A Study to Evaluate the Pharmacokinetics, Efficacy, and Safety of Subcutaneous Administration of the Fixed-Dose Combination of Pertuzumab and Trastuzumab in Combination With Chemotherapy in Participants With HER2-Positive Early Breast Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Evaluate the Pharmacokinetics, Efficacy, and Safety of Subcutaneous Administration of the Fixed-Dose Combination of Pertuzumab and Trastuzumab in Combination With Chemotherapy in Participants With HER2-Positive Early Breast Cancer (NCT03493854), a randomized phase 3 study. The primary outcome is Trough Serum Concentration (Ctrough) of Pertuzumab During Cycle 7 (Pre-Dose Cycle 8).

2. Clinical Question

The endpoint was measured at the pre-dose visit on Cycle 8, Day 1, up to 21 weeks. The registry describes this measurement as following 3 cycles of intravenous pertuzumab and trastuzumab or the subcutaneous fixed-dose combination. Trough concentration is a pharmacokinetic measure: it captures the drug concentration near the end of a dosing interval, before the next dose. Comparing these values addresses whether exposure with the subcutaneous formulation is sufficiently similar under the study’s stated comparison framework.

The registry also lists pathological response and several event-free, recurrence, and survival outcomes. Those outcomes concern different aspects of treatment and follow-up; they do not answer the same question as the primary concentration endpoint. The primary result therefore speaks directly to measured pertuzumab exposure in the specified analysis population, not by itself to long-term clinical outcomes.

3. Trial Design

This was a randomized, parallel, open-label, two-arm Phase III study. It was multicenter and global according to the study summary. The registry lists enrollment of 500 and a completed status. The recorded start date is 2018-06-14 and the primary completion date is 2019-07-04. Masking was listed as none, so participants and study personnel were not blinded according to the registry design fields.

Arm A received intravenous pertuzumab and trastuzumab with chemotherapy. Arm B received the subcutaneous fixed-dose combination of pertuzumab and trastuzumab with chemotherapy. The arm descriptions state that participants received 8 cycles of investigator’s choice of neoadjuvant chemotherapy. They refer to alternative chemotherapy sequences, including anthracycline- and taxane-based medicines, but the arm descriptions in the registry record are incomplete. Specific regimen details beyond what is recorded cannot be established from this record.

The primary endpoint was pertuzumab trough serum concentration before Cycle 8. Its analysis population was narrower than the full study enrollment: the Per Protocol PK population consisted of participants meeting the prespecified pharmacokinetic assessment criteria. The endpoint description specifies several reasons for exclusion, including a missing pre-dose sample, collection outside the stated timing window, dosing deviations, delays, use of a subcutaneous injection site other than the thigh, switched samples, and an assay error affecting the measurement.

The record also reports a pathological complete response outcome and lists pharmacokinetic, event-free, recurrence, and survival outcomes. It does not report the randomization method, allocation concealment procedure, or detailed statistical analysis plan. The protocol’s sample-size justification and any interim analyses are not reported in the registry record. Those omissions limit assessment of how design decisions were made, but they do not change the reported definitions of the arms and primary endpoint.

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 607 patients were screened, 500 of whom were randomized.

Neoadjuvant Treatment Phase (24 Weeks)

Reasons for not completing this period:

Surgery & Adjuvant Treatment (42 Weeks)

Reasons for not completing this period:

Treatment-Free Follow-Up (up to 3 Years)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Per Protocol PK Population: includes only participants who adhered to the pre-specified criteria for the schedule of pharmacokinetic assessments.

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 Trough Serum Concentration (Ctrough) of Pertuzumab During Cycle 7 (Pre-Dose Cycle 8), measured over Pre-dose on Cycle 8, Day 1 (up to 21 weeks).

The primary endpoint was pertuzumab trough serum concentration, or Ctrough, measured before Cycle 8, Day 1, up to 21 weeks. The measure is reported in micrograms per millilitre (μg/mL), using a geometric mean and geometric coefficient of variation. A trough measurement is taken near the end of a dosing interval, before the next dose; as a general pharmacokinetic concept, it can describe how much drug remains in circulation at that point in the schedule.

