Abstract
In higher risk myelofibrosis patients, the only treatment able to improve survival and cure the disease is allogeneic hematopoietic stem cell (HSCT). However, not all potential candidates have a donor and it has been reported that transplantation from an HLA-mismatched donor gives disappointing results as compared to transplantation from an HLA matched donor. In this phase 2 trial (registered at clinicaltrial.com as NCT04728490), transplantation from an haplo-identical donor was proposed in patients without an HLA matched donor using fludarabine 30mg/m2/day (D) for 5 days – treosulfan 10g/m2/D for 3 days– thiotepa 5 mg/kg for one day– post-translant cyclophosphamide 50 mg/kg ion day +3 and day +5 (PTCY) -ciclosporine and mycophenolate mofetil plateform. The trial was sponsored by Hôpital Saint-Louis, APHP, Paris, France, with the financial support of MEDAC pharmaceutical and the scientific support and network of the French Society of Cellular Therapy (SFGM-TC).
We conducted a single-arm phase 2 clinical trial to demonstrate that using an haplo-identical related donor will increase the relapse and rejection-free survival post-HSCT as compared to an historical cohort of patients received 9/10 HLA matched unrelated donor. Our hypothesis was that 1-year relapse and rejection-free survival could reach 55% instead of 30%. Using a 2-sided, one-sample log-rank test, with 90% statistical power and a 5% significance level, 28 patients had to be included. Secondary objectives were incidence of acute and chronic GVHD, hematological recovery, relapse, non-relapse mortality, severe infections and overall survival. Inclusion criteria were age 18-70 years, myelofibrosis, and at least two of the following characteristics: constitutional symptoms, hemoglobin level < 10g/dl, thrombocytopenia < 100G/L, peripheral blast > 1%, white blood cell (WBC) count > 25 G/L, poor cytogenetics (+8; -7/7q-;i(17q);-5;5q-;12p-;inv(3);11q23, and ECOG < 3. Follow-up was scheduled for 12 months and amended later to 24 months after transplantation. Results presented here are at one year endpoint.
We included 29 patients in 15 French centres, one of whom did not receive the transplantation due to a fatal haemorrhage. Further analyses were based on the 28 transplanted patients; all patients are followed > 12 months, 3 patients are still followed-up on July 2025 to reach the 24th month. Median age at inclusion was 65 years (IQR: 58-67), 19 (68%) were men, 11 (39%) had general symptoms, 10 (36%) were transfusion dependant, 21 (75%) had hemoglobin level < 10 g/dl, 13 (46%) had platelet count < 100 G/L, 13 (46%) had peripheral blast >1%, 7 (25%) had WBC count >25 G/L, and 3/16 had a complex cytogenetic. 15 patients had grade 2 fibrosis, 12 grade 3 (and one unknown). Hematopoietic stem Cell Transplantation-Comorbidity Index was 0-1 in 20 (71%) patients and 2 or more in 8 (29%) patients. Among the 25 patients with NGS tested, 15(62%), 5 (21%) and one (4%) had JAK2V617F, CALR, and MPL mutation, respectively. Other genes frequently altered were ASXL1 (71%), TET2 (29%), EZH2 (25%) and RUNX1 (21%). Eighteen patients had primary myelofibrosis, 10 had secondary myelofibrosis. All the 28 patients received peripheral blood stem cells from their donor and cells were cryopreserved in 10 (36%) in the context of COVID-19 pandemia.
One-year relapse and rejection-free survival was 64.3% (95%CI: 48.8-84.7), significantly above the historical rate of 35% in transplantation from a 9/10 unrelated donor. Cumulative incidence of grade 2-4 and grade 3-4 acute GVHD were 64.3% (95%CI: 43-79.4) and 25% (95%CI:10.8-42.2); 11 patients developed chronic GVHD, with 1-year cumulative incidence estimated at 39.3% (95%CI, 21.1-57.1). Cumulative incidence of neutrophils recovery on day 60 was 85.7% (95%CI: 63.9-94.8), and that of platelet recovery was 67.9% (46.4-82.2). One patient relapsed after HSCT, and 10 patients died due to viral pneumonia (SARS-COV-2, n=1 or influenza, n=1), multi-organ failure without documented infection (n=2), graft rejection/non engraftment (n=2), and acute GVHD (n=4). One-year overall survival was 67.9% (95%CI: 52.6 – 87.6). One-year non-relapse mortality (NRM) was 32.1% (95%CI: 15.8 – 49.7).
In conclusion, the primary endpoint of one-year relapse and rejection -free survival was achieved. However, the NRM is the first cause of failure, mainly due to acute GVHD, MOF and infections justifying further investigations to reduce early toxicity.
