Does RSClin provide additional information over classic clinico-pathologic scores (PREDICT 2.1, INFLUENCE 2.0, CTS5)?

Area of Science:

• Oncology
• Genomics
• Biostatistics

Background:

• Early breast cancer (EBC) management benefits from accurate prognostic tools.
• Gene-expression profiling (e.g., Oncotype-Dx) and clinico-pathologic risk calculators (e.g., PREDICT 2.1, INFLUENCE 2.0, CTS5) are used for risk stratification.
• Tools combining both molecular and clinico-pathologic data (e.g., RSClin) may offer enhanced prognostic information.

Purpose of the Study:

• To compare the prognostic performance of gene-expression profiling tests against clinico-pathologic risk calculators and combined tools in postmenopausal women with EBC.
• To evaluate the ability of different risk assessment tools to predict distant metastasis-free survival (DMFS).

Main Methods:

• Analysis of the TEAM pathology cohort including 4736 postmenopausal women with EBC treated with endocrine therapy.
• Generation of Oncotype-Dx-trained risk scores via gene-expression profiling.
• Calculation of recurrence scores using PREDICT 2.1, INFLUENCE 2.0, CTS5, and RSClin.
• Prognostic ability assessed using Cox regression and Harrell's C-statistic in the N0/N1 population.

Main Results:

• Low correlation was observed between clinico-pathologic calculators (PREDICT 2.1, INFLUENCE 2.0, CTS5) and Oncotype-Dx-trained risk scores.
• In N0 patients, RSClin demonstrated superior prognostic ability for DMFS (C-statistic=0.66) compared to PREDICT 2.1 (0.60), INFLUENCE 2.0 (0.57), CTS-5 (0.62), and Oncotype-Dx (0.63).

Conclusions:

• Combining molecular (gene-expression) and clinico-pathologic factors significantly enhances prognostic information for early breast cancer.
• The clinical utility of combined tools like RSClin requires further prospective validation for patient management decisions.