First Results of a New PCR-Based Approach for a Sensitive and Quantitative Monitoring of BCR-ABL Kinase Domain Mutations in Patients with Chronic Myeloid Leukemia.

Mutations of the ABL kinase domain (KD) are the most frequent cause of acquired resistance to imatinib in patients with chronic myeloid leukaemia (CML), most likely due to selection of mutated clones on imatinib treatment. As new Bcr-Abl inhibitors become available, precise quantification of low level mutation will be required to monitor response. Here we report our results evaluating patients with advanced phase or imatinib resistant CML for KD mutations using a newly developed, sensitive and quantitative Ligation PCR (L-PCR) assay in comparison to direct sequencing. Twenty eight CML patients on imatinib (17 male, 11 female) with a median age of 62 (range 20 to 75) years in blast crisis (n=4), accelerated phase (n=12) or imatinib failure (n= 12) were analysed using both approaches. Sequencing of the ABL KD was performed using forward and reverse primers to ABL exons 4 and 7, while the L-PCR analysis focussed initially on the E255K and T315I mutations. Briefly, pairs of probes specific for either wild type (WT) or mutant BCR-ABL were added to the RT-PCR amplified ABL KD, then ligated under conditions optimized for specificity. Ligated probe pairs were than amplified in a quantitative PCR using universal primers. Quantification was performed using internal cell-in-cell dilutions of BaF3 cell lines expressing wt and mutant BCR/ABL and values were expressed as % BCR-ABLmut/ BCR-ABLWT. In our hands, this assay can detect 0.05 – 0.1% T315I and 0.01–0.05% E255K in a BCR-ABL WT background. The inter-assay variation at the lowest detection level was only 6.7 and 4,7% for the mutations T315I and E255K respectively. Results were scored positive only if two independent runs showed amplification exceeding the lowest controls. All patients were treated with a median imatinib dose of 600 (range 500–800) mg for a median of 10.5 (range 1 to 74) months. Dose reductions due to toxicity were necessary in 8 (29%) patients. Direct sequencing revealed E255K or T315I mutations in three patients, each with more than 20% mutated allele. L-PCR revealed these three patients plus three more with lower levels of mutation (T315I, 0.46% and E255K, 0.16, 0.17%). The patient with 0.46% T315I also showed G250E by sequencing. This patient was subsequently treated with Dasatinib but failed to respond. In twenty two patients negative by L-PCR, sequencing of exons 4 and 7 showed 15 (53%) to be WT, 6 (21%) to have KD mutations F317L, F359V (n=2), H396R, M315T with F359V, Y253F with the K247R polymorphism and one to have an 80 base pair deletion 3′ of the p-loop. In conclusion the L-PCR assay was able to detect T315I and E255K mutations in twice as many patients as did direct sequencing. These low-level mutations would most probably also have been missed by the D-HPLC-based screening. The fact that only L-PCR detected this mutation in a patient who failed to respond to dasatinib implies that the assays generate clinically useful information. mutations. The study is currently being expanded to include further mutations and longitudinal monitoring of a larger cohort of patients.