Effects of Optimal Medical Treatment With or Without Coronary Revascularization on Angina and Subsequent Revascularizations in Patients With Type 2 Diabetes Mellitus and Stable Ischemic Heart DiseaseClinical Perspective
Background—In the Bypass Angioplasty Revascularization Investigation 2 Diabetes (BARI 2D) trial, an initial strategy of coronary revascularization and optimal medical treatment (REV) compared with an initial optimal medical treatment with the option of subsequent revascularization (MED) did not reduce all-cause mortality or the composite of cardiovascular death, myocardial infarction, and stroke in patients with type 2 diabetes mellitus and stable ischemic heart disease. In the same population, we tested whether the REV strategy was superior to the MED strategy in preventing worsening and new angina and subsequent coronary revascularizations.
Methods and Results—Among the 2364 men and women (mean age, 62.4 years) with type 2 diabetes mellitus, documented coronary artery disease, and myocardial ischemia, 1191 were randomized to the MED and 1173 to the REV strategy preselected in the percutaneous coronary intervention (796) and coronary artery bypass graft (377) strata. Compared with the MED strategy, the REV strategy at the 3-year follow-up had a lower rate of worsening angina (8% versus 13%; P<0.001), new angina (37% versus 51%; P=0.001), and subsequent coronary revascularizations (18% versus 33%; P<0.001) and a higher rate of angina-free status (66% versus 58%; P=0.003). The coronary artery bypass graft stratum patients were at higher risk than those in the percutaneous coronary intervention stratum, and had the greatest benefits from REV.
Conclusions—In these patients, the REV strategy reduced the occurrence of worsening angina, new angina, and subsequent coronary revascularizations more than the MED strategy. The symptomatic benefits were observed particularly for high-risk patients.
- angina pectoris
- coronary angioplasty
- coronary bypass surgery
- coronary artery disease
- coronary artery bypass
- diabetes mellitus, type 2
The Bypass Angioplasty Revascularization Investigation 2 Diabetes (BARI 2D) trial showed that a strategy of prompt coronary revascularization and optimal medical treatment (REV) was not superior to a strategy of optimal medical treatment with the option of subsequent coronary revascularization (MED) in terms of total mortality and the composite of death, myocardial infarction, or stroke in patients with type 2 diabetes mellitus and stable coronary heart disease.1 Coronary revascularization by either percutaneous coronary interventions (PCI) or coronary artery bypass graft (CABG) surgery has been shown to reduce angina, but to a lesser degree in patients with diabetes mellitus than in patients without diabetes mellitus.2 Although there have been trials comparing coronary revascularization with medical treatment alone in patients with stable coronary artery disease, the numbers of participants with type 2 diabetes mellitus in these studies were limited.3,–,6 The present analysis was undertaken to assess the hypothesis that BARI 2D participants randomized to a prompt coronary revascularization (PCI or CABG) at entry in the trial with an optimal medical treatment would have a lower rate of worsening angina, new angina, and subsequent coronary revascularization and a higher rate of angina-free status than participants randomized to an optimal medical treatment strategy with the option of subsequent coronary revascularization during a 5.3-year follow-up.
Editorial see p 1489
Clinical Perspective on p 1500
The BARI 2D trial was a randomized trial with a 2×2 factorial design to determine whether the REV treatment strategy was superior to the MED strategy and whether a strategy with insulin-sensitizing agents was more efficient than a strategy with insulin-providing agents, both targeting hemoglobin A1c <7.0%, in reducing total mortality and cardiovascular events in patients with type 2 diabetes mellitus and stable coronary artery disease. Details of the BARI 2D trial have been published,1,7,8 and are outlined briefly here.
