Release.endoclab.pt

Clinical Chemistry 50:91650 –1655 (2004) Hyperaldosteronism: Ratio of Plasma Aldosterone to Renin Concentration Determined by Fully Frank Holger Perschel,1* Rudolf Schemer,3 Lysann Seiler,4 Martin Reincke,4 Jaap Deinum,5 Christiane Maser-Gluth,6 David Mechelhoff,1 Rudolf Tauber,1 and Background: The ratio of plasma aldosterone concen-
between 298 and 6756 (pmol/L)/(ng ⅐ mL؊1 ⅐ h؊1) and
tration to plasma renin activity (PAC/PRA) is the most
PAC/PRC ratios between 105 and 2328 pmol/mU.
common screening test for primary hyperaldosteronism
Whereas PAC or PRC showed broad overlap between
(PHA), but it is not standardized among laboratories.
PHA patients and volunteers, the PAC/PRC ratio indi-
We evaluated new automated assays for the simulta-
cated distinct discrimination of these two groups at a
neous measurement of PAC and plasma renin concen-
cutoff of 71 pmol/mU.
tration (PRC).
Conclusion: The PAC/PRC ratio offers several practical
Methods: We studied 76 healthy normotensive volun-
advantages compared with the PAC/PRA screening
teers and 28 patients with confirmed PHA. PAC and
method. The present study offers preliminary evidence
PRC were measured immunochemically in EDTA
that it may be a useful screening test for PHA. Further
plasma on the Nichols Advantage® chemiluminescence
studies are required to validate these results, especially
analyzer, and PRA was determined by an activity assay.
in hypertensive cohorts.
Results: In volunteers, PAC varied from 33.3 to 1930
2004 American Association for Clinical Chemistry
pmol/L, PRA from 1.13 to 19.7 ng ⅐ mL؊1 ⅐ h؊1 (0.215
ng ⅐ mL؊1 ⅐ h؊1 ؍ 1 pmol ⅐ L؊1 ⅐ s؊1), and PRC from 5.70
Primary hyperaldosteronism (PHA)7 is recognized as the to 116 mU/L. PAC/PRA ratios ranged from 4.35 to 494
most common endocrine form of secondary hypertension (pmol/L)/(ng ⅐ mL؊1 ⅐ h؊1) and PAC/PRC ratios from 0.69
[see, e.g., Refs. (1– 6 )] with an estimated prevalence be- to 71.0 pmol/mU. In PHA patients, PAC ranged from 158
tween 5% and 15% in the hypertensive population (7 ).
to 5012 pmol/L, PRA from 0.40 to 1.70 ng ⅐ mL؊1 ⅐ h؊1,
Because diagnosis of PHA facilitates effective therapy, and PRC from 0.80 to 11.7 mU/L. PAC/PRA ratios were
extended screening including normokalemic patients iswidely accepted (7–9 ).
The most common screening test for PHA is the ratio of plasma aldosterone concentration to plasma renin Clinical Chemistry and Pathobiochemistry and 2 Endocrinology, Diabe- tes, and Nutritional Medicine, Charite´-Universita¨tsmedizin Berlin, Campus activity (PAC/PRA) (7, 9 ). Because the measurement of PRA requires special preanalytical prerequisites, is time- 3 Nichols Institute Diagnostics, Bad Vilbel, Germany.
4 consuming, and shows poor interlaboratory reproducibil- Department of Internal Medicine II, University of Freiburg, Freiburg, ity (10 –12 ), this strategy is not well recommended for 5 Department of Medicine, University Medical Center Nijmegen, Nijme- screening of hypertensive patients in primary care cen- ters. The lower limit of detection varies among the differ- 6 Department of Pharmacology, Ruprecht-Karls-University of Heidelberg, *Address correspondence to this author at: Institut fu¨r Klinische Chemie und Pathobiochemie, Charite´-Universita¨tsmedizin Berlin, Campus BenjaminFranklin, Hindenburgdamm 30, D-12200 Berlin, Germany. Fax 49-30-8445- 7 Nonstandard abbreviations: PHA, primary hyperaldosteronism; PAC, 4152; e-mail frank.perschel@charite.de.
plasma aldosterone concentration; PRA, plasma renin activity; PRC, plasma Received February 23, 2004; accepted June 18, 2004.
renin concentration; RLU, relative light unit(s); APA, aldosterone-producing Previously published online at DOI: 10.1373/clinchem.2004.033159 adenoma; and IHA, idiopathic hyperaldosteronism.
