E3 Breast Cancer Taxotere Combination: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for E3 Breast Cancer Taxotere Combination (NCT00494481), a randomized phase 2 study. The primary outcome is Number of Patients With a Disease Progression Event.

2. Clinical Question

This is a question about the observed proportion of participants experiencing the specified event during the reported assessment process. The primary endpoint combines objective tumour progression with death from any cause when objective progression has not occurred. It does not separate those two event types in the posted group totals. Nor does the registry record give a participant-level time-to-event analysis for this endpoint.

The registry lists vandetanib plus docetaxel for 35 participants and placebo plus docetaxel for 29 participants in the primary outcome. The reported event counts are 24 and 18, respectively. The estimated risk difference compares the event proportion in the combination group with the event proportion in the placebo group; its interval includes values on either side of no difference. The result therefore describes an uncertain contrast in this enrolled group, rather than establishing the effect for all people with advanced breast cancer.

3. Trial Design

The registry record describes a phase II, randomized, parallel-group study with quadruple masking. It was placebo-controlled and compared vandetanib plus docetaxel with placebo plus docetaxel. The trial is listed as completed, with an enrollment of 64. The recorded start is 2006-01 and the primary completion is 2007-06.

The two posted primary-outcome groups contain 35 participants assigned to vandetanib plus docetaxel and 29 assigned to placebo plus docetaxel. These group counts sum to the recorded enrollment, but the registry record does not explain whether every enrolled participant was included in the primary analysis or how the analysis population was defined. The population field for the primary outcome is blank, so an intention-to-treat, per-protocol, or other analysis set is not specified.

Quadruple masking indicates that the registry categorized the study as masking four parties, but the record does not identify those parties or describe how masking was implemented. Random allocation is recorded; details such as sequence generation, allocation concealment, stratification, or block design are not reported in the registry record. The original sample-size justification and the trial's design assumptions are also not reported in the registry record.

The primary outcome is the number of participants with an objective disease progression event or death in the absence of objective progression. RECIST tumour assessments were carried out at screening, within 3 weeks before the first dose, and then according to site clinical practice until objective progression. The only additional mandatory RECIST assessment was at the data cut-off. The record does not give the exact data cut-off date or a fixed follow-up duration for all participants. It also does not report posted secondary outcomes or analysis records. These omissions limit what can be inferred about the full statistical analysis plan and the completeness of follow-up.

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:

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 Number of Patients With a Disease Progression Event, measured over RECIST tumour assessments carried out at screening (within 3 weeks before the 1st dose) and then as per site clinical practice until objective progression. The only additional mandatory RECIST assessment is at the point of data cut-off.

The primary endpoint counts participants with objective disease progression or death from any cause before objective progression. Its unit is participants, and the posted outcome is a count for each treatment group. The endpoint therefore treats either of these events as a progression event; the registry totals do not distinguish how many events arose from objective progression and how many from death.

Progression-event counts provide a direct summary of how many participants experienced the defined event within the observation represented by the record. In general, a count or proportion can be easier to interpret than a model-based measure requiring assumptions about event timing. However, this endpoint does not by itself describe when progression occurred, how long participants remained free of progression, or the pattern of repeated events. The posted comparison is not a time-to-event estimate.

The assessment schedule matters for interpreting the endpoint. Screening assessments were performed within 3 weeks before the first dose, subsequent RECIST assessments followed site clinical practice until objective progression, and one additional mandatory assessment occurred at data cut-off. Because the schedule after screening was guided by site practice, the record does not establish that participants had identical assessment intervals. The precise observation period and handling of participants without an observed event are not reported in the registry record.

As a result, the endpoint is best read as a count of participants meeting the registry definition under the recorded assessment process. It should not be interpreted as a complete account of progression timing or as a separately reported measure of objective progression alone.

7. Results Posted in the Registry

• Vandetanib Plus Docetaxel: 68.6% (24/35) versus 62.1% (18/29) on control; risk difference 6.5 percentage points (95% CI -16.9 to 29.9), risk ratio 1.10 (0.77–1.59); unadjusted two-proportion z-test p = 0.586.

8. Statistical Interpretation

9. Statistical Methods Explained

For a binary endpoint, each participant is classified according to whether the defined event occurred during the relevant observation. The registry endpoint is a participant count: objective disease progression or death from any cause in the absence of objective progression. A group count can be expressed as a proportion by relating the number with an event to the group size. The posted comparison includes the group proportions, a risk difference, a confidence interval, and an unadjusted p-value.

What the risk difference represents

A risk difference is the event proportion in one group minus the event proportion in the comparison group, using a stated direction. It is expressed on the probability scale rather than as a relative measure. A value of zero corresponds to equal event proportions under that comparison. The posted estimate is 0.065. Its interval ranges from -0.169 to 0.299, crossing zero. This means the estimated contrast is not precise enough to rule out either direction of difference within the reported interval. The confidence level for this interval is not reported in the registry record.

