A clinical practice-based evaluation of the RIETE score in predicting occult cancer in patients with venous thromboembolism

The association between venous thromboembolism (VTE) and occult cancer is well established. However, the benefit of cancer screening in all VTE patients remains controversial. The Registro Informatizado Enfermedad TromboEmbólica (RIETE) score is a recently proposed risk score to identify VTE patients at high risk of occult cancer. We evaluated the performance of the RIETE score in a routine clinical setting comprising patients presenting with VTE between January 1 and December 31, 2014, at Danderyd University hospital. Out of 488 VTE patients, 47 (9.6%) patients received a new cancer diagnosis during a 24-month follow-up. After exclusion of patients with cancer diagnosed at baseline (≤ 10 days after VTE, n = 16), 472 patients were considered eligible for cancer screening. Among these 472 patients, 31 (6.6%) received a cancer diagnosis during follow-up. The cumulative incidence was high after both unprovoked (8.5%) and provoked (4.8%) VTE. The RIETE score was evaluated in 467 of these patients. Interestingly, a high RIETE score was not significantly associated with cancer diagnosis during follow-up (OR 1.78; 95% CI 0.85–3.63), which was mainly due to a poor performance in women (OR 1.04; 95% CI 0.30–2.83). In summary, we observed a relatively high incidence of occult cancer in both unprovoked and provoked VTE. The RIETE score performed poorly in identifying patients at high risk of occult cancer in our VTE population. Additional risk assessment models are warranted to identify VTE patients who would benefit from extensive cancer screening. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1007/s11239-019-01822-z) contains supplementary material, which is available to authorized users.