Combination Chemotherapy, PEG-Interferon Alfa-2b, and Surgery in Treating Patients With Osteosarcoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Combination Chemotherapy, PEG-Interferon Alfa-2b, and Surgery in Treating Patients With Osteosarcoma (NCT00134030), a randomized phase 3 study. The primary outcome is Event-free Survival (EFS).

2. Clinical Question

3. Trial Design

This was a randomized, phase III, parallel-group trial with no masking reported. The registry names Children's Oncology Group as sponsor and records enrollment of 1334. The recorded start date is 2005-11-14, the primary completion date is 2015-01-31, and the trial status is completed. The official title describes a randomized European and American Osteosarcoma Study Group trial of treatment strategies for resectable osteosarcoma based on histological response to pre-operative chemotherapy.

The registry lists localized and metastatic osteosarcoma. It describes chemotherapy before surgery followed by maintenance strategies that vary by regimen and include PEG-interferon alfa-2b in one described arm. Listed drugs include cisplatin, doxorubicin, etoposide, ifosfamide, and methotrexate, alongside peginterferon alfa-2b. Quality-of-life assessment and questionnaire administration are also listed as interventions. The detailed description of the maintenance therapy group 2 arm II ends mid-regimen, so its full schedule is not reported in the registry record.

The primary outcome is event-free survival. The registry defines the event as the first of death, detection of local recurrence or metastasis, progression of metastatic disease, or detection of a secondary malignancy. It says follow-up per participant will be assessed for up to 10 years and that 3-year event-free survival is provided as a summary. Overall survival and toxicity are listed as secondary outcomes. The registry record does not report the randomization procedure, allocation concealment, eligibility criteria in detail, or the planned analysis populations.

Arm-level primary outcome summaries are posted for MAP-GR, MAPifn, MAP-PR, and MAPIE. The reported primary comparisons pair MAP-GR with MAPifn, and MAP-PR with MAPIE. The registry also reports an RMST comparison for the latter pair. These pairwise results should be read as specific comparisons rather than as evidence that every possible contrast among the named arms was tested or established.

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

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 Event-free Survival (EFS), measured over From date of randomization to date of the event..

Event-free survival (EFS) measures the time from randomization until the first qualifying event. In this trial, those events are death, detection of local recurrence or metastasis, progression of metastatic disease, or detection of a secondary malignancy. A participant who has not had a recorded event by the end of their observed follow-up contributes event-free time until the applicable end of observation; the registry does not describe the censoring rules in further detail.

EFS is used in trials where outcomes other than death are clinically relevant parts of disease progression or recurrence. In general, combining these events can capture disease control earlier than an endpoint based only on death. It also means that the result depends on the trial’s event definition and on how consistently events are detected and recorded. A treatment comparison in EFS therefore addresses the timing of the first defined event, not each component separately. Component-specific EFS results are not reported in the registry record.

The registry presents 3-year EFS as a summary and describes the intended analysis as a log-rank test with hazard ratios and confidence intervals. The posted summaries are MAP-GR 74% (95% CI 69% to 79%), MAPifn 77% (95% CI 72% to 81%), MAP-PR 55% (95% CI 49% to 60%), and MAPIE 53% (95% CI 47% to 59%). These percentages describe event-free survival at the stated time point; they do not show the timing or type of events for individual participants. The record does not report the number or types of EFS events in each group.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Arm-level estimates. The posted EFS percentages describe the estimated proportion of participants who remained event-free at the registry’s stated 3-year summary time. Each estimate is accompanied by a 95% confidence interval. In general, a confidence interval conveys sampling uncertainty around an estimate: narrower intervals indicate greater precision, while wider intervals indicate less precision. It is not a guarantee that the true value lies within the particular interval, and overlap between two separate arm intervals is not itself a formal test of their difference.

Log-rank tests and hazard ratios. The registry describes EFS comparisons using the log-rank test and hazard ratios. In general, a log-rank test compares the observed timing of events between groups across follow-up, rather than comparing only the 3-year percentages. A hazard ratio expresses a relative contrast in event rates over time under the model’s interpretation. A value of 1 represents equal hazards; values on either side indicate different relative event rates, with direction dependent on which group is in the numerator. The registry lists the comparison groups, but does not explain the hazard-ratio coding direction. For that reason, the estimates are best read together with their confidence intervals and p-values rather than assigning a directional benefit from the point estimate alone.

