Management of malignant pleural effusions

INTRODUCTION — Approximately 15 percent of all patients with cancer develop malignant pleural effusions (MPEs) [1], with lung cancer and breast cancer accounting for 50 to 65 percent of MPEs [2]. MPEs can also complicate malignant mesothelioma, metastatic cancer (eg, from distant sites such as stomach or ovary), lymphoma, and other hematologic malignancies. The presence of an MPE usually portends a poor prognosis. More than 90 percent of patients with mesothelioma present with malignant effusion.

The evaluation and management of MPE is discussed in this topic. Much of the content of this topic does not apply to patients with mesothelioma. Our approach is, in general, consistent with that outlined by the American Thoracic Society/Society of Thoracic Surgeons/Society of Thoracic Radiology (ATS/STS/STR), and the European Respiratory Society/European Association of Cardiothoracic Surgery (ERS/EACTS) [1,3]. The diagnosis of MPE and the management of refractory nonmalignant pleural effusions, are discussed in detail separately. (See "Pleural fluid analysis in adults with a pleural effusion" and "Management of nonmalignant pleural effusions in adults" and "Selection of modality for diagnosis and staging of patients with suspected non-small cell lung cancer", section on 'Pleural (T2, T3, M1a)' and "Procedures for tissue biopsy in patients with suspected non-small cell lung cancer", section on 'Suspected pleural metastases'.)

DEFINITION (MALIGNANT/PARAMALIGNANT) — In MPEs, infiltration of cancer cells into pleural tissue is observed; pleural tissue invasion by malignant cells may be seen on pleural biopsy and may result in positive fluid cytology. The diagnosis of MPE is discussed separately. (See "Pleural fluid analysis in adults with a pleural effusion" and "Selection of modality for diagnosis and staging of patients with suspected non-small cell lung cancer", section on 'Pleural (T2, T3, M1a)' and "Procedures for tissue biopsy in patients with suspected non-small cell lung cancer", section on 'Suspected pleural metastases'.)

In contrast, paramalignant effusions result from tumor effects that indirectly act on the pleural space such as by bronchial obstruction, mediastinal lymph node infiltration, thromboembolism, or superior vena cava syndrome. In paramalignant pleural effusions, pleural fluid cytology and pleural biopsy are negative because cancer cells have not invaded pleural membranes. Paramalignant effusions are managed according to the underlying etiology.

TREATMENT GOALS — The presence of an MPE typically signifies advanced stage cancer and usually indicates that death will likely occur within an average of four to seven months, although rare exceptions of prolonged survival exist (eg, lymphoma). Thus, in general, the focus of MPE management is palliative and without survival benefit. Treatment should focus on patient-centered goals of therapy, which include sustained symptom relief, improvement in quality of life, acceptability of an intervention, affordability, and preference for less invasive procedures [4,5]. (See "Assessment and management of dyspnea in palliative care" and "Approach to symptom assessment in palliative care".)

The decision to use a pleural intervention to initially drain an MPE or prevent recurrence depends upon the presence of respiratory symptoms, which occur in 85 percent of patients and among which dyspnea is the most common. Chest discomfort and cough are less common. Determining symptomatology due to a pleural effusion can be challenging, especially when there is severe underlying lung disease or tumor. In most cases, the determination may be assessed clinically and confirmed by demonstrating symptom improvement or relief following initial therapeutic thoracentesis. We suggest the following treatment strategy:

●Symptomatic MPE – Symptomatic MPEs are usually moderate to large in volume although the severity of respiratory symptoms does not correlate with the size of the effusion because other factors, such as coexisting cardiac and pulmonary disorders, contribute to dyspnea [6]. Symptomatic effusions should be drained (eg, large volume thoracentesis) for palliative symptom relief. Although concern exists that prolonged observation of these effusions without drainage results in a nonexpandable lung that may complicate palliative benefit from later drainage, no data establish an association between timing of initial drainage of symptomatic effusions and the occurrence of lung entrapment [3]. Unfortunately, symptomatic malignant effusions usually recur after initial drainage. Subsequent management depends on multiple patient-centered factors, such as prognosis, ability to tolerate various interventions, and patient wishes.

●Asymptomatic MPE – Asymptomatic MPEs are usually small in volume and do not require therapeutic drainage since early drainage does not prevent progression of the effusion to lung entrapment [3]. Two retrospective studies address this concern and found that lung cancer patients with small, asymptomatic effusions do not require pleural interventions during the course of their disease [7,8]. However, these findings may not extend to patients with MPE due to malignancies other than lung cancer. Regardless, nearly all MPEs eventually become symptomatic and require an intervention, such that continued clinical and radiologic surveillance is prudent. This approach is particularly reasonable in MPEs due to some chemoresponsive tumor types, such as breast, ovarian, or prostate cancer, small cell carcinoma of the lung, germ cell tumors, or lymphomas, where antitumor therapy may result in resolution of the effusion without mechanical intervention.

INITIAL TREATMENT — Options for managing an MPE are shown in the table (table 1) [9,10]. Choosing among these options depends upon multiple factors that include patient prognosis, patient performance status, effusion cancer recurrence rate, lung expansion, and responsiveness of the malignancy to therapy. For most patients with symptomatic MPE, we recommend an initial therapeutic thoracentesis rather than no therapy (algorithm 1) [1,3]. Thoracentesis should be performed under ultrasound guidance. Less commonly, chest tube drainage is performed. Regardless of the drainage procedure, patients should also be concomitantly treated for their underlying malignancy, if indicated. (See 'Therapeutic thoracentesis' below and 'Chest tube or catheter drainage' below and 'Treatment of the underlying malignancy' below.)

While many patients initially respond to this approach, symptomatic recurrence of MPE is common. Most patients, therefore, will eventually require a more definitive intervention such as pleurodesis or an intrapleural catheter (IPC) if symptoms due to the effusion decrease quality of life. The high likelihood of recurrence is why some experts combine initial thoracentesis with some form of pleurodesis or prolonged IPC drainage at the time of presentation (eg, pleurodesis during diagnostic thoracoscopy or following catheter placement). However, the difficulty of predicting short term survival supports the majority opinion that a pleurodesis or IPC drainage should be delayed for recurrence of the effusion when patient survivability can be re-evaluated. (See 'Pleural interventions for recurrence' below.)

Therapeutic thoracentesis — Therapeutic thoracentesis is the first-line treatment for symptomatic MPE and is typically performed using a needle/catheter under ultrasound guidance. Thoracentesis determines the symptomatic response to drainage, the ability of the lung to re-expand completely, and the rate of subsequent reaccumulation, all of which inform future more definitive treatments should the effusion reaccumulate.

We generally remove as much fluid as is tolerated up to a limit of 1.5 L because of concern that volumes removed above this amount increases the risk for serious procedure-related complications [11]. Although removal of large volumes is frequently necessary, removal of smaller volumes may provide relief in those with underlying lung disease. We advise monitoring symptoms of chest discomfort and cough carefully during the procedure to guide the volume of fluid to be removed and assess whether or not the underlying lung is expandable. Limited evidence suggests no benefit to routine use of pleural manometry in those with free flowing effusions [12]; thus, pleural manometry may be reserved for those in whom nonexpandable lung is suspected. These data are discussed separately. (See "Large volume (therapeutic) thoracentesis: Procedure and complications" and "Measurement and interpretation of pleural pressure (manometry): Indications and technique".)

