Author + information
- Received August 18, 2016
- Revision received January 6, 2017
- Accepted January 10, 2017
- Published online April 3, 2017.
- Silvia Monticone, MD, PhDa,
- Jacopo Burrello, MDa,
- Davide Tizzani, MDa,
- Chiara Bertello, MDa,
- Andrea Viola, MDa,
- Fabrizio Buffolo, MDa,
- Luisa Gabetti, MDb,
- Giulio Mengozzi, MDc,
- Tracy A. Williams, PhDa,d,
- Franco Rabbia, MDa,
- Franco Veglio, MDa and
- Paolo Mulatero, MDa,∗ ()
- aDivision of Internal Medicine and Hypertension Unit, Department of Medical Sciences, University of Torino, Torino, Italy
- bAzienda Sanitaria Locale, Primary Care Trust, Torino, Italy
- cClinical Chemistry Laboratory, University of Torino, Torino, Italy
- dMedizinische Klinik und Poliklinik IV, Klinikum der Universität München, Ludwig-Maximilians-Universität München, Munich, Germany
- ↵∗Address for correspondence:
Prof. Paolo Mulatero, Division of Internal Medicine and Hypertension, Department of Medical Sciences, University of Torino, Via Genova 3, 10126, Torino, Italy.
Background Despite being widely recognized as the most common form of secondary hypertension, among the general hypertensive population the true prevalence of primary aldosteronism (PA) and its main subtypes, aldosterone-producing adenoma (APA) and bilateral adrenal hyperplasia (BAH), remains a matter of debate.
Objectives This study sought to determine the prevalence and clinical phenotype of PA in a large cohort of unselected patients with hypertension, consecutively referred to our hypertension unit, by 19 general practitioners from Torino, Italy.
Methods Following withdrawal from all interfering medications, patients were screened for PA using the ratio of serum aldosterone to plasma renin activity. PA was diagnosed according to Endocrine Society guidelines. The diagnosis was confirmed or excluded by an intravenous saline infusion test or captopril challenge test and subtype differentiation was performed by adrenal computed tomography scanning and adrenal vein sampling, using strict criteria to define successful cannulation and lateralization of aldosterone production.
Results A total of 1,672 primary care patients with hypertension (569 newly diagnosed and 1,103 patients already diagnosed with arterial hypertension) were included in the study. A total of 99 patients (5.9%) were diagnosed with PA and conclusive subtype differentiation by adrenal vein sampling was made in 91 patients (27 patients with an APA and 64 patients with BAH). The overall prevalence of PA increased with the severity of hypertension, from 3.9% in stage 1 hypertension to 11.8% in stage 3 hypertension. Patients with PA more frequently displayed target organ damage and cardiovascular events compared with those without PA, independent of confounding variables.
Conclusions Our results demonstrated that PA is a frequent cause of secondary hypertension, even in the general population of patients with hypertension, and indicates that most of these patients should be screened for PA.
Primary aldosteronism (PA) is a heterogeneous group of disorders characterized by hypertension and aldosterone overproduction relatively autonomous from the renin-angiotensin system. The importance of a correct diagnosis of PA is demonstrated by the increased risk of cardiovascular (CV) and cerebrovascular complications and the increased rate of metabolic syndrome of patients with PA compared with patients with primary hypertension with matched CV risk profiles (1–3). According to Endocrine Society guidelines (4,5) the diagnosis of PA is a 3-step process comprising screening, confirmatory testing, and subtype differentiation. The latter is necessary to distinguish the 2 most common PA subtypes, unilateral aldosterone-producing adenoma (APA) and bilateral adrenal hyperplasia (BAH), and thereby stratify patients to the recommended therapeutic management of unilateral adrenalectomy or medical therapy with mineralocorticoid receptor antagonists (MRA).
Until the late 1990s, PA was suspected only in the presence of overt and spontaneous hypokalemia, resulting in low estimated prevalence rates from 1% to 2% (6,7). The subsequent widespread application of the aldosterone-to-renin ratio (ARR) as a screening test for unselected patients with hypertension resulted in a 5-to-15-fold increase in the diagnosis of PA (8,9). A large number of studies have investigated the prevalence of PA and have reported wide variations in the estimates of patients referred to hypertension units (1% to 29.8%) and in primary care patients (3.2% to 12.7%) (10), depending on the characteristics of the selected population and the diagnostic method used. In particular, PA prevalence increases with the severity of hypertension, from 2% in patients with grade 1 hypertension (11) to 20% among individuals with resistant hypertension (12). However, most studies were conducted in tertiary hypertension units and referral bias is likely to have influenced the results. For this reason, the true prevalence of PA among the general hypertensive population and the importance of extending the measurement of the ARR in most patients with hypertension remains a matter of debate.
