The comparative accuracy of different pleural biopsy techniques in the diagnosis of malignant mesothelioma.

AIMS: To evaluate the diagnostic accuracy of closed and open pleural biopsies in diagnosing malignant pleural mesothelioma. METHODS AND RESULTS: The autopsy study group comprised 45 malignant mesotheliomas. All prior pleural biopsy investigations were reviewed. Forty-one of 45 (91%) had had an antemortem diagnosis of malignant mesothelioma. In these 41 cases, 57 prior diagnostic pleural biopsies had been performed [36 closed needle biopsies: 31 blind; five computed tomography (CT)-guided and 21 open pleural biopsies]. For definitive diagnosis open pleural biopsy yielded a sensitivity of 95% and specificity of 100%. For definitive diagnosis closed blind pleural biopsies yielded a sensitivity of 16% and specificity of 94%. Thirty-two per cent of 'blind' biopsies were inadequate. CT-guided pleural biopsies yielded a definitive diagnostic accuracy of 100% (5/5). Biopsy specimen size was important in obtaining a positive definitive diagnosis. Diagnosis was attained in 75% of specimens >10 mm in size compared with 8% <10 mm in size. CONCLUSIONS: Overall, all procedures had utility but definitive diagnostic accuracy for 'blind' closed pleural biopsy was low (16%), dependent on biopsy specimen size and tumour subtype. Sarcomatoid subtype malignant mesothelioma yielded the lowest diagnostic accuracy. For all subtypes of malignant mesothelioma, open pleural biopsy produced the highest diagnostic accuracy (100% sensitivity, 95% specificity).

acp. Best practice no 161. Examination of lung specimens.

This article gives guidance for the handling and examination of various types of lung tissue specimens to provide: (1) accurate diagnosis and assessment of severity of disease; (2) sufficient information for the accurate staging of tumours; and (3) an assessment of the contribution of various occupational disorders to the cause of death.

Phraseology in pathology reports. A comparative study of interpretation among pathologists and surgeons.

This questionnaire based study compared the interpretation, use and preferences, among pathologists and surgeons, of descriptive phrases found in surgical reports. The results show that there is a wide variation in individual interpretation of phrases in both groups. The frequency of usage of phrases by pathologists and preference for phrases by surgeons were also diverse. The adoption of a limited number of descriptive phrases that are mutually understood and acceptable for use by both pathologists and clinicians is recommended to avoid interpretive ambiguity in pathology reports.

Mediastinal mixed dendritic cell sarcoma with hybrid features.

This report describes a recurrent sarcoma involving the soft tissues of the posterior mediastinum with features of both follicular dendritic and interdigitating dendritic cells. Histologically, the tumour, which was a recurrent neoplasm 19 years after the initial removal, was composed of bland spindle shaped cells with interspersed inflammatory cells, predominantly lymphocytes. Immunohistochemically, the spindle cells were positive for S100 protein, CD45, CD68, and vimentin, but negative for CD21 and CD35 (markers of follicular dendritic cells). The immunophenotype was in keeping with interdigitating dendritic cells. However, ultrastructural examination demonstrated elongated cell processes joined by desmosome-like junctions-features in keeping with follicular dendritic cells. Follicular dendritic cell sarcoma and interdigitating dendritic cell sarcoma are rare neoplasms and a high index of suspicion is required to make a correct diagnosis. Immunohistochemistry and/or electron microscopy are required for a definitive diagnosis. This case shows that hybrid neoplasms can occur, which have features of both follicular and interdigitating dendritic cells.

Consistency in the observation of features used to classify duct carcinoma in situ (DCIS) of the breast.

AIM: To determine interobserver and intra-observer agreement in the assessment of cytological grade and intraduct necrosis in pure duct carcinoma in situ (DCIS) of the breast. METHODS: Sixty unselected cases with illustrated diagnostic criteria were circulated to 19 practising histopathologists. RESULTS: Overall agreement was moderate for cytological grade in three categories: 71% agreement; weighted kappa (kappa w), 0.36; intraduct necrosis in three categories (absent, present, extensive): 76% agreement; kappa w, 0.57; and the Van Nuys classification system: 73% agreement; kappa w, 0.48. Agreement was no better among observers participating in the National External Quality Assurance Programme. Intra-observer agreement for cytological assessment (69.6% agreement; kappa w, 0.52) and intraduct necrosis (68.3% agreement; kappa w, 0.48) was moderate, suggesting that individual variation rather than precision of criteria contributes to the lack of agreement. CONCLUSIONS: Moderate agreement on observations can be achieved by non-specialist pathologists, with better agreement on necrosis than cytological grade. There was evidence of consistent individual bias towards over or under scoring cytological grade, which could be corrected with adequate and prompt feedback.

