Malignant Ascites

Malignant ascites results when there is an imbalance in the secretion of proteins and cells into the peritoneal cavity and absorption of fluids via the lymphatic system. Greater capillary permeability as a result of the release of cytokines by malignant cells increases the protein concentration in the peritoneal fluid. Recently, several studies have demonstrated higher levels of VEGF, a cytokine known to cause capillary leak, in the sera and effusions of patients with malignancies.

Signs and Symptoms

Patients with malignant ascites usually present with anorexia, nausea, respiratory compromise, and immobility. Complaints of abdominal bloating, heaviness, and ill-fitting clothes are common. Weight gain despite muscle wasting is a prominent sign.

Diagnosis

A malignant etiology accounts for only 10% of all cases of ascites. Nonmalignant diseases causing ascites include liver failure, congestive heart failure, and occlusion of the inferior vena cava or hepatic vein. About one-third of all patients with malignancies will develop ascites. Malignant ascites has been described with many tumor types but is most commonly seen with gynecologic neoplasms (~50%), gastrointestinal malignancies (20% to 25%), and breast cancer (10% to 18%). In 15% to 30% of patients, the ascites is associated with diffuse carcinomatosis of the peritoneal cavity.

Physical examination

Physical examination does not distinguish whether ascites is due to malignant or benign conditions. Patients may have abdominal fullness with fluid wave, anterior distribution of the normal abdominal tympany, and pedal edema. Occasionally, the hepatic metastases or tumor nodules studding the peritoneal surface can be palpated through the abdominal wall, which has been altered by ascitic distention.

Radiologic studies

Radiographs. Ascites can be inferred from plain radiographs of the abdomen. Signs include a ground-glass pattern and centralization of the intestines and abdominal contents.

Ultrasonography. Abdominal ultrasonography has been shown to be the most sensitive, most specific method for detecting and quantifying ascites. It also permits delineation of areas of loculation.

Success at removing peritoneal fluid in patients was markedly better with ultrasonographic assistance, as demonstrated in a randomized trial by Nazeer et al. Ultrasonography improved the physicians’ ability to aspirate ascites from 67% (27 of 44 patients) to 95% (40 of 42 patients).

CT. Abdominal and pelvic CT is effective in detecting ascites. In addition, CT scans may demonstrate masses, mesenteric stranding, omental studding, and diffuse carcinomatosis. Use of IV and oral contrast agents is necessary, thus increasing the degree of invasiveness of this modality.

Paracentesis. After the diagnosis of peritoneal ascites has been made on the basis of the physical examination and imaging, paracentesis should be performed to characterize the fluid. The color and nature of the fluid often suggest the diagnosis. Malignant ascites can be bloody, opaque, chylous, or serous. Benign ascites is usually serous and clear.

Analysis of the fluid should include cell count, cytology, LDH level, proteins, and appropriate evaluation for infectious etiologies. In addition, the fluid can be sent for the determination of tumor markers, such as CEA, CA-125, p53, and human chorionic gonadotropin-β (hCG-β). The hCG-β level is frequently elevated in malignancy-related ascites and has been combined with cytology to yield an 89.5% efficiency in diagnosis. The use of DNA ploidy indices, Decker et al found, allowed a 98.5% sensitivity and a 100% sensitivity in the identification of malignant cells within ascitic fluid. The use of the telomerase assay, along with cytologic evaluation of the ascitic fluid contents, Li et al reported, has a 77% sensitivity in detecting malignant ascites.

Laparoscopy. Several studies have utilized minimally invasive laparoscopy as the diagnostic tool of choice. The fluid can be drained under direct visualization, the peritoneal cavity can be evaluated carefully, and any suspicious masses can be biopsied at the time of the laparoscopy.

Prognosis

The presence of ascites in a patient with malignancy often portends end-stage disease. The median survival after the diagnosis of malignant ascites ranges from 7 to 13 weeks. Patients with gynecologic and breast malignancies have a better overall prognosis than patients with gastrointestinal malignancies.

Treatment

Medical therapy

Traditionally, the first line of treatment is medical management. Medical therapies include repeated paracentesis, fluid restriction, diuretics, chemotherapy, and intraperitoneal sclerosis.

Repeated paracentesis

Repeated paracentesis, probably the most frequently employed treatment modality, provides significant symptomatic relief in the majority of cases. The procedure is minimally invasive and can be combined with abdominal ultrasonography to better localize fluid collections. High-volume paracentesis has been performed without inducing significant hemodynamic instability and with good patient tolerance.

After paracentesis, 78% of all patients reported relief of their symptoms, especially in the areas of abdominal bloating, anorexia, dyspnea, insomnia, and fatigue. In addition, overall quality of life improved after paracentesis.

Significant morbidity occurs with repeated taps and becomes more severe with each tap necessary to alleviate symptoms. Ascitic fluid contains a high concentration of proteins. Routine removal of ascites further depletes protein stores. The removal of large volumes of fluid also can result in electrolyte abnormalities and hypovolemia. In addition, complications can result from the procedure itself. They include hemorrhage, injury to intra-abdominal structures, peritonitis, and bowel obstruction. Contraindications to repeated paracentesis are viscous loculated fluid and hemorrhagic fluid.

