Molecular Drivers of Mesothelioma

Mesothelioma is a cancer that originates from the mesothelium, a tissue cover that protects the lungs (pleura), heart (pericardium), and abdomen (peritoneum). Although still a rare disease, the worldwide incidence of mesothelioma has increased steadily over the last decade and is expected to peak in 2020.

This lesion has greater incidence at the pleural site (1), where two sheets of mesothelium fold over one another to form a serous membrane that covers the lungs. Between these sheets, a virtual cavity containing a small amount of liquid allows regular lung function (2). Pleural mesothelioma is highly correlated with exposure to asbestos, a substance considered a complete carcinogen because it acts as both an initiator (responsible for the neoplastic transformation of target cells) and a promoter (fiber inhalation induces a chronic inflammatory state with reactive oxygen species production). Other etiological factors include erionite, another fibrous mineral used in the construction industry; ionizing radiation; and the SV40 virus (3). Various studies have also highlighted cases of familial mesotheliomas.

The specific genetic factors involved in neoplastic formation have not yet been determined, but many cases show recurrent mutations in BAP1, NF2, CDKN2A, and TP53 – all genes involved in important cell regulatory pathways (4,5,6).

The diagnosis of mesothelioma can be complicated. Formalin-fixed paraffin-embedded (FFPE) tissue sections are stained with hematoxylin and eosin (H&E), followed by additional immunohistochemical investigations fundamental to differentiating between mesothelioma, pulmonary adenocarcinoma, pleural metastasis of extrathoracic carcinoma, and sarcoma (7,8).

From H&E staining, we can distinguish three different histological structures. The most widespread is the epitheliomorphic histotype, which presents globose cells aggregated in tubulopapillary structures and immersed in a dense and fibrous stroma; the much rarer sarcomatous histotype has fusiform cells arranged disorganizedly in a hyalinized collagen stroma; and the biphasic morphology has at least 10 percent of each of the previous components (9).

In our study in particular, selected samples additionally underwent immunohistochemistry (IHC) to investigate their expression of several proteins: PD-L1, PMS2, MLH1, MSH2, and MSH6. PD-L1, the major PD-1 receptor ligand expressed by immune cells, participates in T-cell stimulation as a negative regulator of the immune response (10,11). PMS2, MLH1, MSH2, and MSH6 are all part of the DNA mismatch repair system; the lack or dysfunction of such proteins leads to microsatellite instability and the onset of mutations due to errors in the replicative phase (12,13).

Taking a closer look

My colleagues and I saw the need to better understand the molecular contributors to pleural mesothelioma, so we turned to IHC. Our study took place at the Santo Spirito Hospital in Casale Monferrato, Italy, where we examined 37 selected cases of malignant pleural mesothelioma diagnosed between 2015 and 2017.

Each biopsy specimen was prepared through a multi-step process:

1. Formalin fixation to preserve the morphological and molecular characteristics that the tissue presented in vivo.
2. An automatic processor dehydrated the tissue in an ascending scale of alcohols (from 25 to 100 percent ethanol).
3. Tissue clarification using diaphans (including xylene, benzene and toluene) to give the samples a transparent – or diaphanous – appearance.
4. Paraffin embedding and sectioning at intervals of 1–3 μm using a microtome.
5. De-paraffinizing and rehydrating the tissue sections.
6. Collecting the tissue sections on a slide.

Once the slides were ready for staining, we applied H&E for routine diagnostics. In the case of suspected neoplasia, we carried out further IHC investigations – monoclonal antibodies for the differential diagnosis of mesothelioma and antibodies against PD-L1 and the mismatch repair proteins (14).

For our anti-PD-L1 antibody, positivity was evaluated based on percentage of PD-L1 expression relative to the amount of mesothelial tissue present. According to the International Association for the Study of Lung Cancer (15), <5 percent expression indicates a negative evaluation, whereas > 10 percent indicates low expression and > 50 percent indicates high expression. For our anti-mismatch repair protein antibodies, we evaluated positivity or negativity of staining at the nuclear level in neoplastic cells, but quantification was not necessary because the nuclear positivity of the tumor on IHC slides tested with monoclonal antibodies against mismatch repair proteins means that those proteins are expressed – and that the DNA can repair its own replication errors, an ability that is not dose-dependent. As long as the neoplastic cells retain their DNA repair function, microsatellite instability is not established.

