Effect of reperfusion modality on outcome in nonsmokers and smokers with acute myocardial infarction (a Primary Angioplasty in Myocardial Infarction [PAMI] Substudy)

Effect of Reperfusion Modality on Outcome in Nonsmokers and Smokers With Acute Myocardial Infarction (A Primary Angloplasty in Myocardial Infarction [PAM] Substudy) Terry R. Bowers, MD, Edward F. Terrien, Debra Sachs, MPH, and Cindy L. Grines,

MD, William W. O’Neill, MD, MD, for the PAMI Investigators

We analyzed the 395 patients randomized into the Primary Angioplasty in Myocardial Infarction (PAM) trial to receive tissue plasminogen activator (tPA) or to undergo primary percutaneous transluminal coronary angioplasty (PTCA) fo r acute myocardial infarction (AM). Of these, 168 were current smokers and 128 had never smoked. Univariate analyses of baseline characteristics and outcome, including death, recurrent AM, and recurrent ischemia, were done by chi-square analysis. Multivariate stratified analysis was then performed controlling for age and gender, which were found to be confounders of outcome. The combined in-hospital outcomes of death, recurrent AMI, and recurrent ischemia were similar for smokers and nonsmokers (p = 0.12). When stratified according to treatment modality, nonsmokers treated with PTCA had a lower frequency of death and nonfatal recurrent AMI (7% vs 18%; p =

0.05), in-hospital ischemia (11% vs 33%; p = 0.004), or the combined event (13% vs 40%; p = 0.001). At 6 months, nonsmokers treated with PTCA continued to have a lower incidence of death or nonfatal recurrent AM (11% vs 24%; p = 0.07) compared with tPA. Conversely, in smokers, the treatment strategy did not significantly affect hospital outcomes: recurrent ischemia (12% vs 23%; p = 0.07), death and recurrent AM (6% vs 8%; p = 0.55), or the combined event (15% vs 25%; p = 0.12). The stati st’tea I significance of these associations was maintained when multivariate analysis controlling for age and gender was used. Thus, nonsmokers presenting with AM had a significantly better outcome when treated with primary angioplasty; these differences were not seen in smokers. (Am J Cardiol 1996;78:5 1 l-5 15)

igarette smoking has a multitude of adverse hematologic l-3 and atherosclerotic4’5 effects on the C cardiovascular system, which predispose patients

investigate this issue, we reviewed clinical and angiographic information that was collected prospectively in the Primary Angioplasty in Myocardial Infarction (PAMI) study database.14

who smoke to acute myocardial infarction (AMI). Epidemiologic data have consistently shown that smoking is associated with a higher incidence of AM1 and deam6-’ Paradoxically, thrombolytic trials have reported that patients with AM1 who smoke at the time of presentation have a favorable prognosis compared with nonsmokers.3*9-1’This apparent discrepancy has been postulated by several investigators 3~11-13 to result from favorable baseline demographic and angiographic variables in smokers. Conversely, nonsmokers represent a group who are typically older, with more cardiac risk factors and advanced stages of underlying atherosclerosis leading to AMI.12,13 Because patients treated with primary angioplasty have improved infarct artery patency and Thrombolysis in Myocardial Infarction (TIMI) flow grades compared with those treated with thrombolysis, 14,15we speculated that angioplasty would benefit nonsmokers preferentially. To y, Department of Medicine, William From the Division of Cardiolo Manuscript received OcBeaumont Hospital, Royal Oa &, Michigan. tober 27, 1995; revised manuscript received and accepted April 1, 1996. Address for reprints: Cindy 1. Grines, MD, Division of Cardiology, William Beaumont Hospital, 3601 West Thirteen Mile Road, Royal Oak, Michigan 48073.6769.

0 1996 by Excerpta Medica, All rights reserved.

