Coronary intervention or percutaneous coronary intervention (PCI) is an important therapeutic method for coronary artery disease (CAD) [1], and in-stent restenosis (ISR) is the most common complication of PCI, which occurs in 3%–20% of patients undergoing coronary stent implantation [2]. ISR is a key risk factor that affects the prognosis of PCI and is a difficult problem associated with coronary intervention of CAD [3].

Syphilis is caused by infection with the pathogen Treponema pallidum subsp. pallidum, which is a curable sexually transmitted disease. As economic development and migrants moved to China in recent decades, syphilis re-emerged gradually and widely spread among men who have sex with men, female sex workers, drug users, and HIV-infected populations, so it has become an important public health problem in China [4]. Despite its curability, some individuals are unaware of their infection due to latent syphilis until cardiovascular diseases occur and confirmed by further laboratory examination of Treponema pallidum particle agglutination (TPPA) and titres of the toluidine red unheated serum test (TRUST) [4]. PCI is emergently conducted among CAD patients with syphilis, and ISR may occur in these patients, although antisyphilitic treatment is recommended after performing PCI.

Previous studies focused on finding the risk factors of ISR among general CAD patients receiving PCI [5, 6], but risk factors of ISR among CAD patients with syphilis after PCI are not fully understood. This study aims to elucidate not only risk factors of ISR among CAD patients with syphilis but also the population attributable risk (PAR) percentage (PAR%), which is used to quantify the proportion of ISR that could be eliminated if particular risk factors are not present and reminds physicians to prioritise interventions based on these risk factors to reduce ISR development among CAD patients with syphilis.

ISR is a problematic complication in the treatment of CADs with PCI, as it occurred in 3%–20% of patients following stent implantation [2], and the pathogenesis of ISR has not been fully elucidated. In CAD patients with syphilis, ISR remains a difficult problem in those receiving PCI, and the prevalence of ISR among CAD patients with syphilis receiving PCI is obscure. It was reported that syphilis can affect almost any tissue or organ, including the cardiovascular system, and invasive properties of T. pallidum induced a delayed-type hypersensitivity response [4], resulting in syphilitic damage of the cardiovascular system, if antisyphilitic therapy was unsuccessful. In this study, we found that ISR occurred in 18 patients, with a prevalence of 15.78%. The high prevalence indicated the importance of identifying risk factors of ISR and providing earlier clinical intervention among CAD patients with syphilis who will be undergoing PCI.

In CAD patients undergoing PCI, stent thrombosis is a rare complication that can cause in-stent stenosis [14]. Cutlip et al. [15] reported that stent thrombosis occurred in < 1.0% of patients undergoing PCI and receiving routine DAPT, and Ong et al. [16] reported that 1.2% of patients with bare metal stents, 1.0% with sirolimus-eluting stents, and 1.0% with paclitaxel-eluting stents developed angiographically proven stent thrombosis in the first 30 days after stent implantation. Early stent thrombosis was attributable to mechanical problems, and although drug-eluting stents reduced the occurrence of early thrombosis by introducing local drug delivery, they increased the occurrence of late stent thrombosis due to chronic inflammatory responses, remodeling of the coronary artery vessel wall, and the progression of atherosclerosis [17‐19]. In this observational cohort, generation 2 stents, not generation 1 or bare metal stents, were selected to reduce early thrombosis, and DAPT was recommended to these patients to reduce the development of late stent thrombosis based on expert consensus on PCI [7], which helped reduce the occurrence of ISR.