For MAP-GR versus MAPifn, the reported hazard ratio is 0.83 (95% CI 0.61 to 1.12; p=0.214). For MAP-PR versus MAPIE, it is 0.98 (95% CI 0.78 to 1.23; p=0.86). Both 95% intervals include 1, and neither reported p-value is below 0.05. In general, a p-value measures how unusual data at least as incompatible with a specified null hypothesis would be if that hypothesis and the analysis assumptions were true. It is not the probability that the null hypothesis is true, nor a measure of clinical importance. The registry’s significance threshold and any adjustment for multiple comparisons are not reported.

Restricted mean survival time. The RMST analysis for MAP-PR versus MAPIE reports a mean difference of 0.8 (95% CI -3.3 to 4.9; p=0.69). In general, RMST is the average time spent event-free up to a chosen time horizon. Unlike a hazard ratio, it expresses a difference in time and can provide an absolute summary over that horizon. A confidence interval spanning zero is compatible with no difference, as well as with differences in either direction. The analysis horizon and units for this RMST result are not reported in the registry record, so its magnitude cannot be given a more specific time-based interpretation.

Overall-survival summaries and recalculations. Overall survival is listed as a secondary outcome, with arm-level percentages and 95% confidence intervals. The registry also reports log-rank hazard-ratio analyses for MAP-GR versus MAPifn and MAP-PR versus MAPIE. Separately, simple unadjusted recalculations compare posted binary outcome counts for MAPifn and MAP-PR with MAP-GR. These recalculations use risk differences and their confidence intervals; they are not the same estimand or analysis as a log-rank comparison of event times. They should not be treated as replacements for the sponsor’s time-to-event analyses. No discrepancy is flagged in the registry record.

Planning illustrations. The sample-size calculations are planning illustrations based on the observed effect, not observed power and not the trial’s design justification. They assume a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The listed scenarios plan for 100%, 75%, or 50% of the observed effect; the time-to-event calculations give total events, while the other calculations give participants per arm. In general, planning for a smaller effect requires more information to detect it at a chosen power, all else equal. These hypothetical calculations do not show that the completed trial had a particular power, and the trial’s own sample-size assumptions or justification are not reported in the registry record.

Interpretive limits. Statistical significance, precision, and clinical importance are different considerations. A non-significant result can reflect a small effect, limited precision, or both; it does not establish no effect. Conversely, a statistically notable result does not alone establish clinical importance. The registry does not report a clinically important difference threshold, a multiplicity strategy, or details of model diagnostics, so conclusions should remain tied to the stated estimates and uncertainty intervals.

10. Analyses Reported by the Sponsor

The sponsor-reported primary analyses are log-rank comparisons of event times, expressed as hazard ratios. The simple recalculations for overall survival use posted counts to estimate unadjusted risk differences for particular comparisons. These methods answer related but different questions: a binary risk difference compares the proportions with the outcome at the summarized time, while a log-rank analysis uses the event timing over follow-up. They also do not cover exactly the same pairings: the recalculations include MAP-PR versus MAP-GR, whereas the sponsor-reported secondary log-rank comparison pairs MAP-PR with MAPIE.

For the recalculated MAPifn versus MAP-GR overall-survival comparison, the risk difference is -0.001 (95% CI -0.054 to 0.053; p=0.974). For MAP-PR versus MAP-GR, it is -0.161 (95% CI -0.225 to -0.096; p < 0.001). These are unadjusted recalculations from the posted summary counts, not the sponsor’s time-to-event analyses. The sponsor-reported overall-survival hazard ratios are 0.96 (95% CI 0.69 to 1.33; p=0.804) for MAP-GR versus MAPifn and 1.06 (95% CI 0.81 to 1.39; p=0.674) for MAP-PR versus MAPIE. The differing comparisons and estimands prevent a direct numerical reconciliation.

