Pregnancy-related venous thromboembolism and risk of occult cancer

Cancer increases the risk of venous thromboembolism (VTE) through direct interaction between tumor cells and the hemostatic system and due to tumor pressure.¹  In nonpregnant patients, VTE and superficial venous thromboses are associated with an increased prevalence of occult cancer.^2-5  The association is particularly strong for cancers of the lungs, ovaries, pancreas, brain, and liver, as well as for non-Hodgkin lymphomas.²  It is unclear whether a VTE occurring during pregnancy or the postpartum period is a marker of occult cancer as well.^2-5  Given the trend of postponing childbirth⁶  and the increasing risk of most cancers with age, the number of women giving birth who are at risk of cancer is expected to increase.^7,8  Pregnancy itself is protective against gynecological cancers, most nongynecological cancers, and breast cancer over the long term.^7,9,10  VTE risk increases with advancing gestation and peaks in the days around delivery.^10-14  Obesity, hospitalization, postpartum bleeding, and cesarean section are risk factors for antepartum and postpartum VTE.^11,14 

In one previous Norwegian study based on data from 577 pregnant women with first-time VTE, Wik et al reported a long-term follow-up, age-adjusted hazard ratio of 2.6 (95% confidence interval [CI], 1.3-5.6) for cancer diagnoses compared with a control group of women without VTE who delivered at the same time as the matching cases. However, the cancer risk was not elevated when compared with female, age-specific cancer rates obtained from the Norwegian Cancer Registry (standardized incidence ratio [SIR], 1.0, 95% CI, 0.6-1.7)¹⁵  making the data difficult to interpret.

We conducted a large, nationwide study to examine whether VTE episodes (both incident and recurring VTE, including deep VTE and superficial venous thrombosis) during pregnancy and the postpartum period is associated with an increased risk of a subsequent malignancy diagnosis. The tax-supported Danish health care system ensures equal and income-independent free access to health care services to all residents. Those born in or immigrating to Denmark receive a unique Civil Personal Register number, enabling linkage among Danish registries at the individual level.¹⁶  This nationwide cohort study made use of data from the Danish National Patient Registry (DNPR),¹⁷  the Danish Cancer Registry (DCR),^18,19  the Medical Birth Registry (MBR),²⁰  and the Civil Registration System.¹⁶  The DNPR has recorded data on all admissions and discharges from Danish nonpsychiatric hospitals since 1977 and from emergency and specialist outpatient clinic visits since 1995.¹⁷  Each hospital discharge or outpatient visit is coded in the DNPR according to the International Classification of Diseases, Eighth Revision from 1977 until the end of 1993 and Tenth Revision thereafter. The DCR includes prospectively collected complete and valid data on all incident cases of primary cancer diagnosed in Denmark since 1943.^18,19  The DCR records information on tumor staging at diagnosis according to TNM (tumor, node, metastasis) and Ann Arbor classifications. Reporting to the DCR has been mandatory for all hospital departments from 1987 and for general practitioners from 2004 ensuring national completeness of the register.^18,19  The DCR data are linked to the National Pathology Register for ensuring high validity of the diagnoses; 89% of the malignancy diagnoses in the DCR are with histopathological verification. Furthermore, continuous manual and electronic quality control ensures the high quality of the DCR.¹⁹  The MBR contains prospectively collected data on all deliveries in Denmark since 1 January 1973.²⁰ 

We used the MBR to identify all cancer-free women delivering a child from 1 January 1978 to 30 November 2013. The postpartum period was defined as the first 12 weeks after the delivery date, divided into the early (first 6 weeks after delivery) and the late (6-12 weeks after delivery) postpartum period.¹¹  We linked MBR data with DNPR data to obtain information on VTE (including superficial thrombosis) and with DCR data to identify cancer diagnoses. We included all inpatient and outpatient hospital diagnoses recorded in the DNPR, with the date of discharge defined as the VTE diagnosis date. Cardiovascular diagnoses, including VTE diagnoses, in the DNPR have positive predictive values of 70% to 100%.^21-23  We obtained information from the DNPR on risk factors for VTE (obesity, smoking, surgery, fracture, trauma, cesarean section, and postpartum bleeding) and comorbid conditions known to increase the risk of cancer (diabetes, chronic rheumatological diseases, and inflammatory diseases).

