Bilateral deep vein thrombosis: Time to screen for occult cancer?

Among patients with unprovoked venous thromboembolism (VTE), the benefit of advanced cancer screening with abdomino-pelvic computed tomography (AP-CT) scan ¹ or 18-fluorodeoxyglucose positron emission tomography (18-FDG PET) scan ² was low. Thus, guidelines do not support the use of CT or PET in routine cancer screening. ³ However, the issue of cancer screening according to the clinical presentation of VTE is still debated, especially for bilateral deep vein thrombosis (DVT). This subgroup was outside the scope of the main studies that support current recommendations and previous reports have suggested that bilateral DVT was significantly associated with cancer. ⁴ In this study, we aimed to report the prevalence of occult cancer diagnosed with systematic CT scan in patients admitted for unprovoked VTE according to the presence of bilateral DVT.

We retrospectively reviewed the records of patients admitted to Rouen University Hospital, France, between 2007 and 2016, with a new diagnosis of symptomatic VTE (lower limb DVT, pulmonary embolism [PE], or both). Patients who underwent AP-CT (with thorax if CT pulmonary angiography [CTPA] was not initially performed) for cancer screening were included in the analysis. All patients admitted for PE and/or DVT were routinely investigated with bilateral whole leg colour-Doppler ultrasound (CDUS). A cancer was retrospectively adjudicated as occult if the patient with newly diagnosed cancer within 1 year after the thrombotic event was asymptomatic with normal physical examination and normal first-line test results (blood test, mammography, prostate-specific antigen [PSA]) at the time of VTE.

A total of 891 patients were identified with lower limb DVT or PE. Of these, 435 (48.8%) events were provoked by transient, persistent risk factors or did not have AP-CT. Of the remaining 456 patients, mean age was 67.7 ± 16.3 years and 216 (47.3%) were male. PE was identified in 315 (69.1%) patients and DVT in 303 (66.44%), including bilateral DVT in 47 (10.3%) cases. A cancer was diagnosed in 83 (18.4%, 95% CI: 15.1–22.2%) patients at the time of the thrombotic event or within the 1-year follow-up. Out of the 83 patients, 31 had a malignancy suspected on the AP-CT scan performed for the diagnosis of PE. Multivariate analysis demonstrated a significant association between malignancy and age over 60 years (odds ratio [OR] = 3.4, 95% CI: 1.5–7.6; p < 0.004), active smoking (OR = 3.0, 95% CI: 1.3–6.8; p = 0.01), and bilateral DVT (OR = 3.3, 95% CI: 1.6–6.5; p < 0.001). Incidence of cancer diagnosis over time was more frequent in patients with bilateral DVT at presentation (hazard rate [HR] = 2.8, 95% CI: 1.64–4.85; p < 0.001). This association was previously reported in the large MASTER cohort, with a significantly higher incidence of bilateral DVT in patients with cancer than in patients without cancer (8.5% vs 4.6%; p < 0.01). ⁴ Another study reported an association between cancer and bilateral DVT, with an OR = 2.5 (95% CI: 1.3–4.8%). ⁵ In addition, Le Gal et al. evidenced an association between cancer and bilateral DVT even when the contralateral DVT was asymptomatic. ⁶

In our study, among the 83 patients with a diagnosis of cancer, 18 (21.7%, 95% CI: 14.2–31.7%) were considered occult. Out of the 18 cases, CT scan failed to detect an occult cancer in only two patients. Finally, regarding the whole population, the frequency of occult cancer was 3.9%, and despite a high prevalence of cancer in our population, the rate of occult cancer diagnosed on systematic AP-CT scan was only 2.6% (95% CI: 1.5–4.5%) of cases, which is similar to the results reported in the SOME trial. ¹ However, when we consider patients with bilateral DVT, the rate of occult cancer diagnosed with the CT-based strategy was 12.8% (95% CI: 5.9–24.7%).

Our study has several limitations. This was a retrospective analysis of data from a single centre and the presence of selection and misclassification biases and a loss of information cannot be excluded. However, 90% of patients admitted for VTE were systematically assessed with a CT scan. The population of patients included in our study therefore seems representative of a population admitted for VTE.

Efforts are being made to personalise occult cancer screening. Recently, based on the SOME trial, patient age at unprovoked VTE diagnosis, prior provoked VTE, and active smoking were important predictors of occult cancer diagnosis among patients with VTE. ⁷ Our data suggest that systematic bilateral CDUS could help define the population at high risk of cancer that could benefit from an extensive screening strategy. Further prospective trials on occult cancer should include screening for bilateral DVT.