Circulation Journal
Online ISSN : 1347-4820
Print ISSN : 1346-9843
ISSN-L : 1346-9843

この記事には本公開記事があります。本公開記事を参照してください。
引用する場合も本公開記事を引用してください。

Effect of Smoking Status on Clinical Outcome and Efficacy of Clopidogrel in Acute Coronary Syndrome
Masaki Kodaira, Hiroaki Miyata, Yohei Numasawa, Ikuko Ueda, Yuichiro Maekawa, Koichiro Sueyoshi, Shiro Ishikawa, Takahiro Ohki, Kouji Negishi, Keiichi Fukuda, Shun Kohsaka
著者情報
キーワード: Acute coronary syndrome, Clopidogrel, Percutaneous coronary intervention, Smoking, ST-segment elevation myocardial infarction
ジャーナル フリー HTML 早期公開

論文ID: CJ-16-0032

この記事には本公開記事があります。
Article (80-7 pp. 1590-1599)
詳細
PDFをダウンロード (645K)
メタデータをダウンロード RIS形式

(EndNote、Reference Manager、ProCite、RefWorksとの互換性あり)

BIB TEX形式

(BibDesk、LaTeXとの互換性あり)

テキスト
メタデータのダウンロード方法
発行機関連絡先
Abstract

Background: The “smoker’s paradox” is an otherwise unexplained phenomenon in which the mortality of smokers after acute myocardial infarction is reduced, contrary to expectations. It has been suggested that an association with antiplatelet agents exists, but the true mechanism remains largely unidentified.

Methods and Results: The analysis included 6,195 consecutive patients who underwent percutaneous coronary intervention (PCI) for acute coronary syndrome, registered in the Japanese multicenter PCI registry. Smokers were significantly younger and had less comorbidity than non-smokers. Unadjusted in-hospital mortality rate, general complication rate, and bleeding complication rate were lower in smokers than in non-smokers. After adjustment, the trend persisted and smoking was not associated with overall mortality (odds ratio [OR], 0.90; 95% confidence interval [CI]: 0.61–1.34; P=0.62), and was associated with lower overall (P=0.032) and bleeding complication events (P=0.040). Clopidogrel effectively reduced the occurrence of in-hospital complications and major adverse cardiac events in smokers compared with non-smokers (OR, 0.55; 95% CI: 0.53–0.98 vs. OR, 1.20; 95% CI: 0.87–1.67; and OR, 0.37; 95% CI: 0.20–0.70 vs. OR, 1.48; 95% CI: 0.90–2.43, respectively).

Conclusions: The smoker’s paradox was largely explained by confounding factors related to the lower risk profile of smokers, and they benefited from a positive modification of the efficacy of clopidogrel.

Cigarette smoking is considered to be responsible for one-half of all avoidable deaths in the industrialized world and causes more than 480,000 deaths each year in the USA. Exposure to even low levels of fine particles from secondhand cigarette smoke is sufficient to increase the risk of cardiovascular mortality. The rate of death from ischemic heart disease is 2–3-fold higher among current smokers than among those who have never smoked.1 Past reports, however, have suggested that habitual cigarette smokers have lower unadjusted mortality rates after acute myocardial infarction (MI), a phenomenon termed “smoker’s paradox”.2–4

Editorial p ????

Observational studies have been inconsistent; some reports suggest an increased rate of adverse events in non-smokers,2,5 whereas others refute this association.6–10 More recently, from a secondary data analysis of the CLARITY-TIMI 28 trial, it was suggested that smoking may modify the beneficial effects of clopidogrel, reducing the rate of cardiovascular death, MI, or urgent revascularization.11 Antiplatelet agents remain the cornerstone of acute coronary syndrome (ACS) therapy,12 and further research on the effect of smoking status on the efficacy of these thienopyridine antiplatelet agents is warranted. The CLARITY-TIMI 28 trial was focused on patients with ST-segment elevation myocardial infarction (STEMI); patients presenting with non-ST-segment elevation myocardial infarction (NSTEMI) or unstable angina were not included. Furthermore, there have been few analyses performed within the Asian population, which reportedly has a higher risk of bleeding complications.13 The positive effect of smoking on clopidogrel efficacy is yet to be examined in the whole entire ACS population.

The 2 main objectives of the present study were therefore to (1) evaluate the relationship between smoking status and in-hospital mortality or complications after ACS; and (2) analyze the relative efficacy of clopidogrel according to smoking status. The results of this study will help clinicians better understand the specific characteristics and clinical outcomes of patients presenting with ACS. This study will also aid in determining the relationship between smoking and efficacy of clopidogrel medication.

