Drug treatment

Three main groups of drugs are used in the symptomatic treatment of chronic stable angina: b blockers, nitrates, and calcium channel blockers. Aspirin is used as an antiplatelet drug for the secondary prophylaxis of vascular events in patients with angina. The range of agents and their dosages, contraindications, and side effects are described in the British National Formulary (BNF). All recommendations for treatment apply only in the absence of recognised contraindications, side effects, or interactions as documented in the BNF.

Recommendations

Drug treatment: secondary prophylactic treatment

The aim of secondary prophylactic treatment is to minimise the patient’s risk of subsequent vascular events. It is not clear what to use in those patients who cannot take low dose aspirin.

Recommendations

Statement: aspirin given to patients with stable angina lowers their risk of suffering a subsequent vascular event (I)—Eccles et al summarise the trials of aspirin in patients with stable angina.(79) From six trials the relative risk of myocardial infarction, stroke, or vascular death for patients with stable angina was 0.71 (0.58 to 0.87), and the overall difference in risk was 4.5% (1.9% to 7.1%). The difference in pooled incidence rate, by random effects model, estimated that this translates to a 0.7% reduction in the risk of myocardial infarction, stroke, or vascular death from antiplatelet treatment for one year (number needed to treat=150).

Research question

What is the cost effectiveness of other antiplatelet drugs in those patients who cannot take low dose aspirin?
 

Drug treatment: initial treatment of symptoms

The aim of such treatment is to use the least treatment necessary to minimise symptoms enough to allow patients to live as they want. We identified no trial of glyceryl trinitrate tablets; this is presumably because of the time frame of the literature search.

Recommendations

Statement: sublingual glyceryl trinitrate is effective in the prophylaxis of episodes of angina (I)

We found one supporting paper.(80)
 

Drug treatment: regular symptomatic treatment

Monotherapy: b blockers as first line treatment

The guideline search identified limited evidence on which to base recommendations about drug sequencing. Most of the recommendations are based on extrapolation from studies that were not primarily examining the issue of drug sequencing.

Recommendations

The evidence to support this recommendation is drawn from post-myocardial infarction trials and case-control studies of patients with hypertension and patients treated with b blockers for any reason. The underlying assumption is that patients with angina are at increased risk of myocardial infarction; if they have a myocardial infarction and are taking b blockers then their prognosis is improved.

Statement: patients taking b blockers for the treatment of hypertension are less likely to have a vascular event; patients taking b blockers who then have a myocardial infarction have a subsequently lower mortality (II)—Beevers et al studied patients in a hypertension clinic; those taking b blockers had fewer cardiovascular events than those taking other antihypertensive drugs (diuretics, methyldopa).(81) Nidorf et al looked at the effect of b blockers taken for any reason in patients who subsequently had a myocardial infarction.(82) The 28 day survival after myocardial infarction was 50% higher in patients treated with a b blocker.

Statement: patients who have had a myocardial infarction or currently have angina and are given b blockers have a subsequently lower rate of mortality and morbidity (I)—Yusuf et al carried out a meta-analysis (65 studies) of b blocker use during and after myocardial infarction.(83) They found a 25% lower mortality over one year in those receiving b blockers. In a meta-analysis of nine trials of b blocker use after myocardial infarction, in terms of mortality, benefit was greater in high risk patients (30% lower mortality) than in lower risk patients (20% lower mortality).(84) The ASIST trial compared atenolol with placebo in a randomised trial of 306 patients with documented coronary artery disease but only mild or no angina (and not previously using either b blockers or calcium channel blockers).(85) At one year, 25% of placebo patients compared with 11% of atenolol patients had experienced an adverse cardiac event related to ischaemia (death, resuscitation from ventricular tachycardia/fibrillation, myocardial infarction, admission to hospital for unstable angina, aggravation of angina requiring unblinded treatment, and need for revascularisation; P=0.001).

Statement: b blockers are as effective as other drug groups when used as monotherapy (I)—All the following studies showed the drugs to be effective as first line agents. However, differences in patient selection, study design, and drug dosages all prevent critical comparisons being made.

b blocker compared with dihydropyridines

Egstrup et al studied 42 patients taking metoprolol up to 100 mg twice daily, nifedipine up to 20 mg three times daily, and their combination in a randomised, double blind study.(86) They studied ischaemia measured on ambulatory monitoring. On this end point nifedipine was no better than placebo; metoprolol was significantly better than placebo; and the combination was no better than metoprolol alone.

