A clinicopathologic study of 24 cases of systemic mastocytosis involving the gastrointestinal tract and assessment of mucosal mast cell density in irritable bowel syndrome and asymptomatic patients.
[cutaneous mastocytosis]
Counting mast cells in gastrointestinal (GI) mucosal biopsies is becoming an increasingly common practice. The primary reason for this exercise is to evaluate for possible involvement by systemic mastocytosis (SM). However, the features of mastocytosis in GI biopsies are not well described. In addition, recent studies have suggested that increased mast cells may be involved in the pathogenesis of some cases of diarrhea-predominant irritable bowel syndrome (IBS); the term "mastocytic enterocolitis" has been proposed for such cases. As the baseline mast cell density in colonic biopsies from normal patients has not been established in large cohorts, there is no widely accepted threshold for what constitutes increased mucosal mast cells. The aims of this study were (1) to determine the utility of GI biopsies for the diagnosis of SM, (2) to characterize the clinical, histologic, and immunohistochemical features of mastocytosis in the GI tract, (3) to determine mast cell density in normal colonic mucosa from a large cohort of asymptomatic patients, and (4) to compare these findings with those from patients with diarrhea-predominant IBS. Twenty-four patients with SM involving the GI tract, 100 asymptomatic patients, and 100 patients with IBS (the latter 2 groups with histologically normal colonic biopsies) were included. For the mastocytosis group, 107 biopsies (70 involved by mastocytosis; 67 mucosal, 3 liver) from 20 women and 4 men were evaluated (median age 59 y). The most commonly involved site was the colon (19 patients, 95%), followed by ileum (86%), duodenum (80%), and stomach (54%). In 16 cases (67%), the first diagnosis of SM was made on the basis of GI biopsies. Seventeen patients had documented cutaneous mastocytosis. Fifteen of 17 patients who underwent bone marrow biopsy had marrow involvement by SM. Eighteen patients had indolent disease, and 6 had aggressive disease (including all 3 with liver involvement). The most common GI symptom was diarrhea, followed by abdominal pain, nausea, weight loss, bloating, vomiting, or reflux. Liver disease presented with hepatomegaly and ascites. Endoscopic abnormalities (observed in 62%) included erythema, granularity, and nodules. Histologically, involved biopsies were characterized by infiltrates of ovoid to spindle-shaped mast cells in aggregates or sheets in the lamina propria, sometimes forming a confluent band underneath the surface epithelium; 25% of biopsies had only focal involvement (single aggregate). Prominent eosinophils were seen in 44% of involved colonic/ileal biopsies and 16% of duodenal biopsies. Mast cells were highlighted by diffuse membranous staining for KIT and CD25. In the nonmastocytosis groups, all biopsies contained singly dispersed mast cells with no aggregates. The mean highest mast cell counts (in a single high-power field) for asymptomatic patients and IBS patients were 26 (range, 11 to 55) and 30 (range, 13 to 59), respectively. In summary, GI (especially colonic) biopsies can establish a diagnosis of SM in patients with GI symptoms. GI involvement is usually subtle and is often associated with prominent eosinophils, which may obscure the mast cell infiltrate. KIT and CD25 are invaluable markers for the diagnosis. Mast cell density in colonic mucosa from asymptomatic patients is highly variable. Although patients with diarrhea-predominant IBS on average have mildly increased mast cells, the overlap in range with that of control patients is too great for this difference to be clinically useful. These findings argue against the utility of counting GI mucosal mast cell in patients with chronic diarrhea.
