Search

Introduction: Despite intensive induction and consolidation therapy in newly diagnosed patients with acute myeloid leukemia (AML), approximately half of the younger patients and about 80−90% of the older patients relapse and the majority of patients will succumb to their disease. The prognostic impact of clinical characteristics and genetic abnormalities, which are mostly assessed at initial diagnosis, is less clear. Currently, there is no commonly accepted standard for salvage treatment. Allogeneic hematopoietic cell transplantation (HCT) offers the highest chance of cure in this clinical circumstance. A study in 667 relapsed younger adults (15−60 years) of the HOVON study group revealed a longer relapse-free interval after first complete remission (CR1), presence of a core binding factor (CBF) AML at diagnosis, lower age at relapse and no previous stem-cell transplantation during first-line therapy as factors associated with more favorable prognosis. More recently, additional favorable genetic markers have been reported, such as biallelic CEBPA mutations and the genotype mutated NPM1 in the absence of FLT3-ITD. Methods: In the current study, we evaluated pretreatment characteristics and type of salvage strategy in 1307 adult AML patients enrolled on five prospective first-line treatment trials of the German−Austrian AML Study Group (AMLSG) who experienced a relapse. Between 1993 and 2009, 3324 adults were enrolled on five prospective treatment trials of the AMLSG for newly diagnosed AML (excluding acute promyelocytic leukemia): AMLHD93, AMLHD98A, AMLHD98B, AMLSG 06-04 and AMLSG 07-04. These studies were approved by the institutional review boards of the participating centers. All patients gave informed consent to pretreatment cytogenetic and molecular genetic analyses as well as to treatment within the trials according to the Declaration of Helsinki. The present study included all patients treated in the above protocols who subsequently relapsed. Response assessment followed the standard criteria. However, we report here on complete remission (CR) and CR with incomplete hematological recovery (CRi) as combined response end point, as full hematological recovery was frequently not achieved before the initiation of subsequent treatment (particularly allogeneic HCT). Overall survival was defined here as time between the date of relapse and death or last follow-up. At initial diagnosis chromosome banding analysis and molecular genetics were performed centrally in the AMLSG Laboratory for Cytogenetic and Molecular Diagnosis, and in a subset of patients sequencing data were available. Pairwise comparisons between patient subgroups were performed by the Mann−Whitney or Kruskal−Wallis test for continuous variables and by Fisher’s exact test for categorical variables; multivariable logistic regression models were used to test the influence of covariates on response to salvage therapy. The Kaplan−Meier method was used to estimate the distribution of overall survival and an extended Cox model was used to evaluate prognostic variables. All statistical analyses were performed with the statistical software environment R, version 2.14.0, using the R packages rms, version 3.3-1, and cmprsk, version 2.2-2. Results/discussion: Of the 3324 patients, 2170 (63%) achieved a CR1; a total of 1307 patients relapsed (n=953 after intensive chemotherapy, n=79 after autologous HCT and n=275 after allogeneic HCT). Median duration of CR1 was 274.5 days (range, 31 days−11.4 years). Of 1307 relapsed patients, 1120 patients (median age, 53.6 years; range, 18−82.1) received different salvage regimens and 187 patients (median age, 60.5 years; range, 25.0–85.3) received only palliative care including hydroxyurea. Median and 24-month survival of patients who received salvage therapy versus those who had palliative care were 7.89 months and 27.3% (95% confidence interval (CI), 24.8–30.2%), and 1.58 months and 3.7% (95% CI, 1.7–8.0%), respectively, underlining the very limited prognosis in patients experiencing AML relapse particularly in the absence of specific antineoplastic treatment. Salvage therapy included intensive chemotherapy (n=907, 81.0%), non-intensive chemotherapy including demethylating agents and low-dose cytarabine (n=62, 5.5%), direct allogeneic HCT (n=100, 8.9%), donor lymphocyte infusions (DLI) alone (n=17) or in combination with low-dose chemotherapy (n=13) in patients transplanted in CR1 (n=30, 2.7%) and investigational therapy (n=21 1.9%). Following salvage therapy, response (CR/CRi) was achieved overall in 430 of 1120 patients (38.4%) and according to applied treatment after intensive chemotherapy in 36.8%, after direct allogeneic HCT without prior salvage therapy in 73%, after non-intensive therapy in 11.3%, after investigational therapy in 19.0% and after DLI in 40%. In order to identify prognostic factors that may inform on CR/CRi achievement after intensive salvage therapy, we performed a logistic regression model with the end point CR/CRi after salvage therapy (n=907). This model revealed CR duration >18 months (odds ratio (OR), 1.58; P=0.01), biallelic CEBPA mutation (OR, 2.15; P=0.04) and CBF-AML (OR, 2.20; P0; n=257, CR rate 54%). This simple score may be helpful in decision making, pro or against intensive salvage therapy in relapsed patients. Allogeneic HCT was performed in 537 of 1120 (48%) patients receiving intensive therapy (48%); this includes 100 patients receiving direct allogeneic HCT without prior salvage therapy. The median time interval from diagnosis to allogeneic HCT was 78 days (range, 3−509 days) after relapse, with three patients receiving their transplant beyond 1 year after relapse. The donor types were as follows: n=146 matched related, n=366 matched unrelated and n=22 haplo-identical donors; n=3 were cord blood grafts. Allogeneic HCT was performed with refractory disease after salvage therapy in 247 patients, in CR/CRi after salvage therapy in 190 patients, and in 100 patients directly without salvage. Multivariable analysis on overall survival using an Andersen−Gill regression model taking into account the time dependency of allogeneic HCT on all intensively treated patients (n=1120) revealed CR1 duration 18 months (HR, 0.82, P=0.03), biallelic CEBPA mutation (HR, 0.54, P=0.004), CBF-AML (HR, 0.48, P0. Conclusion: The results of our study in relapsed AML indicate that response to salvage therapy is associated with specific genetic disease entities (CBF-AML, AML with biallelic CEBPA mutation), longer CR1 duration, absence of adverse cytogenetics and FLT3-ITD. A score integrating these factors into three groups is prognostic for the probability of achievement of a CR/CRi after intensive salvage chemotherapy and for survival after relapse in the whole group and in patients receiving direct allogeneic HCT without prior salvage therapy. FLT3-ITD appears to be an unfavorable prognostic marker in all analyses, underlining the need for targeted therapies in these patients.