A Combination Study of Kadcyla (Trastuzumab Emtansine) and Capecitabine in Participants With Human Epidermal Growth Factor Receptor 2 (HER2)-Positive Metastatic Breast Cancer (mBC) or HER2-Positive Locally Advanced/Metastatic Gastric Cancer (LA/mGC): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Combination Study of Kadcyla (Trastuzumab Emtansine) and Capecitabine in Participants With Human Epidermal Growth Factor Receptor 2 (HER2)-Positive Metastatic Breast Cancer (mBC) or HER2-Positive Locally Advanced/Metastatic Gastric Cancer (LA/mGC) (NCT01702558), a non-randomized phase 2 study. The primary outcome is Phase 1 (mBC): Percentage of Participants With Dose-Limiting Toxicities (DLTs).

2. Clinical Question

3. Trial Design

The registry describes a sequential study with Phase 1 and Phase 2 components. Phase 1 assessed capecitabine dose levels in combination with T-DM1. It included an mBC cohort and a separate LA/mGC cohort. The stated Phase 1 aim was to assess the maximum tolerated dose of capecitabine with T-DM1; the mBC cohort’s primary endpoint was the percentage with dose-limiting toxicities during Cycle 1. The gastric cancer cohort is not part of that primary endpoint population.

The Phase 2 component concerned participants with mBC and listed two treatment groups: T-DM1 plus capecitabine at the maximum tolerated dose and T-DM1 alone. The registry’s official title describes this as randomized and open-label, but the allocation field identifies the study as non-randomized. The record also cautions that the arms are not a like-for-like control comparison. Without a consistent account of allocation, the Phase 2 groups should not be interpreted as a randomized causal comparison.

The registry lists enrollment of 182 and status as terminated. It gives a start date of 2012-12-03 and a primary completion date of 2017-05-31. The reason for termination is not reported in the registry record. The study was multicenter according to its summary. Treatment was anticipated to continue until disease progression, intolerable toxicity, withdrawal of consent, or study end; the arm descriptions also specify stopping conditions for Phase 2 treatment.

The record lists 34 outcome measures and one analysis. Those counts do not establish which analyses were prespecified or how the reported outcome measures were prioritized. A detailed statistical analysis plan, allocation procedures, and reasons for the discrepancy about randomization are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: A total of 234 participants were screened, out of which, 182 participants were enrolled into the study. Out of the 182 enrolled participants, 3 participants were excluded from all safety and efficacy analyses because they did not sign correct Informed Consent Form (ICF).

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Analysis was performed on Phase 1 DLT-evaluable population for mBC cohort, which included all enrolled and treated mBC participants who did not experience any major protocol deviation and completed Cycle 1.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: Safety population: Participants who received \>/=1 dose of study drug, analyzed as per actual treatment received. In Phase 2, of 161 participants, 1 was randomized in error(received no treatment) and was excluded from safety analysis and 1 who was randomized to T-DM1 alone Arm received Capecitabine throughout study and was counted in T-DM1+Cape Arm.

6. Primary Endpoint

The registry lists the primary outcome as Phase 1 (mBC): Percentage of Participants With Dose-Limiting Toxicities (DLTs), measured over Continuously during Cycle 1 (up to 3 weeks).

The primary endpoint is the percentage of participants with dose-limiting toxicities (DLTs) in the Phase 1 mBC cohort. The assessment was conducted continuously during Cycle 1, up to 3 weeks. A DLT was defined using a set of treatment-related toxicities and treatment interruptions, including specified severe blood-count abnormalities, febrile neutropenia, diarrhea or hand-foot syndrome, other severe toxicities that prevented the next cycle, and certain prolonged interruptions or dose limitations. The exact criteria varied in some respects by dose level.

This type of endpoint is used in early dose-finding work because it captures clinically important toxicity during a defined initial treatment period. When treatment is being combined or a dose is being explored, a DLT definition gives the study a structured way to describe whether toxicities crossed prespecified severity or treatment-delivery thresholds. The endpoint is a binary participant outcome: each eligible participant either meets the DLT definition or does not.

