Study In Women And Men With Metastatic Breast Cancer That Have Overexpression Of ErbB2: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study In Women And Men With Metastatic Breast Cancer That Have Overexpression Of ErbB2 (NCT00281658), a randomized phase 3 study. The primary outcome is Overall Survival (OS) at 53 Months.

2. Clinical Question

3. Trial Design

This was a randomized, multicenter, double-blind, placebo-controlled Phase III study with a parallel-group design. The registry lists enrollment of 444 and reports that the study is completed. The trial began on 2006-01-02, and primary completion was recorded as 2010-06-18. The sponsor was Novartis Pharmaceuticals.

The principal comparison was lapatinib 1500 mg plus paclitaxel against placebo plus paclitaxel. Randomization assigned participants to the treatment on which the primary Intent-to-Treat analysis was based. The registry describes that population as all randomized subjects, analyzed according to randomized treatment. The primary outcome was overall survival, with the primary analysis cut-off on 18-Jun-2010 and assessment up to 53 months.

The record also lists open-label monotherapy and combination-therapy extension arms. Their presence in the arm descriptions does not establish the rules for entering an extension or how participants were selected; those details are not reported in the registry record. Safety summaries include these extension arms as well as the blinded groups, so their figures should not be mistaken for additional randomized comparison groups.

The registry lists 13 outcome measures and 10 analyses. It does not report the original sample-size justification or the full statistical analysis plan in the record. The planning illustrations associated with the results are not evidence of the assumptions used to design or conduct the trial.

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.

Randomized Phase

Reasons for not completing this period:

Open Label Phase

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent-to-Treat (ITT) Population. The ITT population was comprised of all randomized subjects and was based on the treatment to which the subject was randomized.

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 Overall Survival (OS) at 53 Months, measured over From date of randomization until date of death from any cause, assessed up to 53 months (Primary OS analysis cut-off date = 18-Jun-2010).

The primary endpoint was median overall survival, measured in months from randomization until death from any cause. Participants were followed for this outcome up to 53 months for the primary analysis, whose cut-off date was 18-Jun-2010. The registry defines the analysis population as the Intent-to-Treat population, including all randomized subjects and retaining their randomized treatment assignment for analysis.

Overall survival is a time-to-event outcome: it includes both whether death occurred during follow-up and the timing of that event. A median gives a central time point in the observed survival distribution, rather than an average lifetime or a prediction for an individual. The accompanying hazard ratio describes the relative rate of experiencing the event over the follow-up period under the sponsor’s analysis. These measures complement one another but are not interchangeable.

General statistical context: survival endpoints are useful in trials because they preserve information about event timing and can incorporate follow-up that differs among participants. Interpretation still depends on how the endpoint was defined, the analysis population, the follow-up period, and the assumptions underlying the analysis. Overall survival also counts death from any cause, as specified in this record, rather than limiting the event to a particular cause. The endpoint alone does not explain the mechanisms behind a group difference or establish the balance of benefits and harms.

7. Results Posted in the Registry

• Lapatinib 1500 mg + Paclitaxel vs Placebo + Paclitaxel: hazard ratio 0.74 (95% CI 0.58–0.94), reported p = 0.006 (as reported in the registry).

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Overall Survival (OS) at 190 Months

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor; it was not recalculated.
• Result:
  • Lapatinib 1500 mg + Paclitaxel vs Placebo + Paclitaxel: hazard ratio 0.80 (95% CI 0.64–0.98) (as reported in the registry).
• Interpretation:
  • The 95% confidence interval excludes no difference (a ratio of 1), so the hazard of the event is statistically lower for Lapatinib 1500 mg + Paclitaxel vs Placebo + Paclitaxel than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Progression-free Survival (PFS) by Investigator Assessment

