Summary
Background
Biological diagnostic criteria for diagnosing aldosterone-producing adenoma (APA) are not well-established.
Aim
The aim of the study was to establish the best biological predictors of APA.
Methods
A prospective register was implemented in 17 secondary or tertiary hypertension centres. The inclusion criterion was one of the following: onset of hypertension before 40 years of age; history of hypokalaemia; drug-resistant hypertension (resistant to three drugs); or spironolactone efficiency on BP.
Results
Among the 338 collected cases, 192 patients had two aldosterone-to-renin ratio (ARR) determinations (after 1 hour supine and at least 1 hour upright) on the same occasion. Twenty-five patients (8.2%) had biological hyperaldosteronism and an adrenal adenoma identified by computed tomography. APA was histologically confirmed in all 12 patients who underwent surgery. Histologically proven APAs were used as the ‘gold standard’ in receiver operating characteristic (ROC) curve analysis. ARRs were computed with a minimum renin value set at 5 ng/L to avoid misclassification of so-called ‘low-renin hypertension’. To predict an APA, the ARR area under the ROC curve was 0.93. A supine ARR cut-off value of 32 ng/ng provided the highest sum of sensitivity (92%) plus specificity (92%). On the basis of an ARR ≥ 32 ng/ng in the supine and/or upright position, sensitivity reached 100%.
Conclusion
The proposed cut-off value of 32 ng/ng for ARR (minimum renin value set at 5 ng/L) in one of two determinations had 100% sensitivity and 72% specificity with 20% positive and 100% negative predictive values for diagnosing APA.
Résumé
Contexte
Les critères biologiques pour le diagnostic d’un adénome de Conn ne sont pas bien établi.
Objectif
L’objectif de cette étude est d’établir les seuils biologiques d’aldostérone et de rénine permettant de prédire un adénome de Conn.
Méthodes
Un registre prospectif a été mis en œuvre dans 17 centres de l’hypertension. Le critère d’inclusion dans le registre était l’un des éléments suivants : apparition de l’hypertension avant 40 ans, antécédent d’hypokaliémie, hypertension artérielle résistante ou efficacité trop importante de la spironolactone dans le traitement de l’hypertension.
Résultats
Parmi les 338 cas recueillis, 192 patients ont eu deux déterminations successives de rénine et d’aldostérone (une heure en position couchée et au moins une heure debout). Vingt-cinq patients (8,2 %) avaient un hyperaldostéronisme biologique et un adénome surrénalien identifié par une tomodensitométrie. Une histologie compatible avec un adénome de Conn a été confirmée chez les les 12 patients qui ont été opérés. Un adénome de Conn histologiquement confirmé a été utilisé comme « gold standard » pour l’établissement des courbes ROC. Les rapports aldostérone/rénine (RAR) ont été calculés avec une valeur minimale de rénine fixée à 5 ng/L. Afin de prédire un adénome de Conn, l’aire sous la courbe ROC du RAR était de 0,93. Une valeur de RAR couchée supérieure à 32 ng/ng avait les sensibilités (92 %) et spécificité (92 %) les plus élevées. Sur la base d’un RAR supérieur à 32 ng/ng en décubitus dorsal et/ou en orthostatisme, la sensibilité atteint 100 %.
Conclusion
Le seuil de RAR à 32 ng/ng (valeur minimale de rénine à 5 ng/L) lors de l’une des deux déterminations, soit en décubitus dorsal, soit en orthostatisme a une sensibilité de 100 %, une spécificité de 72 %, une valeur prédictive positive de 20 % et une valeur prédictive négative de 100 % pour le diagnostic d’un adénome de Conn.
Background
Primary aldosteronism (PA) is one of the commonest forms of secondary hypertension and diagnosis of PA has received much attention in recent years . Once PA is confirmed, the subtype must be determined in order to guide treatment. Almost one-third of PA cases are due to aldosterone-producing adenoma (APA) and are surgically curable. Two-thirds of PA cases are related to idiopathic adrenal primary aldosteronism, frequently with nodular adrenal hyperplasia . APA has been reported to carry a higher risk of left ventricular hypertrophy, stroke and chronic kidney disease . A surgical alternative to long-term drug therapy would reduce costs and increase patients’ quality of life. PA was initially defined as hypertension associated with increased aldosterone and low renin concentrations. High concentrations of aldosterone lead to the sodium retention responsible for hypertension and renal potassium loss causing hypokalaemia. However, in half of PA cases, kalaemia is within the normal range and sometimes even blood pressure (BP) is normal. Thus, diagnosis of PA is not easy and screening for possible PA, which was initially restricted to hypertension with hypokalaemia, had to be extended to cases of resistant hypertension . The diagnosis of PA remains biological, associating high concentrations of aldosterone with low concentrations of renin. Many authors have used the aldosterone-to-renin ratio (ARR) to define PA and several cut-off values have been proposed . Furthermore, to avoid diagnosis of low-renin hypertension, minimum values for renin and/or aldosterone have been advocated for calculating the ARR. As noted by Kaplan , there are ‘considerable differences in the definition of an elevated ARR’. Two strategies have been used to define the ARR. The first, representing the upper values obtained in unselected essential hypertensive subjects, led to a lower cut-off value and a consequently higher percentage of PA (up to 39% of hypertensives). The second is based on minimal ARR values in patients with an APA. A more suitable strategy used by Bernini et al. to assess ARR cut-off values is based on receiver operating characteristic (ROC) analysis. In their single centre study, they proposed an aldosterone-to-plasma renin activity cut-off value and reported a high percentage of APA (8.4%) and IHAP (author to define at proof stage) (12.7%) in newly diagnosed hypertensives. In the PAPY study, Rossi et al. reported that ARR based on measurement of active renin is a valuable alternative to that based on plasma renin activity. Using a similar strategy, we conducted a prospective multicentre study to determine an ARR threshold for detecting an APA.
