Bone Marrow Transplantation, (1998) 21, 673–678  1998 Stockton Press All rights reserved 0268–3369/98 $12.00 Leukemia-free survival and mortality in patients with refractory or relapsed acute leukemia given marrow transplants from sibling and unrelated donors HT Greinix 1 , E Reiter 1 , F Keil 1 , G Fischer 2 , K Lechner 3 , K Dieckmann 4 , G Leitner 2 , A Schulenburg 1 , P Hoecker 2 , OA Haas 5 , P Knoebl 3 , C Mannhalter 6 , C Fonatsch 7 , W Hinterberger 1 and P Kalhs 1 1 Department of Medicine I, Bone Marrow Transplantation Unit; 2 Department of Blood Group Serology and Transfusion Medicine; 3 Department of Medicine I, Division of Hematology and Hemostaseology; 4 Department of Radiotherapy, University of Vienna; 5 Children’s Cancer Research Institute, St Anna Kinderspital, Vienna; 6 Clinical Institute for Medical and Chemical Laboratory Diagnostic; 7 Department of Medical Biology, University of Vienna, Vienna, Austria Summary: Between April 1982 and February 1997 39 patients (24 male, 15 female) with refractory acute leukemia and a median age of 31 years (19–51 years) received allogeneic marrow grafts from an HLA-identical sibling (n = 27), HLA-identical unrelated donor (MUD; n = 10) or 1- antigen mismatched unrelated donor (n = 2). Twenty- eight patients had acute myelogenous leukemia and 11 acute lymphoblastic leukemia. For conditioning most patients received total body irradiation combined with cyclophosphamide (n = 23) or etoposide (n = 7). For graft-versus-host disease prophylaxis patients received cyclosporin A (CsA) and methotrexate (MTX) (n = 20), MTX alone (n = 3), CsA and methylprednisone (n = 6), or CsA alone (n = 10), respectively. As of June 1997 probability of leukemia-free survival projected to 3 years after BMT was 14% for patients given sibling marrow grafts and 28% after MUD transplantation. Transplant-related mortality projected to 3 years was 32% after sibling and 37% after MUD marrow grafting. Although not significantly different, probability of relapse projected to 3 years after BMT was lower after MUD at 56% compared to 78% with sibling BMT. Thus, high-dose chemo/radiotherapy followed by allo- geneic marrow infusion has a curative potential for patients with refractory leukemia and offers the chance of long-term disease-free survival for some patients. Keywords: poor risk acute leukemia; sibling and unre- lated marrow transplantation With current chemotherapies 80% of patients with acute myelogenous leukemia (AML) and more than 80% of patients with acute lymphoblastic leukemia (ALL) below the age of 50 years achieve complete hematological remission (CR). 1,2 Patients who are refractory to induction Correspondence: Dr HT Greinix, AKH Wien, Klinik fuer Innere I, Kno- chenmarktransplantation, Waehringer Guertel 18–20, A-1090 Wien, Austria Received 7 August 1997; accepted 24 October 1997 chemotherapy or relapse early have a poor prognosis with death due to disease or complications of therapy within a period of weeks to months. 3 Even when remissions are achieved with aggressive salvage regimens, they tend to be short-lived. 3 High-dose chemoradiotherapy followed by allogeneic marrow infusion has been used to overcome resistance of the refractory leukemic clone and has resulted in long-term survival of some patients. 4–6 Unfortunately, less than 30% of eligible patients have a suitable related donor. Bone mar- row transplantation (BMT) from a human leukocyte antigen (HLA) compatible unrelated donor (MUD) has become a feasible and effective treatment for patients who could benefit from BMT but who lack an HLA-matched sibling. 7–9 Here, we present our experience in patients with primary refractory and refractory relapse of acute leukemia and compare leukemia-free survival and transplant-related mortality in patients given marrow grafts from sibling or unrelated donors. Patients and methods Patients Between April 1982 and February 1997 39 consecutive patients with refractory acute leukemia, 24 males and 15 females, aged 19–51 years (median 31 years), underwent allogeneic marrow transplantation. These patients had either never achieved a complete remission (n = 17) or had relapsed and were unable to attain CR despite induction chemotherapy (n = 22). Their pre-transplant characteristics are shown in Table 1. The conditioning regimens used for transplantation and protocols for GVHD prophylaxis were those used at our institution during the time period patients were transplanted. All patients received one to two doses of methotrexate (12 mg) intrathecally as part of the con- ditioning regimen. Marrow donors were HLA-identical siblings in 27 patients, HLA-identical unrelated donors in 10 patients, and 1-antigen mismatched unrelated donors in two patients. In 25 patients ABO incompatibility and in 14 patients a sex