Molecular tumor markers for asbestos-related mesothelioma: Serum diagnostic markers

Authors


Okio Hino, MD, PhD, Department of Pathology, Juntendo University School of Medicine, 2-1-1 Hongo, Bunkyo-ku, Tokyo 113-8421, Japan. Email: ohino@med.juntendo.ac.jp

Abstract

Mesothelioma is an aggressive tumor arising from the mesothelium, and is usually associated with previous exposure to asbestos. The incubation period of the tumor may be described as 30–40 years, and the prognosis is dismal. In addition to immunohistochemical markers, recently, serum markers for the diagnosis of mesothelioma have been reported as candidates. In contrast, the expression in renal carcinoma (ERC) gene has been discovered in the Eker rat model (Tsc2 gene mutant), which is a homolog of the human mesothelin/megakaryocyte potentiating factor gene, and a novel ELISA system (N-ERC/mesothelin) has been developed. It has also been found that N-ERC/mesothelin is very stable and plentiful in the blood. In the present paper the potential utility of molecular diagnostic markers is reviewed, including ELISA systems for asbestos-related mesothelioma.

Mesothelioma is an aggressive tumor arising from the mesothelium, a membrane lining several body cavities, including the pleura, peritoneum and pericardium, and is usually associated with asbestos exposure. There are mainly three types of mesothelioma according to site of origin: pleural mesothelioma; peritoneal mesothelioma; and pericardial mesothelioma. Pleural mesothelioma is the most common form of mesothelioma. In the classification based on histological categories, mainly three types of mesothelioma are reported: epithelial; sarcomatous; and mixed. Although epithelial mesothelioma has a better prognosis, the histological distinction between epithelial mesothelioma and adenocarcinomas/sarcomatous mesothelioma and sarcomas is often difficult and requires additional immunohistochemial studies.

Asbestos is the commercial name for a group of hydrated magnesium silicate fibrous minerals. Three types are mainly recognized: Chrysotile (white asbestos), Crocidolite (blue asbestos), and Amosite (brown asbestos). Asbestos is valued in the industry for its resistance to heat and combustion, and has been used in the production of cement, ceiling and pool tiles, automobile brake linings, and shipbuilding. Several types of asbestos-related illnesses other than mesothelioma are known: pleural plaques, pleural thickening, asbestosis and lung cancer.

In Japan the amount of asbestos imports increased between 1960 and 1974, and reached a peak in 1974.1 However, an increased risk of developing mesothelioma was later found among people who work with asbestos.

With the increase of asbestos imports, the number of deaths from pleural mesothelioma has grown. Since the early 1990s an exponential increase has been observed: the estimated mean annual number of deaths from malignant pleural mesothelioma (MPM) was 506 between 1995 and 1999, and approximately 1000 in 2005.

Given that the latency of mesothelioma is described as 30–40 years, its incidence is expected to increase in the coming decades.

IMMUNOHISTOCHEMICAL MARKERS FOR MESOTHELIOMA

The most important modalities for the diagnosis of mesothelioma are chest radiography, chest CT, and magnetic resonance imaging (MRI) of the chest. However, it may be difficult to make an early diagnosis of mesothelioma using the currently available diagnostic imaging techniques, and identification of tumor markers is urgently needed. Immunohistochemical diagnosis of epithelioid mesotheliomain pleural biopsy or surgically resected specimens has been actively pursued, using markers such as podoplanin, calretinin, WT-1, cytokeratin 5, thrombomodulin, and mesothelin.2,3 Some of these markers have indeed been helpful for confirming the diagnosis of mesothelioma and distinguishing between mesothelioma and adenocarcinoma, which is one of the challenging problems for surgical pathologists.

Ordonez reported a comparison of the various immunohistochemical markers currently available for the diagnosis of mesothelioma and squamous carcinoma of the lung.4 A total of 30 epithelioid pleural mesotheliomas and 30 non-keratinizing squamous carcinomas of the lung were investigated for the expression of the following markers: podoplanin, calretinin, mesothelin, WT1, keratin 5/6, keratin 7, p63, CEA, MOC-31, Ber-EP4, B72.3, BG-8 (Lewisy), leu-M1 (CD15), and thyroid transcription factor-1 (TTF-1). All 30 (100%) of the mesotheliomas reacted for calretinin, mesothelin and keratin 7, 93% each for podoplanin, WT1 and keratin 5/6, 13% for Ber-EP4, 7% each for p63, MOC-31 and BG-8, and 0% for B72.3, CEA, leu-M1 and TTF-1.

