This page separates the information contained in the ClinicalTrials.gov record from statistical explanation. The registry reports that results have been posted, but the ClinicalTrials.gov record contains no formal statistical analyses.
1. Trial at a Glance
GOG-0213 is a randomized, parallel-design, phase 3 oncology trial with 1052 participants enrolled. The registry identifies two primary time-to-event endpoints: overall survival in patients with recurrent platinum-sensitive epithelial ovarian, peritoneal primary, or fallopian tube cancer, and overall survival evaluating the addition of bevacizumab to second-line paclitaxel and carboplatin.
| Feature | GOG-0213 |
|---|---|
| Trial name | GOG-0213 |
| NCT identifier | NCT00565851 |
| Phase | Phase 3 |
| Status | ACTIVE_NOT_RECRUITING |
| Start | 2007-12-06 |
| Primary completion | 2019-04-30 |
| Lead sponsor | National Cancer Institute (NCI) |
| Sponsor type | NIH |
| Allocation | RANDOMIZED |
| Design model | PARALLEL |
| Masking | NONE |
| Primary purpose | TREATMENT |
| Enrollment | 1052 |
| Number of arms | 4 |
| Registered primary endpoints | 2 |
| Posted outcome measures | 9 |
| Results posted | Yes |
2. Clinical Question
The GOG-0213 registry describes two related primary questions involving overall survival in recurrent platinum-sensitive epithelial ovarian, peritoneal primary, or fallopian tube cancer. One asks whether surgical secondary cytoreduction in addition to adjuvant chemotherapy increases the duration of overall survival. The other asks whether adding bevacizumab increases the duration of overall survival relative to second-line paclitaxel and carboplatin alone.
Population
Patients with recurrent platinum-sensitive epithelial ovarian cancer, peritoneal primary cancer, or fallopian tube cancer. the ClinicalTrials.gov record also lists clear cell, endometrioid, mucinous, serous, transitional-cell, and undifferentiated fallopian tube cancer among the conditions.
Intervention question 1
Whether surgical secondary cytoreduction, in addition to adjuvant chemotherapy, increases the duration of overall survival.
Intervention question 2
Whether addition of bevacizumab increases the duration of overall survival relative to second-line paclitaxel and carboplatin alone.
Primary outcome
Overall survival, analyzed as a time-to-event endpoint. The registry gives a median follow-up time frame of 82.5 months for both primary endpoints.
3. Trial Design
The registry lists four study arms and identifies carboplatin, paclitaxel, gemcitabine hydrochloride, bevacizumab, laboratory biomarker analysis, quality-of-life assessment, and docetaxel among the interventions. The registry-reported extraction does not provide a complete arm-by-arm treatment description sufficient to reconstruct every regimen or assignment, so this page does not infer additional arm structure.
4. Randomization and Stratification
Randomization is explicitly identified in the ClinicalTrials.gov record. For the primary overall-survival endpoints, the registry definition states that treatment regimens will be compared using a logrank procedure including patients according to their randomly assigned treatment.
The registry definition further states that the logrank test will be stratified by secondary surgical debulking status, distinguishing patients randomized to undergo cytoreduction, patients randomized not to undergo secondary cytoreduction, and patients who were not candidates for or did not consent to secondary surgical cytoreduction.
| Design element | What the ClinicalTrials.gov record supports |
|---|---|
| Randomization | Yes; allocation is listed as RANDOMIZED. |
| Design | Parallel. |
| Masking | None. |
| Primary analysis comparison | Randomly assigned treatment regimens. |
| Primary time-to-event test | Logrank procedure. |
| Stratification explicitly identified | Secondary surgical debulking status. |
5. Primary Endpoints
| Endpoint | Registry definition | Time frame | Endpoint type |
|---|---|---|---|
| Overall survival — surgical secondary cytoreduction | To determine if surgical secondary cytoreduction in addition to adjuvant chemotherapy increases the duration of overall survival in patients with recurrent platinum sensitive epithelial ovarian cancer, peritoneal primary or fallopian tube cancer. | The time frame is 82.5 months (median duration of follow-up). | Time-to-event |
| Overall survival — bevacizumab | To determine if the addition of bevacizumab increases the duration of overall survival relative to second-line paclitaxel and carboplatin alone in patients with recurrent platinum sensitive epithelial ovarian, peritoneal primary or fallopian tube cancer. | The time frame is 82.5 months (median duration of follow-up). | Time-to-event |
How the registry defines the primary comparison
For both primary endpoints, the registry-reported definition says that the treatment regimens will be compared with a logrank procedure. The procedure includes all patients categorized according to their randomly assigned treatment and is stratified by secondary surgical debulking status.
