The association between malignancy and thromboembolism (TE) was first described more than 100 years ago as ‘migratory thrombophlebitis’, commonly found in patients with visceral cancer.1 Indeed, TE is a common complication in patients with cancer and the second most common cause of death among such patients.2

Although the association between TE and pancreatic cancer is well established, its effects on overall survival remain unclear. The results of studies conducted in Western countries generally support the notion that TE is associated with worse overall survival.3 4 For example, a recent large retrospective study in France demonstrated a statistically significant decrease in overall survival of 2.9 months among patients with TE, compared with patients who did not exhibit TE.3 In contrast, studies involving Asian populations tend to show similar overall survival in patients with and without TE.5 6 7 Furthermore, among the published retrospective studies concerning the incidence of TE in Asian patients with pancreatic cancer, very few data have focused on the impact of TE in Chinese patients with pancreatic cancer.

In this study, we aimed to investigate the incidence of TE among patients with pancreatic cancer in our centre, where >99% of patients are Chinese; explore risk factors associated with the development of TE; and assess the prognostic impact of TE.

This retrospective study included patients with a histological diagnosis of exocrine pancreatic cancer who were treated at the Department of Clinical Oncology of Prince of Wales Hospital in Hong Kong between 2010 and 2015; eligible patients were identified by a review of electronic medical records. If histological findings were unavailable because of the clinician’s decision to omit biopsy evaluation, patients were identified using clinical diagnoses based on radiological findings and substantial elevation of the level of serum marker carbohydrate antigen 19-9 (CA 19-9) (ie, >500 IU/mL). Patients were excluded if they had an atypical clinical presentation (eg, normal CA 19-9 level) or histological findings of non-exocrine pancreatic malignancies, such as neuroendocrine tumour or metastatic disease.

The following data were extracted from each patient’s electronic and physical medical records: (1) demographics (sex and age); (2) World Health Organization (WHO) performance status score (0: able to perform normal activities without restriction; 1: ambulatory and able to perform light work with limitations on strenuous activities; 2: ambulatory [>50% of waking hours] and capable of self-care but unable to perform any work activities; 3: symptomatic and in a chair or bed for >50% of the day but not bedridden; 4: completely disabled [bedridden] and unable to perform any self-care); (3) disease stage (according to the seventh edition of the American Joint Committee on Cancer tumour-node-metastasis staging system); (4) site of disease (head, neck, body, or tail); (5) CA 19-9 level at diagnosis; and (6) initial treatment (surgery, chemoradiotherapy, chemotherapy, or supportive care). Any occurrences of TE (venous, arterial, or both) were recorded from the time of diagnosis until death or last follow-up; the site of thrombosis (lung, lower limb, multiple, or other) and type of anticoagulation treatment were also recorded. After data entry, all patient data were verified by two authors (LL Chan and KY Lam) under the supervision of the corresponding author (SL Chan). Each patient’s survival status was last updated on 31 October 2019.

Patient factors (eg, age, sex, WHO performance status, and initial treatment), tumour-related factors (eg, histological diagnosis status, CA 19-9 level at diagnosis, stage, and site) and TE-related factors (eg, type and site) were summarised as numbers and percentages for categorical variables, and as medians and interquartile ranges for continuous variables. The Wilcoxon rank-sum test and Chi squared test were used to identify variables associated with the development of TE. Variables that displayed statistical significance in univariate analysis were included in multivariable analysis. Age and sex were included in multivariable analysis as adjustment variables because they are known risk factors for the development of TE in patients with cancer, as well as standard clinical variables commonly included in such analyses.8 9 10 Kaplan-Meier survival analysis and Cox proportional hazards regression analysis were performed to evaluate the relationship between overall survival and TE. P values <0.05 were considered statistically significant. All analyses were performed with R version 3.5.1.11

