This page separates trial information reported in ClinicalTrials.gov from statistical interpretation and educational explanation. The registry reports that results have been posted, but the ClinicalTrials.gov record contains no formal statistical-analysis estimates, confidence intervals, or p-values for the primary endpoint.
1. Trial at a Glance
NSABP B-42 was a randomized, double-masked, parallel phase 3 treatment trial evaluating letrozole versus placebo in postmenopausal women who had received hormone therapy for hormone receptor-positive breast cancer. The registered primary endpoint was disease-free survival at 7 years.
| Feature | NSABP B-42 |
|---|---|
| Trial name | NSABP B-42 |
| NCT identifier | NCT00382070 |
| Brief title | Letrozole in Treating Postmenopausal Women Who Have Received Hormone Therapy for Hormone Receptor-Positive Breast Cancer |
| Condition | Breast Cancer |
| Phase | Phase 3 |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Double |
| Primary purpose | Treatment |
| Enrollment | 3966 |
| Arms | 2 |
| Interventions | Letrozole; Placebo |
| Lead sponsor | NSABP Foundation Inc |
| Sponsor type | Network |
2. Clinical Question
The registered trial question can be framed around whether treatment with letrozole, compared with placebo, affects disease-free survival in postmenopausal women who have received hormone therapy for hormone receptor-positive breast cancer.
Population
Postmenopausal women who have received hormone therapy for hormone receptor-positive breast cancer.
Intervention
Letrozole.
Comparator
Placebo.
Primary question
How does letrozole compare with placebo for the registered disease-free survival endpoint over the 7-year time frame?
3. Trial Design
Letrozole
- Intervention: Letrozole
- Randomized treatment arm
- Included in the double-masked parallel comparison
Placebo
- Intervention: Placebo
- Randomized comparator arm
- Included in the double-masked parallel comparison
The registry identifies the allocation as randomized, the design model as parallel, and masking as double. These three features are statistically important because they establish the basic structure of the comparison: patients are assigned to one of two concurrent treatment groups, and masking is intended to reduce the influence of treatment knowledge during the trial.
4. Trial Timeline and Registry Status
Trial start
The registry lists the study start as August 2006.
Primary completion
The registry lists the primary completion date as August 25, 2016.
Results posted
The ClinicalTrials.gov record indicates that results have been posted and that six outcome measures are posted.
5. Primary Endpoint
| Endpoint | Registered definition | Time frame | Endpoint type |
|---|---|---|---|
| Disease-free Survival | Percentage of patients free from DFS events. DFS events include local, regional, or distant recurrence, second primary cancer or death from any cause prior to recurrence or second primary cancer. | 7 years. | Time-to-event |
This endpoint combines several clinically important events into a single time-to-event outcome. A patient remains disease-free under the registry definition until experiencing one of the specified DFS events. The endpoint therefore incorporates not only recurrence but also second primary cancer and death from any cause prior to recurrence or second primary cancer.
The registry defines the event components and the 7-year assessment frame. The ClinicalTrials.gov record does not provide the numerical event counts, Kaplan-Meier estimates, hazard ratio, confidence interval, or p-value for the DFS comparison.
6. Results Availability
ClinicalTrials.gov indicates that results have been posted for the study and that six outcome measures are posted. However, the ClinicalTrials.gov record contains no formal statistical analyses section. Consequently, the available record does not provide a treatment-effect estimate, confidence interval, or p-value for the registered primary endpoint.
| Results element | Information reported in the registry extract |
|---|---|
| Results posted | Yes |
| Outcome measures posted | 6 |
| Primary endpoint results | Disease-free survival is reported as having results posted |
| Formal statistical analyses | No statistical analyses reported in the extracted data |
| Hazard ratio | Not reported |
| Confidence interval | Not reported |
| P-value | Not reported |
7. Planned Analysis
The registered primary endpoint is a time-to-event measure. For an endpoint of this type, a standard statistical analysis would generally estimate the time-to-event distribution in each randomized group and compare the groups using methods designed for censored survival data.
