Handbook of Cancer Chemotherapy (Lippincott Williams & Wilkins Handbook Series), 8th Ed.

30. Malignant Pleural, Peritoneal, and Pericardial Effusions and Meningeal Infiltrates

Rekha T. Chaudhary

Malignant pleural, peritoneal, and pericardial effusions and malignant meningeal infiltrates are uncommon early in the course of the malignancy. They occur more frequently with disseminated disease and often herald a poor prognosis. Although pleural and peritoneal effusions may initially have little adverse effect on quality of life, when progressive, they can result in incapacitating disability and death. Effusions can denote a poor prognosis; for example, the median survival after a diagnosis of a malignant pleural effusion is 4 months. It is therefore necessary for the clinician to have a high index of suspicion for these problems and to be prepared to take appropriate action and deliver palliative treatment promptly.


A. Causes

Malignant pleural effusions arise in association with malignant cells lining the pleura, exuded into the pleural space, or blocking veins or lymphatics. The most common malignancy associated with pleural effusions in women is carcinoma of the breast; in men, it is carcinoma of the lung. Other causes of malignant pleural effusions include lymphoma, mesothelioma, and carcinomas of the ovary, gastrointestinal tract, urinary tract, and uterus. Malignancy is not the only cause of effusions, even in patients with known neoplastic disease; therefore, it is important to attempt to exclude other possible causes such as congestive heart failure, infection, and pulmonary infarction.

B. Diagnosis

1. Clinical diagnosis. Effusions may be asymptomatic or may be suspected because of respiratory symptoms such as shortness of breath with exertion or at rest, orthopnea, paroxysmal nocturnal dyspnea, or occasionally chest pressure or cough. The patient may feel more comfortable when lying on one side when the effusion is unilateral. On physical examination, dullness to percussion, decreased tactile fremitus, diminished breath sounds, and egophony are typical signs over the area of the effusion.

2. A chest radiograph should be obtained to confirm the clinical impression. If fluid appears to be present, a lateral decubitus film must be obtained to help estimate the volume of the effusion and how free it is within the pleural space.

3. Diagnostic thoracentesis should be performed. Ultrasonographic guidance is helpful if loculation is present. Fluid should be obtained for bacterial, acid-fast, and fungal cultures, for cytologic examination, and for determining protein concentration (greater than 3 g/dL in most exudates), lactate dehydrogenase (LDH) level, specific gravity, and cell count. The cytologic examination is important, because if the results are positive, as in 50% to 70% of patients with malignant effusion, the diagnosis is established. Other parameters of the pleural fluid that may be helpful in establishing that the fluid is an exudate and not a transudate include a specific gravity of more than 1.015, protein concentration that is more than 0.5 times the serum protein concentration, LDH level more than 0.6 times the serum LDH level, and low glucose level. A cytologic examination of fluid from a newly discovered pleural effusion is wise, regardless of whether the patient is known to have malignancy, because for nearly half of all malignant effusions, this finding is the first sign of malignancy. Analyzing pleural fluid for carcinoembryonic antigen (CEA) may be helpful in some patients. Levels higher than 20 ng/mL are suggestive of adenocarcinoma, although they do not substitute for a tissue diagnosis in patients who have no history of malignancy. CEA elevations may be seen in adenocarcinomas from various primary sites including the breast, lung, and gastrointestinal tract. Elevated levels between 10 and 20 ng/mL may reflect malignancy or benign disorders such as pulmonary infection. The role of assessing other tumor markers on a routine basis has not been established. Likewise, the utility of monoclonal antibodies and gene rearrangement studies in patients with lymphomas to distinguish reactive mesothelial or lymphocytic cells from malignant cells has yet to be determined. The routine use of a “panel of tumor markers” is costly, time-consuming, and not recommended.

4. Pleural biopsy may be helpful in establishing the diagnosis in up to 20% of patients for whom the pleural fluid cytology results are negative.

