ABI-007 In Combination With Bevacizumab in Women With Metastatic Breast Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for ABI-007 In Combination With Bevacizumab in Women With Metastatic Breast Cancer (NCT00394082), a non-randomized phase 2 study. The primary outcome is Participants With At Least One Treatment-Emergent Adverse Event (TEAE).

2. Clinical Question

3. Trial Design

This was a phase II, non-randomized, single-group trial with no masking. The registry reports 50 participants enrolled and a completed study. The listed start date is 2006-06-01, and the primary completion date is 2009-09-01. These dates describe the registry record; they do not by themselves establish how long any individual participant was followed.

Participants received ABI-007 plus bevacizumab. ABI-007 was scheduled on days 1, 8 and 15 at 125 mg/m², while bevacizumab was scheduled on days 1 and 15 at 10 mg/kg, within each 28-day cycle. The registry says treatment continued until disease progression or intolerable toxicity. It also describes the possibility of continuing one drug when toxicity affects only the other, but the description of that provision is incomplete in the registry record, so its details cannot be characterized further.

A single-group design can describe outcomes among people who received the combination, but it cannot directly estimate how those outcomes compare with a concurrent untreated group or a different treatment. Because allocation was non-randomized and there was no masking, the design also does not provide the protections against selection and assessment biases that randomization and masking can offer. The registry record does not report a randomized control group, a planned external control, or a statistical plan for comparative efficacy.

The registry’s stated aims include safety and tolerability and progression-free survival in previously untreated advanced or metastatic breast cancer. It lists response, stable disease, duration of response and participant survival among secondary outcomes. Numerical efficacy results for these outcomes are not reported in the registry record.

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: Safety population

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 Participants With At Least One Treatment-Emergent Adverse Event (TEAE), measured over up to 25 months.

The primary outcome is the number of participants in the safety population with at least one treatment-emergent adverse event. The registry defines such an event as one that began or worsened in grade after the start of study drug and through 30 days after the last dose. The timeframe for this outcome is up to 25 months.

This is a participant-level count: a participant is counted if at least one qualifying event occurs. The count does not describe how many events each participant experienced, how severe every event was, or how long events lasted. It also does not, by itself, establish that study treatment caused an event. The registry’s definition specifies when an event qualifies for the endpoint, not whether the drug was judged to be its cause.

As a general statistical principle, a safety endpoint of this kind is useful for describing how commonly participants experienced at least one event during the defined observation period. Its interpretation depends on the population included, the event definition, and the completeness of follow-up and reporting. The registry identifies the safety population but does not report the primary endpoint’s numerical result in the primary outcome entry. The adverse-event information separately reports that 50 participants were at risk for other adverse events and that 50 were affected; the record does not explicitly identify this count as the primary treatment-emergent adverse-event result.

The study summary also names progression-free survival as an evaluation aim, but the registry record does not report a numerical result for that outcome. The safety endpoint and the efficacy outcomes answer different questions and should not be treated as interchangeable.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies the primary outcome as a count of participants with at least one treatment-emergent adverse event in the safety population. It does not report a statistical analysis plan or a numerical result for this outcome in its primary outcome entry. The following statistical concepts explain how such results are generally interpreted; they do not describe unreported methods as methods used in this trial.

What does a participant count measure?

For a binary participant-level outcome, each participant is classified according to whether the event occurred at least once during the specified observation period. The reported count is descriptive: it tells how many participants met the event definition in the population being analyzed. It does not count every episode, distinguish a first event from a recurrence, or summarize the event’s duration. Here, the registry’s endpoint definition uses the start of study drug and 30 days after the last dose as the event window, and identifies the analysis population as the safety population.

In general, a proportion can be calculated from a count and its denominator, but a proportion should not be inferred when the registry does not clearly identify the relevant endpoint result and denominator together. The adverse-event record reports 50 participants at risk and 50 affected for other adverse events, while the primary outcome entry describes a treatment-emergent adverse-event count without a numerical value. Those entries should not be merged unless the record explicitly establishes that they refer to the same measure.

Why does the single-group design matter?

In a randomized comparison, random allocation is intended to make treatment groups comparable, on average, with respect to both measured and unmeasured characteristics. This trial is listed as non-randomized and single-group, so there is no concurrent treatment group to support that type of comparison. A count observed in the treated group therefore cannot isolate the contribution of the combination from other explanations, such as the underlying illness, other care, or differences in participant characteristics. This is a general limitation of uncontrolled comparisons, not a claim about what caused any individual event in this trial.

No masking is listed. In general, knowledge of treatment can influence the reporting, assessment, or classification of outcomes, particularly outcomes that involve judgment. The registry does not report additional procedures that would allow readers to assess whether outcome assessment was blinded or standardized beyond the endpoint descriptions.

How should uncertainty be assessed?

A confidence interval is a statistical range that reflects uncertainty around an estimate under specified assumptions. A narrow interval generally indicates greater precision than a wide interval, while an interval that spans a broad range leaves greater uncertainty about the underlying value. The registry record does not report confidence intervals for the adverse-event information or numerical estimates for the registered efficacy outcomes. The precision of those outcomes therefore cannot be evaluated from the record.

