Malignant pleural effusion (MPE) is a common complication in advanced stages of malignancy and affects approximately 15% of oncological patients.1,2 This condition is associated with poor prognosis, with an overall survival (OS) from 3 to 12 months.3,4 Despite the advances in cancer treatments, MPE management remains palliative and focuses on patient's symptoms relief.4-6 However, there are factors to be taken into account when choosing a treatment, including lung expandability.7

Non-expandable lung (NEL) due to lung entrapment is often found in MPE patients, and may be caused by endobronchial obstruction or malignant involvement of visceral pleura,8 which prevents visceral and parietal pleura apposition following pleural fluid drainage.9 The presence of a hydropneumothorax after thoracocentesis or the inability to completely drain the pleural effusion due to excessive cough or central chest pain during aspiration strongly suggest the presence of lung entrapment. As the lung cannot expand, the pleural pressure becomes more negative, which justifies the continuous transudation of fluid from the parietal pleural capillaries and frequent recurrence of pleural effusion, with a significant impact on the quality of life of these patients. Unfamiliarity with this entity can lead to unnecessary and ineffective interventions, with significant morbidity.10 Intermittent pleural drainage or indwelling pleural catheter (IPC) have been reported to be a better option for the management of malignant lung entrapment.5,11

Thus, early detection and management of MPE should improve outcomes, by decreasing the occurrence of lung entrapment and allowing more patients to benefit from definitive treatment with pleurodesis. Looking for a more patient-centered approach to MPE, this study aimed to identify possible predictors of lung entrapment by the retrospective analysis of patients diagnosed with MPE during a period of six years in our institution.

This study reports a prevalence of nearly 19% of NEL in a large cohort of cancer patients with pathologically confirmed MPE. Only a few studies have previously described the frequency of malignant lung entrapment, with varying rates according to the diagnostic criteria and histological subtypes. In a randomized trial (RCT) of patients with mixed MPE, NEL (defined as <75% pleural apposition) was found in 41 out of 923 patients (4.4%) at screening, and a further 32 of 250 patients (12.5%) were found to have NEL after a 10-day run-in period.15 Another RCT applied the criterion of <90% lung re-expansion to define lung entrapment, and reported a prevalence of nearly 30%.16 Given the inconsistency of NEL definition across studies, we decided to apply a more clinically oriented criterion, considering the recommendations to attempt a pleurodesis whenever ≥50% of visceral and parietal pleura are apposed.14 As such, NEL cases with <50% pleural apposition are, in fact, those that will have a more complicated management of MPE, given the reduced chemical pleurodesis efficacy.6 The treatment decision of MPE in these cases should be based on symptoms, patient's performance status and preferences, along with disease prognosis.7,17 While serial thoracocenteses may be appropriate for slowly reaccumulating pleural effusions and/or in patients with short survival expectancy, IPCs are generally accepted as the frontline approach to most cases of lung entrapment, enabling alleviation of symptoms at the ambulatory basis, without the need of repetitive invasive procedures. Nevertheless, aggressive fluid removal during drainage can induce chest discomfort due to tension on the NEL by negative intrathoracic pressure.18 Although some studies have reported good outcomes with surgical decortication,5 its definitive role in NEL is still uncertain. Results from MesoTRAP, a RCT comparing video-assisted thoracoscopic partial pleurectomy with IPC in patients with trapped lung due to malignant pleural mesothelioma19 may shed some light on this topic.

Considering the difficulties that lung entrapment poses on the symptoms palliation of cancer patients, and the high recurrence rate of MPE, it has been proposed that an early intervention may be beneficial in improving control of the pleural effusion, including a possible reduction of NEL risk.6 Although we did not assess symptoms control or quality of life, we demonstrated that patients with lung entrapment had more hospitalization days and underwent thoracocenteses more frequently than patients with expandable lungs. Considering the dismal prognosis associated with MPE, arguably these findings are inconsistent with the palliative treatment goals. For that reason, it may be useful to determine which patients are more prone to have NEL, in order to better select patients for a more aggressive approach of the MPE, namely early pleurodesis. Although late diagnosis of MPE is often mentioned as a risk factor for lung entrapment,6 we found no statistically significant differences between the median time from primary malignancy diagnosis or from symptoms onset to MPE diagnosis. In fact, when specifically considering lung cancer, the most represented malignancy, MPE was generally diagnosed at the same week as the primary cancer. Similar results were found by Porcel et al.,20 showing that patients with lung cancer are more likely to have pleural effusion at cancer diagnosis, whereas MPE is a late complication in other types of cancer.