The TIBET study randomised 682 patients to atenolol 50 mg twice daily, nifedipine slow release 20 mg twice daily, or their combination.(87)(88) Patients were aged 40-79 years (14% women) and had stable angina, were not being considered for surgical intervention, and had not experienced a myocardial infarction in the past three months. Patients were followed for a mean of two years. Significantly more of the nifedipine (40%) than the atenolol (27%) or combination (29%) groups withdrew from study medication. In the intention to treat analysis, there were no significant differences between the three study groups in terms of fatal and non-fatal cardiovascular events or in performance at exercise test.

The IMAGE study followed 280 patients with stable angina (and positive response to exercise test).(89)(90) Patients were randomised to metoprolol or nifedipine for six weeks and then each arm was further subdivided for an additional four weeks by adding either placebo or the alternative drug to the first drug randomised. At six weeks the metoprolol group had a better response at exercise test than the nifedipine group; at week 10 the groups with the second drug added improved further compared with the placebo groups. However, analysis at the individual patient level showed that most of the improvement was due to patients not responding to monotherapy. There was little evidence of an additive effect.

Rees-Jones and Oliver randomised 102 patients to atenolol 50 mg twice daily and nifedipine 20 mg twice daily or nifedipine 20 mg twice daily alone for three weeks each in a crossover trial.(91) Patients were adults aged under 80 years (39% women) already on either calcium antagonist or b blocker monotherapy, with an average of three or more angina episodes a week during the three months before entry into the study. Patients experienced fewer angina attacks per week on the combination (2.67) compared with the nifedipine only (2.33) treatment (P=0.0001).

The CASIS trial randomised 100 patients to receive amlodipine or atenolol.(92) In each arm, patients received the active drug or placebo in a crossover fashion (double blind). All patients were then given the combination treatment in a single blind (patient blind) manner. Amlodipine but not atenolol delayed onset of ischaemia (as measured by ST segment depression) during treadmill testing (in those patients experiencing angina during exercise testing, both drugs increased time to angina); atenolol but not amlodipine reduced frequency of ischaemia in ambulatory monitoring; in both settings, combination treatment was better than either drug alone.

We found six supporting papers.(93) (94) (95) (96) (97) (98)

b blocker compared with diltiazem or verapamil

The APSIS trial compared metoprolol with verapamil in 809 patients (mean age 59; 31% women) with clinically confirmed stable angina.(99) Patients were excluded if they had had a myocardial infarction in the past three years. Patients were followed for a median of 3.4 years. There were no significant differences between groups in terms of fatal or non-fatal cardiovascular events or in terms of withdrawals for side effects.

We found five supporting papers.(100) (101) (102) (103) (104)

Other comparisons

Stone et al compared the effect of titrated doses of propranolol LA, diltiazem SR, or nifedipine.(105) Anti-ischaemia efficacy was assessed by 48 hour ambulatory electrocardiographic monitoring, exercise treadmill tests, and anginal diaries. Propranolol improved ambulatory ischaemia; diltiazem prolonged exercise time (by 30 seconds in 600); propranolol and diltiazem both reduced angina; nifedipine had no effect.

We found two supporting papers.(106)(107)
 

Choosing a b blocker

Recommendation

· Patients should be treated with the cheapest preparation that they can tolerate and comply with and that controls their symptoms (D)

Statement: within the time frame of our search (1984-97) there was no compelling evidence to choose one b blocker over another (I)—Egstrup et al studied CR metoprolol 200 mg once daily versus standard metoprolol in the same dose.(86) There was a minor difference in adverse reactions but no difference in any outcome measures at 100 mg once daily; at 200 mg once daily there was a slight improvement in exercise tolerance.

Stopping b blockers

Recommendations

  • b blockers should not be stopped suddenly; patients should be warned not to stop taking b blockers suddenly nor to allow them to run out (B)
  • If b blockers need to be stopped, use should be tailed off over a period of four weeks (D)

Statement: acute b blocker withdrawal causes an increase in coronary events (II)—In a study of patients with hypertension, Psaty et al found that acute b blocker withdrawal caused a fourfold increase (1.2 to 16) in coronary events in the month after stopping.(108)