Methods

Study Design

The Japan Cardiovascular Database-Keio interhospital Cardiovascular Studies (JCD-KiCS) is a large, ongoing, prospective, multicenter cohort registry designed to collect clinical data on patients undergoing percutaneous coronary intervention (PCI).14 The JCD-KiCS began enrolling patients in September 2008, and the participating hospitals uploaded patient data into an Internet-based database system with the help of trained clinical research coordinators. Hospitals involved were instructed to record and register data from consecutive hospital visits for PCI using an Internet-based database system. PCI with any commercially available coronary device was included. The data entered were checked for completeness and internal consistency. Quality assurance of the data was achieved through automatic system validation and reporting of data completeness and with education for dedicated clinical research coordinators specifically trained for the present PCI registry. The senior study coordinator (I.U.), along with investigators (S.K. and H.M.) performing exclusive on-site audits, and ensured proper registration of each patient. The clinical variables and in-hospital outcomes for the JCD were defined in accordance with those specified for CathPCI Registry v4.1 (NCDR) in order to enable a direct comparison with the NCDR CathPCI Registry program, in which more than 1,500 centers participate in the USA.15

The participating hospitals were mostly large tertiary-care referral centers (>200 beds; n=12), but a few were midsized satellite hospitals (<200 beds; n=4). The database included patients with all ACS subtypes and stable ischemic heart disease. Patients were enrolled at the time of their first coronary event, with all consecutive PCI procedures during the study period recorded, including cases of failure.

The smoker subgroup included both current and recent smokers, defined as individuals who had smoked cigarettes any time during the year before arrival at the hospital where PCI was performed. Patients were classified as non-smokers if they had reported quitting smoking >1 year before arrival or were never smokers. Clopidogrel users were defined as patients who had received the medication within 24 h before PCI. If the patient was naïve to clopidogrel, use of clopidogrel meant treating with the loading dose of 300 mg according to current guidelines.16,17 For those patients who had a prior history of PCI and had been on home clopidogrel, however, daily dose of 75 mg clopidogrel was continued without loading. Patients with contraindications for clopidogrel were excluded from the study. Ticagrelor was not approved at the time of analysis, and prasugrel users were also excluded from the study.

Complications were defined as any of the following: severe coronary artery dissection or coronary perforation; MI after PCI; cardiac shock or heart failure; cerebral bleeding or stroke; and bleeding complications. Bleeding was defined as the presence of one or more of the following within 72 h of PCI: overt access site bleeding, retroperitoneal hemorrhage, gastrointestinal or genitourinary bleeding, cardiac tamponade, non-bypass surgery-related blood transfusion in patients with hemoglobin (Hb) ≤8 g/dl, or absolute decrease in Hb (≥3 g/dl) after PCI in patients with pre-procedure Hb <16 g/dl. This is equivalent to type 3 according to Bleeding Academic Research Consortium criteria.16

Figure 1 shows the study flowchart. The initial cohort included 15,023 consecutive patients who underwent PCI in the JCD-KiCS institutions between September 2008 and May 2014. Of those, 7,040 had undergone emergency PCI for ACS.

Figure 1.

Study flowchart. ACS, acute coronary syndrome; KiCS, Keio interhospital Cardiovascular Study; PCI, percutaneous coronary intervention.

Study Endpoints

The endpoints of this study were in-hospital outcomes after the index procedure, which included all-cause mortality and procedure-related complications. These endpoints were recorded in the database by the trained coordinators immediately after patient discharge. We studied the effect of smoking status on (1) mortality; (2) major adverse cardiac events (MACE; defined as the combination of death, cardiogenic shock, and heart failure); (3) bleeding; (4) any complication; (5) Thrombosis in Myocardial Infarction (TIMI) grade flow <3; (6) heart failure; and (7) cardiogenic shock. Multivariate logistic regression models were used to analyze the relationship between smoking status and adverse outcomes, after adjustment for baseline demographic and clinical history variables. Adjusted odds ratios (adj. OR) and 95% confidence intervals (95% CI) were derived from logistic regression analysis, adjusting for the following baseline variables: female sex, age >80 years, diabetes mellitus, hyperlipidemia, chronic kidney disease (CKD), STEMI, previous MI, cerebrovascular disease, 3-vessel disease, and transradial intervention. Next, we analyzed the relative effect of clopidogrel on in-hospital general complications and MACE, according to smoking status. In addition, we performed a subgroup analysis on patients who had not undergone PCI prior to admission.

