This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Geyer, J. R. Articles by Shiminski-Maher, T. Search for Related Content PubMed PubMed Citation Articles by Geyer, J. R. Articles by Shiminski-Maher, T. Social Bookmarking                            What's this?

Multiagent Chemotherapy and Deferred Radiotherapy in Infants With Malignant Brain Tumors: A Report From the Children’s Cancer Group

From the Children’s Hospital and Regional Medical Center, Seattle, WA; Children’s Oncology Group, Arcadia; Valley Radiotherapy Associates and University of California, Los Angeles, Los Angeles, CA; Vanderbilt Children’s Hospital, Nashville; St Jude Children’s Research Hospital, Memphis, TN; DeVos Children’s Hospital, Grand Rapids, MI; Children’s Medical Center–Dayton, Dayton, OH; New York University Medical Center; New York Hospital–Cornell University Medical Center, New York, NY; Children’s Hospital of Philadelphia, Philadelphia, PA; Children’s Healthcare of Atlanta at Scottish Rite, Atlanta, GA; Children’s National Medical Center, Washington, DC; and University of Wisconsin Children’s Hospital, Madison, WI

Address reprint requests to J. Russell Geyer, MD, Children’s Hospital and Regional Medical Center, Department of Pediatric Hematology-Oncology, 4800 Sand Point Way NE, MS: CH-29, Seattle, WA 98105; e-mail: russ.geyer{at}seattlechildrens.org

	ABSTRACT

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors’ Disclosures of... REFERENCES

 
PURPOSE: To evaluate response rate, event-free survival (EFS), and toxicity of two chemotherapeutic regimens for treatment of children younger than 36 months with malignant brain tumors and to estimate control intervals without irradiation in children with no residual tumor after initial surgery and induction chemotherapy and with delayed irradiation in patients with residual tumor or metastatic disease at diagnosis.

PATIENTS AND METHODS: Patients were randomly assigned to one of two regimens of induction chemotherapy (vincristine, cisplatin, cyclophosphamide, and etoposide v vincristine, carboplatin, ifosfamide, and etoposide). Maintenance chemotherapy began after induction in children without progressive disease. Children with no residual tumors after induction therapy and no metastatic disease at diagnosis were not to receive radiation therapy unless their tumors progressed.

RESULTS: Two hundred ninety-nine infants were enrolled. Forty-two percent of patients responded to induction chemotherapy. At 5 years from study entry, the EFS rate was 27% ± 3%, and the survival rate was 43% ± 3%. There was no significant difference between the two arms in terms of response rate or EFS. For medulloblastoma, supratentorial primitive neuroectodermal tumor, ependymoma, and rhabdoid tumors, 5-year EFS rates were 32% ± 5%, 17% ± 6%, and 32% ± 6%, and 14% ± 7%, respectively. Fifty-eight percent of patients who were alive 5 years after study entry had not received radiation therapy.

CONCLUSION: Intensified induction chemotherapy resulted in a high response rate of malignant brain tumors in infants. Survival was comparable to that of previous studies, and most patients who survived did not receive radiation therapy.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors’ Disclosures of... REFERENCES

 
Tumors of the CNS account for 17% of childhood malignancies, with the highest incidence during the first 5 years of life.¹ Among children younger than 1 year old, incidence of CNS tumors exceeds that of acute lymphocytic leukemia.¹ Although considerable improvement in survival has been achieved overall in childhood cancer and in CNS cancer in older children,^2,3 little progress has been made in younger children with these tumors. ^4-8 For children younger than 5 years, the 5-year survival rate has declined.¹ Late effects of therapy, in particular radiation therapy, on cognitive function have been substantial among children treated for CNS tumors at young ages.^4,5,9-11

Previous trials found potential for delaying and, in some cases, avoiding radiation therapy in treatment of young children with malignant brain tumors,^12-17 although with survival rates considerably less than the rates of older children treated with early irradiation.^2,3,18 Two studies using intensive chemotherapeutic regimens supported by granulocyte colony-stimulating factor (G-CSF) were piloted; one uses cyclophosphamide and cisplatin,¹⁹ and the second uses ifosfamide and carboplatin,²⁰ with vincristine and etoposide used in each. Both pilot studies had high response rates, particularly in children with medulloblastoma or supratentorial primitive neuroectodermal tumors (PNET). On the basis of those trials, a phase II study was developed to evaluate response rate, event-free survival (EFS), survival, and toxicity of two chemotherapeutic regimens in the treatment of children younger than 36 months old with malignant brain tumors. The study also was designed to estimate disease control intervals without irradiation in children with no residual tumors after initial surgery and induction chemotherapy and with delayed irradiation in patients with residual tumor or metastatic disease at diagnosis.

Patients were randomly assigned between the two regimens to ensure similar patient populations, but there was not an intent to formally compare the two regimens in the original study design. However, because the frequency of gross total resection was greater than expected and because a decision was made to extend the accrual period to address the phase II question within the subset of children with medulloblastoma, it became apparent that a greater number of patients would be entered onto the study than had initially been planned. It was felt to be important to compare the two regimens directly to ensure that a significant difference in EFS between the two arms would be detected promptly. Therefore, the study was redesigned and amended to include a formal phase III randomized comparison.

