Malignant Pleural Mesothelioma

Published on 23/05/2015 by admin

Filed under Pulmolory and Respiratory

Last modified 22/04/2025

Print this page

rate 1 star rate 2 star rate 3 star rate 4 star rate 5 star
Your rating: none, Average: 0 (0 votes)

This article have been viewed 1719 times

Chapter 70 Malignant Pleural Mesothelioma

Malignant mesothelioma is a neoplasm arising from the mesothelial surface of the pleura, peritoneum, pericardium, and tunica vaginalis, with the majority of cases arising in the pleura. Malignant pleural mesothelioma (MPM) has progressively increased in incidence worldwide, associated with increasing exposure to asbestos in the post–World War II era. The disease is preceded by a long latency period, often more than 30 years. Incidence of MPM is expected to peak in the Western world in the next one or two decades because of regulation of asbestos use and increasing public awareness of its health risks since the 1970s and 1980s. However, MPM will continue to increase in many developing countries, where asbestos remains widely used.

Epidemiology

A relationship between asbestos exposure and development of mesothelioma was first reported in 1960 by Wagner et al., who described a cohort of 33 patients who developed diffuse malignant mesothelioma after exposure to asbestos in the North Western Cape Province of South Africa. Subsequent reports in many countries indicated a rising incidence of mesothelioma. Also, it was reported that the risk of developing mesothelioma was dose dependent, with the rate increasing significantly with time from first exposure, duration of exposure, and cumulative exposure. MPM is often preceded by a prolonged latency period of 20 to 60 years.

A review by Price and Ware of mesothelioma time trends in the United States based on the April 2008 release of the Surveillance Epidemiology and End Result (SEER) data estimated that 2400 cases were diagnosed in 2008, with asbestos likely to be the cause in 58%. They also reported a possible shift in the peak incidence in the United States from a peak between 2000 and 2004 to a peak between 2005 and 2010, and that after 2042, mesothelioma cases diagnosed in the United States will be predominantly caused by “background mesothelioma” rather than asbestos-related MPM.

A similar trend is evident in other Western countries and regions, where several decades of rising incidence of MPM have peaked or will peak in the next 20 to 30 years. The regulatory laws introduced progressively since the 1970s restricted asbestos use in various industries, eventually resulting in a total ban in most developed countries and regions. Again, in many Third World and developing countries where asbestos is still widely used and mesothelioma may also be underreported, incidence will increase for decades.

Mesothelioma has also been linked to nonoccupational and environmental exposure to asbestos in multiple epidemiologic studies. Malignant mesothelioma caused by both occupational and nonoccupational asbestos exposure was notably studied in Wittenoom, Western Australia. In areas of Turkey, exposure to tremolite asbestos, commonly used in the white stucco covering walls in houses has been linked to mesothelioma. Another fibrous mineral, erionite, has been associated with MPM in three “cancer villages” in the Cappadocia region of Turkey; later reports suggested that this occurs most often in families with genetic predisposition to mineral fiber carcinogenesis.

The simian virus 40 (SV40) was proposed to act as a “co-carcinogen” with crocidolite asbestos, resulting in malignant transformation, and was shown to induce mesothelioma in hamsters. The same group subsequently demonstrated SV40 viruslike DNA sequences in patients with malignant mesothelioma. The role of SV40 in mesothelioma remains controversial.

Radiation therapy has been linked to MPM in patients followed up after treatment for non-Hodgkin lymphoma. Carbon nanotubes, which are lightweight, strong cylindrical molecules, are increasingly being incorporated into modern manufacturing processes, including electronics, sports equipment, and protective clothing. Initially believed to pose no health risks, they appear to have some similarities to asbestos and were demonstrated to cause mesothelioma when injected into mice. A role for nanotubules in human mesothelioma has not been established, although any long incubation period would make it unlikely that cases would have yet been seen.

Although smoking and asbestos exposure can multiplicatively increase the risk of developing lung cancer, smoking has not been demonstrated to have an independent association with mesothelioma.

Pathogenesis

Asbestos is a term used to describe a group of mineral fibers of two major types: serpentine and amphibole. Chrysotile is the most common serpentine asbestos used commercially and has long, curly and pliable fibers. The amphiboles, which are short, straight, and stiff, include crocidolite (blue asbestos) and amosite. Noncommercial amphiboles include tremolite, which is present in the soil in Turkey, Greece, Afghanistan and other regions.