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor; it was not recalculated.
• Result:
  • Lapatinib 1500 mg + Paclitaxel vs Placebo + Paclitaxel: hazard ratio 0.54 (95% CI 0.44–0.66) (as reported in the registry).
• Interpretation:
  • The 95% confidence interval excludes no difference (a ratio of 1), so the hazard of the event is statistically lower for Lapatinib 1500 mg + Paclitaxel vs Placebo + Paclitaxel than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Overall Response Rate (ORR) by Investigator Assessment

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Lapatinib 1500 mg + Paclitaxel: 68.9% (153/222) versus 50.0% (111/222) on control; risk difference 18.9 percentage points (95% CI 10.0 to 27.9), risk ratio 1.38 (1.18–1.62); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Lapatinib 1500 mg + Paclitaxel than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The secondary results with numerical comparisons generally point in the same direction as the primary overall-survival result: the lapatinib combination has estimates favoring it for investigator-assessed progression-free survival and overall response rate, and the final overall-survival estimate remains in the same direction as the primary analysis. These outcomes address different aspects of the clinical course. Progression-free survival concerns progression or death, while response rate concerns tumor response; neither is identical to survival.

That overall pattern can be viewed as context for the primary result, but it does not make the secondary findings independent confirmations. Outcomes in the same trial may be related, and each additional comparison creates another opportunity for a chance finding. The registry does not report a multiplicity adjustment plan for the secondary analyses. Without that information, secondary p-values or estimates should receive less inferential weight than the prespecified primary question, especially when the complete hierarchy of outcomes and tests is not available.

The registry lists additional outcome measures, including clinical benefit, duration of response, response at specified follow-up assessments, and tumor biomarker evaluations. Numerical findings for all listed outcomes are not included in the posted results summarized in the record. Their absence from the results should not be interpreted as evidence of no effect. Nor can an overall conclusion about every secondary endpoint be drawn from the subset with numerical estimates. The main statistical takeaway is directional consistency among the reported outcomes, paired with less certainty about their confirmatory status and incomplete numerical reporting across the full outcome list.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Overall Survival (OS) at 190 Months
• Progression-free Survival (PFS) by Investigator Assessment
• Overall Response Rate (ORR) by Investigator Assessment
• Clinical Benefit Rate (CBR)
• Duration of Response (DOR)
• Number of Participants With a CR or PR at Weeks 8, 12, 16, 24, 32, 40, 48, 56, 64, and 72
• Number of Tumors Evaluable for PIK3CA Mutations
• Number of Participants With Tumors Evaluable for PTEN

9. Statistical Interpretation

10. Statistical Methods Explained

What does the hazard ratio describe?

The primary result is a time-to-event comparison. A hazard ratio summarizes the relative rate of the event—death in this overall-survival analysis—between the randomized groups over follow-up, as estimated by the sponsor’s analysis. A value below 1 indicates a lower estimated event rate in the lapatinib combination group relative to the placebo combination group under that analysis. It is not a ratio of median survival times, nor does it mean that every participant’s risk changed by that amount. Its interpretation relies on the analysis framework and assumptions; the full model details are not reported in the registry record.

How should the medians be read?

The median is the time by which half of the population represented by the survival distribution has experienced the event, accounting for the time-to-event analysis. It is a summary of the distribution, not an estimate of the average time lived by every participant. Medians can be useful for communicating survival experience, but they do not describe the entire shape of the survival curves. The registry reports median overall survival and confidence intervals for each arm, along with the sponsor’s hazard-ratio comparison.

What does the Log Rank result add?

The primary comparison names the Log Rank method and reports p = 0.006. In general, a Log Rank test compares time-to-event experience between groups across follow-up. A p-value describes how compatible the observed data are with a null comparison under the test’s assumptions; it is not the probability that the null hypothesis is true, nor a measure of clinical importance. The record does not report the full test specification, any stratification factors, or whether other analyses were used alongside this comparison.

How do the binary outcome summaries differ?

