To our best knowledge, this is the first report focused on the occurrence of RISR after DCB angioplasty for ISR lesions in China. RISR was showed occurring in 35.4% of patients at review, with the high RDW, high %DS before procedure, and use of cutting balloon in procedure identified as the independent predictors of RISR after DCB angioplasty according to the multivariate regression analysis.
The existing studies on ISR tend to be focused on initial restenosis, and the limited follow-up time results in the unclear occurrence of RISR in the longer term. Several primary studies of RISR after DCB angioplasty for ISR lesions were carried out in Japan[6, 7], which indicate the emergence of RISR at early follow-up (6 months) after DCB angioplasty in approximately 1/5 of DES-ISR patients. The overall incidence of late RISR in ISR patients was calculated to be higher in our study, mainly due to the relatively prolonged follow-up time and the fact that coronary angiography review is generally recommended only in symptomatic patients. S. Habara et al.[6]have revealed the lower incidence of RISR in BMS-ISR compared to the DES-ISR in both early and late follow-up; In contrast, our data show the opposite, that over half of BMS-ISR developed RISR after DCB angioplasty, which indicate that the DES seems not only advantageous in reducing the incidence of initial ISR.
The predictive value of RDW for the occurrence of primary ISR after PCI has received the wide discussion, and our results further suggest RDW also as an independent predictor of longer-term RISR, which is known to serve as an indicator of the variability of red blood cell volume, with the increased level related to pathological mechanisms such as hypoxia, iron deficiency, and inflammatory response[8], and its several mechanisms may also play a role in the progression of RISR. H. Shimono et al. have analyzed the features of intraluminal imaging of DES-RISR and DES-Non-RISR lesions depending on OCT[9], showing more heterogeneous endometrial hyperplasia and neonatal atherosclerosis of the DES-RISR, which means that although RISR and ISR may share many similar mechanisms, RISR is not a simple repeat of ISR. The high HDL-C levels are previously believed to exert cardioprotective effects, while our data demonstrated no benefit of the higher HDL-C in preventing RISR after DCB angioplasty. Genetic evidence provided by the Mendelian randomized studies also fails to support the idea that HDL-C is "good cholesterol"[10]. In addition, although the TBil in all patients was demonstrated within the normal range, their baseline levels were revealed higher by the non-RISR group, which may be related to the anti-inflammatory, antioxidant, and lipid-lowering effects of bilirubin[11–13]. However, the relatively weak specification of the different laboratory parameters still result in the limited clinical application value.
RISR patients have experienced at least 2 failure of revascularizations. As a result, the lesion characteristics before procedure, the process of DCB angioplasty, and the acute luminal gain may mostly tend to show relation to the occurrence of RISR in the later stage. Our findings further illustrate the strongest correlation of %DS before procedure with the occurrence of RISR in the procedural data. Patients in the RISR group obtained the enlarged acute luminal gain after DCB angioplasty, which could not elucidate the fault of concept "bigger is better" in the pursuit of luminal gain during procedure, but only indicated a higher potential for luminal gain of lesions with high %DS before procedure, considering the worse %DS in RISR group and the statistically significant difference of %DS after DCB angioplasty. The in-stent CTO is characterized in high lesion stenosis rate, high complexity, high difficulty of opening, and special population[14, 15]. The research of K. Miura et al. it as the result of DES-ISR after DCB angioplasty, and the independent risk factors for RISR[7]. In the present study, the incidence of in-stent chronic occlusion was increased in the RISR group, while may be affected by the collinearity with the %DS before procedure that resulted in the negative results of multivariate regression analysis. In addition, we also found a particular higher rate of previous myocardial infarction and a longer time to ISR in the RISR group, both of which may be related to the severity of the disease: the former represents a possible history of severe coronary disease, and the latter as the condition for the formation of worse coronary lesions.
The initial ballooning is performed to obtain the most sufficient luminal gain before applying DCB. Cutting balloons and NSE balloons are the mostly adopted modified balloons to treat target diseases during surgery. It has been believed that NSE balloons show considerable pass capability in coronary lesions, while the curative effect on coronary lesions, especially calcified lesions, was inferior to cutting balloons[16]. Compared to the conventional regimens, the outcomes of cutting balloon or NSE balloon combined with DCB in the treatment of ISR lesions showed no significant difference[17, 18]. Our study indicated the developed RISR in over half of the patients who received cutting balloons in the later stage no matter whether other types of initial balloons were employed or not, and the use of cutting balloons could serve as an independent predictor of RISR. The application of cutting balloons in the pre-expansion process is not the exclusion to specific lesions, nor the preference of cardiovascular specialist. The potential cause remains unclear that requires prospective RCT studies to confirm and explore.
Up to now, the research on the treatment strategy of RISR remains a blank. For patients who require the repeated interventional therapy, DCB seems to serve as a more appropriate approach to avoid more layers of stents. If ISR and RISR exert the curative effects remains unclear, which requires more studies to explore. According to the results from follow-up of 52 RISR patients, only less than half of the RISR patients chose DCB for the second PCI. The revascularization method may be modified due to the more complex lesions and that some patients may have lost confidence in the efficacy of DCB. What is confirmed is the failure of benefitting from the long-term from multiple DCB angioplasty in more than half of patients with RISR.
In conclusion, 3 predictors of RISR after DCB dilation: increased RDW, high preoperative stenosis rate, and intraoperative use of a cutting balloon, were identified from baseline data. Avoiding the use of cutting balloons for predilatation of ISR lesions may reduce the incidence of late RISR.
Limitations
First, this is a retrospective single-center study with a relatively limited sample size. The coronary angiography review is recommended for only symptomatic patients and the follow-up time of the two groups is not completely consistent. Multi-center, large-sample, and prospective studies are required to further clarify the incidence of RISR. Second, patients with BMS were not excluded in order to explore the incidence and predictive factors of RISR in overall ISR patients after DCB angioplasty in the real world; DES-ISR has gradually acted as the main type of ISR lesions, and the analysis for DES-RISR alone requires further research.