American College of Cardiology

BRAUNWALD ET AL., MANAGEMENT OF PATIENTS WITH UNSTABLE ANGINA AND NON-ST-SEGMENT ELEVATION MYOCARDIAL INFARCTION UPDATE
http://www.acc.org/clinical/guidelines/unstable/update_index.htm

ACC/AHA 2002 Guideline Update for the Management of Patients With Unstable Angina and Non-ST-Segment Elevation Myocardial Infarction

A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on the Management of Patients With Unstable Angina)

This is a Guideline Update of the 2000 Unstable Angina Guidelines. To highlight the changes, deleted text is indicated by strikeout, and revised text is presented in brown. A clean version of the document, with changes fully incorporated, is available for download and print.

VI. Special Groups

A. Women

Recommendation

Class I

Women with UA/NSTEMI should be managed in a manner similar to men. Specifically, women, like men with UA/NSTEMI, should receive ASA and clopidogrel. Indications for noninvasive and invasive testing are similar in women and men. (Level of Evidence: B)

Although at any age women have a lower incidence of CAD than men, they account for a considerable proportion of UA/NSTEMI patients, and UA/NSTEMI is a serious and common condition among women. It is important to overcome long-held notions that severe coronary manifestations are uncommon in this population; however, women may manifest CAD somewhat differently than men. Women who present with chest discomfort are more likely than men to have noncardiac causes, as well as nonatherosclerotic cardiac causes, such as coronary vasospasm ^(354-356). Women with CAD are, on average, older than men and are more likely to have comorbidities such as hypertension, diabetes, and CHF ^(32,357-359); to manifest angina rather than AMI; and, among angina and MI patients, to have atypical symptoms.

1. Stress Testing
In general, ECG exercise testing is less predictive in women than in men ^{(261,360-362)}, primarily because of the lower pretest probability of CAD. Breast attenuation may be a problem with thallium-201 stress testing but not with dobutamine echocardiography. Stress echocardiography (dobutamine or exercise) is therefore an accurate and cost-effective technique for CAD detection in women ⁽²⁶¹⁾. Recommendations for noninvasive testing in women are the same as in men (see Section III) ^(26,363). A report of 976 women who underwent treadmill exercise suggests that the Duke Treadmill Score (DTS) provides accurate diagnostic and prognostic estimates in women as well as in men ⁽³⁶⁴⁾. The DTS actually performed better for women than for men in the exclusion of CAD. There were fewer low-risk women than men with any significant CAD (greater than 1 vessel with greater than 75% stenosis: 20% in women vs. 47% in men, p less than 0.001).

Regarding dobutamine stress echocardiography, pilot phase data from the Women's Ischemia Syndrome Evaluation (WISE) study ⁽³⁶⁵⁾ indicated that in women, the test reliably detects multivessel disease (sensitivity 81.8%, similar to that in men) but not 1-vessel disease. Several studies have indicated that women with positive stress tests tend not to be evaluated as aggressively as men ⁽³⁶⁶⁾.

2. Management
In studies that span the spectrum of CAD, women tend to receive less intensive pharmacological treatment than men ⁽³⁵⁷⁾, which is perhaps related in part to a less serious view of the impact of CAD in women. Although the specifics vary regarding beta-blockers and other drugs ^(32,357,366), a consistent (and disturbing) pattern is that women are prescribed ASA as well as other antithrombotic agents less frequently than men ⁽³²⁾.

Although it has been widely believed that women fare worse with PCI and CABG than do men because of technical factors (e.g., smaller artery size, greater age, and more comorbidities) ^(367-371), recent studies cast doubt on this ^(32,358). In the case of PCI, it has been suggested that angiographic success and late outcomes are similar in women and men, although in some series, early complications occur more frequently in women ^{(367,368,372-375)}. However, the outlook for women undergoing PTCA appears to have improved as evidenced by the NHLBI registry ⁽³⁷⁶⁾. Earlier studies of women undergoing CABG showed that women were less likely to receive internal mammary arteries or complete revascularization and had a higher mortality rate (RRs 1.4 to 4.4) than men ^{(367,368,372-380)}. However, more recent studies in CABG patients with ACS show a more favorable outlook for women (see later) ^(32,358) than previously thought.

3. Data on UA/NSTEMI
Considerable clinical information about UA/NSTEMI in women has emerged from the TIMI III trial ⁽³²⁾ (which examined the use of tissue plasminogen activator and invasive strategies in ACS) and the TIMI III registry ⁽³⁵⁷⁾. There were 497 women in the former population and 1,640 in the latter. As in other forms of CAD, women were older and had more comorbidities (diabetes and hypertension), as well as stronger family histories ^(32,357-359). Women were less likely to have had a previous MI or cardiac procedures (Figure 15) ⁽³⁵⁷⁾ and had less LV dysfunction. However, they presented with symptoms of similar frequency, duration, pattern, and ST-segment changes to those of the men. As in other studies, medication use, most particularly ASA, was less in women than in men in the week before the event, during hospitalization, and at discharge. However, no differences between men and women were evident in the results of medical therapy. In the registry, women underwent exercise testing in a similar proportion to that of men. The frequencies of stress test positivity were also similar, although women were less likely to have a high-risk stress test result. However, women were less likely to undergo angiography (RR 0.71, p less than 0.001), perhaps related to the lower percentage with high- risk test results on noninvasive testing ⁽³⁵⁷⁾.

Coronary angiograms in both the TIMI III trial and registry, as well as in other studies ^(285,381), revealed less extensive CAD in women, of whom a higher proportion had no CAD. In the registry, women were also less likely than men to undergo revascularization (RR 0.66, p less than 0.001) ⁽³⁵⁷⁾; in the TIMI III trial (in which angiography was mandated), there was no gender difference in the percent of patients undergoing PTCA, but less CABG was performed in women, presumably because of a lower incidence of multivessel disease. Importantly, gender was not an independent predictor of the outcome of revascularization. Thus, a key observation in the TIMI III trial and registry ^(32,357) was that gender was not an independent prognostic factor, with outcomes of death, MI, and recurrent ischemia similar in women and men.

Two additional studies were consistent with the TIMI data on interventions in ACS. A Mayo Clinic review of 3,014 patients (941 women) with UA who underwent PCI reported that women had similar early and late results to men ⁽³⁵⁸⁾. The BARI trial of 1,829 patients compared PTCA and CABG, primarily in patients with UA, and showed that the results of revascularization were, if anything, better in women than men, when corrected for other factors. At an average 5.4-year follow-up, mortality rates for men and women were 12% and 13%, respectively, but when adjusted for baseline differences (e.g., age, diabetes, and other comorbidities), there was a lower risk of death (RR 0.60, p = 0.003) but a similar risk of death or MI (RR 0.84, p = NS) in women compared with men ⁽³⁸²⁾.

