Construction and verification of a nomogram prediction model for vascular plaque stability in patients with progressive cerebral infarction based on serum MCP-1, MCPIP1 combined with inflammatory factors

Objective‍ ‍To construct and validate a nomogram-based prediction model of vascular plaque stability in patients with progressive cerebral infarction based on serum monocyte chemotactic protein-1 (MCP-1), monocyte chemotactic protein-1 inducible protein 1 (MCPIP1) combined with inflammatory factors. Methods‍ ‍A retrospective cohort study was conducted on 200 patients with progressive cerebral infarction admitted to our department from January to December 2023. All of them were assigned into a modeling group, and were divided into a stable plaque subgroup and the unstable plaque subgroup according to results of carotid multilayer spiral CT angiography. Their general data, results of laboratory tests, and other clinical indicators were collected to identify the influencing factors for vascular plaque instability with single-factor and multifactor analyses. Then a nomogram model for predicting vascular plaque stability was constructed for patients with progressive cerebral infarction. Receiver operating characteristics (ROC) curve was plotted to evaluate the predictive performance of the nomogram model. Subsequently, in a ratio of 7∶3 between the cases in the modeling group and the validation group, another 86 patients with progressive cerebral infarction admitted to our department from January to June 2024 were enrolled and served as the validation group. Their clinical data were collected for external validation of the model. Results‍ ‍In the modeling group, there were 68 patients (34.00%) in the stable plaque subgroup and 132 patients (66.00%) in the unstable plaque subgroup. Univariate analysis showed that there were significant differences between the 2 subgroups in terms of age (65.31±6.74 vs 67.52±7.14 years, t=2.113), comorbid diabetes mellitus [35 (48.53%) vs 80 (60.61%) cases, Chi-square=7.182], MCP-1 (570.67±104.23 vs 693.94±128.45 pg/mL, t=6.836), MCPIP1 (2.93±0.58 vs 4.08±0.75 ng/mL, t=11.051), homocysteine (Hcy, 10.56±2.38 vs 16.04±3.54 μmol/L, t=11.491), C-reactive protein (CRP, 6.16±2.03 vs 8.05±2.67 mg/L, t=5.122) and TNF-ɑ (1.31±0.29 vs 1.79±0.47 ng/mL, t=7.696) (all P˂0.05). Multivariate analysis indicated that age (β=0.103, OR=1.109, 95%CI=1.012~1.215), comorbid diabetes (β=2.135, OR=8.461, 95%CI=1.866~38.353), Hcy (β=0.706, OR=2.026, 95%CI=1.550~2.650), MCP-1 (β=0.011, OR=1.011, 95%CI=1.004~1.018), MCPIP1 (β=1.928, OR=6.875, 95%CI=2.765~17.094), CRP (β=0.327, OR=1.387, 95%CI=1.022~1.883) and TNF-ɑ (β=1.491, OR=4.443, 95%CI=1.389~14.212) were independent influencing factors for vascular plaque instability in the patients with progressive cerebral infarction (all P<0.05). The modeling formula based on these factors was Logit (P)=0.103×(age)+2.135×(combined diabetes)+0.706×(Hcy)+0.011×(MCP-1)+1.928×(MCPIP1)+0.327×(CRP)+1.491×(TNF-ɑ)-34.684. ROC curve analysis revealed that the area under curve (AUC) of the model in the modeling group was 0.956 (95%CI: 0.931~0.981, P<0.001), with a sensitivity of 0.841 and a specificity of 0.926, and the AUC value in validation group was 0.960 (95%CI=0.925~0.996, P<0.001). Conclusion‍ ‍Our nomogram prediction model has a good predictive performance for vascular plaque instability in patients with progressive stroke, and it can be used to identify high-risk patients for vascular plaque instability in clinical practice.