Human Pathology (2015) xx, xxx–xxx

www.elsevier.com/locate/humpath

Progress in pathology

Prognostic factors in malignant pleural mesothelioma Ben Davidson MD, PhD ⁎ Department of Pathology, Oslo University Hospital, Norwegian Radium Hospital, N-0310 Oslo, Norway University of Oslo, Faculty of Medicine, Institute of Clinical Medicine, N-0316 Oslo, Norway Received 17 December 2014; revised 6 February 2015; accepted 11 February 2015

Keywords: Malignant mesothelioma; Tumor markers; Diagnosis; Prognosis; Therapeutic targets

Summary Malignant pleural mesothelioma (MPM) is a clinically aggressive tumor originating from mesothelial cells, which line the serosal cavities. Recent years have seen extensive research aimed at identifying new therapeutic targets, predictive markers and prognostic factors in this disease. These include both serum and tissue markers, and are related to multiple cellular pathways which affect cell survival, proliferation, apoptosis, angiogenesis, interaction with the immune response and DNA repair. Several of these molecules may become relevant for pathologists as part of the effort to select patient sub-populations for targeted therapy in the future. This review summarizes current data in this area and discusses their potential clinical relevance. © 2015 Elsevier Inc. All rights reserved.

1. Introduction Malignant mesothelioma (MM), primary cancer of the serosal cavities, has its origin in the mesothelial cells lining the pleural, peritoneal and the pericardial cavities, as well as the tunica vaginalis. The pleural cavity is the most common site of origin of this tumor. Malignant pleural mesothelioma (MPM) may have epithelioid, sarcomatoid or biphasic morphology (Fig.), and the conditions which enter the differential diagnosis are unique to each of these entities. The main differential diagnosis for epithelioid mesothelioma, the most common morphological variant, is metastatic carcinoma, particularly adenocarcinoma, and the most widely used ancillary method in this context is immunohistochemistry (IHC). A wide array of antibodies is presently available for this purpose, including general carcinoma and mesothelioma markers, as well as carcinoma markers which are more organ-specific (reviewed by Ordóñez [1]). Additional methods, including electron microscopy, measurement of ⁎ Department of Pathology, Norwegian Radium Hospital, Oslo University Hospital, Montebello N-0310 Oslo, Norway. E-mail address: [email protected]. http://dx.doi.org/10.1016/j.humpath.2015.02.006 0046-8177/© 2015 Elsevier Inc. All rights reserved.

soluble molecules in the serum and/or effusion specimen and fluorescent in situ hybridization (FISH) may be employed, depending on laboratory expertise, and assessment of the cytology specimen may be informative of mesothelioma of the epithelioid type [2,3]. MPM is one of the most clinically aggressive malignancies, with the majority of patients succumbing to their disease within 2 years of diagnosis. The combination of surgery with chemotherapy, and the optimization of the latter, has in recent years led to some improvement in the survival and life quality of MPM patients, whereas targeted therapy has to date failed to have major clinical impact in this disease [4]. Despite this improvement, long-term survival is rare in MPM, making it difficult to identify biological factors which may clearly differentiate between patients with poor and improved progression-free or overall survival (PFS; OS). Another limiting factor is the difficulty to outperform clinicopathologic factors shown to have predictive or prognostic role in MPM in different clinical models. The most important of these parameters is the histological subtype of the tumor, with sarcomatoid and biphasic MPM being associated with significantly worse outcome than epithelioid MPM. Other parameters reported to be associated

2

B. Davidson

Fig. The morphology of malignant pleural mesothelioma. Examples of epithelioid (A-C), biphasic (D-E) and sarcomatoid (F) malignant pleural mesothelioma, H&E stain.

with adverse outcome in different scoring systems include male gender, poor performance status, disease stage, high serum lactate dehydrogenase (LDH) level, anemia and leukocytosis (reviewed by Campbell and Kindler [4]). Analysis of a large MPM series recently showed that in patients who underwent pathologic staging (n = 550), advanced TNM stage, non-epithelioid histology, age ≥50 years, male gender, non-curative surgery, absence of adjuvant therapy, platelet counts ≥400 000 and white

blood cell counts ≥15.5 were significantly related to poor survival [5]. Nuclear grade was reported to be a strong independent prognostic marker in epithelioid MM [6], whereas serum c-reactive protein (CRP) was an independent prognostic marker of OS in three studies [7–9], in the latter of which it additionally predicted response to multimodality therapy [9]. Low pleural fluid glucose levels were another independent prognosticator of poor OS in the study by Tanrikulu et al [7], whereas low albumin and high fibrinogen

Malignant pleural mesothelioma prognosis levels were significantly related to poor outcome in 2 other studies [10,11]. Response to chemotherapy was recently shown to be significantly associated with OS in analysis of 526 patients included in 9 clinical trials incorporating 6 European Organisation for Research and Treatment of Cancer (EORTC) studies [12]. MM cells have a variety of well-documented chromosomal and genetic aberrations and express a large array of cancer-associated molecules which facilitate invasion and mediate disease progression, (reviewed by Davidson, Mossman et al, Jean et al [13–15]). The present review summarizes the literature related to predictive and prognostic biological factors in MPM. PubMed search using the terms “mesothelioma" and “prognostic" or “prognosis" generated over 2500 hits, of which b5% have assessed the prognostic role of biological parameters in MPM in series of at least 50 patients, the number regarded by this author as the minimum for reducing the risk of type II error. Papers included in this review were also assessed with respect to whether only epithelioid MPM vs. all histological subtypes of MPM were included, as well as regarding whether multivariate survival analysis was performed. It should be noted that while few of the discussed studies have focused on markers reported to have a diagnostic role in MPM (e.g, WT1, calretinin, p16INK4A deletion, and serum markers such as mesothelin, hyaluronan and fibulin-3), the majority of molecules discussed participate in central cellular pathways in multiple cancers, and as such are relevant only in the context of assessing their association with disease outcome.

2. Receptor tyrosine kinases and other cell membrane proteins Cell surface molecules mediate pro-survival and proliferation signaling through binding of their ligands and cross-talk with the tumor microenvironment, including stromal myofibroblasts, endothelial cells and immune response cells. They additionally play a role in adhesion, motility, invasion and angiogenesis. Edwards and co-workers studied the protein expression of the prototype receptor tyrosine kinases (RTK) epidermal growth factor receptor (EGFR) in 168 MPM using IHC. EGFR was expressed in 74 tumors (44%), and its presence was significantly associated with epithelioid histology, good performance status, absence of chest pain and the presence of tumor necrosis. EGFR expression correlated with longer OS in univariate survival analysis, but significance was lost in multivariate analysis [16]. Another study of 71 MPM patients analyzed serum EGFR levels by enzyme-linked immunosorbent assay (ELISA) and tissue EGFR protein and gene expression by IHC and chromogenic in situ hybridization (CISH). EGFR protein and EGFR gene overexpression were detected in 53 (75%) and 37 (52%) tumors, respectively. Both findings, as well as higher than median (N2.5 ng/ml) serum EGFR levels, assessed in 40 patients, were significantly associated with more advanced T-stage, but unrelated to OS [17].

3 Rena et al found EGFR expression by IHC in 38/83 (46%) MPM specimens, with expression limited to epithelioid and biphasic tumors. Gene amplification by FISH was found in only 3/38 IHC-positive tumors. EGFR protein expression was not significantly related to OS for 67 patients with survival data, although 2-year survival was less common in EGFR-positive cases when analysis was limited to epithelioid MM (n = 47), with significant association retained in multivariate analysis [18]. Schramm and co-workers analyzed EGFR protein expression by IHC in 352 MPM and observed significant association between expression of this receptor and epithelioid histology. No association was seen with survival for 128 patients with complete follow-up [19]. c-MET is another RTK which is commonly overexpressed in cancer, including in MPM. In a collaborative study by the MESOPATH group, c-MET and its active form, phospho-c-MET (p-c-MET), were expressed in 119 (76%) and 77 (49%) of 157 studied tumors, respectively. Higher c-MET staining intensity and its localization to the membrane, compared to coexpression at the membrane and cytoplasm or exclusively cytoplasmic localization, were associated with longer OS in univariate analysis, with no such association observed for p-c-MET. c-MET localization to the cell membrane was an independent prognostic marker in multivariate analysis [20]. Pinato and co-workers recently analyzed the protein expression by IHC of Axl, RTK of the TAM subfamily reported to mediate cell survival and epithelial-to-mesenchymal transition (EMT, see below) and its ligand growth arrest signal6 (Gas-6) in MPM. In analysis of 63 tumors, higher Axl expression was significantly associated with epithelioid histology and longer OS, a finding retained in multivariate analysis, as well as in a validation set of 35 patients. Gas-6 was unrelated to prognosis in both sets [21]. Copy number gain of the PDGFRB gene in N40% of tumor cells by FISH was reported to be associated with improved relapse-free survival (RFS) and OS in analysis of 88 MPM [22]. Syndecan-1 is a member of the Syndecan family, heparan sulfate proteoglycans which mediate adhesion, migration, cytoskeletal organization, angiogenesis, differentiation and proliferation. Syndecan-1 was overexpressed in epithelioid compared to sarcomatoid MPM, and its presence was associated with longer OS in analysis of 52 cases, a finding which lost its significance in analysis limited to 33 epithelioid MM [23]. ERC/mesothelin is a membrane-bound 70 kDa precursor protein that can be cleaved to yield a 31 kDa peptide known as megakaryocyte potentiating factor (MPF or N-ERC/ mesothelin) and a membrane-bound 40 kDa protein (C-ERC/ mesothelin). Mesothelin is expressed by normal mesothelial cells, as well as in MM, ovarian carcinoma and pancreatic carcinoma, and has been extensively studied for its potential role in the early diagnosis of MPM [24] and in monitoring treatment response in this disease [25].

