The patient had a carotid dissection over five years prior to presentation and had a stroke and DVT. To protect her from pulmonary embolism, an IVC filter was placed, she was anticoagulated for a while. She rehabilitated and was doing well when several months prior to consultation she developed severe upper abdominal pain that occurred intermittently without triggers. She underwent a workup which included laboratory testing and endoscopy which were negative, and a CT scan showing the IVC filter had tilted and the legs had eroded through the vena cava to abut the duodenum and abdominal aorta. No other pathology could be found and after discussion operative resection was planned.
As mentioned in an earlier post, 3DVR image processing has become indispensable in planning not only aortic interventions but open surgery as well. An oblique incision was created, basically a high transplant incision, and we stayed in the retroperitoneum. The leg of the filter abutting the duodenum did not penetrate it and there was no leakage of bile. The end of the leg, of which there are two types on this filter, was in the vena cava, but the side had eroded out of the cava. Control of the IVC is always treacherous because of the fragile lumbar tributaries underneath the cava. Once heparinized and clamped, a longitudinal venotomy released the filter. The head was embedded in the cava wall and would not have been easily accessible with a snare.
The tine that was headed towards the aorta had to be removed from the filter to manipulate it out. No bleeding was noted from the aorta.
This is my fifth operative removal of an IVC filter that had eroded into adjacent organs. Recently I removed a filter which had eroded into the aorta -this required a pledgetted suture. In my first case, about 7 years ago, a filter placed in a teen, the victim of polytrauma due to an MVA, caused fevers and an upper GI hemorrhage which was diagnosed on upper endoscopy -somewhere I have a great picture of an IVC filter leg in the duodenum.
Which brings me to my last point. The guidelines for placing these filters has evolved and it is clear now that they are not as benign as once thought. They are not only associated with migration but also iliocaval thrombosis. Stenting across them can push the legs out into organs. Their migration into the retroperitoneum can cause an atypical abdominal pain syndrome -in this patient, the pain was immediately gone. If placed, plan should be made for removal if feasible, anticoagulation maintained if possible, and filter choice limited to those that have long track records.
The treatment of venous thromboembolism remains primarily pharmacologic. When the indication for the filter expires, the filter should be removed unless the risks of removal exceed the risks of leaving it in.