Very limited evidence suggests that sublingual nifedipine improves exercise capacity in people with stable angina compared with placebo or no treatment. There is weak evidence to suggest that sublingual glyceryl trinitrate (GTN) is more effective than sublingual nifedipine. Available evidence suggests that there is no difference in efficacy between GTN spray and GTN tablets.

NICE undertook a review of the evidence to determine the effectiveness of short-acting drugs for the management of angina symptoms [National Clinical Guideline Centre, 2011b]. The review included short-acting nitrates (glyceryl trinitrate) and nifedipine administered via the buccal mucosa.

Sublingual nifedipine versus placebo or no treatment

The review identified two small, low quality randomized crossover trials (RCTs) comparing sublingual nifedipine with placebo (n = 10) or no treatment (n = 10).

Compared with placebo or no treatment, sublingual nifedipine was associated with a significant improvement in exercise capacity.

GTN versus sublingual nifedipine

The review identified two small, low quality RCTs (n = 23) comparing sublingual GTN with sublingual nifedipine for the relief of angina symptoms.

Compared with sublingual nifedipine, sublingual GTN was significantly more effective at reducing pain severity and providing complete symptom relief at 2 and 4 minutes after treatment.

There was no statistically significant difference between sublingual nifedipine and sublingual GTN in the mean exercise time to 1 mm ST depression.

Comparing formulations of GTN

The report identified one open-label RCT (n = 126) comparing sublingual GTN with buccal GTN, and one small RCT (n = 23) comparing GTN spray with GTN tablets.

There was no significant difference reported between the formulations for the following outcomes: mean change in exercise undertaken, time to develop angina, and duration of angina.

Conclusions of the NICE Guideline Development Group (GDG)

The GDG concluded that people with stable angina should be offered a short-acting drug to relieve episodes of angina, and that weak evidence suggests that GTN relieves episodes of angina more effectively than nifedipine [National Clinical Guideline Centre, 2011b].

There is little evidence on the efficacy of either beta-blockers or calcium-channel blockers (CCBs) compared with placebo in the treatment of angina; however, they are both well established treatments for this indication.

Evidence from several randomized controlled trials (RCTs) showed no difference in long-term total or cardiovascular mortality, or in the risk of adverse cardiovascular events between treatment with a CCB or a beta-blocker in people with stable angina. There is evidence that this is consistent across subgroups, including women, people with diabetes, and those aged over 70 years of age.

There is conflicting evidence on the comparative effects of beta-blockers and CCBs on angina symptoms and exercise tolerance. However, differences shown between the two classes of drug were not considered to be clinically significant.

The National Institute for Health and Clinical Excellence (NICE) reviewed the evidence comparing beta-blockers and CCBs in the management of angina, looking in particular at long-term mortality and major adverse cardiovascular events.

For evidence comparing beta-blockers and CCBs with other anti-anginal drugs, see the individual evidence sections on nitrates, nicorandil, ivabradine, and ranolazine.

Long-term mortality and cardiovascular adverse events

The NICE review identified three RCTs that looked at long-term mortality and cardiovascular adverse events. Meta-analysis of these RCTs found no significant difference between beta-blockers and CCBs for:

Total mortality (two RCTs); relative risk (RR) 1.02 (95% CI 0.93 to 1.11).

Cardiovascular death (three RCTs); RR 0.99 (95% CI 0.87 to 1.12).

Non-fatal myocardial infarction (three RCTs); RR 0.99 (95% CI 0.81 to 1.22).

Cardiovascular-related hospitalisations (one RCT); RR 0.97 (95% CI 0.88 to 1.08).

Non-fatal cardiovascular events (one RCT); RR 1.07 (95% CI 0.85 to 1.36).

These results were consistent across subgroups including women, people with diabetes, and people aged over 70 years.

Angina symptoms and exercise tolerance

The NICE review identified five additional RCTs which looked at angina symptoms and exercise tolerance. Meta-analysis showed:

Significantly fewer episodes of angina per week with a CCB compared with a beta-blocker (four RCTs); mean difference 0.11 (95% CI 0.07 to 0.15). However, the Guideline Development Group did not consider this difference (equivalent to a single episode of angina every 9 weeks) to be clinically significant.

No significant difference between CCBs and beta-blockers for prevalence of angina after a mean follow-up of 2.7 years (one RCT); RR 0.87 (95% CI 0.73 to 1.04).

No significant difference between CCBs and beta-blockers for severity of angina, assessed by investigators as moderate or markedly improved (one RCT); RR 0.72 (95% CI 0.51 to 1.02).

No significant difference between CCBs and beta-blockers for severity of angina, assessed by patients as moderate or severe (one RCT); RR 1.33 (95% CI 0.73 to 2.45).

