Influence of doxorubicin dose intensity on response and outcome for patients with osteogenic sarcoma and Ewing's sarcoma.

The goal of this study was to use dose-intensity analyses of published Ewing's sarcoma and osteogenic sarcoma trials to determine which agents were most closely associated with a favorable response. The percentage of patients with more than 90% tumor necrosis following neoadjuvant chemotherapy was the end point for analysis of osteogenic sarcoma trials, and disease-free survival and percentage of patients with distant-only relapse were the end points for analysis of Ewing's sarcoma trials. The data were analyzed using logistic regression analysis to circumvent the distortion of univariate analysis resulting from the correlation between doxorubicin dose intensity and the dose intensity of other agents. Our analysis suggests that doxorubicin dose intensity is an important determinant of favorable outcome for both Ewing's sarcoma and osteogenic sarcoma and that the dose intensities of other agents do not contribute as significantly to outcome as does doxorubicin dose intensity. Increasing dactinomycin dose intensity was associated with a poorer outcome in treatment of osteogenic sarcoma and Ewing's sarcoma, most likely resulting from regimens with a higher dactinomycin dose intensity having a lower doxorubicin dose intensity. While our analysis of osteogenic sarcoma trials is consistent with significant activity for cisplatin and high-dose methotrexate (and likely ifosfamide), a rank ordering of the efficacy of these agents when given with doxorubicin in multiagent regimens is not possible. Our analysis illustrates the importance of analyzing the contributions of individual agents to combination chemotherapy regimens. In the design of future clinical trials for osteogenic sarcoma and Ewing's sarcoma, careful attention should be given to optimizing doxorubicin dose intensity in regimens to be tested.

Potentiation of ifosfamide neurotoxicity, hematotoxicity, and tubular nephrotoxicity by prior cis-diamminedichloroplatinum(II) therapy.

We investigated the relationship between prior therapy and three distinct forms of toxicity that developed during ifosfamide administration (1.6 g/m2/day for 5 days) in 36 children with malignant solid tumors. Of ten therapies that were studied by multiple regression techniques, only the number of doses of cisplatin that patients had received was significantly related to neurotoxicity, hematotoxicity, and tubular nephrotoxicity, with the more severe cases occurring after three or more doses (P less than 0.05). Increased urinary concentrations of the renal tubular enzyme N-acetyl-beta-D-glucosaminidase, measured before each course of ifosfamide, were predictive of neurotoxicity (P = 0.02) and hematotoxicity (P = 0.01). We suggest that cisplatin-induced renal tubular damage, leading to the impaired clearance of ifosfamide metabolites, may account for this added toxicity.

Development and characterization of human ependymoma xenograft HxBr5.

We describe the successful heterotransplantation of a human ependymoma in CBA/CaJ mice immune deprived by infant thymectomy and whole-body irradiation. The xenograft, HxBr5, was established from a fourth ventricular ependymoma, locally recurrent in an 11-yr-old girl who had been treated with radiation therapy to the posterior fossa. HxBr5 retains histological and ultrastructural fidelity to the tumor from which it was derived as does the DNA content, as confirmed by flow cytometric analysis. The karyotype of the xenograft, which is pseudodiploid and exhibits trisomy 1q and deletion of 1p, is the first human ependymoma banded karyotype to be reported. Growth rates of the xenograft tumors are similar to the primary tumor as clinically observed with a doubling time of approximately 42 days. Cell kinetic parameters indicate that this slow-growing tumor has a relatively high growth fraction of 70.8% with a high cell loss of approximately 91%. We anticipate that HxBr5 may be useful as one component of a more complex model for studying the biology and differentiation of human ependymoma.

A novel germline p53 splicing mutation in a pediatric patient with a second malignant neoplasm.

A novel germline p53 splicing mutation was identified in a pediatric patient with two metachronous primary cancers that are constituent tumors of the Li-Fraumeni syndrome. Genomic DNA from the second tumor showed the same mutation and loss of heterozygosity at the p53 locus. The mutant mRNA and protein were present in the tumor tissue. In contrast, in the normal tissues bearing the germline mutation in the heterozygous state, predominantly normal mRNA was expressed and the mutant p53 protein was not detectable. The functional silence and relative lack of mutant p53 mRNA expression in the normal tissues of this patient may be caused by decreased stability or decreased production. If this proves a more general pattern of expression of mutant p53 in individuals with germline mutations, these findings may explain the paucity of tumors in individuals affected with the Li-Fraumeni syndrome.

