Large amounts of VEGF can be produced as tumor cells invade into the pleura, resulting in the acceleration of vascular permeability (13). The pleural fluid VEGF levels in MPE are significantly higher than those in effusions of benign disease, such as congestive heart failure and tuberculosis (2,8,14). In 2011, Fiorelli et al (9) studied 79 patients with unilateral PE. The levels of VEGF were demonstrated to be much higher in malignant than in benign exudates (9). In addition, in a study by Lieser et al (15), a 77-fold higher VEGF expression was observed in MPE compared with that in benign PEs. VEGF has therefore been suggested to be a diagnostic marker for MPE (16). Furthermore, patients with MPE with a high pleural VEGF level have been reported to have a significantly shorter survival than those with normal VEGF level (17,18).

The effects of VEGF are mainly mediated by endothelial cell receptors VEGF receptor-1 (VEGFR-1) and VEGFR-2 (19). VEGF binds to its receptors and induces downstream signaling, such as that of protein kinase C and mitogen-activated protein kinase; this process can induce vascular endothelium differentiation and proliferation, stimulate capillary sprouting and finally produce endothelial fenestrations and loss of junctional integrity, contributing to tumor growth and MPE development (4). Thus, the inhibition of VEGF activity with VEGF inhibitors is regarded as a promising approach for improving the management of MPE.

Bevacizumab (Avastin^®; Roche), the first humanized monoclonal antibody against VEGF, has been used to treat several types of tumor, including lung cancer and gynecological cancers (20-22). It can directly inhibit the proliferation and migration of vascular endothelial cells, promote the apoptosis of endothelial cells and suppress VEGF-induced neoangiogenesis and vascular permeability (23). Several studies have specifically researched the efficacy and safety of bevacizumab for the management of MPE (Table I). Among them, 6 are prospective studies (13,24-28), including two phase II trials conducted by Japanese groups (24,25). Unfortunately, there is still a lack of evidence from large phase 111 trials to confirm the effect of bevacizumab in the management of MPE.

The first retrospective study to investigate the efficacy of bevacizumab plus chemotherapy for patients with MPE was conducted by Kitamura et al (29). The aforementioned study analyzed data from 13 patients with MPE caused by non-small cell lung cancer (NSCLC) who received bevacizumab (15 mg/kg, intravenously) plus chemotherapy as first- or second-line treatment (29). As expected, an MPE control lasting >8 weeks was achieved in 12/13 patients (92.3%) (29). Similar results were obtained in other studies (3,30), which analyzed the records of patients treated for NSCLC-associated MPE, who consequently received bevacizumab (15 or 7.5 mg/kg, intravenously) plus chemotherapy. In one study, a total of 15/21 patients were responders, and a response rate of 71.4% was reported (30), while an MPE response rate of 81.0% was observed in the other study (3).

The delivery of bevacizumab directly to the pleural space is an alternative for MPE treatment. The intrapleural administration of bevacizumab for MPE treatment has been suggested in previous studies. Chen et al (31) demonstrated that the intrapleural infusion of bevacizumab was effective in controlling MPE without apparent toxicity. Jiang et al (32) came to a similar conclusion, with bevacizumab significantly improving the response rate and QoL of patients with MPE without notable adverse events (AEs). Song et al (33) demonstrated that, in patients with MPE treated with an intrapleural infusion of bevacizumab (200 mg) combined with pemetrexed (BP group) or cisplatin (BD group), the response rates in the BP and BD groups were 56.52 and 86.36%, respectively, and the overall survival (OS) time for both groups was both >10 months. These results suggested that the intrapleural infusion of bevacizumab combined with chemotherapy may be effective for patients with MPE (33).

