Surgical and pathologic complications associated with peritoneovenous shunts in management of malignant ascites

Forty-three peritoneovenous shunts have been inserted to palliate malignant ascites in 33 patients. Ascites was controlled for a time in every patient, but 18 shunts eventually blocked. Further shunt revision successfully controlled ascites until death in five of these patients and for prolonged periods in another five. The authors observed a marked difference between the performances of the two available shunts, but emphasize that the two groups of patients were not selected at random and therefore may not be comparable. Twelve postmortem examinations have been performed in the 33 patients to ascertain causes of shunt malfunction and to identify possible evidence of abnormal or accelerated tumor spread. The postmortem findings highlight great variability in the capacity of iatrogenically introduced showers of tumor cells to seed. There was a spectrum of tumor growth in the lung from a complete absence of tumor cells through dormant tumor clumps to developing metastases. The authors found no evidence either clinically or at autopsy, that the procedure had adversely affected the prognosis, except in one patient who died from pulmonary edema immediately after the operation.