Controversy surrounds the role of different types of asbestos fibers and the exposure threshold for malignant transformation, particularly the role of chrysotile asbestos. The general consensus is that the amphibole crocidolite is the most carcinogenic form of asbestos, and that asbestos fibers of greater length and smaller diameter (high length/width ratio) are more carcinogenic, possibly because they can penetrate farther into the pleura.

Asbestos can affect the pleura by four main processes. First, asbestos fibers can penetrate deeply into the lung and irritate the pleura, eventually leading to either scarring (plaques) or malignant transformation. The fibers may also pierce the mitotic spindle and result in chromosomal damage, with loss of chromosome 22 being the most common gross change. Third, DNA can be damaged through the formation of iron-catalyzed free radicals. Fourth, asbestos can increase expression of proto-oncogenes through phosphorylation of mitogen-activated protein (MAPK) and extracellular signal-regulated kinases (ERK) 1 and 2. In addition, the following six alterations in cell physiology that can dictate malignant growth, as described by Hanahan and Weinberg, also appear to be present in mesothelioma:

The SV40 virus may play a role in the pathogenesis of MPM and is a subject of ongoing study to clarify its role. The SV40 large T antigen (SV40Tag) can inhibit p53 and Rb tumor suppressor genes. SV40 infection is also involved in activation of various pathways, including Wnt, ERK, PI3K/Akt, and Notch-1 pathways, and has a co-carcinogenic effect with asbestos.

Pathology

Histology

A definitive diagnosis of MPM is best made with an adequate biopsy in the context of compatible clinical findings. This often represents a diagnostic challenge because of the relatively low frequency of cases outside large tertiary care centers, the often limited amount of tissue available for examination, the heterogeneity of histologic appearance of the same tumor within a patient, and the differential diagnoses of benign pleural disease or other metastatic malignancies. With the increasing availability of immunohistochemical stains, molecular studies, and electron microscopy, diagnostic accuracy has improved. It remains critical for diagnosis to correlate the pathologic findings with the clinicoradiologic features to reach a definitive diagnosis.

The three main subtypes of diffuse MPM are epithelial (epithelioid), sarcomatous (sarcomatoid), and biphasic mesothelioma, with various histologic patterns within each subtype.

Epithelial Subtype

Epithelioid MPMs are the most common subtype and can be further characterized to a number of patterns. The tubulopapillary pattern contains a mixture of tubules and papillary structures (Figure 70-1). Adenomatoid and acinar patterns contain glandlike structures that should be differentiated from adenocarcinomas of other origin (e.g., lung). The solid, well-differentiated pattern has “nests” or sheets of cells that may resemble reactive mesothelium. Differentiating features from reactive mesothelial hyperplasia include stromal invasion, presence of necrosis and expansile nodules, as well as immunohistochemical staining patterns. Other types include solid, poorly differentiated, clear cell, and others.

Sarcomatous Subtype

Sarcomatoid MPMs usually show spindle cells, often have more necrosis and atypical cells than epithelioid patterns, may contain anaplastic or giant cells, and include the lymphohistiocytoid and desmoplastic patterns (Figure 70-2). The desmoplastic pattern contains dense areas of collagen with small amounts of tumor cells, without significant atypia, and can be difficult to differentiate from nonmalignant pleural conditions such as fibrosing pleuritis; differentiating features include stromal invasion and disorganized growth with uneven thickness.

Immunohisto/Cytochemistry

Immunohistochemistry (IHC) is an important diagnostic tool in MPM. The exact makeup of panels depends on the histologic subtype and type of tumor being considered in the differential diagnosis, as well as the available antibodies. The IMIG consensus statement recommends the use of positive and negative markers with sensitivity or specificity greater than 80% and positivity that considers stain localization and percentage staining (>10%), as follows:

Reactive mesothelial hyperplasia and epithelioid MPM may be differentiated by a combination of morphologic features and IHC staining; desmin is often positive in reactive hyperplasia, and epithelial membrane antigen (EMA) and p53 may be positive in epithelial MPM.

Malignant mesothelioma (MM) versus metastatic malignancy (often adenocarcinoma) is a common differential diagnosis. Adenocarcinomas often contain intracellular mucin on periodic acid–Schiff stain after digestion, which is only rarely seen in MM. As a general approach, initial IHC workup may include two mesothelial markers and two markers for the malignancy being considered, as well as pancytokeratin, which is positive in almost all MPMs, except less common sarcomatoid variants. If the tumor is keratin negative, panels for other cancers (e.g., melanoma, lymphoma) would be considered.

