**2. Primary Aldosteronism**

The most prevalent cause of LRH is primary aldosteronism (PA) [10,16]. PA is characterized by hyperaldosteronism that is independent of renin and angiotensin II (thus renin-independent aldosteronism) that results in excessive MR activation, increases intra-vascular volume and blood pressure, and results in renal, vascular, and cardiac disease, and higher mortality [10,11,16–19].

PA is considered the most common form of endocrine hypertension, with an estimated prevalence of 5–10% in the general hypertensive population, at least 6% in the primary care population, and up to 20% in the setting of resistant hypertension [11,17,20]. Since Conn's initial description of the classical PA disease phenotype over 50 years ago, the understanding of the severity spectrum of PA and underlying genetics has greatly expanded [5,21]. First, human studies have shown that there is a broad spectrum of autonomous and renin-independent aldosteronism and MR activation; PA is not only a disease reserved for those with severe and resistant hypertension, rather can be detected in mild to moderate hypertension and also in normotension [20,22–24]. Normotensive individuals with higher aldosterone levels have a higher risk for developing hypertension, an association that is driven by normotensives exhibiting a PA phenotype: renin suppression with increasingly inappropriate aldosterone secretion [22–25]. Thus, it is becoming clearer that dysregulated autonomous aldosterone secretion that is independent of renin, even when it does not meet the classical definitions of overt PA, exists across a large continuum, and therefore, our strict categorization of PA may handicap clinical care by placing focus on only the most severe cases at the expense of ignoring milder disease [23,24,26,27]. Although clinical practice recommendations focus on defining PA using categorical thresholds [16,18], expert opinion is increasingly warning that "the strict definition of primary aldosteronism is no longer tenable," and calling to "recognize the true prevalence of primary aldosteronism to include dysregulated aldosterone secretion and inappropriate aldosterone production" [28]. Second, excessive MR activation in PA contributes to significant cardiovascular and metabolic diseases, independent of blood pressure, such as diabetes and metabolic syndrome, stroke, myocardial infarction, left ventricular hypertrophy, atrial fibrillation, heart failure, and death [11,16,19,29–31]. Collectively, these two important observations have made it clear that recognizing and treating PA as early as possible is critical to prevent long-term adverse outcomes. Finally, our understanding of the pathogenesis of PA has dramatically improved with new genetic and histopathologic discoveries that have shed light on the mechanisms that might underlie PA. These advances will be discussed in more detail below.

The Endocrine Society clinical practice guidelines recommend identifying overt cases of PA by measuring the aldosterone-to-renin ratio (ARR) based on the clinical detection of severe or resistant hypertension, hypokalemia, an adrenal nodule, sleep apnea, or a family history of PA or early cardiovascular disease [16]. The most widely used cut-off for an aldosterone concentration is at least 15 ng/dL (and less frequently 10 ng/dL) with an ARR of at least 30 ng/dL per ng/mL/h [16]. This clinical approach lowers the risk for false positive screening results and, in general, is designed to detect overt and severe cases of PA. Alternatively, to recognize milder forms of PA, and to maximize early case detection in order to mitigate future cardiometabolic disease, more permissive screening criteria have also been proposed: a suppressed renin activity in the context of non-suppressed aldosterone (>6–9 ng/dL) consistent with an ARR >20 [16]. This latter approach may detect milder cases of PA, but will increase the risk of false-positive screening results, and consequently, potentially more costly and/or invasive medical testing. The absence of a single diagnostic criterion is largely propagated by the lack of a histopathologic gold standard for PA diagnosis.

Given the high prevalence of PA, and particularly, the more recent recognition that milder forms of PA may be common even when there are is no radiographic evidence of adrenal neoplasia, a key issue is to understand what may underlie the pathogenesis of PA. The use of specific CYP11B2 antibodies has revealed the presence of aldosterone producing cell clusters (APCCs) in a remarkable proportion of morphologically normal adrenal glands [32,33]. APCCs have been described as non-neoplastic foci of CYP11B2 staining that can co-exist adjacent to aldosterone-producing adenomas, extend into zona fascisculata, and often harbor known somatic mutations in aldosterone driver genes [32,34,35]. Collectively, this evidence strongly suggests that APCCs may represent a common aldosterone-secretory abnormality, which can be detected in normotensives with normal adrenal glands [36], may increase in number with older age, and may thus be a pre-neoplastic and age-dependent precursor to more overt PA [37].

Along with the finding of APCCs, our understanding of the genetics of PA has undergone substantial revision in the past decade. These advances have been described in detail in recent reviews, and therefore, we will summarize them briefly herein. The first described inheritable form of PA is glucocorticoid remediable aldosteronism (GRA), also known as familial hyperaldosteronism type I (FH-I). The estimated prevalence of GRA is only 1% of PA subjects and is secondary to a chimeric gene with recombination of 11β-hydroxylase (*CYP11B1*) and aldosterone synthase (*CYP11B2*) genes [17,38]. The chimeric fusion results in regulation of aldosterone synthesis by adrenocorticotropic hormone (ACTH) and can be mitigated by suppressing ACTH with glucocorticoids. GRA should be considered in patients with early-onset hypertension and family history of PA or early-onset cerebrovascular accidents [17]. GRA has considerable variation in its clinical presentation, some families do not manifest with the classic hypertension and hypokalemia, and therefore, the gold standard diagnosis is sequencing to confirm the chimeric *CYP11B2*/*CYP11B1* gene [39].