What we found

The patients in our study ranged from 44 to 89 years old (with an average age of 73). They included 20 male and 17 female patients. Over two-thirds of cases (26; 70 percent) were the epitheliomorphic histotype, whereas nine (24 percent) were biphasic and two (6 percent) sarcomatous. IHC investigations showed that the four proteins involved in the DNA repair system were all expressed – and presumably functioning – in the samples we analyzed. At the microscopic level, we were able to determine differences in color intensity that indicate variable expression of the proteins under examination (see Figures 1 and 2), but these were not considered in the evaluation of positivity.

PD-L1 was expressed in 20 samples (54 percent), 11 of which showed high expression and nine low (see Figure 3). Of the samples expressing PD-L1, 14 were epithelioid, five biphasic, and one sarcomatoid. Thus, approximately half of mesotheliomas present anti-PD-L1 antigens to stimulate the lymphocyte-mediated immune response.

Spotting the protein culprits

Pleural mesothelioma is an aggressive neoplasm with a poor prognosis. The patients examined in this study all had environmental or occupational exposure to asbestos, which further emphasizes the chemical’s key role in mesothelial carcinogenesis – and explains the cancer’s high incidence in the Casale Monferrato area, where numerous companies have worked on asbestos for decades. Few of the patients reported having smoked and, although mesothelioma is more common in men at the national level (16), the cases studied showed no particular correlation with gender.

Histological classification carries particular importance in mesothelioma because recent studies have determined that patients with PD-L1-positive sarcomatous pleural mesothelioma have lower survival rates (17,18,19). Although PD-L1 expression is associated with a worse prognosis, it has a silver lining; novel immunotherapies that target PD-1 and PD-L1 are permeating the market, so PD-L1 expression in pleural mesotheliomas could lead to targeted treatment that acts on the tumor microenvironment. At the moment, three available antibodies act on PD-1 or PD-L1: pembrolizumab, nivolumab, and atezolizumab. Although none of these three drugs is currently approved for the treatment of mesothelioma, several studies are in progress, and preliminary results suggest slight improvements in lifespan and prognosis for eligible patients (20).

DNA mismatch repair proteins correct errors in the pairing of nucleotides and nucleotide loops, the latter of which form at repeated sequences of nucleotides (microsatellites), where the DNA polymerase can introduce errors by inserting or deleting nucleotides. A deficiency in one or more of these proteins causes microsatellite instability (the progressive elongation or shortening of microsatellite sequences), which is present in many tumor suppressor genes. Defects in the production and function of mismatch repair proteins can therefore influence carcinogenesis, as has been shown in several cancers (21). Mismatch repair proteins are currently understudied in pleural mesothelioma; however, one study has shown tumor cells’ reduced ability to correct DNA defects after platinum-based chemotherapy, demonstrating an increased sensitivity to treatment (22).

In our study, every sample showed expression of the four proteins in the mismatch repair complex, indicating that there is no loss of DNA mismatch repair – so it’s unlikely that these proteins are responsible for pleural mesothelial carcinogenesis. It is therefore also unlikely that the RAS genes play a role; although they are the main culprits in epithelial carcinogenesis, other cancers with RAS oncogene involvement also exhibit mismatch repair complex deletions.

Lessons learned

We still lack effective treatments for pleural mesothelioma; fewer than 10 percent of patients survive five years after diagnosis (23). And that’s why my colleagues and I wanted to identify molecular factors implicated in carcinogenesis. We opted for PD-L1 evaluation because of the increasing availability of immunotherapies targeting immune checkpoint inhibitors, such as PD-1/PD-L1, and explored the mismatch repair proteins because of the epitheliomorphic histotype’s similarity to adenocarcinomas of the gastrointestinal tract – most of which have defects in the mismatch repair complex.

Approximately half of the cases we examined were positive for PD-L1, meaning that immunotherapy is a possible treatment option for these patients. None of the cases showed a DNA mismatch repair deficiency, indicating that, despite morphological and biological similarities between adenocarcinoma and epitheliomorphic mesothelioma, the latter does not follow the epithelial carcinogenesis channels.

The oncogenes mutated in mesotheliomas are numerous; with the rarity of the disease, it’s unsurprising that we have not yet identified the genes or gene complexes involved in its carcinogenesis. With the availability of more cases and the increasing communication between laboratories across the world, though, we anticipate expanding and improving our research to identify the involvement of specific molecular pathways that allow the disease to take hold.