Overestimation of Aldosterone by Immunoassay in Renal Impairment

To the Editor: Accurate measurements of aldosterone are important in the differentiation of various disorders of the renin–angiotensin–aldosterone axis. Aldosterone is commonly measured in clinical laboratories by direct immunoassay without solvent extraction. We report that plasma aldosterone is overestimated by a widely used automated chemiluminescent immunoassay (Diasorin Liaison) in patients with renal impairment. This effect is not confined to end-stage renal failure, but occurs also in graded fashion in patients with mild or moderate renal impairment. We performed aldosterone measurements by the Liaison (DiaSorin) chemiluminescent immunoassay method before and after solvent extraction on 86 outpatient plasma samples. These comprised samples from 41 patients with varying degrees of renal impairment, 21 controls estimated glomerular filtration rate (eGFR) >90 mL · min−1 · (1.73 m2)−1 and 24 patients on hemodialysis. Solvent extraction was performed by adding 3 mL of dichloromethane to 0.6 mL of serum, and the mixture was agitated on a gentle rocker for 10 min. After standing for 30 min, 1.5 mL of the hydrophobic fraction was evaporated to dryness and reconstituted in 0.3 mL Cobas universal diluent (Roche). The typical interbatch analytical CV was 9.4% at 10.5 ng/dL (292 pmol/L) and 6.0% at 32.4 ng/dL (899 pmol/L). In 15 samples following extraction (2 controls, 6 patients with renal impairment, and 7 patients on hemodialysis), aldosterone concentrations were lower than the detection limit of the assay 1.45 ng/dL (40.2 pmol/L) established by the manufacturer in accordance to CLSI EP17-A2. Data were excluded from statistical analysis in these cases. The Student t-test was used for the comparison of ratios. This project was classified as a quality assurance activity using residual tissue, and as such does not require ethical approval in our institution. The New Zealand Health and Disability Ethics Committee considers this project to be out of scope for formal review. As shown in Fig. 1, the fraction of plasma aldosterone immunoreactivity recovered after solvent extraction was markedly decreased in hemodialysis patients 0.31 (0.12); mean (SD) compared with controls 0.71 (0.14); P < 0.0001. Patients with lesser degrees of renal impairment showed decreasing recovery following solvent extraction, in proportion to the severity of the renal impairment. The recovery fraction was positively correlated with eGFR (Pearson correlation coefficient 0.61; P < 0.0001; Fig. 1). In moderate renal failure 12 patients with eGFR between 30–60 mL · min−1 · (1.73 m2)−1 the mean recovery fraction was 0.50 (0.18) (P = 0.003 compared with controls). Even in mild renal impairment 18 patients with eGFR between 60–90 mL · min−1 · (1.73 m2)−1 the mean recovery was significantly decreased to 0.56 (0.17) (P = 0.005 compared with controls). Higher recovery indicates lower overestimation. Linear regression was by the Passing–Bablok method and eGFR is calculated by the CKD EPI (Chronic Kidney Disease Epidemiology Collaboration) equation 2009. Box plots show minimum, maximum, median, and interquartile ranges. The decrease of plasma aldosterone immunoactivity we observed following solvent extraction is most likely a result of the presence of biologically inactive hydrophilic aldosterone metabolites such as aldosterone glucuronide (1) which do not partition into the solvent. With normal renal function, such metabolites are rapidly cleared from plasma, and the solvent-extractable fraction accounts for the major fraction of aldosterone immunoactivity. The decrease in solvent-extractable fraction in renal impairment most likely results from the accumulation of higher concentrations of aldosterone metabolites in the plasma of these patients. The effect of this is that a commonly used automated aldosterone immunoassay overestimates plasma aldosterone in renal impairment, and the magnitude of this bias increases with the worsening of renal function. In comparison to results for individuals with normal renal function, the assay overestimates aldosterone on average by 129% in hemodialysis patients and 42% in moderate renal failure eGFR 30–60 mL · min−1 · (1.73 m2)−1 and 27% in patients with mild renal impairment eGFR 60–90 mL · min−1 · (1.73 m2)−1. Overestimation of aldosterone in patients on dialysis and in end- stage renal failure is not widely recognized although it has been reported previously (1–3). Our study is the first to demonstrate that this problem extends to mild and moderate renal impairment. Although large differences between immunoassays for aldosterone have previously been described (4), it is likely that direct immunoassays for plasma aldosterone other than the one we used would also be subject to interference from aldosterone metabolites in patients with renal impairment. The screening and diagnosis of diseases such as primary aldosteronism depend on the accuracy of aldosterone measurements. Up to one-fifth of patients with newly diagnosed primary aldosteronism may have significant renal impairment (5) and the assay overestimation we describe may increase the likelihood of incorrect diagnosis at both the screening and confirmatory stages of investigation of such patients. We note, however, that the clinical outcomes reported in the existing literature are primarily based on aldosterone immunoassays. Although we have not confirmed this finding, the requirement of solvent extraction in mass spectrometry would eliminate the overestimation observed. Indeed, aldosterone is not overestimated by mass spectrometry in patients on hemodialysis (2). Our results indicate that caution is needed in the interpretation of plasma aldosterone measurements by immunoassay in patients with renal impairment, and, when accurate measurements are required, a liquid extraction step or mass spectrometry should be considered. We would like to thank Linda Henderson (Technical Specialist, Endocrinology, LabPlus) for her technical assistance.