The Treatment of Recurrent Malignant Pleural Effusion

COLLECTIVE REVIEW

The Treatment of Recurrent Malignant Pleural Effusion Erle H. Austin, M.D., and M. Wayne Flye, M.D.

ABSTRACT Effective control of a recurrent malignant pleural effusion can greatly improve the quality of life of the cancer patient. At least a dozen different techniques have been advocated for controlling this common complication of malignant disease. The present review collects and examines the clinical results of all techniques designed to treat this problem. The pathophysiologyand diagnostic evaluation of the effusion are also discussed. On the basis of comparisons involving effectiveness, morbidity, and convenience, we recommend intrapleurally administered tetracycline with thoracostomy drainage as the technique of choice. Instillation of a talc suspension with thoracostomy drainage is also a safe and effective technique and should be employed when tetracycline fails or is contraindicated.

Incidence of Malignant Pleural Effusions Several large retrospective studies [59, 93, 991 (Table 1)have clearly established malignancy as a major cause of pleural effusion. At least one-third and probably closer to one-half of all pleural effusions are secondary to malignancy. Malignant pleural effusions are most commonly secondary to lung cancer, breast cancer, or lymphoma (Hodgkin’s or non-Hodgkin’s) [6,59,72, 991 (Table 2). Metastases from ovarian, genitourinary, and gastrointestinal malignancies as well as less common malignancies such as mesotheliomas, sarcomas, and melanomas also can cause pleural effusion. The likelihood of developing a pleural effusion is quite high in certain metastatic or unresectable primary malignancies. In a retrospective analysis of patients with disseminated breast cancer, Fracchia and associates [331 noted the presence of pleural effusion in 286 (48%) of 601 patients. The effusion in 138 of these patients was sufficiently symptomatic and intractable to warrant intracavitary therapeutic agents. In a postmortem study of patients with lung cancer, Farber [31] noted that 353 (33%) of 1,070 patients had pleural effusions, which correlated with a 30% incidence of effusion as determined from roentgenographic studies made before death. Similar studies in patients with lymphoma also have demonstrated a high incidence of pleural effusion. The incidence ranges from 16 to 29% for Hodgkin’s lymphoma [32, 70, 1011, and from 20 to 33% in other lymphomas [19, 66, 791.

Recurrent pleural effusions are prevalent and vexing problems in patients with advanced malignant disease. Many of these patients experience marked respiratory discomfort although other aspects of the disease would allow them to lead active and productive lives. Frequent thoracenteses inconvenience both patient and physician, can cause the complications of pneumothorax and protein depletion, and at best provide only temporary symptomatic relief. More satisfactory methods for controlling these effusions have been developed. Because the methods are so diverse and provide such variable results, a thorough analysis of the literature was undertaken. The purpose of this review is to present the collected resuIts of all of the published techniques in a fashion that facilitates comparison of their relative values. The Pathophysiology of Pleural Effusion From this comparison the techniques of choice Understanding the pathophysiology behind malignant pleural effusion is helpful in can be determined. evaluating and developing effective therapies. The movement of fluid into and out of the From the Surgery Branch, National Cancer Institute, Na- pleural space is based on principles described tional Institutes of Health, Bethesda, MD. by Starling [891 in 1898. In normal man as a Address reprint requests to Dr. Austin, Surgery Branch, National Cancer Institute, Building 10, Room 10N116, Of net hydrostatic and Osmotic Bethesda, MD 20014. pressures, protein-free liquid filters from the 190

191 Collective Review: Austin and Flye: Recurrent Malignant Pleural Effusion

Table 1. Causes of Pleural Effusion 1946 Tinney, Olsen [991 Cause

No. of Patients

Neoplasms CHF Infection Miscellaneous Indeterminate Total

169 42 40 23 170 444

1955 Leuallen, Cam [591 No. of Patients

Percent

229 44 36 52 75 436

38.1 9.4 9.0 5.2 38.3

Percent

52.5 10.1 8.3 11.9 17.2

1976 Storey et a1 [93] No. of Patients

64 18 7 19 25 133

Percent

48.1 13.5 5.3 14.3 18.8

CHF = congestive heart failure.

Table 2 . Neoplasms That Cause Pleural Effusion 1955 Leuallen, Carr [591

1946 Tinney, Olsen [991 Neoplasm Lung Breast Lymphoma Othera Total

No. of Patients

47 42 28 52 169

Percent

No. of Patients

28 25 16 31

95 53 28 53 229

Percent

42 23 12 23

1974 Anderson et a1 [61 No. of Patients

32 35 32 34 133

Percent

24 26 24 26

“Includes ovary, unknown primary, gastrointestinal tract, mesothelioma, uterus, kidney, sarcoma, melanoma.

systemic capillaries in the parietal pleura into the pleural space and thence into the pulmonary capillaries of the visceral pleura [121. Five to 10 liters of protein-free fluid pass through the pleural space of a normal man in a 24-hour period 13, 611. Normally only a small amount of plasma protein escapes through pleural capillary walls, and a low pleural fluid protein concentration results. The protein that does enter the pleural space can only reenter the circulation by lymphatic drainage [25, 911, which is limited to less than a liter of protein-containing fluid per 24-hour period [90]. Thus, the normal dynamic equilibrium in the pleural space consists of protein-free fluid entering from the parietal pleura and leaving by the visceral pleura while minimal amounts of protein enter from both pleural surfaces and leave by the lymphatics. Normally this equilibrium results in less than 5 ml of fluid in the pleural space [121. However, when the normal equilibrium is

upset, fluid will accumulate until equilibrium is reestablished. The four major forces controlling pleural fluid dynamics-hydrostatic pressure, colloid osmotic pressure, capillary permeability, and lymphatic drainage-are all subject to marked alteration by certain disease states. Congestive heart failure produces a pleural effusion because of increased hydrostatic pressure in systemic and pulmonary venous capillaries. Cirrhosis and nephrotic syndrome produce effusions because plasma colloid osmotic pressure is markedly decreased. Effusions secondary to alterations in hydrostatic pressure or colloid osmotic pressure are defined as transudates, and usually contain less than 3.0 gm per 100 ml of protein (specific gravity, less than 1.016). Increased capillary permeability occurs with many disease processes that involve the pleura. Infection, collagen diseases, pulmonary infarc-

192 The Annals of Thoracic Surgery Vol 28 No 2 August 1979

tion, and neoplastic involvement all produce sufficient pleural capillary breakdown to allow notable protein spillage into the pleural space. Effusions resulting from this process are defined as exudates and usually contain more than 3.0 gm per 100 ml of protein (specific gravity, more than 1.016). Obstruction of lymphatic drainage from the pleural space can also produce pleural effusion. Although protein entry into the pleural space is normal, protein exit is retarded, thereby favoring the influx of additional protein-free fluid. When a new equilibrium is reached, marked effusion results. Because protein cannot readily escape, these effusions usually contain more than 3.0 gm per 100 ml of protein and are thereby usually included among the exudative effusions [12]. This mechanism appears responsible for many of the effusions associated with lymphoma. Thus, the primary mechanisms of pathogenesis of neoplastic effusions are thought to be pleural involvement by tumor or lymphatic obstruction by tumor. Free growth of large numbers of malignant cells in the pleural space, much like cells in experimental ascites tumors, has been proposed as another mechanism of effusion [261. These cells apparently produce an inflammatory reaction in the pleura, and an exudative process evolves. This type of effusion is unusual, but is suggested by cancer cell counts of more than 4,000 cells per cubic millimeter 1263.