In a more recent review of patients with ACS from GUSTO-IIb, an extensive prospective study of anticoagulation in 12,142 patients (3,662 women and 8,480 men) with ACS, the differences in profile between men and women were similar to those previously reported ⁽³¹⁾. As in other studies, women were more likely to have UA than MI (adjusted OR 1.51, 95% CI 1.34 to 1.69, p less than 0.001) and were older and had a higher incidence of CHF (10.2% vs. 6.1%, p less than 0.001) and a different risk factor profile (increased hypertension, diabetes, and cholesterol; less smoking, previous MI, and coronary surgery and procedures). On coronary angiography, women with UA also had fewer diseased arteries than did men, and more had no significant coronary stenosis (30.5% vs. 13.9%, p less than 0.001). The 30-day event rate (death/MI) was significantly lower in women than in men with UA (events OR 0.65, p = 0.003).

The use of HRT in postmenopausal women is discussed in Section V.A.

4. Conclusions
Women with ACS are older and more frequently have comorbidities than men but have more atypical presentations and appear to have less severe and extensive obstructive CAD. Women receive ASA less frequently than do men, but patients with UA/NSTEMI of either sex should receive this agent. Women undergo angiography less frequently, and they have similar use of exercise testing and the same prognostic factors on exercise tests as men. Outcomes of revascularization are similar in women and men, whereas overall outcomes in UA may be similar to that in men or more favorable in women.

B. Diabetes Mellitus

Recommendations

Class I

1. Diabetes is an independent risk factor in patients with UA/NSTEMI. (Level of Evidence: A)
2. Medical treatment in the acute phase and decisions on whether to perform stress testing and angiography and revascularization should be similar in diabetic and nondiabetic patients. (Level of Evidence: C)
3. Attention should be directed toward tight glucose control. (Level of Evidence: B)
4. For patients with multivessel disease, CABG with use of the internal mammary arteries is preferred over PCI in patients being treated for diabetes. (Level of Evidence: B)

Class IIa

1. PCI for diabetic patients with 1-vessel disease and inducible ischemia. (Level of Evidence: B)
2. Abciximab for diabetics treated with coronary stenting. (Level of Evidence: B)

CAD accounts for 75% of all deaths in diabetics ^(32-34), and approximately 20% to 25% of all patients with UA/NSTEMI are diabetic ^{(197,324,357,383-385)}. Among patients with UA/NSTEMI, diabetics have more severe CAD ^{(383,386,387)}, and diabetes is an important independent predictor for adverse outcomes (death, MI, or readmission with UA at 1 year) (RR 4.9) ^(388-391). Also, many diabetics who present with UA/NSTEMI are post-CABG ⁽³⁹²⁾.

Diabetics tend to have more extensive noncoronary vascular comorbidities, hypertension, LV hypertrophy, cardiomyopathy, and CHF. In addition, autonomic dysfunction, which occurs in approximately one third of diabetics, influences heart rate and blood pressure, raises the threshold for the perception of angina, and may be accompanied by LV dysfunction ^(393-395). On coronary angioscopy, diabetic patients with UA have a greater proportion of ulcerated plaques (94% vs. 60%, p = 0.01) and intracoronary thrombi (94% vs. 55%, p = 0.004) than nondiabetics. These findings suggest a higher risk of instability ⁽³⁹⁶⁾.

Although beta-blockers may mask the symptoms of hypoglycemia or lead to it by blunting the hyperglycemic response, they should nevertheless be used with appropriate caution in diabetics with ACS. Diuretics that cause hypokalemia may inhibit insulin release and thereby worsen glucose intolerance.

1. Coronary Revascularization
Approximately 20% of all patients who undergo CABG (397) and PCI ^{(368,369,372,373,386,387)} have diabetes. Data regarding outcomes are complex. In the Coronary Artery Surgery Study (CASS) of CABG, diabetics had a 57% higher mortality rate than nondiabetics. A striking advantage for CABG over PTCA was found in treated diabetics in the BARI trial ⁽³⁸³⁾, a randomized trial of PTCA vs. CABG in 1829 stable patients with multivessel disease, of whom 19% were diabetics (see Section IV). Diabetics, as in other studies, had increased comorbidity rates. Five years after randomization, patients who required treatment for diabetes had a lower survival rate than nondiabetics (73.1% vs. 91.3%, p less than 0.0001), whereas survival rates in nondiabetics and diabetics who did not require hypoglycemic treatment were similar (93.3% vs. 91.1%, p = NS). Outcomes for CABG in treated diabetics were far better than those for PTCA (80.6% vs. 65.5% survival, p = 0.0003). An interesting finding was that the mortality rate during the 5.4 years of the study in diabetics who received SVGs (18.2%) was similar to that of patients who underwent PTCA (20.6%), whereas the mortality rate in patients who received internal mammary arteries was much lower (2.9%). Results of the Emory Angioplasty versus Surgery Trial (EAST) at 8 years showed a similar trend but were less conclusive ⁽³⁹⁸⁾. The increased mortality rate noted in randomized trials in PTCA-treated diabetics has been confirmed in a registry study from Emory University ⁽³²⁷⁾. Uncorrected, there was little difference in long-term mortality rates. The CABG patients had more severe disease, and with correction for baseline differences, there was an improved survival rate in insulin-requiring patients with multivessel disease who were revascularized with CABG rather than with PTCA. That the more severely diseased patients, in a nonrandomized registry, were selectively sent more often for CABG than for PTCA probably represents good clinical decision making.

A 9-year follow-up of the NHLBI registry showed a similar disturbing pattern for diabetics undergoing PTCA. Immediate angiographic success and completeness of revascularization were similar, but compared with nondiabetics, diabetics (who, again, had more severe CAD and comorbidities) had increased rates of hospital mortality (3.2% vs. 0.5%), nonfatal MI (7.0% vs. 4.1%), death and MI (10.0% vs. 4.5%), and the combined end point of death, MI, and CABG (11% vs. 6.7%, p less than 0.01 for all). At 9 years, rates of mortality (35.9% vs. 17.9%), MI (29% vs. 18.5%), repeat PTCA (43.0% vs. 36.5%), and CABG (37.6% vs. 27.4%) were all higher in diabetics than in nondiabetics ⁽³⁸⁶⁾.

