A nomogram risk prediction model for no-reflow after primary percutaneous coronary intervention based on rapidly accessible patient data among patients with ST-segment elevation myocardial infarction and its relationship with prognosis

Machine learning prediction of no reflow in patients with ST-segment elevation myocardial infarction undergoing primary percutaneous coronary intervention

Background

No-reflow (NRF) phenomenon is a significant challenge in patients with ST-segment elevation myocardial infarction (STEMI) undergoing primary percutaneous coronary intervention (pPCI). Accurate prediction of NRF may help improve clinical outcomes of patients. This retrospective study aimed at creating an optimal model based on machine learning (ML) to predict NRF in these patients, with the additional objective of guiding pre- and intra-operative decision-making to reduce NRF incidence.

Methods

Data were collected from 321 STEMI patients undergoing pPCI between January 2022 and May 2023, with the dataset being randomly divided into training and internal validation sets in a 7:3 ratio. Selected features included pre- and intra-operative demographic data, laboratory parameters, electrocardiogram, comorbidities, patients' clinical status, coronary angiographic data, and intraoperative interventions. Post comprehensive feature cleaning and engineering, three logistic regression (LR) models [LR-classic, LR-random forest (LR-RF), and LR-eXtreme Gradient Boosting (LR-XGB)], a RF model and an eXtreme Gradient Boosting (XGBoost) model were developed within the training set, followed by performance evaluation on the internal validation sets.

Results

Among the 261 patients who met the inclusion criteria, 212 were allocated to the normal flow group and 49 to the NRF group. The training group consisted of 183 patients, while the internal validation group included 78 patients. The LR-XGB model, with an area under the curve (AUC) of 0.829 [95% confidence interval (CI): 0.779-0.880], was selected as the representative model for logistic regression analyses. The LR model had an AUC slightly lower than XGBoost model (AUC 0.835, 95% CI: 0.781-0.889) but significantly higher than RF model (AUC 0.731, 95% CI: 0.660-0.802). Internal validation underscored the unique advantages of each model, with the LR model demonstrating the highest clinical net benefit at relevant thresholds, as determined by decision curve analysis. The LR model encompassed seven meaningful features, and notably, thrombolysis in myocardial infarction flow after initial balloon dilation (TFAID) was the most impactful predictor in all models. A web-based application based on the LR model, hosting these predictive models, is available at https://l7173o-wang-lyn.shinyapps.io/shiny-1/.

Conclusions

A LR model was successfully developed through ML to forecast NRF phenomena in STEMI patients undergoing pPCI. A web-based application derived from the LR model facilitates clinical implementation.

No-Reflow Prediction in Acute Coronary Syndrome During Percutaneous Coronary Intervention: The NORPACS Risk Score

BACKGROUND

Suboptimal coronary reperfusion (no reflow) is common in acute coronary syndrome percutaneous coronary intervention (PCI) and is associated with poor outcomes. We aimed to develop and externally validate a clinical risk score for angiographic no reflow for use following angiography and before PCI.

METHODS

We developed and externally validated a logistic regression model for prediction of no reflow among adult patients undergoing PCI for acute coronary syndrome using data from the Melbourne Interventional Group PCI registry (2005-2020; development cohort) and the British Cardiovascular Interventional Society PCI registry (2006-2020; external validation cohort).

RESULTS

A total of 30 561 patients (mean age, 64.1 years; 24% women) were included in the Melbourne Interventional Group development cohort and 440 256 patients (mean age, 64.9 years; 27% women) in the British Cardiovascular Interventional Society external validation cohort. The primary outcome (no reflow) occurred in 4.1% (1249 patients) and 9.4% (41 222 patients) of the development and validation cohorts, respectively. From 33 candidate predictor variables, 6 final variables were selected by an adaptive least absolute shrinkage and selection operator regression model for inclusion (cardiogenic shock, ST-segment-elevation myocardial infarction with symptom onset >195 minutes pre-PCI, estimated stent length ≥20 mm, vessel diameter <2.5 mm, pre-PCI Thrombolysis in Myocardial Infarction flow <3, and lesion location). Model discrimination was very good (development C statistic, 0.808; validation C statistic, 0.741) with excellent calibration. Patients with a score of ≥8 points had a 22% and 27% risk of no reflow in the development and validation cohorts, respectively.

CONCLUSIONS

The no-reflow prediction in acute coronary syndrome risk score is a simple count-based scoring system based on 6 parameters available before PCI to predict the risk of no reflow. This score could be useful in guiding preventative treatment and future trials.