Urinary N-acetyl-beta-D-glucosaminidase and serum creatinine concentrations predict impaired excretion of methotrexate.

We determined the risk of impaired excretion of methotrexate (MTX) in children with osteosarcoma, who also were receiving cisplatin, by analyzing urinary markers of renal tubular damage, as well as serum creatinine measured before each dose of MTX. MTX clearance was impaired in seven of the ten patients studied after cisplatin therapy. Patients with a urinary N-acetyl-beta-D-glucosaminidase (NAG) concentration of greater than 1.5 U/mmol creatinine or a greater than 50% increase in serum creatinine relative to the pretherapy level were approximately 30 times more likely to have MTX half-lives greater than 3.5 hours than were patients with lower values for these markers; MTX clearance was always impaired if both markers were elevated. If neither urinary NAG nor serum creatinine concentrations increased, the risk of impaired MTX excretion was negligible. Our findings demonstrate that urinary NAG and serum creatinine levels, measured before MTX administration, can be used to identify patients who will have difficulty in clearing the drug.

Changing pattern of pulmonary metastases with adjuvant chemotherapy in patients with osteosarcoma: results from the multiinstitutional osteosarcoma study.

The multiinstitutional osteosarcoma study (MIOS), a randomized trial of adjuvant therapy for osteosarcoma with a concurrent control group, registered 113 patients from June 1982 to August 1984. Preliminary analysis of the study indicated a significant event-free survival advantage favoring immediate adjuvant chemotherapy, (P less than .001). For patients treated with surgery alone or with surgery and adjuvant chemotherapy, the lungs were involved in more than 80% of the relapses. Patients relapsing after surgery alone tended to relapse earlier (P less than .01), had more pulmonary nodules (P less than .01), and had more frequent bilateral pulmonary involvement (P less than .01) than those treated with immediate postsurgical adjuvant chemotherapy. However, patients relapsing after treatment with surgery alone experienced a significantly longer interval to further disease progression (P less than .01) and improved survival after relapse (P = .01) when compared with patients who relapsed after treatment with immediate adjuvant chemotherapy. The only factor predictive of survival after relapse was if the patient could be made surgically disease-free after initial relapse (P = .03).

Neurodevelopmental status of infants and young children treated for brain tumors with preirradiation chemotherapy.

In an effort to reduce the severity of late neurotoxicities associated with cranial irradiation, 14 infants and young children with malignant brain tumors were given preirradiation chemotherapy for 2 to 22 months (median, 8 months). Prospective neurodevelopmental evaluations were routinely conducted and now extend from 35 to 60 months (median, 41 months) postdiagnosis, and 10 to 52 months (median, 31 months) postirradiation in the 12 surviving children. At the initiation of chemotherapy, less than one fourth of the patients displayed normal performance status or mental functioning on age-corrected tests; the majority remained stable or declined while receiving chemotherapy. Declining mental development and adaptive behavior were noted in six patients following radiation therapy with only two patients now functioning in the normal range for age. The analysis suggests that neurodevelopmental progress is a function of multiple factors, including neurologic and sensorimotor deficits associated with the tumor, surgical intervention, and chemotherapy that antedated radiation therapy. This implies that delaying irradiation will not necessarily improve the patients' functional status. Whether the interval of postponement of irradiation evidenced in this sample will translate into an ultimately better quality of life remains unknown. Given the probable interaction of multiple risk factors, well-controlled prospective clinical trials are needed to definitively analyze this issue.

Central nervous system toxicity following the treatment of pediatric patients with ifosfamide/mesna.

Ifosfamide/mesna treatment of 50 patients with pediatric malignant solid tumors was associated with the development of neurotoxic signs and symptoms in 11 of these individuals who received 29 courses of treatment. Neurologic toxicity included changes in mental status, cerebellar function, cranial nerve, and cerebellar and motor system function, including seizures. All symptoms, signs, and EEG abnormalities were transient. Some of the affected individuals failed to develop acute neurotoxic signs of symptoms when retreated with ifosfamide. A grading system for scoring these neurologic abnormalities is presented for comparison of acute neurotoxic effects of other agents. Recommendations are made regarding early termination or delay of ifosfamide/mesna treatments in the presence of significant neurotoxicity.

Phase II testing of melphalan in children with newly diagnosed rhabdomyosarcoma: a model for anticancer drug development.

