Prospective evaluation of a telmisartan suppression test as a diagnostic tool for primary hyperaldosteronism in cats

Abstract Background In a previous study, telmisartan suppressed aldosterone secretion in healthy cats but not in cats with primary hyperaldosteronism (PHA). Hypotheses Telmisartan suppresses aldosterone secretion in middle‐aged healthy cat and cats with diseases that may result in secondary hyperaldosteronism, but not in those with PHA. Animals Thirty‐eight cats: 5 with PHA; 16 with chronic kidney disease (CKD), subclassified as hypertensive (CKD‐H) or non‐hypertensive (CKD‐NH); 9 with hyperthyroidism (HTH); 2 with idiopathic systemic arterial hypertension (ISH); and 6 healthy middle‐aged cats. Methods Prospective, cross‐sectional study. Serum aldosterone concentration, potassium concentration, and systolic blood pressure were measured before and 1 and 1.5 hours after PO administration of 2 mg/kg of telmisartan. The aldosterone variation rate (AVR) was calculated for each cat. Results No significant difference in the minimum AVR was observed among groups (median [quartile 1 (Q1); quartile 3 (Q3)]: 25 [0; 30]; 5 [−27; −75]; 10 [−6; −95]; 53 [19; 86]; 29 [5; 78]) for PHA, CKD, HTH, ISH, and healthy cats, respectively (P = .05). Basal serum aldosterone concentration (pmol/L) was significantly higher in PHA cats (median [Q1; Q3]: 2914 [2789; 4600]) than in CKD‐H cats (median [Q1; Q3]: 239 [189; 577], corrected P value = .003) and CKD‐NH cats (median [Q1; Q3]: 353 [136; 1371], corrected P value = .004). Conclusions and Clinical Importance The oral telmisartan suppression test using a single dose of 2 mg/kg telmisartan did not discriminate cats with PHA from healthy middle‐aged cats or cats with diseases that may result in secondary hyperaldosteronism.

Conclusions and Clinical Importance: The oral telmisartan suppression test using a single dose of 2 mg/kg telmisartan did not discriminate cats with PHA from healthy middleaged cats or cats with diseases that may result in secondary hyperaldosteronism.