5. Thoracotomy or pleuroscopy with direct biopsy may be done in patients who have negative cytology and pleural biopsy results but in whom there is still high suspicion of malignancy.

C. Treatment

As malignant pleural effusions are generally a sign of systemic rather than localized disease, the best therapy is treatment that effectively treats the malignancy systemically. Unfortunately, effective systemic treatment is often not possible, particularly when the malignancy is commonly refractory to systemic treatment (e.g., in non–small-cell carcinoma of the lung) or in patients who have previously been heavily treated and in whom systemic therapy is no longer effective. In these circumstances, locoregional therapy is required for palliation of the patient’s symptoms.

1. Drainage. Many malignant pleural effusions recur within 1 to 3 days after simple thoracentesis; about 97% recur within 1 month. Chest tube drainage (closed tube thoracotomy) allows the pleural surfaces to oppose each other and, if maintained for several days, may result in obliteration of the space and improvement in the effusion for several weeks to months. It does not appear to be as effective when used alone as when a cytotoxic or sclerosing agent is added, and therefore, one of these agents is commonly instilled into the space while the chest tube is in place. Repeated thoracentesis is an option in patients who reaccumulate slowly (greater than 1 month).

2. Cytotoxic and sclerosing agents or pleurodesis. The most widely used agents for intrapleural administration are bleomycin, doxycycline, and talc. Other agents, including fluorouracil, interferon-α, and methylprednisolone acetate, have been less commonly used. Randomized studies have suggested that bleomycin may be more effective than doxycycline (in part because doxycycline sometimes requires multiple dose administrations) and that talc is either equal to or slightly better than bleomycin in terms of recurrence. The agents vary in toxicity, ease of administration, and cost. Additionally, institutional experience often determines the agent utilized. Nevertheless, for optimal effectiveness, drainage of pleural fluid as completely as possible is required before instillation.

a. Method of administration. The drug to be used is diluted in 50 to 100 mL of saline and instilled through the thoracostomy tube into the chest cavity after the effusion has been drained for at least 24 hours and the rate of collection is less than 100 mL/24 h. Throughout the procedure, care must be taken to avoid any air leak. The thoracostomy tube is clamped, and the patient is successively repositioned on his or her front, back, and sides for 15-minute periods during the next 2 to 6 hours. The tube is then reconnected to gravity drainage or suction for at least 18 hours to ensure that the pleural surfaces remain opposed and to prevent the rapid accumulation of any fluid in reaction to the instillation. Some clinicians repeat the instillation daily for a total of 2 to 3 days. For most of the agents, this has no proven benefit. Exceptions include methylprednisolone acetate and doxycycline, which appear to be more effective with additional doses. If the drainage is less than 40 to 50 mL over the previous 12 hours, the tube may be removed and a chest radiograph obtained to be certain that pneumothorax has not occurred during removal of the tube. If the thoracostomy tube continues to drain more than 100 mL/24 h after the last instillation, it may be necessary to leave it in place for an additional 48 to 72 hours to ensure that a maximum amount of adhesion between the pleural surfaces has taken place. Because the use of sclerotic agents can be painful, it is prudent for the clinician to consider the use of scheduled narcotic analgesia, particularly during the initial 24 hours.

b. Recommended agents. Efficacy, side effects, cost, and institutional (operator) experience must be considered when choosing a sclerosing agent. Bleomycin, in one prospective study, was shown to be more effective than tetracycline. It is also more expensive per dose than the other agents. Talc is the least expensive, but this must be balanced against the costs of related procedures, including thoracoscopy and anesthesia. Talc is probably superior to bleomycin in terms of recurrence rate of effusions at 90 days and later.

(1) Bleomycin 1 mg/kg or 40 mg/m2 has relatively little myelo-suppressive effect and is highly effective.