Kaplan-Meier estimation
A Kaplan-Meier estimator can describe the proportion of patients remaining free of a DFS event over time while accounting for patients whose event status is not observed through the full follow-up period.
Log-rank comparison
A log-rank test is commonly used to compare time-to-event distributions between randomized groups when the objective is an overall comparison across follow-up.
Cox proportional-hazards model
A Cox model can express the relative event rate between treatment groups through a hazard ratio, together with a confidence interval describing statistical uncertainty.
Censoring
Patients who have not experienced a DFS event at the end of their observable follow-up contribute information up to their censoring time rather than being treated as if an event had occurred.
8. Statistical Methodology
Randomization
Randomization is the fundamental design feature supporting the treatment comparison. In a randomized parallel trial, treatment assignment is determined by the allocation mechanism rather than by the investigator's selection of treatment for an individual participant. This helps balance both measured and unmeasured prognostic factors in expectation across the randomized groups.
Randomization does not guarantee identical groups. Its statistical value comes from making systematic differences in treatment assignment less likely and providing a principled basis for causal comparison.
Double masking
The registry identifies the trial as double-masked. Masking can reduce the influence of treatment knowledge on participant behavior, investigator decisions, outcome assessment, and other aspects of trial conduct. Its importance depends on how the endpoint is measured and how much subjective judgment enters the assessment process.
Parallel-group design
The registry identifies the design model as parallel. Each participant belongs to one randomized treatment group rather than sequentially receiving both trial interventions. The statistical comparison therefore centers on differences between the randomized groups over follow-up.
Time-to-event analysis
Disease-free survival is explicitly identified as a time-to-event endpoint. This matters because ordinary comparisons of percentages at a single time point do not fully use the information generated by different follow-up times and censoring patterns. Survival methods are designed to incorporate the timing of events as well as incomplete follow-up.
Here, di represents the number of events at time ti and ni represents the number of patients at risk immediately before that time. This is the general survival-analysis framework, not a reconstruction of unreported B-42 data.
9. Statistical Methods Explained
Why is disease-free survival a time-to-event endpoint?
The registry does not define DFS merely as whether a patient is disease-free at a single visit. It defines an event-based outcome involving local, regional, or distant recurrence, second primary cancer, or death from any cause prior to recurrence or second primary cancer. The timing of the first qualifying event therefore matters, which makes time-to-event methods appropriate.
Why use Kaplan-Meier estimation?
Clinical trials often have unequal follow-up because participants can enter at different times or stop being observed before an event occurs. Kaplan-Meier estimation allows those patients to contribute information up to the time they are censored. It therefore provides a time-dependent estimate rather than a simple proportion based only on patients observed for the entire study.
What would a hazard ratio mean for this trial?
If a Cox model were used for DFS, its hazard ratio would compare the estimated instantaneous rate of a qualifying DFS event between the letrozole and placebo groups. A hazard ratio below 1 would indicate a lower estimated event hazard in the letrozole group under that model. It would not mean that the same percentage of every individual patient's risk was reduced.
Why is randomization important for the DFS comparison?
Randomization creates the treatment groups through an allocation mechanism rather than allowing treatment choice to depend on prognosis. This reduces the risk that observed differences in DFS are simply consequences of systematic baseline differences in who received each treatment.
Why does double masking matter?
Knowing treatment assignment can influence behavior and assessment. Double masking is intended to reduce these sources of bias. Its value is especially relevant when aspects of trial conduct or outcome assessment could otherwise be affected by knowledge of treatment.
What does a 7-year time frame mean?
The registry specifies a 7-year time frame for the disease-free survival endpoint. This identifies the period associated with the registered outcome measure. It should not be interpreted as saying that every participant necessarily has exactly 7 years of observed follow-up.
Why can the registry say results are posted without supplying a hazard ratio here?