Two phase II trials analyzing the efficacy of bevacizumab in treating MPE were conducted on Japanese patients (24,25). The first study enrolled 23 patients with NSCLC with MPE who were treated with bevacizumab (15 mg/kg, intravenously) and carboplatin-paclitaxel (25). The primary endpoint of this study was an overall response rate (ORR) of 60.8% (25). The disease control rate (DCR), median progression-free survival (PFS) time and median OS time were 87.0%, 7.1 and 11.7 months, respectively (25). Furthermore, the addition of bevacizumab into the treatment protocol increased the MPE control rate from 78.3 to 91.3% (25). Correspondingly, the serum VEGF levels were decreased from 513.6 to 25.1 pg/ml following bevacizumab treatment (25). The second study focused on the treatment of bevacizumab with carboplatin-pemetrexed (24). A total of 28 patients with NSCLC-associated MPE were treated with carboplatin and pemetrexed with bevacizumab (15 mg/kg, intravenously) every 3 weeks (24). The control rate of MPE without pleurodesis at 8 weeks after treatment was defined as the primary endpoint (24). A total of 26/28 (92.8%) patients reached the primary endpoint (24). The median PFS, OS and median pleurodesis-free survival times were 8.2, 18.6 and 13.9 months, respectively (24). A high VEGF (≥100 pg/ml) level in the plasma, indicating a poor prognosis, was also observed (24).

Three Chinese prospective studies investigated the efficacy of intrapleural bevacizumab administration in patients with MPE (26-28). Wang et al (28) enrolled 33 patients with NSCLC with MPE who all received paclitaxel and bevacizumab (5 mg/kg, intrapleurally) once every 3 weeks for 12 consecutive weeks. The total response rate of this study reached 77% (28). The median OS and median PFS times were 22.2 and 8.4 months, respectively (28). Du et al (27) directly compared the efficacy of combined intrapleural therapy with bevacizumab (300 mg) and cisplatin versus cisplatin alone in controling MPE. A total of 70 patients with NSCLC with MPE were included, with 35 patients per group (27). The results revealed that the addition of bevacizumab improved the ORR from 50 to 83.3% (27). Additionally, patients in the bevacizumab + cisplatin group exhibited a higher QoL benefit and a higher decrease in pleural VEGF levels (27). Another group of 24 patients with NSCLC was enrolled in a similar study by Qi et al (26). A total of 14 patients in that study received intrapleural infusion of paclitaxel and bevacizumab (5 mg/kg), and the remaining 10 received paclitaxel alone (26). The results revealed that combination therapy significantly decreased the MPE levels, with an overall efficacy rate of 78.6% (a 29% increase compared with paclitaxel alone) (26). The addition of bevacizumab also improved the 1-year survival rate from 20.8 to 45.8% (26).

In 2017, Zongwen et al (34) performed a meta-analysis that included 11 randomized controlled trials (RCTs) with a total of 769 patients with lung cancer, to investigate the efficacy and safety of bevacizumab in controling MPE. All the studies in the meta-analysis were conducted by Chinese groups and most of them (10/11) were published in Chinese (34). This meta-analysis provided further evidence to support the administration of bevacizumab via intrapleural injection in patients with MPE (34). As compared with platinum alone, the addition of bevacizumab significantly increased the ORR (P=0.003), decreased the incidence of chest pain (P<0.001) and relieved dyspnea in patients (P=0.002) (34).

The aforementioned studies indicated that the combination of bevacizumab and chemotherapy may be effective for controlling MPE, although the sample sizes of these studies were small. For MPE management, bevacizumab can be administered intravenously or intrapleurally. However, the optimal administration route of bevacizumab has not yet been defined. Theoretically, compared with intravenous infusion, intrapleural infusion has some advantages, including the site-specific concentration of therapeutic agents with lower overall doses (10). Furthermore, the closed nature of the pleural space makes it an ideal site for intrapleural infusion (10). A recent randomized clinical study compared the efficiency and safety of intrapleural (7.5 mg/kg) and intravenous (7.5 mg/kg) infusion of bevacizumab in the management of MPE (13). Hypertension, epistaxis and proteinuria are common AEs associated with bevacizumab, which occurred more often in the intravenous group compared with in the intrapleural one (13). A higher ORR and a longer median duration of response were also observed in the intrapleural group, but the difference between the two groups was not statistically significant (13). The reason for this may be that the sample size of the study was too small (n=43). Therefore, large studies are required to confirm these results.

