To assess the place of allogeneic hematopoietic stem cell transplantation (HSCT) in the advanced stage of acute myeloid leukemia (AML), we retrospectively analyzed 379 consecutive patients who underwent allogeneic HSCT for advanced AML. The median follow-up of the entire cohort was 7.5 years. Sixty-nine patients (18%) were transplanted with primary resistant disease. Three hundred and ten (82%) were relapsed patients, 94 (30%) of whom were in untreated relapse, 67 (22%) in refractory relapse and 149 (48%) in 2nd or 3rd complete remission at time of transplantation. The 5-year probabilities of overall survival (OS), disease-free survival (DFS), and transplant-related mortality (TRM) were 22 ± 4% 20 ± 4%, 45 ± 6%, respectively. In multivariate analysis, we demonstrated the favorable impact on OS, DFS and TRM of two factors over which we have no control (age <15 years, complete remission achievement) and three factors over which we have some control (female donor, acute and chronic graft-versus-host disease). The results of this study suggest that the graft-versus-leukemia effect is important in advanced AML and that new HSCT modalities are needed for some patients with this indication. Bone Marrow Transplantation (2000) 26, 1157–1163.
The outcome of treatment in younger patients with de novo acute myeloid leukemia (AML) has improved substantially over the past decade. Complete remission (CR) rates range from 60% to 80% with long-term survival in about 50% of cases. Advances can be attributed to several factors related to supportive care, which has allowed the safer use of more intensive chemotherapy schedules.12 Myeloablative therapy followed by allogeneic hematopoietic stem cell transplantation (HSCT) has also proven to be a very effective therapy.34 Despite these results, patients with AML can still fail to achieve CR or can relapse after first-line therapy within 2 years of CR achievement. These patients have a very poor prognosis and it is generally accepted that few cases can be cured at this stage using chemotherapy alone.5 Indeed, even though the best strategy regarding postremission therapy in first CR is still subject to controversy,6 allogeneic bone marrow transplantation has shown a potential for cure at time of relapse.78 Although allogeneic HSCT has frequently been indicated in recurrent or refractory AML, very few large series allowing prognostic factor analysis have been reported. In this retrospective study, we analyzed the impact of pre- and post-transplantation factors on the outcome of allogeneic transplants in 379 consecutive patients with advanced stage AML whose cases were reported to the Société Française de Greffe de Möelle (SFGM) registry from 1980 to 1993.
Patients and methods
Patients and disease characteristics
Between 1980 and 1993, 379 patients (212 males and 167 females) with advanced AML underwent allogeneic HSCT (Table 1). Sixty-nine patients presented with primary resistant disease (Pr) and 310 were in relapse after having achieved complete remission (CR) by chemotherapy alone. Among these 310 patients, 277 were in first relapse, 31 were in second relapse, and two were in third relapse. The pretransplantation stage of these 310 relapsed AML patients was as follows: untreated relapse (UR) in 94 cases, refractory relapse (RR) in 67 cases, and 2nd or 3rd CR in 149 cases. The median follow-up of the entire cohort was 7.5 years (range 2–15 years). Fifty-six patients received transplants between 1980 and 1984, 178 patients between 1985 and 1989 and 145 patients between 1990 and 1993. The median age was 28 years (range 2 months to 58 years). Three hundred and thirty-two patients were morphologically classified according to the FAB classification.9 There were 63 M1, 103 M2, 31 M3, 57 M4, 62 M5 and 16 M6. All data were prospectively reported to SFGM registry from 33 centers and were completed by additional data obtained from questionnaires sent to each center.
Three hundred and fifty patients received transplants from related donors: 313 from HLA-identical sibling donors, five from twins, one from a phenotypically identical donor and 31 from HLA mismatched related donors (Table 1). Twenty-nine patients received transplants from unrelated donors: 28 from HLA A, B and DRB1 identical donors and one from a one-antigen mismatched donor. Two hundred and ninety-nine patients (79%) received a conditioning regimen based on total body irradiation (TBI). TBI was fractionated in 134 patients. Eighty patients (21%) received an intensive chemotherapy-based conditioning regimen. Two hundred and eighty-six patients (75%) received a standard conditioning regimen of cyclophosphamide associated with either busulfan or TBI, while 93 patients (25%) received an intensified preparatory regimen. Graft-versus-host disease (GVHD) prophylaxis was administered in 363 patients (96%): 32 patients (9%) received cyclosporine alone, 56 patients (15%) received methotrexate alone, and 236 patients (65%) were treated with a combination of cyclosporine and methotrexate. Thirty-nine (11%) patients received a T cell-depleted graft.