No significant difference between CCBs and beta-blockers in the number of glyceryl trinitrate (GTN) tablets used per week (one RCT); mean difference 0.00 (95% CI -0.94 to +0.94).

No significant difference between beta-blockers and CCBs (diltiazem, nifedipine) for exercise duration (mean difference 0.05 minutes; 95% CI -0.82 to +0.92); time to 1 mm ST segment depression (mean difference 12 seconds; 95% CI -35.06 to +59.06); or time to onset of angina (mean difference 0.63 minutes; 95% CI -0.27 to +1.53).

Conclusions of the NICE Guideline Development Group (GDG)

The GDG concluded that there is no evidence to discriminate between a beta-blocker and a CCB for the initial treatment of stable angina.

Limited data are available on the efficacy of long-acting nitrates in the treatment of angina, compared with either placebo or a comparator drug. There is consensus that nitrates can be considered for monotherapy if beta-blockers and calcium-channel blockers (CCBs) are not tolerated or are contraindicated.

Evidence on the use of nitrates in combination with other anti-anginal drugs is poor. Results from small studies suggest that adding a long-acting nitrate to a beta-blocker provides better symptom relief than a beta-blocker alone. Addition of a nitrate to a beta-blocker may be less effective than the combination of a beta-blocker plus a CCB.

Nitrates as monotherapy

The National Institute for Health and Clinical Evidence (NICE) did not review the evidence on nitrates as monotherapy.

CKS identified one systematic review (search date: to 1996) [Sculpher et al, 1998] and one meta-analysis (search date: to 1997) [Heidenreich et al, 1999] that included studies of nitrates as monotherapy for angina.

The systematic review identified one randomized controlled trial (RCT) comparing isosorbide dinitrate with a beta-blocker (propranolol) in people with untreated stable angina. Symptom control was not an outcome of this study.

The meta-analysis identified 12 trials comparing a long-acting nitrate with a CCB and six trials comparing a long-acting nitrate with a beta-blocker.

Long-acting nitrate versus CCB

The two groups did not significantly differ in the frequency of angina episodes, use of glyceryl trinitrate, duration of exercise, and cardiac death or myocardial infarction.

Long-acting nitrate versus beta-blocker

The two groups did not significantly differ in the frequency of angina episodes, use of glyceryl trinitrate, exercise duration, and cardiac death or myocardial infarction.

The authors concluded that there were too few trials comparing long-acting nitrates with CCBs or beta-blockers to draw definite conclusions about their relative efficacy.

Nitrates in combination with other anti-anginal drugs

NICE reviewed the evidence on nitrates as combination therapy in the management of angina, looking in particular at long-term mortality and major adverse cardiovascular events, and symptom severity [National Clinical Guideline Centre, 2011b]. The review identified two RCTs comparing a beta-blocker plus a nitrate with a beta-blocker plus a CCB. NICE identified serious limitations in both studies. NICE found no trials of nitrates in combination with a CCB.

In one 12-week study (n = 46), there were no significant differences between the two groups for exercise time, time to onset of angina, and time to ST segment depression. There were also no significant differences for adverse effects overall, stopping due to adverse events, and headache.

In the second study, the combination of a beta-blocker plus a CCB resulted in greater reduction in the frequency of angina, and improved exercise times than the combination of a beta-blocker plus a nitrate. This study was reported to be of poor quality.

CKS also identified one randomized, double-blind trial comparing beta-blockers and isosorbide mononitrate with beta-blockers and placebo [Uusitalo et al, 1988].

Seventy people reporting five or more attacks of stable angina in a week (41 of whom had a history of myocardial infarction) were included; all were taking beta-blockers.

Compared with placebo, addition of isosorbide mononitrate to the beta-blocker for 2 weeks significantly increased total exercise capacity and time until chest pain (p < 0.01) or ST-segment depression (p < 0.001). The number of angina attacks and use of short-acting nitrate were significantly decreased.

Conclusions of the NICE Guideline Development Group (GDG)

The NICE GDG made a consensus recommendation that long-acting nitrates can be considered for monotherapy if beta-blockers and CCBs are not tolerated or are contraindicated.

The GDG also agreed that the addition of a long-acting nitrate can be considered in people whose symptoms are not controlled by monotherapy with either a beta-blocker or a CCB if the combination of a beta-blocker plus a CCB is not appropriate.

Limited evidence suggests that there is no significant difference in the frequency of angina episodes or in total exercise capacity with nicorandil compared with standard anti-anginal drugs.

Evidence from one large randomized controlled trial (RCT) suggests that there may be some prognostic benefit from adding nicorandil to standard anti-anginal treatment, however, this was largely due to a decrease in the number of hospital admissions for chest pain.

