Abstract & Commentary
Source: Henrickson CA, et al. Chest pain relief by nitroglycerin does not predict active coronary artery disease. Ann Intern Med 2003;139:979-986.
To clarify nitroglycerin’s (NTG) ability to distinguish ischemic chest pain, Henrickson and colleagues studied symptomatic patients in an academic emergency department (ED), evaluating NTG responsiveness as a predictor of coronary artery disease (CAD).
Four hundred fifty-nine consecutive patients seen in Baltimore’s Johns Hopkins Hospital comprised the study population. All reported chest pain, received NTG in the ED, and were hospitalized to rule out myocardial infarction. Average age was 59 years; 54% were female, 81% were Caucasian, and 18% African-American. Prior CAD was present in 31%, diabetes in 24%, hypercholesterolemia in 41%, tobacco abuse in 43%, and family history of CAD in 35%. Electrocardiograms (ECGs) obtained during pain were abnormal in 13%; 5% had ST segment depression, 5% had ST segment elevation, and 3% had left bundle-branch block. Patients whose chest pain resolved prior to ED arrival and those who could not rate pain on a 1-10 scale were excluded.
Each patient received NTG, either 0.4 mg sublingual or 0.4 mg oral spray. Ancillary treatments included oxygen, aspirin, morphine, and beta-adrenergic antagonist in 60%, 33%, 4%, and 3% of patients, respectively. NTG-responsive chest pain was defined as 50% or greater reduction in chest pain within five minutes, as assessed on a 1-10 point severity scale. Subsequent documentation of active CAD was defined as an elevated troponin T, angiography showing greater than 70% coronary artery stenosis, or positive exercise test with confirmatory imaging. Phone interviews were conducted four months after admission to assess interval cardiac events or testing results.
A total of 181 cases (39%) had greater than 50% reduction in chest pain and were termed NTG-responders (NTG-R). In contrast, 278 cases (61%) had insignificant or no response to NTG (NTG-NR). Subsequent cardiac evaluation, as described above, was suggestive of active CAD in 141 (31%) of the 459 patients.
Of those 141 with evidence of CAD and chest pain, 49 (35%) were in the NTG-R group. By comparison, of those in whom active CAD was excluded by testing, 113 of 278 (41%) likewise were in the NTG-R group (p > 0.2). Statistical analysis of this data generated a sensitivity of 34%, and a specificity of 58%, for NTG-R as a predictor of active CAD. All 95% confidence intervals included 1.0 for positive and negative likelihood ratios for NTG relief of chest pain, indicating no significant difference in predicting active CAD. At the four-month follow-up interview, there was no significant difference in NTG-R and NTG-NR groups in terms of myocardial infarction, revascularization, or death. The authors conclude that relief of chest pain with NTG does not correlate with active CAD in the ED setting.
Commentary by Michael W. Felz, MD
Sensitivity and specificity were quite low in predicting CAD, and likelihood ratios indicated that NTG-R (and NTG-NR) are not useful diagnostic signs for confirmation (or exclusion) of myocardial ischemia or infarction. Furthermore, NTG-R did not accurately predict subsequent cardiac events over a four-month follow-up period.
I am uncertain about the applicability of this data. The findings challenge commonly held practice standards for patients whose chest pain resolves with NTG in the ED, office, or hospital. The NTG-R sequence indicates to me either ischemic heart disease or, perhaps, esophageal spasm. Lack of response to NTG, however, speaks against CAD, at least in patients whose ECGs do not suggest acute infarction. In such patients, I search hard for chest wall tenderness reproducing symptomatology, or treat presumptively for GERD. Almost undoubtedly, the presence of ST depression plus chest pain, with reversal of both by NTG administration, reliably still predicts the presence of CAD; in this population, only 13% had abnormal ECGs during pain.
Dr. Felz, Associate Professor, Department of Family Medicine, Medical College of Georgia, Augusta, is on the Editorial Board of Emergency Medicine Alert.