(2) Talc 5 g is given typically as a powder (poudrage). It is highly effective but requires thoracoscopy and general anesthesia. Rarely, adult respiratory distress syndrome has been reported, primarily with doses greater than 10 g. If the patient is a high risk for general anesthesia, talc slurry may be administered at the bedside, though it is probably less effective than the thoracoscopic poudrage.

(3) Doxycycline 500 mg may cause pleuritic chest pain. An injection of 10 mL of 1% lidocaine (100 mg) through the chest tube may reduce this symptom.

c. Alternative agents

(1) Fluorouracil 2 to 3 g (total dose) may have a theoretical advantage in sensitive carcinomas, but whether that advantage has practical significance is not established. Pain is generally minimal. Occasional patients may experience a depressed white blood cell count, especially at the higher dose.

(2) Interferon-α 50 × 106 U typically causes influenza-like symptoms. Lower doses appear to be ineffective. Patients should be premedicated with acetaminophen 650 mg before and then 6 hours after interferon administration. Meperidine 25 mg intravenously by slow push may be given for rigors from interferon.

(3) Methylprednisolone acetate 80 to 160 mg appears to be well tolerated.

d. Responses. Chest tube drainage together with instillation of one of the agents discussed in Section I.C.2.b or c controls pleural effusions more than 75% of the time. The durations of response are often short, with a median between 3 and 6 months unless the patient’s systemic disease comes under adequate control. In that circumstance, the effusion may not recur for years or at least until the systemic disease once more emerges.

e. Side effects common to most agents include chest pain, fever, and occasional hypotension. These effects are usually not severe and may be controlled by standard symptomatic management. Fever after pleurodesis is usually not due to infection.

3. Indwelling pleural catheter placement is another option for patients who have recurrent pleural effusions. It involves placement of a soft, flexible, valved catheter connected to a drainage kit into the pleural space. Patients must be willing to care for the catheter on an outpatient basis but in contrast to a standard chest tube and pleurodesis, it allows the patient to be treated as an outpatient. It has similar efficacy to pleurodesis, but carries an added risk of infection owing to the indwelling catheter. Spontaneous pleurodesis may occur after 1 month or more of pleural catheter placement.

4. Thoracotomy and pleural stripping may be tried subsequently for effusions refractory to other medical treatment, when the prognosis is otherwise good.


A. Causes

Malignant peritoneal effusions usually occur in association with diffuse seeding of the peritoneal surface with small malignant deposits. The impairment of subphrenic lymphatic or portal venous flow may result in peritoneal effusions. Alternatively, it has been postulated that a “capillary leak” phenomenon mediated by tumor cells or immune effector cells could be a contributing factor. Carcinoma of the ovary is the most commonly associated malignancy in women, whereas in men, gastrointestinal carcinomas are most common. Other neoplasms that may cause peritoneal effusions include carcinoma of unknown primary, lymphoma, mesothelioma, and carcinomas of the uterus and breast. Liver metastasis by itself, unless it is far advanced, is not usually associated with symptomatic peritoneal effusions.

B. Diagnosis

1. Symptoms and signs. Patients may be completely symptom-free or have so much fluid that they have severe abdominal distention, abdominal pain, and respiratory distress. In the presence of peritoneal metastases, there may be abnormal bowel motility that at times resembles a paralytic ileus and may result in loss of appetite, early satiety, nausea, and vomiting. On examination, the lower abdomen and flanks bulge when the patient is supine. confirmatory signs include shifting dullness, a fluid wave, diminished bowel sounds, or the “puddle sign” (periumbilical dullness when the patient rests on knees and elbows).

2. Radiographic studies. Ascites may be suggested on a recumbent film of the abdomen, although radiographs are less sensitive than computed tomography (CT) or ultrasound in detecting fluid. CT is also helpful in defining whether there are enlarged retroperitoneal nodes, tumor masses in the abdomen or pelvis, or liver metastases in association with the ascites.

