Nebivolol Vs Atenolol and Placebo in Essential Hypertension: a Double-Blind Randomised Trial

ORIGINAL ARTICLE Nebivolol vs atenolol and placebo in essential hypertension: a double-blind randomised trial

L Van Nueten1, FR Taylor2 and JIS Robertson1 1Janssen Research Foundation, Beerse, Belgium; 2The Surgery, Hexton Road, Barton le Clay, Bedfordshire, UK

A double-blind, randomised, parallel-group trial was cant falls in heart rate. Both active drugs were well toler- conducted in patients with essential hypertension in ated, nebivolol marginally more so. Nebivolol, a long- British general practices, of nebivolol 5 mg, atenolol acting, cardioselective, vasodilating beta-blocker which 50 mg, and placebo each given once daily. Both active acts partly via the l-arginine/nitric oxide mechanism, drugs, in comparison with placebo, caused highly sig- appears potentially valuable for the treatment of hyper- niﬁcant and similar reductions in systolic and diastolic tension. pressures without orthostatic effect, and small signiﬁ-

Keywords: nitric oxide; quality of life; side effects

Introduction daily were begun. After 4 weeks of this single-blind placebo run-in, those with a sitting diastolic press- Nebivolol is a highly cardioselective beta-blocking ure of at least 95 mm Hg and at most 115 mm Hg agent devoid of intrinsic sympathomimetic activity, could enter the trial proper. A previous dose- with vasodilating properties and a prolonged action; response trial of nebivolol had shown that 5 mg it acts in part via the l-arginine/nitric oxide path- once daily in hypertensive patients controlled blood way. At similar anti-hypertensive doses, nebivolol pressure (BP) over 24 h; no additional effect was has been found to exert less beta-1 blockade than do seen with 10 mg.10 Likewise, atenolol 50 mg once standard beta-blockers, as judged by the extent of daily had earlier been found to be as effective as reduction of exercise-induced tachycardia.1–9 This 100 mg.11 Thus the doses of active drugs used were paper reports a double-blind randomised trial of once-daily nebivolol compared with once-daily nebivolol 5 mg and atenolol 50 mg. Patients were atenolol, a well established cardioselective beta- randomised via a computer-generated list to receive, blocker and placebo, in essential hypertension. double-blind, once-daily placebo, or nebivolol 5 mg, or atenolol 50 mg, for the trial duration of 1 month, further visits being scheduled at 2 weeks and at the Patients and methods end of this time. If after 2 weeks of the double-blind Ͼ The trial was conducted in patients under primary treatment the diastolic pressure was 120 mm Hg, Ͼ medical care (general practice) in Britain. Male and or the systolic 220 mm Hg, patients were with- female ambulatory subjects, between the ages of 18 drawn and considered as non-responders. Patients and 71 years, with mild-to-moderate essential could also withdraw at any time should they wish; hypertension, were eligible; those previously or for any reason at the discretion of the investigator, untreated, or who had responded poorly or were or if a serious adverse event occurred. intolerant of previous treatment, could be recruited. All trial medication consisted of tablets identical Exclusion criteria are listed in Table 1. Those in appearance and taste. Patients were instructed to entering were required to give informed consent. At take one tablet once daily at breakfast, except on the the screening visit, initial assessments were perfor- occasion of clinic visits, when medication was omit- med, candidate patients were asked to discontinue ted until after BP had been taken. Thus the BP any current anti-hypertensive drugs within the next values reported (Figures 1–4) are those obtained 23– 2 weeks, while single-blind placebo tablets once 25 h after dosing. Drugs that might interfere with the anti-hypertensive effect of the trial medication were not allowed (see Table 1). Other medications were Correspondence: Dr L Van Nueten, Janssen Research Foundation, permitted as needed provided they remained International Clinical R&D, Internal Medicine, Turnhoutseweg unchanged during the trial; short-term treatment of 30, 2340 Beerse, Belgium Received 20 June 1997; revised 1 October 1997; accepted 10 intercurrent illness was also allowed. Compliance October 1997 was assessed by tablet count at the end of the Nebivolol and atenolol in essential hypertension L Van Nueten et al 136 Table 1 Exclusion criteria