Monotherapy: verapamil for patients who cannot take b blockers

Recommendation

Statement: verapamil used after an infarct reduces the rate of major adverse events (I)—Verapamil is preferred to isosorbide dinitrate on the basis of a subgroup analysis of the DAVIT II trial,(109) which indicated a cardioprotective effect for patients experiencing a myocardial infarction and treated with verapamil that is not seen with isosorbide dinitrate. The two drugs seem similarly effective in the treatment of anginal symptoms. This is based on a subgroup analysis of only one trial and therefore the recommendation is at a lower level. The preference for verapamil, which is more expensive than isosorbide dinitrate, has implications for cost effectiveness. The DAVIT II trial studied the effect of verapamil 360 mg daily versus placebo started in the second week after myocardial infarction. There were fewer major events in the verapamil group. In patients who did not experience heart failure there was also a lower mortality.(109)

Statement: verapamil is as effective as other drug groups when used as monotherapy (I)—Cutler et al compared verapamil (three groups: 180, 360, or 540 mg once daily) with placebo in 278 patients (mean age 61; 11% women) with stable angina.(110) Patients were excluded if they had had a myocardial infarction in the past three months. They had to stop taking b blockers on entry into the study and were followed for four weeks. All verapamil groups had better performance on exercise test than the placebo group (with times to moderate angina increasing by a median of 2 minutes in the verapamil 360 mg and 540 mg groups, 1.5 mins in the 180 mg group, and 0.6 mins in the placebo group). The 360 mg and 540 mg groups had fewer angina attacks than placebo group. No treatment group had a significantly greater reduction in glyceryl trinitrate consumption than the placebo group. Incidence of constipation showed a dose related trend (P<0.01) with 21% of 540 mg patients (6% placebo) experiencing this side effect. The APSIS trial compared metoprolol with verapamil in 809 patients (mean age 59; 31% women) with clinically confirmed stable angina.(99) Patients were excluded if they had had a myocardial infarction in the past three years. Patients were followed for a median of 3.4 years. There were no significant differences between groups in terms of fatal or non-fatal cardiovascular events or in terms of withdrawals for side effects.

We found six supporting papers.(102)(103)(111) (112) (113) (114)
 

Monotherapy for patients who cannot take b blockers or verapamil

Recommendations

All the following studies show the drugs to be effective as first line agents. However, differences in patient selection, study design, and drug dosages all prevent critical comparisons being made.
 

Calcium channel blockers

Statement: calcium channel blockers when used alone are effective (I)—Klinke et al studied diltiazem SR in dosages of 120 mg twice daily and 180 mg twice daily versus placebo. There was no significant difference between the dosages.(115) Theroux et al studied diltiazem SR 180 mg twice daily versus placebo. There were reductions in ST segment depression on stress testing and on Holter monitoring and a 30 second (of 447) increase in exercise tolerance.(116) Thadani et al studied diltiazem SR 60-480 mg daily versus placebo in patients with angina. They found a dose related increase in exercise time, reaching 68 seconds in the highest dose group, and a dose related reduction in angina attacks.(117)

Weiss et al randomised 208 patients to either placebo or twice daily doses of 60, 90, 180, or 240 mg sustained release diltiazem.(118) Patients were aged 34 to 84 years, and 26% were women. All other anti-anginal medication except sublingual glyceryl trinitrate was stopped during the run-in period. Tests were carried out 3 and 12 hours after dose over a seven day period. Compared with baseline, the 90, 180, and 240 mg twice daily groups increased their exercise time significantly at three hours after dose (by a mean of 68, 97, and 82 seconds, respectively). The 180 and 240 mg twice daily groups also showed significant increases 12 hours after dose. Mehta and Lopez gave amlodipine monotherapy glyceryl trinitrate over an eight week single blind period to 226 patients who had not previously used anti-anginal medication except sublingual glyceryl trinitrate.(119) Those who responded to treatment (that is, showed ³ 7% improvement in exercise time) then underwent a randomised four week amlodipine versus placebo (amlodipine withdrawal) phase; 172 patients (mean age 65 years; 17% women) entered this phase. In the double blind phase, the amlodipine group showed a further 3% increase in time to onset of angina, whereas the placebo group showed a 10% decrease (between groups P=0.0001). There was no evidence of adverse effects associated with the abrupt withdrawal of amlodipine.

