This page separates reported trial results from statistical interpretation. The numerical trial information on this page is restricted to the ClinicalTrials.gov record data. ClinicalTrials.gov provides the official trial registry record.
1. Trial at a Glance
DESTINY-Breast01 was a randomized, parallel, unmasked phase 2 study of DS-8201a in metastatic breast cancer previously treated with trastuzumab emtansine (T-DM1). The registry record reports 253 participants, four arms, and a primary endpoint focused on objective response following intravenous administration of 5.4 mg/kg DS-8201a.
| Feature | DESTINY-Breast01 |
|---|---|
| Trial name | DESTINY-Breast01 |
| NCT identifier | NCT03248492 |
| Phase | Phase 2 |
| Condition | Breast Cancer |
| Brief title | A Study of DS-8201a in Metastatic Breast Cancer Previously Treated With Trastuzumab Emtansine (T-DM1) |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None |
| Primary purpose | Treatment |
| Enrollment | 253.0 |
| Arms | 4 |
| Intervention | DS-8201a |
| Lead sponsor | Daiichi Sankyo |
| Sponsor type | Industry |
| Trial status | Terminated |
2. Clinical Question
The registry describes a phase 2 treatment study of DS-8201a in participants with metastatic breast cancer previously treated with trastuzumab emtansine (T-DM1). The primary statistical question recorded for the trial concerns whether participants receiving 5.4 mg/kg DS-8201a achieve an objective response as assessed by independent central review.
Population
Participants with metastatic breast cancer previously treated with trastuzumab emtansine (T-DM1).
Intervention
DS-8201a, including the 5.4 mg/kg administration specified for the registered primary endpoint.
Comparator
The ClinicalTrials.gov record does not identify a separate control treatment for the primary endpoint.
Primary question
What objective response was observed following intravenous administration of 5.4 mg/kg DS-8201a?
3. Trial Design
Randomization is an important feature even when the principal endpoint is a response proportion rather than a time-to-event endpoint. Random allocation can reduce systematic differences in participant characteristics across study groups and provides the foundation for causal comparisons when a valid concurrent comparator is present. In this the ClinicalTrials.gov record, however, the registered primary endpoint is specifically defined for participants receiving 5.4 mg/kg DS-8201a, and the ClinicalTrials.gov record does not provide a formal comparator analysis.
4. Endpoints
| Endpoint | Registry definition | Time frame | Type |
|---|---|---|---|
| Objective Response Rate as Confirmed by Independent Central Review Following Intravenous Administration of 5.4 mg/kg DS-8201a in Participants With Metastatic Breast Cancer (Enrolled Analysis Set) | The number of participants with objective response was assessed every six weeks from Cycle 1 Day 1 through discontinuation of treatment, by independent central imaging facility review based on RECIST version 1.1. | At least 6 months after last participant enrolled received first dose up to 19 months (data cut off) | Binary |
The endpoint is explicitly binary at the participant level: each participant is classified according to whether an objective response was observed under the registered assessment framework. The registry definition also specifies repeated assessment every six weeks from Cycle 1 Day 1 through treatment discontinuation, with confirmation by independent central imaging review using RECIST version 1.1.
5. Statistical Methodology
Binary response endpoint
The primary endpoint is classified as binary. For each participant in the enrolled analysis set receiving the specified 5.4 mg/kg administration, the outcome is whether an objective response occurred under the registry's RECIST version 1.1 assessment framework.
The resulting proportion summarizes the frequency of objective response in the defined analysis population. A confidence interval would quantify statistical uncertainty around that proportion.
Independent central review
The endpoint was assessed by an independent central imaging facility. This is statistically important because response classification depends on radiologic measurements and predefined RECIST version 1.1 rules. Central review provides a standardized assessment process distinct from relying solely on local investigator assessment.