ț secondary hypertension, or malignant hypertension (retinal haemorrhage, exudates or papillary oedema) ț asthma or chronic obstructive airway disease ț bradycardia (less than 60 beats per minute at rest) ț atrial fibrillation or recurrent tachyarrhythmia requiring antiarrhythmic therapy ț insulin dependent diabetes ț history of sensitivity or severe adverse reaction to beta-blockers ț myocardial infarction or cerebrovascular accident within the last 6 months ț heart failure requiring treatment, or valvular disease of haemodynamic significance ț severe renal or hepatic disease (urine protein Ͼ trace, creatinine Ͼ 2.2 mg/dl or Ͼ 200 Tmol/l, ASAT and/or ALAT greater than twice the upper normal limit) ț (possible) pregnancy or nursing ț any concurrent condition that could, according to the investigator, jeopardise participant adherence to the protocol or ability to complete the trial (eg, alcohol or drug abuse, disabling or terminal illness, personality or mental disorders) ț concurrent therapy with medications that could affect blood pressure (therapy with tricyclic antidepressants drugs or MAO inhibitors; corticosteroids or non-steroidal anti-inflammatory drugs, if used regularly; hormonal contraceptives) ț investigational drug treatment within the past 120 days ț predictable lack of co-operation

Figure 1 Diastolic blood pressure, sitting: (means ± s.e.m.). W = Week of trial. The number of patients contributing to each value is given. (– – –) placebo; (------) atenolol; (——) nebivolol.

double-blind trial period. Major dietary changes, eg, taneously, or if recalled in response to the standard for weight reduction or salt restriction, were not per- question, ‘Has treatment upset you in any way?’. mitted during the trial. Positive experiences were recorded if reported spon- Systolic and diastolic pressures were recorded, taneously, or in response to the standard question, using a random-zero sphygmomanometer, taking ‘Did you experience any beneficial effect from treat- Korotkoff phases 1 and 5, respectively. In each ment?’. A quality of life questionnaire comprising patient, measurements were made by the same 91 questions was completed by a subset of patients investigator or member of the investigator’s staff, in at the end of the 4-week double-blind period. A 12- the same room and at the same time of day, prefer- lead electrocardiogram was recorded at screening, at ably in the morning. At the first visit BP was determ- the end of the placebo run-in, and at the end of the ined in both arms; if a difference of Ͼ4 mm Hg was double-blind treatment period. Routine haematolog- found, the arm with the higher pressure was used ical, biochemical, and urine tests were performed at throughout, otherwise the right arm was always screening, at the end of the placebo run-in, and at employed. The sphygmomanometer cuff was at least the end of the double-blind treatment period. 30 cm long and 13 cm wide; if the arm circumfer- On the assumption that a difference in supine ence exceeded 34 cm, an appropriately larger cuff diastolic pressure by a mean ± s.d. of 5 mm Hg ± was used. Sitting BP was taken after the patient had 11.5 would occur between placebo and at least one rested for 5 min; standing pressure after 2 min of of the other treatment groups, it was calculated that standing. Heart rate was counted sitting and stand- a total of 414 patients (138 per treatment group) ing. At each visit patients were weighed partially would be needed to detect the difference with a 5% clothed. two-tailed significance and 95% power. Adverse events were noted if mentioned spon- The primary pre-declared index variable was sit- Nebivolol and atenolol in essential hypertension L Van Nueten et al 137

Figure 2 Diastolic blood pressure, standing: (means ± s.e.m.). W = Week of trial. The number of patients contributing to each value is given. (– – –) placebo; (------) atenolol; (——) nebivolol.

Figure 3 Systolic blood pressure, sitting: (means ± s.e.m.). W = Week of trial. The number of patients contributing to each value is given. (– – –) placebo; (------) atenolol; (——) nebivolol. ting diastolic pressure at ‘trough’ drug level, 23–25 h decrease in sitting diastolic pressure at trough drug from the previous intake of medication, using the level to 90 mm Hg or below. The ␹2 test was used to last available reading in each patient (end-point), detect between-group differences for the percentages whether or not that patient completed the intended of responders. The quality of life questionnaire was full trial duration. An ‘intent to treat’ analysis was evaluated descriptively and by non-parametric test performed, that is, all randomised patients, regard- (Wilcoxon–Mann–Whitney U-test). less of their compliance with the protocol, were included. Treatment effects were evaluated at a two- Results tailed 5% significance level by means of a one-way analysis of variance. If the overall comparison was A total of 366 patients entered the run-in period, ie, significant, multiple comparisons were made by slightly fewer than were intended from the above Dunnett’s t-test to assess differences between nebiv- statistical considerations. Of these, two dropped out olol and placebo and between nebivolol and ateno- before randomisation. Of the remainder, 119 were lol. Response rate (normalisation) was defined as a allocated to nebivolol, 121 to atenolol and 124 to Nebivolol and atenolol in essential hypertension L Van Nueten et al 138