However, as pointed out in Section IV, other data point to less of a differential effect of PCI in diabetics. For example, data from the BARI registry varied from those of the trial. In the registry, there was no significant difference in cardiac survival for diabetics undergoing PTCA (92.5%) and CABG (94%) (NS) ^(328,399). In the Duke University registry, patients with diabetes and PTCA or CABG were matched with the BARI population. The outcome in diabetics was worse than that in nondiabetics with either CABG or PTCA, but there was no differential effect. The 5-year survival rate for PTCA and CABG adjusted for baseline characteristics was 86% and 89% in diabetics and 92% and 93% in nondiabetics, respectively ⁽⁴⁰⁰⁾.

Stents may offer diabetics a much improved outcome for PCI. In a recent study with historical controls, the outcome after coronary stenting was superior to that after PTCA in diabetics, and the restenosis rate after stenting was reduced (63% vs. 36%, diabetics vs. nondiabetics with balloon PTCA at 6 months, p = 0.0002) compared with 25% and 27% with stents (p = NS) ⁽³⁹⁸⁾. On the other hand, diabetics who underwent atherectomy had a substantial restenosis rate (60% over 6 months) ⁽⁴⁰¹⁾.

Finally, 3 recent trials have shown that abciximab considerably improved the outcome of PCI in diabetics. In the EPILOG trial, abciximab resulted in a greater decline in death/MI over 6 months after PTCA in diabetics (hazard ratio 0.36, 95% CI 0.21 to 0.61) than in nondiabetics (0.60, 95% CI 0.44 to 0.829) ⁽⁴⁰²⁾. Similar results have been reported for tirofiban in the PRISM-PLUS trial ⁽³¹⁴⁾. EPISTENT was a randomized trial that compared stent plus placebo with stent plus abciximab and balloon plus abciximab in 2399 patients, of whom 20.5% had diabetes and 20.3% had UA ⁽²⁴⁶⁾. The 30-day event rate (death, MI, urgent revascularization) in diabetics declined from 12.1% (stent plus placebo) to 5.6% (stent plus abciximab) (p = 0.040). At 6 months, the drug reduced revascularization of target arteries in diabetics (16.6% vs. 8.1%, p = 0.02). Death or MI was reduced to a similar degree in diabetics as that of nondiabetics ⁽³¹³⁾. These benefits were maintained at 1 year ⁽⁴⁰³⁾. Thus, in the 6-month data, the drug, as well as stents, considerably improved the safety of PCI in diabetics.

2. Conclusions
Diabetes occurs in about one fifth of patients with UA/NSTEMI and is an independent predictor of adverse outcomes. It is associated with more extensive CAD, unstable lesions, frequent comorbidities, and less favorable long-term outcomes with coronary revascularization, especially with PTCA. It is unclear whether these differences are due to more frequent restenosis and/or severe progression of the underlying disease ⁽³⁸⁶⁾. The use of stents, particularly with abciximab, appears to provide more favorable results in diabetics, although more data are needed. Clinical outcomes with CABG, especially with the use of 1 or both internal mammary arteries, are better than those with PTCA but are still less favorable than in nondiabetics.

C. Post-CABG Patients

Recommendations

Class I

1. Medical treatment for post-CABG patients should follow the same guidelines as for non-post-CABG patients with UA/NSTEMI. (Level of Evidence: C)
2. Because of the many anatomic possibilities that might be responsible for recurrent ischemia, there should be a low threshold for angiography in post-CABG patients with UA/NSTEMI. (Level of Evidence: B)

Class IIa

1. Repeat CABG is recommended for multiple SVG stenoses, especially when there is significant stenosis of a graft that supplies the LAD. PCI is recommended for focal saphenous vein stenosis. (Level of Evidence: C)
2. Stress testing should generally involve imaging in post-CABG patients. (Level of Evidence: C)

Overall, up to 20% of patients presenting with UA/NSTEMI have previously undergone CABG ⁽³⁹²⁾. Conversely, approximately 20% of post-CABG patients develop UA/NSTEMI during an interval of 7.5 years ⁽⁴⁰⁴⁾, with a highly variable postoperative time of occurrence ⁽⁴⁰⁵⁾. Post-CABG patients who present with UA/NSTEMI are at higher risk, with more extensive CAD and LV dysfunction than those patients who have not previously undergone surgery.

1. Pathological Findings
Pathologically, intimal hyperplasia or atherosclerosis may develop in SVGs, and there is a particular tendency for thrombotic lesions to develop in these vessels (in 72% of grafts resected in 1 study) ^(406-409). In addition, post-CABG patients may develop atherosclerosis in their native vessels and this may also lead to UA/ NSTEMI ^(409,410). However, obstructive lesions are more likely to occur in SVGs (53% within 5 years, 76% at 5 to 10 years, 92% at greater than 10 years) ⁽⁴¹¹⁾. Spasm in grafts or native vessels ^(412,413) and technical complications may also play a role in the development of UA/NSTEMI during the early postoperative period ^(404,414). Both angioscopic and angiographic findings indicate that SVG disease is a serious and unstable process. Angioscopically, friable plaques occur uniquely in SVGs (44% vs. 0% in native coronary arteries), whereas rough and white plaques occur in both SVGs and native coronary arteries ⁽⁴¹⁵⁾. Angiographically, the SVGs more frequently have complex lesions (i.e., overhanging edges, irregular borders, ulcerations, or thrombosis), thrombi (37% vs. 12%, p = 0.04), and total occlusions (49% vs. 24%, p = 0.02) ⁽⁴¹¹⁾.

2. Clinical Findings and Approach
Compared with UA/NSTEMI patients without prior CABG, post-CABG patients are more often male (presumably because more men than women have undergone CABG), older, and diabetic. They have more extensive native vessel CAD and more previous MIs and LV dysfunction. Symptomatically, these patients have more prolonged chest pain than ACS patients without prior CABG. More than 30% of post-CABG patients have resting ECG abnormalities, and ECG stress tests are therefore less conclusive ⁽⁴¹⁶⁾. However, a test that becomes positive after having been negative is helpful in the diagnosis of ischemia. Myocardial stress perfusion imaging and dobutamine echocardiography are often helpful diagnostically ⁽⁴¹⁷⁾.

The outcomes of UA/NSTEMI in post-CABG patients are less favorable than those in patients who have not undergone CABG. There is a high rate of embolization of atherosclerotic material from friable grafts at the time of intervention, making these procedures more difficult and associated with higher rates of complications ⁽⁴¹⁸⁾. In 1 matched-control study of UA, the initial course was similar, but post-CABG patients had twice the incidence of adverse events (death, MI, recurrent UA) during the first year. This was attributed to a lower rate of complete revascularization, which was possible in only 9 of 42 post-CABG patients compared with 39 of 52 patients who had not previously undergone CABG (p = 0.001) ⁽⁴⁰⁵⁾. Results were directionally similar in the TIMI III registry of ACS, in which 16% of patients were post-CABG. Here again, early outcomes in post-CABG patients and others were equivalent, but at 1 year, 39.3% vs. 30.2% experienced adverse events (death, MI, recurrent ischemia) (p = 0.002) ⁽⁴¹⁹⁾.

