This page separates reported trial results from statistical interpretation. The numerical results presented here are restricted to the ClinicalTrials.gov record. The registry provides the official trial record.
1. Trial at a Glance
ExteNET was a randomized, parallel-group, quadruple-masked phase 3 trial evaluating neratinib versus placebo in women with early stage breast cancer after adjuvant trastuzumab. The registered primary endpoints were both time-to-event measures of invasive disease-free survival, with one formal primary endpoint analysis reported using a stratified log-rank test and stratified Cox model.
| Feature | ExteNET |
|---|---|
| Trial name | ExteNET |
| Phase | Phase 3 |
| Condition | Breast Cancer |
| Brief title | Study Evaluating The Effects Of Neratinib After Adjuvant Trastuzumab In Women With Early Stage Breast Cancer |
| Design | Randomized, parallel |
| Masking | Quadruple |
| Allocation | Randomized |
| Enrollment | 2840 |
| Interventions | Neratinib and placebo |
| Primary endpoints | 2 registered primary endpoints |
| Primary endpoint type | Time-to-event |
| Lead sponsor | Puma Biotechnology, Inc. |
| Sponsor type | Industry |
| ClinicalTrials.gov | NCT00878709 |
2. Clinical Question
The registered trial question was whether treatment with neratinib, compared with placebo, was associated with a different invasive disease-free survival experience in women with early stage breast cancer after adjuvant trastuzumab.
Population
Women with early stage breast cancer who had received adjuvant trastuzumab, as represented by the registered ExteNET population.
Intervention
Neratinib.
Comparator
Placebo.
Primary question
How does invasive disease-free survival in the neratinib arm compare with the placebo arm at the registered time points?
3. Trial Design
Neratinib
- Neratinib was the registered drug intervention.
- Primary and secondary efficacy comparisons were made against placebo.
- Efficacy analyses used the intention-to-treat population.
Placebo
- Placebo was the registered comparator intervention.
- Primary and secondary efficacy comparisons were made against neratinib.
- Efficacy analyses used the intention-to-treat population.
The ClinicalTrials.gov record identifies the trial as randomized, parallel, quadruple-masked, and treatment-purpose. They do not provide additional treatment-administration details in the ClinicalTrials.gov record, so those details are not reproduced here.
4. Randomization, Stratification, and Analysis Population
The primary iDFS analysis was conducted in the intent-to-treat (ITT) population. The registry-reported analysis description states that this population included all participants who were randomized, regardless of whether they received any study treatment.
| Analysis population | Definition / role |
|---|---|
| Intent-to-treat | All participants who were randomized regardless of whether they received any study treatment; used for the reported efficacy analyses. |
| Safety population | The ClinicalTrials.gov record reports serious adverse events by randomized treatment arm using affected/at-risk counts. |
Stratification factors
The primary analysis used stratification for the prior trastuzumab, nodal status, and ER/PgR status factors specified in the registry analysis text:
- Prior trastuzumab: concurrent or sequential.
- Nodal status: ≤ 3 or ≥ 4.
- ER/PgR status: positive or negative.
The same three factors were identified for the stratified log-rank test and the stratified Cox model in the reported primary analysis and were also identified in the secondary time-to-event analyses.
5. Registered Primary Endpoints
| Endpoint | Registry definition / time frame | Analysis information |
|---|---|---|
| Invasive Disease-free Survival (iDFS) in Neratinib Arm Compared to Placebo Arm at Year 2 | From randomization until time of event up to 2 years. Invasive disease-free survival time is defined as the time from date of randomization until the first disease recurrence of the following events: invasive ipsilateral breast tumor recurrence, invasive contralateral breast cancer, local/regional invasive recurrence, distant recurrence and death from any cause. | Formal analysis posted: stratified log-rank test; hazard ratio estimated by stratified Cox model. |
| Kaplan-Meier Estimates of Invasive Disease-free Survival (iDFS) at Year 2 by Treatment Arms | From randomization until time of event up to 2 years. | Results posted, but the ClinicalTrials.gov record does not provide a separate formal analysis entry or numerical Kaplan-Meier estimates for this endpoint. |
The two registered primary endpoints describe closely related views of the same time-to-event outcome: the first is the treatment comparison, while the second specifies Kaplan-Meier estimates by treatment arm. The ClinicalTrials.gov record provides a formal treatment comparison for the first endpoint but do not provide separate numerical Kaplan-Meier estimates for the second.
