Measurement of the Earth tides with a MEMS gravimeter.

The ability to measure tiny variations in the local gravitational acceleration allows, besides other applications, the detection of hidden hydrocarbon reserves, magma build-up before volcanic eruptions, and subterranean tunnels. Several technologies are available that achieve the sensitivities required for such applications (tens of microgal per hertz(1/2)): free-fall gravimeters, spring-based gravimeters, superconducting gravimeters, and atom interferometers. All of these devices can observe the Earth tides: the elastic deformation of the Earth's crust as a result of tidal forces. This is a universally predictable gravitational signal that requires both high sensitivity and high stability over timescales of several days to measure. All present gravimeters, however, have limitations of high cost (more than 100,000 US dollars) and high mass (more than 8 kilograms). Here we present a microelectromechanical system (MEMS) device with a sensitivity of 40 microgal per hertz(1/2) only a few cubic centimetres in size. We use it to measure the Earth tides, revealing the long-term stability of our instrument compared to any other MEMS device. MEMS accelerometers--found in most smart phones--can be mass-produced remarkably cheaply, but none are stable enough to be called a gravimeter. Our device has thus made the transition from accelerometer to gravimeter. The small size and low cost of this MEMS gravimeter suggests many applications in gravity mapping. For example, it could be mounted on a drone instead of low-flying aircraft for distributed land surveying and exploration, deployed to monitor volcanoes, or built into multi-pixel density-contrast imaging arrays.

Apparatus for dimensional characterization of fused silica fibers for the suspensions of advanced gravitational wave detectors.

Detection of gravitational waves from astrophysical sources remains one of the most challenging problems faced by experimental physicists. A significant limit to the sensitivity of future long-baseline interferometric gravitational wave detectors is thermal displacement noise of the test mass mirrors and their suspensions. Suspension thermal noise results from mechanical dissipation in the fused silica suspension fibers suspending the test mass mirrors and is therefore an important noise source at operating frequencies between ∼10 and 30 Hz. This dissipation occurs due to a combination of thermoelastic damping, surface and bulk losses. Its effects can be reduced by optimizing the thermoelastic and surface loss, and these parameters are a function of the cross sectional dimensions of the fiber along its length. This paper presents a new apparatus capable of high resolution measurements of the cross sectional dimensions of suspension fibers of both rectangular and circular cross section, suitable for use in advanced detector mirror suspensions.

Invited article: CO2 laser production of fused silica fibers for use in interferometric gravitational wave detector mirror suspensions.

In 2000 the first mirror suspensions to use a quasi-monolithic final stage were installed at the GEO600 detector site outside Hannover, pioneering the use of fused silica suspension fibers in long baseline interferometric detectors to reduce suspension thermal noise. Since that time, development of the production methods of fused silica fibers has continued. We present here a review of a novel CO(2) laser-based fiber pulling machine developed for the production of fused silica suspensions for the next generation of interferometric gravitational wave detectors and for use in experiments requiring low thermal noise suspensions. We discuss tolerances, strengths, and thermal noise performance requirements for the next generation of gravitational wave detectors. Measurements made on fibers produced using this machine show a 0.8% variation in vertical stiffness and 0.05% tolerance on length, with average strengths exceeding 4 GPa, and mechanical dissipation which meets the requirements for Advanced LIGO thermal noise performance.

Early testicular biopsy in males with acute lymphoblastic leukemia: lack of impact on subsequent event-free survival.

PURPOSE: Children with acute lymphoblastic leukemia (ALL) who had bulky disease (lymphomatous features) at diagnosis had the highest rate of testicular relapse (20%) of any ALL subgroup on previous Children's Cancer Group (CCG) studies in the late 1980s. To limit curative, but sterilizing, testicular irradiation to those with testicular disease, testicular biopsies were performed to detect occult testicular disease within the first 6 months of treatment. Testicular irradiation then was provided to those with occult disease to increase disease-free survival. Identification of those with occult disease was believed to be a factor that would influence ultimate survival in such patients in that era. PATIENTS AND METHODS: One hundred ninety-nine patients had bilateral testicular wedge biopsies performed during the first maintenance therapy phase of the four different chemotherapy regimens. Patients with positive biopsy results were treated with testicular irradiation and continued on therapy. RESULTS: Eleven of 199 biopsy results (5.5%) were judged positive. Patients with positive biopsy results given testicular radiation had a 45% subsequent adverse event rate, compared with 36% for those with a negative biopsy results (P = 0.4). The survival rates for the two groups were similar. The low rate of positive biopsy specimens resulted in discontinuation of the procedure before closure of the study. CONCLUSION: Positive testicular biopsy results early in remission identified patients at a slightly higher risk of subsequent adverse events but did not influence survival. However, because negative biopsy results (94.5%) did not alter the prescribed treatment, the small number of positive biopsy results did not warrant undertaking the procedure in most male patients with ALL, and this procedure was abandoned.

