Letter to the Editor from Rui Zhu et al: “Performance of the Aldosterone-to-Renin Ratio as a Screening Test for Primary Aldosteronism: A Systematic Review and Meta-Analysis” Free

To the Editor,

The guidelines recommended the aldosterone-to-renin ratio (ARR) for detecting primary aldosteronism (PA). In a meta-analysis, Hung et al (1) found a wide variability of both sensitivity (10%-100%) and specificity (70%-100%) and a sensitivity < 50% in 40% of the studies, which would indicate a scarce ability of the ARR to pinpoint PA patients.

We cannot share their conclusions for the following reasons. First, to gather solid information, a meta-analysis should consider all eligible studies that met well predefined criteria. Oddly, Hung et al fell short in this regard as they ignored the largest study on the ARR published so far (2), besides several others. Moreover, 2 of the 10 studies that they included should have been omitted because they involved normotensive subjects, while in clinical practice, the ARR is used to detect PA in hypertensive patients. Furthermore, data on postcaptopril ARR should by no means be part of a study of diagnostic accuracy of baseline ARR, which they presented in Figure 3.

Second, as sensitivity and specificity depend on disease prevalence, by selecting studies performed in specialized hypertension centers, which recruited referred patients with an increased pretest probability of PA, their estimates of both these operative features were likely biased. Assessment of the prevalence-independent diagnostic odds ratio and positive likelihood ratio would have provided more solid information in this regard.

Third, lack of standardization in terms of proper pharmacological standardized of the patients and testing conditions may have contributed to the seemingly disappointing performance of the ARR found by Hung et al in their meta-analysis. In truth, in experienced hands, the ARR has been shown to display good within-patient reproducibility when repeated within months (3).

Fourth, even more important, assessing the operative features of the ARR should be done by following the STARD (Standards for Reporting Diagnostic Accuracy) methodology by selecting studies that used an unambiguous diagnosis of unilateral PA (4), made by using the 5-corners criteria (5) or the PASO (Primary Aldosteronism Surgery Outcome) criteria (6). Instead, Hung et al used the results of a “confirmatory” test. These tests stand on the fake assumption that the pathology causing PA is unresponsive to angiotensin II, while in vivo and in vitro data showed that even florid aldosterone-producing adenoma respond to angiotensin II and express angiotensin II receptor type 1 (7, 8). Hence, their analysis was affected by a tautology (ie, a “petitio principii” bias), as their reference test by itself is affected by false-positive and false-negative (ie, ambiguous) results.

Fifth, as clearly shown in the Aldosterone-Renin Ratio for Primary Aldosteronism (AQUARR) study (2), the diagnostic performance of the ARR depends critically on its value, with high values (ie, above 50-100 ng/dL/ng/mL/h) being associated with a specificity approaching 100% and a false-positive rate approaching 0%. Hence, the diagnostic performance of the ARR should be evaluated by considering its quantitative value, which will translate into an optimal performance of this key test for the case detection of PA and not simply by categorizing its result as positive or negative (2).

Additional Information

Disclosure Summary: The authors have nothing to disclose.

References

Author notes