4 Higher levels of serum-soluble mesothelin family proteins (SMRP) using the Mesomark ELISA test with a 1 nmol/L cut-off were associated with significantly shorter OS in analysis of 101 MPM patients, a finding retained in multivariate analysis of 52 patients with full clinical data [26]. Comparable results were observed in a study of 96 MPM patients using the same assay, though with a 3.5 nmol/L cut-off, in univariate and multivariate analysis. Pleural mesothelin levels, measured in 58 patients, were unrelated to survival [27]. Schneider et al found significantly higher serum SMRP levels in patients with relapse/ progression of MPM (n = 29) compared to newly diagnosed untreated patients (n = 100). In survival analysis of 91 patients applying a 3.5 nmol/L cut-off, higher SMRP levels were significantly related to OS in univariate and multivariate analysis, although the independent prognostic role of this assay was lost in analysis limited to epithelioid MM [28]. Measurement of serum mesothelin and MPF levels using the Mesomark and MPF ELISA tests, respectively, in 62 MPM patients, showed that both effectively identified patients with stable disease, partial response and progressive disease. However, none of the assays was informative with respect to PFS or OS in survival analysis for 48 patients treated by chemotherapy alone [29]. Serum mesothelin levels dichotomized as high vs. low based on cut-off at 2.7 nmol/L were not significantly related to PFS or OS in analysis of 53 patients, though higher mesothelin levels as continuous variable were associated with shorter PFS [30]. Several other surface markers were studied for their prognostic role in MPM. Alifano and co-workers studied the expression of neurotensin (NTS), regulator of intestinal motility, secretion, smooth muscle activity and epithelial proliferation in physiological conditions, and its receptor NTSR1 in 52 epithelioid MPM using IHC. NTS and NTSR1 were expressed in 37 (71%) and 47 (90%) cases, respectively, although NTSR1 was predominantly expressed at the cytoplasm rather than the cell membrane. NTS expression was associated with poor OS in univariate and multivariate analysis, with no such role for NTSR1 [31]. Tetraspanins are a family of membrane glycoproteins which may have tumor-promoting or -suppressing effects, and which are involved in adhesion, invasion, metastasis and angiogenesis [32]. Expression of the tetraspanin member CD9 by IHC was found in 76/112 investigated MPM, and was significantly related to younger patient age, epithelioid histology and better differentiation. CD9 expression was associated with longer OS in univariate and multivariate analysis [33]. CD26 is a membrane glycoprotein with dipeptidyl peptidase IV activity, which is involved in immune response modulation, primarily via T-cell regulation. The clinical role of CD26 depends on the tumor involved. It is overexpressed in MM compared to normal mesothelial cells [34]. However, analysis of its expression in 79 MPM showed, as with CD9, significant association with epithelioid histology, as well as with longer OS in univariate though not in multivariate analysis for 58 patients who received chemotherapy [35].

B. Davidson Aquaporins are a family of transmembrane channels physiologically involved in water transport, which have been shown to be involved in tumor progression in various malignancies. Aquaporin 1 (AQP1) expression by IHC was analyzed in two MPM cohorts consisting of 80 patients who underwent extrapleural pleuropneumonectomy (EPP) and 56 patients who were treated conservatively. Higher AQP1 expression (score N50%) was associated with significantly longer OS in both cohorts in univariate and multivariate analysis [36]. Analysis of the clinical role of the pro-inflammatory cytokine migration inhibitory factor (MIF) and its receptor CD74 in 135 MPM specimens showed significant association between CD74 protein expression by IHC and longer OS, with no such role for MIF [37]. Caveolae are flask-shaped invaginations of the plasma membrane involved in endosomal transport, adhesion and signaling pathways. Expression of caveolin-1, a major component of caveolae, in peri-tumoral stromal cells of epithelioid MPM was reported to be associated with poor survival compared to tumors with caveolin-1–negative stroma [38]. EMT is a physiological process during embryogenesis which is crucial for formation of the different germ cell layers, structures originating from the neural tube, tubular organs (e.g, the alimentary system) and parenchymal organs developing through budding, such as the liver and pancreas. Pathological EMT in the context of cancer occurs when carcinoma cells lose their epithelial characteristics and acquire certain mesenchymal properties that promote ECM invasion and distant metastasis [39]. Fassina et al analyzed 109 MM, of which 74 were pleural, for the expression of molecules involved in EMT, including cadherins, matrix metalloproteinases (MMP), and the E-cadherin suppressors Snail, Slug, Twist, Zeb1 and Zeb2, using IHC and quantitative PCR (qPCR). Reduced expression of epithelial markers and increase in mesenchymal markers in the transition from epithelioid through biphasic to sarcomatoid MM was seen. Patients with epithelioid MM who had higher E-cadherin expression had longer survival in univariate analysis [40]. In another study of 56 MPM, E-cadherin, Zeb, Twist and Snail expression was studied using IHC. E-cadherin was significantly overexpressed in epithelioid compared to biphasic and sarcomatoid MPM, and its expression was unrelated to that of its three inhibitors. Snail expression was associated with longer OS, retained in multivariate analysis, with no such relationship for the three remaining markers [41]. The latter finding does not concur with the study of Kobayashi and co-workers, in which analysis of 109 MPM showed higher Snail expression and Zeb1 expression in sarcomatoid compared to epithelioid tumors, and significantly poorer survival for patients with Snail-expressing tumors, a finding retained in multivariate analysis [42]. Two proteins investigated as part of the EMT panel chosen by Schramm et al, integrin β1 and integrin-linked

Malignant pleural mesothelioma prognosis kinase (ILK), were expressed by MPM cells [19]. The former was unrelated to survival, whereas the latter was associated with longer OS, albeit when localized to the cytoplasm rather than the cell membrane. ILK expression was analyzed in an additional study of 80 MPM, in which no significant association with survival was observed in the entire cohort, although ILK expression was associated with longer survival in a sub-group of 28 conservatively treated patients [43]. Studies discussed in this section have analyzed MPM of all histological subtypes, with the exception of one study [31], although two additional studies report significant findings in survival analysis limited to epithelioid MPM [18,40].

3. Angiogenic molecules, other growth factors and cytokines MM cells synthesize an array of angiogenic molecules and other growth factors and cytokines. Several studies analyzed the prognostic value of these markers, all including MPM specimens of all histological subtypes. Microvessel density (MVD) was reported to be an independent prognostic marker in a study of 93 MPM [44]. Kumar-Singh et al studied 52 MPM for expression of vascular endothelial growth factor (VEGF), the fibroblast growth factor family members FGF1 and FGF2, and transforming growth factor-β (TGFβ). FGF2 expression was associated with worse OS, with no such role for the other markers [45]. Pleural effusion levels of VEGF are higher in MPM compared to effusions from patients with non-malignant pleural disease [46]. Analysis of serum VEGF levels in 63 patients treated with thalidomide alone or combined with cisplatin/gemcitabine showed a negative predictive and prognostic role for high VEGF levels, the latter retained in multivariate analysis [47]. The serum levels of Angiopoietin-1 (Ang-1), ligand of the endothelial cell RTK Tie-2, were higher in MPM patients compared to individuals with fibrous plaques or no disease following asbestos exposure. For 50 patients with follow-up data, levels N34.5 ng/ml were associated with worse OS in univariate and multivariate analysis [48]. Circulating endothelial cells (CEC), detected by an antiCD105 antibody, were recently suggested to be a surrogate of tumor angiogenesis in analysis of 79 MPM. CEC numbers were significantly related to MVD by CD34 staining, and numbers N50/4 mL were associated with poor survival in univariate and multivariate analysis [49]. Serum levels of platelet-derived growth factor-AB (PDGF-AB) were higher in 93 MPM patients compared to controls exposed to asbestos and healthy controls. Survival analysis of 82 patients with follow-up data showed a nonsignificant trend for shorter survival [50].

5 Expression of Thrombospondin-1 (TSP-1), a secreted glycoprotein which may function as angiogenesis inhibitor or stimulator, was assessed in 78 MPM. High TSP1 mRNA expression was observed in 74 tumors, and mRNA levels correlated with those of VEGF. TSP1 mRNA expression was higher in early stage disease and in patients with no lymph node metastases compared to more advanced-stage tumors, but did not predict survival in the entire cohort [51].

4. Proliferation, apoptosis and cell survival Cancer cells are characterized by sustained proliferation, overriding pro-survival signaling and the ability to avoid apoptosis [52], and several studies have focused on this aspect of MM biology. All studies included tumors of both epithelioid and sarcomatous morphology. O’Kane et al analyzed 54 MPM for protein expression by IHC of p53, anti-apoptotic Bcl-2 family members (Bcl-2, Bcl-XL, Mcl-1) and pro-apoptotic Bcl-2 family members (Bad, Bak, Bax, Bid and Bim). p53 was overexpressed in 81% of tumors, whereas expression of anti- and proapoptotic Bcl-2 family members ranged from 24%-92% and 24%-42%, respectively. None of these proteins was significantly associated with survival [53]. Another study analyzed the protein expression of the cell cycle inhibitors p53, p21 and p27, as well as the prostaglandin synthesis pathway enzyme cyclooxygenase-2 (COX-2), in 77 MPM. High COX-2 and low p21 and p27 expression was associated with significantly shorter OS. None of these markers was an independent prognosticator in multivariate analysis, although the combination of all three did so [54]. Association between p21 and p27 protein expression and longer OS was also reported in the study by Schramm et al, although this was not retained in multivariate analysis [19]. In another study, protein expression of p14, p16, p21, p27, p53 and its negative regulator Mdm-2, and Rb was studied in 55 MPM using IHC. The p14ARF and p16INK4A genes were analyzed for promoter hypermethylation and homozygous deletion. Dysregulation of the studied proteins was observed in about 50% of cases, and altered p14 and p16 expression was associated with genetic aberrations in the corresponding genes. p27 and Rb loss correlated with shorter OS and disease-free survival in multivariate analysis [55]. The association between p27 and longer survival in MPM is supported by another study, in which expression of this protein was significantly higher in tumors of long-term (n = 27) compared to short-term (n = 36) survivors, defined at 24 months cut-off [56]. Mdm-2 protein expression by IHC was recently reported to be associated with poor survival in analysis of 61 MPM [57]. The clinical role of p16INK4A deletion, a frequent molecular event in MPM, was further analyzed in a study of 54 MPM using FISH. Homozygous p16INK4A deletion

