Risk Prediction of Major Adverse Cardiovascular Events After Percutaneous Coronary Intervention Using Inflammatory Biomarkers

Summary

This protocol aims to develop and validate a multidimensional risk prediction model integrating novel inflammatory biomarkers to identify acute coronary syndrome patients at high risk for major adverse cardiovascular events following percutaneous coronary intervention.

Abstract

This study aimed to build and validate a risk prediction model for 1-year major adverse cardiovascular events (MACE) in patients with acute coronary syndrome (ACS) undergoing percutaneous coronary intervention (PCI), utilizing novel inflammatory biomarkers. This single-center retrospective cohort study enrolled 1,337 patients with ACS who underwent PCI between January 2021 and December 2023. Six novel inflammatory indexes (NLR, MHR, NHR, SII, SIRI, AISI) were derived from pre-PCI blood tests. After a 7:3 random split into training (n = 936) and validation (n = 401) cohorts, LASSO regression and multivariable Cox proportional hazards models identified independent predictors, and a combined biomarker-based model was constructed. Age, diabetes, Killip Class ≥ II, reduced LVEF, multivessel disease, no-reflow phenomenon, NHR, and SIRI were identified as independent predictors. The combined model achieved an AUC of 0.81 (95% CI: 0.78–0.84), which remained stable after optimism correction via bootstrapping. This performance was substantially higher than that of any single biomarker (maximum AUC: 0.71) and demonstrated significant improvements in NRI and IDI (all P < 0.001). Risk stratification demonstrated a clear gradient in MACE incidence: 6.3% (low-risk), 15.1% (intermediate-risk), and 25.3% (high-risk), P. < 0.0001, with consistent predictive performance across all evaluated clinical subgroups. The novel inflammatory biomarker-based model substantially improves risk prediction over clinical variables alone, providing a valuable framework for risk stratification and identifying patients at high residual inflammatory risk who may require closer clinical surveillance.

Introduction

Although percutaneous coronary intervention (PCI) techniques and perioperative pharmacotherapy for patients with ACS have made significant progress, these patients remain at high risk for major adverse cardiac events after revascularization¹. Clinical studies have shown that even when low-density lipoprotein cholesterol (LDL-C) levels are reduced to the target range with aggressive lipid-lowering therapy, the persistent hyperinflammatory state continues to contribute to thrombotic recurrence, stent restenosis, and reduced cardiac function. This lipid-independent pathophysiological state is termed 'residual inflammatory risk' and represents a crucial therapeutic target for improving long-term cardiovascular outcomes^2,3.

Traditional clinical indicators commonly used to assess systemic inflammation, such as C-reactive proteins (CRP), white blood cell (WBC), and other routine markers, have severe limitations. Although highly sensitive, CRP is susceptible to confounding by acute infections or non-specific tissue damage, and it fails to reflect the complex interactions among distinct immune cell populations. Simple WBC counts do not distinguish between neutrophils, lymphocytes, and platelets during inflammation. To better characterize the immune-inflammatory homeostasis, new composite inflammatory biomarkers are gradually showing superior prognostic predictive performance⁴. For example, the systemic immune-inflammatory index combines neutrophils, platelets, and lymphocytes to reflect the degree of innate immune activation and prothrombotic state simultaneously. The pan-immune-inflammatory value adds monocytes to better represent the myeloid-lymphoid cell interaction. In addition, a biomarker such as the Neutrophil/Amylase ratio combines inflammatory load and nutritional stress status for prognostic purposes. These new indicators can make up for the bias caused by the fluctuations of single-cell counts through mathematical combinations, and thus achieve a more objective evaluation of patients' biological risk⁵.

Inflammatory responses last throughout the whole course of acute coronary syndrome (ACS). From early lipid deposition and immune cell recruitment to late-stage plaque fibrous cap degradation and eventual plaque rupture, pro-inflammatory factors and immune cells consistently drive the progression. Although interventional procedures resolve macrovascular mechanical obstructions, balloon dilation and stent deployment inevitably cause local vessel wall injury, thereby triggering an acute-phase inflammatory response. The acute reaction, in combination with the patient's chronic inflammation, gradually leads to endothelial dysfunction and even induces microcirculatory perfusion disorder. Since this pathophysiological process is mediated through multiple immunological pathways and involves various cell subsets, a single biomarker can usually only reflect a fraction of the interactions within this complex network⁶.