The primary endpoints of this study were overall survival (OS), disease-free survival (DFS) and transplant-related mortality (TRM). Survival duration was measured from the time of transplantation to the time of death or last follow-up. We calculated DFS in patients who achieved a CR from the time of CR to the time of relapse, death from any cause, or last follow-up. TRM was defined as mortality exclusively related to the transplant. Patients surviving more than 21 days after transplantation with evidence of engraftment were considered at risk for acute GVHD. We diagnosed chronic GVHD as occurring in patients surviving more than 90 days after transplantation. Relapse was defined as a hematological or clinical recurrence at any site. We used the Kaplan–Meier product-limit estimate method to calculate OS, DFS and TRM probabilities.10 Survival curves were compared using the log-rank test.11 Parameters were analyzed for potential prognostic significance for CR achievement, TRM, DFS and OS. We analyzed separately three sets of factors: (1) factors over which we have absolutely no control (age, FAB classification, AML stage, duration of CR); (2) factors over which we have some control (donor sex, HLA compatibility, acute GVHD, and chronic GVHD); and (3) factors over which we have total control (T depletion and conditioning regimen).
In the analysis, qualitative variables were treated as dichotomous with cut-off points at the approximate median value, or were compared using a t-test. Prognostic factors for CR were studied using stepwise multiple logistic regression and prognostic factors for TRM, DFS, and OS were studied using the stepwise Cox proportional hazard model.12 All variables that were statistically predictive for survival rates and their interactions were proposed for entry in the stepwise multivariate model for prediction of the same endpoint. We tested how well the models fit using the likelihood ratio statistics. All calculations were performed using BMDP PC-90 statistical program (BMDP Statistical Software, Los Angeles, CA, USA).
One hundred and forty-nine patients underwent allogeneic HSCT after achieving a second or a third CR with chemotherapy. Allogeneic HSCT was also performed in 69 patients with primary resistant AML, and in 94 and 67 patients respectively with untreated or refractory relapse.
Transplantation allowed CR achievement in 53/69 patients (77%) with primary resistant (Pr) disease, 50/67 patients (75%) with RR, and in 75/94 patients (80%) with UR.
At a median follow-up date 7.5 years after transplant (range 2–15 years), 77 patients were still alive. The 5-year probabilities of OS, DFS, and TRM were 22 ± 4%, 20 ± 4%, 45 ± 6%, respectively. The 5-year probability of survival for patients who had received an allotransplant in CR (35 ± 8%) was significantly better (P = 0.0001) than for patients transplanted in Pr disease (13 ± 8%) or for patients undergoing transplantation in UR or RR (14 ± 6% and 11 ± 6% respectively) (Figure 1). The DFS was better when allogeneic HSCT (5-year probability 32 ± 6%) was performed in CR than when HSCT was performed in Pr, UR, or RR (9 ± 6%, 13 ± 6%, 11 ± 6%) (P = 0.001). We observed a significant difference in terms of OS, DFS and TRM between patients receiving a transplant from HLA-identical sibling donors and patients receiving a transplant from either a related mismatched or an unrelated donor (P = 0.001) (Table 2). Among the 359 evaluable patients, 211 (59%) developed acute GVHD. Grade I acute GVHD was observed in 39% of cases (83 patients), grade II in 29% (61 patients), grade III in 20% (42 patients), and grade IV in 12% (25 patients). Two hundred and thirty patients were evaluable for chronic GVHD. Seventy patients (30%) developed chronic GVHD, 54% of whom had extensive disease. Post-transplant causes of deaths were related to disease progression in 39% of cases and transplant toxicity in 61% of cases. Fifty-five patients were still alive and in CR 5 years after allogeneic HSCT. Death occurred later than 5 years after transplantation in seven cases: two patients died from late relapse, three from viral infections, and two from a second malignancy.