Currently, to the best of our knowledge, there are no prospective phase III studies specifically focusing on the benefit of bevacizumab combined with chemotherapy in MPE, which is a major issue. In addition, most of the aforementioned studies, whether prospective or retrospective, were conducted by Chinese and Japanese groups, which may lead to geographical and ethnical differences in the results. Therefore, data from other ethnic groups are required. Furthermore, there is a lack of data concerning rare tumor-driver mutations in patients with MPE caused by NSCLC. Jiang et al (35) retrospectively investigated 86 patients with NSCLC with MPE who had developed acquired resistance to epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor (TKI) therapy. In the aforementioned study, patients who were treated with bevacizumab and EGFR-TKIs had a longer PFS and higher curative efficacy rate for MPE, indicating that this combination therapy may be a valuable treatment option for patients with MPE (EGFR mutation-positive) caused by NSCLC (35).

Endostatin is an endogenous inhibitor of angiogenesis, which can interfere with the pro-angiogenic effects of growth factors and inhibit angiogenesis in a wide range of tumors, including lung, gastric and colon cancer (36-39). Similarly to VEGF, endostatin is also a potential prognostic factor for MPE (18,36). To enhance the solubility and stability of endostatin, recombinant human endostatin was engineered (a 9-amino acid sequence was added to the N-terminal of the protein) and termed endostar (16). Based on the results of a phase III study (37), endostar was listed as a first-line drug for the treatment of NSCLC in Chinese patients.

For patients with MPE, endostar was effective when administered via intrapleural infusion (37-39). A Chinese group evaluated the efficacy of endostar combined with cisplatin/pemetrexed chemotherapy for elderly patients with MPE (38). A total of 128 patients with lung adenocarcinoma with MPE were randomly assigned to receive chemotherapy plus endostar (treatment group) or chemotherapy alone (control group) (38). Compared with chemotherapy alone, endostar plus chemotherapy significantly improved MPE control rates (93.94 vs. 79.03%; P=0.013) and decreased recurrence rates (9.68 vs. 30.61%; P=0.005) with tolerable side effects (38). Similar results were obtained in a smaller RCT that included 45 patients with MPE or ascites (39). Furthermore, Biaoxue et al (40) conducted a meta-analysis based on 13 RCTs that included 1,066 patients with MPE, demonstrating that the addition of endostar to chemotherapeutic agents can significantly improve ORR and DCR, indicating that endostar is effective in treating MPE (40). However, one of the limitations of the aforementioned meta-analysis was that all included studies were from China. To the best of our knowledge, no data are available from outside China. Therefore, although endostar appears to be promising in controlling MPE, it still requires further investigation before it can be recommended for clinical application.

In addition to VEGF, other pro-angiogenic factors, such as TGF, platelet derived growth factor (PDGF) and fibroblast growth factor, have been associated with the development of PEs (4). Anti-angiogenic TKIs, which block the kinase activity of receptors, have been explored for the treatment of MPE. In in vitro and murine models of MPE, anti-angiogenic TKIs, such as lenvatinib, vatalanib and nintedanib, have been reported to control MPE (41,42).

