Acute lymphoblastic leukemic cells with T (thymus-derived) lymphocyte markers.

Five of nine children with acute lymphoblastic leukemia had lymphoblasts that bound sheep erythrocytes or reacted with antiserum to thymocytes, suggesting involvement of T (thymus-derived) cells. When lymphoblasts from all patients were examined by immunofluorescence they were found to lack a marker for B (bone marrow or bursa-equivalent) cells, that is, the presence of surface immunoglobulins.

Immunotherapy and chemotherapy in children with neuroblastoma.

Recent advances with immunotherapy in animal tumors suggested that trials with a combination of chemotherapy and immunotherapy in human malignant tumors might be worthwhile. A pilot program with Vibrio cholera neuraminidase-treated tumor cells plus BCG was tested in 3 patients who had had chemotherapy for disseminated neuroblastoma. Two of these children were in "complete remission" after radiation therapy and chemotherapy before the administration of immunotherapy. Relapse occurred in 5-6 months in all 3 patients. These disappointing results are discussed in relation to problems of current chemotherapy in disseminated neuroblastoma including results obtained at second-look operations in patients obtaining "complete remission."

Effective dose of L-asparaginase for induction of remission in previously treated children with acute lymphocytic leukemia: a report from Childrens Cancer Study Group.

L-Asparaginase, in the dose of greater than or equal to 6000 IU/sq m three times weekly, was demonstrated to be an effective agent in reinduction of remissions in childhood leukemia. Four hundred thirteen children with acute lymphocytic leukemia were treated with L-asparaginase. Doses i.m. ranged from 300 to 12,000 IU/sq m. None of the patients had received prior asparaginase therapy. 6-Mercaptopurine was given p.o. concurrently. All of the patients had experienced several previous relapses, and their disease was not responsive to 6-mercaptopurine. L-Asparaginase was found to be effective in reinducing remissions at the following rates: 9.5% for 300 IU/sq m; 35.1% for 3,000 IU/sq m; 53.5% for 6,000 IU/sq m; and 62.5% for 12,000 IU/sq m. The drug was given three times weekly for four weeks. Hypersensitivity reactions occurred in 6.5% of patients.

Cytogenetic studies of long-term survivors of childhood acute lymphoblastic leukemia.

Ten long-term survivors of childhood acute lymphoblastic leukemia were studied to determine if cytogenetic abnormalities were present in lymphocytes following discontinuation of therapy. The study included patients diagnosed between 1969 and 1974 who had received radiation therapy and a minimum of 3 years systemic chemotherapy. At study, the patients had been off all therapy from 1.3 to 6.5 years (median, 4 years). Peripheral blood lymphocytes were examined for spontaneous chromosome breakage and sister chromatid exchanges. In addition, G-banded metaphase and prometaphase chromosomes were analyzed. Chromosome breakage was found to be within normal limits for all patients. Likewise, there was no significant difference between patients and controls with respect to sister chromatid exchange frequency. However, seven of the ten patients were found to have one or more cells with nonclonal karyotypic abnormalities. Our results indicate that although long-term survivors of childhood acute lymphoblastic leukemia treated with intensive radiation and combination chemotherapy do not demonstrate chromosome instability or DNA damage as measured by breakage and sister chromatid exchange, a majority of these patients have a subpopulation of lymphocytes with nonclonal chromosome abnormalities years after stopping therapy.

Occupational exposures of parents of children with acute nonlymphocytic leukemia: a report from the Childrens Cancer Study Group.

The Childrens Cancer Study Group conducted a case-control study of occupational exposures of parents of 204 children (under 18 yr of age) with acute nonlymphoblastic leukemia. The most consistent finding was an association of acute nonlymphoblastic leukemia risk with pesticide exposure. Controls matched by date of birth and race were obtained through random digit dialing. Odds ratio (OR) for paternal pesticide exposure in jobs held for longer than 1000 days was 2.7 (95% confidence interval, 1.0 to 7.0; trend, P = 0.06), and seven case mothers and no control mothers had prolonged exposure (trend, P = 0.008). Risk estimates for parental pesticide exposure were substantially increased for children under age 6 at diagnosis (OR for prolonged exposure to either parent = 11.4; trend, P = 0.003) and for those with myelomonocytic and monocytic subtypes (OR, 13.6; trend, P = 0.007). Moreover, there were significantly elevated risks for direct exposure of the child to pesticides in the household (OR for exposure most days = 3.5; trend, P = 0.04) and for maternal exposure to household pesticides at the time of pregnancy (eight case mothers versus no controls for exposure most days; trend, P = 0.05). Paternal exposures to solvents (OR, 2.1; P = 0.003) and petroleum products (OR, 2.4; P = 0.002) were reported more commonly for cases than controls. Other occupational exposures reported significantly more often by case parents were paternal exposure to plastics or lead and maternal exposure to paints and pigments, metal dusts, and sawdust. These data provide further evidence for a role of occupational risk factors in the etiology of childhood cancer.

