The role of prophylactic brain irradiation in limited stage small cell lung cancer: clinical, neuropsychologic, and CT sequelae.

Ninety-four patients with limited stage small cell lung cancer treated between 1981 and 1985 with a regimen including prophylactic brain irradiation (PBI) after combination chemotherapy were assessed for compliance with PBI, brain relapse, and neurologic morbidity. Seventy-seven percent of patients had PBI and of these, 22% developed brain metastases after a median time of 11 months post treatment. The brain was the apparent unique initial site of relapse in 10% of PBI cases but more commonly brain relapse was preceded or accompanied by failure at other sites, especially the chest. Brain metastases were the greatest cause of morbidity in 50% of PBI failures. Twelve of 14 PBI patients alive 2 years after treatment had oncologic, neurologic, and neuropsychological evaluation, and brain CT. All long-term survivors were capable of self care and none fulfilled diagnostic criteria for dementia, with three borderline cases. One third had pretreatment neurologic dysfunction and two thirds post treatment neurologic symptoms, most commonly recent memory loss. Fifty percent had subtle motor findings. Intellectual functioning was at the 38th percentile with most patients having an unskilled occupational history. Neuropsychologic impairment ratings were borderline in three cases and definitely impaired in seven cases. CT scans showed brain atrophy in all cases with mild progression in those having a pre-treatment baseline. Periventricular and subcortical low density lesions identical to the CT appearance of subcortical arteriosclerotic encephalopathy were seen in 82% of posttreatment CT studies, and lacunar infarcts in 54%. Neuropsychologic impairment scores and the extent of CT periventricular low density lesions were strongly associated (rank correlation 0.7, p less than .05). Overall, PBI after intensive combination chemotherapy did not induce gross dementia or neurologic dysfunction but its risk/benefit ratio is not overwhelmingly favorable, with failure to prevent brain relapse in 1/5 patients and subtle but detectable motor findings and neuropsychologic impairment in the majority.