At this week’s annual European Society of Cardiology Congress in Rome, an important randomized clinical trial on stents was presented by Professor Kaare H. Bønaa, MD, PhD of the Clinic for Heart Disease, St. Olav’s University Hospital in Trondheim, Norway. Called NORSTENT, short for the “Norwegian Coronary Stent Trial,” this was the largest stent trial ever conducted, with 9,013 patients followed for six years. That’s serious!
A Reaction to the “Firestorm” of 2006
The NORSTENT trial was started in 2008, shortly after the “firestorm” of the 2006 European Society of Cardiology meeting where, for the first time, serious concerns were raised about the long-term safety of drug-eluting stents (DES). DES originally were designed to reduce restenosis (reblocking of the artery) by eluting an anti-proliferative drug through a plastic polymer. But late stent thrombosis (LST) was being seen in patients who had been implanted with drug-eluting stents more than a year post-procedure. It was thought that the polymer was at fault. This led to a bit of a panic, characterized by the NBC News report of drug-eluting stents as “tiny time bombs in your heart.” The FDA called a two-day hearing (at which I testified as a patient advocate) to determine what the danger, if any, really was.
It turned out that the available data from clinical trials of the first generation DES made by Cordis (Cypher) and Boston Scientific (Taxus) had been designed to get FDA approval and only tracked about 1,000 patients for 9-12 months. And since LST was a relatively infrequent event (although a catastrophic one) it didn’t show up in such a small sample. So LST never got measured. This lack of long-term data was strongly criticized by the FDA panel, including Drs. Eric Topol, Robert Harrington, Steve Nissen, et al. The result of this two-day meeting was a recommendation for patients who got DES to take dual antiplatelet therapy (DAPT) of clopidogrel and aspirin for a full year or more to mediate the potential negative effects from LST. Those patients who received bare metal stents (BMS) with no polymer only needed three months.
New Generation DES
Simultaneously with these issues, both Medtronic and Abbott Vascular were developing their second generation DES, the Endeavor and Xience (which subsequently was also marketed by Boston Scientific as Promus). These newer DES had thinner struts, more biocompatible drug-eluting polymers, and more flexible structures. Medtronic’s Endeavor was soon replaced by an even more efficacious DES, the Resolute.
It is these second generation DES that NORSTENT studied, comparing them to the newest iteration of BMS (stents with no polymer and therefore potentially “safer”). DAPT was given for nine months for both types of stent. (Of note, this week the American College of Cardiology revised its guidelines for DAPT, stating that, for those patients with stable heart disease who received the newer generation DES, six months of DAPT was sufficient.)
The primary outcome for NORSTENT was defined as a composite of death from any cause and nonfatal spontaneous myocardial infarction after a median of 5 years of follow-up. Secondary outcomes included repeat revascularization, stent thrombosis, and quality of life. Interestingly, the original enrollment for NORSTENT was 8,000 patients but, because the event rate was lower than predicted (translate this to “the stents performed better than we thought they would”) the number was expanded to 9,000 patients and followup to six years.
There were a number of excluding factors for enrolling patients, a significant one being no prior stent/PCI. Also patients had to be treatable with either a BMS or DES, which might exclude, for example, diabetic patients with narrow arteries – most interventional cardiologists would opt for a DES, if PCI were the chosen treatment.
The bottom line was that the primary outcome of death or heart attack after six years was 16.6% in the DES group and 17.1% in the BMS group – no statistically significant difference between DES and BMS in terms of mortality or myocardial infarction. The investigators did report a small difference in total repeat revascularization; 16.5% of DES recipients had to get a second procedure, as contrasted with 19.8% in the BMS group.
The full study was published simultaneously in The New England Journal of Medicine.
Total Revascularization or TLR?
