Pb2437: association of wilms tumor gene 1 expression with overall survival after azacytidine and donor lymphocyte infusions: a retrospective single center analysis

Topic: 22. Stem cell transplantation - Clinical Background: In acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) relapse after allogeneic hematopoietic stem cell transplantation (HSCT) single agent Azacytidine (AZA) either alone or with donor lymphocyte infusions (DLI) is reported to be feasible and effective. Aims: The aim of the present study was to investigate the association of patient and disease related variables with overall survival (OS) after AZA±DLI. Methods: A total of 29 clinically and molecularly well characterized adult patients were treated with AZA ± DLI in first relapse after allogeneic HSCT in a single center. 5-Azacytidine was administered in an out-patient setting with 75 mg/m2 from d1-7 or with 100 mg from d1-5 every 28 days. DLI were infused at day 8 of every second cycle of AZA. The dose of DLI was depending on the donor type and increased stepwise with each infusion. Results: The median number of azacytidine cycles was n=5 (range: 1-24). DLI were performed in 17/29 patients starting with a minimum dosage of 1x105 in case of haplo, 1x 106 in case of MUD and 1x107 CD3+ cells/kg in case of sibling donors. When well tolerated, the number of DLI was increased up to a maximum of 0.58 x 108 CD3+ cells/kg. No patient died due to toxicity. Main adverse events occurred due to infectious complications and in 10/29 (34.4%) evolved under disease progression. Overall, after a median of 4.9 months (range: 2.4-29.7), 7/29 patients (24.1%) achieved a complete remission (CR) and 5/29 (17.2%) a partial remission (PR) resulting in an overall response rate of 41.4%. The median duration of response was 19.5 months (range: 11.5-29.7) in CR patients and 9.9 months (range: 1.6-17.4) in PR patients. Of note, dividing the patients according to the disease risk index (DRI), that combines disease risk and stage at transplant as proposed by Armand and colleagues, high/very high risk patients had a significantly shorter median OS of 4.4 months (95% CI 2.0-9.5) than intermediate risk patients with 14.5 months (95% CI 3.3-23.4, p=0.007). Bone marrow Wilms tumor gene 1 expression (WT1), determined by RT-PCR, was available immediately before transplant and at relapse in all patients. With a Cox proportional hazards model a cut-off value for WT1 was determined and resulted significant in association with OS at transplant at 150 WT1 copies (HR 2.64; 95% CI 1.10-6.33, p=0.02). Patients (n=14) with a WT1 value of >150 at transplant had a significantly shorter median OS (5.3 months; 95% CI 3.2-9.5) than patients (n=15) with WT1 ≤150 copies (13.5 months, 95% CI 2.0-18.1). At time of relapse the value of WT1 that maximizes the log-likelihood resulted at WT1≤ 1000 (HR 5.38; 95% CI 2.03-14.24, p<0.001). In fact, patients (n=10) with >1000 bone marrow WT1 copies at relapse had a significantly shorter median OS of 4.4 months (95% CI 2.2-7.2) than patients (n=19) with ≤1000 WT1 copies with 15.3 months (95% CI 3.3-23.4). A Cox regression model with stepwise selection for potential predictors confirmed the statistically independent association between WT1 and OS after AZA±DLI, p=0.02. The same model with pre-transplant variables confirmed the significant association with the DRI, p=0.03. Summary/Conclusion: The present results indicate that WT1 levels at transplant and at relapse may be associated with overall survival after AZA±DLI in post-transplant relapse of MDS and AML. Further studies are requested to investigate, whether WT1 might have a role as a minimal disease marker, especially in recent approaches of treating “only molecularly” relapsed patients. Keywords: Azacitidine, Donor lymphocyte infusion, WT1, Post-transplant