Prolonged vs Short Duration of Dual Antiplatelet Therapy After Percutaneous Coronary Intervention in Patients With or Without Peripheral Arterial Disease: A Subgroup Analysis of the PRODIGY Randomized Clinical Trial

Key Points

Question  What is the ideal duration of dual antiplatelet therapy (DAPT) after coronary stenting in patients with concomitant coronary and peripheral arterial disease?

Findings  In this subanalysis of a randomized clinical trial, peripheral arterial disease emerged as a treatment modifier in terms of DAPT duration in patients undergoing percutaneous coronary intervention, with a prolonged regimen (up to 24 months) associated with a significantly lower risk of death and atherothrombotic events and unaffected risk of bleeding.

Meaning  The significant prognostic effect of concomitant peripheral arterial disease should be taken into account when managing the duration of DAPT after percutaneous coronary intervention.

Importance  Patients with concomitant peripheral arterial disease (PAD) experience worse cardiovascular outcomes after percutaneous coronary intervention (PCI).

Objective  To assess the efficacy and safety of prolonged (24 months) vs short (≤6 months) dual antiplatelet therapy (DAPT) in patients with PAD undergoing PCI.

Design, Setting, and Participants  This subanalysis of the randomized Prolonging Dual Antiplatelet Treatment After Grading Stent-Induced Intimal Hyperplasia Study (PRODIGY) trial assessed unselected patients from tertiary care hospitals with stable coronary artery disease or acute coronary syndromes with or without concomitant PAD from December 2006 to December 2008. Data analysis was performed from January 7 to April 4, 2016.

Interventions  Percutaneous coronary intervention.

Main Outcomes and Measures  Rates of the primary efficacy end point, composite of death, myocardial infarction, or cerebrovascular accidents, and occurrence of the key safety end point, a composite of Bleeding Academic Research Consortium type 2, 3, or 5.

Results  This analysis comprised 246 and 1724 patients with and without PAD, respectively. In the patients with PAD, mean (SD) age was 73.2 (9.2) in the prolonged group and 75.7 (8.7) years in the short DAPT group, and 97 (82.2%) were male in the prolonged group and 92 (71.9%) were male in the short DAPT group. In the patients without PAD, mean (SD) age was 67.1 (11.2) years in the prolonged group and 66.8 (11.3) years in the short DAPT group, and 667 (76.8%) were male in the prolonged group and 655 (76.6%) were male in the short DAPT group. Status of PAD was associated with a higher risk of death and ischemic events (hazard ratio [HR], 2.80; 95% CI, 2.05-3.83; P < .001). Prolonged vs short DAPT conveyed a lower risk of the primary efficacy end point in patients with PAD (19 [16.1%] vs 35 [27.3%]; HR, 0.54; 95% CI, 0.31-0.95; P = .03) but not in patients without PAD (81 [9.3%] vs 63 [7.4%]; HR, 1.28; 95% CI, 0.92-1.77; P = .15), with positive interaction (P = .01). The risk of definite or probable stent thrombosis was significantly lower in patients with PAD treated with prolonged compared with short DAPT (HR, 0.07; 95% CI, 0-1.21; P = .01). Bleeding Academic Research Consortium type 2, 3, or 5 bleeding occurred in 6 patients with PAD (5.2%) receiving prolonged DAPT relative to 8 (6.9%) of those receiving short DAPT (HR, 0.77; 95% CI, 0.27-2.21; P = .62), with a significant interaction (P = .04) compared with patients without PAD.

Conclusions and Relevance  Peripheral artery disease confers a poor prognosis in patients undergoing PCI in the setting of stable coronary artery disease or acute coronary syndromes. Prolonged DAPT lowers the risk of ischemic events with no apparent bleeding liability in this high-risk group.

Trial Registration  clinicaltrials.gov Identifier: NCT00611286.