We followed women from the VTE diagnosis date until a diagnosis of cancer recorded in the DCR, death or emigration documented in the Civil Registration System, or the end of the study period (30 November 2013), whichever came first. We categorized the end points as cancers overall and cancers at selected sites as well as VTE overall and superficial and deep venous thrombosis separately. We obtained data on tumor staging at the time of diagnosis for all cancer cases according to tumor, node, metastasis classification and Ann Arbor classification. We computed the proportion of cancers being localized, with regional or metastatic spread. Using indirect standardization, we computed the SIR for cancer as the ratio of the observed number of cases and the number of cancer cases expected in women from the general population of the same age and in the same calendar period, as drawn from the DCR.²⁴  We investigated the association of cancer within <6 months, 6 to 12 months, and 1 year after VTE diagnosis and stratified the results by age, obesity at the time of VTE diagnosis, and timing of the thrombotic event (antepartum or postpartum). We computed the number needed to screen for cancer occurring within 1 year after the VTE event as (1/[observed cancers/expected cancers]). All statistical analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC). The study was approved by the Danish Data Protection Agency (record number 1-16-02-1-08).

We followed 3934 women diagnosed with a VTE during pregnancy or the postpartum period for a median of 14.8 years (interquartile range, 6.7-24.2 years) (Table 1). In the cohort, 8.8% had a previous history of VTE. During follow-up, 250 women were diagnosed with cancer: 64 (25%) after a VTE occurrence during pregnancy, and 186 (75%) after a VTE occurrence during the postpartum period (Table 2). Table 2 shows the cancer risk estimates for superficial and deep venous thrombosis separately. At the time of cancer diagnosis, 147 (59%) of the cancer cases were localized tumors, 53 (21%) were tumors with regional spread, 19 (8%) were tumors with distant metastases, and 31 (12%) had missing data on tumor staging.

The number of cancer cases observed within 6 months after a VTE event was higher than expected (absolute risk, 0.10%; 95% CI, 0.04-0.25 and SIR, 1.77; 95% CI 0.48-4.53), especially after a postpartum VTE event (SIR, 2.86; 95% CI, 0.78-7.33). Cancer risk in women with VTE during pregnancy remained increased during the 6- to 12-month interval after VTE (SIR, 3.27; 95% CI, 0.67-9.56). This cancer risk was still moderately increased after 1 year (SIR, 1.23; 95% CI, 0.94-1.58) and returned to the expected levels during the remaining follow-up period (absolute risk, 20.13%; 95% CI, 16.55-23.98 and SIR, 1.01; 95% CI, 0.89-1.14). The number needed to screen for detecting the excess cancer cases within 1 year after the VTE was 1689.

The present study suggests that VTE occurring during pregnancy or the postpartum period may be a marker of an underlying occult cancer; the risk of having a cancer diagnosed within the first months after a VTE episode in relation to pregnancy seems increased. However, our data does not suggest any increased cancer risk in the years after a VTE event during pregnancy or the postpartum period.

Our results including a larger number of cases in women with both first-time and recurrent VTE events, with separate risk estimates within a short period after the VTE, support those in the smaller study conducted by Wik et al, although that study did not observe an excess cancer risk in the comparison with female age-adjusted cancer rates from the Cancer Registry of Norway.¹⁵  Our study was based on prospectively collected nationwide data with complete coverage. Still, data for the first year of follow-up were sparse, yielding imprecise risk estimates for site-specific cancers. We showed that the risk of having a diagnosis of malignancy in the first year after a VTE episode during pregnancy or the postpartum period was higher than expected in the general female population of same age and in same calendar years, although the possible excess risk was low in absolute numbers. Given the large number needed to screen, an aggressive search for a hidden cancer in a pregnant woman with VTE is therefore not warranted.

The Program for Clinical Research Infrastructure was supported by the Lundbeck Foundation and the Novo Nordisk Foundation provided salary support for H.T.S.

Contribution: H.T.S. and E.H.-P. contributed substantially to the study’s concept and design, interpretation of data, and revision of the intellectual content and provided final approval of the version to be published; A.T.H. contributed substantially to the study’s concept and design, analysis and interpretation of data, drafting of the paper, and critical review and revision of the intellectual content and provided final approval of the version to be published; K.V. performed the statistical analyses and contributed substantially to interpretation of the data, revision of the intellectual content, and final approval of the version to be published; and V.E. and P.P. contributed substantially to interpretation of the data and critical review and revision of the intellectual content and provided final approval of the version to be published.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Anette Tarp Hansen, Section for Hematology and Coagulation, Department of Clinical Biochemistry, Aalborg University Hospital, Hobrovej 18-22, 9000 Aalborg, Denmark; e-mail address: anette.tarp.hansen@dadlnet.dk.