Author + information
- Received December 9, 2015
- Revision received April 4, 2016
- Accepted April 6, 2016
- Published online July 25, 2016.
- Jennifer Yu, MBBSa,b,
- Usman Baber, MDb,
- Ioannis Mastoris, MDb,
- George Dangas, MD, PhDb,
- Samantha Sartori, PhDb,
- Philippe Gabriel Steg, MDc,
- David J. Cohen, MD, MScd,
- Gennaro Giustino, MDb,
- Jaya Chandrasekhar, MBBSb,
- Cono Ariti, MSce,
- Bernhard Witzenbichler, MDf,
- Timothy D. Henry, MDg,
- Annapoorna S. Kini, MDb,
- Mitchell W. Krucoff, MDh,
- C. Michael Gibson, MDi,
- Alaide Chieffo, MDj,
- David J. Moliterno, MDk,
- Antonio Colombo, MDj,
- Stuart Pocock, PhDe and
- Roxana Mehran, MDb,∗ ()
- aPrince of Wales Clinical School, University of New South Wales, Randwick, Australia
- bIcahn School of Medicine at Mount Sinai, New York, New York
- cHôpital Bichat-Claude Bernard, Paris, France
- dSaint Luke’s Mid America Heart Institute, University of Missouri–Kansas City, Kansas City, Missouri
- eLondon School of Hygiene and Tropical Medicine, London, United Kingdom
- fCharité Hospital, Berlin, Germany
- gMinneapolis Heart Institute Foundation, University of Minnesota, Minneapolis, Minnesota
- hDuke University School of Medicine, Durham, North Carolina
- iHarvard Medical School, Cambridge, Massachusetts
- jSan Raffaele Hospital, Milan, Italy
- kUniversity of Kentucky, Lexington, Kentucky
- ↵∗Reprint requests and correspondence:
Dr. Roxana Mehran, Icahn School of Medicine at Mount Sinai, One Gustave L. Levy Place, Box 1030, New York, New York 10029.
Objectives The aim of this study was to compare the incidence and impact of cessation of dual-antiplatelet therapy (DAPT) in women and men treated with percutaneous coronary intervention.
Background Nonadherence to cardiovascular medications and female sex are associated with worse outcomes. However, the patterns and impact of DAPT cessation in women compared with men following percutaneous coronary intervention have not been studied.
Methods Baseline characteristics, patterns of DAPT cessation, and 2-year clinical outcomes were compared in 5,031 patients (1,279 women, 3,739 men) enrolled following successful percutaneous coronary intervention with stents in the PARIS (Patterns of Non-Adherence to Antiplatelet Regimens in Stented Patients) study. DAPT cessation was adjudicated as physician-guided discontinuation, interruption for surgery, or disruption due to bleeding or noncompliance. Clinical endpoints were major adverse cardiac events (a composite of cardiac death, definite or probable stent thrombosis, spontaneous myocardial infarction, or clinically indicated target lesion revascularization), a second restricted definition of major adverse cardiac events excluding target lesion revascularization, and bleeding.
Results DAPT cessation was more common in women than men (59.1% vs. 55.9%, p = 0.007) and comprised increased rates of discontinuation, disruption for bleeding, and disruption due to noncompliance. The impact of DAPT cessation was similar regardless of sex and varied according the mode; in particular, disruption was associated with increased risk for both ischemic and bleeding events. After adjusting for differences in baseline and treatment characteristics as well as DAPT cessation events, female sex remained an independent predictor of bleeding but not of ischemic events.
Conclusions DAPT cessation was more common in women, but its impact was similar in women and men. Female sex was an independent predictor of bleeding but not of ischemic events after adjustment for differences in DAPT cessation and baseline and treatment characteristics.
Dual-antiplatelet therapy (DAPT) is almost universally prescribed following percutaneous coronary intervention (PCI) in the absence of contraindications (1). Previous studies have suggested lower medication adherence among women compared with men, as well as worse long-term outcomes in nonadherent patients (2–4). Furthermore, women have been shown to have increased adverse events following PCI, including increased risk for bleeding (5,6), which arguably affects both the duration of DAPT prescribed by physicians as well as patient adherence to DAPT. However, the patterns of DAPT cessation among women and men and the contribution of DAPT cessation to the observed risks associated with female sex post-PCI are not known.
The PARIS (Patterns of Non-Adherence to Antiplatelet Regimens in Stented Patients) study demonstrated significant variation in the risk for DAPT cessation according to the specific mode of cessation, with disruptions in therapy (for bleeding or noncompliance) being associated with markedly increased risk for major adverse cardiac events (MACE =) compared with no increased risk with temporary interruptions or physician-guided discontinuations of DAPT (7). We therefore sought to compare the patterns and impact of cessation of DAPT in women and men who have undergone PCI with stents.