Repeated response assessment
The registry specifies assessment every six weeks from Cycle 1 Day 1 through discontinuation of treatment. Repeated assessments are necessary because a participant's disease status can change over the course of treatment. The primary binary endpoint ultimately reduces those longitudinal assessments to whether the participant met the objective-response definition.
Enrolled Analysis Set
The primary endpoint is explicitly defined for the Enrolled Analysis Set. The denominator for an objective response rate should therefore be tied to the prespecified analysis population rather than automatically substituted with the overall trial enrollment of 253.0. The ClinicalTrials.gov record does not provide the enrolled analysis-set count for the 5.4 mg/kg primary analysis.
Randomization and analysis
Randomization is recorded at the trial level, while the primary endpoint is reported as a binary response outcome. If a randomized comparator is being formally evaluated, an appropriate analysis would compare response proportions between randomized groups using a prespecified method for binary outcomes, together with an effect measure and confidence interval. The ClinicalTrials.gov record contains no formal comparison, so no such comparison is reported here.
6. Results
ClinicalTrials.gov reports that results were posted and that 8 outcome measures were posted. The primary endpoint is also marked as having results posted. However, the ClinicalTrials.gov record contains no formal statistical analyses.
Primary Endpoint: Objective Response Rate
Objective response
Primary endpoint: objective response rate confirmed by independent central review following intravenous administration of 5.4 mg/kg DS-8201a.
No formal statistical estimate, confidence interval, or p-value is reported in the ClinicalTrials.gov record.
| Primary endpoint component | Registry information reported |
|---|---|
| Endpoint | Objective Response Rate as Confirmed by Independent Central Review |
| Dose specified | 5.4 mg/kg DS-8201a |
| Analysis set | Enrolled Analysis Set |
| Assessment method | Independent central imaging facility review |
| Response criteria | RECIST version 1.1 |
| Assessment schedule | Every six weeks from Cycle 1 Day 1 through discontinuation of treatment |
| Time frame | At least 6 months after last participant enrolled received first dose up to 19 months (data cut off) |
| Formal statistical analysis reported | No |
| Confidence interval reported | No |
| P-value reported | No |
The endpoint is designed to answer a straightforward descriptive question: what proportion of participants in the specified analysis population achieved an objective response according to independent central review and RECIST version 1.1?
Because the ClinicalTrials.gov record contains no formal comparison, the posted result should not be represented here as evidence from a hypothesis test against a comparator. In particular, there is no registry-reported p-value that can establish statistical significance and no registry-reported confidence interval that can quantify the precision of the reported response proportion.
A response rate is also not the same as the probability of long-term survival, the probability of remaining progression-free, or the proportion of participants who experience durable benefit. It is a binary tumor-response endpoint defined by the specified imaging criteria.
The repeated six-week assessments also matter. Response is determined within a longitudinal assessment process rather than from a single measurement at one arbitrary time point. The registered definition therefore combines a prespecified response criterion with repeated imaging and independent central review.
What would normally be reported for a binary endpoint?
For a binary response endpoint, a complete statistical presentation would normally include the number and proportion of responders, the corresponding denominator, and a confidence interval for the response proportion. If a comparator group were the formal target of inference, the analysis would additionally report an appropriate between-group effect measure and its confidence interval, together with a prespecified hypothesis test when applicable.
Those quantities are not reported in the ClinicalTrials.gov record. Reporting them from memory or reconstructing them from outside publications would violate the numerical source restriction for this record.
7. Safety
The ClinicalTrials.gov record reports serious adverse events by arm. These figures are presented as affected participants divided by participants at risk. They should be interpreted as descriptive safety counts rather than as a formal comparative safety analysis.
| Trial part / arm | Serious adverse events | Affected / at risk |
|---|---|---|
| Part 1: DS-8201a Low Dose | Serious adverse events | 14/50 |
| Part 1: DS-8201a Medium Dose | Serious adverse events | 11/48 |
| Part 1: DS-8201a High Dose | Serious adverse events | 8/21 |
| Part 2a: DS-8201a Low Dose | Serious adverse events | 42/130 |
| Part 2b (Exploratory): DS-8201a Low Dose | Serious adverse events | 1/4 |
What the numerator means
The numerator represents participants affected by a serious adverse event in the corresponding registry arm and trial part.
