Article Text
Abstract
Background Previous reports on smokers' paradox to clopidogrel have only been able to show an association between cigarette smoking and enhanced response to clopidogrel therapy. No study has shown reversal of enhanced clopidogrel response after smoking cessation.
Objective To conduct a prospective observational longitudinal study in order to measure the impact of cigarette smoking on on-clopidogrel platelet reactivity (OPR).
Design From the prospective CROSS-VERIFY cohort, 810 subjects with repeated measurement of OPR at least 1 month apart were analysed. With smoking status ascertained at two time points, baseline and follow-up, study subjects were categorised into never smokers (n=628), smoking quitters (n=77) and persistent smokers (n=105). Dependent variables included OPR measured by the VerifyNow assay and the percentage of subjects with high OPR (HOPR).
Results At baseline, current smokers showed significantly lower OPR compared with never smokers, with no significant differences in OPR between future quitters and future persistent smokers within current smokers. While the OPR of never smokers and persistent smokers did not change significantly during the follow-up, the mean OPR of quitters increased significantly by 19 P2Y12 reaction units (p=0.013). The frequency of HOPR showed similar results, with an 8–10% increase in smoking quitters in contrast to no significant changes in never and persistent smokers. Both mean OPR and the frequency of HOPR showed a linear inverse relationship with the amount of smoking.
Conclusions Enhanced clopidogrel response in smokers is reversed after smoking discontinuation, suggesting a causal relationship in addition to the previously reported association between smoking and enhanced clopidogrel response.
- Antiplatelet treatment
- smoking
- clopidogrel
- coronary intervention
- gene association
- platelets
- atherosclerosis
- interventional cardiology
- SPASM
- molecular biology
- restenosis
Statistics from Altmetric.com
- Antiplatelet treatment
- smoking
- clopidogrel
- coronary intervention
- gene association
- platelets
- atherosclerosis
- interventional cardiology
- SPASM
- molecular biology
- restenosis
Introduction
Cigarette smoking has been shown to be associated with enhanced clopidogrel response, resulting in increased platelet inhibition when taking clopidogrel.1–7 Furthermore, studies have shown that smokers achieve greater clinical benefit compared with non-smokers from clopidogrel therapy,8 9 a phenomenon also known as the ‘smokers’ paradox'. Although the mechanism is unclear, we among others have reported that the altered activity of the hepatic cytochrome P450 (CYP) system, in particular the induced activity of CYP1A2 enzymes by cigarette smoking, may lead to an elevated level of active clopidogrel metabolites and enhanced pharmacodynamic effects of clopidogrel.10 11 However, the major limitation of the previous reports was that they were all based on cross-sectional design that, at best, could only confirm significant correlation between smoking and increased clopidogrel response: that is, in patients treated with clopidogrel, smokers compared with non-smokers had greater pharmacodynamic effects. To further dissect the causal relationship between smoking and clopidogrel response, we need to find out whether the enhanced clopidogrel response in smokers is actually reversed after smoking cessation, or whether non-smokers have an enhanced clopidogrel response when they become active smokers, the latter of which would be an impossible option to study due to ethical issues. Therefore, we conducted a prospective observational longitudinal study to measure the impact of cigarette smoking on clopidogrel response.
Methods
Study design
The CROSS-VERIFY cohort (measuring clopidogrel resistance to assure safety after percutaneous coronary intervention using VerifyNow) is a prospective cohort including all patients undergoing coronary angiography or percutaneous coronary intervention (PCI) who agreed to the measurement of on-clopidogrel platelet reactivity (OPR) with the VerifyNow P2Y12 assay at Seoul National University Hospital (Seoul, Korea). Written informed consent for study participation was obtained from each study patient before enrolment. Exclusion criteria were: contraindication to aspirin, clopidogrel or heparin; the use of intravenous glycoprotein IIb/IIIa inhibitor within 5 days before the clopidogrel reactivity test; the concomitant use of cilostazol; uncontrolled malignancy; bleeding tendency; and ethnicity other than Korean heritage.
The current study was a prospective observational study. From the CROSS-VERIFY cohort, never and current smokers were identified at baseline. We extracted patients who repeated VerifyNow P2Y12 measurements with an interval at least 1 month, from June 2006 to August 2010. As smoking status and amount of smoking were ascertained again at follow-up with questionnaires, current smokers were sequentially subdivided into smoking quitters and persistent smokers. Among the three study arms (never smokers, quitters and persistent smokers), the smoking quitters was the only group that fell into a crossover arm. Data that are not fixed over the course of time such as age, laboratory findings and concomitant medications were collected at each time point.