The pathology associated with therapeutic procedures in malignant mesothelioma.

AIMS: To describe iatrogenic pathological lesions in malignant pleural mesothelioma. METHODS AND RESULTS: All cases of malignant pleural mesothelioma confirmed by antemortem pleural biopsy and undergoing post mortem examination over a 7-year period (1995-2001) formed the study group. This comprised 48 malignant pleural mesotheliomas [epithelioid (n = 21), biphasic (n = 14) and sarcomatoid (n = 13)]. Twenty-eight of 48 (58%) had received chemical (talc) pleurodesis, 30/48 (63%) palliative localized radiotherapy, 6/48 (13%) chemotherapy, and 14/48 (30%) surgery [12/48 (26%) pleural decortication and 2/48 (4%) pleuropneumonectomy]. CONCLUSIONS: Talc pleurodesis induces a marked pseudosarcomatous fibroblastic proliferation which may impart a biphasic pattern to the neoplasm. In more chronic cases, paucicellular fibrosis with a foreign body giant cell reaction is noted. The talc is polarizable and deposited in linear fashion within the tumour. In 2/28 (7%) pleurodesis cases platyform ferruginous bodies were seen in the peripheral alveolated lung parenchyma and these mimicked asbestos bodies. An awareness of this is important to prevent false attribution to asbestos. Talc could be identified by transmission electron microscopic mineral analysis in 5/15 (33%) cases examined. Tumour nodules developing subjacent to iatrogenic wound sites were noted in 8/48 (17%) cases. In 6/8 (75%) of these cases, comparative assessment of the locally irradiated subcutaneous chest wall tumour, with background pleural mesothelioma, showed no morphological difference in architectural tumour growth pattern, extent of necrosis, cytological or nuclear pleomorphism, mitotic activity or tumour immunophenotype. In 2/8 (25%) cases the locally irradiated tumour showed prominent bizarre multinucleated tumour giant cells and intense mixed inflammation, a feature not seen in the background (non-irradiated) tumour. All six malignant pleural mesotheliomas receiving chemotherapy appeared refractory to treatment in that chemotherapy did not appear to have any significant effect on the tumour morphology, cytonuclear pleomorphism, mitotic activity, extent of necrosis or immunophenotype. In the 12 decortication specimens and two pleuropneumonectomy resections, post mortem examination identified evidence of residual malignant mesothelioma of similar morphological subtype and immunophenotype to the resected tumour.

Pathology of malignant mesothelioma.

The diagnosis of malignant mesothelioma can pose several problems to the surgical pathologist. First, the morphological appearances of the tumour are known to be diverse with mimicry of a range of both reactive and neoplastic conditions. Second, due to the relative inaccessibility of the serosa, biopsy material is often scanty and fragmentary, producing a plethora of interpretive ambiguities. Third, adjunct techniques such as mucin histochemistry and immunohistochemistry, whilst useful in excluding malignant mesothelioma have little role in confirming the diagnosis. The accurate diagnosis of diffuse malignant mesothelioma is important for two reasons: (1) In relation to prognosis as it has an almost invariable fatal outcome, which contrasts with the other mesothelial neoplasms such as the benign adenomatoid tumour and the borderline malignant tumours, namely the well-differentiated papillary mesothelioma and multicystic mesothelioma; (2) In relation to occupational-related compensation claims following asbestos exposure. This review summarizes the aetiology of asbestos-induced neoplasia, possible mechanisms of tumour development and highlights potential diagnostic pitfalls.

Primary malignant gonadal mesotheliomas and asbestos.