This endpoint was used to compare exposure under the two administration approaches: intravenous pertuzumab with intravenous trastuzumab, versus the subcutaneous fixed-dose combination of pertuzumab and trastuzumab. According to the registry description, the measurement followed 3 cycles of the assigned pertuzumab/trastuzumab approach. It is therefore a formulation and exposure comparison, not a direct measure of tumour response or survival.

The analysis was restricted to the Per Protocol PK population. Participants could be excluded for specified deviations from the planned sampling schedule or dosing conditions, as well as missing samples and assay problems. These restrictions are relevant because trough concentrations are sensitive to when a sample is collected relative to dosing. As a general principle, timing or dosing differences can make observations less comparable, which is why protocol adherence criteria are used in pharmacokinetic analyses. The endpoint gives information about measured drug concentration under those conditions; it does not alone establish the comparative effect on clinical outcomes.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is a continuous pharmacokinetic measure reported as a geometric mean. Concentration data are often analyzed on a logarithmic scale because values can be right-skewed and because proportional differences may be more meaningful than absolute differences. As a general statistical explanation, a geometric mean is obtained by averaging on the log scale and transforming back to the original scale. The registry reports geometric means of 72.4 μg/mL and 88.7 μg/mL, with geometric coefficients of variation of 34.1% and 33.6%.

What does the ratio express?

A ratio compares the scale of the group geometric means. A ratio of 1.00 would represent equal estimated geometric means. The registry reports a geometric mean ratio of 1.22 and a 90% confidence interval from 1.14 to 1.31. A confidence interval conveys uncertainty in the estimated group contrast under the analysis assumptions; it is not a statement that a specified share of individual participants’ values lie in that interval. Here the interval is for the ratio, not for either arm’s raw concentrations.

The registry labels the analysis non-inferiority. In a non-inferiority design, the estimate is assessed against a prespecified boundary representing the largest acceptable loss on the chosen scale. General statistical practice requires that boundary and the decision rule to interpret the result formally. The boundary, test details, and analysis plan are not reported in the registry record. Consequently, the ratio and interval can be described, but the formal margin-based conclusion cannot be checked from the available record.

Why use a per-protocol pharmacokinetic population?

The endpoint description defines a Per Protocol PK population by adherence to prespecified criteria for pharmacokinetic assessments. It excludes participants with a missing pre-dose Cycle 8 sample, sample timing deviations outside the stated range, dose amounts deviating by more than 20% within 3 cycles from Cycle 5, dose delays of more than 7 days, a subcutaneous injection site other than the thigh, switched Cycle 8 pre-dose and post-dose samples, or an assay error affecting Ctrough. These rules reflect a general concern: concentration measurements are interpretable only in relation to dose and collection timing. Restricting the analysis can improve comparability of measurements, while also making the estimate apply specifically to participants who met the criteria.

The reported primary group sizes are 203.0 for Arm A and 206.0 for Arm B. They are analysis-population sizes, not the registry enrollment figure. The registry does not provide a full accounting of exclusions by reason or explain how missing pharmacokinetic values were handled beyond the stated eligibility exclusions. Imputation methods and sensitivity analyses are not reported in the registry record.

How should the pathological response comparison be read?

The secondary pathological complete response result is reported as a percentage with a 95% confidence interval. The registry gives 59.5% for Arm A and 59.7% for Arm B. The sponsor’s posted comparison is a difference of 0.15 with a 95% confidence interval from -8.67 to 8.97. A simple unadjusted recalculation uses the posted counts of 150 and 148 among group sizes of 252.0 and 248.0. It gives a risk difference of 0.002, a confidence interval from -0.085 to 0.088, and an unadjusted p-value of 0.972. These are unadjusted calculations from summary counts, not a replacement for the sponsor’s own analysis.

The confidence interval is an uncertainty interval for the estimated difference. Because it spans zero, the data are compatible with no difference under this calculation, while also allowing differences in either direction. A non-significant result does not prove the treatments equivalent. Equivalence or non-inferiority for this outcome would require a prespecified margin and an analysis designed for that question; such a margin is not reported for this endpoint.