The study complied with Declaration of Helsinki and was approved by the Institutional Review Board of Seoul National University Hospital.
Platelet function test
The inhibitory effect of clopidogrel on platelet reactivity was measured using the VerifyNow P2Y12 assay (Accumetrics, San Diego, California, USA). A blood sample was obtained 12–24 h after the final dose of clopidogrel in patients who had been on 75 mg of clopidogrel for more than 7 days, and 12–24 h after PCI in patients who were loaded with clopidogrel before catheterisation. A loading dose of 300 mg of clopidogrel was administered to patients who had been taking clopidogrel for <7 days; 600 mg was given to clopidogrel-naïve patients. For follow-up, blood sampling was recommended at 2–4 h after the last dose of clopidogrel. Whole blood was anticoagulated in a sodium citrate bottle used exclusively for the VerifyNow P2Y12 assay. All patients took aspirin at 100 mg per day or 300 mg loading if not taken previously. The VerifyNow assay is a point-of-care test allowing for a rapid platelet reactivity measurement based on a turbidimetric optical detection system. The fibrinogen-coated microparticles aggregate in whole blood in proportion to the number of expressed platelet glycoprotein IIb/IIIa receptors. The P2Y12-mediated pathway was measured by optical signal change, and the results were reported in P2Y12 reaction units (PRU). Technical details and reliability of the assay have been reported previously.12 13 The coefficient of variation for the test was 7.5% in our institution.
Definitions
Study subjects were categorised into three groups according to their smoking status ascertained at baseline and follow-up: ‘never smokers’, ‘quitters’ and ‘persistent smokers’. Never smokers were those who had never experienced cigarette smoking and had not been smoking during the follow-up period. Previous smokers at baseline were excluded in this study. Current smokers were defined as those who had smoked during the year prior to and were currently smoking cigarettes at baseline. Current smokers were divided into quitters and persistent smokers according to the smoking status at follow-up. Patients who smoked during the year preceding the baseline test as well as at any time during the follow-up period were considered persistent smokers. A quitter was defined as someone who had been a current smoker at baseline, but had been abstinent from cigarette smoking for at least 1 month prior to the follow-up. The usual definition of ‘previous smokers’ or ‘quitters’ used in clinical trials requires abstinence from smoking for at least 1 year,14 15 based on the findings that a patient's risk of acute myocardial infarction (AMI) reduces within 1 year of quitting smoking, and that approximately one-third to one-half of AMI patients begin smoking again within 6–12 months of their diagnosis.16 However, for the purpose of this study, we used the definition above since a study reported that enhanced CYP1A2 activity in smokers reaches a steady state within 1 week after smoking cessation (t1/2: 38.6 h).17
Major dependent variables were OPR and the frequency of patients with high OPR (HOPR). OPR was measured using the VerifyNow assay and was reported as PRU. HOPR was defined as OPR ≥235 PRU as previously proposed.18–20
Statistical analysis
Baseline characteristics were compared among the three groups, with both overall and pairwise comparisons. Continuous variables were presented as mean±SD, and compared with the use of one-way analysis of variance (ANOVA). Categorical variables were reported as frequencies, and compared with the use of the χ2 test and logistic regression for pairwise comparison. The generalised estimating equation (GEE) model was used to analyse the relationship between smoking status and the dependent variables.21 Each patient had two repeated measurements at two different time points, which were considered within-subject variables. This statistical method performs complex within-group and between-group comparisons of a 3×2 table with a single procedure, minimising the chance of false positivity due to multiple testing. Multivariable-adjusted analyses were also performed with the use of the GEE model, incorporating the covariates that were either fixed or not fixed over the time course into the model. A 2-sided p value <0.05 was considered significant for all tests. All statistical analyses were performed using R programming language (R Project for Statistical Computing, Vienna, Austria).
Results
Baseline characteristics
Figure 1 shows the scheme of this study population. A total of 810 patients were analysed in this study: 628 never smokers, and 182 current smokers at baseline. Among the latter, 77 eventually quit cigarette smoking, and 105 were smoking persistently at follow-up. Table 1 compares the baseline characteristics of the three groups. Baseline characteristics of quitters and persistent smokers did not differ significantly. Compared to never smokers, current smokers (including quitters and persistent smokers) were younger, and more likely to be male. They had a lower incidence of hypertension and previous coronary intervention history. In terms of laboratory findings, current smokers had higher levels of haemoglobin and triglyceride and a lower serum level of HDL-cholesterol. Administered medications were mostly similar, except for angiotensin receptor blockers and dihydropyridine calcium channel blockers.