Statistical Analysis

Continuous variables are expressed as mean±SD or as median (IQR) for non-normally distributed data. Categorical variables are expressed as frequencies and percentages. Patients were divided into 2 groups based on smoking status. Baseline characteristics were compared using Student’s t-test for normally distributed continuous variables and chi-squared test for categorical variables. All P-values were 2-sided, with a significance threshold of P<0.05. Multivariate logistic regression modeling was performed to determine independent predictors of study objectives with purposeful selection of covariates. Variables associated on univariate analysis (all with P≤0.1) and those judged to be of clinical importance were eligible for inclusion in the multivariate models. Results are reported as unit OR with associated 95% CI and P-values. Statistical analysis was performed using SPSS version 22.0 (SPSS, Chicago, IL, USA).

Results

Patients

Patient exclusions from analysis are shown in Figure 1. We excluded patients with cardiogenic shock (n=290), prasugrel prescription (n=5), and cardiopulmonary arrest on arrival (n=346), and also records with data entry errors or missing data (n=204). Hence, a total of 6,195 patients who underwent PCI for unstable angina, NSTEMI, or STEMI were included. A total of 3,792 patients were not current smokers and 2,403 were current smokers.

Clinical/Angiographic Characteristics at Index Procedure

Baseline clinical data for patients within the investigation cohort are summarized in Table 1. Non-smokers (mean age, 70.6±11.0 years), when compared with smokers (mean age, 62.0±11.7 years) were 8 years older (P<0.0001), more frequently female (P<0.0001), and more likely to have other risk factors for coronary artery disease or comorbidities such as hypertension (P<0.0001), diabetes mellitus (P=0.015), CKD (P<0.0001), or cerebrovascular disease (P<0.0001). Angiographic and procedural characteristics are summarized in Table 2. Among smokers, when compared with non-smokers, the incidence of 3-vessel disease was lower (P<0.0001).

Table 1. Baseline Characteristics vs. Smoking Status
  Smokers Non-smokers P-value
Total group
　n 2,403 3,792
　Age (years) 62.0±11.7 70.6±11.0 <0.0001
　Female 216 (9.0) 1,186 (31.3) <0.0001
　Hypertension 1,585 (66.0) 2,848 (75.1) <0.0001
　Hyperlipidemia 1,503 (62.5) 2,356 (62.1) 0.742
　Diabetes mellitus 830 (34.5) 1,425 (37.6) 0.015
　BMI (kg/m2) 23.98±3.76 24.02±3.70 0.666
　CKD 847 (36.3) 1,662 (45.2) <0.0001
　LVEF (%) 55.75±12.69 56.60±13.55 0.061
　Chronic lung disease 81 (3.4) 107 (2.8) 0.22
　Previous MI 294 (12.2) 613 (16.2) <0.0001
　Previous HF 113 (4.7) 247 (6.5) 0.003
　Hemodialysis 67 (2.8) 171 (4.5) 0.001
　CVD 157 (6.5) 373 (9.8) <0.0001
　PAD 122 (5.1) 218 (5.7) 0.258
　Previous PCI 344 (14.3) 873 (23.0) <0.0001
　Previous CABG 44 (1.8) 177 (4.7) <0.0001
　STEMI 1,238 (51.5) 1,503 (39.6) <0.0001
　NSTEMI 343 (14.3) 580 (15.3) 0.271
　Unstable angina 816 (34.0) 1,695 (44.7) <0.0001
De novo patients
　n 2,059 2,919
　Age (years) 61.4±11.7 70.5±11.3 <0.0001
　Female 185 (8.9) 897 (30.7) <0.0001
　Hypertension 1,293 (62.8) 2,115 (72.5) <0.0001
　Hyperlipidemia 1,224 (59.4) 1,690 (57.9) 0.274
　Diabetes mellitus 752 (36.5) 1,076 (36.9) 0.807
　BMI (kg/m2) 24.04±3.77 23.99±3.62 0.679
　CKD 674 (33.7) 1,189 (42.0) <0.0001
　LVEF (%) 56.04±12.44 56.79±13.33 0.133
　Chronic lung disease 65 (3.2) 79 (2.7) 0.35
　Previous MI 251 (12.1) 437 (15.0) <0.0001
　Previous HF 94 (4.6) 182 (6.2) 0.005
　Hemodialysis 36 (1.7) 84 (2.9) 0.011
　CVD 130 (6.3) 287 (9.8) <0.0001
　PAD 86 (4.2) 140 (4.8) 0.301
　Previous PCI 0 0 1
　Previous CABG 22 (1.1) 80 (2.7) <0.0001
　STEMI 1,158 (56.2) 1,358 (46.5) <0.0001
　NSTEMI 308 (15.0) 464 (15.9) 0.368
　Unstable angina 584 (28.6) 1,087 (37.2) <0.0001