| INTRODUCTION
Despite being increasingly recognized as the most common adrenocortical disorder in cats, 1 definitive diagnosis of primary hyperaldosteronism (PHA) remains a challenge. The disease is defined by autonomous, excessive production of mineralocorticoids (mainly aldosterone) either as a consequence of an adrenal tumor or, presumably less frequently, of idiopathic micronodular hyperplasia in the adrenal cortex. 2 One of the major difficulties in diagnosing this adrenal disease in cats is to differentiate PHA ("low-renin" hyperaldosteronism) from secondary hyperaldosteronism ("high-renin" hyperaldosteronism). The latter is characterized by the presence of an underlying disease (e.g., volume depletion, hyperthyroidism [HTH], or heart disease) that causes activation of the entire renin-angiotensin system, resulting in aldosterone hypersecretion. 1 Diagnostic criteria for PHA historically have relied on the detection of increased plasma or serum aldosterone concentration with consequently suppressed plasma renin activity (PRA) and an increased aldosterone-to-renin ratio (ARR). [2][3][4] However, several limitations to this approach must be considered. First, assays for PRA are not widely available and require stringent sample handling because blood must be rapidly centrifugated, frozen, and transported on dry ice. 3 Second, in human medicine, ARR is considered to be a screening tool with low specificity, and positive ARR results need further confirmation using dynamic endocrine testing. 5 Third, the ability of the ARR to differentiate PHA from secondary hyperaldosteronism has not been thoroughly studied in cats. Therefore, development of a reliable and easily accessible test is needed for the diagnosis of PHA in cats.
Several exploratory tests have been developed to identify autonomous and excessive aldosterone production in PHA. [6][7][8][9][10][11][12] For the diagnosis of PHA in humans, 4 testing procedures currently are recommended by the European Endocrine Society guidelines 13 : the fludrocortisone suppression test, the oral sodium loading test, the saline infusion test, and the captopril challenge test. To date, published evidence is insufficient to recommend 1 test over the others. In cats, oral fludrocortisone and oral sodium loading tests already have been evaluated. [8][9][10] In 19 cats with systemic hypertension caused by PHA (n = 9) or other causes (n = 10), combined evaluation of basal urinary aldosterone-to-creatinine ratio (UACR) and UACR after PO fludrocortisone administration could not substantially discriminate between cats with and without PHA. 8 Surprisingly, 4/9 cats with PHA showed UACR suppression, which may raise concerns about the interpretation of discordant results when performing such tests. Although the urinary fludrocortisone test is of undeniable interest, its disadvantages must be considered. It requires administration of fludrocortisone at home, or in-hospital, twice a day for 4 days, which requires owner and animal compliance. It also entails giving mineralocorticoids to an animal that already suffers from an overproduction of these hormones. As a result, hypokalemia was observed in 7/19 treated cats, and in 1 of these muscle weakness occurred.
A recent study investigated a suppression test using PO telmisartan. It showed that PO administration of 2 mg/kg telmisartan suppressed plasma aldosterone concentration by a minimum of 33% in healthy young cats, but did not decrease aldosterone concentration significantly in cats with PHA; minimal overlap was observed between the 2 groups. 14 Our aims were to extend the results of the previous study with further investigations of the accuracy of the telmisartan suppression test (TST) for PHA diagnosis in groups of cats with different clinical profiles. We hypothesized that telmisartan administration would suppress aldosterone secretion in healthy middle-aged cats as well as in cats with secondary hyperaldosteronism, but not in cats with PHA.
We also hypothesized that the TST would be safe and would not induce clinically relevant variations in blood pressure or serum potassium concentration. A secondary aim was to determine if hypertensive cats without adrenal masses, HTH or azotemia would have TST results consistent with occult PHA. Hypertrophic cardiomyopathy was defined as diffuse or regional increased left ventricular (LV) myocardium thickness with a nondilated LV chamber, according to ACVIM consensus guidelines. 16 Enddiastolic thicknesses of the LV free wall (LVFWd) and the interventricular septum (IVSd) as well as LV end-diastolic and end-systolic internal diameters were measured using the 2-dimensional (2D)guided M-mode as recommended by the ACVIM consensus guidelines, 16 and the LV shortening fraction (SF%) then was calculated. For each cat, all M-mode measurements were compared with the 95% prediction intervals assessed according to body weight from a large population of healthy cats. 17 Extreme LV hypertrophy was defined as LVFWd, IVSd, and subaortic IVSd ≥9 mm. 18 The subaortic interventricular septal thickness also was measured at end-diastole by 2D mode from the right parasternal 5-chamber view at the mitral valve-chordae tendineae interface as previously described and compared with reference ranges. 19 The left atrial (LA) and aortic (Ao) diameters were measured at end-diastole using a 2D method from the right parasternal short axis view, as previously described and the LA:Ao ratio then was calculated, with LA enlargement defined as an LA:Ao >1.2 (upper cutoff obtained from a population of 100 prospectively recruited healthy cats). 19 Continuous-wave Doppler recorded using the left apical 5-chamber view was used to diagnose LV outflow tract obstruction defined as diffuse LV outflow tract turbulence and peak systolic outflow velocity ≥2.5 m/s. 20 Additionally, each HCM phenotype was classified according to the staging system proposed by the ACVIM consensus statement on cardiomyopathies in cats. 15  tion, no findings of concern in blood biochemistry and electrolyte analyses and with serum T4 concentration within the reference internal.

| MATERIALS AND METHODS
Non-inclusion criteria included anemia, receipt of any treatment that could interfere with the renin-angiotensin-aldosterone system (RAAS; e.g., amlodipine, spironolactone, angiotensin II receptor blockers, angiotensin converting enzyme inhibitors, sildenafil, pimobendan, methimazole or diuretics) within the previous 2 weeks, and heart failure.
The TST was performed as previously described, by the same investigator (MK). 14 In brief, serum aldosterone concentration was measured before (T0), 1 hour after (T1), and 1.5 hours after (T1.5) PO administration of 2 mg/kg telmisartan (Semintra, Boehringer Ingelheim). Difficulty in administering the medication to each cat was evaluated subjectively on a scale from 0 (very easy) to 4 (impossible).
Systolic blood pressure (SBP) and serum potassium concentration were measured at T0, T1, and T1.5 to evaluate safety. During the test, cats were kept in a quiet environment. Blood was collected from the jugular vein with cats positioned in sternal recumbency. After centrifugation, serum was stored at À80 C until batch analysis to measure aldosterone concentration by a radio-immunoassay (RIAZENco Aldosterone, ZenTech). The assay was internally validated (intra-and inter-assay coefficients of variation at 3 ranges of concentrations (90, 291, and 585 pmol/L) were between 1.9% to 4.5% and between 5.0%  with CKD-NH and 1 with HTH) that were classified as stage B2 because of gallop sound and severe LV hypertrophy, respectively. Moderate correlation was observed between AVR T1.5 and baseline PRA (r = À.28; P = .09; Figure 5).

| DISCUSSION
We report investigation of an outpatient, dynamic, and short duration endocrine test for the diagnosis of PHA in cats. Unfortunately, the test investigated did not discriminate cats with PHA from other cats. F I G U R E 2 Serum aldosterone concentration at T1.5 in each group. See Figure 1 for legends.