Revascularization with either PCI or reoperation is often indicated and possible in post-CABG patients with UA/NSTEMI. In a randomized control trial that compared stents with PTCA of obstructed SVGs, there was no statistically significant difference in restenosis during 6 months, although a trend favored stents: 34% vs. 46%. Although hemorrhagic complications were higher in the stent group, clinical outcomes (freedom from MI or repeat revascularization) were better (73% vs. 58%, p = 0.03) ⁽⁴²⁰⁾. Reoperation of patients with stenotic SVGs has been successful in reducing symptoms of recurrent ischemia, and it appears to improve survival rates in patients greater than 5 years after surgery, especially with disease in the LAD, for which survival rates were 74% vs. 53% (p = 0.004) (reoperated vs. nonreoperated post-CABG patients) ^(421,422).

3. Conclusions
Post-CABG patients, especially those with only SVGs, are at high risk of UA/NSTEMI. There is a higher likelihood of disease in SVGs than in native arteries and this difference increases with postoperative time. Pathologically and angiographically, disease in SVGs has characteristics associated with instability. There also are difficulties with treadmill ECG testing and less favorable outcomes with repeat revascularization than in patients who have not undergone previous CABG.

D. Elderly Patients

Recommendations

Class I

1. Decisions on management should reflect considerations of general health, comorbidities, cognitive status, and life expectancy. (Level of Evidence: C)
2. Attention should be paid to altered pharmacokinetics and sensitivity to hypotensive drugs. (Level of Evidence: B)
3. Intensive medical and interventional management of ACS may be undertaken but with close observation for adverse effects of these therapies. (Level of Evidence: B)

In this discussion, patients greater than 75 years are considered to be "elderly," although a number of studies have used other cutoff ages, such as 65 or 70 years. Elderly persons constitute about one tenth of ACS patients ⁽³⁵⁷⁾ and present with a number of special problems. They are more likely to have cardiac and noncardiac comorbidities; these include a diminished beta-sympathetic response, increased cardiac afterload due to decreased arterial compliance and arterial hypertension, cardiac hypertrophy, and ventricular dysfunction, especially diastolic dysfunction ⁽⁴²³⁾.

CAD is more common and more severe in elderly persons, who, when they develop UA/NSTEMI, are also more likely to present with atypical symptoms, including dyspnea and confusion, rather than with the ischemic chest pain that is typical for younger patients. There also is a higher incidence of unrecognized prior MI than in younger patients ⁽⁴²⁴⁾. Conversely, comorbid conditions such as hiatus hernia are also more frequent and may be associated with chest pain at rest and may mimic UA. The greater likelihood of comorbidity in elderly persons (e.g., COPD, renal failure, and cerebral disease) also increases the morbidity and mortality rates for cardiac events and interventions in this population.

It may be difficult for elderly persons to perform exercise tests because of muscle weakness and orthopedic problems. In such patients, a pharmacological stress test may be performed (see Section III). The higher prevalence of preexisting resting ECG abnormalities ⁽³⁸⁹⁾, arrhythmias ^(425,426), and cardiac hypertrophy complicates the interpretation of stress ECG and may require the use of imaging.

1. Pharmacological Management
Reductions in cardiac output and in renal and hepatic perfusion and function reduce the elimination of drugs in elderly persons. Drugs such as propranolol that undergo first-pass hepatic metabolism exhibit increased bioavailability ⁽⁴²⁷⁾. Pharmacodynamic responses to drugs are influenced by the lower cardiac output, plasma volume, and vasomotor tone and the blunted baroreceptor and beta-adrenergic responses.

Elderly persons are particularly vulnerable to drugs with hypotensive actions (e.g., nitrates and calcium antagonists) and cerebral effects (e.g., beta-blockers). Responses to beta-blockers are influenced by 2 competing factors. There may be a blunted response to beta-blockers because of decreased adrenergic activity in elderly persons. On the other hand, baseline sympathetic tone may be decreased. Thus, the magnitude of response to beta-blockers is not entirely predictable. Clearance of warfarin may be reduced, and sensitivity to it may be increased with age ⁽⁴²⁸⁾; heparin dosage requirements also appear to be reduced ⁽²¹⁰⁾. Overall, however, it should be emphasized that all of the drugs commonly used in the management of younger patients with UA/NSTEMI are useful in elderly patients, provided these differences are recognized and proper precautions are taken (i.e., beginning with lower doses than in younger patients and, in particular, careful observation for toxicity).

2. Observations in UA/NSTEMI
The TIMI III registry ⁽³⁵⁷⁾ provided data on elderly patients (greater than 75 years old), who (by design) composed 25% of the 3,318 patients. This group had fewer atherosclerotic risk factors (smoking, hypercholesterolemia, family history), more previous angina, and fewer previous procedures (Figure 16) ⁽³⁵⁷⁾, and in other studies, they had more CHF ^(429,430). They were less likely to receive beta-blockers and heparin in the hospital and far less likely to undergo angiography (RR 0.65, p less than 0.001 at 6 weeks) and coronary revascularization (RR 0.79, p = 0.002 at 6 weeks) than younger patients, although when this procedure was carried out, they were found to have more extensive disease. The 6-week mortality (RR 3.76, p less than 0.001) and MI (RR 2.05, p less than 0.001) rates were elevated. Overall, elderly patients were treated less aggressively with both medical therapy and procedures than were their younger counterparts, despite a higher-risk profile.

3. Interventions and Surgery
A high prevalence of cerebral and peripheral vascular comorbidity influences the results of coronary revascularization in elderly persons. However, results of revascularization in elderly persons are improving. A Medicare review of both PCI and CABG (225,915 PCI and 357,885 CABG patients greater than 65 years old) between 1987 and 1990 revealed that revascularization is commonly carried out in patients in this age group and that outcomes have improved compared with earlier periods. The 30-day and 1-year mortality rates during this time period were 3.3% and 8.0% for PCI and 5.8% and 11.0% for CABG, respectively, with lower mortality rates for patients who received internal mammary artery implants. Estimated 30-day and 1-year mortality rates for PCI rose from 2.1% and 5.2% in patients 65 to 69 years old to 7.8% and 17.3% in patients greater than 80 years old, and respective rates for CABG rose from 4.3% and 8.0% to 10.6% and 19.5%. As expected, comorbidities were associated with increased mortality rates ⁽⁴³¹⁾.

