Author + information
- Fernando Alfonso, MD, PhD* ()
- ↵*Reprint requests and correspondence:
Dr. Fernando Alfonso, Interventional Cardiology, Cardiovascular Institute, San Carlos University Hospital, Plaza de Cristo Rey, Madrid 28040, Spain
The treatment of patients with in-stent restenosis (ISR) remains a challenge and represents a major clinical problem (1,2). Bare-metal stents (BMS) are still widely used during coronary interventions, and the treatment of patients with BMS-ISR continues to represent a frequent problem in everyday clinical practice. In addition, drug-eluting stents (DES) are not immune to ISR, especially when used with “off-label” indications in complex clinical and anatomic settings (3). Notably, some registries suggest that despite a considerable increment in DES penetration, the total “restenosis activity” has remained unaltered in most catheterization laboratories around the world, and in fact, the absolute annual number of these procedures continues to rise (3). This scenario is further complicated by recent evidence suggesting that ISR should not be considered just a nuisance without clinical consequences (4–8). Presentation of ISR is no longer considered as benign as previously thought, perhaps partially as a result of the widespread use of highly sensitive markers of myocardial necrosis. Likewise, the management of ISR also appears to pay a price in the long-term clinical outcome of affected patients (4–8).
DES-ISR is quite different from BMS-ISR (4–8). Temporal presentation, morphologic patterns, underlying mechanisms, tissue composition, and response to treatment characteristically differ. As compared with BMS-ISR, the occurrence of DES-ISR tends to spread over a delayed time interval (4,5). The angiographic pattern of ISR is also dissimilar and appears to be related to DES type. In most instances, a “focal” pattern of ISR is detected on angiography whereas diffuse or proliferative patterns are rare (5). Focal patterns are more prevalent after sirolimus-eluting stents (SES) than after paclitaxel-eluting stents (PES). Restenosis confined to the stent edges is quite characteristic, leading to a “candy wrapper” effect at the end of the spectrum (4). This problem persists despite the awareness by the interventional cardiology community of the “geographic miss” phenomenon and systematic efforts to completely cover the entire injured segment.
On quantitative angiography, late lumen loss after BMS follows a characteristic Gaussian distribution whereas distribution of late lumen loss after DES tends to be highly skewed, suggesting potentially different underlying pathophysiological mechanisms (1,2). That creates an additional difficulty when comparing DES studies whenever means, instead of median values, are used to express late angiographic results. Furthermore, mechanistically, DES-ISR tends to be associated with some unique, salient features. Indeed, severe stent underexpansion is frequently detected in these patients (6), whereas this phenomenon appears to play a less relevant role in BMS-ISR. Likewise, strut fractures have been suggested as rare but potentially important underlying causes of DES failure. Moreover, since the dawn of the DES era, the concept of drug failure/resistance has emerged as a new clinical entity (8–10).
Finally, the composition of tissue obstructing the stent also appears to be different. In contradistinction with the monomorphic pattern of neointimal tissue seen after BMS-ISR, a distinct, relatively heterogeneous substrate may be detected in some patients with DES-ISR (4,7). In fact, the presence of true atherosclerotic changes within the stent (including necrotic core development) has been occasionally reported after DES. However, whether these unique findings are inherently related to DES-ISR or merely reflect our current superior diagnostic accuracy to unravel tissue composition, and whether these distinct features have clinical implications, warrant further investigation.
Treatment of patients with BMS-ISR has been extensively studied (1,2). After a decade-long “pilgrimage in the desert”revisiting the potential value of all different mechanical interventional strategies, eventually brachytherapy emerged as a powerful effective tool in this scenario. This technique, however, was cumbersome to use and had inherent limitations and logistics constraints, explaining why it was soon happily abandoned (perhaps prematurely) once DES became available. Subsequent studies clearly demonstrated that DES are not only user-friendly but also much more effective than brachytherapy in this setting and, therefore, currently are enthroned as the cornerstone in the treatment of BMS-ISR (1,2).
