More than 50% of patients (pts.) who relapse from Acute Myeloid Leukemia (AML) have karyotypes different from those seen at primary diagnosis. This phenomenon might mirror clonal evolution or might result from antecedent therapies. In some cases completely different karyotypes might characterize treatment-related AML (t-AML). t-AML is a well recognized complication following treatment of malignancies with a high cure rate, especially primary breast cancer and Hodgkin’s Disease and, according to a recently increasing number of reports, even in Acute Promyelocytic Leukemia. However, as to our knowledge there are no data from larger series concerning therapy-induced AML following chemotherapy for AML also comprising Non-M3-AML. Furthermore we asked if polymorphisms in phase-II-enzymes like Gluthathione S-Transferase (GST) or N-Acetyl-Transferase (NAT) might be risk factors for t-AML. Cytogenetics of 141 (72 male, 69 female) pts. who were presumed to have relapsed from AML on the basis of cytologic criteria were performed. In 99 of them cytogenetic data from initial diagnosis could be made available and were compared to the corresponding results on relapse. Six classes of cytogenetic sequences (initial diagnosis --> relapse) were defined (normal --> normal: 29 pts; normal --> aberrant: 16 pts.; aberrant --> aberrant with identical aberrations: 22 pts.; aberrant --> aberrant with additional anomalies: 19 pts.; aberrant --> aberrant with completely different anomalies: 7pts.; aberrant --> normal: 6 pts.) to characterize the dynamics of genetics from diagnosis to relapse. Among the 7 pts. who developed completely different anomalies from those seen at initial diagnosis 2 had cytogenetic anomalies characteristic for those seen following treatment with alkylating agents, while the aberrations in the remaining pts. fitted neither those typical for alkylating agents nor those for the topoisomerase-II-inhibitor type of treatment related anomalies. Of the 35 pts. in the second and third group together 3 pts. presented chromosomal changes of alkylating agents’ type, 2 of topoisomerase-II-inhibitors’ type, 1 patient a combination of both and in the remaining pts. such aberrations that could not be related to defined groups of chemotherapeutics so far. As far as available data from bone marrow smears, immunophenotyping, Fluorescence-In-Situ-Hybridization and moleculargenetics were used to differentiate clonal evolution from the development of a chemotherapy induced new clone. We conclude that a small proportion (6,93%) of seemingly relapsed AMLs might be treatment related, thus representing new clonal hematologic disorders completely different from the initial disease. Cytogenetic aberrations characteristic of alkylating drugs arose in pts. some of whom never received alkylating agents and at time intervals shorter than those supposed to be characteristic of t-AML following this class of drugs. Perhaps further types of treatment related characteristic anomalies induced by other components of AML-treatment such as Ara C or Thioguanin could be defined soon that so far cannot directly be related to certain drugs. Furthermore in a substantial proportion of pts. (34,65%) it cannot be excluded, that clonal evolution had been modulated by antecedent antileukemic chemotherapy. Preliminary data suggest that at least in t-AML following breast cancer treatment polymorphisms of GST influence the risk of t-AML.

Author notes

Corresponding author

Sign in via your Institution