West Malling, Kent, UK) at 1 ml/hr and subsequently with a modular food comprising hydrolyzed whey (hydrolyzed whey protein multidextrin mixture; SHS, Liverpool, UK), short-chain fat (Liquigen, SHS), and glucose polymer (Maxijule, SHS) (9). Feeding by PN was discontinued on day 18 when total caloric requirements were achieved through the enteral route. Immunosuppression was with tacrolimus with initial posttransplantation levels of approximately 20 ng/ml. The patient had two episodes of mild rejection on days 9 and 13, coinciding with a period when the tacrolimus levels were 12 and 9 ng/ml, respectively. The episodes were successfully treated by increasing the tacrolimus dose and administering a single dose of 30 mg/kg intravenous methylprednisolone. High ileostomy losses were controlled by 1 to 1.5 mg/kg·day loperamide. She was discharged home on day 58 with ileostomy losses of between 35 and 60 ml/kg·day. Small bowel transit time, measured by carmine red vegetable dye, was 4 hours. From discharge to the time of ileostomy closure at 6 months after transplantation, ileostomy losses remained high at approximately 50 ml/kg·day (Fig. 3). Despite enteral replacement of fluid losses with oral rehydration solution at home, the patient became significantly dehydrated twice during this period, necessitating hospital admission and treatment with intravenous fluid.FIG. 3.: Ileostomy and stool losses from time of transplantation until 1 month after ileostomy closure at 6 months after transplantation.♦, stoma losses (ml/kg/d); **, calculated stool loss (ml/kd/d).The ileostomy was closed at 6 months after transplantation, and stool volume decreased markedly after that time. Stool volume was not measured precisely after ileostomy closure because the patient was in diapers, but combined urine and stool loss was measured for 1 month. Average urine output was 350 ml/day before closure of the ileostomy. After ileostomy closure the stool loss was calculated to be the total stool and urine loss (measured by weighing the diapers) minus 350 g of urine. The calculated stool loss was 280 ml/day at 1 week after ileostomy closure and 200 ml/day at 1 month after ileostomy closure, which suggests at least a 50% reduction in stool volume by that time (Fig. 3). She had soft, formed stools from 2 months after ileostomy closure. There have been no further episodes of dehydration since ileostomy closure, and additional enteral fluid replacement was discontinued entirely 3 months after closure. During the 6 months since ileostomy closure the loperamide dose has been reduced from 0.7 mg/kg·day to 0.3 mg/kg·day, with the plan to stop it in the near future. At this writing, the patient is 3 years old, 1 year after small bowel transplantation and 6 months after closure of the ileostomy. She has two to three soft, formed stools per day and is showing signs of catch-up growth (Fig. 4). She is very well and making good developmental progress. Her nutrition is still mainly from nasogastric tube feeding, but she is consuming an increasing amount of the family diet. The current immunosuppressive regimen is with 0.4 mg/kg·day tacrolimus and 5 mg prednisolone on alternate days. Her trough tacrolimus levels are maintained in the range of 8 to 12 ng/ml.FIG. 4.: Height and weight z scores from 450 days before transplantation to 365 days after transplantation. The patient had marked failure to thrive before transplantation, despite adequate nutrition. She did not show catch-up growth after transplantation until fluid and electrolyte balance improved after ileostomy closure. |Z|, height Z scor; ▴, weight Z score.DISCUSSION Intracytoplasmic inclusions lined by intact microvilli present in the absorptive surface epithelial cells of the small and large intestine, associated with poorly developed surface brush border microvilli, are the classic features of MVID. Because the structural abnormalities seen in the small intestine in MVID are also present in the large intestine, it has been assumed that retention of the native colon at the time of intestinal transplantation for MVID may be a source of postoperative secretory diarrhea. Three cases of intestinal transplantation for MVID have been reported previously (5–7). The first case (5) describes a child who had the entire native colon removed at intestinal transplantation, except for 4 cm of sigmoid colon to which the donor ileum was anastomosed. Two years after transplantation the ileostomy had not been closed, and ileostomy output remained high, despite loperamide treatment. She also had persistent carbohydrate intolerance secondary to bacterial overgrowth in the graft, which necessitated antibiotic and sodium citrate treatment. The second patient (6) had the entire native colon removed, but the donor cecum, ascending colon, and half the transverse colon were transplanted, in addition to the small bowel. He had a Bishop–Koop ileostomy, and a perforation occurred at the site 2 weeks after transplantation, necessitating the removal of the distal ileum up to the proximal ascending colon of the graft. A short length of donor colon was therefore left, but no ileocecal valve. The patient had high ileostomy losses that were treated with loperamide and codeine. He had prerenal insufficiency and persistent metabolic acidosis because of high fluid and electrolyte losses, which corrected after closure of the ileostomy at 9 months after transplantation. He began consuming a normal diet without the need for additional fluid replacement. The most recent case reported (7) describes the retention of the native transverse, descending, and sigmoid colon at the time of intestinal transplantation. Similar to the female patient in this present case report, this male patient had high ileostomy losses treated by loperamide before ileostomy closure, but stool volume reduced considerably once ileocolonic continuity was restored. No adverse effects due to the retention of part of the native colon were described, and the child had 2 to 4 stools per day after ileostomy closure. He died at 90 days after transplantation secondary to ischemic graft perforation. Electron microscopic