Salivary cardiac troponin does not correlate with serum levels

Several studies suggest a potential correlation between troponin levels detected in serum and saliva. However, prior investigations have not adequately addressed the critical aspect of collecting samples upon admission, which is essential for timely troponin level determination. This study aimed to evaluate the relationship between troponin levels in serum and saliva among patients admitted for chest pain evaluation.

This observational study was conducted at the Interventional Cardiology Department of the University Medical Center HCMC. Patients presenting with chest pain were enrolled, and unstimulated saliva samples were collected using the Navazesh method simultaneously with the initial blood collection. These samples were then analyzed for levels of salivary troponin I, serum troponin I, and serum high-sensitive troponin T.

Among the 48 patients included, 22 (46%) exhibited myocardial injury, while 12 (25%) were diagnosed with acute myocardial infarction. No significant difference was observed in salivary troponin I levels between the non-myocardial injury and myocardial injury groups (p = 0.425). Moreover, no correlation was found between salivary troponin I levels and either serum troponin T or serum troponin I levels (Pearson correlation p = 0.761, 0.500; Spearman correlation p = 0.857, 0.136, respectively). The ROC curve for salivary troponin I in predicting myocardial injury displayed an AUC of 0.566 (95% CI: 0.402-0.731), indicating poor discriminatory power.

In our investigation, salivary troponin I failed to demonstrate a meaningful correlation with serum troponins, thereby limiting its practical utility in diagnosing myocardial injury or myocardial infarction. Further research is warranted to explore its diagnostic reliability and clinical applicability.