The efficacy and safety of cangrelor in single vessel vs multivessel percutaneous coronary intervention: Insights from CHAMPION PHOENIX

Abstract Background The intravenous, rapidly acting P2Y12 inhibitor cangrelor reduces the rate of ischemic events during PCI with no significant increase in severe bleeding. However, the efficacy and safety of cangrelor compared with clopidogrel in patients treated with single vessel (SV)‐percutaneous coronary intervention (PCI) or multivessel (MV)‐PCI remains unexplored. Methods We studied the modified intention‐to‐treat population of patients from the CHAMPION PHOENIX trial who were randomized to either cangrelor or clopidogrel. We used logistic regression and propensity score matching to evaluate the effect of cangrelor compared with clopidogrel on the primary efficacy outcome (composite of death, myocardial infarction, ischemia‐driven revascularization, or stent thrombosis) at 48 hours. The safety outcome was moderate or severe Global Utilization of Streptokinase and tPA for Occluded Arteries bleeding at 48 hours. Hypothesis Cangrelor is as efficacious and safe as clopidogrel in both SV and MV PCI. Results Among 10 854 patients, 9204 (85%) underwent SV‐ and 1650 (15%) MV‐PCI. After adjustment, cangrelor was associated with similar reductions vs clopidogrel in the primary efficacy outcome in patients undergoing SV‐PCI (4.5% vs 5.2%; odds ratio [OR] 0.81 [0.66‐0.98]) or MV‐PCI (6.1% vs 9.8%, OR 0.59 [0.41‐0.85]; Pint 0.14). Similar results were observed after propensity score matching (SV‐PCI: 5.5% vs 5.9%, OR 0.93 [0.74‐1.18]; MV‐PCI: 6.2% vs 8.9%, OR 0.67 [0.44‐1.01]; Pint 0.17). There was no evidence of heterogeneity in the treatment effect of cangrelor compared with clopidogrel for the safety outcome. Conclusions In patients undergoing SV‐ or MV‐PCI, cangrelor was associated with similar relative risk reductions in ischemic complications and no increased risk of significant bleeding compared with clopidogrel, which highlights the expanding repertoire of options for use in complex PCI.

Methods: We studied the modified intention-to-treat population of patients from the CHAMPION PHOENIX trial who were randomized to either cangrelor or clopidogrel. We used logistic regression and propensity score matching to evaluate the effect of cangrelor compared with clopidogrel on the primary efficacy outcome (composite of death, myocardial infarction, ischemia-driven revascularization, or stent thrombosis) at 48 hours. The safety outcome was moderate or severe Global Utilization of Streptokinase and tPA for Occluded Arteries bleeding at 48 hours.
Hypothesis: Cangrelor is as efficacious and safe as clopidogrel in both SV and MV PCI. There was no evidence of heterogeneity in the treatment effect of cangrelor compared with clopidogrel for the safety outcome.
Conclusions: In patients undergoing SV-or MV-PCI, cangrelor was associated with similar relative risk reductions in ischemic complications and no increased risk of significant bleeding compared with clopidogrel, which highlights the expanding repertoire of options for use in complex PCI.

K E Y W O R D S
cangrelor, clopidogrel, multivessel percutaneous coronary intervention

| INTRODUCTION
Rapid advances in percutaneous coronary intervention (PCI) tools and techniques are allowing operators to tackle progressively more challenging coronary revascularization procedures via a percutaneous approach. Notably, there has been an increase of multivessel (MV)-PCI to address MV-coronary disease as an alternative to coronary artery bypass grafting. 1 However, MV-PCI is associated with increased complexity and risk. 2,3 To mitigate periprocedural ischemic PCI complications, adjuvant pharmacotherapy has evolved.
The CHAMPION PHOENIX trial evaluated the efficacy and safety of cangrelor compared with clopidogrel among patients undergoing PCI for indications ranging from stable angina to all forms of acute coronary syndrome. Cangrelor reduced the rate of ischemic events at 48 hours compared with clopidogrel, without a significant increase in severe bleeding. 4 The efficacy and safety of cangrelor compared with clopidogrel was similar in patients with single vessel (SV)-disease and in patients with  However, the risks and benefits of MV-PCI as a revascularization strategy in combination with either cangrelor or clopidogrel are still unknown. As such, we aimed to evaluate the efficacy and safety of cangrelor compared with clopidogrel among patients actually treated with SV-PCI or with MV-PCI in the CHAMPION PHOENIX trial.

| METHODS
The rationale and design of the CHAMPION PHOENIX trial have been detailed previously 6 and are summarized briefly here.

| Participants
The CHAMPION PHOENIX trial included men and women aged 18 or older who required PCI for stable angina and acute coronary syndrome.
For this study, we included patients enrolled in the CHAMPION PHOE-NIX trial who underwent SV-or MV-index PCI as defined using core angiographic data. Patients in whom procedures were performed in more than one vessel (left main (LM), left anterior descending (LAD), left circumflex (LCX), right coronary artery (RCA)) were defined as having MV-PCI; patients in whom procedures were performed in only one vessel (whether or not they had multivessel disease) were defined as having SV-PCI. If a procedure was staged, it was only considered MV-PCI if multiple vessels were treated in a single setting. Patients who had LM PCI were considered to have MV-PCI. A central clinical events committee adjudicated all angiographic data.