Figure 4 Systolic blood pressure, standing: (means ± s.e.m.). W = Week of trial. The number of patients contributing to each value is given. (– – –) placebo; (------) atenolol; (—–—) nebivolol.

placebo. The characteristics of the patients at given. It can be seen that by 2 weeks, and then randomisation are shown in Table 2. The three throughout the period of active therapy, both nebiv- treatment groups were broadly similar in all relevant olol and atenolol lowered both systolic and diastolic features; the racial composition was overwhelm- pressures significantly (P Ͻ 0.001) in comparison ingly Caucasian. It is apparent from Table 2 that 17 with placebo, but with no appreciable differences in patients at entry showed BP outwith the predefined effect between the two drugs. Both active drugs pre- trial limits; four below and 13 above. Tablet count served the slight rise in diastolic pressure which indicated that all except one patient on atenolol took occurs normally on standing. The percentages of at least 80% of the prescribed medication during the pre-defined responders were nebivolol 59%, ateno- double-blind treatment period. lol 59% and placebo 29% (both active drugs P Ͻ The systolic and diastolic pressures sitting and 0.001 vs placebo). standing through the trial at ‘trough’ drug levels in As expected, resting (sitting) heart rate fell sig- the three treatment groups are shown in Figures 1– nificantly with active treatment; at end-point vs pla- 4. The numbers of patients in each group contribu- cebo by a mean of 7.8 beats per minute on nebivolol ting to the measurements at each time point are also and 7.7 on atenolol (both P Ͻ 0.001 vs placebo).

Table 2 Patient characteristics

Nebivolol Atenolol Placebo (n = 119) (n = 121) (n = 124)

Sex: M/F, n 68/51 57/64 66/58 Age: median (min–max), yrs 55 (21–70) 55 (21–71) 55 (19–71) Weight: median (min–max), kga 78.8 (48–105) 75 (50–119) 75.3 (46–119) Height: median (min–max), cm 168 (145–185) 165 (141–189) 167 (145–198) Body mass indexa median (min–max), kg/m2 27.7 (18–46) 27.2 (21–53) 26.7 (18–41) normal weight, n 24 24 31 overweight, n 94 97 93 Race: black/caucasian/mixed/oriental, n 2/114/0/3 2/115/1/3 5/118/0/1 Low salt diet, na 19 22 20 Alcohol consumption, n 72 74 78 Smoking, n 32 28 31 Severity of hypertension, n DBP Ͻ95 mm Hg 2 2 mild (95 р DBP Ͻ100 mm Hg) 31 32 30 moderate (100 р DBP Ͻ115 mm Hg) 86 84 84 severe (DBP у 115 mm Hg) 5 8

