Article Text

Download PDFPDF

British Thoracic Society Guideline for the investigation and management of malignant pleural mesothelioma
  1. Ian Woolhouse1,
  2. Lesley Bishop2,
  3. Liz Darlison3,
  4. Duneesha De Fonseka4,
  5. Anthony Edey5,
  6. John Edwards6,
  7. Corinne Faivre-Finn7,
  8. Dean A Fennell8,
  9. Steve Holmes9,
  10. Keith M Kerr10,
  11. Apostolos Nakas11,
  12. Tim Peel12,
  13. Najib M Rahman13,
  14. Mark Slade14,
  15. Jeremy Steele15,
  16. Selina Tsim16,
  17. Nick A Maskell17
  1. 1Department of Respiratory Medicine, University Hospitals Birmingham, NHS Foundation Trust, Birmingham, UK
  2. 2Respiratory, Queen Alexandra Hospital, Portsmouth, UK
  3. 3Respiratory Medicine, University Hospitals of Leicester, Leicester, UK
  4. 4Academic Respiratory Unit, North Bristol NHS Trust, Bristol, UK
  5. 5North Bristol NHS Trust, Bristol, UK
  6. 6Sheffield Teaching Hospitals, Sheffield, UK
  7. 7Division of Cancer Services, University of Manchester, Manchester, UK
  8. 8University of Leicester & University Hospitals of Leicester, Leicester, UK
  9. 9The Park Medical Practice, Shepton Mallet, Somerset, UK
  10. 10University of Aberdeen, Pathology, Aberdeen, UK
  11. 11Department of Thoracic Surgery, Glenfield Hospital, Leicester, UK
  12. 12North Tyneside General Hospital, North Shields, UK
  13. 13Oxford NIHR Biomedical Research, University of Oxford, Oxford, UK
  14. 14Papworth Hospital, Thoracic Oncology, Cambridge, UK
  15. 15Cancer, St Bartholomew’s Hospital, London, UK
  16. 16Respiratory Medicine, Queen Elizabeth University Hospital, Glasgow, UK
  17. 17Academic Respiratory Unit, Bristol Medical School, University of Bristol, Bristol, UK
  1. Correspondence to Professor Nick A Maskell, Academic Respiratory Unit, Bristol Medical School, University of Bristol, BS10 5NB, UK; nick.maskell{at}bristol.ac.uk

Statistics from Altmetric.com

Summary of recommendations and good practice points

Section 3: Clinical features which predict the presence of mesothelioma

Recommendations

  • Do not rule out a diagnosis of malignant pleural mesothelioma (MPM) on the basis of symptoms and examination findings alone. Grade D.

  • Offer an urgent chest X-ray to patients with symptoms and signs as outlined in the National Institute for Health and Care Excellence (NICE) NG12. Grade D.

  • Refer all patients with a chest X-ray suggestive of MPM urgently (via the 2-week wait suspected cancer pathway in England and Wales). Consider referral for further investigation in patients with persistent symptoms and history of asbestos exposure despite normal chest X-ray. Grade D.

  • A thorough occupational history should be taken to cover all occupations throughout life. It is important to elicit para exposure by exploring details of relative and/or partner occupations. Grade D.

Section 4: Staging systems

Recommendation

  • Record staging of MPM according to the version 8 of the International Association for the Study of Lung Cancer (IASLC) staging proposals. Grade D.

Section 5: Imaging modalities for diagnosing and staging

Recommendations

  • Offer CT thorax with contrast (optimised for pleural evaluation) as the initial cross-sectional imaging modality in the evaluation of patients with suspected MPM. Grade D.

  • Use of positron emission tomography (PET)-CT for aiding diagnosis of MPM is not recommended in patients who have had prior talc pleurodesis and caution should be employed in populations with a high prevalence of TB. Grade D.

  • In patients where differentiating T stage will change management, consider MRI. Grade D.

  • In patients where excluding distant metastases will change management, offer PET-CT. Grade D.

Section 6: Pathological diagnosis

Recommendations

  • Immunohistochemistry (IHC) is recommended for the differential diagnosis of MPM in both biopsy and cytology-type specimens. Grade D.

  • A combination of at least two positive mesothelial (calretinin, cytokeratin 5/6, Wilms tumour 1, D-240) and at least two negative adenocarcinoma immunohistochemical markers (TTF1, CEA, Ber-EP4) should be used in the differential diagnosis of MPM. (Markers listed in likely order of value). Grade D.

  • Do not rely on cytology alone to make a diagnosis of MPM unless biopsy is not possible or not required to determine treatment due to patient's wishes or poor performance status (PS). Grade D.

  • Pathologists should report the histological subtype of MPM in all cases. Grade D.

Good practice points

  • Biopsies from patients with suspected MPM should be reviewed by a pathologist experienced in the diagnosis of MPM and a second opinion should be sought if there is uncertainty over the diagnosis.

Section 7: Use of biomarkers

Recommendations

  • Do not offer biomarkers in isolation as a diagnostic test in MPM. Grade B.

  • Consider biomarker testing in patients with suspicious cytology who are not fit enough for more invasive diagnostic tests. Grade B.

  • Do not routinely offer biomarker testing to predict treatment response or survival. Grade B.

  • Do not offer biomarker testing to screen for MPM. Grade C.

Research recommendation

Further research is required to identify biomarkers that reliably predict treatment response within clinical practice.

Section 8: Factors determining prognosis and timing of treatment

Recommendations

  • Consider calculating a prognostic score in patients with MPM at diagnosis. Grade D.

  • Prognostic scores can provide useful survival information for patients and doctors, but should not be used in treatment decision-making. Grade D.

  • When calculating a prognostic score use one of the following:

    1. The European Organisation for the Research and Treatment of Cancer (EORTC) prognostic score;

    2. The CALGB score;

    3. The modified Glasgow prognostic score (mGPS);

    4. The LENT score if a pleural effusion is present;

    5. The decision tree analysis.

  • The decision tree analysis scoring systems is likely to be the most useful in routine clinical practice. Grade D.

Section 9: Pleural fluid management

Recommendations

  • Offer either talc (via slurry or poudrage) or indwelling pleural catheters for symptomatic pleural effusion in MPM, informed by patient choice. Grade A.

  • Talc slurry or thoracoscopic talc poudrage pleurodesis should be offered to patients with MPM in preference to a video- assisted thoracoscopic surgery partial pleurectomy (VATS-PP) surgical approach for fluid control in MPM. Grade A.

Section 10: The role of surgery

Recommendations

  • Do not offer VATS-PP over talc pleurodesis in MPM. Grade A.

  • Do not offer Extra-Pleural Pneumonectomy (EPP) in MPM. Grade B.

