Infection Is the Major Cause of Mortality after Nonmyeloablative Unrelated Bone Marrow Transplantation with Alemtuzumab, Fludarabine and Melphalan.

Abstract

Nonmyeloablative (NM) regimens for allogeneic (allo) hematopoietic cell transplantation (HCT) may be effective and associated with a low incidence of acute graft-versus-host disease (GVHD) in high-risk patients (pts) with hematologic malignancies. We report the results of NM bone marrow transplantation (BMT) from HLA-matched volunteer unrelated donors (VUDs) in 14 pts (median age 45 yr; range, 19–58 yr) with relapsed and/or refractory hematologic malignancies (1 ALL, 3 AML, 3 MDS/AML, 3 Hodgkin's disease [HD], 2 multiple myeloma, 2 NHL). Thirteen pts had previous HCT: 10 autologous (auto), 1 related allo, 1 VUD, 1 auto and VUD. The NM regimen included alemtuzumab (20 mg/day x 5, days −8 through −4), fludarabine (30 mg/m2/day x 5, days −7 through −3) and melphalan (140 mg/m2 on day −2), as described by Kottaridis PD et al, Blood 2000; 96: 2419 and Chakraverty R et al, Blood 2002; 99: 1071. All pts received single-agent cyclosporine for prophylaxis of acute graft-versus-host disease (GVHD). Median doses of CD34+ and CD3+ cells were 2.79 (range, 1.35–4.87) x 106/kg and 2.8 (range, 1.99–6.39) x 107/kg, respectively. In 12 evaluable pts, median times to absolute neutrophils >0.5 x 109/L and platelets >20 x 109/L were 15 (range, 9–41) and 21 days (range, 15–41), respectively, after NMBMT. Two pts had primary graft failure (actuarial probability 16.4%; 95% confidence interval [CI], 0–37.6%). Two pts developed acute GVHD (1 grade I, 1 grade II) at 19 and 76 days, respectively, after NMBMT. The actuarial probability of all grades of acute GVHD is 16.1% (95% CI, 0–36.7%) and of grade II or greater GVHD is 9.1 % (95% CI, 0–26.2%). Two pts (1 with grade II acute GVHD) developed extensive chronic GVHD. Five pts developed cytomegalovirus (CMV) reactivation, for an actuarial probability of CMV reactivation of 58.5% (95% CI, 19.9–97.1%). One pt with CMV reactivation died with CMV infection at 122 days after NMBMT. Six pts died with other infections at a median of 74 days (range, 52–209) after NMBMT: 2 adenovirus (1 associated with graft failure), 2 toxoplasmosis, 1 Aspergillus (associated with graft failure), 1 Pseudomonas (associated with chronic GVHD). One pt died with pulmonary failure and chronic GVHD 267 days after NMBMT. Actuarial nonrelapse mortality (NRM) is 64.3% (95% CI, 35.5–93.1), and infection-associated mortality is 52.4% (95% CI 25.0–79.8%). Three pts (2 AML, 1 HD) relapsed at 181, 182 and 202 days, respectively, after NMBMT; actuarial relapse rate is 46.4% (95% CI, 26.4–66.4%). Three pts (1 ALL, 1 AML, 1 NHL) are alive without relapse at 164+, 184+ and 843+ days, respectively, after NMBMT; actuarial event-free survival (EFS) is 10.2% (95% CI, 0–28.8%). We conclude that opportunistic infection is the major cause of NRM after VUD NMBMT with a preparative regimen of alemtuzumab, fludarabine and melphalan and contributes significantly to the poor EFS in pts with relapsed/refractory hematologic malignancies thus treated. Efforts to improve immune reconstitution and infection prophylaxis after NMBMT with this regimen are warranted.