CTA and Cocaine‐associated Chest Pain

Abstract

We appreciate the concerns of Drs. Livshits and Hoffman regarding the use of coronary computed tomography angiography (CTA) in patients with cocaine-associated chest pain. Our study1 was not designed to address appropriateness of use in that patient population. It was designed to evaluate whether the view previously shared by me, Dr. Hoffman, and others that chronic cocaine was associated with accelerated atherosclerosis was, in fact, true. As noted in the article, this widely accepted view was based on small studies without comparator arms. The more extensive observational study we published found that this association does not really exist. The criticisms or concerns pointed out by Drs. Livshits and Hoffman are based on flawed assumptions or misreading of our paper. We will address them one at a time. They claim that “the utility of CTA in patients with low- to moderate-risk chest pain” has not been fully established. Although it is true that nothing is “fully” established, the current literature strongly supports its use. A meta-analysis of 9,592 patients with a median follow-up of 20 months found that the sensitivity was 99% for predicting death, myocardial infarction, or revascularization.2 The pooled negative likelihood ratio from the 18 studies was a remarkably low 0.008.2 Recent guidelines give coronary CTA an appropriateness score of 7 (defined as the test is generally acceptable and is a reasonable approach), for symptomatic low- to intermediate-risk patients without known coronary disease.3 The American Heart Association/American College of Cardiology (AHA/ACC) guidelines specifically state that patients with potential acute coronary syndromes should be evaluated for short-term risk of complications and then subsequently the risk of underlying coronary disease. Although the authors claim that coronary CTA is “unable to determine the presence of coronary disease and/or confirm its absence by accepted criterion standard diagnostic modalities,” they are incorrect. Even using older generation CT scans, Janne d’Othee et al.4 pooled 41 studies of 2,515 patients and found excellent per-patient and per-segment diagnostic accuracy compared to cardiac catheterization. Sixty-four-slice CT had a sensitivity of 98% and a specificity of 92%, far exceeding the predictive properties for stress testing. At many sites, including ours, the radiation exposure from coronary CTA is considerably less than from nuclear stress testing. Similarly, lifetime radiation exposure from a coronary CTA-based approach is expected to be less than from an approach with stress testing. The emergency department–based studies all suggest a lower rate of repeat visits, repeat testing, and repeat admission when patients are plainly told that they do not have coronary artery disease. Coronary CTA allows the physician to do this. A normal stress test does not. It is rare that emergency physicians have a test that has more accuracy at lower cost than the alternative, but a coronary CTA–based approach provides just that.5 Unless Drs. Livshits and Hoffman advocate totally abandoning an evaluation of possible coronary disease in all patients with cocaine-associated chest pain, we believe that a CTA-based strategy has demonstrated a high degree of diagnostic accuracy, strong prognostic ability in both the near and the long term, lower radiation exposure, and more cost-effectiveness than alternative strategies to “rule-out” ACS. Of course, they are correct that patients with recent cocaine use should not be subjected to risks of beta blockade.6 Our article specifically stated, “Beta-blockers … were not used for patients with recent cocaine use ….” Thus, there was no additional risk to patients. It is especially fun to debate the approach to patients with cocaine-associated chest pain with one of the leaders in the field and one of my mentors. Like most things in the scientific community, what we think we know now will continue to be challenged, and we need to continually evaluate “the truth.” With respect to the relationship between chronic cocaine use and development of early atherosclerosis, what we once believed may no longer be true.