Positive mesothelioma markers include calretinin, keratin 5/6, WT-1 protein, and podoplanin (D2-40). Markers that favor lung adenocarcinoma include Ber-EP4, which is positive in 95% to 100% of cases, compared to less than 20% of mesotheliomas. Carcinoembryonic antigen (CEA), TTF-1, MOC-31, BG8, B72.3, and CD15 are often positive in adenocarcinoma but only in up to 10% of mesotheliomas.

When lung squamous cell carcinoma (SCC) is suspected, keratin 5/6 is not useful, being positive in both mesothelioma and lung SCC, but p63 is useful as it is expressed in almost all SCCs and in 7% of mesotheliomas. MOC-31, BG8, and Ber-EP4 are also often positive in lung SCC.

If initial IHC panel results are conflicting, these can be expanded to include additional markers.

Immunocytochemistry (ICC) follows the same logic: determine if the cells are of mesothelial origin (e.g., using calretinin; Figure 70-3), then determine if they are malignant (e.g., using EMA to determine if there is a dense peripheral staining pattern; Figure 70-4). It is essential that the correct anti-EMA antibody is used because some give better results than others, partly contributing to dissent on the role of ICC in diagnosing mesothelioma.

Clinical Presentation

Malignant pleural mesothelioma is most common in men, with age at presentation typically 50 to 70. Presentation at a younger age may also occur in the patient with significant childhood exposure. The insidious onset of symptoms often leads to a delay of 2 to 3 months from initial symptoms to diagnosis, and 25% of patients may not present for evaluation for 6 months or more. Most patients present with a pleural effusion, and the majority develop an effusion at some stage of disease. The most common symptoms at presentation are breathlessness (40%-70%) and chest pain (60%). Constitutional symptoms such as weight loss and fatigue often are not present at diagnosis but develop with progression of disease. Patients may also present with symptoms from local spread causing superior vena cava syndrome, spinal cord compression, pericardial involvement, or chest wall invasion.

The only abnormal physical finding is often of a unilateral pleural effusion, which more frequently (60%) affects the right hemithorax. Tumor spread through the pleural cavity may result in signs of a “fixed” hemithorax. There may be palpable chest wall masses or signs caused by local or distant spread.

There are no specific laboratory features for MPM. Nonspecific findings include anemia, hypergammaglobulinemia, eosinophilia, and thrombocytosis, which is present in 60% to 90% of patients. Pulmonary function tests may show a restrictive pattern, which may result from lung encasement or chest wall involvement.

Distant metastases were present postmortem in 54% to 82% of mesothelioma patients in one large series. However, distant metastases usually are not evident clinically at diagnosis and are not the cause of death. Intracranial metastases may be present infrequently (3%). With the increasing use of positron emission tomography (PET) for diagnostic workup, 5% to 25% of patients may show extrathoracic disease at diagnosis. Miliary spread of mesothelioma has also been described.

Diagnosis

The initial diagnostic workup for patients presenting with an unexplained pleural effusion or pleural thickening is pleural aspiration. Pleural biopsy by sampling an area of the affected pleura with an Abrams needle can be performed, but this blind biopsy is painful and has a yield of only 39% to 60%. In a randomized study, computed tomography (CT) guidance had significantly higher sensitivity than blind biopsy (87% vs. 47%).

Thoracoscopy may be indicated if pleural aspiration is nondiagnostic and for patients who develop recurrent pleural effusions. It has the advantage of a higher sensitivity (98%) and the ability to estimate the extent of involvement of the mediastinum, chest wall and diaphragm, and providing adequate tissue for histologic classification in patients being considered for surgery. If a medical or video-assisted thoracoscopy is not possible, an open biopsy may be necessary. The development of chest wall masses from malignant seeding of biopsy tract sites after thoracocentesis or biopsy is an uncommon complication and may be prevented by radiotherapy to the site after instrumentation, although recent work questions this.

Imaging

The most common finding with MPM on plain chest radiography is a unilateral pleural effusion (Figure 70-5). Pleural thickening may also be seen, or in more advanced cases, an encircling “rind” of tumor or lobulated pleural masses. Signs of asbestos exposure, such as pleural plaques, may be present but are not a precursor to mesothelioma.