Children's Cancer Group trials in childhood acute lymphoblastic leukemia: 1983-1995.

Since 1968, the Children's Cancer Group (CCG) has treated more than 16,000 children with acute lymphoblastic leukemia (ALL). Herein, we report improvements obtained in CCG trials during two successive series of studies (1983-1988 and 1989-1995). Overall, 10-year EFS was 62% +/- 10% for the 1983-1988 series and 72% +/- 1% for the 1988-1995 series (P< 0.0001). Five-year cumulative rates of isolated CNS relapses were 5.9% and 4.4%. Therapy based on the Berlin-Frankfurt-Münster 76/79 study improved outcomes for intermediate and higher risk patients in the first series. For intermediate risk patients, delayed intensification (DI) was most crucial for improved outcome and cranial irradiation was safely replaced with maintenance intrathecal methotrexate, providing patients received intensified systemic therapy. In the second series, randomized trials showed better outcome with one vs no DI phase for lower risk patients, with two vs one DI phase for intermediate risk patients, and with the CCG 'augmented regimen' for higher risk patients with a slow day 7 marrow response. Cranial irradiation was safely replaced with additional intrathecal methotrexate for higher risk patients with a rapid day 7 marrow response. In a subsequent study, substitution of dexamethasone in place of prednisone in induction and maintenance improved outcome for standard risk patients. All patients received dexamethasone in DI. These successful treatment strategies form the basis for our current ALL trials.

Timing and magnitude of decline in alpha-fetoprotein levels in treated children with unresectable or metastatic hepatoblastoma are predictors of outcome: a report from the Children's Cancer Group.

PURPOSE: We analyzed data on 31 children with primary unresectable or metastatic hepatoblastoma (HB) to investigate possible prognostic correlations between the serum level of alpha-fetoprotein (AFP), its changes during treatment, and outcome. PATIENTS AND METHODS: Patients were treated according to the Children's Cancer Group (CCG) protocol 823F, which included an initial surgery before eight courses of chemotherapy that consisted of cisplatin immediately followed by a continuous infusion of doxorubicin. Four courses were given before and four after the second surgery. AFP levels were measured before treatment, before and after second surgery, and at the end of treatment. RESULTS: Twenty-four of 31 patients showed a decline of > or = 1 log in AFP levels before second surgery (early responders). By the end of treatment, there were 16 patients, all early responders, without clinical or radiographic evidence of tumor and with normal AFP levels. Fifteen of those 16 had a decline of > or = 2 logs in AFP before second surgery (large early response). Of the 15 patients who failed to respond to treatment, 10 died, among whom only one patient had a large early response. A large early response was the strongest independent predictor of outcome in a univariate and multivariate Cox regression model, and patients with such a response had the best survival (P < .0001). CONCLUSION: For children with unresectable or metastatic HB, early changes in AFP levels are a reliable predictor of outcome and can be used for identification of poor responders to treatment, ie, patients whose AFP level fails to decrease 2 logs before second surgery should be considered for alternative treatment.

Behavioral adjustment and social functioning of long-term survivors of childhood leukemia: parent and teacher reports.

Obtained parent and teacher reports of behavior and social competence for children who were survivors of acute lymphoblastic leukemia (ALL). At follow-up, children were 5-18 years of age, 48 months postdiagnosis, in first continuous remission, and off chemotherapy. Each child had been randomized to receive either 1,800 cGy whole brain radiation therapy (WBRT) plus intrathecal methotrexate (IT MTX), or IT MTX alone as central nervous system prophylaxis, and one of four chemotherapy regimens that varied in treatment intensity. Scores on standardized measures (CBCL-P/T and PIC) were generally similar to instrument norms. Parents, but not teachers, reported heightened child somatic concerns. There was no effect of WBRT or chemotherapy regimen on ratings of behavioral adjustment. Results indicate minimal psychosocial morbidity among long-term survivors of ALL and suggest that the stressful life events associated with cancer and its treatment do not cause significant behavioral or emotional difficulties.

Morphologic, immunologic, and cytogenetic classification of acute myeloid leukemia and myelodysplastic syndrome in childhood: a report from the Childrens Cancer Group.

