A. Scherpereel [1, 2 and 3], B.-D. Grigoriu [3 and 4], P. Astoul [5 and 6]
[1] Department of Pulmonology and Thoracic Oncology, Calmette Hospital, Lille University Hospital, Lille, France.
[2] Faculty of Medicine Henri-Warembourg, University of Lille II, Lille, France.
[3] INSERM 774 unit, Institut Pasteur de Lille, Lille, France.
[4] University of Medicine and Pharmacy, Department of Pulmonology, Iasi, Romania.
[5] Department of Respiratory Diseases, Thoracic Oncology Unit, Sainte-Marguerite Hospital, Marseille, France.
[6] EMI Inserm 0359, Laboratory of Experimental Cancerology, University of La Méditerranée, Marseille, France.
[7] Department of Pulmonology and Thoracic Oncology, Calmette Hospital, Lille University Hospital, 59037 Lille cedex, France.
summary
Malignant pleural mesothelioma (MPM) is cancer worrying by its increasing incidence and its poor prognosis, despite real progress in the management of patients. The diagnosis is generally made at an advanced stage when radical treatment is no longer possible. A significant increase in survival in MPM requires, in particular, an earlier diagnosis of the disease. Biologically, several potential soluble tumor markers, including mesothelin and osteopontin, have been proposed to aid in the diagnosis, but none have been validated to date. Soluble mesothelin, measured in blood and pleural fluid, seems to be the most interesting candidate. But even if this marker has a good diagnostic and prognostic value, its high specificity for the epithelioid subtype of mesothelioma (the most frequent) limits its usefulness in practice. The other molecules are not very interesting, despite sometimes a good sensitivity, in particular, because of their low diagnostic specificity. In conclusion, the available data do not yet argue for routine use of biology in the diagnosis of MPM, thoracoscopy and histology remaining the references. Soluble mesothelin should continue to be evaluated in clinical studies and look for other tumor markers.
Introduction
Malignant pleural mesothelioma (MPM) is a primary tumor of the pleura, very aggressive and with an overall unfavorable prognosis with a median survival of about one year [1]. The main etiological factor is prior exposure to asbestos, with a typically very long latency period between this exposure and the onset of MPM, in the order of 20 to 40 years [2], [3]. Formerly considered a rare tumor, the incidence of mesothelioma is gradually increasing [2], in parallel with the increasing use of asbestos fibers which, in France, reached its peak in the 70s and 80s (especially in industry and the building). The use of asbestos has been banned in France since 1997. In France, there are 800 to 1,000 cases of MPM per year with an incidence of around 16 cases / million inhabitants/year [4]. Estimates predict a peak incidence of MPM around 2020 with 1140-1300 cases/year [5]. In addition, asbestos is still widely used in emerging countries, suggesting a sustainable increase in the number of mesothelioma cases worldwide.
To date, there is no validated curative treatment for MPM. The majority of patients typically benefit from chemotherapy and/or supportive care. Despite the recent appearance of new chemotherapy regimens including the combination of cisplatin and an antimetabolite [6], the increase in survival remains modest compared to previous protocols. Some authors propose a radical surgical treatment by pleurectomy at a very early stage of the disease (limited involvement of the parietal pleura, IMIG stage Ia) which could improve the survival of these patients (survival in the surgical series from 15% to 5 years) [7]. But these patients currently represent only 1 to 2% of the total, which has so far failed to validate this attitude. Multimodal treatment for MPM, including chemotherapy, broad excision surgery (enlarged pleuropneumonectomy, PEP) and hemithoracic radiotherapy, could also provide benefit at later stages of MPM, but resectable, subject to a very strict selection of candidates [8]. However, if the feasibility of this multimodal approach has been demonstrated [9], there is currently no large randomized clinical study that has validated the interest of this heavy surgery (PEP) or a multimodal treatment for MPM.
Clinical diagnosis and imaging
The clinical and radiological signs of MPM are unfortunately not very specific and often appear at an advanced stage of the disease [7]. Exposure, most often occupational, to asbestos is frequently identified (around 80% of cases in men versus 50% in women) and must always be carefully investigated [3]. Superior to the chest radiograph, multi-bar helical computed tomography (CT) can help in the diagnosis and evaluation of the stage of MPM (grade of recommendation A according to the SPLF expert conference). The fixation of 18FDG during a positron emission tomography (PET) estimated by the "Standard Uptake Value" (SUV) is significantly increased in the case of malignant mesothelioma. A sensitivity of 91% and a specificity of 100% were found to differentiate MPM from benign pleural lesions induced by exposure to asbestos which do not fix 18FDG [7]. But for the diagnosis of MPM, it is recommended not to systematically perform a PET scan (expert opinion) [7]. Thus, despite advances in imaging (CT ± PET…), the majority of MPMs are diagnosed too late for radical treatment.
Pathology diagnosis
Analysis of pleural fluid, conventionally of the exudative type, is not very helpful in diagnosing MPM but can help to eliminate a differential diagnosis. The cytology of the pleural fluid is useful for the diagnosis of cancer but rarely makes the difference between mesothelioma and pleural adenocarcinoma. The diagnosis of MPM is made in only 20-30% of cases by cytology and 20-23% of cases by percutaneous pleural biopsy which does not usually bring enough material to confirm this diagnosis. The combination of the two techniques gives a yield of 35 to 40% [7]. Pleuroscopy or medical thoracoscopy, a simple examination that can be performed in a large majority of patients, is the most sensitive means (> 95%) for obtaining the diagnosis of MPM [7]. It is therefore recommended not to rely on the analysis of pleural fluid for the diagnosis of MPM, to reserve transparietal biopsies with or without detection by CT or ultrasound for patients for whom thoracoscopy cannot be considered, and to always perform thoracoscopy unless there are contraindications related to the patient's condition or a technical impossibility (pleural symphysis) (grade A recommendation) [7]. Diagnosing MPM based on the pathological examination of pleural biopsies can be difficult. For this reason, it is recommended not to diagnose MPM on an extemporaneous examination (A), but to always base it on a morphological examination supplemented with an immunohistochemical analysis (A) [7].