findings were unusual in this patient, in that, although he generally had classic features of MVID, microvillus inclusions, though seen, were not widespread. Therefore, the good outcome in fluid and electrolyte balance after retention of a large part of the colon could not necessarily be extrapolated to all children with MVID. The patient in this present case report had classic MVID findings on periodic acid–Schiff staining and on electron microscopy, both in the small bowel and in the colon. Microvillus inclusions were seen in almost all the colonic enterocytes examined by electron microscopy (Fig. 1). She had a more severe phenotype that manifested early in the neonatal period with extremely severe secretory diarrhea and resultant marked acid-base and electrolyte problems. This suggests that other children with the severe phenotype of MVID also could be expected to benefit from conservation of a short length of ileum, the ileocecal valve, and the entire colon at the time of transplantation. In the three previously reported cases (5–7), improved fluid and electrolyte balance after intestinal transplantation were seen in those children who had some native or transplanted colonic function. The inclusion of the ileocecal valve and at least the ascending colon with the intestinal graft is controversial. There is evidence in adults that the inclusion of the colon adversely affects graft and patient survival, probably because of an increased risk of sepsis and endotoxemia (10). The evidence in pediatric patients is less clear, but there is concern that the same risk may apply (11). It has been observed, however, that the inclusion of the graft colon probably reduces the high fluid and electrolyte losses after intestinal transplantation, thus facilitating fluid and electrolyte management (12). However, if no adverse effects are demonstrated because of the preservation of the native colon, then preserving the native colon is preferential to using a colonic graft, because it would be unaffected by rejection. In addition the preservation of a short segment of native ileum may allow the ileal break feedback system to function. The presence of intraluminal nutrients in the ileum has been shown to slow gastric emptying (13), small bowel transit time (14), and jejunal motility (15), which allow better fluid and electrolyte absorption by the transplanted small bowel. Retaining the ileocecal valve has the theoretical advantage of suppressing retrograde bacterial overgrowth in the small bowel. The patient in this present case report had episodes of sepsis and rejection similar to those of the three reported patients with MVID who underwent intestinal transplantation and resection of part or the entire native colon (5–7). Stool volume decreased markedly at the time of restoration of ileal–colonic continuity, and she no longer needs enteral fluid replacement and is taking only a small dose of loperamide. She did not show catch-up growth after the intestinal transplantation until the ileostomy was closed, despite adequate nutrition. After ileostomy closure, while supported by the same nutritional intake, she showed excellent catch-up growth (Fig. 4), presumed to be due to the improvement in fluid and electrolyte balance. The patient obviously benefited from retention of 15 cm of ileum, the native ileocecal valve, and the entire colon. However, the exact origin of the dramatic improvement in fluid and electrolyte balance after ileostomy closure is unclear. There may have been some element of adaptation of the small intestinal graft, although the ileostomy output had appeared to stabilize at approximately 50 ml/kg·day after transplantation (Fig. 3), and there was no reduction over time before ileostomy closure. Certainly, the preservation of 15 cm of native ileum, may have allowed the ileal break feedback system to function, with a consequent slowing of gastric emptying and transit time, allowing better fluid and electrolyte absorption by the transplanted small bowel. Finally, the native colon may have some absorptive properties despite expression of classic MVID by the colonic enterocytes. It is likely that the improved fluid and electrolyte balance after ileostomy closure was due to a combination of these three factors. Additionally, the presence of the native ileocecal valve theoretically reduced the risk of retrograde bacterial overgrowth of the patients small bowel graft, and in contrast to the case described by Oliva et al. (5), carbohydrate malabsorption and small bowel bacterial overgrowth have not been a significant problem for the patient in our present case report. CONCLUSION We have demonstrated that the retention of the entire native colon, ileocecal valve, and 15 cm of ileum not only had no deleterious effects, specifically no gram-negative sepsis or secretory diarrhea, but improved the fluid and electrolyte balance with associated catch-up growth (Fig. 4). Intestinal transplantation is a life-saving alternative treatment for patients with MVID. There is no evidence either from this case or that reported by Randak et al. (7) that retention of the colon in MVID is a source of posttransplantation secretory diarrhea or sepsis. Because the native colon may have useful water absorption properties, the short section of native ileum can function in the ileal break feedback system, and the native ileocecal valve is likely to reduce bacterial overgrowth of the intestinal graft, we propose that retention of the entire colon, ileocecal valve, and a short length of ileum, at the time of intestinal transplant for MVID, should become the procedure of choice. We would also advocate early closure or conversion of the stoma to a Bishop–Koop type, to use the ileal break feedback system and the native colon's absorptive properties to facilitate posttransplantation fluid management. Acknowledgments: The authors thank Dr. Victor Miller for referring and sharing the care of his patient with us; Prof. Ian W. Booth and Dr. Stephen Murphy for providing advice; and Dr. Faro Raafat for pathology expertise and for providing the electron micrograph of the colonic enterocyte.