ClinicalTrials.gov results can contain outcome measures without the extracted record necessarily providing a formal statistical-analysis entry. In the ClinicalTrials.gov record, results are marked as posted and six outcome measures are reported, but the statistical-analysis array is empty. The appropriate response is to report that distinction rather than infer a treatment effect.
10. Analysis Populations and What Is Known
The ClinicalTrials.gov record establishes an enrollment of 3966 participants and two randomized treatment arms. They do not provide the numerical composition of specific analysis populations, exclusions, protocol deviations, censoring counts, or missing-data handling.
| Analysis component | What the ClinicalTrials.gov record establishes |
|---|---|
| Randomized trial | Yes |
| Parallel treatment groups | Yes |
| Total enrollment | 3966 |
| Number of arms | 2 |
| Primary endpoint type | Time-to-event |
| Formal efficacy analysis population | Not reported |
| Censoring details | Not reported |
| Missing-data/imputation approach | Not reported |
| Statistical stratification factors | Not reported |
| Bayesian methods | Not reported |
11. Safety Results
The ClinicalTrials.gov record includes serious adverse events by randomized arm. These figures describe the number affected and the number at risk for each arm.
| Arm | Serious adverse events | Affected / at risk |
|---|---|---|
| Placebo | Serious adverse events | 87 / 1938 |
| Letrozole | Serious adverse events | 80 / 1951 |
The serious-adverse-event figures should be interpreted as a safety comparison rather than as an efficacy endpoint. The ClinicalTrials.gov record does not provide a formal statistical comparison, confidence interval, p-value, event definition beyond the serious-adverse-event category, or additional safety-event categories.
The reported counts show the number of affected participants and the corresponding numbers at risk in each arm. They do not by themselves establish that one treatment has a higher or lower safety risk, because a formal comparison would depend on the prespecified safety analysis, event definitions, exposure, follow-up, and appropriate denominators.
12. Serious Adverse Events: Statistical Perspective
Safety data illustrate why numerator and denominator information should be reported together. The values posted on ClinicalTrials.gov for NSABP B-42 are expressed as affected / at risk, allowing the reader to see both the event count and the population from which that count arose.
Numerator
The numerator is the number of participants affected by serious adverse events: 87 in the placebo arm and 80 in the letrozole arm.
Denominator
The corresponding numbers at risk are 1938 for placebo and 1951 for letrozole.
Why counts alone are insufficient
Comparing raw event counts without considering the populations at risk can be misleading when the denominators differ.
What is not reported here
The ClinicalTrials.gov record does not provide a formal statistical test or confidence interval for the serious-adverse-event comparison.
13. Understanding the Primary Endpoint Statistically
Disease-free survival is a composite time-to-event endpoint. The registry definition includes several event pathways: local recurrence, regional recurrence, distant recurrence, second primary cancer, and death from any cause prior to recurrence or second primary cancer.
| DFS component | Role in the registered endpoint |
|---|---|
| Local recurrence | Qualifying DFS event |
| Regional recurrence | Qualifying DFS event |
| Distant recurrence | Qualifying DFS event |
| Second primary cancer | Qualifying DFS event |
| Death from any cause prior to recurrence or second primary cancer | Qualifying DFS event |
Because these events can occur at different times, a time-to-event framework preserves information about when the first qualifying event occurs. The composite nature of the endpoint also means that a single overall DFS result would not, by itself, identify which component drove any observed difference.
A DFS analysis is therefore not simply a comparison of the percentage of patients who experienced recurrence. The registered endpoint includes multiple event types and uses their timing. If a formal treatment-effect estimate were available, its interpretation would depend on the complete event definition, censoring rules, analysis population, and statistical model.
14. Hazard Ratios, Confidence Intervals, and P-values
No hazard ratio, confidence interval, or p-value for the primary DFS endpoint is reported in the ClinicalTrials.gov record. These quantities are therefore explained here as statistical concepts rather than reported B-42 results.