Smoking and OPR
Study subjects repeated VerifyNow measurements at an interval of 2.9 months (IQR 1.0 to 6.2 months). Figure 2 shows the distribution of OPR at baseline and follow-up (left and right panels, respectively), both of which followed the normal distribution. OPR at baseline and follow-up in the entire group of patients were highly correlated (R2=0.386; p by linear regression <0.001; see supplementary figure 2), and did not differ significantly (236±86 vs 234±80; p by paired t-test=0.578).
When the patients were divided into three groups according to their smoking status, however, there were significant differences among groups and across times. The results are detailed in table 2 and figure 3A. At baseline, current smokers had lower OPR compared with never smokers (mean difference in OPR, 43±7), with no significant differences between future quitters and future persistent smokers. During the follow-up duration, the means of OPR of never smokers and persistent smokers did not change significantly, while that of quitters increased significantly by 19 PRU. The change in OPR over time of quitters (Δ +22 (95% CI +7 to +38)) differed significantly from that of never smokers (Δ −7 (95% CI −14 to +1); interaction p=0.007) and persistent smokers (Δ −3 (95% CI −16 to +10); interaction p=0.010).
Since gender distribution was unbalanced across the groups among the baseline features and female gender is known to be a significant predictor of high mean OPR,2 4 a separate set of analyses was performed with only the male patients (lower rows of table 2, right panel of figure 3). Similar trends were observed, while the gap in PRU between never smokers versus current smokers decreased to 30. Whereas quitters and persistent smokers had similar means of OPR at baseline, OPR increased by a mean of 22 at follow-up only in smoking quitters.
Multivariable-adjusted analyses showed similar trends (figure 4). After multivariable adjustment, the most significant factors for OPR were shown to be age, sex, hypertension, haemoglobin, and triglyceride along with smoking status, whereas other environmental factors including calcium channel blockers and statins had no significant effects on OPR.
Smoking and frequency of HOPR
As a whole population, the frequency of HOPR was similar at baseline and follow-up (55.3% vs 52.6%; p by McNemar test=0.137). Table 3 compares the frequency of HOPR according to smoking status across time. Concordant to the results of the mean OPR, the rate of HOPR was significantly higher in never smokers compared with current smokers at baseline. At follow-up, the frequency of HOPR increased only in quitters by 7.8%, with no significant change in never smokers and persistent smokers. The amount of increase was significant only when analysed in male subjects. At follow-up, the difference in the rate of HOPR was significant between quitters and persistent smokers, but not between quitters and never smokers.
Relation between smoking amount and OPR
To evaluate the possible association between the amount of smoking and OPR, patients were grouped into non-smokers, smokers with <20 cigarettes per day (1–19 cigarettes a day), and smokers with ≥20 cigarettes per day at each time point. Figure 5A shows that there exists an inverse linear relationship between the amount of smoking and its effect on clopidogrel OPR (p by linear regression <0.001 for both total population and men only). Among quitters, however, the increase in PRU accompanied by smoking cessation did not show significant differences according to their baseline smoking amount (17±71 vs 21±68 for <20 (n=17) and ≥20 cigarettes per day (n=58), respectively; p by t-test=0.829). HOPR frequency also decreased with increased smoking amount and showed an inverse relationship (p for linearity=0.001) (figure 5B).
Discussion
In this study, we found that smoking cessation resulted in reversal of the enhanced response to clopidogrel in smokers. At baseline, current smokers had significantly lower OPR compared with never smokers. Among those who were actively smoking at baseline, those who stopped smoking thereafter (quitters) showed a significant increase in OPR, in contrast to persistent smokers whose OPR did not change significantly at follow-up. Consequently, the OPR of quitters at follow-up approximated to that of never smokers. Accordingly, an additional 8–10% of smokers were categorised as HOPR after quitting smoking, whereas the proportion of HOPR in never smokers and persistent smokers did not change significantly. Furthermore, we found that there was a linear relationship between the amount of cigarette smoking and the antiplatelet effect of clopidogrel.
HOPR while receiving clopidogrel has been shown to be associated with an increased risk of thrombotic events.22 23 A variety of genetic and environmental factors are known to affect response to clopidogrel. Loss-of-function polymorphism of CYP2C19*2 represents a genetic factor that is associated with a more frequent occurrence of HOPR.24 Calcium channel blockers, in addition, are reported to impair platelet inhibition by clopidogrel.25 26 Increased response to clopidogrel found in smokers is another example.1–7
To our knowledge, this is the first study to show the reversal of the enhanced clopidogrel effect seen in smokers after smoking cessation, suggesting that there may be a causal relationship between smoking and enhanced clopidogrel response. The phenomenon that cigarette smoking is associated with enhanced inhibition of platelet aggregation by clopidogrel has been repeatedly shown in previous studies.1–7 Moreover, smokers have been shown to benefit more from clopidogrel therapy than non-smokers in terms of angiographic and clinical outcomes.8 9 By observing serial changes in OPR, this study strengthens the current evidence for the effect of smoking on the response to clopidogrel therapy.