The methods and main findings of the prospective, multicenter PARIS study were previously published (7). In brief, patients who were prescribed DAPT following successful PCI with placement of at least 1 stent between July 2009 and December 2010 were eligible for enrollment. Patients were excluded if they underwent PCI for stent thrombosis (ST) or if they were participants in an investigational drug or device trial. Patients were followed for 2 years, with contact at 1, 6, 12, and 24 months to ascertain the occurrence of any clinical adverse events as well as any cessations of DAPT. All events, including cessations of DAPT, were adjudicated by an external clinical events committee.
In the present study, we compared the baseline characteristics, treatment, DAPT cessation, and clinical outcomes of the women and men enrolled in the PARIS study.
In the present analysis, MACE included cardiac death, definite or probable ST, spontaneous myocardial infarction (MI), or clinically indicated target lesion revascularization. A secondary restricted MACE endpoint comprised cardiac death, definite or probable ST, and spontaneous MI only. Bleeding, unless otherwise specified, referred to bleeding events meeting criteria for Bleeding Academic Research Consortium (BARC) type ≥3 (8).
ST was adjudicated according to the Academic Research Consortium definition (9). Spontaneous MI was defined as raised biomarkers in the presence of supportive clinical or electrocardiographic indicators per the (first) universal definition published in 2007 (10). In addition the BARC criteria, all bleeding events were also adjudicated using the TIMI (Thrombolysis in Myocardial Infarction) and ACUITY (Acute Catheterization and Urgent Intervention Triage Strategy) definitions (11,12).
Each episode of DAPT cessation was adjudicated as 1 of the following 3 “modes”: 1) discontinuation, including cessation of 1 or both components of DAPT by a treating physician because of lack of benefit or increased harm from ongoing therapy; 2) interruption, cessation of ≤14 days’ duration of 1 or both components of DAPT for surgery or an invasive procedure; and 3) disruption, further adjudicated as physician-recommended cessation of antiplatelet treatment (i.e., because of bleeding) or nonrecommended (i.e., patient noncompliance, not directed by the physician). In patients with multiple episodes of DAPT cessation during the 2-year follow-up period, each episode of cessation was considered individually. All episodes of DAPT cessation and clinical endpoints were independently adjudicated by an external committee.
Categorical variables are presented as number (percentage) and were compared using chi-square tests. Continuous variables are presented as mean ± SD and were compared using Student t tests. Time–to–DAPT cessation and time-to-event data are presented as Kaplan-Meier estimates and were compared using log-rank tests. Time to the first DAPT cessation event (of any mode) and time to the first DAPT cessation of each specific mode (i.e., discontinuation, interruption, and disruption) during the 2-year follow-up period were considered separately.
Cox regression models were used to examine the association between DAPT cessation events and adverse events in men and women separately. Covariates included the following baseline and treatment characteristics: age, body mass index, diabetes, hypertension, smoking status, stroke, peripheral vascular disease, prior MI or PCI, education level, region of enrollment, presentation with acute coronary syndrome, and stent type used. DAPT cessation was included as a time-updated categorical variable. Formal interaction testing was then performed to compare these associations in women versus men. Using the same covariates, Cox regression models were also used to examine the relationship between sex and 2-year DAPT discontinuation and 2-year DAPT disruption.
To assess to what extent of DAPT cessation contributed to the sex-associated risks for adverse events following PCI, the association between female (vs. male) sex and adverse events were adjusted in 2 separate Cox regression models with sex forced into the model: first, with the same baseline and treatment characteristics listed previously as covariates; second, with the same covariates as the first and the addition of DAPT cessation as a time-updated covariate.
All p values are 2 tailed. No adjustment for multiple testing was made. Statistical analyses were performed using SAS version 9.3 (SAS Institute, Cary, North Carolina) and Stata version 12.1 (StataCorp LP, College Station, Texas).
Of 5,018 patients enrolled in the PARIS study, 1,279 were women (25.5%) and 3,739 were men (74.5%). The mean duration of follow-up was 677 ± 151 days in women and 686 ± 140 days in men (p = 0.06). There were multiple differences in the baseline characteristics (Table 1). Women were more likely to be older, with a marginally higher body mass index, and to have histories of stroke, transient ischemic attack, peripheral vascular disease, hypertension, or diabetes and family histories of premature coronary artery disease. Men were more likely to have completed tertiary or higher education, to have histories of prior MI and PCI, to be smokers, and to present with silent ischemia.