All 30 (100%) of the squamous carcinomas were positive for p63 and keratin 5/6, 97% for MOC-31, 87% for Ber-EP4, 80% for BG-8, 77% for CEA, 57% for keratin 7, 40% for calretinin and B72.3, 30% for leu-M1, 27% for mesothelin, 15% for podoplanin, and 0% for WT 1 and TTF-1. As a result, a combination of two positive mesothelioma markers (WT1 and calretinin or mesothelin) with two negative mesothelioma markers (p63 and MOC-31) would allow the differential diagnosis to be established between epithelioid mesotheliomas and squamous carcinomas of the lung in nearly all instances (cited in Ordonez4).

Also, Yaziji et al. evaluated the sensitivity and specificity of 12 antibodies for distinguishing epithelioid mesothelioma from adenocarcinoma using immunohistochemistry.5 The antibodies against thrombomodulin and HBME-1 were in addition to those used by Ordonez, aforedescribed. The 143 tumors evaluated included 65 malignant epithelioid mesotheliomas, 22 lung adenocarcinomas, 27 ovarian serous carcinomas, 24 breast carcinomas, and five gastric carcinomas.

Calretinin had the best sensitivity for mesothelioma (95%), followed by HBME-1 (84%), WT-1 (78%), cytokeratin 5 (76%), mesothelin (75%), and vimentin and thrombomodulin (68%). Thrombomodulin had the best specificity for mesothelioma (92%), followed by cytokeratin 5 (89%), calretinin (87%) vimentin (84%), and HBME-1 (45%). When ovarian carcinomas were excluded from the analysis, the specificity of mesothelin and WT-1 for the diagnosis of mesothelioma increased to 90% and 81%, respectively. The specificity of the non-mesothelial antigens for adenocarcinoma was 98% for BG8 and CEA, 97% for CD15, 95% for BerEp4, and 87% for MOC-31 (cited in Yaziji et al.5). Yaziji et al. showed that a three-antibody immunohistochemical panel including calretinin, BG8, and MOC-31, provided >96% sensitivity and specificity for distinguishing epithelioid mesothelioma from adenocarcinoma.5

In the case of sarcomatous mesothelioma, the histological distinction from synovial sarcoma may be difficult because of the combination of epithelioid and spindle cells, potentially shared locations, and antigenic expression. The differential diagnosis of liposarcomas, leiomyosarcomas, rhabdomyosarcoma, angiosarcoma, fibrosarcoma, gastrointestinal stromal tumor, desmoid tumor, Ewing's sarcoma, osteosarcoma and chondrosarcoma can also be difficult, even though a few markers (e.g. CAM 5.2 etc.). are recommended for differential diagnosis by immunostainning.

In immunohistochemical study for mesothelin, 5B2 antibody (Novocastra, Newcastle upon Tyne, UK) is now commonly used. However, the recognition epitope of 5B2 is not defined. We have established polyclonal antimesothelin antibody using the synthetic peptide as immunogen, and performed immunohistochemical study on tissues of mesothelioma patients. As a result, mesothelin is highly expressed in tumor cells (Fig. 1) and membranous reactivity is evident.6 These trials using antibodies with the well-defined epitopes could elucidate varied presentation of mesothelin molecule in mesothelioma and may be helpful in understanding mesothelioma.

Figure 1.

Immunohistochemical staining for mesothelin using antimesothelin antibody (rabbit IgG) in pleural mesothelioma specimens. Mesothelin protein was expressed highly in tumor cells and membranous reactivity was evident.

Thus, immunohistochemical diagnosis using various antibodies for markers is under intense study. However, there are some differences among them. The standardization of protocols may be required.

SERUM MARKERS FOR MESOTHELIOMA

While no reliable serum marker for the diagnosis of mesothelioma has as yet been established, Robinson et al. recently proposed soluble mesothelin-related protein (SMRP) as one of the candidates.7,8 Yamaguchi et al. first reported the purification of megakaryocyte potentiating factor (MPF) protein from the culture supernatant of the human pancreatic cancer cells HPC-Y5.9 Subsequently, Kojima et al. isolated human MPF cDNA from an HPC-Y5 cDNA library by polymerase chain reaction using specific primers derived from the amino acid sequences of purified MPF polypeptides and plaque hybridization methods, and reported the same sequence10 as that reported by Chang and Pastan,11 described here. Sequence analysis predicted that the 622-amino acid protein contains an N-terminal 33-residue signal sequence and four potential N-glycosylation sites.