In total, 365 patients (217 [59.5%] men and 148 [40.5%] women; median age, 65 years [interquartile range=57-72]) were included in the study; baseline characteristics are summarised in Table 1. Of these patients, 268 (73.4%) had WHO performance status score of 0 to 1, whereas 97 (26.6%) scored 2 to 4. Furthermore, 219 patients (60.0%) had a histologically confirmed diagnosis; the remaining 146 patients (40.0%) were diagnosed by radiological and serological modalities. In terms of tumour staging, 171 patients (46.8%) had stage I to III disease; 194 patients (53.2%) had stage IV disease. The tumour location was at the pancreatic head in 203 patients (55.6%) and other sites (neck, body, or tail) in 162 patients (44.4%). Initial treatment was surgery in 78 patients (21.4%), chemotherapy or chemoradiotherapy in 153 patients (41.9%), and supportive care in 134 patients (36.7%). Additional details are provided in Table 1.

Among the 54 patients (14.8%) who developed TE, 32 (59.3%) had venous TE, 18 (33.3%) had arterial TE, and four (7.4%) had both. Lower limbs were the most common sites of thrombosis, with 55.6% of all thromboembolic events. Furthermore, three patients (5.6%) had pulmonary embolism. These findings are summarised in Table 2.

In univariate analysis, non-head pancreatic cancer (P=0.01) and stage IV disease (P=0.01) were significantly associated with TE. Other factors such as age at diagnosis, sex, WHO performance status, elevated CA 19-9 level at diagnosis, and initial treatment were not significantly associated with TE (Table 1). Multivariable analysis showed that stage IV disease was a significant risk factor (odds ratio=1.08, 95% confidence interval [CI]=1.00-1.17; P=0.046) [Table 3]. Median overall survival times in patients with and without TE were 4.88 months and 7.80 months, respectively (Fig 1); the difference between groups was not statistically significant (hazard ratio =1.08, 95% CI=0.80-1.49; P=0.58). Among patients with TE, median overall survival was not affected by anticoagulation treatment (no anticoagulation=4.77 months vs anticoagulation=5.63 months, hazard ratio=0.72, 95% CI=0.40-1.29; P=0.27) [Fig 2].

In the present study, approximately 15% of patients with pancreatic cancer developed TE. Lower limbs were the most frequent sites of TE, and venous TE was the most common type. In univariate analysis, both the site (non-head) and stage (IV) of disease were significantly associated with TE; multivariable analysis revealed that stage IV disease was a significant risk factor for TE.

There is considerable evidence of an association between pancreatic cancer and TE. In the first case series describing the relationship between TE and cancer, the incidence of TE was 60% in patients with pancreatic cancer, whereas it was 15% to 30% among patients with other malignancies.12 Several pathological processes have been implicated in this association.13 First, pancreatic cancer is characterised by high expression levels of tissue factor, which triggers the extrinsic coagulation pathway leading to thrombin formation. Second, the release of tumour-associated microvesicles promotes hypercoagulability and activates platelet aggregation. Third, the establishment of neutrophil extracellular traps secondary to neutrophil activation generates a matrix for platelet and tumour-associated microvesicle adhesion, resulting in blood clot formation.

Thromboembolism incidence of around 15% in our cohort is similar to that reported in other studies of Asian populations5 6 7 but lower than that in most Western populations (Table 4).3 4 14 The figures ranged from 20% to 40% in Western populations and 8% to 18% in Asian populations. Consistent with the findings in other studies of Asian populations, we observed no difference in overall survival between patients with and without TE. However, the literature suggests that, in Western populations, overall survival is affected by TE (Table 4).

Taken together, these findings support the hypothesis that TE incidences and outcomes are influenced by genetic and environmental differences between Western and Asian populations. For example, genetic variants in the clotting cascade (eg, factor V Leiden and thrombin gene G20210A) reportedly increase the risk of TE.15 These variants are much more prevalent in Western populations than in Asian populations.16 The resulting relative hypercoagulability may be one of the main reasons for the higher background incidence of TE in Western populations than in Asian populations.17 Another factor that may contribute to the difference in TE incidence between the two populations is obesity, an established risk factor for TE that is more common in Western populations.18

