Combination of Preoperative D-Dimer and Platelet Distribution width Predicts Postoperative Deep Venous Thrombosis in Patients with Cervical Carcinoma

Background: Deep venous thrombosis (DVT) is associated with severe morbidity and mortality in cancer. Platelet distribution width (PDW), a platelet index, indicates variation in platelet size. We aimed to investigate whether the combination of D-dimer and PDW could have a better performance in predicting DVT in patients with cervical carcinoma. Materials and Methods: In 198 consecutive cervical carcinoma patients without preoperative DVT, preoperative D-dimer and PDW levels were measured. Compression ultrasonography was performed in all cervical carcinoma patients before surgery, as well as one month, three months, six months, and 12 months. Results: During a median period of 12 months, 17 of the 198 patients (8.6 %) developed DVT. PDW levels were reduced and D-dimer levels were increased in patients with DVT events compared to those without DVT. Multivariate Cox analysis revealed that both PDW and D-dimer were independent predictors for DVT events. The area under the ROC curve was 0.628 (95% CI: 0.556 to 0.695, p=0.142) when D-dimer was used alone, whereas it increased to 0.777 (95% CI: 0.712 to 0.833, p<0.011) with the addition of PDW. Incorporation of PDW into the D-dimer model significantly improved the predictive value. Conclusions: The combination of preoperative D-dimer and PDW improves the predictive power of postoperative DVT risk in patients with cervical carcinoma.


Combination of Preoperative D-Dimer and Platelet Distribution width Predicts Postoperative Deep Venous Thrombosis in Patients with Cervical Carcinoma
Na Li 1 , Fu-Bin Zhang 2 , Bing-Jie Li 3 , Rui-Tao Wang 1 * cancer, and laryngeal cancer .
In the present study, we aimed to investigate whether a combination of PDW and D-dimer prior to surgery could better predict the postoperative DVT in cervical carcinoma.

Study population
The present study enrolled 277 consecutive patients who had been diagnosed with cervical carcinoma between 2014 and 2015 at the Harbin Medical University Cancer Hospital, China. All patients undergone complete surgical resection and all diagnoses were based on pathological evidence. None of the patients received preoperative chemotherapy or radiation therapy. The patients with the following characteristics were excluded from the present study: patients who had any coexisting or previous cancers other than cervical carcinoma; patients with concomitant diseases, including hematological disorders, hypertension, diabetes mellitus, atrial fibrillation, and patients who had taken anticoagulant or acetylsalicylic acid drugs in three months.
Our study is a prospective study. A total of 277 patients enrolled in this study. 9 patients were excluded because they refused to give informed consent. 32 patients with hypertension, 38 patients with diabetes mellitus were also excluded. The final analysis included 198 patients. All subjects were followed-up for 12 months. This study complied with the Helsinki Declaration and was approved by the Human Ethics and Research Ethics committees of Harbin Medical University Cancer Hospital. Written informed consent was obtained from all patients.

Biochemical parameters
All participants completed a standardized questionnaire including medical history, lifestyle factors and smoking habits. After a 10-hour overnight fast, blood samples were collected from participants. Platelet-poor plasma was then assayed using second-generation latex immunoassay kit that provides quantitative measure of D-dimer. It was analysed on Sysmex CS-5100 (Siemens Diagnostics, Erlangen, Germany). The standard cut-off value of D-dimer in our hospital is 0.55 mg/L. Platelet count and PDW levels were determined with an autoanalyzer (Sysmex XE-2100, Kobe, Japan). The whole blood samples were collected in EDTA-containing tubes and all samples were processed within 30 minutes after blood collection in order to prevent in vitro platelet activation.

Diagnosis of deep venous thrombosis
An experienced ultrasonographist performed the compression ultrasonography according to standard procedures (grey scale, B-mode, color Doppler) with a high-end ultrasound scanner (DC-8 EXP, Mindray, Shenzhen, China) before surgery, as well as one month, three months, and six months. If the patients were in cases of clinical signs of DVT, an additional compression ultrasonography was performed.