Among CAD patients with syphilis, if antisyphilitic therapy was overlooked, the pathogen T. pallidum invaded the vessel wall, including the coronary artery system, and the inflammatory response resulted in necrosis of the elastic fibers and connective tissue in vessel media and the formation of gumma, which was reactive and granulomatous [20], resulting in syphilitic damage to some degree in the coronary artery system besides atherosclerosis in these patients. In the present observational cohort, 114 CAD patients emergently received PCI and were diagnosed as having coronary atherosclerotic heart disease based on coronary artery angiography results, and the serum screening results of TPPA and TRUST were positive prior to surgery; among those, six patients were found to be complicated with syphilitic cardiovascular diseases. Although antisyphilitic treatment after performing PCI was recommended [21], most patients in this cohort obtained no or unsuccessful antisyphilitic treatment. We found that successful antisyphilitic treatment after PCI was an independent protective factor, which indicated that among CAD patients with syphilis undergoing PCI, successful antisyphilitic treatment is essential among these patients to reduce the occurrence of ISR.

ISR occurred among CAD patients undergoing PCI due to multiple risk factors, including biological, technical, mechanical, and therapeutic factors [17‐19]. The risk factors of ISR among CAD patients with syphilis undergoing PCI are not fully understood. In the present study, we found that average stent length ≥ 35 mm and titres of TRUST > 1:16 were associated with an increased hazard of ISR, while successful antisyphilitic treatment was a protective predictor of ISR among CAD patients with syphilis.

The PAR% measures the burden of exposure factors and helps estimate the level of preventable diseases [22]. It assumes complete causality between a given risk factor and the pathology, so by removing the risk factor, the event probability decreases to the baseline risk. Benichou et al. [22, 23] demonstrated that the PAR% exposure factor was a causal one rather than merely related to diseases. It was reported that a longer stent length is the cause of ISR [5] and that higher titres of TRUST and unsuccessful antisyphilitic treatment [24] result in extensive vascular damage and stent thrombosis after PCI. In the present study, stent length ≥ 35 mm, titres of TRUST > 1:16, and unsuccessful antisyphilitic treatment were causal factors of ISR, which indicated that these three factors were prone to evaluation of the PAR% of ISR among CAD patients with syphilis undergoing PCI.

Chatterjee et al. [25] indicated that risk predictive models could be derived from cohort studies, in which the exposure-specific absolute risks of the disease could be estimated directly. Hu et al. [26] also reported that in a retrospective study, determining the population attributable fraction helped demonstrate the importance of promoting physical activity, a healthy diet, and other healthy lifestyles for reducing obesity among high school students in the United States, which indicated that the PAR% could be used to evaluate the burden of exposure factors in a retrospective study.

In the present study, the PAR% method was used to quantify the proportion of ISR that could be eliminated if particular hazards contributing to ISR based on Cox proportional hazard models were not present or modified, which indicated prioritisation of the clinical intervention or modification to reduce the occurrence of ISR [27]. In the current study, we calculated the PAR% values for the three aforementioned predictors to help determine the appropriate clinical intervention based on the quantitative PAR% to reduce occurrence of ISR. We also strongly recommend not summing the PAR% of a single risk factor to derive the total PAR% of ISR among CAD patients with syphilis undergoing PCI because there was a confounder between these risk factors and conditions under which this method was used [28].

Stent length is a risk factor of ISR development after PCI. Cassese et al. [29] indicated that each 10-mm increase of the stent length was a risk factor of ISR development. Cheng et al. [5] demonstrated that the longer stent resulted in more severe injury of the vessels, a stronger inflammatory reaction, and greater intimal thickness. Herein, we also found that average stent length ≥ 35 mm was an independent risk factor of ISR among CAD patients with syphilis undergoing PCI, which is similar to the findings of the aforementioned studies. We further found that the PAR% was 12.2%, which indicated that 12.2% of the risk of ISR would be reduced if average stent length ≥ 35 mm was not present or modified. This is a reminder to interventional cardiologists that among CAD patients with syphilis undergoing PCI, it is necessary to carefully evaluated the length of the vessel lesion to determine the stent length that can accurately cover the vessel lesion.

