1. What is the relationship between cocaine and Acute Coronary Syndrome/MI?
Cocaine is the second most commonly abused drug in the United States, accounting for the most drug-related visits to the Emergency Department (Finkel, 2011). Chest pain is the most common presenting complaint, representing 40% of cocaine-related visits (Brody, 1990). In fact, up to 25% of all patients presenting to urban EDs with non-traumatic chest pain may have used cocaine (Hollander, 1995).
When evaluating cocaine chest pain (CCP), vasospasm is at the forefront of most clinicians’ minds. Cocaine, however, has multiple hemodynamic and hematologic effects that increase the risk of myocardial ischemia, both acutely after each use and chronically over time.
Acutely, cocaine increases plasma levels of dopamine and norepinephrine through central adrenergic stimulation and inhibition of reuptake at the synapse. The resulting sympathetic outflow manifests as tachycardia, hypertension, and increased myocardial oxygen demand. This occurs in the setting of cocaine-induced coronary vasoconstriction and acute thrombosis, which both decrease myocardial oxygen delivery. Interestingly, this vasoconstriction appears to be more severe in areas of underlying atherosclerosis, and in patients with concomitant tobacco use, but also occurs in otherwise normal coronary arteries (Finkel, 2011; Hollander, 2006). Cocaine also induces a hypercoaguable state with studies demonstrating acute thrombosis after cocaine administration in animal and cadaver models (Dressler, 1990). This occurs via multiple mechanisms, including activation of platelets and elevation of procoagulant factors, including fibrinogen, without compensatory increase in fibrinolytic factors (Siegel, 2002).
Chronically, cocaine has been associated with accelerated atherosclerosis. Autopsy studies of young cocaine users and animal models have supported this theory. (Satran, 2005; Dressler, 1990). While less common in cocaine users with MI than non-cocaine users with MI, atherosclerosis is more common in people who use cocaine compared to controls (Weber, 2002). Recent data however have led some to question this presumption. In 2010, Chang, et al., used coronary computed tomography angiograms (CTA) to evaluate coronary artery disease (CAD) in CCP patients presenting to the ED and found that these patients did not have an increased incidence of atherosclerosis. In this study only low risk patients (normal ECG, negative troponin) were included, possibly explaining this discrepancy (Chang, 2011). Chronic cocaine use has also been associated with left ventricular hypertrophy and coronary artery aneurysms, with one study showing 30% of cocaine users undergoing angiography to have aneurysms (Satran, 2005).
Cocaine use, with resultant increased myocardial oxygen demand, multi-factorial decreased myocardial oxygen delivery and chronic deleterious changes intuitively leads to an increased risk of MI. There are several studies looking at cocaine-associated MI. Overall, the incidence is low, ranging from 0.7%-6% (Finkel, 2011). The COCHPA (Cocaine Associated Chest Pain) study, a multi-center prospective trial looking at patients presenting with chest pain in the setting of cocaine use, found the incidence of MI to be 6% (Hollander, 1994). Compared to the studies with lower incidence of MI, COCHPA used more rigorous inclusion criteria, has been validated by other studies, and is therefore more likely to represent the true incidence of cocaine-associated MI.
With regards to the temporal association between cocaine use and MI, the period of increased risk is unclear. A review of the literature found onset of chest pain to range from one minute to four days after cocaine use (Hollander, 1992). This delayed presentation is inconsistent with the half-life of cocaine of thirty to ninety minutes. However, active metabolites of cocaine are detectable for over forty-eight hours after use, and delayed MIs have been attributed to coronary vasoconstriction from active metabolites (Finkel, 2011). Additionally, cocaine withdrawal has been linked to myocardial ischemia several weeks after use, highlighting a second time window of risk (Hollander, 2006). Despite this extended and variable period of heightened risk, the greatest risk of MI is thought to be within the first few hours after cocaine use. In the first hour, it is estimated that cocaine users are at a full twenty-four-fold increase in risk of MI with rapid decline thereafter (Mittleman, 1999).
