Blood-based biomarkers and major adverse cardiac events after percutaneous coronary intervention: a systematic review with narrative synthesis

Abstract

Major adverse cardiac events (MACE) remain common after percutaneous coronary intervention (PCI) despite advances in stent technology and secondary-prevention pharmacotherapy. Blood-based biomarkers have been proposed as tools for individualised risk stratification, but reported prognostic associations vary widely across studies. This review addressed two prespecified and analytically separate questions. The primary (prognostic) question was: among adults undergoing PCI, are elevated concentrations of blood-based biomarkers associated with subsequent MACE and mortality? The secondary (treatment-effect) question was: among adults undergoing PCI, do specified pharmacotherapeutic strategies reduce MACE? The two questions were addressed using separate eligibility criteria, separate effect measures, separate risk-of-bias instruments and separate syntheses; prognostic associations and treatment effects were not pooled together. MEDLINE (via PubMed), Embase (via Ovid), Web of Science Core Collection and the Cochrane Central Register of Controlled Trials (CENTRAL) were searched from 1 January 2010 to 28 February 2025, supplemented by ClinicalTrials.gov, the WHO ICTRP, OpenGrey and backward and forward citation searching. Records were restricted to English-language reports. Eligible studies for the primary question enrolled adults (≥18 years) undergoing PCI, measured a blood-based biomarker at a defined time point relative to the index procedure, and reported MACE or mortality with an extractable adjusted or unadjusted effect estimate. Risk of bias was assessed with QUIPS for prognostic factor studies, supported by ROBINS-I, Cochrane RoB 2.0 and AMSTAR 2 as appropriate to design. Random-effects meta-analysis was to use the restricted maximum likelihood estimator for between-study variance with Hartung–Knapp–Sidik–Jonkman confidence intervals, with DerSimonian–Laird estimates as a sensitivity analysis. No analyte met the minimum study count required to trigger pooling (Section 3.5), so no meta-analysis was performed. Pooling was restricted to biomarker–outcome pairs with at least three methodologically comparable primary studies. Prediction intervals accompany every pooled estimate derived from five or more studies. Certainty of evidence was rated using a GRADE framework adapted for prognostic factor reviews. Forty-two studies were included, comprising 41 primary reports and one systematic review retained for cross-referencing only, together contributing 37,524 participants to the prognostic synthesis. The evidence base was dispersed across 42 distinct analytes: 38 were evaluated in a single study and four in two studies, so no analyte reached the minimum of three studies prespecified for quantitative pooling. Meta-analysis was therefore not undertaken, and the evidence is reported as a structured narrative synthesis. Elevated concentrations of N-terminal pro-B-type natriuretic peptide, high-sensitivity cardiac troponin T, C-reactive protein and its high-sensitivity assay, interleukin-6, the systemic immune-inflammation index and the neutrophil-to-lymphocyte ratio were each reported as associated with increased risk of MACE, consistently in direction but on the basis of one or two studies per analyte. Evidence for microRNA-497-5p and other emerging analytes rested on single small studies and is reported as preliminary. Thirteen studies were rated at high overall risk of bias, seventeen at moderate and twelve at low. Certainty was very low for the 38 single-study analytes and low for the remainder. Treatment-effect evidence identified under the secondary question is reported separately and was not combined with the prognostic synthesis. Blood-based biomarkers show consistent directional associations with MACE after PCI, but the evidence is too dispersed for quantitative synthesis: no analyte was evaluated in more than two of the included studies. Heterogeneity in assay platform, sampling time, cut-off derivation and endpoint definition compounds this dispersion. These findings support further evaluation of a small number of candidate markers in adequately powered cohorts with standardised measurement, not the incorporation of any biomarker into routine post-PCI decision-making. No included study randomised patients on the basis of a biomarker result, so no biomarker-guided treatment recommendation can be made. https://www.crd.york.ac.uk/PROSPERO/view/CRD420251076074 , PROSPERO CRD420251076074.