Chromosomal abnormalities in acute lymphoblastic leukemia.

Chromosome abnormalities in acute lymphoblastic leukemia (ALL) and their possible clinical significance are briefly reviewed based upon the literature and 60 cases studied at the University of Minnesota. Almost all cases of ALL appear to demonstrate clonal abnormalities; the major abnormal clone is usually hyperdiploid or pseudodiploid. Among cases of non-T, non-B ALL, at least four translocations appear to be present with an increased frequency: t(9;22); t(4;11); t(11;14); and t(1;3). Patients with these translocations appear to have unique clinical and laboratory findings. Although the presence of abnormal clones does not seem to influence remission duration, the nature of the abnormality does. Patients whose leukemias demonstrate predominantly a pseudodiploid abnormal clone or a translocation have significantly shorter first remissions. Most importantly, among patients with non-T, non-B ALL, the presence or absence of translocations may separate poor responders from good responders.

Sanctuary therapy: a randomized trial of 724 children with previously untreated acute lymphoblastic leukemia: A Report from Children's Cancer Study Group.

Between 1972 and 1974, Childrens Cancer Study Group enrolled 724 children with newly diagnosed acute lymphoblastic leukemia on a single randomized clinical trial. Study CCG-101 was designed to test four types of presymptomatic central nervous system and sanctuary therapies consisting of (a) 2400-rad craniospinal radiation therapy (RT) plus 1200-rad extended-field RT, which included the liver, spleen, kidneys, lower abdomen, and gonads; (b) 2400-rad craniospinal RT; (c) 2400-rad cranial RT plus intrathecal methotrexate (i.t. MTX); and (d) i.t. MTX alone. Patients all received a 28-day induction course of vincristine, prednisone, and L-asparaginase and were maintained subsequently on a regimen consisting of daily 6-mercaptopurine, weekly MTX, and monthly pulses of vincristine and prednisone. Patients treated with six doses of i.t. MTX alone had a significantly higher incidence of central nervous system relapse than did patients treated with 2400-rad craniospinal RT plus 1200-rad abdominal RT, 2400-rad craniospinal RT, or 2400-rad cranial RT plus i.t. MTX. There was no significant differences in marrow remission duration or survival of the treatment groups. There appears to be a benefit with regard to length of bone marrow remission and survival for patients with initial white blood counts greater than or equal to 20,000/cu mm treated with cranial RT plus i.t. MTX. The majority of the patients remaining on study have now discontinued maintenance therapy. The 8-year overall estimated survival rate on this study is 56%, and the disease-free survival rate is 52%.

L-Asparaginase, vincristine, and prednisone for induction of first remission in acute lymphocytic leukemia.

L-Asparaginase was added to vincristine and prednisone for induction of first remission in 815 children with acute lymphocytic or acute undifferentiated leukemia. This combination resulted in an overall remission rate of 93%. The addition of L-asparaginse to the standard induction regimen using prednisone and vincristine did not significantly increase the morbidity or mortality rate during the induction period. The most common side effect was transient L-asparaginase-induced hyperglycemia. The safe administration of L-asparaginase i.m. and the dose efficacy of 6000 I.U./sq m were confirmed. For these reasons, L-asparaginase should be combined with vincristine and prednisone for the initial induction of children with acute lymphocytic or acute undifferentiated leukemia.

Use of monoclonal antibodies, morphology, and cytochemistry to probe the cellular heterogeneity of acute leukemia and lymphoma.

