Detection of ABL1 kinase Domain Mutations in Therapy Naive BCR-ABL1 Positive ALL

Background: Mutations in the ABL1 kinase domain are the main mechanism of resistance to tyrosine kinase inhibitors (TKI) in Philadelphia-positive (Ph+) leukemias. In acute lymphoblastic leukemia (ALL), a very early acquisition of mutations can be observed and reports described that mutations may exist already before TKI treatment (Hofmann, Blood, 2003; Pfeifer, Blood, 2007; Soverini, Hematologica, 2011). Initial data came from cloning or mutation-specific PCR assays, which are labor-intensive, do not detect all possible mutations and especially come with the doubt of detecting amplification artefacts. Despite the first description over ten years ago, ABL1 mutation screening is not a standard test for ALL patients before first TKI treatment. Now, the broad availability of highly sensitive next-generation sequencing (NGS) approaches allows overcoming initial limitations.

Aim: To screen initial diagnosis samples of 48 BCR-ABL1 positive ALL patients for ABL1 kinase mutations by NGS.

Patients and Methods: We investigated the diagnosis samples of 48 BCR-ABL1+ ALL patients (21 females, 27 males) with a median age of 62 (range: 19-85) years. Subsequent treatment data was available for 32 patients. The cohort included both patients, who had received stem cell translation (n=11) and only chemo- and/or TKI treatment protocols (n=21). At diagnosis, median blast count was 77% (12%-99%). RNA for BCR-ABL1 detection and mutation analysis was isolated from bone marrow (n=28) or peripheral blood (n=20). Six amplicons for sequencing on the MiSeq (Illumina, San Diego, CA) covered amino acid 184-510 of ABL1 and were generated from pre-amplified BCR-ABL1 . A 1% detection limit was used for analysis (minimal coverage: 400x). Mutations were proven by duplicates to exclude potential artefacts from random PCR errors.

Results: Using NGS, we found well characterized ABL1 kinase domain mutations in three of the 48 (6%) initial diagnosis samples: D276G, T315I and Q252H. The percentage of mutated BCR-ABL1 transcripts were variable and not limited to small clones: D276G: 3% (332/9,725 reads), T315I: 14% (958/6,986 reads) and Q252H: 53% (3,369/7,393 reads with cag>cat, 582/7,393 reads with cag>cac).

The Q252H (cag>cat) mutation had expanded to 97% of BCR-ABL1 transcripts at relapse after five months on treatment with standard protocols including imatinib, to which the mutation causes resistance. The T315I and the D276G mutation also cause resistance to TKIs (especially T315I to four TKIs). Both mutations were found in patients, who had received allogeneic stem cell transplantation and achieved complete remission, thus during follow-up there was no indication for mutation analysis. Interestingly, biochemical assays showed increased oncogenic potency or kinase activity for both mutations without TKIs (Piazza,Leukemia, 2005; Skaggs, PNAS, 2006), which would explain the mutation outgrowth before TKI induced selective pressure. We speculate that mutations, which confer resistance but reduce kinase activity, would not outgrow before TKI treatment and therefore only a subset of resistance mutations can exist in therapy naïve patients. The high turnover rate of ALL cells should allow a much faster selection of mutated clones with a relative growth advantage if compared to the situation in chronic myeloid leukemia.

Interestingly, all three patients with a mutation at the initial diagnosis had a p190 transcript (3/31 [10%]), while none of the 17 patients with a p210 transcript had a mutation at diagnosis.

Of note, in 12 patients we had identified one or two known ABL1 mutations per patient in the later course of TKI treatment as part of routine follow-up. Except for the Q252H mutation, we did not detect any mutation in the diagnosis sample. The relevant bases had a median coverage of 10,341x (4,620x-12,738x), which allows a sensitivity of at least 1%.

Conclusions: Known resistance mutations in the ABL1 kinase domain were detected in 3/48 (6%) therapy naïve diagnosis samples in BCR-ABL1 positive ALL. The mutated clone is expected to expand rapidly if treatment with an insensitive TKI would have been chosen. This might be of clinical relevance as TKIs are one important backbone and guarantor of success for improved outcome. Using NGS allows a direct and sensitive (1%) strategy to identify patients at risk of resistance before any TKI therapy is started and to choose the best drug already in first line.