Other effect measures are also listed in the comparison statistics, including a risk ratio and an odds ratio with their intervals. These measures use different scales and answer related but distinct questions. A risk ratio compares the event proportions as a ratio; an odds ratio compares the odds of the event. Their interpretation depends on the event and on the particular measure. The posted estimates are 1.105 for the risk ratio and 1.333 for the odds ratio, with intervals of 0.769 to 1.587 and 0.473 to 3.755, respectively. Each interval includes the measure's no-association value: one. These relative measures do not change the basic conclusion that the comparison is imprecise.

What a confidence interval conveys

A confidence interval gives a range of values compatible with the observed data under the analysis method and its assumptions. It is not a guarantee that the true value lies within the interval, and it is not a range of individual participant outcomes. In general, wider intervals indicate less precision. Precision depends on factors such as the amount of information in the data and the variability of the outcome. Here, the posted risk-difference interval covers negative and positive values, so the data do not tightly locate the contrast around a single direction or magnitude.

How to read the p-value

The unadjusted p-value is 0.586. In general, a p-value describes the probability, under a specified null model and test assumptions, of obtaining results at least as inconsistent with that model as the observed results. It is not the probability that the null hypothesis is true, and it does not measure the probability that a finding is due to chance. The registry record does not state the exact test used, its assumptions, or whether any analysis other than the posted unadjusted comparison was performed. The p-value should therefore be read as a reported summary, not as a complete account of the analysis.

Why event counts are not time-to-event results

The registry describes a progression event count methodology and reports numbers of participants with an event. A time-to-event analysis uses both whether an event occurred and when it occurred, and may account for participants who have not experienced the event by the end of observation. The posted primary result does not report a time-to-event estimate, a hazard ratio, or a survival curve. The endpoint assessment process included screening, assessments according to site clinical practice until objective progression, and an additional mandatory assessment at data cut-off. The record does not give enough detail to reconstruct event times or a time-to-event analysis.

What over-dispersion means in general

Over-dispersion is a general concern when count data vary more than a chosen statistical model expects. If a model assumes less variability than the data contain, its uncertainty estimates may be too small. This concept is relevant to some count-based analyses, but the posted primary outcome is a count of participants with a binary event, and the registry record does not report an over-dispersion assessment or a count-model analysis. It would not be appropriate to assume that such a method was used.

The record lists no analysis records and does not provide a detailed statistical analysis plan. It does not report model covariates, adjustment methods, missing-data procedures, prespecified subgroup analyses, or a confidence-level specification. The posted summaries can be interpreted on their own terms, but these omissions prevent a fuller audit of the analysis choices.

10. Confidence Intervals and Precision

The risk-difference estimate for vandetanib plus docetaxel compared with placebo plus docetaxel is 0.065. Its reported interval is -0.169 to 0.299. The interval includes zero, as well as estimates in both negative and positive directions. It therefore does not narrowly identify the direction or size of the group contrast. The confidence level associated with this interval is not reported in the registry record.

The posted risk ratio is 1.105, with an interval from 0.769 to 1.587. The odds ratio is 1.333, with an interval from 0.473 to 3.755. Both intervals include one, the no-association value for these relative measures. The different measures have different scales, so their numerical values should not be compared as if they meant the same quantity. Their intervals nevertheless convey the same broad limitation: the estimates are uncertain.

Confidence intervals are useful because they show a range of values compatible with the observed data under the analysis method. In general, narrower intervals indicate greater precision, while wider intervals indicate less precision. An interval does not describe the range of outcomes among participants, and it does not mean that every value inside it is equally likely. Its interpretation also depends on the method and assumptions used to construct it, details that are not fully reported for this comparison.

The unadjusted p-value is 0.586. It does not replace the interval: the interval displays uncertainty on the effect scale, while the p-value summarizes evidence relative to a null model. The relatively small group sizes in the primary outcome, together with the wide reported intervals, counsel against treating the point estimate as a precise estimate of the underlying difference. The registry record does not provide a separate precision analysis or the confidence level.

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 sample-size figures are planning illustrations based on fractions of the observed effect, not the original enrollment target or a report of what the trial was designed to detect. For this binary outcome, the listed values are participants per arm, not total events. The illustration assumes a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Each planning row applies the stated fraction of the observed effect as the effect to detect. The trial's own original design assumptions and sample-size justification are not reported in the registry record.

The 80% power column gives the number of participants per arm that the illustration estimates would be needed to have an 80% chance of detecting the chosen effect under its assumptions, if that effect were the true one. The 90% power column gives the corresponding number per arm for a 90% chance. Power is a property of a planned design under specified assumptions; it is not the probability that a particular result is correct.

