Abstract:

Objective To evaluate the risk stratification value of plasma D-dimer （D-D） levels for venous thromboembolism （VTE） within 6 months in patients with malignant solid tumors. Methods A total of 11 082 patients with pathologically confirmed malignant solid tumors who were first treated at Cancer Hospital of Shantou University Medical College from January 2021 to December 2023 were retrospectively included， 1 222 cases among which completed imaging examinations such as vascular ultrasound， lower-limb venous ultrasound， or CT pulmonary angiography （CTPA） within 6 months. After excluding 95 cases with missing data， 1 127 patients were divided into the VTE group （n=338） and the non-VTE group （n=789） based on the VET diagnosis results. Gender， age， tumor location， baseline D-D measured at first visit （D-D1）， and D-D obtained on or the day before imaging （D-D2） were compared between the two groups. Multivariate logistic regression was used to analyze the independent association between plasma D-D levels and VTE. Based on the results of the multivariate analysis， the “boot” package in R was employed to perform 1 000 Bootstrap resampling iterations. In each iteration， 70% of the samples were randomly selected as the training set for constructing the prediction model， while the remaining 30% served as the validation set. Logistic regression was adopted for model construction， and a nomogram was generated using the R package “nomogram”. The receiver operator characteristic （ROC） curve was constructed to evaluate its diagnostic efficacy. Results There was a statistically significant difference in D-D1 levels between patients of different genders （Z=-5.83， P<0.001）. There were statistically significant differences in the levels of D-D1 and D-D2 in patients of different ages （χ²=585.52， P<0.001； χ²=58.56， P<0.001） and different tumor locations （χ²=1 051.12， P<0.001；χ²=227.64， P<0.001）. There were statistically significant differences in age and tumor location between the VTE group and the non-VTE group （t=-3.70， P<0.001； χ²=3 431.24， P<0.001）. The levels of D-D1 and D-D2 were significantly higher in the VTE group than those in the non-VTE group （Z=9.80， P<0.001； Z=17.12， P<0.001）. Multivariate analysis demonstrated that gender （female：OR=1.87， 95%CI：1.20-2.90， P=0.006）， age （61-70 years old：OR=0.56， 95%CI：0.32-0.98， P=0.042）， tumor location （esophagus：OR=0.30， 95%CI：0.14-0.67， P=0.003； gastrointestinal tract：OR=0.31， 95%CI：0.15-0.68， P=0.003； breast：OR=0.15， 95%CI：0.07-0.33， P<0.001； urinary tract：OR=0.33， 95%CI：0.13-0.86， P=0.023）， and D-D2 levels [551-1 100 μg/L fibrinogen equivalent units （FEU） （OR=2.55， 95%CI：1.31-4.99， P=0.006）， 1 101-4 000 μg/L FEU （OR=9.17， 95%CI：5.06-16.61， P<0.001）， and ≥4 001 μg/L FEU （OR=21.09， 95%CI：11.38-39.08， P<0.001）] were independent influencing factors for VTE in patients with malignant solid tumors. The risk of VTE increased with rising D-D2 levels. A multivariate nomogram was constructed to predict the risk of VET occurrence in patients with malignant solid tumors based on gender， age， tumor location， and D-D2 level. A D-D2 four-tier nomogram was constructed to predict the risk of VET occurrence in patients with malignant solid tumors based on the D-D2 four-tier stratification. ROC curve analysis showed that in the training set， the area under the curve （AUC） of the multivariate nomogram model for predicting VET in patients with malignant solid tumors was 0.828 （95%CI：0.798-0.858）， while the AUC of the D-D2 four‑stratification model （using 1 101-4 000 μg/L FEU as the optimal cutoff interval） was 0.811 （95%CI：0.781-0.840）. The predictive performance of the multivariate model was superior to that of the D-D2 four‑stratification model （Z=3.74， P<0.001）. In the test set， the AUC of the multivariate nomogram model for predicting VET in patients with malignant solid tumors was 0.814 （95%CI：0.763-0.864）， and that of the D-D2 four-stratification model was 0.787 （95%CI：0.733-0.841）， with no statistically significant difference （Z= 1.90， P=0.057）. Conclusions Elevated D-D is an independent risk factor for VTE within 6 months in malignant solid tumor patients. A threshold of ≥4 001 µg/L FEU can trigger intensive thrombotic work-up， facilitating early identification of high-risk patients and improving prognosis.

Key words: D-dimer, Neoplasms, Venous thromboembolism, Risk assessment, Diagnosis