AIMS: The clinicopathological, immunohistochemical and aetiological aspects, with respect to asbestos, of seven primary gonadal mesotheliomas (three intratesticular, four ovarian) are described and compared. These tumours are extremely rare, poorly described and the knowledge of their natural history is very limited. METHODS AND RESULTS: The cases were collated from the UK Health and Safety Executive Mesothelioma Register over a 24-year period (1968-91). Primary mesotheliomas of the tunica vaginalis and ovary comprised 0. 09% (10 cases) and 0.03% (three cases) of mesothelioma deaths, respectively. No primary intratesticular (non-tunica vaginalis) malignant mesotheliomas have been described. In this study, we present seven (three intratesticular, four ovarian) primary malignant gonadal mesotheliomas. In both genders the tumours show a similar age distribution (with median onset in the sixth decade), a similar association with asbestos (in approximately 50% cases), a diverse histological spectrum (with predominantly tubulopapillary epithelial subtype tumours) and an immunophenotype that is comparable with malignant pleural and peritoneal mesothelioma. The clinical course appears variable (mean, 26 months; range, 9-50 months). All tumours in the study presented as localized masses and their prognosis appeared more favourable than that of diffuse pleural and peritoneal cases. CONCLUSIONS: An awareness of the existence of these rare forms of malignant mesothelioma is important to prevent misdiagnosis. Immunohistochemistry has an important role in confirmation of the diagnosis. The accurate diagnosis of primary gonadal mesothelioma has potentially important medicolegal compensation considerations as a significant proportion of these cases are associated with asbestos.

'Pseudomesotheliomatous' carcinomas of the pleura: a 10-year analysis of cases from the Environmental Lung Disease Research Group, Cardiff.

AIMS: To undertake a clinicopathological study of diffuse serosal neoplasms of epithelial histogenesis which clinically and pathologically mimic malignant pleural mesothelioma. METHODS AND RESULTS: Over a 10-year (1990-2000) study period 53 carcinomas mimicking diffuse pleural mesothelioma ('pseudomesotheliomatous' carcinoma) were identified. The study group comprised 50 men and three females, age range 33-77 (median 68) years. In 46 (87%) cases there was a history of smoking and in 40 (76%) cases a history of asbestos exposure. Histologically the pleural 'pseudomesotheliomatous' carcinomas could be divided into two broad groups: primary pulmonary carcinomas with florid pleurotropic growth (n = 47), of which 34 (70%) were adenocarcinomas; and diffuse carcinomatous involvement of the pleura by metastatic tumour (n = 6). This latter group comprised two transitional cell carcinomas of bladder, one renal (clear) cell carcinoma, one ductal pancreatic adenocarcinoma, one prostatic adenocarcinoma and one squamous cell carcinoma of parotid gland origin. Follow-up data were available in 35 cases. Regardless of tumour type, survival was poor (median 8 months) and comparable to diffuse pleural mesothelioma. CONCLUSIONS: Pleural 'pseudomesotheliomatous' carcinomas are uncommon (comprising 6% of referrals), pathologically heterogeneous tumours with poor prognosis. Tissue diagnosis should be obtained in all cases of suspected diffuse pleural neoplasia. By light microscopy and immunophenotype many of the tumours mimicked malignant mesothelioma. In particular, an awareness that all neoplasms exhibiting squamous differentiation may express cytokeratin 5/6 and thrombomodulin is important to prevent misinterpretation. In this respect, calretinin is regarded as the most specific and sensitive mesothelial marker. Misdiagnosis may have medico-legal implications in asbestos-related compensation claims.

Unusual intraparenchymal growth patterns of malignant pleural mesothelioma.

AIMS: Malignant pleural mesothelioma is known to mimic morphologically a number of diverse reactive and neoplastic conditions. We describe three unusual intraparenchymal growth patterns of malignant mesothelioma seen in a series of 200 malignant pleural mesotheliomas. The diagnostic pitfalls associated with these findings are described and their potential medico-legal implications are highlighted. METHODS AND RESULTS: The study group comprised 200 malignant pleural mesotheliomas. In each case diagnosis was morphologically confirmed with ancillary immunohistochemistry using a broad panel of both mesothelial and epithelial markers. The patterns of intraparenchymal growth were documented and grouped as: direct subpleural; lymphangitic; and other. The 200 malignant pleural mesotheliomas comprised 118 epithelioid, 57 biphasic and 25 sarcomatoid, subtyped according to the WHO classification. Direct subpleural invasion was seen in 42 cases, lymphangitic spread in 27 cases. Other less well-defined intraparenchymal patterns included three sarcomatoid subtype malignant mesotheliomas exhibiting an intra-alveolar growth pattern mimicking epithelioid haemangioendothelioma. One epithelioid subtype malignant mesothelioma contained an intraparenchymal tumour nodule microscopically comprising lepidic spread of neoplastic cells over maintained alveolar structures mimicking bronchioloalveolar carcinoma. One epithelioid subtype malignant mesothelioma morphologically had areas in which alveoli were distended by discohesive epithelioid neoplastic cells with no interstitial invasion. The appearances mimicked desquamative interstitial pneumonia. Immunohistochemistry played an important role in the definitive diagnosis of each unusual parenchymal tumour deposit. In 126 malignant mesotheliomas no invasion of the subjacent lung parenchyma was identified. CONCLUSIONS: An awareness of the unusual parenchymal growth pattern in malignant mesothelioma is important to prevent misdiagnosis of other entities. In the medico-legal setting, the presence of epithelioid haemangioendothelioma or bronchioloalveolar carcinoma (in the absence of asbestosis) may be deemed to impact upon the patient's anticipated life expectancy and thereby would decrease the compensation settlement.