In general, women and men received similar treatment, but men were more likely to have thrombotic lesions and to receive marginally longer stents (Table 2). Women were more likely to receive bivalirudin intraprocedurally and to be prescribed clopidogrel (rather than an alternative thienopyridine) and a proton pump inhibitor on discharge.
Compared with men, women were more likely to have a DAPT cessation event during the 2-year follow-up period (Figure 1). There was a spike in cessation in women and men due to increased DAPT discontinuation from 12 months post-PCI. By mode, women had higher rates of discontinuation, which reached borderline significance. Disruption was significantly more common in women than men, whereas the rate of interruption was similar. Both physician-recommended (for bleeding) and nonrecommended (i.e., patient noncompliance) disruption were more common in women than men (Figure 2). Female sex was a significant predictor of both 2-year discontinuation (hazard ratio [HR]: 1.14; 95% confidence interval [CI]: 1.02 to 1.27) and disruption (HR: 1.20; 95% CI: 1.01 to 1.42) after multivariate adjustment. Other variables independently associated with discontinuation or disruption included age, enrollment in Europe compared with the United States, education level, diabetes, smoking status, history of coronary artery disease, presentation with an acute coronary syndrome, and treatment with a drug-eluting (vs. bare-metal) stent (Online Figure 1).
Impact of DAPT cessation in women versus men
Table 3 shows the results of the subgroup analyses comparing the association between DAPT cessation and subsequent clinical events for each sex. In general, the associations were similar in direction and magnitude among women and men, and results of formal interaction testing were not significant.
Disruption was associated with significantly increased risk for ischemic adverse events in both women and men. Women with disruption had almost doubled risk for all-cause mortality (HR: 1.93; 95% CI: 1.10 to 3.39); there was a trend toward increased mortality in men (HR: 1.6; 95% CI: 0.998 to 2.57). Discontinuation was associated with reduced rates of the restricted secondary MACE endpoint in men; the HR was similar among women, but the 95% CI crossed 1.
Interruption was associated with a 6-fold increased risk for subsequent bleeding in women (HR: 6.11; 95% CI: 2.73 to 13.70) and a trend toward increased bleeding in men (HR: 2.2; 95% CI: 0.99 to 4.88). Similarly, disruption was associated with significantly increased bleeding in women (HR: 3.78; 95% CI: 2.03 to 7.02) and a trend toward increased bleeding in men (HR: 1.82; 95% CI: 0.97 to 3.42).
Risk for adverse events associated with sex
Although rates of the primary MACE outcome were similar between men and women, women had significantly higher rates of the more restricted secondary MACE endpoint, all-cause mortality, and bleeding at 2-year follow-up (Online Table 1, Online Figure 2). Women also had higher rates of cardiac death and bleeding events classified as BARC ≥3 (study definition), BARC ≥2, TIMI minor, or ACUITY major, but bleeding events classified as TIMI major occurred similarly in women and men (Online Table 1). Rates of spontaneous MI, target vessel revascularization, target lesion revascularization, and ST were also similar in men and women.
Figure 3 shows the impact of female (vs. male) sex on the risk for adverse events following successful PCI before and after multivariable adjustment for: 1) differences in baseline and treatment characteristics; and 2) differences in baseline characteristics, treatment, and DAPT cessation events. After adjusting for baseline and treatment characteristics, there was no significant association between female sex and MACE or death. Adjusting for DAPT cessation in addition to baseline and treatment characteristics had minimal incremental impact. Female sex remained a predictor of BARC ≥3 bleeding in both multivariate modes (fully adjusted model: HR: 1.39; 95% CI: 1.02 to 1.89; p = 0.04). Results pertaining to other variables independently associated with outcomes are shown in Online Table 2.
In the current sex-based analysis from the large, prospective PARIS study, we found that: 1) women had higher rates of cessation of their DAPT than men in the 2 years following successful PCI, which comprised higher rates of physician-guided discontinuation, disruption due to bleeding, and disruption due to patient noncompliance; 2) the impact of each mode of DAPT cessation on subsequent adverse events was similar among women and men; and 3) female sex remained an independent predictor of BARC ≥3 bleeding but not ischemic events after multivariate adjustment for differences in baseline and treatment characteristics and the occurrence of DAPT cessation events.