This structure is appropriate for an endpoint defined by time from an origin point to death, because patients can have different follow-up durations and may contribute information without having experienced the event by the end of observation.
6. Results and Reporting Status
The ClinicalTrials.gov record states that results have been posted and lists 9 posted outcome measures. However, the ClinicalTrials.gov record contains no formal statistical analyses. Consequently, the source material in the ClinicalTrials.gov record does not contain formal statistical-analysis entries for the primary endpoints.
What would normally be reported for overall survival
For a time-to-event endpoint such as overall survival, a complete statistical-results presentation would ordinarily include a survival-function estimate such as a Kaplan-Meier curve, numbers at risk, event counts, a comparison between randomized groups, and a treatment-effect measure such as a hazard ratio with a confidence interval. A p-value may accompany the prespecified comparison. None of those numerical result fields is in the posted results of the ClinicalTrials.gov record.
| Potential result element | registry GOG-0213 data |
|---|---|
| Formal statistical-analysis entries | None reported. |
| Primary endpoint hazard ratios | Not reported. |
| Primary endpoint confidence intervals | Not reported. |
| Primary endpoint p-values | Not reported. |
| Primary endpoint event counts | Not reported. |
| Median overall survival estimates | Not reported. |
| Time-specific survival estimates | Not reported. |
7. Planned Analysis
Because the registry indicates that results are posted but the registry-reported statistical-analysis section contains no formal analysis entries, the appropriate interpretation is to describe the registered analysis framework without fabricating numerical results.
Primary analysis of overall survival
The registry states that the treatment regimens will be compared using a logrank procedure. The analysis includes patients according to their randomly assigned treatment and stratifies the logrank test by secondary surgical debulking status.
The log-rank procedure compares the observed and expected numbers of events over the follow-up period. Stratification allows the comparison to account for the prespecified surgical-debulking classification rather than treating those categories as if they were ignored.
The registry definition is sufficient to identify the primary comparison and its stratified logrank framework, but it is not sufficient to reproduce the complete statistical analysis or calculate the trial's reported results.
8. Statistical Methodology
Time-to-event analysis
Overall survival is a time-to-event endpoint. Rather than recording only whether a patient died, time-to-event analysis uses the timing of the event and the available follow-up for each participant. This is important because participants can have different observation periods.
A participant who remains alive at the end of available observation has not necessarily experienced the event. Such an observation is generally handled through censoring methods rather than being treated as an observed death. The ClinicalTrials.gov record identifies the endpoint as time-to-event and specifies a logrank comparison, but it does not provide the full censoring rules.
Stratified log-rank procedure
The registered primary definition specifically calls for a logrank procedure stratified by secondary surgical debulking status. Conceptually, a stratified analysis performs the time-to-event comparison within the specified strata and combines the evidence across them.
The logrank framework is designed to compare survival experience over time rather than simply comparing the proportion of deaths at one arbitrary time point.
Randomized-treatment analysis
The registry wording is explicit that patients are categorized by their randomly assigned treatment. This is an important statistical principle: the primary randomized comparison should preserve the treatment assignment generated by the trial design rather than selectively reclassifying participants after treatment has begun.
Overall survival as an endpoint
Overall survival has a particularly direct interpretation because the event is death. The registry endpoint does not define a disease-specific cause of death; therefore, this page describes the endpoint using the registry's overall-survival wording rather than introducing a different event definition.
Follow-up time
Both registered primary endpoints specify a time frame of 82.5 months, described as the median duration of follow-up. This is a follow-up descriptor, not an estimate of median overall survival. Those two quantities should not be confused.
9. Statistical Methods Explained
Why was a logrank test used?
The primary endpoints are time-to-event outcomes. A simple comparison of the percentage of patients who had died would discard information about when deaths occurred and would not naturally accommodate unequal follow-up. The logrank procedure instead compares the event experience across the observed follow-up period.
Why stratify the logrank test by surgical debulking status?
The registry definition identifies secondary surgical debulking status as the stratification variable. Stratification allows the randomized treatment comparison to be evaluated while accounting for these prespecified categories. It prevents the analysis from treating the three identified debulking-status categories as though they were irrelevant to the primary comparison.
What does randomization contribute statistically?
Randomization creates treatment groups through an allocation mechanism rather than through investigators selecting treatment based on observed patient characteristics. The registry's statement that patients are included according to randomly assigned treatment preserves that randomized comparison in the primary analysis framework.
Why is overall survival a time-to-event endpoint?
Because the endpoint concerns the duration until death, both the event status and the amount of observed time matter. A patient followed for a short period without death does not provide the same information as a patient followed for substantially longer without death. Time-to-event methods are designed for this structure.
