Related Experiment Videos
Validation and Exploratory Refinement of the HFA-ICOS Score for Cardiovascular Risk in Proteasome Inhibitor-Treated
Eduardo Pons-Fuster1, Alejandro Riquelme-Perez2, Vyacheslav Shumbar2
1Clinical Pharmacy and Therapeutics Research Group, Virgen de la Arrixaca University Clinical Hospital, Biomedical Research Institute of Murcia Pascual Parrilla (IMIB), 30120 Murcia, Spain.
Cancers
|June 26, 2026
Summary
Proteasome inhibitors for multiple myeloma increase cardiovascular risks. A refined risk score using NT-proBNP and carfilzomib exposure shows improved prediction of cardiotoxicity in these patients.
Area of Science:
- Cardio-Oncology
- Hematology
- Cardiovascular Medicine
Background:
- Proteasome inhibitors (PIs) are crucial for multiple myeloma (MM) treatment.
- PIs are associated with significant cardiovascular adverse events (CVAEs).
- The HFA-ICOS risk score's utility in MM patients is uncertain.
Purpose of the Study:
- To externally validate and refine the HFA-ICOS risk score for predicting CVAEs in MM patients treated with PIs.
- To identify novel predictors of cardiotoxicity in this population.
Main Methods:
- Retrospective analysis of 98 MM or primary amyloidosis patients treated with PIs (2019-2024).
- External validation and refinement of the HFA-ICOS score.
- Independent predictors of CVAEs were identified using Cox regression.
Main Results:
- CVAEs occurred in 22.5% of patients, primarily heart failure.
- The original HFA-ICOS score had modest predictive accuracy (AUC 0.66) and low sensitivity (50%).
- Carfilzomib exposure and elevated NT-proBNP (>300 pg/mL) before cycle 2 independently predicted CVAEs. A refined dynamic model showed improved discrimination (AUC 0.72) and sensitivity (90.9%).
Conclusions:
- The original HFA-ICOS score has limited prognostic value in PI-treated MM patients.
- A refined dynamic model incorporating carfilzomib, NT-proBNP, and age (≥65) shows promise for improved CVAE risk prediction.
- Further validation in larger cohorts is needed before clinical application of the refined model.
