Usefulness of the Columbia score for predicting outcomes in patients with transthyretin amyloid cardiomyopathy. Analysis of the Galician registry of cardiac amyloidosis

Aims: To evaluate the predictive value of the Columbia score in patients with transthyretin amyloid cardiomyopathy (ATTR-CM).

Methods: Observational study based in a prospective, multi-centre registry of patients with ATTR-CM recruited between January-2018 and December-2023 in 7 Spanish hospitals. The Baseline Columbia score was correlated by means of multivariable Cox's regression with study endpoints all-cause death and all-cause death or heart failure (HF) hospitalisation. Discriminative capacity was evaluated by means of Harrell's C statistics and area under 2-year time-dependent receiver-operator curves.

Results: We studied 374 patients with ATTR-CM. Columbia score was independently associated with increased risk of all-cause death (adjusted HR per 1 point = 1.30, 95% CI 1.17-1.45) and all-cause death or HF hospitalisation (adjusted HR per 1 point = 1.38, 95% 1.26-1.50). The score showed moderate discriminative capacity for all-cause death (Harrell's C = 0.653) and all-cause death or HF hospitalisation (Harrell's C = 0.697). The area under the 2-year time-dependent receiver-operator curve was 0.594 for all-cause death and 0.669 for all-cause death or HF hospitalisation. Columbia's score was adequately calibrated for both outcomes.

Conclusions: We studied the prognostic performance of the Columbia score in a Spanish prospective cohort of patients with ATTR-CM. The score showed adequate calibration and moderate discriminative capacity for predicting death and HF hospitalisations.