Utility of Routine Functional Testing After Percutaneous Transluminal Coronary Angioplasty: Results from the ROSETTA Registry

ABSTRACT: Background. There is little consensus regarding the use of functional testing after percutaneous transluminal coronary angioplasty (PTCA). Some physicians employ a routine functional testing strategy, and others employ a symptom-driven strategy. Objective. To examine the effects of routine post-PTCA functional testing on the use of follow-up cardiac procedures and clinical events. Methods. The Routine versus Selective Exercise Treadmill Testing after Angioplasty (ROSETTA) Registry is a prospective multicenter observational study examining the use of functional testing after PTCA. A total of 788 patients (pts) were enrolled in the ROSETTA Registry at 13 clinical centers in 5 countries. The frequencies of functional testing, cardiac procedures and clinical events were examined during the first 6 months following a successful PTCA. Results. Patients were predominantly elderly men (mean age, 61 ± 11 years; 76% male) who underwent single-vessel PTCA (85%) with stent implantation (58%). During the 6-month follow-up, a total of 237 pts were observed to undergo a routine functional testing strategy (100% having functional testing for routine follow-up), while 551 pts underwent a selective (or clinically-driven) strategy (73% having no functional testing and 27% having functional testing for a clinical indication). Patients in the routine testing group underwent a total of 344 functional tests compared with 165 tests performed in the selective testing group (mean, 1.45 tests/patient versus 0.3 tests/patient; p 1 Because it is difficult to predict which patients will develop restenosis, some physicians employ routine functional testing in all their post-PTCA patients.² The rationale for this approach is that routine functional testing will lead to the early identification and treatment of patients with restenosis, which may, in turn, lead to a reduction in clinical events. In contrast to a routine functional testing strategy, many physicians employ a selective or clinically-driven functional testing strategy.² With this strategy, functional testing is only employed in patients who develop symptoms. Selective functional testing will not identify patients with asymptomatic restenosis, and this approach may be associated with a higher rate of follow-up clinical events. The American College of Cardiology (ACC) and the American Heart Association (AHA) have issued guidelines for exercise testing.³ These guidelines suggest that routine post-PTCA functional testing be reserved for selected groups of patients such as those with decreased left ventricular ejection function (LVEF), multivessel coronary artery disease, proximal left anterior descending coronary artery disease and diabetes mellitus. If patients do not have any of these clinical characteristics, a selective functional testing strategy is recommended. Unfortunately, there are few data to support the ACC/AHA guideline recommendations regarding post-PTCA functional testing. To our knowledge, no prospective study has examined the utility of a routine post-PTCA functional testing strategy — in either selected subgroups of patients or in the general post-PTCA population. For this reason, we initiated the Routine versus Selective Exercise Treadmill Test after Angioplasty (ROSETTA) Registry.⁴ The purpose of this registry was to examine the patterns of use and outcomes of the routine and selective functional testing strategies. The primary objective was to examine the effects of routine post-PTCA functional testing on follow-up cardiac procedures and clinical events. Methods Study centers and patient population. Patients were enrolled in the ROSETTA Registry between February 8, 1996 and December 3, 1997. A total of 826 patients were enrolled at 13 clinical centers in 5 countries (Australia, Canada, Israel, Singapore, United States; Appendix 1). Over the 6-month follow-up period, thirty-five patients (4.2%) were lost to follow-up, leaving a total of 791 patients. An additional 3 patients could not be classified into the routine or selective group because of early death post-PTCA. Therefore, data from a total of 788 patients were available for analysis. Patients were enrolled in the study immediately after their PTCA procedure, but before they were discharged from hospital. The study was approved by the Research and Ethics Committee at the institutions involved, and written informed consent was obtained before patients were enrolled in the study. Inclusion criteria included a successful PTCA of at least one coronary lesion, with a successful PTCA being defined as a reduction of at least one lesion from a stenosis of >= 60% to a residual of Baseline clinical and procedural characteristics. At the time of patient entry into the study, baseline and procedural information was collected. Baseline information included demographic data as well as clinical information, such as comorbid conditions, prior procedures and medical therapy. Procedural information included numbers and types of lesions, devices used, pre- and post-procedure stenosis severity, and Thrombolysis in Myocardial Infarction (TIMI) coronary blood flow. Six-month follow-up. Six months after PTCA, the study nurse at the patient’s clinical center contacted the patient to determine whether post-procedure functional testing had taken place. If necessary, the nurse contacted the patient’s cardiologist for 6-month follow-up information. If the patient underwent functional testing during the follow-up period, documentation of the type and results of the testing was obtained. If tests or other procedures occurred at other institutions, the research nurse contacted these institutions to obtain the appropriate