What is the optimal threshold for aberrant lymphoblasts at diagnosis to predict lymphoid transformation in chronic myeloid leukemia?

Abstract Abstract Aim: To explore the prognostic proportion of bone marrow aberrant lymphoblasts (ALB) detected by flow cytometry (FCM) for predicting transformation to lymphoid blast phase (LBP) in newly-diagnosed chronic phase patients with chronic myeloid leukemia (CML) receiving tyrosine kinase inhibitor (TKI) therapy. Method: Bone marrow ALB was determined by FCM testing with the detection sensitivity of 0.01% at diagnosis. The X-tile plots were used to determine the optimal cutoff value of continuous co-variates at diagnosis for predicting transformation-free survival (TFS). Cox, Firth-penalized Cox, Inverse Probability of Treatment Weighting (IPTW)-weighted Cox, and Fine-Gray competing risk regression models were used for univariable and multivariable analyses to identify covariates associated with LBP transformation in CML patients on TKI therapy. Results: 1080 chronic phase CML patients with bone marrow blasts <10% and FCM testing were included in this study. 730 (68%), 287 (26%), and 63 (6%) patients were classified in ELTS low-, intermediate-, and high-risk, respectively. Bone marrow ALB was detected in 53 (5%) patients with a median proportion of 0.3% (IQR, 0.1-1.0%; range, 0.01-9.8%). Initial TKIs included imatinib (n = 776, 72%), nilotinib (n = 135, 12%), dasatinib (n = 42, 4%), and flumatinib (n = 127, 12%). With a median follow-up of 42 months (IQR, 16-69 months), 83 (8%) patients transformed to blast phase including LBP (n = 35, 3%) and myeloid blast phase (MBP, n = 48, 5%). 5-year probabilities of TFS and survival were 96% (95% CI, 95-97%) and 97% (95% CI, 96-98%), respectively. Among the 53 patients with detectable ALB, 52 (98%) achieved CHR; 34 (64%), CCyR; 11 (21%), MMR; 23 (43%) transformed to LBP at a median of 7 months (IQR, 3-11 months) on TKI therapy, none transformed to MBP. X-tile analysis identified 0.4% of the ALB as the optimal cutoff value for predicting lymphoid transformation. All 1080 patients were stratified into low- (ALB = 0), intermediate- (0 < ALB <0.4%), and high-risk (ALB ≥0.4%) groups with significant difference in 5-year cumulative incidences of lymphoid transformation (2% [95% CI, 1-2%] vs. 18% [3-32%] vs. 90% [69-100%], p < 0.001) and 5-year probabilities of TFS (98% [98-99%] vs. 82% [69-98%] vs. 10% [2-51%], p < 0.001). Multi-variable analyses by Fine-Gray competing risk, Cox, Firth-penalized Cox, and IPTW-weighted Cox regression models further confirmed that ALB intermediate-/high-risk groups (vs. low-risk) and higher WBC concentrations at diagnosis were significantly-associated with lymphoid transformation. Conclusion: Low-level ALB detected by FCM at diagnosis in chronic phase patients with CML predicted high possibility of subsequent lymphoid transformation during TKI therapy, especially in those with ALB ≥0.4%.