A p-value, when reported, relates the observed data to a specified statistical model and a null hypothesis. It is not a direct measure of clinical importance, and it does not replace an effect estimate and its uncertainty. No p-values or comparative tests are reported in the registry record. With no randomized comparator, a treatment-versus-control p-value cannot be inferred from the available record.

What do the listed outcome types imply?

The registry lists a confirmed complete or partial overall response according to RECIST, stable disease for >= 16 weeks or complete or partial response according to RECIST, duration of response, and participant survival among secondary outcomes. It also lists Kaplan-Meier estimates for duration of response and participant survival. As a general statistical method, Kaplan-Meier estimation uses observed event times and accounts for participants whose event time is not observed during follow-up by treating their data as censored at the last known time. The registry’s listing of this method as part of outcome descriptions does not establish that a numerical estimate, a particular censoring rule, or a particular analysis was posted. The results and the detailed analysis procedures are not reported in the registry record.

What does the planning illustration tell readers?

Any sample-size values associated with planning illustrations should be read as hypothetical planning quantities, not as observed trial results or observed power. The stated illustration assumptions use a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect defined as a stated fraction of the observed effect. Those assumptions describe an illustration, not this trial’s single-group allocation. The registry record does not report a sample-size justification for the trial or identify a randomized two-arm design. A planning calculation under assumed two-arm conditions cannot establish that this single-group study had a particular probability of detecting an effect, and it cannot replace a reported analysis.

Over-dispersion is a general concern when observed variability exceeds what a statistical model assumes. If a model underestimates variability, its standard errors and uncertainty intervals may be too small. The registry record does not report a model for count variation or an over-dispersion assessment. This concept therefore provides general context only and should not be taken as a description of an analysis conducted in the trial.

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("ABI-007 Plus Bevacizumab")
n      <- c(50)
events <- c(50)
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  ABI-007 Plus Bevacizumab  50  50  100.0  92.9  100.0  92.9  100.0  93.0  100.0

11. Multiplicity, Missing Data and Analysis Populations

The registry lists six outcome measures, including the primary adverse-event outcome and secondary measures for response, stable disease, duration of response and participant survival. It does not report how many statistical hypotheses were formally tested, whether any outcomes were designated for confirmatory inference, or whether a multiplicity adjustment was planned. The registry lists 0 analyses. Without a statistical plan and outcome estimates, it is not possible to assess how multiple outcomes were handled or whether any reported inference accounted for multiplicity.

As a general statistical principle, examining multiple outcomes can increase the chance of finding an apparently notable result by chance when each is tested separately. A prespecified hierarchy or adjustment can address that issue, but the registry record does not report whether either approach was used. No p-values are posted in the registry information described for these outcomes, so multiplicity-adjusted or unadjusted significance claims cannot be evaluated.

The registry identifies the safety population for the primary outcome but does not describe the analysis population for each efficacy outcome. It does not report how many participants contributed data to the response, stable disease, duration of response, survival, or progression-free survival assessments. Enrollment of 50 participants does not establish that all 50 were evaluable for every outcome.

Missing-data procedures are not reported in the registry record. It does not say how participants with incomplete follow-up, unavailable assessments, or discontinuation were handled for each outcome. For time-to-event outcomes, censoring rules and follow-up completeness are also not reported. These omissions make it difficult to determine whether an estimate, if available elsewhere, would be affected by missing observations or differential follow-up.

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: febrile neutropenia (6.0%), neutropenia (2.0%), dehydration (2.0%), electrolyte imbalance (2.0%), hypovolaemia (2.0%).

The registry reports serious adverse events among 50 participants at risk, with 13 affected. The listed serious adverse event terms are febrile neutropenia at 6.0%, and neutropenia, dehydration, electrolyte imbalance, hypovolaemia and chemical pneumonitis each at 2.0%. The registry does not state in this record how the listed terms were selected, whether participants could contribute more than one term, or how the percentages were calculated. The count of deaths is not reported in the registry record.

These entries describe reported events in the treated group; they do not establish that the combination caused each event. Serious adverse events are not the same endpoint as the primary outcome, which counts participants with at least one treatment-emergent adverse event. The registry separately reports 50 participants at risk for other adverse events and 50 affected, but its primary outcome entry does not state a numerical treatment-emergent adverse-event result. The record also does not provide event-level detail about timing, duration, severity beyond the serious classification, or attribution.

As a general principle, adverse-event counts are most informative when read alongside clear definitions, denominators, follow-up, and a comparator. The single-group design allows a descriptive account of the reported events but does not support a controlled comparison of safety between treatments.

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

The registry record supports a descriptive account of a completed, single-group phase II trial of ABI-007 plus bevacizumab, with 50 participants enrolled. The safety information reports 13 participants affected by serious adverse events among 50 at risk. It also lists serious adverse event terms and their reported percentages. These observations characterize events recorded in this group; they do not establish why the events occurred.

