Ruy Felipe Melo Viegas, Sheila Cavalca Cortelli, Davi Romeiro Aquino, Fernando Oliveira Costa, Bruna Hentges, Vitórya Carvalho Pádua de Magalhães, Rodrigo Augusto Foganholi Silva, Gilson Fernandes Ruivo, Jose Roberto Cortelli
Although NT-proBNP, troponin, and hs-CRP are detectable in saliva, salivary measurements of NT-proBNP and troponin showed insufficient agreement with serum levels to support their use as substitutes for blood-based assessment in heart failure. hs-CRP demonstrated comparatively better performance but remains limited by variability. Under the conditions evaluated, serum remains the reference standard for cardiovascular biomarker assessment in heart failure, and saliva should be regarded as a matrix warranting further investigation.
BACKGROUND: Saliva has been proposed as a non-invasive biological matrix for cardiovascular biomarker assessment. However, evidence regarding its agreement with blood-based measurements in patients with heart failure remains inconsistent.
OBJECTIVE: To assess the presence of NT-proBNP, troponin, and high-sensitivity C-reactive protein in saliva and their correlation with serum concentrations in patients with heart failure over a six-month follow-up period.
METHODS: Participants were prospectively evaluated, and paired saliva and blood samples were collected at baseline, 3 and 6 months. The population included 52 individuals (30 males and 22 females) with HFmrEF (LVEF 41-49%) or HFrEF (LVEF ≤40%), aged 49-80 years (mean age of 61.1 ± 6.79 years). Biomarker concentrations were measured in both matrices. Agreement between salivary and serum measurements was assessed using Bland-Altman analysis. Additional analyses explored whether age influenced the relationship between salivary and serum biomarker levels.
RESULTS: All three biomarkers were consistently detectable in whole saliva. However, Bland-Altman analyses showed that salivary NT-proBNP and troponin did not demonstrate clinically acceptable agreement with serum concentrations at any evaluated time point, indicating limited interchangeability between matrices. In contrast, hs-CRP exhibited relatively better agreement between saliva and serum, with mean bias close to zero and moderate limits of agreement, although substantial inter-individual variability persisted. Age and sex did not meaningfully influence the relationship between salivary and serum biomarker concentrations.
CONCLUSIONS: Although NT-proBNP, troponin, and hs-CRP are detectable in saliva, salivary measurements of NT-proBNP and troponin showed insufficient agreement with serum levels to support their use as substitutes for blood-based assessment in heart failure. hs-CRP demonstrated comparatively better performance but remains limited by variability. Under the conditions evaluated, serum remains the reference standard for cardiovascular biomarker assessment in heart failure, and saliva should be regarded as a matrix warranting further investigation.