Seventy-four children ranging in age from 6 months to 17.5 years with acute lymphoblastic leukemia newly diagnosed between 1976 and 1979 were entered on a study incorporating intermittent chemotherapy with or without the addition of bacillus Calmette-Guérin (BCG). The chemotherapy program consisted of induction with vincristine, dexamethasone, and intrathecal methotrexate, intensification with adriamycin and asparaginase, central nervous system treatment with cranial irradiation and intrathecal methotrexate, and continuation treatment with 5-day courses of combination chemotherapy administered every three weeks. The first phase of continuation therapy incorporated vincristine, adriamycin, 6-mercaptopurine, and dexamethasone. In the second phase, oral methotrexate was substituted for the adriamycin in non-T-cell patients; in T-cell patients, cytosine arabinoside or cyclophosphamide and methotrexate in alternating cycles were substituted for the adriamycin and asparaginase was added. Total duration of therapy was approximately 2.5 years. Connaught BCG was administered by Heaf gun on days 8 and 15 of each 3-week cycle for the first 8 months of treatment in approximately one-third of the patients. Actuarial disease-free survival with a median follow-up of 59 months shows no difference in outcome between the BCG and non-BCG poor-risk patients. However, there is an improvement in disease-free survival of BCG-treated good- and average-risk girls (P = 0.04). While patients were actively receiving BCG there was also a trend toward the development of fewer significant infections than when patients were not receiving BCG (P = 0.85). Toxicities from BCG administration included satellite rashes, local tenderness, lymphadenopathy, secondary infection, and residual scars. Overall disease-free survival by actuarial analysis is 60% at 6 years; for patients with unfavorable prognostic features it is 40%. In this trial the addition of BCG prolonged the disease-free survival of girls with good- and average-risk prognostic features and also may have decreased the susceptibility to infection while it was being administered. However, the benefit does not appear sufficient to warrant its routine use, especially in view of the toxicities encountered.
Read full abstract