With respect to TE and pancreatic cancer prognosis, survival appears to be inherently longer in Western populations than in Asian populations (Table 4). Considering the aggressive nature of pancreatic cancer, it is possible that patients with shorter survival (eg, patients in Asian populations) do not live long enough to benefit from treatment of TE, whereas patients with longer survival (eg, patients in Western populations) experience a survival benefit from treatment of TE. Indeed, in a recent systematic review regarding the treatment outcomes of FOLFIRINOX and gemcitabine plus nab-paclitaxel in patients with pancreatic cancer, Lee et al19 showed that, compared with Asian populations, Western populations experienced a greater survival benefit from FOLFIRINOX (ie, standard treatment for metastatic pancreatic cancer) but a smaller survival benefit from gemcitabine plus nab-paclitaxel (which was not available to our patients during the present study). Therefore, a reasonable assumption is that anticoagulation can prolong survival in Western populations among patients treated with FOLFIRINOX. Further studies are needed to determine whether any subgroup of Asian patients with pancreatic cancer can benefit from the treatment of TE.

In univariate analysis, both non-head pancreatic cancer and metastatic disease were associated with the development of TE. However, in multivariable analysis, the association with non-head pancreatic cancer disappeared; metastatic disease was the sole risk factor for TE. This is not surprising—non-head pancreatic cancer is often detected at a late stage because clinical symptoms (eg, biliary obstruction) do not occur until the tumour becomes quite large. Therefore, the association of TE with non-head pancreatic cancer is mainly related to the advanced stage of disease. This finding is also consistent with the results of previous studies in which non-head pancreatic cancer was frequently detected at a later stage of disease.3 20

In the present study, we found that metastatic disease was a risk factor for TE, which is consistent with the results of previous studies.20 21 22 23 The underlying pathophysiological mechanisms involve multiple factors. For example, an advanced stage of disease is often associated with a higher tumour burden and bulky metastases, which can compress blood vessels and inhibit blood flow. Higher tumour burden can also affect WHO performance status, resulting in decreased mobility and bedridden status.

During the present study, most of our patients received low-molecular-weight heparin (LMWH) as treatment for cancer-associated TE, based on the results of the 2003 CLOT (Comparison of Low-molecular-weight heparin versus Oral anticoagulant Therapy for the Prevention of Recurrent Venous Thromboembolism in Patients with Cancer) trial in which LMWH demonstrated superior efficacy in preventing recurrent TE compared with coumarins (eg, warfarin) while maintaining a similar risk of bleeding.24 Recent studies have shown that direct oral anticoagulants (DOACs) such as edoxaban25 and apixaban26 are non-inferior to LMWH as secondary prophylaxis for TE with similar safety profiles. Accordingly, both LMWH and DOACs are valid options for the treatment of TE in patients with pancreatic cancer. This approach is consistent with the latest National Comprehensive Cancer Network 2021 guidelines.27 Although LMWH and DOACs demonstrate similar efficacy in preventing recurrent TE, other factors to consider in drug selection include baseline renal function, patient preference, ease of administration, risk of bleeding (eg, by tumour infiltration into the upper gastrointestinal tract), and availability of antidotes that can reverse anticoagulation.

Considering the overall poor prognosis of pancreatic cancer and the lack of an overall survival benefit associated with anticoagulation treatment of TE, factors such as quality of life should be considered when deciding whether to initiate or discontinue anticoagulation treatment. It is important to have clear discussions with patients regarding the risks and benefits of anticoagulation, particularly during the management of aggressive malignancies such as pancreatic cancer, where the life expectancy is often only months or weeks. Anticoagulation treatment, such as LMWH, may cause subcutaneous injection–related discomfort and carries an increased risk of bleeding, but the therapeutic effects of anticoagulation may relieve symptoms of TE (eg, calf swelling and dyspnoea). In a retrospective study of 128 patients with cancer-associated venous TE, Napolitano et al28 analysed the effects of anticoagulation on quality of life using the EORTC-C30 questionnaire; they found that long-term LMWH was not associated with worse quality of life. However, patients approaching the end of life often prefer to minimise their medication intake.29 In our clinic, we tend not to administer anticoagulation treatment if a patient’s life expectancy is <3 months or whose WHO status score is 3 to 4. This approach is consistent with the patient populations in recent clinical trials comparing the efficacies of DOACs and LMWH in the treatment of cancer-associated TE; patients with poor WHO performance status and short life expectancy were excluded from those trials.25 26