  • Do not offer extended pleurectomy decortication (EPD) outside of a clinical trial. Grade D.

Research recommendation

The role of VATS-PP and EPD in good prognosis patients should be examined further in clinical trials, which should include robust measurement of quality of life.

Section 11: Systemic anticancer treatment

Recommendations

  • Offer patients with MPM with good PS (WHO 0-1) first-line therapy with cisplatin and pemetrexed. Where licensed (not presently in the UK), bevacizumab should be added to this regime. Raltitrexed is an alternative to permetrexed. Grade A.

  • Do not offer pemetrexed or vorinostat as second-line treatment for patients with MPM. Grade A.

Good practice points

  • Where cisplatin is contraindicated, or has adverse risk, offer carboplatin in combination with pemetrexed.

  • First-line clinical trials are an appropriate option for patients with good PS and are recommended above any other option for second-line treatment, providing the patient is of adequate PS.

Research recommendations

The role of immunotherapy in MPM should be further assessed in large phase III randomised controlled trials (RCTs).

Further randomised controlled trials of second-line therapy on MPM are required.

Section 12: Radiotherapy

Recommendations

  • Do not offer prophylactic radiotherapy to chest wall procedure tracts. Grade A.

  • Do not offer preoperative or postoperative radiotherapy in MPM. Grade A.

  • Do not offer hemithorax radiotherapy for MPM. Grade D.

  • Consider palliative radiotherapy for localised pain in MPM, where the pain distribution matches areas of underlying disease. Grade D.

Research recommendations

Prospective clinical trials of preoperative radiotherapy, postoperative radiotherapy after pleurectomy decortication and definitive radiotherapy after chemotherapy in MPM are required.

Further prospective randomised clinical trials are required to determine the role of radiotherapy for symptom control in MPM and the optimal dose fractionation.

Section 13: Symptom control

Good practice point

  • Symptoms in MPM should be managed as per current guidelines for cancer in general (table 18) and early involvement of palliative care specialists is recommended.

Section 14: Care and management

Recommendations

  • Consider referring MPM cases to a regional mesothelioma MDT. Grade D.

  • Offer accurate and understandable information to patients and carers about compensation for MPM. Grade D.

  • Offer patients with MPM and their carers the opportunity to discuss concerns regarding their disease. Grade D.

  • In patients with MPM where accurate determination of radiological progression is required, consider CT with modified Response Evaluation Criteria In Solid (mRECIST) measurement. Grade D.

Good practice points

  • All mesothelioma cases should be discussed in a timely fashion by a MDT that reviews a sufficient number of cases to maintain expertise and competence in the diagnosis and treatment of MPM.

  • The MDT membership should fulfil the requirements set by national cancer peer review (to include a named clinical nurse specialist for MPM).

  • The MDT should maintain an up-to-date portfolio of mesothelioma trials and offer recruitment to all eligible patients.

  • A personalised care approach should be considered for each patient.

Patients should be offered 3–4 monthly follow-up appointments with an oncologist, respiratory physician or specialist nurse according to their current treatment plan. If patients wish to be seen less frequently, offer regular telephone follow-up with specialist nurse with an option to attend clinic in the event of clinical deterioration.

Section 1: Introduction

Aim of the guideline

The key aim of this guideline is to provide detailed, evidence-based guidance for the investigation of suspected MPM and the subsequent care and management of individuals with proven MPM. MPM is a rare cancer where the malignancy affects the pleura, a thin membrane of lubricating cells that lines the lungs and chest wall. The focus of this guideline is MPM as it is far more common than mesothelioma occurring in the abdomen. The 2016 Mesothelioma Audit data reported that in the UK in 2014 pleural mesothelioma accounted for 2179 cases (97%), with 70 peritoneal cases (approximately 3%).1

In 2007, the British Thoracic Society (BTS) statement on mesothelioma was published in response to a request from the National Health Executive in England.2 The BTS has reviewed this statement and is of the opinion that the publication is no longer fit for purpose as an up-to-date reference guide for healthcare professionals. The 2007 statement did not attempt to provide a comprehensive review of all relevant published literature and since the publication of the statement the BTS has achieved NICE accreditation for its guideline production process. The Standards of Care Committee (SOCC) of the BTS established a guideline development working group in 2014, chaired by Professor Nick Maskell and Dr Ian Woolhouse.

The main cause of mesothelioma is breathing in asbestos dust-approximately 85% of all mesotheliomas in males are attributable to occupational asbestos exposures. The use of products containing asbestos was banned in the UK in 1999. The latency period between first exposure and development of the disease is very long, typically 30–40 years.

Cases of mesothelioma were recorded systematically from the late 1960s. The incidence of mesothelioma has been increasing steadily since then, and current predictions suggest there will continue to be approximately 2500 deaths per year for the rest of this decade, before numbers begin to fall. (HSE http://www.hse.gov.uk/Statistics/causdis/mesothelioma/mesothelioma.pdf).

The largest dataset of MPM in the UK comes from the National Lung Cancer Audit report, which described 8740 cases seen in hospitals in England and Wales between 2008 and 2012.3 Eighty-three per cent of patients were male and the median age at diagnosis was 73 years. Sixty-seven per cent of patients received active anticancer treatment (chemotherapy, radiotherapy and surgery) and overall median survival was 9.5 months, with 1-year and 3-year survival rates of 41% and 12%, respectively. The report identified significant variation in treatment and outcomes across the UK which further highlights the need for an evidence-based guideline to facilitate the highest standards of care for all patients with mesothelioma in the UK.

Intended users of the guideline and target patient populations

The guideline will be primarily of interest to healthcare professionals working within the National Health Service (NHS), but the aim is to make the guideline as applicable to international practice as possible so that it may be used across Europe and America as appropriate. Given the nature of MPM, the majority of the guideline will be relevant to secondary care-based specialists; however, symptom recognition, management and follow-up are all relevant to community-based specialties.

Intended users

  • Primary care–general practitioners (GPs) and practice nurses;

  • Hospital specialist teams in respiratory medicine, oncology, thoracic surgery and palliative care;

  • Hospices/community teams;

  • Specialist nurses (including lung cancer and palliative care);

  • Radiologists;

  • Pathologists.

Areas covered by the guideline

Inclusion

  • The epidemiology and incidence of mesothelioma in the UK and worldwide;

  • The preferred investigation pathway of suspected cases of MPM;

  • Consider special situations including:

    • Imaging;

    • Histology/cytology;

    • Frail patient not fit for invasive tests.

  • Biomarkers;

  • Role of mesothelioma MDTs;

  • Outline best practice in oncological management:

  • Role of chemotherapy;

  • Place for radiotherapy;

  • Role of surgery.