We describe events that led to successful testing of melphalan, one of the nitrogen mustard compounds, in children with newly diagnosed, poor-risk rhabdomyosarcoma (RMS). Preclinical studies with xenografts of human RMS, growing in the flanks of immune-deprived mice, had indicated superior oncolytic activity by melphalan compared with other agents commonly used to treat this tumor. However, in a conventional phase II trial, melphalan failed to produce partial responses in 12 of 13 heavily pretreated patients with recurrent tumors. Subsequent comparison of the drug's pharmacokinetics in mice and patients indicated that its poor clinical performance was not the result of interspecies differences in drug disposition. Therefore, we elected to retest melphalan in untreated patients, before they were enrolled in a phase III study. Of 13 children who received the drug for 6 weeks, ten had partial responses, confirming the significant antitumor activity seen in the xenograft system. These findings illustrate the inherent limitations of phase II drug trials in previously treated patients and suggest a useful paradigm for the development of antineoplastic drugs.

Phase I and pharmacokinetic evaluation of all-trans-retinoic acid in pediatric patients with cancer.

PURPOSE: Recent reports of the dramatic antitumor effect of all-trans-retinoic acid (RA) in patients with acute promyelocytic leukemia (APL) have renewed interest in the oncologic indications for retinoids. Furthermore, a variety of pediatric tumors are responsive to RA in vitro, which provides additional rationale for a phase I evaluation of RA in children with cancer that is refractory to standard therapy. PATIENTS AND METHODS: A phase I trial of RA administered orally twice daily for 28-day treatment courses was performed. Cohorts of at least three pediatric cancer patients were entered at successive RA dose levels (from 45 to 80 mg/m2/d) until dose-limiting toxicity (DLT) was consistently observed. RESULTS: The maximum-tolerated dose (MTD) of RA was 60 mg/m2/d. Three of eight patients at the 80-mg/m2/d dose level developed reversible pseudotumor cerebri that necessitated discontinuation of the agent. Both patients with APL achieved complete remission (CR), whereas no patients with solid tumors had objective responses. Pharmacokinetic studies demonstrated a relatively short terminal half-life for RA (45 minutes), with diminution in plasma levels after chronic dosing. CONCLUSIONS: The MTD and recommended phase II dose for RA in children is 60 mg/m2/d given twice daily. Reversible CNS toxicity related to RA-induced pseudotumor cerebri is dose-limiting. Two children with APL achieved a CR to RA, which supports the inclusion of pediatric patients in clinical trials that evaluate the use of RA for patients with APL.

Randomized trial of recombinant human granulocyte-macrophage colony-stimulating factor in pediatric patients receiving intensive myelosuppressive chemotherapy.

PURPOSE: To evaluate whether recombinant human granulocyte-macrophage colony-stimulating factor (GM-CSF) reduces the hematologic toxicities and supportive care requirements of an intensive combination chemoradiotherapy regimen in pediatric and young adult sarcoma patients. PATIENTS AND METHODS: Thirty-seven newly diagnosed patients age 1 to 25 years were randomized to receive 18 cycles of chemotherapy alone or with GM-CSF beginning in cycle 3. GM-CSF (5 to 15 micrograms/kg/d subcutaneously) was begun 24 hours after the completion of chemotherapy and continued through day 19 of each cycle or until the absolute granulocyte count (AGC) was > or = 500/microliter on 2 consecutive days. RESULTS: GM-CSF reduced the median duration of grade 4 granulocytopenia from 9.0 days (range, 2 to 24) to 7.0 days (range, 1 to 21) (P < .0001), but did not significantly affect the grade of granulocyte nadir. No differences were seen in the incidence or types of infectious complications, incidence or duration of hospitalization and antimicrobial therapy, response to chemotherapy, or event-free or overall survival. GM-CSF was associated with more severe and protracted thrombocytopenia (median platelet nadir, 29,500/microliter [range, 3,000 to 288,000] v 59,000/microliter [range, 3,000 to 309,000], P < .0001; median time to recovery > 75,000/microliter, 16.0 days [range, 0 to 61] v 14.0 days [range, 0 to 38], P < .0001). CONCLUSION: GM-CSF does not produce clinically meaningful reductions in the degree or duration of severe granulocytopenia following intensive multiagent chemotherapy, but is associated with worsened thrombocytopenia. GM-CSF also does not reduce the need for hospitalization or the incidence of febrile neutropenia and infectious complications. We conclude that the costs and increased toxicities associated with the use of this agent are not justified by its minimal clinical benefit for regimens of this level of intensity.

Total-body irradiation and autologous bone marrow transplant in the treatment of high-risk Ewing's sarcoma and rhabdomyosarcoma.

