Pathology analysis for mesothelioma study in the United Kingdom: Current practice and historical development

Following up on the largest case-control study of malignant mesothelioma yet performed, investigators at the London School of Hygiene and Tropical Medicine assessed 1732 male and 670 female cases as of May 2013. Epidemiological findings of a subset of these were published previously, excluding patients who died or who refused to be interviewed. Pathology reports were collected for subjects, including those both eligible and ineligible for epidemiology study based on vital status. The current investigation examined 860 cases having pathology reports available. Sixty-one cases were diagnosed using cytology only, often with equivocal diagnoses, while 799 reported at least a biopsy of the tumor. Of these, 748 had pathology sufficiently detailed for evaluation. These reports were examined for basis of diagnosis, differences between study cases and ineligible cases, pathology characteristics, and immunohistochemical and other tests used. The most prominent subtype was epithelioid (64% of study cases but only 49% of ineligible cases). Biphasic subtype was present in 10% of study cases and 16% of those ineligible. Sarcomatoid subtype was present in 7% of study cases and 19% of ineligible cases, most of whom died. Twelve percent of study cases displayed no specified subtype, versus 7% of ineligible cases. Of recorded immunohistochemical stains specific for mesothelial cell origin, calretinin (95%) and CK 5/6 or CK5 alone (84%) were by far the most common. Calretinin and CK 5/6 or CK 5 alone were also most sensitive and positive in 92% of cases presenting with surgical pathology report. Ninety percent of cases had at least one immunohistochemical marker for possible lung carcinoma applied, with BER-Ep4 and TTF-1 the most frequent at 68% and CEA at 58%. TTF-1 and CEA were positive in 1% or less of cases. Patterns of use and positive and negative results for each of these as well as other immunohistochemical stains are presented and discussed, along with a brief historical description of their development and use. Possible effects of the pathologic analysis on the results of previously published and future epidemiological studies are discussed.