Brain CT-scanning and neurological examination in small cell bronchogenic carcinoma.

A brain CT-scan and a neurological examination were performed on forty-nine consecutive patients with small cell bronchogenic carcinoma before the start of chemotherapy and every three months thereafter. Contrast-enhancement was used in 90% of the CT-scans. Ninety percent of the neurologic examinations were performed by the same neurologist. No prophylactic cranial irradiation was given, and cranial irradiation was withheld if a CT-scan indicated metastases, unless the patient was symptomatic. Thirty-five patients are evaluable including 34 with a brain autopsy. CNS-metastases were found in 18 patients. Two of them had not been examined within three months of autopsy and are excluded from the calculations of diagnostic accuracy. Of the remaining 16 patients 10 had a correct diagnosis of cerebral metastases made by CT-scan, while the neurologist made 11 correct diagnoses of CNS-metastases. Seventeen patients did not have CNS-metastases including one patient alive and free of disease. Fifteen and 11 were judged to be free of metastases by the CT-scan and the neurologist, respectively. Two patients with negative autopsies had positive CT-scans turning negative at subsequent examinations. Two had positive CT-scans which became negative and at autopsy CNS-metastases were located at different sites from those initially indicated by CT. The positive predictive value of CT-scan was 71%, while the negative predictive value was 71%. In conclusion, routine CT-scan and neurological examination are equally sensitive but have low yields when there is no clinical suspicion of CNS-metastases. The predictive value of CT-scanning could possibly be higher with the newer generations of equipment.