6. Statistical Methodology
Kaplan-Meier estimation
Kaplan-Meier estimation is the standard descriptive framework for a time-to-event endpoint such as iDFS. It estimates the probability of remaining event-free over time while accounting for participants whose event status is not yet observed because follow-up ends before an event.
where di is the number of events at time ti and ni is the number at risk immediately before that time.
The registry specifically names Kaplan-Meier estimates as a primary endpoint at Year 2, but the ClinicalTrials.gov record does not give the corresponding arm-specific numerical estimates.
Stratified log-rank test
The primary iDFS comparison used a stratified log-rank test. Rather than treating every participant as belonging to a single homogeneous risk set, the analysis accounts for the prespecified stratification factors while comparing the time-to-event experience between randomized treatment groups.
In the registry-reported analysis, the stratification factors were prior trastuzumab, nodal status, and ER/PgR status. This makes the reported comparison consistent with the randomized design and reduces the influence of imbalance across these predefined categories.
Stratified Cox proportional-hazards model
The hazard ratio was estimated using a stratified Cox proportional-hazards model. The model uses the observed event times and censoring information to estimate the relative event hazard associated with neratinib compared with placebo while preserving the specified stratification structure.
A hazard ratio is a relative time-to-event measure. It is not the same thing as a relative risk, an absolute risk difference, or the proportion of participants who benefit.
Intention-to-treat analysis
The primary iDFS efficacy analysis used the ITT population: every randomized participant was included regardless of whether study treatment was received. This preserves the treatment assignment created by randomization and avoids redefining the randomized comparison based on subsequent treatment exposure.
Why stratification matters
Stratification allows the analysis to respect clinically relevant factors used in the trial's statistical comparison. The reported HR therefore represents a treatment comparison from a stratified Cox model rather than an unqualified unstratified estimate.
7. Primary Result: Invasive Disease-free Survival at Year 2
The registry analysis reports a formal comparison of iDFS in the neratinib arm versus the placebo arm from randomization until the time of event up to 2 years.
Year 2 iDFS hazard ratio
95% CI: 0.49–0.90 · P = 0.008
Analysis: stratified log-rank test; HR estimated by stratified Cox model.
| Primary endpoint | Analysis population | Method | Effect measure | Estimate | 95% CI | P-value |
|---|---|---|---|---|---|---|
| iDFS at Year 2 | ITT | Stratified log-rank; stratified Cox model | Hazard ratio | 0.66 | 0.49–0.90 | 0.008 |
The reported hazard ratio of 0.66 means that the estimated instantaneous rate of an iDFS event in the neratinib group was approximately 66% of the corresponding rate in the placebo group under the stratified Cox model. Equivalently, the point estimate corresponds to an estimated 34% lower hazard for the neratinib group relative to placebo.
The HR does not mean that 34% of participants avoided an event, that 34% more participants were cured, or that every participant experienced exactly the same relative reduction. It is a model-based relative measure of event hazard over the analyzed time-to-event experience.
The two-sided 95% confidence interval of 0.49–0.90 describes uncertainty around the estimated hazard ratio under the statistical model and sampling framework. It does not describe the range of individual treatment effects or predict the outcome for an individual participant.
The P = 0.008 value addresses evidence against the null comparison under the reported statistical testing framework. It is not a measure of the size or clinical importance of the treatment effect. Effect size is conveyed by the HR and its confidence interval.
The analysis was based on randomized participants and used stratified log-rank and Cox methods. As with other Cox-model hazard ratios, interpretation also depends on the adequacy of the proportional-hazards representation over the analyzed follow-up. The ClinicalTrials.gov record does not provide a separate assessment of that assumption.
The second registered primary endpoint
The registry also lists Kaplan-Meier Estimates of Invasive Disease-free Survival (iDFS) at Year 2 by Treatment Arms, with a time frame from randomization until time of event up to 2 years. The ClinicalTrials.gov record states that results were posted for this endpoint, but no separate statistical-analysis record with numerical Kaplan-Meier estimates is included in the ClinicalTrials.gov recordset.
