ANGIOLOGÍA 2002; 54 (2): 76-83 N. DE LA FUENTE , ET AL 76 ORIGINAL Servicio de Angiología y Ci- rugía Vascular. Hospital del Mar. Barcelona, España. Correspondencia: Dr. Albert Clará. Servicio de Angiología y Cirugía Vascular. Hospital del Mar. Passeig Marítim, 25-29. E- 08003 Barcelona. E-mail: aclara@imas.imim.es © 2002, ANGIOLOGÍA Introducción La asociación entre trombosis venosa pro- funda (TVP) idiopática y cáncer fue suge- rida hace décadas [1]. No obstante, hasta Cribado de neoplasia oculta en pacientes afectos de trombosis venosa profunda primaria N. de la Fuente, A. Clará, A. Ysa, J. Molina, E. Bou, E. Ortiz, F. Vidal-Barraquer fechas recientes no ha podido demostrarse que, efectivamente, un diagnóstico de en- fermedad tromboembólica primaria se aso- cia a un riesgo de neoplasia oculta superior al esperable en población general [2]. Como DIAGNOSIS OF OCCULT NEOPLASIA IN PATIENTS WITH PRIMARY DEEP VEIN THROMBOSIS Summary. Objectives. To establish the incidence of occult neoplasia (NEO) in patients with primary deep vein thrombosis (DVT) during the first year and evaluate the most effective strategy for diagnosis. Patients and methods. Setting. General hospital, period 1995-1999; 266 consecutive patients with objective diagnosis of DVT, excluding 111 cases of secondary DVT and 16 primary DVT with <1 year follow-up (8 non-neoplastic deaths and 8 lost to follow-up); variables: age, sex, site of DVT, associated pulmonary thromboembolism, haemogram, liver function tests (LF), tumour markers (TM), chest X-ray (RxTx), abdominal CAT scan. Results.In the 139 primary DVT analyzed, 17 (12.2%) patients were diagnosed as having NEO in the first year (3 haematological, 13 adenocarcinomas and 1 carcinoma of the bladder). Sixteen cases (94%) were detected from the differential diagnosis and 9 given early treatment to achieve survival. The clinical data were not related to the NEO. The positive and negative predictive values of the investigations done were: haemogram (16/89%), LF (26/88%), TM (34/90%), RxTx (100/87%) and CAT scan (69/93%). A limited strategy for diagnosis (haemogram, liver function, RxTx and prostatic antigen in men) would give a result suggestive of NEO in 69% of the primary DVT, permitting diagnosis of 13/16 (81%) of the neoplasias detected. With this algorithm other investigations would not have been necessary (CAT, other TM) in 31% of the primary DVT, which made early diagnosis slower in 3/16 (19%) of the cases and early treatment delayed in 2/9 (22%) of the patients. Conclusions. The incidence of NEO in primary DVT makes it necessary to rule them out of the differential diagnosis. A limited strategy for diagnosis allows detection of most, but not all, treatable tumours at the expense of being more economical in the use of investigations in a limited percentage of cases. It would seem therefore preferable to make a fuller investigation of the potential differential diagnosis. [ANGIOLOGÍA 2002; 54: 76-83] Key words. Deep vein thrombosis. Diagnosis. Neoplasia. Screening. Survival. Thromboembolism.