In contrast, Chang et al. isolated an antigen cDNA of MAb K1 that had an 1884 bp open reading frame encoding a 69 kDa protein, and showed that a 69 kDa protein was the precursor of this 40 kDa glycoprotein, with the 40 kDa form released by treatment with phosphatidylinositol-specific phospholipase C.11,12

Scholler et al. reported that the antigen protein recognized by mAb OV569 had an identical N-terminal amino acid sequence to that of the membrane-bound portion of mesothelin and MPF;13 this 42–45 kDa protein was termed a soluble member(s) of the mesothelin/megakaryocyte potentiating factor family related protein (SMRP). SMRP has an 82 bp insertion in the membrane-associated part and codes for a C terminus with a hydrophilic, presumably soluble, tail.

Robinson et al. and Creaney and Robinson reported that the serum concentrations of SMRP measured using a double determinant (sandwich) ELISA could be a useful marker for the diagnosis of mesothelioma and for monitoring the disease progression in cases of mesothelioma.7,8,14

We have previously reported the identification of the expressed in renal carcinoma (ERC) gene, which is strongly expressed in renal cancers in the Eker rat, which develops hereditary renal carcinomas following the induction of two-hit mutations of tumor suppressor genes.15,16 A database search showed that this predicted amino acid from ERC cDNA had 87.4% and 56.1% identity, respectively, to the mouse and human mesothelin gene. Therefore, henceforth, we shall refer to this protein as ERC/mesothelin.

The human mesothelin gene codes for several proteins; the primary product is a 71 kDa precursor protein that can be physiologically cleaved by some furin-like proteases into a 40 kDa C-terminal fragment that remains membrane-bound, and a 31 kDa N-terminal fragment, which is secreted into the blood (Fig. 2). The C-terminal 40 kDa fragment is referred to as mesothelin. In contrast, the N-terminal 31 kDa fragment is a secreted protein identified as MPF.

Figure 2.

Schematic diagram of production and digestion of ERC/mesothelin, and ELISA system. The primary product of the ERC/mesothelin gene is a 71 kDa precursor protein. This protein is cleaved physiologically, releasing its 31 kDa N-terminal fragment, designated as N-ERC/mesothelin, into the blood. The 31 kDa N-terminal fragment is detected by a sandwich ELISA system containing two antibodies against the N-terminal fragment.

In avoiding confusion arising from the use of the term ‘mesothelin’, we shall use the term N-ERC/mesothelin for the N-terminal 31 kDa fragment, which is the same as MPF, and C-ERC/mesothelin for the C-terminal 40 kDa fragment.

ELISA SYSTEMS FOR THE DETECTION OF C-ERC/MESOTHELIN

Although its real physiological role is still unclear, ERC/mesothelin has been identified in the serum of not only normal, but also of cancer patients. So far, two groups have reported on ELISA systems for the detection of C-ERC/mesothelin.

Robinson et al. have established an ELISA system for the detection of SMRP. In their first paper they assayed the serum concentrations of SMRP in serum samples obtained from 44 patients with histologically proven mesothelioma, 68 matched healthy controls, 40 of whom had a history of exposure to asbestos, and 160 patients with other inflammatory or malignant lung or pleural diseases.7 They showed that 84% of the 44 mesothelioma patients had elevated serum concentrations of SMRP, as compared with 2% of the 160 with other lung/pleural diseases. In their second paper they reported that measurement of the concentrations of SMRP in the serum as a marker of mesothelioma had a sensitivity of 83% and specificity of 95% for the diagnosis of mesothelioma in the first 48 mesothelioma patients examined.8

Scherpereel et al. reported their data obtained using this ELISA kit (Mesomark; Fujirebio Diagnostic, Malvern, PA, USA).17 The mean serum SMRP level was higher in patients with mesothelioma (2.05 ± 2.57 nmol/L, n = 28) than inthose with pleural metastasis (1.02 ± 1.79 nmol/L, n = 35) or benign lesions of the pleura (0.55 ± 0.59 nmol/L, n = 28). The pleural fluid SMRP concentrations were higher than the serum SMRP concentrations in all the patient groups (mesothelioma, 46.1 ± 83.2 nmol/L; benign lesions, 6.4 ± 11.1 nmol/L; metastasis, 6.36 ± 21.73 nmol/L).

