Whole Exome Sequencing Reveals Multiple Driver Events in Chronic Lymphocytic Leukemia Patients with Acquired Ibrutinib Resistance

Disease progression in chronic lymphocytic leukemia (CLL) patients receiving single-agent ibrutinib has been associated with either histological transformation or acquired mutations in BTK and PLCG2 genes which are found in up to 85% of patients with progressive CLL (PD-CLL), less commonly in patients with histologic transformation. Variant allele frequencies of these mutations are highly variable and in some patients remain low and thus cannot fully explain the emergence of drug-resistant clones. Cooperation between mutant and "wild-type" clones has been proposed to explain expansion of clones that have no BTK or PLCG2 mutation. Alternatively, additional drivers have been suggested, including 8p deletion. Here we performed a comprehensive analysis of resistant samples' clonal composition using droplet digital polymerase chain reaction (ddPCR) and whole exome sequencing across multiple time points and compartments to reveal additional hits responsible for drug resistance. Eighty-four CLL patients have been enrolled to a phase II study for ibrutinib monotherapy. Eligible patients had either TP53 aberration or age ≥65 years regardless of prior treatment status. All patients received single-agent ibrutinib 420 mg once daily until disease progression or intolerable side effects occurred. At a median follow-up of 5.5 years, 23 (27.4%) patients developed progressive disease (PD) including 17 patients with PD-CLL, five with Richter's transformation and one with prolymphocytic leukemia. Majority of PD-CLL patients had TP53 aberration (88.2%) and relapsed or refractory CLL (58.8%). The median time-to-progression was 38.8 months for all PD patients and was shorter for those with histologic transformations compared to those with PD-CLL (7.0 vs 44.2 months). We tested for BTK/PLCG2 mutations in 17 PD-CLL patients using archival samples collected at baseline, during response to ibrutinib and at progression using variant-specific ddPCR assays. Fifteen PD-CLL patients (88%) had detectable BTK and/or PLCG2 variants at the time of progression. Notably, only 5 PD-CLL patients (29.4%) demonstrated a strong BTK/PLCG2-driven resistance with cumulative clonal fraction (cCF) above 50% at PD and a significant tumor burden in the peripheral blood (PB). Seven patients (41.2%) demonstrated partial BTK/PLCG2-associated resistance (cCF ranging 4-40%) with additional sub-clonal variants (TP53, NFKBIE, SF3B1, BIRC3, KRAS) and chromosomal aberrations (del8p, del6q, amp2p) as putative drivers in concurrently evolving subclones at PD. Finally, 29.4% of patients had very low (<4%, 3 patients) or undetectable (2 patients) frequencies of BTK/PLCG2 variants but often carried multiple chromosomal aberrations (del17p, del8p, del3p, del2p, del13q) and known CLL driver mutations (NOTCH1, SF3B1) in the resistant clone(s). In all PD-CLL patients, the resistant clones were detected in at least one of the 3 compartments: PB, lymph node (LN) or bone marrow (BM). Expanding clones were present in all 3 sites in three patients (18%), in PB and either LN or BM in 10 (59%) patients, and in LN and/or BM but not in PB in 4 patients (23%). In summary, BTK/PLCG2 mutations are present in most patients progressing on ibrutinib but in many cases these mutations cannot completely account for drug resistance. Exome sequencing followed by clonal decomposition analysis revealed additional genetic alterations in ibrutinib-resistant clones including acquisition of multiple copy number alterations and/or additional known CLL driver mutations. In most patients resistant clones are equally detected in blood and tissue sites.