1. Introduction

Essential hypertension is still the most prevalent noncommunicable disease all over the world; back in the year 2010 nothing <31.1% of adult population globally had hypertension.[1] Although true treatment resistant hypertension (TRH) is relatively rare (about 10.3% of all patients with hypertension), almost 35.6% of patients receive suboptimal treatment, and further 50.0% is not adherent, so optimal control of blood pressure is not achieved in every other patient.[2] Starting optimal drug treatment of treatment resistant hypertension from the moment when hypertension was diagnosed or as early as possible during its course is associated with improved cardiovascular outcomes.[3]

Treatment resistant hypertension is defined as casual blood pressure during clinical examination of more than 140/90 mm Hg despite treatment with optimal or best-tolerated doses of 3 or more drugs, which should include a thiazide diuretic, typically an angiotensin-converting enzyme inhibitor (ACE) inhibitor or an angiotensin receptor blocker (ARB), and a calcium channel blocker (CCB).[4,5] Resistant hypertension is treated with add-on therapy with antagonists of mineralocorticoid receptors (spironolactone[SPL]/eplerenone, esaxerenone) or with different types and dosages of diuretics (DIU), alpha 1 adrenergic blockers, alpha-2 adrenergic receptor agonists (A2A), that is, clonidine and methyldopa, or beta 1 blockers (BB).[6,7] Central antihypertensives, which stimulate alpha-2 and/or imidazoline receptors and therefore decrease sympathetic outflow from medulla oblongata, are very effective as add-on therapy in resistant hypertension; urapidil, an antihypertensive that have both alpha 1 adrenergic blocking action and central sympatholytic effect due to activation of serotonin 5HT1A receptors was also successfully combined with first-line antihypertensives.[6,7]

There are several clinical trials comparing effectiveness and safety of add-on therapy of resistant hypertension. Spironolactone was tested in 5 randomized clinical trials.[8] It reduced office systolic blood pressure (SBP) by 15.73 mm Hg and office diastolic blood pressure (DBP) by 6.21 mm Hg as compared to placebo group. The pooled changes of 24 hours ambulatory or home SBP and DBP were −8.7 mm Hg and −4.12 mm Hg, in favor of the spironolactone group. In comparison with alternative drugs for treating resistant hypertension, including beta blocker, candesartan, or alpha methyldopa, spironolactone reduced home SBP by 4.5 mm Hg. Addition of spironolactone provides beneficial effect on blood pressure in patients with resistant hypertension, as confirmed by several meta-analyses and systematic reviews of clinical trials.[9] Although there are also invasive methods of treating resistant hypertension, that is, electrical stimulation of the carotid baroreceptors and baroreflex activation therapy, their efficacy in clinical trials needs to be confirmed with phase-3 studies.[10–12] The catheter-based radiofrequency ablation of renal sympathetic nerves, that is, renal denervation, has emerged as an effective therapeutic option in addition to the nonpharmacological treatment and antihypertensive drugs.[13] When various add-on drugs for resistant hypertension were compared among themselves in clinical trials and meta-analyses, spironolactone showed the best efficacy, but yet the published data are limited,[14] and additional studies are needed to confirm already published studies. It is especially necessary to further explore adverse effects of drugs used in add-on therapy in clinical practice, on patients which are not selected like in clinical trials, but carry the burden of comorbidities and co-medication.[15]

The aim of this study was to compare effectiveness, safety and tolerability of various add-on treatment options in adult patients with treatment resistant hypertension.

2. Patients and methods 2.1. The study design

The study was multicentric, involving 10 specialist centers for treatment of hypertension affiliated with tertiary or secondary care hospitals, and 24 primary care practices; all study sites belong to public healthcare system in Serbia. The resistant hypertension was defined as casual blood pressure during clinical examination of more than 140/90 mm Hg despite treatment with optimal or best-tolerated doses of 3 or more first-line antihypertensive drugs, including a DIU, an ACE inhibitor or an ARB, a CCB and a beta blocker. The add-on therapy was prescribed to the patients within the scope of their routine medical care, independently from the study investigators. Both office and home blood pressure measures were recorded at baseline and then every month for 6 visits. Before, during and after the study the patients were cared for by their chosen general practitioners independently from the study investigators and referred to internal medicine specialists as necessary, according to diagnostic and treatment protocols currently in power at the study sites. Role of the study investigators was limited to observation and collection of data from the patients themselves and from their medical files. The study participants were recruited from June 30th, 2021 to June 30th, 2022, and followed for 6 months up to December the 31st, 2022. The study was approved by the Ethical Committee of Serbia (decision number 515-20-03493-2020-4, date 28.01.2021.) and by Ethical committees of healthcare facilities at the study sites. Protocol of the study was preregistered at ClinicalTrials.gov, with ID number: NCT05087940. The study was sponsored by Galenika a.d. Pharmaceutical Company, Belgrade, Serbia.

