Introduction

In the last decade, the treatment landscape of non-small cell lung cancer (NSCLC) changed dramatically, due to the introduction of targeted therapies and immunotherapy. The success of immune checkpoint inhibitor (ICI) therapy in non-oncogene addicted NSCLC [1–4] has led to investigation of these drugs into other thoracic malignancies such as small cell lung cancer (SCLC), malignant pleural mesothelioma (MPM), and thymic epithelial tumors (TETs).

Clinical trials testing immunotherapy in patients with these thoracic cancers have been conducted or are ongoing [5–7]. Unlike NSCLC, where ICIs have represented a breakdown in the treatment armamentarium and have received approval in different settings of disease [8–10], the improvement in outcomes with ICI therapy in rarer thoracic tumors has been somewhat modest and limited to a small proportion of patients [11].

The peculiar pathogenesis underneath these tumors, along with multiple biases in the design of clinical trials were most probably responsible for delaying the availability of effective immunotherapies. Similar to NSCLC, there might be a subgroup of patients more likely to benefit from ICIs [12], but relevant biomarkers have not been determined yet. In addition, lack of funding for clinical research in rare cancer entities such as MPM and TETs has probably precluded the conduction of robust clinical trials that could address specific research questions [13].

What emerges from these recent approvals and from the disappointing results of single-agent ICIs, is the need to properly combine ICIs with other drugs in order to overcome resistance, improve clinical activity and better handle side effects of ICIs in both SCLC, MPM and TETs.

Other conventional therapies such as chemotherapy and targeted therapy, even the ones with different oncological indications, or more recently developed immunotherapies might synergize with ICIs to counteract the immunosuppressive environment of these tumors (Figure 1).

Display full size

Figure 1. 

Immune modulation of tumor microenvironment mediated by main drugs [platinum-based chemotherapy, pemetrexed, angiogenesis targeted agents (such as bevacizumab and sunitinib), poly (ADP-ribose) polymerase (PARP) inhibitors] used in combination with ICIs in pre-clinical works and clinical trials for SCLC, MPM and thymic cancer. PD-L2: programmed death-ligand 2; DC: dendritic cell; TC: tumor cell; MHC-I: major histocompatibility complex class I; MHC-II: major histocompatibility complex class II; NK-cell: natural killer cell; MDSC: myeloid-derived suppressor cell; Tregs: regulatory T cells

In this review, we discuss the current role of ICI treatment in SCLC, MPM and TETs, by examining the results coming from already published clinical trials and explore future perspectives for novel combination therapies that could improve ICIs efficacy (Tables 1–3), by discussing their biological rationale as well as the importance of properly selecting patients for every treatment.

SCLC

Around 15% of lung cancers is represented by SCLC, a neuroendocrine tumor characterized by a high growth fraction and an early development of distant metastasis. SCLC can be staged using both the American Joint Committee on Cancer (AJCC) TNM and the historical Veterans Affairs (VA) classification system, which has only 2 stages: limited (disease confined to the ipsilateral hemithorax, which could be included in a radiation therapy filed) and extensive (anything beyond limited stage). The clinical implication of VA system made it extremely useful even today in clinical practice as well as research.

SCLC is strictly related to tobacco use. Smoking is known to damage DNA [14] and consequently determine a high quantity of somatic mutations, the so-called as tumor mutation burden (TMB), that is a well-documented biomarker for immunotherapy [15, 16]. This feature can lead to a release of tumor neoantigens capable to stimulate an anti-tumoral immune response, thus representing the rationale for immunotherapy activity in this disease [17].

Unfortunately, ICIs showed only moderate benefit when investigated in SCLC. Although the phase I/II CheckMate-032 trial assessing nivolumab with or without ipilimumab in relapsed SCLC documented a 1-year overall survival (OS) of 42% for nivolumab/ipilimumab combination and 30% for nivolumab alone [18], the phase III CheckMate-331 reported that nivolumab was not superior to topotecan in patients with relapsed or progressed SCLC after a platinum-based treatment in terms of OS, progression-free survival (PFS) and objective response rate (ORR) [19]. Notably, a late separation of the curves and a potential activity in the platinum-refractory setting suggests a possible long-term benefit for a subgroup of patients.

The global, double-blind, phase III study compared nivolumab plus ipilimumab or nivolumab alone vs. placebo as maintenance therapy in patients with extensive SCLC who did not progress to first line platinum chemotherapy. Both nivolumab plus ipilimumab and nivolumab alone did not improve OS compare to placebo but maintenance immunotherapy appeared to improve PFS, with rates of patients who were progression-free at six months of 20% and 21% for nivolumab with or without ipilimumab, respectively, versus 10% for placebo [20].

