Author + information
- Received October 7, 2016
- Revision received December 21, 2016
- Accepted January 27, 2017
- Published online May 1, 2017.
- Lorenzo Azzalini, MD, PhD, MSca,
- Rustem Dautov, MD, PhDb,c,
- Soledad Ojeda, MD, PhDd,
- Susanna Benincasa, MDa,
- Barbara Bellini, MDa,
- Francesco Giannini, MDa,
- Jorge Chavarría, MDd,
- Manuel Pan, MD, PhDd,
- Mauro Carlino, MDa,
- Antonio Colombo, MDa and
- Stéphane Rinfret, MD, SMb,c,∗ ()
- aDivision of Interventional Cardiology, Cardio-Thoracic-Vascular Department, San Raffaele Scientific Institute, Milan, Italy
- bDivision of Interventional Cardiology, McGill University Health Centre, Montreal, Canada
- cDivision of Interventional Cardiology, Quebec Heart and Lung Institute and Laval University, Quebec City, Canada
- dDivision of Interventional Cardiology, Reina Sofia Hospital, University of Córdoba, Maimonides Institute for Research in Biomedicine of Córdoba (IMIBIC), Córdoba, Spain
- ↵∗Address for correspondence:
Dr. Stéphane Rinfret, McGill University, Division of Interventional Cardiology, Cardio-Thoracic-Vascular Department, McGill University Health Centre, Royal Victoria Glen Site (B03.7200), 1001 Boulevard Décarie, Montreal (Québec) H4A 3J1, Canada.
Objectives The study sought to investigate the long-term outcomes and predictors of adverse events of percutaneous coronary intervention (PCI) for in-stent chronic total occlusion (IS-CTO).
Background IS-CTO PCI has traditionally been associated with suboptimal success rates.
Methods We performed a multicenter registry of consecutive patients undergoing CTO PCI at 3 specialized centers. Patients were divided in IS-CTO and de novo CTO. The primary endpoint (major adverse cardiac events [MACE]) was a composite of cardiac death, target-vessel myocardial infarction, and ischemia-driven target-vessel revascularization (TVR) on follow-up. Independent predictors of MACE were sought with Cox regression.
Results We included 899 patients (n = 111 IS-CTO, n = 788 de novo CTO). Baseline clinical and angiographic characteristics were balanced between the 2 groups. Overall mean J-CTO (Japanese-Chronic Total Occlusion) score was 1.88 ± 1.24 and mean PROGRESS-CTO (Prospective Global Registry for the Study of Chronic Total Occlusion Intervention-CTO) score was 1.04 ± 0.88. Antegrade wire escalation was used in 59.0% of IS-CTO and 48.1% of de novo CTO patients (p = 0.08). Procedural success was achieved in 86.5% in both groups (p = 0.99). After a median follow-up of 471 (interquartile range: 354 to 872) days, MACE were observed in 20.8% versus 13.9% in IS-CTO versus de novo CTO (p = 0.07), driven by TVR (16.7% vs. 9.4%; p = 0.03). IS-CTO was an independent predictor of MACE (hazard ratio: 2.16; 95% confidence interval: 1.18 to 3.95; p = 0.01), together with prior surgical revascularization and renal function, CTO PCI indicated for acute coronary syndrome, number of diseased vessels, and PROGRESS-CTO score.
Conclusions Procedural success was high and similar in patients with IS-CTO, as compared with de novo CTO. However, IS-CTO was independently associated with MACE (driven by TVR) on follow-up.
Chronic total occlusion (CTO) percutaneous coronary intervention (PCI) has witnessed a remarkable improvement in success rates (>90%) as well as procedural metrics and safety during the past few years (1).
However, specific patient and lesion subsets still represent a challenge. In particular, in-stent CTOs (IS-CTOs) have traditionally been associated with suboptimal procedural success rates (63% to 71%) (2–4), albeit these figures improved in the recent report from the PROGRESS-CTO (Prospective Global Registry for the Study of Chronic Total Occlusion Intervention-CTO) registry (86%) (5). Additionally, treatment of in-stent occlusive segments has been identified as an independent predictor of the need for revascularization after CTO PCI (6).
