Empiric prednisone therapy for pulmonary toxic reaction after high-dose chemotherapy containing carmustine (BCNU).

STUDY OBJECTIVE To determine pretreatment factors that predict for pulmonary toxic reactions after high-dose chemotherapy containing carmustine (BCNU) and to determine the utility of prednisone in preventing pulmonary toxic reactions. DESIGN Retrospective review. SETTING Tertiary care referral center. PATIENTS Forty-five patients with relapsed or refractory lymphoma and 27 patients with breast cancer with normal cardiopulmonary function were treated with one of two high-dose combination chemotherapeutic regimens containing the same dose of BCNU. MEASUREMENTS Recorded pretreatment patient characteristics included previous chemotherapy or radiation therapy, history of pulmonary metastases, history of chronic obstructive pulmonary disease, and history of smoking. Spirometry and single-breath carbon monoxide diffusing capacity (DCO) were obtained before and after high-dose chemotherapy. INTERVENTIONS Patients were treated with prednisone for a 5% or more drop in postchemotherapy DCO whether or not symptoms were present. RESULTS Fifty-nine patients were evaluable. No pretreatment characteristic predicted for declines in pulmonary function postchemotherapy. The FEV1/FVC ratio did not change significantly after high-dose chemotherapy, but the DCO decreased 12.1% (p < 0.001). Of the 59 evaluable patients, 30 were treated with prednisone for declines in postchemotherapy DCO. Sixteen (53%) of these 30 patients were asymptomatic. The DCO increased 10.3% in patients treated with prednisone compared with a decrease of 2.3% in patients not treated (p = 0.017). There was no statistically significant difference in FEV1/FVC in patients treated with prednisone compared with those not treated. Regression analysis of pretreatment characteristics, type of high-dose chemotherapy received, and treatment with prednisone identified only treatment with prednisone as a significant variable in predicting an increase in DCO (p = 0.03; regression coefficient = +11.5%, SE = +/- 5.2%) after high-dose chemotherapy containing BCNU. CONCLUSIONS High-dose BCNU-containing chemotherapeutic regimens cause decreases in DCO that are often asymptomatic and likely represent subclinical pulmonary toxic reactions. Pretreatment clinical parameters cannot predict which patients will manifest pulmonary toxic reactions after high-dose chemotherapy. Empiric treatment with prednisone will reverse chemotherapy-induced decreases in DCO. Earlier institution of glucocorticoids for evidence of pulmonary dysfunction is recommended.