Diagnosis The development of shortness of breath, cough, or chest pain in a patient with known malignant disease should raise the suspicion of pleural effusion. While physical findings, such as dullness or egophony, often only pick up the larger effusions, a chest roentgenogram will detect as little as 100 ml of pleural fluid if made in the lateral decubitus position [741. Pleural effusion seen on upright chest roentgenogram represents at least 300 ml of fluid [481. Although the development of a pleural effusion in a patient with malignant disease probably represents progression of the tumor, it does not always do so. Careful evaluation of the effusion is warranted in all patients without regard to previous history of malignancy.

The gross appearance of the effusion provides suggestive but nonspecific information. Most effusions are clear, straw-colored, and odorless. Milky-appearing effusions suggest either a chylous effusion or an empyema. Empyema fluid is usually viscid and malodorous and forms a clear supernatant after centrifugation, whereas a chylous effusion is thin and odorless and does not separate with centrifugation. A bloody pleural effusion (red cell count greater than 100,OOO/mma) suggests trauma, tuberculosis, pulmonary infarction, or malignancy. The gross appearance of most malignant effusions, however, resembles that of many benign effusions. Accurate distinction between exudative and transudative pleural effusion is important. If it can be established with certainty that an effusion is secondary to a transudative process, pleural disease as such can be ruled out and appropriate therapy can be directed at the altered hydrostatic or colloid osmotic pressure. If an effusion is clearly secondary to an exudative process, further diagnostic efforts including cytopathology and pleural biopsy must be undertaken to establish the exact nature of the pleural disease. Although a pleural fluid protein concentration of 3.0 gnh per 100 ml is often used to divide transudates from exudates, significant overlap does exist [22, 621. In a recent prospective study, Light and co-workers [621 discovered that 4 out of 47 pleural effusions resulting from congestive heart failure, nephrosis, or cirrhosis contained more than 3.0 gm per 100 ml of protein. Eleven out of 103 effusions resulting from well-defined exudative processes contained less than 3.0 gm per 100 ml of protein. Most disturbingly, 8 (1g0/o) of 43 malignant effusions were incorrectly classified as transudates when 3.0 gm per 100 ml was used as the dividing line. In order to more accurately separate exudates from transudates, Light and colleagues [621 advocated simultaneous measurement of serum and pleural fluid protein and lactic dehydrogenase (LDH). With these values, the presence of an exudate can be determined if one or more of the following criteria exist: (1) a pleural fluid protein to serum protein ratio greater than 0.5; (2) a pleural fluid LDH to serum LDH ratio greater than 0.6; or (3) a pleural fluid LDH

193 Collective Review: Austin and Flye: Recurrent Malignant Pleural Effusion

greater than 200 IU. By these criteria, these researchers were able to correctly classify 148 out of 150 pleural effusions. Once the exudative nature of the effusion has been established, the two most frequently employed studies are pleural fluid cytology and pleural biopsy. Because of its simplicity, cytological examination of pleural fluid has become a part of routine pleural fluid evaluation. Unfortunately, the absence of malignant cells in the pleural fluid does not rule out malignancy. In fact, Jarvi and colleagues 1421 found malignant cells in only 43 (42%) of 103 patients subsequently found to have cancer-related effusions. The importance of repeating cytological examinations was emphasized by Salyer and associates [Sl] in a study of 95 patients known to have cancer involving the pleura. Of these patients, 50 had positive cytological findings with the first specimen (53%), 11 more on the second, 5 more on the third, and 3 more on the fourth, for a total of 69 (73%). This same group of patients also underwent pleural needle biopsy. One biopsy produced positive results in 47 out of 95 patients (49'/0), and a second biopsy added 6 more patients with positive findings, for a total of 53 (56%). As a single method for identifying the malignant nature of a pleural effusion, pleural fluid cytology appears superior to pleural needle biopsy. However, out of the 95 patients in that study, the combination of cytology and pleural biopsy established the diagnosis in 86 patients (90%). The combination of the two techniques thereby minimizes false negative evaluations. The 10% of malignant effusions not correctly identified by these standard techniques probably results from mechanisms other than direct pleural involvement with tumor. Malignant involvement of lymph nodes and channels by lung cancer or lymphoma, for example, can produce effusions without exfoliating cancer cells. Lung cancer can cause bronchial obstruction leading to localized parenchymal and pleural inflammation from which a copious exudate emanates without accompanying malignant cells. Thus, the inability to identify malignant cells by cytology or pleural biopsy does not rule out malignancy if other clinical information (previous histology, roentgeno-

graphic findings, and overall clinical course) strongly favors it. Many patients suffering from such effusions would be denied the palliative relief often afforded with systemic or local therapy if positive cytology or pleural biopsy was an absolute requirement for institution of therapy. Systemic versus Local Therapy Once the malignant nature of the effusion is known and the primary tumor has been identified, appropriate therapy can commence. It is most judicious to begin with the current standard treatment for advanced or unresectable disease. The standard therapy varies with the type of primary. Lung cancer often shows temporary response to localized irradiation, and lymphoma and breast cancer often respond to systemic forms of therapy, including chemotherapeutic agents and hormonal manipulations. In some patients, the standard modalities result in objective improvement of the pleural effusion. Unfortunately, despite recent advances in chemotherapy and radiation therapy, there are still many patients who are disabled by persistent effusions. Fortunately, despite their lack of direct effect on tumor growth, techniques have been developed to effect local control of pleural fluid accumulation and alleviate the associated respiratory distress. Evaluation of the relative effectiveness, convenience, and morbidity of these techniques is the focus of the balance of this review. Criteria for the Use of Local Therapy Regardless of which local technique is ultimately employed, a pleural effusion should meet certain strict criteria before local therapy is instituted. It should be clear to the clinician that the effusion is caused by a malignant condition, that the effusion produces respiratory symptoms that are relieved by removal of the effusion, and that the effusion fails to respond to standard cancer therapies. Also, in the physician's best judgment, the patient should be in relatively satisfactory condition with a life expectancy of several months.

