Prognostic Implication of Framingham Risk Score As a Comorbidity Measure on Treatment Outcomes Following First-Line Tyrosine Kinase Inhibitor in Newly Diagnosed CML Patients

Introduction

Comorbidities, including cardiovascular (CV) disease and risk factors, are crucial in selecting tyrosine kinase inhibitors (TKIs) for chronic myeloid leukemia (CML) patients (pts). About 40-60% of CML pts have at least 1 comorbid condition, and 30% present with CV risk factors at diagnosis. While the Framingham Risk Score (FRS) assesses the 10-year CV risk events in general population and is suggested to be accounted in the TKI selection in CML pts, prospective data on its impact on TKI choice and outcomes is limited.

Patients and method

The TCGA-GTA project, started in November 2020, is an ongoing, prospective, non-interventional registry study of newly diagnosed CML patients in the Greater Toronto Area, Canada. At initial evaluation, comorbidities and FRS were assessed. Pts with low FRS have ≤10% CV risk at 10 years, intermediate risk is 10-20%, and high risk is >20%. Event-free survival (EFS), failure-free survival (FFS), and TKI-switch-free survival were compared between pts with low FRS (FRS^low) and those with intermediate to high FRS (FRS^int/hi). Events included TKI switch, treatment failure (TF), or death, with TF defined by European LeukemiaNet (ELN) 2013 guidelines.

Results

From Nov 2020 to Jun 30, 2024, the study enrolled 101 newly diagnosed CML pts. Of these, 74.3% were in the FRS^low group and 25.7% in the FRS^int/hi group. The median age was 49 years, with the FRS^int/hi cohort being older (67.5 vs. 41 years, p<0.001). Males constituted 59% of the the FRS^low group vs. 81% of the FRS^int/high group (p=0.057). Chronic phase presentation was seen in 96% of the FRS^low group and 92.3% of the FRS^int/hi group. High-risk cytogenetic abnormalities were present in 9.5% of the FRS^low group and 7.7% of the FRS^int/hi group. The median SOKAL score was 0.8 in the FRS^low and 0.9 in the FRS^int/hi group (p=0.198). The median EUTOS score was 1.25 in the FRS^low and 1.80 in FRS^int/hi group (p=0.001). Comorbidities were present in 44% of patients, with hypertension (27.7%) and hyperlipidemia (25.7%) being most common, followed by diabetes mellitus (12.9%), chronic lung disease (11.9%), and vascular disease (9.9%), including coronary artery disease, stroke and peripheral vascular disease. Comorbidities were more prevalent in the FRS^int/hi cohort (80.8% vs. 30.7%, p<0.001), with a median of 3 vs. 0 comorbidities per patient.

Imatinib was more frequently prescribed in the FRS^int/hi group (69.2% vs. 14.9%), while 2^nd generation (2G) TKI were more common in the FRS^low group including nilotinib (40.5%) and dasatinib (36.5%). With a median follow-up of 21.3 months, the FRS^int/hi group had more adverse events (AEs) (56% vs. 38.6%, p=0.161), including gastrointestinal AEs (19.2% vs. 5.4%, p=0.049) and fatigue (15.4% vs. 1.4%, p=0.016).

Overall, 66% of pts achieved BCR::ABL1 <10% at 3 months, 60.5% achieved BCR::ABL1 <1% at 6 months, and 71.9% achieved BCR::ABL1 <0.1% at 12 months. These molecular response rates were higher in the FRS^low group but the differences were not statistically significant. Twenty-seven percent experienced TF, and 31% required a TKI switch, with higher rates in the FRS^int/hi cohort. Resistance was the main reason for switching therapy (17.6% in the FRS^low vs. 23.1% in FRS^int/hi), followed by intolerance (12.2% in FRS^low vs. 15.4% in the FRS^int/hi). At 12 months, EFS and FFS were 69.6% and 75.4% in the FRS^low group vs. 57.2% and 66% in the FRS^int/hi group. The probability of remaining on the 1L TKI was also higher in the FRS^low group (67% vs. 49.8%, p=0.142). Progression to advanced phase was similar between the groups (4-5% at 12 months).

Multivariable analysis showed that the type of TKI was the only independent factor associated with EFS, FFS, and TKI-switch/discontinuation-free survival, with 2G-TKIs demonstrating superior outcomes compared to imatinib.

Conclusion

FRS is a key tool for evaluating CV comorbidities and guiding the choice of 1L TKI in CML practice, with FRS^int/hi group more likely to receive imatinib. However, intolerance and resistance to imatinib remain significant issues in the FRS^int/hi group, and FRS itself does not independently predict treatment outcomes. The type of 1L TKI drug is the most important independent factor influencing treatment outcomes, with 2G-TKIs showing superior results compared to imatinib. Therefore, there is an unmet need for alternative treatments with better efficacy and tolerability, such as asciminib, in the FRS^int/hi group.