Reattempt Percutaneous Coronary Intervention of Chronic Total Occlusions after Prior Failures: A Single-Center Analysis of Strategies and Outcomes

Objective The initial recanalization rate of coronary chronic total occlusions (CTOs) is >85% when performed by experienced operators, but only 10% of prior failed CTO patients receive reattempted recanalization. This retrospective study analyzed the success rate and strategies used in reattempt percutaneous coronary intervention (PCI) of CTOs after prior failures. Methods Overall, 206 patients with 212 CTOs were enrolled. All patients with prior recanalization failures received reattempt PCIs from January 2015 to March 2019 at Zhongshan Hospital, Fudan University. Data on clinical factors (age, sex, comorbidities, left ventricular ejection fraction, history of cigarette usage, and revascularization), angiographic characteristics of CTOs (target lesion, Japanese Chronic Total Occlusion (J-CTO) score, the morphology of CTO lesions, and collateral channel scale), strategies (procedural approach and use of devices), and major adverse events were obtained and analyzed. Results The mean age of enrolled patients was 60.96 ± 12.36 years, with a male predominance of 90.3%. Of the patients, 47.1% had a prior myocardial infarction and 70.4% underwent stent implantation previously, while the in-stent occlusion rate was 6.6%. CTOs were primarily localized in the left anterior descending artery (43.9%) and the right coronary artery (43.9%). 80.7% of lesions were classified as very difficult (J-CTO score ≥3), and the overall success rate was 81.1%. In multivariable regression analysis, J-CTO score, collateral channel scale, application of coronary multispiral computed tomography angiography, dual injection, intravascular ultrasound, active greeting technique, parallel wiring, and CTO morphology were predictors of recanalization success. There were no significant differences in rates of procedural complications between the final recanalization success and failure groups. Conclusions Recanalization of complex CTOs is associated with high success rate and low complication rates when performed by high-volume CTO operators and after multiple reattempts.


Introduction
Chronic total occlusion (CTO) lesions are identified in 18%-33% of all patients referred for coronary angiography [1,2]. Also, the presence of concurrent CTO is a strong predictor for both short-term and long-term mortality [3,4]. In the era of interventional therapy, the indications for CTO revascularization are similar to those for severe stenosis according to European Society of Cardiology guidelines [5]. Compared with optimal medical therapy alone, the combination of CTO revascularization with optimal medical therapy is associated with significant ischemia relief, left ventricular function improvement, and a better quality of life [6][7][8]. In recent years, the development of contemporary techniques and devices has substantially improved the initial success rate >85% for CTO-PCI in unselected clinically indicated cases with ≈3% risk for major in-hospital complications when performed by highly experienced operators [9][10][11]. However, outcomes are less favorable at less-experienced centers with an initial success rate of around 60% [12,13]; the procedural failed CTO patients either receive medical therapy still suffering symptomatic ischemia or are advised to undergo reattempt PCI or more traumatic coronary artery bypass grafting (CABG) surgery [14].
It is unclear whether reattempted PCI for CTO lesions is efficacious and safe by expert operators because the prior failure of percutaneous revascularization of CTO has been identified as an independent predictor for failure at subsequent attempts [15]. To our knowledge, there have been scattered reports focusing on strategies and outcomes for reattempted CTO-PCIs following previously failed procedures. In the present study, we sought to define outcomes and predictors of reattempted CTO-PCI success performed by highly skilled operators at our center.

Study Population and Data
Collection. Between January 1, 2015, and March 31, 2019, a total of 206 consecutive patients with CTO lesions after prior PCI failures at local hospitals or our center with a history of multiple recanalization attempts were included. Patient data on baseline clinical characteristics, coronary angiographic information, procedural strategies, and complications were obtained.