Concomitant peripheral arterial disease (PAD) is increasingly recognized as an important risk factor among patients with coronary artery disease (CAD).^1,2 Clinically relevant atherosclerosis that affects the peripheral vascular circulation is detected in up to 40% of patients with CAD,^3-5 with an adverse effect on clinical outcomes.^6-8 Among patients with CAD who require percutaneous coronary intervention (PCI), those with concomitant PAD have a 2-fold increase in the early and long-term risk of death.²

Antiplatelet therapy is pivotal to prevent cardiovascular events in patients with isolated PAD.^9,10 Evidence suggests that extended duration of dual antiplatelet therapy (DAPT) after PCI provides more effective protection against atherothrombotic events compared with short-term regimens, at the risk of more frequent bleeding.¹¹ Whether the duration of DAPT in patients undergoing PCI should be tailored based on concomitant PAD is currently unknown, to our knowledge. In a survey initiated by the European Association of Percutaneous Cardiovascular Interventions that assessed DAPT prescription patterns after coronary stenting, concomitant PAD was not identified as a clinically important factor affecting the decision regarding treatment duration in 1134 responders from 92 countries.¹² Against this background, we performed a subgroup analysis of the Prolonging Dual Antiplatelet Treatment After Grading Stent-Induced Intimal Hyperplasia Study (PRODIGY) trial¹³ to evaluate the effect of PAD on ischemic and bleeding events and to assess the efficacy and safety of a prolonged (24-month) vs a short (≤6-month) DAPT duration according to the presence of concomitant PAD.

The randomized, multicenter, open-label PRODIGY trial compared the safety and efficacy profile of prolonged vs short duration of DAPT in a minimally selected population of patients undergoing coronary stenting from December 2006 to December 2008.¹³ Data analysis was performed from January 7 to April 4, 2016. All participants gave written informed consent, and their data were deidentified. The Cardiology Department at University of Ferrara (Ferrara, Italy) led the study, and the protocol was independently approved by the ethics committees of the participating centers. The study design and primary results of the PRODIGY trial have been previously described¹⁴ and are summarized in the eMethods in the Supplement.

For the purpose of this intention-to-treat analysis, the study population was stratified according to PAD status. At baseline, patients with intermittent claudication or prior amputation or percutaneous or surgical peripheral revascularization were identified as having PAD. All patients were followed up through clinic visits at 30 days and up to 2 years after PCI.

Definitions of study outcomes have been previously reported.¹³ The primary efficacy end point was a composite of death, myocardial infarction (MI), or cerebrovascular accident (CVA) up to 2 years. The primary key safety end point included a composite of type 2, 3, or 5 bleeding according to the Bleeding Academic Research Consortium (BARC) criteria.¹⁵ Other end points were death, cardiovascular death, MI, and stent thrombosis, defined according to the Academic Research Consortium criteria.¹⁶ A net adverse clinical events (NACE) end point combining the primary efficacy end point of death, MI, or CVA with BARC type 3 or 5 bleeding was additionally analyzed. All study end points were adjudicated by a clinical events committee unaware of treatment allocation.

Categorical baseline variables were presented as frequencies and percentages with P values calculated with unpaired t tests, χ² tests, or Fisher exact tests. Continuous variables were expressed as mean (SD) and compared with the independent-samples t test. The effect of PAD on clinical outcomes was evaluated by Cox proportional hazards regression analyses by adjusting for baseline variables associated with the primary efficacy and key safety end point at the univariate analysis with a significance level of P < .20. The efficacy and safety of prolonged DAPT vs short DAPT for patients with vs without PAD was evaluated at 24 months and from 6 to 24 months. Clinical events were expressed as counts with rates computed according to the Kaplan-Meier method. Cox proportional hazards regression analysis was used to calculate hazard ratios (HRs) with 95% CIs, and an interaction test was provided to evaluate the effect of treatment in patients with vs without PAD. As sensitivity analysis, the efficacy and safety of different durations of DAPT in patients with PAD were evaluated by adjusting for baseline variables that significantly differ between patients randomized to prolonged vs short DAPT. Analyses were performed with STATA statistical software, release 13 (StataCorp).

Of 1970 patients enrolled in the PRODIGY trial, a history of PAD was present in 246 participants (12.5%). Among the PAD group, 118 and 128 patients were randomized to the prolonged and short DAPT groups, respectively. A total of 869 and 855 patients without PAD were randomized to the prolonged and short DAPT groups, respectively (Table).

