SUMMARY A method is described which allows the features seen in images generated during computed tomography (CT) of lungs previously removed at necropsy to be compared with those seen in corresponding thin sections made from the same lungs. After removal from the body, lungs were inflation-fixed using warm formalin vapour before being x-rayed and then scanned in the inflated state. Slices corresponding to the CT scan images were made and x-rayed. Paper mounted sections were then prepared from each slice. Using these methods pathological correlative studies can be used both to validate the interpretation of CT scans of lungs and to assess the sensitivity of this imaging technique.Computed tomography (CT) is increasingly used in the investigation of lung disease.-3 Confidence in the interpretation of the images generated can only be acquired by adequate correlative studies. In this hospital three approaches are currently under intensive exploration. Firstly, we are comparing the iesults of CT scanning with the results of lung function tests.4 5 Secondly, we are documenting the appearances of standardised phantoms used to simulate lung structures.6 Thirdly, we are comparing scan images with the detailed structure of both normal and known pathological lesions in lungs removed at necropsy. This paper presents the experience gained by the third approach, which provides information about the validity of interpretation of lung pathology as seen in the images generated by computed tomography of lungs.

A case is described of a primary cerebral lymphoma which presented with bilateral cerebellopontine angle lesions. No similar case has been reported in the English language in the last 25 years. The literature is reviewed and discussed.

Oestrogen receptor immunocytochemical assay (ER-ICA) was used to determine oestrogen receptor (ER) content of cells in fine needle aspirate (FNA) specimens from 88 breast carcinomas. In 49 of these the radioligand binding assay for oestradiol was available for comparison. The predictive value of ER-ICA staining for a positive radioligand binding assay (greater than 10 fmol/mg protein) was 95%. Although the predictive value of negative staining was only 66%, 34 out of 37 ER-ICA negative tumours had radioligand binding assays below 60 fmol/mg protein. ER-ICA staining showed a strong positive correlation with age of the patient, positivity being rare before the menopause. There was a weak inverse correlation with tumour grade but none with tumour size or lymph node status. The assessment of ER by immunocytochemistry using FNA cytology is a rapid technique, which may easily be repeated and provides a pre-operative assessment of ER status. It allows confirmation that tumour cells are present in the sample and an assessment of tumour heterogeneity.

Oestrogen receptor (ER) status of 77 cases of screen-detected breast cancer has been determined using cytological preparations. In 48% ER status was positive, which was the same proportion as that formed in a control group of age-matched patients with symptomatic breast carcinoma. Since the screen-detected group contained more low grade tumours, the percentage of ER-positive cases would be expected to be higher. The reasons for the discrepancy are discussed. Ki67 score has been determined for 41 cases of screen-detected cancer. Ki67 score showed a positive correlation with histological tumour grade and a negative correlation with ER status. However, there was no correlation with tumour size or lymph node status. The Ki67 scores in the screen-detected cancers were essentially similar to those found in an age-matched symptomatic group, but the very low scores were only found in the screened group.