What the denominator means
The denominator represents participants at risk in that corresponding arm and trial part.
Descriptive evidence
The ClinicalTrials.gov record provides affected/at-risk counts but no formal statistical comparison of serious adverse-event rates.
Arm structure
The safety data distinguish Part 1, Part 2a, and the exploratory Part 2b group, so the counts should not be collapsed into a single unqualified rate.
The arm-specific structure is particularly important statistically. A serious-adverse-event count from one trial part has a different denominator and design context from a count in another part. Combining them without a prespecified analysis could obscure differences in exposure and population.
8. Statistical Methods Explained
Why is the primary endpoint binary?
The registry defines objective response as a binary endpoint. Each participant ultimately contributes a response classification under the RECIST version 1.1 framework. Binary endpoints are useful when the clinical question is whether a participant achieved a predefined response rather than how long the participant survived or how much an imaging measurement changed continuously.
Why was independent central review used?
Radiologic response assessment involves interpretation of imaging findings. Independent central review provides a standardized assessment mechanism and can reduce the influence of local assessment differences. In this trial, the registry specifically identifies an independent central imaging facility and RECIST version 1.1.
Why are repeated assessments made every six weeks?
Tumor response is dynamic. A participant who has not responded at one assessment may meet response criteria later, while a participant's disease status can subsequently change. The registered schedule therefore evaluates participants repeatedly from Cycle 1 Day 1 through discontinuation of treatment rather than relying on a single assessment.
Why does the analysis population matter?
The primary endpoint is defined for the Enrolled Analysis Set. A response rate is always a proportion, so the denominator is fundamental to its interpretation. Using the overall enrollment of 253.0 as the denominator without evidence that it is the relevant enrolled analysis-set size could produce a different and potentially incorrect estimate.
What would a confidence interval add?
A confidence interval would quantify uncertainty around the estimated response proportion. A point estimate alone describes the observed frequency, whereas the interval provides information about the precision of that estimate under the specified statistical framework. The ClinicalTrials.gov record does not provide such an interval for this endpoint.
Why can there be results without a formal statistical analysis?
ClinicalTrials.gov can contain posted outcome information without the ClinicalTrials.gov record containing a corresponding entry in its statistical-analysis structure. That distinction matters here. The trial has posted results, but the ClinicalTrials.gov record does not provide the formal inferential quantities needed to report a statistical comparison.
9. Planned Analysis and Interpretation Framework
Because the registry has posted outcome information but the registry-reported statistical-analysis section is empty, the appropriate educational approach is to distinguish the endpoint measurement from the inferential analysis that would ordinarily accompany it.
| Question | Appropriate statistical framework | What is reported here? |
|---|---|---|
| Did a participant achieve objective response? | Binary classification based on RECIST version 1.1 and independent central review. | Yes; the registry defines the endpoint. |
| What proportion responded? | Response count divided by the prespecified analysis-set denominator. | Primary endpoint results are posted, but the numerical estimate is not present in the ClinicalTrials.gov record. |
| How precise is the response estimate? | Confidence interval for the response proportion. | No confidence interval is reported in the ClinicalTrials.gov record. |
| Is there a statistically significant treatment comparison? | A prespecified comparison between randomized groups, if applicable, with an appropriate binary-outcome test and effect measure. | No formal comparison is reported. |
This distinction prevents an important statistical error: treating a posted descriptive outcome as though it were accompanied by a formal hypothesis test when the registry data posted on ClinicalTrials.gov for analysis do not contain that test.
10. Randomization and What It Contributes Statistically
The trial is classified as randomized with a parallel design. Randomization is one of the central mechanisms used in clinical trials to make treatment assignment independent of measured and unmeasured baseline characteristics in expectation.