Biological diagnosis
The search for specific soluble tumor markers is therefore a major challenge in helping to diagnose, but also to assess the prognosis and response to treatment of patients with MPM. To date, however, there is no established marker for the positive diagnosis of MPM among the candidates considered below [7].
Hyaluronic acid
Hyaluronic acid (HA) has been proposed as a diagnostic marker, but despite good specificity, the sensitivity of its dosage remains modest (between 40 and 70%) both in serum (increase in its level only at an advanced stage of disease) than in pleural fluid [10]. Elevated serum HA values can also be encountered in rheumatoid arthritis and Wilms' tumor as well as in liver disease with fibrosis. In addition, a significant percentage of mesotheliomas do not secrete HA. However, elevated pleural HA levels (> 100 mg / L) are considered specific for the diagnosis of MPM.
ACE and Cyfra 21-1
The other markers studied to date are the carcinoembryonic antigen (CEA), an oncofetal glycoprotein, and the cytokeratin fragment (Cyfra) 21-1. A CEA level> 3 ng / ml in the pleural fluid would exclude MPM [11]. The serum level of Cyfra 21-1 is also high in metastatic pleurisy, but its level measured in pleural fluid is significantly higher in MPM than in other malignant pleurisy [12]. However, elevated levels of this marker can also be measured in patients with lung, breast or pancreatic cancer. Finally, the Cyfra 21-1 dosage would have a prognostic value in the MPM. The CA 125, CA19-9 and CA 15-3 assays have not been shown to be useful in diagnosing MPM. These markers (CA125 and CA19-9) are indeed secreted by normal mesothelial cells, but their rate is increased in inflammatory states, greatly reducing their diagnostic specificity [13]. More recently, soluble mesothelin (also called soluble mesothelin-related peptides, SMRP) [14], osteopontin [15] and the megakaryocyte potentiation factor (MPF) [16] have been suggested as potential tumor markers of MPM.
Soluble mesothelin and the megakaryocyte potentiation factor (MPF)
Mesothelin is a glycoprotein expressed on the surface of normal mesothelial cells, but also strongly in malignant mesothelioma and other neoplasias. It is initially produced in the form of a 69 kDa precursor linked to the membrane by a GPI anchor [17]. This precursor is then cleaved by furin to generate a 40 kDa fragment which remains bound to the membrane (mesothelin) and another soluble fragment of approximately 31 kDa called megakaryocyte potentiation factor (MPF) (Figure 1). Little is known about the biological activity of these molecules. MPF is said to stimulate the growth of megakaryocytes. Soluble mesothelin could be involved in cell adhesion by interaction with the CA125 antigen and thus in the process of peritoneal dissemination of ovarian carcinomatous cells [18].
For the diagnosis of MPM, two uses of mesothelin and MPF were considered: on the one hand, as markers in immunohistochemistry (for the membrane form of mesothelin) and, on the other hand, as soluble markers. The low specificity of membrane mesothelin makes it a poor immunohistochemical marker for distinguishing MPM from different carcinomas. Indeed, it is constantly expressed by normal mesothelial cells. In addition, if mesothelin is expressed by a large majority of epithelioid type MPM (but 20 to 30% of cases are negative), sarcomatoid subtypes are systematically negative by immunostaining [19]. Mixed forms express this marker in a variable percentage (20-30%). Finally, membrane mesothelin is also expressed in almost 40% of adenocarcinomas and 30% of pulmonary squamous cell carcinomas. In ovarian (> 95% positive) and pancreatic cancers, the expression of mesothelin is frequently focal and localized at the cytoplasmic level as opposed to membrane expression by the majority of cells in the epithelioid MPM.
By ELISA assay using a pair of anti-mesothelin monoclonal antibodies, a soluble form of mesothelin has been identified [20]. It is not known exactly how this soluble protein is synthesized and secreted, but two hypotheses are currently plausible. It could be either a mutated form of mesothelin which thus loses its binding capacity to the membrane or a cleavage of the membrane mesothelin by an unknown enzyme. For the moment, the available data suggest that the second hypothesis is the most plausible [21]. It would, therefore, be preferable to date to use the term soluble mesothelin for the circulating form of the molecule, rather than the designation of SMRP initially proposed. To date, there are two pairs of antibodies that can be used for the ELISA assay of mesothelin: that developed by Scholler et al. [20] (MesomarktTM commercial kit - Fujirebio Diagnostics, USA or CisBio International, France) and that developed by Hassan et al. [22].