The registry record does not flag a discrepancy. Accordingly, the recalculations are supplementary descriptive analyses and do not displace the sponsor’s own analyses. The registry does not report enough detail about the sponsor’s analysis populations, adjustment methods, or handling of censoring to reconstruct those analyses independently.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary endpoint and secondary outcomes, but a complete statistical analysis plan is not reported in the registry record. In particular, the record does not say whether or how the primary comparisons, the additional RMST analysis, the secondary outcomes, or any subgroup analyses were accounted for in a multiplicity strategy. Multiplicity matters because repeated testing can increase the chance of at least one nominally small p-value when all null hypotheses are true. Without a stated plan, the reported p-values should be interpreted as results for their individual tests, not as a fully controlled family of comparisons.

Missing-data methods are also not reported in the registry record. The record does not describe how incomplete follow-up, withdrawal, missing endpoint assessments, or incomplete covariate information were handled. For time-to-event outcomes, censoring rules and the amount of follow-up available can affect estimates and their interpretation. The registry states that participant follow-up for EFS will be assessed for up to 10 years, but does not give follow-up completeness or explain censoring in the posted results. These omissions make it difficult to assess how much information supports each estimate.

The analysis populations are not reported in the registry record. The randomized trial description does not establish whether the outcome analyses used all randomized participants, only participants who began treatment, or another defined population. The record also does not report protocol deviations or exclusions from the posted analyses. This information is important because excluding participants after randomization can affect the balance created by random assignment.

Further details not reported in the registry record include the randomization method, allocation concealment, stratification factors, prespecified subgroup analyses, sample-size justification, and any interim analyses. Their absence does not demonstrate that these procedures were not used; it means they cannot be evaluated from the registry record. The planning illustrations based on observed effects should not be confused with the trial’s own design rationale.

12. 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 5% reporting threshold.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: febrile neutropenia (3.2%), mucosal inflammation (1.9%), device related infection (1.7%), sepsis (1.3%), pyrexia (1.1%).

The posted serious adverse event summaries give the number affected and the number at risk in each arm: MAP-GR, 49 of 359; MAPifn, 76 of 357; MAP-PR, 36 of 310; and MAPIE, 43 of 308. These are counts, not adjusted comparisons. The registry record does not report a comparative statistical analysis, the observation window for these counts, or a detailed definition of the serious-event collection period. The death fields in these summaries are not populated, so deaths within the adverse-event reporting summary are not reported in the registry record.

The listed serious-event terms include febrile neutropenia (3.2%), mucosal inflammation (1.9%), device related infection (1.7%), sepsis (1.3%), pyrexia (1.1%), and wound infection (0.8%). The registry does not specify the group attribution or denominator for these term percentages in the posted summary. They should not be assigned to an individual arm or interpreted as rates over a defined exposure period without that information. No other adverse events are recorded in the listed “other” category for the arms, but this does not establish that no non-serious events occurred; the registry record does not explain the scope of that category. These summaries describe reported events and do not by themselves establish comparative safety.

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

The primary results show the estimated 3-year event-free survival in each named arm and the sponsor’s pairwise time-to-event analyses. Within MAP-GR versus MAPifn and MAP-PR versus MAPIE, the reported hazard-ratio confidence intervals include 1 and the p-values do not provide clear evidence of a difference. The RMST comparison for MAP-PR versus MAPIE is also imprecise and its confidence interval includes zero. These findings mean that the posted analyses do not resolve whether there is a difference within either pair; they do not prove that the strategies are equivalent.

The results do not establish a treatment effect for every possible comparison among all named arms. They also do not establish that differences between the two sets of arm summaries are caused by the treatment strategies, because the reported primary comparisons are pairwise and the registry does not report a direct analysis across those sets. The detailed relationship between histological response and the posted arm-level analyses is not reported in the registry record.

Overall-survival summaries and analyses provide additional context, but their interpretation is limited by the secondary status, the differing comparison structures, and the absence of a reported multiplicity plan. The simple recalculations are unadjusted and do not replace the sponsor’s time-to-event analyses. Safety counts describe recorded serious events but do not give a complete comparative safety assessment.

