PURPOSE: Interleukin-2 (IL-2) exerts anti-cancer effect by increasing NK cell activity when the tumor burden is low. Earlier study conducted with high dose intravenous IL-2 exhibited significant toxicities such as capillary leak syndrome, fever, rash, etc. This study was designed to study the effect of low dose IL-2 in children after autologous PBSCT when the cancer is at minimal level. METHODS: A total of 12 patients (6 AML, 6 NBL) were enrolled in this study from May 1997 to Oct 1999. The age of the patients was between 0.9~15 yr (Median age: 4.35 yr). The AML patients were treated with AML-BFM-87 (5 cases) or CCG-2891 (1 case) protocol, and all the patients underwent autologous PBSCT at CR1. The NBL patients were treated with CCG-3891 (4 cases) or '6 in 1' (2 cases) protocol, and they had operation for residual tumor before PBSCT. The conditioning regimen for AML patients was busulfan (16 mg/kg) and cyclophosphamide (120 mg/kg) (4 cases) or BCVAC (2 cases), while NBL patients were conditioned with carboplatin (1200 mg/m2), etoposide (800 mg/m2) and melphalan (180 mg/m2). Infused stem cell dose was MNC: (4.5+/-1.7) 108/kg, CD34 : (8.6+/-3.2) 106/kg. IL-2 (Proleukin , Chiron) was started subcutaneously after neutrophil engraftment (ANC<500/mm3) with the dose of 3~5 MU/m2 for the first 2 days, 1MU/m2 for the subsequent 12 days, then followed by 14 days of rest. IL-2 was restarted with the same regimen for more than 6 cycles as outpatient. The CBC, total eosinophil count (TEC) and T lymphocyte subsets were checked before and after IL-2 therapy. RESULTS: The mean neutrophil engraftment was achieved on 12.0+/-3.4 days, and mean platelet recovery to more than 50,000/mm3 was achieved on 23.7+/-10.3 days. Common toxicities associated with IL-2 were fever and mild tenderness on injection site, but there was no need to discontinue IL-2. A total of 75 cycles of IL-2 therapy was given. During follow-up for 8~30 months (median 21 months), only 1 relapse occurred until now (neuroblastoma stage IV). All parameters of T lymphocyte subsets increased after IL-2 therapy. TEC increased in mean value after IL-2 and it was statistically significant (P<0.05). The absolute count of CD4 and CD8 was significantly increased (CD4 : 410 to 640, P<0.005, CD8 : 720 to 980, P<0.05). CD4/CD8 ratio remained reversed (<1) throughout the course of IL-2 in most patients. The total NK cell count was increased from 510 to 820 (P<0.005). CONCLUSION: Low dose IL-2 therapy was well tolerated as OPD basis and there was a significant change in T lymphocyte subsets, especially in NK cell count. Even though the follow up duration was short, the high relapse free survival indicates the beneficial effect of low dose IL-2. In the setting of low tumor burden, such as after autologous PBSCT, low dose subcutaneous IL-2 seems to provide effective anti-cancer effect.