Frances et al compared once daily sustained release diltiazem 200 mg and 300 mg with placebo in 182 angina patients (30-70 years old; 17 women) in a double blind, parallel group, randomised controlled trial.(120) All anti-anginal medication except sublingual glyceryl trinitrate was tapered off during the washout period before placebo run-in. The active treatment period was one week with exercise tests before and after this. The 300 mg group showed significantly greater improvements in time to ischaemic and anginal threshold and total exercise time than the placebo group; the 200 mg group showed significant improvements in time to anginal threshold only compared with placebo. There were no significant differences between the two diltiazem groups.

Ezekowitz et al compared amlodipine with placebo in a parallel group, crossover trial in 103 patients with stable angina (age range 42-76; 14% women).(121) "Other anti-anginal medication" was discontinued, but it was not clear how many of participants this affected. Patients received either amlodipine (two weeks at 5 mg followed by two weeks at 10 mg once daily) or placebo for four weeks, followed by one week washout and then crossed over for a further four weeks. In period 1, exercise time increased (from 8 minutes) by 42 seconds in the amlodipine group compared with 0.6 seconds in the placebo group (P=0.004); time to onset of angina increased by 56 seconds in the amlodipine group compared with 25 seconds in the placebo group (not significant). The amlodipine group went from a median of three angina attacks a week to one; the placebo group went from three to 1.8 (P=0.016). In period 2, the group changing from amlodipine to placebo had a decrease in exercise time of 33 seconds; the group changing from placebo to amlodipine had a increase of 4.8 seconds. The number of angina attacks did not seem to increase in the group changing from amlodipine to placebo; the group changing from placebo to amlodipine seemed to show a small decrease (about 0.5 attacks a week); but statistics were not given and this difference may not be significant.

We found five supporting papers.(93)(122) (123) (124) (125)

Statement: calcium channel blockers all seem equally effective compared with placebo (I)—We found 10 supporting papers.(126) (127) (128) (129) (130) (131) (132) (133) (134) (135)

Nitrates

When discussing nitrate tolerance the BNF states "in the case of modified release tablets of isosorbide dinitrate (and conventional formulations of isosorbide mononitrate), the second of the two daily doses can be given after about eight hours rather than after 12 hours. Conventional formulations of isosorbide mononitrate should not usually be given more than twice daily unless small doses are used; modified-release formulations of isosorbide mononitrate should only be given once daily." "Some patients on long-acting or transdermal nitrates rapidly develop tolerance (with reduced therapeutic effects). Reduction of blood nitrate concentrations to low levels for four to eight hours each day usually maintains effectiveness in such patients. If tolerance is suspected after the use of transdermal patches they should be removed for several consecutive hours in each 24 hours."

Statement: oral nitrates are effective when used as a sustained release preparation, as an eccentrically dosed twice daily preparation, or as a three times daily preparation (I)—Parker studied asymmetric isosorbide mononitrate 20 mg twice daily (8 am, 3 pm) and found that it led to decreased angina attacks and increased exercise tolerance. There was no rebound or tolerance effect over three weeks.(136)Chrysant et al showed that isosorbide mononitrate SR (once daily dose) worked better than placebo.(137) It prolonged exercise tolerance compared with placebo with no tolerance at 4 or 12 hours after dose. There was no rebound in the morning before the dose. There was, however, little evidence of increased efficacy beyond 120 mg. Thadani et al looked at the effect of asymmetrical isosorbide mononitrate 20 mg twice daily (8 am, 3 pm) in a group of patients 55-60% of whom were on b blockers.(138) There was a 15-20% increase in exercise time and no rebound or early morning angina. They concluded that asymmetrical isosorbide mononitrate 20 mg twice daily was better than placebo in patients both on or off b blockers.

We found one supporting paper.(114)

Statement: nitrate patches are effective in the treatment of angina though dose and dosing interval is important (I)—In a meta-analysis, Colditz et al concluded that transdermal glyceryl trinitrate was effective at four hours with exercise tolerance increased by 76 seconds. There was no effect at 24 hours because of tolerance.(139) Fletcher et al looked at the effect of continuous 5 mg glyceryl trinitrate patches on quality of life using the sickness impact profile.(140) There was no effect on angina rate or use of glyceryl trinitrate. Quality of life was worse on glyceryl trinitrate and headache was more common (causing 5% of patients to be withdrawn). In a study of continuous use of cutaneous glyceryl trinitrate patches at any dose they were shown to be ineffective after four hours or 24 hours within one week.(141) Higher doses failed to overcome tolerance.

