Long‐Term Ticagrelor in Stable Patients With Prior Myocardial Infarction: Bleeding Avoidance First and Foremost

Over 100 000 patients with different manifestations of cardiovascular disease have been studied in large‐scale clinical trials evaluating the safety and efficacy of ticagrelor alone or in combination with aspirin. As a result of these studies, the Food and Drug Administration granted indication to ticagrelor for the secondary prevention of patients presenting with acute coronary syndrome, ¹ prior myocardial infarction (MI), ² and acute noncardioembolic ischemic stroke, ³ as well as for the primary prevention of high‐risk patients with stable coronary artery disease. ⁴ However, given that any advantage in ischemic event prevention may be counterbalanced by an increase in bleeding harm, identifying those patients in whom the ischemia‐bleeding trade‐off of an intensified antithrombotic regimen provides a net clinical benefit remains a major challenge. International guidelines and academic collaborations have repeatedly emphasized the negative prognostic role of bleeding. Risk prediction models and expert consensus documents have been proposed to untangle the complex interplay between ischemic and bleeding risk and guide clinical decisions. ⁵ , ⁶ , ⁷ Nonetheless, several questions remain to be answered. Although no one would argue on the detrimental consequences of bleeding, the correct characterization of various types of hemorrhagic events and their causative role with respect to subsequent mortality are still a matter of debate. Bleeding in itself can be heterogeneous in cause and magnitude, and can thus impact differently on patient prognosis. A specific bleeding site can be more associated with one antithrombotic agent compared with another (eg, gastrointestinal adverse effects of aspirin). Furthermore, the clinical relevance of bleeding may also depend on the patient comorbid conditions and functional status.

In this issue of the Journal of the American Heart Association (JAHA), Magnani at al ⁸ have focused on these subjects by highlighting the results of a post hoc analysis of the PEGASUS‐TIMI 54 (Prevention of Cardiovascular Events in Patients With Prior Heart Attack Using Ticagrelor Compared to Placebo on a Background of Aspirin–Thrombolysis in Myocardial Infarction 54) Trial. ² Of the 21 000 patients with a prior MI who were enrolled and randomized to ticagrelor, 60 mg, ticagrelor, 90 mg, or placebo on top of aspirin, 432 experienced a TIMI minor or major bleeding event over a median follow‐up of 33 months. Presence of anemia at baseline and prior hospitalization for spontaneous bleeding were identified as independent predictors of bleeding but not ischemia and, therefore, were used to stratify the study population into patients with high (4054; 19.4%) and low bleeding risk (16 888; 80.6%). Compared with aspirin alone, ticagrelor, 60 mg, plus aspirin was associated with a similar increase in the risk of bleeding complications in both groups, but the absolute risk difference was nearly tripled in those at high bleeding risk (4.4% versus 1.5%). On the other hand, the composite of cardiovascular death, MI, or stroke was significantly in favor of patients treated with ticagrelor, 60 mg, compared with placebo, with a treatment effect that appeared more pronounced among patients with low bleeding risk (7.1% versus 8.7%; hazard ratio [HR], 0.80; 95% CI, 0.70–0.92) compared with patients with high bleeding risk (10.4% versus 10.4%; HR, 0.98; 95% CI, 0.77–1.26; P interaction=0.154). Consistent results were observed for the net benefit end point that also included intracranial and fatal bleeding. Hence, the authors concluded that absence of 2 simple bleeding risk parameters, such as a history of spontaneous bleeding and anemia, may be sufficient for identifying stable patients after MI who derive greater benefit from prolonged dual antiplatelet therapy with ticagrelor, 60 mg. Some limitations to this study, however, deserve to be mentioned. Anemia is a well‐known predictor of bleeding but also a widely prevalent condition. ⁶ , ⁹ , ¹⁰ Although lower levels of hemoglobin have been shown to correlate with a stepwise increase in bleeding risk, Magnani et al applied a dichotomous and relatively sensitive cutoff (≤12.0 g/dL for women and ≤13.5 g/dL for men). Anemia was present in 18.8% of the study population, a prevalence considerably larger than that of bleeders (2.1%), which on the whole suggests an only modest discrimination ability. Furthermore, the PEGASUS‐TIMI 54 Trial excluded patients with bleeding disorder, history of recent bleeding, or intracranial bleeding at any time, in addition to prior stroke or need for oral anticoagulant therapy. Patients enrolled in the PEGASUS‐TIMI 54 Trial were also at high risk of ischemic events (eg, diabetes mellitus, renal disease, multivessel disease, and recurrent MI). Therefore, the net benefit of prolonged ticagrelor therapy may be particularly enhanced in this population and less generalizable to other patient subsets.