  • Guidance on palliation in MPM;

  • Guidance on providing patients with relevant disease-specific information, including medicolegal/compensation issues;

  • Summary of future therapeutic agents that might be available within next 5 years;

  • Summary of major MPM recommendations.

Areas not covered by the guideline

Non-pleural mesothelioma is excluded from this guideline.

Limitations of the guideline

Healthcare providers need to use clinical judgement, knowledge and expertise when deciding whether it is appropriate to apply recommendations for the management of patients. The recommendations cited here are a guide and may not be appropriate for use in all situations. The guidance provided does not override the responsibility of healthcare professionals to make decisions appropriate to the circumstances of each patient, in consultation with the patient and/or their guardian or carer.

Members of the guideline development group

The guideline development group (GDG) was chaired by two respiratory consultants–Dr Ian Woolhouse and Professor Nick Maskell. The GDG had a wide membership with representation from respiratory medicine, thoracic surgery, medical oncology, radiotherapy, pathology and primary care. A patient representative was on the group for the duration of the process. Those on the group were not required to be BTS members.

Representation

Professor Dean Fennell and Dr Jeremy Steel represented the Association of Cancer Physicians. Dr Anthony Edey represented the British Society of Thoracic Imaging. Professor Corinne Faivre-Finn represented the British Thoracic Oncology Group. Professor Keith Kerr represented the Royal College of Pathologists. Dr Ian Woolhouse represented the Royal College of Physicians. Mr John Edwards and Mr Apostolos Nakas represented the Society of Cardiothoracic Surgeons. Dr Corinne-Faivre-Finn and Dr Anthony Edey represented the Royal College of Radiologists. Dr Tim Peel represented the Association for Palliative Medicine. Dr Steve Holmes represented the Primary Care Respiratory Society UK. Ms Liz Darlison represented the Royal College of Nursing. Dr Graham Abbott, Mr Paul Astle and Mr John Gillies were the patient representatives on the group.

Section 2: Methodology of guideline production

Establishment of guideline development group

The GDG was convened in June 2014, with the first meeting taking place in October 2014. The full GDG met six times during the development of the guideline and kept in close contact by teleconference and email throughout the process. The BTS SOCC reviewed the draft guideline in November 2016. The draft guideline was made available online from 22 March 2017 until 24 April 2017 for public consultation and circulated to all relevant stakeholders. The BTS SOCC reviewed the revised draft in June 2017 and final SOCC approval was granted in September 2017.

Methodology

This guideline is based on the best available evidence and follows the NICE-accredited BTS guideline production process. The methodology used to write the guideline adheres strictly to the criteria as set by the Appraisal of Guidelines, Research and Evaluation (AGREE) II collaboration, which is available online www.agreetrust.org/resource-centre/agree-ii/. The BTS SOCC guideline production manual is available at: https://www.brit-thoracic.org.uk/guidelines-and-quality-standards/

Summary of key questions and literature search

Clinical questions were gathered in the Patient, Intervention, Comparison, Outcome and Time (PICOT) format. The key questions are summarised below.

  • Which clinical features predict the presence of MPM?

  • In patients with suspected MPM (post chest X-ray), which imaging modality is best for diagnosis/staging and what technical factors are important?

  • Should biomarkers (serum/fluid) be measured in MPM?

  • Is there a staging system for MPM that determines management and predicts outcome?

  • What factors determine prognosis and timing of treatment in MPM?

  • What are the appropriate cytopathological approaches which allow diagnosis and subtyping of MPM?

  • Is the care of patients with suspected/proven MPM improved by discussion at a specialist MDT?

  • Where histological confirmation is either not possible or not definite, what are criteria for a clinical diagnosis of MPM?

  • What is the optimum strategy for the management of pleural fluid in MPM?

  • Is there a role for surgery in the management and treatment of patients with MPM?

  • Is there a role for systemic anticancer treatment in MPM?

  • Is there a role for radiotherapy in MPM?

  • What treatment/interventions are effective for symptom control in MPM?

  • What are the nursing care and information needs for patients with suspected and proven MPM?

  • What is the most effective follow-up strategy of patients with MPM?

The PICOT framework was used to define the scope of the guideline and formed the basis of the literature search. The literature search was conducted in December 2014 by York University. Systematic electronic database searches were conducted in order to identify all papers which may potentially be included in the guideline. For each question, the following databases were searched: Cochrane Database of Systematic Reviews (CDSR), Database of Abstracts of Reviews of Effects (DARE), Health Technology Assessment Database (HTA), Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE and MEDLINE In-Process, EMBASE and PUBMED (see online supplementary appendix 1).

Appendix 1: List of guideline group members

The search was limited to papers published in English. The searches identified a total of 6173 abstracts. The full list of abstracts was retained and is kept in an archive. A second search was completed in July 2016 to search for relevant papers published between 2014 and 2016, yielding a further 1038 potentially relevant references. Additional references were included from personal collections as appropriate.

Appraisal of the evidence

An initial screen was completed to remove letters, conference papers and news articles. Dr Woolhouse and Professor Maskell read the remaining abstracts (5129), marked those considered relevant to the scope of the guideline and allocated each relevant abstract to a clinical question(s). 950 abstracts were allocated to clinical question(s). For the second search, the initial screen reduced the abstracts to 582. These were all read by Dr Woolhouse and Professor Maskell and 44 were allocated to clinical question(s). GDG members were allocated to work on the questions in small groups.

Each abstract was read and at least two members agreed whether the paper was relevant to the particular guideline section. Papers were excluded if the following applied:

  • If the paper did not answer the clinical question concerned.

  • If it was a case report of fewer than 20 patients; however, this was not an absolute cut-off. Professional judgement was applied and some smaller case reports were considered, and indeed some case reports of more than 20 patients were excluded.

  • If the language of the full paper was not English.

At least two members of each small group independently appraised each paper using the SIGN critical appraisal checklists. An evidence level was assigned to each study using the SIGN methodology (table 1 and 2).

Full papers were obtained for all relevant, or possibly relevant, abstracts.

Table 1

SIGN levels of evidence

Each relevant paper was read in full by at least two members of the GDG and an evidence table entry was completed for each paper used to support a recommendation/good practice point. The full GDG reviewed each section during the regular meetings and consensus was reached. Evidence tables are available in the online supplementary appendix 2.

Appendix 2: Prognostic scores

From the outset, it was acknowledged that there would be little high-quality evidence for some of the clinical questions identified. In this instance, low-grade evidence was considered, along with expert opinion via consensus at the meetings.

The following parameters were used by the GDG to appraise the evidence:

  • How applicable the obtained evidence was in making recommendations for the defined target audience of this guideline.

  • Whether the evidence was generalisable and relevant to the target population for the guideline.