On univariate analysis, when we considered factors over which we have no control: age ⩾15 years, M3 subtype according to FAB classification, long first CR duration was predictive of better OS and DFS (Table 2). On the other hand, for the factors over which we have total control, we observed that the intensified conditioning regimen and T cell depletion significantly decreased the probability of OS and DFS and significantly increased TRM (Table 2). Between these two situations, we demonstrated that factors over which we have some control, such as donor type (mismatched related and matched unrelated) and acute GVHD ⩾ grade II, significantly decreased OS and DFS and increased TRM (Table 2), although female donor and occurrence of chronic GVHD were associated with better OS and DFS (Table 2).
On multivariate analysis, the variables age <15 years and CR achievement before transplantation showed a favorable impact on OS (Table 3) and age <15 years was associated with lower TRM. Female donor (Figure 2), no acute GVHD or low grade acute GVHD, and chronic GVHD (Figure 3) were associated with better OS and DFS. In addition, no acute GVHD or low grade acute GVHD were associated with lower TRM. Year of transplantation, TBI-based conditioning regimens, and GVHD prophylaxis type other than T cell depletion did not show any prognostic value in terms of OS, DFS, and TRM.
Conventional chemotherapy appears generally unsatisfactory in advanced AML, while myeloablative therapy followed by allogeneic HSCT can produce durable remission in approximately one-third of cases.78 However, few studies reporting a large enough sample of patients to determine prognostic factors have been reported.1314151617
To our knowledge, this study is the largest cohort of patients with an advanced stage of AML undergoing allogeneic HSCT. We have shown that factors over which we have total, some, or no control can influence survival duration, risk of relapse and TRM after transplantation. Multivariate analysis pointed out the importance of factors over which we have no control (younger age, CR achievement before transplant) and factors over which we have some control (HLA-identical sibling donor, female donor, low grade acute GVHD, and chronic GVHD). These factors added substantially to the success of transplantation by increasing overall survival and by decreasing the risk of relapse and TRM.
A previous study of the International Bone Marrow Transplant Registry (IBMTR) has shown a significantly better 3-year probability of leukemia-free survival after transplantation in patients aged under 30 years and in those with a first remission duration longer than 1 year than in patients over 30 years of age and in those with a first remission lasting less than 1 year.8 In our study, we confirmed that younger patients had a better survival rate and a lower TRM. For AML patients who relapsed after conventional therapy, the place of allogeneic HSCT in untreated first relapse or in second remission is still questionable. Allogeneic HSCT in untreated first relapse has proven to be feasible in 126 patients who received transplants at the Fred Hutchinson Cancer Research Center, with a 5-year probability of relapse-free survival of 23%.18 In our current study, survival was significantly lower for patients receiving transplants in UR or in RR when compared to patients undergoing transplantation in CR. These results are quite different from those from Seattle showing no difference between patients who had transplants in untreated relapse and patients who had transplants in second or subsequent remission.19
In our series, we found better results when transplants for advanced AML were performed from HLA-identical sibling donors; transplants from unrelated donors remained disappointing. The importance of genomic typing of class I HLA alleles to the success of transplants from unrelated donors has been recently demonstrated.20 Large prospective studies, including patients with better HLA matching, would be necessary in order to determine the place of such transplants in advanced AML. Other factors over which we have some control – female donor, low grade acute GVHD, and development of chronic GVHD were associated with a lower probability of relapse after transplantation, probably corresponding to a potential graft-versus-leukemia effect. On multivariate analysis, we demonstrated the importance of factors over which we have no control (age, CR achievement) or some control (female donor, acute and chronic GVHD). Two other strategies to reduce relapse rate after allogeneic HSCT involved increasing the myeloablative function of the conditioning regimen or reducing GVHD prophylaxis before transplantation.