The blood test for soluble mesothelin has shown elevated levels in patients with MPM or an ovarian tumor. Serum mesothelin values are higher in ovarian cancer than in benign ovarian tumors and appear to be correlated with disease spread and patient prognosis [12]. The first results published by Robinson et al. [14], based on the technique developed by Scholler [20], suggested that this marker would have good diagnostic sensitivity (> 80%) and very good specificity (> 95%) for epithelioid MPM. In this study, there was no increase in the serum mesothelin level for sarcomatoid MPM, thus confirming the data in immunohistochemistry, as well as in patients with inflammatory pulmonary disease, pleural neoplastic disease outside of MPM or transudative pleurisy. (figure 2). An increase in the serum mesothelin level was observed in 17.5% of healthy subjects exposed to asbestos (n = 7), 3 of whom secondarily developed a BPM, suggesting the possibility of using this marker for screening for BPM [14 ]. The level of serum mesothelin was correlated with tumor size and there was a tendency for this level to gradually increase during the course of the disease [14]. On the other hand, chemotherapy did not significantly influence this rate, while partial surgical resection would make it possible to obtain a decrease in its value [14]. The results of two unpublished studies, one from our team (Grigoriu et al., ERS communication 2007, manuscript in preparation), the other from the American H. Pass team in patients who have undergone chemotherapy and/or surgical excision seem to confirm this hypothesis. Based on these preliminary American results, the commercial kit MesomarktTM however obtained in January 2007 from the Food and Drug Administration (FDA) in the United States the approval of use on a compassionate basis as a follow-up marker for patients with a BPM. In another study including more patients including 60 MPM, we confirmed the diagnostic utility of soluble mesothelin measured in the blood and pleural fluid of patients with MPM compared to those with benign pleural lesions associated with exposure to asbestos or pleural metastases [23]. A threshold value of serum mesothelin of 0.93 nM / l obtained a sensitivity of 80% and a specificity of 83% of the diagnostic test. There were, however, a significant number of pleural metastases from adenocarcinoma also with elevated levels of mesothelin. The pleural dosage of mesothelin did not provide any additional information compared to that of the serum dosage [23], a result discussed by other authors [12].
By extension of this study on a population of 96 MPM, we have also highlighted a prognostic role for serum mesothelin at the threshold of 3.5 nM / l. Patients with a high level of mesothelin had a median survival of 7 months compared to 19 months in the group with a low marker level [24]. Using another pair of anti-mesothelin antibodies, Hassan et al. have confirmed the diagnostic utility of mesothelin in epithelioid-type mesotheliomas and the existence of high rates also in patients with ovarian carcinoma [22].
Like soluble mesothelin, elevated serum MPF levels have been found in patients with ovarian cancer or epithelioid-type mesotheliomas compared to healthy controls and patients with pleural metastases from pulmonary adenocarcinoma or pleural lesions benign [16]. These preliminary data suggest a very good specificity (virtually 100%) with a sensitivity of around 90% of MPF but need to be confirmed by studies including a larger number of patients with MPM or various pathologies (pleural metastases different carcinomas, benign asbestos pleurisy, etc.).
Due to its lack of specificity, the usefulness of mesothelin immunohistochemistry labeling is limited to a few diagnostic "niches". The determination of serum and/or pleural mesothelin appears useful for the diagnosis of MPM, subject to an average sensitivity in the studies (approximately 70% for a specificity of 90%). An "aggressive" exploration strategy for patients with a history of asbestos exposure, pleural abnormalities, and high serum mesothelin levels could be discussed, but its practical utility remains to be validated. Likewise, larger studies are needed to define the use of serum mesothelin as a prognostic and predictive marker in MPM. On the other hand, the use of this marker for screening for MPM in patients exposed to asbestos seems very improbable, if not impossible, due to the low incidence of MPM compared to a large number of patients exposed, its insufficient specificity ( <50% for sensitivity> 90%) and the absence of curative treatment for MPM to date. Studies on this subject are however in progress in Europe as in the USA. Data on MPF are still very limited to suggest any practical use, but the available data suggest a better diagnostic value than for mesothelin [25].
osteopontin
Osteopontin is a pleiotropic molecule, initially described in the bone where its potential role would be to make the link between bone cells and the extracellular matrix. It is also synthesized by malignant cells and acts as a key cytokine in the granulomatous reaction. Osteopontin has also been implicated in tumor progression. Elevated serum levels have been reported in ovarian, colic, mammary, prostate or lung cancers. However, osteopontin is also expressed in benign pleural and extrapleural pathologies (pulmonary tuberculosis, etc.) [12]. Several commercial assays are available (Immunobiological Laboratories (IBL), R & D Systems, Assay design), but comparative studies show that there are significant differences in results between the different kits. In 2005, Pass et al. have, for the first time, reported that osteopontin is expressed in a vast majority (95%) of MPM, regardless of the histological subtype [15]. Serum osteopontin levels were higher in patients with BPM compared to subjects exposed to asbestos with benign pleural and/or pulmonary lesions. In this study, there was no relationship between the tumor stage (IMIG classification) and the serum osteopontin level; as a result, the diagnostic utility of osteopontin remained very high, even in the early stages of the disease (stage I). In patients exposed to asbestos, there was a relationship between the serum osteopontin level and the duration of exposure, but also with the type and intensity of the radiological abnormalities present in these patients (pleural plaques, pleural thickening, pulmonary fibrosis). However, this study did not include patients with other types of tumors (pleural or not) or other non-tumor pleural pathologies. We compared the usefulness of osteopontin and soluble mesothelin in a series of 172 patients with MPM and 112 asymptomatic subjects with occupational exposure to asbestos [24]. This study showed that serum osteopontin values are higher in patients with MPM compared to patients exposed to asbestos, thus confirming the data published by Pass et al. [15]. However, this marker was less useful than mesothelin, with a smaller area under the ROC curve. We also found high values of osteopontin in patients with pleural metastases from carcinoma or benign pleural lesions secondary to asbestos exposure. This same observation is made for the determination of osteopontin in pleural fluid. Despite very good sensitivity, the dosage of osteopontin, therefore, has low specificity for MPM, which makes it difficult to use as a diagnostic marker or screening for MPM in patients highly exposed to asbestos [24].
