Background:Tyrosine kinase inhibitors (TKI) revolutionized treatment and prognosis of chronic myeloid leukemia (CML), but some patients (pts) still suffer from potentially severe TKI‐related toxicity. Of particular relevance is cardiovascular (CV) toxicity, mostly associated to Nilotinib (NILO) and Ponatinib. In a previous retrospective, cross‐sectional, multicenter study we confirmed a higher CV risk for pts treated with NILO, particularly those harboring the unfavorable OLR1 genetic polymorphism and we found a NILO‐associated pro‐inflammatory/oxidative effect, with increased IL6/IL10 and TNFα/IL10 ratios and levels of oxidized LDL (oxLDL) compared to Imatinib (IMA). We here report clinical, genetic and serologic results of a prospective multicenter Italian trial including CML pts treated with any of current first‐line TKIs (IMA, NILO, DASA).Aims:(1) to assess CV toxicity in CML pts in which first‐line TKI choice was made according to current CV risk guidelines (2) to assess pts’ genetic predisposition to CV events (3) to confirm a pro‐oxidative/inflammatory effect induced by NILO during treatment.Methods:A clinical CV risk score (S) was calculated using the algorithm from the European SCORE project. Integrating clinical with genetic information from 12 Single Nucleotide Polymorphisms (SNPs) associated with CV risk (ANRIL, SORT1, MIA3, CXCL12, MRPS6, PHATCR1, WDR12, LDLR, MRAS, CYP17A1, OLR1, IL10) an integrated score (IS) was obtained. CV‐relevant clinical and biological parameters were collected at specific timepoints (diagnosis, +1, +3, +6, +12 months) together with blood samples for IL6, IL10, TNFα and oxLDL plasma levels assessment.Results:Between September 2013 and September 2018, 177 newly‐diagnosed CML pts starting either IMA, NILO or DASA treatment, entered this study. After a median follow‐up of 22.4 months (range: 35 days – 5.4 years), a CV event has occurred in 5 pts, with a median time to event of 15 months (range: 44 days – 31.4 months). This accounts for an incidence of 5.9 per 1000 patient‐year, superior to that of a general age‐ and sex‐matched population (3.85 per 1000 patient‐year, OR 3.27, 95%CI 2.16–4.63; p = 0.013). Pts treated with NILO had lower CV risk at diagnosis (S = 6.9 ± 2.8) compared to those treated with DASA (S = 8.1 ± 3.2; p = 0.042) or IMA (S = 9.9 ± 4.9; p = 0.019). Nonetheless, 3/57 (5.2%) NILO‐treated pts developed CV toxicity, as opposed to 1/83 (1.2%) IMA‐treated (HR 2.9, 95% CI 1.8–4.4; p = 0.021) and 0 out of 37 DASA‐treated pts. One CV event occurred in a patient treated with first‐line IMA 23 months after being switched to Bosutinib, and therefore was not attributed to IMA.Levels of anti‐inflammatory cytokine IL10 were higher after 6 and 12 months of IMA or DASA compared to NILO. Also IL6/IL10 and TNFα/IL10 ratios in IMA‐ and DASA‐treated pts were lower compared to NILO‐treated pts. Increased levels of oxidation marker oxLDL were observed after 12 months of NILO, but not during IMA or DASA treatment. The few CV events registered did not allow for any correlation between genetic polymorphisms and increased CV risk in our population.Summary/Conclusion:This study confirms a higher CV risk for pts treated with NILO, even in a contemporary setting where pts were carefully selected for having a low pre‐treatment CV risk. This may be a consequence of a NILO‐induced pro‐inflammatory/oxidative status, as shown by reduced levels of IL10 and increased levels of oxLDL in pts treated with NILO. A longer follow up is necessary to identify possible specific genetic polymorphisms conferring higher CV risk in this population of CML pts treated with TKI.image
Read full abstract