When I first saw these data, my immediate reaction was “total revascularization?” What about “Target Lesion Revascularization” or TLR – defined as having to re-open the blocked area that was stented? Total revascularization refers to any repeat procedure, which may be an area not previously treated. Coronary artery disease is chronic, and a repeat procedure in another area may be due to disease progression rather than stent failure. So I wrote Prof. Bønaa and he replied that TLR was actually 5.3 % for DES and 10.3 % for BMS, a p-value of <0.001. That means that DES cut the restenosis of the treated blockage by half!
While 10.3% TLR after six years is a pretty good result for any stent, 5.3% is a lot better – twice as good, actually. It turns out that the promise of DES, lower restenosis, is totally born out by NORSTENT. After six years only 5 out of 100 patients needed to have a repeat procedure for the same blockage.
Stent Thrombosis and Quality of Life Outcomes
As for the original concern of late stent thrombosis, the data also favored DES, which showed 0.8% vs. 1.2% for BMS, resulting in a p-value = 0.0498, just edging into statistical significance. Again, these outcomes for both stents are very good and should put to rest the decade-old concerns about increased stent thrombosis with DES. (FYI, stent thrombosis at one year for the newly approved Absorb stent was 1.5% in the ABSORB III trial.)
Another area where DES and BMS were judged “equivalent” was in Quality of Life (QoL), as measured by the Seattle Angina Questionnaire (SAQ). But the SAQ survey doesn’t track those patients who had to have a repeat procedure. In-Stent Restenosis (ISR) is a difficult clinical situation and is not readily fixed. So cutting the ISR rate in half is a significant plus for DES. I would put forth that those patients who had to have a repeat procedure would not be reporting a high QoL score.
As is often the case, this study found its way into the medical and mainstream media with a variety of misleading headlines, or as I’ve called them, “dreadlines.” My favorite was “New generation drug eluting stents show lower efficacy than contemporary BMS, study shows” from News-Medical.Net. Totally wrong. But also MedPageToday, stating “ESC: DES or BMS? No Mortality Difference,” when there is a big difference in TLR. And finally, “Less efficacy than expected in largest drug-eluting stent trial,” when the TLR of 5% is exactly what has been predicted for DES for a decade.
Upon presentation of the results for NORSTENT, which did not highlight the significant reduction in TLR, a very active thread emerged on Twitter where various cardiologists debated the results and the implications of the study. Dr. Sukh Nijjer @sukhnijjer from London tweeted, “Palpable audience reaction to NORSTENT presentation at Hotline session: no mention of TLR. Disingenuous. Harmful. Shameful.” And @Jeddacath tweeted, “Very misleading headline. Can be misinterpreted by patients & doctors Wrong endpoint Ignoring 50% ⬇️TLR.” And more. In fact, you should definitely read the editorial comment in the New England Journal of Medicine from Dr. Eric Bates, “Balancing the Evidence Base on Coronary Stents.”
Bottom line on NORSTENT
The current devices, bare metal stents and drug-eluting stents, are performing very well. And DES, as predicted, is halving the restenosis rate and saving patients from having to undergo a repeat procedure for in-stent restenosis, which has a relatively poor outcome. There are certainly exceptions where BMS are preferable: large arteries, patients who cannot tolerate any DAPT, etc. But, aside from these cases, most interventional cardiologists would agree that a DES is preferable.
The NORSTENT study didn’t deal with cost-effectiveness. Although Dr. Bønaa stated that the number needed to treat (NNT) was 30 (30 patients needed to be treated with a DES to eliminate one revascularization) if you looked at the TLR figures (which are more to the point) the NNT was only 20, basically what was expected from DES. The difference in cost between a DES and BMS is not great, when compared with the total cost of the procedure. So it’s very possible that using the slightly more expensive DES is actually a cost-savings.
In the context of the real world practice, however, the NORSTENT study is useful as a confirmation. Most centers already use DES primarily and many high-volume practices have virtually eliminated the use of BMS. NORSTENT is an important confirmation that current stents do exactly what they were supposed to.
(Updated August 31, 2016 with outcomes for stent thrombosis and exclusion reference.)