A combined immunological, morphological, and cytochemical approach to the study of malignant cells in patients with acute leukemia and lymphoma is presented. Newly produced monoclonal antibodies that bind to antigens of human mononuclear cells (TA-1), or B-lymphocytes (BA-1) were used to study malignant cells from patients with acute lymphoblastic leukemia (ALL). acute myelocytic leukemia, acute myelomonocytic leukemia, and chronic lymphocytic leukemia. Results in lymphoid leukemia-lymphoma patients were compared with other immunological markers and indicate that the major groups of ALL and childhood non-Hodgkin's lymphoma are T-ALL, pre-T-ALL, pre-B-ALL, B-ALL, and non-T, non-B-ALL. In addition, each major group had multiple phenotypes when analyzed with seven immunological markers including the erythrocyte rosette receptor, surface immunoglobulin, cytoplasmic immunoglobulin M, the early lymphocyte-acute lymphoblastic leukemia antigen, monoclonal antibody TA-1, monoclonal antibody BA-1, and a monoclonal antibody against HLA-DR. While immunological heterogeneity was demonstrable within each group, distinct biological behavior was observed, with T-ALL and B-ALL generally presenting as "lymphomas" and the others presenting as "leukemias." Morphological analysis using the French-American-British classification provided independent information in the definition of groups with differing clinical behavior. Cytochemical analyses demonstrated focal paranuclear staining of leukemia cells with acid phosphatase in 73% of T-ALLs and 6% of non-T, non-B-ALLs.

Randomized study of 3 years versus 5 years of chemotherapy in childhood acute lymphoblastic leukemia.

Between 1972 and 1975 the Children's Cancer Study Group conducted two clinical trials for the treatment of newly diagnosed patients with acute lymphoblastic leukemia. Upon achieving 3 yr of continuous complete remission, 316 children and young adults were randomly allocated either to discontinue chemotherapy or to continue chemotherapy for an additional 24 mo. With a median follow-up from the time of randomization of 50 mo, those patients who received 3 yr of therapy have demonstrated a statistically non-significant yet higher incidence of bone marrow relapse as compared to those patients treated for 5 yr (p = 0.09). However, the proportion of patients surviving 5 yr from randomization is 93% for the 3-yr treatment group and 89% for the 5-yr treatment group (p = 0.27). No significant difference was observed between the randomized groups for the occurrence of testicular relapse (p = 0.12), central nervous system relapse (p = 0.17), or first occurrence of relapse or death (p = 0.24). The relapse-free survival of patients treated for 5 yr as compared to those treated for 3 yr was not significantly higher in males (81% versus 75%, p = 0.14) or females (89% versus 89%, p = 0.95). This randomized study did not demonstrate a significant difference between treatment for either 3 or 5 yr.

Successful reinduction of patients with acute lymphoblastic leukemia who relapse following bone marrow transplantation.

At the present time, there is limited information on the outcome of patients with acute lymphoblastic leukemia (ALL) who relapse after bone marrow transplantation (BMT). Intuitively, it might be expected that leukemia recurring after BMT would be refractory to further treatment. In an attempt to improve survival in patients with ALL who relapse after BMT, we used standard chemotherapy for reinduction and maintenance. Of 65 patients who relapsed following allogeneic, autologous, or syngeneic BMT, 12 elected to receive no further chemotherapy, and their median survival from relapse was 36 days (range 13 to 167 days). The 53 patients who received therapy had a significantly longer median survival of 168 days (range 18 days to 4.7 years). With multidrug induction regimens there were 29 of 52 (56%) complete remissions. Six patients are currently alive, with two off therapy. In the patients who received therapy, the following factors were independent predictors of prolonged survival: longer time from BMT to relapse; younger age at diagnosis; and the use of a preparative regimen containing fractionated total body irradiation. In conclusion, leukemia recurring after BMT remains sensitive to standard therapy in many patients. We recommend that patients with ALL who relapse after BMT receive reinduction and maintenance therapy as additional good quality survival time is achieved in patients who attain a remission.

Multimodal therapy for the management of primary, nonmetastatic Ewing's sarcoma of bone: a long-term follow-up of the First Intergroup study.