The present study sought to determine the prevalence and clinical phenotype of PA and its main subtypes, APA and BAH.
The protocol of the PATO (Primary Aldosteronism in Torino) study was approved by the ethical committee of the University of Torino and written informed consent was obtained from all patients participating in the study. To avoid selection bias, the study population was recruited directly from 19 general practitioners (GP) in Torino, Italy. All patients 18 to 60 years of age affected by arterial hypertension (newly diagnosed or previously diagnosed) consecutively presenting to the GP’s office from 2009 to 2014 were invited to participate; 1,672 patients were included in the study and the overall response rate was 51%.
All patients who agreed to participate were contacted by telephone by a physician from our hypertension unit to collect medical and pharmacological history and to schedule the initial visit. In the absence of any medical condition requiring mandatory medication, all drugs interfering with the renin-angiotensin-aldosterone system were withdrawn, as detailed later. During visit 1, patients underwent physical examination including blood pressure (BP) measurement, familial medical and pharmacological history, and the collection of blood samples to assess plasma renin activity (PRA) and measure aldosterone levels.
The group of 19 GPs participating in the PATO study cared for a total of 23,067 patients, 13,090 of whom were ages 18 to 60 years; 3,272 were hypertensive (either newly diagnosed or with pre-existing hypertension) and these patients were invited to participate in the study. A final cohort of 1,672 patients was included in the PATO study. Information about the patients who were not recruited is available in the Online Appendix and Online Table 1.
The diagnosis of hypertension was made by 3 consecutive BP measurements in the sitting position using a mercury sphygmomanometer, according to European Society of Hypertension guidelines (13). The diagnosis of PA was made according to the recommendation of the Endocrine Society guidelines (4). In particular, patients were screened for PA using the serum aldosterone concentration (AC) to PRA ratio (ARR); all interfering medications were withdrawn for at least 4 weeks (6 weeks for diuretics and MRAs) and hypokalemia, if present, was corrected with potassium (K+) supplements. A calcium-channel blocker and/or doxazosin, which are known to have minimal effect on the ARR, were used to control BP when discontinuation of all antihypertensive medications was not feasible because of the risk of uncontrolled hypertension (4). The cutoff level considered to be a positive ARR was 30 (ng/dl per ng/[ml-1 × h-1]) (830 pmol/l per ng/[ml-1 × h-1]), together with an aldosterone level >10 ng/dl (277 pmol/l). The diagnosis of PA was confirmed (visit 2, Figure 1) by an intravenous saline loading test (2 liters of 0.9% sodium chloride infused over 4 h), which was considered positive if post-test aldosterone levels were >5 ng/dl (138.7 pmol/l) (14). Captopril challenge test was used to confirm the diagnosis of PA in those patients with a contraindication to acute volume expansion (e.g., poorly controlled hypertension, chronic kidney failure, heart failure, and previous CV events). Captopril 50 mg was administered orally and PRA and aldosterone were measured after 2 h. PA was confirmed if the post-test ARR was >30 ng/dl per ng/(ml-1 × h-1). Subtype differentiation was performed as previously reported (15). Finally, long polymerase chain reaction amplification was used for glucocorticoid remediable aldosteronism testing, which was performed in all confirmed PA patients.
PRA levels were classified as follows: low renin if PRA was <1 ng/(ml-1 × h-1), normal renin if PRA levels were between 1 and 4 ng/(ml-1 × h-1), and high renin if PRA levels were >4 ng/(ml-1 × h-1). Hypokalemia was defined as serum potassium <3.6 mEq/l. We considered the following CV events: myocardial infarction and unstable angina requiring angioplasty (in the text grouped together under coronary artery disease), stroke or transient ischemic attack (in the text together under stroke), sustained arrhythmias (atrial fibrillation, atrial flutter, sustained ventricular tachycardia, and ventricular fibrillation) demonstrated by electrocardiogram, and heart failure requiring hospitalization (2). Definitions of cutoffs for target organ damage are in the Online Appendix. Because other forms of secondary hypertension were not systematically excluded, patients without PA were considered as non-PA hypertensive patients (apparent essential hypertension).