Data given as mean±SD or n (%). BMI, body mass index; CABG, coronary artery bypass graft; CKD, chronic kidney disease; CVD, cerebrovascular disease; HF, heart failure; LVEF, left ventricular ejection fraction; MI, myocardial infarction; PAD, peripheral arterial disease; PCI, percutaneous coronary intervention; NSTEMI, non-ST-segment elevation myocardial infarction; STEMI, ST-segment elevation myocardial infarction.

Table 2. Procedural Characteristics vs. Smoking Status
  Smokers Non-smokers P-value
Total group
　n 2,403 3,792
　3-vessel disease 409 (17.0) 798 (21.0) <0.0001
　2-vessel disease 727 (30.3) 1,102 (29.1) 0.317
　LMS 42 (1.8) 109 (3.0) 0.004
　LAD 1,008 (43.3) 1,617 (44.7) 0.28
　DES 1,324 (57.5) 2,271 (63.8) <0.0001
　TFI 1,718 (71.5) 2,646 (69.8) 0.149
　TRI 648 (27.0) 1,077 (28.4) 0.219
　IABP 201 (8.4) 308 (8.1) 0.735
　Type C 569 (23.7) 936 (24.7) 0.369
　Bifurcation 571 (23.8) 981 (25.9) 0.062
　Aspirin 2,337 (97.3) 3,645 (96.1) 0.026
　Clopidogrel before PCI 1,766 (73.5) 2,734 (72.1) 0.242
De novo patients
　n 2,059 2,919
　3-vessel disease 333 (16.2) 586 (20.1) <0.0001
　2-vessel disease 599 (29.1) 802 (27.5) 0.212
　LMS 30 (1.5) 70 (2.5) 0.015
　LAD 891 (44.7) 1,316 (47.6) 0.053
　DES 1,110 (56.4) 1,706 (62.6) <0.0001
　TFI 1,503 (73.0) 2,066 (70.8) 0.087
　TRI 530 (25.7) 810 (27.7) 0.116
　IABP 176 (8.5) 246 (8.4) 0.881
　Type C 481 (23.4) 697 (23.9) 0.672
　Bifurcation 495 (24.0) 758 (26.0) 0.123
　Aspirin 2,004 (97.3) 2,799 (95.9) 0.007
　Clopidogrel before PCI 1,488 (72.3) 2,074 (71.1) 0.359

Data given as n (%). DES, drug-eluting stent; IABP, intra-aortic balloon pump; LAD, left anterior descending artery; LMS, left main stem; PCI, percutaneous coronary intervention; TFI, transfemoral intervention; TRI, transradial intervention.

In-Hospital Crude Outcomes

Unadjusted in-hospital event rates are provided in Table 3. In-hospital mortality (P=0.04), general complications (P=0.003), and bleeding complications (P=0.008) were significantly lower in smokers than in non-smokers. No significant differences were observed between smokers and non-smokers with regard to the occurrence of cardiogenic shock (P=0.549), heart failure complication (P=0.476), MACE (P=0.229), or TIMI flow <3 (P=0.448). Mean length of in-hospital stay was not significantly different between smokers (8.6±11.4 days) and non-smokers (8.5±11.4 days; P=0.775).

Table 3. In-Hospital Unadjusted Clinical Outcomes vs. Smoking Status
Outcome Smokers Non-smokers P-value
Total group
　n 2,403 3,792
　Death 41 (1.7) 95 (2.5) 0.04
　General complications 228 (9.5) 451 (11.9) 0.003
　Bleeding complications 65 (2.7) 150 (4.0) 0.008
　Cardiogenic shock 53 (2.2) 94 (2.5) 0.549
　Heart failure 62 (2.6) 110 (2.9) 0.476
　MACE 124 (5.2) 223 (5.9) 0.229
　Final TIMI flow grade <3 83 (3.5) 146 (3.9) 0.448
De novo patients
　n 2,059 2,919
　Death 35 (1.7) 76 (2.6) 0.033
　General complications 203 (9.9) 361 (12.4) 0.006
　Bleeding complications 59 (2.9) 121 (4.1) 0.017
　Cardiogenic shock 46 (2.2) 80 (2.7) 0.262
　Heart failure 58 (2.8) 92 (3.2) 0.496
　MACE 112 (5.4) 185 (6.3) 0.188
　Final TIMI flow grade <3 75 (3.8) 123 (4.5) 0.241

Data given as n (%). MACE, major adverse cardiac events; TIMI, Thrombosis in Myocardial Infarction.