In the current study, the screening of TPPA and titres of TRUST was conducted prior to performing emergent PCI among patients with CAD, and antisyphilitic treatment was recommended after PCI. Despite the availability of effective antisyphilitic treatment, syphilis-related clinical diseases including syphilitic cardiovascular diseases and even cerebral infarction continue to cause comorbidities in Chinese syphilis-infected patients. Some patients were unaware of their syphilis infection until the occurrence of end-organ diseases, such as myocardial or cerebral infarction [30]; some patients neglected to receive antisyphilitic treatment because of mild clinical symptoms or signs; and some patients abandoned treatment because of poor psychosocial factors. In our study, despite the recommended antisyphilitic treatment and periodical follow-up for screening titres of TRUST after PCI, only 43% of patients received successful antisyphilitic treatment; the other 65 patients did not obtain successful antisyphilitic treatment, and their titres of TRUST never became negative at follow-up. The higher titres of TRUST were more prone to cause greater damage to vessels in the coronary artery system [24] and result in ISR, and we found that titres of TRUST > 1:16 was another independent risk factor of ISR among CAD patients with syphilis undergoing PCI. We further found that the PAR% was 24.0%, which indicated that 24.0% of the risk of ISR would be reduced if titres of TRUST were decreased by effective antisyphilitic treatment. We further demonstrated that successful antisyphilitic treatment after PCI was an independent protective factor for ISR, and the PAR% of successful antisyphilitic treatment was -39.6%. The aforementioned results serve to remind interventional cardiologists to attach importance to damage of the coronary artery caused by syphilis and antisyphilitic treatment in CAD patients with syphilis undergoing PCI.

Several variables, including the stent type, HCY level, and histories of diabetes, hypertension, and hyperlipidaemia, have been reported as risk factors of ISR in the general population undergoing PCI [5, 31], but we could not conclude that these variables were associated with ISR in patients with syphilis undergoing PCI. It was widely reported that the bare metal stent was associated with ISR, and experienced interventional cardiologists often selected generation 2 drug-eluting stent to avoid ISR at our hospital. Besides DAPT, clinical interventions for risk factors, including diabetes, hypertension, and hyperlipidaemia, are important and should be recommended to patients with these comorbidities. We did not find that these comorbidities were associated with ISR due to effective clinical intervention, which indicated the importance of clinical intervention for the aforementioned comorbidities in CAD patients with syphilis undergoing PCI. Although we did not find that the HCY level > 15 µmol/L and average stent diameter ≤ 3 mm were associated with ISR in our cohort, we still needed to monitor these two risk factors to alert of the occurrence of ISR.

The present study has some strengths. First, this observational study was firstly conducted to evaluate the risk factors and PAR% values of risk factors of ISR among CAD patients with syphilis undergoing PCI in China, and it is the first to quantify the proportion of ISR that could be eliminated if particular risk factors were not present or modified. Our findings will help remind physicians to prioritise interventions based on the PAR% of risk factors of ISR. Second, the most predominant finding was that successful antisyphilitic treatment was a protective factor of ISR, indicating that interventional cardiologists should attach importance to successful antisyphilitic treatment besides DAPT, which helped reduce the occurrence of ISR among CAD patients with syphilis undergoing PCI.

Our study also have some limitations. First, this study was conducted in a single centre with a small sample size, and its conclusion should be further validated at multiple centres using a prospective cohort study design. Second, potential inherent biases exist in a retrospective and observational study. Third, the duration of syphilis that affected the severity and prognosis of CAD is unknown, because we did not know the exact duration of syphilis and could not investigate the impact of syphilis duration on the occurrence of ISR.

In summary, preventing the occurrence of ISR among CAD patients with syphilis undergoing PCI requires clinical intervention, and our study is the first to present the risk factors and their PAR% values of ISR among CAD patients with syphilis after PCI. We found that average stent length ≥ 35 mm and titres of TRUST > 1:16 are risk factors of ISR, and successful antisyphilitic treatment was a protective predictor of ISR. Therefore, stent length < 35 mm, prioritising clinical intervention for titres of TRUST > 1:16, and successful antisyphilitic treatment could reduce the risk of ISR occurrence by 12.2%, 24.0%, and 39.6%, respectively.