Chronic lymphocytic leukemia : Studies on cardiovascular disease, biomarkers and subclones

Abstract: The treatment of Chronic lymphocytic leukemia (CLL) is rapidly changing, with targeted therapies, e.g. Bruton’s tyrosine kinase inhibitors (BTKi), entering the frontline and relapse setting. However, BTKis have significant adverse effects, most notably cardiovascular disease (CVD). This is of particular concern since the CLL population consists of mainly elderly patients. Data on CVD comorbidity and its possible relation to inflammation in real-world CLL patients is scarce. The aim of this thesis was to study inflammatory biomarkers, the epidemiology of CVD in CLL as well as the occurrence of subclones, in population-based cohorts. In paper I we analyzed the plasma/serum concentrations in of 11 biomarkers with known association to inflammation and/or CVD in 139 patients with CLL and 71 healthy age- and sex-matched controls. Eight out of 11 biomarkers were increased among CLL patients compared to controls. In addition, the pattern of biomarker expression differed between patients and controls. In paper II, we demonstrated that 32% of all patients diagnosed with CLL in Sweden during 2007-2010 had ≥1 CVD <10 years prior to CLL diagnosis which increased to 37% at start of primary therapy. In paper III, a large CLL cohort, n=4261, was studied (all patients diagnosed with CLL during 2007-2015) with the addition of 5 comparators per CLL patient. In line with paper II, a substantial burden of CVD at time of diagnosis was observed, 39% vs 41% among comparators (p<0.05). During follow-up, the incidence of CVD was increased among patients with CLL (HR 1.69, 1.61-1.77) compared to comparators, regardless of previous CVD or need of therapy. All-cause mortality (HR 2.29, 2.14-2.14) but not CVD mortality (HR 1.00, 0.89-1.13) was increased among patients with CLL vs comparators. Finally, approximately 50% of patients with CLL were prescribed antihypertensive drugs. In paper IV, we studied the occurrence of subclones down to a variant allele frequency (VAF) of 1% using a deep-sequencing NGS panel covering 15 recurrently mutated genes in pretreatment samples from 40 patients with CLL. We found that 48% (n=19/40) exhibited clonal mutations (VAF≥10%,) while 25% (n=10/40) had subclonal mutations (VAF<10%). In 6 out of these 10 patients, subclonal and clonal mutations co-occurred revealing that isolated subclonal mutations were not a frequent event. Mutations were significantly enriched in patients requiring therapy (p<0.05). In conclusion, CLL patients have increased blood concentrations of biomarkers associated with low-grade inflammation. However, they do not have an increased risk of CVD at the time of CLL diagnosis. Nonetheless, an increase in CVD during follow-up was observed, the cause of which is currently unknown. Importantly, the CVD burden in CLL is substantial in our population-based studies and of concern in the era of increasing BTKi usage. Finally, subclonal mutations are present among 25% of CLL patients and commonly co-exists with clonal mutations. The prognostic role of these subclonal mutations needs further investigation.