Nicorandil as monotherapy

The National Institute for Health and Clinical Evidence (NICE) reviewed the evidence on nicorandil as monotherapy for stable angina [National Clinical Guideline Centre, 2011b]. The review found no evidence of the effects of nicorandil as monotherapy on long-term mortality or rates of major cardiovascular adverse events.

The NICE review included three RCTs comparing nicorandil with a CCB (diltiazem, amlodipine, or nifedipine) in the management of stable angina; one RCT comparing nicorandil with the beta-blocker, propranolol; and one RCT comparing nicorandil with isosorbide mononitrate. The trials were short-term (follow-up between 2 weeks and 90 days) and of low-to-moderate quality.

There were no significant differences in the frequency of angina episodes or in the total exercise capacity with nicorandil compared with standard anti-anginal drugs.

Nicorandil in combination with usual treatment

NICE also reviewed the evidence on nicorandil in combination with standard anti-anginal treatment [National Clinical Guideline Centre, 2011b]. The review identified one large (n = 5126), long-term RCT (the Impact of Nicorandil in Angina [IONA] study) that compared the use of nicorandil with placebo as add-on therapy to standard anti-anginal treatment (56% beta-blocker, 55% CCB, and 87% nitrate) [IONA Study Group, 2002].

The primary composite outcome of coronary heart disease (CHD) death, non-fatal MI, or unplanned hospital admission for chest pain was significantly reduced in the nicorandil group compared with placebo (relative risk [RR] 0.85; 95% CI 0.74 to 0.97).

There were no significant differences between nicorandil and placebo for CHD death (RR 0.82; 95% CI 0.59 to 1.15), non-fatal myocardial infarction (RR 0.78; 95% CI 0.55 to 1.10), all cause mortality (RR 0.86; 95% CI 0.67 to 1.10), unstable angina (RR 0.90; 95% CI 0.71 to 1.16), or worsening of angina status (RR 0.94; 95% CI 0.85 to 1.04).

There was a significantly greater incidence of gastrointestinal disturbances (RR 1.47; 95% CI 1.18 to 1.82) and headaches (RR 4.49; 95% CI 3.55 to 5.67) in the nicorandil group compared with placebo.

Conclusions of the NICE Guideline Development Group (GDG)

The GDG concluded that there was insufficient evidence to recommend monotherapy with nicorandil in preference to monotherapy with a beta-blocker or a CCB as first line treatment for angina. Nicorandil can be considered as monotherapy for the treatment of stable angina if a beta-blocker and a CCB are not tolerated or contraindicated.

The 2.4% absolute reduction in the rate of the primary composite endpoint in IONA did not justify the routine use of nicorandil as add-on therapy to standard anti-anginal treatment in people with stable angina, particularly as nicorandil is associated with an excess risk of adverse events, including headache and gastrointestinal disturbance. The GDG concluded that addition of nicorandil is an option for people whose symptoms of angina are not controlled by a beta-blocker or a CCB. However, it noted that nicorandil is not currently licensed for this indication.

Short-term randomized controlled trials (RCTs) have shown that ivabradine is more effective than placebo at improving exercise tolerance, reducing the frequency of episodes of angina, and reducing use of short-acting nitrates. Evidence from two further RCTs, both over 3 months, suggests that ivabradine is similarly effective to the beta-blocker atenolol and the calcium-channel blocker (CCB) amlodipine.

Evidence from one RCT suggests that, in people already taking a beta-blocker, the addition of ivabradine appears to provide increased control of angina symptoms.

In one large, long-term morbidity study in people with coronary artery disease, ivabradine had no significant effect on a composite outcome of cardiovascular death, admission to hospital for acute myocardial infarction (MI), and admission to hospital for new-onset or worsening heart failure.

Visual disturbances were commonly reported in these studies in people taking ivabradine. The incidence of adverse effects was higher in people taking ivabradine than in those taking amlodipine or atenolol. Psychiatric adverse effects (including one suicide) were reported in people taking ivabradine.

Monotherapy

Ivabradine versus placebo:

One dose-ranging RCT compared ivabradine with placebo in 360 people with stable angina [Borer et al, 2003]. In the initial phase, participants were randomized to ivabradine (2.5 mg, 5 mg, or 10 mg twice a day) or placebo for 2 weeks; other anti-anginal medication (including beta-blockers, CCBs, and long-acting nitrates) was stopped. This was followed by an open-label extension in which 173 participants received ivabradine 10 mg twice a day for 2–3 months.

Initial phase

Time to limiting angina during an exercise tolerance test (ETT) increased with ivabradine; the change did not reach significance.

Time to 1 mm ST-segment depression during ETT increased with ivabradine treatment; the change was statistically significant relative to placebo in the ivabradine 5 mg and 10 mg groups; a dose effect was seen across all doses.