Predictive effect of triglyceride-glucose index on No-Reflow Phenomenon in patients with type 2 diabetes mellitus and acute myocardial infarction undergoing primary percutaneous coronary intervention

OBJECTIVE

Triglyceride glucose (TyG) index is considered as a new alternative marker of insulin resistance and a clinical predictor of type 2 diabetes mellitus (T2DM) combined with coronary artery disease. However, the prognostic value of TyG index on No-Reflow (NR) Phenomenon in T2DM patients with acute myocardial infarction (AMI) remains unclear.

METHODS

In this retrospective study, 1683 patients with T2DM and AMI underwent primary percutaneous coronary intervention (PCI) were consecutively included between January 2014 and December 2019. The study population was divided into two groups as follows: Reflow (n = 1277) and No-reflow (n = 406) group. The TyG index was calculated as the ln [fasting triglycerides (mg/dL)×fasting plasma glucose (mg/dL)/2].Multivariable logistic regression models and receiver-operating characteristic curve analysis were conducted to predict the possible risk of no-reflow. Net Reclassification Improvement (NRI) and Integrated Discrimination Improvement (IDI) were calculated to determine the ability of the TyG index to contribute to the baseline risk model.

RESULTS

Multivariable logistic regression models revealed that the TyG index was positively associated with NR[OR,95%CI:5.03,(2.72,9.28),p<0.001] in patients with T2DM and AMI. The area under the curve (AUC) of the TyG index predicting the occurrence of NR was 0.645 (95% CI 0.615-0.673; p < 0.001)], with the cut-off value of 8.98. The addition of TyG index to a baseline risk model had an incremental effect on the predictive value for NR [net reclassification improvement (NRI): 0.077(0.043to 0.111), integrated discrimination improvement (IDI): 0.070 (0.031to 0.108), all p < 0.001].

CONCLUSIONS

High TyG index was associated with an increased risk of no-reflow after PCI in AMI patients with T2DM. The TyG index may be a valid predictor of NR phenomenon of patients with T2DM and AMI. Early recognition of NR is critical to improve outcomes with AMI and T2DM patients.

A Prediction Model Based on Systemic Immune-Inflammatory Index Combined with Other Predictors for Major Adverse Cardiovascular Events in Acute Myocardial Infarction Patients

Objective

To evaluate the prognostic value of the systemic immune-inflammatory index (SII) for predicting in-hospital major adverse cardiovascular events (MACEs) in patients with acute myocardial infarction (AMI) and establish a relevant nomogram.

Methods

This study included 954 AMI patients. We examined three inflammatory factors (SII, platelet to lymphocyte ratio (PLR) and neutrophil to lymphocyte ratio (NLR)) to see which one predicts in-hospital MACEs better. The predictors were subsequently screened using bidirectional stepwise regression method, and a MACE nomogram was constructed via logistic regression analysis. The predictive value of the model was evaluated using the area under the curve (AUC), sensitivity and specificity. In addition, the clinical utility of the nomogram was evaluated using decision curve analysis. We also compared the nomogram with the Global Registry of Acute Coronary Events (GRACE) scoring system.

Results

334 (35.0%) patients had MACEs. The SII (AUC =0.684) had a greater predictive value for in-hospital MACEs in AMI patients than the PLR (AUC =0.597, P<0.001) or NLR (AUC=0.654, P=0.01). The area under the curve (AUC) of the SII-based multivariable model for predicting MACEs, which was based on the SII, Killip classification, left ventricular ejection fraction, age, urea nitrogen (BUN) concentration and electrocardiogram-based diagnosis, was 0.862 (95% CI: 0.833-0.891). Decision curve and calibration curve analysis revealed that SII-based multivariable model demonstrated a good fit and calibration and provided positive net benefits than the model without SII. The predictive value of the SII-based multivariable model was greater than that of the GRACE scoring system (P<0.001).

Conclusion

SII is a promising, reliable biomarker for identifying AMI patients at high risk of in-hospital MACEs, and SII-based multivariable model may serve as a quick and easy tool to identify these patients.

