Abstract
Background/Aim: Pancreatic cancer has the highest risk of venous thromboembolism (VTE). Additionally, chemotherapy for cancer patients increases the risk of developing VTE. Due to recent advances in neoadjuvant chemotherapy (NAC) regimens, more patients with resectable pancreatic cancer will receive NAC. However, the incidence, risk, and predictors of developing VTE in these patients have not been fully evaluated. Patients and Methods: We retrospectively evaluated the incidence, risk, and predictors of VTE among 67 consecutive patients with resectable pancreatic cancer who received neoadjuvant combination therapy with gemcitabine+S-1 (NAC-GS) followed by surgery and 45 patients with resectable pancreatic cancer who underwent upfront surgery (Up-S). Results: The incidence of VTE in the NAC-GS and Up-S groups was 10.4% and 6.6%, respectively. Preoperative D-dimer levels were significantly higher in the NAC-GS group, and D-dimer levels were significantly increased during NAC-GS. Preoperative D-dimer level was the only predictor for VTE in patients with resectable pancreatic cancer who received NAC-GS. Conclusion: There is an increased risk of developing VTE during NAC. Screening with D-dimer and taking appropriate measures to suppress critical VTE is essential to provide NAC to patients with resectable pancreatic cancer.
- Pancreatic cancer
- resectable
- neoadjuvant chemotherapy
- venous thromboembolism
- D-dimer
- adverse event
- risk factor
- predictive factor
Venous thromboembolism (VTE) is one of the most critical complications of cancer patients. It affects patient quality of life and the completion of treatment. Cancer increases the risk of developing VTE by at least 4-fold by activating the coagulation systems (1), and 5-20% of cancer patients develop VTE (2, 3). Development of VTE worsens the prognosis of cancer patients (4, 5), with a cancer mortality rate of approximately 20-30% (1), and is the second leading cause of death in these patients (6, 7). The risk of VTE also increases depending on the cancer type, treatment modality, disease stage, and patient-specific factors (1). The risk of VTE in pancreatic cancer is among the highest in all types of cancer along with lung, hematologic, stomach, ovarian, and brain cancer (1-3, 8). The incidence of DVT in patients with pancreatic cancer, including stage IV, was reported to be 28.9% (9). Additionally, chemotherapy is associated with a 6.5-fold increase in the risk of developing VTE in cancer patients (10). It could be stated that patients undergoing neoadjuvant therapy followed by surgery will face a very high risk of developing VTE.
Pancreatic ductal adenocarcinoma (PDAC) is a lethal cancer and the seventh leading cause of cancer death worldwide (11). Surgical resection offers the only curative approach, but surgery alone results in poor outcomes. Many pre- and post-operative adjuvant regimens have been studied (12, 13). Recent advances in adjuvant chemotherapy regimens have improved the outcomes of patients with resected pancreatic cancer, and the use of adjuvant chemotherapy after resection is now strongly recommended (12, 14). In the neoadjuvant setting, some chemotherapy and/or chemoradiotherapy are effective and well-tolerated for borderline resectable pancreatic cancer (BRPC) (15). The broad consensus supports the use of neoadjuvant oncological therapy for patients with BRPC (12, 13, 16). For patients with resectable disease, the role of preoperative therapy is unclear. Some retrospective studies have suggested the potential advantages of neoadjuvant therapy over upfront surgery. Patients with resectable disease but with high-risk features have been recognized as candidates for neoadjuvant therapy (12, 16). The recent Japanese Prep-02/JSAP-05 trial, a multicenter, randomized phase II/III trial, compared neoadjuvant combination chemotherapy with gemcitabine and S-1 (NAC-GS) with upfront surgery in patients with resectable pancreatic cancer, and demonstrated longer median overall survival in patients receiving NAC-GS. This result supports neoadjuvant therapy even for patients with resectable disease.
According to these recent studies, more patients with resectable pancreatic cancer will receive neoadjuvant therapy. To date, increased risk of VTE in patients receiving neoadjuvant therapy compared with surgery alone has been reported in some cancer types, including ovarian, bladder, and esophageal cancer (17). However, the incidence and risk of VTE in pancreatic cancer patients receiving NAC for resectable disease has yet to be thoroughly investigated (18).
