Venous thromboembolism and cancer risk
J Thromb Thrombolysis
Venous thromboembolism and cancer risk
Per Sande´n 0 1 2
Peter J. Svensson 0 1 2
Anders Sja¨lander 0 1 2
0 Department for Coagulation Disorders, University of Lund , Malmo ̈ , Sweden
1 Department of Public Health and Clinical Medicine, Umea ̊ University , Sundsvall , Sweden
2 Department of Clinical Sciences, Karolinska Institutet, Danderyd Hospital , Stockholm , Sweden
Cancer increases the risk of venous thromboembolism (VTE) and about 20 % of all VTE are associated with cancer. VTE can also be used as a marker for occult cancer. The objective was to examine the correlation between VTE and cancer regarding predictors for a subsequent cancer diagnosis. Patients treated for VTE between January 1st 2006 and December 31th 2011 were extracted from the Swedish national quality register AuriculA and crossmatched with the Swedish National Patient Register. In total 7854 patients corresponding to 14284 treatments years were examined. Primary VTE was found in 6451 patients, with 3936 first and 2515 recurrent VTE. There were 1403 patients with secondary VTE. After a first or recurrent primary VTE the incidence of cancer diagnose was high being 9.4-10.0 % the first year compared to 2.7-2.5 % during the second year. Cancer in the digestive organs was the most common type of cancer among those with first primary VTE with 19.2 % of diagnoses. In multivariable analysis age was found to increase the risk of cancer diagnosis after both first and recurrent primary VTE HR 1.02 (CI 1.02-1.03) and HR 1.02 (CI 1.01-1.03). For a first primary VTE anemia HR 2.13 (CI 1.48-3.08) and male sex HR 1.38 (CI 1.09-1.76) increased the risk while hypertension HR 0.74 (0.57-0.96), dementia HR 0.30 (CI 0.10-0.95) and history of major bleeding HR 0.52 (CI
Venous thromboembolism; Warfarin; VTE
0.28–0.97) reduced the risk of a subsequent cancer
diagnosis. There is a substantial proportion of patients being
diagnosed with cancer the first year after a primary VTE,
anaemia and male sex confers an increased risk.
Cancer is known to increase the risk of venous
thromboembolism (VTE) and about 20 % of all VTE are
associated with cancer [1, 2]. Patients with cancer have a four
to seven times increased risk of developing VTE and
chemotherapy increases the risk further . Tumours can
express different procoagulant molecules and modify the
expression of tissue factor which is one explanation to the
increased risk of developing VTE among cancer patients
. Some tumours compress vessels altering blood flow or
create damages to the vessel wall through intravascular
growth increasing the risk for thrombosis.
VTE is also a marker for occult cancer , up to 10 %
of patients with a first VTE have a subsequent cancer
diagnosis within the first year [6, 7]. The risk of VTE differ
between different cancer types , where cancer in the
pancreas, kidney, stomach, ovary, brain and lymphomas
have shown the strongest correlation to venous thrombosis.
After a first VTE recurrence is relatively common and a
concurrent cancer increases that risk . Recurrent VTE
has also previously been shown to increase the risk for a
subsequent cancer diagnosis .
The aim of this study was to examine the correlation
between VTE and cancer regarding predictors for a
subsequent cancer diagnosis.
Materials and methods
This is a retrospective study using secondary databases.
AuriculA  was founded in 2006 and is a Swedish
national quality register for atrial fibrillation (AF) as well as
all patients on oral anticoagulation regardless of treatment
indication. The register now includes over 125.000 patients
and more than 7.000.000 INR values. Of all patients on oral
anticoagulation in AuriculA, 19 % are treated due to a VTE.
Approximately 50 % of all patients on warfarin in Sweden
are included in AuriculA, mainly in whole regions with no
apparent selection bias. Everything done with the patients in
the anticoagulation centres in everyday clinical practice is
recorded and transferred to the quality register automatically
once every 24 h, provided that the patient has not declined to
participate. AuriculA also provides a clinical decision tool,
aiding in the dosage of warfarin using a dosing algorithm.
The Swedish National Patient Register  contains
information about hospital admissions as well as visits in outpatient
clinics in Sweden for all patients with a Swedish personal
identity number. The register was launched in 1964, but
complete coverage began in 1987. Currently, more than 99 % of all
somatic and psychiatric hospital discharges are registered in the
NPR. Information in this register includes dates for admission
and discharge, ICD10 codes for primary and secondary
diagnoses as well as surgical procedures, age and sex category.