With the placement of an intraperitoneal port, used also for the instillation of intraperitoneal chemotherapy, removal of ascitic fluid is possible without the need for repeated paracentesis. Other possible catheters for use in repeated paracentesis include PleurX and Tenckoff catheters (used for intraperitoneal dialysis). Placement of a semipermanent catheter minimizes the risk of injury to intra-abdominal structures. However, the benefits are tempered by increased infectious risks as well as the possibility of a nonfunctioning catheter requiring removal and replacement.

An analysis of the efficacy and cost-effectiveness of PleurX catheter drainage in treatment-resistant, recurrent malignant ascites was performed in the United Kingdom for the National Institute for Health and Care Excellence. The study, reported in 2012 by White and Carolan-Rees, concluded that use of the PleurX catheter in this population was clinically effective, with low complication rates. In addition, the use of PleurX was associated with improved quality of life and decreased overall cost when compared with serial paracentesis.

Diuretics and restriction of fluid and salt. Unlike ascites from benign causes such as cirrhosis and congestive heart failure, malignant ascites responds poorly to fluid restriction, decreased salt intake, and diuretic therapy. The most commonly used diuretics (in patients who may have some response to diuretic treatment) are spironolactone (Aldactone) and amiloride (Midamor). Patients with massive hepatic metastases are most likely to benefit from spironolactone.

The onset of action for spironolactone is delayed (3–4 days), whereas the effects of amiloride are seen after 24 hours. The most common complications associated with these diuretics are painful gynecomastia, renal tubular acidosis, and hyperkalemia.

Chemotherapy. Chemotherapy, both systemic and intraperitoneal, has had some success in the treatment of malignant ascites. The most commonly used agents are cisplatin and mitomycin. Intraperitoneal hyperthermic chemotherapy has been used with some efficacy in gastrointestinal malignancies to decrease recurrence of ascites as well as to prevent the formation of ascites in patients with peritoneal carcinomatosis.

Sclerotherapy. Sclerosing agents include bleomycin (60 mg/50 mL of normal saline) and talc (5 g/50 mL of normal saline). Responses are seen in ~30% of patients treated with these agents.

Theoretically, intraperitoneal chemotherapy and sclerosis obliterate the peritoneal space and prevent future fluid accumulation. If sclerosis is unsuccessful, it may produce loculations and make subsequent paracentesis difficult.

Other therapies. Approved for use to treat malignant ascites in the European Union since 2009, catumaxomab (Removab) is a trifunctional antibody specific for epithelial cell adhesion molecule (EPCAM) on tumor cells, CD3 antigen on T cells, and the Fc regions on accessory cells such as macrophages, dendritic cells, and natural killer cells. A phase II/III clinical trial demonstrated efficacy of the antibody, given as intraperitoneal infusions, in decreasing the number of and timing between paracentesis sessions needed to control the volume and symptoms of malignant ascites.

Catumaxomab is being studied in a randomized phase IIIb trial to improve on the dosing regimen of the antibody in patients with EPCAM+ epithelial carcinomas. The study is ongoing and results are pending.

Experimental models and early clinical trials have shown that an intraperitoneal bolus of tumor necrosis factor (45–350 μg/m²) given weekly may be effective in resolving malignant ascites. Other cytokines, including interferon-α, have had various degrees of success. A randomized, prospective trial definitively addressing the role of cytokines and other biologic treatments in the management of malignant ascites has yet to be completed. Intraperitoneal injection of antibodies directed at VEGF has shown promise in decreasing ascites in early-phase clinical trials, but further studies are needed.

Aflibercept (Zaltrap) has demonstrated the ability to reduce formation of ascites in preclinical models of epithelial ovarian cancer (EOC), as well as in patients with advanced EOC. Aflibercept, a potent angiogenesis inhibitor fusion protein, comprises portions of human VEGF receptor R1+R2 (Flt-1, KDR) extracellular domains fused to the Fc-portion of human IgG. Aflibercept binds VEGF-A and neutralizes all VEGF-A isoforms plus placental growth factor.

A randomized, phase II study by Gotlieb et al determined the efficacy of intravenous aflibercept in the management of symptomatic malignant ascites from ovarian cancer. A total of 26 patients received placebo and 29 patients were treated. Mean time to paracentesis was longer in the aflibercept arm (55.1 days) than in the placebo arm (23.3 days). Side effects included dypsnea, fatigue, dehydration, and bowel perforation.

Surgical techniques

Limited surgical options are available to treat patients who have refractory ascites after maximal medical management, demonstrate a significant decrease in quality of life as a result of ascites, and have a life expectancy of > 3 months.

Peritoneovenous shunts. These have been used since 1974 for the relief of ascites associated with benign conditions. In the 1980s, shunting was applied to the treatment of malignant ascites.

The LeVeen shunt contains a disc valve in a firm polypropylene casing, whereas the Denver shunt has a valve that lies within a fluid-filled, compressible silicone chamber. Both valves provide a connection between the peritoneal cavity and venous system that permits the free flow of fluid from the peritoneal cavity when a 2- to 4-cm water pressure gradient exists.

