The performance of left/right adrenal volume ratio and volume difference in predicting unilateral primary aldosteronism.

Abstract

The role of computed tomography (CT) in the diagnosis of primary aldosteronism (PA) warrants attention, since the success application of adrenal venous sampling (AVS) remains limited. We aimed to investigate the value of CT-based volumetric indicators, including left-versus-right-adrenal-volume ratio (L/Rv) and left-subtract-right-adrenal-volume difference (L - Rv), in the diagnosis of unilateral primary aldosteronism (UPA). A retrospective case-control study included 153 patients with PA and 1272 controls. AVS was used to classify patients into bilateral disease, left-sided disease, and right-sided disease groups. Adrenal gland volume on both sides of PA patients was significantly larger than controls. The optimal cutoff values of L/Rv and L - Rv were 1.417 [area under the curve (AUC) 0.864] and 1.185 (AUC 0.827), respectively, for the diagnosis of left-sided PA, and 1.030 (AUC 0.767) and 0.220 (AUC 0.769), respectively, for the diagnosis of right-sided PA. The mean AUC for subsequent cross-validation ranged from 0.77 ± 0.03 to 0.86 ± 0.02. Based on the optimal cutoff values, the combination of L/Rv and L - Rv detected 69.6% of patients with left-sided PA and 74.3% of patients with right-sided PA, with a specificity of 93.5% and 89.0%, respectively. For a better clinical application, we reported the sub-optimal cutoffs corresponding to a specificity of 95%. A L/Rv higher than 1.431 and a L - Rv higher than 3.185 as sub-optimal cutoff values was detected in 26.1% of patients with left-sided PA (specificity: 97.2%). A L/Rv smaller than 0.892 and a L - Rv smaller than -0.640 could detect 48.6% of patients with right-sided PA (specificity: 97.5%). CT-based L/Rv and L - Rv performed well in predicting UPA. The combination of L/Rv and L - Rv may serve as a potential indicator for guiding surgical decision making in centers without AVS programs.