We found three supporting papers.(142) (143) (144)

Statement: intermittent treatment with high dose patches is more effective than continuous treatment with high dose patches (I)—We found one supporting paper.(145)

Statement: when patches are used with a patch-free interval they are effective, with high dose patches being more effective than lower dose patches (I)—DeMots and Glasser studied placebo and low dose (10 mg/20 mg) or high dose (30 mg/40 mg) patches in patients with stable exercise tests and an acute response to sublingual glyceryl trinitrate.(146) Patches were used 12 hours on, 12 hours off. Active treatment was better than placebo, producing about a one minute increase in exercise tolerance at 29 days. There was no effect on glyceryl trinitrate use or number of episodes of angina. High dose patches were better than low dose patches, which were better than placebo.

We found two supporting papers.(147)(148)

Other drugs

Despite its widespread use we identified only a small number of papers evaluating nicorandil, and those that we did identify were small studies and of limited use in informing clinical decisions.(149)(150)
 

Choosing a second drug

The recommendations for drugs added as second line agents to b blockers can be based on evidence. This is not the case for any of the other drug combinations. However, the evidence for this step conflicts with results of some studies that show an advantage from two drugs while others do not. Some patients probably will benefit and therefore the guideline development group thought that the recommendation to add a second drug was reasonable, although it could not be regarded as a grade A recommendation. A meta-analysis of the studies may resolve this question.

The use of isosorbide mononitrate as second line treatment to verapamil or dihydropyridines is based on extrapolation from effectiveness when added as second line agents to b blockers. The same rationale is true for the use of calcium channel blockers with nitrates.

Five papers showed the effectiveness of adding diltiazem to b blockers(104) (160)(163) (164) (165); if this combination is used the cautions in the BNF should be observed.

Recommendations

Calcium channel blockers as second line treatment

Statement: six studies showed that addition of dihydropyridines to b blockers produces an improvement in a range of parameters (I)—Uusitalo et al studied 62 patients taking metoprolol 100 mg twice daily, nifedipine 10 mg three times daily, and the combination in a randomised double blind crossover study.(151) Metoprolol was better than nifedipine; the combination was better again. DiBianco et al studied b blocker plus amlodipine versus b blocker alone.(152) With dose titration of amlodipine they found a dose response effect, an increase in exercise time (505 v 481 seconds), but no effect on angina attacks or glyceryl trinitrate use.

Foale compared atenolol 50 mg twice daily with atenolol 50 mg once daily plus nifedipine SR 20 mg twice daily.(153) The combination produced 120 seconds (over 320 seconds) extra before pain, and patients preferred the combination. There was no difference in terms of ST depression, exercise duration, adverse reactions, angina attacks, and glyceryl trinitrate use. Dunselman et al (FEMINA study) randomised patients with positive results on bicycle exercise tests, despite optimal b blockade, to either continuation of metoprolol, addition of felodipine to metoprolol, or replacement of metoprolol by felodipine for five weeks.(154) Changes in time to end of exercise and onset of chest pain did not vary across groups but patients taking additional felodipine had better time to 1 mm ST depression results than the metoprolol continuation group or the felodipine replacement group.

Ronnevik et al randomised 128 patients (10% women; age not given) with stable angina already being treated with b blockers to either felodipine or placebo once daily for 12 weeks.(155) The active treatment group was given a dose of 5 mg for the first four weeks; if this had not led to an increase in exercise time of 1 minute by the end of this period then the amount was increased to 10 mg. When they were tested at four weeks, the felodipine group had significantly greater increases than the placebo group in total exercise time and time to onset of angina four hours after dose. At 12 weeks, total exercise time but not time to onset of angina four hours after dose showed greater improvement from baseline in the felodipine than the placebo group. There were no significant differences between groups 24 hours after dose at four or 12 weeks. No changes in quality of life were observed for any group during the study.

The CASIS trial randomised 100 patients to receive amlodipine or atenolol.(92) In each arm, patients received the active drug or placebo in a crossover fashion (double blind). All patients were then given the combination treatment in a single blind (patient blind) manner. Amlodipine but not atenolol delayed onset of ischaemia (as measured by ST segment depression) during treadmill testing (in those patients experiencing angina during exercise testing, all drugs increased time to angina). Atenolol but not amlodipine reduced frequency of ischaemia in ambulatory monitoring. In both settings, combination treatment was better than either drug alone.