3. Paracentesis is used to distinguish malignancy from other causes of peritoneal effusions, including congestive heart failure, hepatic cirrhosis, and peritonitis. Malignant cells are found in about half of patients in whom the effusion is due to malignancy. Other tests are less reliable, and treatment decisions must often be based on incomplete data. Elevated LDH and protein levels, along with a negative Gram stain and cultures, are supportive but nonspecific for malignancy. The use of monoclonal antibodies to identify tumor cells is still experimental. The serum-ascites albumin gradient (SAAG) is also useful. The SAAG is simply calculated by subtracting the ascitic fluid albumin from the serum albumin. If the gradient is at least 1.1 g/dL, the ascites is most likely from portal hypertension, congestive heart failure, or from massive hepatic metastases in a patient with cancer. If the gradient is less than 1.1 g/dL in a patient with cancer, the more likely cause is peritoneal metastasis or another inflammatory condition.

C. Therapy

As with malignant pleural effusions, malignant peritoneal effusions as a rule are optimally treated with effective systemic therapy. (The possible exception to this is peritoneal effusions from carcinoma of the ovary. In this circumstance, there may be an advantage to intraperitoneal therapy, at least as one component of therapy, because most systemic disease is on the peritoneal surface.) If the patient is resistant to all further systemic treatment, regional treatment should be tried, but the likelihood of success is less and the complications greater with peritoneal effusions than with pleural effusions. Success probably is less because of the greater likelihood of loculations to areas inaccessible to therapy and the impossibility of obliterating the peritoneal space in the same way that the pleural space can be obliterated. Complications are greater because of the increase in adhesions caused by instillation therapy and the resultant increase in obstructive bowel problems.

1. Paracentesis may be helpful in acutely relieving intra-abdominal pressure. If the ascites has caused impairment of respiration, paracentesis may give temporary relief. Rapid withdrawal of large volumes of fluid (more than 1 L) can result in hypotension and shock, however, and if frequent paracenteses are performed, severe hypoalbuminemia and electrolyte imbalance may result. Repeated procedures could also subject the patient to increased risk of peritonitis or bowel injury. This procedure thus results in only temporary benefit.

2. Bed rest and dietary salt restriction, although helpful in the treatment of various nonmalignant causes of ascites, are of less benefit in malignant ascites.

3. Diuretics may be helpful in reducing ascites, but care must be taken not to be too vigorous in attempts at diuresis because of the possibility of dehydration and hypotension. A reasonable choice of diuretic is a combination of either furosemide 40 mg or hydrochlorothiazide 50 to 100 mg/day and spironolactone 50 to 100 mg/day.

4. Intracavitary therapy. Radioisotopes, cytotoxic drugs, and sclerosing agents have been used with some benefit for treating malignant ascites, but overall probably fewer than half of patients have a satisfactory response. The utility of these agents has less to do with direct tumor cytotoxicity and more with the induction of a local inflammatory response with subsequent sclerosis. The radioactive isotopes gold-198 and phosphorus-32 should be used only by those with experience and appropriate certification. Cytotoxic agents such as fluorouracil are associated with less risk to the person administering the therapy.

a. Method. The peritoneal fluid should be drained slowly through a Tenckhof catheter over a 24- to 36-hour period. The potential distribution of the therapeutic agent can be determined by instilling technetium-99m–glucoheptonate macroaggregated albumin in 50 mL of saline and obtaining an a bdominal scintigram. Two liters of warmed 1.5% peritoneal dialysate solution is instilled, allowed to remain for 2 hours, and then drained. The chemotherapeutic agent is next mixed with 2 L of fresh 1.5% dialysate solution containing 1000 U of heparin/L. After warming, this solution is instilled through the Tenckhof catheter. For some agents, draining after 4 hours is recommended.

b. Agents

(1) Cisplatin 50 to 100 mg/m2 (especially for carcinoma of the ovary). Drainage is optional. Saline diuresis is recommended. Dosages higher than 100 mg/m2 should not be used without protection by intravenous sodium thiosulfate. Cisplatin is repeated every 3 weeks.