For the full observed effect, the illustration gives 840 participants per arm for 80% power and 1,124 participants per arm for 90% power. For planning for 75% of the observed effect, it gives 1,510 participants per arm for 80% power and 2,021 participants per arm for 90% power. For planning for 50% of the observed effect, it gives 3,433 participants per arm for 80% power and 4,596 participants per arm for 90% power. Smaller effects require more participants because they are harder to distinguish from random variation under the same assumptions. The curve should be read as illustrating this general relationship between the chosen effect size and required sample size, not as evidence about the trial's observed outcomes.

This is not observed power. It is a forward-looking planning exercise using the observed effect as its starting point. Because the observed effect is itself uncertain, the resulting sample-size figures should not be treated as a definitive recommendation or as proof that the completed trial had a particular probability of detecting an effect.

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 = 0.621; treatment proportion p₁ = 0.686; average p̄ = 0.653.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6731 = 1.3192.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6715 = 0.5652.
4. n = (1.3192 + 0.5652)² / (0.0650)² = 839.8, rounded up to 840 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 <- 0.620689655172414                          # control proportion
p1 <- 0.685714285714286                          # 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(494481)
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 = 840 (80% power), 1,124 (90% power)
 75% of observed effect: n per arm = 1,510 (80% power), 2,021 (90% power)
 50% of observed effect: n per arm = 3,433 (80% power), 4,596 (90% power)
power.prop.test: 839.8 
Simulated power at n = 840 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 record lists one outcome measure and no posted secondary outcomes. It also lists no analysis records. These entries do not establish that no other outcomes or analyses were planned or conducted; they mean that additional outcomes and analysis details are not reported in the registry record. The statistical plan for controlling multiplicity, if there were multiple outcomes, time points, subgroups, or interim looks, is not reported in the registry record.

Multiplicity is a general concern when many statistical comparisons are made. The more opportunities there are to find a nominally unusual result, the harder it can be to interpret an individual p-value without knowing which comparisons were prespecified and how they were handled. Here, no multiplicity adjustment is described, but the record also does not provide the information needed to determine whether one was relevant. The unadjusted p-value should be understood as the posted comparison, not as evidence that all possible analyses were considered.

The primary outcome population field is blank. The record does not identify whether the comparison used all randomized participants, participants who received treatment, participants with assessable tumour evaluations, or another analysis set. It also does not report how missing assessments, withdrawals, or participants with incomplete follow-up were handled. The registry gives the primary outcome group sizes, but those counts alone do not establish the analysis rules or the completeness of endpoint data.

Missing data can affect interpretation if the likelihood of an unobserved outcome differs across participants or treatment groups. In general, methods for handling missing outcomes depend on the amount and pattern of missingness and on assumptions about why data are absent. No missing-data method or account of missing primary outcome data is reported in the registry record, so the effect of missingness cannot be evaluated from the posted summary.

The record also does not report interim analyses, stopping rules, covariate adjustment, subgroup methods, or sensitivity analyses. These details should not be inferred from the phase, design, or outcome label. The posted counts and comparison statistics remain interpretable as registry results, but the analysis population and several safeguards needed to assess the full statistical workflow are not documented.

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: neutropenia (12.1%), dyspnoea (10.3%), febrile neutropenia (9.1%), diarrhoea (9.1%), neutropenic sepsis (6.9%).

The registry lists serious adverse events for 14 of 33 participants at risk in the vandetanib-plus-docetaxel group and 12 of 29 participants at risk in the placebo-plus-docetaxel group. It separately lists other adverse events for 33 of 33 participants at risk and 25 of 29 participants at risk, respectively. These are descriptive group counts. They should be read alongside the listed number at risk for each category, which differs from the primary outcome group size in the vandetanib-plus-docetaxel group.

The record lists the following serious adverse event terms and percentages: Neutropenia, 12.1%; Dyspnoea, 10.3%; Febrile Neutropenia, 9.1%; Diarrhoea, 9.1%; Neutropenic Sepsis, 6.9%; and Granulocytopenia, 3.4%. The registry record does not specify the group attribution or denominator for these term percentages in the displayed information, so they should not be assigned to a particular treatment group or reinterpreted as comparative rates.

These summaries do not give event timing, duration, severity grading beyond the serious-event classification, or a statistical comparison between groups. Death counts are not reported in the registry record. The adverse-event figures describe what is posted and do not, by themselves, establish a causal relationship between a treatment and an event.

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

The trial addresses a focused statistical question: how the number of participants with objective progression or death compared between vandetanib plus docetaxel and placebo plus docetaxel. The posted result gives 24 events among 35 participants in the combination group and 18 among 29 in the comparator group. The estimated risk difference is 0.065, with an interval from -0.169 to 0.299, and the unadjusted p-value is 0.586.