Data from clincal trials evaluating the efficacy of EGFR-TKIs in patients with MPE are limited, with only two phase II trials (43,44). The first study enrolled 20 patients with NSCLC to evaluate the efficacy of vandetanib (VEGFR- and EGFR-TKI) for MPE control (43). The patients received oral vandetanib at a dose of 300 mg once a day for a maximum of 10 weeks, and time to pleurodesis was the primary endpoint (43). The results revealed that vandetanib was well tolerated, but it did not significantly decrease time to pleurodesis (43). The second study enrolled 12 patients with malignant ascites or MPE to assess the palliative value of cediranib (VEGF-TKI) (44). The primary endpoint was puncture-free survival, which was defined as the time from the start of the study to the time that paracentesis or thoracentesis were first needed, or the time of death (44). As expected, cediranib treatment significantly increased puncture-free survival with an acceptable toxicity profile (44). This phase II trial was the first study to show the palliative effects of oral VEGFR-TKI in patients with malignant effusions (44). Although sorafenib (RAF-, PDGFR- and VEGFR-TKI) has been reported to decrease MPE in one patient with advanced thyroid carcinoma (45), no further evidence has been acquired to support this treatment strategy.

The clinical trials on TKIs in this context are limited, and more data are urgently required. The other important issue is that most anti-angiogenic TKIs are administrated intraorally. Therefore, whether safe doses of anti-angiogenic TKIs comprise therapeutic concentrations for MPE through oral administration needs to be further investigated.

In addition to bevacizumab, ramucirumab is another anti-VEGF antibody used in a clinical setting. The PLEURAM study is an ongoing phase II trial evaluating the efficacy and safety of ramucirumab in controling MPE (46). The study plans to enroll 15 patients with NSCLC and ramucirumab (10 mg/kg) combined with docetaxel (60 mg/m²) will be administered to each patient every 3 weeks (46). The MPE control rate at 8 weeks will be the primary endpoint (46).

Similar to IL-2, TNF-a has been studied for its potential role in MPE management. In a study by Li et al (59), 102 patients with lung cancer with MPE received a single dose of recombinant human TNF-a (rhu-TNF) following maximum drainage of the pleural cavity. The results revealed that intrapleural infusion of rhu-TNF sufficiently controlled MPE with a response rate of 81.37% and that the AEs of this treatment were well tolerated (59). The data demonstrated that the short-term efficacy of TNF-a treatment was non-inferior to that of anti-angiogenic treatment, such as bevacizumab (59). However, the study was limited in that it was a retrospective trial.

IFNs are a family of cytokine mediators. Type I IFNs, such as IFN-a and IFN-p, are known to stimulate the immune system and inhibit tumor cell proliferation (60). Due to their role in the immune system, IFNs are being investigated and used in various respiratory disorders, including MPE (60). In 1993, Goldman et al (61) evaluated the safety and efficacy of intrapleural IFN-a2b in patients with MPE. A total of 14/20 (70%) evaluable patients exhibited responses lasting for a median of 6 months, including 8 cases of CR and 6 of PR (61). The response rate of IFN-α2b treatment in that study was encouraging. However, a prospective randomized trial published in 2004 revealed that standard bleomycin chemotherapy was more effective than IFN-a2b in MPE, with patients in the bleomycin group exhibiting a higher response rate and longer survival (62). Since then, studies on IFN-a2b in MPE treatment have been rare.

IFN-β is another type I IFN that has been investigated for MPE treatment. In total, 10 patients with malignant pleural mesothelioma (MPM) or MPE were enrolled in a phase I study to evaluate the safety and feasibility of a single-dose intrapleural IFN-β gene transfer using an adenoviral vector expressing IFN-β (Ad.IFN-β) (63). Intrapleural Ad.IFN-β was generally well tolerated, with 7/10 patients responding to this treatment method (63), which indicated that intrapleural Ad.IFN-β may be a potentially useful approach for the treatment of MPE. A follow-up phase I trial was conducted 3 years later, and 7 patients with MPE and 10 with MPM were enrolled to receive 2 doses of intrapleural Ad.IFN (64). After the first dose, the pleural IFN-β levels were significantly elevated; however, the elevated levels were not sustained, falling to <1 ng/ml after the second dose (64). These unsatisfactory results may have been associated with the rapid development of neutralizing antibodies against the adenoviral vector after the second dose (64).

