This page separates registry-reported trial information from statistical interpretation. The ClinicalTrials.gov record provides the official trial registry information.
1. Trial at a Glance
TAPUR is a phase 2, non-randomized, single-group study evaluating FDA-approved drugs that target specific tumor-gene abnormalities in people with advanced cancer. The registry describes cohorts defined by the same tumor type, genomic variant, and study drug, with objective response rate assessed at 16 weeks of treatment.
| Feature | TAPUR |
|---|---|
| Study name | TAPUR: Testing the Use of Food and Drug Administration (FDA) Approved Drugs That Target a Specific Abnormality in a Tumor Gene in People With Advanced Stage Cancer |
| Phase | Phase 2 |
| Status | Recruiting |
| Population | Lymphoma, Non-Hodgkin; Multiple Myeloma; Advanced Solid Tumors |
| Allocation | Non-randomized |
| Design model | Single group |
| Masking | None |
| Primary purpose | Treatment |
| Enrollment | 4200 |
| Number of arms | 12 |
| Lead sponsor | American Society of Clinical Oncology |
| ClinicalTrials.gov | NCT02693535 |
2. Clinical Question
The central statistical question is whether participants in a molecularly defined cohort receiving a study drug experience an objective response or stable disease according to the response criteria specified for their disease at 16 weeks of treatment.
Population
People with advanced solid tumors, non-Hodgkin lymphoma, or multiple myeloma.
Intervention
FDA-approved drugs selected to target a specific abnormality in a tumor gene, administered within molecularly and clinically defined cohorts.
Comparator
There is no randomized comparator. The registered design is non-randomized and single-group.
Primary question
What proportion of participants in each cohort have complete response, partial response, or stable disease at 16 weeks of treatment?
3. Trial Design
TAPUR uses a non-randomized, single-group phase 2 structure rather than a conventional parallel-group randomized comparison. The statistical unit of interpretation is therefore the individual cohort, where participants share the same tumor type, genomic variant, and study drug.
Why the cohort structure matters
A conventional randomized trial generally estimates a treatment contrast because participants are assigned to different treatment groups. TAPUR is different. The registry describes cohorts in which participants have the same tumor type, genomic variant, and study drug. That structure makes the primary statistical question one of estimating response within a defined cohort rather than estimating a randomized treatment-versus-control effect.
This distinction affects nearly every downstream interpretation. A cohort response proportion can describe the observed disease-control experience under the study regimen, but it does not by itself establish how the same participants would have performed under an alternative treatment.
4. Registered Intervention Groups
The registry lists 12 arms. Several are closed to enrollment, while futibatinib, dabrafenib plus trametinib, and fam-trastuzumab deruxtecan-nxki are listed without the closed-to-enrollment designation in the current record.
| Arm | Intervention | Registry status |
|---|---|---|
| 1 | Sunitinib | Closed to enrollment |
| 2 | Trastuzumab and Pertuzumab | Closed to enrollment |
| 3 | Vemurafenib and Cobimetinib | Closed to enrollment |
| 4 | Olaparib | Closed to enrollment |
| 5 | Nivolumab and Ipilimumab | Closed to enrollment |
| 6 | Abemaciclib | Closed to enrollment |
| 7 | Talazoparib | Closed to enrollment |
| 8 | Atezolizumab and Talazoparib | Closed to enrollment |
| 9 | Tucatinib plus Trastuzumab Subcutaneous (SC) | Closed to enrollment |
| 10 | Futibatinib | Not listed as closed to enrollment |
| 11 | Dabrafenib plus Trametinib | Not listed as closed to enrollment |
| 12 | Fam-Trastuzumab Deruxtecan-Nxki (TDxD) | Not listed as closed to enrollment |
The presence of multiple intervention arms does not turn the study into a randomized multi-arm trial. The registered allocation remains non-randomized, and the design model remains single group.