The national quality register AuriculA was matched with
the NPR at the Swedish national board of health and welfare
using personal identity numbers and a common database was
created which was anonymised. From this database patients
treated with warfarin due to VTE between January 1´st 2006
and December 31´th 2011 with available information about
background characteristics in the NPR were extracted.
In total 8831 patients with a first or recurrent VTE were
retrieved. 8112 remained in the study after exclusion of those
with a prior cancer diagnosis (n = 719). We analysed
secondary VTE defined as patients with diagnosis of fracture or
surgery within 6 months prior to warfarin initiation
(n = 1403) separately. Patients with primary VTE with
diagnosis between 90 days and 1 week before warfarin
treatment were excluded from analysis (n = 258).
Statistical analysis plan
A first VTE was defined as a first time diagnosis of a VTE
within 1 week before start of warfarin treatment. Recurrent
venous thrombosis was defined as presence of a previous
VTE diagnosis 90 days or more before start of warfarin
treatment due to the recurrent VTE. During the study
period, almost all patients with VTE were treated with
warfarin in Sweden, new oral anticoagulants were not
available for this indication by then.
ICD10 codes were used for characterising background
factors and new cancer diagnoses (Appendix).
Background characteristics were presented descriptively
and differences between subgroups were tested with t-tests and
X2-tests. Proportion diagnosed with cancer was described as
diagnoses per 100 observation years expressed as percent.
Correlation between background factors and cancer
diagnosis was analysed using multivariate cox regression analysis.
Covariates were selected based on biological plausibility
using ICD 10 codes defined in Appendix with the exemption
of covariates with low a probability of registration in NPR.
Data was analysed using SPSS Statistics (Version 22;
SPSS Inc., IBM Corporation, NY, USA), Confidence
intervals (CI) are 95 % p \ 0.05.
In total 3936 patients with a first, 2515 with a recurrent
primary venous thrombosis and 1403 patients with a VTE
secondary to surgery or fracture were examined, and followed
for 14,284 patient years. Those with a recurrent thrombosis
were older, 68,6 compared to 66,3 years (p = \0.001), and
had a higher proportion of previous diseases (Table 1).
The incidence of cancer diagnosis during the first year
after initiation of warfarin was 9.4 % per year for those
with a first VTE and 10.0 % for those with a recurrent VTE
(Table 2). During the second year the incidence of a new
cancer diagnose was 2.7 and 2.5 % for those with a first or
recurrent VTE, respectively.
After a secondary VTE, 2.5 % per year and 0.8 % per year
had a new cancer diagnosis during the first or second year.
Almost all patients in the study, 96 %, were treated for
venous thrombosis in the leg or lung.
The most common cancer type for those with a first VTE was
cancer in the digestive organs (19.2 %) and cancer in the male
genitals for those with recurrent VTE (23.8 %) (Table 3).
Among men cancer in the male genitals was the most common
type (35.8 %) of which 96 % were prostate cancer and among
women it was cancer in the digestive organs (19.4 %).
Correlation between a new cancer diagnosis the first
year after a VTE and background factors was analyzed
using multivariate Cox regression (Table 4). Factors that
increased the risk of a cancer diagnosis during the first year
for those with a first VTE were age HR 1.03 (1.02–1.03),
anemia HR 2.13 (1.48–3.08) and male sex HR 1.38
(1.09–1.76). Hypertension HR 0.74 (0.57–0.96), dementia HR
0.30 (0.10–0.95) and history of major bleeding HR 0.52
Table 1 Background
characteristics for patients at
initiation of warfarin treatment
presented as n (%)
Table 2 Cancer diagnoses
displayed as proportion (%) per
observed time with 95 % CI,
subdivided in primary VTE
(first or recurrent) and VTE
secondary to fracture or surgery
Table 3 Type of cancer
diagnosed during the first year
after a primary VTE (first or
recurrent). Results given as
numbers and proportion (%) of
Difference between first and recurrent VTE
* Significance with p \ 0.05
** Significance with p \ 0.001
Proportion diagnosed with cancer
(0.28–0.97) reduced the risk of a subsequent cancer
diagnosis. For those with recurrent VTE only age HR
1.02 (1.01–1.03) significantly increased the risk of a new
cancer diagnosis. For patients over 45 years old with a
first VTE, age HR 1.02 (1.01–1.03), anemia HR 2.18
(1.50–3.16) and male sex HR 1.37 (1.07–1.75) increased
the risk of a cancer diagnosis. Hypertension HR 0.73
(0.56–0.95), dementia HR 0.32 (0.10–1.00) and history
Table 4 Multiple cox regression of background characteristics and the occurrence of a new cancer diagnosis during the first year after a first or
recurrent primary VTE
of major bleeding HR 0.53 (0.29–0.99) reduced the risk
of a subsequent cancer diagnosis.
