#NephJC Chat
Tuesday, February 25th, 2025, 9 pm Eastern on BlueSky
Hypertension. 2024 Jun;81(6):1391-1399. doi: 10.1161/HYPERTENSIONAHA.124.22721. Epub 2024 Mar 25.
Subtype Identification of Surgically Curable Primary Aldosteronism During Treatment With Mineralocorticoid Receptor Blockade
Giovanni Pintus, Teresa Maria Seccia, Laurence Amar, Michel Azizi, Anna Riester, Martin Reincke, Jiří Widimský, Mitsuhide Naruse, Tomaz Kocjan, Aurelio Negro, Gregory Kline, Akiyo Tanabe, Fumitoshi Satoh, Lars Christian Rump, Oliver Vonend, Peter J. Fuller, Jun Yang, Nicholas Yong Nian Chee, Steven B. Magill, Zulfiya Shafigullina, Marcus Quinkler, Anna Oliveras, Bo-Ching Lee, Chin-Chen Chang, Vin-Cent Wu, Zuzana Krátká, Michele Battistel, Domenico Bagordo, Brasilina Caroccia, Giulio Ceolotto, Giacomo Rossitto , Gian Paolo Rossi
PMID: 38525605
Introduction
Primary hyperaldosteronism (PA) is a great masquerader in medicine. Contrary to classic teaching (triad of resistant hypertension, hypokalemia, and metabolic alkalosis), PA easily camouflages in the garb of “essential hypertension”, “renoparenchymal hypertension” and now, even normotension! (Kitamoto T et al, Front Endocrinol, 2025). Erroneously presumed to be either too rare or too frequent an entity, estimating the prevalence of PA calls for ‘fuzzy logic’, taking into account ‘uncertainties and degrees of truth’! Various studies estimating the prevalence of PA bring forth intimidating numbers, and an inescapable incredulity over the estimates, which appear to be highly heterogeneous, and “pointless” for settling the question of prevalence (Kayser S et al, J Clin Endocrinol Metab 2016). What we, however, know more certainly now, is that the prevalence is high in resistant hypertension (up to 22%) (Parksook WW et al, J Clin Endocrinol Metab, 2024) and increases with age (elegant molecular demonstration by Nanba K et al, Circulation, 2017). 11% normotensives also appear to bear this high aldosterone signature (Brown JM et al, Hypertension 2025; NephJC Feb 2025). It is then ironic, how the workup of PA remains complicated, expensive, invasive, and consequently, distant. No wonder that less than 2% of all high-risk patients are screened for PA and <1% of these get diagnosed with PA (Vaidya A et al, Am J Hypertension, 2022).
Conn and colleagues (JAMA 1985- originally 1964) described 3 cardinal features of primary hyperaldosteronism which are the basis of the diagnostic workup for this entity to date:
Autonomous aldosterone production causing renin suppression
Blunted renin response to stimulation
Inability to suppress aldosterone despite volume expansion
The diagnostic workflow includes screening of high-risk patients, case detection, case confirmation, and subtype identification (Funder JW et al, J Clin Endocrinol Metab 2016, Faconti L et al, J Human Hypertension, 2024). With evolving evidence, the pendulum is swinging towards testing all patients of hypertension at least once for PA (Vaidya A et al, Am J Hypertension, 2022; McEvoy JW et al, European Heart Journal, 2024). The use of antihypertensive agents, especially those acting on the RAAS, often complicates the interpretation of tests at every step of this workflow.
For screening, the aldosterone renin ratio (ARR), which is often influenced by drugs (and other factors like hypokalemia, posture, circadian rhythm, salt intake, and assays used for detection) is the most commonly recommended and used test. Most guidelines recommend withdrawing agents affecting ARR for 4-6 weeks, a directive that applies to almost all anti-hypertensive agents, except for alpha-blockers, non-DHP CCBs, and hydralazine, making the evaluation a dangerous proposition while dealing with patients with a high requirement of anti-hypertensives. Careful interpretation of the ratio with a rightful emphasis on individual components (and not the ratio alone) is suggested (Faconti L et al, J Human Hypertension, 2024, Endocrine Society, 2016 and Italian Society of Arterial Hypertension, 2020). The EMIRA study (Rossi GP et al, J Clin Endocrinol Metab 2020) supported the continuation of MRA in patients of proven uPA being tested for ARR although AT1 receptor blockers seemed to increase the false negative rate.
