Malignant Pleural Mesothelioma: Factors Influencing the Prognosis
Malignant Pleural Mesothelioma: Factors Influencing the Prognosis
ABSTRACT: Malignant pleural mesothelioma (MPM) is a highly severe primary tumor of the pleura mainly related to exposure to asbestos fibers. The median survival after symptom onset is less than 12 months. Conventional medical and surgical therapies—either as single lines or combined—are not wholly effective. No universally accepted guidelines have yet been established for patient selection and the use of therapeutic strategies. In addition, retrospective staging systems have proved inadequate at improving therapeutic outcomes. Therapy is currently guided by gross tumor characteristics and patient features; however, these seem less accurate than the biological fingerprint of the tumor. A number of clinical prognostic factors have been considered in large multicenter series and independently validated. A series of novel biomarkers can predict the evolution of the disease. Here we summarize the principal and novel factors that influence prognosis and are thus potentially useful for selecting patients for targeted therapy.
Among the primary neoplasms of the chest, malignant pleural mesothelioma (MPM) has benefited less than others from the multidisciplinary advances of the past few decades. MPM is highly aggressive and still has a very poor prognosis. The majority of patients are diagnosed in advanced stages; thus, they are not candidates for surgical care and die within 10 to 17 months of first symptom onset, with a median survival of less than 12 months.
Because of the difficulty in diagnosing and staging the tumor, especially in its initial clinical manifestations, management remains challenging. The unsatisfactory results obtained so far by surgery and multimodality combinations of conventional oncologic treatments have awakened interest in new and intriguing aspects of the disease.
Great efforts have recently been made to improve our knowledge of the molecular pathways and genetic changes involved in the disease, in order to enable researchers to paint a more accurate picture of the tumor—ideally, a true “fingerprint” of the neoplasm, with a high probability of predicting prognosis and facilitating the individualization of targeted therapies.
Here we examine recent advances in the knowledge of this severe and heterogeneous malignancy, and we analyze the clinical significance of prognostic factors.
The clinical factors relevant to management of MPM include basic epidemiologic variables, clinical condition, common blood assays, imaging assessment, and gross tumor features, as well as the anatomic extent of the disease. Many of these variables have been combined into prognostic categories to reinforce their predictive power.
Gender and age
Male gender is conventionally considered to be a predictor of poorer prognosis. Men are more likely to have had occupational exposure and to have had greater asbestos exposure, which results in more aggressive disease.
Age is a potential prognostic factor, but its precise impact is difficult to define. Nonetheless, the onset of MPM in a young person is generally considered the result of an elevated asbestos exposure, thus implying a more aggressive disease. Older age is therefore considered a good predictor of longer survival, with an arbitrary cut-off that varies from 50 to 75 years.[2,3]
Asbestos is the principal etiologic agent of MPM, leading to disease within a period of 10 to 30 years of exposure.[1,4] Although exposure to asbestos fibers should be considered as a risk factor more than a prognosticator, it has been suggested that longer duration and higher intensity of exposure to asbestos can lead to the development of more aggressive tumors.
Severity of symptoms is generally related to extent of disease. Dyspnea and chest pain are simple but reliable clues to lung and chest wall invasion, related to a worse prognosis.[5,6] Loss of appetite and weight, likely resulting from tumor production of cachexia-inducing cytokines, are included as negative factors in the European Organisation for Research and Treatment of Cancer (EORTC) prognostic score.
Performance status and quality of life
Performance status is usually one of the strongest prognosticators of good surgical outcome. It is included in the two main prognostic scoring systems, the Cancer and Leukemia Group B (CALGB) and EORTC systems. Performance status has been measured by a variety of scales: that of the Eastern Cooperative Oncology Group (ECOG), that of the World Health Organization, and the Karnofsky index; the last of these is the most frequently used indicator of prognosis.
The introduction of scientific quantification of quality of life allowed a more precise evaluation of the influence quality of life has on prognosis. Scales of general health, such as the SF-36 Health Survey or the EORTC questionnaires, have been widely used to quantitate quality of life. Significant improvement in the physical component summary of the SF-36 Survey after extrapleural pneumonectomy has been associated with longer survival.
