Diagnostic value of saline infusion test–derived biochemical parameters for primary aldosteronism subtyping

Abstract

Current clinical practice guidelines recommend adrenal venous sampling (AVS) for subtype classification in patients with primary aldosteronism (PA) who are candidates for adrenalectomy; however, AVS may be considered unnecessary in selected young patients with a high probability of unilateral disease. The additional diagnostic information provided by baseline and saline infusion test (SIT)-derived biochemical parameters when evaluated alongside clinical and radiological features remains uncertain. We aimed to compare the diagnostic performance of baseline and SIT-derived biochemical parameters for predicting unilateral PA and to evaluate the additional diagnostic information provided by each parameter when individually added to a common clinical-radiological framework. This retrospective study included 73 patients with PA (41 unilateral and 32 bilateral) who underwent successful AVS for subtype classification. Baseline plasma aldosterone concentration (PAC), baseline aldosterone-to-potassium ratio (APR), post-SIT PAC, SIT-derived APR, and the suppression index were evaluated. Diagnostic performance was assessed using receiver operating characteristic (ROC) curve analysis, while their associations within a common clinical-radiological model were evaluated using multivariable logistic regression analysis. Post-SIT PAC and SIT-derived APR showed the highest numerical discriminative performance among the evaluated biochemical parameters for predicting unilateral PA (AUC = 0.817 and 0.837, respectively). Adding either parameter individually to the common clinical-radiological model significantly improved model fit (likelihood-ratio p  = 0.014 and p  = 0.028, respectively); however, the increases in AUC were not statistically significant (ΔAUC = 0.022, p  = 0.185 and ΔAUC = 0.017, p  = 0.107, respectively). Post-SIT PAC and SIT-derived APR were strongly correlated (Spearman’s ρ = 0.983, p  < 0.001), and their ROC-derived thresholds yielded concordant classifications in 97.3% of patients. Our findings suggest that post-SIT PAC shows useful discriminative performance for predicting unilateral PA and may contribute to subtype assessment when considered alongside clinical and radiological features. These findings require validation in independent cohorts.