Regarding “Recurrent thromboembolism in patients with vena caval filters”☆☆☆★

Article Outline

• Copyright

Please see related article by Greenfield and Proctor on p 510-4 .

In this issue of the Journal of Vascular Surgery, Greenfield and Proctor present their analysis of a large prospectively collected database that includes consecutive patients treated with Greenfield vena cava filters. As the authors note, there are significant limitations to their data set: inconsistencies in the determination of hard end points, incomplete follow-up, and poor quantitative data on the degree and duration of anticoagulation. Although these limitations do constrain the conclusions that can be drawn from the data, their article provides a unique insight into the late results of filter placement in a large number of patients with extended follow-up.

The data show that inferior vena cava (IVC) filters can be placed with a low incidence of technical complications (< 4%) and minimal late morbidity. Their follow-up documents a high (> 99%) incidence of caval patency, and they can show no detrimental effect of IVC filter placement on the incidence of late deep venous thrombosis (DVT). These observations, particularly if they can be generalized, are reassuring at a time when the practice of prophylactic filter placement is increasing. I believe that these conclusions drawn by the authors can be substantiated by their data. The authors go on to suggest that anticoagulation plays a minor role in the incidence of late DVT. This assertion is contrary to the “accepted wisdom” concerning the management of DVT and, if accepted, would have significant clinical implications. As such it needs to be carefully scrutinized. After scrutiny, I do not believe that this conclusion can be drawn on the basis of the data presented. To do so, the authors would need data on the duration and adequacy of anticoagulation, the incidence of symptomatic and asymptomatic DVT, and the underlying cause of the thromboembolic events that occurred. These data are not presented in this article.

Despite the fact that the authors focused on patients with “thromboembolic disease” at the time of filter placement, their population is in many ways unique and distinct from that of patients with primary DVT. In the current study, all patients had an IVC filter placed and therefore, by definition, were inpatients. More than 50% of these patients had an acute intercurrent event (trauma, surgery, or malignancy) associated with filter placement, whereas less than one third presented with thromboembolic disease as the reason for treatment. IVC filters were placed in a large number of patients in the current study for prophylaxis or contraindications to anticoagulation. This clinical profile is quite different from that of patients presenting with primary DVT, who are often outpatients without another intercurrent illness or underlying malignancy. It is easy to imagine that the latter group would have a higher incidence of hypercoagulable conditions. This factor alone might lead to differences in the observed rate of late events when the current report is compared with reports from the literature. The unique characteristics of the current report preclude any conclusions on the effect of anticoagulation on the late incidence of recurrent DVT.

As the authors state in their report, placing an IVC filter cannot be expected to influence the course of DVT, which must be treated “based on an individual risk/benefit assessment.” Good clinical practice dictates that the etiology of DVT in each instance be determined and appropriately addressed. In patients with self-limited problems such as surgery or trauma, which preclude early anticoagulant therapy, anticoagulation may and often should be instituted during follow-up when the risk of hemorrhage has decreased. In those cases the duration of anticoagulation is not established, but may, in fact, be abbreviated. However, when patients present with more chronic conditions such as malignancy or hypercoagulable states, treatment with anticoagulants will need to be more prolonged and may be indefinite.

The authors make several important observations. IVC filters can be placed safely with good long-term patency and no apparent increase in the incidence of DVT. They also observe that although anticoagulation may be initially deferred, it can be initiated later in the patient's course, a fact that is often overlooked. One must not conclude from these data, however, that the issue of long-term treatment of DVT with anticoagulation is of little consequence. The authors have not performed a study designed to test this hypothesis. In clinical practice this is often the conclusion drawn. It is more important, I believe, to treat the whole patient rather than the disease and to be as vigilant in the appropriate treatment of DVT as one is in the prevention of pulmonary embolism.