The pathology diagnosis of mesothelioma can be difficult. Invasion of the stroma continues to be the most reliable criterion for separating benign from malignant mesothelial proliferations. Cytology-only diagnosis on effusion fluids cannot assess invasion and remains controversial, but it can sometimes suggest a diagnosis of epithelioid mesothelioma in conjunction with clinico-radiological and immunohistochemical features. However, cytology diagnosis of sarcomatoid and desmoplastic mesothelioma is rarely achievable. Because usually only the epithelioid component sheds cells, biphasic tumours may be falsely interpreted as epithelioid, which may affect eligibility for surgical treatment. This could also affect a patient’s inclusion into clinical trials, and prognostication.Emerging data suggests that subtyping of epithelioid mesothelioma with regards to architectural features (which require a biopsy for assessment) and nuclear grade are also important to predicting survival. This suggest that a cytology diagnosis of ‘malignant mesothelioma’ (presumably epithelioid in type) might not suffice in the future.
Sensitivity of cytology diagnosis is reportedly variable and (potentially low) - in the range of 16-75% for a 'positive' diagnosis after exclusion of 'suspicious' findings. Studies reporting high sensitivities usually originate from institutions that have a special interest in that area, and some (retrospective) studies did not specify the diagnostic criteria used, or the extent of clinical and radiological information available. Such diagnostic criteria have recently been proposed by Hjerpe et al, but in view of the rarity of mesothelioma, and the variable features at unusual sites, e.g. pericardial, many laboratories would struggle to accumulate sufficient experience.
Immunohistochemical and molecular studies (e.g. BAP1, p16 and 9p21 homozygous deletion) can be helpful but are not universally available, and some of these need further validation.
Whilst there is a requirement for minimal fluid volumes to be submitted for a cell block to be prepared, it is not uncommon that insufficient material is received, especially where the diagnosis is unexpected.
Even if a cell block is available, assessment of prognostic and predictive markers on effusion fluids, whilst possible in principle, has not been widely validated and may be challenging, due to the mixture of reactive benign and malignant cells, which cannot be definitively distinguished by morphology.
When all of the recently proposed cytological criteria for epithelioid mesothelioma are fulfilled in an effusion fluid specimen, a highly probable diagnosis of mesothelioma can be accomplished with a high level of specificity. However, many cases are problematic, and the information that can be gathered is limited. It is appealing to refrain from an invasive procedure, but uncertainty of diagnosis, potential misdiagnosis of morphological subtype, inability to sub-classify epithelioid mesothelioma or diagnose sarcomatoid mesothelioma, lack of data validating predictive and prognostic markers on cytology samples alone and potential delay in treatment are compelling reasons to biopsy when possible.