Sampling variability and concordance between sequential and simultaneous adrenal venous sampling in primary aldosteronism subtype diagnosis
Shaomin Shi, Luhong Li, Shuangyu Shen, Junwei Wen, Qinghe Wang, Junxia Wu, Qingan LiAbstract
Obejective
Accurate subtyping of primary aldosteronism (PA) via adrenal venous sampling (AVS) is pivotal for guiding treatment decisions. However, the optimal AVS approach—simultaneous versus sequential bilateral sampling—remains controversial, necessitating a rigorous evaluation of diagnostic consistency between these two techniques.
Methods
Consecutive PA patients who underwent bilateral simultaneous unstimulated AVS via the antecubital vein between December 2024 and January 2026 were prospectively enrolled. Blood samples were collected from the bilateral adrenal veins and a peripheral vein at baseline (t0), 5 minutes (t5), and 15 minutes (t15) after catheterization. Sequential AVS was simulated by sampling the right adrenal vein (RAV) at t0 and the left adrenal vein (LAV) at either t5 or t15. Cortisol and aldosterone concentrations, the selectivity index (SI), and the lateralization index (LI) were measured for all samples. Diagnostic concordance between sampling approaches was assessed using the intraclass correlation coefficient (ICC) and Kappa statistics.
Results
A total of 55 eligible patients were included in the final analysis. Cortisol levels exhibited a significant gradual decline over time (P<0.05 for t0 vs t5), with substantial temporal variability in LI observed throughout the sampling process. ICC analysis demonstrated excellent agreement for LI among simultaneous AVS measurements at different time points (ICC=0.840, 95% CI: 0.764–0.895), whereas agreement was lower between simultaneous and simulated sequential AVS (ICC=0.680, 95% CI: 0.593–0.756 for t5 vs t0-t5; ICC=0.707, 95% CI: 0.560–0.820 for t15 vs t0-t15). Kappa statistics confirmed superior diagnostic concordance for simultaneous AVS at a 5-minute interval (agreement=85.5%, Kappa=0.768) compared with longer sampling intervals or sequential AVS, with notably poor concordance observed in patients with bilateral PA (BPA). Sequential AVS showed systematic differences in LI estimates, with underestimation in right-dominant unilateral PA (UPA) and overestimation in left-dominant UPA.
Conclusion
Sequential AVS may be subject to bias due to stress-related cortisol changes over time, whereas simultaneous AVS provides better diagnostic concordance. Simultaneous AVS remains the gold standard for optimal precision, though sequential AVS is a viable alternative when simultaneous cannulation is not possible.