The registry defines the analysis population as enrolled and treated mBC participants who had no major protocol deviation and completed Cycle 1. This makes the endpoint specific to a selected, evaluable Phase 1 population rather than everyone enrolled in the overall study. The observed percentage, its numerator and denominator, and an uncertainty interval are not reported in the registry record. The endpoint therefore cannot be quantified from the posted primary-outcome information.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the primary endpoint count?

The endpoint is a percentage of participants who meet the protocol’s DLT definition during Cycle 1. In general, a percentage for a binary outcome is formed by identifying participants with the event and relating that count to the eligible analysis population. For this study, the registry specifies a DLT-evaluable population: enrolled and treated mBC participants without a major protocol deviation who completed Cycle 1. The registry does not report the primary outcome count or denominator, so the percentage cannot be reconstructed from the record details.

A composite definition can include different kinds of events, as this DLT definition does. The advantage is that it can capture several clinically important forms of early toxicity or treatment disruption under one endpoint. The trade-off is that the overall percentage does not show which component occurred most often or whether the events had similar clinical importance. Component-specific DLT results are not reported in the registry record.

How should uncertainty be understood?

In general, the number of observed participants with an event is only one part of an estimate. A confidence interval describes statistical uncertainty around an estimate, with its width influenced by the amount of information and the variability in the data. It does not account for every source of uncertainty: bias from participant selection, outcome measurement, or study conduct may remain. The registry record does not report a confidence interval for the primary endpoint, and no precision estimate can be assessed from the information described.

In general, a binary percentage can be analyzed with methods based on a binomial model, but the appropriate method depends on the analysis plan and data structure. The registry record does not report the specific statistical method used for the primary endpoint, whether any adjustment was made, or how uncertainty was calculated. No particular inferential procedure should therefore be attributed to the study.

What do the Phase 2 groups establish?

The Phase 2 arm descriptions concern a combination group and a T-DM1-alone group. In general, a randomized comparison can support estimation of an average treatment contrast when allocation is implemented as planned and outcomes are analyzed appropriately. Here, the title’s randomized description conflicts with the non-randomized allocation field, and the record warns that the groups are not a like-for-like control comparison. The registry does not clarify allocation or provide a comparative efficacy estimate in the information described. Thus, the listed arms can be described, but a causal interpretation cannot be assumed.

The record includes planning assumptions for sample-size illustrations: two-sided significance level 0.05, equal allocation between the two arms, a normal approximation, and effects defined as stated fractions of the observed effect. These are planning illustrations based on an observed effect, not the trial’s own design justification and not observed power. In general, such calculations ask what sample size might be needed under assumed conditions; they do not show the probability that a completed trial found a true effect. The registry record does not report the specific observed effect or sample-size illustration values in the information, so their numerical implications cannot be assessed.

For time-to-event outcomes, the record says planning illustrations use a total number of events, while other illustrations express participants per arm. In general, event-driven planning reflects that the amount of information in a time-to-event comparison depends on events observed, not simply enrollment. These conventions describe planning calculations; they do not establish that the study used a particular event-driven analysis. The outcome and analysis details needed to interpret such a calculation are not reported in the registry record.

How should serious adverse-event counts be read?

The adverse-event record lists the number at risk and the number affected by serious adverse events for each group, as well as deaths. These are descriptive counts. In general, counts alone do not adjust for differences in treatment exposure, follow-up, participant characteristics, or ascertainment. A comparison of group-level counts is particularly difficult to interpret when allocation is uncertain or non-randomized. The registry does not report an adjusted safety comparison or the follow-up duration underlying these counts.