8. Secondary Time-to-Event Results at Year 2
The ClinicalTrials.gov record contains four secondary time-to-event analyses at Year 2. Each used the ITT population and compared neratinib with placebo. The reported Cox analyses were stratified by prior trastuzumab, nodal status, and ER/PgR status.
| Secondary endpoint | Time frame | Method | HR | 95% CI | P-value |
|---|---|---|---|---|---|
| Overall Survival (OS) | Randomization until death due to any cause (up to 119 Months) | Stratified log-rank; stratified Cox model | 0.952 | 0.747–1.212 | 0.6914 |
| DFS including ductal carcinoma in situ (DFS-DCIS) at Year 2 | From randomization until time of event up to 2 years | Stratified Cox model | 0.61 | 0.45–0.83 | Not reported |
| Distant Disease-free Survival (DDFS) at Year 2 | From randomization until time of event up to 2 years | Stratified Cox model | 0.74 | 0.52–1.05 | Not reported |
| Time to Distant Recurrence (TTDR) at Year 2 | From randomization until time of event up to 2 years | Stratified Cox model | 0.73 | 0.51–1.04 | Not reported |
Overall Survival
OS hazard ratio
95% CI: 0.747–1.212 · P = 0.6914
Time frame: randomization until death due to any cause, up to 119 Months.
The reported OS point estimate is close to 1, and its 95% confidence interval extends on both sides of 1. The reported P-value is 0.6914. These quantities describe the registry-reported statistical comparison; they do not establish an absolute difference in survival probabilities because no arm-specific survival estimates are reported here.
Disease-free Survival Including Ductal Carcinoma in Situ
The hazard ratio was estimated by a stratified Cox proportional-hazards model in the ITT population. The ClinicalTrials.gov record does not report a P-value for this analysis.
Distant Disease-free Survival
The hazard ratio was estimated by a stratified Cox proportional-hazards model in the ITT population. The ClinicalTrials.gov record does not report a P-value for this analysis.
Time to Distant Recurrence
The hazard ratio was estimated by a stratified Cox proportional-hazards model in the ITT population. The ClinicalTrials.gov record does not report a P-value for this analysis.
These secondary endpoints illustrate why confidence intervals should accompany hazard ratios. The Year 2 DFS-DCIS interval is 0.45–0.83, whereas the DDFS and TTDR intervals extend above 1. The ClinicalTrials.gov record does not provide a common multiplicity-adjusted interpretation for these secondary analyses, so they should be read as the registry-reported estimates rather than as a ranked set of confirmatory findings.
9. Five-Year Prespecified Analyses
The ClinicalTrials.gov record includes four other prespecified analyses at Year 5. These analyses retained the ITT population and the same general stratified Cox framework for the time-to-event endpoints.
| Endpoint | Time frame | Method | HR | 95% CI | P-value |
|---|---|---|---|---|---|
| iDFS at Year 5 | From randomization until time of event up to 5 years | Stratified log-rank; stratified Cox model | 0.73 | 0.57–0.92 | 0.008 |
| DFS-DCIS at Year 5 | From randomization until time of event up to 5 years | Stratified Cox model | 0.71 | 0.56–0.89 | Not reported |
| DDFS at Year 5 | From randomization until time of event up to 5 years | Stratified Cox model | 0.78 | 0.6–1.01 | Not reported |
| TTDR at Year 5 | From randomization until time of event up to 5 years | Stratified Cox model | 0.79 | 0.60–1.03 | Not reported |
Invasive Disease-free Survival at Year 5
Year 5 iDFS hazard ratio
95% CI: 0.57–0.92 · P = 0.008
Analysis: stratified log-rank test; HR estimated by stratified Cox model.
The Year 5 iDFS hazard ratio of 0.73 corresponds to an estimated instantaneous iDFS event rate approximately 73% as large in the neratinib group as in the placebo group under the reported stratified Cox model. As a simple relative interpretation, the point estimate corresponds to an estimated 27% lower hazard.
This does not mean that 27% of participants avoided an event, nor does it mean that an individual participant's probability of remaining disease-free was reduced or increased by exactly 27%. The hazard ratio summarizes the relative event hazard estimated by the model.
The 95% CI of 0.57–0.92 conveys the precision of the estimated relative treatment effect. It is narrower than an interval that would provide little information about the direction of the effect, but it remains an interval estimate rather than a guarantee about the effect in another population.
The reported P = 0.008 is evidence against the corresponding null hypothesis under the stated testing framework. It should not be interpreted as a 0.8% probability that the null hypothesis is true, nor as a measure of the magnitude of the treatment effect.
Because the Year 5 estimate is still a Cox-model hazard ratio, its interpretation should remain tied to the model assumptions, censoring process, stratification, and ITT analysis population.
DFS-DCIS at Year 5
The ClinicalTrials.gov record identifies a stratified Cox model and ITT analysis. No P-value is reported for this endpoint in the ClinicalTrials.gov record.
DDFS at Year 5
The ClinicalTrials.gov record identifies a stratified Cox model and ITT analysis. No P-value is reported for this endpoint in the ClinicalTrials.gov record.