The predictive value of laboratory parameters for no-reflow phenomenon in patients with ST-elevation myocardial infarction following primary percutaneous coronary intervention: A meta-analysis

To date, the predictive role of laboratory indicators for the phenomenon of no flow is unclear. Hence, our objective was to conduct a meta-analysis to investigate the association between laboratory parameters and the risk of the no-reflow phenomenon in patients with ST-elevation myocardial infarction (STEMI) following primary percutaneous coronary intervention (PCI). This, in turn, aims to offer valuable insights for early clinical prediction of no-reflow. We searched Pubmed, Embase, and Cochrane Library from the establishment of the database to October 2023. We included case-control or cohort study that patients with STEMI following primary PCI. We excluded repeated publication, research without full text, incomplete information or inability to conduct data extraction and animal experiments, reviews, and systematic reviews. STATA 15.1 was used to analyze the data. The pooled results indicated that elevated white blood cell (WBC) count (odds ratio [OR] = 1.061, 95% confidence interval [CI]: 1.013-1.112), neutrophil count (OR = 1.324, 95% CI: 1.128-1.553), platelet (PLT) (OR = 1.002, 95% CI: 1.000-1.005), blood glucose (OR = 1.005, 95% CI: 1.002-1.009), creatinine (OR = 1.290, 95% CI: 1.070-1.555), total cholesterol (TC) (OR = 1.022, 95% CI: 1.012-1.032), d-dimer (OR = 1.002, 95% CI: 1.001-1.004), and fibrinogen (OR = 1.010, 95% CI: 1.005-1.015) were significantly associated with increased risk of no-reflow. However, elevated hemoglobin was significantly associated with decreased risk of no-reflow. In conclusion, our comprehensive analysis highlights the predictive potential of various parameters in assessing the risk of no-reflow among STEMI patients undergoing PCI. Specifically, WBC count, neutrophil count, PLT, blood glucose, hemoglobin, creatinine, TC,  d-dimer, and fibrinogen emerged as significant predictors. This refined risk prediction may guide clinical decision-making, allowing for more targeted and effective preventive measures to mitigate the occurrence of no-reflow in this patient population.

Prognostic value of systemic immune-inflammation index in CAD patients: Systematic review and meta-analyses

BACKGROUND

Systemic immune-inflammation index (SII) is a novel inflammatory marker based on neutrophils, platelets and lymphocytes counts, which has potential prognostic value among coronary artery disease (CAD) patients as described by some observational studies. We aimed to provide higher-certainty evidence to verify the association of SII with poor outcomes of CAD patients.

METHODS

PubMed, Web of Science, Embase, Ovid and Scopus were searched to find relevant literature exploring the prognostic value of SII among CAD patients. Hazard ratios (HRs) with 95% confidence intervals (CIs) extracted from the literature included were pooled with the fixed-effect or random-effect model. Sensitivity analyses and subgroup analyses were conducted to detect the source of heterogeneity and evaluate the stability of results.

RESULTS

A total of nine studies with 15,832 participants were included. The quantitative synthesis including eight studies with 15,657 participants showed that the high SII was related to the major adverse cardiovascular event in CAD patients (HR with 95% CI: 2.36 [1.67, 3.33]). After eliminating heterogeneity and adjusting for publication bias, the above result was still robust (HR with 95% CI: 1.67 [1.32, 2.12]). Additionally, we also demonstrated the prognostic values of SII for all-cause death, cardiovascular death, myocardial infarction and stroke.

CONCLUSION

Higher SII has prognostic values for adverse outcomes in CAD patients.

Predicting the no-reflow phenomenon in ST-elevation myocardial infarction patients undergoing primary percutaneous coronary intervention: a systematic review of clinical prediction models

BACKGROUND

The no-reflow (NRF) phenomenon is the "Achilles heel" of interventionists after performing percutaneous coronary intervention (PCI) in patients with ST-segment elevation myocardial infarction (STEMI). No definitive treatment has been proposed for NRF, and preventive strategies are central to improving care for patients who develop NRF.

OBJECTIVES

In this study, we aim to investigate the clinical prediction models developed to predict NRF in STEMI patients undergoing primary PCI.

DESIGN

Systematic review.

DATA SOURCES AND METHODS

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were observed. Studies that developed clinical prediction modeling for NRF after primary PCI in STEMI patients were included. Data extraction was performed using the Checklist for Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modeling Studies (CHARMS) checklist. The Prediction Model Risk of Bias Assessment Tool (PROBAST) tool was used for critical appraisal of the included studies.

RESULTS

The three most common predictors were age, total ischemic time, and preoperative thrombolysis in myocardial infarction flow grade. Most of the included studies internally validated their developed model via various methods: random split, bootstrapping, and cross-validation. Only three studies (18%) externally validated their model. Six studies (37%) reported a calibration plot with or without the Hosmer-Lemeshow test. The reported area under the curve ranged from 0.648 to 0.925. The most common biases were in the statistical domain.

CONCLUSION

Clinical prediction models aid in individualizing care for STEMI patients with NRF after primary PCI. Of the 16 included studies, we report four to have a low risk of bias and low concern with regard to our research question, which should undergo external validation with or without updating in future studies.