Computed tomography (CT) scanning
The use of CT scanning as a prognostic tool has recently emerged as a result of the development of software that allows the quantification of tumor volume. The Prospective Mesothelioma Staging Project, as part of their efforts to establish the bases of a new staging system, took into account CT evidence, using as a reference cutoff an estimated tumor volume of 500 cm3.
Positron emission tomography (PET) scanning
An elevated standardized uptake value (SUV) on PET scanning is significantly associated with shorter survival, with a threshold value ranging from 4 to 10. [11,12] Furthermore, a correlation between a decrease in SUV after midterm chemotherapy and slower disease progression has been described.
Three histological subtypes of MPM have been defined. Definitive diagnosis requires a panel of stains. The epithelioid subtype accounts for 50% of cases, whereas the mixed and sarcomatoid types represent 34% and 16%, respectively. A diagnosis of epithelioid subtype has the most clinical relevance. It influences the choice of therapeutic strategy, typically resulting in more aggressive surgical and medical procedures. The presence of an inflammatory stromal response demonstrated an association with improved survival. Nonepithelioid subtypes have a worse prognosis and may be related to a higher amount and longer duration of exposure to asbestos. The sarcomatoid subtype categorically excludes patients from surgery, primarily because of the historic short postoperative survival in this patient subgroup.
The EORTC score includes an increase in the white blood cell count as a negative prognostic factor, due to the production of granulocyte colony-stimulating factors associated with this finding; the threshold used is 8.3 × 109/L. A poor prognosis is also seen in patients with a high blood neutrophil−to-lymphocyte ratio.
Increased platelet aggregation was included as a negative prognosticator in both the CALGB and EORTC scoring systems, with cut-off values of 400 × 109/L and 350 × 109/L, respectively. It is theorized that platelet-derived growth factor (PDGF) and megakaryocyte potentiating factor (MPF) have roles in this phenomenon.
Low hemoglobin level is another variable proposed as a negative predictor. The cut-off value is indicated as 1 g/dL below the sex baseline (ie, 16 g/dL in males and 14 g/dL in females). Anemia (hemoglobin level below 10 g/dL) was recently shown to be a strong negative prognosticator.
A shift of the metabolism towards aerobic glycolysis (resulting in the production of lactate) is characteristic of the most aggressive tumors. As a consequence, an elevated serum lactate dehydrogenase (LDH) level has been correlated with poor prognosis. The CALGB score includes high levels of LDH as an indicator of poor prognosis, using as a threshold a level of 500 IU/L.
Staging describes the anatomical extent of the neoplasm. A variety of staging systems have been proposed.[18-20] To date, two retrospectively-based staging systems have been widely utilized: the Brigham and Women's Hospital system (Table 1), formulated according to the results of extrapleural pneumonectomy in trimodal therapy, and the TNM classification developed by the International Mesothelioma Interest Group (Figure 1). The majority of the studies using these two systems have assigned a prognostic value to the stage.[21-23] In the TNM classification, T status is not as strong a prognostic factor as N status. Mediastinal nodal involvement has been recognized as a critical component of staging, with a detrimental effect on survival.[19-23]
A recent review has proposed limiting use of the TNM classification just to MPM with epithelioid histology ; the authors also proposed adjustments to the T status classification criteria (Table 2). These new criteria improved the distribution of the stages and survival stratification: median survival values were 51, 26, 15, and 8 months in stages I, II, III, and IV, respectively.
A study of a larger group of patients from multiple centers enrolled in a prospective manner is presently ongoing by the International Association for the Study of Lung Cancer (IASLC).