Analysis of Clinical Results All clinical studies evaluating local therapy of malignant pleural effusion published in the En-

194 The Annals of Thoracic Surgery Vol 28 No 2 August 1979

Table 3. Effectiveness of Techniques to Control Malignant Pleural Effusion ~~

Percent Objective Response"

Technique

References

Lung

Breast

Lymphoma

Ovary

Intrapleural radiosotopes

7, 8, 13, 16, 18, 24, 28, 29, 36, 3941, 47, 4951, 65, 68, 69, 78, 82, 83, 86, 87, 92, 103, 104 4, 6, 9, 33, 46, 52, 58, 60, 96, 106 1, 20,37, 45, 73, 75, 85, 88 2, 23

54 yo 1021190

57% 2341413

50% 22144

59yo 32154

66% 35153

48% 100/208

37yo 7/19

93?'o 42145

96% 26127

86Yo 12/14 83Yo 30136

Intrapleural nitrogen mustard Intrapleural talc (gen'l anest) Intrapleural talc (suspension) Intrapleural quinacrine

Range Among Studies Overall

(yo)

48 Yo 851177

55o/' 5361980

25-100

73o/' 8111

46 Yo 11/24

52o/' 177/338

28-87

100% 111

lOO~/O

87% 27131

92% 971105

76-100

111

96% 27128 80YO 37146

100% 313 78Yo 719

88Yo 718 50o/' 316

67% 416 77% 17122

90% 53159 80YO 1021128

83-93

100% 717 0Yo 011

86Yo 617 5270 11121

0o/o

100% 414 0 Yo 012

... ...

... ...

... ...

... ...

71

100% 3/3 100% 111 100% 41141

100% 818 53yo 17132 100% 33133

50% 214

100% 313

... ...

... ...

100% 111 14yo 117 100% 30130

87% 27131 46 Yo 18/39 66Yo 23135 90% 17119 55yo 38169 99 yo 1451147

83-100

011 100% 212

95

... ... ... ... ... ...

15, 27, 35, 38, 44, 53, 76, 97, 100 53, 80, 102

Intrapleural tetracycline Intrapleural thio-TEPA Intrapleural 5fluorouracil Intrapleural bleomycin Chest tube alone

6, 17,39, 56

Pleurectomy

43, 64

5, 10, 33

100% 111

lOO~/O

111

Other

57-100

30-63

... ... ... ... 0-100 95-100

'Total Objective ResponseslTotal Evaluable Patients (objectiveresponse = no requirementfor thoracentesis for at least one month following procedure; evaluable patient = survives at least one month following procedure).

glish literature since the earliest report of intrapleural therapy [461 were reviewed. For each clinical study, the following data were carefully collected: (1)type of agent used for treatment, if any; (2) manner in which the agent was administered; (3) type of tumors producing effusion; (4) number of patients studied who could be evaluated; (5) objective response rate; and (6) morbidity and mortality. Criteria used to determine which patients could be evaluated and which showed an objective response varied considerably from study to study. To minimize this nonuniformity, each study was carefully analyzed and the following

criteria were applied. To be evaluated, the patient must have required frequent and multiple thoracenteses prior to therapy and survived at least one month after therapy. To have obtained an objective response, the patient must have gone for at least one month after therapy without requiring thoracentesis. An overall objective response rate for each particular technique was determined by dividing the total number of objective responses in all studies employing the technique by the total number of patients evaluated. The overall objective response rates for the techniques to be discussed are presented in Table 3.

195 Collective Review: Austin and Flye: Recurrent Malignant Pleural. Effusion

Because in many cases there was significant variability in response rate among studies evaluating the same technique, we have attempted to avoid selection artifact by including all clinical studies containing sufficient data for evaluation. Where possible, objective response rates were also determined within tumor subgroups (see Table 3) to ascertain what effect the type of primary tumor might have on likelihood of objective response. We evaluated a total of 66 clinical studies involving 1,950 patients who could be evaluated. Several reports [21, 34, 54, 57, 77, 841 from which sufficient data could not be reliably discerned were excluded from the evaluation.

Thoracentesis Thoracentesis is clearly the first technique to employ in a patient with an effusion. Of course, it must be done initially for diagnostic purposes and is also important for determining the effect of removal of the effusion on the patient’s respiratory symptoms. Some patients are relieved after the first thoracentesis and require no further local therapy. For the many patients who rapidly have recurrence, further thoracenteses are unlikely to give lasting relief. Anderson and co-workers [61 found that of 97 patients treated with thoracentesis alone, only 4 were free of recurrence for more than 30 days. In their series, the average time to recurrence was only 4.2 days. Repeated and frequent thoracenteses are not without complications. Hypoproteinemia often results and even contributes to an increasing rate of effusion reaccumulation. Pneumothorax, empyema, and fluid loculation threaten each aspiration. For these reasons, thoracentesis has a limited role in the continuing treatment of recurrent malignant effusions. Nevertheless, for diagnosis, for palliation in the terminal patient, or for immediate relief of acute respiratory distress, thoracentesis remains an important procedure. Radiation Therapy In the 1940s, x-ray beam therapy was the only available local treatment, other than thoracentesis, for malignant pleural effusions. Unfortunately, doses sufficient to produce a response were complicated by radiation sickness and

pulmonary fibrosis. Doses low enough to avoid these problems were uniformly ineffective in resolving the effusion [491. In 1973, however, Strober and colleagues [941 reported their results using a moving strip technique and super-voltage therapy. Seven of 10 patients with recurrent malignant effusions that were not lymphomatous were effectively treated without adverse effects. Thus, with today’s advanced equipment and techniques, radiation therapy could become an appropriate modality for safely and effectively treating recurrent malignant effusions. However, at this time, too few patients have been treated to recommend this form of therapy for nonlymphomatous effusions. External beam radiation therapy does have a role, however, in treating lymphomatous pleural effusions. Probably because blockage of lymph channels is a prevalent mechanism for effusion in such patients, radiation therapy to the mediastinum is often associated with resolution of the effusion, regardless of the clinical state of the mediastinal nodes. In fact, in one study, 7 of 10 patients with lymphoma and symptomatic recurrent effusions and no mediastinal adenopathy experienced no further recurrence after 1,400 to 2,600 rads of mediastinal irradiation [191. In reviewing the Mayo Clinic experience with pleural effusion and lymphoma from 1950 through 1964 (159 patients), Weick and associates [lo51 thought that ”radiotherapy to the mediastinum . . . was the most frequent and successful treatment modality employed.” Most oncologists agree that with lymphomas, the treatment of choice for symptomatic pleural effusion is mediastinal irradiation.

lntrapleural Radioisotopes In the 1940s when external beam radiation was found unsatisfactory for treatment of malignant pleural effusions, direct intrapleural radiation was considered. In 1946, Muller [67] demonstrated the feasibility of using intracavitary radioisotopes when he administered radioactive zinc (63Zn)into the peritoneal cavity of a patient with ovarian cancer. By 1951 intrapleurally administered radioisotopes were being used as treatment for malignant pleural ef-