Why can excess variability matter?

Binary outcomes such as pathological response are commonly analyzed using methods based on a binomial model. As a general statistical concept, over-dispersion occurs when outcome variability is greater than the model assumes. If present and ignored, standard errors and confidence intervals may be too optimistic. The registry does not report an over-dispersion adjustment for this outcome, and the unadjusted recalculation uses the posted summary counts rather than participant-level data. It therefore cannot account for covariate adjustment, stratification, or model features that may have been part of the sponsor’s analysis.

For continuous pharmacokinetic values, variation is described using the geometric coefficient of variation. This is not a confidence interval and does not measure uncertainty in the estimated ratio. General statistical interpretation separates within-group variation, which describes spread in observed values, from the confidence interval around a group contrast, which reflects precision of that estimate.

What do the planning illustrations mean?

The registry’s planning illustration is explicitly based on the observed effect, not on the trial’s own design. Its stated assumptions are a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. It presents participants per arm for 80% and 90% power for planning at 100%, 75%, and 50% of the observed effect. These are hypothetical sample-size calculations under those assumptions. They are not the trial’s enrollment target justification, and they are not observed power. The protocol’s sample-size rationale is not reported in the registry record.

General statistical guidance is that power calculations depend on the effect size chosen, the variability or event rate, the significance level, and the design assumptions. A calculation that uses an effect estimated from the same study is especially illustrative rather than a prospective justification. The posted planning values should therefore be read as sensitivity scenarios for future planning, not as evidence that the completed trial had a particular probability of detecting its observed result.

Finally, multiple endpoints create a multiplicity issue. When many outcomes or analyses are examined, the chance of at least one apparently unusual result can increase if each is interpreted independently. The registry lists several secondary outcomes and 9 analyses, but any adjustment strategy, hierarchy, or prespecified testing sequence is not reported in the registry record. Accordingly, unadjusted secondary findings warrant more caution than a primary outcome tested under a fully specified plan.

10. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary comparison as a geometric mean ratio of 1.22 with a 90% confidence interval from 1.14 to 1.31, labelled as a non-inferiority analysis. It does not report a p-value for that comparison. The primary outcome values and this sponsor-reported ratio are the basis for describing the pharmacokinetic result; there is no independently recalculated primary estimate in the posted information. The non-inferiority margin and detailed analysis method are not reported in the registry record, so the formal decision rule cannot be reconstructed.

For pathological complete response, the sponsor reports a difference of 0.15 with a 95% confidence interval from -8.67 to 8.97. The simple recalculation from the posted counts gives a risk difference of 0.002, with a confidence interval from -0.085 to 0.088 and an unadjusted p-value of 0.972. The estimates are expressed on different displayed scales, so their numeric magnitudes should not be compared as if they used identical units. Both intervals include no difference and allow uncertainty in either direction.

The recalculation is unadjusted and uses summary counts. It may not reproduce the sponsor’s analysis choices, and it does not account for participant-level covariates, stratification, or other model details that are not reported. No discrepancy is flagged in the registry record. The recalculation is therefore a transparent check based on posted summaries, not a substitute for the sponsor’s analysis or a claim that the methods were identical.

11. Multiplicity, Missing Data and Analysis Populations

The registry reports a primary pharmacokinetic endpoint and a broad set of secondary outcomes, with 9 analyses recorded. Secondary topics include another drug concentration measure, pathological response, several definitions of event-free or recurrence-free status, and overall survival. Considering several outcomes creates a multiplicity concern: if each is tested separately without adjustment, the chance of at least one false-positive result can increase. This is a general statistical issue, not a claim that any reported result is false.

The registry record does not report a multiplicity adjustment, a hierarchical testing sequence, or which secondary analyses were intended to be confirmatory. It also does not report whether interim analyses were conducted. Without those details, secondary estimates and p-values, where available, should be treated as descriptive or exploratory unless a separate analysis plan establishes otherwise. The primary comparison is identified, but the non-inferiority margin and full decision rules are not reported, which also limits independent appraisal of the primary claim.