No significant intergroup differences. aData on weight, body mass index and low salt diet not available in some patients (no more than one patient in any group). Nebivolol and atenolol in essential hypertension L Van Nueten et al 139 There were no significant changes in body weight Both active drugs were well tolerated, nebivolol during the trial in any of the three treatment groups. marginally more so. Earlier work showed that at During double-blind therapy, adverse events were similar anti-hypertensive doses to standard beta- reported by 33 (28%) of nebivolol patients, 37 (31%) blockers, nebivolol exerted less beta-1 blockade, of atenolol patients and by 31 (25%) on placebo. The assessed by the capacity to limit exercise-induced only notable treatment-related complaint was tachycardia.1,2,4,5 However, resting heart rate, by fatigue, which was reported during the double-blind contrast, was in the present trial, lowered similarly phase in four nebivolol and eight atenolol patients by nebivolol and atenolol. and in one placebo patient. There were at baseline Nebivolol possesses several properties which are differences in the frequency of adverse events in the potentially attractive in the treatment of hyperten- three groups. During the double-blind treatment per- sion. It acts partly through the l-arginine/nitric iod, the increase in fatigue was found to be more oxide pathway, has vasodilating effects, and prominent with atenolol than with nebivolol or pla- improves large arterial compliance.1–10,12 Nebivolol cebo (Table 3). has been shown to enhance left ventricular function During double-blind treatment, the only note- in patients with cardiac impairment, and to be well- worthy positive experience was of enhanced well- tolerated in subjects with cardiac failure.13–17 This being, reported by 47 patients (15 nebivolol, 12 aten- aspect has considerable potential. Many patients olol, 20 placebo). with hypertension have concomitant coronary artery A total of 285 patients, 151 men and 134 women, disease, and the latter is now the most frequent completed the quality of life questionnaire at base- cause of impaired cardiac function.18 Further, line and at treatment end-point. For the total patient arterial endothelial damage and dysfunction are groups, no statistically significant differences were thought possibly to be contributory to the pathogen- found between the three treatments either for the esis of atherosclerosis (atheroma) in hypertension.19 summary scales or composite scales. However, in Drugs capable of lessening such disorders might the atenolol group only there was a significant help to limit the arterial complications of hyperten- deterioration (P Ͻ 0.05) from baseline to end-point sion. in response to the query concerning, ‘Problems in Nebivolol should thus be a valuable addition to maintaining interest during sex’. the antihypertensive therapeutic repertoire. While Electrocardiogram confirmed the slight slowing of several of these aspects are at present speculative, heart rate in those on the two active drugs, but was the drug merits further study regarding its capacity otherwise unremarkable. There were no consistent to limit hypertension-related vascular and cardiac or relevant changes in routine biochemical, haema- disease. tological, or urinary tests at any stage. List of participating physicians Discussion Investigator (Location) The principal aim of the present study was to estab- lish the efficacy, acceptability and safety of nebivo- Agarwala K (Manchester); Agrawal VK (Nuneaton); lol in comparison with the well-established car- Bakhru MN (Hertfordshire); Bayman I (Coventry); dioselective beta-blocker atenolol in the treatment of Berger AB (Milton Keynes); Bhaduri PK (Wigan); essential hypertension in UK primary care (general Bowron P (Bishop Auckland); Brown MR (Glasgow); practice). The doses selected of the active drugs, Burman A (Manchester); Dass BK (Manchester); 5 mg rather than 10 mg of nebivolol,10 and 50 mg Evans TW (Stirling); Fradd SO (Nottingham); Gould rather than 100 mg of atenolol,11 were considered on CHG (Co-Antrim); Guest ND (Cheshire); Hague IU the evidence of the previous dose-response trials (Manchester); Hossain M (London); Huda ZU (Co- quoted, to provide near optimal anti-hypertensive Tyrone); Hughes IW (Liverpool); James IGV (Bolton); efficacy plus acceptability. These aspects were con- Jones BP (Manchester); Kane JA (Surrey); Kapur YP firmed, and nebivolol 5 mg once daily was shown to (Clwyd); Khatri CP (Manchester); Kukula MS be similarly effective to atenolol 50 mg once daily. (Manchester); Lennox B (Ayr); Lone IA

Table 3 Adverse experiences (AE) reported in at least three patients (BL = baseline; DB = double-blind): number of patients studied in each group shown

placebo nebivolol atenolol

BL DB BL DB BL DB (n = 124) (n = 124) (n = 119) (n = 119) (n = 121) (n = 121)