196 The Annals of Thoracic Surgery Vol 28 No 2 August 1979

fusions [49]. Since that time, intrapleural radioisotope therapy has been the most frequently reported technique for local control of malignant effusion (see Table 3). The isotope most often used has been gold 198 (half-life, 2.70 days; p, 0.961 mev; y , 0.412 mev), although more recently C132P04,which is less expensive and less hazardous, has become the radioisotope of choice (half-life, 14.3 days; 0, 1.71 mev) for intrapleural administration. While the size and composition of the patient groups studied were quite variable, summation of all this experience (see Table 3) demonstrates that 55% of a group of almost 1,000 patients were treated effectively. Differences between the types of primary tumor treated were not marked, although breast and ovarian cancer responded best (57 and 59%, respectively). A recent prospective study using C1.32P04with indwelling thoracostomy tube drainage corroborated the retrospective studies showing an objective response rate of 61% (17 out of 28 patients) [391. The morbidity with intrapleurally administered radioisotopes is minimal. Chang and colleagues [24] noted nausea and vomiting in 8 of 40 patients treated. Ariel and associates [8] noted nausea in 25% of patients, vomiting in 5%, and fever in 5%. Evidence of bone marrow depression was rare and insignificant, and no case of severe radiation sickness was reported. Most reports do not even mention complications or side effects. The greatest drawbacks to the use of radioisotopes are their cost and inconvenience. Neither ls8Au nor Cr32P04is readily available. Their short half-lives prevent storage for long periods because of loss of activity. When they are employed, inconvenient precautions are necessary to prevent radiation hazard to personnel and other patients. Because of these problems, any other agent with equal or better effectiveness without significant morbidity is more likely to be employed. Effective and far more convenient methods do exist. In fact, Izbicki and co-workers [39] in one of the only two prospective randomized studies published compared the effectiveness of C1.32P04and pleural drainage (thoracostomy tube for 72 hours) with pleural drainage alone and found no difference.

In trap leu ral Nitrogen Mustard The intrapleural administration of chemotherapeutic agents for malignant effusion was introduced as early as 1948 with nitrogen mustard [46], and in 1954, Albertelli and coworkers [41 reported a clinical series involving intrapleurally administered nitrogen mustard. The experience to date with this technique is summarized in Table 3. The overall effectiveness was %%, with a 66% objective response in patients with lung cancer. Results with lymphoma were relatively poor (37%), while those with ovarian cancer were quite good (73%); but the numbers were very small for these subgroups and are therefore less reliable. Nevertheless, the evidence strongly suggests that intrapleurally administered nitrogen mustard is as effective as intrapleurally administered radioisotopes. Since the expense and inconvenience of nitrogen mustard was much less than that of radioisotopes, many oncologists began to prefer nitrogen mustard as the intrapleural agent of choice. The side effects of nitrogen mustard, however, are not insignificant. Taylor [96] claimed that in 9 out of 10 patients nausea and vomiting developed as well as mild leukopenia (2,300 to 3,000 white blood cells/mm3).In a group of 132 patients with breast cancer, nausea developed in 45, vomiting in 14, and temporary bone marrow depression in 5. While these side effects are relatively minor, evidence that intrapleural administration can result in bone marrow depression is of concern since many patients today do receive systemic combination chemotherapeutic agents as well. Pathological examination of the pleura in patients who were successfully treated with either intrapleurally administered radioisotopes or nitrogen mustard was revealing. Kniseley and Andrews [551 noted a fusion between visceral and parietal pleura and evidence of a dense fibroblastic reaction in patients successfully treated with radioactive gold. Surprisingly, many of these fused pleura contained tumor implants. Virtually identical findings were found in patients whose effusions had been successfully treated with nitrogen mustard [58, 633. Thus, it became apparent that the effectiveness of intrapleural therapy was more related to the creation of a pleurodesis preventing

197 Collective Review: Austin and Flye: Recurrent Malignant Pleural Effusion

effusion reaccumulation than to any antineoplastic effect of the agent administered. Several clinical studies evaluating intrapleurally administered nitrogen mustard demonstrated that the incidence of objective response was improved by adding thoracostomy drainage to the regimen. In the earliest reports, the intrapleural agent was administered by needle after most of the effusion had been evacuated. Taylor [96] introduced the concept of using an indwelling catheter to completely evacuate the pleural space, to administer the agent, and to maintain pleural drainage after instillation of the agent. In patients with breast cancer and recurrent pleural effusions, Fracchia and associates [33] were able to improve the effectiveness of nitrogen mustard given intrapleurally from 28% to 66% by adding thoracostomy drainage to the regimen. Using thoracentesis followed by pleural instillation of nitrogen mustard, Anderson and colleagues [6] were unable to produce a single objective response in 9 consecutive patients. However, with the addition of thoracostomy tube drainage, an objective response was seen in 31 of 66 patients (47%). As more evidence has accumulated suggesting that the effectiveness of intrapleural therapy depends on creating pleural symphysis, more emphasis has been placed on the use of the thoracostomy tube to ensure that the two pleural surfaces are kept adequately apposed for a time after the intrapleural agent is added.

lntrapleural Talc Recognition of the importance of an adhesive obliterative pleuritis in patients treated with radioisotopes or nitrogen mustard led to the introduction of talc as an intrapleural agent to combat recurrent malignant effusions. Since reported by Bethune [ll]in 1935, intrapleural talc had been used for many years to produce adhesions to help prevent recurrent spontaneous pneumothorax. In 1958, Chambers 1231 reported the successful treatment of 15 of 18 malignant pleural effusions with thoracostomy drainage and a sterile suspension of USP talc in 1YO procaine. Further reports of intrapleurally administered talc have demonstrated startlingly good results (see Table 3). Unfortunately, many of these authors used general anesthesia and

administered the talc either directly by a thoracotomy or by means of a large trocar and a specially designed talc aerosol. Of 34 patients treated by Camishion and associates [201 with thoracotomy and talc poudrage, 3 died in the immediate postoperative period. Even though all of the surviving patients were successfully treated, a 9% mortality is unacceptable. Since most reports have advocated the use of general anesthesia, the intrapleural administration of talc has not gained favor despite its excellent response rates. It is to be hoped that the report of Adler and Sayek [2] published in 1976 will help renew interest in the use of intrapleurally administered talc. Using Chambers’ [23] technique of thoracostomy drainage and intrapleural instillation of a talc suspension (a procedure that requires no more than a few milliliters of local anesthesia), these authors successfully treated 38 (93%) of 41 patients. The morbidity with talc given intrapleurally appears to be minimal. Nausea and bone marrow suppression do not occur. Transient temperature elevations apparently are noted but the incidence has not been systematically evaluated. Pleural pain occurs in many patients but is usually controlled with analgesics. Adler and Sayek [2] did report 2 patients in whom transient hypotension developed following talc instillation. Both patients, however, were considered to have been relatively hypovolemic before this therapy and had been on long-term steroids. They both responded immediately to intravenously administered fluid and steroids.

