Mortality Due to Malignancy in Patients With Idiopathic Venous Thromboembolism

Overview

Background Patients with an idiopathic venous thromboembolism (IVTE) appear to have a risk of approximately 10% for symptomatic malignancy within 3 years after the IVTE. It is not clear if extensive screening for malignant disease leads to survival benefit in patients with an IVTE. The SOMIT study learned that it is feasible to screen patients with an IVTE for malignancy and screening by means of a computer tomography (CT) of the chest and abdomen plus a mammography in women had the potential to be most cost-effective. The SOMIT study could not show a survival benefit due to the design of the study. Primary objective: cancer related mortality Methods: The Trousseau study has been designed as a multicenter, prospective concurrently controlled cohort study. Inclusion criteria: 1. Proven first symptomatic deep venous thromboembolic event; 2. Without: known risk factor for venous thromboembolism. Exclusion criteria: 1. Proven deep venous thromboembolic event in the medical history, age under 40 years; 2. Patients without signs of malignancy after routine investigations (medical history, physical examination, laboratory investigations and chest X-ray) were included. Depending on the standard care in the hospital of interest, one group of patients has been screened by means of CT-chest and abdomen plus mammography, the other group had no additional investigations. Follow-up was aimed to be 3 years in both groups (at 3, 6, 12, 24 and 36 months after the thromboembolic event). Data like mortality rate, morbidity due to screening procedures, additional investigations, number of cancer patients detected by the extensive screening, number of cancer patients three years after the IVTE, number and kind of investigations performed and information about cancer treatment and hospitalization was collected. If this information indicate a survival benefit these data enable us to perform a cost-effectiveness analysis. Endpoint: Mortality. Statistics: Based on the prevalence of occult malignancy in VTE patients, the nature and stage of malignancies, the expected mortality, the anticipated detection of cancers and the early treatment related decrease in mortality we needed, in order to detect a true difference of this size with a 80 percent power and a two-tailed certainty of five percent, 750 patients for each group. Therefore, a total of 1500 patients is required for this study.

Study design: Although a randomized design is ideal for most studies we preferred a multicenter, prospective concurrently controlled cohort study design for our study. This is based on the experiences with the SOMIT study, in which two of the members of the executive committee (MH Prins, JMMB Otten) were involved. The SOMIT study originally was supposed to have been conducted in eight countries. Medical ethical committees in most countries however considered it unethical to conduct this randomized study. Either because of the fact that the study contained a control arm, or because the screening arm (and thus the study itself) was considered to be unethical. Patients as well as physicians found it difficult to let fate decide whether or not a patient would be screened for cancer, even though it was not clear if screening was life-saving. Moreover, during the SOMIT study, physicians noticed that patients with IVTE had their cancer detected early if they were in the screening group. This made it even more difficult to withhold additional screening procedures in patients in the routine group. Many physicians themselves showed a strong preference for one of the arms of the study. Therefore they did not include as many patients as they could. With a prospective cohort design we expect to avoid these problems. Per hospital that participates in the Trousseau study the physicians in that hospital will treat the patients according to the local preference for screening or no screening. All hospitals are matched regarding their population as much as possible. Statistics: The prevalence of occult cancer at the time of the thrombotic episode in patients with IVTE can be estimated to be 10%. Based on the nature and stage of malignancies, it is expected that half of these patients with occult malignant disease will die during the 3 years of follow-up, resulting in a cancer-related mortality of 5%. In addition, in approximately half of the patients with malignant disease who survived for 3 years, residual or recurrent cancer will be present. Therefore, cancer-related mortality or residual or recurrent cancer will be present in 75% of the patients with occult malignant disease at presentation, i.e., in 7-8% of the patients of the study cohort. We anticipate that approximately 80% of the occult malignancies will be detected by extensive screening and that early treatment will result in a 50% to 75% reduction of the 3-year incidence of cancer-related mortality or residual or recurrent malignancy.