Patients with high ARR are subjected to one confirmatory investigation, to be chosen from an array of physiology-based tests, that include oral salt suppression test (OSST), saline loading test (SIT), fludrocortisone suppression test (FST)- which has an enhanced version using dexamethasone, captopril challenge test (CCT) and furosemide upright test (FUT) used in Japan. Once PA is diagnosed, subtype identification is done using adrenal imaging followed by adrenal venous sampling (AVS) to differentiate a unilateral disease(adenoma/carcinoma/unilateral hyperplasia) from a bilateral disease (idiopathic hyperaldosteronism). In patients at high risk, often the confirmatory test is skipped and AVS is used both for confirmation and subtype identification (Faconti L et al, J Human Hypertension, 2024).
AVS, a specialized procedure introduced in 1957 (Masoni A, Acta Med Scand 1957), is considered the ‘gold standard’ over other biochemical, imaging, or scintigraphic methods, for detecting the laterality of the disease. The promise of curative unilateral adrenalectomy (laparoscopic>open), and very recently, even a ‘FABULAS’ alternative employing endoscopic, ultrasound-guided, radiofrequency ablation of APAs (Argentesi et al, Lancet 2025), is a better control of hypertension, reduction in the number of anti-hypertensives and mitigation of down-stream effects of MR activation, in addition to the adverse effects of hypertension alone (stroke, myocardial infarction, atrial fibrillation, and heart failure) (Samnani S et al, J Clin Endocrinol Metab. 2023, Huang W-C et al, Front Endocrinol 2021, Hundemer & Vaidya, Eur J Endocrinol 2020). It is indeed fascinating to note that certain teams across the world have followed the dogma-challenging practice of operating even in certain patients with bilateral disease (total or partial adrenalectomies, or ‘one and a half’ adrenalectomy) and the results appear compelling enough ( Williams TA et al. Lancet Diabetes Endocrinol 2022, Sukor N et al, J Clin Endocrinol Metab 2009).
Withdrawal of MRA or other RAAS-interfering drugs AND correction of hypokalemia (counter-intuitive, isn’t it?) is recommended before AVS by most guidelines. This recommendation stands on the premise that unsuppressed renin production due to the use of these drugs, may enhance aldosterone secretion in the non-responsive adrenal and may theoretically blunt the lateralization index (we shall discuss this index in detail in the methods section). The current study challenges this long-held and now-questioned practice, by investigating a large database, based on a multi-centric cohort of consecutive patients with PA undergoing AVS (AVIS-2) (Rossi GP et al, Hypertension 2019) and patients referred to the center for hypertension at the University of Padua (Italy).
The Study
Methods
The study analyzed consecutive patients undergoing adrenal vein sampling (AVS) from 2000-2015 (AVIS-2, 19 centers, 4 continents) (Rossi GP et al, Hypertension. 2019), and from 2016-2023 (University of Padua, Italy). Exclusion criteria included treatments with RAAS-interfering drugs (except MRAs), missing MRA therapy data, and unavailable AVS indices. AVS diagnostic accuracy was assessed using a biochemical cure of unilateral primary aldosteronism (uPA) as the gold standard per STARD guidelines. (Bossuyt PM et al. The BMJ, 2015) Patients without AVS lateralization or biochemical cure post-adrenalectomy were classified as non-uPA.
Ratios defined in the study for adrenal vein sampling interpretation are described in Table S1.
Table S1. Pintus G et al, Hypertension 2024
Primary end-point: unilateral PA (uPA) identification rates by lateralization index (Ll) between MRA-treated and untreated patients.
