First report of real-word outcomes of chronic myeloid leukemia in Uruguay in tyrosine kinase inhibitor era

Abstract Introduction Several reports on chronic myeloid leukemia (CML) real-world outcomes have been published, although limited data is available from South America. In Uruguay, tyrosine kinase inhibitors (TKIs) are universally covered by the Public Health Office “Fondo Nacional de Recursos” (FNR). In this study we report clinical and epidemiological characteristics of CML in Uruguay, evaluate long-term outcomes in the TKIs era, and determine their efficacy and safety. Methods A descriptive retrospective multicenter national cohort study was performed. Data was obtained from FNR database. Patients older than 18 years newly diagnosed (ND) CML, who received TKI as first line (1L) between 2005 and 2023 at all institutions from Uruguay were included. Overall Survival (OS) was analyzed according to the Kaplan-Meier method with log-rank test. In all cases, p<0.005 was considered significant. Results A total of 648 patients ND CML were included. Median age was 55 years (range 18-90) with a slight male predominance (52%). Most patients (69%) were treated in private centers. At diagnosis, 592 (91.4%) patients were in chronic phase (CP), 28 (4.3%) in accelerated phase (AP) and 28 (4.3%) in blast phase (BP). Sokal risk score was available in 587 patients with CP-CML: 20% had low score, 61% had intermediate score, and 19% patients had high score. The median follow-up of the whole cohort of patients was 6 years (range 0-19). TKI therapy was most frequently initiated with imatinib as per policy (99%), while only 1% received dasatinib. No patient received nilotinib or bosutinib in 1L. Second-generation ITKs are not currently approved in our country in a 1L setting. Most patients received the respective TKI standard dose; dose adjustment was required in 34.1%, most frequently dose escalation needed. Responses were exclusively analyzed for 1L imatinib CP-CML cohort, response information was available until October 2020. Cytogenetic and molecular responses were available in 249 (42.3%) and 324 (55%) of CP-CML patients, respectively. Responses at the end of follow-up were: 76.3% complete cytogenetic response (CCyR), 13.9% major molecular response (MMR), and 47.2% of complete molecular response (CMR). MMR and CMR rates at 12 months were 16.4% and 9.9%, and at 18 months 17.9% and 19.1%, respectively. The median time to first MMR and CMR were 21 months (range 1-122) and 31 months (range 2-131), respectively. Switching to second-lines (2L) TKI occurred in 242 patients (37%): 26% due to TKI intolerance, 66% due to treatment failure and 8% for unknown reasons. Switching trends: 80.6% to dasatinib and 19.4% to nilotinib. Sixty-eight patients (28.1%) discontinued 2L treatment, 83.8% and 16.2% in dasatinib and nilotinib groups, respectively (p=0.425). Therapy changes were due to intolerance in 58.8%, treatment failure in 36.8%, and unknown reasons in 4.4%. Subsequently patients received nilotinib or dasatinib. Adverse events to all TKIs matched the known profile. Imatinib had the highest frequency of events (39.7%, mostly GI), followed by dasatinib (29.3%, mostly pleural effusion), and nilotinib (16.5%, mostly thrombocytopenia). A total of 17 CP-CML patients (2.9%) progressed to AP or BP, the median time to progression was 17 months (range 1-116).OS rates at 5 and 10 years were 79.2% (95% CI: 77.5-80.9%) and 65.6% (95% CI: 63.4-67.5%), respectively. Median OS in <50 years not reached (NR), between 50-79 years was 12.4 years (95% CI: 10.5-14.3), and in ≥80 years was 3.5 years (95% CI: 2.8-4.2) (p<0.001). According to phase at diagnosis, median OS in CP or AP-CML patients was NR, and 1.5 years (95% CI: 1.3-1.7) in BP-CML (p<0.001). Median OS in CP-CML patients with high, intermediate and low risk Sokal scores were 10.1 years (95% CI: 4.7-15.6), 15.1 years, and NR, respectively (p<0.001). Mortality rate was 31.3% (n=203). Main causes of death were: acute leukemia 17.2%, solid neoplasm 15.3%, vascular events 9.4%, infections 3%, renal 3.5%, others 4%, and unknown 47.6%. Conclusions We report the first real-world national data on CML. Epidemiological characteristics are similar to other reports. Although the disease follow-up could be biased, imatinib outcomes as 1L treatment was inferior to international reports, which suggest it can be improved with second generation TKI use in 1L, but cost and access could be an issue. OS are acceptable and similar to other reports, with significant differences according to known factors such as age, phase and risk.