(2) Fluorouracil 1000 mg (total dose) in normal saline with 25 mEq of sodium bicarbonate/L. Drainage is optional. Treatment is given on days 1 to 4 monthly.

(3) Mitoxantrone 10 mg/m2. Drainage is optional. This dose has been administered on a weekly basis, although white blood cell counts must be monitored.

(4) Interferon-α 50 × 106 U (for ovarian cancer). Drainage is optional. This dose has been administered weekly for 4 weeks or longer. Patients should be premedicated with acetaminophen before and every 4 hours on the day of therapy.

(5) Floxuridine 3 g in 1.5 to 2 L of normal saline given daily for 3 days every 3 to 4 weeks has been used in colon, gastric, and ovarian cancer.

(6) Other agents that have been used intraperitoneally include carboplatin, paclitaxel, methotrexate, cytosine arabinoside, etoposide, bleomycin, thiotepa, and doxorubicin. High-dose interleukin (IL)-2 with lymphokine-activated killer cells has shown activity in ovarian and colorectal cancer but at the cost of significant toxicity, including peritoneal f brosis, which in general has prevented the administration of more than one or two cycles. Lower-dose IL-2, 6 × 106 IU, on days 1 and 7 has been used successfully.

5. Peritoneal-venous shunts (Denver shunt, LeVeen shunt) may offer palliative relief for refractory ascites because recurrent paracentesis leads to infection and leakage of peritoneal fluid through the paracentesis sites. Potential disadvantages are shunt occlusion, the systemic dissemination of cancer, and disseminated intravascular coagulation.


Although 5% to 10% of patients dying with disseminated malignancy have cardiac or pericardial metastases, far fewer have symptomatic pericardial effusion. However, although malignant pericardial effusions are not particularly common, they are of great importance because of their potential to cause acute cardiac tamponade and death.

A. Causes

The most common neoplasms causing pericardial effusions are carcinomas of the lung and breast, lymphomas, and melanoma.

B. Diagnosis

1. Clinical diagnosis. Patients with developing cardiac tamponade may exhibit a variety of grave symptoms including extreme anxiety, dyspnea, orthopnea, precordial chest pain, cough, and hoarseness. On examination, they are likely to have engorged neck veins, generalized edema, tachycardia, distant heart tones, lateral displacement of the cardiac apex, a low systolic blood pressure and low pulse pressure, and a paradoxical pulse. They may also have tachypnea and a pericardial friction rub.

2. Electrocardiogram (ECG) may show nonspecific low-voltage, T-wave abnormalities, elevation of ST segments, and ventricular alternans or the more specific total electrical alternans. Premature beats and atrial fibrillation also occur.

3. Chest radiograph typically shows an enlarged cardiac silhouette, often with a bulging appearance suggestive of an effusion (“ water-bottle heart”). There is frequently an associated pleural effusion.

4. Echocardiography can confirm the diagnosis and provide important information on the location of the effusion within the pericardium.

5. Pericardiocentesis reveals neoplastic cells on cytologic examination in more than 75% of patients.

C. Treatment

1. Volume expansion and vasopressor support are applied (if necessary) to maintain blood pressure. Adequate oxygenation must be maintained. Diuretics are contraindicated.

2. Pericardiocentesis under ECG and blood pressure monitoring should be done in emergent circumstances. If the patient can be stabilized or in cases of pericardial effusion without tamponade, pericardiocentesis under two-dimensional ECG is preferable because it significantly reduces the incidence of cardiac laceration, arrhythmia, and tension pneumothorax as a complication of the procedure.