Inc

METHODS Study group: The study population was derived from the 395 patients enrolled in the PAM1 trial with evolving AMI, who were randomized to receive tissue plasminogen activator (tPA; Activase) or to undergo primary percutaneous transluminal coronary angioplasty (PTCA) .I4 The PAM1 study protocol, including patient selection, inclusion and exclusion criteria, data collection, definitions of ischemia and infarction, and treatment recommendations, has been described previously.‘4 Subgroup analysis was performed in patients who were current smokers at the time of presentation (regardless of duration) compared with those who never smoked (nonsmokers), with further stratification for reperfusion modality. ‘ ‘Ex-smokers,” or those who quit smoking within 1 month of the index AMI, were excluded from analysis. Baseline characteristics: The baseline characteristics analyzed were age (continuous variable and dichotomous variable of >70 vs 570)) sex, history of hypertension, diabetes mellitus, previous infarction, antecedent angina, smoking status (current smokers vs never-smokers), heart rate at presentation (con0002.9149/96/$15.00 PII SOOO2-9149(96)00354-2

5 11

FIGURE 1. Stratification of patients in the Primary Angioplasty in Myocordial h&r&on trial accordin to smoking status. Ex-smokers who quit smoking > 1 month be!0 re present&ion with acute myocardial infarction were excluded from further analysis. PTCA = percutaneous transluminal coronary angioplasty; tPA = tissue plasminogen activator.

TABLE I Baseline

Data Nonsmokers (n = 128)

Risk factor Age (mean + SD yr) Age 270

Women Systemic hypertension Diabetes mellitus Prior angina Prior AMI Hemodynamic data Resting heart rate (beats/min) Systolic blood pressure

65? 10 44 (34%) 52 (41%) 67 (52%) 18 (14%) 41 (32%) 18 (14%)

Smokers (n = 168)

10

p Value

55

c

17 34 54 13 60 18

(10%)
10.001.

(20%)
L

Nonsmokers

Smokers

I

FIGURE 2. In-hospital clinical outcomes (death, recurrent myocardial infarction, or recurrent ischemia) stratified according to treatment status in the nonsmokers and smokers. Abbreviations as in Figure 1.

for continuous variables. Continuous variables are expressed as mean + 1 SD. Multivariate analyses of the effects of baseline variables, reperfusion strategy, and smoking status on each of the clinical outcomes were performed by stratified analysis (SAS software program), controlling for confounders of outcome. Multivariate statistics are reported as Mantel-Haenszel odds ratios with their 95% confidence limits.

RESULTS

Baseline characteristics of the study population: Of the 395 patients randomized in the trial, 99 had quit 77+ 17 76? 17 0.53 smoking and were excluded from further analysis. Of the remaining patients, 168 were current smokers 134+23 131 224 0.18 and 128 were nonsmokers. Both groups were closely (mm W Killip class > 1 stratified to the treatments (smokers, tPA 50%; 21 (16%) 22 (13%) 0.54 Angiographic data PTCA 50%; nonsmokers, tPA 52%, PTCA 48%) 1-Vessel disease 58 (45%) 84 (50%) 0.48 (Figure 1) . LAD 54 (42%) 64 (38%) 0.82 The baseline characteristics of smokers and nonRCA 59 (46%) 74 (44%) 0.98 smokers differed significantly in terms of age, genLCX 15 (12%) 30 (18%) 0.26 Anterior wall AMI 51 (40%) 62 (37%) 0.1 1 der, and history of hypertension (Table I). NonsAcute coronary angioplasty 61 (48%) 87 (52%) 0.69 mokers were older (65 2 10 vs 55 + 10 years; p Any coronary angioplasty 84 (66%) 111 (66%) 0.94 1)) 90%; p = 0.32) and TIM1 2 (7% vs 8%; p = 0.66) anterior infarct location, l-vessel disease, and ran- flow after primary angioplasty. domization to PTCA. In-hospital and 6-month outcomes: The in-hospital Clinical outcomes: The clinical outcomes analyzed clinical outcomes stratified according to treatment were in-hospital mortality or nonfatal recurrent AMI, status in the nonsmokers and smokers are presented in-hospital recurrent ischemia, and combined 6- as cumulative events of death, recurrent AMI, and month mortality and recurrent AMI. recurrent ischemia (Figure 2). The same percentages Statistical analysis: Univariate analyses of differof both groups had undergone PTCA because of the ences in baseline characteristics and outcomes be- protocol or for recurrent ischemia, which was a nontween smokers and nonsmokers, stratified according protocol indication (nonsmokers 111,66%; smokers to reperfusion strategy, were done by chi-square 84, 66%). Nonsmokers appeared to have improved analysis for categoric variables and Student’s t test outcomes when treated with PTCA compared with 512