These results show an observed difference in the event proportions, but the uncertainty is broad. The interval includes zero and values on both sides of zero; the data therefore do not precisely determine the direction or magnitude of the underlying contrast. The p-value is also not evidence that the groups are equivalent. A result that does not distinguish clearly between groups is not the same as a demonstration that the groups have identical outcomes.

The endpoint combines objective progression and death in the absence of objective progression. Since the record does not separate those components, the posted counts cannot establish which component contributes to the group totals. The endpoint also does not provide information about when events occurred. It should not be read as a time-to-event result, a survival estimate, or an outcome for participants who did not have a recorded event.

The registry record does not specify the analysis population, the handling of incomplete assessments, the confidence level, or the detailed statistical methods. These omissions constrain interpretation of the precision and robustness of the comparison. The record also contains no posted secondary outcomes or analysis records, so conclusions about other clinical outcomes are not available from this entry.

In short, the posted statistics characterize an uncertain comparison in this randomized phase II study. They do not establish a definitive treatment effect, equivalence, or a conclusion about outcomes beyond the registered primary endpoint. The figures should be understood in the context of the recorded group sizes, endpoint definition, assessment process, and limitations in the analysis details.

16. Why This Trial Matters Statistically

This trial illustrates how a randomized comparison can be described with a simple binary endpoint while still leaving important questions about precision and interpretation. The endpoint counts participants with a progression event, defined as objective progression or death before objective progression. The group totals are readily interpretable, but they combine two event types and do not show when events occurred.

The point estimate and its uncertainty should be read together. The risk difference is 0.065, but the interval from -0.169 to 0.299 extends across zero and spans a broad range of possible contrasts. The unadjusted p-value of 0.586 is not a substitute for that interval and does not prove equivalence. This is a useful example of why an observed point estimate should not be treated as a precise or definitive treatment contrast.

The registry record also demonstrates the importance of documenting the analysis population and follow-up. The outcome assessment schedule allowed RECIST assessments according to site clinical practice until objective progression, with an additional mandatory assessment at data cut-off. Without a fixed schedule for all participants, a stated analysis population, and details on incomplete assessments, readers cannot fully assess the relationship between observation and the reported counts.

Finally, planning calculations based on an observed effect are not evidence about the original trial design or its achieved power. The registry's original sample-size assumptions are not reported, while the planning illustration explicitly uses the observed effect and specified assumptions. Distinguishing a retrospective planning illustration from an original design calculation helps prevent an overconfident interpretation of small-trial results.

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.

The registry record has several limitations that are specific to interpreting this trial. Although the study enrollment is 64, the primary outcome groups are 35 and 29, and the record does not describe the analysis population or explain how the outcome totals relate to all randomized participants. It does not report how participants with missing or incomplete assessments were handled.

The endpoint combines objective progression and death from any cause in the absence of objective progression. The posted totals do not separate these components. The assessment schedule is partly dependent on site clinical practice, and the record does not give a common fixed follow-up duration or the exact data cut-off date. Consequently, the group counts do not show event timing or establish identical observation opportunities for all participants.

The registry lists the study as randomized and quadruple-masked, but does not describe allocation concealment, the masking procedures, or which parties were masked. It does not report the original sample-size justification, the detailed statistical analysis plan, the confidence level for the reported intervals, or the test used for the unadjusted p-value. No analysis records or secondary outcomes are posted. The record also does not report interim analyses, multiplicity procedures, covariate adjustment, or sensitivity analyses.

Safety reporting is also limited in the posted information. The at-risk counts differ by adverse-event category and are not fully aligned with the primary outcome group sizes. The serious-event term percentages are listed without their group attribution or denominators in the available record, and death counts are not reported in the registry record. These details limit direct comparison or broader interpretation of the safety summaries.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: E3 Breast Cancer Taxotere Combination — NCT00494481.
• Registry record last updated 2016-09-30; status last verified 2016-08.
• Results first posted 2011-05-24.
• 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 NCT00494481, a completed phase II randomized, parallel-group, quadruple-masked, placebo-controlled study of vandetanib plus docetaxel versus placebo plus docetaxel for advanced breast cancer. Enrollment was 64. The recorded start was 2006-01, and primary completion was 2007-06.

The primary endpoint was the number of participants with objective disease progression or death in the absence of objective progression. The posted counts were 24 of 35 for vandetanib plus docetaxel and 18 of 29 for placebo plus docetaxel. The risk difference was 0.065, with an interval from -0.169 to 0.299; the unadjusted p-value was 0.586.

The registry lists no secondary outcomes or analysis records. The primary analysis population, missing-data procedures, original sample-size justification, confidence level, and detailed statistical plan are not reported in the registry record.