6 was found in 33 (61%) tumors and was associated, together with older age and asbestos exposure, with shorter OS in univariate analysis [58]. The inhibitor of apoptosis (IAP) family constitutes a major negative regulator of apoptosis in cancer, mainly through the inhibition of caspases. Eight human IAPs have been identified: cellular IAP1 (c-IAP1), cellular IAP2 (c-IAP2), neuronal apoptosis inhibitory protein (NAIP), Survivin, X- linked IAP (XIAP), Apollon, testis-specific IAP (Ts-IAP), and Livin [59]. Analysis of the mRNA and protein expression of 5 IAP family members: IAP-1, IAP-2, XIAP, Survivin and Livin in MPM showed frequent expression of these molecules. Survival analysis of protein expression in 66 tumors showed significant association between Livin expression and longer survival [60]. XIAP and Survivin protein expression was studied in 112 MM specimens (77 pleural, 35 peritoneal; 68 solid, 44 effusions) in another study. Pilot analysis of 10 MM effusions by immunoblotting showed frequent XIAP and Survivin expression, with uniform absence of Livin. IHC analysis of the two former proteins showed their expression in the majority of specimens. However, XIAP expression was significantly higher in peritoneal MM compared to MPM and in effusions compared to solid lesions, whereas nuclear Survivin was downregulated in effusions compared to solid tumors. Higher XIAP expression was associated with a trend for poor OS (P = .06) for 62 patients with survival data in univariate analysis [61]. Survivin expression by IHC did not have a prognostic role in a recent analysis of 101 MPM specimens [62]. The phosphatidylinositol 3-kinase (PI3K) pathway is a major regulator of cell metabolism, growth, migration, survival and angiogenesis, and its components, primarily PI3K, AKT and mTOR, are deregulated in multiple cancers [63]. It is negatively regulated by the tumor suppressor PTEN (phosphatase and tensin homolog deleted on chromosome 10) [64]. SV-40 was shown to induce survival in mesothelioma cells via AKT activation [65]. Analysis of p-AKT expression in 80 MPM by IHC did not demonstrate any association with survival [43]. Data regarding PTEN are inconclusive. PTEN localized to the tumor cell cytoplasm was strongly associated with longer OS in univariate and multivariate analysis in the study by Schramm [19]. These results are comparable to previous data from the same group in a comparable series size [66]. The relationship, if any, between these cohorts was not clarified, although both series were from the same period. In contrast, PTEN expression was unrelated to survival in a study from another group [67]. In a third, more recent, study by the above-mentioned group, high expression by IHC of Ki-67 and p-S6 kinase, the latter a downstream molecule in the PI3K pathway, was significantly related to poor OS in chemo-naïve MPM, whereas only Ki-67 had such a role in post-chemotherapy specimens. Reduced cytoplasmic PTEN and increased p-mTOR expression in post-chemotherapy

B. Davidson compared to patient-matched pre-chemotherapy specimens was additionally related to poor OS [68].

5. The immune response The immune response in cancer involves the production of multiple molecules designed to kill cancer cells and suppress tumor growth. In reality, this process often achieves the opposite effect due to the ability of tumor cells to utilize host-produced molecules as growth promoters and to evade immune response–related killing mechanisms. [52] The prognostic role of immune response parameters has been investigated in several studies, of which all but one ([79], see below), included patients with MPM of all subtypes. Cultured MPM cells isolated from pleural effusions and cultured with immunogenic antigens and autologous lymphocytes express molecules required for antigen presentation, including the major histocompatibility complex (MHC) antigen MHCII-DR, intracellular adhesion molecule-1 (ICAM-1) and B7-1 (CD80.3) [69]. They further express interleukin-2 (IL-2) receptor, and IL-2 induces cell growth inhibition in highly proliferating MPM cells in vitro [70]. Expression of another B7 family member, B7-H3, was predominantly observed in epithelioid MPM in analysis of clinical specimens [71]. The expression of B7-H1, a negative co-stimulatory molecule preventing T-cell activation normally expressed on cells of macrophage lineage, was recently studied in 106 MPM. B7-H1 expression on MPM cells, observed in 40% of tumors, was related to sarcomatoid histology and was an independent prognostic factor of poor survival [72]. In a retrospective analysis of 173 MPM patients, including 119 patients undergoing first-line and 54 patients undergoing second- or third-line therapy, white blood cell (WBC) counts b8.3 x 109 and neutrophil-to-lymphocyte ratio (NLR) b5 were associated with significantly longer OS in univariate survival analysis, of which the latter was an independent prognosticator in multivariate analysis [73]. In another study by the same group, which was limited to 85 patients who underwent EPP, NLR was again significantly related to survival in both univariate and multivariate analysis, whereas WBC counts were not. In addition, immunostaining for the mesothelial markers D2-40 and calretinin was associated with longer OS, the latter also in multivariate analysis [74]. In a study of 171 MPM by another group, WBC counts and NLR using the above cut-offs, as well as platelet-to-lymphocyte ratio (PLR) at a cut-off of 300, were significantly associated with survival in univariate analysis, and significance was retained for NLR in multivariate analysis [75]. NLR cut-off at 3 rather than 5 was reported to be informative of prognosis in another study [76]. The role of NLR as a prognostic factor in MPM was disputed by Meniawy et al who found no association between this parameter and OS in analysis of 274 patients [77].

Malignant pleural mesothelioma prognosis Higher blood monocyte and neutrophil counts were associated with poor OS in analysis of 667 MPM patients, and the former parameter was an independent prognosticator in multivariate analysis [78]. Suzuki et al evaluated the inflammatory response semiquantitatively in H&E-stained slides from 175 patients with epithelioid MPM, none of whom received neoadjuvant chemotherapy. Patients with more intense chronic inflammation had longer survival by univariate and multivariate analysis [79]. Expression by IHC of the mast cell enzymes tryptase and chymase was studied in 60 MPM. Higher numbers of tryptase-positive mast cells were associated with longer time to progression and OS in univariate and multivariate analysis, whereas chymase expression was unrelated to outcome [80].

6. DNA repair and chemotherapy response Defective or dysregulated DNA repair mechanisms are frequently observed in cancer cells and affect their biological behavior and their response to chemotherapy [81]. The most extensively studied molecule in this context is excision repair cross-complementation group 1 (ERCC1), a DNA repair protein which belongs to the nucleotide excision repair (NER) family, a group of proteins which remove cisplatininduced DNA adducts, thereby mediating resistance. Righi et al studied the expression of ERCC1 and thymidylate synthase (TS), the cellular target of pemetrexed, in 60 MPM from patients treated by pemetrexed alone or in combination with cisplatin, as well as in 81 MPM not treated by pemetrexed, using qPCR and IHC. Lower TS protein, but not mRNA expression was significantly related to longer time to progression and OS in univariate and multivariate analysis. No such association was seen for ERCC1 for 45 patients treated by cisplatin or for TS in patients not treated with pemetrexed [82]. In another study, TS and ERCC1 protein and mRNA expression, as well as polymorphisms of these genes, were analyzed in 126 MPM, of which 99 were from patients treated with pemetrexed and cisplatin [83]. In the latter group, lower TS protein and mRNA expression was significantly associated with disease control, longer PFS and longer OS in univariate and multivariate analysis, with no such role for ERCC1 protein or mRNA, or for polymorphisms of these genes. Zimling et al immunostained 50 MPM from patients treated with vinorelbine and cisplatin for ERCC1, class III β-tubulin and BRCA. Class III β-tubulin was the only protein related to survival. However, tumors expressing both ERCC1 and class III β-tubulin were considered resistant and were associated with shorter PFS and OS, while those negative for both proteins had longer survival, a difference reproduced in multivariate analysis [84].

7 Two recent studies by one group analyzed the clinical role of alterations in genes related to DNA repair. The first study analyzed polymorphisms in the genes XPD, ERCC1, GSTP1, and deletion of the genes GSTM1 and GSTT1 in 133 MPM, including 97 from patients treated with platinumbased chemotherapy. Wild-type XPD 751Lys/Lys and ERCC1 8092C/C genotypes were significantly associated with reduced risk of progression in platinum-treated patients, whereas GSTT1 polymorphism were significantly related to PFS in patients not treated with platinum. The finding related to ERCC1 8092C/C remained significant in multivariate analysis [85]. The second study assessed the presence of polymorphisms in the genes XRCC1, XRCC3, NBN and RAD51 in 109 MPM. Significant association was found between XRCC1399Gln allele and shorter OS [86]. Ting et al recently studied protein expression of the DNA repair molecules ERCC1, MLH1, MSH2, MSH6, as well as class III β-tubulin, in MPM. ERCC1 codon 118 and C9092A polymorphisms were additionally analyzed by pyrosequencing. In pre-chemotherapy specimens (n = 79), MSH6 protein expression was significantly related to progression under chemotherapy, whereas MLH1 and ERCC1 expression, as well as ERCC1 codon 118 polymorphisms were related to PFS. ERCC1 protein expression was significantly related to shorter OS [87]. Another study related to chemotherapy in MPM analyzed 84 tumors for protein expression by IHC of TS and folylpoly-γ-glutamate synthetase (FPGS), the enzyme that converts pemetrexed to more effective polyglutamated forms. The majority of patients (n = 79) were treated with pemetrexed combined with platinum, whereas the remaining 5 received single-agent pemetrexed. High PFGS expression was significantly associated with better response to chemotherapy and longer PFS, whereas high TS expression was related to shorter OS [88]. These results were recently challenged in a study of 85 MPM from patients who received pemetrexed, in which neither TS nor FGPS were significantly related to disease control rate, time to progression or OS [89]. Two earlier studies analyzed the protein expression by IHC and clinical relevance of glutathione S-transferases (GST), detoxification enzymes that have been related to chemotherapy response and prognosis in different cancers. Analysis of 88 MPM for GST-α, -μ and -π showed expression of at least one GST member in the majority of cases, but no association to survival [90]. Analysis of 57 MPM (56 with survival data) in another study similarly showed frequent expression of these three GST proteins. However, expression of GST-π and GST-μ was significantly associated with longer OS [91]. Another study by the same group identified nuclear size and proliferation index by image analysis as parameters significantly related to survival [92]. All studies discussed in this section included MPM of different subtypes.