2. How do you risk stratify a patient who presents with cocaine chest pain? If low risk, how does the disposition differ from that of standard chest pain patients?
Identification of high-risk patients with CCP parallels that of non-cocaine chest pain. An appropriate clinical scenario accompanied by characteristic ECG changes consistent with a STEMI warrants immediate percutaneous coronary intervention regardless of concomitant cocaine use. The evaluation of non-STEMI CCP however is more challenging, as ECG and serum cardiac markers are often distorted, thus limiting the physician’s ability to assess for ischemia.
ECG interpretation in cocaine users is difficult as these patients often have abnormal ECGs in the absence of ischemia, and conversely can have non-diagnostic ECGs in the setting of true ischemia. Cocaine use promotes left ventricular hypertrophy. Early repolarization is common in the population of patients most frequently presenting with CCP: young males. Both factors hinder the interpretation the ECG, making the diagnosis of MI more complicated. Overall, sensitivity of ECG in the setting of CCP is only 36% (Hollander, 1994). With regard to cardiac biomarkers, false elevations of CK and CK-MB occur secondary to skeletal muscle injury and rhabdomyolysis. Trends should be followed if using these markers; rising levels are more indicative of MI. Preferably, troponin markers should be used as they are more sensitive, particularly in CCP (Hollander, 1998).
The more difficult question, however, is how to manage low risk patients with CCP. While many institutions have moved towards two troponin rule outs, CTA, or observation units for low risk standard chest pain, a more conservative approach has endured for CCP given the limitations above. This approach comes at a cost: an admission for CCP averages three days and costs $83 million annually (Weber, 2003).
There have been several recent studies that have tried to address the disposition of low risk patients with CCP. In 2010, Chang investigated ED CTA for patients with low risk CCP to evaluate for CAD, a practice used at that institution for evaluation of standard chest pain. In this study, CTA did not show an increased incidence of CAD in patients with CCP (Chang, 2011). In standard low risk patients, a negative CTA allowed the patient to be discharged with close follow up. There have been concerns raised however regarding the use of CTA to manage patients with CCP. First, CTA cannot evaluate for vasospasm, one of the major concerns in CCP. Furthermore, CTA often requires use of beta blockade for adequate assessment, which is contraindicated in CCP. Finally, patients with CCP often present to the ED multiple times and may be subjected to a substantial radiation burden if CTA is used (Livshitz, 2011).
Another attractive alternative to admission is the use of observation units. A study of 302 patients looked at a nine to twelve hour observation period for low risk CCP patients and found it to be extremely safe. At thirty days, there were no ventricular dysrhythmmias and no deaths. Four patients had recurrent non-fatal MIs, all of which had continued cocaine use. In this study, just over half of the patients were stressed, four were positive, only two of which had multi-vessel disease (Weber, 2003). These findings are consistent with previous studies showing patients with low risk CCP to be at very small risk for delayed complications, most of which occur within the first twelve hours (Hollander, 1994).
3. What is your medical management of cocaine-associated chest pain?
Medical management of patients with CCP, including those with unstable angina and MI, is similar to the standard ACS equivalent with a few notable exceptions. Aspirin has clear mortality benefit in typical ACS; given the prothrombotic effects of cocaine, likely has benefit specific to cocaine related ischemia as well. It has not been studied in CCP specifically, as it would be unethical to withhold. Aspirin should always be given if not otherwise contraindicated. Nitroglycerin, which reduces infarct size in standard MI, has been shown to relieve chest pain associated with cocaine use, and to reverse vasospasm.
The inclusion of benzodiazepines in the AHA treatment algorithm for CCP distinguishes it from treatment of typical ACS. Patients with cocaine intoxication and chest pain often present extremely agitated and anxious. Animal studies suggest cocaine’s CNS effects are directly related to its cardiovascular manifestations. Treatment of the CNS effects with benzodiazepines improves the latter, reducing heart rate, blood pressure and mortality. Outcomes may actually be worse when these CNS effects are not addressed, despite treatment of peripheral vasoconstriction with nitroglycerin (McCord, 2008; Guinn, 1980).