  • Whether there was a clear consistency in the evidence obtained to support recommendations.

  • What the implications of recommendations would be on clinical practice in terms of resources and skilled expertise.

Cost-effectiveness was not considered in detail as in-depth economic analysis of recommendations falls outside of the BTS guideline production process. However, the GDG were asked to be mindful of any barriers to implementing the recommendations and good practice points (GPPs).

Recommendations were graded from A to D as indicated by the strength of the evidence as shown in table 2. In line with SIGN guidance, ‘minus’ evidence was considered where necessary, but only in such instances when there were no published ‘plus’ papers. In this context, any recommendation based on this evidence was made Grade D. GPPs were included where research evidence was lacking, but the GDG felt it was important to highlight practical points which could improve the care of patients. Research recommendations were also highlighted and passed to the Chair of the SOCC on publication of the guideline.

Table 2

SIGN grades of recommendations

Planned review and updating of the guideline

In line with BTS policy, this guideline will be reviewed by the SOCC within 5 years of publication.

Declaration of interest

BTS Declarations of Interest forms have been completed by all members for each year they were part of the GDG. Details of these forms can be obtained from BTS Head Office. Declarations of Interest was a standing item at each GDG meeting.

Stakeholders

Stakeholders were identified at the start of the process and where appropriate societies and organisations were contacted and asked to nominate a specific person to join the GDG. All stakeholder organisations were notified when the guideline was available for public consultation.

Section 3: Clinical features which predict the presence of mesothelioma

There is a paucity of evidence exploring clinical features specific for MPM. Many of the studies are retrospective questionnaire-based case series, which possess a major inherent recall bias in the diagnosed group making interpretation difficult.

There is consistency in the following risk factors and clinical features:

  • Male preponderance is in keeping with occupational exposure.4

  • High-risk occupations include production of asbestos sheets, brake and clutch linings, construction/demolition work, dock and ship yard workers, electricians, plumbers and launderers.5

  • The predicted lifetime risk of mesothelioma for British men born in the 1940s, who did >10 years of work in the following categories, before the age 30 is as below: 5.9% for carpenters, 2% for plumbers, electricians and painters and 0.8% for other construction workers.6

  • Non-occupational routes of exposure involves: para exposure via a relative or partner, living in the vicinity of an asbestos factory and environmental exposure (low level).4 There is a higher risk of developing MPM from exposure to amphiboles (brown and blue asbestos) rather than chrysotile (white asbestos, the most commonly used form).7 The mean latency between asbestos exposure and developing the disease is 40 years for pleural and 46 years for peritoneal mesothelioma.4

  • There are rare familial cases linked to mutation of the breast cancer-associated protein 1 (BAP1) gene.8

Symptoms

Chest pain and dyspnoea are the most common presenting symptoms, but the relative frequency of these symptoms is not consistent in different studies. Other symptoms include weight loss, fevers and sweats4 9 10 (Table 3).

Table 3

Symptoms at initial presentation in 90 evaluable cases of malignant pleural mesothelioma10

Clinical signs

Pleural effusion is often present. Other signs are variable (eg, palpable lymph nodes).10 Right side predominance of the disease in the order of 1.6:1. This might partially reflect the increased pleural surface area of the right hemithorax.4

Usually the first investigation in patients with suspected mesothelioma will be a chest X-ray. The NICE Guideline on Investigation and Referral for Suspected Cancer gives guidance on when a chest X-ray should be offered in suspected MPM (table 4). The CDG noted that smoking and finger clubbing are more appropriate when considering lung cancer, as opposed to mesothelioma.

Table 4

NICE NG 12. Referral criteria for suspected malignant pleural mesothelioma185

Evidence statements

Occupational exposure to asbestos is recalled in the majority of patients with MPM. High-risk occupations are ship building and construction/demolition work (including boiler repair, and working as a carpenter or electrician). Level 2-.

Symptoms are not specific to MPM. Common symptoms at presentation include chest pain and breathlessness. Less common symptoms at presentation include weight loss, fatigue, fever and cough. Level 2-.

The most common examination finding at presentation is a pleural effusion (with <1 in 10 presenting with lymphadenopathy or clubbing). Level 2-.

Recommendations

  • Do not rule out a diagnosis of MPM on the basis of symptoms and examination findings alone. Grade D.

  • Offer an urgent chest X-ray to patients with symptoms and signs as outlined in NICE NG12. Grade D.

  • Refer all patients with a chest X-ray suggestive of MPM urgently (via the 2-week wait suspected cancer pathway in England and Wales). Consider referral for further investigation in patients with persistent symptoms and history of asbestos exposure despite normal chest X-ray. Grade D.

  • A thorough occupational history should be taken to cover all occupations throughout life. It is important to elicit para exposure by exploring details of relative and/or partner occupations. Grade D.

Section 4: staging systems

In 2016, the International Association for the Study of Lung Cancer (IASLC) Staging Committee published proposals for the revisions of the T, N and M descriptors for the eighth edition of the TNM classification of MPM.11 This was an international, multi-institutional cohort study. The study population was patients with newly diagnosed (cytologically or histologically) MPM. Information was collected on the extent of disease, demographic characteristics, comorbidities, treatment and survival. The dataset included data on 1987 patients with pathologically confirmed MPM from 29 centres on four continents. These comprised 509 cases with only clinical staging information, 836 cases with only pathological staging information (ie, surgical staging) and 642 cases with both clinical and pathological information available. Survival was examined for T, N and M categories according to the seventh edition staging system. Categories were then modified where appropriate to improve prognostic performance. Clinical and pathological T1a and T1b were combined into a single T1 classification. Clinical and pN1 and pN2 categories were collapsed into a single N category comprising ipsilateral, intrathoracic nodal metastases (N1). Nodes previously categorised as N3 were reclassified as N2. M category remained unchanged (see table 5). The proposed TNM groupings are shown in table 6. Figure 1 shows the survival curves for each of the new TNM stage groupings. The prognostic performance comparisons for each stage demonstrated statistically significant HRs for stage IB versus IA, stage IIIA versus II and stage IV versus IIIB.

Figure 1

Overall survival according to best stage (proposed eighth edition). Reprinted from Journal of Thoracic Oncology, Vol 11, No 12, Rusch V.W et al, THE IASLC Mesothelioma Staging Project: Proposals for the M Descriptos and for Revision of the TNM Stage Groupings in the Forthcoming (Eighth) Edition of the TNM Classification for Mesothelioma. 2112-2119 (2016), with permission from Elsevier.