The record does not report numerical efficacy results for the registered response, stable disease, duration of response, participant survival, or progression-free survival outcomes. It also does not provide confidence intervals or comparative treatment-effect estimates. The efficacy aims described in the study summary therefore cannot be evaluated from the registry information available in this record. The primary outcome is defined, but a numerical value is not reported in its primary outcome entry.

The absence of a randomized comparator is important to what can be concluded. A single-group study can document outcomes among participants who received the combination, but cannot directly distinguish treatment effects from other influences on those outcomes. Non-randomized allocation and lack of masking also limit how confidently outcomes can be compared with results from another group or study.

As a general statistical principle, evidence about an outcome includes both an estimate and its uncertainty, interpreted in light of the design and analysis. Counts and percentages alone do not show whether an observed pattern would be similar in a different population or under a different treatment. The registry does not report enough efficacy or analysis detail to assess that uncertainty for the stated efficacy aims.

These limitations do not make the safety entries meaningless: they remain a record of reported events in the study group. They do mean that the entries should be interpreted descriptively rather than as proof of a causal treatment effect or a comparative advantage. No treatment recommendation follows from the registry information alone.

14. Why This Trial Matters Statistically

This trial is statistically informative as an example of the distinction between describing outcomes and estimating comparative effects. Its single-group, non-randomized design can record what happened among participants who received ABI-007 with bevacizumab, but it does not create a randomized counterfactual group against which to compare those outcomes. That design feature shapes what safety and efficacy statements the registry can support.

The record also illustrates why endpoint definitions and numerical results need to be read together. The primary outcome specifies a treatment-emergent event window and a safety population, while its primary outcome entry does not provide a numerical count. Separate adverse-event entries give counts for participants at risk and affected, but they should not automatically be treated as the primary endpoint result without an explicit link in the record.

Outcome completeness matters as well. The registry names response and time-to-event measures, including progression-free survival as a study aim, yet numerical estimates and their uncertainty are not reported. Listing an outcome or an analysis method does not make the result assessable. Without estimates, confidence intervals, and details about analysis populations and missing data, readers cannot judge precision or evaluate how analysis choices might have influenced an efficacy result.

Finally, planning illustrations are not evidence of observed trial performance. Assumptions involving two-arm allocation do not describe this single-group design, and planning quantities do not establish observed power. Keeping design, endpoint definitions, and posted results distinct helps prevent descriptive safety information or hypothetical planning calculations from being mistaken for comparative evidence.

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 most important design limitation is the absence of a randomized comparator. The trial is listed as non-randomized, single-group and unmasked, so the posted outcomes cannot provide a controlled estimate of how the combination compares with another treatment. Differences between this group and participants in other studies could reflect differences in populations, care, follow-up, outcome assessment, or other factors, rather than treatment effects.

The registry reports enrollment of 50, but does not report numerical results for the registered efficacy outcomes. The record therefore does not permit assessment of response, stable disease, duration of response, participant survival, or progression-free survival estimates or their precision. The primary adverse-event outcome is defined, but its numerical result is not reported in its primary outcome entry. The separate adverse-event information should not be assumed to fill that gap without explicit confirmation that it is the same outcome.

The record does not report a statistical analysis plan, sample-size justification, interim analysis procedures, missing-data methods, efficacy analysis populations, or detailed follow-up and censoring rules. These omissions limit reproducibility and assessment of potential analysis choices. Although outcome descriptions list Kaplan-Meier estimates for duration of response and participant survival, the record does not provide the estimates or detailed procedures.

The serious adverse-event entry reports 13 affected among 50 at risk and lists selected terms, but does not provide event-level descriptions, attribution, recurrence, or the death count. The treatment description also contains an incomplete statement about continuing one drug when toxicity affects only the other; the details of that provision are not available in the registry record. These limitations leave important context for interpreting both treatment exposure and reported safety outcomes unresolved.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: ABI-007 In Combination With Bevacizumab in Women With Metastatic Breast Cancer — NCT00394082.
• Registry record last updated 2019-11-25; status last verified 2019-11.
• Results first posted 2013-05-22.
• 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 identifies NCT00394082 as a completed phase II trial of ABI-007 in combination with bevacizumab in women with metastatic breast cancer. It reports a non-randomized, single-group design with no masking and enrollment of 50 participants. The listed study start date is 2006-06-01, and the primary completion date is 2009-09-01.

The primary outcome is the number of participants in the safety population with at least one treatment-emergent adverse event, defined by onset or worsening after study drug began through 30 days after the last dose. A numerical value for this outcome is not reported in its primary outcome entry. The serious adverse-event record reports 13 affected among 50 at risk; the death count is not reported in the registry record.

The registry lists response, stable disease, duration of response, participant survival and progression-free survival among the outcomes or study aims, but numerical efficacy results are not reported in the registry record. It also lists 0 analyses and does not report detailed statistical methods for the outcomes.