Blood samples for AC and PRA measurements were collected in the morning between 7:00 am and 9:00 am at the Città della Salute e della Scienza di Torino University Hospital centralized laboratory with the patients in the sitting position. PRA and AC were measured by radioimmunoassay as described previously (16). AC was assessed by solid-phase radioimmunoassay ALDOCTK-2 (DiaSorin, Saluggia, Italy) and PRA using the RENCTK RIA kit (DiaSorin) according to the manufacturer's instructions. The analytical sensitivity was 0.1 ng/ml/h (16).
IBM SPSS Statistics version 21.0 (IBM Corp., Armonk, New York) was used for statistical analyses. Data were analyzed with the Kolmogorov-Smirnov test to determine their distributions. Results are expressed for continuous variables with a normal distribution as mean ± SD and with non-normal distributions as median (interquartile range). Statistical significance between groups was calculated in normally distributed data by Student t test for independent samples and for non-normally distributed data by Kruskal-Wallis and Mann-Whitney U test. One-way analysis of variance followed by a Bonferroni test was used to compare quantitative variables between groups and chi-square test or the Fisher exact test for qualitative variables. Multivariate logistic regression analysis was performed to correct for confounding variables, including age, male sex, duration of hypertension, systolic blood pressure (SBP), body mass index, high-density lipoprotein cholesterol, smoking habit, diabetes, and PA diagnosis.
Between 2009 and 2014 a total of 1,672 patients agreed to participate and were included in the study. The design of the study, comprising the diagnostic work-up for PA, is shown in Figure 1.
Clinical and demographic features of the included patients are summarized in Table 1. Overall, the average SBP and diastolic BP levels were 147 ± 15 mm Hg and 94 ± 8 mm Hg, respectively; 1,133 (67.8%) patients had stage 1 hypertension, 413 (24.7%) stage 2 hypertension, and 126 (7.5%) stage 3 hypertension. Of the 1,672 included patients, 569 were newly diagnosed with hypertension and 1,103 were previously diagnosed as having hypertension at the time of recruitment. Clinical and biochemical parameters of these 2 subpopulations are shown in Online Table 2. Unsurprisingly, patients with newly diagnosed hypertension were younger and had lower SBP, and prevalence of CV events compared with individuals with established hypertension.
Following diagnostic work-up, comprised of screening, confirmatory testing, and subtype differentiation, a total of 99 (5.9%) patients were affected by PA. Specifically, 33 of the 99 PA diagnoses had newly diagnosed hypertension (5.8% prevalence) and 66 patients with PA had a pre-existing diagnosis of hypertension (6.0% prevalence). A conclusive subtype diagnosis was achieved by adrenal vein sampling (AVS) in 91 of the 99 patients: 27 had an APA and 64 had BAH; in the remaining 8 patients, the subtype was undetermined because AVS was not performed or not diagnostic. No complications occurred during AVS procedures. The overall prevalence of PA and its subtypes was 3.9% in hypertension stage 1 (21% APA, 68% BAH, and 11% undetermined), 9.7% in hypertension stage 2 (30% APA, 62.5% BAH, and 7.5% undetermined), and 11.8% in hypertension stage 3 (40% APA and 60% BAH) (Figure 2). Hypokalemia (defined as serum K+ <3.6 mEq/l) was prevalent in 105 of 1,544 (6.8%) in the overall study population, 77 of 1,447 (5.3%) among the non-PA hypertensives, and 29 of 99 (29.3) in the PA subgroup (52% in APA and 22% in BAH; p = 0.013).
Screening for PA and comparative analysis
Of the 1,672 patients included in the analysis, according to PRA values, 34% displayed low-renin (including PA patients and low-renin essential hypertension [LREH]), 57% normal-renin, and 9% high-renin levels (for further evaluations, normal- and high-renin patients are combined). The LREH category was composed of 2 subgroups of patients: those with PRA <1 ng/ml/h and ARR <30 (negative screening test, n = 331); and those with PRA <1 ng/ml/h and ARR >30 (positive screening test and negative confirmatory testing, n = 133).
