Innovative approaches in chronic myeloid leukemia: from liquid biopsy to genomic characterization beyond BCR::ABL1

Chronic myeloid leukemia (CML) is a paradigm of molecularly targeted cancer therapy. Despite the efficacy of tyrosine kinase inhibitors (TKIs), residual disease persistence and additional genetic abnormalities may prevent the achievement of deep molecular response and treatment-free remission (TFR). To address these clinical challenges, this PhD thesis, encompassing two independent projects, aimed to further elucidate the biological complexity of CML. In the first project, extracellular vesicles (EVs) were explored as a liquid biopsy tool for BCR::ABL1 monitoring. Fifty patients were prospectively monitored by droplet digital PCR (ddPCR): 17 patients (34%) were evaluated from diagnosis onwards and 33 patients (66%) were evaluated during TFR. A significant concordance between EVs and peripheral white blood cells (PWBCs) in BCR::ABL1 detection was found (OR 3.95, 95% CI 1.56–10.0, p = 0.005). In line with PWBC levels, absolute quantification of vesicular BCR::ABL1 showed the highest levels at diagnosis, followed by a gradual reduction at 3 months and a marked decrease by 6 months, with levels remaining low at 12 months (baseline vs 3 months, p=0.116; baseline vs 6 months, p=0.008; baseline vs 12 months, p=0.033; 3 months vs 6 months, p=0.010; 3 months vs 12 months, p=0.010; 6 months vs 12 months, p=0.514). When vesicular BCR::ABL1 levels were normalized to the vesicular RNA plasma concentration, a transient increase was detected at 3 months, suggesting a relative enrichment of BCR::ABL1 transcripts within the vesicular cargo. In patients in TFR, vesicular BCR::ABL1 levels did not differ significantly from those shown by the group of patients monitored longitudinally (TFR vs baseline, p = 0.227; TFR vs 3 months, p = 0.163; TFR vs 6 months, p = 0.643; TFR vs 12 months, p = 0.591). In the second project, the predictive value of cancer-gene variants detected at diagnosis was assessed through next-generation sequencing (NGS) in a retrospective cohort of 50 patients. Overall, 40 Tier I–III variants were identified in 27 patients (54%), most frequently involving ASXL1 and TET2 (20% each), followed by DNMT3A and RUNX1 (7.5% each). Twenty-seven patients (54%) experienced treatment failure due to TKI resistance or disease progression. Within the “failure group”, 18 patients (67%) harbored 28 variants. Patients harboring variants showed a trend toward a shorter failure-free survival (FFS) (log-rank p = 0.054; HR = 2.14, 95% CI 0.96–4.79; Cox p = 0.064) and a significantly shorter event-free survival (EFS) (log-rank p = 0.010; HR = 2.51, 95% CI 1.2–5.2; Cox p = 0.014) compared to those without variants.