The use of immunohistochemistry in distinguishing reactive from neoplastic mesothelium. A novel use for desmin and comparative evaluation with epithelial membrane antigen, p53, platelet-derived growth factor-receptor, P-glycoprotein and Bcl-2.

AIMS: To evaluate the expression of the intermediate filament desmin in reactive mesothelium and malignant mesothelioma and to compare its utility with five other previously reported immunomarkers claimed to be of use in distinguishing reactive from neoplastic mesothelium. METHODS AND RESULTS: Sixty cases of malignant pleural mesothelioma and 40 cases of reactive mesothelial hyperplasia formed the study group. Cases were immunohistochemically stained with desmin, epithelial membrane antigen (EMA), p53, Bcl-2, P-glycoprotein and platelet-derived growth factor receptor (PDGF-R) beta-chain by the avidin-biotin complex method. The cohort of malignant pleural mesotheliomas were immunoreactive to desmin, EMA and p53 in 6/60 (10%), 48/60 (80%) and 27/60 (45%), respectively. In comparison, the cohort of reactive mesothelial hyperplasias were immunoreactive to desmin, EMA and p53 in 34/40 (85%), 8/40 (20%) and 0/40 (0%), respectively. In a smaller cohort (n = 15) of malignant pleural mesotheliomas, Bcl-2, P-glycoprotein and PDGF-R beta-chain were expressed in 0/15 (0%), 2/15 (13%) and 15/15 (100%), respectively. In a small cohort (n = 15) of reactive mesothelial hyperplasias, Bcl-2, P-glycoprotein and PDGF-R beta-chain were immunoreactive in 0/15 (0%), 0/15 (0%) and 6/15 (40%), respectively. CONCLUSIONS: Desmin and EMA appear to be the most useful markers in distinguishing benign from malignant mesothelial proliferations. Desmin appears to be preferentially expressed in reactive mesothelium and EMA appears to be preferentially expressed in neoplastic mesothelium. The complementary use of both markers is advocated in ascertaining the nature of mesothelial proliferations. Immunohistochemical detection of mutated p53 oncoprotein appeared to be of less utility in this study on account of the low marker sensitivity for malignant mesothelioma. However, p53 antibody may be of use as a second-line marker of neoplastic mesothelium within a standard immunohistochemical panel of antibodies. In this study, Bcl-2, P-glycoprotein and PDGF-R beta-chain appear to be of no use in distinguishing reactive from neoplastic mesothelium, although more formal evaluation of these markers is required.

Synchronous diffuse malignant mesothelioma and carcinomas in asbestos-exposed individuals.

AIMS: The development of synchronous diffuse malignant mesothelioma and carcinoma in individuals exposed to asbestos is rare. We report nine cases and discuss the medico-legal implications. METHODS AND RESULTS: Five hundred patients seeking compensation for asbestos-related diffuse malignant mesothelioma were reviewed with access to post-mortem data. The study group comprised cases in which a second (non-mesothelial) neoplasm was identified. The study group comprised eight males, one female, mean age 68 years (range 60-75). All individuals gave a history of asbestos exposure. Synchronous malignant mesothelioma with carcinoma was identified in 9/500 (1.8%). Eight malignant mesotheliomas were pleural, one was primary peritoneal in origin. By morphological subtyping there were four epithelioid, three biphasic and two sarcomatoid mesotheliomas. In 6/9 (67%) the second tumour was a primary bronchogenic carcinoma (three adenocarcinomas, two squamous cell carcinomas and one small-cell carcinoma). In 3/9 (33%) the second tumour was a non-bronchogenic carcinoma (colonic, pancreatic and breast ductal adenocarcinoma). No other neoplasms were identified in the cohort of malignant mesotheliomas studied. Five persons had pathological evidence of asbestosis (four had bronchogenic carcinomas, one colorectal adenocarcinoma). Two persons with non-bronchogenic carcinomas had identifiable asbestos bodies but no interstitial fibrosis. In two cases the second neoplasms (primary bronchogenic squamous cell and small-cell carcinomas) were associated with diffuse interstitial fibrosis but no asbestos bodies were seen on light microscopy. In each case transmission electron microscopic mineral analysis revealed an asbestos fibre burden within the background population range for control subjects and well below that seen in cases of established asbestosis. These cases were considered to represent cryptogenic fibrosing alveolitis in subjects with a history of asbestos exposure. CONCLUSIONS: Synchronous malignant mesothelioma with carcinomas in asbestos-exposed workers is rare and identified in 1.8% of 500 malignant mesotheliomas in this series. In most cases the carcinoma represents a primary bronchogenic neoplasm. Primary lung carcinomas are recognized to be asbestos related only when occurring in association with asbestosis. In this series this combination (bronchogenic carcinoma and asbestosis) was seen in four (0.8%) cases. In post-mortem cases for possible malignant mesothelioma it is important to identify any other neoplasia and determine whether it is related to asbestos. Their presence impact upon anticipated life expectancy and in the presence of malignant mesothelioma will affect the compensation settlement.