A total of 232 patients (13.8%) had an ARR >30 together with an absolute AC >10 ng/dl (positive screening test) and underwent confirmatory testing (187 intravenous saline loading test and 45 captopril challenge test).
Clinical and biochemical parameters of patients according to final diagnosis (PA vs. non-PA) are summarized in Tables 1, 2, and 3 and Online Table 3. Patients with PA displayed higher SBP and diastolic BP than non-PA and, in agreement with the classical PA phenotype, higher serum aldosterone levels and lower PRA and K+ levels. Metabolic syndrome was more frequent in PA patients than non-PA; furthermore, they displayed greater waist circumference, and had lower high-density lipoprotein cholesterol, higher fasting plasma glucose, and more prevalent hypertriglyceridemia. For target organ damage, PA patients presented more frequently with left ventricular hypertrophy (evaluated by electrocardiogram and/or echocardiography) and microalbuminuria; furthermore, PA patients displayed a higher rate of CV events at diagnosis compared with non-PA hypertensives. Interestingly, in a multivariate logistic regression model, the diagnosis of PA was a strong factor associated with the occurrence of CV events in our study population ages 18 to 60 years (odds ratio: 2.1; 95% confidence interval: 1.1 to 3.9), after correction for age, sex, duration of hypertension, SBP, body mass index, high-density lipoprotein cholesterol, smoking habit, and presence of diabetes mellitus (Online Table 4). Accordingly, the diagnosis of PA was a strong factor associated with the occurrence of left ventricular hypertrophy (odds ratio: 2.1; 95% confidence interval: 1.3 to 3.2), after correction for age, duration of hypertension, and SBP (Online Table 5) and with microalbuminuria (odds ratio: 2.6; 95% confidence interval: 1.5 to 4.4), after correction for age, duration of hypertension, SBP, and presence of diabetes mellitus (Online Table 6).
AVS was successfully performed in 91 patients, displaying unilateral aldosterone overproduction in 27 patients (1.6% of the total study population) and bilateral secretion in 64 (3.8% of the study population). Overall, APA patients were characterized by higher serum aldosterone levels, higher ARR, and a trend toward lower K+ levels compared with BAH patients (Table 4). Of note, no significant differences in the total number of CV events at diagnosis were observed between the 2 subgroups. However, the low number of APA patients could have affected the results and limits any conclusions to be drawn.
Comparison of clinical and biochemical parameters of the different subpopulations according to the renin profile are shown in Online Tables 7 and 8. Interestingly, patients with LREH and patients with normal-high renin values display a similar clinical phenotype in terms of K+ levels, prevalence of metabolic syndrome, target organ damage, and CV events, despite higher SBP levels in LREH subpopulation.
Subsequently, we compared the clinical and biochemical features between LREH patients who had positive versus negative ARR screening tests. As expected, PRA levels were lower and aldosterone levels higher in the LREH patients who tested positive at ARR, but these 2 populations did not differ significantly in terms of other clinical parameters and, in particular, displayed a similar prevalence of target organ damage and CV events at diagnosis.
Additional results are in the Online Appendix.
Arterial hypertension affects up to 30% of the Italian population ages 35 to 74 years (17) and represents the single most important contributor to the global burden of disease and mortality (18). Despite this, only a minority of patients with hypertension attain optimal BP control (19), caused in part by the underdiagnosis and targeted treatment of underlying secondary forms of hypertension, such as PA.
The introduction of the ARR as a screening test for PA prompted numerous studies investigating the prevalence of PA in different populations. Some 12 studies have been conducted on patients from primary care centers (10,20,21), representing an apparently unselected hypertensive population; these reported a wide variation in PA prevalence, ranging between 3.2% and 13%. Nevertheless, as suggested by the heterogeneity of the results, these studies could have been biased by the various criteria adopted for patient selection and diagnosis. In fact, only 2 of these studies were published after the 2008 release of the Endocrine Society guidelines for diagnosis and treatment of PA (4). Additionally, only 1 study was performed in more than 1,000 patients (22), subtype differentiation was not performed systematically with AVS, and target organ damage and CV events were not evaluated. Recently, a survey on the knowledge and application of the Endocrine Society PA guidelines by GPs in Italy and Germany demonstrated a low rate of PA diagnosis (1% and 2%, respectively) in primary care patients with hypertension (23).