We found five supporting papers.(156) (157) (158) (159) (160)

Statement: three studies showed that addition of dihydropyridines to b blockers produces no additional effect across a range of parameters (I)—The TIBET study randomised 682 patients to atenolol 50 mg twice daily, nifedipine slow release 20 mg twice daily, or their combination. Patients were aged 40-79 years (14% women), had stable angina, were not being considered for surgical intervention, and had not had a myocardial infarction in the past three months.(87)(88) Patients were followed for a mean of two years. Significantly more of the nifedipine (40%) than the atenolol (27%) or combination (29%) groups withdrew from study medication. In the intention to treat analysis, there were no significant differences between the three study groups in terms of fatal and non-fatal cardiovascular events or in performance at exercise test.

The IMAGE study followed 280 patients with stable angina (and positive response to exercise test).(89)(90) Patients were randomised to metoprolol or nifedipine for six weeks and then each arm was further subdivided for an additional four weeks by adding either placebo or the alternative drug to the first drug randomised. At six weeks the metoprolol group had a better response at exercise test than the nifedipine group; at week 10 the groups with the additional drug improved further compared with the placebo groups; however analysis at the individual patient level showed that most of the improvement was due to patients not responding to monotherapy; there was little evidence of an additive effect.

Madjlessi-Simon et al randomised 84 patients with stable angina who were already being treated with b blockers to either amlodipine 10 mg or placebo once daily for 16 days.(161) Mean age of participants was 60, and 6% were women. There were no significant differences between groups in terms of number or duration of ischaemic episodes during 48 hour ambulatory ECG monitoring. In exercise testing there were no significant differences between groups.

We found one supporting paper.(162)

Statement: addition of diltiazem to b blockers produces a dose dependent improvement in symptom control and exercise tolerance (I)—Heller et al randomised 172 patients (mean age 63; 17% women) with stable angina to placebo or diltiazem 180 mg once daily for two weeks.(163) Patients had symptoms on either a b blocker or long acting nitrate, or both. Mean exercise time improved by 37 seconds in the diltiazem group and 21 seconds in the placebo group; the difference between groups was significant. Time to onset of angina and to moderate angina was also significantly longer in the diltiazem group. The diltiazem group showed a greater reduction in both total angina attacks and angina in relation to exercise.

We found four supporting papers.(104)(160)(164)(165)

Nitrates as second line treatment

Statement: the effectiveness of the addition of nitrates to b blockers or calcium channel blockers may be determined by the preparation used. Addition of isosorbide dinitrate to b blockers or calcium channel blockers produces no additional benefit. Nitrate patches added to b blockers produce no additional benefit. Addition of isosorbide mononitrate to b blockers is effective (I)—Uusitalo et al studied the effect of adding long acting isosorbide mononitrate in patients already taking b blockers. There was an increase in exercise duration (9 minutes to 9.5 minutes or 0.8 minutes median) and a 28% decrease in the use of glyceryl trinitrate tablets.(166) Thadani et al studied the asymmetrical use of isosorbide mononitrate 20 mg twice daily (8 am, 3 pm) in a group of patients, 55-60% of whom were taking b blockers. There was a 15-20% increase in exercise time and no rebound or early morning angina.(138)

Parker et al assessed the effects of three dose levels of transdermal glyceryl trinitrate patches applied for 12 hours daily for 30 days.(167) They randomised 291 patients to placebo, 0.2 mg/hour, 0.4 mg/hour, or 0.8 mg/hour. About 18% were women, and mean age was about 60 years (estimate from details of four groups). About 60% were using b blockers, which were continued; calcium channel blockers were discontinued on run-in. A significantly higher proportion of participants in the 0.8 mg/hour group discontinued the study (half for adverse effects related to glyceryl trinitrate). Patients in all treatment groups generally had better exercise times during treatment than at baseline. However, there were no significant differences between glyceryl trinitrate and placebo groups in frequency of angina attacks or use of sublingual glyceryl trinitrate.

