An Open Label Phase II Trial of BIBW 2992 in Patients With HER2-negative Metastatic Breast Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for An Open Label Phase II Trial of BIBW 2992 in Patients With HER2-negative Metastatic Breast Cancer (NCT00425854), a non-randomized phase 2 study. The primary outcome is Objective Response (OR).

2. Clinical Question

3. Trial Design

This was an open-label phase II study with non-randomized allocation and a parallel model. The intervention was BIBW 2992, described as a high dose once daily; the official title specifies an oral 50 mg once-daily dose. The record describes two cohorts of patients with HER2-negative metastatic breast cancer. It gives enrollment as 50, but the registry information does not clarify how enrollment was distributed between the cohorts or whether all enrolled participants received treatment.

The primary endpoint was objective response, and the record specifies that it applied only to Cohort B. The outcome population was the treated set, defined as patients who received at least one dose of trial medication. Cohort A has no objective-response data in the record because response was not its primary endpoint. This endpoint-specific distinction matters: the fact that both cohorts appear in the trial record does not mean both contribute to the reported primary-endpoint analysis.

Tumour assessments were scheduled at screening, week 8, week 16, week 24, and every 8 weeks thereafter. The registry also lists clinical benefit, time to and duration of objective response, progression-free survival, overall survival, cardiac left ventricular ejection fraction changes, ECOG performance-status change, and a predose plasma concentration measure for afatinib. The record lists safety outcomes as well. It does not give a detailed statistical analysis plan, the assessment rules for these other outcomes, or a description of how missing assessments were handled.

The study was completed, with a start date of 2006-11 and primary completion in 2009-05. Because allocation was non-randomized and masking was absent, the design does not create a blinded, randomized comparison between treatments. The registry record does not report a concurrent control treatment. Consequently, the design is best read as a study of outcomes observed in the treated cohorts rather than as a controlled estimate of a treatment contrast.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Treated Set (TS). TS consisted of all patients who received at least one dose of trial medication. No data for Cohort A as OR was primary endpoint only for Cohort B.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Objective Response (OR), measured over Tumour assessments were performed at screening, week 8, week 16, week 24, and every 8 weeks thereafter..

Objective response (OR) is the primary endpoint named for Cohort B. The record defines it as complete response (CR) or partial response (PR), assessed according to RECIST version 1.0. The endpoint therefore classifies participants according to whether tumour assessment met either of those response categories. It is recorded as a number of participants with OR, rather than as a continuous measure of tumour size.

Response endpoints are used in oncology studies because they offer a defined way to describe tumour shrinkage under prespecified assessment criteria. In general, a binary response outcome makes the number meeting the response definition visible, while interpretation also depends on how many participants were assessed and on uncertainty around the resulting estimate. The response assessment schedule in this trial began at screening and continued at week 8, week 16, week 24, and every 8 weeks thereafter.

The registry record does not report the Cohort B response count or an uncertainty estimate for it. It also states that Cohort A has no data for this endpoint because OR was primary only for Cohort B. Thus, the endpoint definition and assessment schedule are available, but the primary endpoint finding itself is not reported in the registry record. Without the result, the response rate and its precision cannot be described from this record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record gives some design details but does not report a full statistical analysis plan. The points below explain how to interpret the stated endpoint and the kinds of statistical information that would ordinarily matter. General statistical explanations are not claims about methods actually used in this trial.

What does the binary endpoint represent?

Objective response is recorded as participants with an event: in this case, participants meeting the definition of complete or partial response. This is a binary outcome because each assessed participant is classified as having met or not met the response definition. A count alone is not a rate: to interpret the frequency of response, the count must be considered in relation to an appropriate analyzed population or denominator. The registry identifies the treated set as the population for the primary endpoint, but does not report the Cohort B response count or an estimate based on it.

In general, a response proportion is a descriptive estimate for the population analyzed. A confidence interval communicates uncertainty around that estimate: it gives a range of values compatible with the data and the interval method under its assumptions. An interval would not remove design limitations or establish a causal effect. No response estimate or confidence interval is reported in the registry record, so precision cannot be assessed.

Why do allocation and masking matter?

Random allocation is generally used to make treatment groups comparable, on average, with respect to both measured and unmeasured characteristics. This trial is described as non-randomized, and the record does not identify a randomized control treatment. Differences between cohorts could reflect differences in who entered each cohort or other factors, rather than a treatment contrast. The cohort labels should therefore be read as study groups, not as randomized arms in a comparative efficacy trial.

Masking means concealing treatment assignment from participants, clinicians, or assessors, depending on the design. The record says masking was absent. In general, knowledge of treatment can affect subjective assessments or decisions about evaluation, although RECIST supplies defined response criteria. The record does not explain whether image assessments were independently reviewed or otherwise protected from assessment bias, so those details are not reported in the registry record.