While recent studies have increasingly highlighted the prognostic value of various inflammatory biomarker profiles for cardiovascular events in patients undergoing coronary angiography and interventional procedures^7,8,9, studies focusing on the comprehensive combined modeling of multiple novel inflammatory indices in patients with ACS remain relatively scarce. Currently, there is a lack of comprehensive risk assessment tools to directly guide clinical decision-making. Existing risk scoring systems are limited in their ability to integrate dynamic inflammatory burdens, thereby hindering precise individualized risk stratification. Therefore, this study aimed to systematically evaluate the prognostic impact of multiple novel inflammatory biomarkers on post-PCI MACE using real-world clinical follow-up data. By employing advanced statistical screening methods, we sought to identify independent prognostic predictors and construct a robust risk prediction model¹⁰.

Protocol

The study protocol was approved by the Ethics Committee of Jinan Central Hospital. The requirement for written informed consent was waived because the data were anonymized and used retrospectively¹¹. All procedures were conducted in accordance with the Declaration of Helsinki.

1. Study design and patient selection

1. Conduct a single-center retrospective cohort study enrolling patients admitted with acute coronary syndrome (ACS) who undergo percutaneous coronary intervention (PCI).
   NOTE: In the current study, patients admitted between January 2021 and December 2023 were included.
2. Diagnose ACS, including ST-segment elevation myocardial infarction (STEMI), non-ST-segment elevation myocardial infarction (NSTEMI), and unstable angina, in accordance with current clinical guidelines.
3. Obtain ethical approval from the institutional Ethics Committee.
4. Screen patients for inclusion based on the following criteria: age ≥ 18 years; confirmed diagnosis of ACS; successful PCI performed during the index hospitalization; availability of complete blood count (CBC) and biochemistry tests prior to PCI; and availability of complete clinical baseline and follow-up data.
5. Exclude patients presenting with active infections, hematological disorders, malignancies, severe hepatic or renal insufficiency, prior coronary artery bypass grafting (CABG), or missing key laboratory and follow-up data.
6. Randomly partition the final analysis cohort into training and validation sets using a 7:3 ratio with stratified random sampling based on MACE occurrence to ensure balanced distribution of outcome events (Figure 1).
   NOTE: In the current study, the final analysis cohort included 1,337 patients, comprising a training set (n = 936) and a validation set (n = 401).

2. Data collection and definition of combined inflammatory biomarkers

1. Extract and double-check all clinical data using two independent researchers against standard procedures from the electronic medical record (EMR) system, laboratory information system (LIS), and imaging databases.
2. Record baseline demographic and clinical characteristics, including age, sex, body mass index (BMI), smoking status, alcohol consumption, and medical history (hypertension, diabetes mellitus, previous myocardial infarction, previous PCI, and stroke).
3. Collect perioperative parameters, including ACS classification, systolic blood pressure, heart rate, Killip class, left ventricular ejection fraction (LVEF), culprit vessel, number of involved branches, Thrombolysis in Myocardial Infarction (TIMI) flow grade, and stent status.
4. Draw venous blood samples after an overnight fast (≥ 8 h) immediately upon hospital admission.
5. Analyze all laboratory parameters within the 24 h window prior to the PCI procedure. Record the following: white blood cell count, neutrophil count (N), lymphocyte count (L), monocyte count (M), platelet count (P), hemoglobin, albumin, creatinine, high-sensitivity C-reactive protein (hs-CRP), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), cardiac troponin, and N-terminal pro-B-type natriuretic peptide (NT-proBNP)¹².
6. Calculate the novel inflammatory biomarkers using the following standardized formulas, ensuring all blood cell counts are normalized to the same standard unit prior to calculation: 
   1. Neutrophil-to-lymphocyte ratio (NLR) = N / L
   2. Monocyte-to-HDL-C ratio (MHR) = M / HDL-C
   3. Neutrophil-to-HDL-C ratio (NHR) = N / HDL-C
   4. Systemic immune-inflammation index (SII) = (P × N) / L
   5. Systemic inflammatory response index (SIRI) = (N × M) / L
   6. Aggregate index of systemic inflammation (AISI) = (N × P × M) / L