The KEYNOTE-028 and the KEYNOTE-158 trials documented a prolonged durable response for SCLC patients after 2 or more lines [21]. However, these two were a phase Ib and a phase II studies, respectively, and a large randomized controlled trial with pembrolizumab in relapsed SCLC is missing.

Noteworthy, due to the paucity of therapeutic alternatives, nivolumab and pembrolizumab got the FDA approval for the third or later line treatment for SCLC, based on CheckMate-032 and KEYNOTE-028/KEYNOTE-158 trial results. In CheckMate-032 PD-L1 staining seems to not be a predictor of response to nivolumab [22]. Exploratory analysis from the KEYNOTE-158 suggested a role of PD-L1 expression in patients with SCLC who may response to pembrolizumab [21]. Both in CheckMate-032 and in KEYNOTE-158 patients characterized by a high TMB seems to better response to nivolumab or pembrolizumab but further studies are needed to identify new biomarkers of response.

Different mechanisms have been proposed to justify the poor effectiveness of immunotherapy in this disease, such as low PD-L1 expression, downregulation of MHC molecules, immunosuppression induced by SCLC cells and autocrine and paracrine regulation [23, 24].

Therefore, further treatment strategies are required to overcome these mechanisms of ICI resistance, and drug combinations seem to be a promising approach.

MPM

MPM is a rare and aggressive tumor arising from mesothelial cells of the pleura most commonly caused by mineral fibers (such as asbestos and erionite) exposure [58].

Even though surgery and radiotherapy play a role in this pathology, the current established therapy is still systemic chemotherapy with cisplatin and pemetrexed on the basis of the phase III EMPACHIS trial which demonstrated a 3-months survival benefit compared to cisplatin alone (12.1 months vs. 9.3 months) [59, 60].

The limited benefit of first-line chemotherapy and the lack of an effective second-line strategy, with exception of the reintroduction of a pemetrexed-based chemotherapy in patients with durable response to front-line chemotherapy [60] prompt the detection of new therapeutic strategies.

ICIs have been an active field of research in MPM both for their successes in other malignancies and the important role which immune system exerts in the pathogenesis of MPM [61]. Some cases of spontaneous MPM regression likely related to an activation of the immune system have been reported [62] and a worse outcome has been related with high CD163⁺ tumor-associated macrophages and low CD8⁺ tumor infiltrating lymphocytes [63, 64]. Furthermore PD-L1 was shown to be highly expressed in MPM cells and PD-L1 positivity has proven to be an independent risk factor for survival in MPM patients [65].

Therefore, various PD-1/PD-L1 inhibitors have been investigated in patients with disease progression after first-line chemotherapy. In a phase II trial pembrolizumab showed a disease control rate (DCR) of 47% with a PFS of 4.5 months [66] while, in two phase II trials investigating single agent nivolumab, the PFS was 2.6 and 6.1 months, respectively [67, 68]. Similar results were reported for avelumab [69].

A randomized study comparing pembrolizumab with chemotherapy (gemcitabine or vinorelbine) in MPM patients with disease recurrence reported almost equal PFS (2.5 vs. 3.4) and OS (10.7 vs. 11.7) between the two groups [70]. A phase III trial (CONFIRM trial) with randomization to nivolumab or placebo in MPM patients which progressed after at least 2 chemotherapy lines is currently ongoing [71].

Despite initial promising results in phase II trials [72, 73] tremelimumab, a human antibody against CTLA4, failed to demonstrate a benefit compared to placebo in a large randomized trial (mOS of 7.7 months vs. 7.3 months, respectively) [74].

The disappointing results of ICI monotherapy could be due both to the lack of biomarkers able to identify the proper candidate for immunotherapy, since PD-L1 expression seems not to be a reliable prognostic tool [75] and to the immunogenic characteristics of MPM. The limited mutation rate and the resultant low formation of antigens [76, 77] together with the potential upregulation of different inhibitory checkpoints, such as TIM-3 and LAG-3, resulting from anti-PD-1/PD-L1 and anti-CTLA4 drugs use [78], might explain the limited efficacy of ICIs.

