Intensification Of Treosulfan and Fludarabine-Based Conditioning With 4 Gy TBI For Allogeneic Stem Cell Transplantation In Patients With Hematological Malignancies

Abstract

BackgroundHematopoietic stem cell transplantation (HSCT) is so far the only potentially curative option for patients with aggressive hematological malignancies. Since many patients cannot find a suitable HLA-identical (related or unrelated) donor, it is essential to analyze the safety of alternative graft sources, such as haploidentical donors. The use of reduced-intensity conditioning regimens in patients with an advanced disease did not improve the outcome because of a higher incidence of relapse. Here we present an interim analysis of this phase II prospective clinical trial that investigates a new conditioning regimen for allogeneic HSCT based on a well established one (Fludarabine and Treosulfan) intensified adding 4 Gy TBI with the aim of maintaining an acceptable toxicity profile while reducing the incidence of relapse in high-risk patients (TrRaMM4Gy study, Eudract 2011-001534-42). Patients and MethodsNinety-six patients underwent allogeneic HSCT for AML (n=46), ALL (n=13), acute biphenotypic leukaemia (n=2), HD (n=8), NHL (n=7), MDS (n=6), Myelofibrosis (n=5), MM (n=4), CML (n=2), CMML (n=2) or CLL (n=1). The median age was 45 years (range 17-67). At the time of the transplantation most patients (n=76) were in advanced disease phase, while the remaining 20 patients were in early phase. Twenty-eight patients were enrolled for relapse after a previous allogeneic HSCT. Median time from diagnosis to transplantation was 443 days (range 56-5249). Median comorbodity index score (according to Sorror criteria) was 2 (range: 0-8). Sixty-two patients received the graft from haploidentical donors, 17 from MUD, 14 from HLA-identical sibling and 3 patients from a single cord blood unit. Median number of CD34+ and CD3+ cells/kg were 6.46 millions and 241 millions respectively. The conditioning regimen included Treosulfan (14 g/m2 for 3 days), Fludarabine (30 mg/m2 for 5 days) and 4 Gy TBI split in 2 fractions. Patients receiving HSCT from haploidentical donor (arm A) were also treated with ATG-Fresenius (10 mg/kg for 3 days) and Rituximab (200 mg/m2 in single dose), patients receiving HSCT from HLA-identical sibling, MUD or single cord blood unit and in early disease status (arm B) were also treated with ATG-Fresenius (5 mg/kg for 3 days) and Rituximab (200 mg/m2 in single dose), while patients receiving HSCT from HLA-identical sibling, MUD or single cord blood unit and in advanced disease status (arm C) did not receive any ATG or Rituximab. GvHD prophylaxis consisted of Sirolimus (target concentration 8-15 ng/ml, till day +60) and Mycophenolate Mofetil (10 mg/kg tid till day +30). ResultsNeutrophils engraftment was reported at a median time of 17 days (range: 9-46) in all 91 evaluable patients; five patients died of transplant-related causes (n=4) or disease progression (n=1) before day +15. Platelets engraftment occurred after a median time of 17 days (range: 4-153) in 75 patients (82%). At the first marrow evaluation on day +30 all the evaluable patients showed full donor chimerism, 77 patients (90%) were in complete remission and 9 were in stable disease or progression. After a median follow-up of 425 days, cumulative incidence of grade 2-4 and 3-4 aGvHD were 40% and 26% respectively. Cumulative incidence of moderate or severe cGvHD (according to NIH criteria) was 57%. Cumulative incidence of NRM at day +100 and +365 were 23% and 36% respectively; cumulative incidence of relapse at day +365 was 33%. Overall survival and progression-free survival 1 year after HSCT were 51% and 43% respectively. EBV reactivation occurred in 7 patients, but no PTLD was observed. ConclusionsThe new conditioning regimen we investigated proved to be feasible in this high-risk population. We did not observe any major difference in terms of NRM and aGvHD compared to the data from our previous trial (TrRaMM, Eudract 2007-5477-54) with similar inclusion criteria and patient characteristics, but a reduced-intensity conditioning regimen based on Treosulfan and Fludarabine. Interestingly, we observed a trend of lower relapse incidence, resulting in a better OS and PFS. The low toxicity profile of Treosulfan and Fludarabine should be considered as a myeloablative platform for further intensification; more studies are warranted in order to find the association for a conditioning regimen endowed with a low toxicity profile but still a strong anti-tumor activity. Disclosures:Bonini:MolMed SpA: Consultancy.