A smaller but more closely observed matched comparison of CABG vs. PCI in patients greater than 70 years old (a majority of whom had UA) in which the CABG group had more extensive CAD reported that rates for in-hospital mortality (9% vs. 2%), cerebrovascular accidents (5% vs. 0%), and STEMI (6% vs. 1%) were all significantly higher with CABG ⁽⁴³²⁾. However, there was more relief of angina with surgery, and 5-year survival rates were similar between the 2 groups (65% CABG vs. 63% PCI) (p = NS).

Some studies of PCI in patients aged 65 to 75 years have shown that success rates with experienced medical professionals are similar to those in younger patients ^{(429,433-435)}, but with even older patients, success rates decline and complications rates rise. In a recent VA study, in patients greater than 70 years old, the angiographic success rate was 86%, the clinical success rate was 79%, and the in-hospital mortality rate was 11% (all rates were less favorable than those for younger patients), and the urgent CABG rate was less than 1% ⁽⁴³⁶⁾. In 1 report of 26 patients greater than 90 years old, of whom 20 had UA, the procedural success rate for PTCA was 92%, whereas the acute clinical success rate was only 65%, with an in-hospital mortality rate of 23% ⁽⁴³⁷⁾.

On the other hand, a Mayo Clinic review of PCI in patients greater than 65 years old (of whom 75% had UA) revealed an overall success rate of 93.5%, an immediate in-hospital mortality rate of 1.4%, and a need for emergency CABG rate of only 0.7%. Angiographic outcome changed little between the 65- to 69-year-old group and the greater than 75-year-old group, and the 1-year event rate (death, MI, CABG, repeat PCI, or severe angina) was 45.1% in all patients greater than 65 years old ⁽⁴²⁹⁾. Predictors of outcomes (i.e., extent and severity of CAD and comorbidities) after PCI in the elderly were the same as those in younger patients ⁽⁴³⁵⁾. Similarly, a review of coronary stenting in the elderly reported that procedural success rates were high (95% to 98%) and periprocedural complication rates were low (MI 1.2% to 2.8%, urgent CABG 0.9% to 1.8%, repeat PCI 0% to 0.6%) in the elderly with little difference between those greater than 75 years old and those less than 65 years old ⁽⁴³⁰⁾. Subgroup analyses in both TIMI IIIB ⁽¹⁹⁾ and FRISC II ⁽²⁷⁸⁾ showed a greater advantage of the invasive strategy in patients greater than 65 years old.

A review of 15,679 CABG procedures carried out in patients greater than 70 years old from The Toronto Hospital ⁽⁴³⁸⁾ reported encouraging results. Operative mortality rates declined from 7.2% in 1982 to 1986 to 4.4% in 1987 to 1991 (from 17.2% to 9.1% for high-risk patients) but showed little further change in the period of 1992 to 1996. Predictors of operative death (LV dysfunction, previous CABG, peripheral vascular disease, and diabetes) were similar to those in younger patients. When adjusted for these risk factors, age (i.e., a comparison of patients greater than 75 years old with those 70 to 74 years old) was not a significant risk factor.

In octogenarians, early mortality rates with CABG were found to be approximately 2.5 times those in patients 65 to 70 years old ⁽⁴³¹⁾; stroke occurred in approximately 8% ⁽⁴³⁹⁾, and less serious cerebral complications were even more common ^(433,440). However, in a review of patients studied between 1985 and 1989, the 3-year survival rate for octogenarian CABG patients was better than that in comparably aged patients with CAD who did not undergo surgery (77% vs. 55%, p = 0.0294) ⁽⁴⁴⁰⁾, and in another study, the quality of life of patients 80 to 93 years old was improved with CABG ⁽⁴⁴¹⁾.

4. Conclusions
Elderly patients with UA/NSTEMI tend to have atypical presentations of disease, substantial comorbidity, ECG stress tests that are more difficult to interpret, and different responses to pharmacological agents compared with younger patients. Their outcomes with interventions and surgery are not as favorable as those of younger patients, in part because of greater comorbidities. However, coronary revascularization can be performed when the same group of prognostic risk factors that play a role in the younger age group are taken into account. The approach to these patients also must consider general medical and mental status and anticipated life expectancy. Very frail elderly patients represent a high-risk group and should be evaluated for revascularization on a case-by-case basis. In many of these patients, even those with diffuse coronary arterial disease, PCI, with its lower morbidity rates, may be preferable to CABG. In ESSENCE ⁽¹⁶⁹⁾ and TIMI 11B ⁽¹⁷⁰⁾, the benefits or LMWH in patients greater than 65 years old were particularly impressive. In the case of platelet GP IIb/IIIa inhibitors, the relative benefits for older patients were similar to those of younger patients, but with the higher event rate in elderly patients, this translated into a greater absolute benefit.

E. Cocaine

Recommendations for Patients With Chest Pain After Cocaine Use

Class I

1. NTG and oral calcium antagonists for patients with ST-segment elevation or depression that accompanies ischemic chest discomfort. (Level of Evidence: C)
2. Immediate coronary arteriography, if possible, in patients whose ST segments remain elevated after NTG and calcium antagonists; thrombolysis (with or without PCI) if thrombus is detected. (Level of Evidence: C)

Class IIa

1. Intravenous calcium antagonists for patients with ST-segment deviation suggestive of ischemia. (Level of Evidence: C)
2. Beta-blockers for hypertensive patients (systolic blood pressure greater than 150 mm Hg) or those with sinus tachycardia (pulse greater than 100 min-1). (Level of Evidence: C)
3. Thrombolytic therapy if ST segments remain elevated despite NTG and calcium antagonists and coronary arteriography is not possible. (Level of Evidence: C)
4. Coronary arteriography, if available, for patients with ST-segment depression or isolated T-wave changes not known to be old and who are unresponsive to NTG and calcium antagonists. (Level of Evidence: C)

Class III

Coronary arteriography in patients with chest pain without ST-T-wave changes. (Level of Evidence: C)

The use of cocaine is associated with a number of cardiac complications that can produce myocardial ischemia and can cause and present as UA/NSTEMI ^(442-445). The widespread use of cocaine makes it mandatory to consider this cause, because its recognition mandates special management.