Study Definitions and Evaluation of Strategies and
Outcomes. Coronary chronic total occlusions (CTOs) are defined as 100% occlusions with rombolysis in Myocardial Infarction (TIMI) 0 flow with at least a 3-month duration [2,16]. Occlusion duration was estimated based on the first onset of ischemic symptoms, prior history of myocardial infarction (MI) in the target vessel territory, comparison with a prior angiogram, or as the presence of bridging collateral vessels.
Lesion calcification was classified as mild (spots), moderate (≤50% of the reference lesion diameter), or severe (>50% of the reference lesion diameter). Moderate proximal vessel tortuosity was defined as the presence of at least 2 bends >70°or 1 bend >90°, and severe tortuosity was defined as 2 bends >90°or 1 bend >120°in the CTO vessel. Blunt or no stump was defined as the absence of tapering or a funnel shape at the proximal or distal cap.
Collaterals included septal, epicardial, ipsilateral, and saphenous vein graft. Angiographic assessment of a collateral connection (CC) was based on Werner's classification: CC0, no continuous connection between the donor and recipient artery; CC1, continuous, thread-like connection; and CC2, continuous small side branch-like size of the collateral throughout its course [17]. e J-CTO (Multicenter CTO Registry in Japan) score was determined by assigning 1 point to each of the following factors: blunt entry stump, calcification, bend >45 o , occlusion length ≥20 mm, and previous failed attempt. e total number of points was added to stratify lesions into 4 difficulty groups: easy (J-CTO score of 0), intermediate (J-CTO score of 1), difficult (J-CTO score of 2), and very difficult (J-CTO score ≥3) [18].
Recanalization strategies of CTO include two paths (anterograde or retrograde) and two ways: through the true lumen or the subintimal space (with dissection and reentry to the true lumen). e hybrid strategy involved using both antegrade and retrograde approaches or switching from the originally selected approach to the other approach [19,20]. e operators determined which strategy to use. Technical success was defined as the ability to cross an occluded segment and successfully open the artery (restoration of TIMI flow grade of 2 or 3) with residual stenosis of <30%. Procedural success was defined as the achievement of technical success without any in-hospital major adverse cardiac events (MACE) [19].
Procedural complications included coronary artery dissection, hematoma and/or perforation, and tamponade requiring either pericardiocentesis or surgery. Contrastinduced nephropathy (CIN) was defined as an absolute increase of ≥0.5 mg/dL or a relative increase of at least 25% in serum creatinine levels within 48 to 72 h of intravenous administration of an iodinated contrast agent in the absence of other identifiable causes [21]. In-hospital MACE included any of the following before discharge: death, MI (defined using the fourth universal definition of type 4a), and recurrent symptoms requiring immediate repeat target vessel revascularization with PCI or CABG.

Statistical Analysis.
Categorical variables are presented as numbers (percentages) and compared using Pearson chisquare or Fisher exact tests. Continuous variables are presented as mean ± standard deviation if normally distributed or median with an interquartile range if nonnormally distributed and compared using Student's t or Mann-Whitney U test, as appropriate. All indices with a p value <0.1 in the univariate analysis were included in a multivariable logistic regression analysis to test reattempted CTO-PCI success with baseline clinical and angiographic characteristics as well as strategies. All data were analyzed using SPSS 20.0 (IBM SPSS Inc., Armonk, NY). A two-sided p value of 0.05 was considered statistically significant.

Clinical Characteristics.
A total of 206 patients were enrolled in the study. Nearly 90% of the patients for reattempts (178 out of 206 patients) were referred from other centers or other operators. Reattempts of CTOs were required to achieve recanalization in 166 patients, while recanalization could not be achieved in 40 patients even after multiple reattempts. e mean age of patients was 60.96 ± 12.36 years, with a male predominance of 90.3%. Regarding comorbidities, 69.4% had hypertension, 33.5% had diabetes mellitus, 20.9% had hyperlipidemia, 54.9% were smokers, and 47.1% had prior MI. Notably, 70.4% of patients had previously received stent implantation, and 5.8% had received CABG.
ere were no significant differences in the distribution of clinical characteristics between patients with final successful reattempts and final failed reattempts (Table 1).

Angiographic Characteristics.
ere were 212 initially failed CTOs in 206 patients, of which 172 CTOs in 166 patients were successfully recanalized after reattempts, whereas recanalization of 40 CTOs in 40 patients finally failed. Typically, target CTO lesions were primarily located in the left anterior descending coronary artery (43.9%) or right coronary artery (43.9%), and 6.6% were in-stent occlusions. Of the CTOs, 54.2% scaled CC grade 1 and 43.9% scaled CC grade 2, and more than two-thirds were multiple coronary vessel disease. Among the included CTOs, 80.7% were technically very difficult (J-CTO score ≥3, prior failure as part of the score derivation). By detailed evaluation of each characteristic of J-CTO score, 75.5% were occlusion lesions >20 mm, 27.8% were tortuous, 47.2% were blunt proximal cap, 24.1% were blunt distal cap, and 9.4% were severely calcified.

