The study was designed to evaluate the impact of NLR level on long-term clinical outcomes defined as MACE and health status of CTO patients after PCI in the drug eluting stent era. The findings of the study were as follows: 1) patients in the high tertile group (NLR > 3.2) had a higher rate of MACE than those in the low tertile and the intermediate tertile group at long period of follow-up; 2) elevated preprocedural NLR level was independently associated with MACE, mainly driven by TVR, on follow-up; 3) There were no differences in the long-term health status quantified by SAQ scores of CTO patients regardless of the NLR level.
Prior to the advancement of technique and devices for CTO PCI, CTO lesion is still a challenging subset for successful CTO revascularization. PCI for CTO patients was associated with lower procedural success rates and higher risk of operative complications than those with less complex lesions [19]. With the improvement of technique and equipment for CTO revascularization, such as the retrograde approach, CrossBoss catheter and dissection and re-entry techniques, success rates of CTO PCI have remarkably increased [3, 5, 20]. Moreover, new anti-ischemic drugs, drug-eluting stents, potent antiplatelet agents, and high-dose statins have contributed to reduce the progression of cardiovascular disease and atherosclerosis. However, CTO and in-stent restenosis have been shown to be a worsening clinical condition [9, 10, 21]. The successful treatment of CTO lesion was a strong predictor of subsequent adverse events, including cardiac death, ischemia or symptoms-driven TVR, and CTO PCI was not associated with a reduction in long-term MACE [10, 21]. The pathophysiological mechanism of such recurrent events after PCI for CTO is largely unknown. Therefore, attention has shifted to underlie the pathophysiology of atherosclerosis, and there is a need to identify new markers to accurately predict the prognosis of CTO patients and allow better risk stratification.
Indeed, inflammation plays an indispensable role in the pathogenesis of atherosclerosis, and an elevation in inflammatory biomarkers represents the activation of coronary artery damage [14]. Mounting evidence suggests that inflammation, including pre-existing systemic inflammation, local vascular inflammation secondary to PCI, plays a central role in promoting neointimal growth and in-stent restenosis [16, 22]. After stenting for CTO lesions, platelet and fibrin are deposited around the stents. An acute inflammation response secondary to mechanical injury inflicted by revascularization leads to endothelial rupture and dysfunction, plaque rupture and platelet activation [16]. And this initial acute inflammation is followed by granulation tissue reactions, including migration and proliferation of smooth muscle cells, neovascularization and chronic inflammatory cell infiltration occurs within a few weeks, which causes neointimal hyperplasia, atherosclerosis and in-stent restenosis [16, 22, 23]. Because inflammation plays an important role in the atherosclerosis process, there is increasing interest in various inflammatory biomarkers. Inflammation has been associated with many diseases, such as acute myocardial infarction, chronic heart failure, metabolic disorders, cancer and other cardiovascular disease [12, 24, 25]. Currently, many inflammatory biomarkers have been used to investigate the relationship between inflammatory response and CAD, including white blood cell count, hs-CRP, PLR, NLR, MPV and interleukin-1 [12–14, 16]. Furthermore, unlike several comorbidities and cardiovascular risk factors used by the Framingham Coronary Heart Disease Risk Score and the Atherosclerotic Cardiovascular Disease Risk Algorithm [26], some researchers have studied potential novel risk scores that also include inflammatory markers, such as the PLR and NLR values, which have also provided useful information for predicting cardiovascular events in patients with CAD [16, 27, 28].
Recently, NLR has been considered as a potential marker of inflammation in cardiovascular diseases. Neutrophils secrete a variety of inflammatory mediators, which can cause endothelial injury and vascular wall degeneration. Lymphocytes have an anti-atherosclerotic role and regulate the inflammatory response. Consequently, a high NLR represents two immune pathways, making it more predictive than the method used to evaluate either parameter individually [23, 29]. Several studies have shown that the white blood cell count and its subtypes, neutrophils and lymphocytes, were associated with the prevalence and severity of CAD and its prognosis [14, 15, 28, 30]. Meta-analysis shows that NLR was an independent predictor of hospitalization, long-term cardiovascular events and cardiac mortality in patients with ST-segment elevation MI after PCI [31]. Arbel Y et al. reported that in patients with ST-segment elevation MI, higher NLR was independently associated with lower left ventricular ejection fraction and higher all-cause mortality up to 5 years[30]. Therefore, NLR appears to be an effective marker of systemic inflammation and was considered to be a predictor for worse clinical outcomes under different conditions, including acute coronary syndrome, ST-segment elevation MI and in-stent restenosis.
However, few reports have evaluated the impact of preprocedural NLR level on subsequent adverse cardiovascular events at long-term follow-up in CTO patients after stent implantation. Elevated NLR level was reportedly associated with the impaired development of coronary collaterals in CTO patients [32]. Li C et al. observed that preprocedural NLR level was a powerful risk factor for early in-stent restenosis in patients who underwent successful PCI for CTO lesions [16]. To our knowledge, only one study showed that NLR△, defined as the change of NLR level between on admission and post-PCI, was independently associated with higher risks of in-stent restenosis and MACE in 160 CTO patients at 1-year follow-up [33]. However, the NLR△ level may be involved in vascular mechanical injury secondary to PCI. After a median follow-up of 32 months, we found that in CTO patients who underwent successful PCI, the incidence of MACE in high NLR tertile group was significantly higher than those in the other groups, and elevated preprocedural NLR level was independent predictor for long-term MACE in patients with CTO lesions.
No studies have quantified the impact of preprocedural NLR level on the long-term health status quantified by the SAQ scores after PCI for CTO patients. The SAQ, which quantifies five domains to measure the impact of CAD on patients’ health status, has been shown to be effective, reproducible, and sensitive to changes in clinical symptoms [34, 35]. Safley et al. reported that all SAQ scores improved after PCI for CTO and non-CTO patients, and there were no difference in Physical Limitation, Quality of Life and Angina Frequency scores between groups [36]. Ybarra et al. observed that CTO patients with complex features such as those with dissection re-entry techniques, those with high complexity (Japanese CTO ≥ 3) or coronary artery bypass grafting had a similar degree of improvement in health status quantified by the SAQ Scores when compared with those with less complex CTOs [37]. In our study, we first found that there were no significant differences regarding physical limitation, angina frequency, angina stability, treatment satisfaction and quality of life among the 3 groups, and the preprocedural NLR level was not associated with SAQ scores in the five domains. Therefore, the preprocedural NLR level had no effect on the long-term health status quantified by the SAQ scores of CTO patients after PCI.
Our study has several limitations. First of all, it shares all the limitations of observational, single center studies. Second, although there were no statistically significant differences in most potential confounding factors among the 3 groups, we were unable to correct for unmeasured potential variables. Third, this was a retrospective study unpowered to identify all changes to patients’ medical treatment strategies during follow-up. Finally, the sample size was limited, so this study was considerably underpowered, and our results might not be generalizable to other studies.