The 2368 participants of the BARI 2D trial were recruited between January 1, 2001, and March 31, 2005, in 49 clinical sites; 1499 participants were from the United States, 356 were from Brazil, 353 were from Canada, 85 were from Mexico, and 75 were from Austria and the Czech Republic. Participants had to be ≥25 years of age with a diagnosis of type 2 diabetes mellitus, have angiographically documented coronary artery disease (≥50% stenosis in a major epicardial artery), and have myocardial ischemia objectified during a stress test or typical effort angina with an epicardial coronary stenosis ≥70%. All patients had to have suitable coronary anatomy for either PCI or CABG surgery. The main exclusion criteria were a clinical indication for immediate revascularization, coronary revascularization within the 12 months before randomization, left main coronary disease, class III and IV heart failure, serum creatinine level >2.0 mg/dL (177 μmol/L), and hemoglobin A1c >13.0%. Patients with unstable angina who became stable on medical therapy before randomization were included. For the present study, the participants were classified according to their angina status within the 6 weeks before randomization into 3 groups: (1) angina with their functional impairment according to Canadian Cardiovascular Society (CCS) grades of angina,9 stabilized grade IV CCS angina, or unstable angina; (2) anginal equivalents manifested mostly by effort dyspnea; and (3) no angina characterized by neither angina nor anginal equivalent. Among the 2368 participants enrolled in BARI 2D, there was satisfactory information on baseline angina status in 2364; they constituted the population of the study. The protocol was approved by the review board or ethics committee of each institution. All participants provided written informed consent.
Random Assignment, Treatment, and Follow-Up
Randomization to the REV or MED strategy was stratified according to either PCI or CABG as determined by the cardiologist who based his or her decision on the most appropriate therapy for each patient, with the expectation that patients with the most extensive coronary artery disease would be selected for CABG. Participants in the PCI stratum randomized to REV or MED and those in the CABG stratum randomized to REV or MED were also randomized to the insulin-providing strategy or the insulin-sensitizing strategy. Patients randomized to the revascularization procedure had to have their PCI or CABG surgery within 4 weeks of randomization. Optimal medical therapy consisted of lifestyle management targeting smoking cessation, diet, weight loss, and regular physical exercise and pharmacological therapy to maintain hemoglobin A1c <7.0%, low-density lipoprotein cholesterol <100 mg/dL (2.6 mmol/L), and blood pressure ≤130/80 mm Hg. Participants received individualized antianginal medications. Antianginal medication therapy (including β-blockers, calcium channel blockers, and long-acting nitrates) was categorized as none, monoagent, and polyagent (the combination of 2 or 3 agents). For patients incapacitated by angina or anginal equivalent symptoms and for patients with documented worsened ischemia despite optimal medical treatment during follow-up, the treating cardiologist could recommend coronary revascularization, defined as subsequent revascularization. Participants were evaluated on a monthly basis for 6 months and every 3 months thereafter for a mean follow-up of 5.3 years. A standardized questionnaire on angina and anginal equivalents was administered at each visit.
The outcomes for this analysis were worsening angina, freedom from angina, occurrence of new angina, and subsequent coronary revascularization. Angina was characterized by chest pain occurring during physical activity or stress relieved by rest or nitroglycerin. Worsening angina was defined as a changing pattern of angina that distinctly worsened in severity and/or frequency, from no classic angina at entry to CCS grade III or IV angina, or from any status at entry to unstable angina. Freedom from angina was defined as the absence of angina in participants with classic angina at entry. New angina was defined as the first reported classic angina in participants without angina at entry. Finally, subsequent coronary revascularizations were the first PCI or CABG done during the follow-up in the participants randomized to the MED strategy or the first additional revascularization done in participants randomized to the REV strategy.
To assess the effect of the randomized treatment, the comparison throughout this article was on an intention-to-treat basis, ie, comparing the REV strategy with the MED strategy, overall and within the intended revascularization strata, the PCI stratum and CABG stratum. The only exception was the baseline characteristics comparison in Table 1, which was the comparison of the PCI and CABG strata to illustrate the baseline difference between the strata.
The 4 outcomes could be distinguished as 2 types. The worsening angina and freedom from angina were repeated outcomes, whereas the new angina and subsequent revascularization were event outcomes. Thus, for the repeated outcomes, the percentage of patients who had worsening angina or were free from angina during each follow-up year was compared by use of the χ2 test. For the event outcomes, the cumulative incidence rate of new angina or subsequent revascularization was estimated by life-table method or Kaplan-Meier estimator, respectively, and compared by use of the log-rank test. Because there were 4 outcomes of interest, the Bonferroni correction was applied to address the multiple comparisons. Values of P<0.0125 were deemed statistically significant.
The angiographic risk factors were identified as baseline diseased left ventricular regions, myocardial jeopardy index, and history of revascularization. For the subgroup analysis determining which angiographic risk group benefits most from the randomized revascularization (results presented in the online-only Data Supplement), logistic regression models using generalized estimating equations were built for the outcome of worsening angina and freedom from angina to estimate the odds ratio (OR). In the generalized estimating equations model, the OR was estimated across the follow-up year. Cox regression models were constructed for the outcome of new angina and subsequent revascularization to estimate the hazard ratio. In both kinds of models, the interaction effect between the randomized treatment and 1 risk factor was tested for significance. To correct the multiple comparisons, the significance level of 0.0125 was used. SAS 9.2 (SAS Institute Inc) and R (http://R-project.org) were used as the analysis software.