Intensified conditioning regimens have been tested before allogeneic HSCT in phase I–II studies.212223 A standard conditioning regimen combining TBI and cyclophosphamide with etoposide2122 or BCNU23 or busulfan13 and the combination of busulfan with melphalan14 have been examined in advanced-stage AML with encouraging results. A prospective randomized comparison of TBI-etoposide vs busulfan–cyclophosphamide did not show any difference in terms of overall survival and DFS.15 However, a statistically significant difference was observed between AML patients allografted in 2nd CR (standard risk patients) and those allografted in UR or RR (high risk patients).15 In our series, TBI and a GVHD prophylaxis type other than T cell depletion had no real impact on DFS and TRM. However, for these indications a high TRM rate, particularly in refractory and resistant patients, would suggest the need for developing new therapeutic modalities using targeted radiotherapy24 or a reduced myeloablative conditioning regimen.25
In conclusion, recommendations can be drawn from our study taking into account the outcome of treatment and prognostic factors emerging from multivariate analysis: (1) in relapsed patients, CR achievement before transplantation seems an important issue; (2) transplantation from HLA-identical sibling donors using a standard conditioning regimen and GVHD prophylaxis is recommended for patients in second or further remission; (3) unsatisfactory results of allogeneic HSCT for patients with primary refractory AML or refractory relapse require development of new HSCT modalities.
Schiller G, Gajewski J, Nimer S et al. A randomized study of intermediate versus conventional-dose cytarabine as intensive induction for acute myelogenous leukemia Br J Haematol 1992 81: 170–177
Mayer RJ, Davis RB, Schiffer CA et al. Intensive postremission chemotherapy in adults with acute myeloid leukemia New Engl J Med 1994 331: 896–903
Zittoun RA, Mandelli F, Willemze R, for the European Organization for Research and Treatment of Cancer (EORTC) and the Gruppo Italiano Malattie Ematologiche Maligne dell'Adulto (GIMENA) . Autologous or allogeneic bone marrow transplantation compared with intensive chemotherapy in acute myelogenous leukemia New Engl J Med 1995 332: 217–223
Gale RP, Buchner T, Zhang MJ et al. HLA-identical sibling bone marrow transplants vs chemotherapy for acute myelogenous leukemia in first remission Leukemia 1996 10: 1687–1691
Champlin R, Gale RP . Acute myelogenous leukemia: recent advances in therapy Blood 1987 69: 1551–1562
Cassileth PA, Harrington DP, Appelbaum FR et al. Chemotherapy compared with autologous or allogeneic bone marrow transplantation in the management of acute myeloid leukemia in first remission New Engl J Med 1998 339: 1649–1656
Gale RP, Horowitz MM, Biggs JC . Transplant or chemotherapy in acute myelogenous leukemia Lancet 1989 i: 1119–1122
Gale RP, Horowitz MM, Rees JKH et al. Chemotherapy versus transplants for acute myelogenous leukemia in second remission Leukemia 1996 10: 13–19
Bennett JM, Catovsky D, Daniel MT et al. Proposals for the classification of acute leukemias. French–American–British (FAB) co-operative group Br J Haematol 1976 33: 451–460
Kaplan EL, Meier P . Non-parametric estimation from incomplete observations J Am Stat Assoc 1958 53: 457–481
Peto R, Pike MC, Armitage P . Design and analysis of randomized clinical trials requiring prolonged observation of each patient. Part II. Analysis and examples Br J Cancer 1977 35: 1–39
Cox DR . Regression models and life-tables JR Stat Soc (B) 1972 34: 187–220
Petersen FB, Buckner CD, Appelbaum FR et al. Busulfan, cyclophosphamide and fractionated total body irradiation as a preparatory regimen for marrow transplantation in patients with advanced hematological malignancies: a phase I study Bone Marrow Transplant 1989 4: 617–623
Vey N, de Prijck B, Faucher C et al. A pilot study of busulfan and melphalan as preparatory regimen prior to allogeneic bone marrow transplantation in refractory or relapsed hematological malignancies Bone Marrow Transplant 1996 18: 495–499
