PB2339 GASTROINTESTINAL BLEEDING WAS ASSOCIATED WITH WORSE SURVIVAL IN PATIENTS WITH GASTROINTESTINAL ACUTE GRAFT‐VERSUS‐HOST DISEASE: SINGLE CENTER RETROSPECTIVE STUDY

Background: Acute graft‐versus‐host disease (aGVHD) is an important complication after allogeneic stem‐cell transplantation (allo‐SCT). GVHD is caused by alloreactive donor‐derived T lymphocytes and involves various organs such as the skin, liver, and gastrointestinal tract. Especially, gastrointestinal aGVHD (GI‐aGVHD) is associated with severe morbidity and high rate of mortality. Standard initial treatment for GVHD is corticosteroid, and response rate has been reported approximately 40% to 50%. However, patients who did not respond to initial treatment have shown poor survival rate of 5% to 30%, even if those patients received additional treatments. Aims: The aim of this study was to identify risk factors for mortality in patients with GI‐aGVHD after allo‐SCT. Methods: We retrospectively reviewed clinical records of the patients who underwent allo‐SCT for hematological disorders at our center between September 2006 and April 2018. Results: Seventy‐one patients who developed GI‐aGVHD were included in this study. The median age of the patient was 46 years (range; 19‐65 years). Twenty‐four patients had lymphoid malignancy, 42 patients had myeloid malignancy and 5 patients had other hematological disorders. Forty‐two patients were complete remission at allo‐SCT, and the other 29 patients were not in complete remission. Twenty‐five patients received reduced‐intensity conditioning, and 46 patients received myeloablative conditioning. Graft sources were bone marrow in 26 patients, peripheral blood stem cells (PBSC) in 22 patients and umbilical cord blood in 23 patients. Six out of 22 PBSC were obtained from HLA‐haploidentical related donors. Prophylaxes for aGVHD were tacrolimus‐based regimens in 57 patients and cyclosporine‐based regimens in 14 patients. Median onset day of GI‐aGVHD was 33 after allo‐SCT (range; 3‐134). Staging of GI‐aGVHD was as follows: stage 1,n = 44; stage 2, n = 12; stage 3, n = 6; stage 4, n = 9. Forty patients developed grade II aGVHD, 24 patients developed grade III aGVHD and 7 patients developed grade IV aGVHD. Forty‐five patients received systemic corticosteroid treatment for GI‐aGVHD, and 15 of the patients needed additional treatment. At the median follow‐up of 928 days (range; 259‐3434), estimated overall survival (OS) rate at 2 years was 46%. In a multivariate analysis using Cox regression model, patients with GI bleeding [hazards ratio (HR) = 1.26 (95%CI; 0.20‐0.81), p = 0.040] and concomitant with liver aGVHD [HR = 11.3 (95%CI; 1.1‐114), p = 0.040] were independently associated with worse OS. OS was not different between patients who received corticosteroid and patients who did not received corticosteroid by log‐rank test. However, patients who received additional treatment showed significantly worse OS (5‐year OS 43.4% vs. 0.0%, P < 0.001, Figure). Patients with GI bleeding or concomitant with liver aGVHD need to receive additional treatment (odds ratio (OR) = 28.9 (95%CI; 4.62‐340), p<0.001 for GI bleeding; OR = 11.4 (95%CI; 1.74‐134), p = 0.003 for concomitant with liver aGVHD, respectively). Summary/Conclusion: Among patients with GI‐aGVHD, GI bleeding or concomitant with liver aGVHD was associated with refractory to corticosteroid treatment, resulting in worse OS.