Table 5

Eighth edition AJCC/UICC staging for malignant pleural mesothelioma

Table 6

TNM stage groupings proposed for the eighth edition of MPM staging system relative to those used in the seventh edition

The Brigham and Women’s Hospital Group proposed an alternative system to the AJCC/UICC staging system.12 The alternative system is based on patients undergoing EPP, but this has not been accepted widely nor proposals from it included in AJCC/UICC staging group.

The 2016 National Mesothelioma Audit reported that only 42% of patients with MPM diagnosed in 2014 had stage recorded.1

Evidence statements

The proposed eighth edition of the IASLC TNM staging system predicts survival in surgically and non-surgically treated patients with MPM. Level 3.

Recommendation

  • Record staging of MPM according to the version 8 of the IASLC staging proposals. Grade D.

Section 5: imaging modalities for diagnosing and staging

The literature search revealed a large volume of evidence assessing the role of several imaging modalities in the diagnosis and staging of MPM. The use of ultrasound, computed tomography (CT), positron emission tomography (PET) and positron emission tomography-computer tomography (PET-CT) and magnetic resonance imaging (MRI) were all included in the literature review.

A large number of the studies were conducted in mainland Europe or North America. Only a small number of studies were from the UK. The imaging characteristics of MPM are likely to be similar across the world and the demographic profile of patients included is similar to that of patients in the UK (male predominance, mean age >50 years). Therefore, the evidence was considered applicable to the UK population.

Evidence on diagnostic imaging

The majority of diagnostic evidence evaluates the role of imaging in differentiating benign from malignant pleural disease in general, rather than from MPM specifically. Numerous studies have demonstrated the utility of CT, PET-CT and MRI in the assessment of patients with suspected pleural malignancy.13 These studies provide clear guidance on standard morphological characteristics of pleural malignancy using CT and MRI14–17 and are summarised in table 7 along with reported sensitivities and specificities.18–22

Table 7

Diagnostic accuracy of different imaging modalities for diagnosing malignant vs benign pleural disease

Pleural malignancy is typically unilateral. Bilateral involvement is rare, accounting for as few as 3% of cases.15 In 94% of cases of pleural malignancy, there is a pleural effusion on the affected side. However, differentiation between MPM and metastatic pleural malignancy can be challenging. The presence of lung parenchymal involvement or mediastinal or hilar lymph node enlargement may help point towards metastatic pleural disease.21 While the presence of pleural plaques is an indicator of prior asbestos exposure, it is not a marker of malignancy per se and effusions can be found in this context as a result of benign asbestos-related pleural effusion.

PET-CT can be used to provide useful functional information additional to morphology. Typically, areas of abnormal malignant pleural thickening have elevated maximal standardised uptake values (SUVmax).23 24 Thus, using a SUVmax threshold of >2.0 has been found to accurately differentiate malignant from benign pleural disease with a sensitivity of 88%–100% and specificity of 88%–92%.25–27 In a meta-analysis of 11 PET-CT studies, this technique had a pooled sensitivity of 95% (95% CI 92% to 97%) and specificity 82% (95% CI 76% to 88%) for differentiation of malignant from benign pleural disease.28 Causes of false negatives include: small volume tumours and those with a low proliferative index, for instance, early stage epithelioid mesothelioma. In addition, false positives may result from inflammatory diseases, tuberculous pleurisy, parapneumonic effusions and prior talc pleurodesis. One study, which included patients with prior talc pleurodesis, reported significantly lower specificity in comparison to other studies (specificity 35.3%), as a result of the high number of false positives in this group.29

Studies using MRI have highlighted its potential in distinguishing benign from malignant pleural disease. Malignant pleural thickening tends to show inhomogenous hyperintensity on proton-density T2-weighted images and enhancement on T1-weighted images following gadolinium injection, in contradistinction to benign disease that is of low signal on both sequences. When these signal characteristics are combined with morphology and a pleural thickening >1 cm, the accuracy of MRI is very high for differentiation of benign from malignant disease with sensitivity of 100% and specificity of 95% in one study (95% confidence intervals not reported).30 More recent studies have highlighted potential utility for diffusion-weighted MR imaging (DWI-MRI) in differentiating pleural malignancy from benign pleural disease, with lower Apparent Diffusion Coefficient (ADC) values being demonstrated in pleural malignancy.31 32 Coolen et al also performed DWI-MRI in a study of pleural malignancy and reported that inhomogeneous restriction in diffusion of the thickened pleura differentiates malignant from benign pleural disease with a sensitivity of 92.5% (95% CI 84% to 97%) and specificity of 79% (95% CI 62% to 89%).33 Gill et al demonstrated that patients with epithelioid MPM have a significantly higher ADC value than those with non-epithelioid MPM and an ADC threshold of 1.1 could differentiate epithelioid MPM from sarcomatoid MPM with a sensitivity of 60% and specificity of 94% (95% CIs not reported).32 These MRI data appear promising but are yet to be validated prospectively and importantly their added value in disease with atypical or equivocal CT signs is unclear.

Evidence on staging

Seventeen25 34–49 studies were identified that evaluated the role of various imaging modalities when staging MPM. One systematic review50 and one meta-analysis51 were also identified in the literature. To a degree all imaging modalities are limited in accuracy of staging compared with the gold standard of postoperative histological staging and mediastinoscopic sampling of lymph nodes. However, assessment of limitations is made difficult by the relative infrequency of surgical resection and the use of comparator imaging techniques as the reference point in many of the studies.

Despite the overall benefits of CT scanning when initially assessing patients with suspected mesothelioma, CT performs poorly when compared against other modalities for staging of MPM. CT is particularly poor at assessing T4 stage where assessment of invasion through soft tissue such as diaphragm and chest wall is required. CT also performs poorly at lymph node staging, particularly when detecting involved N2 and N3 nodes. In one study, 37% of the patients were upstaged following a PET scan.35

The role of MRI is limited in staging MPM.34 36 37 39 41 42 48 However, MRI does perform better than CT, where tumour-soft tissue delineation is required. For example, MRI has a sensitivity and specificity of 87.5% and 87.5% for stage II disease, and 91% and 100% for stage III disease due to its superiority in detecting invasion into or through chest wall, endothoracic fascia, diaphragmatic muscle and mediastinal fat.36 Table 8 provides a brief summary.

Table 8

Showing the sensitivity and specificity of CT, MRI and positron emission tomography (PET)-CT in mesothelioma staging36

It should also be noted that although Plathow et al36 showed an accuracy of 100% and low interobserver variability when staging patients with MPM with PET-CT, compared with CT and MRI, the results of other smaller studies are mixed.

Pragmatically CT (optimised for pleural enhancement) remains the mainstay of imaging for diagnosis and staging of MPM. A number of centres routinely include the abdomen and pelvis in the initial CT scan whereas others perform completion scanning according to the results of other diagnostic tests.