Patients with PAD were older and more often had hypertension, type 1 or type 2 diabetes, previous MI, and previous coronary artery bypass grafting than patients without PAD. In addition, they were more likely to present with non–ST-segment elevation MI, had more complex CAD, and more frequently underwent multivessel intervention. At 30 days, patients with PAD were more often taking diuretics, whereas the use of β-blockers and statins was higher among patients without PAD (eTable 1 and eTable 2 in the Supplement).

Patients with PAD randomized to prolonged DAPT were younger, had a higher body mass index, and less frequently underwent PCI of the left main coronary artery than those allocated to short DAPT. Otherwise, the baseline and periprocedural characteristics were well matched between the study groups (Table and eTable 3 in the Supplement).

Patients with PAD experienced higher rates of ischemic events at 2-year follow-up compared with patients without PAD: the composite of death, MI, or CVA occurred in 54 patients with PAD (21.9%) vs 144 patients without PAD (8.4%) (adjusted HR, 1.75; 95% CI, 1.26-2.44; P = .001). Mortality was higher among patients with PAD (39 [15.8%] vs 91 patients without PAD [5.3%]; adjusted HR, 1.87; 95% CI, 1.25-2.79; P = .002); cardiac death was specifically reported in 24 patients with PAD (10.0%) compared with 49 patients without PAD (2.9%) (adjusted HR, 2.21; 95% CI, 1.30-3.76; P = .003) (Figure 1 and eFigure 1A in the Supplement).

BARC type 2, 3, or 5 bleeding was not increased among patients with PAD compared with those without PAD (14 [6.0%] vs 93 [5.5%]; adjusted HR, 0.75; 95% CI, 0.42-1.33; P = .32) (Figure 1 and eFigure 1B in the Supplement). Similar results were obtained by applying other bleeding definitions. The risk of NACE was significantly increased in patients with PAD compared with those without PAD (54 [21.9%] vs 168 [9.8%]; adjusted HR, 1.50; 95% CI, 1.08-2.08; P = .02).

As shown in Figure 2 and Figure 3, among patients with PAD, prolonged DAPT was associated with a lower risk of the primary efficacy end point compared with short DAPT (19 [16.1%] vs 35 [27.3%]; HR, 0.54; 95% CI, 0.31-0.95; P = .03). In contrast, no significant difference was found in the risk of the primary end point between the 2 DAPT regimens among patients without PAD (81 [9.3%] in the prolonged-duration regimen vs 63 [7.4%] in the short-duration regimen; HR, 1.28; 95% CI, 0.92-1.77; P = .15), resulting in a significant qualitative interaction (P = .01). This difference was mainly driven by a significant interaction for death (P = .006), which was lower in the group of patients with PAD treated with prolonged DAPT (12 [10.2%] in the prolonged-duration regimen vs 27 [21.1%] in the short-duration regimen; HR, 0.45; 95% CI, 0.23-0.88; P = .02) compared with patients without PAD (53 [6.1%] in the prolonged-duration regimen vs 38 [4.5%] in the short-duration regimen; HR, 1.39; 95% CI, 0.91-2.10; P = .12). In the landmark analysis, the interaction test for the occurrence of the primary efficacy end point remained significant (P = .02), with a lower risk of events in patients with PAD allocated to prolonged DAPT compared with patients in the short DAPT group (15 [13.2%] vs 25 [21.2%]; HR, 0.59; 95% CI, 0.31-1.12; P = .10) (eTable 4 in the Supplement).

Definite or probable stent thrombosis was significantly reduced in patients with PAD randomized to prolonged DAPT compared with short DAPT (0 [0%] vs 7 [6.0%]; HR, 0.07; 95% CI, 0-1.21; P = .01), with significant interaction (P = .03) compared with patients without PAD, in whom prolonged DAPT had no effect (13 [1.5%] vs 8 [1.0%]; HR, 1.62; 95% CI, 0.67-3.90; P = .28) (Figure 3).

