Leukemia (2000) 14, 2059–2063  2000 Macmillan Publishers Ltd All rights reserved 0887-6924/00 $15.00 www.nature.com/leu Prognosis of patients with a second relapse of acute myeloid leukemia B Stoiser 1 , P Kno ¨bl 1 , C Fonatsch 2 , OA Haas 3 , G Mitterbauer 4 , A Weltermann 1 , K Geissler 1 , P Valent 1 , W Sperr 1 , I Pabinger 1 , K Lechner 1 and U Jaeger 1 1 Department of Internal Medicine I (Division of Hematology and Hemostaseology), 2 Institute of Medical Biology, 3 Children’s Cancer Research Institute, St Anna Children’s Hospital, Vienna; and 4 Institute of Laboratory Medicine, Molecular Biology, University of Vienna, Austria Recurrence of the disease is the major problem in the treatment of acute myeloid leukemia (AML). The majority of patients who achieve a second remission will ultimately relapse. In this retro- spective single-center study, we have analyzed the outcome of patients with a second relapse and tried to define the prognos- tic factors in intensively treated patients. Of 534 patients with AML, 62 had a second relapse. Thirty-three received further intensive chemotherapy (CT). Eighteen patients (55%) achieved a third complete remission (CR). The early death (ED) rate was only 9%. The overall survival (OS) of treated vs untreated patients was 6.9 vs 1.3 months, respectively (P = 0.01). The major selection criteria for a third CT were a favourable (t(15;17),t(8;21),inv(16)) or normal karyotype, long (11 months) second CR (P  0.005) and no previous bone marrow transplantation (BMT)(P  0.01). Favorable or normal karyo- type, second CR 11 months, as well as no previous BMT (P  0.01) were associated with the achievement of a third CR. Favorable (P  0.005) or normal karyotype (P  0.01), as well as a second CR 11 months (P  0.005) were associated with prolonged survival after CT. The median OS for patients receiv- ing CT with favorable or normal cytogenetics, a second CR  11 months, but no previous BMT was 26.5 months. Five patients with favorable or normal karyotype achieved a fourth or fifth remission. We conclude that intensive CT is associated with a survival benefit and good quality of life if patients are properly selected. Leukemia (2000) 14, 2059–2063. Keywords: AML; second relapse; karyotype; prognostic factors Introduction With modern consolidation treatment about 60–80% of younger and 20–30% of older patients with de novo acute myeloid leukemia (AML) who achieve a first hematological remission enjoy a long-term remission and are possibly cured. 1,2 Patients who relapse after chemotherapy (CT) have a definite chance to achieve a second remission after treat- ment with intensive CT. 3–5 We and others 6,7 have shown that the probability of a second CR depends mainly on the dur- ation of the first remission. About 20 to 30% of younger patients with a second remission can be cured by allogeneic stem cell transplantation, but only a few (about 10%) by CT alone. 1,2,8,9 Thus, the vast majority of patients with a second remission will have a second relapse. The physician caring for such patients is faced with the problem that no data exist as to the risks and possible benefits of an intensive salvage CT. Therefore, we have analyzed the outcome of patients with a second relapse who were treated with intensive CT or only supportive care and tried to define the prognostic factors in intensively treated patients. Correspondence: U Jaeger, Department of Internal Medicine I, Div of Hematology and Hemostaseology, Wa ¨hringer Gu ¨ rtel 18–20, A-1090 Vienna, Austria; Received 23 May 2000; accepted 28 July 2000 Patients and methods A total of 534 patients with the diagnosis of AML were admit- ted to our institution from 1979 to 1999. Of those patients, 497 were treated with intensive CT, 317/497 (64%) achieved a CR, which lasted for a median time of 8.24 months. Of 317, 196 (62%) patients relapsed. Of these 196 patients, 133 were again treated with CT and 82/133 (62%) achieved a second remission, which had a median duration of 6.6 months. Of 82 patients, 62 in second CR had a second relapse, 33 of those 62 patients underwent a third induction CT. Overall, 108/497 (22%) of patients initially treated underwent autolog- ous or allogeneic bone marrow transplantation in the course of their disease. AML was classified according to the FAB criteria. 10 Karyo- types were available in 59 patients. Immunophenotypic analy- sis was performed according to standard methods. 11–14 For induction CT, the following chemotherapeutic regimens were used: DAV (daunorubicin: 45 mg/m 2 day 1–3; etoposide: 100 mg/m 2 days 1–5; cytosine arabinoside (Ara-C) 2 × 100 mg/m 2 days 1–7); 15 3+ 7 protocol (daunorubicin: 45 mg/m 2 days 1–3; Ara-C 2 × 100 mg/m 2 days 1–7); AIDA protocol (all-trans retinoic acid (ATRA) 45 mg/m 2 /d, idarubicin 12 mg/m 2 days 2, 4, 6, 8). 16 For consolidation CT: intermittent HiDAC (Ara-C 2 × 3 g/m 2 days 1, 3, 5); 17 HAM (Ara-C 2 × 1 g/m 2 days 1–4; mitoxantrone 10 mg/m 2 days 2–5); FLAG (fludarabine 30 mg/m 2 days 1–5; Ara-C 2 g/m 2 days 1–5; G- CSF 300 g/m 2 from day 6); 18 MIDAC (Ara-C 2 × 1 g/m 2 days 1–4; mitoxantrone: 12 mg/m 2 days 3–5). Statistical analysis The Kaplan–Meier technique 19 was used to analyze the prob- ability of overall survival (OS) and continuous complete remission (CCR). Survival was calculated from the time of second relapse until relapse or death from any cause, CCR from the time of CR3 until relapse. For the Kaplan–Meier analysis, all patients including those with BMT were con- sidered. Statistical analysis were performed using parametric (Student’s t-test) or non-parametric tests (Mann–Whitney rank sum test) on the basis of distribution of the data. Differences between means were considered as significant at P  0.05. Data analysis was performed using WinStat for Windows (1996 Version 3.0, G Greulich Software, Germany). Results Demographic and hematological characteristics of patients with a second relapse Sixty-two patients with de novo AML had a second relapse. The median age of these patients was 48 years (range 16–72)