Conversely, data on therapeutic options in patients with DES-ISR remain scarce and based on small, retrospective studies (8–15). The information currently available, however, suggests that this new entity is even more challenging to treat than BMS-ISR. Clinical recurrences after interventions for DES-ISR are roughly 2-fold of those seen after BMS-ISR treatment (8). In addition, the therapy of choice for patients with DES-ISR remains unknown. Fortunately, in this issue of the Journal, Mehilli et al. (16) present the first randomized trial ever performed in patients with DES-ISR. Using a well-designed protocol, patients with SES-ISR were randomly assigned either to repeat SES implantation or to PES. Results were satisfactory and similar in both arms. However, should we just feel satisfied and relaxed by embracing the systematic use of DES (with either a homo- or hetero-DES strategy) in patients with DES-ISR, or are we skipping some important steps in this new journey? In other words, can we do much better from a preventive, diagnostic, and therapeutic point of view?
Previous Observational Studies
Although DES dramatically reduced the risk of ISR, this problem has not been eradicated. Surprisingly, there is a paucity of studies addressing the management of DES-ISR (8–15), explaining why interventional cardiologists frequently face major therapeutic dilemmas. Herewith, we will briefly elaborate on the evidence supporting the use of DES for DES-ISR. The study by Lemos et al. (9) was the first to report high rates of recurrent restenosis after treatment of SES-ISR (43% overall, 29% after repeat DES, 18% in de novo DES-ISR treated with DES). Similar findings were obtained by Kim et al. (10), who suggested that SES for DES-ISR were associated with lower recurrent restenosis rates as compared with conventional strategies. Other studies, however, showed conflicting results, with similar recurrence rates after balloon angioplasty and repeat DES implantation (5). The relatively poor results obtained for patients with DES-ISR (9–11) suggested that DES-ISR elicited a more adverse biologic response as compared with BMS-ISR, thus requiring an aggressive treatment. Therefore, despite the lack of robust data supporting the superiority of repeat DES implantation over alternative modalities, the “sandwich”DES strategy was considered by many as the default therapy for DES-ISR, probably due to inferences from results in BMS-ISR (1,2). Furthermore, the rationale for a “switch”strategy (hetero-DES approach), relies on the concept of drug failure and on the different mechanisms of action of drugs currently available in DES. However, sound scientific evidence on this issue is still missing. Five observational studies (11–15) provided information on the results of the switch strategy as compared with homo-DES implantation. These studies were highly heterogeneous but altogether included 487 patients (279 homo-DES, 208 hetero-DES). Equivalent clinical and angiographic results were found in all the studies, although 1 of them (15) suggested a potential benefit of the switch modality. The GISE-CROSS study, a prospective study currently ongoing in Italy, is specifically addressing this question. Clearly, when assessing the switch strategy, the jury is still out.
In the ISAR-DESIRE 2 (Intracoronary Stenting and Angiographic Results: Drug-Eluting Stents for In-Stent Restenosis) trial (16), 450 patients with clinically significant SES-ISR were randomly assigned to either SES (n = 225; Cypher, Cordis, Miami Lakes, Florida) or PES (n = 225; Taxus, Boston Scientific, Natick, Massachusetts). The primary end point of the study—late lumen loss—was similar in both groups (mean 0.40 ± 0.65 mm and 0.38 ± 0.59 mm, respectively; p = 0.85). Other surrogate angiographic parameters of antirestenotic efficacy, including minimal luminal diameter at follow-up (1.93 ± 0.73 mm vs. 1.94 ±0.67 mm) and binary restenosis rate (19.6% vs. 20.6%) were also similar and translated into equivalent rates of target vessel revascularization (16.6% vs. 14.6%). Safety data were also comparable, and rates of definitive stent thrombosis (0.4%) were identical in the 2 arms. The authors should be commended for this important, prospective, controlled study that provides unique insights into the treatment of patients with DES-ISR.
Some interesting study-related issues should be underlined. First, notably, as many as 90% of potentially eligible patients with SES-ISR (from among 499 screened patients) were actually enrolled by these 2 large Munich centers because of broad inclusion criteria, ensuring the study's external validity. Patients with complex patterns of ISR (occlusive or very diffuse ISR requiring multiple DES) were not excluded, thus enriching the population and reinforcing the representation of “real-world” patients with DES-ISR. Perhaps the only concern in this regard was the inclusion of 2 different SES with ISR, 1 of which (polymer-free ISAR-SES) is not widely available and accounted for two-thirds of the patients in the trial. Nevertheless, when study results were reanalyzed taking into consideration the underlying SES (Cypher vs. ISAR-SES), no differences were detected on the relative efficacy estimates. Furthermore, the study's internal validity is reassured by its rigorous methodology and the consistent findings demonstrated in all pre-specified subgroup analyses.