The strength of this chain depends on how the randomized comparison is defined and analyzed. Randomization does not, by itself, supply a treatment-effect estimate or guarantee numerical balance in every baseline characteristic.
In a parallel trial, participants remain associated with their randomized treatment assignment rather than moving through a sequence of experimental treatments as part of a crossover design. The ClinicalTrials.gov record identifies the study as parallel and unmasked.
For this particular record, the main caution is that the primary endpoint is specified for 5.4 mg/kg DS-8201a and the ClinicalTrials.gov record does not identify a formal comparator analysis. Therefore, the existence of randomization should not be converted into an implied treatment-versus-control statistical result.
11. Interpreting the Four-Arm Structure
The registry identifies four arms, while the ClinicalTrials.gov record distinguishes several dose and trial-part groups. The available information therefore indicates a dose- and part-specific structure rather than a simple two-group randomized comparison.
| Registry feature | What can be stated from the ClinicalTrials.gov record |
|---|---|
| Number of arms | 4 |
| Intervention | DS-8201a |
| Design model | Parallel |
| Masking | None |
| Primary endpoint focus | 5.4 mg/kg DS-8201a |
| Safety groups reported | Part 1 low, medium, and high dose; Part 2a low dose; Part 2b exploratory low dose |
These distinctions matter because an arm label is not merely descriptive. In a dose-oriented study, the denominator, exposure, treatment intensity, and purpose of each trial part can affect the appropriate statistical interpretation. The ClinicalTrials.gov record does not provide enough detail to reconstruct the complete allocation scheme, so the analysis should remain at the level supported by the registry record.
12. Data Cutoff and Timing
The primary endpoint time frame is specified as at least 6 months after the last participant enrolled received the first dose up to 19 months, with the registry identifying this as the data cutoff framework for the endpoint.
Study start
The registry lists 2017-08-25 as the trial start date.
Primary completion
The registry lists 2019-03-21 as the primary completion date.
Response assessment period
The registered primary endpoint uses a period of at least 6 months after the last participant enrolled received first dose up to 19 months, identified as the data cutoff.
A data cutoff is statistically important because a response analysis is tied to the information available by a defined point in time. Without a common cutoff, participants could have materially different opportunities for observation, making the reported outcome difficult to interpret consistently.
13. Limitations
- No formal statistical analyses reported: the statistical-analysis array contains no entries, so no formal confidence interval or p-value can be reported from the ClinicalTrials.gov record.
- No numerical primary result reported in the ClinicalTrials.gov record: the registry marks the primary endpoint as having results posted, but the ClinicalTrials.gov record does not include its numerical response estimate.
- No comparator analysis reported: the available data do not identify a formal comparator result for the primary endpoint.
- Four-arm structure: the trial contains four arms, while the registry-reported primary endpoint focuses specifically on 5.4 mg/kg DS-8201a. The complete allocation structure is not provided.
- Analysis-set denominator: the primary endpoint is defined for the Enrolled Analysis Set, but the ClinicalTrials.gov record does not give that set's numerical size.
- Repeated assessment: response was assessed every six weeks, so the endpoint reflects a longitudinal imaging process rather than a single observation.
- Safety interpretation: serious-adverse-event counts are reported by trial part and arm, and no formal comparative safety analysis is reported.
- Generalizability: the ClinicalTrials.gov record describes a specific metastatic breast cancer population previously treated with T-DM1; it does not establish how the findings apply to populations outside the registered study population.