The design of this study has several advantages over previous studies. First, longitudinal observation confers better evidence for the temporal relationship that smoking cessation preceded the elevation in OPR. Second, the influence of potential confounding factors can be reduced by serially following a study subject from being a current smoker to a quitter. Since baseline characteristics can be quite unbalanced between smokers and non-smokers, as shown in this study as well as in others,5 8 9 a cross-sectional study design is subject to sampling bias that cannot be completely overcome despite multivariable analyses. Third, by comparing with control groups, we could conclude that the temporal increase in OPR was a finding unique in smoking quitters.
We, among others, have reported that the hepatic CYP1A2 enzyme activities induced by smoking could be a mechanism behind the smokers' paradox.1 2 5 6 8 9 Polycyclic aromatic hydrocarbons present in tobacco smoking are known to induce CYP1A2 activity, mediated through the aryl hydrocarbon receptor, a transcriptional factor. Many drugs have been reported to have interactions with smoking, such as theophylline, melatonin and clozapine.27 Although controversial,28 29 several lines of evidence support the assumption that the altered CYP1A2 activity mediates between smoking and the increased response to clopidogrel therapy. First, it is plausible in terms of drug metabolism. While smoking increases the clearance and thus lowers the plasma levels of the drugs listed above, the antiplatelet effect of clopidogrel, a prodrug that needs to be metabolised into active metabolites by the CYP450 system, could be potentiated by cigarette smoking. This study result supports this theory. Second, a dose–response relationship was shown to be present between the smoking amount and the decrease in OPR in this study, as well as in previous ones.1 9 Third, it has an interaction with genetic variation. A study by our own group has shown the ‘smokers' paradox’ was prominent only in CYP1A2 A-allele carriers, but not in the CC genotype.5 The latter two findings are consistent with previous studies on CYP1A2 activities.30–32
This study should not be misinterpreted as discouraging smoking cessation. Smoking is one of the single most important avoidable risk factors of coronary heart disease, and smoking cessation is known to reduce the risk of reinfarction and death within 1 year.33 Current guidelines strongly recommend complete cessation of smoking and no exposure to environmental tobacco smoke for patients with cardiovascular disease.34 This study implies that the OPR measured at baseline could fail to precisely predict the actual risk of thrombotic events in a patient who quits smoking. In these patients, repeat measurement of OPR could be an option. Patients who are recommended to quit smoking need to be closely monitored as to whether or not the behavioural change would lead to a surge in their platelet activity or, in turn, thrombotic events especially during the immediate post-PCI period, a critical time for stent thrombosis. However, considering disappointing results from recent clinical trials, which tested tailored antiplatelet therapy using the platelet function test,35 36 we believe further investigation is warranted to clarify its clinical implication.
Limitations
The major limitation is that this study did not investigate clinical outcomes. It cannot be determined from this study whether the elevation in OPR after smoking cessation could lead to attenuation of the benefits from clopidogrel therapy observed in smokers.8 9 Second, is the observational nature of this study. While all the patients were encouraged to quit smoking, this study was performed in a manner of per-protocol analysis. Third, there could be a selection bias. Only 810 from 1848 candidates met the criteria, and were finally analysed in this study. Fourth, as smoking status was ascertained with self report, and was not verified with a biomarker test such as urinary cotinine levels, there could have been under-reporting. Finally, we did not have a ‘non-smoking → smoking’ arm.
Conclusion
The enhanced inhibition of platelet aggregation by clopidogrel observed in cigarette smokers diminished after smoking cessation. This study provides additional evidence for a possible causal relationship between an environmental factor and the effect of clopidogrel. Further investigation is warranted to elucidate the clinical implications.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Download Supplementary Data (PDF) - Manuscript file of format pdf
- Download Supplementary Data (PDF) - Manuscript file of format pdf
Footnotes
See Editorial, p 963
Funding This study was supported by a grant of the Korea Healthcare Technology R&D Project, Ministry of Health and Welfare, Republic of Korea (A102065), and a grant from the Innovative Research Institute for Cell Therapy, Seoul National University Hospital (A062260), sponsored by the Ministry of Health and Welfare, Republic of Korea.
Competing interests None.
Patient consent Obtained.
Ethics approval This study was conducted with the approval of the Seoul National University Hospital Institutes for Biomedical Research.
Provenance and peer review Not commissioned; externally peer reviewed.