Hazard ratio
A hazard ratio is a relative comparison of event hazards under a specified time-to-event model. It is not an absolute survival probability and is not a percentage of patients who benefit.
Confidence interval
A confidence interval communicates uncertainty around an estimated treatment effect under the statistical model and sampling framework used for the analysis.
P-value
A p-value addresses evidence against a specified null hypothesis under the statistical model. It does not measure the size or clinical importance of an effect.
Absolute estimates
Time-specific DFS estimates can provide an absolute perspective that a relative hazard ratio cannot provide by itself.
This is a general interpretation of a hazard ratio. Because the B-42 hazard ratio is not reported in the ClinicalTrials.gov record, no numerical treatment effect is asserted here.
15. Censoring and the Proportional-Hazards Assumption
Two concepts are particularly important when interpreting a time-to-event endpoint such as DFS: censoring and the assumptions underlying a proportional-hazards model.
Censoring
A participant who has not experienced a qualifying DFS event by the time follow-up ends may be censored. Censoring allows the participant's observed disease-free time to contribute to the analysis without treating the participant as though an event had occurred after the last available observation.
Proportional hazards
If a Cox proportional-hazards model is used, the hazard ratio summarizes a relative hazard under the model's proportional-hazards framework. If relative hazards change materially over time, a single hazard ratio can become less descriptive of the full longitudinal treatment effect.
16. Multiplicity, Interim Analysis, and Other Design Topics
The ClinicalTrials.gov record does not report a multiplicity strategy, interim-analysis plan, alpha-spending method, non-inferiority margin, crossover procedure, Bayesian analysis, or formal missing-data/imputation strategy.
| Design topic | Evidence in the ClinicalTrials.gov record |
|---|---|
| Non-inferiority margin | Not reported |
| Crossover | Not reported |
| Factorial design | Not reported; the registry identifies a parallel design |
| Multiplicity adjustment | Not reported |
| Interim analysis | Not reported |
| Missing-data/imputation strategy | Not reported |
| Stratification factors | Not reported |
| Bayesian methods | Not reported |
These omissions are important because they prevent reconstruction of the full confirmatory statistical framework. For example, knowing that a primary endpoint is time-to-event does not establish whether the trial used a stratified log-rank test, a Cox model, a particular alpha allocation, or another prespecified procedure.
17. Blinding and Randomization as Statistical Controls
The trial's statistical concepts identifies Blinding and Randomization as the key trial concepts. Both are directly connected to the interpretability of the treatment comparison.
Randomization
Randomization determines treatment assignment through the trial's allocation mechanism. Its purpose is to make the randomized groups comparable in expectation and support an unbiased treatment comparison.
Blinding
Double masking reduces the opportunity for knowledge of treatment assignment to influence participant behavior, investigator decisions, or other aspects of trial conduct.
These controls do not eliminate every possible source of bias. Their statistical value is that they address important pathways through which treatment assignment could otherwise become associated with factors that also influence observed outcomes.
18. What Can and Cannot Be Concluded From the registry Results Record
| Question | Supported by the ClinicalTrials.gov record? |
|---|---|
| Was the trial randomized? | Yes. |
| Was it double-masked? | Yes. |
| Was the design parallel? | Yes. |
| Were there two treatment arms? | Yes. |
| Was disease-free survival a registered primary endpoint? | Yes. |
| Was the primary endpoint defined as a time-to-event measure? | Yes. |
| Were results posted? | Yes. |
| Were six outcome measures posted? | Yes. |
| What was the DFS hazard ratio? | Not reported. |
| What was the DFS confidence interval? | Not reported. |
| What was the DFS p-value? | Not reported. |
| Which formal statistical model was used? | Not reported. |
| Was a particular multiplicity strategy used? | Not reported. |
The most important methodological distinction is between results availability and results detail. The registry extract says results are posted, but the registry-reported statistical-analysis section contains no formal comparisons. A rigorous statistical page should preserve that limitation rather than filling the gap with results obtained from memory or an unrelated source.