Overall response rate is reported as a percentage of participants, with a sponsor-reported odds ratio. The posted unadjusted re-calculation uses a risk difference and its confidence interval. These are different ways to express a binary comparison: a risk difference expresses a difference in the proportions, while an odds ratio compares odds. Odds ratios and risk differences are not interchangeable, and their numerical values should not be directly compared as if they shared a scale. The re-calculated p-value is unadjusted; the registry does not report a p-value for the sponsor’s odds-ratio analysis.

Why do analysis populations matter?

The primary overall-survival analysis is explicitly described as Intent-to-Treat. This keeps the randomized assignment as the basis for comparison and includes all randomized subjects in the named population. The registry gives arm-level denominators for posted results, but it does not report detailed information about censoring, missing outcome data, protocol deviations, or handling of incomplete assessments. Those details matter because they affect how the observed follow-up contributes to estimates and how broadly the results can be interpreted.

What is not established by a reported test?

A statistically significant result does not on its own measure the magnitude or practical importance of an effect. Effect estimates and confidence intervals provide information about magnitude and precision, while endpoint relevance and the trial population provide context. The registry does not report a multiplicity adjustment plan for the secondary outcomes, the original statistical analysis plan, or interim-analysis details. Consequently, secondary findings should be treated as less definitive than the stated primary analysis unless additional prespecified analysis information is available. General statistical interpretation cannot fill in those missing trial-specific details.

11. Confidence Intervals and Precision

The primary overall-survival estimate is accompanied by a 95% confidence interval for the hazard ratio: 0.58 to 0.94 around the reported estimate of 0.74. The interval communicates the uncertainty associated with the estimate under the analysis method. It is compatible with a range of relative event-rate values below 1, rather than identifying a single exact value. The interval does not describe the range of outcomes for individual participants.

The median estimates also have 95% confidence intervals. For the lapatinib combination, the primary median was 27.8 months, with an interval from 23.2 to 32.2 months. For placebo plus paclitaxel, the median was 20.5 months, with an interval from 17.9 to 24.3 months. These intervals express uncertainty in each arm’s median; they are not the confidence interval for a direct difference between medians.

General statistical context: a confidence interval is a range produced by a statistical procedure that reflects sampling uncertainty under its assumptions. A narrower interval usually indicates greater precision for the quantity being estimated, while a wider interval indicates more uncertainty. Interval width is not a verdict on whether a result is important, and overlap between separate arm intervals is not a substitute for the formal comparison.

The final overall-survival hazard ratio was 0.80, with a 95% confidence interval from 0.64 to 0.98. Its limits remain below 1, but the interval includes values closer to no relative difference than the primary interval does. The registry reports no p-value for that final estimate. Precision should therefore be read from the reported interval and not supplemented with an unreported test result.

12. 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 illustrations translate assumed effects into the event counts or participant counts needed for a hypothetical study; they do not describe the observed enrollment. For time-to-event outcomes, the entries are total events, not participants per arm. For the binary outcome, the entries are participants per arm. The displayed effect scenarios use 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect.

The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The event-based calculations use the Schoenfeld approach. These are planning illustrations computed from the observed effect, not the original trial’s design assumptions; the trial’s own original design assumptions are not reported in the registry record.

The 80% power column means the hypothetical design is planned to have an 80% chance of detecting the stated effect under those assumptions if that effect is present. The 90% power column uses a higher target chance of detecting it, and therefore generally calls for more information. The required information is expressed as total events for time-to-event outcomes and participants per arm for the binary comparison. As the assumed effect becomes smaller, it is harder to distinguish from random variation, so the planning illustration requires more participants or events.

These figures are not observed power. They are prospective planning illustrations based on effect scenarios, not a measure of how likely the completed trial was to find a significant result after observing its data. They also cannot establish whether the actual trial was adequately planned, since the original design assumptions are not reported in the registry record.

13. 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

d is the total number of events needed (Schoenfeld's formula) with 1:1 allocation and proportional hazards.