Among the 2364 patients, 1434 (61%) had angina (70% CCS grades I and II, 14% CCS grades III and IV, and 16% stabilized unstable angina with medical treatment), 506 (21%) had anginal equivalents, and 424 (18%) had neither angina nor anginal equivalents. There were 1602 patients in the PCI stratum (806 randomized to MED and 796 randomized to REV) and 762 in the CABG stratum (385 randomized to MED and 377 randomized to REV). The baseline characteristics of the 4 groups are shown in Table 1. There was no significant difference in the baseline characteristics between the REV and MED strategies in the PCI and CABG strata. Compared with the patients in the PCI stratum, patients in the CABG stratum had the following significant differences: They were older; were more often men; had higher blood pressure, rates of previous myocardial infarction and 3-vessel disease, and myocardial jeopardy index; and were taking more β-blocker agents but had a smaller body mass index, had less history of coronary revascularization, and were taking fewer diuretics and antiplatelet agents others than aspirin. Patients in the PCI and CABG strata had similar left ventricular ejection fractions (57.1±11.0% versus 57.4±11.0%; P=0.6).
The mean±SD blood pressure at entry was 131.7±20.0/74.5±11.2 mm Hg, and at the 3-year follow-up was 125.4±15.7/70.3±10.6 mm Hg (P<0.001). The mean hemoglobin A1c was 7.7±1.6% at entry and 7.3±1.3% at the 3-year follow-up (P<0.001). There was an increase in the mean body mass index from entry to 3-year follow-up (31.7± 6.0 versus 32.1±6.3 kg/m2; P=0.02). The mean low-density lipoprotein cholesterol was 96±33 mg/dL at entry and decreased to 80±27 mg/dL at the 3-year follow-up (P=0.001). For the same periods, the average high-density lipoprotein cholesterol was 38±10 mg/dL and increased to 41±11 mg/dL (P<0.001). The current smoking rate decreased by 50% (21.8% to 11.2%; P<0.001) between the year before entry into the trial and at the 3-year follow-up. The changes in the risk factor profile during the follow-up between the MED and REV groups were similar, as reported previously.1 Table 2 shows the antianginal medications in the MED and REV groups at baseline and during the annual follow-up. By the end of first year, more patients in the MED arm shifted from none or monotherapy to polytherapy than patients in the REV arm. From the second year to the fifth year of follow-up, the distribution of patient receiving monoagent or polyagent was stable within each randomized arm. Globally, patients in REV arm received fewer antianginal medications during the entire follow-up.
The percentage of worsening angina was computed for each follow-up year. At the first year of follow-up, 160 of the 1086 participants (15%) in the REV arm and 269 of the 1127 (24%) in the MED arm experienced worsening angina (P<0.001). This significant difference was subsequently documented only in the third year of follow-up. In the PCI stratum, 131 of 742 (18%) having the intervention at entry had worsening angina during the first year of follow-up compared with 186 of the 760 participants (24%) randomized to medical treatment with the option of subsequent revascularization (P<0.001). This significant difference was subsequently documented only during the third year of follow-up. In the CABG stratum, 29 of the 344 participants (8%) randomized to REV experienced worsening angina during the first year of follow-up compared with 83 of the 367 participants (23%) randomized to MED (P<0.001), a significant reduction that persisted for 4 years. Overall, compared with the MED strategy, the REV strategy reduced worsening angina by a third (OR, 0.66; 99% confidence interval, 0.53 to 0.81; P<0.0001); within the PCI stratum, the reduction was 22% (OR, 0.78; 99% confidence interval, 0.61 to 0.99; P=0.007); and within the CABG stratum, the reduction was 67% (OR, 0.33; 99% confidence interval, 0.20 to 0.55; P<0.0001; Figure 1).
Freedom From Angina
Freedom from angina focused on 1434 patients with angina at baseline. For each year of follow-up, freedom from angina was determined if the patient did not report any angina during that year. Overall, a higher percentage of freedom from angina was observed for patients randomized to the REV strategy throughout the first 4 years of follow-up. Patients in the CABG stratum randomized to REV had a higher rate of freedom from angina during the 5-year follow-up compared with patients randomized to MED. In the PCI stratum, freedom from angina was significantly higher in the REV group compared with the MED group only in the first year of follow-up (Figure 2).