In-Hospital Adjusted Outcomes

Predictors of in-hospital mortality, general complications, and bleeding complications are listed in Tables 4 – 6. On multivariate logistic regression modeling, in terms of in-hospital mortality, non-smokers and smokers were statistically equivalent (smokers vs. non-smokers, adj. OR, 0.90; 95% CI: 0.61–1.34; P=0.629), but for in-hospital complications (OR, 0.82; 95% CI: 0.69–0.98; P=0.032) and bleeding complications (OR, 0.72; 95% CI: 0.53–0.98; P=0.040), the smoker’s paradox was statistically evident.

Table 4. Indicators of In-Hospital Mortality
  Univariate Multivariate
OR 95% CI P-value OR 95% CI P-value
Total group
　Age >80 years 4.43 3.14–6.26 <0.0001 3.41 2.35–4.95 <0.0001
　Female gender 1.23 0.84–1.81 0.28      
　Hyperlipidemia 0.42 0.30–0.60 <0.0001 0.51 0.36–0.73 <0.0001
　Diabetes mellitus 0.83 0.57–1.19 0.322      
　CKD 3.09 2.14–4.47 <0.0001 2.62 1.78–3.84 <0.0001
　STEMI 2.20 1.55–3.13 <0.0001 2.28 1.58–3.29 <0.0001
　Previous MI 1.45 0.95–2.23 0.084 1.63 1.05–2.53 0.029
　CVD 2.10 1.32–3.35 0.002 1.80 1.11–2.92 0.017
　Smoking 0.67 0.46–0.97 0.038 0.90 0.61–1.34 0.629
　3-vessel disease 1.62 1.10–2.37 0.013 1.37 0.92–2.03 0.114
　TRI 0.36 0.21–0.60 <0.0001 0.46 0.27–0.78 0.004
De novo patients
　Age >80 years 4.65 3.17–6.80 <0.0001 3.65 2.40–5.55 <0.0001
　Female 1.37 0.89–2.10 0.142      
　Hyperlipidemia 0.42 0.28–0.62 <0.0001 0.51 0.34–0.76 0.001
　Diabetes mellitus 0.85 0.57–1.28 0.454      
　CKD 2.93 1.97–4.35 <0.0001 2.37 1.57–3.58 <0.0001
　STEMI 2.35 1.56–3.55 <0.0001 2.62 1.70–4.04 <0.0001
　Previous MI 1.57 0.99–2.49 0.051 1.83 1.13–2.94 0.013
　CVD 2.16 1.29–3.62 0.003 1.91 1.12–3.28 0.018
　Smoking 0.64 0.43–0.96 0.035 0.90 0.58–1.40 0.654
　3-vessel disease 1.65 1.08–2.53 0.020 1.29 0.82–2.02 0.266
　TRI 0.45 0.26–0.76 0.003 0.58 0.33–1.00 0.052

Abbreviations as in Tables 1,2.