Little data exist on the outcomes of this challenging patient subgroup on follow-up. Similarly, predictors of major adverse cardiac events (MACE) on follow-up are poorly characterized. The aim of the present study is to address these important questions.
We queried the CTO PCI database of the 3 participating centers (San Raffaele Hospital, Milan, Italy; Quebec Heart and Lung Institute, Quebec City, Canada; Reina Sofia Hospital, Córdoba, Spain) to identify all consecutive patients with and without IS-CTO who underwent CTO PCI. A total of 991 procedures were performed in 899 patients by experienced CTO PCI operators (>80% success rate) (7) between January 2009 and December 2015. Only the first procedure for each patient was considered for analyses. All procedures were indicated according to the presence of angina, ischemia or both, and were performed electively (ad hoc PCI was discouraged) (1). Baseline, procedural and hospitalization data were recorded. Follow-up was performed by means of phone interview, revision of hospital records or outpatient visit. The study was approved by the institutional review boards of the 3 participating hospitals. Figures 1 and 2 show 2 cases of IS-CTO PCI.
CTO was defined as an occlusive (100% stenosis) coronary lesion with antegrade Thrombolysis In Myocardial Infarction flow grade 0 for at least 3 months (5). A CTO was considered to be IS if the occlusion was located within a previously deployed stent or within the 5 mm proximal and distal to it (5). The J-CTO (Japanese-Chronic Total Occlusion) score (8) and the PROGRESS-CTO score (9) were calculated for each lesion.
A dissection/re-entry crossing was defined by the use of any wire-based (with knuckled or straight wires) or CrossBoss-based (Boston Scientific, Marlborough, Massachusetts) access to the subintimal space, followed by a re-entry technique (either antegrade or retrograde, contrast guided or non–contrast guided, or involving the Stingray [Boston Scientific] system). True-to-true crossing was considered the most likely mechanism with the use of guidewire escalation (either retrograde or antegrade) or when the CrossBoss passed the occlusion reaching the true lumen (i.e., without Stingray-facilitated re-entry). Specifically, the use of CrossBoss for IS-CTO was considered a true-to-true crossing, if not followed by Stingray.
Technical success was defined as a residual stenosis <30% with antegrade TIMI flow grade 3 in the CTO target vessel (5). Procedural success was defined as technical success plus the absence of in-hospital adverse events (all-cause death, Q-wave myocardial infarction [MI], stroke, recurrent angina requiring target vessel revascularization [TVR] with PCI or coronary artery bypass graft [CABG], tamponade requiring pericardiocentesis or surgery) (5).
Major procedural complications included: procedure-related death, stroke, periprocedural type 4a MI (10), major bleeding (bleeding requiring transfusion, vasopressors, surgery or percutaneous intervention), coronary perforation with cardiac tamponade requiring intervention (pericardiocentesis, coiling, covered stent implantation or surgery), and contrast-induced nephropathy (increase in serum creatinine >25% or >0.5 mg/dl at 48 h post-procedure).
MACE on follow-up were defined as the composite of cardiac death, target-vessel MI (Q-wave and non–Q-wave MI) and ischemia-driven TVR, defined as any revascularization in the CTO vessel, including proximal segments or distal branches, driven by angina or ischemia on noninvasive imaging tests.
Continuous variables are presented as mean ± SD and Student t test was used for comparisons. Categorical variables are presented as frequency (percentage) and were compared using a chi-square test.