3. Instillation of chemotherapeutic or sclerosing agents. Because pain may be associated with the intrapericardial therapy, lidocaine (Xylocaine) 100 mg may be administered intrapericardially as a local anesthetic. (Check with the cardiologist on the safety for each patient.) After the cytotoxic or sclerosing agent is instilled, the pericardial catheter is clamped for 1 to 2 hours and then allowed to drain. One of the following agents may be used.

a. Fluorouracil 500 to 1000 mg in aqueous solution as supplied commercially. This dose is generally not repeated.

b. Thiotepa 25 mg/m2 in 10 mL of normal saline may be preferred in tumors deemed sensitive to alkylating agents. Myelosuppression may occur. The dose is usually not repeated.

Complications of intrapericardial therapy include arrhythmias, pain, and fever.

4. Radiotherapy with radioisotopes or 2000 to 4000 cGy of external-beam therapy may help control effusions.

5. Systemic chemotherapy (with standard regimens) after pericardiocentesis is a possible alternative for newly diagnosed, potentially responsive malignancies such as lymphomas. Chemotherapy, i ntrapericardial or systemic, or radiotherapy controls the effusion for at least 30 days in 60% to 70% of patients.

6. Surgery to create a pericardial window may be necessary and can be effective for several months. It is not recommended, however, unless simpler measures fail.


A. Causes

Leptomeningeal involvement with non–central nervous system cancer is an uncommon complication of most neoplasms, although in children with acute lymphocytic leukemia who have not received prophylactic treatment, the incidence approaches 50%. Of the nonleukemic diseases, breast carcinoma and lymphomas (primarily Burkitt and T-cell lymphoblastic) account for about 30% each in cases of malignant subarachnoid infiltrates. Carcinoma of the lung and melanoma account for 10% to 12% each.

B. Diagnosis

1. Clinical diagnosis. Patients commonly present with headache, change in mental status, cranial nerve dysfunction, or spinal root–derived pain, paresthesia, or weakness. Any onset of change in neurologic status, particularly of cerebral, cranial nerve, or spinal root origin, should alert the clinician to the possibility of subarachnoid infiltrates.

2. Diagnostic studies

a. CT or magnetic resonance imaging (MRI) of the head should be done to look for any intracranial mass. If none is present, a lumbar puncture should be done.

b. A lumbar puncture is done to obtain 8 to 12 mL of cerebrospinal fluid (CSF) fluid; 4 mL may be frozen to use for subsequent studies when initial evaluation is inconclusive. The following are evaluated or performed:

bull Opening pressure

bull Cytology of centrifugal specimen for malignant cells

bull Total cell count and differential

bull CSF chemistry, including glucose and protein

bull Microbiologic studies as indicated by the clinical situation: India ink or cryptococcal antigen determination, Gram stain, cultures (routine, acid-fast, fungi), serum toxoplasma titer, CSF viral (e.g., Epstein-Barr virus, cytomegalovirus, HIV, and herpes simplex virus) polymerase chain reaction, and other special studies.

c. MRI of the spine (or less commonly, myelography with CT follow-up) is performed if signs or symptoms of cord compression are present.

C. Treatment

Malignant subarachnoid infiltrates may be treated with radiotherapy, intrathecal chemotherapy, or a combination of the two.

1. Radiotherapy. The radiation field is usually limited to the most involved field ( frequently the brain), and intrathecal chemotherapy is used to control the infiltrates elsewhere. This technique is used even though the entire neuraxis is usually involved because total craniospinal irradiation causes severe myelosuppression, which limits the patient’s tolerance to concurrent or subsequent cytotoxic chemotherapy.

