Abstract
Background and Objectives: Coronary sinus oxygen saturation is a useful indicator of health and disease states. However, it is not routinely used in clinical practice. Cardiovascular magnetic resonance imaging (CMR) oximetry can accurately estimate oxygen saturation in the pulmonary artery. This research aimed to provide a method for calculating coronary sinus oxygen saturation (ScsO(2)) from pulmonary artery oxygen saturation (SpaO(2)) that could be applied to CMR. Materials and Methods: A systematic literature review was conducted to identify prior work that included invasive measures of ScsO(2) and either SpaO(2) or right ventricular oxygen saturation. This revealed one study with appropriate data (ScsO(2) and SpaO(2) measurements, n = 18). We then carried out agreement and correlation analyses. Results: Regression analysis demonstrated a statistically significant, positive relationship between ScsO(2) and SpaO(2), giving a regression equation of ScsO(2) = -31.198 + 1.062 × SpaO(2) (r = 0.76, p < 0.001). A multivariable regression analysis of all reported variables, excluding SpaO(2), independently identified superior vena cava oxygen saturation (SsvcO(2)) and arterial oxygen saturation (SaO(2)) as predictors of ScsO(2) (r = 0.78, p < 0.001), deriving the equation ScsO(2) = -452.8345 + 4.3579 × SaO(2) + 0.8537 × SsvcO(2). Conclusions: In this study, we demonstrated a correlation between coronary sinus oxygen saturation and pulmonary artery oxygen saturation, allowing the estimation of ScsO(2) from SpaO(2). This association enables the estimation of ScsO(2) from purely CMR-derived data. We have also described a second model using arterial and superior vena cava saturation measurements, providing an alternative method. Future validation in larger, independent cohorts is needed.