Predictors of MACE on follow-up were sought using multivariate Cox regression. Variables showing a p < 0.20 in univariate analysis or deemed to be associated with the outcome of interest according to clinical judgment were used as candidate predictors for multivariate analysis. Variable selection was performed by fitting a penalized Cox regression model with LASSO (Least Absolute Shrinkage and Selection Operator) penalty and a tuning parameter selected by cross-validation (lambda = 0.01237) (11). The candidate variables were: IS-CTO, center, age, diabetes, prior MI, prior CABG, estimated glomerular filtration rate (eGFR), acute coronary syndrome presentation, number of diseased vessels, J-CTO score, PROGRESS-CTO score, use of dissection/re-entry techniques, drug-eluting stent (DES) implantation, procedural success, and major procedural complications. The final model obtained after penalization included the following variables (Online Figure 1): IS-CTO, diabetes, prior CABG, eGFR, acute coronary syndrome presentation, number of diseased vessels, J-CTO score, PROGRESS-CTO score, use of dissection/re-entry techniques. The results of this analysis are presented as hazard ratios (HRs) and 95% confidence intervals. Survival curves were plotted to ascertain if the presence of IS-CTO was independently associated with MACE and TVR on follow-up, after adjustment for the independent predictors of MACE previously identified.
For all tests, a p < 0.05 was considered significant. Statistical analysis was performed using SPSS version 20 (IBM Corporation, Armonk, New York) and R 3.1.0 (R Foundation for Statistical Computing, Vienna, Austria). Penalized Cox regression model with LASSO penalty was performed using the glmnet package for R (12).
Baseline clinical characteristics
The study population included 899 patients (n = 111 with IS-CTO and n = 788 with de novo CTO). The prevalence of IS-CTO PCI was therefore 12.3%. Clinical characteristics of the study population are shown in Table 1. There were no differences between IS-CTO and de novo CTO patients with regards to age, gender, prevalence of cardiovascular risk factors, renal function, ejection fraction and indication of CTO PCI. Subjects with IS-CTO had a higher prevalence of prior MI and prior PCI.
Baseline angiographic characteristics
Baseline angiographic data are presented in Table 2. Compared to the de novo CTO group, IS-CTO patients had a lower burden of coronary artery disease. No differences were observed regarding the target CTO vessel, J-CTO and PROGRESS-CTO scores, and the amount of collateralization. The IS-CTO group had lower prevalence of moderate/severe calcifications and bifurcation at the distal cap. However, IS-CTO also had a lower prevalence of adequate landing zone beyond the distal cap, as well as a higher prevalence of occlusion length >20 mm and ostial location.
Table 3 shows procedural data. There was a trend toward lower use of radial access in IS-CTO patients, as compared with de novo CTO subjects. Antegrade wire escalation was the most frequent successful crossing strategy (59.0% in IS-CTO vs. 48.1% in de novo CTO; p = 0.08). The retrograde approach and dissection/re-entry techniques were used less often in IS-CTO patients. The use of CrossBoss was 4-fold higher in IS-CTO (26.1% vs. 6.6%; p < 0.001). Second-generation DES were the preferred stents in both cohorts, although bioresorbable scaffolds were used more frequently in de novo CTO lesions. Despite the fact that dissection/re-entry techniques were utilized in 30% of IS-CTO patients, true-to-true crossing of the occlusion segment containing the restenosed stent was always possible in our experience, and in no case subintimal crushing of the previous stent was required. Total stent length tended to be longer in IS-CTO. Procedural metrics were similar between the 2 groups.
Technical (87.4% vs. 87.2%; p = 0.95) and procedural (86.5% vs. 86.5%; p = 0.99) success rates were similar between the 2 groups. The incidence of major procedural complications was also similar between groups (2.7% vs. 2.4%; p = 0.84). In particular, 1 perforation with tamponade, 1 case of major bleeding, and 1 case of contrast-induced nephropathy were observed in the IS-CTO group. In patients treated for de novo CTO, we observed 8 cases of perforation with tamponade (2 of these patients subsequently died), 1 case of major bleeding, 4 cases of stroke (2 of them later died), 3 cases of contrast-induced nephropathy, and 3 periprocedural MIs.