Biological perspectives in MPM
Diagnostic perspectives
Creaney et al. recently reported a negative result from the evaluation of urinary mesothelin for the diagnosis or screening of MPM (abstract ATS 2008, Toronto - Canada). Given the limited diagnostic capacities of known markers, the question arises of the advantage of combining markers within a panel in order to improve diagnostic performance compared to a marker evaluated in isolation, in this case, the serum or pleural mesothelin. In our experience, the combination of osteopontin with serum mesothelin does not provide any advantages over the measurement of mesothelin alone in the diagnosis of MPM [24]. Other authors have shown that the combination of mesothelin with other potential markers such as Cyfra 21.1 or CA 125 was also not useful in the diagnosis of MPM [11], [12]. We also evaluated the combination of hyaluronic acid (HA) and mesothelin measured in the blood or pleural fluid (manuscript submitted for publication). Adding the dosage of HA to that of mesothelin in the blood could guide the diagnosis in about 15% of cases of MPM when the dosage of mesothelin was negative. The practical interest of this element seems however weak in patients generally receiving a thoracoscopy, histology being to date the key element of the diagnosis of MPM. Interestingly, this additional interest in HA was not limited to patients with epithelioid-type MPM. The evaluation of circulating HA could, therefore, be considered in patients strongly suspected of MPM, but without histological evidence of cancer or radical treatment envisaged and with a low serum mesothelin level. The levels of the two markers (HA and mesothelin) were highly correlated in the pleural fluid, making the HA dosage redundant in practice.
In the current state of knowledge and subject to its sensitivity limited by its specificity to the epithelioid subtype, serum mesothelin is the potential tumor marker of MPM with the best diagnostic value and can, therefore, be considered as the reference marker to which all the other candidates will have to be compared. Panels including soluble mesothelin and other potential diagnostic markers for MPM (survivin, etc.) measured in the pleural fluid are being evaluated. Finally, it should be remembered that no dosage of blood or pleural biological marker, including that of mesothelin, has recognized legal diagnostic value in patients suspected of MPM without histological evidence in order to obtain possible compensation for their pathology under occupational diseases (table n ° 30 of the RG) or with the compensation fund for asbestos victims (FIVA).
Prognostic and predictive markers
There are no significant published data on the usefulness of any marker candidate as a prognostic element or for the evaluation of the response to treatment in MPM. Pleural HA and serum mesothelin [12], [24] have already been studied as prognostic markers, but the available data are still insufficient to draw a clear conclusion and even less to make a recommendation for practice. However, the use of the mesothelin blood test has been validated by the US Food and Drug Administration (FDA) for this latter indication. The analysis of data obtained jointly at the Lille University Hospital and the University of Pennsylvania (USA) supports the interest of monitoring serum mesothelin as a predictive marker of the therapeutic response of MPM (manuscript in preparation).
Therapeutic prospects [12]
Recently, it has been shown that 40% of patients with MPM or ovarian carcinoma have antibodies to mesothelin. This molecule, therefore, seems to be immunogenic and to favor a humoral response. At the same time, other authors have demonstrated that mesothelin comprises several epitopes capable of activating cytotoxic T cells inducing the lysis of tumor cells. Given the very limited expression of mesothelin in normal tissues and its intense expression in tumor tissues, it is then possible to envisage its use, as a tumor-targeting molecule aiming to specifically deliver into the tumor a cytotoxic molecule or to use it as an antigen for immunotherapy. Thus, Hassan et al. have shown that the tissue distribution of the anti-mesothelin K1 antibody occurs specifically in tumors expressing this molecule at the membrane level and that tissue levels are higher than blood levels, even 7 days after the injection of the antibody. This team later built a fusion protein containing the variable part of the K1 antibody and an immunotoxin produced by Pseudomonas aeruginosa, a protein called SS1P. In vitro, this construct showed a potent cytotoxic action against tumor cell lines or against cells from the ascites of patients with peritoneal mesothelioma and expressing mesothelin. In vivo, the injection of SS1P into experimental models of lung, ovarian or mesothelioma cancers in mice made it possible to obtain a specific antitumor action on the primary tumor and on lung metastases. In addition, the combination of radiotherapy or chemotherapy with this immunotoxin has revealed a synergistic anti-tumor action. After promising results in monkeys, two clinical studies on the safety and tolerance of this treatment are underway in patients with cancer expressing this molecule. Intermediate results suggested good tolerance and a partial antitumor response in some patients. Promising preliminary data from a phase I clinical trial testing monoclonal anti-mesothelin antibodies (MORab-009) in the treatment of pancreatic, ovarian, and MPM cancers were recently presented at the ASCO 2008 conference by the team of R. Hassan (Baltimore, USA); phase II clinical trials of these antibodies associated with chemotherapy are being prepared for these same cancers.
In conclusion, unlike the old markers tested, in particular, hyaluronic acid and Cyfra 21.1, the new molecules like mesothelin, MPF and osteopontin, evaluated in the diagnosis of MPM, seem to have a higher sensitivity. However, soluble mesothelin has difficulty detecting non-epithelioid forms of MPM and osteopontin has low specificity. This is why their potential use in the diagnosis of MPM still requires validation by additional studies, in particular with regard to the comparison of these markers or their joint use.
Thus, accompanying real progress in the management of MPM obtained in recent years, biology could help in the diagnosis of this neoplasia, but also in the evaluation of the prognosis of patients and in the treatment of the disease through new therapeutic targets and tumor markers predictive of response to treatment. However, it must always be emphasized that this progress cannot be achieved without multidisciplinary management of both diagnostic and therapeutic MPM.