	PATIENTS AND METHODS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors’ Disclosures of... REFERENCES

 
Study Design
The Children’s Cancer Group (CCG) study CCG-9921 was activated in April of 1993. Children enrolled onto CCG-9921 were younger than 36 months old, untreated, and newly diagnosed with malignant brain tumors. Eligible diagnoses included ependymoma, high-grade astrocytoma, and other high-grade gliomas including anaplastic mixed gliomas or anaplastic glioneuronal tumors, supratentorial PNET, medulloblastoma (ie, infratentorial PNET), choroid plexus carcinomas, atypical teratoid/rhabdoid tumors, malignant germ cell tumors, primary sarcomas, and intrinsic brainstem gliomas. Histologic confirmation of diagnosis was required for entry except for patients with intrinsic brainstem gliomas. Each child was to have as complete a resection as was feasible. Computed tomography or magnetic resonance imaging (MRI) scans with and without contrast enhancement were required before and within 72 hours after surgery, and myelogram or spinal MRI studies were required for all patients. In addition, CSF cytology was required of all patients except those with intrinsic brainstem gliomas or supratentorial high-grade gliomas. Bone scans and bone marrow aspirates were required for patients with medulloblastoma or supratentorial PNET. Patients’ families or legal guardians had to acknowledge in writing that consent to become a study patient had been given, in accordance with institutional policies approved by the US Department of Health and Human Services. All institutional, US Food and Drug Administration, and National Cancer Institute requirements for human studies were met in accordance with the institutional assurance policies of the Department of Health and Human Services.

Children began receiving chemotherapy within 28 days of diagnostic surgery or radiologic diagnosis of brainstem tumor. On entry, patients were assigned randomly to one of two regimens of induction chemotherapy. Induction comprised five cycles of chemotherapy, each 3 weeks in duration. Maintenance chemotherapy began after induction in children without progressive disease. Second surgery was encouraged if residual tumor persisted after induction.

Radiation Therapy
Children with no residual tumor after induction therapy and no metastatic disease at diagnosis were not to receive radiation therapy unless their tumors progressed. In patients who had persistent residual disease after induction or metastatic disease at diagnosis, irradiation was administered when they reached 36 months old or at completion of eight cycles of maintenance chemotherapy, whichever came first. Children whose tumors progressed or recurred were administered radiation therapy at the time of progression or recurrence, regardless of age.

Patients with nonmetastatic tumors, including medulloblastoma and supratentorial PNET in children younger than 18 months old at the time of irradiation, were to have received focal irradiation only, to a total dose of 54 Gy (50.4 Gy for patients younger than 18 months old). Focal radiation was defined as tumor volume plus a 1.5-cm margin. Children with metastatic disease at diagnosis or recurrence and children with medulloblastoma or supratentorial PNET who were older than 18 months at diagnosis were to have received craniospinal (CS) irradiation (18 to 30.6 Gy depending on age and disease status). In patients with infratentorial tumors receiving CS radiation therapy, the local field was defined as the posterior fossa plus a 1-cm margin.

Chemotherapy
Induction regimen A comprised vincristine (0.05 mg/kg) on days 0, 7, and 14; cisplatin (3.5 mg/kg) on day 0; cyclophosphamide (55 mg/kg) on days 1 and 2; and etoposide (2.5 mg/kg) on days 0, 1, and 2. Mesna was administered after cyclophosphamide. Induction regimen B consisted of vincristine (0.05 mg/kg) on days 0, 7, and 14; carboplatin (10 mg/kg) on days 0 through 1; ifosfamide (60 mg/kg) on days 0 through 4; and etoposide (1.5 mg/kg) on days 0 through 4. Mesna was administered after cyclophosphamide and ifosfamide, and G-CSF (5 µg/kg) was begun 24 hours after completion of chemotherapy. In both regimens, induction consisted of five cycles, with each cycle being 3 weeks in duration.

Maintenance chemotherapy was the same for both regimens and comprised vincristine (0.05 mg/kg) on days 0, 7, 14, and 21; etoposide (2.5 mg/kg) on days 0, 1, 28, and 29; carboplatin (18 mg/kg) on day 0; and cyclophosphamide (65 mg/kg) on day 28. Mesna was administered after cyclophosphamide. G-CSF was not administered routinely during maintenance therapy. Maintenance consisted of eight cycles of chemotherapy, and each cycle was 49 days in duration.

Blood counts and chemistries and renal and audiologic evaluations were performed regularly. Imaging of the primary site and neuroaxis was performed at completion of induction, every 2 months during maintenance, and every 3 months for 2 years and every 6 months for the subsequent 2 years after completion of therapy. Imaging of the neuroaxis (except brainstem gliomas and supratentorial malignant gliomas) was performed every 4 months during maintenance and every 6 months for 2 years after completion of therapy.

Central Radiology Review and Assessment of Response
Pre- and postoperative computed tomography or MRI were reviewed centrally by two investigators (J.R.G. and L.A.H.). Postoperative scans were compared with those performed after induction to assess response.