A total of 342 previously untreated eligible children were entered into the first Intergroup Ewing's Sarcoma Study (IESS) between May 1973 and November 1978. In group I institutions, patients were randomized between treatment 1 (radiotherapy to primary lesion plus cyclophosphamide, vincristine, dactinomycin, and Adriamycin [doxorubicin; Adria Laboratories, Columbus, OH] [VAC plus ADR]) or treatment 2 (same as treatment 1 without ADR), and group II institutions randomized patients between treatment 2 or treatment 3 (same as treatment 2 plus bilateral pulmonary radiotherapy [VAC plus BPR]). The percentages of patients relapse-free and surviving (RFS) at 5 years for treatments 1, 2, and 3 were 60%, 24%, and 44%, respectively. There was strong statistical evidence of a significant advantage in RFS for treatment 1 (VAC plus ADR) versus 2 (VAC alone) (P less than .001) and 3 (P less than .05) and also of treatment 3 versus 2 (P less than .001). Similar significant results were observed with respect to overall survival. Patients with disease at pelvic sites have significantly poorer survival at 5 years than those with disease at nonpelvic sites (34% v 57%; P less than .001). Among pelvic cases, there was no evidence of differing survival by treatment (P = .81), but among nonpelvic cases, there was strong evidence of differing survival by treatment (P less than .001). The overall percentage of patients developing metastatic disease was 44%; the percentages by treatments 1, 2, and 3 were 30%, 72%, and 42%, respectively. The overall incidence of local recurrence was 15%, and there was no evidence that local recurrence rate differed by treatment. Patient characteristics related to prognosis, both with respect to RFS and overall survival experience, were primary site (nonpelvic patients were most favorable) and patient age (younger patients were more favorable).

Correlation of chromosome abnormalities with patient characteristics, histologic subtype, and induction success in children with acute nonlymphocytic leukemia.

Cytogenetic analyses of bone marrow cells were performed in 195 children with acute nonlymphocytic leukemia (ANLL) at diagnosis, as part of Childrens Cancer Study Group Study No. 251. Ninety-six patients (49%) exhibited clonal abnormalities, including trisomy 8 in 18 patients, t(8;21) in 11, t(15;17) in seven, loss of a sex chromosome in seven, monosomy 7 in seven, and the Philadelphia chromosome in four. Clonal abnormalities were found significantly more often in younger patients. Furthermore, recurring cytogenetic abnormalities tended to correlate with specific ages. For example, t(8;21) was associated significantly with children over four years of age, while -7 associated with overall loss of genetic material from the long arm of chromosome 7 (7q) and 11q- were associated significantly with younger children. Recurring chromosome abnormalities also correlated with specific ANLL histologic subtypes, such as t(8;21) with acute myelogenous leukemia and t(15;17) with acute promyelocytic leukemia. Presence or absence of cytogenetic abnormalities was compared with the ability of patients to achieve remission. Individuals exhibiting clonal abnormalities in bone marrow cells had an equally likely chance of achieving remission (74%) as those individuals with normal karyotypes (75%). Nonrandom chromosome abnormalities associated with a high induction success rate included +8 with a 94% induction success rate (P = .13) and t(8;21) with a 91% success rate (P = .46). Patients exhibiting the -7 abnormality associated with overall loss of 7q had a significantly less successful induction outcome, with only 28% achieving remission (P = .02); three of seven patients with t(15;17) died during induction therapy.

Long-term evaluation of a CNS prophylaxis trial--treatment comparisons and outcome after CNS relapse in childhood ALL: a report from the Childrens Cancer Study Group.

The current status of children with acute lymphoblastic leukemia (ALL) who had developed CNS disease while being treated on protocol CCG-101 was investigated. Seven hundred thirty-six eligible patients were entered into the study between June 1972 and July 1974. All children who were greater than 18 months of age were eligible for randomization to a CNS prophylaxis trial for which one regimen gave only a short course of intrathecal methotrexate (IT MTX) as prophylaxis. All other regimens included radiation therapy as prophylaxis. Current follow-up (median, greater than 10 years) shows no significant difference by standard life-table analysis for ultimate survival, although a substantial excess of CNS episodes occurred on the IT MTX regimen. Of the 675 patients who completed induction therapy and achieved remission in the study, 100 (14.8%) developed CNS disease as the first evidence of relapse. Fifty-five of these 100 had no subsequent CNS episodes. Only 17 of these 55 patients are surviving without further relapses since the CNS episode. The median time to isolated CNS relapse was 457 days. Time to the initial CNS relapse was found to be the most important factor for predicting outcome. Thirty-five of the 55 patients with isolated relapse subsequently relapsed in the bone marrow, and of these, 32 have died. Twenty patients of the 100 with CNS disease as the first evidence of relapse developed two episodes of CNS involvement and 17 of these 20 patients subsequently relapsed in the bone marrow; only one patient survived. Twenty-five patients of the 100 have shown a pattern of chronic CNS disease with multiple CNS relapses. The overall disease-free survival for the 100 patients who developed one or more relapse was only 16%. These data demonstrate that the occurrence of a CNS relapse is an indicator of poor subsequent outcome. Comparison of results of groups receiving different CNS prophylaxis required careful consideration of the entire pattern of relapses and mortality.