5. Primary Endpoint
| Endpoint | Definition | Time frame | Disease-specific assessment |
|---|---|---|---|
| Objective Response Rate | % of participants in a cohort with complete or partial response or with stable disease according to standard response criteria | Assessed at 16 weeks of treatment | Solid tumors: RECIST criteria; non-Hodgkin lymphoma: Lugano Criteria; multiple myeloma: International Uniform Response Criteria for Multiple Myeloma |
A terminology point worth noticing
The registry defines the primary endpoint as “Objective Response Rate” while explicitly including stable disease alongside complete and partial response. That definition is important because stable disease is ordinarily distinguished from objective tumor response in many conventional oncology response frameworks. The statistical interpretation on this page therefore follows the registry's stated endpoint definition rather than substituting a different definition.
6. How the Endpoint Is Constructed
For a given TAPUR cohort, the registered endpoint can be represented conceptually as a proportion:
The registry expresses the endpoint as a percentage of participants in a cohort. The clinical response category is determined using the disease-specific response criteria identified in the registry.
This is fundamentally different from a hazard ratio. There is no time-to-event ratio in the registered primary endpoint. The principal quantity is a proportion assessed at a fixed treatment time point: 16 weeks.
Because the cohort is defined by tumor type, genomic variant, and study drug, the denominator and interpretation belong to that particular cohort. Combining cohorts with different diseases, genomic abnormalities, or drugs would change the scientific question and would require an explicit statistical justification.
7. Statistical Methodology
Analysis of a fixed-time proportion
For an endpoint defined as the percentage of participants meeting a specified response category at 16 weeks, the basic statistical object is a binomial proportion. A typical analysis reports the observed proportion together with a confidence interval describing uncertainty around the cohort estimate.
where x is the number of participants meeting the registered endpoint definition and n is the relevant cohort denominator.
The appropriate confidence interval depends on the prespecified statistical analysis. For a cohort with a modest number of participants or a proportion near 0% or 100%, methods designed for binomial proportions can behave differently from the simple normal approximation.
Why the analysis is cohort-specific
The cohort definition is part of the statistical design. A response proportion among participants sharing a tumor type, genomic variant, and study drug answers a narrower question than an overall response percentage across all 12 arms. The narrower cohort estimate is more directly aligned with the biological and treatment-specific question being studied.
No randomized treatment effect is estimated
Because allocation is non-randomized and the design model is single group, a standard randomized treatment-effect estimate such as a between-arm hazard ratio, risk ratio, or treatment-group mean difference is not the primary statistical quantity described by the registry.
A comparison with an external historical benchmark would require that benchmark to be explicitly defined, clinically comparable, and statistically prespecified. The ClinicalTrials.gov record does not identify such a comparator or non-inferiority margin.
Disease-specific response criteria
The registry uses different response frameworks for different disease classes: RECIST for solid tumors, the Lugano Criteria for non-Hodgkin lymphoma, and the International Uniform Response Criteria for Multiple Myeloma. This matters statistically because the observed endpoint depends on how response and stable disease are classified within each disease-specific framework.
8. Planned Analysis
No formal statistical analyses are posted to ClinicalTrials.gov for this record. The registry nevertheless identifies the endpoint and its assessment framework, which permits the statistical structure of the planned measurement to be described.
What is measured?
The percentage of participants in each cohort with complete response, partial response, or stable disease at 16 weeks.
What is the natural statistical scale?
A proportion or percentage within a defined cohort rather than a randomized treatment-effect estimate.
What uncertainty is relevant?
Sampling uncertainty around the observed cohort proportion, appropriately expressed with a confidence interval when formal results are available.
What defines the cohort?
The same tumor type, genomic variant, and study drug.
For a fixed-time binary endpoint, a formal statistical analysis would normally report the number meeting the endpoint, the relevant denominator, the resulting percentage, and an interval estimate for the proportion. The registry does not post those numerical analyses for TAPUR.
9. Statistical Methods Explained
Why is a proportion the central statistical quantity?
The primary endpoint is explicitly defined as a percentage of participants in a cohort meeting one of the registry's specified response categories at 16 weeks. That makes the observed cohort proportion the direct summary measure.