Worked calculation for this trial

1. Hazard ratio HR = 0.740, so ln(HR) = -0.3011.
2. d = 4 × (1.96 + 0.842)² / [ln(0.740)]² = 346.3, rounded up to 347 events in total 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
hr <- 0.74                              # observed hazard ratio
events_needed <- function(power, frac = 1) ceiling(round(4 * (z_a + qnorm(power))^2 / log(hr^frac)^2, 6))   # Schoenfeld, 1:1 allocation
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: events = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(events_needed(0.80, f), big.mark = ","), format(events_needed(0.90, f), big.mark = ",")))

# Simulation: given d total events under proportional hazards, events in the treatment arm are Binomial(d, hr / (1 + hr)).
# The score (log-rank type) statistic is (D1 - d / 2) / sqrt(d / 4).
set.seed(281658)
d    <- events_needed(0.80)
reps <- 20000
D1 <- rbinom(reps, d, hr / (1 + hr))
z  <- (D1 - d / 2) / sqrt(d / 4)
cat(sprintf("Simulated power with %s events: %.3f (%d replicates)\n", format(d, big.mark = ","), mean(abs(z) > z_a), reps))

Output from running the code

100% of observed effect: events = 347 (80% power), 464 (90% power)
 75% of observed effect: events = 616 (80% power), 825 (90% power)
 50% of observed effect: events = 1,386 (80% power), 1,855 (90% power)
Simulated power with 347 events: 0.810 (20000 replicates)

The simulation uses the fact that, with d events in total and proportional hazards, the number of events in the treatment arm is binomial. It is an approximation that conditions on the total number of events, and it should land close to the target power.

14. Group-Sequential Design Analysis

Many time-to-event trials, especially in oncology, stop or report early when interim data are overwhelming, so a single fixed-sample number understates the real design problem. The calculation below is an illustrative group-sequential design built from this trial's posted control median and hazard ratio. The registry record does not state the planned number of looks or the spending function, so the design is a planning illustration, not a reconstruction of the protocol. Assumed: one-sided α = 0.025, 90% power, three equally spaced looks on information, Lan-DeMets O'Brien-Fleming spending, 24 months of accrual, 60 months in total and a 1% monthly dropout hazard.

Under these assumptions the design needs about 749 participants in total, with the boundaries below.

Fixed design versus group-sequential alternatives

Group-sequential designs cost a small increase in maximum sample size and events compared with the fixed design, in exchange for the chance to stop early for overwhelming efficacy. O'Brien-Fleming spending keeps early boundaries very conservative and the final boundary close to the fixed-design value, whereas Pocock spending spends more alpha early.

Sensitivity to a smaller true effect

Planning for exactly the observed effect is optimistic, because observed effects in published trials tend to be somewhat larger than the true effect. The table repeats the main design (three looks, O'Brien-Fleming spending, 90% power) for smaller true hazard ratios.

For context, this trial enrolled 444 participants.

Cross-check, upper boundaries: gsDesign 3.71, 2.511, 1.993 | rpact 3.71, 2.511, 1.993. Agreement between the two packages confirms the boundary calculation.

R code and output

suppressWarnings(suppressMessages({ library(gsDesign); library(rpact) }))
med_control <- 20.5          # control-arm median, months (converted from the posted unit)
hr          <- 0.74          # hazard ratio used for planning
lambdaC     <- log(2) / med_control
eta         <- 0.01               # dropout hazard per month (assumption)
acc_months  <- 24                 # accrual duration (assumption)
study_months<- 60                 # total study duration (assumption)

design <- function(k, sfu, beta, h = hr) gsSurv(k = k, test.type = 1, alpha = 0.025, beta = beta, sfu = sfu,
  lambdaC = lambdaC, hr = h, hr0 = 1, eta = eta, gamma = 1, R = acc_months, T = study_months, minfup = study_months - acc_months)