2.2. The patients

The study population consisted of subjects with therapy-resistant essential hypertension to whom secondary causes of hypertension were previously excluded. The inclusion criteria were: age 18 years or older, having resistant hypertension defined as casual blood pressure during clinical examination of more than 140/90 mm Hg despite treatment with optimal or best-tolerated doses of 3 or more first-line antihypertensive drugs, and inadequate control of hypertension confirmed with both office and home blood pressure measurement (morning and evening measurements after 5 minutes of relaxation in sitting position, for 6 days). The exclusion criteria were: pregnancy, anticipated pregnancy, breastfeeding, incarceration or institutionalized living which may prohibit measurement of home blood pressure, participation in another intervention study that may affect blood pressure, life expectancy < 4 months, anticipated living donor kidney transplant within 4 months, systolic blood pressure value over 220 mm Hg requiring immediate adjustments of therapy, moderate to severe renal insufficiency (acute or chronic) with glomerular filtration rate below 30 mL/minutes, active bronchospastic disorders, heart failure classes III and IV, severe bradycardia (heart rate below 50/minutes), 2nd and 3rd degree AV block, and history of hypersensitivity to any of the drugs under study. The study sample was of convenient type, and consecutive, that is, the investigators were enrolling all patients referred by general practitioners to the study sites who satisfied inclusion and did not have exclusion criteria, and who signed the informed consent for participation in the study.

2.3. The study treatments and outcomes

The study was prospective, observational, with cohort design, therefore treatment groups were formed independently from the study investigators, according to the add-on therapy prescribed by the internal medicine specialists to whom the patients were referred by general practitioners. Primary outcomes of the study were decrease of arterial blood pressure at both office blood pressure measurement (OBP) and home blood pressure self-measurement (HBPM) to levels below 140/90 mm Hg after 1, 2, 3, 4, 5, and 6 months, reduction of systolic arterial blood pressure at both OBP and HBPM for 10 mm Hg or more after 1, 2, 3, 4, 5, and 6 months, and reduction of diastolic arterial blood pressure at both OBP and HBPM for 5 mm Hg or more after 1, 2, 3, 4, 5, and 6 months. The secondary study outcomes were absolute values of systolic and diastolic blood pressure after 1, 2, 3, 4, 5, and 6 months, obtained by both office and home measurements, serum level of creatinine and presence of proteinuria, type and frequency of adverse events, and health-related quality of life.

2.4. The study procedures

The study was conducted through screening, baseline and 6 consequent monthly visits. All centers performed OBP and HBPM following the same 2017 guidelines issued by American College of Cardiology/American Heart Association Task Force.[16] The OBP was measured by mercury sphygmomanometer, and 3 readings with 3 minutes apart were taken from each patient: a mean of the 2nd and 3rd measurements was taken as actual office blood pressure. The HBPM was measured by patients themselves, by means of automated digital oscillometric devices validated after production and supplied exclusively through state-owned community pharmacies. The patients recorded all home measurements in their diaries, which were then handed to the study investigators at the study visits.

The patients adherence to prescribed treatment was checked at each study visit by the study investigators who interviewed the patients. Demographic information, relevant medical history, including history of hypertension, and concomitant medication use (dose, route and frequency of administration) were recorded at baseline visit and updated at the 6 follow-up study visits. Comorbidities were summed-up by means of Charlson Comorbidity Index which is calculated by addition of points for each disease a patient is suffering from; the value of Charlson Comorbidity Index correlates with probability of 10-year survival.[17] Information regarding occurrence of adverse events was captured throughout the study, and standard clinical laboratory measurements were documented at the study visits. Health-related quality of life was measured at each visit by previously validated[18,19] visual analogue scale with a span from 0 to 100.