Computed tomography of the chest may show only a pleural effusion (74%). Pleural-based masses with or without thickening of the interlobular septa are not seen in most patients at presentation but develop later in the disease course (Figure 70-6). Lower lobe collapse may result from pleural encasement, pulmonary nodules (60%), or lymphadenopathy affecting the hilar and mediastinal lymph nodes.

Magnetic resonance imaging (MRI), although not commonly used, may be useful in determining the extent of tumor spread, particularly diaphragmatic invasion or assessment of suspected spinal cord metastases.

Positron emission tomography with fluorodeoxyglucose (FDG-PET) may be helpful to distinguish MPM from benign pleural disease (Figure 70-7). PET may also be useful in excluding extrathoracic disease in patients being considered for surgery, but is less accurate for defining locoregional spread to mediastinal lymph nodes. Integrated PET-CT may be more reliable in defining extent of disease than PET, CT, or MRI alone. Changes in PET results may predict response to chemotherapy.

Biomarkers

Two biomarkers, soluble mesothelin-related peptides (SMRPs) and osteopontin, have been extensively investigated as potential diagnostic or treatment monitoring markers. Mesothelin is a 40-kDa glycoprotein expressed on the surface of mesothelial cells and is overexpressed in mesothelioma. Serum SMRPs were increased in 84% of 44 patients with MPM, and in only three (2%) of 160 control patients. SMRPs measured in pleural effusions have also been investigated. In 192 patients presenting to respiratory clinics, elevated SMRP levels in effusions had a sensitivity of 67% and specificity of 98% for a diagnosis of mesothelioma, often several months before a definitive diagnosis was eventually made (Figure 70-8). Their utility has been validated, and SMRPs are now widely available. Of note, the terms mesothelin and SMRP are often used interchangeably in the literature.

Osteopontin is a glycoprotein that is overexpressed in several cancers. In a report comparing osteopontin levels in MPM patients and in asbestos-exposed patients, a high level of osteopontin had a sensitivity of 77.6% and specificity of 85.5% for mesothelioma, suggesting that it may be useful for distinguishing these two groups. However, osteopontin lacks specificity in other patients with pleural disease compared with SMRPs and thus is not recommended.

Prognosis

The median survival of patients with malignant mesothelioma receiving best available treatment is 12 months, or as short as 4 months in patients receiving supportive care alone. The most useful clinical prognostic scoring systems are those derived by the Cancer and Leukaemia Group B (CALGB) and the European Organisation for Research and Treatment of Cancer (EORTC), both validated in subsequent studies. The CALGB index used Cox survival models and regression trees to examine the effects of pretreatment prognostic factors on the survival of 337 patients treated in seven Phase II CALGB clinical trials. Multivariate Cox analysis showed increasing age over 75, pleural involvement, elevated LDH, poor performance status, chest pain, thrombocytosis, and nonepithelial histology to predict a poor prognosis, and six prognostic groups were identified based on these factors. The EORTC identified prognostic factors in 204 patients enrolled in five Phase II trials and reported that a poor performance status, high white blood cell count, male gender, sarcomatous histology, and a probable/possible histologic diagnosis were independent factors of poor prognosis. Patients were classified into a good-prognosis group, with a 40% 1-year survival, and a poor-prognosis group, with a 12% 1-year survival.

An elevated neutrophil/lymphocyte ratio (NLR) at baseline was reported to be an independent prognostic factor in a group of 173 patients undergoing systemic therapy, and normalization of NLR after one cycle of treatment was a predictor of improved survival on subgroup analysis.

The standardized uptake value (SUV) on PET correlated with survival. Multivariate analysis of 65 patients with MPM demonstrated a longer survival for the group with low-SUV tumors than those with high-SUV lesions.

Gene expression array technology has been investigated as a potential prognostication tool. A molecular predictive test that stratifies patients with good or poor outcomes using relative expression of four genes has been prospectively validated in a group of patients undergoing surgery. Patients with good prognosis had a median survival of 16.8 months, versus 9.5 months for those with poor prognosis. There are no validated tests at present for patients not receiving therapy or for those receiving chemotherapy.

Management

The prognosis of the patient with MPM remains poor, with almost all patients eventually succumbing to their illness. Management options include surgery, chemotherapy, and radiotherapy, with the goal of palliating symptoms or prolongation of survival.