8

B. Davidson Table 1 Ref. 16 17 18 19 19 19 43 20 20 21 21 22 23 26 27 28 29 30 31 31 33 35 36 37 38 40 41 41 42 41 41

Membrane molecules Molecule EGFR EGFR EGFR EGFR β1 integrin ILK ILK c-MET p-c-MET Axl Gas-6 PDGFRB Syndecan-1 Mesothelin d Mesothelin d Mesothelin d Mesothelin d Mesothelin d,f NTS NTSR1 CD9 CD26 AQP1 CD74 g Caveolin-1 h E-cadherin E-cadherin Snail Snail Twist Zeb1

Expression level Protein Protein + Gene Protein Protein Protein Protein Protein Protein Protein Protein Protein Gene Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein

No. 168 71 67 128 128 128 80 157 157 63 63 61 52 101 96 91 48 53 52 52 112 58 80 135 63 58 56 56 63 56 56

Histology a

All All All All All All All All All All All All All All All All All All Epithelioid Epithelioid All All All All All Epithelioid All All All All All

Univariate

Multivariate

Prognosis

P = .01 NS NS c NS NS P = .022 NS P = .02 NS P = .001 NS P = .029 P = .033 P b .001 P = .009 P = .003 NS NS P = .015 NS P = .025 P = .008 P b .001 P b .001 P = .002 P = .042 NS P = .046 P = .0016 NS NS

NS NS NP b NP NP NS NP P = 0.043 c NP P = .01 NP P = .02 NP P = .02 e P = .004 P = .025 NP NP P = .007 NP P = .026 NS P = .02 P = .003 NS NP NS P = .036 P = .0142 NS NS

Good Good Good Good Good Good Poor Poor Poor Poor Good Good Good Good Poor Good Good Poor -

Abbreviations: EGFR, epidermal growth factor receptor; NS, not significant; NP, not performed; ILK, integrin-linked kinase; PDGFRB, platelet-derived growth factor receptor B; NTS, neurotensin; NTSR1, neurotensin receptor 1; AQP1, aquaporin 1. a All = epithelioid, biphasic and sarcomatous histology. b Significant for patients with epithelioid histology (P = .029 and P = .04 in univariate and multivariate analysis, respectively). c Limited to c-MET–positive specimens, with exclusive membrane expression vs. cytoplasmic or cytoplasmic and membrane localization as analyzed factor. d Serum levels by ELISA. e For 52 patients. f In addition, association with shorter PFS as continuous variable (P = .013). g In addition, calretinin associated with longer survival, MIF unrelated to survival. h Stromal expression; tumor cell expression unrelated to survival.

7. Miscellaneous Several other markers have been assessed for their clinical role in MPM, of which the most well-studied is osteopontin (OPN). OPN, a member of the SIBLING (small integrin-binding ligand N-linked glycoprotein) family, is a secreted phospoprotein present in all body fluids. An intracellular form of OPN (iOPN) exists as part of a membrane complex including CD44 and the ERM family members ezrin/radixin/moesin. Alternative splicing and posttranslational modifications of OPN result in a variable molecular weight of 41-75 kDa. OPN contains an integrin-binding site, a thrombin cleavage site, and sites susceptible for cleavage by MMP. Transcrip-

tional regulators of OPN include OCT-1, AP-1, Myc, p53 and the TGF-β pathway. OPN is involved in adhesion, motility, metastasis, inflammation and angiogenesis, as well as in regulation of apoptosis and the immune response [93,94]. OPN was initially reported to be useful in the differentiation between individuals with MPM and those with asbestos exposure who did not have cancer [95], although later studies did not support a role for this marker in differentiating these patient groups [27] or in monitoring disease response [29]. Assessment of the prognostic value of OPN in the study by Grigoriu et al showed, as for mesothelin, significant association between higher OPN serum levels at a 350 ng/mL cut-off, but not pleural levels, and poor survival in univariate and

Malignant pleural mesothelioma prognosis

9

Table 2

Cytokines and growth factors

Ref.

Molecule

Expression level

No.

Histology

Univariate

Multivariate

Prognosis

44 45 45 45 45 47 48 49 50 51

CD34 (MVD) FGF1 FGF2 VEGF TGFβ Serum VEGF Serum Ang-1 CEC Serum PDGF-AB TSP-1

Protein Protein Protein Protein Protein Protein Protein Cells Protein mRNA

93 52 52 52 52 63 50 79 82 69

All All All All All All All All All All

P = .02 NS P = .001 NS NS P b .05 P = .029 P = .03 NS NS

P = .01 NP NP NP NP P = .025 P = .036 P b .01 NP NP

Poor Poor Poor Poor Poor -

Abbreviations: MVD, microvessel density; FGF, fibroblast growth factor; NS, not significant; NP, not performed; VEGF, vascular endothelial growth factor; TGFβ, transforming growth factor β; Ang-1, angiopoietin 1; PDGF-AB, platelet-derived growth factor AB.

multivariate analysis [27]. Using different cut-offs, Hollevoet et al also found significant association between higher OPN serum levels and poor PFS (862.78 ng/mL cut-off) and OS (1141.11 ng/mL cut-off) in survival analysis for 48 patients treated by chemotherapy alone [29]. In another study, OPN tissue expression by IHC was compared between 32 MPM from long-term survivors and 69 MPM from short-term survivors at a 24-month cut-off [96]. OPN expression was significantly higher in tumors from long-term survivors, and high expression was significantly related to shorter survival in univariate and multivariate analysis. Other markers evaluated for their prognostic role in MM were predominantly in single studies. High levels of serum cytokeratin 19 fragment (CK19) using the Cyfra-21 assay and of tissue polypeptide antigen (TPA) were associated with poor survival in univariate and multivariate analysis in analysis of 51 MPM patients [97]. COX-2 expression by IHC was associated with longer survival in univariate and multivariate analysis in a series of 86 MPM [98]. Pinton et al similarly reported an association between estrogen receptor-β (ERβ) and longer survival in IHC analysis of 78 MPM [99]. In another study, expression of ERβ and the steroid receptor coactivators (SRC) SRC-1, SRC-2/TIF-2 and SRC-3/AIB-1 was studied in 89 MPM using the same method [100]. Higher SRC-2/TIF-2 expression was associated with longer OS, while the remaining proteins had no prognostic value. GATA-6, a member of the GATA family of zinc finger transcription factors, was more highly expressed in MPM compared to lung adenocarcinoma in IHC analysis. Survival analysis for MPM patients (n = 63) did not show association with survival, although nuclear GATA-6 expression was associated with longer OS in analysis limited to 42 epithelioid MM [101]. Hyaluronan levels in pleural effusions, measured by liquid chromatography, were associated with longer survival in analysis of 100 MPM patients, a finding retained in multivariate analysis at a cut-off of 225 mg/L [102]. This was

reproduced in a recent study of 96 MPM, in which effusion hyaluronan levels was both useful in differentiating MPM from adenocarcinoma and benign conditions, and prognostic, the latter analysis showing significant association between higher hyaluronan levels and longer survival [103]. The levels of another MPM marker, fibulin-3, performed less well than those of mesothelin in both plasma and effusion samples as a diagnostic marker. However, effusion fibulin-3 levels provided prognostic information, with higher levels associated with shorter survival [104]. Expression of HtrA1, a member of the HtrA (high temperature requirement A) serine protease family was studied in 70 MPM using IHC. High HtrA1 expression was associated with significantly better OS in univariate and multivariate analysis [105]. Protein expression of periostin, a matricellular glutamatecontaining connective tissue protein expressed primarily in tissues subjected to mechanical stress, such as bones, tendons and ligaments, was studied by Schramm et al [19]. Cytoplasmic periostin expression in MPM cells was associated with shorter OS in univariate and multivariate analysis. Expression of the endoplasmic reticulum stress markers GADD34, BiP1 and GADD153/CHOP1 was recently studied in 135 MPM using IHC. Higher (upper quintile) CHOP1 score was associated with worse OS, with no prognostic role for the two remaining proteins [106]. MMP-14, identified as an overexpressed gene in MPM (n = 9) compared to benign pleura (n = 4) in gene expression analysis, was reported to be a marker of poor prognosis in a validation set of 70 patients using IHC [107]. Methylation of the p15 INK4B , p16INK4A , RASSF1A, NORE1A genes was found in 5%-20% of 79 MPM and was not significantly associated with survival [108]. Wilms' tumor antigen (WT1) expression was associated with longer OS in analysis of 52 MPM, whereas calretinin expression was unrelated to survival [109]. Higher serum levels of high-mobility group box 1 (HMGB1), a transcriptional regulator involved in proliferation, DNA repair and inflammation, were related to shorter OS based on cut-off at 9 ng/ml in analysis of 61 MPM [110].

10 Table 3

B. Davidson Apoptosis, cell cycle and proliferation

Ref.

Molecule

Expression level

No.