Chest tube or catheter drainage — Chest tube or small-bore catheter drainage (ie, the tube or catheter is left in place) is an alternative to large volume thoracentesis and may be performed in those who need especially large volumes of fluid removed slowly (over hours or days) for symptom relief.

Treatment of the underlying malignancy — In most patients, MPE does not respond to, or recurs despite, antitumor therapy. However, in some tumor types, treating the underlying malignancy may be effective in reducing the size of or resolving an MPE and preventing its recurrence (eg, chemotherapy, protein kinase inhibitors, monoclonal antibodies, and radiation). Examples of antitumor-responsive malignancies include breast, ovarian, and prostate cancer, germ cell tumors, lymphoma, and small cell lung cancer. However, if symptoms from the effusion are burdensome, definitive pleural interventions, such as drainage or pleurodesis, should not be delayed for systemic cancer therapy.

●Systemic antitumor drug therapy – With the exception of chemoresponsive tumors, the response of MPEs to systemic chemotherapy is disappointing for most malignancies, especially MPEs due to non-small cell lung cancer (NSCLC). However, preliminary reports suggest that some MPEs due to NSCLC may respond to bevacizumab, a recombinant monoclonal antibody directed against vascular endothelial growth factor (VEGF) [13-15] or to epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs) [16,17]. In one study of 39 patients with the EGFR mutation who presented with an MPE, time to recurrence was the same with EGFR-TKI therapy alone compared with EGFR-TKIs plus pleurodesis (11.7 versus 9.9 months) [18]. However, most patients who present with EGFR mutant NSCLC develop resistance to therapy within one year [16], after which rapid recurrence of the MPE can occur; the addition of bevacizumab to ongoing EGFR-TKI therapy may result in MPE control in this group [19]. Zoledronic acid, initially felt to have some potential role in MPEs [20] was not found to be useful in a small feasibility randomized controlled trial [21].

●Radiation therapy – Radiation therapy directed at the primary tumor may be helpful in resolving MPEs when mediastinal lymph node disease predominates (eg, lymphoma).

●Intrapleural agents – Preliminary phase I and II studies with nonsclerosant intrapleural chemotherapeutic agents have shown a partial or complete response of MPE in the short-term but this approach has not been directly compared with standard therapy [22-27]. Further randomized trials are needed with comparisons to standard care before such an approach can be recommended.

PLEURAL INTERVENTIONS FOR RECURRENCE — Despite initial thoracentesis and antitumor therapy, recurrence of symptomatic pleural effusions is common. Over half of MPEs recur and among those, up to two-thirds recur rapidly within one month while the remainder recur more slowly [28]. Options include repeat thoracentesis, indwelling pleural catheter (IPC) drainage, pleurodesis by various techniques (bedside via chest tube or catheter, thoracoscopic), combination therapies, complete or partial pleurectomy with decortication, and pleuroperitoneal shunt (table 1 and algorithm 1). Definitive procedures (other than repeat thoracentesis) generally result in fewer subsequent pleural procedures, emergency department visits, hospital admissions, and pneumothoraces [28,29].

Choosing among the options — The optimal option for patients with recurrent MPE is unclear. Considerable variation exists in practice with guideline groups and many other experts suggesting proceeding directly to IPC drainage or to pleurodesis rather than repeated aspiration [1,3,30]. Our general preference is for IPC for the reasons outlined below (see 'Indwelling pleural catheter (IPC)' below). However, in reality several factors weigh in to the decision such that choosing the optimal option should be tailored to the individual patient. These factors include the following:

●Presence of expandable or nonexpandable lung – Patients with expandable lung are candidates for several therapies (IPC drainage and/or pleurodesis) while those with nonexpandable lung are typically limited to IPC drainage only. (See 'Expandable lung with rapid reaccumulation' below and 'Nonexpandable lung' below.)

●Rate of reaccumulation – Patients with rapid reaccumulation typically need drainage and/or pleurodesis while those with slower rates of reaccumulation may be suitable for repeat thoracentesis or IPC drainage. Predictors are unclear but one study suggested that a large pleural effusion and high daily drainage may predict reaccumulation [31]. (See 'Expandable lung with rapid reaccumulation' below and 'Expandable lung with slow reaccumulation' below.)

●Predicted responsiveness of the MPE to antitumor therapy – Avoiding pleurodesis may be appropriate in those with tumor types that are typically responsive to antitumor therapy (eg, breast, ovarian, and prostate cancer, germ cell tumors, lymphoma, and small cell lung cancer). IPC drainage or pleurodesis may be appropriate should the patients fail antitumor therapy. (See 'Treatment of the underlying malignancy' above.)

●Patient prognosis and functional status – Minimally invasive approaches are often preferred in those with poor prognosis or poor functional status. Predicting survival, including using the LENT score, is discussed below. (See 'Prognosis' below.)

●Severity of symptoms – Patients with severe symptoms should be drained while those who are asymptomatic may not need drainage.

●Institutional resources, expertise, and physician preference – IPC requires interventional expertise and may not be available in some institutions. In addition, practice may vary with IPC placement being more popular in the US while pleurodesis may be more popular in Europe.

●Patient preferences – Patient preferences weigh heavily on the decision when choosing between IPC drainage and chemical pleurodesis. As examples, patients with a relatively longer anticipated survival (eg, longer than two months) may prefer chemical pleurodesis due to a desire for an effective treatment that can be carried out in a single definitive procedure, rather than living with the inconvenience of an indwelling catheter that requires intermittent drainage and is in place for several weeks. Other patients may choose indwelling catheter drainage to avoid risks of pleurodesis (eg, local pain, low-grade fever, chest infections, and rarely respiratory distress) and the increased time in the hospital [32]. Some patients may choose to perform talc pleurodesis via an indwelling catheter as soon as it is inserted to limit the duration of catheter insertion and minimize return visits or hospitalizations. Informing patients of local facility and professional charges for each approach help patients state a preference based on their copayments and other financial considerations.

Expandable lung with rapid reaccumulation — For patients with rapidly recurrent MPE (ie, within one month of the initial thoracentesis) who have expandable lung, we prefer an IPC. However, pleurodesis is also an acceptable alternative. Repeat thoracentesis is not typically considered unless the prognosis is grim (eg, <2 weeks) or the patient declines IPC and pleurodesis. Choosing among these options is discussed above. (See 'Choosing among the options' above.)

