Comparative Study of Different Classification Models in Renal-Cell Carcinoma

Area of Science:

• Oncology
• Clinical Research
• Biostatistics

Background:

• Accurate classification of advanced renal cell carcinoma (aRCC) is crucial for treatment planning.
• Several prognostic models exist, including Memorial Sloan-Kettering Cancer Center (MSKCC), Cleveland Clinic Foundation (CCF), and others derived from clinical trials and expert consensus.

Purpose of the Study:

• To compare the prognostic and predictive performance of various classification models for advanced renal cell carcinoma (aRCC) patients receiving first-line systemic therapy.
• To identify the optimal model for patient stratification based on survival outcomes.

Main Methods:

• Retrospective observational study of 88 aRCC patients treated between 2008-2011.
• Evaluation of classification models using median overall survival (mOS) and median progression-free survival (mPFS).
• Comparison of model performance using time-dependent area under the ROC curve (AUC) analysis.

Main Results:

• The Escudier model demonstrated superior short-term (1-year) prognostic value, followed by the Heng model.
• For long-term prognosis, the Hudes and CCF models showed higher capacity, with the Heng model closely following.
• The Heng model exhibited slightly better predictive ability compared to other models.

Conclusions:

• The Heng model appears to be the most suitable for classifying aRCC patients.
• It offers a balanced approach to short- and long-term prognostics, enhanced predictive accuracy, and better patient distribution.
• Findings support the Heng model's utility in clinical practice, aligning with European Society for Medical Oncology (ESMO) guidelines.