Histology

Univariate

Multivariate

Prognosis

55 55 58 19 54 55 19 54 55 53 54 55 57 53 53 53 53 53 53 53 53 54 55 57 55 60 60 60 60 61 b 60 61 b 62 43 19 66 67 68 68

p14 p16 p16 p21 p21 p21 p27 p27 p27 p53 p53 p53 p53 Bcl-2 Bcl-XL Mcl-1 Bad Bak Bax Bid Bim COX-2 Mdm-2 Mdm-2 Rb IAP-1 IAP-2 Livin XIAP XIAP Survivin Survivin Survivin AKT PTEN PTEN PTEN PTEN Ki-67 c

Protein Protein DNA Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein Protein

55 55 54 128 77 55 128 77 55 54 77 55 61 54 54 54 54 54 54 54 54 77 55 61 55 66 66 66 66 62 66 62 101 80 128 129 86 75 75

All All All All All All All All All All All All All All All All All All All All All All All All All All All All All All All All All All All All All All All

NS NS P = .039 P = .006 P = .0001 NS P = .023 P = .001 ND NS NS NS NS NS NS NS NS NS NS NS NS P = .0001 NS P = .005 ND NS NS P = .044 NS NS NS NS NS NS P b .001 P b .001 NS NS P = .02

NS NS NP NS NS NS NS NS P = .011 NP NP NS NP NP NP NP NP NP NP NP NP NS NS NP P = .003 NS NS NP NS NS NS NS NS NP P = .007 P = .003 NP NP NP

Poor a Good Good Good Good Good Poor Poor Good Good Good Good Poor

Abbreviations: NS, not significant; NP, not performed; ND, not detailed; COX-2, cyclooxygenase 2; Mdm-2, mouse double minute 2 homolog; Rb, retinoblastoma protein; IAP, inhibitor of apoptosis; XIAP, X-linked inhibitor of apoptosis; PTEN, phosphatase and tensin homolog. a Poor survival when homozygous deletion was present. b Including both pleural and peritoneal MM. c In post-chemotherapy specimens; P = .03 for PFS. In addition, association between high Ki- 67 (P = .04) and pS6 and (P = .02) poor PFS in pre-chemotherapy specimens.

Low levels of EMX2, a member of the homeobox family of transcription factors, by qPCR, were significantly related to shorter PFS in analysis of 50 MPM, a finding retained in multivariate survival analysis [111]. Protein expression of the tumor suppressor BRCA1associated protein (BAP1) was reported to be related to worse survival in analysis of 123 MPM [112]. As in the majority of reports discussed in this review, studies of miscellaneous molecules included all MPM subtypes.

Data relating the above-discussed molecules and OS are summarized in Tables 1–6.

8. High-throughput methodology As in practically any other cancer form, the possibility to harness high-throughput methodology for better diagnosis and prognostication of MPM has been explored by several

Malignant pleural mesothelioma prognosis Table 4

11

Immune response

Ref.

Parameter

No.

Histology

Univariate

Multivariate

Prognosis

72 73 73 74 74 75 75 75 76 77 78 78 79 80 80

B7-H1 WBC count b8.3 x 109 NLR b5 WBC count b8.3 x 109 NLR b5 WBC count b8.3 x 109 NLR b5 PLR b 300 NLR b3 NLR b5 High monocyte count High neutrophil count Chronic inflammation Tryptase Chymase

106 173 173 85 85 171 171 171 155 274 667 667 175 60 60

All All All All All All All All All All All All Epithelioid All All

P b .0001 P = .008 P b .001 NS P b .01 P = .04 P = .006 P = .03 P = .004 NS P = .0002 P = .02 P = .01 P = .02 a NS

P = .04 NS P b .001 NP P = .04 NS P = .008 NS P = .009 NS P b .0001 NP P = .02 P = .02 NS

Poor Good Good Good Good Good Good Good Poor Poor Good Good -

Abbreviations: WBC, white blood cells; NS, not significant; NP, not performed; NLR, neurophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio. a P = .01 for time to progression in univariate analysis.

groups. Studies in which large series of tumors were analyzed are listed below. Gordon et al analyzed the gene expression profiles of 17 MPM and identified a prognostic profile consisting of 46 genes, of which the ratios of 4 genes (KIAA0097, GD1A1, L6-related EST, CTHBP) were found to provide the most accurate prognostic information. This was validated in a set of 29 tumors in the same study [113]. A subsequent study from this group applied the 4-gene set, as well as a set of the 20 best differentiating genes, to a series of 39 uniformly treated patients, and found these signatures to be informative in this cohort as well [114]. Analysis of the new cohort led to identification of an 8-gene set, including 4 genes associated with good outcome (EST DKFZp586J2118, CD9, DLG5, C3) and 4 related to poor outcome (KIAA1199, 2 copies of CD24, THBD), which was then validated in a set of 52 tumors. The 4-gene signature from the first study was found to be prognostic in a third study in which 120 MPM were analyzed using qPCR [115]. López-Ríos et al performed gene expression array analysis of 99 MPM from a cohort in which advanced stage, sarcomatous histology and the presence of p16/CDKN2A deletions were found to be independent poor prognostic parameters in multivariate analysis [116]. A 20-gene set differentiated between epithelioid and sarcomatous MPM, and a 35-probe set representing 29 genes provided the best prognostic information. Among the genes identified in the prognostic set were several related to the cell cycle, including those coding for Aurora kinases A and B, cyclin D1, Survivin and CDC25A. The latter test, at 65% accuracy, was nevertheless weaker than the above-mentioned clinicopathologic variables, as were the above-discussed gene sets identified by Gordon et al and another set obtained in a smaller series by Pass et al [117].

de Reyniès et al recently applied microarray technology to primary MPM cultures and identified 2 MPM sub-groups, termed C1 and C2. The latter group contained all tumors with sarcomatoid and desmoplastic histology, whereas epithelioid tumors were found in both groups. Mutations in CDKN2A and BAP1 were more frequent in C1 tumors, whereas C2 tumors had EMT characteristics. Patients with C2 tumors had worse prognosis than C1 tumors irrespective of their histology. The PPL, UPK3B and TFPI genes were identified as classifiers between the C1 and C2 group in this paper [118]. MicroRNA array analysis of 37 MPM and a test set of 92 MPM identified has-miR-29c* as a marker of longer time to progression and overall survival [119].

9. Future directions In MPM, as in other cancers, whole genome analyses and more focused analyses of commonly mutated genes are likely to play a central role in the choice of targeted therapy and clinical management in general in the near future. Several studies using these molecular platforms have been published in recent years. Sequencing of a biphasic MPM specimen revealed multiple genomic changes in the form of deletions, inversions, and translocations compared to normal tissue. mRNA expression of one of the genes in which a deletion was found, DPP10, encoding dipeptidyl-peptidase 10, was found to be related to longer survival in analysis of 53 MPM [120]. Shukuya et al recently analyzed 42 MPM of mixed histology for mutations in EGFR, KRAS, BRAF, PIK3CA, NRAS, MEK1, AKT1, PTEN, and HER2 and amplifications in EGFR, MET, PIK3CA, FGFR1, and FGFR2. One KRAS mutation and 3 PIK3CA amplifications were found. Analysis

12 Table 5

B. Davidson DNA repair and chemotherapy response

Ref.

Molecule

Expression level

No.

Histology

Univariate

Multivariate

Prognosis

82 82 83 83 88 82 82 83 83 84 84 84 87 87 87 87 87 87 88 89 89 90 91

TS TS TS TS TS ERCC1 ERCC1 ERCC1 ERCC1 ERCC1 class III β-tubulin BRCA ERCC1 MLH1 d MSH2 MSH6 class III β-tubulin ERCC1 e FPGS TS FPGS GST-α, −μ, −π GST-α, −μ, −π

mRNA Protein mRNA Protein Protein mRNA Protein mRNA Protein Protein Protein Protein Protein Protein Protein Protein Protein Gene Protein Protein Protein Protein Protein

60 60 99 99 84 60 60 99 99 50 50 50 79 c 79 79 79 79 79 84 85 85 88 56

All All All All All All All All All All All All All All All All All All All All All All All

NS P = .003 a P b .001 b P = .041 b P = .007 NS NS NS NS NS P = .01 NS NP NP NP NP NP NP NS NP NP NP P not provided

Poor Poor Poor Poor Poor Poor Good

92

Mean nuclear diameter DNA index

56

All

NS P = .019 a P b .001 b P = .022 b P = .004 NS NS NS NS NS NS NS P = .032 NS NS NS NS NS NS f NS NS NS α: NS, π: P = .012, μ: P = .024 P = .0006 P = .007

P not provided

Poor

Abbreviations: TS, thymidylate synthase; NS, not significant; ERCC1, excision repair cross-complementing group 1; BRCA, breast cancer gene; NP, not performed; MLH1, MutL homolog 1; MSH2, MutS homolog 2; MSH6, MutS homolog 6; FGPS, folylpoly-γ-glutamate synthetase; GST, glutathione S-transferase. a P = .02 for time to progression in univariate and multivariate analysis. b P = .017 and P b .001 for PFS in univariate analysis for protein and mRNA, respectively; P = .025 and P b .001 for PFS in multivariate analysis for protein and mRNA, respectively. c Pre-treatment specimens. d Lower MLH1 levels related to longer PFS (P = .0205). e Codon 118 polymorphisms related to shorter PFS (P = .0002). f Not significant for OS, significant for PFS in univariate (P = .023) and multivariate (P = .034) analysis.

of 21 MPM from this series using amplicon-based massively parallel sequencing targeting 48 genes identified mutations in 1–3 genes in 9 tumors (8 epithelioid, 1 biphasic) which involved the TP53, APC, PIK3CA, VHL, KRAS, FGFR3, ALK and PTPN11 genes [121]. Whole-exome sequencing of 22 MPM of mixed histology and matched blood samples identified 517 somatic mutations across 490 mutated genes. Frequent genetic alterations were found in the BAP1, NF2, CDKN2A, and CUL1 genes [122]. Whole-exome analysis of 8 cultures from MPM of different histology with focus on histone acetylation levels identified mutations in multiple transcriptional regulators, including the SWI/SNF chromatin regulating family [123].

10. Conclusions In their review from 2005, Kumar and Kratzke provided an overview of the molecular characteristics of MM, and predicted that targeted therapy, with emphasis on EGFR, VEGF and cell cycle proteins, was likely to change patient care in this disease [124]. Ten years later, this prediction appears to be rather remote from reality. These disappointing results are not unique to MM, and reflect numerous factors, including the ability of cancer cells to use alternative pathways once a given one is blocked, difficulties in drug delivery, side effects, and other influencing factors. As shown by the literature cited in this review, even the more modest aim of predicting treatment response and identifying reliable prognostic factors meets with difficulties in MPM. Many of the molecules discussed were

Malignant pleural mesothelioma prognosis

13

Table 6

Miscellaneous

Ref.

