A Study of Avastin (Bevacizumab) in Combination With Chemotherapy in Patients With Primary Breast Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Avastin (Bevacizumab) in Combination With Chemotherapy in Patients With Primary Breast Cancer (NCT02005549), a non-randomized phase 2 study. The primary outcome is Percentage of Participants With Pathological Complete Response (pCR).

2. Clinical Question

3. Trial Design

This was a phase II, non-randomized, single-group study of neoadjuvant bevacizumab with docetaxel and capecitabine in patients with primary breast cancer. The registry lists an enrollment of 18 and a completed status. It reports a start of 2006-02 and primary completion of 2008-04. With no random assignment and no separate comparator arm, the design describes participants receiving the combination rather than a controlled treatment contrast.

The brief study summary describes bevacizumab at 15mg/kg intravenously every 3 weeks during the first 5 chemotherapy cycles. The anticipated time on study treatment was 3-12 months. These details describe the registry summary; further protocol information, including any treatment modifications or adherence details, is not reported in the registry record.

The primary endpoint was assessed at final surgery, with the listed timeframe of baseline and 20-24 weeks. The record says surgery was performed 2 to 4 weeks after the last chemotherapy cycle, identified as Week 18. The endpoint definition relies on the local pathologist's judgment of the final surgical sample.

The registry also lists outcomes for clinical complete or partial response and breast-conserving surgery. It reports 3 outcome measures and 0 analyses. The record does not state a planned sample-size justification, interim-analysis approach, or detailed statistical analysis plan. Those features should not be inferred from the phase designation or the fact that enrollment was 18.

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

Analysis population for the primary outcome, as defined in the registry: ITT population; participants evaluable for response included those participants who received a minimum of 3 cycles of treatment (9 weeks on study) with final surgery performed and the samples and reports available.

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.

Population: Intent-to-treat (ITT) population. All participants who received at least one dose of study drug.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants With Pathological Complete Response (pCR), measured over Baseline, 20-24 weeks (final surgery, performed 2 to 4 weeks after the last chemotherapy cycle [Week 18]).

The primary endpoint was the percentage of participants with pathological complete response, or pCR, at final surgery. The registry defines pCR as the absence of signs for invasive tumor in the final surgical sample, as judged by the local pathologist. Surgery was scheduled 2 to 4 weeks after the last chemotherapy cycle, with the timeframe also described as 20-24 weeks and Week 18 for the last cycle.

General statistical context: a binary percentage expresses how often an outcome is observed in the population being assessed. pCR is a surgical and pathological endpoint: it depends on the examination of tissue obtained at surgery, rather than on clinical assessment alone. It is used to describe response in the resected tumor after pre-operative treatment. This endpoint does not by itself describe every aspect of a participant's subsequent health or establish a comparison with another treatment.

The registry identifies an intention-to-treat population and separately describes participants evaluable for response as those who received a minimum of 3 cycles, had final surgery, and had samples and reports available. The number meeting that evaluability description and the pCR percentage are not reported in the registry record. The listed 95% confidence interval is also not accompanied by a numerical result.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a percentage endpoint describe?

The primary outcome is a percentage of participants with pCR, a binary outcome: each assessed participant either meets the stated definition or does not. The registry defines the qualifying finding as absence of signs for invasive tumor in the final surgical sample, as judged by the local pathologist. General statistical context: a percentage summarizes the observed frequency of an outcome within a specified population. Its meaning depends on both the outcome definition and who is included in the denominator.

The record names an intention-to-treat population, and then describes response-evaluable participants as those who received a minimum of 3 cycles, had final surgery, and had the samples and reports available. Those criteria may result in a different analysis population from everyone enrolled. The registry record does not report how many participants met the evaluability criteria or give a numerical primary outcome result. Therefore, the denominator and the observed pCR percentage cannot be reconstructed from the record.

What does a 95% confidence interval add?

The endpoint dispersion is identified as a 95% confidence interval. General statistical context: an interval communicates uncertainty around an estimate, rather than replacing the estimate itself. A wider interval indicates less precision than a narrower interval under the same interpretation, while the endpoints indicate a range of values compatible with the data and method. The interval's numerical limits are not reported in the registry record, so neither its width nor the precision of the pCR estimate can be assessed.

A confidence interval is not a guarantee that the true value falls within the particular reported interval, and it does not correct limitations of the study design. In particular, an interval around a single-group percentage would describe uncertainty in that group's result; it would not create a controlled comparison or remove potential selection and measurement limitations.

How should the single-group design be read?

The record identifies allocation as non-randomized, masking as none, and the model as single-group. General statistical context: randomization helps balance known and unknown prognostic factors between treatment groups, while masking can reduce some influences on assessment. Neither feature is present as a comparison mechanism in the registry description. Because there is no separate control arm, the study can describe outcomes among participants receiving the combination but cannot isolate the regimen's effect from other explanations for those outcomes.

