Treatment of AML in First Remission (CR1) with Allogeneic Hematopoietic Stem Cell Transplantation (HSCT) Using Unrelated Donors (UD)

Introduction/Methods: A key component of improving the success rate of allogeneic HSCT for AML in CR1 is to reduce non-relapse mortality (NRM), which, if excessive, will deny the benefit of the graft-versus-leukemia effect. UD HSCT has traditionally been associated with high NRM rates. We reported previously on the significant reduction of NRM using the conditioning regimen of fludarabine 40mg/m², IV busulfan 130 mg/m² for 4 days and thymoglobulin. Here we analyze the outcomes of all patients (n=37) with AML in CR1 treated with this regimen and UD HSCT in our institution from January 2002 to December 2007. High-resolution allele level HLA typing was performed for all donor-recipient pairs for HLA-A, -B, -C, DRB1 and DQB1; up to one mismatch was allowed (9/10).  Median follow up is 30 months (range, 9-72).

Results: Median age was 48 years (range, 13-68); 30% (n=11) were older than 54 years and 51% (n=19) were male. Eleven patients (30%) had secondary AML. Prognostic cytogenetics classification was poor and intermediate in 53% and 47% of the cases, respectively. Stem cell source was bone marrow (BM) in 68% (n=25) and peripheral blood (PB) in 32% (n=12). Graft-versus-host disease (GVHD) prophylaxis was tacrolimus and mini-methotrexate with and without pentostatin (1 or 1.5 mg/m² on days 8, 15, 22 and 30) in 46% (n=17) and 54% of the cases (n=20), respectively. Donor-recipient HLA match was 9/10 and 10/10 in 14% and 86% of the cases. Median infused total nucleated and CD34+ cells was 3.72 x 10⁸ (range, 0.57 – 11.78) and 3.77 x 10⁶ (range, 0.45 – 12.4), respectively. Median time to neutrophil and platelet engraftment was 12 days (range, 8-18) and 14 days (range, 8 -101), respectively. All but one patient engrafted. Grade II-IV acute (a) GVHD rate was 13% (n=2) and 50% (n=10) for patients that received and not received pentostatin-based prophylaxis. Grade III-IV aGVHD rate was 0% versus 15% (n=3) for patients receiving and not receiving pentostatin. Chronic GVHD was diagnosed in 55% (n=18) of all patients (extensive in 10). 100-day and 3-year NRM rate was 11% (n=4) and 20% (n=4), respectively, and was due to engraftment failure (n=1) and aGVHD (n=3). Eight patients (22%) have relapsed, and 8 (22%) have died (4 of relapse, and 4 of NRM causes). Relapse rate was 18% (3/17) and 25% (5/20) for patients that received and not received pentostatin as part of GVHD prophylaxis. Actuarial 3-year event-free and overall survival (figure) is 68% and 78%, respectively.  Actuarial 3-year overall survival for patients receiving BM and PB is 80% and 75%, respectively.

Conclusion: Long-term disease control can be achieved in a significant fraction of high-risk AML patients undergoing UD transplants as described in this abstract. Use of pentostatin in this context deserves further prospective evaluation.