Missing-data handling is only partly described. For the primary endpoint, the Per Protocol PK population excludes participants with a missing pre-dose sample and several other departures from the planned assessment or dosing conditions. The record does not give the number excluded for each reason, describe imputation, or report sensitivity analyses addressing missing pharmacokinetic measurements. Because the analysis is restricted to protocol-adherent measurements, its results apply to that defined population and may not represent all randomized participants in the same way.

The registry reports the pathological response outcome in groups of 252.0 and 248.0, while the primary PK analysis groups are 203.0 and 206.0. These are endpoint-specific denominators. The record does not provide a complete participant flow or explain all differences between analysis populations. It also does not report the handling of missing pathological response data. A clear accounting of analysis sets and missing outcomes would help assess whether exclusions could influence interpretation.

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: febrile neutropenia (4.0%), cardiac failure (2.4%), neutropenic sepsis (1.2%), neutrophil count decreased (1.2%), neutropenia (1.2%).

The safety record reports serious adverse events for groups at risk of 252 and 248. Serious adverse events affected 52 participants in Arm A and 49 in Arm B. Other adverse events affected 251 and 248 participants, respectively. The record also lists 12 deaths in Arm A and 14 in Arm B. These are reported counts; the registry summary does not establish that a particular event or death was caused by the study treatment.

The most frequently listed serious adverse event terms include febrile neutropenia at 4.0%, cardiac failure at 2.4%, and neutropenic sepsis, neutrophil count decreased, neutropenia, and pyrexia at 1.2% each. These terms provide a snapshot of events recorded as serious, but the registry information does not give the full event-level context needed to assess timing, duration, severity distribution, or attribution.

Safety counts should be read in relation to their stated groups at risk and outcome definitions. The registry record does not report additional details such as exposure-adjusted rates or a causal assessment for the deaths. Comparisons of raw counts alone do not establish that one regimen is safer, particularly when the record does not give the complete analysis framework for safety outcomes.

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

The primary result describes pertuzumab trough concentration before Cycle 8 in the protocol-defined Per Protocol PK population. The reported geometric mean ratio is 1.22, with a 90% confidence interval from 1.14 to 1.31. This is evidence about measured drug exposure under the study’s dosing and sampling conditions. It is not itself a measure of tumour response, recurrence, survival, or quality of life.

The registry labels the primary analysis non-inferiority. A formal non-inferiority interpretation requires a prespecified margin and a decision rule. Because the margin is not reported in the registry record, readers cannot independently determine from this information alone whether the interval meets that criterion. The ratio and interval remain interpretable as the posted estimate and its uncertainty, but they do not replace the missing design detail.

The pathological complete response outcome provides a separate clinical result. Its reported percentages are close, and the sponsor’s difference interval spans no difference. The simple unadjusted recalculation is also uncertain. This does not prove equivalence, because no equivalence margin is reported and the interval allows effects in either direction. Nor does it establish a difference. The result is best described as an estimate with uncertainty rather than as proof of identical outcomes.

The listed time-to-event outcomes and safety findings add context, but their interpretation is limited by incomplete reporting of analyses, populations, and missing-data methods. The absence of an estimate for an endpoint does not imply that no events occurred or that the arms were alike. Similarly, adverse event counts and deaths are recorded observations, not proof of causation.

In general, a clinical trial result is most informative when the endpoint, analysis population, estimate, uncertainty, and prespecified decision rule are considered together. In this study, the concentration result informs a pharmacokinetic comparison, while the other endpoints address different clinical and safety questions. Conclusions should remain within those boundaries and should not treat a surrogate exposure measure as a complete account of clinical benefit or harm.

14. Why This Trial Matters Statistically

This trial raises a useful statistical distinction between comparing drug exposure and comparing clinical outcomes. Its primary endpoint is a measured trough concentration, analyzed as a geometric mean ratio. That type of endpoint can provide a direct comparison of pharmacokinetic exposure between formulations, while pathological response and time-to-event outcomes address different questions. Keeping these estimands separate prevents a concentration result from being interpreted as though it were itself a survival or response result.