[1] Department of Pulmonology and Thoracic Oncology, Calmette Hospital, Lille University Hospital, Lille, France.
[2] Faculty of Medicine Henri-Warembourg, University of Lille II, Lille, France.
[3] INSERM 774 unit, Institut Pasteur de Lille, Lille, France.
[4] University of Medicine and Pharmacy, Department of Pulmonology, Iasi, Romania.
[5] Department of Respiratory Diseases, Thoracic Oncology Unit, Sainte-Marguerite Hospital, Marseille, France.
[6] EMI Inserm 0359, Laboratory of Experimental Cancerology, University of La Méditerranée, Marseille, France.
[7] Department of Pulmonology and Thoracic Oncology, Calmette Hospital, Lille University Hospital, 59037 Lille cedex, France.
summary
Malignant pleural mesothelioma (MPM) is cancer worrying by its increasing incidence and its poor prognosis, despite real progress in the management of patients. The diagnosis is generally made at an advanced stage when radical treatment is no longer possible. A significant increase in survival in MPM requires, in particular, an earlier diagnosis of the disease. Biologically, several potential soluble tumor markers, including mesothelin and osteopontin, have been proposed to aid in the diagnosis, but none have been validated to date. Soluble mesothelin, measured in blood and pleural fluid, seems to be the most interesting candidate. But even if this marker has a good diagnostic and prognostic value, its high specificity for the epithelioid subtype of mesothelioma (the most frequent) limits its usefulness in practice. The other molecules are not very interesting, despite sometimes a good sensitivity, in particular, because of their low diagnostic specificity. In conclusion, the available data do not yet argue for routine use of biology in the diagnosis of MPM, thoracoscopy and histology remaining the references. Soluble mesothelin should continue to be evaluated in clinical studies and look for other tumor markers.
Introduction
Malignant pleural mesothelioma (MPM) is a primary tumor of the pleura, very aggressive and with an overall unfavorable prognosis with a median survival of about one year [1]. The main etiological factor is prior exposure to asbestos, with a typically very long latency period between this exposure and the onset of MPM, in the order of 20 to 40 years [2], [3]. Formerly considered a rare tumor, the incidence of mesothelioma is gradually increasing [2], in parallel with the increasing use of asbestos fibers which, in France, reached its peak in the 70s and 80s (especially in industry and the building). The use of asbestos has been banned in France since 1997. In France, there are 800 to 1,000 cases of MPM per year with an incidence of around 16 cases / million inhabitants/year [4]. Estimates predict a peak incidence of MPM around 2020 with 1140-1300 cases/year [5]. In addition, asbestos is still widely used in emerging countries, suggesting a sustainable increase in the number of mesothelioma cases worldwide.
To date, there is no validated curative treatment for MPM. The majority of patients typically benefit from chemotherapy and/or supportive care. Despite the recent appearance of new chemotherapy regimens including the combination of cisplatin and an antimetabolite [6], the increase in survival remains modest compared to previous protocols. Some authors propose a radical surgical treatment by pleurectomy at a very early stage of the disease (limited involvement of the parietal pleura, IMIG stage Ia) which could improve the survival of these patients (survival in the surgical series from 15% to 5 years) [7]. But these patients currently represent only 1 to 2% of the total, which has so far failed to validate this attitude. Multimodal treatment for MPM, including chemotherapy, broad excision surgery (enlarged pleuropneumonectomy, PEP) and hemithoracic radiotherapy, could also provide benefit at later stages of MPM, but resectable, subject to a very strict selection of candidates [8]. However, if the feasibility of this multimodal approach has been demonstrated [9], there is currently no large randomized clinical study that has validated the interest of this heavy surgery (PEP) or a multimodal treatment for MPM.
Clinical diagnosis and imaging
The clinical and radiological signs of MPM are unfortunately not very specific and often appear at an advanced stage of the disease [7]. Exposure, most often occupational, to asbestos is frequently identified (around 80% of cases in men versus 50% in women) and must always be carefully investigated [3]. Superior to the chest radiograph, multi-bar helical computed tomography (CT) can help in the diagnosis and evaluation of the stage of MPM (grade of recommendation A according to the SPLF expert conference). The fixation of 18FDG during a positron emission tomography (PET) estimated by the "Standard Uptake Value" (SUV) is significantly increased in the case of malignant mesothelioma. A sensitivity of 91% and a specificity of 100% were found to differentiate MPM from benign pleural lesions induced by exposure to asbestos which do not fix 18FDG [7]. But for the diagnosis of MPM, it is recommended not to systematically perform a PET scan (expert opinion) [7]. Thus, despite advances in imaging (CT ± PET…), the majority of MPMs are diagnosed too late for radical treatment.
Pathology diagnosis
Analysis of pleural fluid, conventionally of the exudative type, is not very helpful in diagnosing MPM but can help to eliminate a differential diagnosis. The cytology of the pleural fluid is useful for the diagnosis of cancer but rarely makes the difference between mesothelioma and pleural adenocarcinoma. The diagnosis of MPM is made in only 20-30% of cases by cytology and 20-23% of cases by percutaneous pleural biopsy which does not usually bring enough material to confirm this diagnosis. The combination of the two techniques gives a yield of 35 to 40% [7]. Pleuroscopy or medical thoracoscopy, a simple examination that can be performed in a large majority of patients, is the most sensitive means (> 95%) for obtaining the diagnosis of MPM [7]. It is therefore recommended not to rely on the analysis of pleural fluid for the diagnosis of MPM, to reserve transparietal biopsies with or without detection by CT or ultrasound for patients for whom thoracoscopy cannot be considered, and to always perform thoracoscopy unless there are contraindications related to the patient's condition or a technical impossibility (pleural symphysis) (grade A recommendation) [7]. Diagnosing MPM based on the pathological examination of pleural biopsies can be difficult. For this reason, it is recommended not to diagnose MPM on an extemporaneous examination (A), but to always base it on a morphological examination supplemented with an immunohistochemical analysis (A) [7].