Patients classified as having had progressive disease or treatment failure during induction by treating institutions were classified the same way for response analysis. However, in patients whose disease did not progress or in whom treatment did not fail otherwise during induction, response assessment was based only on central review of films, and patients for whom central review of response was not possible were noted.

Central Pathology Review
Eligibility was based on institutional pathologic diagnosis. The study neuropathologist (D.C.M.) reviewed all tumor specimens for which tissue was collected to ensure consistent classification based on contemporary WHO guidelines. A variety of immunohistochemical stains were used in central review to aid tumor classification. Medulloblastoma and PNET were identified, when possible, according to identified patterns of differentiation based on histologic patterns and a standard panel of immunostains. PNET-like tumors with rhabdoid tumor elements (atypical teratoid tumor) were identified similarly. Central pathology review was performed for 80% of patients, including 238 (79.6%) of 299 patients and 227 (79.9%) of 284 eligible patients.

Central Review of Radiation Therapy
Patients who received radiation therapy were reviewed (M.L.T.) for adherence to protocol-specified dose, fields, and timing. Review of target volume was not performed. Radiation treatment data were obtained from institutional reports. Dose to the primary tumor and to the CS axis was classified as within protocol limits (within ± 5% of nominal dose), minor deviation (within ± 10% of nominal dose), and major deviation (in excess of ± 10% of nominal dose).

Statistical Methods
The primary end points for statistical analysis were tumor response to chemotherapy and time to a treatment failure event, which was defined as the minimum time from study entry to disease progression, disease relapse, second malignant neoplasm, or death from any cause. Secondary end points included time to death from any cause. The study was originally designed as a phase II study of induction response to treatment, with patients assigned equally to the two treatments via random assignment. The study was redesigned to include a formal phase III comparison of EFS. The phase II design required that response would be evaluated in the first 14 patients within each stratum and treatment to ensure that at least one patient via a two-stage rule obtained a true objective response (complete response [CR] plus partial response [PR]) consistent with a rate of 20% or higher. If none of the first 14 assessable patients in each regimen responded, then accrual to that regimen for infants with that histology would be terminated. This rule had at least 95% power to detect true response rates of 20% and greater. The strata for phase II evaluation were medulloblastoma, nonposterior fossa PNET, high-grade astrocytoma, and other. No stratum was closed because of lack of response because at least one objective response was observed for each stratum/treatment combination in the first 14 assessable patients.

Nonparametric estimates of EFS and survival probabilities were calculated using the product-limit estimate, with SEs computed using the Greenwood formula.²¹ All analyses were based on intent-to-treat analysis, whereby patients were not censored for reasons of deviations from assigned protocol therapy. The primary analysis of the phase III question was based on a two-sided log-rank test²¹ with 5% type I error stratified by tumor type, extent of resection, and M stage. On the basis of 284 randomly assigned patients accrued over 4 years, with minimum follow-up of 1 year and current median follow-up 6.6 years, this comparison had in excess of 85% power to detect a reduction in treatment failure rate that would correspond to an increase in long-term EFS rate from 28% to 44% under the proportional hazards assumption.²² Other comparisons of EFS and survival between groups also were based on the log-rank test.²¹ In some analyses, patients were classified into the following three prognostic groups: M0 patients with minimum residual disease (defined as > 90% resection and < 1.5 cm of residual tumor) who were to have received radiation therapy per protocol only in cases of tumor progression; other M0 patients who were to have received radiation only if they did not have a CR to induction; and M1+ patients (disseminated). A Cox regression model with the time to start radiation therapy as a time-dependent covariate²¹ was used to investigate the efficacy of radiation therapy. The percent of patients who ultimately received radiation therapy, after reaching their scheduled radiation therapy time without progression, was estimated using the product-limit estimate to correctly account for patient censoring. A Cox regression analysis in which all patients were censored at 1 year also was used to test differences in EFS during the first year to make inference about early disease control while controlling for prognostic factors.

	RESULTS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors’ Disclosures of... REFERENCES

 
Study Patients
Between April 1993 and June 1997, 299 infants with malignant brain tumors were enrolled onto CCG-9921. Fifteen patients were excluded for eligibility reasons; three had ineligible pathologic diagnoses (one low-grade astrocytoma, one desmoplastic infantile ganglioglioma, and one inconclusive diagnosis based on CSF cytology only), eight had inadequate staging of spinal or metastatic disease (four had inadequate spinal imaging and four had no CSF examinations), three had more than a 28-day delay in start of chemotherapy, and one child was older than 36 months.

Demographic and disease characteristics of the 284 eligible patients are listed in Table 1. Twenty-five percent of patients were younger than 1 year old. Medulloblastoma was the most prevalent tumor type (32%), and ependymoma was the second most prevalent type (26%). Most tumors were infratentorial (68%). There were no significant imbalances in any patient or tumor features between treatment groups. Median follow-up time among patients who had no treatment failure events was 6.6 years, with at least 4.8 years of follow-up available for 80% of patients (actuarial estimates).