Familial Hyperaldosteronism type II (FH-II) has been described in 6% of PA cases but is by definition a familial form of PA with yet unknown genetic loci [17]. Whether the discovery of newer inheritable forms of PA (below) results in a reclassification of FH-II remains to be seen.

Most recently, a new Familial Hyperaldosteronism named type III (FHIII), was described secondary to a gain-of-function germline mutation in the *KCNJ5* gene [40,41]. The *KCNJ5* gene mutation results in loss of potassium selectivity in a zona glomerulosa potassium channel, consequent increased influx of sodium resulting in a higher cell membrane potential and lower depolarization threshold, and therefore increased aldosterone synthesis and secretion [17,38,41]. Interestingly, although germline mutations in *KCNJ5* are rare, somatic mutations of *KCNJ5* have been described in nearly half of aldosterone-producing adenomas [38]. In addition, somatic (and rarely germline) mutations in other zona glomerulosa channels that result in increased cell membrane potential and decreased depolarization thresholds have been described in *CACNA1D* gene (codes for calcium channel, voltage-dependent, L-type, α-1d subunit), *CACNA1H* gene (codes for T-type voltage dependent calcium channel Cav3.2) [42], while somatic mutations have been found in the *ATP1A1* gene (Na+, K+-ATPase), and *ATP2B3* gene (calcium transporting ATPase 3) [38,43].

Treatment of PA should be tailored according to the severity of disease, age of the patient, anatomic type of disease (unilateral adenoma versus bilateral hyperplasia) and desire for surgery. Laparoscopic surgery is the recommended and ideal therapeutic intervention if PA is unilateral since it can cure aldosterone excess and improve long-term cardiometabolic outcomes and blood pressure control [16,19,44]. In most cases of PA, the source of autonomous aldosterone is bilateral or surgery is not pursued due to other complicating factors, and therefore, medical therapy with MR antagonists (such as spironolactone and eplerenone) is recommended [16]. Although medical therapy is often assumed to be equally efficacious to surgical therapy if blood pressure is normalized, a recent study suggested that this assumption may not be correct. In this large cohort study, patients with PA treated with MR antagonists had 2–3 times higher risk for incident cardiovascular events and death, when compared to age-matched essential hypertensives, even though they had similarly controlled (and relatively normalized) blood pressure while on MR antagonist therapy [19]. In contrast, PA patients who developed a longitudinal increase in renin activity while being treated with MR antagonists had the same risk for incident cardiovascular events and death as patients with essential hypertension, suggesting that the excess risk in medically treated PA may be due to inadequate MR blockade as manifested by a persistently suppressed renin. Future prospective studies will be needed to determine the optimal approach for medical therapy in PA, and how it compares with surgical adrenalectomy [19].

#### **3. Hypercortisolism**

Endogenous hypercortisolism, with or without the overt manifestations of Cushing syndrome, can result in chronic stimulation of the glucocorticoid receptor and also potentially the MR, with consequent development of hypertension, insulin resistance, diabetes, and cardiovascular disease and mortality [45–49].

There are several potential mechanisms for developing the LRH phenotype with hypercortisolism. Activation of the glucocorticoid receptor by excess cortisol can induce a direct vasopressor effect and elevations in blood pressure [50]. However, cortisol-mediated activation of the renal MR can also play a role in developing LRH. Cortisol and aldosterone are similarly potent MR agonists, but cortisol is inactivated to cortisone by 11β-hydroxysteroid dehydrogenase Type 2 (11βHSD2), thereby "protecting" the renal MR from abundant cortisol stimulation and permitting a high-affinity aldosterone-MR interaction. However, in states of severe hypercortisolism, excess cortisol can overwhelm 11βHSD2 activity and result in direct cortisol-mediated MR activation and subsequent intravascular volume expansion with suppression of renin and aldosterone [50]. Thus, the LRH phenotype with hypercortisolism is unique from PA in that it is a hyporeninemic hypoaldosteronism manifesting with Cushing syndrome and apparent MR overactivation (i.e., hypertension, hypokalemia and increased kaliuresis) [47,51,52]. Studies have also shown that chronic hypercortisolism can also result in activation of ENaC and increases in angiotensinogen [50,51,53].