Table 5. Indicators of In-Hospital General Complications
  Univariate Multivariate
OR 95% CI P-value OR 95% CI P-value
Total group
　Age >80 years 1.76 1.45–2.13 <0.0001 1.49 1.21–1.82 <0.0001
　Female gender 0.96 0.79–1.17 0.722      
　Hyperlipidemia 0.79 0.67–0.93 0.006 0.86 0.73–1.02 0.089
　Diabetes mellitus 0.95 0.80–1.12 0.545      
　CKD 1.65 1.40–1.94 <0.0001 1.54 1.30–1.83 <0.0001
　STEMI 1.81 1.54–2.13 <0.0001 1.86 1.57–2.20 <0.0001
　Previous MI 1.21 0.98–1.50 0.073 1.23 0.99–1.53 0.057
　CVD 1.76 1.38–2.24 <0.0001 1.64 1.28–2.11 <0.0001
　Smoking 0.77 0.65–0.91 0.003 0.82 0.69–0.98 0.032
　3-vessel disease 1.66 1.38–1.99 <0.0001 1.53 1.27–1.85 <0.0001
　TRI 0.49 0.41–0.61 <0.0001 0.57 0.46–0.71 <0.0001
De novo patients
　Age >80 years 1.81 1.46–2.23 <0.0001 1.61 1.28–2.02 <0.0001
　Female 0.97 0.78–1.20 0.79      
　Hyperlipidemia 0.85 0.71–1.02 0.82      
　Diabetes mellitus 0.97 0.81–1.16 0.773      
　CKD 1.66 1.39–1.99 <0.0001 1.52 1.26–1.83 <0.0001
　STEMI 1.84 1.54–2.21 <0.0001 1.92 1.58–2.32 <0.0001
　Previous MI 1.23 0.97–1.55 0.075 1.24 0.97–1.58 0.074
　CVD 1.86 1.42–2.42 <0.0001 1.73 1.31–2.29 <0.0001
　Smoking 0.77 0.64–0.93 0.006 0.86 0.71–1.05 0.158
　3-vessel disease 1.71 1.40–2.10 <0.0001 1.58 1.28–1.95 <0.0001
　TRI 0.53 0.42–0.66 <0.0001 0.60 0.47–0.76 <0.0001

Abbreviations as in Tables 1,2.

Table 6. Indicators of In-Hospital Bleeding Complications
  Univariate Multivariate
OR 95% CI P-value OR 95% CI P-value
Total group
　Age >80 years 1.69 1.23–2.33 0.001 1.36 0.98–1.91 0.065
　Female 1.15 0.84–1.57 0.376      
　Hyperlipidemia 0.90 0.68–1.19 0.480      
　Diabetes mellitus 0.95 0.71–1.26 0.744      
　CKD 1.79 1.35–2.36 <0.0001 1.64 1.24–2.19 0.001
　STEMI 1.55 1.18–2.04 0.002 1.65 1.24–2.19 <0.0001
　Previous MI 1.02 0.69–1.49 0.918      
　CVD 1.92 1.30–2.83 0.001 1.73 1.17–2.57 0.006
　Smoking 0.67 0.50–0.90 0.009 0.72 0.53–0.98 0.040
　3-vessel disease 1.55 1.14–2.11 0.005 1.38 1.01–1.89 0.040
　TRI 0.56 0.39–0.80 0.001 0.64 0.44–0.91 0.015
De novo patients
　Age >80 years 1.72 1.21–2.43 0.002 1.45 1.00–2.10 0.047
　Female 1.16 0.82–1.64 0.398      
　Hyperlipidemia 0.86 0.63–1.16 0.327      
　Diabetes mellitus 1.10 0.81–1.49 0.539      
　CKD 1.84 1.36–2.50 <0.0001 1.67 1.23–2.29 0.001
　STEMI 1.48 1.09–2.01 0.01 1.61 1.17–2.22 0.003
　Previous MI 1.09 0.72–1.63 0.678      
　CVD 1.72 1.10–2.68 0.016 1.51 0.95–2.38 0.078
　Smoking 0.68 0.49–0.93 0.018 0.76 0.54–1.06 0.116
　3-vessel disease 1.54 1.09–2.17 0.013 1.38 0.97–1.97 0.069
　TRI 0.60 0.40–0.87 0.009 0.66 0.44–0.98 0.042

Abbreviations as in Tables 1,2.

This was consistent in the STEMI subgroup for both in-hospital general complications (OR, 0.82; 95% CI: 0.69–0.98; P=0.032) and bleeding complications (OR, 0.64; 95% CI: 0.43–0.95; P=0.030), but it did not apply to the NSTEMI-ACS subgroup (in-hospital mortality: OR, 0.82; 95% CI: 0.62–1.09; P=0.179; bleeding complication: OR, 0.82; 95% CI: 0.51–1.32; P=0.429).

The adjusted risk of MACE (death, cardiogenic shock, and heart failure), stratified by smoking status, is shown in Figure 2A. Among smokers, clopidogrel reduced the incidence of MACE (OR, 0.37; 95% CI: 0.20 to 0.70; P=0.002). Among non-smokers, however, clopidogrel had no effect on the incidence of MACE (OR, 1.48; 95% CI: 0.90–2.43; P=0.121). The adjusted risk of in-hospital general complications, stratified by smoking status, is shown in Figure 2B. Among non-smokers, clopidogrel had no effect on the rate of general complications (OR, 1.20; 95% CI: 0.87–1.67; P=0.257). In contrast, among smokers, clopidogrel reduced the rate of general complications (OR, 0.55; 95% CI: 0.33–0.90; P=0.018).