The subgroup analysis of new angina was constructed for 930 patients without classic angina at entry. Cumulatively, the MED strategy group had a significantly higher rate of new angina than the REV strategy group at 5 years (59% versus 46%; P<0.001). In the CABG stratum, the REV group compared with the MED group had a lower rate of new angina at the 5-year follow-up (35% versus 61%; P<0.0001). However, in the PCI stratum, the reduced rate of new angina in the REV group compared with the rate in the MED group reached borderline statistical significance (50% versus 58%; P=0.053; Figure 3).
The First Subsequent Revascularization
By 5 years, the first subsequent coronary revascularization was performed in 452 of the 1191 participants (38%) of the MED strategy (275 had PCI and 177 had CABG) and in 240 of the 1173 participants (20.5%) of the REV strategy (179 had PCI and 61 had CABG). The main reasons for the first revascularizations were acute coronary syndrome, severe angina, or documentation of worsening myocardial ischemia (Table 3). Over the 5-year follow-up, the cumulative rate of subsequent revascularizations was significantly higher in the MED arm than in the REV arm. The magnitude of the difference between the REV and MED strategies was much greater in the CABG stratum than in the PCI stratum (Figure 4).
Impact on the 4 Outcomes
At the 3-year follow-up, although the rates of the outcomes were relatively elevated in the REV group, their differences compared with the event rates in the MED group were significant: for worsening angina, 8% versus 13% (P<0.001); for freedom from angina, 66% versus 58% (P=0.003); for new angina, 37% versus 51% (P=0.001); and for subsequent coronary revascularizations, 18% versus 33% (P<0.001). Subgroups analyses showed a greater benefit on each of the 4 outcomes in the REV group in patients without prior revascularization compared with those who have had such an intervention and in the REV CABG group patients with triple-vessel disease compared with those with single- or double-vessel disease (Figures I through IV in the online-only Data Supplement).
Impact of Glucose-Lowering Strategies
The angina outcomes and the subsequent revascularizations did not differ by randomized treatment with insulin sensitizers versus insulin providers. For insulin providers versus insulin sensitizers, the OR of worsening angina was 1.03 (P=0.7); the OR of freedom from angina was 1.01 (P=0.9) after adjustment for follow-up years. The 5-year cumulative rate of new angina was 0.51 for insulin sensitizers versus 0.55 for insulin providers (P=0.3). The 5-year cumulative rate of subsequent revascularizations was 0.32 for insulin sensitizers versus 0.34 for insulin providers (P=0.08).
Prompt REV strategy compared with MED strategy was associated with a lower rate of worsening angina, new angina, and subsequent coronary revascularizations and with a higher rate of patients without angina. For the entire cohort, the symptomatic benefits were significant and may have been attenuated by the higher rates of coronary revascularization in patients in the MED strategy than in those with the REV strategy during the follow-up. However, the option of coronary revascularization in such patients reflects clinical practice and was part of the study design. Although the BARI 2D was not a trial comparing PCI and CABG, the patients in the CABG stratum were at higher risk than the patients in the PCI stratum and had a greater and more sustained benefit on the 4 outcomes with the REV strategy than did the patients in the PCI stratum with the REV strategy.
The clinical presentation of the BARI 2D population reflects the characteristics encountered in patients with diabetes mellitus: angina, anginal equivalents, and pain-free ischemia. During the course of the trial, the medical management in both strategy groups reduced glycemia, blood pressure, lipids, and tobacco consumption to a similar level, but participants had a slight weight increase. Patients in the REV strategy were globally taking fewer antianginal medications than patients in the MED strategy. Although 70% of the population had CCS grade I or II angina and the others had stable or no symptoms at randomization, the participants were nevertheless at high risk; the 5-year mortality in the MED strategy was 10.2% in the PCI stratum and 16.4% in the CABG stratum.