THE AMERICAN JOURNAL OF CARDIOlOGYa

VOL. 78

SEPTEMBER1, 1996

those treated with thrombolytic agents ( 13% vs 40%; p = 0.001) . However, no significant differences in outcome were observed in smokers treated with PTCA or tPA (15% vs 25%; p = 0.12). Further segregating the outcomes into combined in-hospital death and recurrent AMI-the primary end point of the PAM1 trial-nonsmokers treated with PTCA continued to have a significantly lower incidence than those treated with thrombolytic agents (7% vs 18%; p = 0.05) (Figure 3A). At 6 months, the difference in outcome lost statistical significance, although there was a trend in favor of nonsmokers treated with angioplasty (PTCA 11% vs tPA 24%; p = 0.07) (Figure 4). Recurrent ischemia occurred much less frequently in nonsmokers treated with PTCA than with tPA ( 11% vs 33%; p = 0.004) (Figure 3B). Smokers had no significant differences in combined outcomes of death and recurrent AM1 with regard to reperfusion modality either in the hospital (PTCA 6% vs tPA 8%; p = 0.55) (Figure 3A) or at 6 months (PTCA 9% vs tPA 13%; p = 0.47) (Figure 4). Recurrent ischemia was not significantly different, but there was a trend toward smokers treated with PTCA rather than tPA (12% vs 23%; p = 0.07) (Figure 3B ) . Multivariate analysis revealed that the baseline characteristics of age (in a continuous or categoric fashion) and gender were associated with the various outcomes in the stratified analysis. All other variables fell out of the analysis as confounders of outcome. After accounting for age and gender, the decrease in the combined end points of death, recurrent AMI, and recurrent ischemia in nonsmokers treated with PTCA remained highly significant (p
DISCUSSION This study examines the effect of the reperfusion modality in patients with AM1 stratified according to smoking status on in-hospital and 6-month outcome. It has been well documented in the past decade that patients who smoke at the time of presentation with AM1 have a favorable prognosis compared with nonsmokers.3~g-‘1The seeming “smoker’s paradox” was explained by several investigators to be based on the observation that nonsmokers have a less favorable risk profile with regard to several cliniand, more recently, angiographic ‘2x13varical 3,7,g-12 ables. In the PAM1 data, nonsmokers treated with primary angioplasty appeared to benefit in terms of less in-hospital ischemia and the combined end point of death and recurrent AMI. Explanation

for the treatment effect seen in nonsmok-

The differences in end points between patients treated with primary angioplasty and tPA occurred only in the nonsmoking cohort. A plausible expla-

ers:

CORONARY

A20 16

16%

m 0

tPA PTCA

16 14 12 Oh 10 6 6 4 2 0

Nonsmokers

B

Smokers

40 33%

E

tPA PTCA

30 23% % 2u

0

Nonsmokers

Smokers

FIGURE 3. in-hospital clinical outcomes divided into subgroups of death and recurrent acute myocardial infarction (the 2 rimary end points of the Primary Angioplasty in Myocardial In L rction trial) (A) and recurrent ischemia (ft) stratified according to treatment status in nonsmokers and smokers. Abbreviations as in Figure 1.

I

30

I I 0

24%

tPA PTCA

20 % 10

0

Nonsmokers

Smokers

FIGURE 4. The 6-month cumulative outcomes of death and recurrent acute myocardial infarction. Abbreviations as in Figure 1.

nation is that in addition to having the adverse prognostic features of advanced age and female gender, 11-‘3~‘6-18 controlled for in this analysis, nonsmokers may represent a high-risk subgroup based on their atherosclerotic disease. Recently, angiographic data have revealed that nonsmokers have more atherosclerotic burden than smokers before reperfusionl’and in the period after thrombolysis in association with less early complete reperfusion.12p20 Incomplete reperfusion (
ARTERY DISEASE/SMOKING

AND REPERFUSION MODALIW OUTCOMES

513

Nonsmokers

In-Hospital Death,Recurrent MI

0

Smokers FIGURE 5. Effect of the reperhsion modality (percutaneous transluminal coronary angioplasty vs tissue plasminogen activator) on outcome after control~~-~n~e$dd;~d

In-Hospital Recurrent lschemia

Ye&a/ lines represent the odbs bars depict 95% ratio; horizontal confidence intervals (Cl). Angioplasty resulted in a significant decrease in events in nonsmokers, but did not affect the outcome in smokers. MI = myocardial infarction.

