Management of non-ovarian cancer malignant ascites through indwelling catheter drainage

From March 2008 to March 2011, 345 cancer patients were admitted to the Integrated Therapy Department (also known as Palliative Care Unit, PCU) of Fudan University Shanghai Cancer Center (FUSCC). 145 patients were diagnosed with MA. The inclusion criteria consisted of the following: 1. Patients’ age was not less than 18 years at the time of enrollment. 2. Patients had histologic or cytologic proof of original malignancy and/or MA. 3. Patients had no chance to receive anti-cancer treatments for malignancy, including systematic chemotherapy, radiotherapy and etc. 4. Patients received intra-peritoneal catheterization and drainage during the interval to relieve symptom sufferings. 5. A coagulation screen, comprised of platelet count, prothrombin time (PT) and activated partial thromboplastin time (aPTT), was performed prior to the procedure to exclude contraindications. Of the 145 patients diagnosed with MA, 78 patients met the inclusion criteria and were included in this study. The international Classification of Disease-9 (ICD-9) coding system was used to code the medical records of all patients.

Medical records were reviewed by three attending physicians through the Union Medical System (UMS) of FUSCC to gather general characteristics, cancer-related information, symptoms, and ascites-related records. The UMS is the electronic medical records system containing all the medical information of patients in the FUSCC, including demographic and clinical data.

Informed consent for paracentesis and drainage was obtained before the procedure was performed. The procedure was carried out by patients’ attending physicians with more than 5 years’ clinical experience bedside with the patient supine. Patients were asked to urinate before the procedure. The CVC set used (Arrow Raulerson Syring, USA, REF ES-04306), which contains multiple side-holes in addition to an end-hole, was initially developed for deep vein indwelling and allowed for improved drainage and decreased risk of catheter blockage. The puncture site was located by ultrasonographic imaging prior to the procedure. 5–10 ml 2 % lidocaine was used for local anesthesia. An introducer needle was inserted into the peritoneal cavity, and a spring wire guide was placed in the peritoneal cavity through the introducer needle. The needle was then extracted. The indwelling catheter was placed in the peritoneal cavity following the spring wire guide after the abdominal wall was dilated. The spring wire guide was then extracted gradually with the catheter pushing forward into the peritoneal cavity. The catheter outlet was attached to a Heparin Cap (Heatlth Care Ltd) and a governor (part of the Infusion Sets with Precision Filters for Single Use, Wuhan W.E.O. Science&Technique Development Co, Ltd), then connected to a drainage bag (Fig. 1). Finally, the catheter was fixed to the abdominal wall by Statlock (Catheter Stabilization Device, Bard Access Systems, Inc). The governor for intravenous infusion was used to help control the rate of fluid removal and prevent sudden rapid fluid loss (Fig. 2). Especially for patients with tense ascites, the high tension in the peritoneal cavity can result in rapid drainage. We controlled the rate at about 300 ml per hour (85drops/min) for patients. 1000–1500 ml ascites were aspirated on the first day. The fluid of the first day was used for biochemical tests and fluid cytology. Drainage was continued or stopped according to the patients’ suffering and specific situation. Drainage was discontinued if disease progression or intolerable toxicities were confirmed or patients refused further treatments. In order to avoid hypovolemia and paracentesis-induced circulatory dysfunction (PICD), infusion of albumin and synthetic plasma expanders was used for all patients during the drainage interval. The diuretics were used and adjusted according to patients’ weight loss and electrolyte levels. The volume of extracted ascites and immediate and late complications including hypotension, haemorrhage, tube blockage, dislodgment and sepsis, were recorded in chart.

The symptoms related with ascites were assessed before paracentesis and then at the end-point of catheter removal. The most frequent ascites-related symptoms included fatigue, abdominal swelling, anorexia, abdominal pain, constipation, dyspnea, nausea and vomiting, insomnia, early satiety, and dizziness. The severity of the symptoms was assessed on a scale from 0 to 10, with zero denoting no symptom sufferings and ten representing symptoms as severe as the patients could imagine. The mean score of ascites-related symptoms was calculated and pre- and post-treatment scores were compared. Patient data was retrospectively reviewed through the charts.