lntrapleural Quinacrine Another intrapleural agent thought to control malignant effusion by effecting pleural symphysis is quinacrine. Based on a favorable effect on Ehrlich ascites carcinoma in mice administered quinacrine intraperitoneally, Gellhorn and co-workers 1351 employed quinacrine clinically in 1956 as an intracavitary agent (see Table 3). Its overall effectiveness appears to be quite good (80%). The toxic side-effects of intrapleurally administered quinacrine have been well documented. Ultman and associates [loo] noted fever in 31 of 60 patients (52%), and pain in 14 of 60 (23%). Other retrospective studies reported fever in from 50 to 90% of patients [15, 27, 35, 38, 521 and pain in from 35 to 78% of

198 The Annals of Thoracic Surgery Vol 28 No 2 August 1979

patients [15, 27, 35, 38, 531. In addition, Borja and Pugh [15] reported that nausea developed in 11 out of 29 patients (38%), transient hypotension in 4 of 29 (14%), and hallucinations in 3 out of 29 (loo/,) after quinacrine instillation. While quinacrine has no bone marrow suppressive effects, it is known to have cerebrocortical stimulating effects [30]. Borda and Krant [14] reported 2 patients in whom convulsions developed following intrapleural administration of quinacrine. Status epilepticus developed in 1of these patients and resulted in death. Though quinacrine is effective in 80°/o of patients, reservations do exist in its use due to the high incidence of minor side effects and the small but real potential for serious complication or death.

lntrapleural Tetracycline A readily available and relatively benign agent for local treatment of pleural effusion was introduced by Rubinson and Bolooki [80] in 1972. Drawing from animal studies showing that tetracycline produces pleural adhesions [981, these authors were able to effectively control 10 out of 12 (83%) malignant pleural effusions with thoracostomy drainage and tetracycline instillation. Because this technique can be used inexpensively and conveniently in virtually any hospital setting, tetracycline is popular at present. Unfortunately, the published experience (see Table 3), though encouraging, is relatively small. It includes a prospective randomized trial in which thoracostomy drainage and tetracycline is compared with thoracostomy and quinacrine [531. Although this study is ongoing, a preliminary report [53] confirmed that tetracycline and chest tube drainage controlled better than 80% (10 out of 12) of malignant pleural effusions. While no difference in effectiveness could be seen this early in the study (8 out of 9 patients were controlled with quinacrine), the tetracycline could be given in one dose (versus five doses for quinacrine), produced less pleuritic pain (5 out of 12 versus 7 out of 9 for quinacrine), and caused less fever (4 out of 12 versus 8 out of 9) than quinacrine.

The experience is small for each, however, and at least for thio-TEPA [5, 10, 331 and 5fluorouracil[951the response rates were inferior to other currently used agents. For bleomycin, on the other hand, the experience of Paladine and co-workers [711 demonstrated a marked effectiveness (17 out of 19 patients or 89%) and minimal toxicity that could establish it as an appropriate agent. Further comparison with a more commonly used agent is needed.

Other Techniques of Local TherapyChest Tube Only While most surgeons now agree that thoracostomy drainage is a necessary part of intrapleural therapy of malignant effusion, some claim that chest tube drainage alone is sufficient (see Table 3). In a retrospective analysis, Lambert and colleagues [561 claimed that 19 out of 22 (86%) malignant effusions were successfully controlled by a period of chest tube drainage alone. Anderson and co-workers [6] found this difficult to reproduce when 7 consecutive patients failed to respond. Probably the closest approximation to the true effectiveness of chest tube drainage alone is seen in the prospective study by Izbicki and associates [391. Fifteen (50%) out of 30 effusions in patients with breast cancer were successfully controlled by this technique. Pleural drainage alone, therefore, is partially effective, but the intrapleural administration of certain agents (e.g., talc, quinacrine, tetracycline, or bleomycin) appears to improve the likelihood of an objective response.

Pleurectomy Since clinical evidence suggests that chemically induced pleurodesis will prevent recurrence of malignant pleural effusion, it is natural to expect surgically induced pleurodesis to be equally effective. In fact, since many surgeons consider parietal pleurectomy to be far more effective than chemical pleurodesis in the treatment of recurrent pneumothorax, pleurectomy might also be more effective in controlling recurrent malignant pleural effusion. The experiences of Jensik [43], Martini [641, and their colleagues appear to support this contention Other lntrapleural Agents (see Table 3). There was a 99% objective reOther agents have been introduced intrapleu- sponse rate in their combined experience of 147 rally to relieve malignant effusions (see Table 3). patients. It must, however, be emphasized that

199 Collective Review: Austin and Flye: Recurrent Malignant Pleural Effusion

the most effective means of local control is not necessarily the best. Patients incapacitated by recurrent malignant effusions are, in general, poor operative risks. Indeed, the mortality experience in the studies of Jensik [43], Martini [64], and their associates was 6% and lo%, respectively. In addition, postoperative complications occurred in 23% of patients [641. Pleurectomy, therefore, is hardly a therapy of choice and should not be considered for most patients. Nevertheless, it is conceivable that there might exist a very small subset of patients whose health is otherwise exceptionally good, whose malignancy is otherwise under good control, and in whom several intrapleural agents have been ineffective. In such a rare situation, pleurectomy would be appropriate.

Recommendations The ideal treatment of a recurrent pleural effusion should be 100% effective, safe, and convenient. On the basis of this review, we can rationally choose among the many modalities available. Table 4 is designed to help make this choice. Radiation therapy and bleomycin have not been included due to the currently small published clinical experience. Toxicity and convenience, though very pertinent, are difficult to compare quantitatively from study to study. To facilitate this comparison we have chosen to score the toxicity and convenience of each technique on a scale of 1 to 5. On the basis of effectiveness alone it appears

that intrapleural administration of isotopes or nitrogen mustard, and chest tube drainage alone no longer have a role in the local therapy of recurrent malignant effusion. Pleurectomy on the other hand, while very effective, is associated with undue morbidity and mortality. The remaining three therapies are virtually equal in effectiveness, and the choice of which intrapleural agent to use must be based on comparison of their relative toxicities and inconvenience. Because of its minimal inconvenience and toxicity we favor tetracycline as the intrapleural agent of choice. Although intrapleural instillation of a talc suspension is almost as convenient as that of a tetracycline solution, talc is not as readily available and the side effects are not as well documented. Nevertheless, should initial therapy with intrapleurally administered tetracycline fail or should the patient be allergic to tetracycline, we advocate the use of a talc suspension. Because of its frequent side effects and rare serious complications, we do not recommend the intrapleural use of quinacrine. Whichever agent is used, the effectiveness of local therapy is correlated with the degree of pleurodesis achieved. To maximize the chance of this obliterative process, thoracostomy drainage is mandatory. For the intrapleural agent to be most effective in producing the necessary chemical pleuritis, the pleural space must be fully evacuated before the agent is instilled. After the agent has been fully allowed to distribute over the pleural surfaces, subsequent

Table 4. Overall Comparison of Techniques for Controlling Malignant Pleural Effusions

Technique Tetracycline Talc Chest tube alone Quinacrine Nitrogen mustard Isotopes Pleurectomy

Published Experience

Effectiveness

No. of Studies

% Objective Response

3

No. of Patients

Morbidity"

Inconvenience"

87

2+

2+

90

2+

3+

4

31 59 69

55

1+

1+

8 10

128 338

80 52

4+ 3+

3+

27 2

980 147

55 99

2+ 5+

4+ 5+

2b

aDetermined from studies cited in text and rated on scale from 1 to 5. bExcludesstudies using general anesthesia.

