OUTCOME OF PATIENTS WITH RELAPSED OR REFRACTORY NON-RHABDOMYOSARCOMA SOFT TISSUE SARCOMAS ENROLLED IN PHASE II COOPERATIVE GROUP CLINICAL TRIALS: A REPORT FROM THE CHILDREN’S ONCOLOGY GROUP

Abstract

Purpose:

We aimed to estimate event-free survival (EFS) of children and young adults with relapsed or refractory NRSTS treated in non-randomized phase II studies conducted by Children’s Oncology Group (COG) and predecessor groups to establish a benchmark EFS for future phase II NRSTS trials evaluating the activity of novel agents.

Methods:

A retrospective analysis of recurrent or refractory NRSTS patients prospectively enrolled in non-randomized phase II COG and predecessor group trials between 1994 and 2015 was conducted. The EFS was defined as disease progression/relapse or death and calculated using the Kaplan-Meier Method. The Log-rank test and relative-risk regression were used to compare EFS distribution by age at enrollment, sex, race, NRSTS histology, prior lines of therapy, calendar year of trial and type of radiographic response.

Results:

In total, 137 patients were enrolled in thirteen phase II trials. All trials used radiographic response rate as a primary outcome, and none of the agents utilized were considered active based on trial-specified thresholds. The estimated median EFS and 6-month EFS of entire study cohort was 1.5 months [95% Confidence Interval (CI) 1.3-1.8] and 19.4% (95% CI: 12.7%-26%), respectively. No difference in EFS was observed by age at enrollment, sex, race, NRSTS histology subtype, prior lines of therapies and trial initiation year. EFS significantly differed by radiographic response.

Conclusion:

EFS for children and young adults with relapsed or refractory NRSTS remains suboptimal. Established EFS can be referenced as a benchmark for future single-agent phase II trials incorporating potentially active novel agents in this population.

Precis:

The survival outcomes of relapsed/refractory children and young adults with non-rhabdomyosarcoma soft tissue sarcoma (NRSTS) remain unknown. Our results show that the survival remains suboptimal for these patients and provide the benchmark survival estimates for future phase II trials testing novel agents in this population.

INTRODUCTION

Non-rhabdomyosarcoma soft tissue sarcomas (NRSTS) are a heterogeneous group of soft tissue sarcomas accounting for 4% of all childhood neoplasms.¹ More than 50 distinct types of NRSTS histologies are described among children with varied clinical presentation, biology and treatment. Due to the extensive histological heterogeneity and rarity of NRSTS among children, only a few prospective clinical trials have been conducted in the Cooperative Group setting for newly diagnosed NRSTS among children.^2-4

The two recently completed Children’s Oncology Group (COG) NRSTS trials, ARST0332 and ARST1321, employed a risk-adapted multimodality strategy and examined the role of pazopanib in combination with chemotherapy, respectively, to treat children and young adults with NRSTS. The 5-year event-free survival (EFS) and overall survival (OS) of patients treated in these trials were 52.5-68% and 75.7%-79.4%, respectively.³ Patients with metastatic disease had dismal outcomes. Similar to North American trials, the children and young adults with non-metastatic NRSTS had 5-year EFS and OS of 73.7% and 83.8%, in the European Pediatric Soft Tissue Sarcoma Study Group (EpSSG) NRSTS 2005 study and those with metastatic disease had unfavorable outcomes.^2,5

These prospective trials demonstrated that nearly 25% to 30% of children and young adults with NRSTS suffer a relapse or progression following first-line treatment.^2,3 Unfortunately, despite the development of newer agents, minimal salvage treatment options exist, resulting in poor survival.⁶ Furthermore, most data for relapsed/refractory NRSTS tumors come from adult trials.^7-9 To date, no Cooperative Group clinical trials have solely focused on treating children with relapsed or refractory NRSTS. Most of these patients have been included among children and young adults with other relapsed/refractory cancers in prospective phase II trials. Consequently, the data on the outcomes within this relapse/refractory population are limited.¹⁰

To design future phase II trials to improve the outcomes of children and young adults with relapsed or refractory NRSTS, it is imperative to understand their survival outcomes on previously conducted studies. COG has recently completed similar analyses for patients with relapsed/refractory osteosarcoma and Ewing sarcoma. These analyses have been critical in developing trials for relapsed/refractory osteosarcoma and Ewing sarcoma by providing the historical progression rates and EFS rates.^11,12 Thus, in this analysis, we sought to pool the data for the children and young adults with relapsed or refractory NRSTS that were treated on previous non-randomized phase II COG, Pediatric Oncology Group (POG) and Children’s Cancer Group (CCG) trials to establish benchmark EFS rates that can be used to design future phase II NRSTS trials exploring the activity of potentially novel agents among this population. Our secondary objectives were to compare the EFS by the pre-defined trial-, patient- and disease-related characteristics and type of radiographic response and to estimate the OS.