Evidence statements

Overall reported diagnostic accuracy of CT in the detection of pleural malignancy is 68%–97%, with specificity of 78%–89%. Level: 3.

CT and ultrasound features of malignant pleural disease include pleural thickening >1 cm, nodular pleural thickening, mediastinal pleural thickening and interlobar fissural nodularity. Level: 3.

Features favouring MPM over metastatic pleural malignancy are the presence of pleural plaques, involvement of the interlobar fissure and the absence of lung parenchymal involvement. Level: 3.

Overall reported diagnostic accuracy of PET-CT in the detection of pleural malignancy–sensitivity 88%–95%, specificity 35%–100%. Level: 2+.

False positives at PET-CT are common in TB pleuritis, inflammatory disorders of the pleura and previous talc pleurodesis. Level: 3.

Overall reported diagnostic accuracy of MRI in the detection of pleural malignancy–sensitivity 60%–100%, specificity 73%–95%. Level: 2−.

CT has limited accuracy for staging MPM using current staging systems. Level: 3.

MRI is better than CT at detecting invasion through diaphragm and T3 disease (invasion through muscle, bone, mediastinal fat), but has limited sensitivity in nodal staging. Level: 3.

Integrated PET-CT has the highest accuracy for staging MPM. It has better sensitivity across all three criteria T, N and M compared with CT and MRI. Level: 2+.

Recommendations

  • Offer CT thorax with contrast (optimised for pleural evaluation) as the initial cross-sectional imaging modality in the evaluation of patients with suspected MPM. Grade D.

  • Use of PET-CT for aiding diagnosis of MPM is not recommended in patients who have had prior talc pleurodesis and caution should be employed in populations with a high prevalence of TB. Grade D.

  • In patients where differentiating T stage will change management, consider MRI. Grade D.

  • In patients where excluding distant metastases will change management, offer PET-CT. Grade D.

Section 6: Pathological diagnosis

A diagnosis of MPM can be challenging because the tumour has a wide range of morphological appearances and may mimic many other epithelial or sarcomatoid malignancies. The best method for obtaining pleural tissue is already covered in the current BTS pleural disease guidelines. For this reason, this topic was not covered in the PICOT questions used in our initial mesothelioma literature search. The BTS pleural disease guideline can be downloaded at the following website: https://www.brit-thoracic.org.uk/standards-of-care/guidelines/

In summary, the BTS pleural guideline states:

  1. In patients with a symptomatic exudative pleural effusion where a diagnostic pleural aspiration is negative or inconclusive, thoracoscopy (either by local anaesthetic thoracoscopy or VATS) is suggested as the next choice investigation since the procedure is relatively uncomplicated and pleurodesis can be performed at the same time if indicated.

  2. If a contrast-enhanced thoracic CT scan of a patient shows a focal area of abnormal pleura (with or without a pleural effusion), an image-guided cutting needle biopsy has a high yield and low complication rates. This technique is particularly useful in patients who are unsuitable for thoracoscopy.

The morphological features of MPM are well described elsewhere in the WHO classification of pleural tumours,52 and the guidelines of the International Mesothelioma panel,53and are beyond the scope of this guideline. The importance of histological subtyping of MPM is highlighted in the national mesothelioma audit report, which demonstrates that non-epithelioid histology was associated with significantly shorter overall survival in this cohort.1 Table 9 highlights the main subtypes of mesothelioma and the different morphological features that might be present within each group.

Table 9

Mesothelioma subtypes

The literature search identified 176 papers related to the use of ancillary techniques to improve the diagnosis of malignant mesothelioma. Several were rejected due to study age, the applicability of the diagnostic tests, small numbers of cases, or an inability to extract data, resulting in 70 papers being selected for review. All were retrospective case series. Case numbers varied greatly, from 23 up to 596 cases, and were often very heterogeneous case mixtures. Immunohistochemistry (IHC) was by far the most frequently considered ancillary diagnostic technique. Other approaches used included electron microscopy, chromosomal analysis, microRNA expression, DNA methylation, mRNA expression array, fluid chemistry assay, cytofluorimetry, flow cytometry and fluorescence in situ hybridisation (FISH).

The quality of the evidence reviewed was highly variable. Some of the papers were unique descriptions of unusual diagnostic approaches without comparators. In some studies, the origin of the tumour tissue was not clear and others used autopsy material. Many of the older studies, especially those published prior to 1990, use clones of primary antibody or other immunohistochemical techniques that are no longer used or available. More recent studies typically used contemporary reagents that are available and applicable in the UK.

Summary of individual immunohistochemistry evidence

A large number of IHC markers have been reviewed and are summarised in table 10, with sensitivity and specificity values where available. It should be noted that the sensitivity and specificity of many of these markers are reduced in sarcomatoid MPM, which frequently does not express any of the typical ‘mesothelial’ markers. In this scenario, expression of keratins may be the only demonstrable feature, which is helpful but non-specific. Additionally, discriminating malignant from benign mesothelial proliferations is not reliable using IHC markers.

Table 10

Summary of immunohistochemistry markers

Additional techniques

Wu et al54 examined p16 FISH to discriminate reactive from malignant mesothelium in 60 patients. Hemizygous or homozygous deletion of p16 was not seen in fibrous pleurisy (FP), but was detected in 66.7% of epithelioid MPM, 87.5% of biphasic MPM and 100% of sarcomatoid cases, highlighting potential utility in the differentiation of MPM from FP. Hida et al55 performed BAP1 and p16 FISH in 40 cases of MPM and 20 cases of inflammatory pleuritis. All inflammatory cases and only three mesothelioma cases were negative for both. The presence of BAP1 and or p16 FISH may therefore be helpful in differentiating MPM from benign mesothelial proliferation.

Diagnosis in cytology

This remains a controversial subject. The reliability of an MPM diagnosis on effusion cytology is highly variable (sensitivity ranging from 16% to 73%56 57) and is very much dependent on cytologist experience. Some centres will send clot/cell block sections for the homozygous deletion of the 9p21 band (p16), which can increase diagnostic certainty.

Evidence statements

Glut1 IHC and p16 FISH have potential for discriminating benign from malignant mesothelium. Level 3.

The sensitivity of pleural fluid cytology for the diagnosis of MPM is highly variable and is dependent on the cytologist’s experience. Level 3.

Positive IHC markers for MPM include calretinin, thrombomodulin, CK5/6, CAM5.2, EMA, vimentin, GLUT-1, HBME-1, WT-1, p53. Overall sensitivity is 45%–100%. Level 3.

Negative IHC markers for MPM include Ber-Ep4, MOC-31, CEA, Leu-1, CD15, TTF-1, B72.3. Overall specificity is 53%–100%. Level 3.

