Cisplatin-Based Chemotherapy and/or Surgery in Treating Young Patients With Adrenocortical Tumor: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Cisplatin-Based Chemotherapy and/or Surgery in Treating Young Patients With Adrenocortical Tumor (NCT00304070), a non-randomized phase 3 study. The primary outcome is Five Year Event-free Survival (EFS).

2. Clinical Question

The study addresses how outcomes under different management pathways compare with prespecified event-free survival targets. The registry describes Stratum I as primary tumor resection and retroperitoneal lymph node sampling followed by observation, with observation alone for some patients who had prior surgery without nodal sampling. Stratum II involved primary tumor resection and extended regional lymph node dissection followed by observation, or exploratory surgery and extended dissection after prior simple resection. Stratum III involved combination chemotherapy and mitotane, with surgery after chemotherapy for some patients and possible additional surgery after chemotherapy ended.

Because treatment pathways differed by stratum and allocation was not randomized, the clinical question is not whether one assigned treatment outperformed another in a controlled comparison. The registry instead describes a one-sample comparison of observed event-free survival with hypothesized models. It does not report an actual five-year event-free survival estimate, confidence interval, or test result. The definition of an event for this endpoint is not reported in the registry record, so the specific events counted in event-free survival cannot be identified from the record.

3. Trial Design

This was a completed phase III study sponsored by Children's Oncology Group. The registry lists enrollment of 78 and a parallel, non-randomized design with no masking. The recorded start date is 2007-05-03, and the primary completion date is 2015-12-31. Eligible disease stages listed in the record range from stage I through stage IV adrenocortical carcinoma. A more specific age range is not reported in the registry record.

The treatment pathways were organized into strata rather than randomized arms. Stratum I involved primary tumor resection and retroperitoneal lymph node sampling followed by observation; patients with prior surgery without nodal sampling underwent observation only. Stratum II involved primary tumor resection with extended regional lymph node dissection and observation, or exploratory surgery with extended dissection after prior simple resection. Stratum III involved combination chemotherapy with filgrastim for up to 30 weeks (10 courses), followed by mitotane alone for an additional 2 months. Some patients underwent surgery after chemotherapy course 2 or 4, and some had additional surgery after completing chemotherapy.

The listed interventions include cisplatin, doxorubicin hydrochloride, etoposide, mitotane, filgrastim, pegfilgrastim and conventional surgery. Their listing does not establish that every participant received every intervention. Treatment assignment procedures beyond the non-randomized classification, reasons for treatment selection, and detailed eligibility criteria are not reported in the registry record. The registry describes three analyses and seven outcome measures, but the actual endpoint estimates and analysis results are not reported there. The design therefore supports description of distinct management pathways and comparison with specified survival targets, not a randomized head-to-head treatment effect.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Five Year Event-free Survival (EFS), measured over Up to five years after enrollment.

The primary endpoint is five-year event-free survival (EFS), described as an estimated probability over a period of up to five years after enrollment. In general, event-free survival is a time-to-event measure: it tracks how long participants remain free of events defined by the study. The specific events that counted for this endpoint are not reported in the registry record, so the measure's precise clinical definition cannot be reconstructed.

Survival endpoints are useful when both the occurrence and timing of an event matter. Unlike a simple count at one time point, a time-to-event analysis can use information about when events occur and account for follow-up that ends before an event is observed. Those general advantages depend on appropriate event definitions, follow-up information and analysis assumptions. The registry record does not give the event definition, censoring rules, or details about handling missing follow-up.

The planned comparison was against hypothesized exponential models with different failure rates by stratum. The record gives a rate of 0.053 per year for Stratum I and 0.347 per year for Stratum II, and a rate of 0.602 per year for Stratum III and IV, during the first two years; it specifies a rate of 0 after that. These are model assumptions, not observed event rates. The record does not report the observed EFS probability or its 95% confidence interval, so it does not establish how closely observed survival matched the target models.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes a one-sample, one-sided log-rank test comparing observed event-time data with a hypothesized model. In general, a one-sample test compares one set of observations with a reference value or distribution rather than comparing two randomized groups. Here, the reference is an exponential model with stratum-specific failure rates. The record states that the test has size 0.05. This is the planned significance level, not a reported p-value or evidence that the data did or did not meet the target.

What does the exponential model assume?

In general, an exponential time-to-event model represents a constant event rate over a specified period. The registry's target model uses a constant failure rate during the first two years and sets the rate to 0 after that. It gives a rate of 0.053 per year for Stratum I, 0.347 per year for Stratum II, and 0.602 per year for Stratum III and IV during the first two years. Those values define the reference models; they are not study results. A model that sets the later rate to 0 makes a strong assumption about the event process after the specified period. Whether the observed data support that assumption cannot be assessed without the reported results and relevant follow-up details.