| Study Treatment
Prior to PCI but after angiography, patients were randomized to receive either cangrelor or clopidogrel in a double-dummy, doubleblind manner. Patients received either a cangrelor infusion (initial bolus 30 mcg/kg) plus placebo capsules or placebo followed by a loading dose of clopidogrel (300 or 600 mg, either before or after PCI, per site investigator). After loading, patients in the cangrelor arm received a 4 mcg/kg/min infusion of cangrelor for at least 2 hours or until the procedure was complete, whichever was longer.
All patients received aspirin (75 to 325 mg). All patients received 75 mg clopidogrel as maintenance therapy. Periprocedural use of bivalirudin, unfractionated heparin, low molecular weight heparin, or fondaparinux was allowed at the discretion of the site investigator, with a glycoprotein IIb/IIIa inhibitor as an option only for rescue therapy.

| Efficacy outcomes
We considered the same primary efficacy outcome for this analysis as in CHAMPION PHOENIX, defined as the composite of death from any cause, myocardial infarction (MI), ischemia-driven revascularization (IDR), or stent thrombosis (ST) at 48 hours after randomization.
We also evaluated the same composite outcome at 30 days as the secondary efficacy analysis. A central clinical event committee adjudicated all suspected events of the efficacy outcome. 5

| Safety outcomes
The primary safety outcome for the purposes of this analysis was defined as moderate or severe Global Utilization of Streptokinase and tPA for Occluded Arteries bleeding at 48 hours after randomization (due to the small number of events of the pre-specified primary safety endpoint of severe Global Utilization of Streptokinase and tPA for Occluded Arteries bleeding in the original trial). According to prespecified definitions, bleeding end points were reported by the blinded site investigators and not centrally adjudicated. 7

| Statistical analysis
The analyses were conducted in the prespecified modified intentionto-treat (mITT) population, which included all randomized patients who underwent PCI and received at least one dose of study drug.
Univariable analysis was used to compare demographic and clinical characteristics between the two procedure groups (SV-vs MV-PCI). We used logistic regression techniques to assess efficacy at 48 hours and 30 days. Our primary analysis was logistic regression of outcome on treatment group, procedure group (SV-or MV-PCI), the interaction between treatment group and procedure group, and with adjustment for selected characteristics in the original cohort. We selected baseline demographic and clinical characteristics that were significantly different (two-sided significance level of 0.01) between participants undergoing SV-vs MV-PCI (sex, history of diabetes, access, bare-metal stent, drugeluting stent, clopidogrel loading dose (600 vs 300), region, clinical presentation (stable angina vs acute coronary syndrome), bifurcation treated), and characteristics that were deemed clinically relevant (time to PCI procedure from study drug administration, lesion calcification, lesion tortuosity, and PCI duration).
We performed a propensity analysis to minimize variance in baseline demographic and clinical characteristics of patients in the two procedure groups, since MV or SV-PCI was a post randomization event. We estimated the propensity score using logistic regression where we regressed MV-PCI on sex, history of diabetes, bifurcation treated, presentation, and clinically relevant characteristics listed above; we did not match on access, use of bare metal or drug-eluting stent, or periprocedural clopidogrel dose because they were significantly associated with region. Patients were matched in a 1:4 ratio on propensity score; we did an exact match for region and used a 5% caliper matching for propensity score for the other variables.

| 48-hour outcomes
In the unadjusted analysis, there was no heterogeneity in treatment effect associated with cangrelor compared with clopidogrel observed

| Safety Outcomes
For moderate/severe Global Utilization of Streptokinase and tPA for Occluded Arteries bleeding, rates were low (<1%) with no heterogeneity in the treatment effect associated with cangrelor compared with clopidogrel for SV-vs MV-PCI either in the unadjusted, the adjusted, or the propensity score matched analysis ( Figure 3).

| DISCUSSION
This study is the first to provide two notable findings: While recent studies suggest that MV-PCI is safe, 9,[11][12][13][14]  Note that these findings are tempered by the (a) overall low event rates, and (b) potential treatment effect: patients with bleeding complications during their first PCI may not proceed to a second PCI.
There are limitations to this study. First, this was not a prespecified subgroup analysis. PCI was a postrandomization variable; as such, while we performed propensity matching to attempt to account for numerous potential confounders, it is possible that other confounders persist. For example, if an ischemic complication occurred during the first PCI, the decision might be made to not proceed to a second PCI. As a result, we could be underestimating the treatment effect in planned MV-PCI. Secondly, while extensive core angiographic data did provide details on tortuosity and calcification, some measures of the complexity of PCI or the overall clinical risk of each procedure (ranging from stable angina to ST elevation Myocardial Infarction presentation) were not systematically assessed. Fractional flow reserve was not used to assess the functional significance of disease in each vessel, which has been shown to differentiate necessity of intervention better than angiographic appearance alone. 17 Additionally, while 45% of patients who received SV-PCI had MV-disease, we did not systematically collect information why exactly MV-PCI was not pursued in those patients and the degree to which this may have resulted in incomplete revascularization. 18 As such, our results reflect a cumulative summary of outcomes rather than the likely heterogeneity within each PCI procedure group.
These findings among a contemporary cohort of patients across the full spectrum of acute coronary syndrome suggest that cangrelor compared with clopidogrel is associated with a similar reduction in the composite of death, MI, IDR, and ST at 48 hours in patients undergoing PCI with consistent results in patients treated with SV-and MV-PCI without an associated increased risk of severe bleeding.