Clinical Chemistry 50, No. 9, 2004 ent PRA assays; therefore, the effect on the PAC/PRA concentrations, respectively, yielded renin values be- ratio and the resulting cutoff values for PHA can be dramatic (12 ). The validity and usefulness of the PAC/PRA ratio for screening for PHA has therefore been Limits of detection and functional sensitivity. The limits of detection, defined as the means plus 3 SD in two series of Recently developed immunoassays for measurement 20 runs of the sample diluent, were 0.013 and 0.094 of circulating renin in plasma (PRC) may overcome this mU/L, respectively. The functional sensitivity, defined as limitation. However, studies evaluating these assays with the sample renin concentration at which the CV for four regard to the diagnosis of PHA are rare (16, 17 ). An series of five assay runs is Ͻ20%, was 2.65 mU/L; 1 mU/L additional problem is that the proposed cutoff values is equivalent to 0.6 ng/L (18, 19 ).
published to date are derived from measurements of The aldosterone assay on Nichols Advantage is a aldosterone in either plasma (2, 3, 6, 11, 13, 15, 17 ) or se- competitive one-site immunometric assay that uses a rum (1, 4, 16 ) and are based on different assays. Because biotinylated monoclonal antibody bound to streptavidin- both PRA and PRC are measured in plasma and because coated magnetic particles. Acridinium-ester-labeled aldo- a simple screening test should be done from only a single sterone competes with sample aldosterone for the limited specimen, aldosterone should also be measured in plasma.
amount of biotinylated antibody. The cross-reactivity to We therefore compared the “new” ratio between plasma any steroid hormones is negligible. According to the aldosterone concentration and plasma renin concentration manufacturer’s package insert, the assay has an analytical (PAC/PRC) with the established PAC/PRA screening test sensitivity of 33.3 pmol/L, and the dynamic range is in patients with PHA and in healthy volunteers to evalu- 0 –3330 pmol/L. The within-run imprecision (CV) at 119, ate its clinical and diagnostic significance.
228, 547, 835, and 1990 pmol/L is 14.0%, 5.4%, 4.1%, 4.4%,and 2.9%, respectively. The total CV at 119, 228, 547, 835, and 1990 pmol/L is 18.6%, 8.5%, 5.2%, 6.3%, and 4.9%, respectively. Addition and dilution study results are The PAC and PRC were measured on a fully automated between 88% and 110%. Parallelism results are between chemiluminescence analyzer (Nichols Advantage®; Ni- 91% and 116%. A method comparison with a commer- chols Institute Diagnostics). This system incorporates sen- cially available RIA gave a correlation coefficient of 0.96 sitive acridinium ester detection technology in combina- and a slope of 1.04; 1 pmol/L is equivalent to 0.36 ng/L.
tion with magnetic particles as solid phase. Samples, PRA was measured as described previously (10, 11 ).
reagents, and magnetic particles are pipetted into dispos- The intraassay variation of this assay was 3.9% at 11.7 able cuvettes and incubated at 37 °C until the reaction is ng ⅐ mLϪ1 ⅐ hϪ1; the interassay variation was 3.6% at 11.0 stopped by a washing step, and the emitted light is ng ⅐ mLϪ1 ⅐ hϪ1 and 6.2% at 3.29 ng ⅐ mLϪ1 ⅐ hϪ1, respec- measured in relative light units (RLU). The system is tively. The detection limit was 0.40 ng ⅐ mLϪ1 ⅐ hϪ1 (0.215 calibrated by a two-point recalibration against a stored ng ⅐ mLϪ1 ⅐ hϪ1 ϭ 1 pmol ⅐ LϪ1 ⅐ sϪ1).
All laboratory testing was done by qualified staff of the PRC was measured with the Nichols Advantage Direct Charite´. The study was not performed as a blind study.
ReninTM assay. This two-site immunometric assay uses an Data analysis and estimation of cutoff values by ROC acridinium-ester-labeled monoclonal antibody, a second curve analysis were done with the Analyze-It software biotinylated monoclonal antibody, and streptavidin- coated magnetic particles. To avoid prorenin activationduring the assay, the incubation time is limited to 30 min at 37 °C. This assay is calibrated to the WHO reference We tested 76 healthy, normotensive volunteers [employ- material (National Institute for Biological Standards and ees and students of the Charite´; 23 males (age range, 22–75 years) and 53 females (age range, 16 – 69 years)], who were In an evaluation study (18 ) this PRC assay showed the not taking any relevant medications. Smoking and use of oral contraceptives were not exclusion criteria. Bloodsamples were drawn in an upright sitting posture.
