Treatment of Relapse after Allogeneic Hematopoietic Stem Cell Transplantation with Venetoclax, Hypomethylating Agents and DLI - a Retrospective Multi Center Study

Introduction:

The most common cause of treatment failure after allogeneic hematopoetic stem cell transplantation (aHSCT) is relapse. The combination of venetoclax and the hypomethylating agents (HMA) azacitidine (AZA) or decitabine (DAC) have shown promising efficacy in elderly patients with AML. We here present clinical data on 32 patients, who were treated with an HMA/venetoclax combination therapy (HMAClax) for relapse of a myeloid malignancy after aHSCT, collected retrospectively from 11 German centers.

Results:

Sixteen patients (50%) were male, median age was 54 years (30.8-71.5). Diagnoses at aHSCT were 25 AML (17 primary, 8 emerging from MDS, CMML or OMF), 5 MDS, 1 CMML and one atypical CML. Twenty six patients were treated for relapse after their 1^st and 6 after their 2^nd aHSCT. Only 9 patients were in CR at aHSCT. The majority received a graft of a matched unrelated donor (21), 4 from an HLA-identical sibling and 7 from a haploidentical relative. Conditioning was myeloablative in 15 and RIC in 17patients. Median time from aHSCT to last relapse was 5.7ms (1.1-67.8). Five patients had molecular (MR) and 23 had hematologic relapses (HR), 4 patients had extramedullary manifestations 3 with concurrent HR and 1 with MR. Twenty-one patients were treated for 1^st and 5 for 2^nd relapse after 1st aHSCT. Four patients were treated for 1^st and 2 for 2^nd relapse after 2^nd aHSCT. HMAClax was first line therapy for relapse in 8, 2^nd line in 22, 3^rd line 1 and 4^th line in 1 patient. In 21 patients relapse had been refractory to HMA (+/- DLI, +/- lenalidomide). Median time from relapse to HMAClax was 1.8 ms (0.3-42.9). Twelve patients received AZA and 19 DAC with venetoclax. One patient was switched from AZAClax to DAClax because of rising MRD after 6 cycles and back to AZAClax after another 7 cycles. Six patients received DLI. Median number of cycles was 2 (1-15). Six patients are still on therapy.

In total 75 cycles were given. Three patients had non-fatal tumor lysis syndrome. All but one patient had grade 3/4 neutropenia and 25 patients (78%) had grade 3/4 thrombocytopenia. Hospital admission for grade 3/4 infections was necessary in 23 patients (72%), 5 of these infections (22%) were fatal. Overall response rate was 43% (12/28, 2 CR MRD-, 4 CR, 2 CRi, 3 PR, 1 MLFS). Two patients died of infection before first response evaluation and in another 2 response has not been evaluated yet. ORR for patients who received first line HMAClax was 80% (4/5) and 35% (8/23) for salvage treatment. Three of 5 patients with MR reached CR, 2 received HMAClax first line. Time to best response was 1.2ms (0.7-3.8). Six patients lost best response after 1ms (0.4-3.3.) 2 underwent second transplant in remission, 4 have ongoing responses (0.4, 0.7, 3.1 and 8.8ms at last follow up). On July 25th 2019, median follow up was 3.3 ms (0.9-17.3), 20 patients (63%) had died and 12 were alive. Six were continuing HMAClax. One patient developed cGvHD and 4 underwent second aHSCT (2 in remission). Estimated median overall survival was 3.7ms (CI 2.9-4.7). Four responders are continuing treatment with HMAClax. Patients, who responded had an estimated OS of 11.1ms (2 underwent second aHSCT in remission). Median survival of patients with HMAClax first line therapy was 5.8ms and of patients with HMAClax salvage therapy 3.7ms.

Conclusion:

For patients relapsing after aHSCT, venetoclax plus AZA or DAC seems to be an effective, but also highly hematotoxic therapy. Responses occurred fast and were more frequently seen during 1^st line treatment for relapse. Duration of response was short, especially in patients receiving HMAClax as 2^nd, 3^rd or 4^th line therapy. Therefore HMAClax should be explored as 1^st line therapy for relapse after aHSCT in combination with DLI or as a bridge to 2^nd transplant.