Treatment Failure Events
There were 209 treatment failure events overall in the study; 121 were local progression or recurrence, 20 were distant progression without local progression, 45 were combined local and distant progression, one was an extraneural recurrence, and 10 were deaths attributed to progressive disease or complications of disease, such as severe neurologic and respiratory dysfunction, but without a reported distinct antecedent disease progression event. In addition, seven patients died from infections (six on regimen A and one on regimen B), and five patients had second neoplasms as a first event.

Table 3 lists failure types within each of the tumor types. Overall, 11% (20 of 186 patients) of intraneural relapses (local, distant, or local plus distant) of known type were not local recurrences, and those relapses were most frequent among medulloblastomas, supratentorial PNETs, and rhabdoid tumors (P = .06, ² test of homogeneity among diagnoses of distant v local or local plus distant). The five second neoplasms comprised myelodysplastic syndrome (MDS)/acute myelogenous leukemia (AML) at 40 months from study entry (regimen A), AML at 51 months (regimen B), MDS at 20 months (regimen A), non-Hodgkin’s lymphoma at 10 months (regimen A), and a rhabdoid kidney tumor at 5 months in a disseminated pineal rhabdoid PNET patient.

View larger version (4K): [in this window] [in a new window]   Fig 1. Event-free survival and survival of 284 patients.

EFS and Overall Survival for Medulloblastoma, Supratentorial PNET, Ependymoma, and Rhabdoid Tumors
EFS and survival rates are shown in Figures 2 and 3 for the four largest homogeneous groups of patients (medulloblastoma, supratentorial PNET, ependymoma, and rhabdoid tumors). Five-year EFS rates in the medulloblastoma, supratentorial PNET, ependymoma, and rhabdoid tumor groups were 32% ± 5%, 17% ± 6%, 32% ± 6%, and 14% ± 7%, respectively. Differences in outcomes between groups were statistically significant overall (EFS, P = .0003; survival, P = .0008, log-rank test) and after controlling for differences in extent of resection (< 90% v 90%), amount of residual disease ( 1.5 cm v > 1.5 cm), and M stage (M0 v M1+; EFS, P = .0028; survival, P = .0027, stratified log-rank test).

View larger version (4K): [in this window] [in a new window]   Fig 2. Event-free survival by pathology. Supra, supratentorial; PNET, primitive neuroectodermal tumors.

View larger version (4K): [in this window] [in a new window]   Fig 3. Survival by pathology. Supra, supratentorial; PNET, primitive neuroectodermal tumors.

View larger version (4K): [in this window] [in a new window]   Fig 4. Event-free survival by regimen in patients with medulloblastoma.

Outcome in Ependymoma
The 1- and 5-year EFS rates for ependymoma were 72% ± 5% and 32% ± 6%, respectively. Survival rate at 5 years was 59% ± 6%. Incomplete resection was prognostically significant for EFS (RFR = 2.3; 95% CI, 1.2 to 4.5; P = .020). No other factor reached statistical significance, although disseminated patients had treatment failures at a nominally higher rate (RFR = 1.3; 95% CI, 0.59 to 2.8).

Outcome in Rhabdoid Tumors
The 1- and 5-year EFS rates for rhabdoid tumors were 32% ± 9% and 14% ± 7%, respectively. Survival rate at 5 years was 29% ± 9%. No factor was significantly associated with prognosis in this small subgroup of patients

Outcomes in Other Histologic Tumor Types
Seven of nine patients with choroid plexus carcinoma had treatment failure (five patients within 1 year of study entry). Four patients have died. EFS and survival rates at 3 years were 33% ± 16% and 63% ± 17%, respectively. The other two children were alive at 70 months.

Five of eight patients with high-grade glioma had treatment failure, and all died (3-year EFS and survival rates: 38% ± 17% and 33% ± 18%, respectively). Thirteen of 14 patients with intrinsic brainstem glioma had treatment failures, and only two patients were alive at last contact (EFS rate, 7% ± 7%; survival rate, 29% ± 12% at 3 years). Among 13 patients with other tumor types, nine had treatment failures, and seven died (3-year EFS rate, 38% ± 13%; and 3-year survival rate, 46% ± 14%).

Radiation Therapy
Table 4 also lists data on radiation treatment in this study. Among 284 patients on the study, 125 received radiation therapy, 27 before disease progression (at a median 55 weeks from initial diagnosis) and 98 after progression (at a median of 53 weeks from initial diagnosis and a median of 4 weeks from progression). Among patients who were event free at 5 years, 77% had not received radiation therapy, and among patients alive 5 years from study entry, 58% had not received radiation therapy. The percentage of patients who did not receive radiation therapy did not differ by tumor type (P = .36). In medulloblastoma and ependymoma patients, 83% and 63%, respectively, of 5-year event-free survivors had not received radiation.