Baseline clinical characteristics of patients experiencing stent thrombosis with PAD or without PAD were comparable (eTable 5 in the Supplement), although multivessel disease and multiple treated lesions were more frequently observed in patients with PAD (eTable 6 in the Supplement).

The cumulative time-to-event curves for the key safety end point are shown in Figure 2. BARC type 2, 3, or 5 bleeding was not significantly affected by DAPT duration in the PAD cohort (6 [5.2%] in the prolonged-duration regimen vs 8 [6.9%] in the short-duration regimen; HR, 0.77; 95% CI, 0.27-2.21, P = .62), whereas patients without PAD receiving prolonged DAPT experienced higher rates of bleeding events according to BARC and other prespecified criteria compared with patients in the short DAPT group (Figure 4). These findings were confirmed in the landmark analysis with focus between 6 and 24 months (eTable 4 in the Supplement).

Cumulative incidence curves of NACE according to DAPT duration among patients with and without PAD are shown in eFigure 2 in the Supplement. The risk of NACE was significantly decreased in the prolonged DAPT group compared with the short DAPT group in the PAD population (19 [16.1%] in the prolonged-duration regimen vs 35 [27.3%] in the short-duration regimen; HR, 0.54; 95% CI, 0.31-0.95; P = .03). Among patients without PAD, NACE occurred in 97 patients (11.2%) in the prolonged DAPT group and 71 patients (8.3%) in the short DAPT group (HR, 1.36; 95% CI, 1.00-1.85; P for interaction = .047).

As reported in eTable 7 in the Supplement, among patients with PAD presenting with stable CAD, no significant difference in the rate of the primary efficacy end point at 2 years was found between those receiving prolonged and short DAPT (4 [13.8%] vs 6 [15.8%], P = .82). Among patients with PAD and acute coronary syndromes at presentation, the primary end point arose in 15 patients (16.9%) randomized to prolonged DAPT compared with 29 (32.2%) of those randomized to short DAPT (HR, 0.46; 95% CI, 0.25-0.87; P = .02; P for interaction = .36). BARC type 2, 3, or 5 bleeding were not significantly different between the prolonged and short DAPT groups in patients with PAD with stable CAD (2 [6.9%] vs 2 [5.4%], P = .77) or acute coronary syndromes (4 [4.6%] vs 6 [7.7%], P = .44).

After adjustment for age, body mass index, and left main PCI, the adjusted HR for the primary efficacy end point was 0.66 (95% CI, 0.37-1.16; P = .15), with positive interaction testing (P = .05). Furthermore, prolonged DAPT in patients with PAD was still associated with a reduction in the composite of death or MI (16 [13.6%] in the prolonged-duration regimen vs 35 [27.3%] in the short-duration regimen; adjusted HR, 0.51; 95% CI, 0.28-0.94; P = .02) and overall stent thrombosis (6 [5.2%] in the prolonged-duration regimen vs 21 [17.2%] in the short-duration regimen; adjusted HR, 0.33; 95% CI, 0.13-0.82; P = .01).

The salient findings of the present analysis are as follows. First, in patients undergoing PCI, concomitant PAD was associated with a 2-fold increased risk of ischemic events, whereas the risk of bleeding was unaffected. Second, prolonged DAPT duration of 24 months after PCI reduced the risk of the primary efficacy end point of death, MI, or CVA compared with short DAPT of 6 months or less in patients with PAD. The improved efficacy of prolonged DAPT in patients with PAD was not offset by an increased risk of actionable bleeding episodes.

