Key result
Adding relative CRP change improves post-AMI in-stent restenosis prediction to an AUC of ~0.82.
Why the study?
The associations of the post-procedural CRP-triglyceride-fasting glucose index (post-CTI), change in CTI (ΔCTI), and relative change in CRP (ΔlnCRP) with in-stent restenosis (ISR) in patients with acute myocardial infarction (AMI) undergoing drug-eluting stent (DES) implantation needed evaluation.
Does the relative change in CRP (ΔlnCRP), post-procedural CRP-triglyceride-fasting glucose index (post-CTI), or change in CTI (ΔCTI) predict in-stent restenosis in patients with acute myocardial infarction undergoing drug-eluting stent implantation?
Cohort (n=397)
No
Does the relative change in CRP (ΔlnCRP), post-procedural CRP-triglyceride-fasting glucose index (post-CTI), or change in CTI (ΔCTI) predict in-stent restenosis in patients with acute myocardial infarction undergoing drug-eluting stent implantation?
Answers are grounded in this paper and cite their sources.
Relative change in CRP (ΔlnCRP) and the CRP-triglyceride-fasting glucose index (post-CTI and ΔCTI) are independently associated with in-stent restenosis after drug-eluting stent implantation in acute myocardial infarction patients, with ΔlnCRP offering the best predictive model performance (AUC 0.818).
OBJECTIVE: To evaluate the associations of the post-procedural CRP-triglyceride-fasting glucose index (post-CTI), the change in CTI (ΔCTI), and the relative change in CRP (ΔlnCRP) with in-stent restenosis (ISR) in patients with acute myocardial infarction (AMI) undergoing drug-eluting stent (DES) implantation. METHODS: This single-center retrospective cohort study included 397 AMI patients treated with DES. Three indices were evaluated: ΔlnCRP, post-CTI, and ΔCTI. Four models were built: base model, base + ΔlnCRP, base + ΔCTI, and base + post-CTI. Model performance was assessed using AUC, NRI, IDI, Brier score, calibration, and decision curve analysis (DCA). The shape of the association between each marker and ISR was evaluated using progressively adjusted models and restricted cubic splines (RCS). Subgroup and sensitivity analyses were also performed to assess the robustness of the findings. RESULTS: The ISR group had significantly higher monocyte count, HbA1c, post-procedural CRP and glucose, post-CTI, ΔCTI, and ΔlnCRP (all p < 0.05), with more diabetes and three-vessel disease (p < 0.05). In the fully adjusted model, all three markers were independently associated with ISR, with a significant trend across quartiles (all P for trend < 0.001) and effect magnitudes ordered as post-CTI > ΔlnCRP > ΔCTI. post-CTI showed an approximately linear association (P for non-linearity = 0.299). ΔCTI exhibited a U-shaped association (P for non-linearity = 0.011), with the lowest risk at ≈ -0.59 (OR = 0.56) and a significant increase at 1.00 (OR = 3.14). ΔlnCRP showed a threshold effect (P for non-linearity = 0.002). Model 2 (basic model + ΔlnCRP) achieved the highest AUC (0.818), outperforming Model 1 (0.735, p = 0.003) and Model 3 (0.762, p = 0.027), with the greatest NRI and IDI improvements. All models were well calibrated (Brier scores < 0.25; Model 2 lowest at 0.108), and Model 2 provided the highest net benefit at thresholds of 0.20-0.40. All interaction p values exceeded 0.05. Sensitivity analysis shows that the conclusions are independent of the type of medication used after surgery. CONCLUSIONS: ΔlnCRP, post-CTI, and ΔCTI were all independently associated with ISR in AMI patients after DES implantation. ΔlnCRP provides the best model performance, while post-CTI shows the strongest effect size and ΔCTI captures the inflammation-metabolism dynamic changes. ΔlnCRP may serve as a simple, low-cost complementary biomarker for ISR risk stratification.
No takes yet. Share an insight, caveat, or question.
郑庆华 et al. (2026) conducted a cohort in Acute myocardial infarction undergoing drug-eluting stent implantation (n=397). Relative change in CRP (ΔlnCRP), post-procedural CRP-triglyceride-fasting glucose index (post-CTI), and change in CTI (ΔCTI) vs. Base clinical model / lower biomarker quartiles was evaluated on In-stent restenosis (ISR) (AUC 0.818, p=<0.001). Adding relative CRP change (ΔlnCRP) to a base model best predicted in-stent restenosis after DES in AMI patients.
Synapse has enriched 10 related papers. Consider them for comparative context: