A number of publications recently maligned the use of morphine for treating acute coronary syndrome (ACS). The hypothesis is that morphine slows the absorption of the oral P2Y12 inhibitors such as ticagrelor, delaying the dual-platelet inhibition desired by cardiologists.
The most recent data supporting this hypothesis was in a study appropriately entitled the IMPRESSION trial. (Eur Heart J 2016;37:245.) Its results propagated a concept that appears valid on the surface but may not accurately represent the underlying reality.
This single-center trial by Kubica, et al., randomized patients presenting with acute myocardial infarction destined for cardiac catheterization to receive 5 mg of IV morphine or placebo prior to receiving a 180 mg loading dose of ticagrelor. Drug concentrations and platelet function were measured over the next 12 hours of each patient's hospital course.
The authors found just what they set out to: The use of morphine inhibited the absorption of ticagrelor and delayed its platelet inhibitory effects. The serum levels of ticagrelor were generally higher in the patients who received the placebo. The time to reach peak serum drug levels was also significantly delayed in the patients who received IV morphine. These pharmacokinetic shortcomings directly translated into delays in platelet inhibition measured using platelet function assays. The time to reach therapeutic platelet inhibition was one hour in the placebo group compared with two hours in the patients who received IV morphine.
This reasonably robust evidence demonstrates the pharmacokinetic and pharmacodynamic influence that morphine has on ticagrelor's platelet inhibitory capabilities. At first glance, this seems to condemn yet another staple in the emergency department's management of ACS. But as with so many questions involving the use of P2Y12 inhibitors, the pharmaceutical companies, in this case, AstraZeneca, have purposely and cunningly changed the question, deflecting attention from the more important inquiry: What benefits, if any, do P2Y12 inhibitors provide in the emergency department when managing ACS?
If we are to believe that morphine is harmful when utilized in patients experiencing myocardial infarction, then we first have to believe that the early upstream use of P2Y12 inhibitors provides some clinically important benefit despite the preponderance of literature demonstrating the exact opposite. P2Y12 inhibitors have failed to prove themselves capable of providing clinically important outcomes in patients presenting with ACS as far back as the CURE trial 16 years ago. (N Engl J Med 2001;345:494.) The authors demonstrated a statistically significant benefit through the use of statistical bullying tactics. The small benefits found were achieved through large sample sizes and the use of questionable composite endpoints. The majority of the difference resulted from the reduction of type 4a MIs of questionable clinical consequence. (N Engl J Med 2001;345:494.)
Achieving these paltry benefits has never been shown to be dependent on the upstream administration of these medications in the emergency department. In fact, multiple trials have demonstrated that an identical level of mediocrity can be achieved once stent-appropriate anatomy has been defined.
The CREDO trial examined the upstream use of clopidogrel prior to cardiac catheterization, and found no clinical benefit to the early dual-antiplatelet effects this prompt administration provided. (JAMA 2002;288:2411.) The 2013 ACCOAST trial by Montalescot, et al., examined early loading of prasugrel in patients requiring urgent catheterization after experiencing an NSTEMI. (N Engl J Med 2013;369:999.) These authors randomized patients to receive prasugrel upstream in the emergency department or in the cath lab, only after anatomy appropriate for stenting was defined. Similar to the CREDO trial, these authors found no benefit to the upstream use of prasugrel, reporting no difference in the rate of cardiovascular death, myocardial infarction, stroke, urgent revascularization, or glycoprotein IIb/IIIa bailout (10.8% vs 10.8%). Not surprisingly, the pretreatment group was found to have an increase in major bleeding of approximately one percent, mostly due to an increase in CABG-related bleeding (20.7% vs 13.7%).
Finally, the ATLANTIC trial, which was published in the New England Journal of Medicine by Montalescot, et al., examined the use of prehospital ticagrelor in patients presenting with ST-elevation MI. (2014;371:1016.) The authors randomized 1,862 patients to receive a prehospital 180 mg loading dose of ticagrelor or a delayed dose in the cath lab prior to the start of the procedure. And just like the multitude of trials before it, these authors found no improvement in the rate of cardiovascular deaths, MIs, strokes, urgent revascularizations, or definitive stent thrombosis (4.5% vs 4.4%) between the groups. Nor did they find a difference in the rate of 70 percent or greater resolution of ST-segment elevation before PCI or in the proportion of patients who did not meet the criteria for TIMI flow grade 3 in the infarct-related artery at angiography before PCI.
In an attempt to save face, the authors noted the one subgroup that demonstrated a statistically significant finding in favor of the patients receiving prehospital ticagrelor. In patients who did not receive IV morphine, the rates of 70 percent or greater resolution of ST-segment elevation before PCI were more frequently observed in patients randomized to upstream ticagrelor use. This observation did not translate into a difference in the rate of TIMI flow grade 3 in this subset of patients.
Despite the well-known dangers of multiple outcomes and subset analyses, those in support of the hypothesis proposed by the IMPRESSION authors cling to this statistical lifeboat, suggesting that the use of morphine was responsible for the failure of the ATLANTIC trial to find benefit in the upstream administration of P2Y12 inhibition. That this statistical anomaly was discovered while dredging a sea of insignificance may, in fact, represent a true clinically important interaction, but another explanation is far more likely. The majority of trials examining the utility of P2Y12 inhibition have found trivial improvements, and no trial has ever demonstrated that these marginal benefits are dependent on their timely administration. Despite claims of harm, the authors of the IMPRESSION trial reported no difference in the rates of patient-oriented outcomes, death, MI, stent thrombosis, or stroke. And while statistically underpowered, these results are consistent with the remainder of the literature. As such, the negative results demonstrated in the ATLANTIC trial are likely because the upstream use of P2Y12 inhibitors provide no clinically meaningful benefits. This latter explanation seems far more plausible when the literature is examined in whole.
But plausibility and truth do not sell medications. For that purpose, all that is required is composite endpoints, stacked on surrogate outcomes, and piled on top of methodologically manufactured p-values served to us under a physiologically palatable fiction such as dual-antiplatelet therapy.