Biological diagnosis
The search for specific soluble tumor markers is therefore a major challenge in helping to diagnose, but also to assess the prognosis and response to treatment of patients with MPM. To date, however, there is no established marker for the positive diagnosis of MPM among the candidates considered below [7].
Hyaluronic acid
Hyaluronic acid (HA) has been proposed as a diagnostic marker, but despite good specificity, the sensitivity of its dosage remains modest (between 40 and 70%) both in serum (increase in its level only at an advanced stage of disease) than in pleural fluid [10]. Elevated serum HA values can also be encountered in rheumatoid arthritis and Wilms' tumor as well as in liver disease with fibrosis. In addition, a significant percentage of mesotheliomas do not secrete HA. However, elevated pleural HA levels (> 100 mg / L) are considered specific for the diagnosis of MPM.
ACE and Cyfra 21-1
The other markers studied to date are the carcinoembryonic antigen (CEA), an oncofetal glycoprotein, and the cytokeratin fragment (Cyfra) 21-1. A CEA level> 3 ng / ml in the pleural fluid would exclude MPM [11]. The serum level of Cyfra 21-1 is also high in metastatic pleurisy, but its level measured in pleural fluid is significantly higher in MPM than in other malignant pleurisy [12]. However, elevated levels of this marker can also be measured in patients with lung, breast or pancreatic cancer. Finally, the Cyfra 21-1 dosage would have a prognostic value in the MPM. The CA 125, CA19-9 and CA 15-3 assays have not been shown to be useful in diagnosing MPM. These markers (CA125 and CA19-9) are indeed secreted by normal mesothelial cells, but their rate is increased in inflammatory states, greatly reducing their diagnostic specificity [13]. More recently, soluble mesothelin (also called soluble mesothelin-related peptides, SMRP) [14], osteopontin [15] and the megakaryocyte potentiation factor (MPF) [16] have been suggested as potential tumor markers of MPM.
Soluble mesothelin and the megakaryocyte potentiation factor (MPF)
Mesothelin is a glycoprotein expressed on the surface of normal mesothelial cells, but also strongly in malignant mesothelioma and other neoplasias. It is initially produced in the form of a 69 kDa precursor linked to the membrane by a GPI anchor [17]. This precursor is then cleaved by furin to generate a 40 kDa fragment which remains bound to the membrane (mesothelin) and another soluble fragment of approximately 31 kDa called megakaryocyte potentiation factor (MPF) (Figure 1). Little is known about the biological activity of these molecules. MPF is said to stimulate the growth of megakaryocytes. Soluble mesothelin could be involved in cell adhesion by interaction with the CA125 antigen and thus in the process of peritoneal dissemination of ovarian carcinomatous cells [18].
For the diagnosis of MPM, two uses of mesothelin and MPF were considered: on the one hand, as markers in immunohistochemistry (for the membrane form of mesothelin) and, on the other hand, as soluble markers. The low specificity of membrane mesothelin makes it a poor immunohistochemical marker for distinguishing MPM from different carcinomas. Indeed, it is constantly expressed by normal mesothelial cells. In addition, if mesothelin is expressed by a large majority of epithelioid type MPM (but 20 to 30% of cases are negative), sarcomatoid subtypes are systematically negative by immunostaining [19]. Mixed forms express this marker in a variable percentage (20-30%). Finally, membrane mesothelin is also expressed in almost 40% of adenocarcinomas and 30% of pulmonary squamous cell carcinomas. In ovarian (> 95% positive) and pancreatic cancers, the expression of mesothelin is frequently focal and localized at the cytoplasmic level as opposed to membrane expression by the majority of cells in the epithelioid MPM.
By ELISA assay using a pair of anti-mesothelin monoclonal antibodies, a soluble form of mesothelin has been identified [20]. It is not known exactly how this soluble protein is synthesized and secreted, but two hypotheses are currently plausible. It could be either a mutated form of mesothelin which thus loses its binding capacity to the membrane or a cleavage of the membrane mesothelin by an unknown enzyme. For the moment, the available data suggest that the second hypothesis is the most plausible [21]. It would, therefore, be preferable to date to use the term soluble mesothelin for the circulating form of the molecule, rather than the designation of SMRP initially proposed. To date, there are two pairs of antibodies that can be used for the ELISA assay of mesothelin: that developed by Scholler et al. [20] (MesomarktTM commercial kit - Fujirebio Diagnostics, USA or CisBio International, France) and that developed by Hassan et al. [22].
