Mycophenolic acid (cellcept and myofortic) induced injury of the upper GI tract.

BACKGROUND: Mycophenolic acid (MPA) is an immunosuppressant drug commonly used in patients undergoing solid organ transplant. Although its pattern of inducing injury in the colon is well-known and features prominent crypt apoptosis that mimics graft-versus-host-disease, the injury pattern in the upper gastrointestinal (GI) tract is less extensively documented. We studied the pattern of upper GI tract injury in symptomatic patients taking MPA.
DESIGN: Twenty solid organ transplant patients who were taking MPA and had concurrent upper GI tract biopsies were identified on a laboratory information system search. From these 20 patients, 17 duodenal, 18 gastric, and 7 esophageal biopsies were examined. All patients were symptomatic. Apoptosis and patterns of chronic and active injuries were assessed on standard hematoxylin and eosin, periodic acid-Schiff/Alcian blue, and Diff-Quik stained slides. To measure the significance of apoptosis, we standardized the apoptotic counts in normal biopsies using duodenal, gastric, and esophageal biopsies from 45 normal cases and performed statistical analysis. For the purposes of this study, we regarded apoptotic counts higher than the mean plus 2 SDs as significant. Thus the cut-off values for apoptosis were > or =2 apoptotic bodies/100 crypts for duodenum, > or =3/100 glands for stomach, and > or =2/10 high-power field for esophagus.
RESULTS: GI-related symptoms and abnormalities manifested between 1 month and 10 years posttransplant and included diarrhea (55%); nausea (45%); abdominal pain (35%); vomiting (25%); GI bleeding (15%); dysphagia (10%); dyspepsia, anemia, and hematemesis (5% for each). Most (14 out of 17, 82%) duodenal biopsies showed apoptotic counts of > or =2/100 crypts, 28% (5 out of 18) of gastric biopsies showed apoptotic counts of > or =3/100 glands, and 57% (4 out of 7) of esophageal biopsies showed apoptotic counts of > or =2/10 high-power field. Four gastric biopsies showed a previously undescribed injury pattern of parietal cells resembling the ballooning degeneration. Additional pathologic findings included: chronic peptic duodenitis (6 out of 17, 35%), active duodenitis (1 out of 17, 6%), and celiac-like features (2 out of 17, 12%) in the duodenum; chemical gastropathy (3 out of 18, 17%), active chronic gastritis without Helicobacter pylori (2 out of 18, 11%), and erosion (1/18,6%) in the stomach; reactive epithelial change (3 out of 7, 43%), active esophagitis (3 out of 7, 43%), ulceration (2 out of 7, 29%), and erosion (1 out of 7, 14%) in the esophagus. Serum MPA levels were available in 7 patients, 6 of whom had abnormal duodenal apoptotic counts. On follow-up, available for 16 patients, symptoms improved in all the patients whose dose was decreased or whose medication was withdrawn (10 out of 10) and symptoms persisted in all the patients whose dose was not altered (6 out of 6). Follow-up biopsies after reduction of the medication dose were available for 1 patient and showed substantial reduction in apoptosis. In contrast, follow-up biopsies from 1 patient whose dosage was not altered showed persistent abnormal apoptotic counts in the duodenum.
CONCLUSIONS: As noted by others (Parfitt JR, Jayakumar S, Driman DK. Am J Surg Pathol. 2008; 32:1367-1372), mycophenolate mofetil-associated injury of the upper GI tract, like that in the colon, is characterized by prominent apoptosis similar to that of mild or grade I graft-versus-host-disease injury. We offer apoptotic count guidelines, which we hope will facilitate recognition of mycophenolate mofetil-associated injury in the upper GI tract.