In this study, we prospectively investigated the prevalence of PA and its main subtypes, APA and BAH, in a large cohort of 1,672 unselected patients with hypertension consecutively recruited by 19 GPs in Torino, Italy. We observed an overall PA prevalence of 5.9% and an APA was found in 27% of the patients diagnosed with PA (1.6% of the total study population) (Central Illustration). Interestingly, similar estimates of PA prevalence were observed between the newly diagnosed and the established hypertensive populations, thereby excluding the occurrence of any referral or selection bias for patients with established hypertension. Importantly, PA prevalence was not rare even in patients with relatively mild hypertension (3.9% in stage 1 and 9.7% in stage 2, with a consistent proportion of the potentially curable unilateral PA). It should be noted that, in addition to the hypertensive population with overt PA (5.9%), another 133 (8%) patients displayed inappropriate aldosterone production for their renin status, despite being still suppressible with a saline load and 231 (13.8%) patients had an ARR between 20 and 30. This population of patients with hypertension has been shown to respond particularly well to therapy with MRAs (24,25).
In the present study, we demonstrated that the prevalence of PA is largely underestimated in the general hypertensive population, and a wider application of the simple ARR screening test would benefit a consistent proportion of patients with hypertension by offering the possibility of surgical cure or targeted pharmacotherapy. Spontaneous or diuretic-induced hypokalemia, traditionally considered a prerequisite for pursuing diagnostic testing for PA (26), was detected in 6.8% of the hypertensive population and in 28.9% of our PA cohort, making normokalemic hypertension the most common presentation of the disease, as previously observed in a study including centers from 5 continents (9).
In our study, patients affected by PA presented a more severe clinical phenotype, as demonstrated by the increased prevalence of the disease according to the severity of hypertension, in agreement with earlier reports (2,11). We also observed an increased rate of CV and cerebrovascular events in PA patients compared with non-PA patients with hypertension that were at least partly independent of BP levels, consistent with previous studies in selected hypertensive populations (2,27,28). Similarly, target organ damage (left ventricular hypertrophy and microalbuminuria) was also associated with PA, in part, independent of confounding variables.
The present study suggested that most patients with hypertension should be screened with the simple ARR test, in agreement with the Japanese Endocrine Society guidelines (29). Some concerns could arise from the consequences of widespread use of the ARR; confirmatory or exclusion tests are poorly standardized; plus, it may lead to an increase of expensive and invasive subtype diagnostic procedures (e.g., computed tomography and AVS). However, it has been shown that PA diagnosis and treatment is feasible and economical (30,31). Furthermore, an alternative strategy with patients displaying high ARR and relatively mild hypertensive disease could be strict adherence to a low-sodium diet (32) and low-dose MRA therapy (5).
Strengths of our study included our use of Endocrine Society guidelines to diagnose PA (4,5), the screening of all patients using the ARR, and confirmation of a PA diagnosis in all patients. The ARR is the most reliable means to screen for PA, but a unique cutoff is difficult to define because of the wide variability of laboratory assays and population-speciﬁc characteristics (4,5). To increase the test’s sensitivity, we adopted 30 ng/dl per ng/(ml-1 × h-1) as the cutoff to define a positive screening test, in addition to a minimum AC of 10 ng/dl, instead of the less sensitive but more specific cutoff routinely used in our hypertension unit (ARR >40 and AC >15 ng/dl). Moreover, AVS was performed in all PA patients and strict criteria were used to define both successful cannulation of the adrenal glands and lateralization of aldosterone production (15,33); our study was also the first to investigate in a population of primary care patients with hypertension the presence of target organ damage and the rate of CV events. Another strength of the present study was that the distribution of the different pattern of PRA levels (low-, normal-, and high-renin hypertension) in our patients with hypertension did not differ significantly compared with another extensively studied general hypertensive population (34); therefore, our PA prevalence findings were confidently representative of the general unselected hypertensive population.
Limitations of the study were that 24-h ambulatory BP monitoring levels were not available and therefore BP level subdivision could be affected by the white coat effect; only 51% of all patients with hypertension were recruited for study participation and only information from about 31% of the patients not recruited was available; and, because other secondary forms of hypertension were not excluded, patients who did not have PA possibly had, in a minority of cases, another secondary cause of hypertension.