Forty patients who either had metastatic disease at diagnosis or who had insufficient response to induction therapy were observed on protocol and were event free long enough to reach the date at which radiation therapy was due (eg, minimum time to third birthday or end of chemotherapy). Six patients received radiation therapy a median of 10 weeks before the theoretical due date (range, 1 to 15 weeks). Of the remaining 34 patients, 11 progressed or died after the due date but before receiving radiation therapy, 14 were last seen before having received radiation therapy, and nine were treated with radiation therapy a median of 5 weeks after the theoretical due date (range, 0 to 68 weeks). There was a 2.2-fold lower risk of recurrence or death in patients who received radiation therapy compared with patients who did not, but this effect was not statistically significant (P = .19, Cox analysis with radiation therapy as time-dependent covariate). Including patients who received early radiation therapy and correcting for patient censoring, an estimated 40% of patients who reached the scheduled radiation therapy time without disease progression were ultimately treated with radiation therapy at a median time of 54 weeks. This compliance rate did not differ between patients with metastatic disease and patients without metastatic disease. Overall, 62% ± 8.2% of these patients were event free 5 years after the due date, and five of 12 patients who were event free at that time had not received radiation therapy.

Seventy-two patients were to have received radiation therapy because of metastatic disease at diagnosis. Thirty-eight ultimately received radiation therapy (11 patients before disease progression). Among patients who were alive and event free at 5 years, 54% had not received radiation therapy, and of those patients who were alive at 5 years, 50% had not received radiation therapy.

All patients whose disease progressed and who had not previously received radiation therapy were to have proceeded immediately to radiation therapy in this study. Among 187 patients who had clearly defined disease progression, six patients had received radiation before progression, 98 received radiation after progression (average, 9 weeks; range, 0 to 70 weeks), and 83 had not received radiation before death or last contact. Among the 181 patients who had not received radiation before progression, the 3-year postprogression survival (PPS) rate was 24% ± 3% overall. PPS differed significantly by age at the time of progression; 3-year PPS rate was 15% ± 5% for patients younger than 24 months old, 24% ± 6% for patients between 24 and 35 months old, and 37% ± 7% for patients 36 months old or older (P < .001, log-rank test for trend). Patients who progressed who received radiation therapy had a 1.5-fold lower risk of subsequent death compared with patients who did not receive radiation therapy at a comparable time after progression (P = .040, Cox analysis with radiation therapy as time-dependent covariate). Of the patients who received radiation therapy within 30 days of progression, 23% ± 6% were alive 3 years after progression. Among 37 patients alive and in follow-up 3 years after progression, eight (22%) had never received radiation. This fraction differed according to age at progression (five of eight patients younger than 24 months old, one of 12 patients between 24 and 35 months old, and two of 17 patients 36 months or older; P < .01).

In patients with medulloblastoma, the 3-year PPS rate was 20% ± 6%. All nine patients who were alive and still in follow-up at 3 years had received radiation therapy. Those patients who received radiation therapy had a 2.5-fold lower risk of death (P = .017), with 33% ± 12% of patients treated immediately alive at 3 years. For ependymoma, 3-year PPS rate was 43% ± 8%. Of 14 patients still alive and being observed at 3 years, 12 (86%) had received radiation therapy.

One hundred twenty-one of 125 patients who received radiation therapy were reviewed for adherence to protocol radiation therapy guidelines. Forty-three patients (35.5%) were to receive local radiation alone, and 78 (64.5%) were to receive local plus CS radiation; radiation field data were available for review in 115 patients (42 and 73 patients, respectively). Forty (95.2%) of 42 patients received the prescribed local radiation only, with two patients receiving local plus CS radiation. Sixty-three (86.3%) of 73 patients received prescribed local plus CS irradiation, with nine patients receiving local radiation only and one patient receiving cranial irradiation. Hence, overall, there was 89.6% compliance (103 of 115 patients) with prescribed radiation field among patients who received radiation therapy.

Local radiation dose data were reviewed in 115 patients. Eighty-five patients (73.9%) received the prescribed dose, 13 (11.3%) received a minor deviation from this dose, and 17 (14.8%) received a dose that deviated in a major way. CS dose data were reviewed for 67 patients; 49 patients (73.1%) received the prescribed dose, three patients received a minor deviation from this dose, and 15 patients (22.4%) received a dose that deviated in a major way.

Toxicity and Complications of Treatment
As expected, there was significant toxicity associated with this intensive chemotherapeutic regimen. Most patients on both arms had grade 3 or 4 hematologic toxicity during induction (Table 5). Twice as many patients on regimen A compared with regimen B (14% v 7%, respectively) had significant infections during induction. Six patients on regimen A (five during induction) and one patient on regimen B (one during induction) died from infections.

Median time to complete induction was 19 weeks for regimens A and B. Ninety percent of patients in regimens A and B completed induction within 25 and 23 weeks, respectively. Second malignant neoplasms occurred in three patients on regimen A (AML and MDS) and one patient in regimen B (MDS).

	DISCUSSION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors’ Disclosures of... REFERENCES

 
CCG-9921 was designed as a phase II randomized trial to evaluate the response rates and toxicities of two promising regimens for infants with malignant brain tumors in a larger cohort of patients. In addition, we wished to evaluate the strategy of withholding radiation therapy before tumor progression in children without metastasis at diagnosis and with no residual tumor after induction chemotherapy.