This study had a few limitations. First, its retrospective nature may have permitted bias related to missing data and the possibility of asymptomatic TE. However, TE tends to be symptomatic in patients with cancer; thus, it is unlikely that events were missed. Additionally, analyses of symptomatic TE are more relevant to real-world clinical practice. Second, the overall survival time of patients in the present study was worse than the survival times reported in randomised clinical trials of patients with metastatic pancreatic cancer.30 31 This discrepancy may have occurred because our study cohort was representative of real-world patients who more frequently have reduced liver function and worse WHO performance status. It may also be related to the absence of more effective chemotherapy (eg, nab-paclitaxel) during the study period.

In conclusion, this study demonstrated that the incidence of TE was around 15% in Chinese patients with pancreatic cancer. Notably, the presence of TE was not associated with worse overall survival, and metastatic disease increased the risk of TE.

Concept or design: LL Chan, KY Lam, SL Chan.
Acquisition of data: All authors.
Analysis or interpretation of data: LL Chan, DCM Lam, KY Lam, SL Chan.
Drafting of the manuscript: LL Chan, SL Chan.
Critical revision of the manuscript for important intellectual content: All authors.

The authors have no conflicts of interest to disclose.

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

This study protocol was approved by the Joint Chinese University of Hong Kong–New Territories East Cluster Clinical Research Ethics Committee (Ref No.: 2016.730). Informed patient consent was waived by the Committee due to the retrospective nature of the research.

1. Metharom P, Falasca M, Berndt MC. The history of Armand Trousseau and cancer-associated thrombosis. Cancers (Basel) 2019;11:158. Crossref

2. Khorana AA. Cancer and thrombosis: implications of published guidelines for clinical practice. Ann Oncol 2009;20:1619-30. Crossref

3. Frere C, Bournet B, Gourgou S, et al. Incidence of venous thromboembolism in patients with newly diagnosed pancreatic cancer and factors associated with outcomes. Gastroenterology 2020;158:1346-58.e4. Crossref

4. Kruger S, Haas M, Burkl C, et al. Incidence, outcome and risk stratification tools for venous thromboembolism in advanced pancreatic cancer—a retrospective cohort study. Thromb Res 2017;157:9-15. Crossref

5. Suzuki T, Hori R, Takeuchi K, et al. Venous thromboembolism in Japanese patients with pancreatic cancer. Clin Appl Thromb Hemost 2021;27:10760296211051766. Crossref

6. Chen JS, Hung CY, Chang H, et al. Venous thromboembolism in Asian patients with pancreatic cancer following palliative chemotherapy: low incidence but a negative prognosticator for those with early onset. Cancers (Basel) 2018;10:501. Crossref

7. Yoon SY, Lee MY, Yun J, et al. The incidence of venous thromboembolism is not low in Korean patients with advanced pancreatic cancer [corrected]. Blood Res 2018;53:227-32. Crossref

8. Khorana AA, Francis CW, Culakova E, Kuderer NM, Lyman GH. Frequency, risk factors, and trends for venous thromboembolism among hospitalized cancer patients. Cancer 2007;110:2339-46. Crossref

9. Eichinger S. Cancer associated thrombosis: risk factors and outcomes. Thromb Res 2016;140 Suppl 1:S12-7. Crossref

10. Abdol Razak NB, Jones G, Bhandari M, Berndt MC, Metharom P. Cancer-associated thrombosis: an overview of mechanisms, risk factors, and treatment. Cancers (Basel) 2018;10:380. Crossref

11. R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing; 2018.

12. Sproul EE. Carcinoma and venous thrombosis: the frequency of association of carcinoma in the body or tail of the pancreas with multiple venous thrombosis. Am J Cancer. 1938;34:566-85.