2+

200 The Annals of Thoracic Surgery Vol 28 No 2 August 1979

exudation must be immediately removed so that both pleural surfaces remain apposed. On the basis of our review- of the literature and our own clinical experience, we advocate the following technique for tube thoracostomy and chemical pleurodesis: The lightly sedated patient is placed on his side and the eighth intercostal space is identified at the midaxillary line. After several milliliters of local anesthesia has been injected, a 24F thoracic catheter is inserted at this point and advanced until the last hole in the tube (an extra hole cut before insertion) is inside the pleural space. The tube is sutured in place, and the chest cavity is drained into a waterseal suction device. Complete evacuation of fluid and full lung expansion should be verified before proceeding with the intrapleural agent. Because instillation of either tetracycline or talc may be painful, an injection of meperidine or morphine is administered 1 hour before instillation to minimize this discomfort. If tetracycline is to be used, 500 mg is diluted in 30 to 50 ml of saline solution [go]. For talc 10 gm of USP talc powder previously gas sterilized and aerated is suspended in 250 ml of sterile saline solution (this can easily be done in a urological irrigation set) [21. Just before instillation, the chest tube is clamped. Administration of the tetracycline is done by needle directly into the chest tube. Ten to 20 ml of saline solution are then injected to clear the tube. The talc suspension is instilled with a catheter-tipped syringe directly into the temporarily disconnected chest tube. Regardless of the agent used, the chest tube is then left clamped for 2 hours. During this 2-hour period the patient is placed in several positions including supine and prone in order to distribute the agent over the entire pleural surface. The tube is then unclamped and left on low suction. After instillation, pleural drainage is carefully monitored. When drainage ceases (usually 3 or 4 days), the chest tube is removed.

Summary Pleural effusion is a common complication of malignancy especially in breast and lung cancer, and in the lymphomas. When the dynamic equilibrium of free water and protein in the pleural space is upset by the spillage of

protein from pleural metastasis or by obstruction of lymphatic drainage, large pleural effusions result and marked respiratory embarrassment can ensue. The exudative nature of the effusion can be accurately determined by analyzing the relative protein and LDH concentrations of pleural fluid and serum. The malignant nature of these effusions can be identified in 90% of patients by a combination of pleural fluid cytology and pleural biopsy. Treatment of symptomatic recurrent pleural effusion begins with standard therapy for the advanced malignancy in question. When these treatments fail to control the effusion, local therapy can provide notable palliation. Mediastinal radiation constitutes the best local therapy for effusions secondary to lymphoma, while intrapleural therapy is best for all other malignant effusions. The effectiveness of intrapleural therapy depends more on the sclerotic properties of the instilled agent than on its antineoplastic effect. Between 45 and 55% of patients will be effectively treated by intrapleurally administered isotopes or nitrogen mustard, or simply by chest tube drainage alone. However, 80 to 90% of patients will benefit if tetracycline, talc, or quinacrine are instilled, in addition to use of thoracostomy drainage. Because of its effectiveness, ready availability, convenience, and minimal toxicity, intrapleural administration of tetracycline plus chest tube drainage is the therapy of choice for symptomatic recurrent malignant pleural effusion.

References 1. Adler RH, Rappole BW: Recurrent malignant pleural effusions and talc powder aerosol treatment. Surgery 62:1000, 1967 2. Adler RH, Sayek I: Treatment of malignant pleural effusion: a method using tube thoracostomy and talc. Ann Thorac Surg 22:8, 1976. 3. Agostoni E: Mechanics of the pleural space. Physiol Rev 52:57, 1972 4. Albertelli J, Monaco H, O’Connor T, et al: Nitrogen mustard in serosal cavities in treatment of advanced malignant tumors. Prensa Med Argent 41:201, 1954 5. Anderson AP, Brincker H: Intracavitary thioTEPA in malignant pleural and peritoneal effusions. Acta Radio1 [Ther] (Stockh) 7:369, 1968 6. Anderson CB, Philpott GW, Ferguson TB: The

201 Collective Review: Austin and Flye: Recurrent Malignant Pleural Effusion

treatment of malignant pleural effusions. Cancer 33:916, 1974 7. Andrews GA, Root SW, Kerman HD, et al: Intracavitary colloidal radiogold in the treatment of effusion caused by malignant neoplasms. Ann Surg 137:375, 1953 8. Ariel lM, Oropeza R, Pack GT: Intracavitary administration of radioactive isotopes in the control of effusions due to cancer. Cancer 19:1096, 1966 9. Bass BH: Nitrogen mustard in the palliation of lung cancer. Br Med J 1:617, 1960 10. Bateman JC, Moulton B, Larsen NJ: Control of neoplastic effusion by phosphoramide chemotherapy. AMA Arch Intern Med 95:73, 1955 11. Bethune N: Pleural poudrage-a new technique for the deliberate production of pleural adhesions as a preliminary to lobectomy. J Thorac Surg 4:251, 1935 12. Black LF: The pleural space and pleural fluid. Mayo Clin Proc 47:493, 1972 13. Bonte FJ, Storaasli JP, Weisberger AS: Comparative evaluation of radioactive colloidal gold and nitrogen mustard in the treatment of serous effusions of neoplastic origin. Radiology 67:63, 1956 14. Borda 1, Krant M: Convulsions following intrapleural administration of quinacrine hydrochloride. JAMA 201:173, 1967 15. Borja ER, Pugh RP: Single-dose quinacrine (Atabrine) and thoracostomy in the control of pleural effusion in patients with neoplastic disease. Cancer 31:899, 1973 16. Botsford TW: Experiences with radioactive colloidal gold in the treatment of pleural effusion caused by metastatic cancer of the breast. N Engl J Med 270:552, 1964 17. Braun EM, Schlang HA: Large-tube drainage of malignant pleural effusion. NY State J Med 75:1550, 1975 18. Brennan MJ, McGinnis KD, Preuss L: Experience with intracavitary radiogold at Henry Ford Hospital. Henry Ford Hosp Med Bull 4539,1956 19. Bruneau R, Rubin P: The management of pleural effusions and chylothorax in lymphoma. Radiology 85:1085, 1965 20. Camishion RC, Gibbon JH, Nealon TF Jr: Talc poudrage in the treatment of pleural effusion due to cancer. Surg Clin North Am 42:1521,1962 21. Card RY, Cole DR, Henschke UK: Summary of ten years of the use of radioactive colloids in intracavitary therapy. J Nucl Med 1:195, 1960 22. Carr DT, Power MH: Clinical value of measurements of concentration of protein in pleural fluid. N Engl J Med 259:926, 1958 23. Chambers JS: Palliative treatment of neoplastic pleural effusion with intercostal intubation and talc instillation. West J Surg Obstet Gynecol 66:26, 1958