Figure 2.

(A) Adjusted risk for in-hospital major adverse cardiovascular events (MACE), subclassified by smoking status for (a) the total group and (b) de novo patients. (B) Adjusted risk for in-hospital general complications, subclassified by smoking status for (a) the total group and (b) de novo patients. ORs are adjusted for age, chronic kidney disease, ST-segment elevation myocardial infarction, cerebrovascular disease, 3-vessel disease and transradial intervention. Horizontal line, 95% CI.

When we restricted the analysis to 4,978 de novo patients, after excluding 1,217 patients who had a prior history of PCI, the findings were consistent with the overall analysis (Tables 1 – 6; Figure 2).

Discussion

The present study has 3 major findings. First, smokers had significantly better in-hospital outcomes than non-smokers, consistent with previous studies that demonstrated the smoker’s paradox. Second, after adjusting for confounding variables, smoking was not an independent predictor of adverse in-hospital outcome, but, in contrast, was associated with a lower incidence of in-hospital complications, including bleeding complications. And third, smokers benefited more from medication with clopidogrel than non-smokers.

The present results indicate that the smoker’s paradox does exist and that non-smokers have poorer short-term outcome, but this could be largely ascribed to the higher comorbidity of non-smokers. Smokers and non-smokers have completely different backgrounds. Smokers present more frequently with STEMI than non-smokers. This is consistent with past reports and is believed to be related to the heightened thrombotic response of smokers.7,18,19 Smokers have lower rates of other risk factors of coronary artery disease, such as diabetes mellitus, hypertension, and CKD.20 Smokers with ACS present an average of 8 years earlier with fewer risk factors and a lower incidence of multivessel disease, in agreement with past reports.7,9,21 The Million Women Study observed a 2-fold increase in in-hospital mortality from all causes in smokers compared with that in individuals who had never smoked, with expected decrease in life expectancy estimated to be a decade.1,22–24 This indicates that there may have been a selection bias. Of interest, according to Pirie et al, vascular mortality decreases steeply with age, whereas mortality from chronic lung disease and lung cancer increases gradually.23 This could be the reason for the observed age difference in the present study.

A recent study investigating the effect of smoking status on prognosis in patients with coronary disease came from a post-hoc analysis of the SYNTAX trial.10 In that study, smoking was associated with an increased risk of death, MI, stroke, as well as MACE. The SYNTAX trial, however, was a randomized, prospective trial on patients with complex stable coronary artery disease, differing substantially from the present ACS cohort. The strict inclusion and exclusion criteria used in randomized trials may not be generalizable to a real-world clinical population. Most studies focusing on the smoker’s paradox have been post-hoc analyses on randomized trials, meaning that the present analysis has unique strength in being based on a large cohort of real-world patients. Furthermore, the present study is incongruent with the SYNTAX-based study, which was conducted on a Japanese population. Tobacco consumption in the Japanese population is the highest in the world,22 and, further to this, individuals of Japanese descent have a substantially higher bleeding risk.13

No reflow from thrombotic complications remains a concern in treating patients with ACS. This led to the development of the concept of pretreatment with clopidogrel,25 but delayed onset of action and its lack of consistency among individuals still pose a challenge. The GRAVIS trial demonstrated an association between on-clopidogrel reactivity <208 P2Y12 reaction units at 12–24 h after PCI or during follow-up and a lower risk for cardiovascular events. Thus, they proposed the importance of targeting a level of reactivity below this threshold.26 The present analysis demonstrated a reduced benefit from clopidogrel therapy in non-smokers, in agreement with the Desai et al and Ueno et al studies.11,27 The present results indicate that cigarette smoking modifies the beneficial effect of clopidogrel on clinical outcome. Clopidogrel, a thienopyridine antiplatelet agent, is a prodrug that inhibits platelet activation and aggregation by binding to the P2Y12 class of adenosine diphosphate receptors on platelets. As a prodrug, clopidogrel is dependent on metabolic activation through CYP-dependent oxidation involving CYP2C19.28 Metabolic activation is achieved by cytochrome P450 enzymes in 2 steps.29 Cigarette smoking is an inducer of CYP1A2,30 and it has been reported to be associated with increased platelet inhibition in response to clopidogrel.31 Past studies on pharmacodynamics have also reported that this relationship between cigarette smoking and enhancement of platelet inhibition by clopidogrel, increased with dose of cigarette consumption.27,31,32