Hassan et al. have pursued the development of mAb against ERC/mesothelin. Some of these antibodies have been applied for the treatment of mesothelioma by the immunotoxin technology.18 Recently, they have produced new mAb against C-ERC/mesothelin19 and an ELISA system for the detection of C-ERC/mesothelin using these mAb, and they reported elevated serum levels of mesothelin in 40 of 56 (71%) mesothelioma patients and 14 of 21 (67%) ovarian cancer patients.20 The serum levels of mesothelin were increased in 80% of patients who had positive tumor immunohistochemistry for mesothelin. They also observed a rapid decrease in the serum mesothelin levels after surgery in patients with peritoneal mesothelioma.

Thus, several ELISA systems have been evaluated for measurement of the serum levels of C-ERC/mesothelin as a possible diagnostic marker for mesothelioma. However, some questions on the measurement of C-ERC/mesothelin remain.

First, the mechanisms related to increase of C-ERC/mesothelin release from the mesothelioma cell surface have not yet been elucidated. While N-ERC/mesothelin represents the secretory form, the remaining C-ERC/mesothelin part of the molecule is still membrane-bound. The release of mesothelin from the cell surface is probably mediated by phospholipase C. There might be mechanisms by which this portion of the molecule is also released to induce tumor progression.

Second, the ELISA system for SMRP could detect higher levels of serum SMRP in mesothelioma patients. However, the existence of SMRP (82 bp inserted splicing isoform and other spliced forms) has not yet been demonstrated in the serum of mesothelioma patients. Recently, Hellstrom et al. reported the establishment of specific mAb and ELISA systems for each of three variants of ERC/mesothelin, 1, 2 and 3, which had an 82 bp insert within the C-ERC/mesothelin region.21 In that study, purification and analysis of the ERC/mesothelin variants from the ascitic fluid of ovarian cancer patients using immunoaffinity chromatography with mAb 569 showed ERC/mesothelin variants 1 and 2, but not 3. The same approach should be used in mesothelioma patients.

NOVEL ELISA SYSTEM FOR N-ERC/MESOTHELIN

The 31 kDa N-ERC/mesothelin is secreted into the blood after cleavage by furin-like proteases, and ERC/mesothelin digestion is inhibited by furin inhibitors (Fig. 2). A mutated variant of ERC/mesothelin, in which an amino acid in the furin recognition site was substituted by alanine, was not digested by furin, even when expressed in mammalian cells (data not shown). These data have confirmed the processing pathway of the ERC/mesothelin protein. Therefore, we have focused on N-ERC/mesothelin, because N-ERC/mesothelin can probably be detected earlier and with higher sensitivity in cases of mesothelioma than C-ERC/mesothelin.

We have developed a novel ELISA system for the detection of N-ERC/mesothelin in the serum of mesothelioma patients, and have begun to examine its clinical usefulness.6 Mouse monoclonal antibodies (clone 7E7) and rabbit polyclonal antibodies (PoAb-282) were raised against human N-ERC/mesothelin and used for the establishment of the ELISA system. Serum samples from seven patients with mesothelioma, four patients with carcinomatous pleuritis pleural metastasis (lung cancer), three with benign asbestos pleuritis, and 13 healthy volunteers were evaluated. The diagnosis in patients with mesothelioma was confirmed immunohistochemically (all epithelial types).

The median serum concentrations in the seven mesothelioma patients, 13 healthy controls, four pleural metastasis (lung cancer) patients and three benign asbestos pleuritis patients were 25.12 ng/mL (range: 10.32–53.13 ng/mL), 1.4 ng/mL (range: 0.67–7.60 ng/mL), 1.31 (range: 0.68–4.33 ng/mL), and 1.10 ng/mL (range: 0.3–1.19 ng/mL), respectively (Fig. 3). Until now, our ELISA system has detected much higher serum levels of N-ERC/mesothelin in mesothelioma patients than in healthy controls or patients with other lung/pleural diseases.

Figure 3.

Scatter plot of serum N-ERC/mesothelin. The serum concentrations of N-ERC/mesothelin were much higher in mesothelioma patients than in the healthy controls, carcinomatous pleuritis (lung cancer) patients and benign asbestos pleuritis patients (cited in Shiomi et al.6).

We propose to conduct a prospective study to establish the clinical utility of our sandwich ELISA system and to determine the relationship between the serum N-ERC/mesothelin concentrations and the clinicopathological features of the patients, cut-off values, and the sensitivity/specificity of the measurement, to explore the possibility of early diagnosis.