The value of lymph node imprint cytodiagnosis: an assessment of interobserver agreement and diagnostic accuracy.

The aim of this study was to assess the reliability of cytodiagnosis of lymph node imprints without fixed tissue sections. One hundred randomly selected archival cases were used in the study. These air-dried May-Grünwald-Giemsa imprint slides were assessed independently and blind by three pathologists. Cases were assigned to one of four diagnostic categories: reactive changes, non-Hodgkin's lymphoma (NHL), Hodgkin's disease (HD) and secondary malignancy. Each broad diagnosis was compared with the 'correct' reviewed histological diagnosis to calculate interobserver agreement and diagnostic accuracy. The overall kappa score (+0.59) was indicative of moderate agreement. The mean pathologist diagnostic accuracy was 78%, with complete agreement with the histological diagnosis in 61% of cases. The main diagnostic difficulties were in the distinction between reactive changes and NHL and distinguishing NHL from HD. Further diagnostic classification, e.g. typing of lymphomas and subclassification of Hodgkin's disease, was not found to be reliable using the imprints alone. With these limitations in mind, pathologists should be able to use lymph node imprints for cytodiagnosis in selected cases. The study also emphasized the utility of imprints as a corollary to the histology and as a tool for cytology training and continuing education.

Primary sarcomas of the lung: a clinicopathological and immunohistochemical study of 14 cases.

The clinicopathological features and immunohistochemical findings in 14 primary sarcomas of the lung collected over a 30-year-period are presented. This represents one sarcoma per 550 bronchogenic carcinomas undergoing resection in this centre. The study group comprised six leiomyosarcomas, five malignant peripheral nerve sheath tumours, two haemangiopericytomas and one epithelioid haemangioendothelioma. The majority of cases occurred in men (nine males: five females), with mean age at presentation of 54 years for men and 47 years for women. All leiomyosarcomas were seen in men, whereas malignant peripheral nerve sheath tumours showed no particular sex preponderance. Leiomyosarcomas were larger tumours than malignant peripheral nerve sheath tumours, mean tumour diameter 15 cm (range 10-25 cm) compared to 9.5 cm (7-15 cm), respectively. All leiomyosarcomas were situated intraparenchymally whereas two of the five malignant peripheral nerve sheath tumours were endobronchial in site. Extrathoracic metastates were seen at death in two of the six leiomyosarcomas but not in any of the malignant peripheral nerve sheath tumours. Overall survival was 28 months although for the leiomyosarcoma/malignant peripheral nerve sheath tumour group alone survival was 8 months. Tumour grading appeared to be a more useful prognostic factor than tumour site (endobronchial/parenchymal) or tumour size. Haemangiopericytoma and epithelioid haemangioendothelioma were associated with a more favourable prognosis.

Thrombomodulin as a marker of vascular and lymphatic tumours.

The aim of this study was to evaluate the utility of a new commercially available antibody to thrombomodulin as an endothelial marker in formalin-fixed paraffin-embedded tissue. The expression of thrombomodulin in a variety of 50 vascular and lymphatic neoplasms and malformations was compared to the expression of von Willebrand factor, QBend 10 (CD34) and JC70 (CD31). We showed that thrombomodulin was the best marker of lymphatic endothelium and also stained a higher percentage of malignant vasoformative tumours when compared to the other markers. We recommend the assessment of thrombomodulin expression in the differential diagnosis of malignant vasoformative neoplasms and in the detection of lymphatic endothelium for evidence of tumour permeation.