Frequency of angina attacks and use of short-acting nitrates were reduced with ivabradine treatment; the change did not reach statistical significance.

Extension phase

The improvements in time to 1 mm ST-segment depression and time to limiting angina seen with ivabradine treatment in the double-blind phase were maintained in the open-label extension.

The frequency of angina attacks decreased from 4.14 (+/– 5.59) to 0.95 (+/– 2.24) attacks per week (p < 0.001).

Use of short-acting nitrates decreased from 2.28 (+/– 3.74) to 0.50 (+/– 1.14) units per week (p < 0.001).

Adverse effects

During the initial phase, visual symptoms were reported by one person in each of the ivabradine 2.5 mg and 5 mg groups and by 13 people (14.8%) in the ivabradine 10 mg group.

During the extension phase, 31 people (17.9%) experienced visual symptoms, three of whom withdrew from the study because of these symptoms.

All visual symptoms were reported to resolve spontaneously during or after drug discontinuation.

Ivabradine versus amlodipine:

One double-blind RCT with 1195 participants compared the efficacy of ivabradine with that of amlodipine [Ruzyllo et al, 2007]. People with stable angina were randomized to ivabradine 7.5 mg twice a day, ivabradine 10 mg twice a day, or amlodipine 10 mg once a day for 3 months.

Total exercise duration during ETT increased by 27.6 (+/– 91.7) seconds in the ivabradine 7.5 mg group, 21.7 (+/– 94.5) seconds in the ivabradine 10 mg group, and 31.2 (+/– 92.0) seconds in the amlodipine group. Both doses of ivabradine were non-inferior to amlodipine (p < 0.001).

Similar results were obtained for time to angina onset and time to 1 mm ST-segment depression. Both doses of ivabradine were non-inferior to amlodipine.

People in all three groups showed a reduction of about 60% (about three attacks per week) in the self-reported frequency of angina attacks. There was no evidence of a significant difference between either dose of ivabradine and amlodipine.

People in all three groups showed a reduction of 50–60% (about 3 units per week) in the self-reported frequency of nitrate use. There was no evidence of a significant difference between either dose of ivabradine and amlodipine.

Adverse effects

A total of 30 people (7.5%) in the ivabradine 7.5 mg group, 24 (6.1%) in the ivabradine 10 mg group, and 21 (5.2%) in the amlodipine group withdrew from the study because of adverse effects (details not reported).

The incidence of adverse events was higher in the ivabradine 7.5 mg and 10 mg groups (48% and 55% respectively) than in the amlodipine group (38%). The incidences of visual symptoms and sinus bradycardia were greater in the ivabradine groups. The incidence of peripheral oedema was greater in the amlodipine group.

Ivabradine versus atenolol

One double-blind RCT with 939 participants compared the efficacy of ivabradine with atenolol [Tardif et al, 2005]. People with stable angina were randomized to ivabradine 5 mg twice a day for 4 weeks then 7.5 mg twice a day for 12 weeks; ivabradine 5 mg twice a day for 4 weeks then 10 mg twice a day for 12 weeks; or atenolol 50 mg once a day for 4 weeks then 100 mg once a day for 12 weeks.

Total exercise duration during ETT increased by 86.8 (+/– 129.0) seconds in the ivabradine 7.5 mg group, 91.7 (+/– 118.8) seconds in the ivabradine 10 mg group, and 78.8 (+/– 133.4) seconds in the atenolol group. Both doses of ivabradine were non-inferior to atenolol (p < 0.001).

Similar results were obtained for time to limiting angina, time to angina onset, and time to 1 mm ST-segment depression. Both doses of ivabradine were non-inferior to atenolol.

The frequency of angina attacks decreased by 2.2 (+/– 4.3) attacks per week in the ivabradine 7.5 mg group, 2.3 (+/– 4.2) attacks per week in the ivabradine 10 mg group, and 2.7 (+/– 12.3) attacks per week in the atenolol group.

Use of short-acting nitrates decreased by 1.6 (+/– 4.1) units per week in the ivabradine 7.5 mg group, 1.4 (+/– 4.7) units per week in the ivabradine 10 mg group, and 1.2 (+/– 3.4) units per week in the atenolol group.

Ivabradine in combination with other anti-anginal drugs:

A double-blind RCT compared the efficacy of ivabradine with placebo in 889 people who were already receiving a beta-blocker [Tardif et al, 2009]. People with a history of at least 3 months of stable angina were given atenolol 50 mg once a day for a 6–8 week run-in period, after which they were randomized to ivabradine 5 mg twice a day for 2 months, followed by 7.5 mg twice a day for 2 months, or to placebo for 4 months.

Total exercise duration during ETT increased by 24.3 (+/– 65.3) seconds in the ivabradine group, compared with 7.7 (+/– 63.8) seconds with placebo (p < 0.001).