To carefully assess, prevent and treat VTE in these patients undergoing neoadjuvant chemotherapy, more clinical information on this subject is needed. Therefore, this study aimed to evaluate the incidence, risk, and predictive factors for developing VTE in patients with resectable pancreatic cancer who underwent NAC-GS.
Patients and Methods
Patients. Sixty-seven consecutive pancreatic cancer patients with resectable disease according to the NCCN criteria of the multidisciplinary cancer board, and who met the inclusion criteria for the Prep-02/JSAP-05 study protocol (19), received adjuvant chemotherapy between April 2019 and December 2020, and were classified into the neoadjuvant group (NAC-GS group). The critical eligibility criteria included age <80 years, Eastern Cooperative Oncology Group (ECOG) performance status 0-1, and adequate hepatic, renal, cardiopulmonary, and bone marrow functions. Forty-five patients who met the above diagnostic and eligibility criteria and who underwent up-front surgery between April 2018 and March 2019 were classified into the up-front surgery group (Up-S group) as a historical control.
NAC-GS administration. NAC-GS was administered based on the protocol of the PREP-02/JSAP05 study. In brief, intravenous gemcitabine (1,000 mg/m2 on days 1 and 8), and oral S-1 [80/100/120 mg/day/body (according to the body surface area) on days 1-14] was repeated every three weeks for two cycles. These chemotherapies were reduced, suspended, or discontinued as needed (19).
Evaluation of VTE. Computed tomography (CT) from the chest to the pelvis was performed at the time of staging and after NAC-GS in all cases. The D-dimer level was calculated before and after NAC-GS or before upfront surgery. Lower extremity venous ultrasonography was applied in cases with elevated D-dimer levels (≥1.3) or if the patient had any symptoms that raised the suspicion of DVT.
Assessment of the risk of VTE and preventive measures. Risk assessment and preventive measures for VTE were taken according to the guidelines for the diagnosis, treatment, and prevention of pulmonary thromboembolism and deep vein thrombosis (20). Patients undergoing pancreatic cancer surgery are considered to be at high risk for VTE. Compression stockings, intermittent pneumatic compression (IPC), and intraoperative and postoperative anticoagulant therapies were used for the prevention of VTE in the high-risk group. In addition, preoperative anticoagulant therapy was also used for the highest-risk group. When VTE was found, we treated these patients as the highest-risk group, and oral anticoagulant therapy with edoxaban was administered. Subsequently, we repeated D-dimer measurement and lower extremity venous ultrasonography as needed.
Statistical analysis. In this study, comparisons between two groups were performed using the Fisher’s exact test, Mann–Whitney U-test, and Wilcoxon signed-rank test, as appropriate. Changes from before to after NAC-GS were analyzed by paired t-test. Multivariate logistic regression analyses were performed to identify risk factors for VTE. Backward elimination was used to select a model. All statistical analyses were two-sided, and p-values of <0.05 were considered to indicate statistical significance. A receiver operating characteristic (ROC) curve was constructed to analyze the sensitivity-specificity relationship of D-dimer levels, and the area under the curve (AUC) was calculated. The EZR software program (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna Austria), was used for all statistical analyses (21).
Ethical review. The present exploratory analysis was conducted in accordance with the Declaration of Helsinki, and the study protocol was approved by the Kanagawa Cancer Center Research Ethics Committee. After a full explanation of the study, written informed consent was obtained from all patients before enrollment.
Results
Patient characteristics and clinical features (Table I). A total of 112 patients were enrolled in this study. Among them, 45 underwent upfront surgery (Up-S group), and 67 received NAC-GS followed by surgery (NAC-GS group). There were no differences in age, sex, ECOG-PS, tumor location, or type of surgical approach between the two groups. The hemoglobin levels were significantly lower in the NAC-GS group (p=0.006), whereas there were no significant differences between the two groups in body mass index (BMI), albumin, WBC count, or platelet count.
Patient characteristics and clinical features.