In general, over-dispersion describes greater variability in data than a simple model assumes. It matters in outcomes such as counts when participants may have differing event rates or repeated events, because a model that assumes less variability can give overly narrow uncertainty intervals. The primary endpoint is binary at the participant level, and the registry record does not report a count model or an over-dispersion adjustment. This general issue should not be mistaken for a method used in this study.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Phase 1 (mBC) Cohort 1 (DL 1): T-DM1 + Cape", "Phase 1 (mBC) Cohort 1 (DL -1): T-DM1 + Cape")
n      <- c(6, 5)
events <- c(2, 0)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Phase 1 (mBC) Cohort 1 (DL 1): T-DM1 + Cape  6  2  33.3  4.3  77.7  9.7  70.0  9.9  71.0
ARM  Phase 1 (mBC) Cohort 1 (DL -1): T-DM1 + Cape  5  0  0.0  0.0  52.2  0.0  43.4  0.4  45.9

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 34 outcome measures, including the primary Phase 1 mBC DLT endpoint and secondary outcomes covering response and pharmacokinetic measurements. Multiple outcomes can create multiplicity: when many hypotheses are tested, the chance of at least one apparently unusual result can increase if each is judged separately. In general, a prespecified hierarchy or adjustment can address some multiplicity concerns. The statistical analysis plan, any multiplicity strategy, and the status of the listed outcomes as confirmatory or exploratory are not reported in the registry record. No claim about adjustment or unadjusted testing is warranted.

The primary endpoint’s analysis population is defined as enrolled and treated mBC participants who completed Cycle 1 and had no major protocol deviation. This is a DLT-evaluable population, not necessarily every participant who entered Phase 1. The registry does not report how many participants were excluded from this population, why any were not evaluable, or whether another analysis population was used. The primary endpoint numerator and denominator are also not reported.

Missing data handling is not reported in the registry record. In general, missing outcome data can affect interpretation if participants with missing assessments differ from those with observed data, especially when exclusion is related to toxicity or treatment discontinuation. The record does not describe imputation, sensitivity analyses, or rules for participants who did not complete Cycle 1. It is therefore not possible to assess how missingness or incomplete observation was handled.

The study is listed as terminated, but the reason for termination and any consequences for outcome collection or analysis are not reported in the registry record. Termination status alone does not establish whether results were incomplete, whether planned analyses changed, or why the study ended.

12. 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: muscular weakness (20.0%), pathological fracture (20.0%), brain oedema (20.0%), cerebral cyst (20.0%), radiation necrosis (20.0%).

The serious adverse-event figures are counts of affected participants, with the number at risk listed for each group. In Phase 1, the mBC dose-level groups list 0 among 7 and 2 among 5 affected; the LA/mGC group lists 4 among 6. In Phase 2, the combination group lists 11 among 82 and the T-DM1-alone group lists 10 among 78. The registry also lists deaths by group, but these counts do not establish that a death was caused by study treatment. Causality is not reported in the registry record.

These figures describe recorded safety experience; they are not, on their own, a fair comparison of treatment risks. Follow-up duration, exposure time, participant characteristics, and the way events were collected can affect counts. Interpretation of the Phase 2 contrast is further limited by the record’s non-randomized allocation field and warning that the groups are not like-for-like.

The listed serious-event terms include muscular weakness, pathological fracture, brain oedema, cerebral cyst, radiation necrosis, and myocardial infarction, with percentages reported for those terms. The term-specific denominators and their relationship to the group-level counts are not reported in the registry record. These percentages should not be read as overall rates for the study or as evidence of treatment causation.

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

The registry record documents a study designed to examine early DLTs with T-DM1 plus capecitabine and then explore efficacy and safety in an mBC Phase 2 component. The primary endpoint has a defined period and an evaluable population, which makes clear what the planned primary measurement concerns. However, the numerical primary result is not reported in the registry record, so no conclusion about the observed DLT percentage or its precision can be drawn from that endpoint information.

The Phase 2 treatment groups and safety counts provide descriptive context, but the record does not resolve whether allocation was randomized. The official title describes randomization, whereas the allocation field says non-randomized and the record cautions against treating the arms as like-for-like. Consequently, an apparent difference between group counts cannot be attributed to treatment on the basis of these descriptions alone. A comparative efficacy result is not reported in the registry record.

The adverse-event counts indicate that serious adverse events and deaths were recorded in the listed groups. Counts do not explain severity, timing, exposure, or causality. The registry’s term-specific percentages also lack reported denominators in the information described, limiting interpretation of those terms. Neither the counts nor the presence of a treatment arm establishes a benefit-harm balance.

