Recent advances in understanding and managing malabsorption: focus on microvillus inclusion disease

Diagnosis of microvillus inclusion disease

In 1978, Davidson et al. presented a case report of five infants with persistent severe diarrhea from birth and marked abnormalities of absorption associated with failure to thrive, leading to death in four infants¹. The common histological abnormalities in duodenal mucosa from those infants were villus atrophy, crypt hypoplasia (without an increase in mitoses or inflammatory cell infiltrate in the lamina propria) and absence of a brush border in villus enterocytes, and an increase in lysosome-like inclusions^2,3. Originally referred to as Davidson's disease, congenital microvillus atrophy, and intestinal microvillus dystrophy, the disease was named microvillus inclusion disease (MVID) in 1989 by Cutz et al.⁴.

As with all rare genetic diseases, the diagnosis of MVID was quite challenging until recently and required histological evaluation for confirmation. It is important to note that MVID has a very low incidence, making it extremely difficult to investigate its pathophysiology⁵. The morphological anomalies observed in the enterocytes of patients with MVID are widely utilized in disease diagnosis². Until recently, the gold standard in diagnosing MVID was combined light and electron microscopy of small bowel biopsy samples of patients. The abnormalities are mainly observed in the small intestine and less frequently in the colon⁶. However, some studies have shown that the colon and rectum biopsies may also contain characteristic features which would be useful in diagnosing MVID^2,6,7.

The key hallmarks which aid in the differential diagnosis include blunted or absent microvilli, accumulation of secretory granules, and microvillus inclusions (MIs) in the epithelial cells^2,8. As depicted in Figure 1, these granules, in most cases, are positive for periodic acid Schiff (PAS) stain and CD10 with an intracellular PAS or CD10 positive line in enterocytes that is commonly detected. Another apical marker which may aid in the identification of the trademark MIs is villin, an apical surface marker of enterocytes⁹. An important factor which should be accounted for during histological evaluation of biopsies is sampling variability and patient-to-patient variability. The diagnosis is confirmed further by genetic testing, which can specifically identify the genetic anomaly of each patient. In this instance, currently, there is a registry which tracks each genetic variation observed in MVID patients to facilitate ease of access to patient-related data for clinicians involved in the management of this rare genetic disorder¹⁰.

Figure 1. Characteristic histological features of microvillus inclusion disease reprinted with permission from Ruemmele et al.².

MVA, microvillous atrophy; PAS, periodic acid Schiff

Differential diagnosis

There are several features that differentiate MVID from other diarrheal conditions with similar clinical presentation including the onset at birth, absence of inflammation, presence of vacuoles containing granules with the characteristic PAS and CD10 positive stain observed under light microscopy, and presence of MIs (Table 1). Other congenital disorders such as chloride and sodium diarrhea can be easily excluded from biochemical assays or genetic testing^6,11. Tufting enteropathy is a disorder with similar onset and blunted villi; however, the presence of surface apical tufts as opposed to apical inclusion bodies in enterocytes distinguishes tufting enteropathy from MVID. Enteroendocrine cell dysgenesis can be differentiated from MVID by the lack of enteroendocrine cells and the presence of normal microvilli. Finally, abetalipoproteinemia is distinguished from MVID by the presence of fat vacuoles and a foamy cytoplasm²⁰.

Clinical manifestations of microvillus inclusion disease

Earlier studies in patients with MVID showed high stool volume (150 to 300 mL/kg/day) with remarkably elevated sodium content (approximately 100 mmol/L)^2,21. The diarrhea present in MVID is considered to be non-osmotic in nature (i.e. fecal ion gap <100 mOsm) and is persistent even when the patient is unfed²². This type of diarrhea is categorized as electrolyte transport-related diarrhea caused by mechanisms involving net secretion of anions (chloride, bicarbonate, or potassium) and/or net inhibition of sodium or chloride absorption^11,23–25. Steatorrhea and impaired glucose absorption have also been reported in MVID patients^26,27. Various studies have shown mislocalization of apical membrane-targeted proteins such as sucrase isomaltase, alkaline phosphatase, and sodium hydrogen exchanger 3 (NHE3) in MVID, which might partly explain the pathophysiology of malabsorption and diarrhea²⁸. Due to the high-volume and persistent diarrhea observed in these patients, the main life-saving treatment option remains life-long total parenteral nutrition (TPN). The use of life-long TPN poses many complications including sepsis and worsening cholestatic liver disease that may require intestinal transplantation. However, outcomes of intestinal transplantation remain poor^2,27. Owing to the immature nature of the enterocytes present in infants with MVID, absorption of essential nutrients is hampered and, therefore, recent studies are directed at developing therapeutic agents which are capable of increasing the maturity of the enterocytes to ultimately recuperate the loss of absorptive capacity of the small intestine²⁹. Although little progress has been made in developing treatment options, the growing research has certainly highlighted relevant mechanisms linking perturbation in cellular trafficking and signaling pathways to functional physiological defects leading to malabsorption and chronic diarrhea³⁰.