Clinical Outcomes on Follow-Up
Follow-up was available for 807 of 899 (89.8%) patients. Median follow-up was 471 (interquartile range: 354 to 872) days. Table 4 shows clinical outcomes on follow-up. IS-CTO patients tended to have a higher incidence of MACE (20.8% vs. 13.9%; p = 0.07) driven by TVR (16.7% vs. 9.4%; p = 0.03). No differences were observed regarding cardiac death and target-vessel MI.
Table 5 shows univariate and multivariate predictors of MACE on follow-up. Treatment of IS-CTO remained independently associated with MACE (HR: 2.16; 95% confidence interval: 1.18 to 3.95; p = 0.01), in addition to post-CABG status (HR: 1.73; p = 0.05), eGFR (HR: 0.92; p = 0.04), acute coronary syndrome presentation (HR: 1.92; p = 0.008), number of diseased vessels (HR: 1.44; p = 0.03), and the PROGRESS-CTO score (HR: 1.41; p = 0.01). Figure 3 presents adjusted 2-year curves of survival free from MACE and TVR. Patients with IS-CTO had a >2-fold increase in the risk of adverse outcomes, with events accumulating throughout follow-up.
The main findings of our study are as follows: 1) IS-CTO PCI is performed in about 12% of consecutive CTO interventions in large CTO PCI programs; 2) these patients and lesions do not present marked differences regarding clinical and procedural characteristics, and therefore success rates are high and similar to de novo CTO PCI in the hands of experienced operators using the hybrid algorithm; 3) despite this observation, IS-CTO is independently associated with a 2-fold increase in the adjusted risk of MACE (driven by TVR) on follow-up.
Prior to the development and widespread adoption of the hybrid algorithm (13) and dedicated devices (e.g., CrossBoss), PCI of IS-CTO lesions had traditionally been associated with suboptimal success rates (2–4). Abbas et al. (2) studied the outcomes of a cohort of patients undergoing IS-CTO PCI during the bare-metal stent (BMS) era (25% of the study population). Success rates were similar for IS-CTO versus de novo CTO (63% vs. 70%). Although inability to wire the occlusion was the primary mechanism of failure in both groups, inability to advance or fully dilate the balloon was more frequent in the IS-CTO group. Werner et al. (4) reported that only 5% of all CTO treated at their institution were IS-CTO. Success rate was 70%, lower than for de novo CTO (85%), despite the availability of retrograde techniques. In the small series by Abdel-Karim et al. (3) success rate was 71%. Failure was due to inability to cross the occlusion in all cases. de la Torre et al. (14) recently reported a procedural success rate of 82% in a single-arm study on 233 patients with IS-CTO. However, patient inclusion spread across a decade, and this study is therefore not representative of a modern approach to IS-CTO PCI.
Following systematic implementation of the hybrid algorithm, IS-CTO PCI success rates have remarkably improved. In an all-comer patient population from the PROGRESS-CTO registry (5), the incidence of IS-CTO PCI was 10.9%. The retrograde approach was used in approximately one-third of procedures, similar to our data. Procedural success rate was high and did not differ between groups (86.0% and 90.3% in IS-CTO vs. de novo CTO, respectively), which mirrors our results (86.5% in both groups). Figure 4 compares procedural success rates of IS-CTO PCI in the literature over time.
The advancement of wires and microcatheters is greatly impaired by the presence of stent struts in close proximity with the subadventitial space. In such setting, the use of CrossBoss might represent a very efficient alternative (Central Illustration), as it allows for an easy, within-stent, true-to-true occlusion crossing, due to its higher crossing profile combined with its effective dissecting action. In a small single-arm study of selected patients with IS-CTO, CrossBoss showed a 90% success rate, with good procedural metrics (median crossing time of 8 min, despite a mean CTO length of 39 mm) (15). The role of this device for the treatment of IS-CTO deserves further investigation.
In comparison with the aforementioned reports (2–5,14,15), our study has several strengths, including a larger sample size, comparison with a de novo CTO PCI group, a detailed angiographic analysis, and, remarkably, the availability of data on long-term clinical outcomes. Additionally, we performed an adjusted analysis that identified IS-CTO as an independent predictor of MACE (driven by TVR) on follow-up. This latter observation had already been reported (6), and parallels similar findings from the non-CTO interventional practice, where PCI for in-stent restenosis had been identified as a predictor of future restenotic events (16).
