Early reappearance of intraclonal proliferative subpopulations in ibrutinib-resistant chronic lymphocytic leukemia

Abstract

The Bruton's tyrosine kinase (BTK) inhibitor ibrutinib represents an effective strategy for treatment of chronic lymphocytic leukemia (CLL), nevertheless about 30% of patients eventually undergo disease progression. Here we investigated by flow cytometry the long-term modulation of the CLL CXCR4^dim/CD5^bright proliferative fraction (PF), its correlation with therapeutic outcome and emergence of ibrutinib resistance. By longitudinal tracking, the PF, initially suppressed by ibrutinib, reappeared upon early disease progression, without association with lymphocyte count or serum beta-2-microglobulin. Somatic mutations of BTK/PLCG2, detected in 57% of progressing cases, were significantly enriched in PF with a 3-fold greater allele frequency than the non-PF fraction, suggesting a BTK/PLCG2-mutated reservoir resident within the proliferative compartments. PF increase was also present in BTK/PLCG2-unmutated cases at progression, indicating that PF evaluation could represent a marker of CLL progression under ibrutinib. Furthermore, we evidence different transcriptomic profiles of PF at progression in cases with or without BTK/PLCG2 mutations, suggestive of a reactivation of B-cell receptor signaling or the emergence of bypass signaling through MYC and/or Toll-Like-Receptor-9. Clinically, longitudinal monitoring of the CXCR4^dim/CD5^bright PF by flow cytometry may provide a simple tool helping to intercept CLL progression under ibrutinib therapy.

Fig. 1. Proliferative fraction dynamics in the IOSI-EMA-001 trial.

A Gating strategy for proliferative (PF, CXCR4^dim/CD5^bright, red) and resting (RF, CXCR4^bright/CD5^dim, green) populations by flow cytometry of a representative CLL case in the IOSI cohort. Populations were tracked using a fixed-gate strategy. Reported are percentage of each population over the entire CD19 + /CD5 + CLL population and the month of sampling under ibrutinib. Trend of the proliferative fraction (PF, B) and resting fraction (RF, C) over time by month under ibrutinib. ***p ≤ 0.001 by two-sided Mann–Whitney rank-test compared to month zero. Only significant comparisons are reported. Red dots indicate samples collected from patients close to clinical progression. D Heatmap summarizing the row-normalized median value of investigated features over the cohort: percentage of parent population for PF/RF, mean fluorescence intensity for CXCR4, CD5, and CD20.

Fig. 2. Proliferative fraction dynamics under long-term ibrutinib in the CRO cohort.

A Box-and-whiskers plot of the PF in samples from the CRO cohort before (blue) and after (green) ibrutinib initiation. Samples are grouped by number of years under ibrutinib; pre-ibrutinib samples have been collected within 6 months prior to therapy initiation. Dotted lines are added for visual reference. ****p ≤ 0.0001, ***p ≤ 0.001, **p ≤ 0.01 by two-sided Mann–Whitney rank-test compared to pre-ibrutinib. Only significant comparisons are reported. B Representative flow cytometry contour/dot plot of patient RM113 showing reppearance of the PF after four years of ibrutinib therapy. Contours denote 10% steps of event density. C Left panel: binned dot and density distribution of 64 samples collected within 12 months from ibrutinib discontinuation (“near-discontinuation samples”); right panel: distribution of the PF of the near-discontinuation samples, split by cause of discontinuation, compared to the 22 on-treatment samples (overall n = 86). Dotted lines are added for visual reference. ***p ≤ 0.001, by two-sided Mann–Whitney rank-test. Only significant comparisons are reported. D Trend for beta-2-microglobulin (B2M) serum levels (boxplots and gray area) and PF (red lines) or absolute lymphocyte count (ALC, blue lines) in five patients progressed under ibrutinib. Dotted line represents the 3.5 mg/L threshold for B2M. E Cumulative incidence of progression, split by PF above (red) or below (blue) the ROC criterion in the near-discontinuation cohort (n = 86; PF > 3%). Reported is p value of Gray’s test. F Curve of cumulative incidence of progression split by PF above (red) or below (blue) the ROC criterion (PF > 3%) in BTK-unmutated cases only (n = 49). Reported is p value of Gray’s test.

Fig. 3. Elevated PF associated with BTK mutations.

A Flow cytometry dot plot of cell sorting strategy for PF/RF fractions. B Box-and-whiskers and dot-and-line plot for variant allele frequency (VAF) of BTK/PLCG2 mutations in 11 CLL samples collected near clinical progression. ***p ≤ 0.001, **p ≤ 0.01, by paired Wilcoxon rank test. C Bar plot of PF over RF VAF ratios of each BTK (red) or PLCG2 (blue) mutation (n = 39 for 11 patients) found in all sequenced fractions. Box plots summarize the overall distribution of the ratios. D Box-and-whiskers plot of the PF in the near-discontinuation samples (n = 86), split by presence of BTK/PLCG2 mutations. ***p ≤ 0.001, **p ≤ 0.01 by two-sided Mann–Whitney rank-test. Only significant comparisons are reported.

Fig. 4. Transcriptomic programming of the PF/RF subpopulations.

A Clustered heatmap of 476 differentially expressed genes between paired PF and RF in 16 fractions from 8 CLL patients before ibrutinib initiation. Gene list is reported in Table S2. B Principal component analysis (PCA) of RNA-seq of all sorted fractions (n = 40) using the 476 genes signature, splitting PF (red) from RF (green) samples. C Summary dot-plot of Gene Set Enrichment Analysis (GSEA) on PF/RF-related gene sets curated from literature (Calissano et al. ref. [13], Bartholdy et al. ref. [15], Cadot et al. ref. [46], Seda et al. ref. [17]). Color represents the normalized enrichment score, size is proportional to –log10(FDR q-value), shape denotes significance (circle if q < 0.05). Enrichment plots of selected gene sets are reported separately for samples at pre-ibrutinib and at progression. GSEA summary is reported in Table S3A. D Enrichment plot from GSEA of the HALLMARK_MYC_TARGETS_V2 and HALLMARC_MTORC1_SIGNALING of BTK-unmutated (BTK-wt) versus BTK-mutated PF at progression. GSEA summary for whole Hallmark collection is reported in Table S4C. E Heatmap of Gene Set Variation Analysis (GSVA) of BTK-unmutated (BTK-wt) versus BTK-mutated PF at progression, run with selected gene sets.