Occult cancer detection in patients with hemostatic disorder and venous thromboembolism

CT imaging for occult malignancy in patients with unprovoked venous thromboembolism (VTE) in a tertiary centre: is it worthwhile?

BACKGROUND

Investigating patients with unprovoked venous thromboembolism (uVTE) for occult malignancy can prove a diagnostic dilemma and imaging is often used extensively in this patient group.

AIMS

The primary objective of this study was to determine the incidence of malignancy on CT and other imaging over a 10-year period. A secondary objective was to evaluate the role of laboratory and other non-imaging tests performed.

METHODS

A retrospective key word search of our hospital's imaging system was performed to identify patients with unprovoked DVT/PE over the last 10 years. All imaging, histology, endoscopy, laboratory tests, and clinical follow-up over 2 years were analysed. Patients with provoked VTE were excluded.

RESULTS

150 patients had uVTE. 9 patients were diagnosed with occult malignancy by different investigations on index hospital admission (3 patients) or subsequently on clinical follow-up (6 patients). Mean age of patients was 62 years. 116 patients had CT body imaging. The incidence of malignancy diagnosed by initial CT imaging was 1.7% with a sensitivity of 22%, specificity 87%, and PPV 12.5%. Overall incidence of malignancy identified by imaging alone during the index hospital admission was 2%. Total incidence of malignancy including index admission and follow-up was 6%. Median time to cancer diagnosis was 12 months.

CONCLUSION

CT imaging had a low yield for diagnosing malignancy. Extensive imaging strategies increase cost and radiation exposure without improving mortality. Clinical follow-up, history taking, and physical examination guiding appropriate investigations remain the best tool for unmasking occult malignancy in patients with uVTE.

Cancer Screening in Patients with Unprovoked Thromboembolism: How to do it and Who Benefits?

Introduction: Unprovoked venous thromboembolism (uVTE) may be the first manifestation of cancer. The main objectives of this study were to compare limited screening (LS) and extended screening (ES) and to make a protocol to approach these patients.

Methods: This is a retrospective, unicentric observational study that included 245 patients with venous thromboembolism (VTE) admitted to an Internal Medicine Service for five years. The incidence of cancer and mortality during hospitalization, and at one and three years after admission were calculated in both LS and ES groups and compared.

Results: Of the 245 patients with VTE, 59 (24.1%) had uVTE: 35 (59.3%) were submitted to LS and 24 (40.7%) to ES, with 10 (4.1%) diagnosis of cancer. In the following three years, 10 more patients were diagnosed. There were no statistically significant differences in inpatient diagnosis rates (8.6% vs. 4.2%; p=0.51) or in-hospital mortality (2.9% vs. 4.2%; p=0.79) or mortality at one year (8.6% vs. 8.3%; p=0.97) and three years (20.0% vs. 20.8%; p = 0.94) between LS and ES groups respectively. The Computerized Registry of Patients with Venous Thromboembolism (RIETE) score was equal or superior to 3 in 69.5% (N=41) of the population with uVTE.

Discussion: The results of our study are consistent with the literature; there are no differences between screenings, as the difference in the number of diagnoses does not reflect on mortality.

Conclusion: There were no statistically significant differences between the two types of screening in this population. We suggest a protocol that includes the RIETE score to better select the patients who might benefit the most from an ES.