Atherosclerosis of peripheral vessels portends a higher burden of CAD, accounting for impaired survival and worse cardiovascular outcomes. In a pooled analysis of 19 867 patients from 8 randomized clinical trials, PAD was a significant predictor of the composite of death, MI, and target-vessel revascularization 6 months after PCI (HR, 1.17; 95% CI, 1.05-1.31; P = .005) and was independently associated with an increased risk of mortality throughout follow-up.² Among outpatients with stable CAD included in the REACH (Reduction of Atherothrombosis for Continued Health) registry, the 1-year rate of cardiovascular death, MI, stroke, or hospitalization for atherothrombotic events was 23.1% in patients with concomitant PAD compared with 13% among patients with isolated PAD and 17% among patients with isolated CAD.¹⁷ Consistently, the GRACE (Global Registry of Acute Coronary Events) registry reported higher in-hospital mortality among patients with acute coronary syndromes and concomitant PAD, with a 6-month rate of major cardiovascular events of 14.6% compared with 7.2% in patients without PAD.⁸

Several pathophysiologic and clinical observations support the detrimental association of CAD and PAD. A multivessel coronary involvement, along with a proinflammatory status, is more common among patients with concomitant CAD and PAD.^18,19 Moreover, the onset of clinical symptoms and the identification of CAD are frequently camouflaged by PAD owing to the impaired functional status of patients.^6,20 Finally, the underestimation of this high-risk condition in clinical practice accounts for a well-described tendency to withhold pharmacologic therapies for secondary prevention in these patients.²¹ Our study lends further support to the notion of an increased risk of ischemic events in patients with CAD and concomitant PAD. More specifically, the risks of NACE and cardiovascular mortality were more than doubled among patients with concomitant PAD.

Although patients with PAD experienced higher rates of ischemic events, they did not have a parallel increase in the risk of bleeding. These results are difficult to interpret because they conflict with other studies,^6,22,23 hence, potentially reflecting a power issue. Nonetheless, because we assessed bleeding events starting from 30 days after PCI, it may be possible that the exclusion of periprocedural events, which are more common in patients with PAD,²⁴ may explain the lack of association noted in our study between PAD status and bleeding risk. In addition, the high burden of mortality observed in patients with PAD may have camouflaged the occurrence of bleeding events. In 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) study, PAD status was not associated with bleeding, and ticagrelor compared with placebo did not result in a higher risk of major or minor bleeding at both the 60-mg and 90-mg doses.²⁵ Furthermore, there is evidence demonstrating an increased platelet reactivity in patients with multisite atherosclerosis as a consequence of the larger amount of diseased endothelium and diffuse high shear stress.²⁶ An inverse association between platelet reactivity and bleeding exists.^27,28

Antiplatelet therapy has the potential to mitigate the risk of atherothrombotic events in patients with PAD. In the CAPRIE (Clopidogrel vs Aspirin in Patients at Risk of Ischemic Events) trial,²⁹ clopidogrel monotherapy conveyed higher cardiovascular ischemic protection and low major bleeding in patients with stable PAD vs aspirin. The CHARISMA (Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization, Management, and Avoidance) trial¹⁰ tested the benefit of DAPT with aspirin and clopidogrel in patients with established stable atherosclerotic vascular disease or multiple atherothrombotic risk factors. Overall, DAPT was not superior to aspirin monotherapy for the primary end point of death, MI, or stroke (534 of 7802 [6.8%] vs 773 of 7801 [7.3%], P = .22). However, DAPT was associated with lower rates of MI and hospitalization for ischemic events without increasing the risk of major bleeding among patients with PAD. The addition of vorapaxar, an antagonist of protease-activated receptor 1 of thrombin, to aspirin and clopidogrel significantly reduced acute limb ischemia and peripheral revascularization but increased the risk of serious (intracranial) bleeding among patients with PAD included in the TRA2°P-TIMI 50 trial (Trial to Assess the Effects of SCH 530348 in Preventing Heart Attack and Stroke in Patients With Atherosclerosis).³⁰

To our knowledge, this is the first study to evaluate the safety and efficacy of different DAPT durations in patients with concomitant CAD and PAD after PCI. Although in the overall population of the PRODIGY trial prolonged DAPT was not more effective, the relative risk of the primary efficacy end point was approximately halved by prolonged DAPT in patients with PAD with an interaction testing suggesting heterogeneity of treatment effect according to PAD status. Despite a high residual burden of risk in this patient population (the composite end point still occurred in roughly 16% of patients allocated to prolonged DAPT), a positive interaction was found for all-cause mortality, cardiovascular mortality, and definite or probable stent thrombosis. These results should be interpreted in the context of the DAPT study¹¹ that demonstrated a significantly reduced risk of stent thrombosis and major adverse cardiovascular and cerebrovascular events among 9961 patients randomly assigned to continue thienopyridine treatment or to receive placebo beyond 1 year after PCI.

