Microvillous Inclusion Disease: An Evolving Condition

Abstract

Microvillous inclusion disease (MVID) is a congenital defect of the intestinal epithelial brush border leading to severe intractable diarrhea of infancy. Lifelong parenteral nutrition (PN) is necessary from diagnosis, and the outlook is poor. In one review, 74% of affected infants died before 9 months of age (1). Because of the very poor prognosis, small bowel transplantation is recommended as a therapeutic option (2), although the best time to attempt this has yet to be clearly established. We report a child with MVID who, at the age of 5 years, is thriving on a normal unrestricted diet and in whom the most recent small bowel biopsy specimens showed pathologic abnormalities significantly less marked than those found at diagnosis. CASE REPORT A girl weighing 3.51 kg was born at term to nonconsanguineous parents and received cow's milk formula from birth. She was admitted at the age of 12 days to the local hospital with drowsiness, vomiting, weight loss (from the 50th to the 10th percentile) and a 1-day history of watery green diarrhea. Stool analysis by microscopy, culture, electron microscopy, and rotavirus antigen were all negative. Other extensive investigations did not provide a clear diagnosis, and because of persisting diarrhea and weight loss, she needed 7 days of PN. Subsequently, she was weaned onto a hydrolyzed protein formula (Pregestimil; Mead-Johnson, Hounslow, Middlesex, UK), which she tolerated well, and she was discharged home 4 weeks after admission. Her weight had increased by 500 g, and she was passing three seedy stools per day. She subsequently had reasonable weight gain along the 10th percentile, reportedly passing two to three seedy stools per day until 3 months of age, when she was admitted with fever, vomiting, and recurrence of green watery stools (six to eight stools per day). She was found to be anemic (hemoglobin, 6.5 g/dL) with a white cell count of 30.7 × 109/mm3. Serum ferritin, folate, vitamin B12, and red cell folate levels were all within the normal range. Stools, blood, swabs, and urine examination did not reveal any pathogens. Two days after admission, spontaneous bleeding into the left hip and buttock occurred, and coagulopathy was identified and corrected with fresh-frozen plasma and vitamin K. Profuse watery diarrhea (6 to 8 times per day) persisted. A diagnosis of intractable diarrhea of infancy was made, and she was transferred to the Royal Hospital for Sick Children, where extensive investigations revealed no evidence of enteropathogens, disaccharidase deficiency, cystic fibrosis, pancreatic insufficiency, or immunodeficiency. Duodenal biopsies (using a Crosby capsule) were then performed. The specimens were collected and processed, using routine methods for light and electron microscopy. Staining with hematoxylin and eosin and periodic acid–Schiff (PAS) was performed. Examination of these specimens confirmed the diagnosis of microvillous inclusion disease (3). Noted were subtotal villous atrophy (Fig. 1A); abnormal PAS staining of the brush border and apical cytoplasm, which was first noted in the upper crypt epithelium (Fig. 1A); microvillous inclusions (Fig. 2A); disordered and patchy surface microvilli (Fig. 2B); and epithelial cells with abnormal accumulation of secretory granules (Fig. 2C). The microvillous inclusions were readily visible throughout the sections. It was also noted that lateral membrane microvilli were readily found (Fig. 2D), which have been described in late-onset MVID with a milder clinical phenotype (1).FIG. 1.: Periodic acid–Schiff (PAS) staining of the original and recent biopsy specimens. (A) Staining (original magnification, × 100) of the original diagnostic specimens in the patient aged 3 months. The jejunal mucosa showed partial villous atrophy and foci of enterocyte cytoplasmic vacuolation most prominent at the apices of villi. Staining (with diastase) showed patchy loss of the enterocyte brush border and positive staining of the apical cytoplasm of the enterocytes, which was first noted in the upper crypts. (B) Staining (original magnification, ×100) of the subsequently obtained specimens at 3 years. In contrast with the original specimens, there was a marked change with longer villi, normal brush borders, no staining of the apical cytoplasm of the enterocytes, and only small foci of enterocyte vacuolation.Figure 1: ContinuedFIG. 2.: Electron micrographs of original diagnostic specimens show (A) a microvillous inclusion (original magnification, ×20,000), (B) areas of virtual absence or disordered microvilli in the brush border (original magnification, ×4,500), (C) abnormal accumulation of secretory granules (original magnification, ×22,500), and (D) lateral membrane microvilli (original magnification, ×28,000).Figure 2: ContinuedFigure 2: ContinuedFigure 2: ContinuedAttempts at enteral feeding with polymeric and elemental formulae were unsuccessful, and total PN was initiated when the child was 5 months of age, at which stage her weight had declined below the 3rd percentile. She subsequently thrived, stabilizing at the 25th percentile at 20 months and subsequently (Fig. 3). Other than central venous catheter infections and catheter changes, she remained in good health, although diarrhea continued unchanged at seven to eight loose stools per day. She had no evidence of significant liver or cardiopulmonary disease. During this period she also received an oral rehydration solution (Dioralyte) and oral bicarbonate supplements with snacks very rarely