What does the log-rank test assess?

Generally, a log-rank test uses event timing and the pattern of participants remaining at risk over follow-up to assess differences between observed and expected event experience. In the one-sample form described in the registry, the expectation comes from the hypothesized model. The test's result is a summary of how inconsistent the observed event pattern is with that model under the test assumptions. It does not quantify the size of a treatment effect, and it does not prove that the reference model is true when a test is not statistically significant. The registry record does not state the observed test statistic or p-value.

Why does the record mention independent increments?

In general, independent increments means that information accruing over separate portions of follow-up can be treated as incrementally independent under the model. The record says this property will be used with the Lan and DeMets method to derive p-values for the testing procedure. This indicates a planned approach to sequential testing boundaries or p-value calculation. It does not establish how many interim looks occurred, whether the analysis plan was implemented as described, or what the resulting p-values were; those details are not reported in the registry record.

How should a confidence interval be read?

A confidence interval expresses uncertainty around an estimated quantity. In general, a narrower interval indicates greater precision, while a wider interval indicates more uncertainty. The primary endpoint names a 95% confidence interval as its dispersion measure, but the registry record does not give the actual interval or explain its construction. Therefore, no statement about the precision of the observed five-year EFS estimate can be made from the record.

What do planning illustrations mean?

Planning illustrations are not observed power and do not describe the trial's actual enrollment rationale. The record explicitly characterizes its sample-size assumptions as an illustration computed from the observed effect, not the trial's own design. It specifies a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and effects expressed as stated fractions of the observed effect. For time-to-event outcomes, the illustrations give total events; for other outcomes, they give participants per arm. These are planning conventions, not trial outcomes, and they do not change the fact that the study itself is non-randomized and organized by strata. The specific illustration values are not reported in the registry record.

Statistical interpretation also depends on the underlying data and analysis choices. In general, time-to-event analyses require defined events, follow-up and censoring rules, and a clear analysis population. The registry's description does not report the EFS event definition, the censoring method, handling of missing observations, or which participants were included in each analysis. Without these details and the numerical results, the planned method can be explained, but its execution and the resulting evidence cannot be independently evaluated from the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry describes one primary endpoint and a one-sided one-sample log-rank test of size 0.05. It also states that the Lan and DeMets method will be used to derive p-values because the test has independent increments. That is the multiplicity-related procedure described for the planned testing approach. The record does not report how testing across the primary and secondary outcomes was handled, whether any adjustments were made across outcomes or strata, or whether the planned procedure was implemented as described. Those details are not reported in the registry record.

Missing-data handling is also not reported in the registry record. For event-free survival, interpretation generally depends on when follow-up ends for each participant and how participants without an observed event are treated analytically. The record does not specify censoring rules, loss-to-follow-up procedures, or how incomplete outcome information was handled. It also does not describe an imputation method or sensitivity analyses for missing information.

The analysis populations are not reported in the registry record. It is therefore unclear from this record which participants were included in each endpoint analysis, how prior surgery or eligibility issues were handled, and whether analyses were based on enrollment, treatment received, or another population definition. The registry lists enrollment of 78, but does not provide endpoint denominators or participant flow through the analyses. These omissions limit assessment of the amount of missing information and the applicability of the endpoint results, which themselves are not reported.

11. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: pneumonitis (2.6%).

The serious adverse-event entry is limited to Stratum 3. It lists 39 at risk and 1 affected in the serious-event fields. The other-event fields list 39 at risk and 38 affected. Pneumonitis is listed as a leading serious term, with 2.6% recorded. These are the registry's reported figures; the record does not explain the relationship between the term-level percentage and the group-level entries, so they should not be combined or treated as interchangeable measures.

In general, an at-risk count describes the group considered for a safety measure, while an affected count describes the recorded occurrence for that measure. Interpretation depends on the registry's event definitions, reporting window, and whether counts represent people or events. Those details are not fully described in the record. The deaths count is not reported in the registry record.

These safety figures do not establish causation. They describe events recorded in a study involving different treatment pathways, without a randomized control group identified for a like-for-like comparison. The listed figures should therefore be read as descriptive registry entries, not as estimates of comparative risk or proof that an intervention caused or prevented an adverse event.

12. What the Result Does, and Does Not, Mean

The record establishes that the study was a completed, non-randomized phase III trial with a stated primary endpoint of five-year event-free survival and a specified model-based testing plan. It describes three management strata and gives the hypothesized failure rates that define the reference models. Those details make the intended statistical question understandable: whether observed event-time data were consistent with stratum-specific targets.