Nevertheless, the implementation of prolonged DAPT in clinical practice remains controversial because of the higher risk of bleeding invariably reported in studies^11,14 comparing different DAPT durations. Although prolonged DAPT increased the bleeding risk compared with short DAPT among patients without PAD, we did not observe a significant effect of prolonged DAPT on bleeding risk among patients with PAD. The disconnect between PAD status and bleeding risk remains difficult to explain and could have been in part related to the exceedingly high incidence of ischemic events and the modest sample size. However, it is consistent with previous reports.^10,25 Overall, our results should be considered hypothesis generating for future dedicated randomized clinical trials.

We acknowledge the following limitations. First, this was not a prespecified analysis and therefore has all the potential shortcomings of such studies, including limited number of patients and events. Furthermore, because the randomization was not stratified by PAD status, our results may be attributable to differences in baseline variables. Second, PAD status was ascertained solely on the basis of clinical history; therefore, underreporting cannot be excluded. Third, we were unable to indicate the peripheral vascular district primarily diseased and the rates of subsequent limb-related outcomes. Furthermore, we did not ascertain the safety and efficacy of different DAPT durations according to the clinical presentation (claudication, previous peripheral revascularization) of PAD. Fourth, DAPT in the PRODIGY trial was based on the use of aspirin and clopidogrel administration. Therefore, whether new P2Y12 receptor inhibitors may further improve clinical outcomes in this high-risk subset remains unanswered. This issue will be ascertained in part by the ongoing EUCLID (Examining Use of Ticagrelor in PAD) trial, testing the superiority of ticagrelor compared with clopidogrel monotherapy in reducing the risk of cardiovascular events among 13 500 patients with PAD (clinicaltrials.gov, NCT01732822).³¹

The coexistence of PAD and CAD conveyed higher risks of cardiovascular morbidity and mortality among patients undergoing PCI. A prolonged DAPT duration (up to 24 months) resulted in a lower risk of atherothrombotic events, including a mortality benefit, than short DAPT. The apparent neutral effect of longer DAPT on bleeding risk of patients with PAD requires further evaluation in adequately powered studies.

Corresponding Author: Marco Valgimigli, MD, PhD, Department of Cardiology, Bern University Hospital, 3010 Bern, Switzerland (marco.valgimigli@insel.ch).

Accepted for Publication: June 26, 2016.

Published Online: August 30, 2016. doi:10.1001/jamacardio.2016.2811

Author Contributions: Dr Valgimigli had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Franzone, Piccolo, Ariotti, Santucci, Baldo, Valgimigli.
Acquisition, analysis, or interpretation of data: Piccolo, Gargiulo, Marino, Magnani, Moschovitis, Windecker, Valgimigli.
Drafting of the manuscript: Franzone, Valgimigli.
Critical revision of the manuscript for important intellectual content: Piccolo, Gargiulo, Ariotti, Marino, Santucci, Baldo, Magnani, Moschovitis, Windecker, Valgimigli.
Statistical analysis: Franzone, Piccolo, Gargiulo.
Obtaining funding: Windecker.
Administrative, technical, or material support: Franzone, Valgimigli.
Study supervision: Marino, Baldo, Windecker, Valgimigli.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Windecker has reported receiving research grants to the Department of Cardiology, Bern University Hospital, from Abbott, Biotronik, Boston Scientific, Biosensors, Medtronic, Edwards, and St Jude. No other disclosures were reported.

Funding/Support: Dr Franzone is supported by a research grant from Società italiana di Cardiologia/Merck, Sharp, and Dohme Corporation. Dr Piccolo is supported by a research grant from Fondazione Umberto Venoresi. Dr Gargiulo is supported by a research grant from the European Association of Percutaneous Coronary Intervention.

Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.