Indwelling pleural catheter (IPC) — Placement of an IPC (also known as a tunneled pleural catheter), with intermittent outpatient drainage by the patient or a patient attendant, is our preferred initial step for most patients with recurrent MPE, including those with expandable or nonexpandable lung. Several randomized trials [32-36] and meta-analyses [37,38] report that IPC provides good symptom relief and can induce spontaneous pleurodesis; while successful pleurodesis is reported to occur in up to 70 percent, this success rate has not been consistent among studies and in our experience and that of others, successful pleurodesis occurs in up to 40 percent [39,40]. One network meta-analysis of three studies reported that patients treated with an IPC were less likely to require a repeat invasive pleural intervention when compared with patients treated initially with talc slurry (odds ratio 0.25, 95% CI 0.13-0.48) [38]. Additional advantages are that IPC requires little if any time in the hospital, as the catheter may be placed during an outpatient procedure [37,41-47]. IPC is best suited to those with a shorter survival (eg, <2 months) and for patients who prefer outpatient management and a minimally invasive intervention. An IPC may also be used in patients who fail pleurodesis as an initial therapy, and it does not preclude subsequent pleurodesis if desired. In addition, combination therapy with IPC plus talc pleurodesis is possible; this procedure results in higher rates of pleurodesis compared with IPC alone, and can be performed on an outpatient basis without risks of IPC obstruction with talc [39]. The main disadvantage of IPC placement is the longer dwell time of the catheter (ie, weeks as compared with days for pleurodesis) and infection. In addition, it requires that the patient or a caregiver perform drainage at home. IPC also requires specific expertise and is typically performed by a thoracic surgeon, interventional pulmonologist, or radiologist.

In most patients, an IPC remains in place for a few weeks (eg, two to six weeks) but may be left in place for longer, provided no indications for removal are present (eg, empyema, blockage). Following placement, patients or their caregiver perform daily drainage. However, drainage every other day, or drainage as needed may be appropriate in some patients (eg, catheter removal is unlikely or unimportant). The volume or frequency of drainage is expected to dwindle over the ensuing few weeks (typically two to six weeks). There is no difference in symptom relief with an aggressive compared with an every other day drainage strategy, although spontaneous pleurodesis may be more rapid with an aggressive strategy [40]. If drainage ceases either due to a response to antitumor treatment and/or spontaneous pleurodesis from the IPC, the device can be removed. While arbitrary, the traditional criterion for IPC removal is <50 mL drainage for three consecutive occurrences, typically >24 hours apart. For patients who demonstrate a continued need for symptomatic drainage, continued drainage through a functioning IPC is appropriate. Pleurodesis (via the IPC itself) may be offered to those with continued drainage who decline prolonged treatment with an IPC. (See 'Pleurodesis (alternative to IPC or failed IPC)' below.)

IPCs appear to be appropriate in patients undergoing chemotherapy [48,49] and in those with MPEs due to hematologic malignancies [50].

Postinsertion management of the IPC requires a careful treatment plan based on the goals of therapy. A clinical practice guideline outlining management recommendations has been published [51].

Silver-nitrate impregnated IPC have been proposed for pleurodesis in patients with MPE, but data suggest no difference in successful pleurodesis when compared with IPC alone [52].

Efficacy — Data to support IPC as the preferred therapy for recurrent MPE include the following:

●Length of stay and reduced interventions – A meta-analysis of five trials that included 545 patients confirmed that compared with pleurodesis (mostly bedside via chest tube), IPCs resulted in a shorter hospital length of stay and reduced frequency of repeat interventions without any survival benefit [37]. However, potentially diminishing the importance of these outcomes, a shorter length of hospital stay may not be clinically meaningful to patients and the advantage of reduced interventions does not take in to account the frequency of drainage performed at home with IPC. In addition, three major randomized trials (included in the meta-analysis) found no difference in post-procedure dyspnea, chest pain, or quality of life scores between the groups [32-34].

●Spontaneous pleurodesis – Spontaneous pleurodesis may occur in 27 to 70 percent of patients after 2 to 12 weeks of IPC drainage [53-56]. However, our experience and that of others is in keeping with rates at the lower to mid end of this range (approximately 40 percent) [39,40]. As an example, a randomized trial of 149 patients with MPEs treated with IPCs demonstrated that daily catheter drainage of 1 liter of fluid, as compared with every other day drainage of 1 liter resulted in more frequent occurrence of spontaneous pleurodesis (47 versus 24 percent, respectively) and a shorter median time to pleurodesis (54 versus 90 days) [40]. Another study reported higher rates of pleurodesis with lymphoma and ovarian cancer, but reduced rates with gastrointestinal cancers and hydropneumothorax [56].

●Cost effectiveness – One study examined the cost-effectiveness of IPCs as compared with pleurodesis and found that indwelling catheters were more cost-effective in patients with survival less than three months while talc pleurodesis was more cost-effective in those with longer survival [33]. These findings require validation in other studies that include patients with a wide spectrum of underlying tumors and in different health care settings.

Complications — Complications include bleeding, catheter blockage, catheter fracture on removal, tract metastases, and, in particular, infection at the point of insertion, along the catheter tract, or within the pleural space [33,34,37,57-59].

●Infection – In one meta-analysis of five trials, IPCs resulted in an increased risk of cellulitis compared with pleurodesis (risk ratio 5.83; 95% CI 1.56-21.8) [37]. When suspected, pleural fluid should be sent for microbiological Gram stain and culture; whether pleural fluid is obtained via direct pleural fluid sampling or via the IPC is unknown. In general, most patients with infected IPCs can be treated with antibiotics (with coverage for typical skin pathogens [eg, S. aureus, Gram negative rods]) without removal of the catheter. The catheter may need to be removed in patients whose catheter infection fails to respond to antibiotics, patients with concomitant tunnel tract infection, or patients with evidence of empyema or catheter-related sepsis. One retrospective study demonstrated a low incidence of infection from IPCs (5 percent) with an overall mortality risk from infections of 0.29 percent [46]. Fifty-four percent of the infections in this study were treated with antibiotics without removal of the chest catheter. Pleurodesis occurred in 62 percent of patients who experienced a catheter-related infection [46]. (See "Diagnosis and management of pleural causes of nonexpandable lung".)

●Loculations – Some patients develop catheter infection-related or disease-related loculations that may be amenable to intrapleural fibrinolytic therapy, the details of which are discussed below. (See 'Nonexpandable lung' below.)

●Catheter fracture – Catheter fracture can occur during removal. One study reported an incidence of 10 percent (6 of 61 patients) [60]. Four of the six patients had retained catheter fragments in the pleural space, no complications were reported during a mean follow up of 459 days suggesting that aggressive retrieval may not be necessary. In our experience and that of others [61], this is a rare complication (<1 percent); advances in catheter design, appropriate technique in placement (eg, cuff 1 cm from the exit site and tunnel of 5 cm long), and experience with IPC removal have likely reduced this complication.

●Other – Other complications include those generally associated with catheter placement including bleeding and pneumothorax. Suspected catheter obstruction should be treated by flushing the catheter and in some cases, instillation of fibrinolytics. Failure of the latter to improve IPC flow is an indication for removal. (See "Management and prognosis of parapneumonic pleural effusion and empyema in adults", section on 'Intrapleural tPA +/- additional drainage'.)