In-Hospital Death,Recurrent MI, Recurrent lschemia 6 Month Death,Recurrent MI

I ! I 0.5

1.0

I 1.5

2.0

Odds Ratlo (ss%Cl)

* p
** p
l

*+ p=O.O69

high-grade residual stenoses prone to platelet depo- of ex-smokers was e:YCIluded from this analysis to sition due to shear forces. avoid a possible selection bias. lack of treatment effect in smokers: Interestingly, the Study limitations: The randomization in this trial to smoking cohort did not have a statistically significant a reperfusion strategy was not stratified based on difference in outcome with the different reperfusion smoking status; therefore, the conclusions are limtherapies. This may be due to a different mechanism ited by the inherent shortcomings of this type of analof AM1 in smokers, resulting from a more throm- ysis. The small numbers of patients potentially acbotic occlusion of a less atherosclerotic lesion.3*‘2*26,27count for the lack of angiographic and clinical Based on this hypothesis, it follows that high rates differences between smokers and nonsmokers seen of reperfusion in smokers 21-23,27 would be achievable in other studies.11-14The necessity to control for the with both reperfusion modalities; high patency rates confounding variables of age and gender, in addition and normal perfusion with primary angioplasty oc- to the small sample size, limit the statistical power curred in PAM1 in both nonsmokers and smokers. of this study. Nevertheless, the association of the Explanation for loss of statistical significance at six treatment effect of PTCA on in-hospital outcomes months: The benefits seen in the primary angioplasty achieved statistical significance. group with regard to decreases in mortality and reThe folkwing institutions and investigators participated current AM1 lost statistical significance at 6-month in theAppendix Primary Angioplasty in Myocardial Infarction Study Group: William follow-up. The differences that did exist approached Beaumont Hospital, Royal Oak, MI: W. O’Neill, C. Grines, M. Brodsky, N. statistical significance and may reflect an insufficient Choksi, J. Cieszkowski, M. Elliott, H. Friedman, V. Gangadharan, R. Levin, D. M. May, B. Meany, G. Pavlides, S. Puchrowicz-Ochocki, R. Ramos, sample size. However, the treatment effect (PTCA Marsalese, R. Safian, V. Savas, T. Schreiber, and M. Strzelecki; Lakeland Regional Medical vs tPA) more likely became diluted at follow-up by Center, Lakeland, FL: K. Browne, R. Vlietstra, and R. Roy; Clinique Pasteur, France: J. Marco, M. Vandonnael, and G. Roberts; St. Vincent Hosthe aggressive strategy of the investigation to per- Toulouse, pital, Indianapolis, IN: D. Rothbaum and M. Klette; El Camino Hospital, Mounform revascularization in patients with ischemia; tain View, CA: G. Stone, C. Bavor, R. Constantine, M. Klughaupt, 1. Saab, F. 50% of the patients assigned to the tPA arm under- St. Gear, and E. Bough; Mid-America Heart Institute, Kansas City, MO: G. J. O’Keefe, and C. Dreiling; St. Mary of the Plains, Lubbock, TX: P. went PTCA or bypass surgery before follow-up. Fi- Hartzler, Overlie and M. Quijada; Allegheny General Hospital, Pittsburgh, PA: B. Donally, this finding is not likely to be due to a sampling nohue, R. Begg, and L. Zahren; United Hospital, Grand Forks, ND: N. Cbelliab, bias, because 95% of the patients were available for R. Wolf, and N. Endres; Heart Institute of St. Joseph Hospital, Atlanta, GA: C. Gates, W. Knopf, and J. Sheftel; Florida Hospital South, Orlando, FL: R. Ivanfollow-up, with all the clinical events confirmed. hoe and C. Simpkiss; and North Central Heti Clinic, Wausau, WI: J. Freeman. Rationale for exclusion of ex-smokers: Smoking cesThe following are members of the Safety and Data Monitoring ComG. Timmis, William Beaumont Hospital, Royal Oak, MI; B. Pitt, Unisation has been shown to reduce the risk of first AM1 mittee: versity of Michigan Medical Center, Ann Arbor, MI; and T. Ryan, University to the same level as in those who have never smoked, Hospital, Boston, MA. regardless of the duration of smoking or the time elapsed since quitting.** Yet a recent large subgroup analysis3 revealed that ex-smokers have an inter1. Fitzgerald GA, Oates JA, Nowak J. Cigarette smoking and hemostatic funcmediate outcome after AM1 compared with smokers tion. Am Heart J 1988;15:267-271. and nonsmokers. Furthermore, smokers with severe 2. Faster V, Chesebro JH, Frye RL, Elveback LR. Platelet survival and the dedisease who stopped smoking would thereby in- velopment of coronary artery disease in the young adult: effect of cigarette smokstrong family history and medical therapy. Circulation 1981;63:546-551. crease the mortality in the nonsmoking group if they ing, 3. Barbash GI, White HD, Modan M, Diaz R, Hampton JR, Heikkila J, Kriswere combined. Thus, the relatively small subgroup tinsson A, Moulopoulos S, Paolasso EAC, Van der Werf T, Pehrsson K, Sandoe 514

