Baseline and early changes in cardiac biomarkers as predictors of global longitudinal strain deterioration during breast cancer therapy: a PRADAII substudy
TL;DR
Baseline biomarker levels were not associated with GLS deterioration, and no biomarker–treatment interactions were observed (all p for interaction >0.27).Early post-anthracycline changes in hs-cTnI, hs-cTnT, or NT-proBNP were not predictive of subsequent GLS decline, nor were peak or persistent elevations (Table 1, Figure 1).
Abstract
Cardiac biomarkers are recommended for risk stratification before and during potentially cardiotoxic cancer therapy, but the prognostic value of baseline levels and early changes for subsequent left ventricular (LV) dysfunction remains unclear. To evaluate wheter baseline and early changes in high-sensitivity cardiac troponins (hs-cTnI, hs-cTnT) and N-terminal pro–B-type natriuretic peptide (NT-proBNP) predict changes in global longitudinal strain (GLS) over 18 months. In the randomized PRADAII, 138 women treated with anthracyclines (±trastuzumab) received sacubitril/valsartan or placebo and underwent serial biomarker sampling and cardiac imaging at baseline, post-anthracycline, and 18 months. Longitudinal changes in GLS and left ventricular ejection fraction (LVEF) were analyzed using adjusted linear mixed models, including exploratory analyses of combined, persistent, and peak biomarker elevations. Mean age was 54.0±9.4 years. At baseline, no patients had hs-cTnI values above the sex-specific 99th percentile, 6 (4%) had hs-cTnT >99th percentile, and 16 (12%) had NT-proBNP >125 ng/L. Mean baseline GLS was −19.5±2.1% and CMR- and echocardiography-derived LVEF was 60.3±4.8% and 58.3±6.2%, respectively. GLS declined modestly over follow-up, with less deterioration in the sacubitril/valsartan group than in the placebo group (ΔΔ −1.26, 95% CI −2.19 to −0.33). Both hs-cTnI and hs-cTnT increased markedly after anthracycline therapy, with 16.7% and 71.7% exceeding the sex-specific 99th percentile at peak, whereas NT-proBNP showed modest changes with wide inter-individual variability. Transient and persistent troponin elevation occurred in 60.1% and 11.6%, respectively; corresponding NT-proBNP elevations in 5.1% and 7.3%, and combined troponin and NT-proBNP elevation in 9.4%. Baseline biomarker levels were not associated with GLS deterioration, and no biomarker–treatment interactions were observed (all p for interaction >0.27). Early post-anthracycline changes in hs-cTnI, hs-cTnT, or NT-proBNP were not predictive of subsequent GLS decline, nor were peak or persistent elevations (Table 1, Figure 1). No significant associations were observed between baseline or early biomarker levels and changes in LVEF assessed by echocardiography or CMR. In our breast cancer cohort with preserved baseline LV function, baseline levels and early changes in hs-cTnI, hs-cTnT, and NT-proBNP were not associated with subsequent deterioration in GLS over 18 months. Larger studies including higher-risk patients and clinical endpoints are needed to clarify the role of cardiac biomarkers in individualized cardiotoxicity risk stratification.