Several studies have reported higher rates of nonadherence to prescribed medications in women compared with men (13–17). One study that examined this issue in a PCI population (PCI for acute coronary syndrome) using a validated adherence questionnaire found no association between female sex and medication adherence (18). However, most of these studies measured nonadherence by a medication possession ratio (the ratio of the days with medication in possession [calculated using dates prescriptions were filled] to the total number of days observed) <80%, and in contrast to the present study, nonadherence was not differentiated by cause (e.g., recommended by a physician, due to side effects of therapy, or representative of patient noncompliance).
To our knowledge, the present analysis is the first study to compare the rates of cessation of DAPT in women compared with men following successful PCI with stents. Consistent with previous studies of adherence to other cardiovascular medications, we found that women had higher rates of cessation compared with men. Of note, because all cessation events in the current study were adjudicated as specific modes according to the duration of and the reason for cessation, we were able to identify multiple contributing factors for higher cessation among women, including increased noncompliance, increased bleeding events on DAPT and physician-related factors.
We can postulate several reasons for the higher rates of noncompliance with DAPT in women compared with men in our study (10.9% vs. 8.5%; p = 0.01). The most common response among women surveyed regarding barriers to undertaking cardiovascular preventive measures was family obligations (19). Lower socioeconomic status, education level, and lack of social support may also contribute (20,21). Sex differences in the psychological reaction to the diagnosis of coronary artery disease may also play a role. A study comparing illness behavior in 1,300 women and men following cardiac surgery (for coronary or valvular disease) identified significantly higher rates of anxiety and depression among women, whereas disease denial was more common in men (22).
Several studies have shown that female sex is an independent predictor of increased bleeding following PCI (5,6,23–25). Similarly, the present study also demonstrated an association between female sex and 2-year bleeding, which persisted after adjustment. Of note, the present study identified that this increased risk for bleeding among women following PCI resulted in increased DAPT disruption (because of bleeding; 6.2% vs. 4.6%; p = 0.03) as well as indirectly contributing to higher rates of DAPT discontinuation compared with men (42.3% vs. 39.7%; p = 0.051).
The association between nonadherence (grossly defined as a medication possession ratio <80%) to cardiovascular medications and worse long-term outcomes is well established (2–4,14). Chowdhury et al. (2) performed a meta-analysis of 44 prospective epidemiological studies that reported the risk estimates of adherence to cardiovascular medications. Adherence to cardiovascular medications was protective with respect to cardiovascular disease (composite of fatal or nonfatal coronary heart disease, stroke, or sudden cardiac death; adherence to statins, HR: 0.85; 95% CI: 0.81 to 0.89; adherence to antihypertensive agents, HR: 0.81; 95% CI: 0.76 to 0.86) and all-cause mortality (adherence to statins, HR: 0.55; 95% CI: 0.46 to 0.37; adherence to antihypertensive agents, HR: 0.71; 95% CI: 0.64 to 0.78). Substudies of the CHARM (Candesartan in Heart Failure Assessment of Mortality and Morbidity) program showed that female sex was an independent predictor of nonadherence (i.e., women were less likely to be adherent) and that adherence was associated with lower mortality in all patients regardless of sex (HR: 0.65; 95% CI: 0.57 to 0.75) (4,14).
The important caveat to this from the primary analysis of the PARIS paper was that the clinical impact varied according to the mode of cessation, with the most harm associated with DAPT disruption (7). In the present study, we found that the impact of each mode of DAPT cessation was similar in men and women. Disruption was associated with similar magnitudes of increased risk in subsequent ischemic events and mortality regardless of sex.
The present study was a post hoc analysis of an observational, albeit large, prospective study, and our findings should be considered hypothesis generating. Furthermore, there were multiple differences in the baseline and treatment characteristics between men and women, and although we attempted to adjust for these differences, the effect of residual confounding cannot be excluded. Because the PARIS study enrolled patients in 2009 and 2010, the majority of patients were prescribed DAPT with aspirin and clopidogrel, and our findings may therefore not be generalizable to patients prescribed more novel P2Y12 inhibitors. Although PARIS is the largest study to date to prospectively examine DAPT nonadherence, we did not have enough cessation events to examine the predictors of interruption or DAPT discontinuation in the first year following PCI. We did not collect data regarding illness behavior, social supports, or family caregiver or financial responsibilities, or the data elements required to estimate frailty or bleeding risk by the use of a score, and were missing data related to anemia, renal and left ventricular function, and completeness of coronary revascularization, which may have provided further insight regarding sex-based differences in DAPT cessation patterns.