After a first primary VTE a there is a high rate of cancer
diagnosis with 9.4 % per year during the first year and
additionally 2.7 % during the second year. For recurrent
primary VTE the rate diagnosed with cancer is similar with
10.0 % during the first year, this surpasses the 2.5 % per
year after a VTE secondary to fracture or surgery.
These results are well in line with those previously
published where Sørenssen et al.  found an increased
standardized incidence ratio of 2.2–3.2 for a cancer
diagnosis the first year after a first or recurrent primary VTE,
Prandoni et al. a frequency of 7.6–17.1 %  and Sun
et al.  showing a clearly higher risk of cancer diagnosis
after unprovoked VTE compared to a provoked VTE.
Except for rising age, which is a predictor for cancer
diagnosis both after a first and recurrent VTE, we found anaemia to
be a risk factor for a new cancer diagnosis after a first primary
VTE in accordance to results by Trujillo-Santo et al. .
Anaemia has also been shown to increase the risk of VTE after
cancer diagnosis  further supporting this connection.
However, anaemia was not identified to be a risk factor for
those with recurrent primary thrombosis, which to our
knowledge has not been studied before. Those with a previous
VTE have had treatment with anticoagulants where potential
bleeding sources like tumours in the gastrointestinal tract
might already have been discovered, rendering a new cancer
diagnosis emerging after a recurrent VTE less likely. The same
was true for patients with a history of major bleeding, where the
risk of a new cancer diagnosis was reduced after a first but not
after recurrent VTE, possibly due to the same mechanism as
proposed above for anaemia. Major bleeding could cause
immobilization and prolonged hospital stay, which increases
the risk of VTE by itself, which also might explain the lower
prevalence of new cancer diagnoses in this group.
Male sex increased the risk of obtaining a cancer
diagnosis after a first primary VTE and a trend towards an
increased risk for males with recurrent primary VTE. This
has not previously been shown. Ihaddadene et al. 
found no association between sex and occult malignancy,
Sørenssen et al.  found no particular gender difference
and Trujillo-Santos et al.  found an odds ratio of 1.3 for
male sex but did not reach significance. The explanation
for the increased risk of a cancer diagnosis associated with
male sex found here is not fully understood. One reason
could be that there are other causes of venous thrombosis in
women like pregnancy or hormonal treatment for which we
lack information in our material. However, limiting our
analysis to patients above 45 years of age where
hormonally induced VTE in women is unlikely, male sex is still a
strong risk factor for a subsequent cancer diagnose. One
recent study  found an increased risk for first time VTE
among men compared to females if female reproductive
factors are taken into account. Here they discussed
differences in genetic factors as one possible explanation. The
increased risk of new cancer diagnosis among males
identified in our study could be an alternative explanation
for part of the higher risk of VTE associated with male sex.
Hypertension reduced the risk of a cancer diagnosis after
a first VTE and there was a trend towards a reduced risk
after a myocardial infarction in both first and recurrent
VTE. Interestingly, Lind et al.  have published an
increased prevalence of myocardial infarction after VTE,
showing a link between cardiovascular disease and VTE
risk. With cardiovascular disease as an alternate cause of
VTE concurrent cancer might less often be the cause of
VTE among these patients.
Dementia was shown to reduce the risk of a cancer
diagnosis for a first but not for recurrent VTE. This may be
due to a reluctance of physicians to search for malignancies
in patients with severe dementia.
The most common type of cancer diagnosed after a
recurrent VTE was in the male genitals with 23.8 %. Of
these 96 % were prostate cancers. Previous studies have
shown a stronger correlation between prostate cancer and
venous thrombosis compared to breast cancer , which
could contribute to the risk increase for males.
With the high rate of cancer among VTE patients
limited screening for occult cancer have been advised,
including medical history, physical exam, basic bloodwork
and chest radiograph. More extensive screening programs
including CT scans or PET scans (fludeoxyglucose positron
emission tomography) have not been found to be beneficial
in identifying more occult cancers or reducing mortality
[20–22] and is not advised as routine for all VTE patients.
