Limited diagnostic utility of partially successful adrenal vein sampling for primary aldosteronism subtyping.

Abstract

Failure of adrenal vein sampling (AVS) due to difficult cannulation of the right adrenal vein (AV) frequently precludes subtyping of patients with primary aldosteronism (PA) before adrenalectomy. According to a recent study, lateralized PA could be accurately predicted from partial AVS data based on the gradient of the aldosterone-to-cortisol ratios (ACR) between left AV and inferior vena cava (IVC) (LAV/IVC index). We aimed to validate the diagnostic utility of this index for PA subtyping in our cohort. A retrospective cross-sectional study included all patients who underwent bilaterally successful AVS at our centre and were diagnosed with either bilateral PA according to AVS or with lateralized PA after successful adrenalectomy from November 2004 to the end of 2019. Final diagnoses were compared to originally suggested LAV/IVC index cut-offs of ≥ 5.5 for ipsilateral disease and of ≤ 0.5 for contralateral disease, respectively. The inclusion criteria were met in 168 patients: 46 women and 122 men, aged 54 years on average (range 32-72 years); 67 with lateralized and 101 with bilateral PA. LAV/IVC index using cut-offs of ≥ 5.5 or ≤ 0.5 anticipated ipsilateral (left lateralized) PA with a sensitivity of 32% and specificity of 97%, while a sensitivity of 47% and specificity of 95%, were found for contralateral (right lateralized) PA in our cohort. The overall inappropriate adrenalectomy rate was 29.7% (p = 0.314 for comparison between sides). When ascertaining ipsilateral disease (LAV/IVC index ≥ 5.5), 4 out of 16 patients (25%) would have been incorrectly sent to left adrenalectomy. Inappropriate right adrenalectomy would have occurred in 7 out of 21 patients (33.3%) when predicting contralateral disease (LAV/IVC index of ≤ 0.5). Thus, 11 patients with bilateral PA (6.5% of the entire cohort) would have been misclassified as lateralized PA and referred to surgery. Failed lateralization would have occurred in 61.2% of patients overall (53.3% for overlooked contralateral disease, 67.6% for missed ipsilateral disease; p = 0.723 for comparison between sides). Based on our cohort, we conclude that application of the suggested LAV/IVC index cut-offs did not predict lateralized PA with the high accuracy previously reported.