Dynamic Risk Model Improves Outcome Prediction After Fixed-Duration CLL Therapy

Pretreatment prognostic indices for chronic lymphocytic leukemia (CLL) do not incorporate treatment-related changes in measurable residual disease (MRD). Researchers developed CIRI2-CLL, a refined continuous individualized risk index incorporating MRD at interim assessment, end of treatment, and 12 months afterward. Model parameters were developed using CLL8/10/11 and MURANO trial data, with independent validation in CLL14 and CLL13, which included fixed-duration venetoclax-obinutuzumab regimens and comparator treatments. Assessments covered calibration, risk stratification, and predictive performance.

Observed and predicted progression-free survival (PFS) event probabilities differed by approximately 5%. CIRI2-CLL achieved C-statistics of 0.82–0.98 across assessment time points and outperformed individual indices, including MRD and CLL-IPI. The authors reported a 14%–37% improvement in PFS prediction over 2–5 years. Three-year PFS rates were 100.0%, 70.2%, and 10.8% in the model-defined low-, intermediate-, and high-risk groups, respectively. Predictive performance was also maintained for overall survival.

These findings support longitudinal MRD integration into relapse-risk estimation after limited-duration CLL therapy and may inform future risk-adapted trials. However, predictive accuracy does not establish that model-guided treatment decisions improve outcomes. The available abstract does not report sample sizes, confidence intervals, or detailed subgroup performance; full-text review is needed to assess those aspects.