Abstract

Introduction: B-cell acute lymphoblastic leukemia (B-ALL) with Philadelphia (Ph) chromosome or BCR::ABL1 fusion transcript is a high-risk subtype in hematological malignancy. It is crucial to identify this cytogenetic abnormality because it can help clinicians decide the best treatment regimen quickly and prolong overall survival for patients. The gold standard for detecting the BCR::ABL1 fusion transcript and Ph chromosome is real-time quantitative polymerase chain reaction (RQ-PCR) and fluorescence in situ hybridization (FISH) techniques, respectively. In this study, we aimed to screen B-ALL with BCR::ABL1 fusion transcript by aberrant immunophenotype during the routine diagnosis process.

Methods: We divided forty B-ALL patients into two groups based on the presence of the BCR::ABL1 fusion transcript and Ph chromosome. Immunophenotype analysis and white blood cell (WBC) count were performed for all patients. Comparing the expression level of each antigen between the two groups, we recorded the aberrant immunophenotype related to BCR::ABL1 rearrangement. Bayesian Model Averaging (BMA) was used to find an optimal model to predict the BCR::ABL1 fusion transcript in B-ALL disease.

Results: Besides WBC count (p = 0.004), the presence of three aberrant immunophenotype markers—CD66c (p < 0.001), CD25 (p = 0.002), and CD38 (p = 0.003)—was related to BCR::ABL1 rearrangement. BMA analysis proposed a model to predict B-ALL with BCR::ABL1 fusion transcript using three variables: WBC count and the expression intensity of CD38 and CD66c. The predictive model using WBC count, CD38, and CD66c showed an Area Under the Curve (AUC) value of 0.938 (CI95%: 0.868–1.000), a sensitivity of 85% (CI95%: 70%–100%), and a specificity of 90% (CI95%: 75%–100%) (p < 0.001).

Conclusion: The combined model using WBC count, CD38, and CD66c proposed in this study was useful in predicting B-ALL with BCR::ABL1 fusion transcript.