The role of prophylactic cranial irradiation in the management of small cell lung cancer

Abstract

Those patients who benefit most in terms of increased survival from the current treatment of small cell lung cancer (SCLC) are also those who run the greatest risk of developing brain metastases. In a compilation of 11 studies with 1688 patients, Pedersen (1) calculated that about 10% had central nervous system (CNS) metastases at diagnosis of SCLC, while a further 20% (range 7-30%) developed CNS metastases during therapy. At autopsy the frequency was found to be about 50% (1, 2). An increase in frequency of brain metastases in SCLC with increased survival was predicted as early as 1973 (3), and the prediction was confirmed in 1979 by Nugent et al. (4), who demonstrated an actuarial probability of developing brain metastases of 80% 2 years after diagnosis. Similar tendencies were subsequently demonstrated by others (5), and in a series from the Finsen Institute in Copenhagen the cumulative risk of brain metastases reached 50% at 2 years survival (1). An early and widely disseminated hypothesis explaining the increase in CNS metastases in patients with prolonged survival describe the brain as a pharmacologic sanctuary, where metastases can grow shielded from cytostatic agents by the blood-brain-barrier (BBB). This hypothesis has to some degree been corroborated by the clinical experience with childhood leukemia, where prophylactic cranial irradiation (PCI) is an integrated part of most treatment programmes. When CNS metastases are diagnosed by intravenous iodine contrast-enhanced CT-scan, the BBB is not intact in the tumor areas, and such areas therefore should be accessible to drugs according to the same principles that govern extracranial metastases. During the last decade, several reports about the effect of systemic chemotherapy on brain metastases from SCLC have been published, see (6, 7) for a review. Thus, the concept of the brain as a pharmacologic sanctuary site for metastases is inconsistent with newer clinical observations. In spite of the fact that clinical brain metastasis is an ominous prognostic indicator, it is disputable what influence the presence of brain metastases per se, if actively