BACKGROUND: Modern intensive chemotherapy has dramatically improved the prognosis of acute lymphoblastic leukemia in children. However, quality of life and even survival may be threatened by infection. Immunosuppression is experted due to disease itself or therapy, and sometimes, immunosuppression itself may lead to reactivation of latent viral infections in these patients. Often the viruses involved in the most severe infections suggest that patients suffer from defect in the cellular immunity. The principal defects that predispose leukemia patients to infection are defects of T cell, B cell, stem cell, complement, and macrophage. These contributing factors interact in a complex manner resulting in spectrum of problems. But these may result from a T cell defect and, in this study, 7 cell responsiveness of patients at diagnosis, remission induction, maintenance chemotherapy and after chemotherapy for leukemia has been investigated. Studies of the immune competence of patients undergoing chemotherapy for leukemia is in progress, but results are different from each other. METHOD: Between July 1994 and May 1996, seventy patients with childhood ALL were enrolled in this study. In order to expect frequency and depth of infection and prognosis, we investigated concentrations of immunoglobulins G, A, M, peripheral total lymphocyte count, 7 cell subsets, phytohemmaglutinin responsiveness, interleukln-2(IL-2), gamma-interferon(gamma-INF), and natural killer cell activity. RESULTS: 1) IgA concentrations were often markedly raised at diagnosis, and IgG, IgM concentrations both were within normal limits. During and after chemotherapy, IgA had fallen significantly but IgG, IgM are within normal limits. 2) Total lymphocyte count had fallen during chemotherapy, and returned to normal levels after chemotherapy. CD4+ T cell were markedly decresed at diagnosis, during chemotherapy and returned to normal levels after chemotherapy. 3) In vitro proliferative response of peripheral blood lymphocytes to the T cell mitogen phytohenagglutinin were impaired at diagnosis, during chemotherapy but did not returned to normal levels. 4) Interlekin-2, gamma interferon were normal levels at diagnosis, and had fallen in the induction of remission and quickly returned to normal levels with the swish to maintenance chemotherapy. But Interleukin-2 had fallen during and after chemotherapy. Natural killer cell activity had fallen at diagnosis, during chemotherapy and returned to normal levels after chemotherapy. CONCLUSION: It is assumed that evidence of impaired T cell responses is somewhat definite. These observations suggest that proliferative responses to phytohemagglutinin, CD4+ T cell, natural killer cell activity defects are due to leukemia itself but others more likely are generalizable defects caused by chemotherapy. Further investigations, however, have suggested a persisting defect in IgA, proliferative reponses to phytohemagglutinin, and interleukln-2. Our observations also show that despite normal immunoglobulin levels, most of these children have nonprotective levels for common childhood bacterial or viral disease. These results support to the praxis to withdraw prophylactic antibiotics after discontinuation of intensive chemotherapy and to start the immunization. It is expect to try to use cytokine on treatment and to improve mortality and morbidity for children of acute leukemia also.