THE AMERICAN JOURNAL OF CARDIOLOGY”

VOL. 78

SEPTEMBER1, 1996

E, Simes J, Wilcox RG, Verstraete M, von der Lippe G, Van der Werf F, for the Investigators of the International Tissue Plasminogen Activator/Streptokinase Mortality Trial. Significance of smoking in patients receiving thromholytic therapy for acute myocardial infarction. Circulation 1993;87:53-58. 4. Garrison RJ, Kannel WB, Feinlieb M, Castelli WP, McNamera S. Cigarette smoking and HDL cholesterol. The Framingham offspring study. Atherosclerosis 1978;30: 17-251. 5. Ramsdale DR, Faragher EB, Bray CL, Bennett DH, Ward C, Beton DC. Smoking and coronary artery disease assessed by routine coronary angiography. Br Med J 1985;290: 197-200. 6. Doll R, Peto R. Mortality in relation to smoking: twenty years’ observations of male British doctors. Br Med .I 1976;2:1525-1536. 7. Wilhelmsen L. Coronary heart disease: epidemiology of smoking and intervention studies of smoking. Am Heart J 1988;115:242-249. 8. LaCroix AZ, Lang J, Scherr P, Wallace RB, Comoni-Huntley J, Berkman L, Curb D, Evans D, Hennekens CH. Smoking and mortality among older men and women in three communities. NE& JMed 1991;324:1619-1625. 9. AIMS Trial Study Group. Long-term effects of intravenous anistreplase in acute myocardial infarction: final report of the AIMS study. Lancer 1990;335:427-431. 10. Gruppo Italian0 per lo Studio della Sopravvivenza nell’Infarto Miocardico. GISSI-2. A factorial randomized trial of alteplase versus streptokinase and heparin versus no heparin among 12,490 patients with myocardial infarction. Lancet 1990;336:65-71. 11. Mueller HS, Cohen LS, Braunwald E, Forman S, Feit F, Ross A, Schweiger M, Cabin H, Davison R, Miller D, Solomon R, Knatterud GL, for the TIM1 Investigators. Predictors of early morbidity and mortality after thrombolytic therapy of acute myocardial infarction: analysis of patient subgroups in the TIM1 trial, phase II. Circukztion 1992;85: 1254-1264. 12. Grines CL, Top01 EJ, O’Neill WW, George BS, Kereiakes D, Phillips HR. Leimberger JD, Woodlief LH, Califf RM, for the TAM1 Investigators. Effect of cigarette smoking on outcome after thrombolytic therapy for myocardial infarction. Circulation 1995;91:298-303. 13. Barbash GI, Reiner JS, Granger CB, White HD, Lundergan CF, Ross AM, for the GUSTO Investigators. Smoking and acute myocardial infarction survival: demographic and quantitative angiographic foundations for the “smoker’s paradox” (abstr). Circulation 1994;90:221. 14. Grines CL, Browne KF, Marco J, Rothbaum D, Stone GW, O’Keefe J, Overlie P, Donahue B, Chelliah N, Timmis GC, Vlietstra RE, Strzelecki M, Puchrowicz-Ochrocki S, O’Neill WW, for the Primary Angioplasty in Myocardial Infarction Study Group. A comparison of immediate angioplasty with thrombolytic therapy for acute myocardial infarction. N Engl J Med 1993;328:673-679.