given as treats.FIG. 3.: Weight chart of the patient from the age of 2 years shows rapid increase in weight after the introduction of an unrestricted diet and continuing weight gain after the cessation of parenteral nutrition.At the age of 3.3 years, after we held a discussion with her parents, because of her general good health and increasing interaction with other children at a nursery, the patient's diet was liberalized to allow her to eat freely. It was thought that if she was closely monitored and her diarrhea did not significantly increase, an oral diet was unlikely to cause her harm. During this period, PN was continued unchanged. Unexpectedly, her weight rapidly climbed to above the 50th percentile (Fig. 3) suggesting that she was absorbing significant energy and nutrients enterally. Her mother did not report significantly increased diarrhea, and a 3-day fat balance study revealed fat intake of 36 g/24 hr, and fecal fat output of 0.74 g/24 hr (absorption index, 98%), which confirmed that she did not have fat malabsorption. Stool electrolytes at this time showed a secretory pattern (108 mmol/L sodium, 11.9 mmol/L potassium, 55 mmol/L chloride and 330 mOsm/kg osmolality). A duodenal mucosal biopsy was performed endoscopically at the beginning of the liberalization of her diet, and examination of the tissue showed a marked morphologic improvement over that shown in the original diagnostic specimens (Figs. 1B and 4). The villi were slightly short (villus-to-crypt ratio 2:1, Fig. 1B). Surface enterocytes over large parts of the mucosa appeared entirely normal with well-preserved brush borders and no abnormalities of the enterocyte cytoplasm detected by PAS staining (Fig. 1B). Electron microscopic examination also showed a significant change, with minimally disordered microvilli (Fig. 4A) and no internalization of microvilli within cytoplasmic vacuoles. Lateral membrane microvilli continued to be found but less frequently than in the original set of biopsy specimens (Fig. 4B). Abnormal accumulations of secretory granules were not present.FIG. 4.: Electron micrographs of recent specimens. (A) A marked improvement in the morphology of the enterocyte microvilli (original magnification, ×3,000) was found and (B) displacement of microvilli along the lateral cell borders of the enterocytes (original magnification, ×32,000) continued to be found, but in lower numbers.Figure 4: ContinuedFurther biopsies (both endoscopic and using Crosby capsule) were performed simultaneously 6 months after the patient's diet had been liberalized. The specimens showed no changes in comparison with the specimens obtained 6 months before. There were also no differences between the specimens obtained in endoscopic biopsies (distal duodenal) and those obtained at the same time using a Crosby capsule (jejunal). This confirmed that any changes between the original diagnostic specimens (obtained by Crosby capsule) and the more recent specimens (obtained endoscopically) were not due to the method of sampling. After this, PN was gradually discontinued. At this writing, she is growing along the 50th percentile for weight without requiring PN (Fig. 3). She remains well, and after 1 year without need for PN, the central catheter has been removed. DISCUSSION This is the first report of a child with proven MVID in whom PN has been discontinued and whose condition has shown both a clinical and pathologic improvement. Her pathologic findings at diagnosis were pathognomonic of the disease (3), and review of histology by experts in this disease confirmed the diagnosis. The dramatic improvement in the PAS-stained specimens and the reduced (but persisting) abnormalities in the samples examined by electron microscopy, allied with her ability to thrive on an unrestricted enteral diet are, as far as we are aware, unique. It is well recognized that there is variation in the clinical severity of MVID. Phillips and Schmitz (1) reviewed 23 cases, 19 of which occurred within a week of birth and were described as congenital. The remaining 4 occurred between 6 and 9 weeks and were described as late onset. The late-onset group appeared to have a better prognosis; three were alive at the time of publication. The oldest was then 5 years of age, and two tolerated some oral feeding, but all needed PN. In our patient, symptoms appeared at 12 days but appeared to improve with hydrolyzed feeding until 3 months of age when she experienced diarrhea and evidence of malabsorption (coagulopathy responsive to vitamin K). Her clinical presentation was thus more in keeping with the late-onset group, and the disease was clearly at the mild end of the spectrum. Despite this, she continued to have secretory diarrhea, possibly relating to an abnormal sodium transport system (4). One case has been reported in which there was a clinical improvement (reduction in the output of stool), which followed an episode of shock (5). Although the appearance of the small bowel specimens were unchanged before and after this episode, colonic mucosa collected afterward was normal. Unfortunately, no colonic specimens were obtained before the episode of shock, and it is therefore unknown whether this was an improvement. The abnormal accumulation of PAS material in the tissue specimens of our patient was initially seen in the epithelial cells of the upper crypt. In Phillips' series this was found to be consistent with the severe congenital group. In those with late-onset disease, changes were first noted in the low villus epithelium. However the presence of a larger number of lateral