A Phase II Combination of Trastuzumab and Ixabepilone Versus Trastuzumab and Docetaxel in Patients With Advanced and/or Metastatic Breast Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Phase II Combination of Trastuzumab and Ixabepilone Versus Trastuzumab and Docetaxel in Patients With Advanced and/or Metastatic Breast Cancer (NCT00490646), a randomized phase 2 study. The primary outcome is Percentage of Participants With Objective Response (OR; Assessed by Response Evaluation Criteria in Solid Tumors [RECIST] Version 1.1).

2. Clinical Question

3. Trial Design

This was a randomized, open-label, parallel-group Phase 2 study. The registry record lists enrollment of 50 and records the study as completed. Its stated purpose was to assess response in participants with HER2-positive locally advanced and/or metastatic breast cancer. The brief summary specifies that participants had not previously received chemotherapy or trastuzumab.

Participants were assigned to trastuzumab plus ixabepilone or trastuzumab plus docetaxel. Each treatment group lists 25 participants. Trastuzumab was given as a 4 mg/kg loading dose, followed by 2 mg/kg weekly. Ixabepilone was administered at 40 mg/m^2 intravenously over 3 hours once every 21 days; docetaxel was administered at 100 mg/m^2 intravenously over 1 hour once every 21 days. Both regimens were described as continuing until disease progression or unacceptable toxicity.

The study had no masking, so treatment assignment was not concealed from participants or study personnel. This matters particularly for an endpoint assessed by the investigator: knowledge of treatment assignment can be a potential source of influence on assessments, even when response criteria are specified. The primary endpoint description states that response was evaluated using RECIST version 1.1.

The registry lists a start date of 2008-02 and primary completion in 2011-06. These record dates provide context for the study but do not establish when each participant was assessed. The registry record does not report the original sample-size justification, the planned statistical analysis in detail, or whether interim analyses were planned. Those features should not be inferred from the reported enrollment or the posted response comparison.

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: All randomized participants.

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 Percentage of Participants With Objective Response (OR; Assessed by Response Evaluation Criteria in Solid Tumors [RECIST] Version 1.1), measured over Assessed every 6 weeks from initiation of study therapy up to 12 months; then every 3 months until disease progression (maximum time that any participant was on therapy was 108 weeks).

The primary endpoint was the percentage of participants with an objective response: a best overall response of complete response or partial response under RECIST version 1.1, as determined by the investigator. A complete response required disappearance of all evidence of target and non-target lesions. A partial response required at least 30% reduction from baseline in the sum of the longest diameters of all lesions. The response criteria had to be met again after 4 weeks and before 6 weeks after the initial assessment.

Response was assessed every 6 weeks from the start of study therapy up to 12 months, then every 3 months until disease progression. The registry record notes that the maximum time any participant was on therapy was 108 weeks. These are assessment and treatment-duration details; they do not mean that every participant was followed for the same length of time.

Objective response is a binary outcome at the participant level: each participant is classified as having met the response definition or not. Reporting the percentage in each group makes the observed response frequency interpretable, while a confidence interval conveys the uncertainty around each estimate. This endpoint is not a measure of how long a response lasted or how long participants lived without progression. Progression-free survival, time to response, and duration of response are listed separately as secondary outcomes.

7. Results Posted in the Registry

• Trastuzumab 2 mg/kg + Docetaxel 100 mg/m^2 IV: 52.0% (13/25) versus 60.0% (15/25) on control; risk difference -8.0 percentage points (95% CI -35.4 to 19.4), risk ratio 0.87 (0.53–1.42); unadjusted two-proportion z-test p = 0.569.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is a binary response outcome. For each participant, the investigator classified the best overall response as complete response, partial response, or neither of these response categories. The reported percentage is the proportion classified as having an objective response. General statistical principle: binary endpoints are often summarized by a proportion because it directly describes the frequency of participants meeting the endpoint definition.

How should the confidence intervals be understood?

The registry record states that two-sided confidence intervals for the response percentages were computed using the Clopper-Pearson method. This is an exact binomial interval method. General explanation: a confidence interval describes the range of values reasonably compatible with the observed data under the method’s assumptions. It is not a guarantee that the true value lies inside a particular interval, and it does not measure the probability that one treatment is better than the other.