Incidence of VTE and D-dimer levels (Table I). VTE occurred in 3 of 45 patients (6.6%) of the Up-S group and 7 of 67 patients (10.4%) of the NAC-GS group, which did not amount to a statistically significant difference (p=0.74) (Table I). All VTE cases involved deep venous thrombosis; no cases of pulmonary embolism were observed. The preoperative D-dimer levels in the NAC-GS group were significantly higher than those in the Up-S group (p=0.02). Using the Khorana score (10), 7 patients (15.5%) were classified as high risk in the Up-S group and 14 patients (20.8%) were classified as high risk in the NAC-GS group, and there was no statistical difference between the two groups.
D-dimer levels before and after neoadjuvant therapy (Figure 1). In the NAC-GS group, the D-dimer levels were significantly increased from 0.84 μg/ml (mean±0.42) before NAC-GS to 1.84 μg/ml (mean±2.12) after NAC-GS (p=0.001).
D-dimer levels before and after neoadjuvant chemotherapy (NAC).
ROC analysis (Figure 2). The D-dimer levels before surgery in patients who received NAC-GS were used to construct the ROC model for predicting the development of VTE. Using a cutoff level of 3.8 μg/ml, the sensitivity of D-dimer for predicting VTE was 94.1%, and the specificity was 57.1%. The AUC was 0.7633 (95%CI=0.511-1.000).
Receiver operating characteristic (ROC) curve analysis of D-dimer levels.
Predictors for the development of VTE during neoadjuvant therapy. Among 67 patients enrolled in the NAC-GS group, 7 patients developed VTE. The univariate analysis showed that preoperative D-dimer positivity (cutoff value; 3.8) was the only factor for which a significant difference was observed between the VTE-positive and VTE-negative cases (Table II). The Khorana score showed no significant difference between VTE-positive and VTE-negative cases (Table II). The multivariate analysis identified preoperative D-dimer positivity (higher than 3.8) as the only significant and independent predictor of VTE (odds ratio=72.4; 95%CI=3.57-1470; p=0.01) (Table III).
Risk factors for venous thromboembolism (VTE) after neoadjuvant combination therapy with gemcitabine+S-1.
Multivariate analysis of predictive factors for venous thromboembolism (D-dimer cut-off=3.8).
Discussion
We retrospectively evaluated the incidence, risk, and predictors of the development of VTE in patients with resectable pancreatic cancer who underwent NAC-GS before surgery. The incidence of VTE in the Up-S and NAC-GS groups did not differ to a statistically significant extent. However, the preoperative D-dimer levels were significantly higher in the NAC-GS group than those in the Up-S group. Additionally, the D-dimer levels significantly increased during NAC-GS, suggesting that they had an increased risk of developing VTE during NAC-GS. Preoperative D-dimer positivity was indicated as the only predictor of the development of VTE in patients receiving NAC for resectable pancreatic cancer.
Chemotherapy increases the risk of VTE up to 6.5-fold (22, 23). Chemotherapy causes inappropriate activation of hemostasis through various mechanisms, including direct damage to endothelial cells and amplification of the prothrombotic potential of cancer cells (24). A large observational study cohort of unselected patients who were receiving chemotherapy, and were registered in the United States IMPACT health care insurance claims database, revealed that the incidence of VTE was 7.3% at 3.5 months after the initiation of chemotherapy and increased to 13.5% at 12 months after the initiation of chemotherapy (1). Patients with pancreatic cancer showed the highest incidence (11.6% at 3.5 months after the initiation of chemotherapy and 21.3% at 12 months after the initiation of chemotherapy) (1).
While the risk of VTE in patients receiving chemotherapy for advanced cancer is well established, that in patients receiving neoadjuvant treatment who are intending to undergo surgery remain unclear. In a systematic review of more than 7,800 patients including various types of cancers, the overall incidence of VTE was 7% during neoadjuvant therapy (25). A high incidence and increased risk of development of VTE during neoadjuvant treatment and the perioperative period have been reported in some cancer types, including ovarian, bladder, and esophageal cancer (17).