The pathophysiology of such in-stent occlusions is largely unknown. Although neointima formation may be involved, we also know that such phenomenon is dramatically reduced with DES. We hypothesize that a significant proportion of IS-CTOs were the results of other factors (e.g., stent thrombosis, stent fracture, neoatherosclerosis) (Central Illustration). It is reasonable to say that the patient’s biological factors that were involved in the first episode of stent occlusion are likely again triggered in the newly implanted stent in the same segment, and may involve abnormal local inflammation, adverse reaction to polymer, resistance to antiplatelet agents, stent malapposition or underexpansion, or others. In a seminal study in the BMS era, Mehran et al. (17) indicated that up to one-half of IS-CTO cases had suffered a prior episode of in-stent restenosis and that IS-CTO was an independent predictor of target-lesion revascularization after recanalization.
In some cases, stent occlusion might have been the result of thrombosis rather than restenosis. Indeed, stent thrombosis can develop asymptomatically several months after implantation (18). This phenomenon might be especially relevant after recanalization of chronically occluded arteries. In registries where multidetector computed tomography was routinely performed 6 months after bioresorbable scaffold implantation, asymptomatic restenosis or reocclusion was detected in up to 5.7% of patients (19,20).
The improvement in clinical outcomes for IS-CTO patients will likely stem from the development of newer-generation DES, with better antiproliferative characteristics as well as eliciting lower inflammatory response. Additionally, intravascular ultrasound (IVUS) can provide useful information about the mechanisms underlying IS-CTO (after pre-dilatation with a small balloon) and, more importantly, to optimize stent implantation to minimize the risk of adverse outcomes on follow-up (21). Indeed, recent data from the Multicenter Korean CTO Registry indicated that IVUS-guided CTO PCI was associated with lower risk of stent thrombosis, and possibly target-lesion revascularization in the subset of long lesions (22). Additionally, pharmacologic strategies to deal with this challenging patient population deserve special attention. The use of more potent antiplatelet therapy might improve patient outcomes, although such a hypothesis has not been tested yet. Also, systemically administered colchicine can decrease neointimal hyperplasia following stent implantation and has been associated with a decreased in-stent restenosis rate (23). Finally, CABG might be considered for patients with multivessel disease and IS-CTO to reduce the risk of TVR, provided that full arterial revascularization can be achieved (24,25).
Our adjusted analysis also identified prior CABG, worse renal function, procedural indication for acute coronary syndrome, higher number of diseased vessels, and higher PROGRESS-CTO score as independent predictors of MACE on follow-up. While most of these variables are well known to be associated with adverse events, the observation that an angiographic tool such as the PROGRESS-CTO score is able to predict clinical outcomes in patients undergoing CTO PCI (similar to the SYNTAX score in all-comers) (26) is novel. As compared with the J-CTO score, the PROGRESS-CTO score does not account for the operator-dependent component (i.e., retry CTO PCI) in the prediction of successful revascularization using the hybrid algorithm (9). Therefore, the PROGRESS-CTO score exclusively evaluates occlusion complexity, which in our study was associated with higher risk of MACE on follow-up. This observation warrants further research.
First, it shares the limitations of all observational studies. Second, no data was available on the specific type of stent (DES vs. BMS) that had suffered occlusive restenosis in the IS-CTO group and the time frame between stent implantation and development of IS-CTO, due to the fact that most patients were referred to our CTO PCI programs from other centers. However, we do not think that knowing this would have altered clinical decision making because drug-eluting platforms were implanted in all IS-CTO cases. Third, IVUS use was low in the IS-CTO group, in order not to prolong these already lengthy procedures and due to cost-related issues. Additionally, IVUS findings were not captured by our database, which prevented identification of the mechanisms leading to IS-CTO in our study. Finally, our observations might not be generalizable to other institutions where expert hybrid CTO PCI operators and dedicated devices are not available.