Pleurodesis (alternative to IPC or failed IPC) — For patients with rapidly recurrent MPE (ie, within one month), pleurodesis is an appropriate alternative to IPC alone. It is particularly suitable for patients with expandable lung who have a reasonable survival (eg, >2 to 6 months) and/or who prefer a more definitive and rapid procedure (ie, a "one and done") that minimizes future interventions. Pleurodesis may also be a first-line option in institutions without expertise in IPC placement or in patients who decline or fail IPC. Pleurodesis is not suitable for those with a short anticipated survival (eg, <2 months).

Pleurodesis refers to obliteration of the pleural space by the induction of pleural inflammation and fibrosis using a sclerosant (usually talc) or manual abrasion (mechanical). Chemical pleurodesis may be performed through an IPC or chest tube (ie, medical pleurodesis), or thoracoscopically (ie, surgical pleurodesis). In contrast, mechanical pleurodesis is only performed thoracoscopically.

There is significant variation in practice regarding the type of pleurodesis (chemical versus mechanical; medical versus surgical), method of delivery (eg, via an IPC or chest tube), chest tube size (large- versus small-bore), sclerosing agent used (talc versus other), choice of analgesic agent (opiate versus nonsteroidal anti-inflammatory [NSAID]), and duration of chest tube drainage following pleurodesis. Choosing among these options is often institution- and operator-dependent. Based upon available data derived from retrospective studies, limited randomized trials, and one meta-analysis, we prefer chemical pleurodesis using talc slurry delivered at the bedside via a small-bore chest tube, provided a functioning IPC is not already in place. This approach is preferred since, in our experience, it is as effective, less invasive, and better tolerated than surgical pleurodesis [62-66]. Surgical pleurodesis may be administered during a diagnostic thoracoscopy and sclerosant, typically talc, may also be delivered at the time of IPC placement or following IPC placement (via the IPC device).

The main advantage of pleurodesis is that it is typically more effective within a shorter timeframe than IPC alone. The main disadvantage is significant postprocedural pain and a small but potentially serious risk of respiratory complications, particularly when talc is used as the sclerosant. The complications of pleurodesis (eg, transient hypoxemia, fever, pain, and gastrointestinal symptoms, and rarely acute respiratory distress syndrome) are discussed separately. (See "Chemical pleurodesis for the prevention of recurrent pleural effusion", section on 'Complications'.)

Predictors of success have been described. Emerging data suggest that bedside ultrasound may play a role in predicting successful pleurodesis by comparing pre- and postpleurodesis imaging findings [67-69]. Select characteristics of MPE fluid may predict the likelihood of success or failure associated with pleurodesis but none have sufficient predictive properties to be used to exclude individual patients with MPE from consideration for pleurodesis [70]. For example, characteristics that make successful pleurodesis less likely include pleural pH less than 7.20 or 7.30, pleural glucose less than 60 mg/dL, and lactate dehydrogenase greater than 600 U/L [71-74]. A meta-analysis found that these criteria predicted pleurodesis failure with a sensitivity of only 55 to 59 percent and a specificity of 65 to 78 percent [75]. Among patients with MPE due to non-small cell lung cancer, epidermal growth factor receptor mutations predicted a high success rate for pleurodesis (>90 percent) [18,76].

Data that support our approach are discussed below. (See 'Chemical pleurodesis alone (bedside or thoracoscopic)' below and 'Talc pleurodesis plus an indwelling catheter' below and 'Mechanical pleurodesis' below.)

Chemical pleurodesis alone (bedside or thoracoscopic) — We prefer chemical pleurodesis using talc. Talc can be administered via a chest tube at the bedside (talc slurry) or at the time of thoracoscopy or thoracotomy (talc poudrage). Our preference is for talc slurry delivered at the bedside via a small-bore chest tube since, in our experience, it is as effective, less invasive, and better tolerated than surgical pleurodesis (ie, thoracoscopic delivery of talc poudrage) [62-66]. Best illustrating this is one open-label randomized trial of 320 patients with MPE who were treated with either 4 g of talc slurry via small-bore chest tube or 4 g of talc poudrage via thoracoscopy followed by chest tube drainage [66]. The failure rate at 90 days was no different between the groups (24 percent for slurry versus 22 percent for poudrage). In addition, there were no differences observed in any other outcome measured including failure at 30 or 180 days, pain, hospital length of stay, dyspnea, and quality of life. However, the study may have been underpowered to detect a smaller difference than the 15 percent rate set at the trial outset.

Although we prefer chemical pleurodesis via a small-bore chest tube/catheter, choosing between talc slurry or poudrage also usually depends upon the medical circumstances, goals and preferences of the patient, and institutional practice. As an example, thoracoscopy may be preferred in those with a concomitant indication for thoracoscopy or those in whom diagnosis of pleural malignancy is made during thoracoscopy. In contrast, administration via a chest tube may be preferred in patients with cardiopulmonary compromise who cannot tolerate thoracoscopy or via an IPC for those in whom a functioning IPC is already in place. (See "Chemical pleurodesis for the prevention of recurrent pleural effusion" and "Medical thoracoscopy (pleuroscopy): Diagnostic and therapeutic applications".)

Importantly, the pleural space needs to be dry for optimal results; so, as much fluid as is possible should be removed prior to the application of sclerosant. The data that support our approach are discussed in this section while technical aspects and complications of chemical pleurodesis are discussed in greater detail elsewhere. (See "Chemical pleurodesis for the prevention of recurrent pleural effusion".)

Additional details regarding chemical pleurodesis include the following:

●Chest tube size – No convincing data have demonstrated superior efficacy of chemical pleurodesis via a standard large-bore (eg, 24 French) compared with a small-bore (eg, 12 French) chest tube [77-79]. However, in many but not all studies, large-bore chest tubes cause more patient discomfort. While the European Respiratory Society/European Association of Cardiothoracic Society (ERS/EACTS) prefer large-bore chest tubes, we prefer small-bore tubes since they are also better tolerated by patients with adequate analgesia [1,77-81].

In a randomized trial of 320 patients with MPE, smaller chest tubes (12 French) were associated with a similar rate of pleurodesis failure at three months when compared with patients in whom large-bore chest tubes were placed (24 French; 30 versus 24 percent) [77]. However, pain scores were significantly lower among patients with smaller chest tubes (mean visual analogue score, 22 versus 27 mm), and although the complication rate was higher with the insertion of smaller chest tubes it was not significantly different (24 versus 14 percent). Methodologic flaws including small sample size for group comparisons and uncertainty in some studies whether the small-bore chest tubes were inserted under imaging guidance limit the interpretation of these data.

●Sclerosing agent – Talc (slurry or poudrage) is typically the agent preferred by many experts given its relative superiority for reducing the rate of MPE recurrence, when compared with other agents in patients with a variety of cancer cell types [38,79,82-87]. Talc is also cheap and readily available. General principles of choosing an agent are described separately. (See "Chemical pleurodesis for the prevention of recurrent pleural effusion", section on 'Our general approach'.)

