DISCRIMINATION, CALIBRATION, AND STRATUM-SPECIFIC LIKELIHOOD RATIOS FOR THE ALDOSTERONOMA RESOLUTION SCORE: A META-ANALYSIS

Abstract

Objective: The Aldosteronoma Resolution Score (ARS) is the most studied scoring system for predicting the high likelihood of complete resolution of hypertension (CRH) after adrenalectomy for primary aldosteronism (PA). We aimed to perform a meta-analysis of the accuracy, discrimination, and calibration of the ARS using a novel approach to calculation of stratum-specific likelihood ratios (SSLR). Design and method: We searched PubMed, Embase (Ovid), the Cochrane CENTRAL, Web of Science, the Bielefeld Academic Search Engine (BASE) to November 2021 according to the Preferred Reporting Items for Systematic Review and Meta-analysis (PRISMA) statement. The quality assessment used adapted TRIPOD (Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis) and PROBAST (Prediction model Risk Of Bias ASsessment Tool) criteria. Results: Thirteen studies comprising 2,158 PA patients from North America (43%), Europe (32%), Asia (22%), and other continents, were included. The pooled estimate of the area under the receiver operating characteristic curve for all studies was 0.77 (95% CI 0.73–0.81), and the ratio of the observed to expected CRH for all studies was 0.9 (95% CI 0.8–1.0). The summary estimates of the SSLR for all studies were 0.3, 0.9, and 3.1, for the low (ARS 0–1), medium (ARS 2–3), and high-likelihood group (ARS 4–5) of CRH, respectively. For studies with a mean follow-up closest to 6, 12, and 36 months, the likelihood ratios were 0.3, 0.8, and 2.6; 0.3, 0.7, and 3.9; and 0.3, 1.8, and 7.0, respectively. When we compared studies that had performed adrenal vein sampling (AVS) in at least 50% of the cohort (mean 77% vs. 38%), the ARS model’s accuracy doubled in the high-likelihood group (ARS 4–5) and tripled in the only study in which all patients were underwent AVS (Figure). Conclusions: In patients with unilateral PA, the ARS is currently the easiest, cheapest, and accurate prediction tool for identifying the long-term high likelihood of CRH after adrenalectomy, particularly when the adrenalectomy is AVS-guided.