3. Outcome definition and follow-up strategy

1. Define the primary endpoint as the first occurrence of a major adverse cardiovascular event (MACE) within 1 year following PCI.
2. Classify MACE as the occurrence of any of the following: cardiac death, non-fatal myocardial infarction, ischemia-driven revascularization of target vessels or lesions, re-hospitalization due to heart failure, or re-hospitalization for unstable angina pectoris associated with confirmed myocardial ischemia. Define all-cause mortality as a secondary endpoint.
3. Conduct patient follow-up via outpatient clinic records, re-hospitalization data, and telephone interviews.
4. Initiate follow-up immediately post-PCI and terminate at 12 months postoperatively or upon the first occurrence of an endpoint event.
5. Evaluate disputed endpoint events independently by two cardiovascular experts. Involve a third senior physician to review and adjudicate any persistent disagreements.
6. Treat patients lost to follow-up as right-censored observations, censoring their follow-up duration at the date of their last known event-free status.

4. Statistical analysis

1. Perform all statistical analyses using R software (version R 4.3.1) and SPSS software (version SPSS 26.0). Set the significance level at a 2-sided p-value &lt; 0.05.
2. Test continuous variables for normality. Express normally distributed data as mean &plusmn; standard deviation and compare using the independent samples t-test. Express non-normally distributed data as median (interquartile range, IQR) and compare using the Mann-Whitney U test.
3. Present categorical data as frequencies and percentages. Compare groups using the chi-square test or Fisher's exact test.
4. Handle missing data by performing complete case analysis, as patients with missing key parameters were excluded during screening, and the remaining missingness proportion was &lt; 1.5% (assumed missing completely at random, MCAR).
5. Conduct all variable selection and model training exclusively within the training cohort (n = 936) to prevent data leakage. Ensure the validation cohort (n = 401) remains completely isolated for unbiased performance evaluation.
6. Perform collinearity diagnostics prior to modeling by calculating the Variance Inflation Factor (VIF) and Pearson correlation matrices to assess cellular overlap among inflammatory indices.
7. Screen candidate predictors using univariate Cox proportional hazards regression.
8. Apply Least Absolute Shrinkage and Selection Operator (LASSO) regression with 10-fold cross-validation to select features and mitigate overfitting.
9. Construct the final multivariable Cox regression model using variables with non-zero LASSO coefficients to identify independent predictors and calculate the combined risk score.
10. Stratify patients into distinct risk levels based on the calculated model scores. Compare the cumulative incidence of MACE across strata using Kaplan-Meier analysis and the log-rank test.
11. Evaluate model discrimination using Harrell’s concordance index (C-index) and time-dependent receiver operating characteristic (ROC) curve analysis at a 1-year prediction horizon.
12. Assess model calibration via calibration curves and clinical utility via Decision Curve Analysis (DCA).
13. Perform internal validation utilizing bootstrapping with 1,000 resamples to calculate the optimism-corrected C-index and verify model stability.

Results

Baseline characteristics

A total of 1,337 patients with ACS undergoing PCI were included in this study; 208 patients experienced a MACE during follow-up, and 1,129 did not. Compared to the non-MACE group, patients in the MACE group were older and exhibited a higher prevalence of comorbidities, including hypertension, diabetes, prior myocardial infarction (MI), prior PCI, and stroke. The MACE group also presented with a higher Killip class, lower left ventricular ejection fract...

Discussion

This study successfully developed and internally validated a risk prediction model incorporating novel composite inflammatory biomarkers for major adverse cardiovascular events (MACE) in patients with acute coronary syndrome (ACS) undergoing percutaneous coronary intervention (PCI)¹³. The combined model demonstrated a robust discrimination index (AUC = 0.81), significantly outperforming the peak predictive value of any single inflammatory biomarker (AUC = 0.71). Furthermore, the integrated model y...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Coronary angiography records	Jinan Central Hospital	Catheterization laboratory database	Collection of PCI-related procedural variables
Echocardiography system	Jinan Central Hospital	Hospital imaging system	Assessment of LVEF and cardiac structure/function
Electronic medical record system	Jinan Central Hospital	Hospital database system	Collection of demographic, clinical, and follow-up data
Laboratory information system	Jinan Central Hospital	LIS database	Retrieval of blood routine and biochemical test results
R software	R Foundation for Statistical Computing	Version 4.3.0	Statistical analysis and model construction
SPSS software	IBM Corp.	Version 27.0	Data processing and statistical analysis