In general, the success rate of talc in preventing recurrence at 30 days after pleurodesis ranges from 60 to 90 percent [34,79,83,86,88-90]. One network meta-analysis reviewed 62 randomized trials of pleurodesis reported that thoracoscopic talc poudrage was the most effective method, compared with several other methods [79]. Extensive heterogeneity between studies and risk for bias, however, limited the interpretation. It is difficult to determine success rates at 30 days because many studies exclude those patients who die before the 30-day assessment mark. If patients who die before 30 days after lung reinflation are included in analyses, success rates of talc fall to 50 to 60 percent [62]. Moreover, it appears that the longer the patient survives after pleurodesis, the greater the probability of recurrence with some studies reporting that up to 50 percent of patients undergoing talc pleurodesis experience recurrence at six months [62,91]. A network meta-analysis determined similar rates of pleurodesis success between talc slurry and talc poudrage when homogeneous studies were entered into analyses [38]. Data that discuss the efficacy of talc delivered at the time of IPC placement or through an IPC are discussed below. (See 'Talc pleurodesis plus an indwelling catheter' below.)

Intrapleural doxycycline is somewhat less effective than talc and may be associated with more pain. As a result, it may be used in settings where talc is not available or settings where talc should be avoided. Reported success rates are approximately 80 percent although an accurate assessment of effectiveness is limited by the small number and low quality of existing studies [79,88,92,93].

Other sclerosing agents, such as bleomycin, silver nitrate, iodopovidone, mepacrine, and Corynebacterium parvum have also been used successfully [79,87,94-98]. Small randomized trials and one meta-analysis of patients with pleural effusion suggest no difference in success rates between silver nitrate and talc or tetracycline derivatives [94,99,100]. Silver nitrate has been shown in a small case series of 17 patients to produce pleurodesis in 89 percent of patients for whom prior talc pleurodesis had failed [101]. Observational and small randomized trials report that iodopovidone induced successful pleurodesis in 86 to 96 percent of patients and was no different to talc in its efficacy [97,102,103]. Another randomized trial compared two concentrations of iodopovidone (1 versus 2 percent) in 60 patients with MPEs mainly due to breast cancer with the purpose of systematically identifying adverse effects [95]. Adverse effects were common and included pain, hyper- and hypotension, and subclinical hypothyroidism that resolved without treatment. Bleomycin is used less commonly for pleurodesis because of systemic toxicity (eg, lung fibrosis) and cost.

●Analgesia – The parietal pleural membrane contains a high percentage of pain receptors such that the induction of inflammation by the intrapleural instillation of a sclerosant is often intensely painful. NSAIDs and opiates are frequently prescribed for pain control. The choice of agent is typically individualized and dependent upon factors including level of pain, known sensitivity to opiates or NSAIDs, history of gastrointestinal bleeding, opiate drug abuse, liver or renal failure, and expected poor survival.

While animal studies had suggested that NSAIDs decrease the efficacy of pleurodesis, in a randomized study of patients with MPE, a similar rate of pleurodesis failure was noted in patients treated with opiates (10 to 20 mg oral morphine four times daily) when compared with patients treated with NSAIDs (ibuprofen 800 mg three times daily; 20 versus 23 percent) [77]. Although pain scores were no different between each group, twice as many patients who used NSAIDs required rescue analgesia with intravenous morphine.

●Duration of chest tube drainage – No consensus exists regarding the duration of chest tube drainage required to achieve pleurodesis after delivery of the sclerosing agent. One study found that a shorter duration of chest tube placement of 24 hours as compared with 72 hours did not decrease the likelihood of a successful pleurodesis regardless of the rate of pleural fluid drainage [104]. We individualize management with a goal of removing the chest tube after 24 hours of instillation of the sclerosant. However, for patients who have exceptionally large volumes of ongoing drainage, we sometimes leave the chest tube in place for up to 72 hours. For patients with persistent large volume drainage, we assess for complicating factors, such as chylothorax or nonexpandable lung. (See 'Nonexpandable lung' below and "Etiology, clinical presentation, and diagnosis of chylothorax".)

One randomized trial demonstrated that serial chest sonograms in patients with MPE may identify patients for chest tube removal earlier than waiting for diminution of chest tube drainage, but the reduction in hospital stay was negligible and there was no difference in successful pleurodesis at three months [69].

Studies differ regarding the influence of tumor type on the success rate of pleurodesis. Some studies report no differences [75,105]. In contrast, one retrospective study of 447 patients undergoing talc pleurodesis by thoracoscopy demonstrated a lower success rate for mesothelioma (76 percent) as compared with breast cancer (93 percent) [88]. The authors also reported that pleurodesis success was negatively influenced by high intrapleural tumor burdens, which are especially common in patients with mesothelioma.

Talc pleurodesis plus an indwelling catheter — Another option that may be offered to patients with MPE with expendable lung is talc pleurodesis combined with pleural fluid drainage with an IPC. This approach combines the efficacy and rapidity of pleurodesis with the safety and good patient tolerance of indwelling catheters. No randomized prospective studies have examined comparative outcomes with this combination approach as compared with IPC or pleurodesis alone. In addition, many patients who have IPCs in place undergo spontaneous pleurodesis without the instillation of talc (up to 40 percent), which in essence eliminates the need for the latter step. Thus, we believe that this option may be considered in patients who have a strong preference for a rapid and effective response who have a reasonable life expectancy (eg, >2 months).

There is no consensus on how talc pleurodesis is performed in patients with an IPC and either of the options below are acceptable; ultimately the choice may depend upon available expertise and operator experience:

●Thoracoscopic drainage and talc instillation at the time of IPC placement (in-patient procedure) – One approach is to drain the pleural fluid under thoracoscopic guidance, place an IPC, and insufflate sterile talc (3 to 5 g) via the thoracoscope at the time of IPC placement [55]. At the end of the procedure, a 24 gauge chest tube is placed through the port created for the thoracoscope, attached to suction overnight and removed 24 hours later. The indwelling catheter is drained three times a day on the first postoperative day, twice on the second and third postoperative days, and then once daily until the output is less than 150 mL per day. A chest radiograph is then obtained and the tunneled catheter removed if no reaccumulation of pleural fluid is noted. Details regarding the performance of thoracoscopy are provided separately. (See "Medical thoracoscopy (pleuroscopy): Diagnostic and therapeutic applications".)

This procedure was assessed in a series of 30 patients [55]. All of the patients reported improved dyspnea [55]. Successful pleurodesis was noted in 24 of the 26 patients who were still alive at six months (92 percent). The indwelling catheter was removed at a median duration of 7.5 days following the procedure. Adverse effects included fever in two patients, need for replacement of the IPC due to pleural loculations in one patient, and an empyema in one patient. Another consecutive case series of 30 patients found that pleuroscopy combined with IPC placement as compared with historical controls of conventional pleuroscopic pleurodesis had a similar success rate for pleurodesis but shorter hospital stay and shorter time to pleurodesis [106]. An uncontrolled study of 29 patients who underwent talc pleurodesis at the time of IPC placement found that patients were discharged after a median of two days and had the catheter removed after a median of 10 days and only four of 29 patients had a recurrence of pleural effusions [107]. One retrospective case series of 45 patients reported a 78 percent rate of successful pleurodesis after six months for patients treated with ambulatory thoracoscopic poudrage and IPC insertion [108].