Secondary end-points: Assess MRA effects on AVS indices (Ll, relative aldosterone secretion index) and AVS accuracy in subtyping PA based on renin suppression.
Diagnostic criteria of uPA included biochemical cure i.e. normalization of hypokalemia without potassium supplementation, of ARR, and of renin suppression after adrenalectomy of the involved side.
For the primary endpoint, only bilateral selective AVS procedures were included. Bilateral selective AVS refers to cases where blood samples were successfully obtained from both the right and left adrenal veins in a manner that meets the protocol. AUROC analysis was used to evaluate the accuracy. Propensity score matching was applied to mitigate the confounding with four models built based on sampling methods, covariates, and missing data handling. Sensitivity analysis examined the impact of renin suppression on uPA identification and LI accuracy.
Funding: The study was partly supported by research grants from the Departments of Medicine Dotazione, Ordinaria della Ricerca, University of Padua, and The Foundation for Advance Research in Hypertension and Cardiovascular Diseases. The sponsors had no role in the study design, data collection, analysis, interpretation, or article writing.
Results
The study included 1746 patients with PA undergoing AVS (1625 from AVIS-2, 121 from Padua). After excluding patients with incomplete data or RAAS-interfering drugs, the final cohort included 840 patients: 61 patients on MRA, and 779 patients without MRA at AVS.
Figure 1. Flowchart of the study, from Pintus G et al, Hypertension, 2024
No significant differences in plasma aldosterone, renin, aldosterone-to-renin ratio, or BP were found between groups. MRA-treated patients had higher potassium levels in the Padua cohort (3.7 vs 3.4 mEq/L), required more antihypertensive treatment (2.25 vs 1.5 defined daily dose), and had more adrenal nodules (82 vs 67%).
(Defined daily dose as per WHO is the assumed average maintenance dose per day for a drug used for its main indication in adults.)
Table 1. Baseline demographics, from Pintus G et al, Hypertension, 2024
Adrenalectomy rates were higher in the MRA group (72% vs 54%), but biochemical cure rates were identical (94%) (P.S.: post-surgical follow-up data present only for 64% of patients who underwent adrenalectomy).
Bilaterally Selective AVS
The table compares AVS selectivity rates between patients on MRAs and those not on MRAs, showing no significant differences between the groups. Unstimulated selectivity rates were 83% on the right and 87% on the left, while post-cosyntropin rates improved to 87% and 99%, respectively. Left-side selectivity was consistently higher, likely due to technical challenges in right adrenal vein cannulation. The increase in selectivity post-cosyntropin highlights its role in AVS. Overall, MRA use did not impact selectivity, questioning the necessity of discontinuation before AVS.
Table S3. Rate of selective procedures per side, from Pintus G et al, Hypertension, 2024
Diagnostic accuracy
AUROC for uPA identification was similar between MRA (0.960, 95% CI 0.821-0.998) and non-MRA (0.937, 95% CI, 0.908-0.959) groups (P = 0.59). Postcosyntropin AUROC also showed no differences (P = 0.938).
Figure 2. Receiving operating characteristic (ROC) curves by MRA treatment. A- Unstimulated adrenal vein sampling (AVS) accuracy. B- Postcosyntropin AVS accuracy. from Pintus G et al, Hypertension, 2024
AVS indices were comparable across groups.
Table 2. AVS indices, from Pintus G et al, Hypertension, 2024
Propensity Score Matching
Four PSM models showed no significant Ll differences between MRA and non-MRA groups. A slight increase in lateralization in the MRA group (P= 0.03) in model 1 was not confirmed in models 2-4.
Figure S1. Propensity score matching flowchart from Pintus G et al, Hypertension, 2024
Renin Suppression and AUROC
There were no significant differences in AUROC for uPA identification between undetectable, suppressed, and unsuppressed renin groups.
Table S4. Comparison of AUROC for LI under unstimulated conditions by active renin (DRC) levels subgroups, from Pintus G et al, Hypertension, 2024
Ll and uPA identification rates were similar between MRA and non-MRA groups across ronin suppression subgroups.
