Prognostic models to predict complete resolution of hypertension after adrenalectomy in primary aldosteronism: A systematic review and meta-analysis.

ABSTRACT

Complete resolution of hypertension (CRH) after adrenalectomy for primary aldosteronism is far from a certainty. Although several prognostic models have been proposed to predict outcome after adrenalectomy, studies have not clarified which of the available models can be used reliably in clinical practice. To identify, describe and appraise all prognostic models developed to predict CRH, and meta-analyse their predictive performances. We searched MEDLINE, Embase and Web of Science for development and validation studies of prognostic models. After selection, we extracted descriptive statistics and aggregated area under the receiver operator curve (AUC) using meta-analysis. From 25 eligible studies, we identified 12 prognostic models used for predicting CRH after total adrenalectomy in primary aldosteronism. We report the results for 3 models that had available data from at least 3 external validation studies the primary aldosteronism surgical outcome (PASO) score (AUC 0.81; 95% confidence interval [CI] 0.74-0.86; 95% predictive interval [PI] 0.04-1.00), Utsumi nomogram (AUC 0.79; 95% CI 0.72-0.85; 95% PI 0.03-1.00) and the aldosteronoma resolution score (ARS) model (AUC 0.77; 95% CI 0.74-0.80; 95% PI 0.59-0.86 for all studies and AUC 0.80; 95% CI 0.75-0.85; 95% PI 0.57-0.93 for the studies with the same adrenal vein sampling-guided adrenalectomy rate compared to the models meta-analysed). The PASO score, Utsumi nomogram and ARS model showed comparable discrimination performance to predict CRH in primary aldosteronism. Unlike the ARS model, the number of external validation studies for the PASO score and the Utsumi nomogram was relatively low to draw definite conclusions.