●Talc instillation after IPC placement (outpatient procedure) – Other experts instill talc through an IPC after IPC drainage and demonstration of expandable lung. Studies suggest that this is efficacious and cost-effective [39,109]. A randomized trial of 154 patients with MPE compared IPC drainage alone with pleurodesis by talc instilled through the IPC [39]. After initial fluid drainage during catheter insertion, patients underwent three drainages daily (limited to 1 L of fluid removed with each drainage) at home over the subsequent 10 days. Patients who achieved full lung re-expansion (ie, expandable lung) underwent randomization to continued drainage versus instillation of talc via the indwelling catheter (4 g in 50 mL normal saline slurry). Successful pleurodesis occurred more frequently in patients treated with IPC plus talc pleurodesis as compared with indwelling pleural drainage alone at 35 days (43 versus 23 percent) and at 70 days (51 versus 27 percent). No difference in adverse events or death was noted. Patients treated by pleurodesis via the IPC had improved symptom and quality of life scores, compared with patients treated with IPC alone. The study was limited by the large number of run-in exclusions; among the 250 patients enrolled, 96 were excluded with 32 having lung entrapment, 23 were too ill to continue in the study, 13 had problems with IPC insertion, and 28 had other reasons to be excluded. Moreover, nearly 10 percent of those assigned to treatment groups were excluded from the data analysis and the IPCs were inserted before a trial of therapeutic thoracentesis to determine the rapidity of fluid re-accumulation, which would bias the results toward positive study effects. In addition, the success rate for pleurodesis at 43 percent was lower than that published historically (ie, 60 to 90 percent) [79] (see 'Chemical pleurodesis alone (bedside or thoracoscopic)' above). However, the study does demonstrate, the safety and relative efficacy of instilling talc via an IPC.

Mechanical pleurodesis — Mechanical pleurodesis such as pleural abrasion may be an option in patients with recurrent MPE. However, this procedure requires thoracoscopy or thoracotomy, which may be technically difficult if previous attempts at nonsurgical pleurodesis were partially successful or the effusion is significantly loculated [110]. Some clinicians may perform mechanical pleurodesis during a diagnostic thoracoscopic procedure. Mechanical pleurodesis can also be combined with chemical pleurodesis (eg, abrasion plus talc) but whether this increases the efficacy of the procedure is unclear.

Expandable lung with slow reaccumulation — IPC or repeat thoracentesis are typical options in this population. Pleurodesis may not be necessary, especially when the rate of accumulation is very slow. Choosing among them depends upon the actual rate of reaccumulation and additional factors that are discussed above. (See 'Choosing among the options' above.)

IPC — IPC is the preferred option in this population since it is effective and may result in spontaneous pleurodesis, thereby avoiding the need to perform a separate procedure for pleurodesis. (See 'Indwelling pleural catheter (IPC)' above.)

Repeat thoracentesis — The role for repeated thoracenteses has decreased with the advent of IPCs. However, multiple repeat therapeutic thoracentesis is an alternative to IPC for managing MPEs that reaccumulate slowly (eg, >1 month). We prefer repeat thoracentesis for patients with a very short life expectancy (eg, two to six weeks) or poor performance, and/or for those who decline an IPC or pleurodesis. Manometry is encouraged during repeat thoracentesis, if it was not performed previously and/or if imaging findings suggest the new development of nonexpandable lung. (See "Ultrasound-guided thoracentesis" and "Measurement and interpretation of pleural pressure (manometry): Indications and technique".)

Repeat thoracentesis may induce pleural adhesions that can complicate thoracoscopic pleurodesis if that procedure becomes indicated, although the precise number of repeat attempts associated with this complication is unknown. Other complications of therapeutic thoracentesis are the same as those for diagnostic thoracentesis. (See "Ultrasound-guided thoracentesis", section on 'Complications'.)

Nonexpandable lung — Nonexpandable lung may be suggested when a therapeutic thoracentesis does not improve dyspnea and when fluid is replaced by air after thoracentesis on chest imaging. It can also be identified on pleural manometry. One study suggested that over half of patients with MPE had nonexpandable lung as identified by chest radiograph or pleural manometry [111]. However, among those with a radiograph that had complete re-expansion, pleural manometry identified non expandable lung in 28 percent.

Examples of nonexpandable lung include patients with entrapped or trapped lung (eg, due to pleural tumor, septations, loculations, or scar tissue) or patients with lung collapse from endobronchial obstruction by tumor (see "Diagnosis and management of pleural causes of nonexpandable lung", section on 'Diagnosis'). Since full lung expansion with pleural-pleural apposition is necessary for successful pleurodesis, patients with irremediable nonexpandable lung are not typically suitable for any type of pleurodesis [112]. Thus, options in this population are limited to drainage with an IPC as the preferred intervention; ultrasound may guide IPC placement into the largest pocket of fluid to maximize symptom relief. However, partial pleurodesis may be performed if sufficient pleural-pleural apposition can be achieved after IPC drainage. In patients who decline an IPC or in whom a short life expectancy is anticipated, repeat thoracentesis may be an alternative. The ideal drainage schedule is unknown and should be guided by the individual response. (See 'Indwelling pleural catheter (IPC)' above and 'Repeat thoracentesis' above.)

Other options in this population include:

●Intrapleural thrombolytic agents – In rare patients with septations evident on chest computed tomography and poor drainage from a chest tube or IPC, intrapleural thrombolytic agents may be considered, although this approach is poorly studied and of unclear efficacy and risk in patients with MPE. In such cases, when continued drainage is preferred, the instillation of an intrapleural fibrinolytic agent (tissue-plasminogen activator, urokinase, or streptokinase) may be performed [113-116]. While possible benefit was suggested by a multicenter observational study [113], a randomized trial that compared urokinase with placebo in patients with nondraining MPEs intrapleural thrombolytic agents did not improve dyspnea or pleurodesis success [114]. More data are needed before intrapleural fibrinolytics can be routinely recommended for patients with loculated MPEs. (See 'Indwelling pleural catheter (IPC)' above and "Management and prognosis of parapneumonic pleural effusion and empyema in adults", section on 'Intrapleural tPA +/- additional drainage'.)

●Loculations may be mechanically debrided using pleuroscopy or thoracoscopy. (See "Medical thoracoscopy (pleuroscopy): Diagnostic and therapeutic applications" and "Overview of minimally invasive thoracic surgery".)

●Partial pleurodesis – Some patients experience sufficient lung re-expansion (ie, partial pleural apposition) after an interval of IPC drainage to allow pleurodesis that results in enough symphysis of visceral and parietal membranes to improve respiratory symptoms [3]. (See 'Chest tube or catheter drainage' above.)

●Decortication, pleurectomy, and shunt – Some patients with nonexpandable lung may theoretically achieve adequate drainage with these procedures, although they are almost never performed, especially when the prognosis is poor. (See 'Shunt' below and 'Refractory malignant pleural effusion' below.)

REFRACTORY MALIGNANT PLEURAL EFFUSION

Repeat or combination procedures — There is no consensus or guidelines on how to manage patients with MPE who fail indwelling pleural catheter (IPC) and pleurodesis. In most cases of refractory MPE, repeat attempts at thoracentesis (with or without fibrinolytics), prolonged IPC, or pleurodesis (chemical, mechanical, or both) are often tried before resorting to pleuroperitoneal shunt or pleurectomy. The efficacy of such an approach is unknown.

