Peritoneovenous shunting for nongynecologic malignant ascites

BACKGROUND. The development of malignant ascites has been associated with a poor prognosis. Previous reports have documented high morbidity rates asso ciated with placement of palliative peritoneovenous shunts (PVS). Most stud y series have included gynecologic malignancies in their analysis, and wide variations in survival time have been reported. Reported data from nongyne cologic malignancies and identification of preoperative factors associated with improved outcome were the concerns of the current study, which attempt ed toidentify patients with malignant ascites who might have benefitted fro m PVS. METHODS. A retrospective chart review was performed and data including age, gender, weight, preoperative laboratory values, cytology on peritoneal flu id aspirates, and complications within 30 days of the operative procedure w ere obtained and recorded. Discharge date and follow-up status were obtaine d for all patients. Statistical analysis was done far categorical values by comparing survival times from date of procedure with follow-up times using the log rank test. Significance for numeric Values was determined with Cox regression analysis. Multivariate analysis using Cox regression was perfor med for those values found to be significant on univariate analysis. RESULTS. Fifty- five patients who had undergone PVS from 1980-1996 for asci tes on the Gastric and Mixed Tumor service at the Memorial Sloan-Kettering Cancer Center were identified. Two patients with benign disease and two pat ients with ovarian malignancies were excluded. The remaining 51 patients un derwent placement of 53 PVSs for palliation. Median survival time for the e ntire group was 52 days. Univariate analysis identified preoperative blood urea nitrogen (BUN), creatinine (Cr), BUN to Cr ratio, and diagnosis as sig nificant factors. Preoperative BUN emerged as an independent predictor of s urvival by multivariate analysis, and those patients who had a BUN value of less than or equal to 17 demonstrated a survival advantage over those with a BUN of > 17. The assessable palliation factors were hospital discharge ( 80% of patients) and weight loss after shunting (68% of patients lost > 1 k g). Ninety-six percent of patients (24 of 25) with a preoperative BUN of le ss than or equal to 17 were discharged. CONCLUSIONS. The development of nongynecologic malignant ascites is an end stage event for most patients. The placement of PVS for those patients with nongastrointestinal tumor etiologies, a BUN of < 17, a Cr of <less than or equal to> 1.1, and a BUN to Cr ratio of < 19 yielded the best results. In the current study, palliation was difficult to assess accurately, although most patients were discharged or lost > 1kg of weight after shunting. (C) 2 001 American Cancer Society.