The REV strategy compared with the MED strategy was associated with greater reductions in worsening angina, new angina, and subsequent coronary revascularization and an increased rate of freedom from angina. Freedom from angina and reduction in CCS grades of angina have been associated with an improved quality of life.10,11 This observation has also been confirmed in the BARI 2D trial.12
The benefits of REV versus MED in the PCI stratum were documented mostly during the year after the intervention for worsening angina and freedom from angina. Two previous studies compared a strategy of optimal medical treatment plus prompt PCI and an optimal medical strategy alone; the second Randomized Intervention Treatment of Angina (RITA-2) trial enrolled 1018 patients who were followed up for a median of 7 years,4 and the Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation (COURAGE) trial enrolled 2287 participants who were followed up for a median of 4.6 years.6 However, only 9% of the RITA-2 participants and 33.5% of the COURAGE participants had diabetes mellitus. The RITA-2 and COURAGE studies showed that compared with optimal medical treatment strategy, the early PCI strategy reduced angina and subsequent coronary revascularization more but failed to significantly reduce mortality and myocardial infarction. At the 1-year follow-up, BARI 2D PCI stratum patients had a lower rate of freedom from angina than the COURAGE participants (40% in PCI and 24% in MED compared with 57% in PCI and 50% in MED). The significant benefit lasted only 1 year in BARI 2D compared with 2 years in COURAGE. At the 5-year follow-up, 72% of the participants in the MED arm and 74% in the PCI arm in COURAGE compared with 59% and 61%, respectively, in BARI 2D were free from angina. Revascularization at the 4-year follow-up was lower in COURAGE than in BARI 2D in the MED and PCI arms (30.5% and 19.8% in COURAGE and 38.9% and 27.0% in BARI 2D, respectively). The explanation for lesser and shorter-term benefits in BARI 2D compared with COURAGE is most likely the diabetes mellitus in the BARI 2D cohort, known to be associated with more rapid progression of coronary disease.13 In the original BARI trial, patients with type 2 diabetes mellitus compared with those without diabetes mellitus had the highest rate of new lesions on follow-up angiograms.14
The significant long-term reduction in each of the 4 outcomes was driven mostly by the benefits of REV versus MED in the CABG stratum (Figures 1 through 4). Although it is known that CABG reduces angina more than medical management alone, the present observations constitute new findings in a population like BARI 2D because previous studies enrolled more symptomatic patients, most of them without diabetes mellitus.3 In addition, the finding of a long-term reduction in worsening angina in the CABG arm is consistent with the previously reported reduction of myocardial infarction with this REV strategy compared with the MED strategy.1 These findings have clinical implications. The benefits for the entire cohort should also be considered in the context of the main findings of BARI 2D showing that the REV strategy with CABG reduced cardiovascular events but not mortality.1 Patients, particularly those with triple-vessel disease, randomized to the CABG REV strategy had more relief from symptoms and fewer subsequent revascularizations than those with single- or double-vessel disease. Such patients may benefit from prompt revascularization. For the other patients, optimal medical management is a reasonable initial strategy, and revascularization can be done subsequently if patients are limited by their symptoms despite an optimal medical regimen.
Our study has some limitations. Although the outcomes are relatively soft compared with death or myocardial infarction, they were well defined and monitored during the trials, and the results were consistent. In contrast to RITA-2 and COURAGE participants, all patients in our study did not have angina; the BARI 2D inclusion criteria focused on patients with mild or no symptoms and evidence of coronary disease suitable for revascularization and documented myocardial ischemia. Although with these characteristics, our population is highly selective, it represents the clinical features of several patients with type 2 diabetes mellitus. Finally, in the PCI stratum, nearly one third of the participants assigned to undergo revascularization received a drug-eluting stent; this represents a low rate for such stents in patients with coronary artery disease. However, this type of stent has not been shown to reduce major cardiovascular events but reduces target vessel repeat revascularization.15
Our data show that selected patients with type 2 diabetes mellitus, stable coronary artery disease, and mild or no symptoms but with documented myocardial ischemia had less angina and subsequent coronary revascularization from a strategy of prompt CABG and PCI and optimal medical treatment than from a strategy of optimal medical treatment with delayed or no revascularization. Patients in the CABG stratum at higher risk than patients in the PCI stratum had a greater and more sustained benefit from the prompt intervention and optimal medical treatment than from the optimal medical treatment.
