On-Label and Off-Label Use of Drug-Eluting Stents

Abstract

In this retrospective study, we compared the in-hospital and long-term outcomes of the on-label and off-label uses of drug-eluting stents.

From April 2003 through June 2007, 1,538 patients underwent percutaneous coronary intervention with a drug-eluting stent (sirolimus or paclitaxel) at Tehran Heart Center. Off-label implantation of the drug-eluting stent was as implemented on the basis of specific clinical and procedural characteristics set forth in our text. There were 708 patients in the on-label group and 830 in the off-label group.

Baseline characteristics were not significantly different between the groups. Histories of non-ST-segment-elevation myocardial infarction, percutaneous coronary intervention, and coronary artery bypass grafting were more prevalent in the off-label group. Both groups had similar procedural and in-hospital complications. The follow-up rate at 1 year was 93.1% in the on-label group and 93.3% in the off-label group. During that period, the occurrence of major adverse cardiac events was not significantly different between the groups. After 1 year between the respective on- and off-label uses of the sirolimus-eluting and paclitaxel-eluting stents, and after adjustment for diabetes mellitus, myocardial infarction, percutaneous coronary intervention, and coronary artery bypass grafting, there was no remarkable difference in the occurrence of major adverse cardiac events (hazard ratio, 0.688; 95% confidence interval, 0.365–1.295; P=0.2463) or target-vessel revascularization (hazard ratio, 0.69; 95% confidence interval, 0.291–1.636; P=0.3993).

We found that off-label use of drug-eluting stents was safe after 1 year and that such use was not associated with increased in-hospital myocardial infarction or death.

Use of drug-eluting coronary stents (DESs) has improved the outcomes in patients undergoing percutaneous coronary intervention (PCI) and resolved the limitations of balloon coronary angioplasty and the use of bare-metal stents.

The sirolimus-eluting coronary stent (SES) is indicated for improving the coronary luminal diameter in patients with symptomatic ischemic disease that is due to discrete de novo lesions of length ≤30 mm in native coronary arteries with a reference-vessel diameter of ≥2.5 and ≤3.5 mm. Indications for the paclitaxel-eluting stent (PES) differ only in terms of lesion length (<28 mm) and vessel diameter (>2.5 to <3.75 mm).^1,2

Over the years, implantations of DESs have extended beyond on-label indications to what are called off-label uses. The off-label use of DESs has become common in clinical practice and has been estimated to account for at least 60% of DES usage.³ The DES is now commonly used off-label in long lesions (>30 mm), bypass grafts, in-stent restenotic lesions, acute myocardial infarction (MI), left main coronary artery stenosis, and ostial lesions.^4–10 Some studies have concluded that off-label use of the DES is safe and associated with a low rate of short-term adverse events,^11,12 whereas other studies have shown that safety is poor in primary and long-term outcomes.¹³

In this retrospective study, we compared the in-hospital and 1-year outcomes of off-label and on-label DES implantations.

Patients and Methods

This study involved a consecutive series of 2,965 patients who had presented with coronary artery disease and who were treated in our hospital with a DES in on- or off-label implantation (subject to stent availability) from April 2003 through June 2007. A total of 1,538 patients who were exclusively treated with either a SES or a PES were eligible for inclusion in the analysis. The study protocol was approved by our institutional review board.

The data were extracted from a prospective computerized registry of patients in Tehran Heart Center. This registry comprised detailed information on baseline, angiographic, and procedural characteristics in addition to in-hospital outcomes. We obtained information about major adverse cardiac events (MACE), including target-vessel revascularization (TVR), nonfatal MI, and cardiac death, in formal telephone interviews or during patients' visits to our clinic. The patients in this study were monitored at 1, 6, and 12 months after their stent implantations.

Patients who had undergone PCI immediately after sustaining an MI and those in whom PCI was unsuccessful were excluded. Off-label use was defined as implantation of a DES for left main coronary artery lesions; occlusions; ostial, bifurcation, and restenotic lesions; lesions in bypass grafts; a left ventricular ejection fraction <0.30; a history of MI; lesion lengths >30 mm and reference-vessel diameters outside a range of 2.5 to 3.5 mm for an SES, and lesion lengths >28 mm with vessel diameters <2.5 mm or >3.75 mm for a PES.