Time to limiting angina during ETT increased by 26.0 (+/– 65.7) seconds in the ivabradine group, compared with 9.4 (+/– 63.8) seconds with placebo (p < 0.001).

Time to onset of angina during ETT increased by 49.1 (+/– 83.3) seconds in the ivabradine group, compared with 22.7 (+/– 79.1) seconds with placebo (p < 0.001).

Time to 1 mm ST-segment depression during ETT increased by 45.7 (+/– 93.0) seconds in the ivabradine group, compared with 15.4 (+/– 86.6) seconds with placebo (p < 0.001).

Frequency of angina attacks decreased from 1.8 (+/– 3.3) to 0.9 (+/– 2.4) attacks per week in the ivabradine group, and from 1.6 (+/– 2.4) to 0.9 (+/– 2.1) attacks per week with placebo (between-group difference not significant).

Adverse effects

A total of 13 people (2.9%) in the ivabradine group and four people (0.9%) in the placebo group withdrew from the study because of adverse effects. The most frequent reasons for withdrawal were bradycardia (five people in the ivabradine group) and unstable angina (three people in the ivabradine group and one person in the placebo group).

Bradycardia was reported in 19 people (4.2%) in the ivabradine group and in two people (0.5%) in the placebo group.

Visual disturbances were reported by nine people (2%) in the ivabradine group and four people (0.9%) in the placebo group.

There was one suicide in the ivabradine group.

Effects on morbidity and mortality:

In the morBidity-mortality EvAlUaTion of the I_f inhibitor ivabradine in patients with coronary disease and left-ventricULar dysfunction (BEAUTIFUL) study, 10,917 people with evidence of coronary artery disease were randomized to ivabradine 5 mg twice a day (increased to 7.5 mg twice a day after 2 weeks if resting heart rate remained above 60 beats per minute) or placebo [Fox et al, 2008]. Participants continued to take other cardiovascular medication as considered appropriate, including beta-blockers, angiotensin-converting enzyme inhibitors or angiotensin II receptor antagonists, lipid-lowering drugs, and aspirin or other antiplatelet or antithrombotic drugs. The primary endpoint was the composite of cardiovascular death, admission to hospital for acute myocardial infarction (MI), and admission to hospital for new-onset or worsening heart failure. The median duration of follow up was 19 months.

Treatment with ivabradine did not affect the primary endpoint. Cardiovascular death or admission to hospital for acute MI or new-onset or worsening heart failure occurred in 844 people (15.4%) in the ivabradine group and 832 people (15.3%) in the placebo group (hazard ratio [HR] 1.00, 95% CI 0.91 to 1.10; p = 0.94).

In a prespecified subgroup of people with heart rate of 70 beats per minute or greater, ivabradine treatment did not significantly affect the primary outcome, but it did reduce admission to hospital for MI (HR 0.64, 95% CI 0.49 to 0.84; p = 0.001), admission to hospital for MI or unstable angina (HR 0.78, 95% CI 0.62 to 0.97; p = 0.023), and coronary revascularization (HR 0.70, 95% CI 0.52 to 0.93; p = 0.016).

Adverse effects

A total of 1233 people (23%) in the ivabradine group and 1239 people (23%) in the placebo group reported adverse effects rated as being serious.

Significantly more people in the ivabradine group than the placebo group reported psychiatric disorders (17 compared to 5, p = 0.01).

Long-term safety and efficacy

In a 12-month study, 386 people with stable angina were randomized to ivabradine 5 mg twice a day or ivabradine 7.5 mg twice a day [López-Bescós et al, 2007]. After 12 months:

The frequency of angina attacks decreased from baseline by more than 50% in both groups, from 2.9 (+/– 5.3) to 1.0 (+/– 2.7) attacks per week in the 5 mg group and from 2.3 (+/– 3.9) to 1.1 (+/– 3.8) in the 7.5 mg group.

Adverse effects

A total of 56 people (14.5%) withdrew from the study because of adverse effects or lack of efficacy.

Visual symptoms were the most commonly reported adverse effects, occurring in 23 people (12%) in the 5 mg group and 43 people (23%) in the 7.5 mg group. In total, four people (1%) withdrew because of visual adverse effects.

Serious cardiac adverse effects (other than death) occurred in 13 people (7%) in the 5 mg group and 12 people (6%) in the 7.5 mg group. These included unstable angina, myocardial infarction, rhythm disturbances, and heart failure.

Conclusions of the NICE Guideline Development Group (GDG)

The GDG concluded that monotherapy with ivabradine should not be used as an alternative to monotherapy with a beta-blocker or a CCB, but that it can be considered when beta-blockers and CCBs are contraindicated or not tolerated.