documentation. Routine functional testing was defined according to the reason for the first functional test performed after the index PTCA. If the first test was performed as a routine follow-up, the patient was considered to be in the routine functional testing group. If the patient had their first functional test for a clinical reason (e.g., typical or atypical symptoms) or if the patient underwent no functional testing during the 6-month follow-up, he was considered to be in the selective functional testing group. Data analysis. Continuous data are presented as mean ± standard deviations and dichotomous data are presented as percentages. The primary data analysis examined the proportions of patients undergoing functional tests and cardiac procedures among the two testing groups. We then examined the occurrence of clinical events, including hospitalization for unstable angina, myocardial infarction, death and a composite endpoint including all three outcomes. To examine the role of the testing strategy and other potential predictors of the composite endpoint, we performed both univariate and multivariate analyses. Dichotomous predictors were examined using Fisher’s Exact tests, and continuous predictors were examined using logistic regression. If a particular variable was associated with the composite endpoint with a p-value Clinical and procedural characteristics. Table 1 shows baseline clinical characteristics according to whether patients subsequently underwent a routine or selective functional testing strategy. In general, patients were elderly men with multiple risk factors for coronary artery disease. Because patients were not randomized into the routine and selective functional testing strategies, imbalances in clinical differences between the two groups are evident. Patients who subsequently underwent routine functional testing were younger, were more likely to be men, and were more likely to have a family history of coronary artery disease. In contrast, patients who subsequently underwent selective functional testing were more likely to have Canadian Cardiovascular Society Class III-IV angina at the time of their PTCA and were more likely to have hypertension. There were also imbalances in the procedural characteristics of the patients who subsequently underwent the two functional testing strategies (Table 2). Patients who underwent routine functional testing were more likely to have ACC/AHA lesion types A or B1, while patients who underwent selective functional testing were more likely to receive a stent, to have undergone atherectomy, and to have a slightly greater residual stenosis after PTCA. Functional testing. Patterns of functional testing use among patients enrolled in the ROSETTA Registry have been previously described.4 Among the 788 patients, a total of 237 (30%) were observed to undergo a routine functional testing strategy. In contrast, a total of 551 patients (70%) underwent a selective functional testing strategy in which they either underwent no functional testing during the 6-month follow-up (73%) or they underwent one or more functional tests in which the first was done for a clinical indication (27%). Patients in the routine functional testing group were more likely to undergo two or more functional tests than patients in the selective group (35% versus 2%, respectively; p Follow-up cardiac procedures. There was little difference in the use of follow-up cardiac procedures among patients in the 2 groups (Figure 1). Patients in the routine and selective groups had similar rates of follow-up cardiac catheterization (13.9% versus 17.5%; p = NS), PTCA (8.4% versus 8.7%; p = NS) and CABG (2.1% versus 3.3%; p = NS). Among the 131 patients who had repeat cardiac catheterization procedures during the 6-month follow-up, the major indications were unstable angina (36%), recurrent chest pain (35%) and stable angina (12%). A positive functional test was an infrequent reason for repeat cardiac catheterization (8%). Thus, cardiac procedures were not associated with a reduction in clinical events. Reduction in clinical events could be due to more intensive medical therapy in response to functional test results or it could be due to PCI/CABG on patients who really need it. In other words, even though the number of procedures did not increase, they were done in the patients with demonstrable ischemia. Because of the large difference in the number of functional tests used in the routine and selective groups, the yield of functional testing (expressed as the number of functional tests over the number of follow-up cardiac procedures) was more favorable among patients who underwent a selective functional testing strategy. As an example, patients in the routine testing group underwent 10.4 functional tests for every cardiac catheterization performed, compared to 1.7 in the selective group (p Clinical outcomes. In spite of similar rates of cardiac procedures, patients in the routine functional testing group were much less likely to experience a clinical event during the 6-month follow-up period, including unstable angina (6.1% versus 14.4%; p = 0.001), myocardial infarction (0.4% versus 1.6%; p = NS), death (0% versus 2.2%; p = 0.02) and composite clinical events (6.3% versus 16.3%; p Relationship between functional testing and clinical outcomes. In order to examine the impact of functional testing strategy on clinical outcome, we performed regression modeling to control for clinical and procedural differences between the 2 groups and to examine whether routine functional testing was independently associated with a reduction in composite clinical events. Univariate analyses demonstrated that a number of clinical and procedural variables were associated with the composite endpoint (Table 3). Functional test results were included in the