RESUMO
Adrenocorticotrophic hormone (ACTH) concentrations were measured in the plasma and cerebrospinal fluid (CSF) of 107 consecutive patients with known or suspected central nervous system (CNS) metastases secondary to small cell carcinoma of the lung. The combined results of computerized tomography scans, neurologic examination, and autopsy were used to determine the presence or absence of CNS metastases. On the basis of such an assessment, definitive conclusions were possible in 77 patients. CNS metastases were present in 52 cases and absent in 25. The median CSF ACTH level was 30 ng/ml in both groups. None of five patients with very high CSF ACTH concentrations had elevated ACTH concentrations in plasma. Considering the 95th percentile of patients without CNS metastases as the upper limit of normal, 12 patients with metastases and one without had an elevated CSF ACTH value. Eleven patients with leptomeningeal carcinomatosis (MC) did not constitute a special subgroup in this respect. The median ratio of CSF ACTH and plasma ACTH was 1.0 in patients with CNS metastases and 0.4 in those without (P less than 0.05). Patients with MC had a median ratio of 1.3, which was significantly different from that of both of the other groups (P less than 0.05). Ten patients with CNS metastases (one with MC) and one without exceeded the upper 95th percentile of the CSF/plasma (ACTH) ratio in patients without CNS metastases. The significance levels of these findings disappeared, however, when patients with signs of an elevated ACTH concentration in plasma were excluded. Patients with ectopic ACTH production into CSF do not necessarily have ectopic ACTH production outside the CNS, despite the presence of extracerebral metastases. With the criteria employed in this study, an elevated level of CSF ACTH diagnosed too few patients for the authors to recommend its determination as a single test in diagnosing CNS metastases or MC secondary to small cell carcinoma of the lung.