Molecule

Expression level

No.

Histology

Univariate

Multivariate

Prognosis

19 27 29 96 74 74 97 97 98 99 100 100 100 100 101 102 103 104 105 106 106 106 107 108 108 108 108 109 109 110 111 112

Periostin OPN OPN OPN D2-40 Calretinin CK19 TPA COX-2 ERβ ERβ SRC-1 SRC-2/TIF-2 SRC-3/AIB-1 GATA-6 Hyaluronan Hyaluronan Fibulin-3 HtrA1 CHOP1 GADD34 BiP1 MMP14 p15INK4B p16INK4A RASSF1A NORE1A WT1 Calretinin HMGB1 EMX2 c BAP1

Protein Protein a Protein a Protein Protein Protein Serum protein Serum protein Protein Protein Protein Protein Protein Protein Protein Pleural effusion protein a Pleural effusion protein a Pleural effusion protein a Protein Protein Protein Protein Protein Gene methylation Gene methylation Gene methylation Gene methylation Protein Protein Serum protein mRNA Protein

128 96 48 101 85 85 51 51 86 78 89 89 89 89 63 100 96 82 70 135 135 135 70 70 70 70 70 52 52 61 50 123

All All All All All All All All All All All All All All All All All All All All All All All All All All All All All All All All

P b .001 P = .002 P b .05 b P b .0001 P = .02 P b .001 P = .002 P = .003 P b .001 P = .002 NS NS P = .036 NS NS P = .03 P = .004 P = .012 P b .0001 P = .005 NS NS P b .0001 NS NS NS NS P = .013 NS P = .03 NS P = .0014

P = .002 P = .034 P b .05 b P b .0001 NS P = .03 P = .018 P = .004 P = .002 P = .006 NS NS P = .024 NS NP P = .04 ND NP P b .001 P = .014 NS NS P = .002 NS NS NS NS P = .01 NS P = .03 NS NP

Poor Poor Poor Poor Good Good Poor Poor Good Good Good Good Good Poor Good Poor Poor Good Poor Poor

Abbreviations: OPN, osteopontin; NS, not significant; CK19, cytokeratin 19; TPA, tissue polypeptide antigen; COX-2, cyclooxygenase 2; ER, estrogen receptor; NP, not performed; ND, not detailed; HtrA, high temperature requirement A; MMP14, matrix metalloproteinase 14; WT1, Wilms’ tumor antigen 1; HMGB, high-motility group B; BAP1, BRCA-associated protein 1. a Serum levels by ELISA. b Additionally observed for PFS. c Lower EMX2 levels related to shorter PFS (P = .002), retained in multivariate analysis (P = .004).

studied in single reports, some of which may have been of insufficient power. For other biomarkers, which have been studied by several groups, results are often conflicting. Differences in case distribution based on gender, histology, stage and treatment, as well as different methodology, have likely contributed to these discrepancies. The solution to this problem lies in performing multiinstitutional studies of large cohorts, in which sub-groups of patients with similar histology and treatment may be analyzed separately. Pathologists have a central role in this context, as their expertise is critical in correctly diagnosing MM and making sure that tissue submitted to molecular analyses is representative and contains a sufficient number of tumor cells, as well in evaluating the results of IHC or molecular tests in studies assessing targeted therapies.

MPM presents a significant health problem and the need for improved diagnosis, prognostication and treatment in this cancer is evident. The body of work reviewed in this paper documents the growing effort to understand the biology of this disease and its genetic make-up, and to define the biological characteristics that affect the survival of MPM patients. It is hoped that expanded knowledge will in the future allow us to achieve more success in treating this highly lethal tumor.

References [1] Ordóñez NG. Application of immunohistochemistry in the diagnosis of epithelioid mesothelioma: a review and update. HUM PATHOL 2013; 44:1-19.

14 [2] Henderson DW, Reid G, Kao SC, van Zandwijk N, Klebe S. Challenges and controversies in the diagnosis of mesothelioma: Part 1. Cytology-only diagnosis, biopsies, immunohistochemistry, discrimination between mesothelioma and reactive mesothelial hyperplasia, and biomarkers. J Clin Pathol 2013;66:847-53. [3] Betta PG, Magnani C, Bensi T, Trincheri NF, Orecchia S. Immunohistochemistry and molecular diagnostics of pleural malignant mesothelioma. Arch Pathol Lab Med 2012;136:253-61. [4] Campbell NP, Kindler HL. Update on malignant pleural mesothelioma. Semin Respir Crit Care Med 2011;32:102-10. [5] Pass HI, Giroux D, Kennedy C, et al. IASLC Staging Committee and Participating Institutions. Supplementary prognostic variables for pleural mesothelioma: a report from the IASLC staging committee. J Thorac Oncol 2014;9:856-64. [6] Kadota K, Suzuki K, Colovos C, et al. A nuclear grading system is a strong predictor of survival in epitheloid diffuse malignant pleural mesothelioma. Mod Pathol 2012;25:260-71. [7] Tanrikulu AC, Abakay A, Kaplan MA, et al. A clinical, radiographic and laboratory evaluation of prognostic factors in 363 patients with malignant pleural mesothelioma. Respiration 2010;80:480-7. [8] Nojiri S, Gemba K, Aoe K, et al. Survival and prognostic factors in malignant pleural mesothelioma: a retrospective study of 314 patients in the west part of Japan. Jpn J Clin Oncol 2011;41:32-9. [9] Ghanim B, Hoda MA, Winter MP, et al. Pretreatment serum Creactive protein levels predict benefit from multimodality treatment including radical surgery in malignant pleural mesothelioma: a retrospective multicenter analysis. Ann Surg 2012;256:357-62. [10] Yao ZH, Tian GY, Yang SX, et al. Serum albumin as a significant prognostic factor in patients with malignant pleural mesothelioma. Tumour Biol 2014;35:6839-45. [11] Ghanim B, Hoda MA, Klikovits T, et al. Circulating fibrinogen is a prognostic and predictive biomarker in malignant pleural mesothelioma. Br J Cancer 2014;110:984-90. [12] Blayney JK, Ceresoli GL, Castagneto B, et al. Response to chemotherapy is predictive in relation to longer overall survival in an individual patient combined-analysis with pleural mesothelioma. Eur J Cancer 2012;48:2983-92. [13] Davidson B. New diagnostic and molecular characteristics of malignant mesothelioma. Ultrastruct Pathol 2008;32:227-40. [14] Mossman BT, Shukla A, Heintz NH, Verschraegen CF, Thomas A, Hassan R. New insights into understanding the mechanisms, pathogenesis, and management of malignant mesotheliomas. Am J Pathol 2013;182:1065-77. [15] Jean D, Daubriac J, Le Pimpec-Barthes F, Galateau-Salle F, Jaurand MC. Molecular changes in mesothelioma with an impact on prognosis and treatment. Arch Pathol Lab Med 2012;136:277-93. [16] Edwards JG, Swinson DE, Jones JL, Waller DA, O'Byrne KJ. EGFR expression: associations with outcome and clinicopathological variables in malignant pleural mesothelioma. Lung Cancer 2006; 54:399-407. [17] Gaafar R, Bahnassy A, Abdelsalam I, et al. Tissue and serum EGFR as prognostic factors in malignant pleural mesothelioma. Lung Cancer 2010;70:43-50. [18] Rena O, Boldorini LR, Gaudino E, Casadio C. Epidermal growth factor receptor overexpression in malignant pleural mesothelioma: prognostic correlations. J Surg Oncol 2011;104:701-5. [19] Schramm A, Opitz I, Thies S, et al. Prognostic significance of epithelial-mesenchymal transition in malignant pleural mesothelioma. Eur J Cardiothorac Surg 2010;37:566-72. [20] Levallet G, Vaisse-Lesteven M, Le Stang N, et al. Plasma cell membrane localization of c-MET predicts longer survival in patients with malignant mesothelioma: a series of 157 cases from the MESOPATH Group. J Thorac Oncol 2012;7:599-606. [21] Pinato DJ, Mauri FA, Lloyd T, et al. The expression of Axl receptor tyrosine kinase influences the tumour phenotype and clinical outcome of patients with malignant pleural mesothelioma. Br J Cancer 2013; 108:621-8.