19. Limitations
- Incomplete statistical-analysis detail: the ClinicalTrials.gov record contains no formal statistical-analysis entries for the primary endpoint.
- No numerical efficacy estimates reported: the data do not provide a DFS hazard ratio, confidence interval, p-value, event count, or time-specific DFS estimate.
- Unreported analysis populations: the ClinicalTrials.gov record does not identify the precise populations used for efficacy or safety analyses beyond the overall enrollment and reported serious-adverse-event denominators.
- Unreported censoring rules: the registry extract does not specify the censoring conventions used for DFS.
- Unreported multiplicity: the ClinicalTrials.gov record does not establish how the primary endpoint and other posted outcomes were incorporated into a type I error-control strategy.
- Unreported interim monitoring: no interim-analysis or alpha-spending method is reported.
- Unreported stratification: the statistical stratification factors, if any, are not provided in the ClinicalTrials.gov record.
- Safety interpretation: serious-adverse-event counts are reported, but no formal comparative safety analysis is provided.
- Registry status: the ClinicalTrials.gov record lists the study status as UNKNOWN, even though results are marked as posted.
20. Why This Trial Matters Statistically
NSABP B-42 provides a useful teaching example because its registered primary endpoint is a time-to-event outcome within a randomized, double-masked, parallel phase 3 design. The trial therefore connects several foundational concepts in clinical-trial statistics.
| Concept | How it appears in NSABP B-42 |
|---|---|
| Randomization | The registry identifies the allocation as randomized. |
| Blinding | The registry identifies the study as double-masked. |
| Parallel design | The registry identifies a parallel design with two arms. |
| Time-to-event analysis | Disease-free survival is the registered primary endpoint type. |
| Composite endpoint | DFS includes local, regional, or distant recurrence, second primary cancer, and specified death events. |
| Censoring | Relevant to interpretation of any time-to-event analysis, although the registry extract does not provide the actual censoring rules. |
| Hazard ratio | A standard possible summary measure for a time-to-event treatment comparison, but no B-42 hazard ratio is reported here. |
| Confidence interval | A standard way to quantify uncertainty around an estimated treatment effect, but no B-42 confidence interval is reported here. |
| Safety comparison | Serious adverse events are reported as affected participants over participants at risk for each arm. |
The trial is also a useful reminder that a statistical analysis page should distinguish what the registry establishes from what statistical theory suggests would ordinarily be done. The first is evidence about this specific trial; the second is methodological education.
21. Related Tutorials
Learn more about the methods used in this trial:
22. Related Calculators
The ClinicalTrials.gov record identifies disease-free survival as a time-to-event endpoint and randomization as a trial-design concept. The calculator links therefore point to the general statistical-calculator hub rather than asserting a specific B-42 calculation that is not documented in the ClinicalTrials.gov record.
23. Sources
- ClinicalTrials.gov: NCT00382070.
- PubMed: PMID 39047219.
- PubMed: PMID 30509771.
Continue through Clinical Biostats
Explore broader statistical tutorials and calculators for clinical-trial design, survival analysis, randomization, and related methods.
24. Record Summary
NSABP B-42 is a randomized, double-masked, parallel phase 3 trial with an enrollment of 3966 and two intervention arms: letrozole and placebo. Its registered primary endpoint is disease-free survival over a 7-year time frame, with DFS defined as the percentage of patients free from events including local, regional, or distant recurrence, second primary cancer, or death from any cause prior to recurrence or second primary cancer.
The ClinicalTrials.gov record states that results have been posted and that six outcome measures are posted, but it contains no formal statistical-analysis entries. As a result, this page deliberately does not supply a hazard ratio, confidence interval, p-value, event count, or other efficacy estimate that is absent from the ClinicalTrials.gov record. Instead, it explains how a randomized time-to-event endpoint such as DFS is ordinarily analyzed and how randomization, double masking, censoring, and model-based effect measures affect interpretation.