TETs

TETs are a heterogeneous group of thoracic cancers, with an annual incidence of about 1.3 to 3.2 per million [96]. The World Health Organization classification stratified TETs into A, AB, B1, B2, and B3 thymomas and TC, taking into account characteristics of malignant epithelial cells and the percentage of non-neoplastic lymphocytes [97, 98]. TC represent about 10–12% of TETs and show a more aggressive clinical behavior and worse overall prognosis compared to thymomas [99]. If complete resection is possible due to early stage, surgery represents the first treatment choice. Locally advanced TETs require a multimodality approach characterized by chemotherapy and radiotherapy in addition to surgery. First-line platinum-based chemotherapy should be considered for advanced non-resectable or metastatic TETs [96]. Treatment options for relapsed or refractory disease after first-line are limited. Imatinib, a KIT TKI, may represent a therapy option for patients with TC who have progressed after first-line chemotherapy, in case of detection of c-KIT gene activating mutations [100, 101]. Sunitinib, an anti-angiogenic multikinase inhibitor, could be used with the same indication, regardless from KIT status [102]. The thymus detains a key role in the development of immune tolerance and, through complex steps, leads to the development of central T-cell tolerance, necessary to avoid the onset of autoimmune diseases [103]. In the thymic medulla, the autoimmune regulator (AIRE) gene and the transcription factor FEZ family zinc finger 2 (FEZF2) promote the expression of tissue-specific antigens (TSAs) in order to develop T-cells. Those cells which react against TSAs undergo apoptosis [104, 105]. About 30% of thymoma patients develop autoimmune disorders, such as myasthenia gravis (the most frequent), pure red cell aplasia, hypogammaglobulinemia, systemic lupus erythematosus and pemphigus [106, 107]. This is due to the downregulation of AIRE and FEZF2, the overthrow of the normal thymic histological architecture, the deficient expression of MHC class II molecules by thymoma cells. These mechanisms lead to the failure of central immune tolerance and susceptibility to auto-immunity [108, 109].

Results obtained from several clinical trials with ICIs as monotherapy are encouraging for TETs showing promising outcome with a response rate (RR) of approximately 20% in pre-treated patient populations [110–113].

Giaccone et al. [110], assessed the activity of pembrolizumab in patients with advanced TC who had progressed after at least one line of chemotherapy, in a single-arm phase 2 study. The results were promising, in fact 22.5% of patients achieved a response; one (3%) patient showed a complete response, eight (20%) patients showed partial responses, and 21 (53%) patients stable disease. Six (15%) patients developed severe autoimmune toxicity, including two (5%) patients with myocarditis. No deaths due to toxicity were observed.

Another phase II study evaluated the efficacy and safety of pembrolizumab in a cohort of 26 patients affected by TC and seven patients affected by thymoma, progressed after at least one line of platinum-based chemotherapy. Of seven thymomas, two (28.6%) obtained partial response, and five (71.6%) obtained stable disease. Of 26 TC, five (19.2%) obtained partial response and 14 (53.8%) stable disease. Taking into account immune-related adverse events (irAEs), five (71.4%) of seven patients with thymoma and four (15.4%) of 26 patients with TC reported grade ≥ 3 irAEs, including myocarditis (three; 9.1%) [111].

As shown by these trials, due to the physiological function of the thymus and the fact that TETs, mainly thymomas, are associated with defective immune tolerance, the use of immunotherapy could expose TETs patients to an increased risk for developing irAEs compared with patients with other malignancies. The incidence of irAEs is high in thymoma patients but also those with TC detain a greater risk of developing severe irAEs. In fact, about 15% of patients, compared to 6% of patients affected by other cancers, develop grade ≥ 3 irAE, including potentially fatal myocarditis [111]. For this reason, an immune baseline check-up and a close monitoring of autoimmunity should be performed in TETs patients treated with ICIs, and their inclusion in clinical trials should be preferred. The detection of high tumor cells PD-L1 expression represents a strong rationale for using ICIs for treatment of TETs [114]. On the other hand, TMB, an emerging potential biomarker for ICIs efficacy, is very low in TETs [115, 116]. In a wide cohort of 100 thymomas and 69 TCs tissue expression of PD-L1, IDO and FOXP3 was analyzed and higher PD-L1 staining was detected in 36% of cases of both thymomas and TCs [114]. A work by Padda et al. [117], showed that TETs with higher PD-L1 expression detained a more aggressive histology (B3 thymomas and TCs) and worse prognosis. A meta-analysis about PD-L1 expression levels in TETs according to histological grade confirmed a significant higher PD-L1 positive rate in type B2/B3 thymoma and TC compared to the type A/AB/B1 thymoma group, suggesting that ICIs might be more effective for the former [118]. Interestingly, in a cohort of 43 patients affected by TCs, the PD-L1 tumor tissue staining increased after induction chemotherapy treatment [119]. This fact emphasizes that chemotherapy, but also targeted and epigenetic therapies [102, 120], holds immunomodulatory activities into the tumor microenvironment and that drug combinations with ICIs could be effective in a specific group of TETs patients. For example, belinostat, a pan-histone deacetylase inhibitor, detained immunomodulatory activity, leading to reduction in blood circulating Tregs and exhausted CD8⁺ T-cell population of TETs patients included in the phase II trial [120], suggesting a potential synergy between epigenetic drugs and immunotherapy.