24. Chang CH, Janzen AH, Skorneck AB, et al: Treatment of malignant effusions by intracavitary injection of radioactive colloidal gold. Am J Roentgen01 77:486, 1957 25. Courtice FC, Simmonds WJ: Absorption of fluids from the pleural cavities of rabbits and cats. J Physiol (Lond) 109:117, 1949 26. Dollinger MR: Management of recurrent malignant effusions. CA 22:138, 1972 27. Dollinger MR, Krakoff IH, Karnofsky DA: Quinacrine (Atabrine") in the treatment of neoplastic effusions. Ann Intern Med 66:249, 1967 28. Dybicki J, Balchum OJ, Meneely GR: Treatment of pleural and peritoneal effusion with intracavitary colloidal radiogold (Au'"')). AMA Arch Intern Med 104:140, 1959 29. El-Mahdi A, Levene A, Lott S: Observations and management of malignant pleural effusions in breast carcinoma. Johns Hopkins Med J 132:44, 1973 30. Engel GL: Quinacrine effects on the central nervous system. JAMA 197:515, 1966 31. Farber SM: Lung Cancer. Springfield, Thomas, 1954 32. Fisher AMH, Kendall B, Van Leuven BD: Hodgkin's disease: a radiological survey. Clin Radiol 13:115, 1962 33. Fracchia AA, Knapper WH, Carey JT, et al: Intrapleural chemotherapy for effusion from metastatic breast carcinoma. Cancer 26:626, 1970 34. Fullerton CW, Reed PI: Nitrogen mustard in treatment of pleural and peritoneal effusions. Can Med Assoc J 79:190, 1958 35. Gellhorn A, Zaidenweber J, Ultmann J, et al: The use of Atabrine (quinacrine) in the control of recurrent neoplastic effusions. Dis Chest 39:165, 1961 36. Hahn PF, Meneely GR, Carothers EL: The use of gold and silver-coated radioactive gold colloids in the palliation of ascites and pleural effusions. Br J Radiol 31:240, 1958 37. Haupt GJ, Camishion RC, Templeton JY 111, et al: Treatment of malignant pleural effusion by talc poudrage. JAMA 172:918, 1960 38. Hickman JA, Jones MC: Treatment of neoplastic pleural effusions with local instillation of quinacrine (mepacrine) hydrochloride. Thorax 25:226, 1970 39. Izbicki R, Weyhing BT 111, Baker L, et al: Pleural effusion in cancer patients-a prospective randomized study of pleural drainage with the addition of radioactive phosphorous to the pleural space vs pleural drainage alone. Cancer 36: 1511, 1975 40. Jacobs ML: Radioactive colloidal chromic phosphate to control pleural effusion and ascites. , JAMA 166:597, 1958 41. Jaffe HL: Treatment of malignant serous effu-

202 The Annals of Thoracic Surgery Vol 28 No 2 August 1979

42.

43.

44.

45.

46.

47.

48. 49.

50.

51.

52.

53.

54.

55.

56.

57.

sions with radioactive colloidal chromic phosphate. Am J Roentgenol 74:657, 1955 Jarvi OH, Kunnas RJ, Laitio MT, et al: The accuracy and significance of cytologic cancer diagnosis of pleural effusions. Acta Cytol (Baltimore) 16:152, 1972 Jensik R, Cagle JE Jr, Milloy F, et al: Pleurectomy in the treatment of pleural effusion due to metastatic malignancy. J Thorac Cardiovasc Surg 46:322, 1963 Johnson LP, Riokin SE, Weber EL: Palliation in malignant pleural effusion. Am Surg 41:529, 1975 Jones GR: Treatment of recurrent malignant pleural effusion by iodized talc pleurodesis. Thorax 24:69, 1969 Karnofsky DA, Abelmann WH, Craver LF, et al: The use of the nitrogen mustards in the palliative treatment of carcinoma: with particular reference to bronchogenic carcinoma. Cancer 1:634, 1948 Karras BG, Moss WT: Radioactive gold in the control of malignant serous effusions. Q Bull Northwest Univ Med Sch 36:57, 1962 Kaunitz J: Landmarks in simple pleural effusions. JAMA 113:1312, 1939 Kent EM, Moses C: Radioactive isotopes in the palliative management of carcinomatosis of the pleura. J Thorac Surg 22:503, 1951 Kent EM, Moses C, Ford WB, et al: Radioactive isotopes in management of carcinomatosis of serous body cavities. Arch Intern Med 94:334, 1954 King ER, Spicer DW, Dowda FW, et al: The use of radioactive colloidal gold (AU''~)in pleural effusions and ascites associated with malignancy. Am J Roentgenol 68:413, 1952 Kinsey DL, Carter D, Klassen KP: Simplified management of malignant pleural effusion. Arch Surg 89:389, 1964 Kisner D, Bayly T, Sybert A, et al: Tetracycline and quinacrine in the control of malignant pleural effusion: a randomized clinical trial. ASCO Abstracts C-93, 1977 Kligerman MM, Habif DV: The use of radioactive gold in the treatment of effusion due to carcinomatosis of the pleura and peritoneum. Am J Roentgenol 74:651, 1955 Kniseley RM, Andrews GA: Pathological changes following intracavitary therapy with colloidal Au-198. Cancer 6:303, 1953 Lambert CJ, Shah HH, Urschel HC Jr, et al: The treatment of malignant pleural effusions by closed trocar tube drainage. Ann Thorac Surg 3:1, 1967 Lambrethsen E, Sell A: Palliative treatment of carcinomatosis effusion in the pleura and peritoneal cavities with radioactive gold. e t a Radio1 (Stockh) 56:33, 1961