Precision. The intraassay variation for three samples with The patients studied included 28 patients with con- various concentrations of renin was 1.7–5.3%. Interassay firmed PHA [16 males and 12 females; mean (SD) age, 51 variation in seven samples with renin concentrations from (13.1) years; age range, 26 –72 years]: 9 patients had 10 to 466.5 mU/L was between 2.7– 8.2%.
aldosterone-producing adenomas (APAs), 18 had idio-pathic hyperaldosteronism (IHA), and 1 patient had glu- Accuracy. Dilution of four plasma samples with the sam- cocorticoid-suppressible hyperaldosteronism. The charac- ple diluent gave parallel lines whose slopes did not differ.
teristics of these PHA patients are summarized in Table 1.
Renin measurements in two series of mixtures prepared The screening of our patients was done according to a from two plasma samples with low and high renin generally accepted protocol (7–9 ), in which plasma aldo- Perschel et al.: Rapid Screening Test for PHA Table 1. Characterization of 28 patients with confirmed PHA.a a For details on test protocols, see Quinkler et al. (8 ) and Young (12 ). Only specific medications, which were used after confirmatory and differential diagnostic tests, were given. Most patients with IHA needed additional antihypertensive drugs.
b Saline infusion test: ϩ, plasma aldosterone Ͼ230 pmol/L (Ͼ85 ng/L) after 2 L of isotonic saline; ND, not done.
c 24-h Urinary aldosterone: ϩ, urinary aldosterone Ͼ39 nmol/day and urinary sodium Ͼ200 mmol/day; ND, not done; NU, not usable (urinary sodium Ͻ200 d Posture test: 2, decrease in plasma aldosterone; 1, increase in plasma aldosterone; NU, not useable because of concomitant increase in serum cortisol.
e CT, computed tomography; MRT, magnetic resonance tomography; Spiro., spironolactone; GSH, glucocorticoid-suppressible hyperaldosteronism; Dexa., f CT or MRT: ϩ, unilateral tumor; Ϫ, no tumor.
g Adrenal vein sampling: lateral., lateralization; ND, not done; NU, not usable because of unsuccessful catheterization.
h Glucocorticoid-suppressible hyperaldosteronism was confirmed by genetic testing.
sterone was measured with a commercially available RIA was done routinely in the outpatient clinics of the depart- (not with the new immunometric assay) and had to be ments of endocrinology at the Universities of Freiburg Ͼ416 pmol/L (150 ng/L). The ratio between PAC and PRA had to be Ͼ555 (pmol/L)/(ng ⅐ mLϪ1 ⅐ hϪ1) [Ͼ(200 For this evaluation study, separate blood samples were drawn from all PHA patients in an upright sitting posi- Diagnosis was confirmed by successful surgery (all tion at 0800. The results presented in Tables 2– 4 and Fig.
APAs), pathologic saline infusion test (7–9 ), and in- 1 are from this standardized blood sampling. Although creased 24-h urinary aldosterone excretion on a high- several authors have recommended PAC/PRA screening sodium diet (12 ) (for details see Table 1). Seven of the 28 without discontinuing hypertensive medications (20 –22 ), patients were normokalemic [mean (SD) serum potas- our study protocol for the evaluation of PAC/PRC was sium, 3.4 (0.5) mmol/L; range, 2.0 – 4.3 mmol/L]. Systolic the following: Most of the patients were on antihyperten- blood pressure was between 140 and 220 mmHg [mean sive therapy, but ␤-blockers, angiotensin-converting en- (SD), 179 (21) mmHg]; diastolic blood pressure was be- zyme inhibitors, angiotensin antagonists, and diuretics tween 80 and 140 mmHg [mean (SD), 104 (12.5) mmHg].
were stopped for at least 3 days and spironolactone for at The described characterization of the PHA patients least 4 weeks before blood sampling.
Clinical Chemistry 50, No. 9, 2004 Fig. 1. PAC, PRA, PRC, and the resulting PAC/PRA and PAC/PRC ratios for 76 healthy volunteers and 28 patients with PHA.
The data points for PAC, PRA, PRC, and the ratios are plotted logarithmically; ᭛, healthy volunteers; F, patients with PHA.