The response intervals are reported separately for the two treatment groups. They describe uncertainty in each group’s response percentage. A direct comparison requires its own estimate and uncertainty interval; the difference between two percentages is not assessed merely by looking at whether their separate intervals overlap. The registry record also reports a risk difference, a risk ratio, and an odds ratio for the comparison.

What does the risk difference express?

The risk difference is an absolute contrast between the response proportions in the two groups. The posted estimate is -0.08, with a 95% confidence interval from -0.354 to 0.194. The interval includes zero, the value corresponding to equal proportions. Its width indicates that the comparison is not estimated precisely. General explanation: an interval around an absolute difference is useful because it expresses uncertainty on the same scale as the difference itself.

Why are ratio measures also reported?

A risk ratio compares the proportions as a ratio, while an odds ratio compares the odds of response. The registry reports a risk ratio of 0.867, with a 95% confidence interval from 0.529 to 1.421, and an odds ratio of 0.722, with a 95% confidence interval from 0.235 to 2.215. For these ratio measures, the null value is one; both reported intervals include that value. General explanation: ratios can convey relative contrasts, but their interpretation depends on the measure and they do not replace the response percentages or the absolute difference.

What does the p-value add?

The unadjusted p-value for the comparison is 0.569. General explanation: a p-value describes how unusual data at least as incompatible with a specified null model would be if that model and the test assumptions held. It is not the probability that the null hypothesis is true, nor does it quantify the size or clinical importance of an effect. The value should be read alongside the point estimate and confidence interval, rather than as a stand-alone verdict.

The registry record does not report a detailed statistical analysis plan, a named comparison test, or whether adjustment was made for any factors. The posted p-value is identified as unadjusted. It is therefore appropriate to describe it as an unadjusted comparison, without attributing unreported model features to the trial.

Why can small group sizes leave substantial uncertainty?

General explanation: a proportion estimated from a limited number of participants can change noticeably when the observed outcome status of a few participants differs. This can produce wide intervals, particularly when the aim is to estimate a group contrast. The response confidence intervals and comparison interval in this record illustrate the need to consider precision as well as the point estimates. Statistical uncertainty is not removed by the fact that the study was randomized.

The endpoint used the investigator’s assessment under RECIST version 1.1. The response definition included confirmation of the criteria after the initial assessment. The open-label design means that treatment assignment was not masked, while the registry does not report whether independent review of response was used. No independent-review procedure should be assumed.

10. Confidence Intervals and Precision

The primary response estimates are 60.0% for trastuzumab plus ixabepilone and 52.0% for trastuzumab plus docetaxel. Their 95% confidence intervals are 38.7% to 78.9% and 31.3% to 72.2%, respectively. Both intervals cover a wide range of plausible response percentages, so the point estimates should not be read as highly precise descriptions of the underlying response rates.

The direct risk-difference estimate is -0.08, with a 95% confidence interval from -0.354 to 0.194. This range includes zero and extends in both directions, indicating considerable uncertainty about the direction and size of the group contrast. It is more informative to consider this interval than to focus only on which group’s point estimate is larger. The unadjusted p-value of 0.569 is consistent with the uncertainty reflected by the interval, but does not establish equality between treatments.

General explanation: precision concerns how narrowly a quantity is estimated. A narrower confidence interval indicates greater precision under the chosen method; a wider one leaves more uncertainty. Precision is different from whether a result is statistically significant and different again from whether an effect is important to patients. The registry record does not report a threshold for a clinically meaningful difference, so the clinical importance of particular values within the comparison interval cannot be determined from the posted results alone.

The registry states that the response intervals were calculated with the Clopper-Pearson method. These are two-sided 95% intervals. The analysis population for the primary endpoint is described as all randomized participants. The registry record does not provide further detail about handling missing response assessments, so the reported precision cannot be fully evaluated with respect to any missing-outcome assumptions.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are illustrations based on the observed effect, not the trial’s original design assumptions. For this binary outcome, the figures are participants per arm, not total events. The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect. The trial’s own original sample-size assumptions and justification are not reported in the registry record.

The 80% power column gives the illustrated number of participants per arm for a design intended to have an 80% chance of detecting the specified effect under those assumptions, if that effect and the model assumptions were correct. The 90% power column describes the corresponding planning target for a 90% chance. These are prospective planning quantities, not a description of the chance that the completed trial detected an effect.

