Fig. 4

Exploration of protein distribution in plasma and development of a multi-marker score. (A-C) The expression levels of PTMA, ADAMTS8, and CD36 proteins were significantly higher in AAD patients than in control subjects. ADAMTS8 levels were elevated to 898.00 pg/mL (median, IQR: 575.50 – 1241.00 vs. 484.00 pg/mL (median, IQR: 392.00 – 726.50) in healthy controls (P < 0.0001), CD36 levels were elevated to 17.59 ng/mL (median, IQR: 10.72 – 24.39) vs. 6.25 ng/mL (median, IQR: 4.54 – 8.91) in healthy controls (P < 0.0001), and PTMA levels were elevated to 0.76 ng/mL (median, IQR: 0.46 – 1.28) vs. 0.57 ng/mL (median, IQR: 0.48 – 0.74) in healthy controls (P = 0.004). (D-F) ROC curve analysis of individual proteins: ADAMTS8 (AUC=0.782, 95%CI: 0.733-0.831), CD36 (AUC=0.881, 95%CI: 0.842-0.920), and PTMA (AUC=0.596, 95%CI: 0.527-0.665). (G) PCA plot based on the top two principal components distinguishing AAD group from healthy controls. (H) Selection of five core predictive factors (D-dimer, ADAMTS8, height, SBP and age) based on PCA loading values. (I) Optimal cut-off values of the five core factors for multi-marker score construction: D-dimer ≥ 500 ng/mL, ADAMTS8 ≥ 802.5 pg/mL, height ≥ 166.5 cm, SBP ≥ 140 mmHg, age ≥ 65 years. (J) Multi-marker scores were significantly higher in the AAD patients than in the controls. (K) ROC curve analysis of the multi-marker score for AAD diagnosis (AUC = 0.921, 95%CI: 0.889 - 0.952; sensitivity 77.5%, specificity 96.5%). (L) Restricted cubic spline regression showing the relationship between multi-marker score and AAD events, with exponentially increased AAD risk when the score ≥ 3 points.