Surgery

The role of surgery in MPM, particularly radical surgery, remains controversial. No randomized controlled trials (RCTs) support radical surgery, and there is considerable difficulty in conducting such trials. Four types of surgery are performed as treatment for mesothelioma: extrapleural pneumonectomy (EPP), pleurectomy/decortication (P/D), limited pleurectomy, and thoracoscopy with pleurodesis. EPP usually involves an en bloc resection of the lung, pleura, pericardium, and diaphragm, whereas P/D involves resection of the parietal and visceral pleurae, pericardium, and diaphragm but spares the lung.

Although medium-term to long-term survival after multimodality treatment with EPP and radio/chemotherapy has been reported in surgical series from specialized centers, the optimal procedure for patients undergoing resection for early MPM is controversial. Most studies have included exclusively either P/D or EPP as part of multimodality therapy, and selection bias plays a major role in determining the type of surgery, or indeed whether surgery is considered. A retrospective case-control study comparing EPP and P/D for pathologic N2 disease in 57 patients suggested that P/D did not compromise survival for these patients. Retrospectively analyzing outcomes for 663 patients undergoing EPP or P/D at the Memorial Sloan-Kettering Cancer Center, National Cancer Institute, and Karamanos Cancer Institute, Flores et al. reported that patients who underwent P/D had better survival than those undergoing EPP, but acknowledged that the reasons were multifactorial and subject to selection bias, suggesting that therapy should be tailored for individual patients. Operative mortality was 4% for P/D and 7% for EPP.

An important unanswered question is whether radical surgery offers any advantage over nonsurgical therapies. Some patients survived long term after surgery, but these patients might have had slow-growing disease and may have had similar survival without surgery. In most surgical series, the survival benefit could not be specifically attributed to the procedure rather than other components of multimodality therapy. Furthermore, survival analysis for patients who were well enough postoperatively to receive radiotherapy and chemotherapy introduces further selection bias. The Mesothelioma and Radical Surgery (MARS) trial, a feasibility study randomizing patients to trimodality therapy (EPP, chemotherapy, radical radiotherapy) or identical chemotherapy, followed by appropriate nonsurgical treatment, established that randomization to surgical or nonsurgical resection was feasible, but the study was not powered to address effectiveness and may not provide a definitive answer on the role for surgery. The MARS investigators have proposed a further trial of surgery (MARS-2), that may evaluate P/D as well as respiratory function and quality of life.

For the small proportion of patients presenting with early, potentially resectable MPM, thorough staging with PET-CT, bilateral thoracoscopy, mediastinoscopy, and laparoscopy may avoid major surgery in patients with disease outside the ipsilateral hemithorax. These patients should be counseled on the risks and benefits of radical surgery, including the limitations of available evidence, and referred to a thoracic surgeon with experience in radical surgery, if appropriate.

Limited pleurectomy or thoracoscopy with talc pleurodesis are palliative procedures that can be effective for control of pleural effusions (discussed later).

Chemotherapy

Systemic platinum-based chemotherapy is the only treatment for the patient with MPM that has a demonstrated survival benefit from RCTs. In 2003 a landmark study of 226 patients with advanced MPM randomized to cisplatin and pemetrexed combined or cisplatin alone. Median survival improved from 9.3 months for the cisplatin arm to 12.1 months for the cisplatin/pemetrexed group (p = 0.02). Radiologic objective response rate, time to progression, and quality of life and symptom distress measures also improved in the two-drug group. In another RCT, the addition of raltitrexed to cisplatin showed a similar overall survival benefit: 11.4 months for cisplatin/raltitrexed versus 8.8 months for cisplatin.

Neither of these trials compared chemotherapy to best supportive care, an issue investigated in the MS01study, which compared “active symptoms control” (ASC) to vinorelbine or to mitomycin, vinblastine, and cisplatin (MVP). Because of slow accrual, the two chemotherapy arms were combined, and no overall survival benefit was demonstrated over ASC. However, the two chemotherapy arms were significantly different, with vinorelbine demonstrating a 2-month survival advantage that approached significance, suggesting that MVP was not an active regimen in this setting.

The cisplatin/pemetrexed combination has become the standard first-line therapy for patients with MPM in many centers, but carboplatin is often substituted for cisplatin for simpler administration and less toxicity, with Phase II evidence of activity and response rates of 20% to 30%. An expanded-access program demonstrated a lower response rate for carboplatin/pemetrexed but similar 1-year survival and time to progression as with cisplatin/pemetrexed. Other regimens with a good response rate include cisplatin and gemcitabine, with good response rates in Phase II trials, but this has not been tested in Phase III trials.