In general, trial results estimate average outcomes for the studied participants under the trial conditions. They do not guarantee an individual outcome or determine what treatment is appropriate for a particular person. The registry does not report detailed eligibility criteria, participant characteristics, analysis populations, missing-data methods, or the full statistical plan. Those omissions restrict conclusions about how broadly the results apply and how robustly the reported estimates can be evaluated. No clinical recommendation follows from these registry results alone.

14. Why This Trial Matters Statistically

This trial is statistically informative because it reports both arm-level EFS estimates and time-to-event comparisons, rather than relying on a single percentage alone. The distinction matters: the 3-year estimates summarize status at a specified time, while log-rank tests and hazard ratios use event timing across follow-up. For MAP-PR and MAPIE, an RMST comparison supplies another summary that expresses a difference in average event-free time over a horizon, although the registry does not report that horizon.

The results also illustrate why uncertainty intervals matter. Point estimates can differ, but intervals that include the no-difference value indicate that the data remain compatible with a range of possible effects. A non-significant test does not demonstrate identical outcomes, and a point estimate should not be read as definitive evidence of benefit or harm when uncertainty is substantial. The posted EFS results do not establish equivalence within either reported pair.

The registry also makes visible a common interpretive challenge: arm-level percentages, time-to-event analyses, and unadjusted binary recalculations are not interchangeable. They use different outcome summaries and, in the overall-survival results, not all comparisons match. Keeping those distinctions explicit prevents a descriptive contrast from being mistaken for the trial’s prespecified or sponsor-reported inference.

Finally, the record shows why reporting of the analysis plan, missing data, populations, and multiplicity matters. These details determine how fully the estimates can be assessed and how much weight to place on secondary findings. The posted findings are useful for understanding the comparisons that were reported, while leaving important questions unresolved about design and analysis implementation.

15. 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.

Several limitations are specific to the information available for this trial. The study is described as open label, with no masking. The registry record does not explain whether outcome assessment was blinded or how event ascertainment was standardized. This is relevant because recurrence, progression, and secondary malignancy are components of EFS whose detection depends on assessment and recording.

The population information is limited to localized and metastatic osteosarcoma, while the official title refers to resectable disease and strategies based on histological response. The registry does not report participant characteristics, the distribution of disease categories, histological response definitions, or subgroup-specific results. It is therefore not possible to determine from the posted record how the reported pairwise estimates vary across those clinical groups.

The detailed regimen description is incomplete: the maintenance therapy group 2 arm II description ends mid-sentence. The record lists treatment components and schedules for other descriptions, but does not give a complete account of all regimen details in that arm. The relationship between the abbreviated arm names and all the treatment schedules is also not fully explained. This limits the ability to interpret the contrasts as exact treatment-component effects.

For EFS, the record posts 3-year summaries and says follow-up per participant will be assessed for up to 10 years, but does not report event counts, follow-up completeness, censoring details, or component-specific outcomes. It does not report the RMST horizon or its units. The analysis populations, missing-data handling, multiplicity plan, randomization details, and protocol deviations are not reported in the registry record. These gaps limit independent assessment of precision, risk of bias, and generalizability; they do not establish that the unreported procedures were absent.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Combination Chemotherapy, PEG-Interferon Alfa-2b, and Surgery in Treating Patients With Osteosarcoma — NCT00134030.
• Registry record last updated 2023-06-07; status last verified 2023-05.
• Results first posted 2023-06-07.
• 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.

19. Record Summary

The registry record describes a completed randomized phase III parallel-group trial sponsored by Children's Oncology Group, with enrollment of 1334 and no masking. The trial studied chemotherapy and surgery strategies for osteosarcoma, including regimens with and without PEG-interferon alfa-2b. The primary outcome was EFS, defined by the first occurrence of death, local recurrence or metastasis, progression of metastatic disease, or secondary malignancy; the posted summary is 3-year EFS.

For the reported MAP-GR/MAPifn and MAP-PR/MAPIE comparisons, the sponsor’s log-rank analyses do not establish a clear EFS difference. An RMST analysis for the latter pair is also compatible with no difference. Overall survival is posted as a secondary outcome, while detailed toxicity results and several analysis-plan details are not reported in the registry record. The results should be interpreted as pair-specific estimates with uncertainty, not as proof of equivalence or a comparison across every arm.