PURPOSE: In an effort to improve outcome in patients with metastatic or high-risk localized Ewing's sarcoma family of tumors (ESF) and rhabdomyosarcoma (RMS), we explored the role of consolidation therapy with total-body irradiation (TBI) plus autologous bone marrow transplantation (ABMT). PATIENTS AND METHODS: Ninety-one patients were entered onto one of three consecutive protocols from 1981 to 1986. Induction therapy consisted of four or five cycles of vincristine, doxorubicin, and cyclophosphamide (VAdriaC); in the earlier series, patients received one or two cycles with dactinomycin instead of doxorubicin. Irradiation of the primary site was used for local control. Patients who attained a complete response (CR) to induction therapy were eligible for consolidation with 8 Gy TBI plus VAdriaC and ABMT. RESULTS: Nineteen patients were ineligible for consolidation after failing to achieve or maintain a CR following induction therapy; all 19 are dead of disease. Seven eligible patients elected to forgo consolidation; three of seven are long-term event-free survivors. Sixty-five patients received consolidation therapy; 20 of 65 are long-term event-free survivors. A local control rate of 83% was achieved using radiation therapy as the primary modality of local control. Patients with metastatic disease at diagnosis fared substantially worse than did patients with localized tumors (6-year event-free survival [EFS] rate, 14% v 38%; two-sided P [P2] = .008). CONCLUSIONS: Consolidation of patients with metastatic or high-risk localized pediatric sarcomas with 8 Gy TBI plus ABMT has failed to improve the outcome of this group of patients. Metastatic disease at diagnosis continues to confer the poorest prognosis. New therapeutic strategies are needed to consolidate more effectively the remissions that can be achieved in the majority of these patients.

Therapy for childhood soft-tissue sarcomas other than rhabdomyosarcoma: a review of 62 cases treated at a single institution.

The rarity and diverse characteristics of the nonrhabdomyosarcomatous soft-tissue sarcomas (NRSTS) in children have hindered study of their clinical presentations and response to therapy. Here we describe the findings of a retrospective analysis of 62 cases of NRSTS seen in a single institution from 1962 through 1983. The most common histopathologic diagnosis was synovial sarcoma, occurring in 18 patients, followed by malignant schwannoma in 12. The median age at diagnosis was 11 years (range, 2 months to 20 years). Anatomic sites of primary tumors were the trunk (28), extremity (24), and head and neck (10). Of the 31 patients whose tumors were completely resected, 26 (84%) survive with no evidence of disease. Postoperative chemotherapy, administered to nearly one half of this group, did not produce any demonstrable gains in survival. Only one of the 26 patients with local or metastatic gross tumor after resection survives. We conclude that an aggressive surgical approach is imperative in patients with NRSTS and that the contribution of other treatment modalities needs to be defined in a collaborative group trial.

Preirradiation cisplatin and etoposide in the treatment of high-risk medulloblastoma and other malignant embryonal tumors of the central nervous system: a phase II study.

Medulloblastoma, pineoblastoma, and cerebral neuroblastoma are malignant embryonal tumors of the CNS that may demonstrate similar histologic features, a propensity for neuraxis dissemination and sensitivity to radiation therapy and, in certain cases, chemotherapy. To evaluate the activity of preirradiation chemotherapy in such tumors, 11 newly diagnosed children with measurable residual disease and characteristics indicative of poor prognosis were treated postoperatively with cisplatin (CDDP) and etoposide (VP-16). Responses graded on the basis of radiographic findings in areas of either macroscopic residual tumor or metastatic disease included two complete responses (CRs), eight partial responses (PRs), and one stable disease (SD). Acute and subacute toxicity consisted of high-frequency hearing loss in four patients, reversible signs and symptoms of increased intracranial pressure in two patients, and transient neutropenia. Seven of eight patients with high-risk medulloblastoma and two of two with pineoblastoma remain free of tumor progression following neuraxis irradiation at 8 to 48 months postdiagnosis (median, 18 months). CDDP and VP-16 is a highly active drug combination when given before irradiation in children with high-risk medulloblastoma and other malignant embryonal tumors of the CNS, producing objective responses in at least one site of measurable disease in 10 of 11 newly diagnosed patients, including all of five with gross neuraxis dissemination.

Second malignancies after Ewing's sarcoma: radiation dose-dependency of secondary sarcomas.

