Prognostic impact of white blood cell counts on clinical outcomes in patients with chronic renal insufficiency undergoing percutaneous coronary intervention

Objective To determine whether the inclusion of white blood cell (WBC) counts in the SYNTAX score (SS) or SS II models could improve the models’ performance for risk stratification in individuals with chronic renal insufficiency (CRI) following percutaneous coronary intervention (PCI). Methods In total, 2,313 patients with CRI, who were subjected to PCI and had data available on in-hospital WBC (ih-WBC) counts, were recruited. Patients were divided into 3 groups as per their ih-WBC counts (low, medium, and high). The primary endpoints were all-cause mortality (ACM) and cardiac mortality (CM). The secondary endpoints incorporated myocardial infarction, stroke, unplanned revascularization, and major adverse cardiovascular and cerebrovascular events (MACCEs). Results During a median follow-up of 3 years, the high WBC group had the highest incidences of CM (2.4% vs. 2.1% vs. 6.7%; p < 0.001), ACM (6.3% vs. 4.1% vs. 8.2%; p < 0.001), unplanned revascularization (8.4% vs. 12.4% vs. 14.1%; p < 0.001), and MACCEs (19.3% vs. 23.0% vs. 29.2%; p < 0.001) among the three groups. Multivariable Cox regression analysis depicted that the risk of ACM and CM in the high WBC group was 2.577 (95% confidence interval [CI]: 1.504–4.415, p < 0.001) and 3.850 (95% CI: 1.835–8.080, p < 0.001) times that in the low WBC group after adjusting for other confounding factors. A combination of ih-WBC counts with SS or SS II significantly improved the risk assessment and prediction of ACM and CM. Conclusion The ih-WBC counts was associated with the risk of occurrence of ACM, CM, unplanned revascularization, and MACCEs in individuals with CRI following PCI. It provides an incremental predictive value for the occurrence of ACM and CM when included in SS or SS II models.


Introduction
Inflammation exerts a critical function in the progression as well as plaque destabilization in atherosclerosis (1). Studies have found that downstream biomarkers of inflammation, such as interleukin-6 and the high-sensitivity C-reactive protein (hs-CRP) are linked to a greater risk of cardiovascular events (2,3).
The white blood cell (WBC) counts are a widely-used and easily available marker of inflammation in clinical practice. Its predictive value as a marker for mortality in individuals with acute coronary syndrome (ACS) is well-established (4)(5)(6). Recently, it has also been demonstrated that total, as well as differential in-hospital white blood cell (ih-WBC) counts, are independent prognostic factors for long-term deaths and major adverse cardiovascular and cerebrovascular events (MACCEs). Including ih-WBC counts in SYNTAX score (SS) or SYNTAX score II (SS-II) models can improve mortality predictions in individuals with triple-vessel coronary artery disease (CAD) (7). However, there have been no assessments of the predictive value of ih-WBC counts in patients with CRI after percutaneous coronary intervention (PCI). Chronic renal insufficiency (CRI), which is defined as the presence of kidney damage or reduced kidney function (estimated glomerular filtration rate (eGFR) <90 mL/min/1.73 m 2 ) for ≥3 months (8), is high-risk comorbidity that increases the risk of cardiovascular mortality and morbidity and is known to be associated with poorer clinical outcomes in patients following PCI (9). In this investigation, we sought to evaluate the utility of including ih-WBC counts as a factor in SS or SS II models for anticipating long-term clinical outcomes in individuals with CRI following PCI.

Study population
The study design of the risk evaluation of CRI patients following PCI has been described previously (10). Briefly, a total of 2,468 patients with creatinine clearance rates (CrCl) <90 mL/ min per 1.73 m 2 who underwent PCI between January 2014 and September 2017 in Cangzhou Central Hospital, Hebei Medical University, were retrospectively enrolled in the study. After excluding 155 patients for whom ih-WBC counts data were not available, 2,313 patients were included in this investigation. Patients were classified into three groups as per tertiles of ih-WBC counts as follows: low WBC group [WBC counts ≤6.18*10 9 /L (n = 776)], medium WBC group [WBC counts >6.18*10 9 /L but ≤8.14*10 9 /L (n = 768)], and high WBC group [WBC counts >8.14*10 9 /L (n = 769)] ( Figure 1). The study protocol was subjected to approval by the ethics committee of Cangzhou Central Hospital. Written and informed consent was obtained for all subjects.
The clinical and interventional data of the participants were collected from the electronic medical record (EMR) system. The CrCl values were calculated utilizing the simplified Modification of Diet in Renal Disease (MDRD) equation. The ih-WBC counts were defined as the first WBC counts value from the EMR system. Two of three trained cardiologists (who were blinded to the clinical data as well as outcomes) calculated the SS (11) and SS II (12) through the dedicated website. 1 In case of any disagreement, the opinion of a third observer was obtained and resolved by consensus.