Sources of Funding
BARI 2D is funded by the National Heart, Lung, and Blood Institute and the National Institute of Diabetes and Digestive and Kidney Diseases (U01 HL061744, U01 HL061746, U01 HL061748, U01 HL063804). BARI 2D received significant supplemental funding provided by GlaxoSmithKline, Collegeville, PA; Lantheus Medical Imaging Inc. (formerly Bristol-Myers Squibb Medical Imaging, Inc.), North Billerica, MA; Astellas Pharma US Inc., Deerfield, IL; Merck & Co Inc., Whitehouse Station, NJ; Abbott Laboratories Inc., Abbott Park, IL; and Pfizer Inc., New York, NY. Generous support is given by Abbott Laboratories Ltd, MediSense Products, Mississauga, Canada; Bayer Diagnostics, Tarrytown, NY; Becton, Dickinson and Co, Franklin Lakes, NJ; J.R. Carlson Labs, Arlington Heights, IL; Centocor Inc, Malvern, PA; Eli Lilly and Co, Indianapolis, IN; LipoScience Inc, Raleigh, NC; Merck Sante, Lyon, France; Novartis Pharmaceuticals Corporation, East Hanover, NJ; and Novo Nordisk, Inc, Princeton, NJ. A full listing of all sponsors can be found in the supplementary Appendix at http://www.NEJM.org (N Engl J Med. 2009;360:2503–2515). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Heart, Lung, and Blood Institute; the National Institute of Diabetes and Digestive and Kidney Diseases; or the National Institutes of Health.
Dr Dagenais received lecture fees from GlaxoSmithKline and consultant fees from Abbott. Dr Faxon received grant support from Sanofi-aventis and consultant fees from Boston Scientific and owns stock option from RIVA Medical. Dr Slater has received grant and honoraria for lecture fees from Medtronic. The other authors report no conflicts.
Continuing medical education (CME) credit is available for this article. Go to http://cme.ahajournals.org to take the quiz.
Guest Editor for this article was Robert Eckel, MD.
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.110.978247/DC1.
- Received July 15, 2010.
- Accepted January 28, 2011.
- © 2011 American Heart Association, Inc.
- Rihal CS,
- Raco DL,
- Gersh BJ,
- Yusuf S
- Katritsis DG,
- Ioannidis JPA
- Boden WE,
- O'Rourke RA,
- Teo KK,
- Hartigan PM,
- Maron DJ,
- Kostuk WJ,
- Knudtson M,
- Dada M,
- Casperson P,
- Harris CL,
- Chaitman BR,
- Shaw L,
- Gosselin G,
- Nawaz S,
- Title LM,
- Gau G,
- Blaustein AS,
- Booth DC,
- Bates ER,
- Spertus JA,
- Berman DS,
- Mancini GBJ,
- Weintraub WS
- Weintraub WS,
- Spertus JA,
- Kolm P,
- Maron DJ,
- Zhang Z,
- Jurkovitz C,
- Zhang W,
- Hartigan P,
- Lewis C,
- Veledar E,
- Bowen J,
- Dunbar SB,
- Deaton C,
- Kaufman S,
- O'Rourke RA,
- Goeree R,
- Barnett PG,
- Teo KK,
- Boden WE
- Brooks MM,
- Chung SC,
- Helmy T,
- Hillegass WB,
- Escobedo J,
- Melsop KA,
- Massaro EM,
- McBane RD,
- Hyde P,
- Hlatky MA
The Bypass Angioplasty Revascularization Investigation 2 Diabetes (BARI 2D) trial showed that a strategy of prompt coronary revascularization and optimal medical treatment (REV) was not superior to a strategy of optimal medical treatment with the option of subsequent coronary revascularization (MED) in reducing total mortality and the composite of death, myocardial infarction, or stroke in patients with type 2 diabetes mellitus and stable coronary heart disease. However, does the strategy of prompt revascularization compared with the other strategy reduce worsening angina, new angina, and subsequent coronary revascularization? To answer this question, we evaluated the 2364 BARI 2D participants followed up for 5.3 years. At entry in the trial, 1191 were randomized to MED and 1173 to REV preselected in the percutaneous coronary intervention (n=796) and coronary artery bypass graft (n=377) strata. Patients in the coronary artery bypass grafts stratum had more severe coronary disease and were at higher risk than patients in the percutaneous coronary interventions stratum. Compared with the MED strategy, the REV strategy at the 3-year follow-up had a lower rate of worsening angina (8% versus 13%), new angina (37% versus 51%), and subsequent coronary revascularizations (18% versus 33%) and a higher rate of angina-free status (66% versus 58%). Coronary artery bypass graft stratum patients had the greatest benefits from REV. Thus, patients in the REV strategy had less angina and fewer subsequent coronary revascularizations than patients in the MED strategy. The symptomatic benefits were observed particularly for high-risk patients.