TTDR at Year 5
The ClinicalTrials.gov record identifies a stratified Cox model and ITT analysis. No P-value is reported for this endpoint in the ClinicalTrials.gov record.
10. Statistical Methods Explained
Why was a stratified log-rank test used?
The primary iDFS comparison is a time-to-event question, so the analysis must account for both the timing of events and participants who are censored. A log-rank test compares the event experience between randomized groups over follow-up. In ExteNET, the test was stratified by prior trastuzumab, nodal status, and ER/PgR status, matching the factors specified in the ClinicalTrials.gov record.
What does an iDFS hazard ratio of 0.66 mean?
A hazard ratio of 0.66 is a relative estimate of the instantaneous event rate. Under the fitted stratified Cox model, the estimated event hazard for neratinib was 66% of the placebo hazard. The corresponding point estimate can be described as an estimated 34% lower hazard. It does not represent a 34-percentage-point difference in disease-free survival.
Why is the confidence interval important?
The point estimate is only one estimate of the treatment effect. The 95% confidence interval provides an interval describing statistical uncertainty around that estimate. For the Year 2 iDFS HR, the interval is 0.49–0.90. For the Year 5 iDFS HR, it is 0.57–0.92. The intervals therefore provide information about both the estimated direction and the precision of the treatment comparison.
Why does the P-value not measure effect size?
A P-value addresses the compatibility of the observed data with a specified null hypothesis under the statistical testing framework. It does not tell us how large the treatment effect is. The HR provides the relative effect estimate, while its confidence interval describes uncertainty around that estimate. The Year 2 iDFS result illustrates this distinction: HR 0.66 and P = 0.008 answer different statistical questions.
Why use an ITT population?
The ITT principle keeps participants in the groups to which they were randomized. In the registry-reported ExteNET analysis, the ITT population included all randomized participants regardless of whether they received study treatment. This preserves the randomized comparison rather than redefining groups after randomization.
What does stratification change in the Cox model?
A stratified Cox model allows the baseline hazard to differ across the specified strata while estimating a common treatment hazard ratio across those strata. Here, the registry-reported analysis identifies prior trastuzumab, nodal status, and ER/PgR status as the stratification factors.
Why should Year 2 and Year 5 hazard ratios not be treated as independent trials?
Both estimates arise from the same randomized trial and address the same broad time-to-event treatment comparison at different follow-up horizons. The later estimate adds information from longer follow-up; it does not represent a new randomization. Comparisons across time points therefore need to account for their shared participants and event history.
11. Interpreting Time-to-Event Endpoints
Time-to-event analysis differs from a simple comparison of proportions because the timing of an event matters. Two participants who both experience an event do not necessarily contribute the same information if their events occur at very different times.
Event
For the registered iDFS definition, an event is the first occurrence of one of the specified invasive recurrence events or death from any cause.
Censoring
Participants without an observed event by the end of their available follow-up contribute information up to the time at which they are censored.
Kaplan-Meier
Provides an estimate of the event-free survival function over time.
Cox model
Provides a relative hazard estimate while allowing the analysis to account for the specified stratification structure.
The distinction is important when interpreting the ExteNET results. A hazard ratio summarizes relative event rates over time; it is not interchangeable with an absolute survival probability at a particular year.
12. Comparing the Reported Hazard Ratios
| Endpoint | Time point / frame | HR | 95% CI | P-value |
|---|---|---|---|---|
| iDFS | Year 2 | 0.66 | 0.49–0.90 | 0.008 |
| OS | Up to 119 Months | 0.952 | 0.747–1.212 | 0.6914 |
| DFS-DCIS | Year 2 | 0.61 | 0.45–0.83 | Not reported |
| DDFS | Year 2 | 0.74 | 0.52–1.05 | Not reported |
| TTDR | Year 2 | 0.73 | 0.51–1.04 | Not reported |
| iDFS | Year 5 | 0.73 | 0.57–0.92 | 0.008 |
| DFS-DCIS | Year 5 | 0.71 | 0.56–0.89 | Not reported |
| DDFS | Year 5 | 0.78 | 0.6–1.01 | Not reported |
| TTDR | Year 5 | 0.79 | 0.60–1.03 | Not reported |
The table should be read as a collection of reported estimates, not as a ranking of endpoints. Different endpoints represent different definitions of disease control and death, and the ClinicalTrials.gov record does not provide enough information to treat their hazard ratios as directly interchangeable measures of one underlying outcome.