For planning for 100% of the observed effect, the figures are 604 participants per arm for 80% power and 807 participants per arm for 90% power. For planning for 75% of the observed effect, they are 1,068 participants per arm for 80% power and 1,429 participants per arm for 90% power. For planning for 50% of the observed effect, they are 2,389 participants per arm for 80% power and 3,198 participants per arm for 90% power.

Smaller effects are harder to distinguish from random variation, so they require more participants under the same planning assumptions. These figures are not observed power; they are sample-size planning illustrations calculated using the observed effect and the stated assumptions.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.600; treatment proportion p₁ = 0.520; average p̄ = 0.560.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7020 = 1.3759.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6997 = 0.5889.
4. n = (1.3759 + 0.5889)² / (-0.0800)² = 603.2, rounded up to 604 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.6                          # control proportion
p1 <- 0.52                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(490646)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 604 (80% power), 807 (90% power)
 75% of observed effect: n per arm = 1,068 (80% power), 1,429 (90% power)
 50% of observed effect: n per arm = 2,389 (80% power), 3,198 (90% power)
power.prop.test: 603.2 
Simulated power at n = 604 per arm: 0.800 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists six secondary outcomes, including progression-free survival, time to response, and duration of response, as well as safety outcomes. The primary endpoint is objective response. Because several outcomes are listed, interpreting multiple statistical tests would require knowing whether and how the analysis accounted for multiplicity. A multiplicity adjustment or testing hierarchy is not reported in the registry record.

The posted comparison includes an unadjusted p-value. No information is reported about adjustment for multiple comparisons, covariate adjustment, or a prespecified sequence for testing outcomes. Therefore, the p-value should be treated as an unadjusted result rather than as a multiplicity-controlled finding. The registry record does not report whether interim analyses were planned or conducted.

The primary endpoint population is all randomized participants. This identifies the population specified for the endpoint, but does not explain how participants without an evaluable response were handled. The registry record does not report the amount or pattern of missing response data, the rules used for classifying participants with missing assessments, or any sensitivity analyses for missingness.

These omissions matter because conclusions can depend on the handling of missing outcomes and on whether multiple endpoints or comparisons are tested. They do not demonstrate that an inappropriate method was used; they mean the corresponding procedures cannot be evaluated from the registry record. The record also lists zero analysis entries, so no additional endpoint-specific analysis details are available there.

14. 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 (20.8%), neutropenia (12.5%), dyspnoea (8.3%), circulatory collapse (4.2%), ataxia (4.2%).

The safety figures count participants with events, rather than the total number of event episodes. In the docetaxel group, 13 of 24 participants at risk were recorded with a serious adverse event; in the ixabepilone group, 6 of 24 were recorded with a serious adverse event. For other adverse events, the corresponding counts were 23 of 24 and 24 of 24. These are descriptive counts and should be read separately from the primary response comparison.

The registry also lists serious adverse-event terms and percentages: febrile neutropenia, 20.8%; neutropenia, 12.5%; dyspnoea, 8.3%; circulatory collapse, 4.2%; ataxia, 4.2%; and pericardial effusion, 4.2%. The listed terms do not identify an arm for each percentage in the posted summary, so an arm-specific interpretation of those percentages is not reported in the registry record.

Deaths are not reported in the registry record. These safety summaries do not establish causation, and the counts alone do not describe event severity beyond the serious-event classification or the time at which events occurred. They should not be used as a substitute for the full adverse-event results.

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

The posted results describe objective response in a randomized Phase 2 comparison of two trastuzumab combinations. The observed response estimates were 60.0% with ixabepilone and 52.0% with docetaxel. The estimated risk difference was -0.08, and its confidence interval extended from -0.354 to 0.194. The interval includes zero and is wide, so the observed estimates do not precisely determine the direction or magnitude of the difference.

These results do not show that the combinations are equivalent. A failure to demonstrate a clear difference is not the same as evidence that the treatments have the same effect. Equivalence or non-inferiority would require an appropriate design and prespecified margin; neither is reported in the registry record. Nor do the posted response estimates, on their own, establish whether either treatment is beneficial compared with no treatment or another reference, because this trial compared two active combinations.