Recent clinical studies have compared nitroglycerin to benzodiazepines, with conflicting results. Baumann found both nitroglycerin and diazepam to relieve chest pain and to decrease cardiac output and index. There was no significant difference between the two, however, and no additional benefit with combined treatment (2000). Honderick compared lorazepam plus nitroglycerin to nitroglycerin alone and found greater reduction in chest pain using the combination (2003). Of note very few patients in either study had ACS, thus limiting the applicability of the results.
Calcium channel blockers are also considered appropriate in CCP, but the data are limited and mixed. Some cardiac catheterization studies show reversal of vasospasm but outcomes have been worse in large scale studies of ACS, raising questions of morbidity or mortality benefit in CCP. Similar to typical ACS, calcium channel blockers should be avoided if there is evidence of heart failure. Calcium channel blockers are currently recommended as second line agents in CCP after nitroglycerin and benzodiazepines. (Finkel, 2011; McCord, 2008).
Phentolamine, an alpha-1 blocker, has theoretical benefit for alleviation of adrenergic alpha-1 mediated vasoconstriction in CCP. Reduction in chest pain has been reported in case reports and reversal of vasospasm has been documented in cath lab studies only (McCord, 2008). Phentolamine is not considered standard of care as more evidence is needed.
4. Do you ever use beta-blockers in patients with cocaine chest pain? What about labetalol?
Conventional teaching forbids the use of beta blockers in the management of cocaine toxicity, and with good reason: beta blockade permits unopposed alpha adrenergic stimulation, causing coronary vasoconstriction and hypertension. This has been demonstrated in animal models and illustrated in two placebo-controlled patient trials. In the first, by Lange, thirty patients receiving elective catheterization were randomized to receive intranasal saline or cocaine. These groups were then further randomized to receive intracoronary propranolol or saline. In the cocaine group, Lange found a 19% increase in coronary vascular resistance and a significant decrease in coronary sinus blood flow after propranolol administration. Furthermore, five patients showed at least a 10% constriction in a single coronary artery segment (1990). The second randomized controlled clinical trial of beta blockers and cocaine examined labetalol but unfortunately, in this small study, labetalol administration failed to reverse cocaine induced vasoconstriction (Boehrer, 1993).
Periodically, the controversy is reinvigorated with a new study. Most recently, in 2008, a retrospective cohort study claimed to find a decreased incidence of MI in patients admitted to telemetry or the ICU with a positive urine tox screen for cocaine (Dattilo, 2008). Critics pointed out that fewer than half of the patients in this study had chest pain, and that it was not applicable as it included patients admitted for any reason who also happened to have recently used cocaine. Additionally, the mortality benefit from beta blockade appears to derive, in large part, from continued usage after hospital discharge, something unwise to advocate in patients who will most likely continue using cocaine (Hoffman, 2008).
Despite the above data, beta blockers represent a cornerstone in treatment of traditional MI with a proven morbidity and mortality benefit. Given this, efforts to find a safe alternative and to optimize care of cocaine-associated MI has led to investigation of the combined alpha and beta blocker labetalol. The results, however, are far from promising. This is not surprising given that labetalol is primarily a beta adrenergic antagonist with very little alpha activity. While some animal studies have demonstrated absence of coronary vasoconstriction and improved hemodynamics others have shown an increase in mortality when cocaine exposed animals are treated with labetalol (Smith, 1991). More recently, Hoskins compared labetalol to diltiazem in a non-randomized study of ninety patients with CCP and found improvements in biomarkers and hemodynamic profiles in both groups with no adverse events (Hoskins, 2010).
The limited data on CCP and beta blockade suggests the combination to be harmful. The risk-benefit ratio should be considered carefully when deciding to use beta blockers in CCP, remembering the risk of MI in these patients is low and outcomes are typically good.