13. Multiplicity and Multiple Endpoints
ExteNET has two registered primary endpoints, followed by multiple secondary and other prespecified time-to-event analyses. This creates an important statistical distinction between the existence of a reported estimate and the confirmatory interpretation of that estimate.
| Analysis family | Role in the ClinicalTrials.gov record | Interpretation issue |
|---|---|---|
| Year 2 iDFS treatment comparison | Primary endpoint with formal analysis | Reported using stratified log-rank testing and stratified Cox estimation. |
| Year 2 Kaplan-Meier iDFS estimates | Primary endpoint | Results are identified as posted, but numerical estimates are not included in the ClinicalTrials.gov record. |
| OS, DFS-DCIS, DDFS, TTDR | Secondary endpoints | Multiple time-to-event comparisons require attention to the broader testing framework. |
| Year 5 iDFS, DFS-DCIS, DDFS, TTDR | Other prespecified analyses | Longer-term estimates should be distinguished from the Year 2 primary analysis. |
Multiplicity matters because the probability of obtaining at least one apparently unusual result increases when many statistical hypotheses are evaluated. The ClinicalTrials.gov record identifies hypothesis types for the analyses, but they do not provide a complete multiplicity-adjustment procedure. Accordingly, this page does not assign an adjusted confirmatory interpretation to the secondary or other prespecified estimates.
14. Safety Results
the ClinicalTrials.gov record reports serious adverse events by treatment arm using affected participants and participants at risk.
| Safety measure | Neratinib | Placebo |
|---|---|---|
| Serious adverse events, affected / at risk | 103 / 1408 | 85 / 1408 |
These are arm-specific counts as reported in the ClinicalTrials.gov record. They should not be confused with an efficacy event count: serious adverse events are safety outcomes, whereas iDFS events are defined by the registered disease-free survival endpoint.
15. Longitudinal Trial History
Trial start
The ExteNET trial began on 2009-07-09 according to the registry profile.
Primary completion
The registry profile gives 2014-08-21 as the primary completion date.
Primary iDFS comparison
The reported Year 2 primary iDFS analysis produced HR 0.66 with a 95% CI of 0.49–0.90 and P = 0.008.
Longer-term prespecified endpoints
The ClinicalTrials.gov record reports Year 5 iDFS HR 0.73 with a 95% CI of 0.57–0.92 and P = 0.008, alongside additional DFS-DCIS, DDFS, and TTDR analyses.
16. What the Hazard Ratio Does — and Does Not — Mean
A reported iDFS hazard ratio of 0.66 at Year 2 means that, under the stratified Cox model, the estimated instantaneous rate of an iDFS event in the neratinib group was approximately 66% of that in the placebo group. The point estimate can therefore be described as an estimated 34% lower hazard.
It does not mean that 34% of participants avoided recurrence or death, that 34% of patients were cured, or that every participant experienced exactly a 34% reduction in risk.
The Year 2 95% CI of 0.49–0.90 quantifies statistical uncertainty around the HR estimate under the model and sampling framework. The Year 5 interval of 0.57–0.92 provides the corresponding uncertainty statement for the longer follow-up analysis.
The Year 2 and Year 5 iDFS analyses both report P = 0.008. A P-value is evidence relative to a null hypothesis under the specified testing procedure; it is not a measure of treatment-effect magnitude. The HR and confidence interval should be used to describe the estimated effect and its precision.
17. Statistical Interpretation of the Secondary Results
The secondary analyses illustrate several distinct statistical situations.
OS
The reported HR is 0.952 with a 95% CI of 0.747–1.212 and P = 0.6914. The confidence interval spans 1, so the point estimate should not be interpreted without its uncertainty interval.
DFS-DCIS
The Year 2 HR is 0.61 with a 95% CI of 0.45–0.83. The ClinicalTrials.gov record does not provide a P-value for this analysis.
DDFS
The Year 2 HR is 0.74 with a 95% CI of 0.52–1.05. The interval extends above 1, and no P-value is in the ClinicalTrials.gov record.
TTDR
The Year 2 HR is 0.73 with a 95% CI of 0.51–1.04. The interval extends above 1, and no P-value is in the ClinicalTrials.gov record.
For the Year 5 analyses, the reported HRs are 0.73 for iDFS, 0.71 for DFS-DCIS, 0.78 for DDFS, and 0.79 for TTDR. Their corresponding confidence intervals should be read separately because each endpoint has its own event definition and uncertainty.