The study by Magnani et al ⁸ is of great value in that it provides important insights on the causes of bleeding and its association with mortality. Nearly half (46%) of patients with a TIMI major or minor bleed were treated conservatively (without transfusion or intervention to reverse bleeding), and 9.7% died within a few days of causes directly related to bleeding. Interestingly, among patients who survived the bleeding event, another 9.7% died after a median time of 207.5 days, a rate of mortality that far exceeds the one observed in the overall PEGASUS‐TIMI 54 Trial population. By looking at the causes of late death in bleeding survivors, about half were cardiovascular, to which bleeding may contribute by prompting therapy discontinuation, hemodynamic compromise, and inflammatory reactions. ¹¹ Nonetheless, another 37% were adjudicated as related to malignancy. Notably, the excess in noncardiovascular death associated with prolonged dual antiplatelet therapy was attributed to cancer in both the DAPT (Dual Antiplatelet Therapy) and PEGASUS‐TIMI 54 Trials, ² , ¹² thereby challenging the role of bleeding as a mere disease manifestation or an actual trigger of the downstream fatal outcome.

TIMI major or minor bleeds were increased with the adjunctive use of ticagrelor, regardless of the baseline bleeding risk status. Most bleeds were spontaneous (71%), and similar in location and cause between treatment arms. Of every 4 bleeding events, 3 arose from the gastrointestinal tract, and ticagrelor plus aspirin was associated with a 3‐fold increase in their occurrence compared with aspirin alone. Yet, only 26% of patients were treated with proton pump inhibitors, despite their established indication for gastrointestinal prophylaxis among those with cardiovascular disease on antiplatelet therapy. ¹³ Furthermore there were no between‐group differences in intracranial and fatal bleeding. All these findings should be interpreted in the context of prior studies using different referent treatment arms. No difference in bleeding risk was observed among patients randomized to monotherapy with ticagrelor or aspirin after a cerebrovascular accident. ³ Bleeding risk was also similar between ticagrelor and clopidogrel monotherapy in patients with symptomatic peripheral artery disease. ¹⁴ A paradoxically greater proportion of gastrointestinal bleeds was found in patients with an acute coronary syndrome treated with dual antiplatelet therapy with clopidogrel compared with ticagrelor, despite a higher absolute risk with the latter. ¹⁵ Taken together, these data suggest that ticagrelor alone does not portend an excess in major bleeding complications if compared with other antiplatelet monotherapies. Nonetheless, when used in combination with aspirin, ticagrelor exerts a significant additive prohemorrhagic effect, especially on the gastrointestinal tract. The hypothesis that an aspirin‐free strategy with ticagrelor, or P2Y₁₂ inhibitors in general, would guarantee adequate ischemic protection and improved safety has been originally tested among high bleeding risk subgroups (such as those with atrial fibrillation requiring dual antiplatelet therapy) and subsequently extended to broader patient populations undergoing percutaneous coronary intervention. ¹⁶ , ¹⁷ Whether this treatment strategy is safe and effective also for the long‐term secondary prevention of stable patients with established cardiovascular disease is yet to be proved, although signals of such benefit have already been shown. ¹⁸ Finally, findings by Magnani et al are consonant with prior analyses supporting the preeminent role of bleeding risk assessment in guiding intensified or prolonged antithrombotic regimens. ¹⁹ Ongoing studies will help refine the available tools that can inform clinicians of the balance between bleeding and ischemic risk of various treatment combinations relatively to the individual patient profile.

Disclosures

Dr Mehran reports grants from Abbott Laboratories, AstraZeneca, Bayer, Beth Israel Deaconess, Bristol Myers Squibb, CSL Behring, DSI, Medtronic, Novartis Pharmaceuticals, and OrbusNeich; personal fees from Abbott Laboratories, Boston Scientific, Medscape/WebMD, Siemens Medical Solutions, PLx Opco Inc/dba PLx Pharma Inc, Roivant Sciences, Sanofi, Medtelligence (Janssen Scientific Affairs), and Janssen Scientific Affairs; other from Abbott Laboratories, Abiomed, Bristol Myers Squibb, Claret Medical, Elixir Medical, The Medicines Company, Spectranetics/Philips/Volcano Corp, and Watermark Research Partners; and nonfinancial support and other from Regeneron Pharmaceuticals and Idorsia Pharmaceuticals Ltd. Dr Dangas reports consulting fees or honoraria from AstraZeneca, Biosensors, Boston Scientific, and Medtronic; reports research grants to the institution from Biotronik and Abbott Laboratories; and has equity (entirely divested) with Medtronic. Dr Cao has no disclosures to report.

(J Am Heart Assoc. 2021;10:e019889. DOI: 10.1161/JAHA.120.019889.)