The present study demonstrated that PA is a frequent cause of secondary hypertension even in the general hypertensive population and indicated that most hypertensive patients should be screened for PA. Early diagnosis is critical to obtain cure in patients who are adrenalectomized for unilateral PA and fundamental to prevent the development of target organ damage. Our results also supported the view that aldosterone excess determines an increase in organ damage and CV events, supporting the extensive screening of the hypertensive population to ensure PA patients receive therapy, preferably adrenalectomy for unilateral forms and pharmacological therapy with MRA for bilateral forms.
COMPETENCY IN MEDICAL KNOWLEDGE: Primary aldosteronism, including potentially curable aldosterone-producing adenomas, is a common form of secondary hypertension encountered in primary care practice and is associated with relatively high rates of target organ damage, cardiovascular complications, and metabolic syndrome.
TRANSLATIONAL OUTLOOK: Further studies are warranted to explore whether earlier detection of primary aldosteronism or detecting milder forms of the disease reduces the risk of CV events.
The authors thank all the general practitioners who participated in the PATO study, and in particular Beatrice M. Amosso, Luisa M. Gabetti, Donata Tarello, Carmen Usai, Laura Morselli, Manuela Ruffa, Denise Marri, Marco Giovannone, Maria L. Bracco, Giuseppe Tafuri, Piero Verdi, Claudio Nocentini, Laura Cecchetto, Teresita M. Bracci, Maurizio Milletarì, Maria C. Mercadante, Paola Livorno, Vincenzo Surgo, and Claudia Matrella. They also thank the society EVOLUS Srl for providing patient data using the NEXT MMG software.
For a supplemental Methods section as well as tables, please see the online version of this article.
This study was supported by grants from Regione Piemonte 2008-I and 2008-II; and the Italian Ministry of the Instruction, University and Research (ex-60% 2013 and 2014, to Dr. Mulatero; ex-60% 2014 and 2015, to Dr. Monticone). All authors have reported that they have no relationships relevant to the contents of this paper to disclose. Drs. Monticone and Burrello contributed equally to this work.
- Abbreviations and Acronyms
- aldosterone concentration
- aldosterone-producing adenoma
- aldosterone-to-renin ratio
- adrenal vein sampling
- bilateral adrenal hyperplasia
- blood pressure
- general practitioner
- low-renin essential hypertensives
- mineralocorticoid receptor antagonist
- primary aldosteronism
- plasma renin activity
- systolic blood pressure
- Received August 18, 2016.
- Revision received January 6, 2017.
- Accepted January 10, 2017.
- 2017 American College of Cardiology Foundation
- Milliez P.,
- Girerd X.,
- Plouin P.-F.,
- Blacher J.,
- Safar M.E.,
- Mourad J.-J.
- Gordon R.D.,
- Klemm S.A.,
- Stowasser M.T.,
- Terry I.,
- Storie W.I.,
- Rutherford J.C.
- Käyser S.C.,
- Dekkers T.,
- Groenewoud H.J.,
- et al.
- Mancia G.,
- Fagard R.,
- Narkiewicz K.,
- et al.
- Monticone S.,
- Viola A.,
- Rossato D.,
- et al.
- Burrello J.,
- Monticone S.,
- Buffolo F.,
- et al.
- ↵Palmieri L, Donfrancesco C, Vannucchi S, Lo Noce C, Dima F, Giampaoli S, Vanuzzo D. Epicentro, Il portale dell'epidemiologia per la sanità pubblica. Ipertensione: i dati dell’Osservatorio epidemiologico cardiovascolare/Health Examination Survey. Available at: http://www.epicentro.iss.it/approfondimenti/2013/whdDatiOEC.asp. Accessed April 4, 2013.
- Banegas J.R.,
- López-García E.,
- Dallongeville J.,
- et al.
- Mulatero P.,
- Monticone S.,
- Burrello J.,
- Williams T.A.,
- Veglio F.,
- Funder J.W.
- Ori Y.,
- Chagnac A.,
- Korzets A.,
- et al.
- Sywak M.,
- Pasieka L.
- Sato M.,
- Morimoto R.,
- Seiji Y.,
- et al.
- Williams J.S.,
- Williams G.H.,
- Raji A.,
- Jeunemaitre X.,
- Brown N.J.,
- Hopkins P.N.