Blume KG, Kopecky KJ, Henslee-Downey JP et al. A prospective randomized comparison of total body irradiation–etoposide versus busulfan–cyclophosphamide as preparatory regimens for bone marrow transplantation in patients with leukemia who were not in first remission: a Southwest Oncology Group Study Blood 1993 81: 2187–2193
Ostendorf P, Ehninger G, Kallmayer ML et al. Bone marrow transplantation for acute leukemia in second or subsequent remission Klin Wochenschr 1984 62: 1081–1085
Tomas F, Gomez-Garcia de Soria V, Lopez-Lorenzo JL et al. Autologous or allogeneic bone marrow transplantation for acute myeloblastic leukemia in second complete remission. Importance of duration of first complete remission in final outcome Bone Marrow Transplant 1996 17: 979–984
Clift RA, Buckner CD, Appelbaum FR et al. Allogeneic marrow transplantation during untreated first relapse of acute myeloid leukemia J Clin Oncol 1992 10: 1723–1729
Appelbaum FR, Clift RA, Buckner CD et al. Allogeneic marrow transplantation for acute non lymphoblastic leukemia after first relapse Blood 1983 61: 949–953
Sasazuki T, Juji T, Morishima Y et al, . for the Japan Marrow Donor Program Effect of matching of class I HLA alleles on clinical outcome after transplantation of hematopoietic stem cells from unrelated donor New Engl J Med 1998 339: 1177–1185
Brown RA, Wolff SN, Fay JW et al. High-dose etoposide, cyclophosphamide and total body irradiation with allogeneic bone marrow transplantation for patients with acute myeloid leukemia in untreated first relapse: a study by the North American Marrow Transplant Group Blood 1995 85: 1391–1395
Long GD, Amylon MD, Stockerl-Goldstein KE et al. Fractionated total-body irradiation, etoposide and cyclophosphamide followed by allogeneic bone marrow transplantation for patients with high-risk or advanced-stage hematological malignancies Biol Blood Marrow Transplant 1997 3: 324–330
Wu DP, Milpied N, Moreau P et al. Total body irradiation and high-dose cyclophosphamide, BCNU and VP-16 (CBV) as new preparatory regimen for allogeneic bone marrow transplantation in patients with advanced hematologic malignancies Bone Marrow Transplant 1994 14: 751–757
Giralt S, Estey E, Albitar M et al. Engraftment of allogeneic hemapoietic progenitor cells with purine analog-containing chemotherapy: harnessing graft-versus-leukemia without myeloablative therapy Blood 1997 89: 531–536
Dana C, Matthews DC . Immunotherapy in acute myelogenous leukemia and myelodysplastic syndrome Leukemia 1998 12:: (Suppl. 1) S33–S36
We thank the following institutions: Hôpital Universitaire d'Angers (Angers), Hôpital Jean Minjoz (Besançon), Hôpital Haut Lévêque (Bordeaux), Höpital Universitaire de Caen (Caen), Hôpital de Percy (Clamart), Hôpital de l'Hôtel-Dieu (Clermont-Ferrand), Hôpital Henri Mondor (Créteil), Hôpital de Dijon (Dijon), Hôpital Michallon (Grenoble), Hôpital Huriez (Lille), Hôpital Debrousse (Lyon), Institut Paoli-Calmettes (Marseille), Hôpital Lapeyronie (Montpellier), Hôpital de Brabois (Nancy), Hôpital de l'Hôtel-Dieu (Nantes), Hôpital de l'Archet (Nice), Hôpital Cochin (Paris), Hôpital Hôtel-Dieu (Paris), Hôpital Necker Adultes (Paris), Hôpital Necker Enfants (Paris), Hôpital de la Pitié Salpêtrière (Paris), Hôpital Saint Louis (Paris), Institut Gustave Roussy (Paris), Hôpital Jean Bernard (Poitiers), Hôpital de Pontchaillon (Rennes), Hôpital Becquerel (Rouen), Hôpital Charles Nicolle (Rouen), Hôpital Nord (Saint-Etienne), Hôpital de Hautepierre (Strasbourg), Hôpital de Purpan (Toulouse), Hôpital Paul Brousse (Villejuif).
About this article
Prognostic index for patients with relapsed or refractory acute myeloid leukemia who underwent hematopoietic cell transplantation: a KSGCT multicenter analysis
Bone Marrow Transplantation (2019)
Allogeneic hematopoietic cell transplantation for patients with a history of multiple relapses of acute myeloid leukemia
Annals of Hematology (2019)
Efficacy of haploidentical hematopoietic stem cell transplantation compared to HLA-matched transplantation for primary refractory acute myeloid leukemia
Annals of Hematology (2018)
Similar outcome of allogeneic stem cell transplantation after myeloablative and sequential conditioning regimen in patients with refractory or relapsed acute myeloid leukemia: A study from the Société Francophone de Greffe de Moelle et de Thérapie Cellula
American Journal of Hematology (2018)