# Main design: 3 looks, Lan-DeMets O'Brien-Fleming spending, one-sided alpha 0.025, 90% power
x <- design(3, sfLDOF, 0.10)
cat(sprintf("N_TOTAL|%s\n", format(ceiling(sum(x$gamma * x$R)), big.mark = ",")))
for (i in 1:3) cat(sprintf("GS|%d|%.1f|%.0f|%.3f|%.5f|%.5f\n", i, x$T[i], x$n.I[i], x$upper$bound[i], pnorm(-x$upper$bound[i]), cumsum(x$upper$spend)[i]))
cat("\n"); print(gsBoundSummary(x))

# Comparison of designs
cmp <- list(list("Fixed design, one analysis", 1, sfLDOF), list("Two looks, O'Brien-Fleming", 2, sfLDOF),
            list("Three looks, O'Brien-Fleming", 3, sfLDOF), list("Three looks, Pocock", 3, sfLDPocock))
cat("\n")
for (b in c(0.20, 0.10)) for (d in cmp) {
  y <- design(d[[2]], d[[3]], b)
  cat(sprintf("DES|%s|%d|%.0f|%s|%.1f\n", d[[1]], round(100 * (1 - b)), tail(y$n.I, 1), format(ceiling(sum(y$gamma * y$R)), big.mark = ","), tail(y$T, 1)))
}

# Sensitivity: what if the true effect is smaller than the observed one? (3 looks, O'Brien-Fleming, 90% power)
cat("\n")
for (fr in c(1, 0.75, 0.5)) {
  y <- design(3, sfLDOF, 0.10, exp(log(hr) * fr))
  cat(sprintf("SENS|%.2f|%.3f|%.0f|%s\n", fr, exp(log(hr) * fr), tail(y$n.I, 1), format(ceiling(sum(y$gamma * y$R)), big.mark = ",")))
}

# Independent cross-check of the boundaries with rpact
r <- getDesignGroupSequential(kMax = 3, alpha = 0.025, beta = 0.10, sided = 1, typeOfDesign = "asOF", informationRates = x$timing)
cat(sprintf("\nCross-check, upper boundaries: gsDesign %s | rpact %s\n", paste(round(x$upper$bound, 3), collapse = ", "), paste(round(r$criticalValues, 3), collapse = ", ")))

Method: LachinFoulkes 
    Analysis               Value Efficacy
   IA 1: 33%                   Z   3.7103
      N: 656         p (1-sided)   0.0001
 Events: 157        ~HR at bound   0.5521
   Month: 21    P(Cross) if HR=1   0.0001
             P(Cross) if HR=0.74   0.0338
   IA 2: 67%                   Z   2.5114
      N: 750         p (1-sided)   0.0060
 Events: 313        ~HR at bound   0.7525
   Month: 34    P(Cross) if HR=1   0.0060
             P(Cross) if HR=0.74   0.5603
       Final                   Z   1.9930
      N: 750         p (1-sided)   0.0231
 Events: 469        ~HR at bound   0.8318
   Month: 60    P(Cross) if HR=1   0.0250
             P(Cross) if HR=0.74   0.9000



Cross-check, upper boundaries: gsDesign 3.71, 2.511, 1.993 | rpact 3.71, 2.511, 1.993

15. Analyses Reported by the Sponsor

The primary time-to-event comparison is reported by the sponsor as a hazard ratio of 0.74, with a 95% confidence interval from 0.58 to 0.94 and p = 0.006 by Log Rank. The final overall-survival analysis reports a hazard ratio of 0.80 with a 95% confidence interval from 0.64 to 0.98, without a p-value in the registry. For these time-to-event outcomes, the interpretation uses the sponsor-reported hazard ratios; they are not independent re-calculations from the posted median values.

For overall response rate, the sponsor reports an odds ratio of 2.30 with a 95% confidence interval from 1.54 to 3.47. The posted re-calculation instead expresses the comparison as an unadjusted risk difference and reports p < 0.001. Since the measures use different scales and analyses, their estimates need not match numerically. The re-calculated test should not be presented as the sponsor’s analysis or as an adjusted analysis.