What does the 82.5-month time frame mean?
The registry describes 82.5 months as the median duration of follow-up. It describes how long participants were followed, not the median time until death. Without a reported median overall-survival estimate, it would be incorrect to substitute 82.5 months for median survival.
Can a p-value or hazard ratio be calculated from the enrollment number alone?
No. The enrollment of 1052 participants does not contain enough information to reconstruct an overall-survival treatment effect. A survival comparison requires information about event timing, censoring, treatment assignment, and the specified analysis framework. None of the formal numerical primary-endpoint analyses is included in the posted results.
10. Serious Adverse Events
| registry safety entry | Affected / at risk |
|---|---|
| Arm I (no Surgery; Carboplatin and Pacli | 37/316 |
| Arm II (no Surgery; Carboplatin, Paclita | 104/456 |
| Arm III (Surgery; Carboplatin and Paclit | 1/33 |
| Arm IV (Surgery; Carboplatin, Paclitaxel | 24/165 |
| Arm V (no Surgery; Carboplatin and Gemci | 0/6 |
| Arm VI (no Surgery; Carboplatin, Gemcita | 1/34 |
| Additional registry-reported entry | Arm VI… |
11. What the Statistical Design Can and Cannot Tell Us
What it can address
The registered primary analysis can compare overall-survival experience between the randomized treatment regimens using a stratified logrank procedure.
What the enrollment number cannot address
The sample size alone cannot establish the magnitude, precision, or statistical significance of a treatment effect.
What the follow-up value tells us
The 82.5-month value describes median follow-up duration for the primary endpoints.
What remains unavailable here
The ClinicalTrials.gov record does not contain formal effect estimates, confidence intervals, p-values, or survival summaries.
12. Limitations
- Missing formal statistical analyses: No formal statistical analyses were posted despite the registry profile stating that results are posted.
- No numerical primary results reported: Hazard ratios, confidence intervals, p-values, event counts, median overall survival, and survival probabilities are not available in the ClinicalTrials.gov record.
- Incomplete arm information: The profile states that the trial has 4 arms, while the registry-reported serious-adverse-event field contains incomplete labels extending through Arm VI. The arm structure cannot safely be reconstructed from that field.
- Limited censoring information: The ClinicalTrials.gov record identifies overall survival as a time-to-event endpoint and specifies a logrank procedure but does not provide the complete censoring rules.
- No reconstructed results: Numerical results have not been derived from enrollment, follow-up duration, or other partial information because doing so would create statistics not contained in the registry-reported source.
- Posted results versus the ClinicalTrials.gov record: The distinction matters. A registry may contain posted outcome data while the particular statistical-analysis extraction in the ClinicalTrials.gov record contains no formal analysis records.
13. Why This Trial Matters Statistically
GOG-0213 is a useful teaching case because the registry explicitly links a randomized clinical-trial question to a time-to-event endpoint and a stratified logrank analysis. It also illustrates why a statistical analysis page should distinguish between trial design, registered analysis methodology, and reported numerical results.
| Concept | How it appears in GOG-0213 |
|---|---|
| Randomization | The allocation is listed as RANDOMIZED, and the primary analysis definition uses randomly assigned treatment. |
| Parallel design | The design model is listed as PARALLEL. |
| Time-to-event analysis | Both registered primary endpoints are identified as time-to-event outcomes. |
| Overall survival | Both primary questions use duration of overall survival as the outcome. |
| Logrank testing | The registry specifies a logrank procedure for comparing treatment regimens. |
| Stratification | The registry-reported definition specifies stratification by secondary surgical debulking status. |
| Follow-up | The primary endpoint time frame is 82.5 months, described as median duration of follow-up. |
| Analysis reporting | Results are posted in the registry profile, but no formal statistical-analysis entries are reported in the source data used here. |
14. Interpreting the Primary Endpoint Correctly
Overall survival is evaluated over time. The relevant statistical question is not simply how many participants died, but how the timing of death differs between randomized treatment groups while accounting for the available follow-up.
The logrank procedure evaluates evidence of a difference in survival experience across the follow-up period. It is therefore conceptually different from comparing survival percentages at one selected time point.
Because the registry specifies secondary surgical debulking status as a stratification variable, the primary comparison is not an unstratified logrank test. The analysis is structured to account for the specified debulking-status categories.
The randomized design and prespecified analysis method describe how evidence should be evaluated. They do not, by themselves, establish the direction or magnitude of the observed treatment effect. Those quantities require the actual event-level or formal statistical results.
15. Results Reporting: A Statistical Reading Guide
If the complete posted outcome-analysis information were available, the primary results should be read in a structured order rather than focusing on a single p-value.