The purposes of this report are to reaffirm concordance difficulties with the acute myeloid leukemia (AML) French-American-British (FAB) classification, to present the frequency of previously delineated AML syndromes in pediatric patients and to describe additional characteristic AML profiles utilizing composite morphologic, cytogenetic and immunophenotypic data. Profiles of 124 children with acute myeloid leukemia (AML) and 13 children with myelodysplastic syndrome entered on the Childrens Cancer Group (CCG) pilot study CCG-2861 were examined. Concordance between institutions and reviewers for FAB designation was 65%. Discordance was found principally between M1 and M2, M2 and M4, and M4 and M5. In 49% of marrow specimens, leukemic blasts expressed at least one T lineage-related antigen; 24% expressed the B lineage-related antigen CD19. CDw14 correlated with FAB M4 or M5 morphology and was the only surface antigen associated with a specific FAB subtype. Normal karyotypes were found for 15% of the 75 children with satisfactory karyotype preparations. Recurring aberrations, found in 76% of children, included t(15;17)(q22;q11), t(8;21)(q22;q22), inv(16)(p13q22), rearrangements of band 11q23, t(6;9) (p23;q34), trisomy 8 and monosomy 7. Results from this pilot study and from the current CCG randomized trial correlating morphology, immunophenotyping and cytogenetics, will help to classify AML into unique subgroups with differing clinical consequences or therapy requirements.

Low numbers of CSF blasts at diagnosis do not predict for the development of CNS leukemia in children with intermediate-risk acute lymphoblastic leukemia: a Childrens Cancer Group report.

PURPOSE: This study was designed to evaluate the effect on CNS relapse (CNSR) and overall relapse rates of blast cells in the CSF containing < or = 5 cells/microL at the time of diagnosis of intermediate-risk acute lymphoblastic leukemia (ALL) in children entered onto a large randomized multicenter prospective therapeutic trial (Childrens Cancer Group [CCG]-105). PATIENTS AND METHODS: We studied outcome in terms of CNSR and event-free survival (EFS) in 1,544 patients who successfully completed remission-induction therapy and had been randomized to one of four systemic chemotherapy regimens and to one of two CNS prophylaxis regimens. We compared outcome between 1,450 patients who had varying degrees of pleocytosis but no blasts in the CSF at diagnosis (blast-negative group) with 94 who had blasts detected in the CSF after cytocentrifugation but had a total CSF WBC count of < or = 5/microL (blast-positive group). RESULTS: No statistically significant differences in overall CNSR or EFS rates were observed between the two groups and no differences were found when analyzed according to age or WBC count at diagnosis, sex, or type of CNS prophylaxis (intrathecal [IT] methotrexate [MTX] alone v IT MTX plus 18 Gy cranial irradiation [CXRT]). CONCLUSION: In intermediate-risk ALL, there was no significant difference in CNSR and systemic relapse rates after standard presymptomatic CNS therapy between patients with a CSF WBC count < or = 5/microL and those without identifiable blasts in the CSF. These findings suggest that certain approaches to therapy, such as that used in this study, may eliminate the need for any additional special treatment directed at this subset of patients with CSF blasts.

Phase I study of tumor necrosis factor-alpha and actinomycin D in pediatric patients with cancer: a Children's Cancer Group study.

In preclinical studies, synergy was observed between tumor necrosis factor-alpha (TNF-alpha) and agents that interact with DNA topoisomerase II, such as actinomycin D (Act D). Based upon this, a Phase I study was conducted in pediatric patients utilizing an escalating dose of recombinant TNF (rTNF) in combination with a fixed dose of Act D. Act D (15 micrograms/kg/day) was administered daily by intravenous push immediately followed by intravenous rTNF daily for 5 consecutive days. Thirty-three patients with refractory malignancies were entered in the study, of whom 28 patients could be evaluated for toxicity. Malignancies included sarcomas (16), Wilms' tumor (6), leukemias (3), and others (3). The starting dose for rTNF was 40 micrograms/m2/day x 5 and was escalated in subsequent patient groups until nonhematopoietic, dose-limiting toxicity occurred. At 240 micrograms/m2/day of rTNF, three of six patients experienced grade 4 toxicity consisting of hypotension, hemorrhagic gastritis, and renal and liver biochemical abnormalities. Evidence of antitumor response was observed in two patients: one with metastatic Ewing's sarcoma and one with Wilms' tumor. We conclude that the maximum tolerated dose of rTNF when combined with Act D is between 200 and 220 micrograms/m2/day x 5 for pediatric patients.