There is no second-line therapy with a proven survival advantage that can be considered the standard of care in the MPM patient. It may be reasonable to use pemetrexed in the second-line regimen when not used in first-line therapy, although there is only low-level evidence for this strategy. The timing of initiating chemotherapy (early vs. delayed) in one small randomized study showed an advantage for the early-treatment strategy, but used MVP, not currently considered an active regimen. Other strategies being investigated include the role of maintenance therapy after first-line treatment and the addition of a third cytotoxic agent to initial two-drug chemotherapy.

Controversies and Pitfalls

Patients with MPM often have an insidious onset of symptoms that may lead to a delay in diagnosis. A high index of clinical suspicions and a comprehensive history, including occupational and environmental exposure to asbestos, as well as appropriate investigation may reduce this delay and enable patients to receive potentially beneficial therapies.

Some centers experienced in cytology methods are able to make a definitive diagnosis of MPM on cytologic samples, using specific antibodies, but centers with less experience continue to rely on tissue samples. Definitive diagnosis of MPM may be challenging because of limited available tissue and heterogeneous histologic appearance. Adequate tissue specimens examined by pathologists experienced in diagnosing MPM and using appropriate immunohistochemical panels to differentiate MPM from benign or other malignant disorders, as well as correlation with clinicoradiologic findings, enable a definitive diagnosis in most patients.

Novel biomarkers such as serum or pleural effusion levels of mesothelin may help differentiate benign from malignant causes of pleural effusions that would benefit from further investigation, but these are not in widespread clinical use. Mesothelin has high specificity in blood and pleural fluid and should probably be used more often in the workup of patients suspected of having pleural malignancy.

The role of surgery in the treatment of MPM remains controversial, with improved survival in select patients, who may have had a similar outcome with nonsurgical therapies. For MPM patients considered surgical candidates, comprehensive staging and discussion of available options can avoid a major, potentially futile procedure.

Palliative chemotherapy remains the only treatment modality with a proven survival benefit in patients with advanced MPM able to undergo treatment. Radiotherapy may play a role in palliating symptoms or as a component of multimodality therapy. Because cures are rare, future studies will seek to generate new strategies to circumvent the chemoresistance of MPM.

Malignant pleural mesothelioma research faces many unanswered questions, with limited effective therapies for an aggressive disease with a poor prognosis that will continue to increase in incidence for decades.

Suggested Reading

Cagle PT, Allen TC. Pathology of the pleura: what the pulmonologists need to know. Respirology. 2011;16:430–438.

Creaney J, Francis RJ, Dick IM, et al. Serum soluble mesothelin concentrations in malignant pleural mesothelioma: relationship to tumor volume, clinical stage and changes in tumor burden. Clin Cancer Res. 2011;17:1181–1189.

Creaney J, Yeoman D, Demelker Y, et al. Comparison of osteopontin, megakaryocyte potentiating factor, and mesothelin proteins as markers in the serum of patients with malignant mesothelioma. J Thorac Oncol. 2008;3:851–857.

Edwards JG, Abrams KR, Leverment JN, et al. Prognostic factors for malignant mesothelioma in 142 patients: validation of CALGB and EORTC prognostic scoring systems. Thorax. 2000;55:731–735.

Flores RM, Pass HI, Seshan VE, et al. Extrapleural pneumonectomy versus pleurectomy/decortication in the surgical management of malignant pleural mesothelioma: results in 663 patients. J Thorac Cardiovasc Surg. 2008;135:620–626. e1–e3

King J, Thatcher N, Pickering C, et al. Sensitivity and specificity of immunohistochemical antibodies used to distinguish between benign and malignant pleural disease: a systematic review of published reports. Histopathology. 2006;49:561–568.

Price B, Ware A. Time trend of mesothelioma incidence in the United States and projection of future cases: an update based on SEER data for 1973 through 2005. Crit Rev Toxicol. 2009;39:576–588.

Robinson BW, Lake RA. Advances in malignant mesothelioma. N Engl J Med. 2005;353:1591–1603.

Robinson BW, et al. Mesothelin-family proteins and diagnosis of mesothelioma. Lancet. 2003;362:1612–1616.

Whitaker D. The cytology of malignant mesothelioma. Cytopathology. 2000;11:139–151.