Strategies between Successful and Failed Cases.
Of all reattempted CTO procedures, second-attempt recanalization was achieved in 165 lesions, third-attempt recanalization was acquired in 6 lesions, and fourth-attempt recanalization was successful in 1 lesion, and the overall success rate was 81.1% (172/212 lesions; Figure 1).
e intervals for a second recanalization attempt between finally successful and failed cases were comparable (4.08 ± 5.26 months vs. 3.85 ± 5.34 months, p � 0.799). Compared with final failed cases, final successful cases involved greater apply of coronary multispiral computed tomography angiography (CCTA, 27.1% vs. 7.5%, p < 0.001) and dual injection (73.8% vs. 42.5%, p � 0.003) to acquire comprehensive anatomical information of target lesions. Meanwhile, implementation of IVUS, R-CART, and AGT was more common in final successful cases. ere were no significant differences in comparison to the procedural approach, collaterals used, rotational atherectomy and anterograde dissection and reentry (ADR; use of CrossBoss and/or Stringray CTO reentry devices) maneuvers (Table 3). Table 4, reattempts were accompanied by greater contrast and X-ray doses than the initial attempts. Compared to finally recanalized lesions, final failed cases required more contrast (317.38 ± 184.42 ml vs. 297.38 ± 148.96 ml, p < 0.05) and increased dose of X-ray (2491.28 ± 1178.14 mGy vs. 2319.11 ± 1397.31 mGy, p < 0.05) and were time consuming (48.82 ± 33.66 min. vs. 46.89 ± 34.10 min., p < 0.05).

Safety and Complications. As is presented in
7.5% of patients in the final failed group experienced tamponade during the procedure, while other procedurerelated complications and MACE were comparable between final successful and failed cases. Notably, the contrast-induced nephropathy (CIN) rate was considerably higher in the final successful group patients than in the final failed group patients (Table 5).

Reasons for Initial Procedural Failure and Predictors of Success.
e reasons for initial procedural failures are summarized in Figure 2; wire failure was the predominant reason to terminate the procedure (84.5%), followed by a contrast limit (6.1%). Rare reasons included radiation limit, operator or patient fatigue, and microcatheter failure.
To determine which indexes were mainly affecting procedural outcomes, we conducted correlation and logistic regression analyses. Multivariable regression analysis revealed that J-CTO score, CC scale, CCTA, dual injection, IVUS, AGT, parallel wiring, and CTO morphology were predictors of recanalization success. In contrast, sex, age, target lesion location, time of interval, procedural approach, ADR/RDR, rotational atherectomy, bridging collaterals, in-stent occlusion, CTO bend, comorbidities, smoke, LVEF, renal function, and previous histories of MI and revascularization appeared less important (Table 6). Correlation analysis revealed that higher J-CTO score, collaterals existed in proximal or distal lesions, lesion length ≥20 mm, and severe lesion calcification were negatively correlated with procedural success. In contrast, higher CC scale, preoperation CCTA application, dual injection, IVUS-guided wiring, AGT, and tapered cap existed in both proximal and distal lesions were positively correlated with successful recanalization (supplementary table).