Pathophysiology of microvillus inclusion disease

Extraintestinal manifestations in microvillus inclusion disease

The MYO5B gene is expressed in all epithelial tissues, but the most prominent phenotype is observed in the intestine. However, several extraintestinal pathologies have also been reported in other tissues. In this regard, pathologies identified include renal Fanconi syndrome, cholestasis, hematuria, and pneumonia^27,46. Therefore, animal models of MVID could be useful to study these conditions that may be missed in humans owing to the complications associated with disease diagnosis, the very early onset, and lack of survival. With respect to biliary dysfunction, a recent study found cholestasis in 30% of their patient cohort, which was characterized by a low level of serum gamma-glutamyl transpeptidase (GGT)⁴⁷. The study reported abnormalities in the recycling of MYO5B and RAB11A and mistargeting of bile salt export pump (BSEP) to the canalicular membrane of hepatocytes. Although cholestasis in MVID patients was previously thought to be solely due to TPN-related toxicity, evidence has emerged supporting cholestasis in the absence of TPN due to apical trafficking defects in MVID hepatocytes⁴⁸. In this regard, the investigators noted that the unexpected low levels of GGT in MVID patients contrasted with the high levels of this surrogate in cases of liver failure associated with TPN. In a very recent preliminary study conducted in MYO5B null mice and pigs with Navajo mutation (published in an abstract form), the authors demonstrated an interference with apical membrane trafficking in hepatocytes. Specifically, multidrug resistance associated protein-2 (MRP2) and BSEP were mislocalized to subapical compartments. In addition, dipeptidyl peptidase-4 (DPPIV) enzyme was mistrafficked and the liver bile canaliculi lacked branching, highlighting the importance of MYO5B in studying liver dysfunction associated with MVID patients⁴⁹.

Conclusions

Malabsorptive disorders lead to retarded growth and nutritional deficiencies. The complex nature of these disorders poses a challenge for treatment options⁵⁰. Understanding the pathophysiological mechanisms of malabsorption should improve current management protocols and immensely enhance our knowledge regarding intestinal physiology. In this regard, increased understanding of the intriguing malabsorptive disorders of childhood such as MVID should offer new insights at the cellular and molecular levels to unravel the link between cellular trafficking and epithelial absorptive processes. The research in the field of MVID has considerably progressed over the last decade. The generation of novel mouse models with MYO5B deletion has been successful in recapitulating various hallmark features of MVID. So far, the utilization of these models has not only substantiated the role of MYO5B and trafficking machinery in the disease’s pathogenesis but also underscored the importance of cellular trafficking mechanisms in maintaining optimal function of nutrient and electrolyte transporters such as SGLT1 and NHE3. Unlike the in vitro and in vivo mouse models, where loss of MYO5B ideally disrupts intracellular trafficking in all cells, the manifestation of abnormalities in MVID patients is patchy and sometimes confined to a few enterocytes⁵. In addition, although some studies described the presence of abnormalities in the colon and rectum of MVID patients, most animal models focused only on the duodenum and upper small intestine^35,38,39,42. More studies in the distal parts of the small intestine and colon should broaden our understanding of the compensatory mechanisms that the intestine may employ to adapt in consequences of MYO5B mutations. The mechanisms underlying lipid malabsorption associated with MVID remain elusive. Therefore, investigations to explore the molecular basis for dysregulation of lipid absorption in MVID patients and mouse models are warranted. The inducible MYO5B-deficient mouse models have the additional advantage of studying the consequences of time- and age-dependent occurrences of disease-specific hallmarks^29,33,38. Although MVID is a rare disorder, the organoids derived from MVID patients can provide unique opportunities to model the disease and modify the mutated genes by state-of-the-art approaches, including the CRISPR/Cas9 gene editing system, for rescuing the defective phenotype²⁹.