Am Fam Physician. 1998;58(5):1196
The pathogenic factors for thromboembolic disease in patients with cancer include hypercoagulability related to clotting activation by malignant cells, vessel-wall injury and venous stasis. Several studies suggest the need to look for occult cancer in patients who have deep venous thrombosis or pulmonary embolism. To examine the possible association between thromboembolic disease and occult cancer, Sørensen and colleagues performed a large, population-based study of the frequency of cancer in patients with deep venous thrombosis and pulmonary embolism.
The authors identified cases of deep venous thrombosis and pulmonary embolism reported to the Danish National Registry of Patients from 1977 to 1992. Patients were excluded if they were pregnant, had undergone a surgical procedure within six months of the diagnosis of thromboembolism or had preexisting cancer. A total of 11,305 patients with pulmonary embolism and 15,348 patients with deep venous thrombosis were identified. All of the patients were linked by a national registration number to the nationwide Cause of Death Registry and to the Cancer Registry.
The patients were divided into three groups according to age (younger than 60 years, 60 to 74 years and older than 74 years). Each group consisted of approximately similar proportions of men and women and contained approximately one third of the patients studied. Another cohort consisted of patients who had a recurrent thrombotic event.
An incidence ratio of 1.3 for all types of cancer was found in the cohort with deep venous thrombosis and in the cohort with pulmonary embolism. Among patients with deep venous thrombosis, there were 1,737 cases of cancer, but only 1,372 cases would have been expected based on the population size. In patients with pulmonary embolism, there were 730 cases of cancer, but only 556 cases would have been expected.
The cancer risk was highest in the first six months following the occurrence of thromboembolic disease. The risk was approximately three times the expected risk. However, the risk then declined to a constant level of slightly more than 1.0 at one year after the thrombotic event and for the remainder of the follow-up period.
In both groups during the first year of follow-up, the strongest association of thromboembolic disease and cancer occurred with cancers of the pancreas, ovary, liver and brain. No association was observed in either group with melanoma or with breast, bladder or rectal cancer. Interestingly, patients younger than 60 years had the highest estimated risk of all types of cancer in the first year. The standardized incidence ratio in this cohort was 3.6 during the first year, compared with 2.2 in the 60-to 74-year group and 1.8 in the group older than 74 years of age. After one year of follow-up, the slight but significant increase of cancer incidence was evenly distributed among all cancer types and sites. In a subgroup of 3,762 patients with more than one thromboembolic event, the risk of all types of cancer combined was 3.2 during the first year of follow-up but declined to 1.3 thereafter.
The authors conclude that the association between cancer and thromboembolic disease seems to be most significant during the first year after the thrombotic event. However, they do not believe that an extensive evaluation for occult cancer is justified in patients with deep venous thrombosis or pulmonary embolism, especially in light of the fact that early screening for most of the primary cancers found in this population has not been shown to improve prognosis. The authors advocate using only currently recommended standards for screening and basing further testing on the presence of signs or symptoms of cancer.