14. Why This Trial Matters Statistically
DESTINY-Breast01 is useful as a statistical teaching case because its registry structure illustrates how randomized trial design and binary tumor-response endpoints fit together without requiring every clinical trial to use a time-to-event primary endpoint.
| Concept | How it appears in DESTINY-Breast01 |
|---|---|
| Randomization | The registry classifies allocation as randomized. |
| Parallel design | The trial is classified as a parallel study. |
| Binary endpoint | The primary endpoint is classified as binary. |
| Objective response | The primary endpoint evaluates objective response. |
| Independent central review | Response is assessed by an independent central imaging facility. |
| RECIST v1.1 | The registry specifies RECIST version 1.1 for response assessment. |
| Repeated assessment | Response is assessed every six weeks from Cycle 1 Day 1 through discontinuation. |
| Analysis population | The primary endpoint specifies the Enrolled Analysis Set. |
| Data cutoff | The primary endpoint has a defined assessment window ending at 19 months. |
| Safety denominators | Serious adverse events are reported as affected participants divided by participants at risk for specific arms. |
The statistical lesson is that the endpoint definition is inseparable from its denominator, assessment method, timing, and analysis population. A response proportion without those elements is incomplete. Conversely, a well-defined endpoint does not automatically imply that a formal inferential comparison has been performed.
15. What an Objective Response Rate Does — and Does Not — Mean
An objective response rate is a proportion of participants who meet a prespecified response definition. In this trial, the registry specifies independent central imaging review and RECIST version 1.1, with assessments every six weeks from Cycle 1 Day 1 through discontinuation.
It does not mean that every responding participant has the same magnitude or duration of benefit. A binary response endpoint collapses a participant's response status into a yes-or-no classification for the defined analysis.
The primary endpoint is explicitly tied to the Enrolled Analysis Set. Therefore, the number of participants in that analysis set is essential for reproducing the response-rate calculation. The overall enrollment of 253.0 should not automatically be treated as the primary endpoint denominator.
A descriptive response proportion and a treatment-effect test answer different questions. A response proportion describes what was observed. A confidence interval describes uncertainty around that estimate. A comparative hypothesis test addresses whether the observed difference between defined groups is compatible with a prespecified null hypothesis. The ClinicalTrials.gov record contains no such formal comparison.
16. Clinical Interpretation vs Statistical Interpretation
Statistical interpretation
The trial uses a randomized parallel design and defines a binary primary endpoint based on independent central review of RECIST version 1.1 assessments. The ClinicalTrials.gov record does not contain a formal statistical analysis of that endpoint.
Clinical interpretation
The study evaluates DS-8201a in metastatic breast cancer previously treated with T-DM1, with objective response selected as the registered primary outcome for the specified 5.4 mg/kg administration.
Keeping these two levels separate is important. The statistical design tells us how evidence is intended to be generated and evaluated. The clinical question tells us what outcome is considered relevant. Neither should be expanded beyond the information actually documented in the ClinicalTrials.gov record.
17. Statistical Concepts in This Trial
Learn more about the methods used in this trial:
18. Related Statistical Calculators
19. Sources
- ClinicalTrials.gov: NCT03248492 — DESTINY-Breast01.
- PubMed: PMID 36342619.
- PubMed: PMID 35964548.
- PubMed: PMID 35642907.
- PubMed: PMID 34263665.
- PubMed: PMID 33999422.
Continue through the Clinical Biostats tutorials and calculators
Explore statistical methods and tools related to randomized clinical-trial design and binary outcomes.
20. Record Summary
DESTINY-Breast01 provides a useful example of how a randomized phase 2 oncology trial can define a binary objective-response endpoint using repeated imaging, independent central review, and RECIST version 1.1. The registry identifies 253.0 enrolled participants, four parallel arms, DS-8201a as the intervention, and a primary endpoint focused on objective response following intravenous administration of 5.4 mg/kg DS-8201a in the Enrolled Analysis Set.
The most important statistical distinction is between endpoint measurement and formal statistical inference. The registry defines how objective response was assessed and indicates that results were posted, but the ClinicalTrials.gov record contains no formal analysis, confidence interval, or p-value. The serious-adverse-event information is likewise registry-reported as arm-specific affected/at-risk counts without a formal comparative safety analysis.