Study endpoints and follow-up
Clinical follow-up was performed via clinic visits or telephone conversations. The primary endpoints included all-cause mortality (ACM) as well as cardiac mortality (CM). Deaths that could not be attributed to non-cardiac causes were considered CM. The secondary endpoints included myocardial infarction (MI), stroke, unplanned revascularization, and major adverse cardiovascular and cerebrovascular events (MACCEs), defined as a composite of ACM, MI, stroke, and unplanned revascularization. MI was defined following the consensus document on the fourth universal definition (13). Stroke was defined as neural dysfunction due to a sudden rupture or blockage of a blood vessel, and was diagnosed based on signs of brain dysfunction or imaging evidence (14). Revascularization of PCI or coronary artery bypass grafting (CABG) driven by ischemic symptoms or cardiac events was defined as unplanned revascularization. All endpoints were confirmed by two independent clinicians.

Statistical analysis
Continuous variables are reported as mean ± standard deviation (SD) or medians with interquartile ranges (IQR). Categorical variables are presented as frequencies and percentages. Continuous variables were compared utilizing one-way ANOVAs or Kruskal-Wallis tests, when necessary. Chi-square or the Fisher exact tests, on the other hand, carried out comparisons of categorical variables. Patients who were lost to follow-up were deemed at risk until they were censored at the date of the last contact. The cumulative event rates were measured utilizing Kaplan-Meier curves and compared across groups via the log-rank test. We assessed the prognostic value of ih-WBC grouping for predicting clinical outcomes using multivariable Cox regression models. While Log[−logS(t)] plots were used to test for proportional hazard assumption. All potential confounders (with p < 0.1 in the univariate analyses) were incorporated in the multivariate analyses. By combining ih-WBC counts with SS or SS II values, we assessed the improvements in model performance, risk classification, and discrimination; this was done by comparing the AUC of the two nested models employing the nonparametric deLong approach and computing the net reclassification improvement (NRI) as well as the integrated discrimination improvement (IDI) indices. A two-sided value of p < 0.05 was statistically significant. SPSS 24.0 (IBM Corp., Armonk, NY, United States) and R Software Version 3.6.0 (The R Foundation for Statistical Computing, Vienna, Austria) conducted all the statistical analyses of this investigation.

Patients' baseline characteristics
The ih-WBC counts ranged from 2.6*10 9 /L to 32.7*10 9 /L. The SS values ranged from 1.0 to 47.0, and the SS-II values ranged from 9.7 to 59.6. Patients with high ih-WBC counts were younger and more likely to have current smoker status; clinical presentation of MI; reduced eGFR; lower left ventricular ejection fraction (LVEF); higher left ventricular end-diastolic diameter (LVEDD); and elevated levels of serum creatinine, blood glucose, total cholesterol, triglyceride, and low-density lipoprotein (p < 0.05 or p < 0.001) than those with low and median ih-WBC counts. Patients with high ih-WBC counts also had higher baseline SS values and were more likely to have thrombus lesions and undergo primary PCI (p < 0.001 for all) (Tables 1, 2).

Association of ih-WBC counts with clinical outcomes
The median follow-up period was 3 years (IQR = 1.5-5.0). Among the three groups, the high ih-WBC counts group had the highest 5-year cumulative event rates of ACM (6.3% vs. 4

Combination of ih-WBC counts with SS or SS II values for prediction of ACM and CM
The analyses of time-dependent AUCs for ACM showed that the AUCs for SS in combination with ih-WBC counts were significantly larger than those of SS alone (p < 0.05) during the 30-month follow-up period. However, the degree of increase tended to decrease with time. There was no remarkable difference between the predictive value of the SS plus ih-WBC counts model and the SS model (p > 0.05) when the follow-up times were longer than 30 months. A similar result was observed for the SS-II models, albeit with a cutoff value of 28 months instead of 30 months. In the models used to predict the incidence of CM, though the extent of increase also tended to decrease with time, there was a remarkable difference between the predictive value of the SS plus ih-WBC counts model and the SS model (p < 0.05) during the whole follow-up period, and for the SS-II models, the cutoff value was 33 months ( Figure 3).
Furthermore, by combining the ih-WBC counts with SS or SS II models, the metrics for risk reclassification and discrimination were significantly improved. The respective NRIs of the SS plus ih-WBC counts model over the SS model were 0.121 for ACM and 0.188 for CM; the NRIs of the SS-II plus ih-WBC counts model over the SS-II model were 0.025 for ACM and 0.135 for CM. The IDI indices of the SS plus ih-WBC counts model over the SS model were 0.019 (p < 0.01) for ACM and 0.022 (p < 0.001) for CM and the IDI indices of the SS-II Study flow chart. PCI, percutaneous coronary intervention; eGFR, estimated glomerular filtration rate; CABG, coronary artery bypass graft; WBC, white blood cell.