58. Leininger BJ, Barker WL, Langston HT: A simplified method for management of malignant pleural effusion. J Thorac Cardiovasc Surg 58:758, 1969 59. Leuallen EC, Carr DT: Pleural effusion-a statistical study of 436 patients. N Engl J Med 252:79, 1955 60. Levison VB: Nitrogen mustard in palliation of malignant effusions. Br Med J 1:1143, 1961 61. Light RW: Pleural effusions. Med Clin North Am 61:1339, 1977 62. Light RW, MacGregor MI, Luchsinger PC: Pleural effusions: the diagnostic separation of transudates and exudates. Ann Intern Med 77:507, 1972 63. Mark JBD, Goldberg IS, Montague AC: Intrapleural mechlorethamine hydrochloride therapy for malignant pleural effusion. JAMA 1872358, 1964 64. Martini N, Bains MS, Beattie EJ Jr: Indications for pleurectomy in malignant effusion. Cancer 35:734, 1975 65. Millar 0, MacDonald JCF: Radioactive gold in malignant effusions. Can Med Assoc J 74:783, 1956 66. Molander DW, Pack GT: Lymphosarcoma: choice of treatment and end-results in 567 patients: role of surgical treatment for cure and palliation. Rev Surg 20:3, 1963 67. Muller JH: Zur medizinisch-therapeutischen: Verwendung der Kunstlichen Radioaktivitat. Bull Schweiz Akad Med Wiss 5:484, 1949 68. OBryan RM, Talley RW, Brennan MJ, et al: Critical analysis of the control of malignant effusions with radioisotopes. Henry Ford Hosp Med J 16:3, 1968 69. Osbourne MP, Copeland BE: Intracavitary administration of radioactive colloidal gold (AU'~') for treatment of malignant effusions. N Engl J Med 255:1122, 1956 70. Pack GT, Molander DW: The surgical treatment of Hodgkin's disease. Cancer Res 26:1254, 1966 71. Paladine W, Cunningham TJ, Sponzo R, et al: Intracavitary bleomycin in the management of malignant effusions. Cancer 38:1903, 1976 72. Papac RJ: Treatment of malignant disease in closed spaces, in Cancer, A Comprehensive Treatise: Chemotherapy. Edited by FF Becker. New York, Plenum, 1977, vol5, p p 207-222 73. Pearson FG, MacGregor DC: Talc poudrage for malignant pleural effusion. J Thorac Cardiovasc Surg 51:732, 1966 74. Rigler LG: Roentgen diagnosis of small pleural effusions: new roentgenographic position. JAMA 96:104, 1931 75. Roche G, Delanoe Y, Moayer N: Talcage de la plhvre sous pleuroscopie: resultats, indications, technique. J Fr Med Chir Thyrac 1:677, 1963 76. Rochlin DB, Smart CR, Wagher DE, et al: The

203 Collective Review: Austin and Flye: Recurrent Malignant Pleural Effusion

control of recurrent malignant effusions using quinacrine hydrochloride. Surg Gynecol Obstet 118:991, 1964 77. Rosato FE, Wallach MW, Rosato EF: The management of malignant effusions from breast cancer. J Surg Oncol 6:441, 1974 78. Rose RG: Intracavitary radioactive colloidal gold: results in 257 cancer patients. J Nucl Med 3:323, 1962 79. Rosenberg SA, Diamond HD, Jaslowitz B, et al: Lymphosarcoma: a review of 1269 cases. Medicine 40:31, 1961 80. Rubinson RM, Bolooki H: Intrapleural tetracycline for control of malignant pleural effusion: a preliminary report. South Med J 65:847, 1972 81. Salyer WR, Eggleston JC, Erozan YS: Efficacy of pleural needle biopsy and pleural fluid cytopathology in the diagnosis of malignant neoplasm involving the pleura. Chest 67:536, 1975 82. Seal SH, Crosignani S, Valrassori G, et al: The treatment of malignant effusions with radioactive colloidal gold (Au"'). Am J Obstet Gynecol 75:1027, 1958 83. Seaman WB, Sherman AI, Bonebrake M: Radioactive gold i n the treatment of malignant effusions. JAMA 153:630, 1953 84. Shea FP: Nitrogen mustard for control of malignant effusions. Harper Hosp Bull 20:239, 1962 85. Shedbalkar AR, Head JM, Head LR, et al: Evaluation of talc pleural symphysis in the management of malignant pleural effusion. J Thorac Cardiovasc Surg 61:492, 1971 86. Siege1 EP, Hart HE, Brothers M, et al: Radioyttrium (""Y) for the palliative treatment of effusions due to malignancy. JAMA 161:499, 1956 87. Simon N: Radioactive gold treatment: results in 85 effusions due to cancer. J Mt Sinai Hosp 22:96, 1955 88. Starkey GW: Recurrent malignant pleural effusions. N Engl J Med 270:436, 1964 89. Starling EH: On the absorption of fluids from the connective tissue spaces. J Physiol (Lond) 19:312, 1898 90. Stewart PB: The rate of formation and lymphatic removal of fluid in pleural effusions. J Clin Invest 42:258, 1963 91. Stewart PB. Bureen AS: The turnover of fluid in "

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102. 103.

104.

105.

106.

the dog's pleural cavity. J Lab Clin Med 52:212, 1958 Storaasli JP, Bonte FJ, King DP, et al: The use of radioactive colloidal gold in the treatment of serous effusions of neoplastic origin. Surg Gynecol Obstet 96:707, 1953 Storey DD, Dines DE, Coles DT: Pleural effusion: a diagnostic dilemma. JAMA 236:2183, 1976 Strober SJ, Klotz E, Kuperman A, et al: Malignant pleural disease-a radiotherapeutic approach to the problem. JAMA 226:296, 1973 Suhrland LG, Weisberger AS: lntracavitary 5-fluorouracil in malignant effusions. Arch Intern Med 116:431, 1965 Taylor L: A catheter technique for intra-pleural administration of alkylating agents: a report of ten cases. Am J Med Sci 244:706, 1962 Taylor SA, Hooton NS, MacArthur AM: Quinacrine in the management of malignant pleural effusion. Br J Surg 64:52, 1977 Thorsrud GK: Pleural reactions to irritants-an experimental study with special references to pleural adhesions and concrescence in relation to pleural turnover of fluid. Acta Chir Scand [Suppl] 355:9, 1965 Tinney WS, Olsen AM: The significance of fluid i n the pleural space: a study of 274 cases. J Thorac Surg 14:248, 1946 Ultmann JE, Gellhorn A, Osnos M, et al: The effect of quinacrine on neoplastic effusions and certain of their enzymes. Cancer 16:203, 1963 Vieta JO, Craver LF: Intrathoracic manifestations of lymphomatoid diseases. Radiology 37:138, 1941 Wallach HW: Intrapleural tetracycline for malignant pleural effusions. Chest 68:510, 1975 Walter J: Radioactive gold in malignant effusions. Proc R SOCMed 46:466, 1953 Walton RJ, Sinclair WK: Intracavitary irradiation with radioactive colloidal gold in the palliative treatment of malignant pleural and peritoneal effusions. Br Med Bull 8:165, 1952 Weick JK, Kiely JM, Harrison EG Jr, et al: Pleural effusion in lymphoma. Cancer 31:848, 1973 Weisberger AS, Levine B, Storaasli JP: Use of nitrogen mustard in treatment of serous effusions of neoplastic origin. JAMA 159:1704, 1955