ROC analysis (Table 3) showed the superiority of the evaluation of pac assay on nichols advantage ratios compared with PAC, PRA, and PRC alone. The The intraassay imprecision (CV) at concentrations of 114, PAC/PRC ratio performed at least as well as the PAC/ 566, and 1906 pmol/L was 8.2%, 4.1%, and 4.4%, respec- PRA ratio for differentiating PHA patients from healthy tively. The interassay CV at 128, 611, and 1956 pmol/L volunteers. With respect to the requirements of a screen- was 20%, 6.6% and 4.2%, respectively.
ing test, we evaluated the cutoff values at a sensitivity of The linearity on dilution was evaluated in two samples with concentrations of 525 and 1141 pmol/L. The mea-sured concentrations were 76 –100% of the expected. Par- allelism in two samples of different concentrations was Because PHA is a common cause of secondary hyperten- sion and its diagnosis can lead to cure or improvement of Calibration stability was evaluated with eight different calibrations. The CV was 3.0% for calibrator A (mean of60 127 RLU) and 3.2% for calibrator B (mean of 23 062 Table 2. PAC, PRA, PRC, and the resulting PAC/PRA and RLU). The analytical sensitivity was 25 pmol/L.
PAC/PRC ratios for 76 healthy volunteers and 28 patients All methods tested (Fig. 1 and Table 2) differed signifi- cantly between PHA patients and healthy volunteers.
Whereas single measurements of PAC, PRA, or PRC concentrations showed broad overlap between both groups, the ratios PAC/PRA and PAC/PRC provided distinct discrimination between these two groups.
PRC and PRA showed good correlation [Pearson re- gression coefficient (r) ϭ 0.72]. Because the same PAC values served for calculation of the ratios, PAC/PRC and PAC/PRA correlated similarly (r ϭ 0.73). As expected, differences occurred more frequently at the ends of lower renin concentration and activity ranges near the detection Perschel et al.: Rapid Screening Test for PHA confirmed the assay specifications given by the manufac- Table 3. Calculated cutoff values and resulting specificity with respect to 100% sensitivity to differentiate patients Because this is the first study with a fully automated system and only two studies with manual PRC assays for screening for PHA have been published to date (16, 17 ), the comparison of our data with data from the literature is limited. Trenkel et al. (16 ) suggested a cutoff value of 50 (aldosterone measured in serum by RIA, expressed in ng/L; PRC measured by IRMA, expressed in ng/L), which corresponds to 83 when using the units in Table 2 (PAC expressed in pmol/L, PRC in mU/L). Ferrari et al.
(17 ) recommended a cutoff of 150 (aldosterone measuredin plasma by RIA, expressed in ng/L; PRC measured by hypertension (APA) or targeted pharmacotherapy (IHA), IRMA, expressed in ng/L) corresponding to 90 when validated and cost-effective routine screening protocols using the units in Table 2. Our data analysis (Tables 2 and 3; Fig. 1) produced a tentatively proposed cutoff value of At present, the most common screening test is the ratio 71. Subsequent studies including essential hypertensive between serum or plasma aldosterone (PAC) and PRA.
cohorts may necessitate readjustment, although there is Because several authors have determined PAC/PRA cut- good concordance with the recommendations made by offs to screen for PHA (1–9 ), this procedure is widely Trenkel et al. (16 ) and Ferrari et al. (17 ).