Second, the authors suggest that drug resistance at the individual patient level could have played some role in the higher than expected late loss found after SES. However, in fact, the study was initially designed under this hypothesis, considering the superiority design in favor of PES (sample size calculation with an estimated late loss of 0.60 and 0.40 mm after SES and PES, respectively). This indicates that a major reduction in the efficacy of SES was already expected in this scenario. That might be considered as surprising, firstly, because previous studies by these investigators consistently demonstrated a significantly lower late loss after SES than after PES, and secondly, because of the lack of data supporting the drug resistance issue in the clinical setting. In particular, we should keep in mind that in the ISAR-DESIRE 1 study (1), in-stent medianlate loss was only 0.10 and 0.26 mm after SES and PES, respectively. Therefore, the most likely explanation for this assumption would had been the “fear” of sirolimus hyporesponsiveness in patients already presenting with SES failure. As suggested, additional efforts are required to identify, beforehand, potential nonresponders to specific drugs.
Third, when the 2 ISAR-DESIRE studies (1,16) are compared, it is clear that both SES and PES produce worse in-stent long-term angiographic results in patients with DES-ISR (including late loss, diameter stenosis, and minimal lumen diameter). Although the results of the in-segment analysis were also slightly worse after DES-ISR, it should be noted that in the ISAR-DESIRE 2 trial, in-stent late loss was larger than in-segment late loss, whereas the opposite was true in the ISAR-DESIRE 1 trial (1,16). Whether this finding reflects technical subtleties in quantitative angiographic analyses or, rather, indicates real differences in stenting techniques or, more importantly, in the elicited vascular response remains to be determined. Anyway, every effort should be made to ensure optimization of the final results—including the stent edges—in patients with DES-ISR.
Fourth, in this study, intravascular ultrasound (IVUS) was not used to exclude the presence of severe DES underexpansion or the rare occurrence of stent fracture as the harbinger of DES failure. Many investigators, however, suggest that IVUS is a valuable tool to unravel underlying mechanical causes—which demand specific treatments—and to optimize final results (4,6). Accordingly, the routine use of IVUS has been advocated during these repeated interventions to allow individualized, tailored therapies. In this regard, one wonders whether the final pressures used in the study (mean maximal pressure 15.5 atm) might have been too conservative to tackle potentially underexpanded stents. Likewise, data on pre-dilation/post-dilation strategies would have been of major interest. However, the excellent angiographic results obtained immediately after the procedure (11% of diameter stenosis in both arms) argues against the possibility of suboptimal deployment. Nevertheless, we should keep in mind that angiography may be completely misleading in this regard. Accordingly, further studies should investigate whether IVUS-guided, high-pressure, repeat DES implantation is associated with improved clinical and angiographic outcomes as compared with angiographic guidance alone.
Fifth, angiographic follow-up was performed at 6 to 8 months. After DES implantation, the possibility of ongoing erosion of luminal caliber beyond this time frame should be considered. However, the 1-year clinical follow-up did not suggest any significant catch-up phenomenon. In fact, after the scheduled late angiography, the event rate was very low, and curves tended to flatten and run parallel in both arms.
Finally, the protocol stated that clopidogrel therapy had to be maintained for at least 6 months only. This could be considered insufficient in the light of current recommendations to prolong this therapy for at least 1 year. Nonetheless, two-thirds of patients continued this therapy for 1 year. In addition, the low rate (0.4%) of definitive stent thrombosis in the 2 arms is highly reassuring. This information is particularly valuable because some have warned against the use of DES sandwich strategies, suggesting that this might delay endothelialization (as the result of excessive concentrations of the same drug or the synergistic effects of different drugs) potentially increasing thrombotic risks (8,14). Fortunately, all of these concerns were virtually dissipated by the present study. Nevertheless, a longer follow-up would be of value to confirm the safety and efficacy of the sandwich DES approach.