13. Campello E, Ilich A, Simioni P, Key NS. The relationship between pancreatic cancer and hypercoagulability: a comprehensive review on epidemiological and biological issues. Br J Cancer 2019;121:359-71. Crossref

14. Ouaissi M, Frasconi C, Mege D, et al. Impact of venous thromboembolism on the natural history of pancreatic adenocarcinoma. Hepatobiliary Pancreat Dis Int 2015;14:436-42. Crossref

15. Blom JW, Doggen CJ, Osanto S, Rosendaal FR. Malignancies, prothrombotic mutations, and the risk of venous thrombosis. JAMA 2005;293:715-22. Crossref

16. Jun ZJ, Ping T, Lei Y, Li L, Ming SY, Jing W. Prevalence of factor V Leiden and prothrombin G20210A mutations in Chinese patients with deep venous thrombosis and pulmonary embolism. Clin Lab Haematol 2006;28:111-6. Crossref

17. Wang KL, Yap ES, Goto S, Zhang S, Siu CW, Chiang CE. The diagnosis and treatment of venous thromboembolism in Asian patients. Thromb J 2018;16:4. Crossref

18. Yang G, De Staercke C, Hooper WC. The effects of obesity on venous thromboembolism: a review. Open J Prev Med 2012;2:499-509. Crossref

19. Lee YS, Lee JC, Kim JH, Kim J, Hwang JH. Pharmacoethnicity of FOLFIRINOX versus gemcitabine plus nab-paclitaxel in metastatic pancreatic cancer: a systematic review and meta-analysis. Sci Rep 2021;11:20152. Crossref

20. Lee JC, Ro YS, Cho J, et al. Characteristics of venous thromboembolism in pancreatic adenocarcinoma in East Asian ethnics: a large population-based observational study. Medicine (Baltimore) 2016;95:e3472. Crossref

21. Dickmann B, Ahlbrecht J, Ay C, et al. Regional lymph node metastases are a strong risk factor for venous thromboembolism: results from the Vienna Cancer and Thrombosis Study. Haematologica 2013;98:1309-14. Crossref

22. Khorana AA, Kuderer NM, Culakova E, Lyman GH, Francis CW. Development and validation of a predictive model for chemotherapy-associated thrombosis. Blood 2008;111:4902-7. Crossref

23. Cronin-Fenton DP, Søndergaard F, Pedersen LA, et al. Hospitalisation for venous thromboembolism in cancer patients and the general population: a population-based cohort study in Denmark, 1997-2006. Br J Cancer 2010;103:947-53. Crossref

24. Lee AY, Levine MN, Baker RI, et al. Low-molecular-weight heparin versus a coumarin for the prevention of recurrent venous thromboembolism in patients with cancer. N Engl J Med 2003;349:146-53. Crossref

25. Raskob GE, van Es N, Verhamme P, et al. Edoxaban for the treatment of cancer-associated venous thromboembolism. N Engl J Med 2018;378:615-24. Crossref

26. Agnelli G, Becattini C, Meyer G, et al. Apixaban for the treatment of venous thromboembolism associated with cancer. N Engl J Med 2020;382:1599-607. Crossref

27. Streiff MB, Holmstrom B, Angelini D, et al. Cancer-associated venous thromboembolic disease, version 2.2021, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw 2021;19:1181-201. Crossref

28. Napolitano M, Mansueto MF, Raso S, Siragusa S. Quality of life in patients with cancer under prolonged anticoagulation for high-risk deep vein thrombosis: a long-term follow-up. Clin Appl Thromb Hemost 2020;26:1076029620918290. Crossref

29. Huisman BA, Geijteman EC, Arevalo JJ, et al. Use of antithrombotics at the end of life: an in-depth chart review study. BMC Palliat Care 2021;20:110. Crossref

30. Von Hoff DD, Ervin T, Arena FP, et al. Increased survival in pancreatic cancer with nab-paclitaxel plus gemcitabine. N Engl J Med 2013;369:1691-703. Crossref

31. Vaccaro V, Sperduti I, Milella M. FOLFIRINOX versus gemcitabine for metastatic pancreatic cancer. N Engl J Med 2011;365:768-9; author reply 769. Crossref