The action of cocaine is to block presynaptic reuptake of neurotransmitters such as norepinephrine and dopamine, producing excess concentrations at the postsynaptic receptors that lead to sympathetic activation and the stimulation of dopaminergic neurons ⁽⁴⁴⁶⁾. There may also be a direct contractile effect on vascular smooth muscle ⁽⁴⁴³⁾. Detoxification occurs with plasma and liver cholinesterases, which form metabolic products that are excreted in the urine. Infants, elderly patients, and patients with hepatic dysfunction lack sufficient plasma cholinesterase to metabolize the drug ⁽⁴⁴⁷⁾ and therefore are at high risk of adverse effects with cocaine use.

1. Coronary Artery Spasm
The basis for coronary spasm has been demonstrated in both in vitro ⁽⁴⁴⁷⁾ and in vivo ^{(443,448-452)} experiments in animals and humans. Reversible vasoconstriction of rabbit aortic rings has been demonstrated with cocaine in concentrations of 10^-3 to 10^-8 mol per L. Pretreatment with calcium antagonists markedly inhibits the cocaine-induced vasoconstriction. Coronary injection of cocaine produces vasoconstriction in miniswine with experimentally induced nonocclusive atherosclerotic lesions ⁽⁴⁵³⁾.

Nademanee et al. ⁽⁴⁵⁴⁾ performed 24-h ECG monitoring in 21 male cocaine users after admission to a substance abuse treatment center and found that 8 had frequent episodes of ST-segment elevation, most during the first 2 weeks of withdrawal. In cocaine users with prolonged myocardial ischemia, coronary arteriography may reveal coronary artery spasm with otherwise normal appearing coronary arteries or with underlying minimally obstructive coronary atherosclerosis ^{(443,445,448)}. The cocaine-induced increase in coronary vascular resistance is reversed with calcium antagonists ^(449,455). Cocaine increases the response of platelets to arachidonic acid, thus increasing thromboxane A2 production and platelet aggregation ⁽⁴⁵⁶⁾. In addition, reversible combined reduction in protein C and antithrombin III has been observed in patients with cocaine-related arterial thrombosis ⁽⁴⁵⁷⁾. All of these effects favor coronary thrombosis ^{(443,450,458)}. Coronary thrombus may also develop as a consequence of coronary spasm.

Cocaine users may develop ischemic chest discomfort that is indistinguishable from the UA/NSTEMI secondary to coronary atherosclerosis. The patient who presents with prolonged myocardial ischemia should be questioned about the use of cocaine. In a study by Hollander et al., the presence or absence of cocaine use was assessed in only 13% of patients who presented to the ED with chest pain ⁽⁴⁵⁸⁾. Table 22 lists the clinical characteristics of a typical patient with cocaine-related chest pain or MI ⁽⁴⁴⁵⁾.

Most patients who present to the ED with cocaine-associated chest pain do not develop MI ⁽⁴⁶⁰⁾; MI development has been reported to occur in 6% of such patients ⁽⁴⁴⁵⁾.

Accelerated coronary atherosclerosis has been reported in chronic users of cocaine ^(461,462); coronary artery spasm is more readily precipitated at sites of atherosclerotic plaques ⁽⁴⁴⁸⁾. Cocaine causes sinus tachycardia, an increase in blood pressure, and myocardial contractility, thereby increasing myocardial oxygen demand ⁽⁴⁴⁹⁾. These increases may precipitate myocardial ischemia and UA/NSTEMI in both the presence and absence of obstructive coronary atherosclerosis and coronary spasm.

Aortic dissection ⁽⁴⁶³⁾ and coronary artery dissection ^(443,463) have been reported as consequences of cocaine use. Other reported cardiac complications are myocarditis ⁽⁴⁶²⁾ and cardiomyopathy ^(464,465).

2. Treatment
When a patient with or suspected of cocaine use is seen in the ED with chest pain compatible with myocardial ischemia and ST-segment elevation, NTG and a calcium antagonist (e.g., 20 mg diltiazem) should be administered intravenously ^(443,452). If there is no response, immediate coronary arteriography should be performed, if possible. If thrombus is present, thrombolytic agents are administered if there are no contraindications ^(466,467). If catheterization is not available, thrombolytic agents may be considered.

If the ECG is normal or shows only minimal T-wave changes and there is a history of chest pain compatible with acute myocardial ischemia, the patient should receive NTG and an oral calcium antagonist and be observed. After cocaine use, increased motor activity, skeletal muscle injury, and rhabdomyolysis can occur, causing CK and even CK-MB elevation in the absence of MI ⁽⁴⁶⁸⁾. TnI or TnT is more specific for myocardial injury and therefore is preferred. Blood should be drawn twice for serum markers of myocardial necrosis at 6-h intervals. If the ECG shows ST-segment changes and the biochemical markers are normal, the patient should be observed in the hospital in a monitored bed for 24 h; most complications will occur within 24 h ⁽⁴⁵⁹⁾. If the patient's clinical condition is unchanged and the ECG remains unchanged after 24 h, the patient can be discharged ⁽⁴⁶⁹⁾.

Many observers believe that beta-blockers are contraindicated in cocaine-induced coronary spasm, because there is evidence from a single double-blind, randomized, placebo-controlled trial that beta-adrenergic blockade augments cocaine-induced coronary artery vasoconstriction ⁽⁴⁷⁰⁾. Others believe that if the patient has a high sympathetic state with sinus tachycardia and hypertension, then beta-blockers should be used ⁽⁴⁴³⁾. Labetalol, an alpha/beta-blocker, has been advocated, but in the doses commonly used, its beta-adrenergic-blocking action predominates over its alpha-adrenergic-blocking activity ⁽⁴⁷¹⁾. Therefore, in cocaine-induced myocardial ischemia and vasoconstriction, NTG and calcium antagonists are the preferred drugs. Both NTG and verapamil have been shown to reverse cocaine-induced hypertension and coronary arterial vasoconstriction ^(452,470) and tachycardia (verapamil).