Activity of both regimens was confirmed in our study, particularly in patients with medulloblastoma, supratentorial PNET and ependymoma. Although the response rate for regimen A was nominally higher than for regimen B, it was not a statistically significant difference.

For all patients, the 2-year EFS and survival rates (32% and 60%, respectively) were comparable to a prior Pediatric Oncology Group (POG) study (35% and 55%, respectively).¹⁷ EFS was also similar between the POG study and this study for medulloblastoma (34% v 39%, respectively) and ependymoma (42% v 45%, respectively). However, almost all surviving patients in the POG study received radiation, whereas 58% of patients who survived at least 5 years in our study did not receive radiation.¹⁷ Neurocognitive assessment of surviving patients continues and will be the subject of a future publication.

This study was not designed initially to compare outcomes between regimens, which differed only in the initial 15 weeks of therapy. However, as described in the introduction, the study was redesigned to include a formal randomized phase III comparison between the treatments. The study had abundant power to detect an outcome difference but did not find significant differences in outcomes between arms.

Although the study was not designed to compare treatment regimens within tumor types, there is a suggestion that regimen A was more effective in early disease control in medulloblastoma. Sixty-one percent of medulloblastoma patients treated with regimen A had no disease progression 1 year after diagnosis compared with 41% treated with regimen B (P = .059) and 42% for the POG study. The early improved tumor control with regimen A did not translate into significant long-term survival improvement, but it does suggest that the regimen might be considered for induction, before additional consolidation therapy.

As was found in the POG study,¹⁷ extent of resection was of prognostic significance. Patients with medulloblastoma, ependymoma, and rhabdoid tumors who had minimal residual disease after resection had nominally better EFS, although this was statistically significant only for patients with ependymoma. Differing from the POG study, most patients with complete resections in our study ultimately had tumor progression.¹⁷ Also differing from the POG study in which patients with postoperative residual tumor who had a CR to chemotherapy had similar results to patients with complete surgical resection, there was no difference in outcome in our study between those patients who had a CR to chemotherapy and those patients who completed induction but had less than a CR.

In most patients, recurrence was at least local; isolated leptomeningeal recurrence occurred in only 15% of medulloblastoma and supratentorial PNET patients and rarely among patients with other tumor types. This suggests that more effective local tumor control might increase survival. Recent improvements in radiation-targeting techniques may improve local tumor control with acceptable sequelae, even in young children, and this is under investigation in nonmetastatic medulloblastoma in an ongoing Children’s Oncology Group study.

Supratentorial PNETs and medulloblastomas are quite similar histopathologically, and it has been suggested that the only difference between them is location. However, in this study, as in the POG study, children with supratentorial PNETs had significantly worse outcomes than children with medulloblastoma, which is a difference not readily attributable to other known prognostic factors and which suggests that they are different biologic entities.

The atypical teratoid/rhabdoid tumor, which has been described only recently^23-25 and is associated with mutations of INI1,²⁶ also was associated with poor prognoses. In our study, in which histopathologic tumor diagnoses were centrally reviewed, atypical teratoid/rhabdoid tumors were approximately one fifth as frequent as medulloblastoma and supratentorial PNET. Primary tumor site and frequency of metastatic disease at diagnosis did not differ between medulloblastoma and supratentorial/PNET patients. However, median age at diagnosis and median time to tumor progression did differentiate these patients, as did survival, compared with medulloblastoma. It is likely that these tumors were classified as medulloblastoma or supratentorial PNET in previous studies.

Although this study mandated radiation therapy before tumor progression in patients who had metastatic disease at diagnosis or residual tumor after induction, compliance was poor. Only 40% of such patients received radiation therapy as intended before progression. That suggests reluctance among treating physicians to recommend radiation therapy for these patients.

After tumor progression, 55% of nonirradiated patients received radiation therapy. Although the death rate after tumor progression was lower for patients who received irradiation, the benefit of immediate postprogression radiation may not be accurately estimated in this analysis. Some children might have been considered poor candidates for irradiation at the time of progression, which would cause an overestimation of the benefit. However, many patients who received postprogression radiation therapy had substantial intervals of further therapy before radiation, which could cause underestimation of the benefit. Even so, in patients with medulloblastoma, a relatively radiosensitive tumor, only one third of patients receiving radiation therapy at the time of progression were treated successfully with salvage radiotherapy, which is consistent with other studies of salvage radiotherapy in patients with this tumor.

Unfortunately, the intensified chemotherapy in this protocol did not result in a higher rate of survival compared with previous studies. However, survival was comparable, and most surviving patients did not receive irradiation. Clinical prognostic factors, such as histology, extent of resection, and metastatic disease, were insufficient to distinguish accurately this subset of infants. Biologic differences between tumors of the same histologic type, such as Trc C and MYL expression in medulloblastoma,^27,28 recently have been correlated with outcome. It is possible that further study of such biologic tumor differences will identify patients who can be treated successfully without irradiation. However, new treatment strategies are needed to improve prognoses of most infants with malignant brain tumors, perhaps including earlier use of focal radiation therapy.