Statement: nitrate related headache is less frequent with a low initial dose of nitrate compared with a high initial dose (I)—Cleophas et al randomised 89 patients with stable angina to either 60 mg isosorbide mononitrate once daily for two weeks or 30 mg isosorbide mononitrate for one week followed by 60 mg isosorbide mononitrate for one week in a crossover study.(168) After a two week washout period the alternative treatment was given; 91% of the patients were treated with b blockers or calcium channel blockers, mean age was 62, and 19% were women. Headaches were more common on high doses than low doses and during the first time period than the second (irrespective of dose)

We found six supporting papers.(107)(114)(169) (170) (171) (172)

Statement: in patients taking nitrates add any calcium channel blocker (I)—Cutler et al randomised 189 patients (mean age 62; 19% women) with stable angina to placebo or diltiazem at doses of 120, 240, or 480 mg once daily for two weeks.(173) All anti-anginal medication except long acting nitrates was discontinued on run-in. Total exercise time 24 hours after dose was significantly greater in the 240 mg and 480 mg diltiazem groups than the placebo group (change from baseline 49 and 56 seconds, respectively); total exercise time eight hours after dose improved by a similar amount in these two groups. All three treatment groups experienced a reduction in angina attacks; this was not reflected in use of glyceryl trinitrate.

Research question

What is the most appropriate choice of second line drug?

Choosing a third drug

Recommendations

There are a limited number of studies evaluating the effect of adding a third drug and the effectiveness of triple therapy cannot therefore be evaluated. Some studies identified have added calcium channel blockers to b blockers and nitrates. In two of the studies, the nitrate used was isosorbide dinitrate, and the evidence from dual therapy studies suggests that it is of no additional benefit when added to a b blocker. Read in that light these papers confirm the benefit of adding a calcium channel blocker to a b blocker. The third paper does not state which nitrate was used, but in this study the addition of the calcium channel blocker made no difference.

Statement: the effectiveness of adding a third anti-anginal drug is not clear (I)—We found one supporting paper.(174)

Non-drug treatment

Recommendation

Statement: in one randomised controlled trial supplemental vitamin E reduced rates of non-fatal myocardial infarction (I)—Stephens et al compared vitamin E with placebo in 2002 patients with angiographically proved coronary atherosclerosis (90% with angina, evidence of reversible cardiac ischaemia, or both).(175) The study was a parallel group trial with randomisation stratified by sex, blood pressure, age, BMI, total cholesterol concentration, smoking habits, and planned treatment. Patients in the active treatment group initially received 800 IU daily; it was subsequently found that lower doses would exceed physiological values, so patients recruited from that point on were allocated 400 IU daily. Patients were followed for a median of 510 days. The active treatment group had a significantly lower rate of non-fatal myocardial infarction (relative risk 0.23, 0.11 to 0.47), but there was no difference in cardiovascular deaths. There was also no difference in side effects.

Statement: there is no evidence that patients with angina benefit from acupuncture—Ballegaard et al randomised 49 patients with stable angina to either genuine or sham acupuncture.(176) Patients received 10 treatments over a three week period. There were no differences between groups in exercise test results, changes in rates of angina attacks, or consumption of glyceryl trinitrate. Both groups reported significant reductions in angina attack rates and use of glyceryl trinitrate.

Research question

Is there a role for complementary therapies in the treatment of stable angina?

Review of patients with stable angina

Recommendation

Patients with stable angina should be reviewed periodically; the guideline development group could find no basis for recommending one interval over another and thought that at least annually was appropriate.
 

Referral of patients

Referral for exercise testing

Referral for exercise testing will be influenced by whether or not open access facilities are available locally. If they are not then referral for exercise testing will have to be via a cardiologist (see below).

Referral to a cardiologist

Recommendations

Referral from a general practitioner to a cardiologist will be for one of three reasons: to identify whether they fall into a group that would benefit from prognostic investigation and treatment; to establish a diagnosis; or for management advice. Identification of whether patients fall into a group that would benefit from prognostic investigation and treatment will apply in patients with a clear history of angina; they will usually be on medical treatment. The purpose of referral is to confirm the presence and severity of disease by further investigation with a view to identifying those suitable for revascularisation. Patients in whom the diagnosis is unclear should be referred to a cardiologist, not for an open access exercise test. While referral of these patients may lead to coronary angiography, the purpose is to clarify the diagnosis, potentially excluding significant cardiac disease.

Referral for management advice will apply to those patients in whom the diagnosis is certain but in whom symptom control is suboptimal. Patients whose condition is not controlled on maximal medical treatment should be referred for consideration for angiography not for open access exercise testing.

What influences the decision to refer?

In all patients considered for referral the decision will be influenced by:

These factors represent a spectrum for most patients, and their effect on the decision to refer will be additive. The referral decision cannot be taken in isolation and needs to be set in the current context of the patient. Not all patients need to be referred. Those who will benefit are:

Research questions

There is little analysis on cost effectiveness data available for any of the strategies recommended within the guideline; such analysis would be valuable.