The primary endpoint concerns best overall response under RECIST version 1.1. It does not answer how long participants remained progression-free, how quickly responses occurred, how long responses lasted, or how participants fared on other outcomes. Those outcomes are listed separately as secondary measures. The safety summaries likewise describe recorded events, not a causal comparison.

The study was open-label, and response was determined by the investigator. This context is relevant when considering assessment, although the registry record does not report whether independent review was conducted. The sample-size illustrations are planning examples based on the observed effect; they do not change the uncertainty in the completed trial’s estimates or provide evidence that the study had sufficient power for a particular target effect.

The most defensible reading is therefore descriptive and uncertainty-aware: the groups had the reported response estimates, but the confidence intervals leave substantial uncertainty about the underlying response rates and their contrast. The record does not support a definitive claim of superiority, equivalence, or non-inferiority.

16. Why This Trial Matters Statistically

This study illustrates why a point estimate is not enough to interpret a randomized comparison. The response estimates differ numerically, but the direct comparison has a broad confidence interval that includes zero. That uncertainty is central to the statistical reading: the observed ordering alone cannot establish a dependable difference between the regimens.

It also shows the value of reporting both group-specific outcomes and a direct contrast. The response percentages describe each group, while the risk difference expresses the absolute comparison. The risk ratio and odds ratio provide relative measures, but none removes the uncertainty evident in their intervals. Readers can see how conclusions may vary with the scale used while still recognizing that the estimates are imprecise.

The open-label design and investigator-assessed endpoint make the distinction between randomized allocation and masked outcome assessment important. Randomization can support a fair comparison of groups, but it does not by itself eliminate possible influences on an unmasked assessment. The record specifies RECIST response criteria, yet does not report independent review or a detailed plan for handling missing assessments.

Finally, the sample-size illustrations demonstrate a general planning principle: targeting smaller effects requires larger studies under the same assumptions. Because these illustrations use the observed effect rather than the trial’s original design assumptions, they should not be mistaken for evidence about the trial’s original planning or its observed power. The registry record does not report that original rationale, leaving an important distinction between the completed trial and hypothetical future planning.

17. 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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The study was a Phase 2 trial with 25 participants listed in each group. This limits the precision of the response and safety estimates, as reflected in the confidence intervals. The registry record does not report the original sample-size rationale, so it is not possible to assess the intended precision or target effect from the record.

The trial was open-label, and the primary endpoint was determined by the investigator. The record describes response criteria and confirmation requirements but does not report whether outcome assessment was independently reviewed. This leaves uncertainty about the safeguards, if any, used to reduce potential influence of knowledge of treatment assignment.

Although the primary endpoint population is stated as all randomized participants, the registry record does not describe missing response assessments, exclusions from the analysis, or methods for handling participants without an evaluable endpoint. It also does not report a detailed statistical analysis plan, multiplicity procedures, interim-analysis plans, or adjustment for other factors. These are limits on what can be evaluated from the registry record, not proof that the trial did not use such procedures.

The posted comparison is unadjusted, and its confidence interval is broad. The record does not report an equivalence or non-inferiority framework, so the findings should not be interpreted as demonstrating that the regimens have the same effect. Secondary outcomes are listed, but the registry record does not provide analysis details for them in the posted summary. Finally, the listed serious adverse-event terms do not identify arm-specific percentages, and deaths are not reported in the registry record.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: A Phase II Combination of Trastuzumab and Ixabepilone Versus Trastuzumab and Docetaxel in Patients With Advanced and/or Metastatic Breast Cancer — NCT00490646.
• Registry record last updated 2016-03-10; status last verified 2016-02.
• Results first posted 2012-07-18.
• 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.

21. Record Summary

The registry record describes a completed, randomized, open-label, parallel-group Phase 2 study of trastuzumab plus ixabepilone versus trastuzumab plus docetaxel in HER2-positive locally advanced and/or metastatic breast cancer. Enrollment was 50, with 25 participants listed in each group.

The primary endpoint was the percentage with investigator-assessed objective response under RECIST version 1.1. The response estimates were 60.0% and 52.0%, respectively, with broad 95% confidence intervals. The posted risk-difference estimate was -0.08, with a 95% confidence interval from -0.354 to 0.194 and an unadjusted p-value of 0.569.

The record also lists secondary efficacy and safety outcomes. Important details—including the original sample-size rationale, multiplicity procedures, missing-data handling, and further analysis-plan information—are not reported in the registry record.