The registry record does not indicate a discrepancy between the reported and re-calculated results. The sponsor’s reported analyses remain the trial’s own results; the re-calculations offer a separate, simple comparison based on posted summary statistics and do not replace the sponsor’s analysis plan.

16. Multiplicity, Missing Data and Analysis Populations

The registry record does not report how the trial handled multiplicity across secondary endpoints or multiple analyses. Multiplicity arises when several outcomes or comparisons are assessed: without an appropriate plan, the chance of at least one apparently notable result can increase. A prespecified hierarchy or adjustment method can help control that issue, but the relevant trial-specific procedure is not reported in the registry record. The secondary estimates should therefore be interpreted as supportive or exploratory context rather than assumed to have the same confirmatory status as the primary outcome.

The record does not report interim analyses or stopping rules. It also does not provide the full statistical analysis plan, including details about stratification, covariate adjustment, censoring rules, or sensitivity analyses. These omissions limit the ability to evaluate whether each analysis followed a prespecified approach or how robust the results were to alternative assumptions. They do not establish that such methods were absent; they mean the details are not reported in the registry record.

The primary population is identified as Intent-to-Treat and defined as all randomized subjects analyzed according to randomized assignment. This is an important point of clarity for the primary comparison. However, detailed information on missing data, loss to follow-up, censoring, and the handling of incomplete outcome assessments is not reported in the registry record. Such details are relevant to survival estimates because participants may have different lengths of observed follow-up and may be censored before an event is recorded.

For secondary outcomes, the analysis populations and missing-data handling are not specified in the available descriptions. In addition, the registry lists 13 outcome measures and 10 analyses but does not provide numerical findings for every listed outcome. This incomplete reporting and the absence of multiplicity details constrain broad claims across the outcome family. Statistical interpretation should distinguish the clearly named primary analysis from the less fully documented secondary analyses.

17. Safety Results

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

Deaths

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 (16.2%), ejection fraction decreased (5.9%), diarrhoea (4.5%), leukopenia (3.2%), febrile neutropenia (2.7%).

The serious adverse event summaries report 67 affected participants among 222 at risk in the lapatinib 1500 mg plus paclitaxel group, and 30 among 221 at risk in the placebo plus paclitaxel group. The registry also lists 10 deaths in the first group and 15 in the second in its adverse-event summary. These are counts in the posted safety record, not an adjusted comparison of risk, and the death counts should not be interpreted as deaths caused by treatment.

The most frequently listed serious adverse event terms are neutropenia at 16.2%, ejection fraction decreased at 5.9%, diarrhoea at 4.5%, leukopenia at 3.2%, febrile neutropenia at 2.7%, and granulocytopenia at 1.8%. The record’s listed terms and percentages should be read as descriptive safety information; the available summary does not establish a causal relationship or provide a formal comparative analysis for each term.

The extension-arm safety figures are also included in the registry, but the record’s arm descriptions do not make them equivalent to the randomized blinded comparison. Differences in the groups or their follow-up cannot be resolved from these counts alone. A full benefit-harm assessment would require more detail than these summary adverse-event results.

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

The posted primary analysis reports longer median overall survival in the lapatinib combination group than in the placebo combination group, alongside a hazard ratio below 1 and a 95% confidence interval below 1. The reported Log Rank p-value is 0.006. Taken together, these statistics describe a difference in the randomized groups’ observed time-to-death experience during the primary analysis follow-up, with uncertainty quantified by the reported interval.

This result does not mean that every participant lived longer, nor does it provide an individual prediction. A group-level median is not a promise about a particular person’s outcome. The hazard ratio is a summary of relative event rates, not a direct percentage change in survival time or an estimate of the number of additional months for each participant. The endpoint includes death from any cause, as defined in the record.