analyses, and none were found to be significant in predicting reduced events. We then performed a multivariate logistic regression analysis, which demonstrated that after simultaneously controlling for pre-determined clinical and procedural characteristics, routine functional testing had a persistent independent association with a reduction in the composite clinical event rate (odds ratio, 0.45; 95% confidence interval, 0.24–0.81; p = 0.008). Discussion Our study was designed to examine the effects of post-PTCA functional testing on the use of follow-up cardiac procedures and clinical events. We found that there was a five-fold difference in the intensity of functional testing between patients who underwent the routine and selective testing strategies. Patients in the routine testing group underwent an average of 1.5 functional tests during the 6-month follow-up period versus 0.3 tests for patients in the selective testing strategies. Although routine functional testing was not associated with an increase in the use of cardiac procedures, it was associated with a reduction in cardiac events. The association between routine functional testing and a reduction in follow-up events may be attributable to the early identification and treatment of patients at risk for follow-up events, or it may be due to clinical differences between patients who undergo routine and selective testing. The major benefit of routine testing was reassurance for patients and physicians, and that chest pain was less likely to result in hospitalization. These results are significant because very few prospective studies have been conducted to examine the clinical value of routine functional testing after PTCA. For this reason, there is little consensus on the appropriate functional testing strategy to use. As demonstrated in this study, some physicians employ a routine testing strategy in the hopes of identifying restenosis early and then instituting therapy to avoid clinical events.² Other physicians employ a “watchful waiting” approach, and employ functional testing only if symptoms develop.² To our knowledge, this is the first study to prospectively examine this issue. Previous studies. Previous studies have primarily examined post-PTCA functional testing with respect to its ability to identify restenosis.^5–10 We previously showed in a meta-analysis that ETT alone is poorly diagnostic of restenosis and that the addition of nuclear or echocardiographic imaging significantly improves the diagnosis.¹¹ ETT alone has an estimated sensitivity of 46% and a specificity of 77% for the identification of post-PTCA restenosis.¹¹ The use of stress nuclear imaging increases the sensitivity to 87% and the specificity to 78%, and the use of stress echocardiographic imaging increases the sensitivity to 63% and the specificity to 87%.¹¹ The BENESTENT and STRESS trials suggest that routine stenting decreases restenosis rates by 10%.^12,13 Because of the increasing use of coronary stenting and because predictive values are highly dependent on the pretest probability of restenosis, the diagnostic ability of routine functional testing is declining substantially.¹¹ Depending on the diagnostic accuracy of functional testing, it has been estimated that routine functional testing with exercise thallium scans would cost $8,200–22,400 to identify an asymptomatic patient with restenosis following PTCA.¹⁴Study limitations. Several potential limitations of our study should be noted. First, misclassification of patients with respect to functional testing strategy could have occurred. For example, patients whose first functional test was done for a clinical reason were included in the selective functional testing group. If these patients had remained asymptomatic, however, their physicians might have sent them for routine testing. Similarly, patients who developed a clinical event and then underwent a repeat cardiac catheterization without functional testing might have had routine testing if they had remained asymptomatic. If misclassification occurred, however, if would lead to a reduction in the differences found between the routine and selective testing groups. A second potential limitation of our study is its observational design, because patients were not randomized to the routine or the selective groups. In addition, different strategies were used at different centers with respect to the routine/selective type of functional test as well as timing. Nevertheless, our study reflects current clinical practice and is thus a “snapshot” of what is taking place in the real world. Finally, although we found that routine functional testing is associated with a reduction in clinical events, these results may well be explained by clinical differences between patients in the routine and selective groups. Although we controlled for a wide array of clinical and procedural differences in our multivariate analysis, it is still possible that other confounding variables were present that may explain our findings (i.e., selection bias). Conclusion. In conclusion, our study was designed to examine the effects of post-PTCA functional testing on the use of follow-up cardiac procedures and clinical events. We found that although there was an almost five-fold difference in the use of functional testing between the routine and selective groups, this difference did not translate into an appreciable difference in the use of follow-up cardiac procedures. However, routine functional testing after PTCA is associated with a reduction in the frequency of follow-up clinical events. This association may be attributable to the early identification and treatment of patients at risk for follow-up events, or it may be due to clinical differences between patients who are referred for routine and selective functional testing.