Statistical analysis
All statistical analyses were performed using SPSS Statistics version 22.0 (SPSS Inc., Chicago, IL, USA). Values for continuous variables are given as means ± SD or median (range). Values for categorical data are specified as frequency (percent). Student t test was used for the continuous variables, while the Chi-square test was used for the categorical variables. The Kaplan-Meier method was used to obtain the cumulative incidence proportion of DVT while statistical significance was compared between groups using the log-rank test. Uni-and multivariate analyses were performed using Cox's proportional hazards regression model. Hazard ratios (HRs) estimated from the Cox-regression analysis were reported as relative risks with corresponding 95% confidence intervals. Receiver operating characteristic (ROC) curves were constructed for quantitative variables to determine the best cut-off point values. The area under the ROC curve (AUC) was calculated for the D-dimer model and for the combination of D-dimer and PDW. The differences in AUC were detected by using MedCalc version 15.0. For all analyses, a two-sided P-value of 0.05 or less was considered statistically significant.

Results
Data from the 198 remaining patients were analyzed. 17 (8.6 %) patients had death events. The baseline characteristics of the patients with DVT or without DVT events are summarized in Table 1. Compared to the patients without DVT, those with DVT were older, had a longer duration of postoperative immobilization, and higher frequency of venous thrombo-embolism. Additionally, PDW levels were reduced and D-dimer levels were increased in DVT patients.
A ROC curve for DVT prediction was plotted to verify the optimal cut-off value for PDW, which was 15.8 % (Figure 1). It demonstrated that PDW predicts DVT with a sensitivity of 29.4 % and a specificity of 94.5 % (AUC = 0.629, 95 % CI: 0.558-0.696).
In the univariate Cox regression model, PDW, D-dimer, duration of postoperative immobilization, and past history of venous thrombo-embolism showed significant associations with a higher risk of DVT ( Table  2). The multivariate Cox regression analysis revealed that PDW was an independent predictor for poor prognosis in patients with cervical carcinoma (HR, 4.945; 95% CI, 1.113-21.978; p = 0.036) ( Table 3).
The estimated cumulative incidence of DVT events by the Kaplan-Meier curve was shown in Figure 2. Patients with PDW ≤ 15.8 % had increased cumulative incidence of DVT compared with the patients with PDW > 15.8 % (33.3% vs. 6.6%, p < 0.001).
As shown in Table 3, D-dimer could independently predict future DVT events in cervical carcinoma patients. To evaluate whether the combination of PDW and D-dimer could improve the prediction of DVT, ROC analysis was performed (Fig 3). The area under the ROC curve was 0.628 (95% CI: 0.556 to 0.695, p = 0.142) when D-dimer was used alone, whereas it increased to 0.777 (95% CI Figure 1. Optimized Cut-Off was Determined for PDW Using Standard ROC Curve Analysis DVT in patients with cervical carcinoma. In addition, the combination of D-dimer and PDW has a better performance in predicting DVT. An increased preoperative D-dimer was associated with a higher cumulated incidence of postoperative deep venous thrombosis in patients with colorectal cancer (Stender et al., 2009). However, the nonspecific 0.712 to 0.833, p < 0.001) when the combination of PDW with D-dimer was used.

Discussion
In this study, we observed that preoperative PDW has independent prognostic value for future postoperative Past history of VTE (n, %) 1 (5.9) 0 (0) 0.001  elevations of D-dimer in cancer patients limit its wider applicability (Schaefer et al., 2017). Our results showed that PDW confers an important additive effect to D-dimer. One possible explanation is that reduced PDW indicates platelet activation which exerts extensive effects on atherogenesis, inflammation, and atherothrombosis (Linden and Jackson, 2010). Bone marrow cells (including megakaryocytes) dys-regulation may contribute to changed PDW. PDW is a measure of platelet heterogeneity caused by heterogeneous demarcation of megakarocytes (Paulus,1981). Megakaryocytic maturation, platelet production and platelet size could be regulated by cytokines, such as interleukin-6 (IL-6), granulocytes colony stimulating factor (G-CSF) and macrophage colony stimulating factor (M-CSF) (Kaushansky, 1998). IL-6 promotes tumor angiogenesis, metastasis and metabolism (Kumari et al., 2016). Moreover, the cytokines G-CSF and M-CSF that be secreted by tumor cells could stimulate megakaryopoiesis and subsequent thrombopoiesis in cancer (Kowanetz et al., 2010).