Cell surface expression of c-kit receptors by childhood acute myeloid leukemia blasts is not of prognostic value: a report from the Childrens Cancer Group.

The prognostic significance of c-kit receptor expression on leukemic blast cells was determined in 122 children with acute myeloid leukemia (AML) entered onto Childrens Cancer Group protocol 213. Clinical and laboratory characteristics as well as outcome were analyzed according to the percentage of blast cells expressing c-kit receptors and the relative number of c-kit receptors per cell as determined by indirect immunofluorescence. c-kit receptor expression was strongly associated with the expression of the CD34 antigen. However, contrary to findings in adult patients with AML, c-kit receptor expression by childhood AML blast cells was not predictive of a poor response to therapy.

What caused my child's cancer? Parents' responses to an epidemiology study of childhood cancer.

When a child is diagnosed with cancer, parents try to understand why the cancer developed. Although usually it is not possible to explain what caused an individual child's cancer, clinical experience has shown that parents do form theories about the origins of their child's illness although, or perhaps because, no one knows the actual cause. A parent-completed epidemiology questionnaire (EQ), designed to provide a comprehensive and general epidemiology data base for studies conducted by the Childrens Cancer Group, included an open-ended item ("Do you have any additional comments or concerns about anything that could have caused or contributed to your child's illness?"). A convenience sample of 500 EQs containing responses to the open-ended question was reviewed independently by two experienced pediatric oncology nurses. Statements contained in the responses were categorized into 12 major themes according to content: concern about environmental exposures (n = 303), concern about family health history (n = 270), specific causality attribution (n = 39), puzzlement (n = 24), concern with cancer "clusters" (n = 23), concern with stress (n = 22), altruism (n = 15), specific feedback requests (n = 11), myths/misconceptions (n = 5), advocation of preventive education/screening (n = 4), active information-seeking (n = 6), and parental self-blame (n = 4). These themes or concerns provide useful information that can be applied in planning educational and supportive clinical interventions, as well as further research.

A phase II study of carboplatin in children with recurrent or progressive solid tumors. A report from the Childrens Cancer Group.

BACKGROUND: Carboplatin is an analogue of cisplatin with less nonhematologic toxicity than the parent compound. It has been demonstrated previously to have activity against a spectrum of pediatric brain tumors. This Phase II study was undertaken to assess the activity of carboplatin in children with various solid tumors. METHODS: Between October 1985 and March 1988, the Childrens Cancer Group entered 117 patients with drug-resistant, recurrent lymphomas and solid tumors, excluding primary central nervous system tumors, into a Phase II trial of carboplatin given intravenously at a dosage of 560 mg/m2 over 1 hour every 4 weeks. RESULTS: A complete response was seen in 1 of 15 evaluable patients with Ewing's sarcoma. Partial responses were seen in 2 of 17 evaluable patients with neuroblastoma, 1 of 16 with soft tissue sarcoma, 2 of 5 with Wilms' tumor, and 1 with an endodermal sinus tumor of the testis. Objective responses were not seen in patients with malignant lymphoma, osteosarcoma, or hepatoma. Four of 7 patients who responded to carboplatin had previously received cisplatin. Sixty-two percent of patients had a platelet count nadir of less than 50,000/mm3, and 41% had an absolute neutrophil count nadir of less than 1,000/mm3. Ototoxicity, nephrotoxicity, hypomagnesemia, hypertransaminasemia, and nausea and vomiting each were seen in fewer than 10% of patients. CONCLUSIONS: Carboplatin has some activity against Wilms' tumor, Ewing's sarcoma, neuroblastoma, soft tissue sarcoma, and endodermal sinus tumor of the testis. Activity was not demonstrated against osteosarcoma, malignant lymphoma, hepatoma, and miscellaneous other tumors. Myelosuppression was seen commonly, and nonhematologic toxicity was infrequent.

Treatment of overt isolated testicular relapse in children on therapy for acute lymphoblastic leukemia. A report from the Childrens Cancer Group.

