Work

Gastric Antral Vascular Ectasia As a Manifestation of Liver Graft Versus Host Disease Responsive to Ruxolitinib

Year: 2024

Type: article

Abstract: Gastric antral vascular ectasia (GAVE) or “watermelon stomach” is a rare angio-dysplastic disorder causing gastric bleeding, which can be severe and prolonged. The relationship between GAVE and chronic GVHD is unknown. The underlying pathology of HSCT-GAVE is poorly understood with no established therapy. Treatment of GAVE can include medical, endoscopic, and surgical interventions with varied efficacy. This report presents a pt s/p allogeneic HSCT who developed GAVE refractory to endoscopic intervention, ultimately found to be a manifestation of liver GVHD. A 69 yo female with a history of myelofibrosis s/p matched unrelated male donor HSCT. Her post-tx course was complicated by steroid responsive acute upper GI GVHD at D+27. She had a CR and was tapered off prednisone at D+191 and off tacrolimus D+240. She developed mild ocular and oral chronic GVHD managed topically and remained off systemic IS. Months later, she developed acute GIB with Hgb 5.9. EGD showed diffusely erythematous mucosa without bleeding. But GIB with severe anemia requiring transfusions continued. Repeat upper EGD showed striped erythema with oozing, treated with argon plasma coagulation (APC), was concerning for severe GAVE. She continued to have transfusion dependence, requiring EGD every 2-4 wks treated with APC, radiofrequency ablation, EGD band ligation, hemospray, and spray cryotherapy with limited success. She was referred for consideration for antrectomy. Liver ultrasound showed hepatic steatosis with coarse echotexture concerning for cirrhosis. Liver biopsy demonstrated chronic hepatitis with interface and lobular activity and cirrhosis, most consistent with autoimmune hepatitis. Surgery was deferred. Five months later, the pt established with the OSU GVHD clinic for evaluation and further assessments. CD19+ B cell chimerism was 100% donor. Autoimmune hepatitis antibodies were negative. A liver biopsy taken 4 months previously was further analyzed and showed the vast majority of inflammatory cells in periportal regions (panels A, B) were CD3+CD8+ T-cells (CD3 in panel C) compared to CD20+ B-cells (panel D) that were mostly XY (donor) cells by XY FISH (panel G). CK7 immunostaining highlighted septal bile ducts with mild epithelial damage and bile ductular proliferation. These findings were supportive of hepatic involvement by GVHD. The pt was initiated on ruxolitinib with close monitoring. Overt GIB improved with last transfusion 2 months after start of ruxolitinib. Repeat liver biopsy after 4 months on ruxolitinib showed mild inflammatory infiltrate with no interface activity and no significant lobular inflammation. Repeat EGD showed only mild GAVE without bleeding. Findings on both liver biopsy and EGD were consistent with improved liver GVHD and GAVE. Referral to a multidisciplinary GVHD program enhanced complex patient diagnosis and care.

Source: Transplantation and Cellular Therapy

DOI: 10.1016/j.jtct.2023.12.607