What does the assessment schedule add?

The trial planned tumour assessments at screening, week 8, week 16, week 24, and every 8 weeks thereafter. Scheduled assessments define when response could be evaluated, but they do not by themselves describe how missing visits, unevaluable scans, or confirmation of a response were handled. Those rules are not reported in the registry record. In general, the timing and handling of assessments can influence which response is recorded and when it is observed.

How should planning illustrations be read?

The registry context describes sample-size planning illustrations computed from an observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The stated effect to detect is a fraction of the observed effect. For time-to-event outcomes, the illustrations refer to a total event count; for other outcomes, they refer to participants per arm. These are planning illustrations, not observed power and not evidence that the trial was designed using those assumptions.

In general, a planning calculation estimates the sample size or event information needed under chosen assumptions. Its result depends on the assumed effect, significance level, power target, allocation, and statistical model. A calculation based on an observed effect is especially dependent on that effect estimate and is not a substitute for a prospective design justification. The registry record does not report a trial-specific sample-size justification, a target power, or evidence that these planning assumptions were used when the study was designed.

What statistical details remain unknown?

The record does not report the test or interval method for objective response, any adjusted analysis, or a model for the time-to-event outcomes. It also does not report interim analyses, a missing-data strategy, or a multiplicity procedure. In general, testing multiple outcomes can increase the chance that at least one result appears statistically notable by chance; adjustment methods can address that issue but may alter the interpretation of individual tests. Whether any such method was used in this trial is not reported in the registry record.

Over-dispersion is generally relevant when count data vary more than a chosen model expects; allowing for it can prevent uncertainty from being understated. The primary endpoint is a binary participant count, and the record gives no dispersion parameter or model specification. It would not be appropriate to infer an over-dispersion adjustment from the endpoint label. More broadly, statistical methods should not be imputed from common practice when the registry does not describe them.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Cohort B")
n      <- c(21)
events <- c(0)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Cohort B  21  0  0.0  0.0  16.1  -0.0  15.5  0.1  15.4

11. Multiplicity, Missing Data and Analysis Populations

The registry lists objective response as primary only for Cohort B and identifies additional outcomes, including clinical benefit, time to and duration of response, progression-free survival, overall survival, cardiac measures, ECOG performance status, and plasma concentration. It also reports ten outcome measures and two analyses. The record does not explain how the analyses correspond to particular outcomes, nor does it report a hierarchy or a multiplicity adjustment.

In general, evaluating multiple outcomes can increase the chance of a chance-positive result if each is tested separately. A prespecified hierarchy or adjustment can clarify how evidence is interpreted across outcomes. The registry record does not report whether either approach was used, so the degree to which multiplicity was addressed is unknown. The absence of that information does not prove that no adjustment occurred; it means the method is not reported in the registry record.

The primary endpoint population is the treated set, defined as all patients who received at least one dose of trial medication. The record does not specify how many participants contributed to the Cohort B endpoint analysis, how missing tumour assessments were handled, or whether participants with unevaluable assessments were included in a particular category. It also does not report the missing-data approach for the listed secondary outcomes or the populations used for their analyses.

These omissions matter because the analyzed population and missing-data rules affect the interpretation of an endpoint estimate. The endpoint result itself is not reported, so there is no estimate to evaluate for sensitivity to missing data. Detailed analysis populations, missing-data procedures, and multiplicity methods are not reported in the registry record.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: general physical health deterioration (17.2%), diarrhoea (10.3%), nausea (9.5%), malignant neoplasm progression (9.5%), vomiting (6.9%).

The registry lists serious-event information separately for Cohort A and Cohort B. For Cohort A, the serious-event fields report 13 affected among 29 at risk; for Cohort B, they report 7 affected among 21 at risk. The other-event fields report 28 among 29 at risk for Cohort A and 20 among 21 at risk for Cohort B. These are cohort-level counts as recorded, not results from a randomized comparison.

The registry also lists serious event terms with percentages: general physical health deterioration, 17.2%; diarrhoea, 10.3%; nausea, 9.5%; malignant neoplasm progression, 9.5%; vomiting, 6.9%; and dehydration, 6.9%. The denominator and whether these terms count participants or events are not reported in the registry record. The terms should therefore be read as listed, not converted into participant counts or added together. In general, participants can experience more than one event, so event categories need not be mutually exclusive.

The deaths fields contain no reported values. These entries describe recorded safety information, but do not establish causation or a comparative safety effect. The record does not provide further details on event duration, severity grading beyond the serious-event designation, or analysis procedures.

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

The registry record documents a phase II study of BIBW 2992 in patients with HER2-negative metastatic breast cancer, with objective response designated as the primary endpoint for Cohort B. It also gives a response definition, an assessment schedule, descriptions of the treated set, and selected safety counts. Those details help establish what the study intended to evaluate and what kinds of outcomes were recorded.