Recently, Onda et al. also reported the generation of anti-MPF monoclonal antibodies and the development of ELISA to MPF (N-ERC/mesothelin).22 Their ELISA data showed that MPF levels were elevated in 91% of patients (51 of 56) with mesothelioma compared with healthy controls. Furthermore, MPF levels decreased in patients after tumor debulking surgery. Thus, N-ERC/mesothelin may represent a promising tumor marker for mesothelioma.

NOVEL MOLECULAR MARKER: PERSPECTIVE FOR THE FUTURE

Although mesothelioma is typically associated with asbestos exposure, the basic roles or physiological functions and relationship with carcinogenesis of ERC/mesothelin remain to be resolved. It was recently reported that mesothelioma is highly likely to be caused by asbestos exposure in genetically predisposed individuals in a certain limited area.23 There is also the possibility of existence of mesothelioma susceptibility genes.

Ramos-Nino et al. reported that the simian virus 40 (SV40) T-antigen (Tag) is present in a large percentage of human mesotheliomas.24 Approximately half of mesotheliomas in the USA are positive for SV40 Tag. Moreover, in recent epidemiological studies, the hazard ratio for risk of developing mesothelioma due to asbestos exposure alone, or SV40 alone, was compared with the hazard ratio due to asbestos exposure plus SV40 infection. The combination of SV40 infection plus asbestos exposure produced a risk of developing mesothelioma 27-fold higher than that of subjects with asbestos exposure alone. That study provides epidemiological support for a possible cocarcinogenic link between SV40 infection and asbestos exposure in the development of mesothelioma (cited in Ramos-Nino et al.24).

To clarify potential molecular and pathological pathways in mesothelioma, large-scale transcriptional studies using microarrays have been reported.25,26 Gordon et al. reported on transcriptional profiling using high-density oligonucleotide microarrays containing approximately 22 000 genes to elucidate potential molecular and pathobiological pathways in MPM using discarded human MPM tumor specimens (n = 40).25 Microarray gene expression data were confirmed using quantitative reverse transcriptase–polymerase chain reaction and protein analysis for three novel candidate oncogenes (NME2, CRI1, and PDGFC) and one candidate tumor suppressor (GSN).

The NME2 gene is a transcriptional regulator with diverse functions that binds and cleaves DNA via covalent bond formation and catalyzes phosphoryl transfer. The high levels of NME2 are mutagenic. NME2 protein has been shown to bind single-stranded telomeric DNA and the RNA component of telomerase. CRI1 is a cyclic AMP response element-binding protein (CREB)-binding protein and potential oncogene that antagonizes the action of pRb, p300, and CREB-binding protein (CBP) histone acetylase activity. Platelet-derived growth factor-C (PDGFC) is described as a novel transforming growth factor that participates in an autocrine signaling loop. Gelsolin (GSN) is known to be downregulated in multiple neoplasms and is capable of suppressing tumorigenicity by inhibiting protein kinase C activation (cited in Gordon et al.25).

Combination of these results may explain the tumorigenesis and/or pathology of mesothelioma. Moreover, these high-throughput gene expression data on mesothelioma could provide novel candidates for diagnostic markers for mesothelioma.

In clinical practice, pathological diagnosis by trained pathologists using authorized markers is still conducted in pleural biopsy specimens obtained by thoracoscopy. However, it is absolutely essential to establish a method for the early diagnosis or screening of malignant mesothelioma, because of the long incubation period and absence, until now, of effective therapeutic modalities for the tumor. ERC/mesothelin may be a probable candidate for an early serum marker of mesothelioma.

In a recent report, serum osteopontin was evaluated as a marker for mesothelioma.27 However, the concentrations of osteopontin in the serum do not represent the actual serum concentrations of osteopontin, because osteopontin is cleaved by thrombin, a component of the blood coagulation. Furthermore, elevation of the serum osteopontin level is not specific for mesothelioma.

Several molecular markers in immunohistochemical and ELISA study for mesothelioma have been developed and must be evaluated in parallel or systematic studies. Retrospective and prospective studies using molecular markers should be conducted in collaboration with imaging diagnosis.

ACKNOWLEDGMENTS

This work is supported by a Grant-in-Aid for Cancer Research and Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports and Science and Technology of Japan and the Ministry of Health, Labor and Welfare of Japan.

Ancillary