Fatigue 3 1 4 4 1 8 Headache 5 7 4 3 3 4 Dyspnoea 0 3 0 1 0 4 Insomnia 1 4 0 2 0 1 Dizziness 2 2 2 3 0 2 Paresthesia 0 1 0 3 0 2 Per cent of patients with AE 23 25 22 29 17 31 Nebivolol and atenolol in essential hypertension L Van Nueten et al 140 (Middlesbrough); Martin GDR (Nottingham); mechanism. J Pharmacol Exp Ther 1995; 274: 1067– Mayhew SR (Cambridgeshire); Mc Gregor GS 1071. (Bishop Auckland); Miller S (London); Mukherjee 8 Bowman AJ, Chen CPL-H, Ford GA. Nitroc oxide SK (Merseyside); Multani SS (Falkirk); Nagpal IC mediated venodilator effects of nebivolol. Br J Clin Pharmacol 1994; 38: 199–204. (Mexborough, Yorkshire); Nankani JPN (Wrexham); ´ 9 Van Rooy P, Van Nueten L, De Cree J. Nebivolol blood Patel NH (Sheffield); Pepperman MA pressure reduction in relation to beta-blockade in heal- (Northamptonshire); Pitalia PL (Wigan); Rau UBN thy volunteers. Drug Invest 1991; 3 (Suppl 1): 174– (Staffordshire); Sagar D (Nottinghamshire); Scuda- 176. more JA (Bedfordshire); Selfridge DI (Grangemouth); 10 Van Nueten L et al. A dose-response trial of nebivolol Selvam A (Staffordshire); Sheldon DM (Newton in essential hypertension. J Hum Hypertens 1997; 11: Aycliffe); Snook R (Neston, Cheshire); Stirling AW 139–144. (Cheshire); Tandon AK (Preston); Taylor FR 11 Scott AK et al. Atenolol and metoprolol once daily in (Bedfordshire); Vilath E (London); Whitton ADC hypertension. Br Med J 1982; 284: 1514–1516. (Lancaster). 12 Van Merode T et al. Verapamil and nebivolol improve carotid artery distensibility in hypertensive patients. J Hypertens 1989; 7 (Suppl 6): S262–S263. 13 Stoleru L et al. Effects of d-nebivolol and l-nebivolol on left ventricular systolic and diastolic function: com- References parison with d-l nebivolol and atenolol. J Cardiovasc Pharmacol 1993; 22: 183–190. 1 Fitzgerald JD (Ed). First International Nebivolol Inves- 14 Goldstein M et al. Administration of nebivolol after tigators’ Meeting. Drug Invest 1991; 3 (Suppl 1): 1–203. coronary bypass graft in patients with altered left ven- 2 Van de Water A et al. Pharmacological and hemodyn- tricular function. J Cardiovasc Pharmacol 1993; 22: amic profile of nebivolol, a chemically novel, potent 253–258. and selective ␤1-adrenergic antagonist. J Cardiovasc 15 Wisenbaugh T et al. Long-term (3 month) effects of a Pharmacol 1988; 11: 552–563. new beta-blocker (nebivolol) on cardiac performance 3 Pauwels PJ et al. The receptor binding profile of the in dilated cardiomyopathy. J Am Coll Cardiol 1993; 21: new antihypertensive agent nebivolol and its stereo- 1094–1100. isomers compared with various ␤1-adrenergic block- 16 Uhlir O et al. Nebivolol in the treatment of cardiac fail- ers. Mol Pharmacol 1988; 34: 843–851. ure: a double-blind controlled clinical trial. J Card Fail 4 Janssens WJ et al. Nebivolol is devoid of intrinsic sym- 1998; (in press). pathomimetic activity. Eur J Pharmacol 1988; 159: 17 Rousseau MF et al. Medium term effects of beta-block- 89–95. ade on left ventricular mechanics: a double-blind, pla- 5 Van de Water A, Xhonneux R, Reneman RS, Janssen cebo-controlled comparison of nebivolol and atenolol PAJ. Cardiovascular effects of dl-nebivolol and its in patients with ischemic left ventricular dysfunction. enantiomers—a comparison with those of atenolol. Eur J Card Fail 1996; 2: 15–23. J Pharmacol 1988; 156: 95–103. 18 Robertson JIS, Ball SG. The heart in hypertension. 6 Gao Y et al. Nebivolol induces endothelium-depen- Hypertension for the Clinician. WB Saunders: dent relaxations of canine coronary arteries. J Cardiov- London/Philadelphia, 1994, pp 77–79. asc Pharmacol 1991; 17: 964–969. 19 Robertson JIS, Ball SG. Endothelium-derived relaxing 7 Cockroft JR et al. Nebivolol vasodilates human forearm factor (EDRF; nitric oxide); Arterial changes in hyper- vasculature: Evidence for an L-arginine/NO-dependent tension. Ibid, 48 and 60–68.