accepted as the best validated screening protocol. How- Although the simultaneous measurement of PAC/PRC ever, simultaneous measurement of PAC and PRA has in plasma samples from 28 patients with known PHA some disadvantages: The method for measuring PRA showed no overlap with healthy volunteers (Fig. 1), there requires cooling of the specimen during transport and are certain limitations to our study: The PAC/PRC ratio storage, is time-consuming, and shows weak interlabora- seems to be superior to the PAC/PRA ratio. Although tory reproducibility (10 –12 ). Because aldosterone is usu- the PAC/PRA ratio of our PHA patients in the initial ally measured in serum, determination of the PAC/PRA outpatient characterization was clearly above the cutoff ratio requires an additional, simultaneously drawn, blood value of 555 (pmol/L)/(ng ⅐ mLϪ1 ⅐ hϪ1) [200 (ng/L)/ tube. In addition, PAC, PRA, and the PAC/PRA ratio (ng ⅐ mLϪ1 ⅐ hϪ1)], the standardized reexamination of show large intra- and interpatient variations in patients these patients by use of a second blood sample showed with PHA (15 ); therefore, several authors have cast doubt that some patients would have failed this diagnostic on the validity and sensitivity of the PAC/PRA ratio criterion. Our proposed cutoff value (Table 3), which was chosen to achieve a 100% sensitivity with the new auto- Alternative screening procedures that overcome these mated assay, was therefore somewhat lower than the disadvantages seem to be necessary. Measurement of PRC values recommended in the literature. In addition, some instead of PRA may reduce some preanalytical and ana- patients with confirmed PHA had PACs in the standard- lytical problems and, therefore, improve intra- and inter- ized reexamination that were Ͻ414 pmol/L (150 ng/L) laboratory reproducibility (23, 24 ). Major advantages of and therefore would also fail the second diagnostic crite- the new PAC and PRC assays evaluated in our study include the ease and performance of the test (no cooling, Because there is no doubt in the correct characteriza- a single plasma sample, and good interlaboratory repro- tion of our patients as having PHA, we have two expla- ducibility because the assays are automated) and the nations for these findings: (a) a recent study by Tanabe et rapid availability of results (automated immunochemilu- al. (15 ) has demonstrated high intraindividual variability minometric assays), which is a prerequisite for extensive of PAC, PRA, and the PAC/PRA ratio in patients with screening of hypertensive patients for PHA.
PHA; and (b) we measured plasma aldosterone with the The performance of the new PAC assay was evaluated new immunochemiluminometric assay, these data show with a preliminary reagent lot. However, our results that each PAC and PRC assay requires separate validationof cutoff values.
Evaluating diagnostic tests in a group of patients Table 4. Proposed cutoff values for the PAC/PRC ratio already known to have the disease and in a group of with different units used in the literature.
healthy volunteers can lead to overestimation of diagnos- tic accuracy (25 ). Thus, our data necessarily need to be complemented by further studies in hypertensive cohorts, e.g., in groups of patients with essential hypertension and The clinical conditions necessary for testing of the PAC/PRC or PAC/PRA ratios, such as discontinuing Clinical Chemistry 50, No. 9, 2004 drug therapy, recording of dietary sodium intake, or time 8. Quinkler M, Lepenies J, Diederich S. Primary hyperaldosteronism.
of day for sampling, are not yet sufficiently standardized, Exp Clin Endocrinol Diabetes 2002;110:263–71.
which contributes to the different published cutoff values 9. Montori VM, Young WF Jr. Use of plasma aldosterone concentra- and complicates the use of the screening test in clinical tion-to-plasma renin activity ratio as a screening test for primaryaldosteronism. A systematic review of the literature. Endocrinol routine. An adequate standardization that meets general Metab Clin North Am 2002;31:619 –32.
10. Oelkers W, Diederich S, Bahr V. Diagnosis and therapy surveil- lance in Addison’s disease: rapid adrenocorticotropin (ACTH) test In summary, our results suggest that the simultaneous and measurement of plasma ACTH, renin activity, and aldoste- measurement of PAC and PRC with automated immuno- rone. J Clin Endocrinol Metab 1992;75:259 – 64.
chemiluminometric assays may be useful in screening for 11. Oelkers W, Diederich S, Bahr V. Primary hyperaldosteronism PHA. Nevertheless, subsequent studies are required for without suppressed renin due to secondary hypertensive kidneydamage. J Clin Endocrinol Metab 2000;85:3266 –70.
further comparison of the PAC/PRC ratio vs the PAC/ 12. Young WF Jr. Primary aldosteronism: management issues. Ann PRA ratio used for screening at present. These studies will have to deal with the following questions: (a) Because we 13. Montori VM, Schwartz GL, Chapman AB, Boerwinkle E, Turner ST.
tested only for the differentiation of patients from normo- Validity of the aldosterone-renin ratio used to screen for primary tensive volunteers, has the PAC/PRC ratio similar sensi- aldosteronism. Mayo Clin Proc 2001;76:877– 82.
tivities for screening of large hypertensive populations? 14. Schwartz GL, Chapman AB, Boerwinkle E, Kisabeth RM, Turner ST.
Prospective studies are needed. (b) Is the newly devel- Screening for primary aldosteronism: implications of an increased oped assay for PAC really a method that offers reliable plasma aldosterone/renin ratio. Clin Chem 2002;48:1919 –23.
