A simplified prognostic score for T-cell large granular lymphocyte leukaemia

Area of Science:

• Hematology
• Oncology
• Clinical Prediction Modeling

Background:

• T-cell large granular lymphocyte leukaemia (T-LGLL) typically has a favorable prognosis.
• A subset of T-LGLL patients experiences significantly shorter survival times.
• Identifying prognostic factors is crucial for effective therapeutic strategies.

Purpose of the Study:

• To identify clinical factors associated with overall survival (OS) in T-LGLL patients.
• To develop a predictive model for OS in T-LGLL.
• To guide therapeutic decision-making for T-LGLL management.

Main Methods:

• Retrospective analysis of 120 T-LGLL patients.
• Lasso regression for feature selection.
• Univariate and multivariate Cox regression analyses.
• Decision tree algorithm for OS prediction model construction.

Main Results:

• Median follow-up was 75 months; 5-year OS rate was 82.2%, 10-year OS rate was 63.8%.
• Poor Eastern Cooperative Oncology Group performance status (≥2) and low platelet count (<100 × 10^9/L) independently predicted worse OS.
• A validated decision tree model stratified patients into low, intermediate, and high-risk groups with distinct survival outcomes.
• Conventional immunosuppressive therapy was insufficient for high-risk T-LGLL patients.

Conclusions:

• A simplified, clinically applicable scoring system for predicting OS in T-LGLL has been developed.
• The model effectively predicts survival outcomes in T-LGLL.
• External validation is recommended prior to widespread clinical implementation.