The primary endpoint uses local pathologist judgment of surgical samples. General statistical context: an outcome that depends on human assessment can be affected by variation in interpretation. The record does not report centralized pathology review or masking of pathologists, so neither should be assumed. The lack of masking is listed for the study; further details about outcome-assessment procedures are not reported in the registry record.

What do the planning calculations mean?

Any sample-size illustrations associated with the posted results are planning calculations, not observed study findings and not a measure of the study's observed power. Their stated assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is described as a stated fraction of the observed effect. This is a hypothetical planning setup, not the actual single-group, non-randomized design of this study.

General statistical context: planning calculations ask how much information might be needed to detect a specified effect under chosen assumptions. Changing the target effect or the statistical assumptions changes the required sample size. These calculations do not show that a study with a particular enrollment was adequately powered, nor do they establish that an effect exists. In this record, the underlying primary response estimate is not reported, and no numerical planning result is stated in the registry record. The registry also does not report the study's own sample-size justification, so a planning illustration cannot be treated as the protocol rationale.

Why does over-dispersion matter for other outcome types?

General statistical context: over-dispersion arises when observed variation is greater than a statistical model expects. For binary percentage outcomes, the assumed model and uncertainty calculation should reflect the data structure and analysis population. If variability is understated, uncertainty can appear smaller than it is. No over-dispersion assessment or modeling approach is reported in the registry record, and the missing primary estimate and interval prevent independent assessment. This is a general consideration, not a claim that over-dispersion occurred in this study.

The registry's outcome and analysis descriptions do not identify a specific model, adjustment set, or hypothesis-testing procedure. Those details are not reported in the registry record. Accordingly, the numerical methods behind the listed confidence interval cannot be evaluated from the entry, and no alternative recalculation is available from the posted information.

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("Bevacizumab+Docetaxel+Capecitabine")
n      <- c(18)
events <- c(4)
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  Bevacizumab+Docetaxel+Capecitabine  18  4  22.2  6.4  47.6  9.0  45.2  9.1  45.6

11. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary pCR outcome and secondary outcomes for clinical complete response, clinical partial response, and breast-conserving surgery. It reports 3 outcome measures and 0 analyses. The numerical results for the primary and secondary outcomes are not reported in the registry record. The record does not say whether the outcomes were tested as separate hypotheses, whether any multiplicity adjustment was planned, or which outcome was designated as the focus of any statistical testing beyond identifying pCR as primary.

General statistical context: when several outcomes or hypotheses are tested, the chance of at least one apparently unusual result can increase if each is assessed separately without accounting for the number of tests. A prespecified testing hierarchy or adjustment can address some multiplicity concerns. The registry entry does not describe either approach, so no conclusion about multiplicity control is possible. The absence of analyses in the record should not be interpreted as evidence that no analysis was performed; it means analysis results and methods are not reported there.

The primary outcome description combines an intention-to-treat population label with criteria for response evaluability: at least 3 cycles of treatment, final surgery, and available samples and reports. The record does not give the number of participants meeting those criteria, explain how participants without evaluable surgery or pathology were handled, or report missing endpoint data. A missing-data method, imputation approach, or sensitivity analysis is not reported in the registry record.

These omissions limit assessment of how closely the analyzed population may correspond to all enrolled participants and whether missing outcome information could affect the reported estimate. Because the endpoint result itself is not present, the extent or direction of any such effect cannot be determined.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: deep vein thrombosis (11.1%), intestinal perforation (5.6%), general physical health deterioration (5.6%), impaired healing (5.6%), neutropenic infection (5.6%).

The registry reports safety figures for the bevacizumab, docetaxel, and capecitabine group. For serious adverse events, 18 participants were at risk and 5 were affected. For other adverse events, 18 participants were at risk and 18 were affected. These are registry counts for the listed categories; they do not give the duration of observation or explain whether participants could contribute more than one event.

The record lists deep vein thrombosis at 11.1% and intestinal perforation at 5.6% among the top serious adverse-event terms. It also lists general physical health deterioration, impaired healing, neutropenic infection, and contralateral breast cancer, each at 5.6%. These term-specific percentages should be read as reported, not added together: the record does not say that event categories are mutually exclusive or provide a basis for summing them.

General statistical context: adverse-event figures describe events observed and recorded in the study group. In a single-group study, they do not establish whether the frequency is higher or lower than it would be under another treatment. The death figure is not reported in the registry record. Details such as event timing, duration, severity grading beyond the serious-event categorization, and attribution to treatment are not reported in the registry record.