The record does not report the primary objective-response result. It therefore cannot establish from the information available whether Cohort B met the response definition, how frequently that occurred, or how precise an estimate would be. This is not evidence of a null result; the endpoint outcome is absent from the registry record. The same caution applies to the listed secondary outcomes: their presence as named outcomes is not equivalent to reporting their findings.

The study was open-label and non-randomized, and the registry does not describe a concurrent randomized control treatment. In general, this design can describe outcomes among treated participants, but it cannot by itself isolate the effect of treatment from differences in participant characteristics, disease course, or other influences. Cohort differences should not be interpreted as a causal treatment comparison.

The safety counts and event terms provide descriptive information about events recorded in the cohorts. They do not establish that BIBW 2992 caused an event, and the listed percentages have no reported denominator in the record. No comparative safety conclusion can be drawn from these entries alone. The registry does not provide enough information to assess missing observations, analysis methods, or uncertainty around these safety summaries.

Accordingly, the record supports a careful account of trial design and reported safety entries, but not an efficacy conclusion. The distinction between endpoint specification and endpoint result is essential: a trial can define what it plans to measure without the registry record presenting the observed outcome. Statistical interpretation must remain within what is actually recorded.

14. Why This Trial Matters Statistically

This trial is a useful example of why reading a registry requires separating design information from outcome information. The record names objective response as the primary endpoint for Cohort B and defines the response categories using RECIST version 1.0. Yet it does not report the primary result. A clearly specified endpoint can identify the research question, but it cannot answer that question without the observed estimate and information about its uncertainty.

The cohort structure also illustrates why allocation details matter. Although the record describes two cohorts, allocation is non-randomized and masking is absent. These design features affect what kinds of comparisons are credible. Cohort-specific counts can be described, but they should not be treated as randomized evidence that one treatment or group outperformed another. The registry does not name a randomized comparator.

There is also a distinction between event counts and comparative safety evidence. The safety entries describe serious events and other events in each cohort, while the list of event terms gives percentages without a reported denominator. Without definitions and analysis context, adding or directly comparing these terms would be misleading. In general, an event may occur more than once or in the same participant as another event.

Finally, the record illustrates the importance of transparency about analysis populations, missing data, and multiplicity. It identifies the treated set for the primary endpoint but does not report the response estimate or the methods used to address these statistical issues. Recognizing what is absent is part of sound interpretation: it prevents endpoint descriptions and planning assumptions from being mistaken for findings.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The central limitation of the registry information is that the primary objective-response result for Cohort B is not reported. No response count, estimate, or confidence interval is available, so neither the outcome frequency nor its precision can be assessed. The record explicitly provides no Cohort A objective-response data because OR was primary only for Cohort B.

The design was open-label and non-randomized. The registry does not identify a randomized control treatment, so it does not support a like-for-like causal comparison. Cohort outcomes may be influenced by differences in participants or other factors, and the record does not give enough information to assess those differences. The absence of masking is also relevant to assessment, while independent review procedures are not reported in the registry record.

Eligibility wording is inconsistent: the official title refers to failure of no more than two chemotherapy regimens, whereas the brief summary refers to no more than three. The record does not reconcile these statements. It also does not provide detailed demographic or clinical characteristics, cohort-specific enrollment totals, or the number included in the primary endpoint analysis.

Other limitations include unreported methods for missing assessments, endpoint analysis, multiplicity, and time-to-event outcomes. The registry names additional outcomes but does not report their results in the information. For safety, the denominator for the listed event-term percentages is not reported, and the deaths fields are empty. These gaps limit interpretation of both efficacy and safety beyond the descriptive entries recorded.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: An Open Label Phase II Trial of BIBW 2992 in Patients With HER2-negative Metastatic Breast Cancer — NCT00425854.
• Registry record last updated 2013-12-31; status last verified 2013-08.
• Results first posted 2013-10-14.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

NCT00425854 is an open-label, non-randomized phase II trial of BIBW 2992 in HER2-negative metastatic breast cancer. The intervention is described as once-daily oral dosing, with the official title specifying 50 mg; the arm description calls it a high dose once daily. Enrollment is listed as 50. The study is completed, with a start date of 2006-11 and primary completion in 2009-05.

Objective response, defined as complete or partial response by RECIST version 1.0, was the primary endpoint only for Cohort B in the treated set. Tumour assessments were scheduled at screening, week 8, week 16, week 24, and every 8 weeks thereafter. The primary response result is not reported in the registry record. Selected serious-event counts and event terms are listed, but the record does not give their full analysis context. It also does not report a detailed statistical plan or procedures for missing data and multiplicity.