The aforementioned Ad.IFN-β treatment is an example of using intrapleural immunogene therapy to treat MPE (63,64). For intrapleural immunogene therapy, viral vectors are often used, functioning as an 'in-situ vaccination' (10). Based on the expression of the coxsackie-adenovirus receptor on the tumor cell surface, adenovirus-based viral vectors can selectively infect tumor cells (10). After reaching the pleural space, Ad.IFN efficiently transfects tumor cells. As a result, large concentrations of IFN are produced, serving a role in immunity stimulation and tumor inhibition (65).

Another example of the use of intrapleural immunogene therapy for the treatment of MPE is from a phase I study conducted by Aggarwal et al (66). A total of 19 patients with MPE caused by MPM, NSCLC and breast cancer were enrolled to evaluate the tolerability and efficacy of intrapleural gene-mediated cytotoxic immunotherapy (GMCI) (66). GMCI is an immune strategy that consists of two steps: First, an adenovirus-mediated herpes simplex virus thymidine kinase gene (Ad.V-tk) is administered via intrapleural delivery; next, the anti-herpetic drug valacyclovir is administrated the day after Ad.V-tk infusion (66). Following intrapleural injection, Ad.V-tk efficiently transfected tumor cells to express the thymidine kinase gene, resulting in the production of a large amount of nucleotide analogs, which can interrupt normal DNA replication and trigger tumor cell death, consequently releasing tumor neoantigens (66). It has been revealed that tumor neoantigens from dying tumor cells can also induce antitumor immune responses (10). The results of Aggarwal et al (66) demonstrated that GMCI was safe and well tolerated in patients with MPE, with an encouraging efficacy (DCR of 71%).

Currently, ICI treatment is the mainstay of anticancer immunotherapy. Agents that target cytotoxic T lymphocyte-associated protein 4 (CTLA-4), such as ipilimumab, and PD-1, such as nivolumab, are two major classes of ICIs (67). PD-1, a receptor on the surface of T cells, binds to its ligand (PD-L1 or PD-L2) to decrease T-cell activity, causing apoptosis in cytotoxic T cells (68). CTLA-4 is a key negative regulator of T-cell responses, which can restrict the antitumor immune response (69). ICIs can discrupt these two immunosuppressive signaling pathways, resulting in improved survival outcomes for patients with solid tumors, such as NSCLC, melanoma and colorectal cancer (70-72).

However, there is currently almost no available data on the response of patients with MPE to ICIs. The efficacy of ICIs in MPE remains to be determined by clinical trials. Recently, two retrospective studies suggested that the presence of MPE in patients is a negative predictor of anti-PD-1 antibody efficacy (73,74). However, these results did not suggest that ICIs are not valid for MPE treatment. In a study by Grosu et al (75), a high concordance in PD-L1 expression was identified between histological specimens and matched pleural fluid from patients with NSCLC, suggesting that, if the primary tumor is sensitive to anti-PD-1 treatment, MPE may also be affected by this treatment. The role of the PD-L1/PD-1 pathway in MPE development has also been explored previously (76). In patients with lung cancer with MPE, Prado-Garcia et al (76) found that tumor-responding CD8+ T cells were not completely differentiated into effector cells, which were negative regulated by PD-L1, so that the PD-L1/PD-1 pathway could promote the dysfunction of tumor-responding T cells from MPE.

