Third-generation drug-eluting stents and beyond: Where were we, where are we, and where we are heading

Michael Maeng

Second-generation drug-eluting stents resolved the issues of late and very late stent thrombosis that were seen with the first-generation devices and have good safety and efficacy data. However, there is still room for improvement with these devices. Michael Maeng explores the journey from first-generation to second-generation drug-eluting stents, looking at if future devices have the potential to be superior to current treatments.

The first drug-eluting stents revolutionised the field of interventional cardiology. The sirolimus-eluting Cypher stent (Cordis), the paclitaxel-eluting Taxus stent (Boston Scientific), and the zotarolimus-eluting Endeavor stent (Medtronic) constituted the first-generation drug-eluting devices, and millions of patients were treated with them.  Some of the lessons learned with these stents were that Cypher and Taxus were very effective tools to decrease in-stent restenosis and that their drug elution—over approximately one month—was efficient. However, they were also associated with late and very late stent thrombosis; probably related to adverse polymer-vessel wall interactions. Also, the Endeavor device found to elute its drug too fast, which explains why it was less effective and was associated with higher rates of early stent thrombosis.

By contrast, the second-generation drug-eluting stents—such as the everolimus-eluting Xience/Promus stents (Abbott Vascular/Boston Scientific), the biolimus-eluting Nobori/BioMatrix stents (Terumo/Biosensors), and the zotarolimus-eluting Resolute stent (Medtronic)—used thinner struts, better polymers, and eluted the limus drug of choice (paclitaxel disappeared with Taxus) over a period of approximately one month.

Current status of drug-eluting stents

With the appearance of the second-generation (and the disappearance of the first-generation) drug-eluting stents, non-inferiority head-to-head comparisons replaced superiority trials. A long series of non-inferiority trials have subsequently shown what they were designed to do: namely, non-inferiority between stents.

However, there have been two exceptions to this rule. The first exception was the SORT OUT V trial, which failed to show that the Nobori biolimus-eluting stent was superior to the first-generation Cypher stent. The second exception was BioFlow V (N=1334 patients), which compared the thin-strut sirolimus-eluting Orsiro stent (Biotronik) to the durable-polymer Xience stent. It found that 6% of patients in the Orsiro group and 10% in the Xience group met the 12-month primary endpoint of target lesion failure (p=0.0399). However, the difference in target lesion failure was primarily driven by a difference in target vessel myocardial infarction (4.7% vs 8.3%, respectively) with no difference in definite stent thrombosis. Superiority could not be confirmed in the SORT OUT VII trial (N=2525) comparing Orsiro vs Nobori (target lesion failure 3.8% vs 4.6%).

Consequently, we now have many newer-generation drug-eluting stents that have been compared to at least one other newer-generation drug-eluting stent and all of these do principally show non-inferiority. At first glance, this may suggest stagnation in stent development rather than progress. But, this may not be entirely true.

First, if we look at the outcomes with the biolimus-eluting drug-eluting stents (Nobori, BioMatrix, BioMatrix NeoFlex)—which have been used in four consecutive SORT OUT studies—then the outcomes of target lesion failure (from 5.3% to 4.4%) and target lesion revascularisation (from 3.3% to 2.5%) have been improving with each consecutive study. So even with almost identical drug-eluting stents, results are getting better over time. Second, both Synergy (a thin-strut, everolimus-eluting stent with a biodegradable polymer, Boston Scientific) and Orsiro have target lesion failure rates around 3.8–4.0%. Although these improvements (compared with other devices) have not reached statistical significance, the results suggest that we—in a period with non-inferiority trials—are slowly but surely improving our results.

The future of drug-eluting stents

Theoretically, we can improve drug-eluting stents on at least five parameters: strut thickness, drug, polymer, inactivation of the luminal side of the stent, and deliverability.

  • Strut thickness: the relative thick-strut stainless-steel biolimus-eluting stents will soon be history. Nobori has been replaced with the cobalt-chromium sirolimus-eluting Ultimaster device (Terumo), and BioMatrix NeoFlex is being replaced by a new cobalt-chromium version. This means that all major drug-eluting stent brands will have almost identical strut thickness based on a cobalt-chromium or platinum-chromium alloys. There is, thus, probably not much to gain with regard to this parameter
  • Drug: all major drug-eluting stent brands use limus drugs and major improvements are not expected in this field
  • Polymer: the currently available durable polymer drug-eluting stents (Xience/Promus/Resolute) have shown results that are equal to biodegradable polymers. Within the biodegradable polymer field, there seems to be very similar results with polymers that are absorbed after three (Synergy/Ultimaster), eight (Nobori/BioMatrix), or 18 (Orsiro) months. Polymer-free stents are interesting and may be a preferable strategy, but more data are needed
  • Inactivation of the luminal stent side: this may be particularly relevant for patients at high risk of bleeding events as it may allow shorter duration of dual antiplatet therapy. It also may lower the risk of stent thrombosis. The Combo (OrbusNeich) and Cobra PzF (CelaNova) stents use such strategies, but we are awaiting more data from studies, such as SORT OUT X, that compare Combo vs Orsiro
  • Deliverability: stent delivery failure remains an issue, at least with the thick-strut drug-eluting stents, but thin-strut third-generation drug-eluting stent trials all report low rates of stent delivery failure. Still, factors such as the ability of the stent to stay on the balloon in complex percutaneous coronary intervention (i.e. not get lost proximal to the lesion), pushability, and flexibility remain important factors for operators.


We still have a dream of a bioabsorbable scaffold that delivers the drug to the vessel, maintain radial strength for a sufficient period of time to avoid constrictive vascular remodelling, and then disappears when the job is done. At this moment, however, we have an array of drug-eluting stents that perform extremely well.

In all-comer trials, the rates of target lesion failure are now down to 4% in the first pivotal year. Since target lesion failure includes cardiac death (1.4% in the latest SORT OUT trial), this leaves very little space for reduction of true stent-related failures. Identification of high-risk patients and optimisation of adjuvant therapies is important and may be a more cost-effective strategy to achieve better outcomes for our patients. However, an important function of the non-inferiority studies is to monitor the continuous modifications of available drug-eluting stents, and to function as an independent quality control of new drug-eluting devices being developed and sold by new companies on the market. Since superiority studies in the drug-eluting stent area are no longer being performed, an important function of drug-eluting comparisons is to capture and avoid the entry of truly inferior drug-eluting stents. To do so, and to ensure the best available treatment of our patients, we must continue to perform independent testing of drug-eluting stent modifications and new arrivals.

Michael Maeng is at Department of Cardiology, Aarhus University Hospital, Skejby, Aarhus, Denmark


Please enter your comment!
Please enter your name here