Surgical radical resection
The completeness of surgical resection is an important prognostic factor. Classic extrapleural pneumonectomy with en bloc hemidiaphragm and pericardium removal and mediastinal lymph node sampling is generally regarded as the most radical procedure. Although associated with greater morbidity and perioperative mortality, it may result in longer survival, especially when combined with multimodality therapies. More recently, the National Comprehensive Cancer Network guidelines stated that radical pleurectomy/decortication represents the first surgical option for tumor not growing within the parenchyma. This lung-sparing procedure entails resection of involved pleura, pericardium, and diaphragm, along with mediastinal lymph node sampling, and provides good symptomatic relief.
To date, there has been a general consensus that these two operations are not comparable because they target different patients with different degrees of neoplastic local extension. However, we can assume that any surgery able to achieve radical eradication of all gross tumor is likely to have a survival benefit. Patients suitable for surgical resection should be at earlier stages, with good preoperative performance status and few comorbidities; such patients form a group that by definition has a better prognosis. The removal of all microscopic tumor is difficult to achieve. The presence of occult disease can be demonstrated by immunohistochemistry and has been significantly associated with rapid macroscopic recurrence and a worse prognosis.
Complete eradication of microscopic disease has been shown to improve survival. National Comprehensive Cancer Network  and European Society of Medical Oncology  guidelines indicate that multimodality treatment carried out by an experienced multidisciplinary team can achieve better outcomes. Patients who received complete adjuvant therapy demonstrated a longer survival compared with those who underwent surgery alone. Chemotherapy and radiotherapy alone have very poor results, underscoring the pivotal role of surgery. However, when surgical resection of all gross tumor is feasible, adjuvant and/or neoadjuvant therapies must always be planned. The use of intraoperative hyperthermic intrapleural chemotherapy has been proposed as well and has been shown to have a significant survival advantage.
The prognostic impact of including radiotherapy in the multimodality approach to MPM is currently under investigation. To improve local control after radical resection, it has been shown that both 3D conformal and intensity-modulated radiotherapy can be delivered at doses of > 45 Gy.
It has been documented that use of cisplatin as single-agent chemotherapy has unsatisfactory results. Improvement in survival, along with a benefit in symptom control, has been achieved when platinum derivatives have been combined with new antifolate drugs such as pemetrexed (Alimta) or raltitrexed. Chemosensitivity varies widely among individuals, and this may represent an important prognostic factor. High expression of thymidylate synthase predicts a poor response to pemetrexed, and it has been documented as a prognosticator of shorter survival. Excision repair cross-complementation group 1 (ERCC1) is a protein involved in DNA nucleotide repair. Low expression of ERCC1 might predict increased sensitivity to platinum-based chemotherapy, thus implying a longer survival.
Prognostic scoring systems
The difficulty in finding a single prognostic predictor has favored the development of clinical scoring systems. The CALGB system, the EORTC system, and the modified version of the latter are summarized in Table 3. These systems identify low- and high-risk subgroups with different survival rates. A new combined prognostic system is currently being developed by IASLC using a database of more than 3000 surgically managed patients. It considers an initial group of prognostic factors called “core” factors, together with supplementary prognostic variables (Table 3). Apart from staging, the IASLC found that the most significant variables on multivariate analysis were histology, sex, age, white blood cell count, and platelet count.
Serum biomarkers are used primarily for early diagnosis, but they should also be used to monitor the evolution of the disease and to predict prognosis. Serum mesothelin is a three-form (variants 1, 2, and 3) differentiation antigen in mesothelial cells that is highly expressed in MPM. Abnormal serum levels are associated with a large tumor volume and have proved effective as a prognostic indicator. MPF, mentioned above, is a soluble protein produced by proteolytic cleavage of the amyloid precursor protein mesothelin. It is secreted by MPM cells and has recently been shown to have a negative prognostic value. Osteopontin is a glycoprotein involved in the adhesion of cells to the bone matrix. Serum levels of osteopontin are higher in MPM and are related to the duration of exposure to asbestos. Low serum levels of osteopontin are associated with a better overall survival. More diffuse oncomarkers, such as soluble cytokeratin fragments (CYFRA) 21-1 and tissue polypeptide antigen (TPA) have been shown to have only a limited diagnostic or prognostic role.