	Authors’ Disclosures of Potential Conflicts of Interest

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors’ Disclosures of... REFERENCES

 
The author or immediate family members indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. For a detailed discription of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.

Authors Employment Leadership Consultant Stock Honoraria Research Funds Testimony Other Joel W. Goldwein IMPAC Medical Systems, Inc (N/R) Richard Sposto Merck (A)

Dollar Amount Codes (A) < $10,000 (B) $10,000–99,000 (C) $100,000 (N/R) Not Required

	NOTES

 
Supported by National Cancer Institute Grant No. CA 10382.

Authors’ disclosures of potential conflicts of interest are found at the end of this article.

	REFERENCES

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors’ Disclosures of... REFERENCES

 
1. Smith MA, Gurney JG, Linet M, et al (eds): Cancer incidence and survival among children and adolescents: United States SEER Program 1975-1995, National Cancer Institute, Bethesda, MD, 1999

2. Zeltzer PM, Boyett JM, Finlay JL, et al: Metastasis stage, adjuvant treatment and residual tumor are prognostic factors for medulloblastoma in children: Conclusions from the Children’s Cancer Group 921 randomized phase III study. J Clin Oncol 17:832-845, 1999[Abstract/Free Full Text]

3. Cohen BH, Zeltzer PM, Boyett JM, et al: Prognostic factors and treatment results for supratentorial primitive neuroectodermal tumors in children using radiation and chemotherapy: A Children’s Cancer Group randomized trial. J Clin Oncol 13:1687-1696, 1995[Medline]

4. Cohen BH, Packer RJ, Siegel KR, et al: Brain tumors in children under 2 years: Treatment, survival, and long-term prognosis. Pediatr Neurosurg 19:171-179, 1993[Medline]

5. Di Rocco C, Iannelli A, Ceddia A: Intracranial tumors of the first year of life: A cooperative survey of the 1986-1987 Education Committee of the ISPN. Childs Nerv Syst 7:150-153, 1991[CrossRef][Medline]

6. Di Rocco C, Iannelli A, Papacci F, et al: Prognosis of medulloblastoma in infants. Childs Nerv Syst 13:388-396, 1997[Medline]

7. Ilveskoski I, Pihko H, Sankila R, et al: Improving outcome of malignant brain tumours in very young children: A population-based study in Finland during 1975-93. Acta Paediatr 86:724-729, 1997[Medline]

8. Massimino M, Gandola L, Cefalo G, et al: Management of medulloblastoma and ependymoma in infants: A single-institution long-term retrospective report. Childs Nerv Syst 16:15-20, 2000[CrossRef][Medline]

9. Kiltie A, Lashford L, Gattamaneni H: Survival and late effects in medulloblastoma patients treated with craniospinal irradiation under three years old. Med Pediatr Oncol 28:348-354, 1997[CrossRef][Medline]

10. Walter A, Mulhern RK, Gajjar A, et al: Survival and neurodevelopmental outcome of young children with medulloblastoma at St Jude Children’s Research Hospital. J Clin Oncol 17:3720-3728, 1999[Abstract/Free Full Text]

11. Kellie SJ: Chemotherapy of central nervous system tumours in infants. Childs Nerv Syst 15:592-612, 1999[CrossRef][Medline]

12. Duffner PK, Kun LE, Burger PC, et al: Postoperative chemotherapy and delayed radiation in infants and very young children with choroid plexus carcinomas. Pediatr Neurosurg 22:189-196, 1995[Medline]

13. Geyer J, Zeltzer PM, Boyett JM, et al: Survival of infants with primitive neuroectodermal tumors or malignant ependymomas of the CNS treated with eight drugs in 1 day: A report from the Children’s Cancer Group. J Clin Oncol 12:1607-1615, 1994[Abstract/Free Full Text]

14. Duffner PK, Krischer JP, Burger PC, et al: Treatment of infants with malignant gliomas: The Pediatric Oncology Group experience. J Neurooncol 28:245-256, 1996[Medline]

15. Geyer J, Finlay JL, Boyett JM, et al: Survival of infants with malignant astrocytomas: A report from the Children’s Cancer Group. Cancer 75:1045-1050, 1995[CrossRef][Medline]

16. Duffner P, Krischer JP, Sanford RA, et al: Prognostic factors in infants and very young children with intracranial ependymomas. Pediatr Neurosurg 28:215-222, 1998[CrossRef][Medline]

17. Duffner P, Horowitz ME, Krischer JP, et al: Postoperative chemotherapy and delayed radiation in children less than three years of age with malignant brain tumors. N Engl J Med 328:1725-1731, 1993[Abstract/Free Full Text]

18. Albright A, Wisoff JH, Zeltzer PM, et al: Effects of medulloblastoma resections on outcome in children: A report from the Children’s Cancer Group. Neurosurgery 38:265-271, 1996[CrossRef][Medline]

19. Geyer R, Berger M, Allen J, et al: Intensive chemotherapy pilot for infants with malignant brain tumors. Proc Am Soc Clin Oncol 12:417, 1993 (abstr 1431)