15. Tanabe A, Naruse M, Takagi S, Tsuchiya K, Imaki T, Takano K.
results if performed by different laboratories in clinical Variability in the renin/aldosterone profile under random and routine? (c) How is the PAC/PRC ratio influenced by standardized sampling conditions in primary aldosteronism. J Clin clinical conditions, especially drug effects? Regardless of Endocrinol Metab 2003;88:2489 –94.
the answers to these questions, we expect that the new 16. Trenkel S, Seifarth C, Schobel H, Hahn EG, Hensen J. Ratio of ratio method will evolve as a rapid and easily performed serum aldosterone to plasma renin concentration in essential hypertension and primary aldosteronism. Exp Clin EndocrinolDiabetes 2002;110:80 –5.
17. Ferrari P, Shaw SG, Nicod J, Saner E, Nussberger J. Active renin versus plasma renin activity to define aldosterone-to-renin ratio for We would like to acknowledge J. Wilde for statistical primary aldosteronism. J Hypertens 2004;22:377– 81.
calculations and P. Exner, R. Go¨ber, K. Hanusa, and M.
18. De Bruin RA, Bouhuizen A, Diederich S, Perschel FH, Boomsma F, Schro¨dter for excellent technical assistance.
Deinum J. Validation of a new rapid renin assay. Clin Chem;inpress.
19. Simon D, Hartmann DJ, Badouaille G, Caillot G, Guyenne TT, 1. Fardella CE, Mosso L, Gomez-Sanchez C, Cortes P, Soto J, Gomez Corvol P, et al. Two-site direct immunoassay specific for active L, et al. Primary hyperaldosteronism in essential hypertensives: renin. Clin Chem 1992;38:1959 – 62.
prevalence, biochemical profile, and molecular biology. J Clin 20. Gallay BJ, Ahmad S, Xu L, Toivola B, Davidson RC. Screening for primary aldosteronism without discontinuing hypertensive medi- 2. Loh KC, Koay ES, Khaw MC, Emmanuel SC, Young WF Jr.
cations: plasma aldosterone-renin ratio. Am J Kidney Dis 2001; Prevalence of primary aldosteronism among Asian hypertensive patients in Singapore. J Clin Endocrinol Metab 2000;85:2854 –9.
21. Seifarth C, Trenkel S, Schobel H, Hahn EG, Hensen J. Influence of 3. Calhoun DA, Nishizaka MK, Zaman MA, Thakkar RB, Weissmann antihypertensive medication on aldosterone and renin concentra- P. Hyperaldosteronism among black and white subjects with tion in the differential diagnosis of essential hypertension and resistant hypertension. Hypertension 2002;40:892– 6.
primary aldosteronism. Clin Endocrinol (Oxf) 2002;57:457– 65.
4. Mosso L, Carvajal C, Gonzalez A, Barraza A, Avila F, Montero J, et 22. Mulatero P, Rabbia F, Milan A, Paglieri C, Morello F, Chiandussi L, al. Primary aldosteronism and hypertensive disease. Hypertension et al. Drug effects on aldosterone/plasma renin activity ratio in primary aldosteronism. Hypertension 2002;40:897–902.
5. Lim PO, MacDonald TM. Primary aldosteronism, diagnosed by the 23. Cartledge S, Lawson N. Aldosterone and renin measurements.
aldosterone to renin ratio, is a common cause of hypertension.
Clin Endocrinol (Oxf) 2003;59:427–30.
24. Morganti A, Pelizzola D, Mantero F, Gazzano G, Opocher G, 6. Stowasser M, Gordon RD, Gunasekera TG, Cowley DC, Ward G, Piffanelli A, on behalf of the Italian Multicenter Study for Standard- Archibald C, et al. High rate of detection of primary aldosteronism, ization of Renin Measurement. Immunoradiometric versus enzy- including surgically treatable forms, after ‘non-selective’ screening matic renin assay: results of the Italian Multicenter Comparative of hypertensive patients. J Hypertens 2003;21:2149 –57.
Study. J Hypertens 1995;13:19 –26.
7. Young WF Jr. Minireview: primary aldosteronism-changing con- 25. Lijmer JG, Mol BW, Heisterkamp S, Bonsel GJ, Prins MH, van der cepts in diagnosis and treatment. Endocrinology 2003;144: Meulen JHP, et al. Empirical evidence of design-related bias in studies of diagnostic tests. JAMA 1999;282:1061– 6.

Source: http://release.endoclab.pt/UploadImages/Downloads/DOWNLOAD_59.pdf