Routine Versus Selective Functional Testing after Percutaneous Coronary Intervention in Patients with Diabetes Mellitus
- Volume 17 - Issue 1 - January, 2005
- Posted on: 8/1/08
- 0 Comments
- 2966 reads
Patients with diabetes mellitus have higher adverse event rates1–4 and a higher risk of restenosis following percutaneous coronary intervention (PCI) compared to non-diabetic patients.5–10 The American College of Cardiology (ACC) and the American Heart Association (AHA) guidelines for exercise testing suggest that functional testing (FT) should not be performed routinely following PCI.11 However, these guidelines leave open the possibility that routine FT may be of benefit in selected high-risk patient groups, including patients with diabetes mellitus. These recommendations are based on the assumption that routine FT will lead to earlier detection and management of restenosis, with a subsequent reduction in adverse events. Few data support the current ACC/AHA recommendations,12 and no randomized trial has examined the utility of a routine post-PCI FT strategy in patients with diabetes. We examined this issue among patients with diabetes enrolled in the Aggressive Diagnosis of Restenosis (ADORE) trial.
Patient population. A total of 348 patients were enrolled at 23 clinical centers in 9 countries between April 25, 1997 and June 30, 2000. The 9-month follow-up period was completed by April 7, 2001. Patients were included in the trial if they had undergone complete coronary revascularization by a percutaneous technique. Patients were randomized immediately following their PCI and before they were discharged from the hospital. Baseline clinical and procedural characteristics were collected at the time of randomization. Diabetes mellitus was defined as hyperglycemia treated with oral hypoglycemic agents or insulin. Patients were identified by review of medical records and by patient self-report. Diabetic status was undetermined in 10 patients. For the purposes of this study, these patients were excluded from the analysis (n = 338).
Patients randomized to the routine functional testing group underwent a regular exercise treadmill test at 1.5 months after PCI and an exercise treadmill test with nuclear perfusion imaging at 6 months. Patients randomized to the selective functional testing arm only underwent FT if they developed a clinical indication. The results of FT were forwarded to the patients’ physicians. The Ethics Committee at each of the institutions involved approved the study, and written informed consent was obtained before patients were randomized.
Endpoints. Nine months post-PCI, patients returned to their clinical center for a maximal endurance exercise treadmill test and an interview with the study nurse. The following questionnaires were administered: the Duke Activity Status Index (DASI),13 the Seattle Angina Questionnaire (SAQ),14 and the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36).15 If tests or events occurred at other institutions, the research nurse contacted that institution to obtain appropriate documentation. Cardiac procedures included cardiac catheterization, repeat PCI, and coronary artery bypass graft surgery (CABG). Clinical events included unstable angina, myocardial infarction (MI), and death. Unstable angina was defined as angina requiring repeat hospitalization after the index PCI. The diagnosis of MI required at least two of the following: clinical symptoms, new ECG evidence of Q waves, and significant elevation in cardiac enzymes. Members of the Endpoints Evaluation Committee, who were blinded to the patient’s treatment assignment, performed endpoint confirmation. The Data Safety and Monitoring Board was responsible for the identification of potential adverse outcomes.
Statistical analysis. The primary analysis examined clinical procedures and event rates in diabetic patients compared to non-diabetic patients. The secondary analysis examined cardiac procedures and event rates in patients with diabetes randomized to the routine FT strategy versus patients with diabetes randomized to the selective arm of the trial. Continuous data are presented as the mean ± standard deviation and were analyzed using the Student’s unpaired t-test or a Mann-Whitney-U test in the case of skewed data. Clinical events and procedures were evaluated using logistic regression to assess the impact of routine FT among diabetics. Dichotomous data are presented as percentages and were analyzed using the Chi-Square test. All statistical tests were two-tailed, and analyses were performed according to the intention-to-treat principle. A p value <= 0.05 was considered significant.
Baseline characteristics. Of 338 patients included in the final analysis, 61 had diabetes mellitus (Table 1). In general, patients were middle-aged men who underwent single-vessel PCI with stent deployment. Approximately one-third of the patients had prior MIs, and 17% had prior PCI. Diabetic patients had a significantly higher prevalence of obesity, hypertension, history of angina and CABG than did non-diabetic patients. Lesion location, complexity and severity were well balanced between the two patient groups. However, there was a trend toward a lower rate of stent use in the diabetic patients (78.7% versus 87.7%, p = 0.10). Relatively small numbers of patients received glycoprotein IIb/IIIa inhibitors at the time of their index PCI procedure.
Outcomes. Nine months after the index PCI, the diabetic patients achieved significantly less metabolic equivalents (METS) on maximal endurance exercise treadmill testing than did the non-diabetic patients (8.5 ± 2.6 versus 9.9 ± 3.1, p = 0.01). Results from the DASI revealed a statistical trend toward worse functional status in the diabetic patients (33.0 ± 16.4 versus 38.2 ± 15.2, p = 0.07) (Table 2). The patients with diabetes mellitus also scored significantly worse than non-diabetic patients on multiple items of the SF-36 health survey, including physical functioning, general health and vitality. There was also a statistical trend toward worsened social functioning in the diabetic patients (Table 2). There were no significant differences among patients with diabetes mellitus and non-diabetic patients in any of the SAQ categories (Table 2).
The composite event rate of death, MI, and unstable angina was significantly higher among diabetic patients than in non-diabetic patients (19.6% versus 8.7%, p = 0.016), though individual event rates did not differ significantly between the two groups (Figure 1). Individual procedure rates including cardiac catheterization (21.4% versus 15.4%, p = 0.27), repeat PCI (10.7 % versus 6.2%, p = 0.23), and CABG (1.8% versus 2.3%, p = 0.81) as well as composite procedures (catherization, PCI and CABG) (23.2% versus 15.5%, p = 0.16) were not significantly different between the two groups (Table 3).
Routine versus selective functional testing. The baseline clinical characteristics of patients with diabetes mellitus randomized to the routine (n = 35) and selective FT arms (n = 26) were similar, although there was a trend toward a higher prevalence of angina among patients randomized to the selective FT strategy (96.2% versus 78.1%, p = 0.06). When compared to the diabetic patients who underwent routine FT, functional status at 9 months was no different among diabetic patients undergoing the selective FT strategy (Table 4). The SF-36 questionnaire detected a borderline significant improvement in general health in diabetic patients undergoing routine compared with selective FT (65.1 ± 21.0 versus 50.1 ± 25.9, p = 0.05). However, there were no other significant differences in any of the quality-of-life measures at 9 months between diabetic patients undergoing routine versus selective FT. There were also no significant differences in any of the quality-of-life measures at 9 months between the non-diabetic patients in the routine versus selective arms of the trial. Diabetic patients in the routine FT arm had a higher composite clinical event rate than those randomized to the selective treatment arm (29.0% versus 8.0% p = 0.05), though individual event rates (Figure 2) and composite procedural rates (25.8% versus 20.0%, p = 0.61) did not differ significantly between the two groups. Logistic regression showed no benefit with routine functional testing among diabetic patients with respect to clinical event rates (OR 2.8; 95%; CI 0.4 – 18.0; p = 0.28).