BACKGROUND: The testes are one of the most common extramedullary sites of relapse in boys with acute lymphoblastic leukemia(ALL). The reported incidence of isolated testicular relapse varies from 3 to 40%. If these patients are treated exclusively with testicular irradialion, a systemic relapse occurs within a few months. Recently, the use of intensive chemotherapy and testicular irradiation improved the survival rate for boys with testicular leukemia. So, we performed this study to identify clinical manifestations, disease free survival and prognostic factors of testicular leukemia in children. METHODS: We reviewed 33 patients of testicular leukemia among total 410 boys with ALL diagnosed at the Department of Pediatrics, Seoul National University Children's Hospital from Jan. 1970 to Aug. 1996. Testicular leukemia was confirmed by testicular biopsy in all 33 patients. These patients were treated with combined local testicular irradiation(2,400~2,500 cGy/8~12fractions) and systemic chemotherapy. Two patients, in whom testicular relapse was diagnosed before 1979, unilateral orchiectomy of the involved site and testicular irradiation of the opposite site were performed. Probability estimates of disease free survival (DFS) were calculated by the method of Kaplan and Meier, and the relationship of prognostic factors to DFS was compared using the chi-square test in survival analysis. RESULTS: In 410 boys with ALL, testicular leukemia occurred in 33 patients(8%). Of 33 patients, 6 patients presented with testicular involvement at initial diagnosis, 16 patients had testicular relapse while still receiving chemotherapy and 11 patients had testicular relapse 3 to 57 months(median : 15 months) after cessation of chemotherapy. The median age of 33 patients was 7.4 yrs(9 months~18 yrs) and median WBC count 7,600/ L(2,700~270,000/L). All patients presented with painless testicular enlargement and testicular leukemia was confirmed by testicular biopsy. Among 33 patients, 2 had prior CNS relapse and 11 had concomitant bone marrow and/or CNS relapse. Twenty nine patients were treated with combined local testicular irradiation and systemic chemotherapy. Eleven had second relapse(6 bone marrow, 3 CNS, 2 opposite testis). Seventeen have been followed until now: 6 patients on chemotherapy and 11 patients(37.9%) in complete remission for 48.5+/-22.3 months(19~86 months). The 3 year DFS for 29 patients was 55.3%+/-10.1%. The following prognostic factors showed no significant association with DFS in testicular relapse : age and WBC count at initial diagnosis, age at testicular relapse, and concomitant relapse. Whether testicular relapse occurred on initial therapy or off initial therapy has prognostic value in predicting DFS. The 3 year DFS for boys with testicular relapse on and off initial therapy were 40.0%+/-12.9% and 78.8%+/-13.4%, respectively(P: 0.046). CONCLUSION: With the use of chemotherapy and testicular irradiation, prolonged second re mission can be achieved in many patients with testicular leukemia. The patients with testicular relapse off initial therapy fared significantly better than patients on therapy. So, to improve the DFS for boys with testicular leukemia, a better understanding of its biology and prognostic factors is needed.