BACKGROUND: An excess risk of second malignancies has been reported in survivors of Ewing's sarcoma. We examined a multiinstitutional data base to reevaluate the risk among survivors of Ewing's sarcoma and to identify possible causal factors. METHODS: Information was derived from a data base that included 266 survivors of Ewing's sarcoma. Cumulative incidence rates of second malignancies were calculated. Contributions of clinical features, type and dose of chemotherapy, and cumulative radiation dose to the risk of second malignancies were evaluated. RESULTS: After a median follow-up duration of 9.5 years (range, 3.0 to 30), 16 patients have developed second malignancies, which included 10 sarcomas (five osteosarcomas, three fibrosarcomas, and two malignant fibrous histiocytomas) and six other malignancies (acute myeloblastic leukemia, acute lymphoblastic leukemia, meningioma, bronchioalveolar carcinoma, basal cell carcinoma, and carcinoma-in-situ of the cervix). The median latency to the diagnosis of the second malignancy was 7.6 years (range, 3.5 to 25.7). The estimated cumulative incidence rates at 20 years for any second malignancy and for secondary sarcoma were 9.2% (SD = 2.7%) and 6.5% (SD = 2.4%), respectively. The cumulative incidence rate of secondary sarcoma was radiation dose-dependent (P = .002). No secondary sarcomas developed among patients who had received less than 48 Gy, while the absolute risk of secondary sarcoma was 130 cases per 10,000 person-years of observation among patients who had received > or = 60 Gy. CONCLUSION: The overall risk of second malignancies after Ewing's sarcomas is similar to that associated with treatment for other childhood cancers. The radiation dose-dependency of secondary sarcomas justifies modification in therapy to reduce radiation doses.

Randomized trial of the cardioprotective agent ICRF-187 in pediatric sarcoma patients treated with doxorubicin.

PURPOSE: We conducted an open-label, randomized trial to determine whether ICRF-187 would reduce doxorubicin-induced cardiotoxicity in pediatric sarcoma patients. METHODS: Thirty-eight patients were randomized to receive doxorubicin-containing chemotherapy (given as an intravenous bolus) with or without ICRF-187. Resting left ventricular ejection fraction (LVEF) was monitored serially with multigated radionuclide angiography (MUGA) scan. The two groups were compared for incidence and degree of cardiotoxicity, response rates to four cycles of chemotherapy, event-free and overall survival, and incidence and severity of noncardiac toxicities. RESULTS: Eighteen ICRF-187-treated and 15 control patients were assessable for cardiac toxicity. ICRF-187-treated patients were less likely to develop subclinical cardiotoxicity (22% v 67%, P < .01), had a smaller decline in LVEF per 100 mg/m2 of doxorubicin (1.0 v 2.7 percentage points, P = .02), and received a higher median cumulative dose of doxorubicin (410 v 310 mg/m2, P < .05) than did control patients. Objective response rates were identical in the two groups, with no significant differences seen in event-free or overall survival. ICRF-187-treated patients had a significantly higher incidence of transient grade 1 serum transaminase elevations and a trend toward increased hematologic toxicity. CONCLUSION: ICRF-187 reduces the risk of developing short-term subclinical cardiotoxicity in pediatric sarcoma patients who receive up to 410 mg/m2 of doxorubicin. Response rates to chemotherapy, event-free and overall survival, and noncardiac toxicities appear to be unaffected by the use of ICRF-187. Additional clinical trials with larger numbers of patients are needed to determine if the short-term cardioprotection afforded by ICRF-187 will reduce the incidence of late cardiac complications in long-term survivors of childhood cancer.

Is neuro-ectodermal differentiation of Ewing's sarcoma of bone associated with an unfavourable prognosis?

Among Ewing's sarcoma (ES) of bone and related entities are tumours with neuro-ectodermal features that could represent a biologically distinct type. In order to assess the prognostic significance of the various forms of ES, a retrospective joint study involving three cancer centres in Europe and the U.S.A. was initiated. The material from 315 primary ES was reviewed by a panel of five pathologists and classified as typical ES (220 cases), atypical ES (48 cases) or ES with neuro-ectodermal features (47 cases). Prognostic factor analysis on treatment failure-free survival was performed using the Cox model. It included histopathological classification, initial patient characteristics, clinical presentation and treatment type. After multivariate analysis, in addition to treatment type (P < 0.001), metastases (P = 0.003) and proximal tumour location (P = 0.006), two histopathological parameters correlated with poor treatment failure-free survival, the presence of filigree pattern (P = 0.044) and dark cells (P = 0.043). We conclude that ES with neuro-ectodermal features does not appear to have a different outcome to the other subtypes.