The GDG concluded that ivabradine can be considered as add-on therapy to a beta-blocker for people whose symptoms are not controlled by monotherapy and the addition of a CCB is contraindicated or not tolerated. Ivabradine can also be combined with a CCB (if a beta-blocker is contraindicated or not tolerated).

Short-term randomized controlled trials (RCTs) have shown that ranolazine is more effective than placebo at improving exercise tolerance, reducing the frequency of episodes of angina, and reducing use of short-acting nitrates. Evidence from one RCT suggests that ranolazine is similarly effective to the beta-blocker atenolol. There is no evidence on the effects of ranolazine monotherapy on long-term outcomes in people with stable angina.

Evidence from moderate to high quality RCTs suggests that the addition of ranolazine to current anti-anginal therapy produces further improvements in exercise capacity and reduces the frequency of angina episodes and GTN use compared with placebo. However, the clinical significance of these improvements are unclear.

In one large, long-term morbidity study in people with recent ischaemic symptoms, ranolazine had no significant effect on a composite outcome of cardiovascular death, myocardial infarction (MI), or recurrent ischaemia.

There are concerns regarding the safety of ranolazine including QT prolongation, drug interactions with several other cardiovascular drugs, and safety in renal and hepatic disease.

Ranolazine as monotherapy

NICE found no evidence on the effects of ranolazine monotherapy on long-term outcomes in people with stable angina [National Clinical Guideline Centre, 2011b].

CKS identified two RCTs on the effects of ranolazine as monotherapy. These were excluded from the NICE review because they are both of short duration.

Ranolazine versus placebo

Two double-blind RCTs compared ranolazine with placebo in people with stable angina [Chaitman et al, 2004a; Chaitman et al, 2004b].

The Monotherapy Assessment of Ranolazine In Stable Angina (MARISA) trial was a dose-ranging study that randomized 191 people to modified-release ranolazine 500 mg, 1000 mg, 1500 mg, or placebo twice a day for 1 week in a four-period crossover design; it was not stated whether there was a washout period between crossover periods [Chaitman et al, 2004b]. All other anti-anginal medication was stopped.

Total exercise duration during exercise tolerance testing (ETT) was significantly increased compared with placebo in all ranolazine groups. The increase above placebo was 23.8 (+/– 7.2) seconds with ranolazine 500 mg (p = 0.003), 33.7 (+/– 8.02) seconds with ranolazine 1000 mg (p < 0.001), and 45.9 (+/– 8.0) seconds with ranolazine 1500 mg (p < 0.001).

Similar results were obtained for time to onset of angina and time to 1 mm ST-segment depression during ETT; all doses of ranolazine were significantly better than placebo (p < 0.001).

Adverse effects

Adverse events led to withdrawal from the study in 11 people during treatment with ranolazine 1500 mg twice a day, two people with ranolazine 500 mg twice a day, one person with ranolazine 1000 mg twice a day, and one person with placebo.

Adverse events were reported in 15.6% of people whilst taking placebo, 16.0% whilst taking ranolazine 500 mg twice a day, 21.7% whilst taking ranolazine 1000 mg twice a day, and 34.2% whilst taking ranolazine 1500 mg twice a day.

The most common dose-related adverse effects with ranolazine were dizziness, nausea, asthenia (muscle weakness), and constipation.

Ranolazine versus atenolol

One RCT compared ranolazine with atenolol in 158 people with chronic stable angina [Rousseau et al, 2005].

After withdrawal from their current beta-blocker treatment over 7–10 days, participants were randomized to immediate-release ranolazine 400 mg three times a day (formulation not available in the UK), atenolol 100 mg once a day, or placebo for 1 week in a three-period crossover design; there was no washout period between treatments.

Time to angina onset during ETT was significantly longer during treatment with ranolazine and atenolol therapy compared with placebo. The mean time to angina onset was longer with ranolazine than with atenolol, but the difference was not significant (mean difference 11.4 seconds, 95% CI –5.4 to +28.2; p = 0.18).

Time to 1 mm ST-segment depression during ETT was significantly longer during treatment with ranolazine and atenolol therapy than with placebo. The mean time to 1 mm ST-segment depression was longer with ranolazine than with atenolol, but the difference was not significant (mean difference 1.6 seconds, 95% CI –16.3 to +19.6; p = 0.86).

Total exercise duration during ETT was significantly longer during treatment with ranolazine and atenolol therapy compared with placebo, and during treatment with ranolazine compared with atenolol (mean difference 21.1 seconds, 95% CI 6.2 to 36.0; p = 0.006).

Adverse effects

A total of 29% of people experienced an adverse event during treatment with ranolazine, 25% during treatment with atenolol, and 17% during treatment with placebo.