Intrapleural chemotherapy remains investigational.

Shunt — Pleuroperitoneal shunting is a rarely used option for patients with refractory MPE. It is also a rarely used option in those who have nonexpandable lung or malignant chylothorax [117-119]. While the complication rate is high (15 percent usually due to occlusion, shunt fracture, and infection), pleuroperitoneal shunting, however, may palliate 95 percent of patients and provide nutritional advantages for patients with malignant chylothorax and high volumes of pleural fluid drainage; this is because shunting of nutrient-rich chyle to the peritoneal space allows its reabsorption. (See "Etiology, clinical presentation, and diagnosis of chylothorax" and 'Indwelling pleural catheter (IPC)' above.)

Usually, shunt placement requires thoracoscopy and general anesthesia but shunts may also be placed using interventional radiological techniques [118]. The shunt catheter (Denver pleuroperitoneal shunt), has one or two one-way valves that allow unidirectional flow away from the pleural space. One end is inserted into the pleural cavity and the other through a subcutaneous tunnel into the peritoneum; the shunt pumping chamber is placed in a subcutaneous pocket overlying the costal margin.

In the hands of experienced operators, placement of a pleuroperitoneal shunt is reasonably safe, although shunt-related complications occur in about 15 percent [117]. The major problems have been shunt failure, most commonly due to occlusion of the catheter and infection. In a retrospective review of 160 patients who received a pleuroperitoneal shunt for MPE, 12 developed shunt occlusion (requiring revision in five and replacement in seven) and another seven developed infection [117]. One patient developed malignant seeding on the chest wall at the site of shunt insertion, but peritoneal seeding was not observed. It is not known whether patients who have had shunt occlusion are at greater risk for occlusion after a new shunt is placed.

Limited data suggest that palliation of the pleural effusion is achieved in 73 to 90 percent of properly selected patients [117].

Pleurectomy — Total or partial pleurectomy (resection of visceral and parietal pleura) with decortication (removal of fibrous pleural rind) is a last resort because it is highly invasive and malignant adhesions, septations and loculations are not easily amenable to surgical removal. The exception is for patients with MPE due to mesothelioma, the management of which is discussed separately. (See "Initial management of malignant pleural mesothelioma".)

Total pleurectomy with decortication is a highly invasive procedure for which thoracotomy is required; it is technically difficult, has significant morbidity and mortality, a long recovery time, and a paucity of weak evidence to support it [120-123]. If selected, patients must be good surgical candidates, be willing to accept the risks of surgery including death, and have a reasonably long expected survival (eg, >12 months). Partial pleurectomy/decortication can be accomplished thoracoscopically; however, it too, is a technically difficult procedure, although recovery time may be less than with a thoracotomy.

Existing studies report efficacy of pleurectomy as initial therapy with no studies reporting their benefit for patients who fail IPC with pleurodesis [124]. One observational study reported outcomes from single incision thoracoscopic pleurectomy in 19 patients with MPEs who were highly selected as "suitable" for surgery [125]. Pleural effusions did not recur in 91 percent of patients and no mortality or complications were observed.

PROGNOSIS — The prognosis of patients with an MPE is generally poor with an estimated survival between 3 and 12 months after diagnosis [1,3,126]. The prognosis depends on a variety of factors, such as age, persistent breathlessness, performance status, tumor type, tumor stage, comorbidities, composition of the pleural fluid, and responsiveness of the underlying cancer to antitumor therapy. Overall, observational studies demonstrate that mortality is higher for patients with an MPE compared with patients with metastatic cancer who don’t have a malignant effusion. As examples:

●Analysis of the Nationwide Readmissions Database revealed that among the 25 percent of patients who are readmitted to hospital within 30 days of discharge, approximately 17 percent die during readmission [29]. Predictors of early readmission include having Medicaid insurance status, treatment with thoracentesis only, and discharge to a care facility or home health care.

●Among patients with advanced non-small cell lung cancer who have distant metastases, one study demonstrated that the presence of an MPE was an independent predictor of a lower overall survival at one and two years (hazard ratio [HR] 1.36, 95% CI 1.30-1.43) [127].

●In a systematic review of 417 patients with MPE, the median survival was four months, although some patients may survive longer than that [128].

●A series of 278 patients referred to a thoracic surgery clinic for management of an MPE reported a median postoperative survival of 211 days [129].

●A series of 789 patients with MPE reported variable median survival, ranging from 74 days for patients with lung cancer to 339 days for patients with mesothelioma (table 2) [130].

Multiple factors have been examined to predict survival of patients with MPEs. These factors include clinical performance status, tumor type, pleural fluid protein-to-chloride ratios, and blood or pleural fluid neutrophil-to-lymphocyte ratios [131,132]. Although pleural fluid lactate dehydrogenase [130] and pH [128] correlate with survival, they are both weak predictors of survival for individual patients.

These factors have been examined in predictive models:

●The largest series with prospective data collection that assessed prognostic factors derived and validated a risk stratification system called the LENT score (pleural fluid lactate dehydrogenase, Eastern Cooperative Oncology Group [ECOG] performance score, blood neutrophil-to-lymphocyte ratio, and tumor type) (table 3) [130]. The four factors combined into the LENT score had greater prognostic value than individual factors. Final scores separated patients into low-, moderate-, and high-risk groups that had a median survival of 319, 130, and 44 days respectively. The receiver operating curve area, however, ranged from 0.77 to 0.85 for prediction of survival at one-month and six-month respectively, so uncertainty regarding prognosis remains when applying the LENT score or other prognosticators to individual patients [38], especially considering the score groups patients into only three possible predicted median survivals. Moreover, various tumor types with varying biological behaviors are classified into only three groups [38]. Thus, the LENT score is infrequently used in clinical practice and is mostly used for research purposes.

●Similarly, another score, holds promise but is not yet ready for clinical use (PROMISE) [133].

●The Breast and Lung Effusion Survival Score (BLESS score) represents an effort to use disease-specific models to provide more accurate survival predictions than LENT or PROMISE. Initial encouraging reports from a single analysis [38] require external validation.

A case series of 117 patients with MPE reported that those who had a neutrophil-to-lymphocyte ratio >0.745 measured in pleural fluid had a shorter median survival of 130 (95% CI 0-282) days as compared to 312 (95% CI 195-428) days for patients with a ratio <0.745 [132].

The tumor subtype affects survival. For example, MPE associated with cancer of the lungs and gastrointestinal system have the worst survival (eg, <6 months) while patients with MPE due to chemoresponsive tumors such as lymphoma or breast cancer are more likely to have prolonged survival. A retrospective analysis of 277 patients with MPE due to non-small cell lung cancer found that patients with squamous cell versus adenocarcinoma pleural fluid cytopathology had a shorter survival after diagnosis (median survival 112 [interquartile range (IQR) 44-220] versus 194 [IQR 54-523] days, respectively) [134].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Pleural effusion".)