B. Davidson [22] Tsao AS, Harun N, Fujimoto J, et al. Elevated PDGFRB gene copy number gain is prognostic for improved survival outcomes in resected malignant pleural mesothelioma. Ann Diagn Pathol 2014;18:140-5. [23] Kumar-Singh S, Jacobs W, Dhaene K, et al. Syndecan-1 expression in malignant mesothelioma: correlation with cell differentiation, WT1 expression, and clinical outcome. J Pathol 1998;186:300-5. [24] Robinson BW, Creaney J, Lake R, et al. Mesothelin-family proteins and diagnosis of mesothelioma. Lancet 2003;362:1612-6. [25] Hassan R, Remaley AT, Sampson ML, et al. Detection and quantification of serum mesothelin, a tumor marker for patients with mesothelioma and ovarian cancer. Clin Cancer Res 2006;12: 447-53. [26] Cristaudo A, Foddis R, Vivaldi A, et al. Clinical significance of serum mesothelin in patients with mesothelioma and lung cancer. Clin Cancer Res 2007;13:5076-81. [27] Grigoriu BD, Scherpereel A, Devos P, et al. Utility of osteopontin and serum mesothelin in malignant pleural mesothelioma diagnosis and prognosis assessment. Clin Cancer Res 2007;13:2928-35. [28] Schneider J, Hoffmann H, Dienemann H, Herth FJ, Meister M, Muley T. Diagnostic and prognostic value of soluble mesothelinrelated proteins in patients with malignant pleural mesothelioma in comparison with benign asbestosis and lung cancer. J Thorac Oncol 2008;3:1317-24. [29] Hollevoet K, Nackaerts K, Gosselin R, et al. Soluble mesothelin, megakaryocyte potentiating factor, and osteopontin as markers of patient response and outcome in mesothelioma. J Thorac Oncol 2011; 6:1930-7. [30] Linch M, Gennatas S, Kazikin S, et al. A serum mesothelin level is a prognostic indicator for patients with malignant mesothelioma in routine clinical practice. BMC Cancer 2014;14:674. [31] Alifano M, Loi M, Camilleri-Broet S, Dupouy S, Régnard JF, Forgez P. Neurotensin expression and outcome of malignant pleural mesothelioma. Biochimie 2010;92:164-70. [32] Richardson MM, Jennings LK, Zhang XA. Tetraspanins and tumor progression. Clin Exp Metastasis 2011;28:261-70. [33] Amatya VJ, Takeshima Y, Aoe K, et al. CD9 expression as a favorable prognostic marker for patients with malignant mesothelioma. Oncol Rep 2013;29:21-8. [34] Amatya VJ, Takeshima Y, Kushitani K, Yamada T, Morimoto C, Inai K. Overexpression of CD26/DPPIV in mesothelioma tissue and mesothelioma cell lines. Oncol Rep 2011;26:1369-75. [35] Aoe K, Amatya VJ, Fujimoto N, et al. CD26 overexpression is associated with prolonged survival and enhanced chemosensitivity in malignant pleural mesothelioma. Clin Cancer Res 2012;18:1447-56. [36] Kao SC, Armstrong N, Condon B, et al. Aquaporin 1 is an independent prognostic factor in pleural malignant mesothelioma. Cancer 2012;118:2952-61. [37] Otterstrom C, Soltermann A, Opitz I, et al. CD74: a new prognostic factor for patients with malignant pleural mesothelioma. Br J Cancer 2014;110:2040-6. [38] Righi L, Cavallo MC, Gatti G, et al. Tumor/stromal caveolin-1 expression patterns in pleural mesothelioma define a subgroup of the epithelial histotype with poorer prognosis. Am J Clin Pathol 2014; 141:816-27. [39] Thiery JP, Acloque H, Huang RY, Nieto MA. Epithelial-mesenchymal transitions in development and disease. Cell 2009;139:871-90. [40] Fassina A, Cappellesso R, Guzzardo V, et al. Epithelial-mesenchymal transition in malignant mesothelioma. Mod Pathol 2012;25:86-99. [41] Merikallio H, Pääkkö P, Salmenkivi K, Kinnula V, Harju T, Soini Y. Expression of snail, twist, and Zeb1 in malignant mesothelioma. APMIS 2013;121:1-10. [42] Kobayashi M, Huang CL, Sonobe M, et al. Snail expression is associated with a poor prognosis in malignant pleural mesotheliomas. Ann Thorac Surg 2013;95:1181-8. [43] Watzka SB, Setinek U, Stubenberger EB, et al. Integrin-linked kinase, phosphorylated AKT and the prognosis of malignant pleural mesothelioma. Eur J Cardiothorac Surg 2011;39:180-4.

Malignant pleural mesothelioma prognosis [44] Edwards JG, Cox G, Andi A, et al. Angiogenesis is an independent prognostic factor in malignant mesothelioma. Br J Cancer 2001;85: 863-8. [45] Kumar-Singh S, Weyler J, Martin MJ, Vermeulen PB, Van Marck E. Angiogenic cytokines in mesothelioma: A study of VEGF, FGF-1 and −2, and TGFβ expression. J Pathol 1999;189:72-8. [46] Strizzi L, Catalano A, Vianale G, et al. Vascular endothelial growth factor is an autocrine growth factor in human malignant mesothelioma. J Pathol 2001;193:468-75. [47] Kao SC, Harvie R, Paturi F, et al. The predictive role of serum VEGF in an advanced malignant mesothelioma patient cohort treated with thalidomide alone or combined with cisplatin/gemcitabine. Lung Cancer 2012;75:248-54. [48] Tabata C, Hirayama N, Tabata R, et al. A novel clinical role for angiopoietin-1 in malignant pleural mesothelioma. Eur Respir J 2010; 36:1099-105. [49] Yoneda K, Tanaka F, Kondo N, et al. Circulating endothelial cell (CEC) as a diagnostic and prognostic marker in malignant pleural mesothelioma (MPM). Ann Surg Oncol 2012;19:4229-37. [50] Filiberti R, Marroni P, Neri M, et al. Serum PDGF-AB in pleural mesothelioma. Tumour Biol 2005;26:221-6. [51] Ohta Y, Shridhar V, Kalemkerian GP, Bright RK, Watanabe Y, Pass HI. Thrombospondin-1 expression and clinical implications in malignant pleural mesothelioma. Cancer 1999;85:2570-6. [52] Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell 2011;144:646-74. [53] O’Kane SL, Pound RJ, Campbell A, Chaudhuri N, Lind MJ, Cawkwell L. Expression of bcl-2 family members in malignant pleural mesothelioma. Acta Oncol 2006;45:449-53. [54] Mineo TC, Ambrogi V, Cufari ME, Pompeo E. May cyclooxygenase2 (COX-2), p21 and p27 expression affect prognosis and therapeutic strategy of patients with malignant pleural mesothelioma? Eur J Cardiothorac Surg 2010;38:245-52. [55] Bahnassy AA, Zekri AR, Abou-Bakr AA, et al. Aberrant expression of cell cycle regulatory genes predicts overall and disease free survival in malignant pleural mesothelioma patients. Exp Mol Pathol 2012;93:154-61. [56] Bongiovanni M, Cassoni P, De Giuli P, et al. p27(kip1) immunoreactivity correlates with long-term survival in pleural malignant mesothelioma. Cancer 2001;92:1245-50. [57] Mairinger FD, Walter RF, Ting S, et al. Mdm2 protein expression is strongly associated with survival in malignant pleural mesothelioma. Future Oncol 2014;10:995-1005. [58] Chung CT, Santos Gda C, Hwang DM, et al. FISH assay development for the detection of p16/CDKN2A deletion in malignant pleural mesothelioma. J Clin Pathol 2010;63:630-4. [59] Nachmias B, Ashhab Y, Ben-Yehuda D. The inhibitor of apoptosis protein family (IAPs): an emerging therapeutic target in cancer. Semin Cancer Biol 2004;14:231-43. [60] Gordon GJ, Mani M, Mukhopadhyay L, et al. Expression patterns of inhibitor of apoptosis proteins in malignant pleural mesothelioma. J Pathol 2007;211:447-54. [61] Kleinberg L, Lie AK, Flørenes VA, Nesland JM, Davidson B. Expression of inhibitors of apoptosis (IAP) family members in malignant mesothelioma. HUM PATHOL 2007;38:986-94. [62] Hmeljak J, Erčulj N, Dolžan V, et al. Is survivin expression prognostic or predictive in malignant pleural mesothelioma? Virchows Arch 2013;462:315-21. [63] Rodon J, Dienstmann R, Serra V, Tabernero J. Development of PI3K inhibitors: lessons learned from early clinical trials. Nat Rev Clin Oncol 2013;10:143-53. [64] Song MS, Salmena L, Pandolfi PP. The functions and regulation of the PTEN tumour suppressor. Nat Rev Mol Cell Biol 2012;13: 283-96. [65] Cacciotti P, Barbone D, Porta C, et al. SV40-dependent AKT activity drives mesothelial cell transformation after asbestos exposure. Cancer Res 2005;65:5256-62.

15 [66] Opitz I, Soltermann A, Abaecherli M, et al. PTEN expression is a strong predictor of survival in mesothelioma patients. Eur J Cardiothorac Surg 2008;33:502-6. [67] Agarwal V, Campbell A, Beaumont KL, Cawkwell L, Lind MJ. PTEN protein expression in malignant pleural mesothelioma. Tumour Biol 2012;34:847-51. [68] Bitanihirwe BK, Meerang M, Friess M, et al. PI3K/mTOR signaling in mesothelioma patients treated with induction chemotherapy followed by extrapleural pneumonectomy. J Thorac Oncol 2014;9: 239-47. [69] Mutti L, Valle MT, Balbi B, et al. Primary human mesothelioma cells express class II MHC, ICAM-1 and B7-2 and can present recall antigens to autologous blood lymphocytes. Int J Cancer 1998;78: 740-9. [70] Porta C, Danova M, Orengo AM, et al. Interleukin-2 induces cell cycle perturbations leading to cell growth inhibition and death in malignant mesothelioma cells in vitro. J Cell Physiol 2000;185: 126-34. [71] Calabrò L, Sigalotti L, Fonsatti E, et al. Expression and regulation of B7-H3 immunoregulatory receptor, in human mesothelial and mesothelioma cells: immunotherapeutic implications. J Cell Physiol 2011;226:2595-600. [72] Mansfield AS, Roden AC, Peikert T, et al. B7-H1 expression in malignant pleural mesothelioma is associated with sarcomatoid histology and poor prognosis. J Thorac Oncol 2014;9:1036-40. [73] Kao SC, Pavlakis N, Harvie R, et al. High blood neutrophil-tolymphocyte ratio is an indicator of poor prognosis in malignant mesothelioma patients undergoing systemic therapy. Clin Cancer Res 2010;16:5805-13. [74] Kao SC, Klebe S, Henderson DW, et al. Low calretinin expression and high neutrophil-to-lymphocyte ratio are poor prognostic factors in patients with malignant mesothelioma undergoing extrapleural pneumonectomy. J Thorac Oncol 2011;6:1923-9. [75] Pinato DJ, Mauri FA, Ramakrishnan R, Wahab L, Lloyd T, Sharma R. Inflammation-based prognostic indices in malignant pleural mesothelioma. J Thorac Oncol 2012;7:587-94. [76] Abakay O, Tanrikulu AC, Palanci Y, Abakay A. The value of inflammatory parameters in the prognosis of malignant mesothelioma. J Int Med Res 2014;42:554-65. [77] Meniawy TM, Creaney J, Lake RA, Nowak AK. Existing models, but not neutrophil-to-lymphocyte ratio, are prognostic in malignant mesothelioma. Br J Cancer 2013;109:1813-20. [78] Burt BM, Rodig SJ, Tilleman TR, Elbardissi AW, Bueno R, Sugarbaker DJ. Circulating and tumor-infiltrating myeloid cells predict survival in human pleural mesothelioma. Cancer 2011;117: 5234-44. [79] Suzuki K, Kadota K, Sima CS, et al. Chronic inflammation in tumor stroma is an independent predictor of prolonged survival in epithelioid malignant pleural mesothelioma patients. Cancer Immunol Immunother 2011;60:1721-8. [80] Alì G, Boldrini L, Lucchi M, Mussi A, Corsi V, Fontanini G. Tryptase mast cells in malignant pleural mesothelioma as an independent favorable prognostic factor. J Thorac Oncol 2009;4: 348-54. [81] Curtin NJ. DNA repair dysregulation from cancer driver to therapeutic target. Nat Rev Cancer 2012;12:801-17. [82] Righi L, Papotti MG, Ceppi P, et al. Thymidylate synthase but not excision repair cross-complementation group 1 tumor expression predicts outcome in patients with malignant pleural mesothelioma treated with pemetrexed-based chemotherapy. J Clin Oncol 2010;28: 1534-9. [83] Zucali PA, Giovannetti E, Destro A, et al. Thymidylate synthase and excision repair cross-complementing group-1 as predictors of responsiveness in mesothelioma patients treated with pemetrexed/ carboplatin. Clin Cancer Res 2011;17:2581-90. [84] Zimling ZG, Sørensen JB, Gerds TA, Bech C, Andersen CB, SantoniRugiu E. A biomarker profile for predicting efficacy of cisplatin-