membrane microvilli in this case, both at diagnosis and in later specimens, is consistent with the findings in the less severe late-onset group reported by Phillips and Schmitz (1). Thus, this case seems to have disparate data for the clinical and pathologic phenotypes. The PAS-stained polysaccharides, glycoproteins, and glycolipids and the abnormal accumulation in the epithelial cells are thought to be due to abnormal localization of the brush border enzymes (6) and have been related to the presence of secretory granules (7). This agrees with our finding of virtually normal PAS staining and the absence of accumulations of secretory granules in the follow-up biopsy specimens. This case and the review by Phillips and Schmitz (1) suggest that the presence of lateral membrane microvilli detected by electron microscopy can be related to a mild clinical phenotype. The frequency of cytoplasmic inclusions has not previously been related to the clinical outcome; however, three cases labeled intestinal microvillous dystrophy in which no inclusions were seen all had a poor prognosis (8). By obtaining mucosal specimens from two sites at the same time, we found no regional differences in the abnormalities seen. A pathologic study of resected tissue in a child with MVID who underwent multivisceral transplantation showed that the histologic abnormalities extend throughout the small intestine (9). The apparent improvement in the appearance of the mucosal specimens was a great surprise. We had expected that the features originally identified would persist largely unchanged. It has been suggested that MVID is a congenital abnormality of a transport mechanism in the exocytosis of brush border–related material (10). However Phillips et al. (11), in studying the exocytotic pathway for sucrase-isomaltase was unable to identify an abnormality in the constitutive pathway. Despite clinical and pathologic improvement, this child continues to pass loose stools six to eight times per day, and abnormalities persist in the intestinal biopsy specimens (Fig. 4) that suggest that there is an underlying continuing pathologic process. There are three possible ways in which this improvement may have occurred. The first possibility is related to the genetic basis of the disease. Microvillous inclusion disease is considered to be an autosomal recessive condition, although the molecular abnormality has not been identified. If MVID is an abnormality of a single gene, the most severe congenital cases could have a mutation of that particular gene that does not alter with age. In normal human development, there are a number of examples of changes of gene regulation with age (e.g., hemoglobin chain synthesis). It could be hypothesized that the less severe subjects, such as this case, may have some defect in the regulation of the gene, which could then alter with age leading to the improvement seen. Next is the possibility that the marked abnormalities seen in the original specimens were the consequence of an acquired disease yet to be described. Infection is possible, although no infective organisms were identified. Whether environmental agents could cause this disease has not been explored, although experimental data have produced microvillous inclusions in the epithelium. In vitro, this has been shown in organ culture of fetal intestinal epithelium exposed to cytochalasin, which disrupts microfilaments by binding to their elongation ends (7). In vivo, displacement of microvilli along the lateral cell border of the enterocytes, as seen in this patient and in less severe cases reported by Phillips and Schmitz (1), has also been induced in rats by using colchicine, an antimicrotubular agent (12). The final way that the improvement may have occurred is as a consequence of the dietary exclusion that the child underwent after total PN was initiated when she was 5 months of age. This is not the case, in that after the reintroduction of a normal diet for more than 6 months, we have not found any clinical deterioration or change in the morphology of the tissue specimens. Therefore, the suggestion that food may be a factor in the development of the microscopic abnormalities (13) is questionable, and we are encouraged to attempt repeated enteral challenges in other children with this condition. Because more of these children survive for longer periods (with improving management of PN) we believe that the use of early intestinal transplantation for the treatment of MVID should be reviewed. We suggest that in the few patients with features of late-occurring MVID, associated with the increased presence of lateral membrane microvilli on electron microscopy, PN should be the mainstay of treatment, but regular review of small intestinal morphology and function, to include enteral challenges, should be undertaken. Small bowel transplantation will continue to have a role in the management of this disease, but as the outcome from transplantation continues to improve, there may be a temptation to list children with MVID for transplantation before the development of significant PN-associated liver disease. The case presented here illustrates the need for caution in considering early transplantation in children with late-occurring or clinically mild MVID. Acknowledgments: The authors thank Dr. Mary Loudon of the Monklands and Bellshill Hospitals National Health Service Trust for referring this case, and Dr. Alan Phillips of the Department of Paediatric Gastroenterology, Royal Free Hospital, for his helpful comments on the original biopsy specimens.