METHODS

We reviewed the study protocols, final study reports and publications of all-phase II trials conducted by CCG, POG and COG between 1994 and 2015 for the inclusion of patients aged 0 to 39 years with relapsed or refractory NRSTS irrespective of histology, for this analysis. The local institutional research ethics boards approved these phase II studies. We collected the following trial-level information for each eligible trial: eligibility criteria, calendar year of trial initiation, number of patients with NRSTS enrolled, NRSTS histology, name and dose of study drug, primary study endpoint, radiographic response criteria [Response Evaluation Criteria in Solid Tumors (RECIST) vs. World Health Organization (WHO)], timing of the response assessment, response rate and study-defined drug activity. The patient-level variables extracted for each eligible patient included the age of the patient at trial enrolment, sex, race, number of prior lines of therapy, NRSTS histology, type of radiographic response, treatment-related toxicity, type and date of outcome (relapse, progression, death or alive), cause of death and date of last follow up.

Statistical Analysis

We included all NRSTS patients enrolled in the identified non-randomized phase II trials irrespective of their evaluability of the primary study endpoint. The enrolled patients were observed for relapse/progression and death until loss to follow-up or at least five years after enrollment (whichever occurred first). The EFS was defined as the time from study enrollment until the date of last contact, disease progression, relapse, or death, whichever occurred first. We used the Kaplan-Meier Method to calculate EFS and the Peto-Peto method to calculate the 95% Confidence Interval (CI). The study records were reviewed to determine the reason for terminating the protocol therapy. Patients with disease progression, relapse or death, or those dying after stopping the study treatment due to the patient or caregiver preference or unacceptable toxicity, were considered to have an event. Based on available data, we were unable to determine if unacceptable toxicity was protocol-defined or patient/caregiver-declared. Patients not experiencing an event were censored at the date of the last follow-up. The Log-rank test was used to compare the EFS distribution between prespecified subgroups based on age at enrollment, sex, race, NRSTS histology, prior lines of therapy, calendar year of trial initiation and type of best radiographic response. A relative risk regression model was run to evaluate the risk for EFS events across these groups.¹³ The OS was defined as the time from study enrollment to death due to any cause. For OS, patients were censored at their last contact date. A two-sided p-value < 0.05 was considered statistically significant.

RESULTS

Eligibility and Patient Characteristics

We identified 15 non-randomized phase II trials conducted by POG, CCG and COG, enrolling 152 NRSTS patients between 1994 and 2015. Patient level data was not available for 14 patients treated on two POG trials (POG9361 and POG9464), and for one patient, the reason for coming off treatment and death was not apparent; therefore, the final study cohort included 137 patients [median age 17 years (range 1.6-26.9 years)] with relapsed/refractory NRSTS patients treated on 13 non-randomized phase II trials (Table 1). All but one trial (ADVL1221) tested the activity of a single agent in this population. Only two patients were enrolled in more than one phase II trial; one was enrolled first in the POG9963 trial, followed by A09713, and the second patient was enrolled in POG9963 first and then ADVL0421. We considered each enrollment as an independent observation for this analysis. Patients > 18 years old represented 36.5% of the study population. Specific NRSTS histology was not reported for 57 (71.2%) patients; for the remaining 80 patients, synovial sarcoma (SS) (N=41, 29.9%) was the most common diagnosis, followed by desmoplastic small round cell tumor (DSRCT) (N=16, 11.7%) and malignant peripheral nerve sheath tumor (MPNST) (N=13, 9.5%). The RECIST and WHO criteria assessed the radiographic response in 8 (61.5%) and 5 (38.5%) trials, respectively. Of the 121 patients with evaluable radiographic response, 96 (70.1%) had progressive disease (PD), 22 (16.1%) had stable disease (SD), 3 (2.2%) achieved complete response (CR), while no patient achieved a partial response. None of the agents utilized in the trials were considered active based on the trial-specified response rates (Table 1). The median duration of the follow-up for patients who remained event-free at the end of the follow-up was 3.1 years, whereas the median follow-up time for those who were alive at the end of the follow-up was 4.9 years.

