Patients With Philadelphia-Positive Leukemia With BCR-ABL Kinase Mutations Prior To Allogeneic Transplantation Predominantly Relapse With The Same Mutation

Tyrosine kinase inhibitors (TKIs) remain the front-line therapy for chronic myeloid leukemia (CML). TKIs also improve remission rates when incorporated into induction and maintenance regimens for Philadelphia-positive acute lymphoblastic leukemia (Ph+ ALL). Unfortunately, resistance to TKIs can occur, and is commonly associated with a point mutation in the ABL kinase domain of BCR-ABL. Allogeneic hematopoietic stem cell transplant (HSCT) is indicated for patients with advanced phase CML or for those who fail TKI therapy, and is also often used in Ph+ ALL patients with suitably matched donors. Sadly, relapse after HSCT is fairly common in both advanced phase CML and Ph+ ALL. Studies have explored the benefit of TKIs given post-HSCT to prevent relapse, but there is limited data as to guide their selection and administration. An important consideration is that the “second generation” TKIs may be associated with more toxicity, particularly cytopenias, in the post-HSCT setting. The question arises whether pre-HSCT mutation status should guide prophylactic TKI selection. Given that resistance can occur with a mix of wild type and mutant clones, which clones “win” in the post-transplant setting? Thus, the purpose of this study is to investigate if specific ABL kinase domain mutations persist or recur in CML and Ph+ ALL patients after HSCT.

In this retrospective analysis, subjects were at least 18 years of age, had undergone allogeneic HSCT at our center between 2000 and 2010 for CML or Ph+ ALL, and had a positive p210 or p190 BCR-ABL transcript by polymerase chain reaction (PCR) in both the pre- and post-HSCT settings. Patients without available records for chart review and those without available pre- and post-HSCT RNA were excluded. Bone marrow and peripheral blood samples obtained prior to HSCT were analyzed for mutations in BCR-ABL. Total RNA was extracted using a TRIzol reagent (Invitrogen, CA, USA) method. A nested PCR, including an initial RT-PCR step for p210 and/or p190 transcripts, was used to amplify a 928-bp product spanning exons 4-9 of the ABL kinase domain. Sanger sequencing of the PCR product was performed using an Applied Biosystems (ABI, CA, USA) 3730xl Analyzer. In patients with a mutation identified, post-HSCT samples were then sequenced.

A total of 95 CML patients and 20 Ph+ ALL patients who underwent HSCT were included in the study. History was notable for pre-HSCT TKI therapy in 64.2% of CML and 90.0% of Ph+ ALL patients, correlating with a mean duration of pre-HSCT TKI exposure of 12.3 and 10.6 months, respectively. At a mean of 2.1 years of follow-up, 30.5% of CML and 70.0% of Ph+ ALL patients had received a TKI post-HSCT either for prophylaxis or relapsed/refractory disease.

Sequencing revealed pre-HSCT ABL kinase mutations in 10 (10.5%) of CML and in 4 (20.0%) of Ph+ ALL patients. All 14 harbored at least one mutation known to be associated with resistance to one or more TKIs. Mutations occurred across the kinase domain, including the TKI-binding site, P-loop and A-loop: L248V, G250E, Q252H, Y253H, T315I, F317L, M351T, F359V, R362G, E450K, E459K and F486S. In 9 of the 14 (64.2%), the mutation conferring TKI resistance was also detectable post-HSCT at an average of day +191 (range +25 to +559). Seven of the 14 patients had refractory or relapsed disease by last follow-up and, of these, 5 (71.4%) received a TKI known to be resistant to the pre-HSCT mutation and failed to respond.

We have shown that resistance mutations often continue to be detectable in patients with molecular relapse or persistently positive BCR-ABL transcripts after HSCT, attributable to persistence or relapse of a TKI-resistant leukemic clone. This finding suggests that in these patients at greatest risk for disease relapse, resistance patterns may be greatly influenced by any drug-resistant clone present before transplant. Thus, in choosing a TKI for use in the prophylactic setting, if no other clinical features (such as co-morbidities) force the exclusion of a particular TKI, the selection of the TKI with predicted activity against the mutant clone might “hedge the bet.” Notably, our overall rate of ABL mutations in CML is somewhat lower than expected, though this may be attributable to the number of patients without TKI exposure. To our knowledge, this is the largest analysis of ABL kinase domain mutations in a transplant population. Further analysis of additional samples is also underway.

No relevant conflicts of interest to declare.