F. Variant (Prinzmetal's) Angina

Class I

1. Coronary arteriography in patients with episodic chest pain and ST-segment elevation that resolves with NTG and/or calcium antagonists. (Level of Evidence: B)
2. Treatment with nitrates and calcium antagonists in patients whose coronary arteriogram is normal or shows only nonobstructive lesions. (Level of Evidence: B)

Class IIa

Provocative testing in patients with a nonobstructive lesion on coronary arteriography, the clinical picture of coronary spasm, and transient ST-segment elevation. (Level of Evidence: B)

Class IIb

1. Provocative testing without coronary arteriography. (Level of Evidence: C)
2. In the absence of significant CAD on coronary arteriography, provocative testing with methylergonovine, acetylcholine, or methacholine when coronary spasm is suspected but there is no ECG evidence of transient ST-segment elevation. (Level of Evidence: C)

Class III

Provocative testing in patients with high-grade obstructive lesions on coronary arteriography. (Level of Evidence: B)

Variant angina (Prinzmetal's angina) is a form of UA that usually occurs spontaneously, is characterized by transient ST-segment elevation, and most commonly resolves without progression to MI ⁽⁴⁷²⁾. The earliest stages of AMI may also be associated with cyclic ST-segment elevations. Although Prinzmetal was not the first to describe this condition ⁽⁴⁷³⁾, he was the first to offer the hypothesis that it is caused by transient coronary artery spasm; this was subsequently proved with coronary arteriography ⁽⁴⁷⁴⁾. The spasm is most commonly focal and can occur simultaneously at greater than 1 site ⁽⁴⁷⁵⁾. Even coronary segments that are apparently normal on coronary angiography often have evidence of mural atherosclerosis on intravascular ultrasonography ⁽⁴⁷⁶⁾. This can result in localized endothelial dysfunction and coronary spasm.

Patients with Prinzmetal's angina frequently have coronary artery plaques that can be nonobstructive or produce significant stenosis ⁽⁴⁷⁷⁾. Walling et al. ⁽⁴⁷⁸⁾ reported in 217 patients that coronary arteriography showed 1-vessel disease in 81 (39%) patients and multivessel disease in 40 (19%) patients. Rovai et al. ⁽⁴⁷⁹⁾ found a similar high prevalence of obstructive disease in 162 patients with variant angina.

1. Clinical Picture
Although chest discomfort in the patient with variant angina can be precipitated by exercise, it usually occurs without any preceding increase in myocardial oxygen demand; the majority of patients have normal exercise tolerance, and stress testing may be negative. Because the anginal discomfort usually occurs at rest without precipitating cause, it simulates UA/NSTEMI secondary to coronary atherosclerosis. Episodes of Prinzmetal's angina often occur in clusters with prolonged weeks to months of asymptomatic periods. However, attacks can be precipitated by hyperventilation ⁽⁴⁸⁰⁾, exercise ⁽⁴⁸¹⁾, and exposure to cold ⁽⁴⁸²⁾. There tends to be a circadian variation in the episodes of angina, with most attacks occurring in the early morning ⁽⁴⁸³⁾. Compared with patients with chronic stable angina, patients with variant angina are younger and, except for smoking, have fewer coronary risk factors ^(484,485). Some studies have shown an association of variant angina with other vasospastic disorders such as migraine headache and Raynaud's phenomenon ⁽⁴⁸⁶⁾.

Most often, the attacks resolve spontaneously without evidence of MI. However, if the coronary vasospasm is prolonged, MI, a high degree of AV block, life-threatening ventricular tachycardia, or sudden death may occur ^(487,488).

2. Pathogenesis
The pathogenesis of focal coronary spasm in this condition is not well understood. The probable underlying defect is the presence of dysfunctional endothelium that exposes the medial smooth muscle to vasoconstrictors such as catecholamines, thromboxane A2, serotonin, histamine, and endothelin ⁽⁴⁸⁹⁾. Endothelial dysfunction may also impair coronary flow-dependent vasodilatation due to the decreased production and release of nitric oxide ⁽⁴⁹⁰⁾ and enhance phosphorylation of myosin light chains, an important step for smooth muscle contraction ⁽⁴⁹¹⁾. There may be an imbalance between endothelium-produced vasodilator factors (i.e., prostacylin, nitric oxide) and vasoconstrictor factors (i.e., endothelin, angiotensin II), to favor the latter ⁽⁴⁹²⁾. There also is evidence for involvement of the autonomic nervous system with reduced parasympathetic tone and enhanced reactivity of the alpha-adrenergic vascular receptors ^{(490,493,494)}. Regardless of the mechanism, the risk for focal spasm is transient but recurrent.

3. Diagnosis
The diagnosis of variant angina is made by demonstrating ST-segment elevation in a patient during transient chest discomfort (which usually occurs at rest) that resolves when the chest discomfort abates. On coronary arteriography, if the artery is found to be angiographically normal or exhibits only nonobstructive plaques, then coronary artery spasm is the most likely explanation. If the patient has a spontaneous episode of pain and ST-segment elevation in the course of coronary arteriography, severe focal spasm of an epicardial coronary artery may be visualized. If the spasm is persistent, MI can occur; this is a rare complication in patients with variant angina who have normal or near-normal coronary arteries on arteriography. However, MI is common when coronary spasm complicates multivessel obstructive CAD ⁽⁴⁷⁸⁾. Coronary arteriography can show obstructive lesions, and increased arterial tone at a site of stenosis can precipitate total occlusion and the picture of impending infarction that is reversed on resolution of the increased vasomotor tone.

When the coronary arteriogram is normal or shows only nonobstructive plaques and if transient ST-segment elevation can be demonstrated at the time at which the patient has the discomfort, the presumptive diagnosis of Prinzmetal's angina can be made and no further tests are necessary. The key is to observe the ECG at the time of the attacks. If attacks occur frequently, a 24-h ambulatory ECG may be helpful in establishing the diagnosis.

In the absence of ST-segment elevation that accompanies chest discomfort, a number of provocative tests (methylergonovine, acetylcholine, and methacholine) can precipitate coronary artery spasm that can be visualized angiographically and is accompanied by ST-segment elevation in patients with Prinzmetal's variant angina ⁽⁴⁹⁵⁾. The patients should be withdrawn from nitrates and calcium antagonists before provocative testing. Hyperventilation performed for 6 min in the morning alone or after exercise is another test for coronary artery spasm ⁽⁴⁹⁶⁾. Patients with a positive hyperventilation test are more likely to have a higher frequency of attacks, multivessel spasm, and a high degree of AV block or ventricular tachycardia than are patients with a negative hyperventilation test ⁽⁴⁹⁶⁾. Because these provocative tests can occasionally cause prolonged intense and even multivessel spasm that requires intracoronary NTG or calcium antagonists for relief, the tests that require intravenous injections should be conducted in a catheterization laboratory with the catheter positioned in the coronary artery to deliver these drugs ⁽⁴⁹⁷⁾. The aforementioned drugs that are used to precipitate coronary artery spasm are not always readily available, so hyperventilation may be used.

