[Primary central nervous system lymphomas. Diagnostic and prognostic effect of steroid-induced remission].
PURPOSE: Non-AIDS primary central nervous system lymphomas may respond totally or partially to corticosteroids. These corticoid-induced remissions seems to be very specific for this disease. They have been proposed as diagnostic test. The effect of these remissions on prognosis remains unknown. METHODS: A retrospective study was conducted. Corticosteroid sensibility, duration of survival, duration of disease free interval and type of treatment were compared between two groups of patients. The first one (group 1) included 44 patients in which exact diagnosis was made by stereotactic biopsy or surgery. The second group (group 2) included 5 patients exhibiting typical neuroradiological aspects of primary lymphoma in whom corticosteroid therapy produced a total regression of the lesions. RESULTS: In group 1, 29.1% of the lymphomas exhibited cortico-sensitivity. Duration of free interval of disease accounted for 75% of the overall duration of survival. Age was the only significant factor predicting remission (p = 0.019). The sole factor influencing total duration of remission was the type of treatment (p = 0.03). Duration of remission was significantly shorter in group 2 versus group 1 patients (p = 0.007). CONCLUSIONS: Duration of the first remission is of paramount importance on survival as well as the quality of first line therapy. In these conditions the absence of precise diagnosis due to corticoid-induced remissions could be dramatically deleterious.