Mitotic recombinatory evolution in acute leukemia.

A cohort of leukemia cases is presented with ancillary testing that includes microarray studies, karyotyping, FISH, and RNA sequencing to illustrate clonal evolution. Common evolution etiology with each case is apparent homologous mitotic recombination (HMR). The cohort includes: four cases of Pre B-cell acute lymphoblastic leukemia (B-ALL) with a single translocation derivative (19)t(1;19)(q23.3;p13.3), an acute myelogenous leukemia (AML) case with a paracentric inversion of 11q13.3q23 in both homologues confirmed as a rare KMT2A-MAML2 gene fusion, and a transplant patient in AML relapse with a t(6;11)(6q27;q23) and evolution to an additional derivative 6 chromosome. The PBX1-TCF3 fusion in the t(1;19) B-ALL subgroup has long been associated with clones that show either the balanced translocation (∼25%) or the unbalanced single derivative 19 (∼75%).  Evidence from the CMAs and FISH is consistent with HMR initiating at either the PBX1 translocation breakpoint or at a more proximal long arm site that mediates the evolution to the unbalanced form. This is contrary to the previous assumptions of either nondisjunction duplication of the normal homologue with loss of the translocation derivative 1 or an original trisomy 1 that loses the translocation derivative 1. Relapse from an unrelated transplant donor created unique allele dosage ratios in the microarray of the AML patient with the t(6;11) KMT2A-AFDN fusion.  An HMR-based evolution initiation site proximal to the 6q27 AFDN fusion gene is evident in the microarray of chromosome 6, the known oncogenic fusion derivative. The HMR selection driver in both AML cases is very likely associated with the DNA doubling of the oncogenic fusions in 6q and 11q, respectively. Since the oncogenic derivatives in the 1;19 cases are clearly the retained derivative 19, selection for the HMR clonal evolution in 1q is apparently based on the known proliferative advantage of extra copies of 1q in B-ALL and other malignancies. Although selection-based HMR can effectively initiate at any site proximal to a driver gene fusion, it appears that the translocation breaksite is common for many translocations. In addition, evidence from HMR evolution related distal 11q mutations, numerous unbalanced CCND1/IGH translocations, and the double MAML2/KMT2A presented in this study suggest that a recombinatorial "hot spot" exists near the CCND1 gene in many rearrangements or mutations within 11q.