Table S5. Lateralization index (LI) in MRA versus non-MRA patients by active renin (DRC) level subgroups, from Pintus G et al, Hypertension, 2024
Discussion
PA is notorious for difficult-to-control hypertension and hypokalemia, at times even leading to hypokalemic paralytic episodes. Clinical practice guidelines (Endocrine Society, 2016 and Italian Society of Arterial Hypertension, 2020) recommend withdrawal of MRA for 4-6 weeks in patients with a florid clinical phenotype of PA when planning subtyping by AVS. The risk of precipitating a hypertensive crisis by withdrawing drugs active on RAAS is justified only for the expectation of a surgical cure of hypertension and regression of hypertension/MR activation-mediated organ damage. In this study despite MRA continuation, 22.9% of patients required potassium supplementation. Additionally, systolic BP was much higher than standard targets in both groups (153 mm Hg in the MRA group vs 149 mm Hg in the MRA withdrawal group).
Following several small studies (Haase M et al, J Clin Endocrinol Metab. 2014; Nanba AT et al, J Clin Endocrinol Metab. 2018; Nagasawa M et al, J Hypertension 2019; Ganesh M et al, Endocrine Prac 2020; Torresan et al, Clin Sci 2020), this large study, although a retrospective analysis, appears to be the crowning evidence showing that MRA treatment does not endanger the identification of unilateral PA. Recreating similar results in randomized controlled trials is desirable but probably not feasible since there may be a class of patients who may not be appropriate for safe MRA withdrawal. The MRA-treated patients in this study were biochemically similar to the non-MRA treated patients, suggesting no differences in the underlying metabolic signature of PA (PAC, active renin, ARR, serum potassium values, and even blood pressure values) between the two groups. However, results from the Padua study, the Japanese Primary Aldosteronism study (JPAS study: Nagasawa M et al, J Hypertension 2019), and other studies hint at the possibility of a more clinically florid disease in the MRA-requiring group. The MRA group appeared to require more anti-hypertensive drug therapy and had a propensity towards hyperkalemia when MRAs were uptitrated to the maximum tolerated dose in these studies. They were also more likely to show adrenal nodules at imaging.
The authors undertook a more nuanced sensitivity analysis and looked at the identification of unilateral disease in 3 scenarios—with renin undetectable, suppressed, or unsuppressed. No significant differences were identified, thus reinforcing the primary message that suppressing renin was not truly necessary while planning an AVS. Furthermore, MRAs did not affect the results of the cosyntropin-stimulated test. The conclusions from the study make a great piece of news because they do serve to take off a load of recommendations from the pre-procedure checklists followed for decades. (Funder JW, et al. J Clin Endocrinol Metab. 2016) The only hiccup is that the overall number of patients in the MRA group was small (n = 61), especially the unsuppressed renin group had a meager five patients, limiting the generalisability of the observations. None of the study patients were on ACEi/ARBs, and hence a wider generalization to these drug classes as well cannot be made from the results of this study. However, the JPAS study (Nagasawa M et al, J Hypertension 2019) did provide sufficient evidence to support the continuation of ACEi/ARBs as well.
Limitations of the study include its retrospective nature and propensity score matching instead of the more desirable randomization (may have created an imbalance in the confounders). Also, having only 61 patients on MRAs, a consequence of strict exclusion criteria, brings in a higher chance of accepting the null hypothesis due to a limited power for the analysis. Another disadvantage of a small sample size is not enough representation from different ethnicities to draw inferences on the matter at hand.
Conclusion
The study is an important milestone and serves to encourage physicians to order an AVS in those deemed high risk for the procedure, without the fear of rocking the boat. All said and done, surgical adrenalectomy does achieve better CV outcomes in PA than medical therapy, hence justifying the need to identify the potentially curable hypertensives in our clinic. For the rest, RASi and MRAs can continue working their customary magic.