Abstract Background: Despite the significant improvement in outcomes for chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL) patients with Bruton's tyrosine kinase inhibitors (BTKi), a subset of patients still experience treatment failure, including non-response, disease progression, or Richter transformation (RT). Currently, there is a lack of predictive models based on real-world data to identify high-risk patients. Methods: We conducted a retrospective analysis of CLL/SLL patients treated with ibrutinib or second-generation BTK inhibitors (orelabrutinib and zanubrutinib) across three centers in China. These patients were stratified using block randomization into a development set and a validation set. Logistic regression analysis was used to identify covariates associated with BTKi-therapy failure. The goodness-of-fit of the model was assessed using the Hosmer-Lemeshow test. Model performance was evaluated through discrimination and calibration. Decision curve analysis (DCA) was utilized to quantify the net benefit of the model. Results: A total of 139 patients received BTKi therapy were identified in this retrospective analysis, covering the period from January 2018 to July 2025. 61.87% were male and the median age at diagnosis was 65 years (range 32-86). Additionally, the median age at the initiation of BTK inhibitor therapy was 68 years. Among these patients, 25 died (17.99%), 23 exhibited treatment non-response and disease progression (16.55%), 5 developed RT (3.60%), and 4 were diagnosed with subsequent hematologic malignancies (2.88%). Treatment was discontinued due to toxicity in 8 patients (5.76%). For the remaining patients, dose adjustments were implemented based on treatment response and/or adverse events. Considering treatment non-response and progression as treatment failure events, Logistic regression identified four independent predictors of treatment failure: Lipoprotein (a) 0.3g/L; High-risk FISH (del(11q)/del(17p)); Complex karyotype. A novel predictive model given above risk factors was developed based on the regression coefficient. The predictive model was internally validated via random split (training set: n=97; validation set: n=42). The receiver operating characteristic (ROC) curves of both the training set and validation set demonstrated favorable discriminative performance of the model (area under the curve AUC: 0.725 in the training cohort; AUC: 0.714 in the validation cohort). Furthermore, calibration curves of both cohorts confirmed satisfactory agreement between predicted probabilities and observed outcomes. Additionally, DCA revealed achievable clinical net benefits across various risk thresholds. Conclusion: This multi-center study developed a validated predictive model for BTK inhibitor treatment failure in CLL/SLL, integrating genetic and metabolic factors. The model aids in early identification of high-risk patients, facilitating individualized treatment decisions and optimal resource allocation.
Xinting Hu (Mon,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: