This page separates registry-reported information from statistical interpretation. The numerical trial information on this page is restricted to the ClinicalTrials.gov record. The registry provides the official trial record: NCT01275677.
1. Trial at a Glance
NSABP B-47 was a randomized, phase 3, parallel-group trial evaluating chemotherapy with or without trastuzumab after surgery in women with invasive breast cancer. The registry reports one primary time-to-event endpoint: the percentage of patients alive and free from invasive disease at 5 years.
| Feature | NSABP B-47 |
|---|---|
| Trial name | NSABP B-47 |
| NCT identifier | NCT01275677 |
| Brief title | Chemotherapy With or Without Trastuzumab After Surgery in Treating Women With Invasive Breast Cancer |
| Phase | Phase 3 |
| Status | COMPLETED |
| Allocation | RANDOMIZED |
| Design model | PARALLEL |
| Masking | NONE |
| Primary purpose | TREATMENT |
| Enrollment | 3270 |
| Arms | 2 |
| Lead sponsor | National Cancer Institute (NCI) |
| Sponsor type | NIH |
| Start | 2011-02-08 |
| Primary completion | 2017-07-31 |
2. Clinical Question
The trial's central clinical question can be stated statistically as whether adding trastuzumab to chemotherapy after surgery changes the time-to-event outcome defined as invasive disease-free survival in the enrolled population.
Population
Women with invasive breast cancer represented by the registry's listed conditions, including HER2/Neu Positive, Progesterone Receptor Positive, recurrent breast carcinoma, and specified stage IA, IB, IIA, IIB, IIIA, and IIIC breast cancer categories.
Intervention
Chemotherapy with trastuzumab as the biologic intervention evaluated in the trial.
Comparator
Chemotherapy without trastuzumab.
Primary question
How does the randomized treatment assignment relate to the percentage of patients alive and free from invasive disease at 5 years?
3. Trial Design
Chemotherapy
- Cyclophosphamide
- Docetaxel
- Doxorubicin
- Doxorubicin hydrochloride
- Paclitaxel
- Quality-of-life assessment
- Laboratory biomarker analysis
Chemotherapy, Trastuzumab
- Cyclophosphamide
- Docetaxel
- Doxorubicin
- Doxorubicin hydrochloride
- Paclitaxel
- Trastuzumab
- Quality-of-life assessment
- Laboratory biomarker analysis
The registry describes the allocation as randomized, the design model as parallel, and masking as none. These features identify the trial as an open-label randomized comparison rather than a masked treatment comparison.
4. Trial Timeline
Trial start
The registry lists 2011-02-08 as the study start date.
Primary completion
The registry lists 2017-07-31 as the primary completion date.
Registry status
The trial is listed as completed.
5. Primary Endpoint
| Endpoint | Registry definition | Time frame | Endpoint type |
|---|---|---|---|
| Percentage of Patients Alive and Free From Invasive Disease (IDFS) | Percentage of patients free from an invasive disease-free survival event where events include any invasive recurrence, contralateral invasive breast cancer, second non-breast primary cancer (excluding squamous or basal cell carcinoma of the skin), or death from any cause. | 5 years | Time-to-event |
This definition is important because IDFS is a composite time-to-event endpoint. A patient can experience an event through several distinct pathways: invasive recurrence, contralateral invasive breast cancer, a qualifying second non-breast primary cancer, or death from any cause. Consequently, the endpoint is not simply a measure of recurrence.
For a time-to-event analysis, the clinically relevant quantity is not merely whether an event eventually occurs, but the time from the relevant starting point to the first qualifying event, subject to the registry's prespecified analysis rules.
6. Statistical Methodology
The ClinicalTrials.gov record classifies the primary endpoint as time-to-event and state that results have been posted. However, the ClinicalTrials.gov record contains no formal statistical analyses. Therefore, this page does not attribute a particular hazard ratio, confidence interval, test statistic, or p-value to NSABP B-47.
How a time-to-event endpoint is typically analyzed
A time-to-event endpoint such as IDFS is commonly summarized using Kaplan-Meier estimation. This method accounts for patients whose event status is not observed through the end of follow-up by treating them as censored according to the applicable censoring rules.
Here, di is the number of events at event time ti, while ni is the number of individuals at risk immediately before that time.
Potential comparative measures
For a randomized comparison of two time-to-event groups, investigators commonly report an absolute survival estimate at a clinically relevant time point, a hazard ratio from a Cox proportional-hazards model, and a corresponding confidence interval. A log-rank test is also commonly used for comparing survival distributions.
7. Planned Analysis
Because the ClinicalTrials.gov record contains no formal statistical analyses, there is no numerical treatment-effect result to reproduce here. The registry nevertheless identifies the primary endpoint and its time frame, so the appropriate statistical framework can be described without implying an outcome.
What is measured?
The percentage of patients alive and free from an invasive disease-free survival event at 5 years.
Why time-to-event analysis?
Patients can experience the first qualifying event at different times, while others may remain event-free for the available observation period.
Typical estimator
Kaplan-Meier estimation is commonly used to estimate the event-free survival function over time.
Typical treatment comparison
A randomized comparison of time-to-event distributions can commonly use a log-rank test and a Cox model for a relative hazard estimate, when appropriate.
ClinicalTrials.gov reports that results have been posted, but the ClinicalTrials.gov record does not provide a formal statistical comparison for the primary endpoint. Accordingly, no result is inferred from the trial design, sample size, or adverse-event counts.
8. Statistical Methods Explained
What does randomization contribute statistically?
Randomization assigns participants to treatment groups using a chance-based allocation process. Its central statistical role is to reduce systematic differences in treatment assignment, allowing subsequent differences in outcomes to be interpreted in the context of the randomized comparison. Randomization does not guarantee identical groups in every baseline characteristic.
Why is IDFS a time-to-event endpoint rather than just a percentage?
The registry defines IDFS as a time-to-event endpoint with a 5-year time frame. Time-to-event methods retain information about when events occur rather than reducing the entire follow-up experience to a simple event proportion. They can also account for participants whose complete event history is not observed.
What does Kaplan-Meier estimation provide?
Kaplan-Meier estimation describes the estimated probability of remaining free of the event through successive follow-up times. It is especially useful when participants enter the analysis with different amounts of observable follow-up or when some participants are censored before an event occurs.
What would a hazard ratio mean if one were reported?
A hazard ratio would describe the relative instantaneous event rate between the randomized groups under a specified survival model. A value below 1 would indicate a lower estimated event hazard in the numerator group, while a value above 1 would indicate a higher estimated event hazard. It would not directly equal a difference in 5-year IDFS percentages.
Why does the confidence interval matter?
A confidence interval communicates statistical uncertainty around an estimated treatment effect. A narrow interval indicates greater precision than a wide interval under the same statistical framework. It does not describe the range of effects experienced by individual patients.
Why can censoring affect interpretation?
Time-to-event methods rely on censoring assumptions. A censored participant contributes information up to the censoring point, but the validity of the resulting estimate depends on how censoring relates to the event process and on the analysis rules used. The ClinicalTrials.gov record does not specify the censoring or imputation procedures for B-47.
Why should a p-value not be treated as an effect size?
A p-value addresses evidence against a specified null hypothesis under a statistical model. It does not tell the reader how large or clinically important an observed treatment difference is. Effect estimates, absolute event-free survival estimates, and confidence intervals are needed to describe magnitude and precision.
9. Results Reporting Status
ClinicalTrials.gov indicates that results have been posted for NSABP B-47 and that 9 outcome measures are posted. The ClinicalTrials.gov record, however, do not include the numerical statistical analyses associated with those outcomes.
| Registry information reported | Status |
|---|---|
| Results posted | Yes |
| Outcome measures posted | 9 |
| Primary endpoints registered | 1 |
| Primary endpoint type | Time-to-event |
| Statistical analyses reported | None |
| Formal numerical treatment-effect estimates reported | None |
10. Safety Results
The ClinicalTrials.gov record provides serious adverse-event counts by randomized arm. These counts are reported as affected participants over participants at risk.
| Arm | Serious adverse events | At risk |
|---|---|---|
| Arm I — Chemotherapy | 31 | 1615 |
| Arm II — Chemotherapy, Trastuzumab | 96 | 1625 |
The raw serious-adverse-event counts differ between the two arms, with 31 affected participants reported in the chemotherapy arm and 96 in the chemotherapy-plus-trastuzumab arm. The denominators also differ: 1615 participants were listed as at risk in Arm I and 1625 in Arm II.
11. What the Primary Endpoint Means
The 5-year IDFS endpoint combines several clinically distinct events into one time-to-event outcome. A patient is counted as having an IDFS event if the patient experiences any invasive recurrence, contralateral invasive breast cancer, a qualifying second non-breast primary cancer, or death from any cause.
This means that a difference in IDFS cannot automatically be interpreted as a difference in recurrence alone. The composite definition determines which first events contribute to the endpoint.
A 5-year IDFS percentage is not the same as overall survival, recurrence-free survival under a narrower event definition, or the probability of avoiding every possible cancer-related event. Its meaning is tied specifically to the registry's composite event definition.
The registry specifies a 5-year time frame. A time-specific estimate answers the question of how many patients remain alive and free from a qualifying IDFS event through that point under the prespecified analysis framework. It does not describe what happens after 5 years.
12. Composite Endpoints and Statistical Interpretation
Composite endpoints can increase the number of events available for statistical analysis, but they also require careful interpretation. The individual components may differ substantially in clinical frequency, severity, and relationship to treatment.
| IDFS event component | Included in registry definition? | Statistical implication |
|---|---|---|
| Invasive recurrence | Yes | Contributes an event to the composite endpoint. |
| Contralateral invasive breast cancer | Yes | Contributes an event even though it differs biologically from recurrence of the original disease. |
| Second non-breast primary cancer | Yes | Included subject to the registry's exclusion of squamous or basal cell carcinoma of the skin. |
| Death from any cause | Yes | Counts as an event regardless of the cause of death. |
Because the endpoint is defined as the first qualifying event, an individual who experiences one component generally cannot subsequently contribute a later component as the first IDFS event. This is one reason that the component composition of a composite endpoint is important when interpreting an overall treatment effect.
13. Statistical Interpretation of the Serious Adverse-Event Counts
The ClinicalTrials.gov record reports 31 affected participants among 1615 at risk in the chemotherapy arm and 96 among 1625 at risk in the chemotherapy-plus-trastuzumab arm.
These figures are descriptive counts. They do not establish a causal treatment effect or provide a statistical test because the ClinicalTrials.gov record does not include an analysis of the serious-adverse-event endpoint.
The denominators should also be retained when interpreting the counts because the arm sizes are not identical.
14. Randomization as the Core Statistical Design Feature
Randomization is the principal design feature identified in the registry. With a randomized parallel-group design, participants are assigned to one of the two treatment strategies before the comparative outcome is assessed.
The registry identifies the allocation as RANDOMIZED.
The registry identifies the design model as PARALLEL.
The registry identifies the masking as NONE.
The registry identifies the primary purpose as TREATMENT.
Randomization supports a clean comparison of treatment assignment, but statistical analysis still depends on the prespecified endpoint definition, follow-up rules, censoring, analysis population, and treatment of missing information. The ClinicalTrials.gov record does not document those additional analytical specifications.
15. Missing Data, Censoring, and Analysis Population
These issues are particularly important for a 5-year time-to-event endpoint, but the ClinicalTrials.gov record does not specify the statistical handling of missing observations, censoring rules, or analysis populations for the primary endpoint.
Censoring
The registry classifies IDFS as time-to-event, but the ClinicalTrials.gov record does not state the precise censoring rules.
Missing data
No imputation method is specified in the ClinicalTrials.gov record.
Analysis population
The ClinicalTrials.gov record does not identify whether the primary analysis used an intention-to-treat, modified intention-to-treat, or another formally defined population.
Follow-up
The primary endpoint is evaluated at 5 years, but the ClinicalTrials.gov record does not provide the distribution of follow-up times or the number censored.