Smokers would theoretically have more bleeding complications, because the effect of clopidogrel is enhanced. In the present study, however, smokers had significantly less bleeding complications, with smoking associated with lower bleeding complication rate on multivariate logistic regression analysis. Japan has a unique approved loading dose for clopidogrel, with the common loading dose being 300 mg. An initial loading dose of 600 mg is not authorized.33 A 300-mg dose may be insufficient for any substantial effect in non-smokers, but an initial loading dose of 300 mg may be appropriate for smokers in Japan, considering the enhanced anti-thrombotic effect. We have previously demonstrated in the Japanese cohort registry-based study that patients undergoing PCI with dual antiplatelet therapy had a lower risk of post-procedural cardiac events, without any increased bleeding risk.34 The present findings suggest that adjustment of clopidogrel dose may be required based on smoking status. A recent post-hoc analysis of the GRAVITAS clinical trial, which assessed the pharmacodynamic effects of smoking on clopidogrel response after PCI, supports this theory. In that study, the influence of smoking on clopidogrel effect differed according to clopidogrel dose. Standard-dose clopidogrel smokers had a greater antiplatelet effect than non-smokers. In contrast, double-dose clopidogrel eliminated the difference due to smoking status, resulting in similar levels of platelet activity between smokers and non-smokers.35

We do not encourage cigarette smoking based on the present study results. Smoking induces hypercoagulability and results in early MI, as shown in the present study. Thun et al noted that the relative risk of death from ischemic heart disease for current smokers, as compared with those who had never smoked, was 2.86 (95% CI: 2.65–3.08) for women and 2.50 (95% CI: 2.34–2.66) for men.36

Study Limitations

The present study used data from a PCI database and hence has limitations inherent to a retrospective study design. Studies comparing smokers and non-smokers are subject to selection bias and confounding bias, specifically with regard to the procedural risk of the patient. To minimize these biases, multivariate hazards regression analysis was performed. Hidden bias, however, may remain owing to the influences of unmeasured confounders. Some information, such as smoking intensity, was not collected in the registry and therefore could not be investigated as an explanatory value. This study also lacked ischemic endpoints, such as recurrent MI or recurrent ischemia leading to urgent revascularization. Clinical outcomes within this study were limited to a hospital setting, and analysis of long-term outcomes was not performed. In addition, given that the subject group was carefully selected using the described inclusion and exclusion criteria to best fit the purpose of the study, they included only those who underwent PCI, thus excluding patients who were not eligible for coronary revascularization. Furthermore, the subject group was limited to Japanese patients and the present results may not be directly applicable to other ethnic groups.

Recently, oral P2Y12 antagonists, ticagrelor and prasugrel, have become available as alternatives to clopidogrel.16,37 These drugs provide more consistent and powerful platelet inhibition than clopidogrel and their safety and effectiveness in Asian patients has been demonstrated in recent clinical trials.38–40 Ticagrelor, however, remains unavailable in Japan currently, and prasugrel has been on the market only since March 2014. Thus, ticagrelor was not used in the present cohort, and the 5 patients who were prescribed prasugrel were excluded from the study. It is likely that the present study may be the last to evaluate the smoker’s paradox in patients treated with clopidogrel. It will be interesting to compare results from future studies on ticagrelor and prasugrel, because in both cases the association between cigarette smoking and CYP activity would be ameliorated.

Conclusions

The smoker’s paradox was largely explained by confounding factors related to the lower risk profile of smokers. Smokers benefit from a positive modification of the efficacy of clopidogrel.

Disclosures

The JCD-KiCS Steering Committee was responsible for overall study guidance, including the study protocol, data analysis, and interpretation of results. The Department of Healthcare Quality Assessment at Tokyo University independently managed the database. The KiCS Scientific Committee managed the participating sites and provided monthly on-site monitoring services to assure data accuracy and completeness throughout the study. During the planning, implementation, and reporting of this study, there were no issues such as conflicts of interest, conflicts of responsibility, or intellectual property right concerns.

Acknowledgments

S.K. received an unrestricted research grant for the Department of Cardiology, Keio University School of Medicine from Bayer Pharmaceutical and Pfizer Japan.

References
 
© 2016 THE JAPANESE CIRCULATION SOCIETY
feedback
 Top