Using registers make the results dependant on the validity
of data in the registers. The validity of the NPR has
previously been shown to be good. In somatic care the register
lacks information in only 0.5–0.9 % of hospital admissions
. However, information about diagnoses from the
primary health care is not included in the NPR, and therefore
some concomitant diseases may have been underestimated.
Cancer however is almost exclusively diagnosed and at
least initially managed in specialised healthcare.
The inclusion of gender specific cancer complicates the
analysis and may bias the results. However being a large
proportion of all cancer the exclusion of gender specific
cancer may also have created bias.
We also lack information about some factors that may
influence both the cancer risk and risk for venous
thrombosis like smoking, obesity and medical treatment. These
could all be factors confounding the results. On the other
hand, we here present a large real-life cohort of 7854 VTE
patients corresponding to 14,284 observation years.
There is a substantial proportion of patients being
diagnosed with cancer the first year after a primary VTE,
anaemia and male sex confers an increased risk. Screening
for cancer after a VTE has previously not been advised, but
physicians caring for patients with VTE should be aware of
symptoms and findings indicating occult cancer, especially
among men and elderly without cardiovascular morbidity.
Author contribution AS and PJS designed the study. PS extracted
and analysed the data. PS and AS drafted the manuscript. All authors
critically reviewed the manuscript, contributed to its revision, and
approved the final version submitted.
Compliance with ethical standards
Conflicts of interest None declared.
Ethical approval This study was approved by the regional ethical
review board in Umea˚, Sweden (EPN nr 2011-349-31 M and
2014-191-32 M) and conformed to the declaration of Helsinki.
Open Access This article is distributed under the terms of the
Creative Commons Attribution 4.0 International License (http://crea
tivecommons.org/licenses/by/4.0/), which permits unrestricted use,
distribution, and reproduction in any medium, provided you give
appropriate credit to the original author(s) and the source, provide a
link to the Creative Commons license, and indicate if changes were
1. Heit JA , O'Fallon WM , Petterson TM , Lohse CM , Silverstein MD , Mohr DN , Melton LJ 3rd ( 2002 ) Relative impact of risk factors for deep vein thrombosis and pulmonary embolism: a population-based study . Arch Intern Med 162 ( 11 ): 1245 - 1248
2. Caine GJ , Stonelake PS , Lip GY , Kehoe ST ( 2002 ) The hypercoagulable state of malignancy: pathogenesis and current debate . Neoplasia 4 ( 6 ): 465 - 473 . doi:10.1038/sj.neo. 7900263
3. Khorana AA , Connolly GC ( 2009 ) Assessing risk of venous thromboembolism in the patient with cancer . J Clin Oncol 27 ( 29 ): 4839 - 4847 . doi:10.1200/JCO. 2009 .22.3271
4. Langer F , Bokemeyer C ( 2012 ) Crosstalk between cancer and haemostasis. Implications for cancer biology and cancer-associated thrombosis with focus on tissue factor . Hamostaseologie 32 ( 2 ): 95 - 104 . doi:10.5482/ha- 1160
5. Sorensen HT , Svaerke C , Farkas DK , Christiansen CF , Pedersen L , Lash TL , Prandoni P , Baron JA ( 2012 ) Superficial and deep venous thrombosis, pulmonary embolism and subsequent risk of cancer . Eur J Cancer 48 ( 4 ): 586 - 593 . doi:10.1016/j.ejca. 2011 .10. 032
6. Prandoni P , Falanga A , Piccioli A ( 2005 ) Cancer and venous thromboembolism . Lancet Oncol 6 ( 6 ): 401 - 410 . doi:10.1016/ S1470-2045(05)70207-2
7. Otten HM , Prins MH ( 2001 ) Venous thromboembolism and occult malignancy . Thromb Res 102 ( 6 ): V187 - V194
8. Blom JW , Doggen CJ , Osanto S , Rosendaal FR ( 2005 ) Malignancies, prothrombotic mutations, and the risk of venous thrombosis . JAMA 293 ( 6 ): 715 - 722 . doi:10.1001/jama.293.6. 715