Discussion
This study shows that the ih-WBC counts is associated with the risk of occurrence of ACM, CM, unplanned revascularization, and MACCEs in individuals with CRI following PCI. The ih-WBC counts can be utilized for risk reclassification, especially in secondary prevention among patients with CRI post-PCI. Integrating ih-WBC counts into SS or SS II models also improves the predictive performance of these models and facilitates the identification of patients at risk for future ACM and CM. Therefore, the ih-WBC counts may also be used to flag patients at risk for adverse cardiac events post-PCI, who may warrant more intensive follow-up and preventive treatment.
Inflammation is known to exert an important function not only in atherogenesis but also in atherosclerotic plaque rupture resulting in acute coronary syndrome (ACS) (15)(16)(17)(18). Studies have found that biomarkers of inflammation including hs-CRP and interleukin-6 are independent risk factors for cardiovascular events (19). Preprocedural hs-CRP elevation has been linked to a greater risk of adverse cardiac events in people undergoing PCI (20)(21)(22). The WBC counts, which can be easily and repeatedly obtained in clinical practice, are one of the most viable inflammatory biomarkers. Cannon et al. (4) have   The SS model is a well-established tool used to predict adverse clinical outcomes to help clinicians decide on optimum revascularization strategies in individuals with complex CAD (11,24,25). The SS-II model incorporates the anatomical variables in the SS model with other clinical variables (age, sex, LVEF, CrCl, chronic obstructive pulmonary disease, and peripheral vascular disease), and can predict 4-year mortality with higher accuracy. The SS-II model is also a better guide than the SS model for decisions on PCI and CABG in complex CAD cases (12). Subsequently, several studies have demonstrated the predictive value of the SS-II model for predicting outcomes in different cohorts such as three-vessel and/or unprotected left main coronary artery disease (ULMAD) following PCI (26, 27), ACS (28, 29), and cardiogenic shocks after primary PCI (30).
The SS-II model was created according to a Cox proportional hazards model utilizing the SYNTAX trial findings (12). The baseline features that were strongly associated with 4-year mortality were added to the model. However, the ih-WBC counts, a marker of the inflammatory state, were not included in the SS-II model. A recent study observed that a combination of differential WBC (eosinophil, monocyte, and lymphocyte) counts enhanced the success of risk prediction and reclassification for mortality when combined with SS        (7). Patients with CRI have a hallmark feature of persistent, low-grade inflammation, which is involved in the development of ACM (31). Inflammation plays an important role in the initiation and progression of kidney disease. Recent studies have reported that plasma proinflammatory biomarkers, such as soluble TNF receptors 1 and 2 (TNFR-1 and TNFR2) were associated with the increased risk of progression of diabetic kidney disease, even after adjustment for established clinical risk factors (32, 33). However, not much is known about the predictive value of ih-WBC counts for predicting clinical outcomes in patients with CRI post-PCI. This study demonstrated that elevated ih-WBC counts were associated with increased incidence of ACM, CM, unplanned revascularization, and MACCEs in individuals with CRI following PCI. Furthermore, the addition of ih-WBC counts to the SS or SS II models also improved the predictive performance of these models in predicting ACM and CM events in individuals with CRI following PCI, although the degree of improvement in predictive performance tended to decrease with time.

Limitations
Despite these promising results, our results should be viewed in the light of multiple limitations. First, this study is based on data from a single center and is retrospective and observational in nature; therefore, it can only identify associations and cannot ascribe causality to related events. Second, differences in blood collection periods from the occurrence of the index event were not controlled for in this analysis. Third, we did not collect information on differential WBC counts and hs-CRP levels, both of which may be important for clinical outcomes. Finally, in some patients, we believe that the ih-WBC counts data may have been affected by undetected infections or other conditions for which we have no information.

Conclusion
In patients with CRI following PCI, an elevated ih-WBC counts was found to be associated with the risk of occurrence of ACM, CM, unplanned revascularization, and MACCEs. A combination of ih-WBC counts with SS or SS II models significantly improved these models' performance in predicting ACM and CM.

Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement
The studies involving human participants were reviewed and approved by the Ethics Committee of Cangzhou Central Hospital, Hebei Medical University. The patients/participants provided their written informed consent to participate in this study.

Author contributions
WY, LY, and ML provided the conception of the idea for the study and analyzed the acquired data. LY, WY, ML, SZ, and YL contributed to the development of the methodology and wrote the manuscript. FL, JW, NY, and MC were responsible for the interpretation of statistical results. XC revised the manuscript. All authors contributed to the article and approved the submitted version.

Funding
This work was supported by the Natural Science Foundation of Hebei Province, China (H2021110008) and Hebei Provence Key Research Projects (172777163).