As a promising therapeutic modality for the treatment of cancer, oncolytic virotherapy has attracted more attention in recent years. Oncolytic virotherapy infects malignant tissues with tumor-specific viruses, causing the lytic destruction of solid tumors (77). Tumor cell lysis caused by viral infection can induce antitumor immune responses (78). A total of 13 patients with MPM were enrolled in a phase I/IIa trial to receive intrapleural oncolytic herpes simplex virus 1716; the systemic immune responses observed further indicated that an intrapleural oncolytic virus could induce antitumor immune responses, which may serve a role in MPE treatment (79). In a previous preclinical study, oncolytic virotherapy using a tumor-specific vaccinia virus represented a novel and promising treatment modality for the treatment of MPE in tumor mouse models (77). However, clinical evidence on this remains limited. Certain phase I/II clinical trials of oncolytic virotherapy in MPE are ongoing (Table II). NCT01766739 is an ongoing phase I study on the intrapleural administration of GL-ONC1, a genetically modified vaccinia virus, in patients with MPE, aiming to assess the safety and efficacy of this treatment method. NCT03597009 is an ongoing phase I/II trial to evaluate the feasibility of adminstering oncolytic virus talimogene laherparepvec (via intrapleural perfusion) and nivolumab (intravenously) in patients with MPE; it is the first clinical trial to combine oncolytic virotherapy and ICIs in MPE treatment.

In the immune system, DCs are the dominant antigen-presenting cells (80). DC-based immunotherapy is a promising cancer treatment method in various types of cancer, including MPE (81). A previous study enrolled 8 patients with late-stage lung cancer who were then injected with autologous DCs generated by culturing adherent mononuclear cells from MPE (81). No grade II/III toxicity was observed (81). Following DC vaccination, an increase in T-cell responses against tumor antigens was observed in 6/8 patients (81). Another 5 patients with MPE or malignant ascites, who were resistant to standard chemotherapy, were treated with combined immunotherapy using monocyte-derived DCs, activated lymphocytes and low-dose OK-432 (a streptococcal preparation) (82). Effusion production was decreased in all of the patients and the mean OS time was >9 months (82). Furthermore, the presence of inflammatory DCs (infDCs) in patients with NSCLC with MPE has been recently observed (83). infDCs represent a distinct human DC subset that can induce T helper 1 cell differentiation in the presence of Toll-like receptor agonists (83). This promising finding may provide a new approach for MPE treatment.

CAR T-cell treatment is another research hotspot in cancer immunotherapy. This technique involves CAR T-cell receptors being specifically engineered to eradicate tumors by recognizing surface proteins expressed on tumor cells (84). In previous studies, CAR T-cell treatment has had substantial clinical success in treating patients with hematological malignancies (85-87). Currently, a growing number of clinical trials of CAR T-cell treatment have focused on solid tumors, targeting surface proteins, including carcinoembryonic antigen, mesothelin, fibroblast activation protein (FAP) and human epidermal growth factor receptor 2 (HER2) (84,88-90). However, to the best of our knowledge, no clinical trial has specifically targeted MPE, despite the local application of CAR T-cell treatment being an attractive approach. A phase I study of intrapleural CAR T cells directed against FAP in patients with MPM with MPE has been completed, but the results have not yet been published (NCT01722149). NCT02414269 and NCT03054298 are two ongoing trials on intrapleural and/or systemic mesothelin CAR T-cell delivery for patients with MPM (Table II). As patients with MPM often present with PE (91), the aforementioned studies will also provide a foundation for controlling MPE via CAR T-cell delivery. FAP and mesothelin are tumor-specific antigens (TSAs) expressed on the surface of MPM cells (92). In order to investigate the efficacy of CAR T-cell delivery for MPE secondary to NSCLC or breast cancer, other targeted TSAs, such as EGFR and HER2, should be investigated.

TILs are cell clusters with an antigen effect resulting from tumorigenesis (93). T lymphocytes, B lymphocytes and natural killer lymphocytes are the main components of TILs (93). Among them, CD8⁺ T lymphocytes mainly exert an anticancer activity (94). In a retrospective study, 27 patients with MPE and ascites were treated with either cisplatin (60 mg) or TILs (100 ml) (95). Compared with patients who received cisplatin, patients who received TILs had a higher ORR (83.33 vs. 33.33%) and DCR (71.43 vs. 28.57%), without severe adverse effects (95). Therefore, TILs may represent a promising treatment method for MPE, and should be investigated further.