18. Important Limitations and Interpretation Issues
- Incomplete numerical endpoint reporting: the ClinicalTrials.gov record identifies two primary endpoints, but separate numerical Kaplan-Meier estimates for the Year 2 arm-specific iDFS endpoint are not included in the ClinicalTrials.gov record.
- Hazard-ratio interpretation: an HR is a model-based relative measure and should not be substituted for an absolute event probability or an absolute risk difference.
- Proportional-hazards assumption: Cox-model interpretation relies on a proportional-hazards framework. The ClinicalTrials.gov record does not provide a formal diagnostic or test of this assumption.
- Multiple endpoints: the trial includes two primary endpoints and multiple secondary and prespecified analyses. The ClinicalTrials.gov record does not provide enough information to reconstruct a complete multiplicity-adjustment strategy.
- Different endpoint definitions: iDFS, DFS-DCIS, DDFS, TTDR, and OS are not interchangeable outcomes. A hazard ratio for one should not automatically be treated as a surrogate for another.
- Secondary P-values: the ClinicalTrials.gov record provides a P-value for OS and the two reported iDFS analyses, but not for the other secondary and prespecified Cox analyses. Missing P-values are not reconstructed from the confidence intervals.
- Safety detail: the ClinicalTrials.gov record is limited to serious adverse-event affected/at-risk counts by arm and does not provide a detailed safety analysis.
- Generalizability: the analysis describes the randomized ExteNET population represented in the registry and should not automatically be generalized to populations outside the trial's eligibility framework.
19. Why This Trial Matters Statistically
ExteNET is a useful teaching case because its registry data bring together several core concepts in survival analysis: randomized treatment assignment, quadruple masking, ITT analysis, prespecified stratification, Kaplan-Meier estimation, stratified log-rank testing, Cox proportional-hazards modeling, hazard ratios, confidence intervals, multiple time horizons, and multiple endpoint definitions.
| Concept | How it appears in ExteNET |
|---|---|
| Randomization | The trial is registered as randomized with two parallel treatment arms. |
| Blinding | The registry identifies the trial as quadruple masked. |
| ITT analysis | The reported efficacy analyses include all randomized participants regardless of whether they received study treatment. |
| Kaplan-Meier estimation | The Year 2 primary endpoints include Kaplan-Meier estimates of iDFS by treatment arm. |
| Log-rank testing | The formal Year 2 iDFS analysis and Year 5 iDFS analysis use a stratified log-rank test. |
| Hazard ratio | Relative treatment effects are reported as hazard ratios from stratified Cox models. |
| Confidence intervals | Reported HRs are accompanied by two-sided 95% confidence intervals. |
| Stratified analysis | Prior trastuzumab, nodal status, and ER/PgR status are used as stratification factors. |
| Multiple endpoints | The registry contains two primary endpoints plus multiple secondary and prespecified analyses. |
| Longer follow-up | The registry reports both Year 2 and Year 5 time-to-event analyses. |
20. Related Tutorials
Learn more about the methods used in this trial:
21. Related Statistical Calculators
22. Sources
- ClinicalTrials.gov: NCT00878709 — ExteNET.
- Linked PubMed record: PMID 36898233.
- Linked PubMed record: PMID 31140297.
- Linked PubMed record: PMID 30867034.
- Linked PubMed record: PMID 30813966.
- Linked PubMed record: PMID 30689703.
Continue through Clinical Biostats
Connect this trial's time-to-event methods with statistical tutorials, calculators, and additional clinical-trial analyses.
23. Record Summary
ExteNET provides a compact example of how a randomized clinical trial can be analyzed through a sequence of time-to-event methods. The ClinicalTrials.gov record defines iDFS as time from randomization to the first qualifying invasive recurrence event or death, use an ITT population for efficacy, and apply stratified log-rank and Cox methods incorporating prior trastuzumab, nodal status, and ER/PgR status. The reported Year 2 iDFS comparison has an HR of 0.66 (95% CI 0.49–0.90; P = 0.008), while the reported Year 5 iDFS comparison has an HR of 0.73 (95% CI 0.57–0.92; P = 0.008).
The secondary analyses demonstrate why a statistical analysis should distinguish endpoint definitions and uncertainty rather than reducing a trial to a single number. OS, DFS-DCIS, DDFS, and TTDR have different event definitions and reported hazard ratios, and the ClinicalTrials.gov record does not provide P-values for every secondary or prespecified analysis. The registry also reports serious adverse-event counts of 103/1408 for neratinib and 85/1408 for placebo, providing safety information that should be considered separately from efficacy endpoints.