Angioplasty Procedures

We previously reported a procedural description and detailed definition of variables.^14,15 In summary, patients were given 325 mg of aspirin before the procedure, and coronary stenting was performed by routine methods. No glycoprotein IIb/IIIa blocker was given to any patient. In case of dissection or incomplete coverage of a lesion, additional DESs were used as necessary. Stents were deployed with or without predilation, in accordance with standard techniques. After stent implantation, patients were prescribed lifelong aspirin (80–325 mg/d) and clopidogrel (75 mg/d) for 6 to 24 months.

Procedural success was defined as residual stenosis <20% and angiographically determined Thrombolysis in Myocardial Infarction grade 3 flow without major complications (death, MI, or emergency bypass surgery) during hospitalization. A MACE was defined as cardiac death, MI, or TVR. Target-vessel revascularization was defined as clinically driven percutaneous revascularization or bypass of the target lesion or any segment of the epicardial coronary artery that contained the target lesion. Target-lesion revascularization was defined as any repeat revascularization procedure (percutaneous or surgical) of the original target-lesion site.¹

Statistical Analysis

Results are reported as mean ± SD for quantitative variables and are summarized by absolute frequencies and percentages for categorical variables. Continuous variables were compared using the independent 2-sample t test or nonparametric Mann-Whitney U test when the presumption of normality was violated. The Kolmogorov-Smirnov test was applied to test for a normal distribution. We compared categorical variables with use of the χ² or Fisher exact test when we observed more than 20% of cells in the output table with an expected count of fewer than 5 patients. The Kaplan-Meier method of survival analysis was used to estimate time-related events of MACE or TVR. The log-rank test was applied to test for the similarity of event-free survival curves across the 2 groups of on- and off-label patients.

Multivariate Cox proportional hazards models for comparing long-term events between the 2 groups were constructed, with consideration of confounders affecting outcome variables, including diabetes mellitus and histories of MI, PCI, and coronary artery bypass grafting (CABG). Covariates were selected by inter methods and those that were considered biologically relevant irrespective of their statistical significance. The association was expressed as hazard ratios with 95% confidence intervals (CIs).

For the statistical analysis, SPSS version 13.0 for Windows (SPSS, part of IBM Corporation; Somers, NY) and SAS version 9.1 for Windows (SAS Institute, Inc.; Cary, NC) were used. P values less than 0.05 were considered statistically significant.

Results

The final study population consisted of 1,538 patients who were treated with SES or PES implantation: 708 patients (46%) had on-label and 830 (54%) had off-label indications. Table I shows that baseline characteristics were not significantly different between the groups except for hypertension, which presented more frequently in the on-label group, and histories of non-ST-segment-elevation MI, PCI, and CABG, which were more prevalent in the off-label group.

TABLE I. Demographic and Clinical Characteristics of 1,538 Patients Who Underwent On- and Off-Label Implantation of Drug-Eluting Stents

Procedural and Angiographic Characteristics

Table II shows the procedural and angiographic characteristics. Significant differences were noted between on- and off-label circumstances, and it is reasonable that these differences were driven predominantly by the characteristics of the lesions: bifurcated, diffuse, occluded, ostial, preprocedurally stenotic, and type B2/C of the American College of Cardiology/American Heart Association classification. Lesions tended to be longer and of larger diameter in the off-label group; stent sizes also showed corresponding differences.

TABLE II. Angiographic and Procedural Characteristics of 1,538 Patients Who Underwent On- and Off-Label Implantation of Drug-Eluting Stents

In-Hospital Event Rates

In-hospital death, non-Q-wave MI, and stent thrombosis were compared between the on- and off-label uses of the SES and PES (Table III); there was no significant difference between the uses.

TABLE III. In-Hospital and Long-Term Clinical Outcomes of Patients Who Underwent On- and Off-Label Implantation of Drug-Eluting Stents

Long-Term Outcomes

The occurrence of MACE upon 1-year follow-up did not differ between the groups. The follow-up rate was 93.1% in the on-label group and 93.3% in the off-label group. During a mean follow-up period of 11.67 ± 1.4 months, 3 patients died in the on-label group (0.4%) versus 2 in the off-label group (0.2%).