Table 1.

Characteristics and Outcomes of the Patients with Relapsed/Refractory NRSTS tumors Treated on Non-randomized Phase II Trials Conducted through the Children’s Oncology Group and Predecessor Groups from 1994-2015 (N=13).

Study	Study (Years Open)	Eligibility age in years	Prior lines of therapy	Study drug	Study’s Endpoint (Criteria)	No. of NRSTS patients enrolled	NRSTS histology (No. of patients)	No. of patients with evaluable response	Type of best response (No. of patients)	Drug activity as per study endpoint
ADVL1522 ( NCT02452554)14	2015-2016	1≤ to ≤ 30	1 or more lines	Lorvotuzumab Mertansine	Response rate (RECIST)	15	SS(10), MPNST(5)	15	PD(15)	No activity
ADVL1221 (NCT01614795)15	2012-2013	1≤ to ≤ 30	1 or more lines	Cixutumumab and temsirolimus	Response rate (RECIST)	12	Not specified	8	PD(5), SD(3)	No activity
ADVL0921 (NCT01154816)16	2011-2013	1< to <22	1 or more lines	Alisertib	Response rate (RECIST)	8	Not specified	7	PD(5), SD(2)	No activity
ADVL0821 (NCT00831844)17	2009-2012	1≤ to ≤ 31	1 or more lines	Cixutumumab	Response rate (RECIST)	11	SS(11)	10	PD(6), SD(4)	No activity
ADVL0221 (NCT00070109)18	2008-2010	1≤ to ≤ 21	1 or more lines	Trabectedin	Response rate (RECIST)	11	Not specified	10	PD(8), SD(2)	No activity
ADVL0524 (NCT00331643)19	2006-2007	1≤ to ≤35	1 or more lines	Ixabepilone	Response rate (RECIST)	10	SS(6), MPNST(3), Spindle cell sarcoma(1)	10	PD(6), SD(4)	No activity
ADVL0421 (NCT00091182)20	2004-2006	≤ 21	1 or more lines	Oxaliplatin	Response rate (RECIST)	11	Not specified	11	PD(7), SD(4)	No activity
ADVL0122 (NCT00030667)21	2002-2004	< 30	1 or more lines	Imatinib	Response rate (RECIST)	19	DSRCT (13), SS (6)	15	PD (14), SD(1)	No activity
P9963 (NCT00006102) 22	2000-2004	< 22	1 or 2 lines	Rebeccamycin	Response rate (WHO)	12	DSRCT(2), SS(3), Undifferentiated sarcoma(2), spindle cell sarcoma(1), Fibrosarcoma(1), GCMFH (1), MPNST(1), Not specified(1)	12	PD(12)	No activity
POG9761 (NCT00004078) 23	1999-2005	1< to <22	1 or 2 lines	Irinotecan	Response rate (WHO)	5	DSRCT(1), ES(1), MPNST(1), STS NOS(1), SS(1)	5	CR(3), MR(2)	No activity
A09713 (NCT00003745)24	1999-2003	≤ 21	1 or more lines	Topotecan	Response rate (WHO)	3	Not specified	3	PD(2), unclear(1)	No activity
A09705 (NCT00003234)25	1998-2002	≤ 21	1 or 2 lines	Vinorelbine	Response rate (WHO)	9	MPNST(1), STS NOS(6), SS (2)	9	PD(7), SD(2)	No activity
COG0962 (NCT00002825)26	1997-2001	≤ 21	1 or 2 lines	Docetaxel	Response rate (WHO)	11	MPNST(2), Not specified (4), Undifferentiated sarcoma(3), SS(2),	11	PD(9), NR(2)	No activity

Abbreviations: CR: Complete Response; DSRCT: Desmoplastic Small Round Cell Tumor; ES: Epithelioid Sarcoma; GCMFH: Giant cell malignant fibrous histiocytoma MPNST: Malignant Peripheral Nerve Sheath Tumor; MR: Minor Response; PD: Progressive Disease; RECIST: Response Evaluation Criteria in Solid Tumors; SS: Synovial Sarcoma; STS-NOS: Soft Tissue Sarcoma – Not Otherwise Specified; SD: Stable Disease; WHO: World Health Organization