A combination of two positive mesothelial markers and two negative adenocarcinoma markers increases diagnostic accuracy. Level 3.

Diagnostic accuracy of IHC markers is reduced in sarcomatoid MPM. Level 3.

Accurate subtyping of IHC markers is reduced in sarcomatoid MPM. Level 3.

Recommendations

  • IHC is recommended for the differential diagnosis of MPM in both biopsy and cytology type specimens. Grade D.

  • A combination of at least two positive mesothelial (calretinin, cytokeratin 5/6, Wilms tumour 1, D-240) and at least two negative adenocarcinoma immunohistochemical markers (TTF1, CEA, Ber-EP4) should be used in the differential diagnosis of MPM. (Markers listed in likely order of value). Grade D.

  • Do not rely on cytology alone to make a diagnosis of MPM unless biopsy is not possible or not required to determine treatment due to patient wishes or poor PS. Grade D.

  • Pathologists should report the histological subtype of MPM in all cases. Grade D.

Good practice points

  • Biopsies from patients with suspected MPM should be reviewed by a pathologist experienced in the diagnosis of MPM and a second opinion should be sought if there is uncertainty over the diagnosis.

Section 7: use of biomarkers

The literature search revealed a large volume of evidence, exploring different biomarkers that may have a role in MPM. Literature on at least 20 markers tested in serum, plasma, pleural fluid and exhaled breath were reviewed. A number of markers were assessed in exploratory studies with no further validation, and such markers have not been considered further given the lack of validation studies.

Several markers such as Mesothelin, Fibulin-3, Osteopontin and Megakaryocyte Potentiating Factor (MPF) have been extensively studied internationally. Individual studies and controlled meta-analyses specifically looking at these markers were identified and reviewed. Significant heterogeneity was noted between study populations. In particular, there was wide variability in comparator groups and disease prevalence. For example, comparator groups include normal controls, asbestos-exposed well individuals, patients with benign effusions and patients with non-mesothelioma malignant effusions. In some areas, the prevalence of mesothelioma in the sampled population was >30%, in others <5%. The cut-off value for markers varied in most studies.

Although most studies included sarcomatoid mesothelioma, this made up only a small proportion of the overall cohort of any single study.

Evidence on diagnostic markers

The most robust body of evidence at present for diagnosis of MPM is for Soluble Mesothelin Related Peptides (SMRP) and Osteopontin, as summarised below:

  • A meta-analysis by Cui et al58 reviewed 28 publications totalling 7550 patients (1562 MPM and 5988 non-MPM), which confirmed serum SMRP to have an overall sensitivity of 60% and a specificity of 81%, with an area under the curve (AUC) of 0.734.

  • The same review also demonstrated that pleural fluid SMRP has an overall sensitivity of 75%, specificity 76% and AUC of 0.809 (total number of patients 1506; 460 MPM and 1046 non-MPM).

  • Summary sensitivities and specificities for SMRP and Osteopontin–from two meta-analyses by Hu et al,59 reviewing six publications with a total of 906 patients, and Lin et al60 reviewing seven publications with a total of 1096 patients (Table 11).

Table 11

Summary sensitivities and specificities for Soluble Mesothelin Related Peptides (SMRP) and Osteopontin (OPN)

There were a number of studies on Fibulin-3, representing a smaller body of evidence than that above for SMRP and Osteopontin. These are summarised in table 12.

Table 12

Summary sensitivities and specificities for Fibulin-3

Markers for disease monitoring and assessment of progression

Sixteen61–76 papers were reviewed in relation to above. Again, SMRP is the most widely studied marker but other biomarkers such as Fibulin-3, Osteopontin, Megakaryocyte potentiating factor (MPF and Hyaluronic acid (HA) were also assessed. Study populations are heterogeneous with regards to their management. Disease progression/stability in these studies has in general been assessed by the use of the modified response evaluation criteria in solid tumours (RECIST).

Overall:

  • SMRP shows a positive correlation with tumour bulk.62

  • In patients who had EPP, there was a significant drop in SMRP levels (on average 54%). Despite the relationship with tumour bulk, there is no significant correlation with increasing disease stage.

  • Mean and median SMRP levels for those with progressive disease showed a significant difference compared with patients with partial/complete response and stable disease.62

  • A falling SMRP level between baseline and two cycles of chemotherapy was associated with a longer ‘time to progression’ of disease. Fibulin-3 failed to show a similar relationship.75

  • Low Fibulin-3 at diagnosis is associated with a prolonged survival.61

Outcome prediction

Four studies71 72 74 75 assessed the independent predictive value of biomarkers on overall survival in MPM, accounting for the recognised prognostic indicators of histological subtype, age and PS. These demonstrate:

  • The modified Glasgow Prognostic Score (mGPS) (serum, C reactive protein (CRP) and albumin level at baseline) and blood neutrophil-to-lymphocyte ratio (NLR) are independent predictors of overall survival (HR 2.6 and 2.0, respectively).71

  • Pleural fluid HA level (<225 mg/L) is independently associated with overall survival (RR 0.63).72

  • Resection specimen staining for smoothened SMO transmembrane receptor (HR 1.06) was an independent predictor of overall survival.74

A fall in SMRP between baseline and an interval of 6–8 weeks (post two cycles of chemotherapy) is predictive of radiographic stability of disease. A falling SMRP level at completion of chemotherapy is strongly associated with a longer survival.75 Baseline SMRP was unable to predict survival. Apart from SMRP in the SWAMP study,75 none of the other markers have been prospectively validated.

Biomarkers for screening

Five studies77–81 explored the potential role of biomarkers in screening for MPM. All five studies looked at SMRP but two studies also looked at Osteopontin, CA-125 and cytokeratin fragment 19.58 75 Studies were heterogeneous particularly with regards to the cut-off value of SMRP, duration of follow-up and the patient populations (other cancers/control groups). Despite these differences, SMRP tended to be higher in those with asbestos-related disorders such as asbestosis or diffuse pleural thickening, and in renal impairment. One study found SMRP levels are also elevated in other cancers such as lung, ovarian, pancreatic and endometrial cancer, but the populations of patients with these cancers were small.

Evidence statements

Diagnosis

There is no diagnostic biomarker, which is able to consistently diagnose MPM with a sensitivity and specificity above 90%. Level 2+.

The diagnostic value of biomarkers in sarcomatoid mesothelioma is lower than that for epithelioid, but small numbers mean that accuracy of sensitivity and specificity are difficult to derive. Level 2+.

Serum SMRP has a relatively high specificity in the diagnosis of MPM across a large number of studies (81%). Level 2+.

Serum and pleural fluid Osteopontin has a relatively high specificity in the diagnosis of mesothelioma across a modest number of studies (81%). Level 2++.

