[Criteria for utilization and indications for use of permanent and short- and medium term temporary endocaval filters. Personal experience and review of the literature].

Screening and prophylaxis of the population at risk is the most correct approach to thromboembolic disease. Caval filters play a major role in the prevention of pulmonary embolism, but their use remains widely controversial, even if they are an alternative or additional method to medical therapy and not antithetical to it. July, 1990, to September, 1995, seventy-seven permanent vena cava filters (59 LGM, 13 Filcard and 5 titanium-Greenfield), 22 short-duration temporary vena cava filters (11 LGT, 6 Filcard, 3 Filcard-Emanuelli and 2 Lysofilter) and 10 mid-duration temporary vena cava filters (Tempofilter) were inserted in 109 patients (55 men and 54 women) aged 17-88 years. An inferior vena cavogram was always made before filter insertion to "map" the inferior vena cava and its tributaries (renal veins) and to assess caval size. This is a mandatory step to exclude or confirm the presence of endocaval thrombi. Sixty-eight filters were inserted percutaneously via a right internal transjugular approach, 40 via a common transfemoral approach (34 right and 6 left) using the same route as inferior cavography and digital pneumoangiography. One temporary filter was inserted via a right transhumeral approach. Sixty-three patients (57.8%) (44 with a permanent filter, 14 with a temporary filter and 5 with a Tempofilter) had medical diseases, 7 patients (6.4%) (4 with a permanent filter, one with a temporary filter and 2 with a Tempofilter) had surgical conditions; 10 patients (9.2%) (7 with a permanent filter and 3 with a Tempofilter) were surgical-orthopedic patients. Twenty-seven patients (24.8%) (22 with a permanent and 5 with a temporary filter) had cancer. Two patients (1.8%) with a temporary filter were breast feeding. All the patients had deep venous thrombosis of the lower limbs, including the inferior vena cava and the right renal vein in two patients. A proximal floating thrombus and a pulmonary thromboembolism was diagnosed in 41 (37.6%) and 34 (31.2%) cases, respectively. The temporary filters remained in situ 1-4 days and the Tempofilters 3-5 weeks. A pelvic hematoma developed (because of heparin?) after the insertion of a temporary filter and a Tempofilter branch detached and migrated into a lower branch of the left pulmonary artery were reported as immediate and short-term consequences. There were no early or late consequences for the patients. An axillary hemorrhage was observed in the site of previous surgery, during fibrinolysis with a Lysofilter, as well as the incomplete opening of a titanium-Greenfield filter and of two permanent LGM filters, partly trapped within endocaval thrombi. No further consequences to the patients due to permanent filters were observed, nor any case of symptomatic pulmonary thromboembolism, in patients with permanent or temporary filters. With reference to the relative literature and to their own experience, the Authors propose a detailed and rational synthesis of the diagnostic-instrumental approach protocol to thromboembolic disease.