2. Chemotherapy may be administered by lumbar puncture or preferably into a surgically implanted (Ommaya) reservoir that communicates with the lateral ventricle. The latter has the advantages of being easily accessible in patients who require repeated treatments and of giving a better distribution of drug than can be obtained through lumbar puncture. When the Ommaya reservoir is used, a volume of CSF equal to that to be injected (6 to 10 mL) should be removed through the reservoir with a small-caliber needle. The chemotherapy should then be given as a slow injection. When the chemotherapy is given through lumbar puncture, the volume of injection (usually 7 to 10 mL) should be greater than that of the CSF withdrawn, so as to have a higher closing than opening pressure. This method facilitates distribution of the drug and minimizes postlumbar puncture headache. The most commonly used drugs for intrathecal therapy are the following.

a. Methotrexate 12 mg/m2 (maximum 15 mg) twice weekly until the CSF clears of malignant cells, then monthly.

b. Cytarabine 30 mg/m2 (maximum 50 mg) twice weekly until the CSF clears of malignant cells, then monthly.

c. Liposomal cytarabine 50 mg (total dose) is given every 14 days for two doses. If the CSF clears, give 50 mg every 14 days for two additional doses. Then give 50 mg every 4 weeks for two additional doses (total of six doses).

d. Thiotepa 2 to 10 mg/m2 twice weekly until the CSF clears of malignant cells, then monthly.

Each of the agents is given in preservative-free saline or, if available, buffered preservative-free diluent similar to Elliot B solution. Any subsequent flush solution should be of similar composition. Other drugs used to treat effusions (e.g., fluorouracil, mechlorethamine, or radioisotopes) must not be used to treat meningeal disease.

D. Response to treatment

Most patients with meningeal leukemia or lymphoma respond to a combination of radiotherapy and intrathecal chemotherapy. Carcinomas are less likely to improve, but mild to moderate improvement may be seen in up to 50% of patients.

E. Complications

Aseptic meningitis or arachnoiditis, seizures, acute encephalopathy, myelopathy, leukoencephalopathy, and radicular neuropathy may result from intrathecal chemotherapy with or without radiotherapy. Bone marrow suppression is not usually severe unless the patient undergoes spinal irradiation or systemic chemotherapy as well. Oral leucovorin can be given after the intrathecal methotrexate (10 mg leucovorin by mouth every 6 hours for six to eight doses, starting either at the same time or 24 hours after the methotrexate) to prevent marrow toxicity. Serious complications are infrequent, however, and in patients with advanced metastatic disease, they usually are not a major problem.

Selected Readings

Pleural Effusions

Andrews CO, Gora W. Pleural effusions: pathophysiology and management. Ann Pharmacother. 1994;28:894–903.

de Campos JR, Vargas FS, de Campos Werebe E, et al. Thoracoscopy talc poudrage: a 15-year experience. Chest. 2001;119:801–806.

Diacon AH, Wyser C, Bollinger CT, et al. Prospective randomized comparison of thoracoscopic talc poudrage under local anesthesia versus bleomycin instillation for pleurodesis in malignant pleural effusions. Am J Respir Crit Care Med. 2000;162:1445–1449.

Fuller DK. Bleomycin versus doxycyclines: a patient-oriented approach to pleurodesis. Ann Pharmacother. 1993;27:794.

Herrington JD. Chemical pleurodesis, with doxycycline 1 g. Pharmacotherapy. 1996;16:290–295.

Johnson WW. The malignant pleural effusion: a review of cytopathologic diagnoses of 584 specimens from 472 consecutive patients. Cancer. 1985;56:905.

Kessinger A, Wigton RS. Intracavitary bleomycin and tetracycline in the management of malignant pleural effusions: a randomized study. J Surg Oncol. 1997;36:81–83.

Patz EF Jr, McAdams HP, Erasmus JJ, et al. Sclerotherapy for malignant pleural effusions: a prospective randomized trial of bleomycin vs. doxycyline with small-bore catheter drainage. Chest.1998;113:1305–1311.

Putnam JB Jr, Walsh GL, Swisher SG, et al. Outpatient management of malignant pleural effusion by a chronic indwelling catheter. Ann Thorac Surg. 2000;69:369–375.

Stefani A, Natali P, Casali C, Morandi U. Talc poudrage versus talc slurry in the treatment of malignant pleural effusion. A prospective comparative study. Eur J Cardiothorac Surg. 2006;30:827–832.

