Feasibility and Outcome of Sequential Therapy in the Setting of HLA-Haploidentical Transplantation Utilizing Reduced-Intensity Conditioning, T-Cell-Replete Grafts and Post-Transplantation High-Dose Cyclophosphamide in the Treatment of High-Risk AML and MDS

Hematopoietic stem cell transplantation is a potent curative treatment option for patients suffering from high-risk AML and MDS. However, not in all of our patients a suitable HLA-matched donor could be identified in time.

To evaluate the feasibility and outcome of T-cell-replete (TCR) HLA-haploidentical hematopoietic stem cell transplantation (haplo-HSCT) utilizing post-transplantation high-dose cyclophosphamide (PTCY) in the context of sequential therapy in patients with high-risk, relapsed/refractory and progressive AML and MDS, we retrospectively analyzed the course of 64 patients (AML n=61, MDS n=3; median age: 50 years; male n=30) transplanted between 2009 and 2015 at our center. Disease was advanced in 55 patients including 24 patients with relapse after a first allogeneic transplantation. 54 patients presented with active disease at time of haplo-HSCT. Donors (median age: 38 years; male n=30) were fully HLA-haplotype mismatched (5/10 HLA-loci) in 48 cases (77%). All patients received sequential therapy prior to haplo-HSCT combining cytoreductive chemotherapy (clofarabine n=34; FLAMSA n=25; FLAG-Ida n=2; others n=3) and reduced-intensity conditioning (RIC) which was started after three days of rest thereafter. Conditioning was drug-based in 42 patients receiving fludarabine, cyclophosphamide (CY) and melphalan (110 mg/m²) while it was TBI-based in the others (n=22) consisting of fludarabine and CY plus either 4 (n=20) or 2 Gy TBI (n=2). Post-grafting immunosuppression was high-dose CY given on day + 3 and +4, tacrolimus and mycophenolate mofetil (both started on day +5) in all patients. Unstimulated bone marrow was the graft source in 37 patients.

One graft rejection was observed after conditioning with 2 Gy TBI. Neutrophil/platelet engraftment was achieved in 49/58 evaluable patients at a median of 16 (range: 14-27) and 20 (range: 13-74) days, respectively. Acute GvHD grade II-IV occurred in 19 patients (30 %) while it was severe (grade III-IV) in only 3 (5 %). Chronic GvHD was most frequently mild (n=9) to moderate (n=8); one severe chronic GvHD occurred. Severe (grade III-IV) mucositis, hemorrhagic cystitis and hand-foot syndrome/rash were observed in 10, 5 and 2 patients, respectively; no patient developed VOD. Kidney failure requiring hemodialysis occurred in 7 patients. CMV reactivation was observed in 28 of 47 patients at risk (59 %) and EBV in 3, while only one patient developed CMV disease (pneumonia) and no patient developed PTLD. Probable or proven (n=2) invasive aspergillosis was diagnosed in 10 patients. One-year non-relapse mortality was 27.5 % (95% CI 17-41). After a median follow up of 21 months (range: 3-64), estimated one-year overall survival (OS) was 55 % (95 % CI 41-66), and one-year disease-free survival (DFS) was 43 % (95 % CI 26-51). At two-years after sequential haplo-HSCT OS and DFS were both 39 % (95 % CI 26-51).

In summary, sequential therapy in the setting of RIC-TCR haplo-HSCT using PTCY is well tolerated with low rates of GvHD and acceptable NRM in patients with high-risk AML and MDS, while providing an effective anti-leukemic activity in advanced disease. Results are comparable to data of a historical cohort of patients with high-risk AML and MDS undergoing sequential therapy using the FLAMSA-RIC protocol in a HLA-matched setting, as reported by our group previously. Thus, we suggest that donor availability can be expanded in patients with high-risk and advanced AML/MDS who lack a conventional donor or need promptly access to a donor due to aggressive disease.