Thiotepa-Fludarabine-Treosulfan (TFT) Conditioning for 2nd Allogeneic Peripheral Blood Hematopoietic Cell Transplantation (HCT) from a Second Unrelated Donor in Patients with Acute Myeloid Leukemia (AML) Relapsed after Prior Allogeneic HCT: A Prospective Multicenter Phase II Trial

Introduction: Relapse of AML after allogeneic HCT has a dismal prognosis. Long-term survival after 2^nd allogeneic HCT has been described in selected patients. Here we tested a specific protocol with a fixed drug combination for myeloablative conditioning and GvHD prophylaxis for 2^nd allogeneic HCT from a different unrelated donor. (EudraCT no.: 2012-005414-18, German Clinical Trials Registry no.: DRKS00005126)

Methods:

Aim of the trial was to show efficacy and safety of a 2^nd alloHCT from an unrelated donor after a uniform conditioning with Treosulfan 3x12gm/m2, Fludarabin 3x30mg/m2, and Thiotepa 3x 5mg/kg (TFT), and GvHD prophylaxis with cyclosporine A (CyA) /Mycophenolate and ATG-F (Neovii) 3x10mg/kg. Eligible were adult patients with AML, ECOG ≤ 2, with sensitive or refractory hematologic relapse (≥ 20% blasts) > 6 months after a prior allogeneic HCT , including secondary (s) and/ or tAML. CR prior 2^nd HCT was no prerequisite.

The primary endpoint of the study was disease-free survival (DFS) defined as being alive and free of disease at 1 year post 2^nd HCT. Secondary endpoints were relapse, relapse mortality (RM), NRM, overall survival (OS), acute GvHD, chronic GvHD, engraftment, and adverse events. According to the Fleming one-stage design, 50 evaluable patients had to be included. If 16 or more patients were alive and free of disease at 1 year post 2^nd SCT, the regimen could be considered as successful for evaluation in further trials. With this decision rule, it can be shown at one-sided α=0.1 that the probability of DFS at 1 year post 2^nd HCT is higher than 23% with a power of 90%, when it is at least 40%. This is a first analysis of the study covering the first year after 2^nd HCT of each patient. The analysis is based on the full analysis set, which includes all patients, for whom the conditioning regimen TFT and the GvHD prophylaxis regimen CyA, MPA/MMF, ATG-F has started, and for whom allogeneic HCT from an unrelated donor has been performed.

Results: Fifty-two patients were registered for the study from 25^th March 2014 up to 10^th March 2017 from 9 German centres. The full analysis set includes 50 patients (median age 53.5 years). ECOG was median 1. Donors for 1^st allo HCT had been related (n=11 (22.0%)) or unrelated (n=39 (78.0%)) (n=48 PBSCT, n=2 bone marrow). Conditioning for 1^st HCT was myeloablative in 23 (46.0%) patients. After 1^st HCT, the rate of acute GvHD I-IV was 34.0%, and of chronic GvHD was 40.0%. Median time from 1^st HCT to relapse was 17.2 months and from relapse after 1^st HCT to 2^nd HCT 2.5 months. Thirty-six (72%) patients had received induction chemotherapy for relapse prior to 2^nd HCT, 11 (22.0%) patients had received azacytidine or decitabine, and 11 (22.0%) had received donor lymphocyte infusions (DLI). Remission status prior to 2^nd HCT was complete remission in 16 (32.0%) patients, chemo-refractory relapse in 33 (66.0%), one patient was in partial remission.

With regard to the primary endpoint, 23 (46%, 95%-CI (31.8-60.7%) of the patients were alive and free of relapse at 1 year after 2^nd SCT. With regard to the secondary endpoints at 1 year, the cumulative incidence of relapse (95%-CI) was 26 (17-42)%, 9/50 patients (18 (10-33)%) died after relapse of AML, NRM was 14/50 patients (28 (18-44)%, cause: infection n=6, infection after aGvHD n=6, PTLD n=2), OS was 54 (39-66)%. Four patients are alive after relapse. aGvHD rate was 54 (42-70)%, aGvHD III-IV 26 (16-42)%, cGvHD 26 (16-42)%, extensive cGvHD 20 (12-35)%, engraftment rate with ANC > 1.0 x 10³/µl was 92 (85-100)%, with platelets > 20 x 10³/µl was 80 (70-92)%, and with platelets > 100 x 10³/µl was 66 (54-81)%. After 1 year, 4 patients had received DLI for prophylaxis, and 5 patients for mixed chimerism as per protocol.

Conclusion:

Second alloHCT with an ablative double alkylator containing conditioning regimen with Thiotepa, Fludarabine, and Treosulfan is feasible and can result in sustained disease control in patients with AML relapse after a first alloHCT, and therefore seems to be a valid option in this otherwise detrimental setting.