Peritoneovenous shunt occlusion. Etiology, diagnosis, therapy.

Electronic pressure testing of every LeVeen valve has practically eliminated mechanical malfunction as a cause of shunt failure. Nonetheless, failures do occur and in a series of 240 cases, early or late shunt failure occurred in 29 patients. Thirty-five additional cases of failures were either referred by other physicians over a period of 6 years or information and x-rays were accumulated by direct contact. Shunt failure becomes manifest by a sudden reaccumulation of ascites in patients with a previously functioning shunt. In immediate failure, the ascites may fail to disappear after surgery or reaccumulate if removed. Ideally, caval clotting should be first excluded by x-ray visualization of the superior vena prior to injection of the shunt with contrast agent. Shuntograms are done with fine-bore needles. The venous pressure is also measured. The entry of contrast into the vena cava without pooling indicates a patent venous limb. The contrast will empty from the venous tubing with forceful inspiration if the entire system is patent. The venous tube will not clear if the valve or peritoneal collecting tubes are blocked. Only the valve and collecting tube need then be replaced if contrast enters the cava but does not leave the venous tubing. Occluded valves must not be flushed to restore patency since inflammatory exudate and cellular debris are erroneously identified as "fibrin flecks." Histology and culture are mandatory. Immediate and early failure are often caused by malposition of the venous tubing. Malplacements can often be diagnosed simply by chest x-rays. Intraoperative injection of methylene blue into the venous tubing establishes a satisfactory washout prior to wound closure. Fresh clots in the vena cava can be dissolved by the slow injection of streptokinase into the venous tubing. Other patent veins are chosen for access. Patients having repeat surgery after clotting must be heparinized to prevent a similar recurrence. Flushing blood clots from the cava can be fatal.

Evaluation of peritoneovenous shunt patency with Tc-99m labeled microspheres.

The LeVeen peritoneovenous shunt (PVS) was investigated in 40 cirrhotic patients with refractory ascites. Five millicuries of Tc-99m-tagged human albumin microspheres (15-36 microns) were injected into the peritoneal cavity between the umbilicus and the left anterior superior iliac spine. The radiotracer was always detectable by scintigram in the lungs when the shunt was patent. In case of malfunction, by contrast, the radioactivity was either restricted to the venous tube or confined below the diaphragm for at least 4 hr. In the presence of complete obstruction, whereas the tube was not visualized after peritoneal injection, it was outlined by direct injection of 2 mCi of Tc-99m albumin microspheres into its subcutaneous tract, where it crossed the 12th rib, immediately above the valve. This technique sufficed to establish whether the site of obstruction was at the valve or in the tubing itself. In one patient, poor visualization of the tube and a delayed image of the lungs was caused by partial occlusion of the valve with fibrinoid debris. This radiotracer method proved simple, quick, and led to an immediate selective replacement when the shunt was not patent. Therefore, the use of this test is recommended for a definitive diagnosis, since there were neither false negatives nor false positives. No complications such as embolism or bacterial infection were encountered with Tc-99m human albumin microspheres, which are excellent tracers.