2.5. Statistics

The sample size was calculated on the basis of target statistical power of 80% and type I error probability of 0.05. Before processing the collected data statistically, screen/enrollment ratio and dropout ratio were reported. The unit of analysis were individual patients. The data then were first described by measures of central tendency (mean and median) and variability (standard deviation and interquartile range), depending on the normality of the data distribution. Normality of the data distribution was checked by the Kolmogorov-Smirnov test. The significance of differences in values of continuous variables among the repeated measurements was tested by the Student t test for dependent samples and 1-way analysis of variance (when the data distributions were normal) or by Wilcoxon Signed Rank test and Friedman test (when the data distributions were not normal). The significance of differences in values of continuous variables among the treatment groups was tested by the Student t test for independent samples and 1-way analysis of variance (when data distributions were normal) or by Mann–Whitney U test and Kruskall Wallis analysis of variance (when data distributions were not normal). The differences in values of categorical variables (e.g., frequencies) were tested by Chi-square or Fisher test. The results were considered significant if probability of null hypothesis was 0.05 or less. Sensitivity analysis was not included. All calculations were performed by Statistical Package for Social Sciences (SPSS, IBM) software, version 18.0 for Windows.

3. Results

In total 731 patients were enrolled, but after excluding screen failures, 522 patients actually entered the study, and 515 completed the study protocol; 7 patients were excluded at baseline visit due to protocol violation. The study flowchart is shown at Figure 1. There were 268 female and 247 male patients, with average age of 64.7 ± 10.8 years; detailes of the study cohort are given in the Table 1.

Table 1 - Characteristics of the study sample (n = 515). Parameter Value (median and interquartile range, or frequency and percentage) Age (yr) 66.0 [17.0] Sex Female 268 (52.0%) Body weight (kg) 85.0 [17.0] Charlson comorbidity index 2.0 [2.0] Office systolic pressure at screening (mm Hg) 160.0 [16.7] Office diastolic pressure at screening (mm Hg) 95.0 [8.3] Office heart rate (beats/min) 78.3 [13.7] Antihypertensive regimen at baseline BB + ACE/ARB + CCB + DIU 218 (42.3%) BB + ACE/ARB + CCB 52 (10.1%) BB + ACE/ARB + DIU 68 (13.2%) ACE/ARB + CCB + DIU 124 (24.1%) BB + CCB + DIU 1 (0.2%) BB + ACE/ARB + CCB + DIU + A2A 50 (9.7%) Smoking Yes 170 (33.0%) Drinking alcohol Yes 96 (18.6%) Drinking coffee Yes 454 (88.2%) Heart failure Yes 26 (5.0%) Coronary disease Yes 84 (16.3%) COPD Yes 33 (6.4%) Asthma Yes 10 (1.9%) Stroke Yes 12 (2.3%)

A2A = alpha-2 adrenergic receptor agonists, ACE = angiotensin-converting enzyme inhibitor, ARB = angiotensin receptor blocker, BB = beta blocker, CCB = calcium channel blocker, DIU = diuretics.

Figure 1.:

The study flowchart.

Before starting with add-on antihypertensive therapy for resistant hypertension, the patients were taking one of the following combinations of antihypertensives: BB + ACE/ARB + CCB + DIU (218 patients, or 42.3%), BB + ACE/ARB + CCB (52 patients, or 10.1%), BB + ACE/ARB + DIU (68 patients, or 13.2%), ACE/ARB + CCB + DIU (124 patients, or 24.1%), BB + CCB + DIU (1 patient) and BB + ACE/ARB + CCB + DIU + A2A(50 patients, or 9.7%).

At the study visits (baseline visit and visits 1–6), the patients were given add-on antihypertensive drugs according to the preferences of the researchers. At the baseline visit, add-on antihypertensives were not prescribed to 387 patients, 61 of them received an additional drug from the group of first-line antihypertensives, 60 patients received spironolactone, and 7 patients one of the central antihypertensives. Depending on the patient’s response, at subsequent visits the add-on therapy was corrected as necessary. The number of patients by add-on therapy groups, systolic and diastolic pressures and heart rate measured in the physician’s office and at home after all visits are shown in the Table 2. Systolic and diastolic pressures measured at home after visits, depending on the add-on therapy are shown in Figure S1 and S2, Supplemental Digital Content, https://links.lww.com/MD/J84 of the Supplementary file. While at the baseline and visit 1 patients with added spironolactone had the highest SBP (P = .000 and 0.003, respectively), and patients with 1st line antihypertensive added had the highest DBP (P = .018 and 0.017, respectively), these differences were lost at later visits, because the patients were switching add-on therapy according to a patient’s response. Number of patients with specific type of add-on therapy at visit 6 reflects its relative efficacy, that is, for 89.3% of patients spironolactone was optimal add-on therapy, followed by SPL + A2A (3.1%), a 1st line antihypertensive (3.1%), and A2A (1.9%).