Methods
A prospective register was implemented in 17 secondary or tertiary hypertensive centres. The inclusion criterion was one of the following: onset of hypertension before 40 years of age; history of hypokalaemia; drug-resistant hypertension (resistant to three drugs); or spironolactone efficiency on BP. The register opened in October 2006 and closed in June 2007 when 300 consecutive case reports were recorded as specified in the study protocol.
For each patient, the following data were collected: anthropomorphic variables and lifestyle (weight, height, smoking habits, alcohol consumption, liquorice consumption); hypertension history (duration, family history, efficiency of spironolactone if available and potassium supplementation; sitting BP [measured three times at 2-minute intervals after a 5-minute rest; the averages of the three BP measurements were calculated; validated automatic sphygmomanometers were used in all centres]); cardiovascular complications (myocardial infarction, arrhythmia and stroke); biological variables (kalaemia, creatininaemia, plasma bicarbonate, 24-hour urinary excretion of creatinine, sodium, potassium, albumin and aldosterone); plasma aldosterone and active renin (both expressed in ng/L), determined after 1 hour in a supine position and after at least 1 hour in an upright position.
A minimum plasma renin value of 5 ng/L was set to calculate the ARR. Treatments that might interfere with the renin-angiotensin system were not allowed; however, oral contraceptive and hormonal replacement therapies were not stopped or modified. A wash-out period prior to ARR screening was 2 weeks for angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, beta-blockers and diuretics (except for spironolactone) and 4 weeks for spironolactone; however, in 22 patients with ischaemic myocardial disease, beta-blockers could not be interrupted. When necessary, authorized treatments were calcium channel blockers (non-dihydropyridine or long-acting dihydropyridine), central-acting agents (rilmenidine) and alpha-blockers (slow-release prasozin, urapidil). Sodium diet and potassium supplementation (if applicable) were not modified. Cases reports containing unauthorized drugs were discarded ( n = 6).
Using complete case report forms, a blind outcome committee (J.-P.F. and J.-P.B.) established the diagnosis of PA. The ARR cut-off value of 23 ng/ng was used to suspect a PA, with the minimal renin concentration set at 5 ng/L. This cut-off value was chosen because it was the lowest reported in the literature . However, in five case reports, the ARR was greater than 23 ng/ng and computed tomography (CT) scans were not available; the diagnosis could not be made and these cases were excluded for incomplete data. Actually, in the remaining patients, the minimum ARR value used by the investigators for high-resolution angiographic CT-scan was 14.5 ng/ng. Thus, a CT-scan was also performed in 99 patients (70%) in whom the ARR was lower than 23 ng/ng. Twenty-two other case reports were also excluded due to incomplete data.
Statistical analysis
Data were expressed as mean values ± standard deviations. Mean values between groups were compared using Student’s t test. Categorical variables were compared using the Chi 2 test. ROC analysis was used to evaluate the diagnostic power of different ARR levels. To estimate classification accuracy using ROC analysis, the disease status of each patient had to be validated without error. Thus, we chose the histologically proven cases of APA as the gold standards for the ROC analysis. ARR is a continuous-scale measurement. Once the diagnosis of APA for each patient was determined, a 2 × 2 contingency table was made. Sensitivity (i.e. true positive rate) and specificity (i.e. true negative rate) were calculated for each ARR cut-off value. A ROC curve that plotted sensitivity against (1–specificity) was drawn. The closer the area under the ROC curve was to 1, the more accurate the diagnostic test. The Younden score (sensitivity + specificity–1) provided the optimal ARR cut-off value to diagnose APA. Data analysis used Medcalc ® 11.5.1.0 software.
Results
As shown on the flow-chart ( Fig. 1 ), among the 338 collected cases, 305 met all the inclusion criteria, 192 patients had two ARR determinations (supine and upright) and 25 patients (8.2%) had biological PA and an adrenal adenoma. Among them, 12 patients chose adrenalectomy and 13 preferred pharmacological treatments. Five patients (two among those who chose adrenalectomy) who had biological PA and an adrenal adenoma were taking a beta-blocker. Among the 12 operated patients, only six had adrenal venous sampling (AVS); in each case, it confirmed the unilateral aldosterone secretion. No technical failure was reported by the investigators. In all those who underwent unilateral adrenalectomy, an APA was confirmed by histology. Patients treated by spironolactone achieved either normalization ( n = 9) or significant reduction ( n = 4) in BP (BP decrease > 20% or use of fewer antihypertensive drugs).
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