Patient follow up took place 1 week after discharge and monthly for 6 months, then once every 3 months until death. The data were obtained from phone call records of the follow-up information of the Department. The overall survival (OS) was defined as the interval post-disease diagnosis till death (months). Since-paracentesis survival (SP-survival) was defined as the interval post-initial paracentesis till death (days).

Continuous variables were compared using the t-test for samples, while frequencies and rates were compared using the chi-square test. Symptom scores were presented as mean ± standard error. Pre-and post-treatment symptom scores were compared using a t-test. Values of p < 0.05 were considered to be statistically significant. The relationship between each variable and survival time was analyzed by the Kaplan-Meier method and Cox proportional hazards models. Statistical analysis was performed using Statistical Package for the Social Sciences (SPSS) software Version 16.0 for Windows (SPSS Inc., Chicago, USA).

During the study period, 178 patients were diagnosed with MA. Eighty two non-ovarian cancer patients accepted paracentesis and indwelling catheter drainage. Seventy eight patients were eligible for this study, with 43 male and 35 females. The mean age of those included was 58 years old (95 % CI: 55–62). The most frequent original cancers causing ascites in this study were hepatic (15, 19.2 %), pancreatic (12, 15.4 %) and colorectal cancer (12, 15.4 %). 59 (75.6 %) patients had visceral metastasis. The liver (59.0 %) was the most prevalent metastatic site, and 38 patients (48.7 %) had more than three sites of metastasis. 33 (42.3 %) patients were also diagnosed with pleural effusion. Thirty four patients had previously received chemotherapy, with a median of three previous courses of chemotherapy. The most frequent KPS was 40, indicating that patients were mainly bed bound, needing assistance with care.

Insertion of the catheter was technically successful in all 78 cases. 16 patients had two catheters inserted in separated locations. The peritoneal catheters remained in situ for a median of 13 days (range 5–18 days). The mean depth of the ascites before drainage scanned by B-type ultrasound was 93.6 mm (range: 35–143). The medial total serum albumin at admission was 30.1 g/L (range: 21–36). The median volumeof drainages per patient was 5750 ml and the mean volume was 8538 ml (range: 750–53300 ml). Diuretics were used in 70 patients concomitant with the paracentesis. Among these patients, 35 patients used spironolactone alone, ten used furosemide alone, and 25 patients used both.

No immediate complication such as perforation of a viscus or excessive bleeding was encountered during the placement procedure. Seven patients had continuous leakage of peritoneal fluid from the needle site during the drainage. The general characteristics and ascites-related characteristic are shown in Table 1.

The most common symptom before paracentesis was fatigue, reported in 73 patients (93.6 %). The second most common symptom was abdominal swelling in 72 patients (92.3 %), then anorexia in 59 (75.6 %), and abdominal pain in 36 (46.2 %) patients. For 62 patients, more than 3 symptoms were reported. All the ascites-related symptoms were alleviated after drainage compared with the baseline. There was a statistically significant alleviation in the mean score for abdominal swelling (p < 0.001), anorexia (p = 0.023) and constipation (p = 0.045). Details are shown in Table 2.

Follow-up was conducted until death or at the time of the study (December, 2013). Fourteen patients received indwelling catheter drainage twice. Seventy patients died within the follow-up period and 8 patients were still alive at the time of survival analysis. The median OS was 13 months (95 % CI: 11.2–14.8). Patients with ascites of gastrointestinal origin group had the worst OS (p = 0.001) (Fig. 3a, b). The median SP-Survival was 36 days (95 % CI: 29.9–43.0). Serum albumin concentration significantly affected survival with a serum albumin of greater than 30 g/L being associated with an improved SP-Survival (p = 0.02) (Fig. 3c, d). Multivariate Cox regression showed cancer type to be an independent prognostic factor for OS and serum albumin to be an independent prognostic factor for SP-Survival. Details are shown in Table 3.

Annonymised data is held at Department of Integrated Therapy, FUSCC. It is not possible to share the data due to regulations states in the ethical approval.

This study does not contain any individual person data.

This study protocol was approved by the ethical review board of FUSCC.