Tremblay A, Mason C, Michaud G. Use of tunnelled catheters for malignant pleural effusions in patients fit for pleurodesis. Eur Respir J. 2007;30(4):759–762.

Van Hoff DD, LiVolsi V. Diagnostic reliability of needle biopsy of the parietal pleura: a review of 272 biopsies. Am J Clin Pathol. 1975;64:200.

Walker-Renard PB, Vaughan LM, Sahn SA. Chemical pleurodesis for malignant p leural effusions. Ann Intern Med. 1994;120:56–64.

Peritoneal Effusions

Lacy JH, Wieman TJ, Shivley EH. Management of malignant ascites. Surg Gynecol Obstet. 1984;159:397.

Muggia FM, Liu PY, Alberts DS, et al. Intraperitoneal mitoxantrone or floxuridine: effects on time-to-failure and survival in patients with minimal residual ovarian cancer after second-look laparotomy: a randomized phase II study by the Southwest Oncology Group. Gynecol Oncol. 1996;61:395–402.

Speyer JL, Beller U, Colombo N, et al. Intraperitoneal carboplatin: favorable results in women with minimal residual ovarian cancer after cisplatin therapy. J Clin Oncol. 1990;8:1335–1341.

Sugarbaker PH, Gianola FJ, Speyer JC, et al. Prospective, randomized trial of intravenous versus intraperitoneal 5-fluorouracil in patients with advanced primary colon or rectal cancer. Surgery. 1985;95:414.

Pericardial Effusions

Buzaid AC, Garewal HS, Greenberg BR. Managing malignant pericardial effusion. West J Med. 1989;150:174–179.

Callahan JA, Seward JB, Nishimura RA, et al. Two-dimensional echocardiographically guided pericardiocentesis: experience in 117 consecutive patients. Am J Cardiol. 1985;55:476–479.

Helms SR, Carlson MD. Cardiovascular emergencies. Semin Oncol. 1989;16:463.

Liu G, Crump M, Gross PE, et al. Prospective comparison of the sclerosing agents doxycycline and bleomycin for the primary management of malignant pericardial effusion and cardiac tamponade. J Clin Oncol. 1996;14:3141–3147.

Maher ER, Buckman R. Intrapericardial installation of bleomycin in malignant pericardial effusion. Am Heart J. 1986;111:613–614.

Runyon BA, Montano AA, Akriviadis EA, Antillon MR, Irving MA, McHutchison JG. The serum-ascites albumin gradient is superior to the exudate-transudate concept in the differential diagnosis of ascites. Ann Intern Med.1992;117(3):215–220.

Shepherd FA, Ginsberg JS, Evans WK, et al. Tetracycline sclerosis in the management of pericardial effusion. J Clin Oncol. 1985;3:1678–1682.

Malignant Subarachnoid Infiltrates

Glantz MJ, Jaeckle KA, Chamberlain MC, et al. A randomized controlled trial comparing intrathecal sustained-release cytarabine (Depocyt) to intrathecal methotrexate in patients with neoplastic meningitis from solid tumors. Clin Cancer Res. 1999;5:3394–3402.

Grossman SA, Krabak MJ. Leptomeningeal carcinomatosis. Cancer Treat Rev. 1999;25:103–119.

Grossman SA. Advances in the treatment of central nervous system metastases: treatment of leptomeningeal metastasis. In: American Society of Clinical Oncology, Educational Book. Alexandria, VA: American Society of Clinical Oncology; 2001:598–604.

Gutin PH, Levi JA, Wiernik PH, et al. Treatment of malignant meningeal disease with intrathecal thiotepa: a phase II study. Cancer Treat Rep. 1977;61:885–887.

Jaeckle KA, Phuphanich S, van den Bent MJ, et al. Intrathecal treatment of neoplastic meningitis due to breast cancer with a slow-release formulation of cytarabine. Br J Cancer. 2001;84:157–163.