9. Levitan N , Dowlati A , Remick SC , Tahsildar HI , Sivinski LD , Beyth R , Rimm AA ( 1999 ) Rates of initial and recurrent thromboembolic disease among patients with malignancy versus those without malignancy. Risk analysis using medicare claims data . Medicine 78 ( 5 ): 285 - 291
10. Sorensen HT , Mellemkjaer L , Steffensen FH , Olsen JH , Nielsen GL ( 1998 ) The risk of a diagnosis of cancer after primary deep venous thrombosis or pulmonary embolism . N Engl J Med 338 ( 17 ): 1169 - 1173 . doi:10.1056/NEJM199804233381701
11. UCR UCRC ( 2015 ) Auricula Annual Report 2013 [Auricula a˚rsrapport 2013] . http://www.ucr.uu.se/auricula/index.php/arsrap porter. Accessed 26 Aug 2015
12. site TNBoHaWSw ( 2015 ) Swedish National Patient Register . http://www.socialstyrelsen.se/register/halsodataregister/patientre gistret/inenglish. Accessed 15 Nov 2015
13. Prandoni P , Lensing AW , Buller HR , Cogo A , Prins MH , Cattelan AM , Cuppini S , Noventa F , ten Cate JW ( 1992 ) Deep-vein thrombosis and the incidence of subsequent symptomatic cancer . N Engl J Med 327 ( 16 ): 1128 - 1133 . doi:10.1056/ NEJM199 210153271604
14. Sun LM , Chung WS , Lin CL , Liang JA , Kao CH ( 2016 ) Unprovoked venous thromboembolism and subsequent cancer risk: a population-based cohort study . J Thromb Haemost 14 ( 3 ): 495 - 503 . doi:10.1111/jth.13251
15. Trujillo-Santos J , Prandoni P , Rivron-Guillot K , Roman P , Sanchez R , Tiberio G , Monreal M , Investigators R ( 2008 ) Clinical outcome in patients with venous thromboembolism and hidden cancer: findings from the RIETE registry . J Thromb Haemost 6 ( 2 ): 251 - 255 . doi:10.1111/j.1538- 7836 . 2007 .02837.x
16. Khorana AA , Francis CW , Culakova E , Kuderer NM , Lyman GH ( 2007 ) Frequency, risk factors, and trends for venous thromboembolism among hospitalized cancer patients . Cancer 110 ( 10 ): 2339 - 2346 . doi:10.1002/cncr.23062
17. Ihaddadene R , Corsi DJ , Lazo-Langner A , Shivakumar S , Zarychanski R , Tagalakis V , Solymoss S , Routhier N , Douketis J , Le Gal G , Carrier M ( 2016 ) Risk factors predictive of occult cancer detection in patients with unprovoked venous thromboembolism . Blood 127 ( 16 ): 2035 - 2037 . doi:10.1182/blood2015- 11 - 682963
18. Roach RE , Lijfering WM , Rosendaal FR , Cannegieter SC , le Cessie S ( 2014 ) Sex difference in risk of second but not of first venous thrombosis: paradox explained . Circulation 129 ( 1 ): 51 - 56 . doi:10.1161/CIRCULATIONAHA.113.004768
19. Lind C , Flinterman LE , Enga KF , Severinsen MT , Kristensen SR , Braekkan SK , Mathiesen EB , Njolstad I , Cannegieter SC , Overvad K , Hansen JB ( 2014 ) Impact of incident venous thromboembolism on risk of arterial thrombotic diseases . Circulation 129 ( 8 ): 855 - 863 . doi:10.1161/CIRCULATIONAHA.113.004168
20. Carrier M , Lazo-Langner A , Shivakumar S , Tagalakis V , Zarychanski R , Solymoss S , Routhier N , Douketis J , Danovitch K , Lee AY , Le Gal G , Wells PS , Corsi DJ , Ramsay T , Coyle D , Chagnon I , Kassam Z , Tao H , Rodger MA , Investigators S ( 2015 ) Screening for occult cancer in unprovoked venous thromboembolism . N Engl J Med 373 ( 8 ): 697 - 704 . doi:10.1056/NEJM oa1506623
21. Robertson L , Yeoh SE , Stansby G , Agarwal R ( 2015 ) Effect of testing for cancer on cancer- and venous thromboembolism (VTE)-related mortality and morbidity in patients with unprovoked VTE . Cochrane Database Syst Rev . doi:10.1002/ 14651858.CD010837.pub2
22. Gheshmy A , Carrier M ( 2016 ) Venous thromboembolism and occult cancer: impact on clinical practice . Thromb Res 140 (Suppl 1): S8 - S11 . doi:10.1016/S0049-3848(16)30091- 3
23. Ludvigsson JF , Andersson E , Ekbom A , Feychting M , Kim JL , Reuterwall C , Heurgren M , Olausson PO ( 2011 ) External review and validation of the Swedish national inpatient register . BMC Public Health 11 :450. doi:10.1186/ 1471 - 2458 - 11 - 450