The most frequent adverse effects were asthenia (12.3% during ranolazine treatment compared to 16.9% during atenolol treatment), dizziness (1.3% with ranolazine compared to 5.8% with atenolol), dyspepsia (4.5% with ranolazine compared to 0% with atenolol), headache, and nausea (each 3.9% with ranolazine compared to 0% with atenolol).

Ranolazine in combination with standard anti-anginal treatment

The Combination Assessment of Ranolazine in Stable Angina (CARISA) trial randomized 823 people to modified-release ranolazine 750 mg or 1000 mg twice a day or placebo for 12 weeks [Chaitman et al, 2004a]. At enrolment, 354 people (43%) were taking atenolol, 256 (31.1%) amlodipine, and 213 (25.9%) diltiazem.

Total exercise duration during ETT was significantly increased compared with placebo in both ranolazine groups. The increase from baseline with placebo was 91.7 (+/– 8.3) seconds; the increase above placebo was 23.7 (+/– 10.9) seconds with ranolazine 750 mg (p = 0.03) and 24.0 (+/– 11.0) seconds with ranolazine 1000 mg (p = 0.03).

Time to onset of angina during ETT was significantly increased compared with placebo in both ranolazine groups. The increase from baseline with placebo was 114.1 (+/– 9.2) seconds; the increase above placebo was 29.7 (+/– 12.1) seconds with ranolazine 750 mg (p = 0.01) and 26.0 (+/– 12.2) seconds with ranolazine 1000 mg (p = 0.03).

The frequency of angina attacks reduced from an average of 4.5 attacks per week at baseline to 3.3 attacks per week for placebo, 2.5 attacks per week for ranolazine 750 mg (p = 0.006), and 2.1 attacks per week for ranolazine 1000 mg (p < 0.001). There were similar reductions in the frequency of use of short-acting nitrates.

Adverse effects

Adverse effects were reported in 26.4% of people in the placebo group, 31.2% in the ranolazine 750 mg group, and 32.7% in the ranolazine 1000 mg group.

The most common dose-related adverse effects experienced with ranolazine were constipation, dizziness, nausea, and asthenia.

Five people in the ranolazine 1000 mg group reported syncope.

One RCT investigated the efficacy of ranolazine in people with persisting symptoms of angina despite maximum doses of amlodipine [Stone et al, 2006]. A total of 565 people were randomized to 1000 mg ranolazine twice a day or placebo for 6 weeks.

Treatment with ranolazine significantly reduced the number of angina attacks per week compared with placebo (2.88 versus 3.31, p = 0.028).

Treatment with ranolazine significantly reduced the weekly use of short-acting nitrate compared with placebo (2.03 versus 2.68 units per week, p = 0.014).

The scores on the angina frequency dimension of the Seattle Angina Questionnaire were significantly improved in people receiving ranolazine compared with placebo. None of the other dimensions of the questionnaire differed significantly between treatment groups.

Adverse effects

Adverse effects occurred in 39.9% of people treated with ranolazine and 35.3% treated with placebo.

The most frequently reported adverse effects were constipation, peripheral oedema, dizziness, nausea, and headache.

Seven people (three in the ranolazine group and four in the placebo group) withdrew from the study because of adverse effects.

Morbidity and mortality:

In the Metabolic Efficiency With Ranolazine for Less Ischemia in Non-ST Elevation Acute Coronary Syndromes–Thrombolysis in Myocardial Infarction (MERLIN-TIMI) 36 study, 6560 people with ischaemic symptoms in the previous 48 hours were randomized to ranolazine 1000 mg twice a day (with an initial intravenous dose) or placebo [Morrow et al, 2007]. Dose adjustments were made for people with renal insufficiency, QT interval prolongation, and other specific adverse effects. Participants continued to take other cardiovascular medication as considered appropriate. The primary outcome was the first occurrence of any element of the composite of cardiovascular death, MI, or recurrent ischaemia. The median duration of follow up was 348 days.

Treatment with ranolazine did not affect the primary outcome. Cardiovascular death, MI, or recurrent ischaemia occurred in 696 people (21.8%) in the ranolazine group compared with 753 people (23.5%) in the placebo group (hazard ratio [HR] 0.92, 95% CI 0.83 to 1.02; p = 0.11).

Recurrent ischaemia was reduced in the ranolazine group (13.9%) compared with the placebo group (16.1%) (HR 0.87, 95% CI 0.76 to 0.99; p = 0.03).

There was no evidence of a difference in total mortality between people in the ranolazine group (172) and those in the placebo group (175) (HR 0.99, 95% CI 0.80 to 1.22; p = 0.91).

Longer-term safety

A total of 746 people who completed the MARISA and CARISA studies were entered into an open-label study to assess the long-term safety of ranolazine [Koren et al, 2007]. Ranolazine was titrated to an optimal dose between 500 mg and 1000 mg twice a day, based on clinical response. Average follow up was 2.82 years (range 6 days to 6.5 years).