Table 2 - Blood pressure by visit according to the groups formed by antihypertensive added at certain visit (the data are presented as median and interquartile range in parenthesis. Visit Blood pressure No antihypertensive added 1st line antihypertensive added Spironolactone added Central antihypertensive added Spironolactone + central antihypertensive added Probability of null hypothesis (p) Baseline No of patients prescribed with add-on drug at baseline n = 387 n = 61 n = 60 n = 7 n = 0 Systolic–office 156.6 [14.0] 158.3 [11.7] 161.0 [19.0] 153.0 [18.8] - 0.310 Systolic–home 156.8 [17.0]* 162.2 [12.8] 163.3 [16.2]* 155.6 [9.0] - 0.000* Diastolic–office 94.7 [7.7] 95.0 [8.1] 93.3 [10.0] 90.0 [10.7] - 0.647 Diastolic–home 95.0 [6.8] 97.3 [8.3]* 95.7 [10.1] 93.0 [4.7]* - 0.018* Heart rate at office 76.7 [12.7] 76.2 [10.1] 75.7 [13.1] 73.3 [16.7] - 0.737 Heart rate at home 78.0 [12.9] 78.3 [9.6] 78.8 [11.7] 80.4 [18.1] - 0.729 Visit 1 No of patients prescribed with add-on drug at visit 1 n = 49 n = 29 n = 417 n = 15 n = 5 Systolic–office 145.0 [18.0] 143.0 [12.8] 150.1 [22.3] 137.2 [24.2] - 0.166 Systolic–home 146.3 [18.2]* 152.2 [16.5] 151.7 [21.3]* 141.2 [20.9] - 0.003* Diastolic–office 90.0 [9.2] 90.0 [7.3] 87.7 [10.3] 86.7 [20.0] - 0.789 Diastolic–home 90.0 [9.9]* 92.8 [8.2]* 87.7 [14.4] 82.8 [17.4] - 0.017* Heart rate at office 74.7 [10.6] 73.8 [7.7] 74.3 [8.7] 69.7 [15.0] - 0.665 Heart rate at home 75.2 [10.8] 76.2 [8.2] 75.2 [9.7] 74.5 [16.2] - 0.869 Visit 2 No of patients prescribed with add-on drug at visit 2 n = 26 n = 21 n = 447 n = 14 n = 7 Systolic–office 143.3 [22.7] 142.5 [19.6] 137.0 [15.7] 134.2 [28.9] 131.2 [20.0] 0.083 Systolic–home 144.3 [18.3] 144.7 [18.6] 138.0 [15.1] 136.6 [20.8] 135.1 [17.3] 0.061 Diastolic–office 88.3 [10.0] 86.7 [9.5] 85.0 [10.0] 82.8 [10.3] 83.3 [12.5] 0.051 Diastolic–home 87.2 [10.5] 87.1 [9.5] 85.4 [10.1] 83.8 [15.4] 78.1 [17.7] 0.135 Heart rate at office 78.3 [10.5]* 70.7 [7.3]* 72.0 [8.3] 73.0 [8.9] 73.3 [14.0] 0.006* Heart rate at home 77.0 [9.9] 71.7 [9.7] 73.3 [8.5] 71.8 [7.1] 71.6 [19.2] 0.062 Visit 3 No of patients prescribed with add-on drug at visit 3 n = 19 n = 16 n = 457 n = 12 n = 11 Systolic–office 139.2 [23.6] 139.2 [14.7] 133.3 [13.3] 133.3 [25.5] 128.3 [11.7] 0.117 Systolic–home 137.9 [16.9] 139.8 [15.1] 135.6 [13.8] 131.9 [19.7] 131.7 [9.7] 0.272 Diastolic–office 83.3 [10.0] 83.3 [8.3] 82.7 [8.2] 82.0 [13.7] 80.0 [15.7] 0.357 Diastolic–home 86.8 [13.4] 84.5 [8.0] 83.2 [8.9] 82.6 [18.6] 76.5 [11.8] 0.106 Heart rate at office 74.0 [7.8] 72.3 [11.7] 71.7 [9.0] 73.3 [11.3] 68.7 [17.0] 0.827 Heart rate at home 74.6 [10.0] 74.3 [8.0] 72.5 [8.4] 69.4 [6.8] 68.2 [7.4] 0.206 Visit 4 No of patients prescribed with add-on drug at visit 4 n = 14 n = 16 n = 459 n = 13 n = 13 Systolic–office 131.7 [21.7] 133.3 [10.7] 132.7 [13.3] 130.0 [22.0] 123.3 [16.7] 0.069 Systolic–home 132.9 [18.9] 137.5 [10.8] 133.2 [11.6] 134.2 [20.0] 130.0 [7.7] 0.280 Diastolic–office 83.3 [13.3] 87.7 [11.7] 81.7 [7.3] 83.3 [9.7] 76.7 [11.7]* 0.025* Diastolic–home 84.6 [14.7] 84.8 [9.1] 82.2 [8.3] 83.0 [8.3] 73.7 [7.9]* 0.001* Heart rate at office 71.3 [13.7] 69.3 [9.0] 71.7 [8.3] 70.7 [9.7] 67.7 [12.7] 0.317 Heart rate at home 71.4 [8.7] 72.0 [7.3] 72.3 [7.8] 70.7 [5.1] 69.4 [9.9] 0.346 Visit 5 No of patients prescribed with add-on drug at visit 5 n = 13 n = 16 n = 460 n = 10 n = 16 Systolic–office 126.2 [12.6] 130.8 [10.4] 130.7 [11.7] 138.3 [21.0] 128.7 [13.0] 0.056 Systolic–home 129.0 [12.6] 132.4 [10.9] 131.8 [10.8] 134.9 [18.4] 129.0 [9.5] 0.391 Diastolic–office 79.3 [4.1] 83.8 [4.2] 80.3 [7.3] 83.3 [9.3] 76.7 [12.0] 0.069 Diastolic–home 79.6 [5.8] 84.1 [7.3] 81.2 [7.6] 84.7 [7.6] 74.7 [14.0] 0.030 Heart rate at office 72.0 [12.7] 70.3 [7.4] 71.3 [8.3] 70.7 [10.3] 66.7 [12.8] 0.586 Heart rate at home 70.0 [8.0] 72.5 [9.6] 71.3 [7.6] 69.2 [13.1] 69.6 [9.3] 0.472 Visit 6 No of patients prescribed with add-on drug at visit 6 n = 13 n = 16 n = 460 n = 10 n = 16 Systolic – office 122.5 [7.5] 128.5 [18.3] 129.3 [10.7] 131.7 [14.9] 130.7 [15.0] 0.078 Systolic–home 125.9 [8.3] 129.1 [10.7] 129.7 [9.4] 128.9 [12.7] 128.6 [12.6] 0.246 Diastolic–office 79.2 [4.6] 80.7 [5.0] 80.0 [5.3] 83.3 [8.2] 80.0 [16.7] 0.370 Diastolic–home 78.6 [8.8] 80.7 [7.3] 80.1 [6.9] 84.1 [6.8] 75.9 [19.5] 0.170 Heart rate at office 71.0 [11.8] 69.0 [7.2] 70.7 [8.3] 71.2 [10.4] 71.7 [16.0] 0.819 Heart rate at home 70.1 [9.0] 70.0 [10.8] 71.3 [7.5] 69.8 [8.2] 69.4 [12.3] 0.900

* Significant difference.

Patients treated by add-on therapy with SPL, A2A or SPL + A2A achieved more rapid complete blood pressure control than patients treated with other therapeutic options. The first primary endpoint of the study (decrease of arterial blood pressure to levels < 140/90 mm Hg) was achieved in much larger fraction of patients as early as at visit 2 (≥45%, P = .049) and at visit 3 (>57%, P = .011). At the last visit (visit 6) more than 80% of patients with SPL add-on therapy, and 73.3% of patients with SPL + A2A add-on therapy achieved complete blood pressure control; the details for all therapy groups and visits are shown in the Table 3, and in the Figure 2.

Table 3 - Percent of patients achieving blood pressure Visit Row legend No antihypertensive added First-line AH drug added Spironolactone added Central antihypertensive added Spironolactone + central antihypertensive added Null hypothesis probability 1 No of patients with added antihypertensive at previous visit 387 61 60 7 0 0.176 Percent of responders 20.40% 11.50% 11.70% 14.30% 0.00% 2 No of patients with added antihypertensive at previ