16. Multiplicity and Other Design Topics
The ClinicalTrials.gov record does not provide a non-inferiority margin, crossover specification, factorial structure, interim-analysis plan, multiplicity strategy, stratification factors, or Bayesian analysis. These topics therefore cannot be attributed to NSABP B-47 from the ClinicalTrials.gov record.
| Design topic | Supported by the ClinicalTrials.gov record? | Interpretation |
|---|---|---|
| Randomization | Yes | Registry identifies the allocation as RANDOMIZED. |
| Parallel design | Yes | Registry identifies the design model as PARALLEL. |
| Non-inferiority margin | Not reported | No margin should be inferred. |
| Crossover | Not reported | No crossover claim should be made. |
| Factorial design | Not reported | The registry identifies a parallel design, but no factorial structure is reported. |
| Interim analysis | Not reported | No interim-analysis method is reported in the ClinicalTrials.gov record. |
| Multiplicity strategy | Not reported | No alpha-allocation or multiplicity procedure is reported. |
| Stratification | Not reported | No stratification factors are reported. |
| Bayesian methods | Not reported | No Bayesian analysis is reported in the ClinicalTrials.gov record. |
17. Limitations
- Incomplete statistical-analysis detail: the registry record contains no formal statistical analyses, even though results are posted.
- No numerical efficacy estimates reported: the available data do not provide an IDFS estimate, treatment-effect estimate, confidence interval, or p-value.
- No formal primary-endpoint comparison reported: a treatment effect cannot be inferred from the trial design or enrollment alone.
- Composite endpoint: IDFS includes several event types, so an overall composite result would require attention to the contribution of each component.
- Unknown censoring rules: the ClinicalTrials.gov record does not specify how censoring was handled for the 5-year time-to-event analysis.
- Unknown analysis population: the ClinicalTrials.gov record does not identify the formal population used for the primary efficacy analysis.
- Safety interpretation: serious-adverse-event counts are reported, but no formal statistical comparison or exposure-adjusted analysis is provided.
- No subgroup results reported: the available data do not provide subgroup-specific treatment effects or interaction analyses.
18. Why This Trial Matters Statistically
NSABP B-47 is useful as a statistical teaching case because its registry record combines a randomized parallel-group design with a clinically meaningful composite time-to-event endpoint and publicly posted results that, in the ClinicalTrials.gov record, are not accompanied by formal statistical-analysis entries.
| Concept | How it appears in NSABP B-47 |
|---|---|
| Randomization | The registry identifies the study as randomized. |
| Parallel-group design | The registry identifies the design model as parallel. |
| Time-to-event analysis | The primary endpoint is explicitly classified as time-to-event. |
| Composite endpoint | IDFS combines invasive recurrence, contralateral invasive breast cancer, qualifying second primary cancer, and death. |
| Fixed time horizon | The primary endpoint is reported at 5 years. |
| Censoring | A relevant methodological issue for any time-to-event analysis, although the ClinicalTrials.gov record does not state the exact censoring rules. |
| Effect estimation | A formal treatment-effect estimate would normally accompany a comparative time-to-event analysis, but none is in the posted results. |
| Safety comparison | Serious adverse-event counts are provided for both arms. |
19. Related Tutorials
Learn more about the method directly identified in this trial's learning pathway:
20. Related Statistical Calculators
21. Sources
- ClinicalTrials.gov: NSABP B-47, NCT01275677.
- PubMed: PubMed record, PMID 31821109.
Continue with Clinical Biostats
Explore statistical tutorials, calculators, and clinical-trial analyses covering the methods used to design, analyze, and interpret randomized studies.
22. Record Summary
NSABP B-47 is a completed randomized phase 3, parallel-group trial with 3270 enrolled participants and two arms. Its registered primary endpoint is the percentage of patients alive and free from invasive disease at 5 years, classified as a time-to-event outcome and defined using a composite of invasive recurrence, contralateral invasive breast cancer, qualifying second non-breast primary cancer, or death from any cause.
The ClinicalTrials.gov record also report serious adverse-event counts of 31 among 1615 participants at risk in the chemotherapy arm and 96 among 1625 participants at risk in the chemotherapy-plus-trastuzumab arm. ClinicalTrials.gov indicates that results have been posted and that 9 outcome measures are posted, but the registry record contains no formal statistical-analysis entries.