Allogeneic bone marrow transplantation for acute lymphoblastic leukemia in remission: prolonged survival associated with acute graft-versus-host disease.

Forty remission patients with high-risk acute lymphoblastic leukemia (ALL) underwent matched allogenic bone marrow transplantation (BMT) following preparation with cyclophosphamide and fractionated total body irradiation (TBI). As of March 1987, the median follow-up is more than 3 1/2 years. Thirteen patients are alive (11 relapse free) between 2 and 4 1/2 years post-BMT. Neither age, sex, remission number, prior extramedullary leukemia, nor WBC at diagnosis of ALL was statistically significant as a predictor of relapse-free survival. The development of acute graft-v-host disease (GVHD) in 17 patients was found, with time-dependent Cox regression analysis, to be associated with a significant reduction in post-BMT relapse risk (P = .04) and improved disease-free survival (P = .11). A prospective, randomized trial of maintenance chemotherapy with oral methotrexate and mercaptopurine did not demonstrate improvement in relapse risk or survival for those assigned maintenance chemotherapy (P = .7). These results suggest that allogeneic BMT can result in extended relapse-free survival for some patients with high-risk ALL. More effective preparative chemoradiotherapy and exploitation of the apparent graft-v-leukemia effect may be useful in future trials.

Effects of radiation on ovarian function in long-term survivors of childhood acute lymphoblastic leukemia: a report from the Childrens Cancer Study Group.

The Childrens Cancer Study Group has assessed serum follicle-stimulating hormone (FSH), luteinizing hormone (LH), and pubertal development in 97 long-term female survivors of childhood acute lymphoblastic leukemia (ALL). All patients received identical induction and maintenance therapy with either 18 or 24 Gy of radiation therapy (RT) to one of the following fields: cranial, craniospinal, or craniospinal plus 12 Gy abdominal RT including the ovaries. Thirty-six percent (35 patients) were found to have above normal levels of FSH and/or LH. The percentages of elevated values for RT fields were 93% for craniospinal plus abdominal RT, 49% for craniospinal RT, and 9% for cranial RT (P less than .001). A dose-response relationship was observed between 18 Gy and 24 Gy in females receiving only craniospinal RT (P = .01). Craniospinal plus abdominal RT and abnormal FSH/LH levels were significantly associated with lack of pubertal development and delayed onset of menses. Duration of maintenance chemotherapy was not associated with abnormal gonadotropin levels or the development of secondary sexual characteristics. Additional follow-up of this cohort is needed to establish the ultimate pubertal development and fertility of these patients.

Multimodal therapy for the management of localized Ewing's sarcoma of pelvic and sacral bones: a report from the second intergroup study.

A total of 59 eligible patients with localized Ewing's sarcoma of the pelvic and sacral bones were entered into a multimodal Intergroup Ewing's Sarcoma Study (IESS-II) (1978 to 1982) and compared with a historical control series of 68 patients entered into an earlier multimodal Intergroup Ewing's Sarcoma Study (IESS-I) (1973 to 1978). High-dose intermittent multiagent chemotherapy (vincristine, cyclophosphamide, Adriamycin [doxorubicin; Adria Laboratories, Columbus, OH], and dactinomycin) was given to all patients for 6 weeks before and for 70 weeks following local therapy. All patients who had a tumor biopsy or incomplete resection performed received a dose of 55 Gy to the tumor bed. With a median follow-up time of 5.5 years, two of 59 patients (3%) had a local recurrence, five patients (8%) had a local recurrence and metastases, and 17 patients (29%) developed metastases only. There was significant statistical evidence of an advantage in relapse-free survival (RFS) and survival (S) for patients on IESS-II versus IESS-I, P = .006 and P = .002, respectively. At 5 years, the comparison between IESS-II versus IESS-I was 55% versus 23% for RFS and 63% versus 35% for S.