CORONARY

15. Gibbons RJ, Holmes DR, Reeder GS, Bailey KR, Hopfensperger MR, Gersh BJ. Immediate angioplasty compared with the administration of a tbrombolytic agent followed by conservative treatment for myocardial infarction. N Eng[ J Med 1993;328:685-691. 16. Sparrow D, Dawber TR, Colton T. Influence of cigarette smoking on prognosis after a first myocardial infarction. J Chronic Dis 1978;31:425-432. 17. O’Neill WW, Brodie B, Knopf W, Ivanhoe R, O’Keefe J, Taylor G, Berdan L, Woodlief L, Califf R, Duke Medical Center, Durham, NC. Initial report of the Primary Angioplasty Revascularization (PAR) Multicenter Registry (abstr). Circulation 1991;84(suppl 2):11-536. 18. O’Keefe J, Rutherford BD, McConahay DR. Ligon RW, Johnson WL Jr, Giorgi LV, Crockett JE, McCallister BD, Conn RD. Gura GM, Good TH, Steinhaus DM, Bateman TM, Shimshak TM, Hartzler GO. Early and late results of coronary angioplasty without antecedent thrombolytic therapy for acute myocardial infarction. Am J Cardiol 1989;64:1221-1230, 19. Vlietstra RE, Kmnmal RA, Frye RL, Seth AK, Tristani FE, Killip T III. Factors affecting the extent and severity of comnary heart disease in patients enrolled in the Coronary Artery Surgery Study. Atherosclerosis 1982;2:208-215. 20. Laster SB, O’Keefe J, Gibbons RJ. The incidence and importance of TIM1 3 flow following primary infarct angioplasty (abstr). Circulation 1994;90:1-221. 21. O’Neill WW, Timmis GC, Bourdillon PD, Lai P, Ganghadarhan V, Walton J, Ramos R, Laufer N, Gordon S, Schork A, Pitt B. A prospective randomized clinical trial of intracoronary streptokinase versus coronary angioplasty for acute myocardial infarction. N Engl J Med 1986;314:812-818. 22. Zijlstra F, de Boer MJ, Hoomtje JCA, Reiffers S, Reiber JHC, Suryapranata H. A comparison of immediate coronary angioplasty with intravenous streptokinase in acute myocardial infarction. N Engl J Med 1993;328:680-684. 23. Dittrich H, Gilpin E, Nicod P, Cali G, Henning H, Ross J Jr. Acute myocardial infarction in women: influence of gender on mortality and prognostic variables. Am J Cardiol 1988;62:1-7. 24. Weisfeldt ML, Lakatta EG, Gerstenblith G. Aging and the heart. In: Braunwald, ed. Heart Disease, a Textbook of Cardiovascular Medicine. 4th ed. 1992:1656-1666. 25. Zahger D, Cercek B, Cannon CP, Jordan M, Davis V, Braunwald E, Shah PK, for the TIMIInvestigators. How do smokers differ from nonsmokers in their response to tbrombolysis? (The TIMI- Trial.) Am J Cardiol1995;75:232236. 26. How does smoking harm the heart? [editorial] Br Med J 1980;281:573574. 27. Gomez MA, Karagounis LA, Allen A, Anderson JL, for the TEAM-2 Investigators. Effect of cigarette smoking on coronary patency after thrombolytic therapy for myocardial infarction. Am J Cardiol 1993;72:373-378. 28. Nyboe J, Jensen G, Appleyard M, Schnohr P. Smoking and the risk of first acute myocardial infarction. Am Heart J 1991;122:438-444.

ARTERY DISEASE/SMOKING

AND REPERFUSION MODALITY OUTCOMES

515