The study also illustrates why the analysis population matters. The primary pharmacokinetic result comes from a Per Protocol PK population with specific rules for sampling time, dosing, injection site, and assay validity. These conditions aim to make concentration measurements comparable, but they define a subset of participants and make the result specific to that subset. By contrast, the registry reports pathological response with different group denominators. Endpoint-specific populations need to be considered rather than assumed interchangeable.

Non-inferiority comparisons also depend on more than the estimated ratio and confidence interval. The margin is central to the interpretation, yet it is not reported in the registry record. The posted estimate can be described, but a reader cannot independently verify the formal margin-based conclusion without that design information. This makes explicit reporting of the margin and analysis rules important for transparent evaluation.

Finally, the many secondary outcomes show why multiplicity and missing results matter. A collection of endpoints can describe response, exposure, recurrence, event-free status, survival, and safety, but it does not automatically form a single confirmatory conclusion. A clearly reported testing strategy, participant flow, and missing-data approach would help readers distinguish planned evidence from exploratory patterns. These are general lessons in evaluating complex randomized trials, grounded in the endpoint and reporting structure of this study.

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.

Several trial-specific details needed for a fuller statistical assessment are not reported in the registry record. The non-inferiority margin, the full primary analysis method, and the decision criteria are absent. Without these, the reported geometric mean ratio and 90% confidence interval can be described, but the formal non-inferiority assessment cannot be independently reconstructed.

The primary pharmacokinetic analysis is restricted to a Per Protocol PK population. The record lists detailed exclusion criteria but does not quantify exclusions by reason, provide a complete flow from enrollment to each analysis set, or describe sensitivity analyses. The primary endpoint’s group sizes differ from those reported for pathological response. These distinct denominators reflect endpoint-specific analysis populations, but their full relationship is not explained.

The registry does not report the randomization procedure, allocation concealment, or detailed statistical analysis plan. The arm descriptions are incomplete, so the exact chemotherapy regimens and their distribution cannot be fully established from the record. The protocol’s sample-size justification and any interim analyses are also not reported. A planning illustration based on the observed effect is not a substitute for the trial’s original design rationale.

Secondary reporting is incomplete for some two-arm comparisons, and the multiplicity strategy is not reported. The record does not describe missing-data handling for the pathological response outcome or provide full methods for the listed time-to-event analyses. Safety results are reported as counts and selected serious-event terms, without a complete causal assessment or all context needed for comparative safety interpretation. These limitations do not negate the posted estimates, but they narrow what can be concluded from the registry record alone.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study to Evaluate the Pharmacokinetics, Efficacy, and Safety of Subcutaneous Administration of the Fixed-Dose Combination of Pertuzumab and Trastuzumab in Combination With Chemotherapy in Participants With HER2-Positive Early Breast Cancer — NCT03493854.
• Registry record last updated 2024-06-26; status last verified 2024-06.
• Results first posted 2020-06-25.
• 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

NCT03493854 is a completed randomized, open-label, parallel Phase III study in HER2-positive early breast cancer. It compared intravenous pertuzumab and trastuzumab with a subcutaneous fixed-dose combination of pertuzumab and trastuzumab, both in combination with chemotherapy. The registry lists enrollment of 500 and records the study start as 2018-06-14 and primary completion as 2019-07-04.

The primary endpoint was pertuzumab trough concentration before Cycle 8 in the Per Protocol PK population. The registry reports a geometric mean ratio of 1.22 with a 90% confidence interval from 1.14 to 1.31 and labels the analysis non-inferiority. The non-inferiority margin is not reported in the registry record. Secondary outcomes include pathological response and several pharmacokinetic and time-to-event measures. Safety counts and selected serious adverse event terms are also recorded.

Interpretation is constrained by missing analysis-plan details, incomplete reporting for some secondary comparisons, and endpoint-specific analysis populations. The posted primary result describes pharmacokinetic exposure; it does not by itself establish comparative clinical outcomes.