4. Treatment
Coronary spasm is usually very responsive to NTG, long-acting nitrates, and calcium antagonists ^(498-500). Smoking should be discontinued. Usually, a calcium antagonist at a high dose (e.g., 240 to 480 mg per d verapamil, 120 to 360 mg per d diltiazem, 60 to 120 mg per d nifedipine) is started. If the episodes are not completely eliminated, a second calcium antagonist from another class or a long-acting nitrate should be added. Alpha-receptor blockers have been reported to be of benefit, especially in patients who are not responding completely to calcium antagonists and nitrates ⁽⁴⁹¹⁾. In patients who develop coronary spasm (with or without provocation) during coronary angiography, 0.3 mg of NTG should be infused directly into the coronary artery that is involved.

5. Prognosis
The prognosis is usually excellent in patients with variant angina who receive medical therapy. Yasue et al. ⁽⁵⁰¹⁾ reported an 89% to 97% overall 5-year survival rate. The prognosis is especially favorable in patients with normal or near-normal coronary arteries on arteriography. In a 7-year follow-up in approximately 300 patients, the incidence of sudden death was 3.6%, and the incidence of MI was 6.5% ⁽⁵⁰¹⁾. Patients with coronary artery vasospasm superimposed on a fixed obstructive CAD have a worse prognosis. In a study of 162 patients with variant angina by Rovai et al. ⁽⁴⁷⁹⁾, the patients with normal coronary arteries and 1-vessel disease had a 5-year survival rate of 95% compared with a rate of 80% for those with multivessel disease. Almost identical survival rates were reported in an earlier study by Walling et al. ⁽⁴⁷⁸⁾.

G. Syndrome X

Recommendations

Class I

1. Reassurance and medical therapy with nitrates, beta-blockers, and calcium antagonists alone or in combination. (Level of Evidence: B)
2. Risk factor reduction. (Level of Evidence: C)

Class IIb

1. Intracoronary ultrasound to rule out missed obstructive lesions. (Level of Evidence: B)
2. If no ECGs are available during chest pain and coronary spasm cannot be ruled out, coronary arteriography and provocative testing with methylergonovine, acetylcholine, or methacholine. (Level of Evidence: C)
3. HRT in postmenopausal women unless there is a contraindication. (Level of Evidence: C)
4. Imipramine for continued pain despite Class I measures. (Level of Evidence: C)

Class III

Medical therapy with nitrates, beta-blockers, and calcium antagonists for patients with noncardiac chest pain. (Level of Evidence: C)

1. Definition and Clinical Picture
The term "syndrome X" is used to describe patients with angina or angina-like discomfort with exercise, ST-segment depression on treadmill testing, and normal or nonobstructed coronary arteries on arteriography. This entity should be differentiated from the "metabolic syndrome X," which describes patients with insulin resistance, hyperinsulinemia, dyslipidemia, hypertension, and abdominal obesity. It should also be differentiated from noncardiac chest pain. Syndrome X is more common in women than in men ^(354,502). Chest pain can vary from that of typical angina pectoris to chest pain with atypical features to chest pain that simulates UA, secondary to CAD ⁽⁵⁰²⁾. Other atypical features can be prolonged chest pain at rest and chest pain that is unresponsive to NTG ⁽⁵⁰³⁾. Most often, the chest pain occurs with activity and simulates angina pectoris due to stable CAD. However, because chest pain may accelerate in frequency or intensity or occur at rest, the patient may present with the clinical picture of UA. Therefore, this syndrome is discussed in this guideline.

The cause of the discomfort and ST-segment depression in patients with syndrome X is not well understood. The most frequently proposed causes are impaired endothelium-dependent arterial vasodilatation with decreased nitric oxide production, increased sensitivity to sympathetic stimulation, or coronary vasoconstriction in response to exercise ^{(355,504,505)}. There is increasing evidence that these patients frequently also have an increased responsiveness to pain and an abnormality in pain perception.

The diagnosis of syndrome X is one of the exclusion of critical obstruction of an epicardial coronary artery in patients with exertional chest discomfort who have ST-segment depression on treadmill exercise. Other causes of angina-like chest discomfort not associated with cardiac disease, such as esophageal dysmotility, fibromyalgia, and costochondritis, must also be eliminated. In addition, in patients with a clinical presentation consistent with variant angina, coronary spasm must be ruled out by the absence of ST-segment elevation with the anginal discomfort or by provocative testing. Myocardial perfusion scanning may be abnormal due to a patchy abnormal response to exercise of the microvasculature that may lead to reduced coronary flow to different regions of the myocardium ⁽³⁵⁵⁾.

The intermediate-term prognosis of patients with syndrome X is excellent ^{(502,503,506)}. The CASS registry reported a 96% 7-year survival rate in patients with anginal chest pain, normal coronary arteriograms, and an LV EF of greater than 0.50 ⁽⁵⁰⁷⁾. Long-term follow-up shows that ventricular function usually remains normal ⁽⁵⁰³⁾, although there have been reports of progressive LV dysfunction, and many patients continue to have chest pain that requires medication ⁽⁵⁰⁸⁾.

2. Treatment
Because the long-term prognosis is excellent, the most important therapy consists of reassurance and symptom relief. However, persistence of symptoms is common, and many patients do not return to work ⁽⁵⁰³⁾. The demonstration of normal coronary arteries on angiography can be reassuring. In 1 study, after a normal coronary arteriogram, there was a reduced need for hospitalization as well as a reduced number of hospital days for cardiac reasons ⁽²⁸⁶⁾.

Both beta-blockers and calcium antagonists have been found to be effective in reducing the number of episodes of chest discomfort ^(509,510). Beneficial effects with nitrates are seen in about one half of patients ⁽⁵¹¹⁾. The use of alpha-adrenergic blockers would appear to be a rational therapy, but the results of small trials are inconsistent ⁽⁵¹²⁾. Imipramine 50 mg daily has been successful in some chronic pain syndromes, including syndrome X, reducing the frequency of chest pain by 50% ⁽⁵¹³⁾. Estrogen replacement in postmenopausal women with angina and normal coronary arteriograms has been shown to reverse the acetylcholine-induced coronary arterial vasoconstriction, presumably by improving endothelium-dependent coronary vasomotion ⁽⁵¹⁴⁾. In a double-blind, placebo-controlled study, Rosano et al. ⁽⁵¹⁵⁾ found that cutaneous estrogen patches in 25 postmenopausal women with syndrome X reduced the frequency of chest pain episodes by 50%.

It is recommended that patients be reassured of the excellent intermediate-term prognosis and treated with long-acting nitrates. If the patient continues to have episodes of chest pain, a calcium antagonist or beta-blocker can be started ⁽⁵¹⁰⁾. Finally, 50 mg of imipramine daily has been successful in reducing the frequency of chest pain episodes. If symptoms persist, other causes of chest pain, especially esophageal dysmotility, should be ruled out ^{(9,95,97-99,102,104,105,318)}.