20. International Society of Pediatric Oncology: SIOP XXIII Meeting. Med Pediatr Oncol 19:406, 1991

21. Cox D, Oakes D: Analysis of Survival Data. London, United Kingdom, Chapman and Hall, 1984

22. Sposto R, Sather H: Determining the duration of comparative clinical trials while allowing for cure. J Chronic Dis 38:683-690, 1985[CrossRef][Medline]

23. Hilden J, Watterson J, Longee DC, et al: Central nervous system atypical teratoid tumor/rhabdoid tumor: Response to intensive therapy and review of the literature. J Neurooncol 40:265-275, 1998[CrossRef][Medline]

24. Bhattacharjee M, Hicks J, Langford L, et al: Central nervous system atypical teratoid/rhabdoid tumors of infancy and childhood. Ultrastruct Pathol 21:369-378, 1997[Medline]

25. Biegel J, Rorke LB, Packer RJ, et al: Monosomy 22 in rhabdoid or atypical tumors of the brain. J Neurosurg 73:710-714, 1990[Medline]

26. Biegel JA, Zhou JY, Rorke LB, et al: Germ-line and acquired mutations of INI1 in atypical teratoid and rhabdoid tumors. Cancer Res 59:74-79, 1999[Abstract/Free Full Text]

27. Grotzer M, Hogarty MD, Janss AJ, et al: MYC messenger RNA expression predicts survival outcome in childhood primitive neuroectodermal tumor/medulloblastoma. Clin Cancer Res 7:2425-2433, 2001[Abstract/Free Full Text]

28. Grotzer M, Janss AJ, Fung K, et al:, TrkC expression predicts good clinical outcome in primitive neuroectodermal brain tumors. J Clin Oncol 18:1027-1035, 2000[Abstract/Free Full Text]

Submitted August 30, 2004; accepted July 26, 2005.

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This article has been cited by other articles:

				G. Dhall Medulloblastoma J Child Neurol, November 1, 2009; 24(11): 1418 - 1430. [Abstract] [PDF]

				S. Zacharoulis and L. Moreno Ependymoma: An Update J Child Neurol, November 1, 2009; 24(11): 1431 - 1438. [Abstract] [PDF]

				M. L. Garre, A. Cama, F. Bagnasco, G. Morana, F. Giangaspero, M. Brisigotti, C. Gambini, M. Forni, A. Rossi, R. Haupt, et al. Medulloblastoma Variants: Age-Dependent Occurrence and Relation to Gorlin Syndrome--A New Clinical Perspective Clin. Cancer Res., April 1, 2009; 15(7): 2463 - 2471. [Abstract] [Full Text] [PDF]

				L. Lafay-Cousin and D. Strother Current Treatment Approaches for Infants with Malignant Central Nervous System Tumors Oncologist, April 1, 2009; 14(4): 433 - 444. [Abstract] [Full Text] [PDF]

				S. Rutkowski, N. U. Gerber, K. von Hoff, A. Gnekow, U. Bode, N. Graf, F. Berthold, G. Henze, J. E.A. Wolff, M. Warmuth-Metz, et al. Treatment of early childhood medulloblastoma by postoperative chemotherapy and deferred radiotherapy Neuro-oncol, January 1, 2009; 11(2): 201 - 210. [Abstract] [Full Text] [PDF]

				R. J. Packer and G. Vezina Management of and Prognosis With Medulloblastoma: Therapy at a Crossroads Arch Neurol, November 1, 2008; 65(11): 1419 - 1424. [Abstract] [Full Text] [PDF]

				R. J. Packer Childhood Brain Tumors: Accomplishments and Ongoing Challenges J Child Neurol, October 1, 2008; 23(10): 1122 - 1127. [Abstract] [PDF]

				N. G. Gottardo and A. Gajjar Chemotherapy for Malignant Brain Tumors of Childhood J Child Neurol, October 1, 2008; 23(10): 1149 - 1159. [Abstract] [PDF]

				I. F. Pollack Diagnostic and Therapeutic Stratification of Childhood Brain Tumors: Implications for Translational Research J Child Neurol, October 1, 2008; 23(10): 1179 - 1185. [Abstract] [PDF]

				T. J. MacDonald Aggressive Infantile Embryonal Tumors J Child Neurol, October 1, 2008; 23(10): 1195 - 1204. [Abstract] [PDF]

				L. Ridley, R. Rahman, M.-A. Brundler, D. Ellison, J. Lowe, K. Robson, E. Prebble, I. Luckett, R. J. Gilbertson, S. Parkes, et al. Multifactorial analysis of predictors of outcome in pediatric intracranial ependymoma Neuro-oncol, January 1, 2008; 10(5): 675 - 689. [Abstract] [Full Text] [PDF]

				V. Larouche, M. Capra, A. Huang, U. Bartels, and E. Bouffet Supratentorial Primitive Neuroectodermal Tumors in Young Children J. Clin. Oncol., December 10, 2006; 24(35): 5609 - 5609. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Geyer, J. R. Articles by Shiminski-Maher, T. Search for Related Content PubMed PubMed Citation Articles by Geyer, J. R. Articles by Shiminski-Maher, T. Social Bookmarking                            What's this?