PREDICTIVE VALUE OF QUANTITATIVE RT-QPCR BCR::ABL1 AT DIAGNOSIS AND EUTOS SCORE FOR IMATINIB RESPONSE IN CML PATIENTS

Abstract

The primary objective was to evaluate the association between the levels of BCR::ABL1 at diagnosis (BCR0), as measured by quantitative RT-qPCR, and the EUTOS score in patients with chronic myeloid leukemia (CML), and their association with BCR::ABL1 transcripts at 3 months after tyrosine kinase inhibitor therapy (BCRM3). We conducted a cross-sectional study evaluating 35 patients diagnosed with CML, using quantitative RT-qPCR to measure BCR0. The relationship between the EUTOS score, BCR0, and BCRM3 was evaluated using Spearman's rank correlation coefficient (Rho). Linear regression modelling was employed to examine the predictive capacity of the EUTOS score and BCR0 on BCRM3. Model fit was assessed using the adjusted coefficient of determination (R2). Statistical significance was determined for p-values less than 0.05. A total of 35 patients with CML were analyzed, with BCR::ABL1 activity assessed at diagnosis. Of these, 68.6% were male. The median age was 39.1 years (IQR: 23.5). The median EUTOS score was 58 (IQR: 69.5), the median percentage of blasts in the bone marrow at diagnosis was 6% (IQR: 8), and the median percentage of CD34 positive cells was 1.25% (IQR: 2). The median BCR0 level was 96% (IQR: 70.5). The correlation between BCR0 and BCRM3 was 0.63 (p-value < 0.01), demonstrating a strong positive correlation. Similarly, the correlation between EUTOS score and BCRM3 was 0.55 (p-value < 0.001), also indicating a strong positive correlation. In the linear regression analysis, using EUTOS and BCR0 as independent variables, the regression model for BCR0 had a slope coefficient of 0.05 with an R2 of 0.6 (p-value < 0.01), indicating that BCR0 significantly predicts BCRM3. The EUTOS regression model had a slope coefficient of 0.3 with an R2 of 0.45 (p-value < 0.01), also indicating a significant predictive value. In the multivariate model, the coefficients were 0.2 for EUTOS (p-value: 0.03) and 0.03 for BCR0 (p-value < 0.001). The multivariate model was more robust, with an R2 of 0.7 (p-value < 0.01), indicating that the combination of EUTOS and BCR0 provides a better prediction of BCRM3. The observed median BCR0 level of 96% underscores the high disease burden at diagnosis. The correlation analysis indicated a weak, non-significant association between BCR0 and EUTOS score (Rho = 0.317, p = 0.06), suggesting that while both are prognostic markers, they may reflect different aspects of disease biology. In contrast, the strong positive correlations between BCR0 and BCRM3 (Rho = 0.63, p < 0.01) and between EUTOS score and BCRM3 (Rho = 0.55, p < 0.001) underscore their relevance in predicting early treatment response. Linear regression analysis further demonstrated the predictive value of BCR0 and EUTOS scores. The BCR0 regression model, with a slope coefficient of 0.05 and R2 of 0.6, and the EUTOS model, with a slope coefficient of 0.3 and R2 of 0.45, both indicated significant predictive power. Notably, the multivariate model, incorporating both BCR0 and EUTOS scores, was more robust (R2 = 0.7), suggesting that a combined approach provides a superior predictive framework for assessing early response to imatinib therapy. These findings highlight the importance of integrating multiple prognostic markers to enhance predictive accuracy in CML management. The significant associations and robust multivariate model support the use of BCR0 and EUTOS scores as complementary tools in clinical practice.