The blood test for soluble mesothelin has shown elevated levels in patients with MPM or an ovarian tumor. Serum mesothelin values are higher in ovarian cancer than in benign ovarian tumors and appear to be correlated with disease spread and patient prognosis [12]. The first results published by Robinson et al. [14], based on the technique developed by Scholler [20], suggested that this marker would have good diagnostic sensitivity (> 80%) and very good specificity (> 95%) for epithelioid MPM. In this study, there was no increase in the serum mesothelin level for sarcomatoid MPM, thus confirming the data in immunohistochemistry, as well as in patients with inflammatory pulmonary disease, pleural neoplastic disease outside of MPM or transudative pleurisy. (figure 2). An increase in the serum mesothelin level was observed in 17.5% of healthy subjects exposed to asbestos (n = 7), 3 of whom secondarily developed a BPM, suggesting the possibility of using this marker for screening for BPM [14 ]. The level of serum mesothelin was correlated with tumor size and there was a tendency for this level to gradually increase during the course of the disease [14]. On the other hand, chemotherapy did not significantly influence this rate, while partial surgical resection would make it possible to obtain a decrease in its value [14]. The results of two unpublished studies, one from our team (Grigoriu et al., ERS communication 2007, manuscript in preparation), the other from the American H. Pass team in patients who have undergone chemotherapy and/or surgical excision seem to confirm this hypothesis. Based on these preliminary American results, the commercial kit MesomarktTM however obtained in January 2007 from the Food and Drug Administration (FDA) in the United States the approval of use on a compassionate basis as a follow-up marker for patients with a BPM. In another study including more patients including 60 MPM, we confirmed the diagnostic utility of soluble mesothelin measured in the blood and pleural fluid of patients with MPM compared to those with benign pleural lesions associated with exposure to asbestos or pleural metastases [23]. A threshold value of serum mesothelin of 0.93 nM / l obtained a sensitivity of 80% and a specificity of 83% of the diagnostic test. There were, however, a significant number of pleural metastases from adenocarcinoma also with elevated levels of mesothelin. The pleural dosage of mesothelin did not provide any additional information compared to that of the serum dosage [23], a result discussed by other authors [12].
By extension of this study on a population of 96 MPM, we have also highlighted a prognostic role for serum mesothelin at the threshold of 3.5 nM / l. Patients with a high level of mesothelin had a median survival of 7 months compared to 19 months in the group with a low marker level [24]. Using another pair of anti-mesothelin antibodies, Hassan et al. have confirmed the diagnostic utility of mesothelin in epithelioid-type mesotheliomas and the existence of high rates also in patients with ovarian carcinoma [22].
Like soluble mesothelin, elevated serum MPF levels have been found in patients with ovarian cancer or epithelioid-type mesotheliomas compared to healthy controls and patients with pleural metastases from pulmonary adenocarcinoma or pleural lesions benign [16]. These preliminary data suggest a very good specificity (virtually 100%) with a sensitivity of around 90% of MPF but need to be confirmed by studies including a larger number of patients with MPM or various pathologies (pleural metastases different carcinomas, benign asbestos pleurisy, etc.).
Due to its lack of specificity, the usefulness of mesothelin immunohistochemistry labeling is limited to a few diagnostic "niches". The determination of serum and/or pleural mesothelin appears useful for the diagnosis of MPM, subject to an average sensitivity in the studies (approximately 70% for a specificity of 90%). An "aggressive" exploration strategy for patients with a history of asbestos exposure, pleural abnormalities, and high serum mesothelin levels could be discussed, but its practical utility remains to be validated. Likewise, larger studies are needed to define the use of serum mesothelin as a prognostic and predictive marker in MPM. On the other hand, the use of this marker for screening for MPM in patients exposed to asbestos seems very improbable, if not impossible, due to the low incidence of MPM compared to a large number of patients exposed, its insufficient specificity ( <50% for sensitivity> 90%) and the absence of curative treatment for MPM to date. Studies on this subject are however in progress in Europe as in the USA. Data on MPF are still very limited to suggest any practical use, but the available data suggest a better diagnostic value than for mesothelin [25].
osteopontin
Osteopontin is a pleiotropic molecule, initially described in the bone where its potential role would be to make the link between bone cells and the extracellular matrix. It is also synthesized by malignant cells and acts as a key cytokine in the granulomatous reaction. Osteopontin has also been implicated in tumor progression. Elevated serum levels have been reported in ovarian, colic, mammary, prostate or lung cancers. However, osteopontin is also expressed in benign pleural and extrapleural pathologies (pulmonary tuberculosis, etc.) [12]. Several commercial assays are available (Immunobiological Laboratories (IBL), R & D Systems, Assay design), but comparative studies show that there are significant differences in results between the different kits. In 2005, Pass et al. have, for the first time, reported that osteopontin is expressed in a vast majority (95%) of MPM, regardless of the histological subtype [15]. Serum osteopontin levels were higher in patients with BPM compared to subjects exposed to asbestos with benign pleural and/or pulmonary lesions. In this study, there was no relationship between the tumor stage (IMIG classification) and the serum osteopontin level; as a result, the diagnostic utility of osteopontin remained very high, even in the early stages of the disease (stage I). In patients exposed to asbestos, there was a relationship between the serum osteopontin level and the duration of exposure, but also with the type and intensity of the radiological abnormalities present in these patients (pleural plaques, pleural thickening, pulmonary fibrosis). However, this study did not include patients with other types of tumors (pleural or not) or other non-tumor pleural pathologies. We compared the usefulness of osteopontin and soluble mesothelin in a series of 172 patients with MPM and 112 asymptomatic subjects with occupational exposure to asbestos [24]. This study showed that serum osteopontin values are higher in patients with MPM compared to patients exposed to asbestos, thus confirming the data published by Pass et al. [15]. However, this marker was less useful than mesothelin, with a smaller area under the ROC curve. We also found high values of osteopontin in patients with pleural metastases from carcinoma or benign pleural lesions secondary to asbestos exposure. This same observation is made for the determination of osteopontin in pleural fluid. Despite very good sensitivity, the dosage of osteopontin, therefore, has low specificity for MPM, which makes it difficult to use as a diagnostic marker or screening for MPM in patients highly exposed to asbestos [24].
