Objective:
Guidelines recommend use of the aldosterone-renin ratio (ARR) for the case detection of primary aldosteronism (PA) and “confirmatory” tests in patients with a baseline ARR value above a given cut-off to exclude false positive results from further diagnostic work-up.
We tested the hypothesis that the ARR carries important quantitative information, which could avoid confirmatory tests in a substantial proportion of the patients.
Design and method:
The study was carried out in two prospectively recruited large cohorts of hypertensive patients who underwent measurement of the ARR and the captopril challenge test, in which a conclusive diagnosis of aldosterone-producing adenoma (APA) was made by the “four corners” criteria. We analyzed the exploratory dataset based on ARR cut-off values corresponding to predefined false positive rates for the diagnosis of APA. The results obtained in the exploratory dataset were then tested in a sensitivity analysis of the validation cohort. To determine if the captopril challenge test yielded any diagnostic gain over the cut-off value, which maximized accuracy, the AUC of ROC curve of the post-captopril ARR was analyzed.
Results:
The exploratory dataset was analyzed at five predefined ARR cut-offs, that yielded a 1% stepwise decrease of false positive rates from 5% to 1%, showing that the positive and negative likelihood ratio progressively increased along with the ARR cut-off value. ARR values above 115.4 (ng/dl)/(ng/ml/h), e.g. the cut-off associated with a false positive rate of 2%, provided no incremental positive likelihood ratio. The diagnostic odds ratio progressively increased up to a peak at the cut-off value corresponding to a 2% false positive rate. After selection of patients with an ARR >115.4 the post-captopril ARR ROC curve (using the APA as diagnosis status) showed no diagnostic gain of the captopril challenge test (0.550 95% CI 0.369–0.732, p = 0.5 vs. the AUC identity line). These results were confirmed by the analysis of the validation dataset.
Conclusions:
Our study showed that the ARR carries quantitative information and that values above the 115.4 cut-off identify patients with a high probability of APA, in whom the diagnostic yield of the captopril challenge test is null.
