Interim Analysis of a Prospective Multicentre Study Using Next Generation Sequencing for Kinase Domain Mutational Analysis in CML Patients on First or Subsequent TKI Therapy

Introduction

Kinase domain mutations in the BCR-ABL1 gene are associated with resistance to tyrosine kinase inhibitors (TKI) in chronic myeloid leukaemia (CML). Next-generation Sequencing (NGS) allows detection of low-level kinase domain mutations as well as quantification of Variant Allele Frequency (VAF).We have previously demonstrated that NGS consistently detects early appearance of kinase domain (KD) mutations in CML patients (Kizilors et al. Lancet Haematology 2019)and highlighted the need for regular monitoring for KD mutations.To that end a multicentre prospective non-interventional study of centralised NGS screening to detect KD mutations was launched in the UK and Republic of Ireland to evaluate the use of NGS in clinical practice (ClinicalTrials.gov, number NCT03647215, INCB 84344-401).

Methods

Patients with CML on first or subsequent TKI therapy who meet the ELN 2013 criteria for warning or failure are eligible for this prospective study. NGS assay was performed on illumnia MiSeq with a single round PCR using our ISO15189 accredited assay as previously described. Results were communicated to the treating haematologist within 10-12 working days. As this is a non-interventional study, clinical intervention was left to the discretion of treating physician. Repeat mutational analysis was/is encouraged until achievement of a sustained MR3/BCR-ABL<0.1%IS.

Results

Between December 2017 and June 2019, 192 CML patients from 31 institutions were included in this prospective study (median age 56 yrs, 19-91). 184 patients had chronic phase CML, 5 de novo accelerated phase and 2 de novo blastic phase (missing data n=1). 84 (43.8%) patients were on first line TKI (imatinib n=71, nilotinib n=9, dasatinib n=4), while 58 (30.2%) patients were treated with second line TKI (imatinib followed by 2GTKI n= 52, 2GTKI followed by another 2GTKI n=5), 27 (14.1%) patients were treated with 3^rdline TKI (including 4/27 patients on ponatinib) and 23 (12.0%) patients were on 4^thline (11/23 patients on ponatinib). At the time of study entry, 78 (40.6%) were in MR2 (BCR-ABL 0.1-1%IS), 50 (26.0%) were in MR1 only (BCR-ABL 1-10%IS), and 54 (28.1%) had a BCR-ABL>10%IS. Of note 10 (5.2%) had a BCR-ABL<0.1%IS.

Only 93 (48.9%) patients had previous KD mutational analysis performed by Sanger Sequencing (SS) before study entry.

23 of 192 patients (12.0%) were found to have a KD mutation by NGS while in CP after one (n=12), two (n=3) three (n=2) or four (n=6) lines of TKI therapy. The ongoing TKI therapy was imatinib (n=10), dasatinib (n=4), nilotinib (n=2), bosutinib (n=3) and ponatinib (n=5). Incidence of KD mutation was 8/78 (10.2%) in the MR2 group, 8/50 (16%) in the MR1 and 7/54 (12.9%) in patient with BCR-ABL>10%IS. A single mutation was found in 19 patients while two patients had two mutations and two patients had 3 or more mutations. The most frequent mutations found were T315I (n=11), F317L (n=3), G250E (n=3), V299L (n=2) and E459K (n=2). A low-level mutation was found in 8/23 (35%) patients and would otherwise not be detectable by SS. A mutation conferring resistance to the ongoing TKI ('clinically relevant mutations') was found in 16/23 patients (69%). TKI switch was made in at least 7 patients with response obtained in 5/7 patients at last follow up. Update on the remaining 16 patients is currently being collected and interim updated results will be presented.

Serial samples from patients tested negative on the first KD mutational analysis were obtained for 27 patients and remained negative on repeat analysis, of whom 7 patients had reduction in BCR-ABL transcript levels in the interim (2 increase and 21 without change). Four patients found with KD mutation(s) also underwent repeat testing for monitoring of VAF showing a reduction in clone size/VAF and BCR-ABL transcript levels in two patients.

Conclusions

This interim analysis demonstrates the clinical importance of KD mutational analysis using NGS. The high proportion of clinically relevant mutations -ie conferring resistance to the ongoing TKI treatment-highlights the potential clinical impact of early detection of KD mutation by NGS providing guidance for a rationale switch of TKI therapy. KD mutation distribution was similar in patients in MR2 compared to those with higher disease burden suggesting the importance of using NGS while disease burden is low in order to increase the success of early TKI switch. Interim updated results will be presented.