Intraocular retinoblastoma group V: an analysis of prognostic factors.

A retrospective analysis of the University of Minnesota (Minneapolis) experience with retinoblastoma is presented. Seventy-five patients were diagnosed with retinoblastoma between 1958 and 1983, of which 53 (71%) had at least one Reese-Ellsworth group V eye. Nineteen group V patients and one group II patient developed extraocular disease recurrence. The cumulative actuarial rate of recurrence at 12 years was 36% for patients with group V disease. The median time from diagnosis to recurrence for unilateral patients was seven months and for bilateral patients 28 months (P = .001). Patients developing extraocular disease had a 10-year actuarial survival rate postrecurrence of 34%. The four long-term survivors of extraocular recurrences had had isolated orbital or local soft tissue recurrences only. Features of group V patients associated with extraocular recurrences were identified by univariate life table analyses. Clinical poor-risk factors included the nongenetic form of the disease (P = .03) and male sex (P = .02). Pathologic poor risk factors included rubeosis (P = .01), undifferentiated histology (P = .03), large tumor size (P = .05), and intraocular extension to the anterior segment (P = .02), retinal pigment epithelium (P = .03), choroid (P less than .001), and optic nerve beyond the lamina cribrosa (P = .02). Treatment-associated poor-risk factors included an optic nerve length of less than 5 mm removed at enucleation (P = .003). Multivariate life table analyses demonstrated the following parameters to be independent poor-prognostic factors: optic nerve length of less than 5 mm removed at enucleation (P = .001), optic nerve involvement (P = .004), and large tumor size (P = .01). These results will help to identify patients with retinoblastoma who are at greatest risk for extraocular recurrence.

Ewing's sarcoma in bones of the hands and feet: a clinicopathologic study and review of the literature.

Review of current data from the Intergroup Ewing's Sarcoma Study (IESS) shows that Ewing's sarcoma (ES) is rare in bones of the hands and feet. Only 12 of 377 evaluable patients in the first two IESS studies had a primary tumor in these small, distal bones. The age distribution was typical for that seen in patients with ES at other sites. Males were affected twice as often as females, and tumors in the bones of the feet were much more common than those in the hands. All signs and symptoms were local in distribution. As in other sites, the dominant histologic pattern was categorized as diffuse. With the exception of those patients with lesions in the calcaneus, the prognosis for disease-free survival was excellent. A literature review of cases of ES reported in bones of the hands and feet showed generally comparable results.

Late mortality experience in five-year survivors of childhood and adolescent cancer: the Childhood Cancer Survivor Study.

PURPOSE: Survivors of childhood and adolescent cancer are at risk for long-term effects of disease and treatment. The Childhood Cancer Survivor Study assessed overall and cause-specific mortality in a retrospective cohort of 20,227 5-year survivors. PATIENTS AND METHODS: Eligible subjects were individuals diagnosed with cancer (from 1970 to 1986) before the age of 21 who had survived 5 years from diagnosis. Underlying cause of death was obtained from death certificates and other sources and coded and categorized as recurrent disease, sequelae of cancer treatment, or non-cancer-related. Age and sex standardized mortality ratios (SMRs) were calculated using United States population mortality data. RESULTS: The cohort, including 208,947 person-years of follow-up, demonstrated a 10.8-fold excess in overall mortality (95% confidence interval, 10.3 to 11.3). Risk of death was statistically significantly higher in females (SMR = 18.2), individuals diagnosed with cancer before the age of 5 years (SMR = 14.0), and those with an initial diagnosis of leukemia (SMR = 15.5) or CNS tumor (SMR = 15.7). Recurrence of the original cancer was the leading cause of death among 5-year survivors, accounting for 67% of deaths. Statistically significant excess mortality rates were seen due to subsequent malignancies (SMR = 19.4), along with cardiac (SMR = 8.2), pulmonary (SMR = 9.2), and other causes (SMR = 3.3). Treatment-related associations were present for subsequent cancer mortality (radiation, alkylating agents, epipodophyllotoxins), cardiac mortality (chest irradiation, bleomycin), and other deaths (radiation, anthracyclines). No excess mortality was observed for external causes (SMR = 0.8). CONCLUSION: While recurrent disease remains a major contributor to late mortality in 5-year survivors of childhood cancer, significant excesses in mortality risk associated with treatment-related complications exist up to 25 years after the initial cancer diagnosis.