Success rates vary with shunting, depending on the nature of the ascites and the pathology of the primary tumor. Patients with ovarian cancer, for example, do very well, with palliation achieved in ≥ 50% of cases. However, ascites arising from gastrointestinal malignancies is associated with a poorer response rate (10% to 15%).

Candidates for shunt placement should be carefully selected. Cardiac and respiratory evaluations should be performed prior to the procedure. Shunt placement is contraindicated in the presence of the following:

• a moribund patient whose death is anticipated within weeks

• peritonitis

• major organ failure

• adhesive loculation

• thick, tenacious fluid.

Complications of shunting. Initial concerns about the use of a shunt in the treatment of malignant ascites centered on intravascular dissemination of tumor. In practice, there has been little difference in overall mortality in patients with and without shunts.

Disseminated intravascular coagulation. During the early experience with shunting, particularly in cirrhotic patients, symptomatic clinical disseminated intravascular coagulation (DIC) developed rapidly and was a major source of morbidity and mortality. However, overwhelming DIC occurs infrequently in the oncologic population.

The pathophysiology of DIC has been studied extensively and is thought to be multifactorial. The reinfusion of large volumes of ascitic fluid may cause a deficiency in endogenous circulating coagulation factors by dilution. Secondarily, a fibrinolytic state is initiated by the introduction of soluble collagen (contained within the ascitic fluid) into the bloodstream, leading to a DIC state. Infrequently, full-blown DIC is the result and requires ligation of the shunt or even shunt removal. Discarding 50% to 70% of the ascitic fluid before establishing the peritoneovenous connection may prevent this complication but may increase the risk of early failure due to a reduced initial flow rate.

Commonly, coagulation parameters are abnormal without signs or symptoms. In some institutions, these laboratory values are so consistently abnormal that they are used to monitor shunt patency. Abnormalities most commonly seen include decreased platelets and fibrinogen and elevated prothrombin time, partial thromboplastin time, and fibrin split products.

Other common complications include shunt occlusion (10%–20%), heart failure (6%), ascitic leak from the insertion site (4%), infection (< 5%), and perioperative death (10% to 20% when all operative candidates are included).

Shunt patency may be indirectly correlated with the presence of malignant cells. One study found that patients with positive cytology results had a 26-day shunt survival, as compared with 140 days in patients with negative cytology results. Other studies have failed to demonstrate a correlation between ascites with malignant cells and decreased survival.

Clearly, shunting is not a benign procedure, but in carefully selected patients who have not responded to other treatment modalities and who are experiencing symptoms from ascites, it may provide needed palliation. Because of the limited effectiveness of peritoneovenous shunts, patients should be carefully selected prior to shunt placement.

Radical peritonectomy. Other surgical procedures used to treat malignant ascites have been proposed. They include radical peritonectomy combined with intraperitoneal chemotherapy. This is an extensive operation with significant morbidity, although initial results appear to demonstrate that it decreases the production of ascites. To date, no randomized trial has demonstrated that radical peritonectomy increases efficacy or survival. However, there is an emerging body of literature supporting use of intraperitoneal chemotherapy in the management of malignant ascites. Although different chemotherapeutic agents have been studied for intraperitoneal use, Mitomycin-C is most often used.

Combined fluid complications. Combinations of pleural and pericardial effusion, ascites and pleural effusion, or even ascites combined with pleural and bilateral pleural effusion are not uncommon. Management is complex in these cases, and in the authors’ experience, survival is more limited.

Management of patients with ovarian cancer, ascites, and pleural effusion is particularly challenging. Two reports published in 2010 are illustrative.

Kim et al, from Seoul National University College of Medicine, Korea, studied 38 patients with ovarian cancer with pleural effusion on CT scan who had undergone thoracentesis before treatment. The investigators assessed the amount of ascites and pleural effusion, as well as lymph node enlargement and presence of pleural nodules or thickening. A total of 16 patients (42%) had a malignant pleural effusion. In patients with malignant pleural effusion, the volumes of pleural effusions were larger than in those with nonmalignant effusions. Pleural nodules were found more frequently in the malignant group (50% vs 0%). Supradiaphragmatic lymph node enlargement also was more frequent in the malignant group (25% vs 9%). The investigators concluded that the probability of malignant pleural effusion was correlated with the volume of pleural effusion, the presence of pleural nodules, and supradiaphragmatic lymph node enlargement on CT.

Diaz et al, from Memorial Sloan Kettering Cancer Center, performed VATS of pleural effusions in patients with advanced ovarian cancer. They studied 42 patients with a median age of 58 years and a median CEA-125 of 1,747 units/mL. Effusions were right-sided in 30 patients (71%). Macroscopic pleural disease was found in 29 (69%) of patients. Of 11 patients with negative cytology, macroscopic pleural disease was found in 4 (36%). Six of 18 patients had intrathoracic cytoreductive surgery after VATS. A total of 29 of 42 patients (69%) underwent attempted primary abdominal surgical debulking. VATS investigation prompted changes in management in 43% of cases.