After adjustment for the confounding variables, there was also no remarkable difference in the occurrence of MACE (hazard ratio, 0.688; 95% CI, 0.365–1.295; P=0.2463) or TVR (hazard ratio, 0.69; 95% CI, 0.291–1.636; P=0.3993) upon 1-year follow-up between the on- and off-label uses of the SES and PES. Figures 1 and 2 show the Kaplan-Meier survival curves for freedom from TVR and MACE.

Fig. 1 Kaplan-Meier curves for freedom from target-vessel revascularization (TVR) at 1 year after percutaneous coronary intervention (PCI) after the on- and off-label use of drug-eluting stents.

Fig. 2 Kaplan-Meier curves for freedom from major adverse cardiac events (MACE) at 1 year after percutaneous coronary intervention (PCI) after the on- and off-label use of drug-eluting stents.

We compared on- and off-label outcomes in SES and PES separately. After 1 year, there was no significant difference in MACE in either subgroup of SES and PES recipients: 3.3% for on-label use versus 1.4% for off-label use in the PES group and 2.7% versus 2.8% in the corresponding SES groups.

Discussion

This study compared the clinical outcomes of patients who underwent on-label versus off-label intracoronary DES implantation. Off-label indications accounted for more than 60% of the PCI procedures in routine clinical practice; however, the appropriate therapeutic approach for patients requiring coronary revascularization in terms of on- or off-label use of DESs is still unclear because of conflicting evidence.¹⁶

Our main finding was that off-label implantation of a DES did not lead to an increased risk of procedural and in-hospital complications or of MACE upon 1-year follow-up. These findings were not significantly different between the on- and off-label uses of the DES with respect to the univariate analysis and after adjustment for the confounding factors.

The frequency of off-label DES implantation was more than 50% in our institution; this value is compatible with the proportion reported by other data registries worldwide.^13,17–19

Pivotal trials have reported that DESs significantly decrease rates of restenosis, clinical events, repeat revascularization, and MACE when compared with bare-metal stents.^1,20–24 Several studies have compared the outcomes of DESs and bare-metal stent implantation in ostial lesions, bifurcated lesions, and chronic occlusions, and a reduction has been shown in the rates of TVR and MACE in the DES groups.^14,25,26 Investigators have also compared the DES and bare-metal stent in off-label indications and concluded that off-label DES use yields lower rates of death and TVR than does off-label use of the bare-metal stent.^18,19,27

The clinical trial from the American College of Cardiology National Cardiovascular Data Registry³ reported on off-label use in 4 off-label indications (MI, in-stent restenosis, saphenous vein grafts, and chronic occlusion) in clinical practice. This multicenter data registry reported that off-label implantation is common and increasingly frequent, and that off-label DES use appears to yield low rates of adverse events.¹² Qasim and colleagues¹³ also documented that off-label DES use was safe and not associated with increased stent thrombosis, MI, or death. In accordance with these observations, our findings showed that the occurrence of in-hospital major complications and MACE during a 1-year follow-up period did not differ in respect to on- and off-label implantations of the DES. The rates of in-hospital death, MI, and stent thrombosis were low in each group, although the prevalence of death and MI trended higher in the off-label group. In the year after the procedure, rates of MI and TVR were lower in the off-label group, although not to a significant level. Several reports have indicated that off-label DES use increases the risk of adverse events in comparison with on-label DES use.^13,16,28 After comparing off-label and untested uses of the DES with standard uses, Beohar and colleagues¹⁷ reported that relative early safety was lower in off-label use and that long-term effectiveness was lower in both off-label and untested uses; however, the event rates were low (<1%) in both approaches. Be that as it may, despite all the reports of a higher risk of adverse events in the short term, the rates of absolute events are low in the off-label use of the DES.^11,12,16

Study Limitations

Chief among the limitations of the present study is that it was conducted in a single center. Patients who underwent multiple-lesion PCI had to be excluded, because otherwise the attribution of MACE to a particular lesion site would not have been possible. In addition, the complication rate was generally low, so a larger sample is required if any meaningful difference is to be observed.

Conclusion

In conclusion, our main finding was that the off-label implantation of DESs did not yield an increased risk of procedural complications or in-hospital events. Furthermore, the rate of MACE upon 1-year follow-up was not significantly different between the on- and off-label implantations of DESs.

Acknowledgment

We thank Neda Karimi for her assistance with data entry.