Identification of Patients at High Risk of Cancer after Venous Thromboembolic Disease: A Prospective Multicenter Observational Study

Introduction:

Recent studies have shown lower rates of cancer (2 – 3%) following a single episode of venous thromboembolism (VTE) than previously described. The absence of any proven impact of early diagnosis has revived the debate about systematic screening for cancer in these patients. The aim of this study was to identify whether any sub-groups of patients were at excessive risk of cancer.

Methods:

Based on available data at baseline and during a 3 years longitudinal follow-up of the OPTIMEV cohort, a French multicenter prospective epidemiological study (N° NCT00670540), we performed a comparative prospective epidemiological study, whose main objective was to reassess the occurrence of cancer after an episode of VTE. We sought to determine whether among the VTE positive (VTE+) patients there was any sub-group of patients at greater risk of cancer and to define classification criteria that would allow them to be identified.

Results:

Among 1565 VTE+ patients, 2.7% and 4.7% developed cancer during the 1^stand 3 years of follow-up respectively.

The probability of cancer before 3 years was similar whether the DVT was proximal or distal. In contrast, the probability of cancer before 3 years depended on the number of sites affected by VTE, 4.2 % [3.1 - 5.6] for those who had presented with unilateral DVT without PE, and 8.5% [4.6 - 15.2] for those with bilateral DVT with or without PE. The probability also depended on whether the thrombosis was idiopathic 7.1% [5.1 - 9.9] vsa probability of 4.1% [3.1- 5.5] for those who had non-idiopathic DVT. The probability of cancer was also linked to whether or not the patient was on anticoagulant treatment at the time of VTE recurrence. This was 24.8 % [13.7 - 42.5] for those who had relapsed whilst on anticoagulant treatment, 13.1% [7.4 -22.5 ] for those who have relapsed when not on anticoagulants, compared with 4.0% [3.1 - 5.2] for those who did not relapse (p < 0.01). After adjustment, in VTE+ patients, there was a significantly higher adjusted HR for patients over 50 years of age (HR 11.1 [2.7 - 45.5] (p < 0.01), those with idiopathic VTE (HR 1.8 [1.1-2.8] (p = 0.02), bilateral DVT with or without PE (HR 1.9 [1.01 - 3.7] (p = 0.05), those who had a recurrence during follow-up whilst on anticoagulant treatment (HR 6.5 [3.2 - 13.4] (p<0.01) and for those with a recurrence when not on anticoagulant treatment (HR 3.5 [1.9 - 6.5] (p<0.01).

For VTE+ patients over 50 (n = 1169), we sought to classify these patients according to their risk of cancer by weighting risk factors according to adjusted HR from the multivariate model. Three points were given for recurrence while on anticoagulants, one point for a recurrence when not on anticoagulant treatment, one point for cases of idiopathic VTE and one point for cases of bilateral DVT. According to the scoring, the probability of developing cancer during follow-up was depicted table 1.

Discussion

The utility of systematically screening for cancer during an episode of VTE is debatable. The issue of the cost of screening in relation to its benefit in terms of reducing mortality and the problem of the physical and psycho-social consequences caused by such screening are also increasingly questioned. Based on the lower incidence of cancer observed in this prospective cohort, it might be more accurate to focus further studies regarding cancer screening on high risk population.

Conclusion:

Bilateral DVT, idiopathic but especially recurrent VTE appear to be good criteria for selecting patients over 50 years of age for whom the issue of cancer testing should be posed.