Fibulin-3 shows variable performance in diagnosis of MPM (sensitivity range 22%–100%). Level 2+.

Disease response

SMRP level is correlated with tumour bulk and falls post-EPP, but baseline level does not predict pathological stage in mesothelioma. Level 2+.

In assessing response to therapy, SMRP levels are higher in those with progressive disease compared with those with partial response, complete response or disease stability. Level 3.

During chemotherapy, a falling level of SMRP from baseline to interval, or a falling level at completion of palliative chemotherapy is associated with a longer survival. Level 3.

Outcome prediction

There is no prospectively validated biomarker which independently predicts overall survival in MPM. Level 2-.

Markers of inflammation, pleural fluid HA and cell staining patterns may predict survival but further studies are required to validate this. Level 2-.

Recommendations

  • Do not offer biomarkers in isolation as a diagnostic test in MPM. Grade B.

  • Consider biomarker testing in patients with suspicious cytology who are not fit enough for more invasive diagnostic tests. Grade B.

  • Do not routinely offer biomarker testing to predict treatment response or survival. Grade B.

  • Do not offer biomarker testing to screen for MPM. Grade C.

Research recommendation

Further research is required to identify biomarkers that reliably predict treatment response within clinical practice.

Section 8: factors determining prognosis and timing of treatment

There is a large body of evidence on this topic in the literature. The great majority of it is of poor quality, being retrospective case series. Some of these are taken from patients enrolled into clinical trials, where the consistency and quality of the data collected is higher.

A large number of baseline patient variables have been studied seeking prognostic factors. These include demographic factors (age, sex, race), disease features (histological subtype and grade, site of disease, disease stage using various staging systems), Eastern Cooperative Oncology Group (ECOG) PS or Karnofski performance score (KPS), symptoms (particularly chest pain and weight loss, usually not further defined), markers of inflammation (total white blood count (WBC), platelet count, neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), C-reactive protein level (CRP) and blood test markers of systemic disease such as haemoglobin (Hb) level, Hb difference from a population ideal value (160 g/L in men, 140 g/L in women), serum albumin.

Several prognostic scores have been developed for mesothelioma, combining groups of prognostic variables derived from derivation cohorts of patients with mesothelioma and subsequently validated in different test cohorts. The following scores are described in more detail below; the EORTC prognostic score (EPS), the CALGB score,82–87 mGPS have been studied retrospectively in a cohort of patients with mesothelioma,71 the LENT prognostic score88 and a prognostic model using decision tree analysis was published by Brims et al89 in 2016.

Evidence from very large studies

Three retrospective studies were identified, which included more than 5000 patients from population-level registries.90–92 These consistently demonstrate that increasing age, male sex, advanced stage and non-epithelioid histology are prognostic of worse overall survival. Although this evidence is of low quality, being retrospective, the size of the datasets studied and the absence of any contradictory evidence increases the confidence in these findings.

Findings from the National Lung Cancer Audit

In 2015, Beckett et al published data from 8740 mesothelioma cases included in the National Lung Cancer Audit.3 This is the largest prospectively collected case series in the literature. It has the advantage of reflecting the characteristics of unselected incident cases. In this respect, it differs from the populations of clinical trial recruits who have been used to derive, for example, the EORTC and CALGB prognostic scores (see below). Poorer PS and non-epithelioid histology were associated with shorter overall survival in this cohort. Survival by sex is not reported.

The EORTC prognostic score

This was derived by Curran et al83 in 1998 based on maximum-likelihood parameter estimates of the prognostic factors retained in a multivariate model derived from a population of 204 patients (89% male) entered into clinical trials of chemotherapy in Europe. All patients were PS 0–2. More detail on the score can be found at appendix 2.

CALGB prognostic groups

Herndon et al studied prognostic factors in a group of 337 patients with MPM not previously treated with chemotherapy who were entered into phase II trials of chemotherapy.87 Cox survival and exponential regression trees were used to determine prognostic importance of pretreatment patient characteristics. Terminal nodes were amalgamated to form six distinct prognostic subgroups.

The derived prognostic groups are complex, and continuous variables are dichotomised differently for different subgroups (eg, Hb and WBC). Edwards et al validated the CALGB groups in a retrospective study conducted in a UK population.86

Meniawy et al have validated the CALGB prognostic group method in a recent, large study in Western Australia, in a population of patients where 62% received chemotherapy. This is considerably higher than the proportion of patients currently receiving chemotherapy for mesothelioma in the UK and therefore the median survival estimates derived from the validation study are likely to be considerably better than those observed in the UK.

The neutrophil-to-lymphocyte ratio

Five studies have considered the NLR in mesothelioma. The evidence on the prognostic utility of NLR was reviewed by Meniawy et al.85 They concluded that the cut-off value chosen for NLR is variable, the independent predictive effect inconsistent and the NLR has not been validated in a prospective study. More information about the studies can be found in appendix 2.

The modified Glasgow prognostic score

The mGPS stratifies patients with cancer according to CRP and serum albumin. This was found to be an independent predictor of overall survival in MPM in one study71 (HR 2.6, 95% CI 1.6 to 4.2, P<0.001) but has not been the subject of prospective validation.

Prognostic model using decision tree analysis

Brims et al derived a prognostic model using classification and regression tree analysis from an unselected population of 482 patients newly diagnosed with MPM in Western Australia, of whom 274 were collected retrospectively and 208 prospectively.89 Unlike the cohorts used to derive the CALGB and EORTC models, which were of participants in chemotherapy trials, this paper included all patients with a confirmed diagnosis of MPM within the inclusion period. The model was validated in a cohort of 177 patients with MPM prospectively collected in Bristol, UK. The validation cohort is likely to be highly representative of typical new patients with MPM presenting in the UK. The model was used to predict death at 18 months. The variable with the greatest influence on survival in the derivation cohort was weight loss, defined as any weight loss considered significant by the medical team. The decision tree for classifying patients into prognostic groups in this study is shown in table 13. The variables having an influence on prognosis within this model are histological subtype, weight loss, PS, Hb and serum albumin. The C-statistic for the derivation cohort was 0.76 and the sensitivity 94.5% (95% CI 91.4% to 96.7%) and the specificity 38.2% (95% CI 30.6% to 46.3%). The positive predictive value for death at 18 months was 76% (95% CI 71.5% to 80.1%). The C-statistic for model performance in the validation cohort was 0.68 (95% CI 0.60 to 0.75).

Table 13

Brim decision tree classification

The model can be found in appendix 2.

The LENT score

Clive and others derived the LENT score, for predicting survival in patients presenting with malignant pleural effusion (MPE).88 The LENT score uses pleural fluid LDH (&g