The findings apply to the trial population described: women and men with ErbB2-amplified metastatic breast cancer who had not received prior therapy for metastatic disease. The registry record does not support extending the result to people with other disease characteristics, treatment histories, or settings without further evidence. The record also does not provide all details needed to assess treatment delivery, missing follow-up, the complete analysis plan, or multiplicity across secondary outcomes.

Statistical evidence of a group difference is not the same as a complete assessment of clinical importance or safety. The safety summaries are descriptive, and the overall balance of outcomes requires considering both benefits and harms alongside patient circumstances and other evidence. These results describe the trial’s posted data; they do not constitute medical advice or establish that a treatment is appropriate for any individual.

19. Why This Trial Matters Statistically

This trial illustrates why time-to-event results should be read through more than one statistic. The median overall-survival estimates communicate a familiar time point, while the hazard ratio summarizes the relative event rate over follow-up. Confidence intervals add a view of precision, and the Log Rank p-value reports evidence from the named comparison. None of these quantities alone gives a complete account of the data.

The comparison also shows the importance of separating a primary endpoint from secondary outcomes. The primary overall-survival analysis is clearly identified, with its analysis population, cut-off, and sponsor-reported comparison. Secondary outcomes provide useful context about progression and response, but the registry does not report a multiplicity plan and does not provide numerical results for every listed outcome. This distinction helps prevent a collection of related outcomes from being treated as if each were an independent confirmatory test.

Finally, the record illustrates the difference between trial results and planning calculations. The sample-size illustrations translate observed effects into hypothetical event or participant requirements under stated assumptions. They do not recover the original design rationale or measure observed power. Keeping those purposes separate supports more careful interpretation of both completed-trial evidence and future study planning.

20. 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 does not report the original sample-size justification, so the planning illustrations cannot be used to determine the trial’s initial design assumptions. Nor does it report interim-analysis details, stopping rules, or the full statistical analysis plan. These gaps limit assessment of prespecification and the way the final analyses were selected.

For the primary outcome, the ITT population is defined, and the cut-off is given. However, detailed censoring rules, missing-data handling, loss to follow-up, and sensitivity analyses are not reported in the registry record. These details would help assess how incomplete follow-up affected the survival estimates. The record also does not provide the shape of the survival curves in the information, so the hazard ratio cannot be independently evaluated for the proportionality assumptions or time variation in effects.

The final overall-survival result has no reported p-value, and secondary outcomes have no reported multiplicity adjustment plan. Numerical findings are not available for every listed outcome. The serious adverse-event summary is descriptive and does not provide a formal comparative safety analysis; extension arms also appear in the safety listings without enough information to treat them as randomized comparisons. These limitations do not negate the posted primary result, but they narrow what can be concluded beyond the reported estimates and their stated analyses.

21. Related Tutorials

22. Related Statistical Calculators

23. Sources

• ClinicalTrials.gov: Study In Women And Men With Metastatic Breast Cancer That Have Overexpression Of ErbB2 — NCT00281658.
• Registry record last updated 2023-02-10; status last verified 2023-02.
• Results first posted 2011-06-01.
• 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.

24. Record Summary

The registry describes a completed, randomized, double-blind, placebo-controlled, multicenter Phase III study of lapatinib plus paclitaxel versus placebo plus paclitaxel in women and men with ErbB2-amplified metastatic breast cancer who had not received prior therapy for metastatic disease. Enrollment was 444, and the primary endpoint was overall survival in the Intent-to-Treat population.

At the primary analysis cut-off, the sponsor reported median overall survival of 27.8 months and 20.5 months, with a hazard ratio of 0.74, a 95% confidence interval from 0.58 to 0.94, and p = 0.006. The final overall-survival analysis reported a hazard ratio of 0.80 with a 95% confidence interval from 0.64 to 0.98, without a reported p-value. Secondary results include progression-free survival and response outcomes, while multiplicity and several analysis details are not reported in the registry record.