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

The registry record establishes that this was a completed phase II, non-randomized, single-group study with enrollment of 18. It describes neoadjuvant bevacizumab with docetaxel and capecitabine and defines pCR by the absence of signs for invasive tumor in the final surgical sample, judged by the local pathologist. It also records selected serious and other adverse-event figures for the treated group.

It does not report the numerical pCR percentage or its 95% confidence interval. It also does not report numerical findings for the listed secondary outcomes. As a result, the posted information cannot show the magnitude or precision of the primary endpoint result. A missing result is not evidence for either a high or low pCR percentage.

General statistical context: an observed outcome in a single treatment group can describe what occurred among enrolled and assessed participants, but it does not by itself show what would have occurred under a different treatment or without treatment. Without a comparator, differences due to patient characteristics, disease course, assessment, and other factors cannot be separated from a treatment effect. The study's non-randomized status reinforces that limitation.

The safety figures describe reported events in the study group, not comparative risks. They do not establish that an event was caused by the regimen, and the registry does not provide a comparator for judging relative frequency. The listed percentages are not a complete description of all possible safety outcomes.

The record also does not provide a statistical analysis plan, sample-size justification, missing-data approach, or multiplicity procedure. Those gaps restrict assessment of how endpoints were analyzed and how robust the estimates might be. Accordingly, the study record supports a careful description of the design and recorded outcomes, but not a conclusion about comparative efficacy, causation, or the regimen's effect relative to another option.

14. Why This Trial Matters Statistically

This study is statistically informative as an example of the difference between an outcome definition and a reported outcome estimate. The registry specifies what counted as pCR and gives a confidence-interval type, but it does not report the pCR percentage or interval limits. A clearly defined endpoint is important, yet without its estimate and uncertainty, readers cannot assess the result's size or precision.

The single-group design also determines what conclusions the data can support. It can describe outcomes among participants who received the regimen, but there is no like-for-like control comparison. Without randomization and a comparator, treatment-associated outcomes cannot be interpreted as a controlled estimate of benefit or harm relative to another approach.

The record highlights why population definitions matter. It labels the primary population as intention-to-treat while describing response-evaluable participants through treatment exposure, surgery, and availability of samples and reports. Since the number meeting those criteria and the missing-data handling are not reported, the relationship between enrollment and the analyzed endpoint population remains unclear.

Finally, the listed safety counts and percentages demonstrate that event reporting has its own interpretive limits. They describe what was recorded in this group, but do not provide comparative risk or establish attribution. Together, these features make the record useful for understanding design and measurement, while underscoring why endpoint estimates, analysis details, and an appropriate comparator are necessary for stronger statistical interpretation.

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 principal limitation for interpreting efficacy is that the registry record does not report the primary pCR percentage or its 95% confidence interval. Numerical results for the secondary outcomes are also not reported. This prevents assessment of the magnitude, precision, and pattern of the reported response outcomes.

The design is non-randomized and single-group, with no separate control arm. It therefore cannot provide a controlled estimate of how outcomes compare with another treatment. The registry lists masking as none. It does not report whether pathology assessment was centrally reviewed or otherwise standardized beyond judgment by the local pathologist.

The enrollment is 18, but the number included in the response-evaluable population is not reported. The primary outcome description requires treatment exposure, final surgery, and available samples and reports for response evaluability. The record does not explain how missing samples, unavailable reports, or participants without final surgery were handled. It also does not report a missing-data method or sensitivity analysis.

The registry lists 3 outcome measures and 0 analyses, but does not provide the statistical analysis plan, multiplicity approach, sample-size justification, interim-analysis details, or the method used to calculate the confidence interval. Safety information gives selected event counts and terms, but details on event timing, attribution, and observation duration are not reported. These limitations concern what can be interpreted from the registry entry; they do not establish that unreported procedures or analyses were absent from the study.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study of Avastin (Bevacizumab) in Combination With Chemotherapy in Patients With Primary Breast Cancer — NCT02005549.
• Registry record last updated 2014-07-14; status last verified 2014-05.
• Results first posted 2014-06-06.
• 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

The registry describes a completed phase II study of neoadjuvant bevacizumab with docetaxel and capecitabine for primary breast cancer. It was non-randomized and single-group, with enrollment of 18. The listed primary outcome was the percentage of participants with pathological complete response at final surgery, defined by absence of signs for invasive tumor in the final surgical sample as judged by the local pathologist.

The registry does not report the numerical primary outcome or its 95% confidence interval, and it does not report numerical results for the listed secondary outcomes. It reports serious adverse events in 5 of the 18 participants at risk and other adverse events in 18 of the 18 participants at risk, with selected serious-event terms and percentages. The single-group design does not support a like-for-like treatment comparison.