F I G U R E 3
Aldosterone variation rate at T1.5 and minimal aldosterone variation rate between T1 and T1.5 in each group. See Figure 1 for legends.
When trying to develop a confirmatory test for an endocrine disease characterized by hypersecretion of a hormone, it is natural to consider a suppression test. These tests aim to confirm that production of the hormone is not downregulated by an exogenous inhibitor, because it is being secreted autonomously. 8-10 A suppression test using losartan, an angiotensin II receptor antagonist, has been evaluated in humans. 12 The test is rapid and practical, requiring a single blood sample 2 hours after PO administration of losartan. Moreover, a study that included patients > 50 years of age suggested that the losartan suppression test was conservative than traditional exogenous mineralocorticoid-based tests in patients with cardiac or renal impairment. 12 Such comorbidities are thought to be frequent in cats with PHA. 1,2,4 In veterinary medicine, telmisartan, another angiotensin receptor antagonist, is already available commercially (Semintra, Boehringer Ingelheim) and approved in cats for the management of proteinuria in CKD and systemic hypertension. In a pilot study, the TST accurately discriminated between 10 healthy cats and 6 cats with F I G U R E 4 Absolute variation in aldosterone concentration between T0 and T1.5, and minimal absolute variation in aldosterone concentration between T1 and T1.5 in each group. See Figure 1 for legends.
F I G U R E 5 Aldosterone variation rate at T1.5 plotted against baseline plasma renin activity for the whole studied population.
PHA. 14 These preliminary data indicated that a dose of 2 mg/kg was more effective at differentiating healthy cats from cats with PHA than a dose of 1 mg/kg, and that both T1 and T1.5 sampling times were necessary.
In our study, although telmisartan did not significantly suppress serum aldosterone concentration in cats with PHA, the TST was unable to differentiate between these cats and cats with diseases that potentially result in secondary hyperaldosteronism (e.g., CKD, HTH).
Furthermore, telmisartan did not suppress serum aldosterone concentration in any of the healthy cats. From our results, we could speculate that suppression of serum aldosterone concentration of at least 20% excludes PHA. However, our results contradict the results of the previous study, in which all healthy cats showed at least 33% aldosterone suppression. 14 Because the main difference in the healthy cat population between the 2 studies was age (median 3 years in the pilot study vs 8 years in the our present study), we speculate that telmisartan has a different pharmacokinetic or pharmacodynamic profile in older cats, or that older cats are better able to resist its suppressive effects on aldosterone secretion. An "aldosterone breakthrough" effect is described in the human and veterinary medical literature, 26,27 when serum aldosterone concentrations do not remain suppressed despite ongoing treatment with an angiotensin converting enzyme inhibitor or angiotensin receptor antagonist; the serum concentration of circulating aldosterone may even increase above pretreatment concentrations. However, this escape phenomenon usually is reported after months of treatment in humans, and was detected after at least 7 days of telmisartan administration in dogs. 28,29 Because the cats in our study received a single dose of telmisartan, it is unlikely that they exhibited aldosterone breakthrough in the strictest sense. From a pharmacodynamic perspective, it seems reasonable to assume that homeostatic mechanisms in a healthy animal initially would compensate for a pharmacological perturbation, and that accordingly our cats may have failed to suppress aldosterone. It is possible that measuring serum aldosterone concentration after several days of telmisartan administration would lead to more marked aldosterone suppression, yielding more accurate results than the protocol we have used here.
This approach was not taken because our aim was to develop a confirmatory test that is easier to perform than the fludrocortisone suppression test.
Another explanation for our unexpected results may be related to variability in the pharmacokinetics of telmisartan. We used sampling times for aldosterone (T1 and T1.5) on the basis of the pilot study, 14 which was itself informed by pharmacokinetic data for telmisartan in 12 young laboratory cats. Our current results, however, have identified unexpected high variability in serum telmisartan concentration at T1.5. Moderate variability in maximal serum telmisartan concentration (C max ) also was observed in the Semintra licensing study (coefficient of variation between 33% and 45%) but was lower than in our population (118%). 30

| CONCLUSIONS
The TST based on measurement of serum aldosterone concentration before, and 1 and 1.

OFF-LABEL ANTIMICROBIAL DECLARATION
Authors declare no off-label use of antimicrobials.

INSTITUTIONAL ANIMAL CARE AND USE COMMITTEE (IACUC) OR OTHER APPROVAL DECLARATION
The work described in this manuscript involved the use of nonexperimental (owned or unowned) animals and procedures that differed from established internationally recognized high standards ("best practice") of veterinary clinical care for the individual patient.
The study had prior ethical approval from the Comité d'éthique en Recherche Clinique (ComERC) from the Ecole nationale vétérinaire d'Alfort, Number 2020-05-24-1. Written informed consent was obtained from the owner or legal custodian of all animals described in this work for all procedures undertaken.

HUMAN ETHICS APPROVAL DECLARATION
Authors declare human ethics approval was not needed for this study.