Women were more likely than men to have cessation of their DAPT in the 2 years following successful PCI, including discontinuation of DAPT by their treating physicians, disruption due to bleeding, and disruption due to patient noncompliance. The clinical impact of DAPT cessation was similar in men and women, with disruption being most significantly associated with subsequent ischemic and bleeding events. Adjusting for DAPT cessation in addition to differences in baseline and treatment characteristics had only a marginal impact on the associations between female sex and adverse outcomes, and female sex remained an independent predictor of bleeding but not of ischemic events in both multivariate models.
WHAT IS KNOWN? Women undergoing PCI have more comorbidities and higher complication rates following PCI, including bleeding, compared with men. Previous studies have also suggested lower medication adherence among women, which is also associated with worse long-term outcomes. However, the patterns of cessation of DAPT among women and men and the contribution of DAPT cessation to the observed risks associated with female sex post-PCI are not known.
WHAT IS NEW? Women had higher rates of DAPT cessation than men in the 2 years following successful PCI, which comprised higher rates of physician-guided discontinuation, disruption for bleeding, and disruption due to patient noncompliance. The impact of each mode of DAPT cessation was similar among women and men; in particular, disruption was associated with increased risk for both ischemic and bleeding events. Female sex was an independent predictor of bleeding but not of ischemic events after adjustment for differences in baseline and treatment characteristics as well as the occurrence of DAPT cessation events.
WHAT IS NEXT? The present study was a post hoc analysis, and our findings should be considered hypothesis generating. Large prospective studies are required to confirm these findings and to further explore the possible reasons for the increased DAPT cessation we observed in women compared with men.
The PARIS study was supported by research grants from Bristol-Myers Squibb and Sanofi. Dr. Steg has received research grants (to INSERM U1148) from Sanofi and Servier; has received speaking or consulting fees from Amarin, AstraZeneca, Bayer, Boehringer-Ingelheim, Bristol-Myers Squibb, CSL-Behring, Daiichi Sankyo, GlaxoSmithKline, Janssen, Eli Lilly, Novartis, Pfizer, Regeneron, Roche, Sanofi, Servier, and The Medicines Company; and holds stock with Aterovax. Dr. Cohen has received research grant support from Eli Lilly and AstraZeneca, consulting fees from Eli Lilly and AstraZeneca, and speaking honoraria from AstraZeneca. Dr. Gibson has received consulting fees from Bayer, Janssen Pharmaceuticals, Johnson & Johnson, and Portola Pharmaceuticals. Dr. Moliterno has received consulting fees from Janssen Pharmaceuticals and research grant support from AstraZeneca. Dr. Mehran has received research grant support from Eli Lilly, AstraZeneca, The Medicines Company, Bristol-Myers Squibb/Sanofi; has received consulting fees from AstraZeneca, Bayer, CSL-Behring, Janssen Pharmaceuticals, Merck, Osprey Medical, and Watermark Research Partners; and serves as a scientific advisory board member for Abbott Laboratories, Boston Scientific, Covidien, Janssen Pharmaceuticals, The Medicines Company, and Sanofi. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- Bleeding Academic Research Consortium
- confidence interval
- dual-antiplatelet therapy
- hazard ratio
- major adverse cardiac event(s)
- myocardial infarction
- percutaneous coronary intervention
- stent thrombosis
- Received December 9, 2015.
- Revision received April 4, 2016.
- Accepted April 6, 2016.
- American College of Cardiology Foundation
- Levine G.N.,
- Bates E.R.,
- Blankenship J.C.,
- et al.
- Chowdhury R.,
- Khan H.,
- Heydon E.,
- et al.
- Mehran R.,
- Pocock S.J.,
- Nikolsky E.,
- et al.
- Rao S.V.,
- McCoy L.A.,
- Spertus J.A.,
- et al.
- Mehran R.,
- Rao S.V.,
- Bhatt D.L.,
- et al.
- Cutlip D.E.,
- Windecker S.,
- Mehran R.,
- et al.
- Thygesen K.,
- Alpert J.S.,
- White H.D.,
- et al.
- Wiviott S.D.,
- Antman E.M.,
- Gibson C.M.,
- et al.
- Rolnick S.J.,
- Pawloski P.A.,
- Hedblom B.D.,
- Asche S.E.,
- Bruzek R.J.
- Mathews R.,
- Peterson E.D.,
- Honeycutt E.,
- et al.
- Mosca L.,
- Mochari-Greenberger H.,
- Dolor R.J.,
- Newby L.K.,
- Robb K.J.
- Mosca L.,
- Benjamin E.J.,
- Berra K.,
- et al.
- Hess C.N.,
- McCoy L.A.,
- Duggirala H.J.,
- et al.