Does cranial irradiation reduce the risk for bone marrow relapse in acute myelogenous leukemia? Unexpected results of the Childhood Acute Myelogenous Leukemia Study BFM-87.

PURPOSE One of the goals of study AMA-BFM-87 was to test prospectively in acute myelogenous leukemia (AML) patients if cranial irradiation could be replaced by late intensification therapy with high-dose cytarabine (Ara-C) and etoposide (VP-16). PATIENTS AND METHODS Patients with a low risk of CNS relapses (ie, no initial CNS disease, WBC count at diagnosis < or = 70.000/microL) were randomized for irradiation (group A, 31 patients). In 25 patients (group B), randomization was refused. As interim results showed no increase of CNS relapses in nonirradiated patients, prophylactic irradiation was discontinued after 2 1/2 years to prevent unnecessary CNS toxicity. Forty-four patients (group C) entered the study after randomization had been stopped. RESULTS In all patients with a low risk of CNS recurrences (n = 100), a significantly higher probability of relapse-free interval (pRFI) of 5 years was found in irradiated patients (pRFI = .78) compared with nonirradiated patients (pRFI = .41) (P = .007). Moreover, a slightly higher incidence of CNS relapses was observed in nonirradiated patients. Due to the small number of patients, this was not observed when randomized patients only were analyzed. In accordance with these findings, the favorable outcome of low-risk patients in the preceding study, AML-BFM-83 (pRFI > .80), could only be reproduced in study AML-BFM-87 in patients who had received cranial irradiation. CONCLUSION These results indicate that cranial irradiation should be an integral part of the treatment of all AML patients not undergoing bone marrow transplantation. Residual blasts in the CNS may escape systemic chemotherapy and lead to recurrence of the initial disease not only in the CNS, but also in the bone marrow.