The incidence and risk of development of VTE in patients with resectable pancreatic cancer who are undergoing NAC have yet to be fully investigated (18, 26, 27). In patients with pancreatic cancer, including resectable and borderline resectable diseases, 9.3% developed VTE during neoadjuvant therapies (18). Among patients who received neoadjuvant therapy with various chemotherapy regimens before surgical resection, 6% developed VTE during neoadjuvant chemotherapy, and 20% developed VTE during neoadjuvant therapy and during the perioperative and postoperative periods (26). These rates were in line with 10.4% in our patient cohort. In addition, neoadjuvant therapy for patients with pancreatic cancer increased the risk of developing VTE after surgery, and VTE development was associated with poor outcomes (27). These results indicate an increased risk of developing VTE even in the adjuvant setting, and VTE should be carefully assessed and prevented, and that prophylactic anticoagulation should be considered in these patients.
In our study, the D-dimer levels significantly increased during NAC, and preoperative D-dimer levels in the NAC-GS group were significantly higher than those in the Up-S group. These suggest an increased potential risk of developing VTE during NAC, as reported in esophageal cancer (28). In our patient who received NAC-GS, an elevated D-dimer level was the only significant predictor of developing VTE during NAC. However, decreased hemoglobin levels and increased platelet counts have been reported to predict the development of VTE during NAC (26). In addition, prognostic nutritional index, blood urea nitrogen, and D-dimer have been reported as predictors of DVT in pancreatic surgery (29). In general, D-dimer is the most common biomarker for the prediction of VTE in cancer patients (24, 30). Therefore, it is essential to monitor D-dimer levels to assess VTE in patients undergoing NAC for resectable pancreatic cancer.
The Khorana score is one of the most well-known indicators of VTE in ambulatory outpatients with advanced cancer who are undergoing chemotherapy (10). The score includes the cancer site, platelet count, leukocyte count, hemoglobin level, and BMI. A total score of 0 to 1 indicates a low risk, 2 indicates an intermediate risk, and ≥3 indicates a high risk of VTE (10). In our study, the Khorana score could not predict the development of VTE in the patients undergoing NAC, as previously reported in ovarian, bladder, and pancreatic cancer (26, 31, 32). The predictive value of the Khorana score in the neoadjuvant setting remains to be elucidated.
The present study was associated with some limitations. First, this was a single-center retrospective study with a relatively small sample size. Second, because of the change in treatment strategy for resectable pancreatic cancer in our institution, based on the results of Prep-02/JSAP-05 study, patients who underwent upfront surgery were enrolled as historical controls. Nonetheless, to our knowledge, this study is the first to report the incidence, risk, and predictors of VTE in patients receiving NAC for strictly categorized resectable pancreatic cancer.
In conclusion, VTE occurred in 10.4% of patients who underwent NAC-GS for resectable pancreatic cancer. D-dimer levels significantly increased during NAC-GS, and the preoperative D-dimer levels in the NAC-GS group were significantly higher in comparison to those in the Up-S group, suggesting an increased risk of developing VTE during NAC-GS. Furthermore, an elevated D-dimer level was the only significant and independent predictor of VTE in patients who underwent NAC-GS. Therefore, the monitoring of D-dimer levels may play an essential role in carefully assessing, preventing, and treating VTE in patients undergoing neoadjuvant therapy for resectable pancreatic cancer.
Acknowledgements
This work was supported by non-governmental organizations: the Kanagawa Cancer Center and the Yokohama Surgical Research Group (Yokohama, Japan).
Footnotes
Authors’ Contributions
All Authors contributed to the study conception and design. Data collection of the clinical was performed by MK, SK, YK, MM, TA, SK, NY, TO, MU, NY, YR, MM and SM. Data analysis and references collection were performed by MK. The first draft of the manuscript was written by MK. The meaning of the manuscript was discussed and revised by MK and SM. All Authors read and approved the final manuscript.
Conflicts of Interest
The Authors declare no competing interests in association with the present study.
- Received January 15, 2023.
- Revision received January 30, 2023.
- Accepted February 2, 2023.
- Copyright © 2023 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