BACKGROUND: Fifty-seven children with acute lymphoblastic leukemia (ALL) receiving therapy who experienced overt isolated testicular relapse while in bone marrow remission were entered into a study that featured an intensive retreatment regimen. METHODS: The objective was to determine whether a change in chemotherapy and local irradiation would prevent subsequent marrow relapse and increase the survival rate. The regimens used (modified Berlin-Frankfurt-Munster or modified New York) delivered acceptable therapy based on analyses of toxicity data. RESULTS: Overall survival at 5 years from the time of testicular relapse was 47%, with an event-free survival of 43%. Events have been documented in 28 of 55 evaluable children. Analysis of these patients revealed that 23 children had bone marrow relapse, 4 children had central nervous system relapse, and 1 child had testicular relapse. In addition, two patients were removed from the study for toxicity, one child for infection (mucormycosis), and five children had a bone marrow transplantation while in remission and became ineligible to continue in the study. Two children were removed at the request of their parents, and nine children were lost to follow-up. CONCLUSION: Because of their high risk of developing systemic relapse, boys with ALL who experience isolated overt testicular relapse during active therapy should be retreated with intensive treatment.

Molecular rearrangements of the MLL gene are present in most cases of infant acute myeloid leukemia and are strongly correlated with monocytic or myelomonocytic phenotypes.

Cytogenetic studies have previously identified abnormalities of chromosome band 11q23 in many cases of infant acute leukemia. Recent studies by ourselves and others have demonstrated breakpoint clustering in acute leukemias bearing translocations involving 11q23, and a Drosophila trithorax gene homologue (called MLL, HRX, or ALL-1) has been shown to span the 11q23 breakpoints of these translocations. To determine if this gene is affected in infant acute myeloid leukemia (AML), we have analyzed 26 infant AML cases for molecular alterations of this 11q23 gene. 15 out of 26 cases studied (58%) showed rearrangement of the MLL gene at the molecular level, and these rearrangements were clustered within an approximately 11-kb region containing nine exons of this gene. Moreover, 14 of the 15 cases with 11q23 rearrangements (93%) had myelomonocytic or monocytic phenotypes (M4 or M5 FAB subtypes, respectively), both of which are associated with a poor prognosis in childhood AML. In contrast, only 1 of 11 nonrearranged cases had an M4 or M5 phenotype (P = 0.00002). Rearrangement also correlated significantly with hyperleukocytosis (P = 0.02), another clinical parameter associated with poor outcome in this disease. Our results demonstrate that molecular rearrangements of MLL are common in M4 or M5 infant AML, and suggest that alteration of this gene may result in abnormal control of proliferation and differentiation in monocytic progenitor cells.

Magnetic resonance scans should replace biopsies for the diagnosis of diffuse brain stem gliomas: a report from the Children's Cancer Group.

Children's Cancer Group Protocol CCG-9882 was designed to determine the effectiveness of hyperfractionated radiation for the treatment of children and young adults with brain stem gliomas. The study opened for the accrual of patients on September 21, 1988, and was closed on June 30, 1991. The first 54 children in the study were treated with irradiation doses of 100 cGy given twice daily to a total dosage of 7200 cGy. The next 66 children were treated with a similar daily regimens to a total of 7800 cGy. Tumors were diagnosed by clinical and radiographic criteria. Decisions about the need for surgery were left to the discretion of the treating neurosurgeon; tissue diagnosis did not alter the therapy in patients with diffuse infiltrating tumors. We reviewed the neuroradiology and neurosurgery reports as well as the pathological specimens of children entered on the study. By magnetic resonance (MR) imaging criteria, tumors involved the majority of the brain stem in 76% of cases; only three patients had tumors localized to the midbrain or medulla. Operations were performed on 56 of 120 patients (47%). Cerebrospinal fluid shunts were inserted in 27 (23%) of the children; insertion of a shunt was the only operation in 11, and a shunt was inserted in conjunction with a tumor operation in 16. Tumor operations were performed in 45 (38%) of the patients; 24 had stereotactic biopsies, and 21 had craniotomies. Of the 21 patients who had craniotomies, only biopsies were performed in 11; partial tumor resections were performed in 5 patients and subtotal resection in 5.(ABSTRACT TRUNCATED AT 250 WORDS)

The role of radiation therapy in the treatment of acute lymphoblastic leukemia with lymphomatous presentation: a report from the Childrens Cancer Group.