Biological perspectives in MPM
Diagnostic perspectives
Creaney et al. recently reported a negative result from the evaluation of urinary mesothelin for the diagnosis or screening of MPM (abstract ATS 2008, Toronto - Canada). Given the limited diagnostic capacities of known markers, the question arises of the advantage of combining markers within a panel in order to improve diagnostic performance compared to a marker evaluated in isolation, in this case, the serum or pleural mesothelin. In our experience, the combination of osteopontin with serum mesothelin does not provide any advantages over the measurement of mesothelin alone in the diagnosis of MPM [24]. Other authors have shown that the combination of mesothelin with other potential markers such as Cyfra 21.1 or CA 125 was also not useful in the diagnosis of MPM [11], [12]. We also evaluated the combination of hyaluronic acid (HA) and mesothelin measured in the blood or pleural fluid (manuscript submitted for publication). Adding the dosage of HA to that of mesothelin in the blood could guide the diagnosis in about 15% of cases of MPM when the dosage of mesothelin was negative. The practical interest of this element seems however weak in patients generally receiving a thoracoscopy, histology being to date the key element of the diagnosis of MPM. Interestingly, this additional interest in HA was not limited to patients with epithelioid-type MPM. The evaluation of circulating HA could, therefore, be considered in patients strongly suspected of MPM, but without histological evidence of cancer or radical treatment envisaged and with a low serum mesothelin level. The levels of the two markers (HA and mesothelin) were highly correlated in the pleural fluid, making the HA dosage redundant in practice.
In the current state of knowledge and subject to its sensitivity limited by its specificity to the epithelioid subtype, serum mesothelin is the potential tumor marker of MPM with the best diagnostic value and can, therefore, be considered as the reference marker to which all the other candidates will have to be compared. Panels including soluble mesothelin and other potential diagnostic markers for MPM (survivin, etc.) measured in the pleural fluid are being evaluated. Finally, it should be remembered that no dosage of blood or pleural biological marker, including that of mesothelin, has recognized legal diagnostic value in patients suspected of MPM without histological evidence in order to obtain possible compensation for their pathology under occupational diseases (table n ° 30 of the RG) or with the compensation fund for asbestos victims (FIVA).
Prognostic and predictive markers
There are no significant published data on the usefulness of any marker candidate as a prognostic element or for the evaluation of the response to treatment in MPM. Pleural HA and serum mesothelin [12], [24] have already been studied as prognostic markers, but the available data are still insufficient to draw a clear conclusion and even less to make a recommendation for practice. However, the use of the mesothelin blood test has been validated by the US Food and Drug Administration (FDA) for this latter indication. The analysis of data obtained jointly at the Lille University Hospital and the University of Pennsylvania (USA) supports the interest of monitoring serum mesothelin as a predictive marker of the therapeutic response of MPM (manuscript in preparation).
Therapeutic prospects [12]
Recently, it has been shown that 40% of patients with MPM or ovarian carcinoma have antibodies to mesothelin. This molecule, therefore, seems to be immunogenic and to favor a humoral response. At the same time, other authors have demonstrated that mesothelin comprises several epitopes capable of activating cytotoxic T cells inducing the lysis of tumor cells. Given the very limited expression of mesothelin in normal tissues and its intense expression in tumor tissues, it is then possible to envisage its use, as a tumor-targeting molecule aiming to specifically deliver into the tumor a cytotoxic molecule or to use it as an antigen for immunotherapy. Thus, Hassan et al. have shown that the tissue distribution of the anti-mesothelin K1 antibody occurs specifically in tumors expressing this molecule at the membrane level and that tissue levels are higher than blood levels, even 7 days after the injection of the antibody. This team later built a fusion protein containing the variable part of the K1 antibody and an immunotoxin produced by Pseudomonas aeruginosa, a protein called SS1P. In vitro, this construct showed a potent cytotoxic action against tumor cell lines or against cells from the ascites of patients with peritoneal mesothelioma and expressing mesothelin. In vivo, the injection of SS1P into experimental models of lung, ovarian or mesothelioma cancers in mice made it possible to obtain a specific antitumor action on the primary tumor and on lung metastases. In addition, the combination of radiotherapy or chemotherapy with this immunotoxin has revealed a synergistic anti-tumor action. After promising results in monkeys, two clinical studies on the safety and tolerance of this treatment are underway in patients with cancer expressing this molecule. Intermediate results suggested good tolerance and a partial antitumor response in some patients. Promising preliminary data from a phase I clinical trial testing monoclonal anti-mesothelin antibodies (MORab-009) in the treatment of pancreatic, ovarian, and MPM cancers were recently presented at the ASCO 2008 conference by the team of R. Hassan (Baltimore, USA); phase II clinical trials of these antibodies associated with chemotherapy are being prepared for these same cancers.
In conclusion, unlike the old markers tested, in particular, hyaluronic acid and Cyfra 21.1, the new molecules like mesothelin, MPF and osteopontin, evaluated in the diagnosis of MPM, seem to have a higher sensitivity. However, soluble mesothelin has difficulty detecting non-epithelioid forms of MPM and osteopontin has low specificity. This is why their potential use in the diagnosis of MPM still requires validation by additional studies, in particular with regard to the comparison of these markers or their joint use.
Thus, accompanying real progress in the management of MPM obtained in recent years, biology could help in the diagnosis of this neoplasia, but also in the evaluation of the prognosis of patients and in the treatment of the disease through new therapeutic targets and tumor markers predictive of response to treatment. However, it must always be emphasized that this progress cannot be achieved without multidisciplinary management of both diagnostic and therapeutic MPM.
Enregistrer un commentaire
Enregistrer un commentaire