DES have reached the status of antirestenosis “icons” in the mindset of most cardiologists. Therefore, the appearance of DES-ISR in a given patient represents the paradigm of “failure.” It constitutes a painful reminder that our knowledge remains largely limited, and that we should keep fighting ISR from bench to bedside. New insights at the molecular level are needed to better understand the pathophysiology of drug failure in DES-ISR. Meanwhile, efforts to optimize final results, pursuing ideal geometrical models, should be maintained, considering the high prevalence of underlying mechanical problems. Hopefully, these pathways (mechanistic vs. pathophysiologic) will eventually converge and lead to improved results.
Many DES are currently available, thus broadening the landscape of coronary interventions. Polymer-mediated inflammatory reaction may act as a potential stimulus for neointimal growth, and therefore DES with biodegradable polymers or polymer-free might be of value. The number of potential combinations of DES to treat ISR of specific DES will grow exponentially. In the study of Mehilli et al. (16), only patients with SES-ISR were included and only 2 first-generation DES were analyzed. Whether ISR of new DES will benefit from a switch strategy (as proof of concept) and whether second-generation DES will be more effective in this setting remain to be elucidated.
Finally, alternatives to recurrent implantation of metallic layers in the vessel wall (“onion skin” strategy) to address recalcitrant ISR, are eagerly awaited. Concerns about stent underexpansion are particularly worrisome in this scenario (17). Although initial experiences with the implantation of a third metal layer in patients with recurrent ISR are associated with favorable mid-term clinical and angiographic outcomes (17), larger series of patients with extended follow-up are required to confirm these results. Drug-eluting balloons are highly appealing in this regard, but their efficacy has only been demonstrated after BMS-ISR (18,19). Two randomized studies, also from Germany, demonstrated the impressive antirestenotic efficacy of paclitaxel-eluting balloons in patients with BMS-ISR (18,19). Further, 2 ongoing randomized clinical trials—ISAR-DESIRE 3 (in patients with limus-DES-ISR), and RIBS-IV (in patients with any DES-ISR)—will establish the value of drug-eluting balloons in patients with DES-ISR. The final resource to avoid escalating toward a multilayered metallic vessel wall will be bioabsorbable DES. Currently, however, that should be considered a dream rather than a reality because of the lack of data from patients with ISR.
Where are we going now? To avoid chaotic wandering and eventually getting lost in winding roads, pilgrimsshould follow clear recommendations. Only carefully designed studies and, particularly, randomized controlled trials will be able to shed the required light on the elusive and challenging clinical problem of DES-ISR. The study of Mehilli et al. (16) provides a secure platform to advance our knowledge and to guide our next steps in this new, impassioned, scientific journey.
↵* Editorials published in the Journal of the American College of Cardiologyreflect the views of the authors and do not necessarily represent the views of JACCor the American College of Cardiology.
- American College of Cardiology Foundation
- Alfonso F.,
- Pérez-Vizcayno M.J.,
- Hernández R.,
- et al.
- Baz J.A.,
- Pinar E.,
- Albarrán A.,
- Mauri J.,
- Spanish Cardiac Catheterization and Coronary Intervention Registry
- Alfonso F.,
- Perez-Vizcayno M.J.,
- Cruz A.,
- et al.
- Cosgrave J.,
- Melzi G.,
- Biondi-Zoccai G.G.L.,
- et al.
- Fujii K.,
- Mintz G.S.,
- Kobayashi Y.,
- et al.
- Lemos P.A.,
- van Mieghem C.A.,
- Arampatzis C.A.,
- et al.
- Mishkel G.J.,
- Moore A.L.,
- Markwell S.,
- Shelton M.C.,
- Shelton M.E.
- Sardella G.,
- Colantonio R.,
- De Luca L.,
- et al.
- Mehilli J.,
- Byrne R.A.,
- Tiroch K.,
- et al.,
- ISAR-DESIRE 2 Investigators
- Alfonso F.,
- García J.,
- Pérez-Vizcayno M.J.,
- et al.
- Unverdorben M.,
- Vallbracht C.,
- Cremers B.,
- et al.