The record does not report the observed EFS estimates, confidence intervals or p-values. It therefore cannot establish whether the data were consistent with the target models, how precisely EFS was estimated, or how much observed survival differed from a target. No conclusion about the magnitude or direction of endpoint results can be drawn from the information recorded.

The study's non-randomized structure also sets a limit on causal interpretation. The strata describe different treatment pathways, rather than a randomized contrast between equivalent groups. Even if outcome estimates for each stratum were available, differences between strata would not automatically be attributable to their treatment strategies. Patient selection, disease characteristics and other factors could contribute; the information needed to evaluate such factors is not reported in the registry record.

In general, a planned statistical test and a completed study are not substitutes for reported results. A test addresses compatibility with a specified model under its assumptions; it does not show that a treatment works or does not work, and it does not establish causation. The endpoint's event definition, analysis populations, missing-data methods and actual numerical findings are not reported. The sound interpretation is therefore limited to the design and planned analysis, along with the narrow safety entries documented for Stratum 3.

13. Why This Trial Matters Statistically

This study raises a useful statistical distinction between comparing treatments and comparing observed outcomes with a reference model. Its primary analysis was described as a one-sample log-rank test against hypothesized event-time patterns, rather than as a randomized comparison between treatment groups. That distinction matters because the two questions have different interpretations: a model-based test evaluates consistency with a target, while a randomized comparison is designed to estimate a treatment contrast under random allocation.

The stratified management pathways also illustrate why clinical context matters in analysis. Surgery and observation pathways differed from the chemotherapy and surgery pathway, and the registry does not describe the strata as like-for-like control groups. A reader should not interpret stratum differences as treatment effects without a design and analysis that address the reasons participants entered different pathways.

The registry specifies assumptions that shape the primary test, including stratum-specific failure rates during the first two years and a rate of 0 after that. Making assumptions explicit helps readers understand what the planned test treats as its benchmark. It also highlights why observed estimates, confidence intervals and test results are needed: without them, the statistical question is visible, but the evidence bearing on it is not.

Finally, the record shows why trial completion and enrollment do not alone communicate statistical findings. The actual endpoint results, analysis populations and missing-data methods are not reported. Clear reporting of those elements would allow readers to assess precision, assumptions and the limits of inference, rather than relying only on a planned analysis description.

14. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The most direct limitation is that the registry record does not report the primary EFS estimate, its 95% confidence interval, or the p-value. As a result, the main statistical question cannot be answered from the record. The event definition for EFS is also not reported, leaving uncertainty about which events contributed to the endpoint.

Non-randomized allocation limits comparison among the strata. The management pathways differ, and the record does not present them as a like-for-like control comparison. It does not report how participants were assigned to strata, what baseline characteristics differed, or whether the analysis adjusted for potential confounding. Consequently, the design does not support a simple causal reading of any differences that might be observed.

The primary model relies on stratum-specific failure rates during the first two years and assumes a rate of 0 after that. The record does not provide information to assess the fit of those assumptions or sensitivity to alternative models. Details about censoring, follow-up completeness, analysis populations, and missing-data handling are not reported in the registry record, limiting assessment of the survival analysis.

The safety information is narrow: the reported group is Stratum 3, and the record does not report a deaths count. The relationship between the group-level serious and other-event entries and the term-level Pneumonitis percentage is not explained. The record also does not provide enough detail to interpret the safety entries as comparative or causal findings. These limitations apply to interpretation of the registry record; they do not establish what may appear in other sources.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Cisplatin-Based Chemotherapy and/or Surgery in Treating Young Patients With Adrenocortical Tumor — NCT00304070.
• Registry record last updated 2024-02-28; status last verified 2024-02.
• Results first posted 2017-06-12.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

18. Record Summary

The registry describes NCT00304070, a completed phase III study sponsored by Children's Oncology Group, with enrollment of 78 and a non-randomized, unmasked, parallel design. It studied surgery, observation and cisplatin-based combination chemotherapy pathways for young patients with stage I through stage IV adrenocortical cancer. The recorded start date was 2007-05-03, and the primary completion date was 2015-12-31.

The primary endpoint was five-year event-free survival, assessed against hypothesized exponential models using a one-sample, one-sided log-rank test. The record describes a Lan and DeMets p-value approach but does not report the actual primary endpoint estimates, confidence interval or test result. It also does not report the EFS event definition, analysis populations or missing-data methods. The available safety entry concerns Stratum 3 and includes serious and other adverse-event counts, with Pneumonitis listed at 2.6% among the leading serious terms.