Considering the overexpression of VEGFA and VEGFR-1 and -2 [121] and the frequent PD-L1 expression both in TETs [122], anti-VEGF agents and ICIs represent an alternative combination strategy.

In a cohort of TETs patients treated with sunitinib (a multiple receptor TKI active against VEGFR-1, -2 and -3, PDGFR-α, PDGFR-β and fibroblast growth factor receptor 1), higher expression of ICIs (i.e. CTLA4 and PD-1) was seen in circulating lymphocytes after sunitinib administration [102], suggesting its immuno-modulatory effect and a potential synergism with ICIs. That represented the rationale for phase II trials that are currently evaluating pembrolizumab plus sunitinib and axitinib plus avelumab, respectively, for treatment of patients with TC. Another phase I/II trial is ongoing to assess the safety, tolerability and efficacy of nivolumab and vorolanib in combination in patients with thoracic tumors including TC. Vorolanib is a new selective VEGFR/PDGFR TKI designed to minimize side effects compared to other anti-angiogenic multikinase inhibitors [123, 124].

Finally, a phase II trial is available for TC patients to evaluate the efficacy of epacadost and pembrolizumab. Epacadostat is a potent indoleamine 2,3-dioxygenase 1 (IDO1) enzyme inhibitor [125]. IDO1 plays a central role in immune regulation through the catabolism of tryptophan to kynurenine [126]. Kynurenine and others tryptophan metabolites induce suppression of effector CD8⁺ T cells function, increase activity of Tregs, inhibit NK-cells and promote expansion of DCs and MDSCs. Therefore, cancers that express high levels of IDO1 may elude immunosurveillance. In TETs, high expression of PD-L1 and IDO was observed in higher-grade forms of tumor histology [114] and for these reasons is reasonable to hypothesize a synergistic effect between epadacostat and pembrolizumab in patients affected by TC.

In conclusion, immunotherapy seems to be a promising therapeutic weapon in TETs, especially for TC. It is important to bear in mind that targeted therapies [127, 128] can also induce severe adverse effects and, due to their immune-modulation activity, potentially exacerbate irAEs linked to ICIs.

In order to better monitor TETs patients receiving immunotherapy, it would be appropriate to include patients in clinical trials and develop an appropriate and accurate monitoring plan that would allow early recognition of irAEs.

Discovering predictive factors, able to discriminate at baseline, before therapy start, patients more likely to develop potentially severe irAEs, would ultimately guide clinicians to the best therapeutic choice.

Conclusions

Since the introduction of PD-L1 and PD-1 inhibitors into the field of NSCLC, several clinical trials attempted to readapt these drugs in other thoracic malignancies. Although a modest single-agent activity has been seen in SCLC, MPM, and TETs, the benefits of ICI monotherapy have remained below expectations. The subtle but substantial differences between these tumor types and within the same tumor type are most probably responsible for delaying the strong affirmation of ICI monotherapy in this setting. Last evidences showed that the molecular profile and the prognosis of these rarer thoracic malignancies might be better explained by a continuous model rather than by a canonical categorical subdivision [129, 130]. Similar to NSCLC, in which most of oncogene-driven tumors are highly resistant to ICIs, it is then improbable that all patients with SCLC, MPM, TETs will benefit from the same ICI strategy. In addition, it is now clear that the interaction between the tumor and the immune counterparts is considerably more complex and cannot be explained by just one immune checkpoint receptor. Spatial [131, 132], temporal, and therapeutic [133] factors can affect checkpoints (both PD-1, PD-L1 and PD-L2) expression in tumor microenvironment. Taking into account these factors in the context of robust international trials supported by a proper outcome selection and strong translational analyses will be mandatory in the next future. As first approvals of anti-PD-1/PD-L1 in SCLC and MPM and also results from early-phase trials have recently shown, combining ICIs with chemotherapy, different ICIs (such as those targeting CTLA4) or even with targeted therapies, might represent a valid solution to overcome resistance and to broaden the population of thoracic cancer patients who may benefit from immunotherapy. However, further insight into potential biomarkers (both at tumor and at patient level) [12, 130–133] as well as into the ideal timing/setting for immunotherapy is needed to get an upfront identification of patients who are likely to respond to these combination strategies and to avoid potentially harmful treatments. The reported toxicity profile of drug combinations draws clinical attention to both ongoing trials and real-world practice, with preexisting paraneoplastic disorders precluding the administration of ICIs combinations in some of these patients. Nevertheless, exploiting all advantages of ICIs in SCLC, MPM, and TETs is crucial, as very few other therapeutic options proved clinically relevant in these diseases.