In the present study, we confirmed the dismal prognosis associated with MPE,4,7 with a median OS of 3 months, but failed to show that NEL is associated with shorter survival.21-23 Although the presence of a pleural effusion reflects more advanced disease, most cases of lung entrapment could not be explained by diffuse or bulky metastatic spread along the pleural surface. Instead, we hypothesized that a thickened visceral pleura peel could be caused by local factors of the effusion, as explained below.

Most references in the literature point to the elastance of the pleural space as predictor for lung entrapment.24,25 These studies are usually small and do not investigate the impact of other clinical and pleural effusion features on the development of NEL. According to our analysis, loculated pleural effusion, complete white-out of a hemithorax, lung cancer and higher effusion-serum LDH ratio were all independent predictors for the occurrence of lung entrapment in patients with MPE. Septated or loculated pleural effusion are an obvious signs of NEL, as they limit the complete drainage of the pleural space.5 Likewise, complete hemithorax opacification usually result from a combination of pleural effusion, caused by direct tumor infiltration or metastatic invasion of the visceral pleura, and endobronchial obstruction causing distal lung collapse or atelectasis, which are both hallmarks of the NEL. Malignant airway obstruction may not be reversible, even when rigid bronchoscopy is available. We recently published our experience, showing that in patients with endobronchial obstruction causing lung atelectasis that were initially considered candidates for bronchoscopic intervention (selected based on clinical, endoscopic and radiological pre-assessment), the procedure was unsuccessful in 22% of cases.26 Technical failure was more often seen when distal airway patency was absent on thoracic CT, as is often the case in complete lung white-out. While loculations may be managed with intrapleural fibrinolytics administration,5 hemithorax opacification is hardly solved with a chest tube, and measures should be taken, when possible, to prevent the progression to this stage. Hence, we propose early intervention in MPE of lung cancer patients and/or with high effusion-serum LDH ratio on the first diagnostic thoracocentesis. Although increased pleural fluid protein levels were associated with lung entrapment, only high effusion-serum LDH ratio kept a significant association after multivariate analysis. Importantly, the finding of elevated LDH in the pleural fluid is an important indicator for the presence of active pleural disease, while protein concentration is more dependent on microvascular permeability.27 Lower glucose level was found to be a lung entrapment predictor in a previous study,24 but this association was not retrieved from our analysis.

Current guidelines5,6 recommend placement of IPCs in patients with symptomatic MPE with NEL, which in some cases could lead to pleural apposition overtime and enable pleurodesis. In the present cohort of MPE patients diagnosed between 2012-2018, serial thoracocenteses were the usual approach, and “definitive” pleural procedure, such as IPCs, were less commonly applied, even in cases of NEL. One large retrospective study of 23,431 patients with MPE demonstrated that only 24% underwent a definitive pleural procedure, talc pleurodesis or IPC placement, as opposed to repeat thoracentesis, after rapid fluid reaccumulation.28 Patients undergoing definitive pleural approach experienced fewer additional pleural procedures, including those performed in the emergency department, and fewer complications than cases managed with serial thoracocenteses, underlining the importance of definitive pleural intervention at the appropriate time. We admit that the issuance of guidelines may have an impact on the local habits regarding MPE management in the following years, and more cases with NEL will have an IPC. In our opinion, therapeutic drainage through thoracocentesis serves essentially 3 purposes: to confirm symptomatic improvement after fluid removal (thus, determining benefit at definitive pleural treatment), to identify NEL (to prevent futile attempts at pleurodesis), and to palliate symptoms in patients with poor performance status, with an estimated survival of less than one month.

Some limitations can be identified in our study. The present analysis is based on a single center, where local MPE management practices might influence the results presented. Moreover, its retrospective nature has the potential to introduce some information bias, and we lack data concerning the evolution of symptoms over time, which would be important to assess patients' quality of life and its relationship with the management of MPE. Despite that, we provide one of the largest cohort studies of patients with MPE and NEL in the recent literature. Prospective studies with patients with malignant lung entrapment should seek validation of our results and assess how the delay to MPE definitive treatment may compromise lung expandability and the patient's quality of life.

This study gives an overview of our local MPE epidemiology and management options, and demonstrates that NEL is a relatively common phenomenon in patients with lung cancer and/or with higher effusion-serum LDH ratio. Moreover, the presence of loculations and complete hemithorax opacification are independent predictors for lung entrapment. Late diagnosis of MPE was not associated with a higher risk of NEL in our cohort, and no statistically significant difference regarding pleurodesis success or OS was found between NEL and expandable lung group.