Mesothelioma-binding antibodies: thrombomodulin, OV 632 and HBME-1 and their use in the diagnosis of malignant mesothelioma.

The aim of this study was to examine the expression of three putative mesothelioma-binding antibodies, thrombomodulin, OV 632 and HBME-1 in 42 malignant mesotheliomas (27 pleural and 15 peritoneal) and 32 pulmonary adenocarcinomas. Evaluation of their use in differentiating between the mesotheliomas and pulmonary adenocarcinomas was assessed. Thrombomodulin was expressed by 22 of 42 (52%) mesotheliomas but was seen in eight of 12 pure epithelial-type mesotheliomas of the pleura and in all four papillary epithelial peritoneal mesotheliomas. For pure epithelial mesotheliomas thrombomodulin was 75% sensitive. Only two of 32 pulmonary adenocarcinomas were immunoreactive yielding a 94% specificity for thrombomodulin. In comparison, OV 632 and HBME-1 showed 67% and 62% antibody sensitivity, respectively, for malignant mesothelioma but this was accompanied by low specificity (OV 632, 37%; HBME-1, 28%). Both OV 632 and HBME-1 are considered unsuitable for use in differentiating between mesotheliomas and pulmonary adenocarcinomas. We advocate the use of thrombomodulin as a mesothelioma-binding antibody in the standard panel of antibodies used in the evaluation of malignant mesothelioma.

CD44H expression in reactive mesothelium, pleural mesothelioma and pulmonary adenocarcinoma.

Malignant mesotheliomas are known to produce hyaluronic acid, in contrast to most pulmonary adenocarcinomas which produce neutral mucin. CD44H is the major cell surface receptor for hyaluronic acid. The aim of this study was to investigate immunohistochemically the expression of this antigen in reactive mesothelium, pleural mesothelioma and pulmonary adenocarcinoma and to assess its diagnostic utility in distinguishing the two tumours. Diffuse and intense membranous CD44H immunoreactivity was seen in 15 of 20 (75%) mesotheliomas and in all 20 biopsies of reactive mesothelium. In contrast, focal (< 10% tumour) expression of CD44H was seen in only three of 20 (15%) pulmonary adenocarcinomas. We advocate the use of CD44H as a positive mesothelial marker for incorporation alongside other established immunohistochemical markers used to distinguish mesothelioma from adenocarcinoma.

Anti-mesothelial markers in sarcomatoid mesothelioma and other spindle cell neoplasms.

AIMS: To undertake a comparative evaluation of three antimesothelial markers (thrombomodulin, cytokeratin 5/6 and calretinin) with broad spectrum cytokeratin (AE1/AE3) in differentiating between sarcomatoid mesothelioma and a spectrum of spindle cell neoplasms. METHODS AND RESULTS: Thirty-one malignant sarcomatoid mesotheliomas were studied. Calretinin expression was focally identified in 12 (39%) tumours and thrombomodulin and cytokeratin 5/6 immunoreactivity was seen in nine (29%) cases. In comparison there was strong diffuse cytoplasmic reactivity with the broad spectrum cytokeratin (AE1/AE3) in 24 of 31 (77%) tumours. Thirty mixed spindle cells neoplasms were studied. No calretinin expression was identified in any case. Thrombomodulin immunoreactivity was identified in four (16%) cases (two angiosarcomas, two high-grade sarcomas, not otherwise specified). Cytokeratin 5/6 expression was seen in one high-grade pulmonary sarcoma originally termed malignant fibrous histiocytoma. None of the antimesothelial markers was expressed in the four spindle cell carcinomas studied. In contrast, broad spectrum cytokeratin was diffusely expressed in all four spindle cell carcinomas (three pulmonary, one renal), both synovial sarcomas, both malignant mixed Müllerian tumours, one of three pulmonary leiomyosarcomas and two of nine sarcomas, not otherwise specified. CONCLUSIONS: Immunohistochemistry has a more limited role in the diagnosis and distinction of sarcomatoid mesothelioma from other spindle cell neoplasms. The combination of a broad spectrum cytokeratin with calretinin combines both high sensitivity (77% for AE1/AE3) with high specificity (100% for calretinin) for sarcomatoid mesothelioma and can be diagnostically useful. The mesothelial markers, thrombomodulin and cytokeratin 5/6, are not useful alone in the diagnosis of sarcomatoid mesothelioma as each shows insufficient antibody sensitivity, although together they complement calretinin.