Clinical studies of ifosfamide/mesna at St Jude Children's Research Hospital, 1983-1988.

A phase-II study of ifosfamide with mesna, given intravenously daily for five days by bolus injection, has demonstrated the activity of ifosfamide against a spectrum of childhood malignant solid tumors. Ifosfamide presently is being investigated in alternative phase-I schedules, daily times three or every other day times three with the aim of delivering comparable amounts of ifosfamide without increasing toxicity--specifically, neurotoxicity. Additionally, response following ifosfamide treatment is being evaluated for previously untreated children with osteosarcoma and rhabdomyosarcoma after 6 weeks of treatment, and for previously untreated patients with Ewing's sarcoma after 9 weeks of treatment with ifosfamide/VP-16 (etoposide) given in combination.

Sarcomas of the hand and foot: analysis of local control and functional result with combined modality therapy in extremity preservation.

PURPOSE: The records of 28 patients with sarcomas of the hand and foot treated at the National Cancer Institute (NCI) between 1977 and 1992 were reviewed to assess local control and functional results. METHODS AND MATERIALS: Histologic types included 15 cases of the Ewing's sarcoma family of tumors, 7 cases of alveolar rhabdomyosarcoma, and 6 cases of nonrhabdomyosarcoma soft tissue sarcomas. Median age of all patients was 18 years (range 4-61), with a median potential follow-up of 114 months following diagnosis. Surgery varied from incisional biopsies for Ewing's Sarcoma and rhabdomyosarcoma lesions to complete excision when possible for nonrhabdomyosarcoma soft tissue sarcoma lesions. Amputation was not primarily performed, except in two patients who underwent ray resections of hand lesions (patients 13 and 24). Radiotherapy generally consisted of 50 Gy/25 fractions (fx)/5 weeks for Ewing's Sarcoma, 54 Gy/30 fx/6 weeks for rhabdomyosarcoma, and 63 Gy/35 fx/7 weeks for nonrhabdomyosarcoma soft tissue sarcomas. Chemotherapy was administered on various NCI protocols. RESULTS: Actuarial local control for Ewing's Sarcoma was 84% at 5 and 10 years. All but one survivor are capable of hand/foot function for routine activities without orthotic requirements. Five of six patients (83%) who died of metastatic disease had functional distal extremities. Actuarial local control for rhabdomyosarcomas was 100%, with equivalent function. No patient developed a second malignancy in the treatment field. CONCLUSIONS: Although equivalent local control may be achieved in these lesions with either amputation or radiotherapy, a prudent management course would be to defer amputation for management of local recurrences. Many patients with these lesions fail in distant sites only and die without local failure. For these patients and for those who remain long-term survivors, we believe a functional hand and foot provides a better quality of life than a prosthesis.

Long-term follow-up of Ewing's sarcoma of bone treated with combined modality therapy.

Between 1968 and 1980, 107 consecutive patients with Ewing's sarcoma of bone were entered on three sequential combined modality treatment protocols (S2, S3, S4) at the National Cancer Institute (NCI). Protocol treatment involved 4 cycles of two drug [cyclophosphamide (CTX) and vincristine (VCR)] or three drug [CTX and VCR with either actinomycin-D (ACT-D) or doxorubicin (ADR)] regimens and local irradiation (50 Gy) to the involved bone. Eighty patients presented with localized disease and 27 patients had metastatic disease at presentation, including 11 patients with multiple metastatic sites. With a median potential follow-up of greater than 15 yrs (range 8-20 yrs), 28 pts (27%) remain alive. Disease-free (DFS) and overall survival (OS) decreased most rapidly during the initial 5 yrs of follow-up with 5-yr DFS of 29% and 5-yr OS of 39%. Only two patients with metastases at presentation are long term (greater than 5 yr) survivors. For localized disease patients, the 2, 5, 10, and 15 yr DFS and OS are 52%, 37%, 35%, and 33% DFS and 68%, 51%, 39%, and 34% OS, respectively. Eleven patients relapsed locally as the first site of failure. Using the Cox proportional hazards model, four significant variables for both DFS and OS were recognized, including metastatic disease at presentation, age greater than 25 yrs, high LDH in localized disease patients, and central primary tumor in localized disease patients in decreasing order of significance. We conclude that a majority of these patients with Ewing's sarcoma of bone relapsed within 5 yrs of presentation although late relapse (5-15 yrs) did occur. Local failure occurred in 20% of patients using these combined modality treatments but had no impact on overall survival.