Event-Free Survival

Of the 137 patients, 128 patients had an event (progression/relapse = 109 and death =19). None of the patients had death related to treatment-related toxicity; however, five patients (3.8%) came off protocol due to treatment-related toxicities. The median EFS of the whole study cohort was 1.5 months [95% CI (1.3-1.8)]. Overall, 81% of the events occurred within 6 months of trial enrolment. The estimated 6-month EFS of the entire cohort was 19.4% [95%CI: 12.7%-26%)] (Figure 1A), while the 6-month EFS of the patients aged ≤ 18 years at enrolment was 16.6% [95% CI: 8.7%-24.6%]. Supplemental Tables 1 and 2 summarize the 3-month EFS of the whole study cohort and 3-month and 6-month EFS of the patients aged ≤ 18 years stratified by specific NRSTS histologies. Sensitivity analysis by excluding the patients who came off the trials due to the patient or caregiver preference, unanticipated toxicity or where the primary reason for coming off study treatment was not available (N=19, 13.9%) yielded a 6-month EFS of 14.5% [95% CI: 8.2%-21%]. Sensitivity analysis of restricting the pooled analysis to only the 12 single-agent phase II trials demonstrated a 6-month EFS of 18.8% [95% CI:11.9%-25.8%].

Figure 1A.

Event-Free Survival of Relapsed/Refractory NRSTS Patients Enrolled in Thirteen Phase II Clinical Trials

No significant difference in EFS was observed by age at enrollment, sex, race, histology subtype, prior lines of therapies and trial initiation calendar year (p> 0.05) (Table 2 & Figure 1B). The EFS significantly differed among the trials (p=0.0019) and by the type of best radiographic response (PD vs. SD vs. CR: 11.5% vs. 47.8% vs. 33.3%, p= <0.0001). Black patients had non-statistically significantly higher EFS compared to others. Exploration of the key disease characteristics between Black patients and White patients did not reveal any significant differences in the distribution of age at enrollment, sex, NRSTS histology, and prior lines of therapies (Supplemental Table 3). A relative risk regression model, including age at enrolment, sex, race, NRSTS histology and prior lines of treatment, demonstrated that the risk of EFS was not statistically different from the reference group (Supplemental Table 4).

Table 2.

Event-Free Survival by Trial, Patient and Disease Characteristics

Variable	No. of Patients (%)	First Event			EFS at 6 months (95% Confidence Interval)	p-value
		None	Progression/ Relapse	Death
All Eligible Patients	137	9	109	19
Trial						0.0019
ADVL1522	15 (10.9%)	1	13	1	20% (0%, 40.2%)
ADVL1221	12 (8.8%)	2	9	1	25% (0.5%, 49.5%)
ADVL0821	11 (8%)	0	10	1	18.2% (0%, 41%)
ADVL0221	11 (8%)	1	8	2	9.1% (0%, 26.1%)
ADVL0524	10 (7.3%)	1	9	0	12.0% (0%, 34.1%)
ADVL0421a	11 (8.1%)	2	5	4	36.4% (7.9%, 64.8%)
ADVL0122	19 (13.9%)	0	16	3	26.3% (6.5%, 46.1%)
P9963	12 (8.8%)	0	12	0	0%
COG0962	11 (8%)	0	9	2	18.2% (0%, 41%)
Combined POG9761, A09705, A09713, ADVL0921 (trials enrolling < 10 patients)	25 (18.2%)	2	18	5	21.5% (4.8%, 38.3%)
Age at enrolment						0.73
≤ 18 years	87 (63.5%)	7	68	12	16.6% (8.7%, 24.6%)
>18 years	50 (36.5%)	2	41	7	24% (12.2%, 35.8%)
Sex						0.77
Female	51 (37.2%)	3	45	3	17.6% (7.2%, 28.1%)
Male	86 (62.8%)	6	63	16	20.4% (11.8%, 29.1%)
Race						0.20
Black	18 (13.1%)	2	12	4	47.6% (23.7%, 71.5%)
Other	9 (6.6%)	1	8	0	22.2% (0%, 49.4%)
Unknown	17 (12.4%)	1	15	1	11.8% (0%, 27.1%)
White	93 (67.9%)	5	74	14	15.3% (7.9%, 22.6%)
Histology type						0.98
Desmoplastic small round cell tumor	16 (11.7%)	0	13	3	18.7% (0%, 37.9%)
Malignant peripheral nerve sheath tumor	13 (9.5%)	2	10	1	17.1% (0%, 38.7%)
Synovial sarcoma	41 (29.9%)	1	37	3	21.9% (9.3%, 34.6%)
Other NRSTS Types	67 (48.9%)	6	49	12	18.3% (8.9%, 27.7%)
Prior lines of therapy						0.60
1 or 2 lines	37 (27%)	1	30	6	18.9% (6.3%, 31.5%)
1 or more lines	100 (73%)	8	79	13	19.5% (11.6%, 27.4%)
Calendar year of trial initiation						0.32
Initiated after 2006	67 (48.9%)	5	56	6	15.2% (6.5%, 23.9%)
Initiated before 2006	70 (51.1%)	4	53	13	23.3% (13.3%, 33.3%)
Type of best response						<.0001
Complete response	3 (2.2%)	0	3	0	33.3% (0%, 86.7%)
Stable disease	22 (16.1%)	5	15	2	47.8% (26.4%, 69.3%)
Progressive disease	96 (70.1%)	2	86	8	11.5% (5.1%, 17.8%)

