Clinico-pathological correlation of asbestos-related disease in ex-miners

Introduction
It is important to make accurate diagnoses of asbestos-related diseases (ARDs) not only because they cause respiratory impairment and death, but also for compensation purposes. In the clinical setting, diagnoses are defensibly made without the aid of pathology. American studies have reported that 44% of chest radiographs in biopsy-proven asbestosis were read as negative. Recent clinico-pathological studies are sparse.

Methods
We compared the in-life clinical, and post-mortem pathological, diagnoses of asbestosis, mesothelioma and lung cancer. Consecutive autopsies of all deceased Asbestos Relief Trust claimants whose cardio-respiratory organs were submitted to the National Institute for Occupational Health for examination and reported on from May 2010 to May 2011 were studied. Sensitivities, specificities and related values were calculated.

Results
The 94 cases coming to autopsy had been assessed in-life using chest radiographs and, in known malignancies, biopsies. Pathological ARDs were diagnosed at autopsy in 78 (83%) of the cases: 47 (50%) asbestosis cases, 20 (21%) mesotheliomas and 13 (14%) lung cancers.

Sensitivity, specificity and accuracy rates for the clinical diagnoses were 47%, 83% and 65% for asbestosis; 65%, 96% and 89% for mesothelioma and 40%, 100% and 90% for lung cancer respectively. Using an ILO grading of 1/0 and above for the clinical diagnosis of asbestosis, there were 25 (53%) false negative cases. Of the mesotheliomas, 3 had been misdiagnosed as lung cancer and 7 cases had been missed in life. No false positive diagnoses of lung cancer were made but 9 (60%) of the cases had not been diagnosed in life.

Discussion
The accuracy of clinical assessment for asbestosis was 65% and ~90% for malignancies. Despite prior contact with the Trust, 16 cases of malignancy went undiagnosed in life. These findings underline the importance of the autopsy service.