PURPOSE: Childrens Cancer Group 123 was a trial of intensive multidrug chemotherapy as well as cranial irradiation and bulk disease irradiation in children with acute lymphoblastic leukemia with lymphomatous presentation (bulk disease and either T-cell phenotype, high white blood count, or absence of anemia), a poor prognostic group with an increased risk of central nervous system (CNS) and other extramedullary recurrence. METHODS AND MATERIALS: Three hundred eight patients without CNS disease were randomized among three regimens: A--BFM chemotherapy (designed for high risk ALL patients) with 1800 cGy cranial irradiation; B--LSA2L2 chemotherapy (designed for non-Hodgkins lymphoma patients) with 1800 cGy cranial irradiation and 1500 cGy to nonabdominal bulk disease; C--Reg B without cranial irradiation. All patients received intrathecal methotrexate throughout therapy. Radiation treatment records were reviewed. RESULTS: With a minimum 52-month follow-up, Regimen B and C patients had 5-year actuarial CNS relapses of 7% and 17% (p = 0.01) and event-free survivals of 53% and 39% (p = 0.04). Patients with white blood count < 50,000/mm3 did not benefit from cranial irradiation. Regimen A patients had the same CNS relapse rate as Regimen B patients but an improved event-free survival. Regimen B and C patients with large mediastinal masses who received their assigned chest radiation had a lower event rate than those who did not (p = 0.06). Patients whose cranial fields did or did not encompass the entire meningeal surface had equivalent CNS relapse rates. CONCLUSION: Patients treated with LSA2L2 chemotherapy, a less than optimal regimen, benefited from cranial and mediastinal irradiation. Compliance with radiation volume guidelines was not essential for patients to receive the benefit of cranial irradiation.

A phase I study of carboplatin in children with acute leukemia in bone marrow relapse. A report from the Childrens Cancer Group.

BACKGROUND: Carboplatin is an analogue of cisplatin with less nonhematologic toxicity and a similar spectrum of antineoplastic activity as the parent compound. Although cisplatin has not been found to be an active agent in leukemia, carboplatin induced complete remissions in adults with acute myelogenous leukemia (AML). Therefore, a pediatric Phase I study in acute leukemia was performed. METHODS: Between January 1988 and April 1990, the Childrens Cancer Group performed a Phase I study of carboplatin administered by a 5-day continuous intravenous infusion to children with acute leukemia in bone marrow relapse. RESULTS: Mild to moderate glomerular and tubular nephrotoxicity was seen in most patients treated at the initial dose level of 336 mg/m2/day. Therefore, patients at the second dose level were treated at 270 mg/m2/day. At this level, one patient died of acute hepatic necrosis and hepatic encephalopathy, and a second patient had presumed hemorrhagic cystitis develop. The third dose level tested, 216 mg/m2/day, was not associated with unacceptable toxic effects and was considered the maximum tolerated dose (dose-limiting toxicity was not observed). Within the confines of this Phase I study, antileukemic activity was shown in patients with acute lymphoblastic leukemia (ALL) and AML. CONCLUSIONS: In this pediatric Phase I trial of carboplatin in acute leukemia, glomerular and tubular nephrotoxicity was considered dose-limiting. In addition, hepatotoxicity and hemorrhagic cystitis were observed. Antileukemic activity was shown in patients with ALL and AML. The recommended Phase II dose is 216 mg/m2/day by 5-day continuous intravenous infusion.

Hyperfractionated radiation therapy (72 Gy) for children with brain stem gliomas. A Childrens Cancer Group Phase I/II Trial.

BACKGROUND: Most children with brain stem gliomas (BSG) die within 18 months of diagnosis. Early experience suggested that hyperfractionated radiation therapy (RT) at a dose of 72 Gy, administered in 1-Gy fractions twice daily, possibly improved disease-free survival for children with BSG. METHODS: To better characterize the toxicity and possible efficacy of this dose and fractionation of RT, 53 assessable children with diffuse intrinsic or malignant BSG were treated. Survival figures also were combined with outcome in 36 patients treated in a previous pilot study. RESULTS: An objective response to treatment was observed in 28 of 53 (53%) patients; a partial response occurred in 7. No child died of treatment-related brain necrosis, although 7 of 53 did have intralesional cystic/necrotic changes within 6 weeks of completion of RT. The overall survival rate for patients in the study was 38% (+/- 6.5) at 1 year, 14% (+/- 5.4) at 2 years, and 8% (+/- 6.5) at 3 years. Leptomeningeal dissemination was observed in 4 of 48 (8%) children who had relapses. A greater than 2-month duration of symptoms before diagnosis was related to a better prognosis. There was no statistical association between any other clinical parameter, neuroradiographic finding, or pathologic finding and outcome. Combined with that in 35 patients treated in the pilot study, the survival rate in 88 children was 14% (+/- 5) at 3 years. CONCLUSIONS: The radiographic response rate is encouraging; however, it cannot be concluded that hyperfractionated RT, at this dose schedule and total dose, is superior to conventional RT.