Discussion
e main findings of the present study were that prior failed CTO lesions were associated with higher complexity of morphology; however, 81.1% of CTOs could be recanalized safely and effectively by experienced operators at repeat attempts. ere was a definite relationship between lesion complexity and the increasing need for multiple approaches and technologies during CTO-PCI to achieve success.
Multicenter CTO Registry in Japan proved that "retry" was an unsuccessful predictor for initial failed CTOs [15]; on the other hand, CTO patients were less likely to choose for re-PCI potentially due to low expectations of procedural success and concerns regarding complications; thus, merely 10% patients were referred for PCI other than optimal medical therapy or CABG [22]. A meta-analysis of 25 studies compared successful (71%) with failed (29%) CTO-PCIs in 28486 patients. During a mean follow-up of 3.11 years, successful CTO-PCI was associated with lower mortality, less residual angina, lower risk for stroke, and less need for subsequent CABG with comparison to failed procedures  [23]. According to related guidelines and from our experience, successful opening of CTOs, even repeated attempts, would greatly benefit patients with fewer complications. Generally, initial failure of recanalization for CTOs indicates that lesion morphologies are more unfavorable for wire or microcatheter crossing, which required prolonged procedural time, increased probability of complications, and reduced reimbursement. All of these factors rendered reattempt recanalization of CTO lesions considerably difficult and failed. As was demonstrated in our analysis and the study by Tanabe et al., the inability of guidewire passage through CTO lesions was the most common reason for the failure of re-PCI of CTOs [24]. e main reason of wire failure lied in lesion anatomy, as is shown in  Given increased complexity and risk, the retrograde approach was usually performed when antegrade crossing attempts fail or carried more risk. In the antegrade approach era, the low success rate of 65%-70% (even lower rate for reattempts) and high complication rate were the main barriers to CTO-PCIs [25]. e advent of the retrograde approach circumvented the limitations of the antegrade approach. Moreover, increasing knowledge and expertise could improve complex CTO-PCIs' success rate when the conventional antegrade approach was deemed unsafe or inefficient [1,26]. However, the choice between antegrade and retrograde approaches using wire escalation or dissection reentry methods depended on the CTO anatomy and the operator's experience, favoring timely use of a hybrid algorithm if the procedure did not progress smoothly [27]. It was very important to determine the strategy for a reattempt CTO-PCI with reference to detailed lesion information by preprocedural CCTA and dual injection during the operation. In this study, preprocedural CCTA occupied 27.1%, and dual injection was employed in 73.8% of patients in final successful reattempts to provide detailed lesion characteristics, including lesion length, proximal and distal cap  Values are mean ± SD. mL, milliliter; mGy, milligray. a, compared to the index attempt, p < 0.01; b, comparison of failed reattempt vs. successful reattempt, p < 0.05. morphology, and the extension and morphology of the collateral branches, which helped drive the strategy choice or switch. CTO crossing is often easier in the retrograde direction because the distal CTO cap is often less fibrocalcific, more tapered, and exposed to lower pressure than the proximal cap [28]. In addition to choosing an appropriate procedural approach and strategy, it is noteworthy that a suitable combination of techniques and devices was also of high importance. In our analysis of final successful reattempts, IVUS-guided CTO wiring to resolve proximal cap ambiguity, R-CART, AGT, and CrossBoss and Stingray systems were used at a high percentage. Particularly, retrograde wire externalization facilitated by AGT appeared feasible and safe at our center [29]. e optimal time threshold for change continued to be debated and heavily dependent on operator expertise. e operators decided the use of all available crossing strategies and prompt changes in techniques and equipment. Expertise in all CTO crossing strategies was crucial for achieving the best possible procedural outcomes, especially for complex lesions. Antegrade wire escalation was often associated with lower complication rates and was often preferred as the initial crossing strategy if technically feasible. However, for complicated CTO lesions, antegrade wiring was related to a low success rate. Retrograde and antegrade dissection and reentry remained essential for the successful recanalization of more complex CTOs. In subsequent reattempts, all operators who had performed a career minimum of 300 CTO cases mastered all the available techniques and the appropriate combination of devices (antegrade dissection reentry (ADR) and retrograde dissection reentry (RDR)) in the hybrid algorithm. Referral to a more experienced operator could be helpful, especially if the CTO is of high complexity or there was a previous failed attempt. e overall incidences of procedure-related complications and MACE were low in this study, similar to previous reports [24,30], except that the CIN rate was higher in both finally successful and failed reattempted patients. Consequently, CTO-PCI in reattempted lesions was as safe as initially attempted CTO-PCI.
It should be noted that forty CTO lesions in forty patients remained unopened despite multiple attempts, twenty-four patients were referred to CABG, and the rest chose medical therapy with residual symptoms. ese lesions were compromised with high J-CTO scores and lacked fair interventional collaterals, which means no chance for an antegrade or retrograde approach, and successful interventional revascularization might be limited even though expert operators endeavored reattempted CTO-PCIs.

Study Limitations.
Although we aimed to take an allcomer population, most operators were reliant upon referrals from other cardiologists, and the treatment policy was largely determined by the individual operator. erefore, these results would be unlikely generalizable to other lessexperienced CTO-PCI operators.   Figure 2: Distribution of procedural failures.

Conclusions
Considering lesion complexity and prior failure, reattempted CTO-PCIs could still achieve an overall success rate of 81.1% by high-volume operators with acceptable complications.

Data Availability
e primary data are available on request from the author.

Disclosure
Mingqiang Fu and Shufu Chang are the co-first authors.

Conflicts of Interest
e authors declare that they have no conflicts of interest.