Haploidentical Hematopoietic Cell Transplantation Using G-CSF Mobilized T-Cell Replete Grafts for for Acute Leukemia and MDS

The use of T-cell replete grafts from MHC-haploidentical donors with post-transplant cyclophosphamide (PTCy) to selectively deplete alloreactive T cells is efficacious in patients who do not have ready access to an HLA identical donor. The Johns Hopkins group that pioneered this regimen used non-myeloablative (NMA) conditioning and bone marrow (BM) grafts to help mitigate the risk of GvHD. Their results showed low rates of GvHD and treatment related mortality but outcomes were limited by a high relapse rate. However, for patients with active disease or other high risk characteristics new approaches are needed to help mitigate post-transplant relapse. For other transplant modalities, peripheral blood stem cell (PBSC) grafts are associated with lower relapse rates, faster engraftment, and improved overall survival when compared to BM derived grafts. To date, there are few small studies evaluating the outcomes G-CSF mobilized PBSC grafts for haplo-HCT with PTCy.

To report our experience with T cell replete PBSC grafts with PTCy for haplo-HCT, we retrospectively analyzed the outcomes of patients with leukemia and MDS who underwent this transplant regimen. Between 2009 and 2015, 124 patients were transplanted at Washington University School of Medicine with this transplant platform and their outcomes were measured by retrospective chart review through June 2016. Donors were mobilized with G-CSF 10 mcg/kg daily for 5 days prior to beginning pheresis and no grafts were subjected to ex vivo T-cell depletion. Myeloablative conditioning was used for 54 patients and NMA regimens were used for 70 patients, after which, patients were infused with a median of 5.0 x 10⁶ CD34+ cells/kg and 18.0 x 10⁷ CD3+ cells/kg. For GvHD prophylaxis, all patients received PTCy 50 mg/kg on days +3 and +4, tacrolimus from day +5 to 180, and mycophenolate mofetil from day +5 to 35, as previously described.

The median age at time of haplo-HCT was 53 years (range 19-73). Ninety three patients were transplanted for AML, 16 for ALL, and 15 for MDS. This was a high risk population with a median Hematopoietic Cell Transplantation-specific Comorbidity Index (HCT-CI) score of 3.5 (≥3 indicates high risk). Thirty two patients had prior transplants and 49 patients had active disease prior to their transplant.

Median follow up for surviving patients was 553 days. Neutrophil count recovery occurred after a median of 17 days and platelet recovery after 29 days. 74 patient deaths occurred by June 2016. Relapse accounted for 36 deaths, infection for 17, acute GvHD for 9, graft failure for 3, and 9 patients died from other causes. The 1 year overall survival (OS) was 47.5% (95% CI 38.2-56.2%) and 1 year event free survival was 40.2 (31.2-48.9%). For patients transplanted in remission, 1 year OS was 50.8% (38.7-61.6%), while for patients transplanted with active disease 1 year survival was 42.9% (28.4-56.5%). Treatment related mortality at 6 months and 1 year was 20.2% (13.7-27.8%) and 31.4% (23.1-40.0%). The relapse rate at 1 year was 39.5% (28.7-50.1%). For patients transplanted in remission, the 1 year relapse rate was 31.4% (18.7-44.9%) and for those transplanted with active disease it was 50.7% (32.5-66.3%). Univariate analysis showed increased risk of relapse when going into transplant with active disease or receiving NMA regimens.

The 180 day cumulative incidence of grades II-IV acute GvHD was 39.3% (30.0-48.9%) and grade III-IV acute GvHD was 10.8% (3.1-23.7%). 1 year cumulative incidence of chronic GvHD was 49.0 (35.8-60.9%) and severe chronic GvHD was very low at 4.0% (1.0-10.4%).

In conclusion, using G-CSF mobilized grafts from PBSC for haplo HCT with PTCy is feasible for patients with leukemia and MDS who do not have ready access to an HLA identical donor. The use of PBSC grafts with higher numbers of CD3+ cells led to rapid engraftment, low rates of graft failure, and acceptable rates of acute and chronic GvHD. The 1 year OS was almost 1 year, which is encouraging considering most patients had a HCT CI score > 3 and many had active disease at transplant or already failed a prior transplant. Disease relapse post-transplant remains the primary cause of post-transplant mortality for these leukemia patients. In this cohort relapse was associated with having active disease at time of transplant and receiving NMA conditioning. Our results show using PBSC may be a valid alternative to bone marrow grafts for haplo HCT with PTCy.