^aOne patient in this trial had no information on progression or relapse but remained alive on the last follow up date

Figure 1B.

Event-Free Survival by NRSTS Histology Subtype

Overall Survival

The median OS of the entire study population was 7.5 months [95% CI (5.5-10.1)], with 1-year and 2-year OS of 33.8% (95% CI: 25.5% to 42.1%) and 16.1% (95% CI: 8.9% to 23.4%), respectively (Figure 2). When the analysis was restricted to the pediatric age group (≤ 18 years at the time of enrolment), the median OS was 7 months [95% CI (5.2-9.7)], with 1-year and 2-year OS of 33.2% (95% CI: 22.8% to 43.6%) and 16.4% (95% CI: 7.1% to 25.7%), respectively

Figure 2.

Overall Survival of Relapsed/Refractory NRSTS Patients Enrolled in the Thirteen Phase II Clinical Trials

DISCUSSION

Our analysis demonstrates the poor survival of children and young adults with relapsed/refractory NRSTS with measurable disease treated in the prospective non-randomized phase II clinical trials conducted in a 21-year span with an estimated 6-month EFS of 19.4%. The EFS among patients with advanced/metastatic STS studies has been considered a suitable surrogate for OS, highlighting the importance of specifying this benchmark for the next generation of phase II clinical trials selecting novel drugs for further testing.^27-29 Prior research also suggests that although associated with OS in advanced/metastatic STS patients, the radiographic response rate should not be used as an exclusive endpoint to determine the disease activity of novel agents in phase II trials, especially those investigating the role of cytostatic agents, as these agents can attain significant disease control without improving documented PR or CR.^27,30,31 This has been clearly demonstrated in the clinical trials analyzing the activity of pazopanib and regorafenib in advanced/relapsed STS, where these drugs exhibited disease activity despite having low radiographic response.^30,31 Consequently, future non-randomized phase II trials determining the potential active agents in relapsed or refractory pediatric NRSTS patients can employ the established benchmark of 6-month EFS of 19% and 95%CI (12.4%, 25.6%) and median EFS for recognizing novel agents, meriting additional testing.

Our results also reveal that the EFS of children and young adults with relapsed/refractory NRSTS is dismal, underscoring the urgent need for developing novel drugs for this subgroup of patients. Additionally, this study highlights the lack of phase II trials focusing primarily in this population. All included trials enrolled only a few patients with diverse NRSTS histologies, complicating the interpretation of the investigational agent’s activity for a specific NRSTS histology. The number of overlapping histologies and the older median age at diagnosis supports inclusion of pediatric patients on histology-specific adult trials when possible. Collaborative efforts in this regard are ongoing through organizations such as the National Clinical Trials Network. Ideally, these collaborations would start early in trial development rather than after trial activation. While an overall lumping strategy may not be ideal moving forward, there may be histology-agnostic opportunities (e.g. shared molecular or biological alternations) that may be best studied in a limited lumping trial, once again including both the pediatric and adult population.

Concerns could be raised about how the EFS for this heterogenous group of histological diagnoses will be utilized in the current era of biomarker-selected trials. Unfortunately, most pediatric and young adult NRSTS histologies do not have targetable genomic alterations/mutations. In a recent publication, of 18 patients with synovial sarcoma, none had any actionable alterations. Similarly, only 20% of patients with desmoplastic small round cell tumor and 15.4% with “undifferentiated sarcomas” had actionable alterations.³² Therefore, since conducting biomarker-selected trials may not be feasible for all NRSTS histologies, our EFS benchmarks are still relevant for this population.