SUMMARY AND RECOMMENDATIONS

●Definition – Malignant pleural effusions (MPEs) are caused by infiltration of cancer cells into pleural tissue that may result in positive fluid cytology and/or be seen on pleural biopsy. MPEs commonly complicate patients with lung cancer and malignant mesothelioma but can also complicate metastatic cancers (from lung or distant sites), lymphoma, and other hematologic malignancies. The recommendations made in this topic do not apply to patients with MPE from mesothelioma. (See 'Definition (malignant/paramalignant)' above.)

●Goals – Management of MPE is typically palliative and without survival benefit. Treatment should focus on patient-centered goals of therapy, which include sustained symptom relief and improvement of quality of life. Asymptomatic patients with MPEs do not require treatment until they become symptomatic, which is common. (See 'Treatment goals' above.)

●Initial treatment – The following is a reasonable approach to initial management (algorithm 1 and table 1) (See 'Initial treatment' above.):

•For most patients initially presenting with symptomatic MPE, we recommend therapeutic large-volume thoracentesis as the initial intervention rather than proceeding straight to IPC (algorithm 1) (Grade 2C). (See 'Initial treatment' above and 'Therapeutic thoracentesis' above and 'Chest tube or catheter drainage' above.)

•In all cases, if indicated, the underlying malignancy should be simultaneously treated; select tumors may respond to antitumor therapy including breast, ovarian, and prostate cancer, germ cell tumors, lymphoma, and small cell lung cancer. Thoracentesis determines the symptomatic response to drainage, the ability of the lung to re-expand completely, and the rate of subsequent reaccumulation, all of which inform future more definitive treatments should the effusion reaccumulate. (See 'Treatment of the underlying malignancy' above.)

●Recurrent MPE – Over half of patients experience recurrent MPE, and among those, up to two-thirds recur rapidly within one month. The optimal option for patients with recurrent MPE is unclear (table 1 and algorithm 1). Considerable variation exists in practice with guideline groups and many other experts suggesting proceeding directly to intrapleural catheter (IPC) drainage or to pleurodesis. In reality, several factors weigh in to the decision including the presence of expandable lung, rate of reaccumulation, predicted responsiveness to antitumor therapy, prognosis, functional status, severity of symptoms, institutional resources and expertise, and patient preferences. (See 'Pleural interventions for recurrence' above and 'Choosing among the options' above.)

•Expandable lung with rapid recurrence – Our approach is the following (see 'Expandable lung with rapid reaccumulation' above):

-IPC – For patients with rapidly recurrent symptomatic MPE who have expandable lung, we suggest drainage via an IPC rather than pleurodesis (Grade 2C). However, pleurodesis is a reasonable alternative to IPC in some patients. Use of thoracentesis in this setting is generally limited to patients in whom the prognosis is grim (eg, <2 weeks) or when IPC and pleurodesis are declined. (See 'Indwelling pleural catheter (IPC)' above.)

Our preference for IPC is based upon randomized trial data with important limitations that suggest IPC may provide symptom relief, can induce spontaneous pleurodesis (in up to 40 percent of cases), and may reduce length of stay compared with pleurodesis. In addition, unlike pleurodesis, IPC can be used in patients with expandable or nonexpandable lung and can be performed as an outpatient procedure. IPC is ideal for patients with a shorter anticipated duration of survival (less than two months). However, the patient or a caregiver need to be able to perform drainage at home and maintain catheter sterility and the longer dwell time of the tube (ie, weeks as compared with days for pleurodesis) is not palatable to some patients.

-Pleurodesis – Pleurodesis may be suitable for patients who desire a rapid and effective way to prevent further recurrence ("a one and done"). Pleurodesis may be the only option in institutions without expertise in IPC placement or in patients who decline or fail IPC. Pleurodesis is not suitable for patients with a short anticipated survival (eg, <2 months) or those with nonexpandable lung. (See 'Pleurodesis (alternative to IPC or failed IPC)' above.)

When pleurodesis is chosen, we suggest chemical pleurodesis using talc slurry delivered at the bedside via a small-bore chest tube rather than talc poudrage administered thoracoscopically (Grade 2C). Our preference is based upon one randomized trial with limitations and retrospective data suggesting that bedside chemical pleurodesis is as effective, less invasive, and better tolerated than surgical pleurodesis via thoracoscopy However, there is considerable variation in practice Such that choosing among these options is often institution- and operator-dependent. (See 'Chemical pleurodesis alone (bedside or thoracoscopic)' above.)

However, surgical pleurodesis may be administered during a diagnostic thoracoscopy. In addition, sclerosant, typically talc, may also be delivered at the time of IPC placement or following IPC placement (via the IPC device). (See 'Talc pleurodesis plus an indwelling catheter' above and 'Mechanical pleurodesis' above.)

The main advantage of pleurodesis is that it is typically more effective within a shorter timeframe than IPC drainage alone. The main disadvantage is significant postprocedural pain, chest infections, and a small but potentially serious risk of respiratory complications, particularly when talc is used as the sclerosant.

•Expandable lung with slow recurrence – Our approach is the following (see 'Expandable lung with slow reaccumulation' above):

-IPC – For patients with slow reaccumulation of their MPE (eg, longer than one month), we suggest IPC drainage rather than repeat thoracentesis (Grade 2C). IPC drainage is generally the first choice regardless of the presence of expandable or nonexpendable lung since it is effective and may result in spontaneous pleurodesis. (See 'IPC' above.)

-Repeat thoracentesis – Repeat therapeutic thoracentesis is reasonable, if patients decline IPC or have a grim prognosis (eg, <2 weeks). Pleurodesis may not be needed in this population especially when the rate of accumulation is very slow. (See 'Repeat thoracentesis' above.)

•Nonexpandable lung – For patients with recurrent symptomatic MPE who have irremediable nonexpandable lung, options are limited to IPC, when prolonged drainage is indicated. Since full lung expansion with pleural-pleural apposition is necessary for successful pleurodesis, patients with irremediable nonexpandable lung are not typically suitable for any type of pleurodesis. However, partial pleurodesis may be performed if sufficient pleural-pleural apposition can be achieved after IPC drainage. In patients who decline an IPC or in whom a short life expectancy is anticipated, repeat thoracentesis may be an alternative. (See 'Nonexpandable lung' above.)

●Refractory MPE – For patients who fail IPC drainage and pleurodesis, options include repeat attempts at thoracentesis, prolonged IPC drainage, repeat attempts at pleurodesis, or some combination thereof; choosing among these is individualized. Pleuroperitoneal shunt or pleurectomy with decortication is rarely performed. (See 'Refractory malignant pleural effusion' above.)

●Prognosis – The prognosis of patients with an MPE is generally poor but depends on a variety of factors, such as age, performance status, tumor type (table 2), tumor stage, comorbidities, composition of the pleural fluid, and responsiveness of the underlying cancer to antitumor therapy. The LENT score (pleural fluid lactate dehydrogenase, Eastern Cooperative Oncology Group [ECOG] performance score, blood neutrophil–to-lymphocyte ratio, and tumor type) (table 3) may be a useful prognostication tool but is mostly used for research purposes. Other prognostic tools are in progress. (See 'Prognosis' above.)