Confirm what was measured
Here, the primary outcome is overall survival and the registered endpoint is time-to-event.
Confirm who contributed to the comparison
The registry-reported endpoint definition says patients are categorized by their randomly assigned treatment.
Confirm the comparison method
The registry specifies a logrank procedure stratified by secondary surgical debulking status.
Examine the magnitude
A complete report would normally provide a treatment-effect estimate and its uncertainty. Those numerical values are not included in the ClinicalTrials.gov record.
Separate statistical evidence from clinical interpretation
A statistical comparison establishes evidence about the randomized groups; interpretation of clinical importance requires the magnitude and precision of the observed effect together with the endpoint definition and treatment context.
16. Statistical Questions Raised by GOG-0213
Why randomize?
Randomization provides the framework for comparing treatment groups while preserving the assigned-treatment analysis specified by the registry.
Why use time-to-event methods?
Patients can have different observation times, and the timing of death contains information that a simple event proportion would discard.
Why stratify?
The registry explicitly identifies secondary surgical debulking status as the stratification variable for the logrank procedure.
A hazard ratio requires actual survival-event information or a reported model result. The ClinicalTrials.gov record contains neither.
17. Clinical Biostats Interpretation
The GOG-0213 registry establishes two randomized overall-survival questions and specifies a stratified logrank framework for comparing treatment regimens. That is the documented statistical design. It does not, in the ClinicalTrials.gov record, provide the numerical effect estimates needed to state how large the observed difference was.
If a complete analysis reported a hazard ratio, it would summarize the relative event rate under the specified survival model. It would not mean that the same percentage of individual patients experienced a reduction in their personal probability of death, and it would not replace absolute survival estimates.
A confidence interval would describe statistical uncertainty around the estimated treatment effect under the analysis framework. Its width would provide information about precision that a point estimate alone cannot provide.
A p-value would quantify the evidence against the specified null hypothesis under the test procedure. It would not measure the size of the treatment effect, the probability that the treatment is beneficial, or the clinical importance of an observed difference.
The ClinicalTrials.gov record specifies a logrank analysis but does not provide a Cox-model result or document a proportional-hazards assessment. Therefore, this page does not assume or report a proportional-hazards model as a documented GOG-0213 result.
18. What Is Known From the ClinicalTrials.gov record
| Question | Answer supported by the ClinicalTrials.gov record |
|---|---|
| Was the trial randomized? | Yes. |
| Was it a phase 3 trial? | Yes. |
| How many participants were enrolled? | 1052. |
| How many arms are listed in the ClinicalTrials.gov record? | 4. |
| Were the primary endpoints time-to-event outcomes? | Yes. |
| What were the primary endpoints? | Two overall-survival questions: secondary cytoreduction plus adjuvant chemotherapy, and addition of bevacizumab relative to second-line paclitaxel/carboplatin alone. |
| What was the stated follow-up time frame? | 82.5 months, described as median duration of follow-up. |
| What statistical test was registered? | Logrank procedure. |
| Was the logrank analysis stratified? | Yes, by secondary surgical debulking status. |
| Were formal statistical-analysis entries reported? | No; no formal statistical analyses were posted. |
| Does the ClinicalTrials.gov record contains primary endpoint HRs? | No. |
| Does the ClinicalTrials.gov record contains primary endpoint p-values? | No. |
19. Related Tutorials
Learn more about the methods used in this trial:
20. Related Statistical Calculators
21. Sources
- ClinicalTrials.gov: GOG-0213 (NCT00565851). Trial profile, registered primary endpoints, design information, and posted-results status.
- PubMed: PMID 37185961.
- PubMed: PMID 31722153.
- PubMed: PMID 28438473.
Continue through the Clinical Biostats statistical pathway
Explore the underlying survival-analysis concepts and statistical calculators used to understand randomized clinical-trial time-to-event endpoints.
22. Record Summary
GOG-0213 is a randomized phase 3 trial with 1052 enrolled participants and four listed arms. Its two registered primary endpoints are time-to-event measures concerning overall survival: one evaluating surgical secondary cytoreduction in addition to adjuvant chemotherapy, and the other evaluating addition of bevacizumab relative to second-line paclitaxel and carboplatin alone. The registered primary analysis uses a logrank procedure stratified by secondary surgical debulking status.
The central statistical lesson is the distinction between a prespecified analysis framework and the numerical evidence produced by that analysis. The ClinicalTrials.gov record documents randomization, the parallel design, the overall-survival endpoints, the 82.5-month median follow-up time frame, and the stratified logrank methodology. It does not supply formal statistical-analysis entries or numerical treatment-effect estimates.