The record’s terminated status is relevant context, but its reason is not reported. It does not by itself establish that the study stopped for safety, lack of efficacy, recruitment, or any other cause. More complete outcome results, analysis methods, follow-up information, and clarification of allocation would be needed to evaluate the study’s findings in detail. The registry record supports describing the design and listed safety counts, but not a definitive treatment-effect conclusion.

14. Why This Trial Matters Statistically

This study illustrates why a trial’s endpoint must be interpreted in the context of its phase and population. The primary endpoint is a Cycle 1 DLT measure in the Phase 1 mBC cohort, while the study’s later component was intended to explore efficacy and safety in mBC. These address different questions. Early tolerability information cannot be substituted for a comparative efficacy result, and the primary Phase 1 endpoint does not describe the separate LA/mGC cohort.

It also illustrates how registry design fields affect interpretation. A title can describe a randomized study while another registry field records non-randomized allocation. Because the record additionally cautions against a like-for-like comparison, readers need to preserve that uncertainty rather than assume the arms were balanced by random assignment. This matters whenever group outcomes are considered as evidence of a treatment contrast.

The DLT-evaluable definition is another important statistical feature. Restricting an endpoint to participants who completed Cycle 1 and had no major protocol deviation defines a specific analysis population. It also makes the number excluded and the handling of incomplete observation essential to judging how representative the estimate would be. Those details and the endpoint result are not reported in the registry record.

Finally, the record shows why safety counts and planning calculations need careful framing. A count is descriptive unless the comparison accounts for design and observation differences; a sample-size illustration is a planning tool, not evidence that a study had a particular probability of success. The registry’s limited statistical detail means that design description, observed results, and causal interpretation must remain distinct.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The primary endpoint result is not reported in the registry record. Without the number of DLTs, the evaluable denominator, or an uncertainty interval, the key Phase 1 mBC endpoint cannot be quantified. The record defines the evaluable population but does not state how many participants met that definition or how participants with incomplete Cycle 1 assessments were handled.

The allocation description is inconsistent: the official title describes a randomized Phase 2 study, while the allocation field says non-randomized. The record also states that the arms are not a like-for-like control comparison. No clarification of this inconsistency or details of allocation procedures are reported. This prevents a confident interpretation of the Phase 2 groups as a randomized comparison.

The posted adverse-event information is limited to group counts, deaths, and selected serious-event terms with percentages. Follow-up duration, exposure, term-specific denominators, event timing, and a comparative safety analysis are not reported in the registry record. Counts therefore cannot account for differences in observation or establish treatment causality.

The record lists the study as terminated, but the reason for termination and its consequences for planned enrollment, follow-up, and analyses are not reported. A statistical analysis plan, missing-data methods, multiplicity approach, and primary endpoint precision estimates are also not reported. These omissions limit assessment of prespecification, potential selection, and the strength of any inferential conclusions.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Combination Study of Kadcyla (Trastuzumab Emtansine) and Capecitabine in Participants With Human Epidermal Growth Factor Receptor 2 (HER2)-Positive Metastatic Breast Cancer (mBC) or HER2-Positive Locally Advanced/Metastatic Gastric Cancer (LA/mGC) — NCT01702558.
• Registry record last updated 2021-01-22; status last verified 2021-01.
• Results first posted 2019-04-08.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

The registry describes a sequential Phase 1 and Phase 2 study of T-DM1 and capecitabine in HER2-positive mBC or LA/mGC. The primary endpoint is the percentage of participants with DLTs during Cycle 1 in the Phase 1 mBC DLT-evaluable population. The primary endpoint result is not reported in the registry record.

The Phase 2 mBC groups are the combination and T-DM1 alone. Although the official title describes randomization, the allocation field says non-randomized and the record cautions that the groups are not a like-for-like comparison. Serious adverse-event counts are listed, but they are descriptive and do not establish causation. The study status is terminated; the reason is not reported in the registry record.