IS-CTO PCI is frequently performed with excellent success rates in the hands of experienced operators using the hybrid algorithm. Although, as compared with de novo CTO, IS-CTO was not associated with lower procedural success, it was found to be independently associated with a higher risk of MACE (driven by TVR) on follow-up. Novel strategies and stents are therefore needed to improve the outcomes of IS-CTO PCI.
WHAT IS KNOWN? IS-CTOs have traditionally been associated with suboptimal success rates of PCI. Little data exist on the outcomes of this challenging patient subgroup on follow-up.
WHAT IS NEW? This multicenter registry analyzed the outcomes of a large cohort of patients with IS-CTO versus de novo CTO. Procedural success was high and similar in the 2 groups. At long-term follow-up, the rate of MACE was higher in IS-CTO versus de novo CTO, driven by TVR. IS-CTO was an independent predictor of MACE.
WHAT IS NEXT? The development of newer-generation DES, with better antiproliferative characteristics as well as eliciting lower inflammatory response, warrants further research, to improve the long-term outcomes of this challenging patient population.
For a supplemental figure, please see the online version of this article.
Dr. Rinfret has served as a consultant for Boston Scientific and SoundBite medical; has received honoraria for proctorship and lectures for Boston Scientific, Abbott Vascular, and Terumo; and has received research funding from Medtronic and Abbott Vascular. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- bare-metal stent(s)
- coronary artery bypass graft
- chronic total occlusion
- drug-eluting stent(s)
- estimated glomerular filtration rate
- hazard ratio
- in-stent chronic total occlusion
- intravascular ultrasound
- major adverse cardiac event(s)
- myocardial infarction
- percutaneous coronary intervention
- target vessel revascularization
- Received October 7, 2016.
- Revision received December 21, 2016.
- Accepted January 27, 2017.
- 2017 American College of Cardiology Foundation
- Abbas A.E.,
- Brewington S.D.,
- Dixon S.R.,
- Boura J.,
- Grines C.L.,
- O’Neill W.W.
- Morino Y.,
- Abe M.,
- Morimoto T.,
- et al.
- Christopoulos G.,
- Kandzari D.E.,
- Yeh R.W.,
- et al.
- Thygesen K.,
- Alpert J.S.,
- Jaffe A.S.,
- et al.
- ↵Friedman J, Hastie T, Simon N, Tibshirani R. glmnet: Lasso and elastic-net regularized generalized linear models. R package version 2.0–5. Available at: https://cran.r-project.org/web/packages/glmnet/index.html. Accessed December 23, 2016.
- Brilakis E.S.,
- Grantham J.A.,
- Rinfret S.,
- et al.
- ↵de la Torre Hernandez JM, Rumoroso JR, Subinas A, et al. Percutaneous intervention in chronic total coronary occlusions caused by in-stent restenosis. Procedural results and long term clinical outcomes in the TORO (Spanish registry of chronicTOtal occlusion secondary to an occlusive in stent RestenOsis) multicenter registry. EuroIntervention 2016 Dec 20 [E-pub ahead of print].
- Wilson W.M.,
- Walsh S.,
- Hanratty C.,
- et al.
- Singh M.,
- Gersh B.J.,
- McClelland R.L.,
- et al.
- Mehran R.,
- Dangas G.,
- Abizaid A.S.,
- et al.
- Vaquerizo B.,
- Barros A.,
- Pujadas S.,
- et al.
- Elgendy I.Y.,
- Mahmoud A.N.,
- Elgendy A.Y.,
- Bavry A.A.
- Hong S.J.,
- Kim B.K.,
- Shin D.H.,
- et al.
- Deftereos S.,
- Giannopoulos G.,
- Raisakis K.,
- et al.
- Azzalini L.
- Azzalini L.,
- Torregrossa G.,
- Puskas J.D.,
- et al.
- Wykrzykowska J.J.,
- Garg S.,
- Girasis C.,
- et al.