Dissection of Clonal Evolution By Temporal Mutation Profiling in Chronic Lymphocytic Leukemia Patients Treated with Ibrutinib

Introduction: The irreversible Bruton tyrosine kinase (BTK) inhibitor ibrutinib is changing the paradigm of the treatment of chronic lymphocytic leukemia (CLL) with remarkable outcomes in first line as well as in relapsed CLL, including high risk patients with TP53 aberration. While the majority of patients demonstrate durable responses, with a median follow up of 3 years, approximately 18% of them develop resistance and relapse. Mutations in the BTK and PLCG2 genes emerged as predominant mechanisms conferring secondary ibrutinib resistance with BTK and/or PLCG2 mutations detected in the majority of resistant patients, often even 10 months before the clinical relapse. Given the very poor outcome of patients discontinuing ibrutinib, comprehensive understanding of the mechanisms of ibrutinib resistance and changes in the clonal architecture induced by the selective pressure of the drug are of major clinical importance. To dissect the clonal evolution in the context of all relevant mutation targets in CLL, we performed a temporal mutation profiling of 31 genes in paired pre- and post-treatment samples of patients receiving ibrutinib therapy.

Methods: Sequential samples from 18 patients treated with single-agent ibrutinib were included in this study. Pre-treatment peripheral blood samples were available in all patients, with post-treatment samples at 18 months for 6 patients, at 12 months for 7 patients, at 8 months for 3 patients and at 3 months for 2 patients, with a median follow up of 12 months. Targeted deep NGS analysis of 31 recurrently mutated target genes, including ATM, BCOR, BIRC3, BRAF, BTK, CHD2, DDX3X, EGR2, EIF2A, EP300, FBXW7, HIST1H1E, IGLL5, KLHL6, KMT2D, LRP1B, MAPK1, MED12, MGA, MYD88, NFKBIE, NOTCH1, PLCG2, POT1, RIPK1, RPS15, SAMHD1, SF3B1, TP53, XPO1 and ZMYM3 was performed using the TruSeq Custom Amplicon approach (Illumina) on genomic DNA specimens extracted from peripheral blood mononuclear cells. Variant calling was performed using the VariantStudio3.0 application (Illumina) and only variants supported with more than 100 mutant reads were considered. Our cohort represented a heavily pretreated patient group with a median of 2.5 (range: 1-5) lines of prior therapies.

Results: The ultra-deep NGS analysis revealed a total of 473 somatic mutations in the 18 paired samples with a median allelic depth of 25.550x across the 31 genes analyzed. The post-treatment samples carried a considerably higher number of mutations compared to the pre-treatment samples (114 vs 359 mutations) with more than half of the variants representing subclonal alterations with variant allele frequencies (VAF) of <10%. The most frequently mutated genes at baseline included TP53 (50%) , NOTCH1 (50%) and POT1 (28%) with similarly high mutation frequencies in the post-treatment samples. MYD88, BIRC3 and SF3B1 mutations were enriched in the post-ibrutinib samples with mutation frequencies of 11% vs 33%, 11% vs 39% and 11% vs 50%, respectively. Interestingly, XPO1 mutations emerged only in the post-treatment samples of five (28%) patients. As for the BTK and PLCG2 mutations, none of the pre-treatment samples carried a mutation in either of these genes, however, we detected emerging subclonal BTK and PLCG2 mutations (with VAFs between 1.5% and 11%) in 28% (5/18) and 33% (6/18) of the post-treatment samples, respectively. A subset of cases harbored multiple mutations in both genes with some of these representing novel BTK and PLCG2 mutations affecting previously unreported amino acid residues. Scrutiny of the clonal evolution revealed that all individual cases were characterized by almost unique combinations of mutations as well as patterns of emerging and disappearing mutant subclones upon the selective pressure of ibrutinib.

Conclusions: The ultra-deep NGS analysis performed on paired samples of CLL patients treated with ibrutinib provided a significant insight into the mutational repertoire as well as subclonal dynamics of the leukemic compartment. Furthermore, it has revealed a profound subclonal heterogeneity that seems to be further enhanced by ibrutinib therapy, as demonstrated by a higher frequency of subclonal variants in the post-treatment samples. The clinical significance and the potential role of the novel BTK and PLCG2 variants identified in our study will require a longer follow up time, as currently all patients harboring these subclonal variants are still on ibrutinib therapy.