16

[85]

[86]

[87]

[88]

[89]

[90]

[91]

[92]

[93] [94] [95]

[96]

[97]

[98]

[99]

[100]

[101]

[102]

[103]

[104]

B. Davidson vinorelbine therapy in malignant pleural mesothelioma. Cancer Chemother Pharmacol 2012;70:743-54. Erčulj N, Kovač V, Hmeljak J, Dolžan V. The influence of platinum pathway polymorphisms on the outcome in patients with malignant mesothelioma. Ann Oncol 2012;23:961-7. Erčulj N, Kovač V, Hmeljak J, Franko A, Dodič-Fikfak M, Dolžan V. DNA repair polymorphisms and treatment outcomes of patients with malignant mesothelioma treated with gemcitabine-platinum combination chemotherapy. J Thorac Oncol 2012;7:1609-17. Ting S, Mairinger FD, Hager T, et al. ERCC1, MLH1, MSH2, MSH6, and βIII-tubulin: resistance proteins associated with response and outcome to platinum-based chemotherapy in malignant pleural mesothelioma. Clin Lung Cancer 2013;14:558-67. Christoph DC, Asuncion BR, Mascaux C, et al. Folylpoly-glutamate synthetase expression is associated with tumor response and outcome from pemetrexed-based chemotherapy in malignant pleural mesothelioma. J Thorac Oncol 2012;7:1440-8. Lustgarten DE, Deshpande C, Aggarwal C, et al. Thymidylate synthase and folyl-polyglutamate synthase are not clinically useful markers of response to pemetrexed in patients with malignant pleural mesothelioma. J Thorac Oncol 2013;8:469-77. Dejmek A, Brockstedt U, Hjerpe A. Immunoreactivity of pleural malignant mesotheliomas to glutathione S-transferases. APMIS 1998;106:489-94. Segers K, Kumar-Singh S, Weyler J, et al. Glutathione S-transferase expression in malignant mesothelioma and non-neoplastic mesothelium: an immunohistochemical study. J Cancer Res Clin Oncol 1996; 122:619-24. Segers K, Singh SK, Van Daele A, et al. Evaluation and prognostic value of DNA content and of morphometric parameters in malignant mesothelioma using digital image analysis. Lung Cancer 1996;14:229-37. Wai PY, Kuo PC. Osteopontin: regulation in tumor metastasis. Cancer Metastasis Rev 2008;27:103-18. Wang KX, Denhardt DT. Osteopontin: role in immune regulation and stress responses. Cytokine Growth Factor Rev 2008;19:333-45. Pass HI, Lott D, Lonardo F, et al. Asbestos exposure, pleural mesothelioma, and serum osteopontin levels. N Engl J Med 2005; 353:1564-73. Cappia S, Righi L, Mirabelli D, et al. Prognostic role of osteopontin expression in malignant pleural mesothelioma. Am J Clin Pathol 2008;130:58-64. Schouwink H, Korse CM, Bonfrer JM, Hart AA, Baas P. Prognostic value of the serum tumour markers Cyfra 21–1 and tissue polypeptide antigen in malignant mesothelioma. Lung Cancer 1999;25:25-32. O'Kane SL, Cawkwell L, Campbell A, Lind MJ. Cyclooxygenase-2 expression predicts survival in malignant pleural mesothelioma. Eur J Cancer 2005;41:1645-8. Pinton G, Brunelli E, Murer B, et al. Estrogen receptor-beta affects the prognosis of human malignant mesothelioma. Cancer Res 2009; 69:4598-604. Jennings CJ, O'Grady A, Cummins R, et al. Sustained expression of steroid receptor coactivator SRC-2/TIF-2 is associated with better prognosis in malignant pleural mesothelioma. J Thorac Oncol 2012;7:243-8. Lindholm PM, Soini Y, Myllärniemi M, et al. Expression of GATA-6 transcription factor in pleural malignant mesothelioma and metastatic pulmonary adenocarcinoma. J Clin Pathol 2009;62:339-44. Thylén A, Hjerpe A, Martensson G. Hyaluronan content in pleural fluid as a prognostic factor in patients with malignant pleural mesothelioma. Cancer 2001;92:1224-30. Creaney J, Dick IM, Segal A, Musk AW, Robinson BW. Pleural effusion hyaluronic acid as a prognostic marker in pleural malignant mesothelioma. Lung Cancer 2013;82:491-8. Creaney J, Dick IM, Meniawy TM, et al. Comparison of fibulin-3 and mesothelin as markers in malignant mesothelioma. Thorax 2014;69: 895-902.

[105] Baldi A, Mottolese M, Vincenzi B, et al. The serine protease HtrA1 is a novel prognostic factor for human mesothelioma. Pharmacogenomics 2008;9:1069-77. [106] Dalton LE, Clarke HJ, Knight J, et al. The endoplasmic reticulum stress marker CHOP predicts survival in malignant mesothelioma. Br J Cancer 2013;108:1340-7. [107] Crispi S, Calogero RA, Santini M, et al. Global gene expression profiling of human pleural mesotheliomas: identification of matrix metalloproteinase 14 (MMP-14) as potential tumour target. PLoS One 2009;4:e7016. [108] Destro A, Ceresoli GL, Baryshnikova E, et al. Gene methylation in pleural mesothelioma: correlations with clinicopathological features and patient's follow-up. Lung Cancer 2008;59:369-76. [109] Cedrés S, Montero MA, Zamora E, et al. Expression of Wilms' tumor gene (WT1) is associated with survival in malignant pleural mesothelioma. Clin Transl Oncol 2014;16:776-82. [110] Tabata C, Shibata E, Tabata R, et al. Serum HMGB1 as a prognostic marker for malignant pleural mesothelioma. BMC Cancer 2013;13:205. [111] Giroux Leprieur E, Hirata T, Mo M, et al. The homeobox gene EMX2 is a prognostic and predictive marker in malignant pleural mesothelioma. Lung Cancer 2014;85:465-71. [112] Arzt L, Quehenberger F, Halbwedl I, Mairinger T, Popper HH. BAP1 protein is a progression factor in malignant pleural mesothelioma. Pathol Oncol Res 2014;20:145-51. [113] Gordon GJ, Jensen RV, Hsiao LL, et al. Using gene expression ratios to predict outcome among patients with mesothelioma. J Natl Cancer Inst 2003;95:598-605. [114] Gordon GJ, Rockwell GN, Godfrey PA, et al. Validation of genomics-based prognostic tests in malignant pleural mesothelioma. Clin Cancer Res 2005;11:4406-14. [115] Gordon GJ, Dong L, Yeap BY, et al. Four-gene expression ratio test for survival in patients undergoing surgery for mesothelioma. J Natl Cancer Inst 2009;101:678-86. [116] López-Ríos F, Chuai S, Flores R, et al. Global gene expression profiling of pleural mesotheliomas: overexpression of aurora kinases and P16/CDKN2A deletion as prognostic factors and critical evaluation of microarray-based prognostic prediction. Cancer Res 2006;66:2970-9. [117] Pass HI, Liu Z, Wali A, et al. Gene expression profiles predict survival and progression of pleural mesothelioma. Clin Cancer Res 2004;10:849-59. [118] de Reyniès A, Jaurand MC, Renier A, et al. Molecular classification of malignant pleural mesothelioma: identification of a poor prognosis subgroup linked to the epithelial-to-mesenchymal transition. Clin Cancer Res 2014;20:1323-34. [119] Pass HI, Goparaju C, Ivanov S, et al. hsa-miR-29c* is linked to the prognosis of malignant pleural mesothelioma. Cancer Res 2010;70: 1916-24. [120] Bueno R, De Rienzo A, Dong L, et al. Second generation sequencing of the mesothelioma tumor genome. PLoS One 2010; 5:e10612. [121] Shukuya T, Serizawa M, Watanabe M, et al. Identification of actionable mutations in malignant pleural mesothelioma. Lung Cancer 2014;86:35-40. [122] Guo G, Chmielecki J, Goparaju C, et al. Whole-exome sequencing reveals frequent genetic alterations in BAP1, NF2, CDKN2A, and CUL1 in malignant pleural mesothelioma. Cancer Res 2015;75: 264-9. [123] Yoshikawa Y, Sato A, Tsujimura T, et al. Biallelic germline and somatic mutations in malignant mesothelioma: multiple mutations in transcription regulators including mSWI/SNF genes. Int J Cancer 2015;136:560-71. [124] Kumar P, Kratzke RA. Molecular prognostic markers in malignant mesothelioma. Lung Cancer 2005;49(Suppl. 1):S53-60.