Endogenous hypercortisolism is most commonly due to a benign pituitary ACTH secreting tumor (Cushing disease), and less frequently due to benign or malignant adrenal tumors (ACTH-independent hypercortisolism), and non-pituitary ACTH-secreting tumor (ectopic ACTH secretion). Current guidelines recommend that the diagnosis of overt hypercortisolism be confirmed using two distinct tests: elevated late night salivary cortisols, elevated 24 h urinary free cortisols, and/or incomplete suppression of cortisol following overnight dexamethasone [47]. Beyond overt hypercortisolism, the concept of "subclinical hypercortisolism" or "autonomous cortisol secretion" in association with adrenocortical adenomas, whereby there is excess cortisol without the hallmark signs of Cushing syndrome, is being recognized as a prevalent phenotype that is associated with higher risk for cardiometabolic disease [54]. For example, studies using mass spectrometry to analyze steroid metabolites have shown that even apparently nonfunctional adrenal adenomas secrete higher concentrations of glucocorticoids when compared to patients with no adenomas [55]; therefore, it is possible that many or most adrenocortical neoplasms secrete at least miniscule amounts of glucocorticoid, and that the categorization of "nonfunctional adenoma" may be misleading or a misnomer. In parallel, cohort studies in patients with incidentally discovered benign adrenocortical tumors have observed that patients with subclinical hypercortisolism have a higher risk for incident cardiovascular disease and death when compared to those with "nonfunctional" tumors [49]. Further, patients with apparently "nonfunctional" adrenocortical tumors have a higher risk for incident diabetes, a risk that was related to the degree of autonomous cortisol secretion within the normal range (normal defined as cortisol < 1.8 mcg/dL following overnight 1 mg dexamethasone suppression), compared with patients with no adrenal tumors [48]. Thus, as with PA, there is increased recognition that adrenal hypercortisolism is not a categorical phenotype, rather exists over an expanded continuum ranging from mild and subclinical to overt Cushing syndrome, with a parallel risk profile of increasing cardiometabolic risk. Future and ongoing studies will be necessary to determine when intervention to treat the autonomous cortisol is indicated with respect to abrogating incident adverse outcomes.

The genetics of hypercortisolism, both related to pituitary and adrenals tumors, has undergone dramatic change in the recent decade. Concerning pituitary tumors, germline mutations that cause Cushing disease have been previously ascribed to *MEN1* gene mutations and *AIP* mutations in Familial isolated pituitary adenomas (FIPA). More recently, next generation sequencing of pituitary adenomas identified somatic driver mutations in ubiquitin-specific protease 8 (*USP8*), leading to ACTH excess and hypercortisolism [56]. In adrenal hypercortisolism, novel genes have also been described beyond the classic germline mutations of the Carney complex (*PRKAR1A* gene, regulatory subunit of protein kinase) or McCune-Albright syndrome (activating somatic mutations in *GNAS1* oncogene) [56]. Germline mutations in *ARMC5* (Armadillo repeat-containing protein 5) have been identified in familial cases and are present in approximately 50% of sporadic cases of macronodular adrenal hyperplasia [57]. In addition, somatic events in *ARMC5* and *PRKACA* (encodes for catalytic subunit α of protein kinase A) are frequently observed in cortisol-producing adrenal adenomas [38,58,59].

## **4. Apparent Mineralocorticoid Excess Syndrome**

The syndrome of apparent mineralocorticoid excess (AME) is a rare disease, first described in the late 1970s, as a syndrome of severe pediatric LRH [60]. The AME syndrome is an autosomal recessive condition due to loss of function mutations in 11βHSD2. The insufficient activity of 11βHSD2 permits normal cortisol concentrations to activate the renal MR, resulting in a syndrome of MR-mediated LRH, low aldosterone, hypokalemia, alkalosis, and usually failure to thrive and poor weight gain.

The diagnosis of AME is usually suspected in the setting on non-aldosterone dependent LRH with classic features of MR activation and confirmed by a high cortisol/cortisone (F/E) ratio in the serum or urine, and/or genetic sequencing of 11βHSD2 [61,62]. These infrequent cases of classic AME are treated with low dose dexamethasone to suppress endogenous ACTH and cortisol (since dexamethasone is not metabolized by 11βHSD2) in combination with an MR antagonist, and in extreme cases, renal transplantation [63].

Although the classic AME syndrome is rare, recent research suggests that the spectrum of cortisol-mediated MR activation may be more expansive than currently recognized in that milder forms of AME may be common. For instance, Ulick et al. described a milder version of AME and named it the type 2 variant, caused by a decrease in the cortisol clearance rate but not related to cortisone conversion [64]. More recently, cross-sectional human studies have shown that both cortisone levels and renin activity decline with older age, suggesting a potential age-dependent decline in the activity of 11βHSD2 [65]. Since higher F/E ratio and the low-renin phenotype have been correlated with higher blood pressure in both adults and children [66,67], new evidence suggests that a proportion of LRH may be explained by a less severe, or "non-classical", phenotype of AME that may respond to MR antagonists [68]. Interestingly, milder phenotypes of AME may be explained by less severe inactivating mutations or heterozygosity with partial activity of 11βHSD2, but also by consumption of exogenous inhibitors of 11βHSD2such as licorice or grapefruit [69]. In a recent study, we observed that lower cortisone levels (in combination with higher F/E ratio) were strongly associated with higher MR activity (lower renin activity and higher urinary potassium excretion) in patients suspected to have mild or non-classical AME (Tapia-Castillo, Baudrand, Vaidya, et al. personal communication). The summary of available data suggest that beyond the rare classical phenotype of AME, milder forms of non-classical AME may contribute to LRH and may be best detected by recognition of high F/E ratio in combination with low cortisone levels.
