Review of the NORSTENT trial
Drug-Eluting or Bare-Metal Stents for Coronary Artery Disease
N Engl J Med 2016;375:1242-1252
Background: The first drug-eluting stent (DES) was approved by the FDA in 2003 following the publication of the RAVEL trial. Since then, newer generations of DES were developed and were tested in clinical trials. The majority of trials comparing DES to bare-metal stents (BMS) showed reduction in repeat revascularization with DES but no significant reduction in death or myocardial infarction. Following these publications, the use of DES grew rapidly and was used in more than two thirds of percutaneous coronary interventions (PCI) by 2010.
These trials, however, were very selective, had short follow up time (TAXUS-IV followed patients for 9 months and SPRIT IV followed patients for 12 months), and had limited power to assess hard outcomes.
The NORSTENT trial investigators sought to compare DES to BMS in a more pragmatic design and follow patients for longer time.
Patients: All patients who were undergoing PCI in Norway were assessed for enrollment. Patients had stable angina or acute coronary syndrome. Lesions were in native coronary arteries or bypass grafts.
Patients were excluded if they had prior coronary stents, bifurcating lesions requiring a two-stent technique or life expectancy less than 5 years due to a medical condition other than coronary artery disease. Patients were also excluded if they had contraindications to dual antiplatelets or were taking warfarin.
Baseline characteristics: The trial randomized 9,013 patients – 4,504 randomized to receive a DES and 4,509 to receive a BMS.
The average age of patients was 63 years and 75% were men. Approximately 42% had hypertension, 54% had hyperlipidemia, 10% had prior myocardial infarction, 7% had prior CABG, 12% had diabetes, and 35% were current smokers.
The indication for PCI was stable angina in 29% of the patients, unstable angina in 12% and STEMI or NSTEMI in 58%.
Procedures: The study was open-label but outcomes assessment was blinded. Patients were randomly assigned in a 1:1 ratio to receive DES or BMS. Patients could receive several stents as clinically indicated but can only receive the assigned stent type during the index procedure.
In all patients, aspirin 75 mg daily was given indefinitely while clopidogrel 75 mg daily was given for 9 months.
Follow up visits were done as clinically appropriate without specification from the study protocol. Similarly, no routine follow up coronary angiography was performed.
Endpoints: The primary outcome was a composite of all-cause death or spontaneous myocardial infarction. Secondary outcomes included repeat revascularization, stent thrombosis, major bleeding and health status based on the Seattle Angina Questionnaire.
Clinical outcomes were collected by linking each patient unique national identification number to the Norwegian national patient registry.
Analysis was performed based on the intention-to-treat principle. The study planned to enroll 8,000 patients to be followed for a median of 5 years. Assuming the 5-year event rate of the primary outcome to be 17%, the study would provide 93% power to detect 3% absolute risk difference between the study groups (rate ratio: 1.18). Due to lower than expected mortality, the sample size was increased to 9,000 patients
Results: Among the 20,663 patients who were assessed for eligibility, 12,425 met inclusion criteria. Among patients who met inclusion criteria, 9,013 were randomized. Figure 1 in the manuscript provides details for excluding patients and for not randomizing patients who met eligibility criteria. The most common reason for exclusion was prior PCI.
The number of stents implanted per patient was 1.7 and more than 98% received the assigned stent type. The median follow up time was 5 years.
The primary composite outcome of all-cause death or nonfatal spontaneous myocardial infarction was not significantly different between both treatment arms (16.6% with DES vs 17.1% with BMS, HR: 0.98; 95% CI: 0.88 - 1.09; p= 0.66).
For the secondary outcomes – Hospitalization for unstable angina was similar between treatment groups (5.2% vs. 5.7%; p= 0.21). Stent thrombosis was lower with DES (0.8% vs 1.2%; p= 0.05). Target-lesion revascularization was also lower with DES (5.3% vs 10.3%; p< 0.001). Bleeding Academic Research Consortium (BARC) 3, 4 or 5 was similar between groups (5.5% vs 5.6%; p= 0.88).
There was no significant difference in health status based on the Seattle Angina Questionnaire.
There were no significant subgroup interactions.
Conclusion: In patients undergoing PCI, the use of DES did not reduce the composite endpoint of death or spontaneous myocardial infarction compared to BMS. Target-lesion revascularization was reduced with DES with a number needed to treat of 20 patients.
The findings of this study align with the results of other trials comparing DES to BMS. We have reviewed several key trials and included links to additional studies in this field below. Overall, DES significantly reduce target-lesion revascularization without significant effect on all-cause mortality or myocardial infarction.
An important consideration in this and other related trials is that both stent types were studied using similar durations of dual antiplatelet therapy (DAPT) following PCI. For patients with stable angina, BMS typically require only one month of DAPT, while DES often necessitate three to twelve months. Since shorter durations of DAPT are generally safer for patients, a trial comparing DES with three to twelve months of DAPT compared to BMS with one month of DAPT would be insightful.
A final teaching point is that less than 50% of screened patients were ultimately enrolled in this pragmatic trial, which had minimal exclusion criteria. It’s not uncommon for trials to enroll less than 5% of screened patients which limits their external validity.
Other trials of DES vs BMS
The advent of DES was spurred by the need to address the Achilles heel of BMS (ie. neointimal hyperplasia). It’s essentially a focal recurrence of “stable CAD” via a different mechanism than atherosclerosis.
Knowing what we know now of stable CAD, it is completely congruent that DES would offer no advantage on hard outcomes over BMS, and that benefits would only accrue to reduction in TLR.
It is interesting that there was no signal of increased stent thrombosis with DES (which is its own Achilles heel) with 9 mos of DAPT, which is generally halfway btw the “recommended” duration for stable CAD PCI (6 mos) vs ACS PCI (12 mos). Perhaps this simply reflects the patient population.