The majority of people (76.7%) completed at least 2 years of treatment. Adverse events were the most common reason for stopping treatment; 72 people withdrew because of adverse events in the first 2 years.

The most common adverse effects were dizziness (11.8%), constipation (10.9%), and peripheral oedema (8.3%).

There were no significant changes from baseline in either the PR or QRS interval. However, the QTc interval was prolonged by a mean of approximately 2.4 milliseconds compared with baseline (p < 0.001). Sixteen electrocardiogram readings showed prolongation of the QTc interval exceeding 500 milliseconds in 10 people (1.2%). Ten of these occurrences were recorded during treatment with the 1000 mg dose, and six were recorded during treatment with the 750 mg dose. There were no withdrawals from the study because of QTc interval prolongation.

Conclusions from the NICE Guideline Development Group (GDG)

The GDG concluded that there is insufficient evidence to recommend routine use of ranolazine, but that it may have a role in people with stable angina who are inadequately controlled or intolerant of first-line anti-anginal therapies.

There is no evidence of a difference in cardiac mortality or rate of non-fatal myocardial infarction (MI) between people treated with the combination of a beta-blocker and a calcium channel blocker (CCB) compared with a beta-blocker alone or a CCB alone.

There is some evidence from short-term randomized controlled trials (RCTs) that the combination of a beta-blocker and a CCB increases exercise time and time to 1mm ST-segment depression compared with a beta-blocker alone or a CCB alone. This beneficial effect of combination treatment was not matched by evidence of improved symptom control, as assessed by the frequency of episodes of angina and the use of glyceryl trinitrate (GTN).

Beta-blocker versus beta-blocker plus calcium-channel blocker

The NICE review included five RCTs comparing a beta-blocker with a beta-blocker plus a CCB. The review found that:

Evidence from short-term RCTs suggests that the combination of a beta-blocker and a CCB may improve exercise capacity compared with a beta-blocker alone, but there was no significant difference between the two groups in frequency of angina or use of GTN.

Evidence from one long-term RCT (mean follow-up 2 years) found no significant difference between the two groups in cardiac deaths or non-fatal MI.

Calcium-channel blocker versus beta-blocker plus calcium-channel blocker

The NICE review included four RCTs comparing a beta-blocker with a beta-blocker plus a CCB. The review found that:

Evidence from short-term RCTs suggests that the combination of a beta-blocker and a CCB may improve exercise capacity compared with a CCB alone, but there was no significant difference between the two groups in frequency of angina or use of GTN.

Evidence from one long-term RCT (mean follow-up 2 years) found no significant difference between the two groups in cardiac deaths or non-fatal MI.

Calcium-channel blocker plus standard anti-anginal treatment

The NICE review also included one long-term RCT comparing the effects of adding a CCB or placebo to usual anti-anginal treatment. A significant proportion of participants (about 80%) were also taking a beta-blocker. After 4.9 years follow up:

There was no significant difference between CCB and placebo groups for all-cause mortality, cardiovascular or unknown death, or MI.

This was true across the subgroups of age greater than 65 years, gender, and people with diabetes.

Significantly more people in the CCB groups withdrew from the trial because of adverse effects compared with the placebo group (relative risk 2.27; 95% CI 1.91 to 2.7).

Conclusions of the NICE Guideline Development Group (GDG)

The GDG concluded that evidence that combination therapy with a beta-blocker and a CCB is superior to treatment with a beta-blocker or a CCB alone is weak, and is largely confined to short-term improvements in exercise tolerance. The GDG considered that people who are not controlled on monotherapy (with a beta-blocker or a CCB) should be offered a change to the other drug class, or combination therapy with both drug classes. The GDG recommended that this should be decided on a case-by-case basis.

Evidence on combining three drugs for symptom relief in angina is very limited. Evidence from a small trial suggests that treatment with a calcium-channel blocker (CCB), beta-blocker, and long-acting nitrate is no more effective than treatment with a long-acting nitrate and either a CCB or beta-blocker.

In a randomized trial, people with a diagnosis of stable angina based on exercise testing were assigned to receive amlodipine (116 people), atenolol (116 people), or both (119 people) for 10 weeks. All groups also received a long-acting nitrate [Pehrsson et al, 2000].

Long-acting nitrate plus amlodipine or atenolol was as effective as the combination of all three drugs in terms of outcomes, including time to onset of 1 mm ST-segment depression, time to onset of angina symptoms, and total exercise time.

Of people who withdrew from the trial because of potential adverse effects, eight were in the amlodipine group, six in the atenolol group, and six in the amlodipine plus atenolol group.