There are several limitations to our analysis. One is the heterogeneity of the study populations. Since none of the trials were NRSTS or histology subtype-specific, we could only analyze the association of a few variables with EFS, including the estimated histology-specific 6-month EFS for just three common subtypes of pediatric NRSTS. More granular data on additional histotypes of NRSTS, previous lines of therapies, time to relapse, molecular information, and extent of the disease at enrollment were not included on available trial individual case report forms, preventing our ability to explore further the impact of these key variables on the survival. Furthermore, the disease assessment time points differed among included trials, with most trials evaluating radiological response at the end of one or two 21–28-day cycles. Since variation in disease evaluation can influence the EFS estimate, to overcome this limitation, we concentrated on the 6-month EFS, a time when 81% of the patients had an event of interest.

Although we did not find any significant difference in the EFS between the patients treated before and after 2006, this analysis cannot control for the influence of the change in patients' management, diagnostics or supportive care over time on the EFS. The apparent higher EFS of Black patients could be due to the small sample size of the study cohort. Additionally, although none of the investigational drugs in the included trials was deemed active based on the radiographic response rate endpoint, the estimated 6-month EFS varied significantly between the trials. Differences in the EFS between the included trials are intriguing and could have resulted from the heterogeneity of the study population, actual superiority of the investigational agent in terms of EFS despite not meeting the criteria for objective response rate or inclusion of death as the first event rather than relapse/progression. For instance, the patients with relapsed/refractory NRSTS treated on ADVL0421 trial testing oxaliplatin activity had a 6-month EFS of 36%, which is superior to other included trials.²⁰ Notably, for four patients in this trial, the event was death rather than progression/relapse, and for one patient, relapse/progression data were not available, and this patient was alive at the last follow-up date; both of these factors could have contributed to higher estimated EFS in this trial. Including the trials of drugs with potential activity in NRSTS and patients with death as an event could have resulted in overestimating the benchmark EFS; thus, the agents found to be active based on the identified benchmark EFS are more likely to exhibit activity in future trials. Since the trials did not capture the subsequent lines of therapies for the participants off-trial or post-progression period, the effect of these therapies on the estimated OS remains unclear. Finally, our findings do not apply to older adults with relapsed/refractory NRSTS tumors due to the differences in the NRSTS histologies and survival outcomes between the pediatric and older adult populations.¹

In conclusion, this study demonstrates the dismal outcome of children and young adults with relapsed/refractory NRSTS enrolled in 13 closed non-randomized phase II studies for refractory/recurrent solid tumors conducted by COG and its predecessor groups. The baseline EFS determined by this study for this population can be utilized as a benchmark for the design of future phase II trials testing novel agents in children and young adults with NRSTS tumors.

Data Sharing Statement

The Children’s Oncology Group Data Sharing policy describes the release and use of COG individual subject data for use in research projects in accordance with National Clinical Trials Network (NCTN) Program and NCI Community Oncology Research Program (NCORP) Guidelines. Only data expressly released from the oversight of the relevant COG Data and Safety Monitoring Committee (DSMC) are available to be shared. Data sharing will ordinarily be considered only after the primary study manuscript is accepted for publication. For phase 3 studies, individual-level de-identified datasets that would be sufficient to reproduce results provided in a publication containing the primary study analysis can be requested from the NCTN/NCORP Data Archive at https://nctn-data-archive.nci.nih.gov/. Data are available to researchers who wish to analyze the data in secondary studies to enhance the public health benefit of the original work and agree to the terms and conditions of use. For non-phase 3 studies, data are available following the primary publication. An individual-level de-identified dataset containing the variables analyzed in the primary results paper can be expected to be available upon request. Requests for access to COG protocol research data should be sent to: datarequest@childrensoncologygroup.org. Data are available to researchers whose proposed analysis is found by COG to be feasible and of scientific merit and who agree to the terms and conditions of use. For all requests, no other study documents, including the protocol, will be made available and no end date exists for requests. In addition to above, release of data collected in a clinical trial conducted under a binding collaborative agreement between COG or the NCI Cancer Therapy Evaluation Program (CTEP) and a pharmaceutical/biotechnology company must comply with the data sharing terms of the binding collaborative/contractual agreement and must receive the proper approvals.