Abstract
OBJECTIVE: The aim of this study was to assess the predictive value of 256-detector-row helical computed tomographic pulmonary angiography (CTPA) parameters, combined with serum D-dimer levels, for risk stratification and 30-day prognostic assessment in patients diagnosed with acute pulmonary embolism (APE). METHODS: This retrospective cohort study included clinical data from 77 patients with APE, stratified into low-risk (n = 38), intermediate-low-risk (n = 17), intermediate-high-risk (n = 14), and high-risk (n = 8) groups, as well as 89 control individuals, treated at Baoshan People's Hospital between January 2021 and December 2024. Clinical characteristics, CTPA-derived metrics [pulmonary artery (PA)-to-ascending aorta diameter ratio, right-to-left ventricular diameter ratio (RVD/LVD), and superior vena cava diameter], and D-dimer levels were analyzed and compared across groups. Receiver operating characteristic curve analysis and multivariate logistic regression were used to assess predictive performance. RESULTS: The incidence of bilateral pulmonary embolism increased with escalating risk stratification, from 44.7% in the low-risk group to 87.5% in the high-risk group. Similarly, the frequency of thrombus involvement in the main PA and its branches rose from 28.9% to 100%. The RVD/LVD ratio demonstrated a significant positive correlation with risk stratification (ρ = 0.492;95%CI: 0.30-0.64; p < 0.001) and showed statistically significant differences across groups (H = 26.093, p < 0.001); median values were higher in the intermediate-high-risk (1.55) and high-risk (1.52) groups compared with the low-risk group (0.99). D-dimer concentrations increased progressively with risk stratification (H = 45.593, p < 0.001) and effectively differentiated patients with APE from control individuals (p < 0.001), although no significant differences were observed among APE subgroups. The combined model incorporating RVD/LVD and D-dimer yielded an area under the curve (AUC) of 0.917 (95% CI: 0.852-0.983) for predicting high-risk APE, surpassing the performance of individual indicators (RVD/LVD AUC = 0.874; D-dimer AUC = 0.716). Multivariate logistic regression identified both RVD/LVD (OR = 124.605, p = 0.001) and D-dimer (OR = 1.313, p = 0.032) as independent predictors of poor short-term prognosis. CONCLUSION: The integration of CTPA-derived imaging parameters with D-dimer levels shows promise for risk stratification and prognostic assessment in APE, with the RVD/LVD ratio emerging as a key imaging biomarker. However, given the single-center retrospective design and limited sample size, these findings should be considered exploratory. The combined model requires external validation in larger, multicenter cohorts before its routine clinical applicability, especially in diverse healthcare settings, can be established.