Screening for occult cancer in patients with acute venous thromboembolism

Abstract

One in every five episodes, venous thromboembolism (VTE) occurs in patients with known cancer [1], but VTE may also be the first manifestation of an underlying (previously not known) malignancy [2,3]. This association is greatest within the first few months, thus suggesting these cancers are present at the time of diagnosis. A number of studies have been published on this issue, but there still exists controversy over the adequacy of performing some screening workup in patients with acute VTE in order to detect cancer in earlier phases. What a number of studies have already demonstrated is: (i) that a diagnosis of cancer is more likely to arise in patients without identified risk factors for thrombosis and apparently idiopathic VTE than in those in whom VTE occurs postoperatively or after immobilization [4,5]; (ii) that the variation in reported incidence reflects the different definitions for idiopathic and secondary cases, and variation in the intensity of cancer surveillance in each study; and (iii) that these cancers are more often in early stages of the disease [6,7]. However, there still remain some important unresolved questions: (i) is there any specific subgroup of patients with idiopathic VTE who may benefit more from an aggressive workup?; (ii) what is the most cost-effective workup to perform in these patients?; (iii) do early detection of cancer influence the survival of these patients? Firstly, as one in every two patients with VTE has idiopathic VTE, it would be interesting to know if hidden cancer is more frequent in any specific subgroup of patients. There is some evidence that patients with recurrent idiopathic VTE [5], bilateral DVT [8], upper-extremity DVT [9] or increased D-dimer levels at presentation [10] have an increased risk for hidden cancer. Besides, there is some controversy over the role of patient’s age, with some reports finding occult cancer most commonly in the younger, others in the elderly. A large study is warranted in which all these findings might be validated. Secondly, although intuitively a search for occult malignancy is attractive, there is no evidence that such an approach identifies all occult malignancies and ultimately affects the course of the disease. The mere knowledge of the presence of a substantial risk for occult cancer has led to the common practice of performing at least some extra tests in the attempt to identify occult malignancies in addition to the standard clinical evaluation. In a prospective observational study, we demonstrated that in patients with acute VTE a routine clinical evaluation results in the identification of approximately half of the patients who harbor an undiagnosed malignancy at the time of the thrombotic event [6]. A further limited diagnostic work-up, including abdominopelvic ultrasound and tumor markers, identified a malignancy in an additional number of patients. However, follow-up revealed that approximately half of the remaining patients with occult malignancies were missed. Up to now, the only randomized trial comparing extensive screening vs. no screening in patients with idiopathic VTE has demonstrated that most cancers are potentially diagnosable at presentation [7]. However, a study including ultrasound and computed tomography of the abdomen and pelvis, stool guaiac examination, gastroscopy, colonoscopy, sputum cytology, mammography, and measurement of tumor markers is unlikely to be cost-effective. In this issue of the Journal, Di Nisio et al. [11] provide some evidence that some screening strategies may be cost-effective, comparing the total costs of a number of screening strategies, and the expected Live Years Gained with each of them. This is a useful study that has the limits of the original study (mainly, the small number of patients and the unknown life expectancy in patients with concurrent cancer and VTE), as acknowledged by the authors. Finally, there is no evidence to date that routine cancer screening would improve the prognosis of these patients. Any benefit associated with an extensive strategy of investigation should be considered in conjunction with the potential for