Why does the 16-week time point matter?
The endpoint is not simply the proportion of participants who ever experience a response or stable disease. The registry specifies that it is assessed at 16 weeks of treatment. The time point therefore becomes part of the endpoint definition.
Why should each cohort be analyzed separately?
Each cohort includes participants with the same tumor type, genomic variant, and study drug. These characteristics define the clinical question. Pooling unlike cohorts would combine different diseases, molecular abnormalities, and treatments into a statistic whose interpretation would be less specific.
Why is a confidence interval useful for a response percentage?
An observed percentage is calculated from a finite number of participants. A confidence interval communicates the statistical uncertainty around that estimate. A wider interval generally indicates less precision than a narrower interval, although the interval is not a range containing individual patient outcomes.
Why is this not the same as a randomized treatment effect?
Randomization creates balance between treatment groups in expectation and supports a causal treatment comparison. TAPUR is registered as non-randomized and single group. Its primary endpoint is therefore a cohort-level outcome rather than a randomized comparison between intervention and control groups.
Why do the response criteria differ by disease?
Solid tumors, non-Hodgkin lymphoma, and multiple myeloma have different disease manifestations and response-assessment frameworks. The registry therefore specifies RECIST for solid tumors, the Lugano Criteria for non-Hodgkin lymphoma, and the International Uniform Response Criteria for Multiple Myeloma.
10. Interpreting a Cohort Response Percentage
If a cohort's endpoint percentage were reported as a particular value, that percentage would represent the fraction of participants classified as having complete response, partial response, or stable disease at 16 weeks under the registry definition.
It would not automatically represent the probability that a randomly selected future patient will experience the same outcome, the probability of long-term survival, or a causal treatment effect relative to another therapy.
The percentage cannot be interpreted independently of its cohort denominator. A percentage based on a small cohort can have substantially more statistical uncertainty than the same percentage based on a much larger cohort.
Because the TAPUR registry definition includes stable disease along with complete and partial response, the percentage should not be silently relabeled as a conventional tumor-shrinkage response rate. The wording of the endpoint determines what the statistic represents.
11. What the Registry Does Not Establish About the Primary Endpoint
The registry identifies the primary endpoint, its 16-week assessment time, its cohort structure, and the disease-specific response criteria. It does not post numerical statistical analyses for that endpoint in the record described here.
Consequently, the primary endpoint should be understood as a prespecified measurement framework rather than as a currently reported treatment-effect estimate. In particular, the registry information does not provide a posted numerical response percentage, confidence interval, or p-value for the endpoint.
12. Design Features With Statistical Consequences
| Design feature | Registered TAPUR characteristic | Statistical consequence |
|---|---|---|
| Allocation | Non-randomized | The design does not create a randomized treatment-versus-control comparison. |
| Design model | Single group | The primary endpoint is naturally summarized within a cohort rather than through a randomized between-arm contrast. |
| Masking | None | Treatment assignment is not masked. |
| Phase | Phase 2 | The study is positioned as a phase 2 evaluation rather than a randomized phase 3 comparison. |
| Primary purpose | Treatment | The study evaluates treatment activity within the defined clinical and molecular cohorts. |
| Endpoint timing | 16 weeks of treatment | The assessment time is part of the primary endpoint definition. |
| Cohort definition | Same tumor type, genomic variant, and study drug | Interpretation should remain aligned with the individual molecular and treatment cohort. |
Non-randomization and causal interpretation
In a randomized trial, treatment allocation is deliberately constructed so that systematic differences between groups can be separated from treatment assignment under the design assumptions. In a non-randomized study, that protection is absent. Differences between observed outcomes and an external expectation can therefore reflect patient selection, disease characteristics, prior treatment, prognostic factors, or other differences in addition to treatment activity.
Single-group interpretation
The single-group structure makes the endpoint especially suitable for describing the activity observed within a defined cohort. It does not, by itself, quantify how much better the study drug performs than another available therapy.
13. Multiple Arms Are Not the Same as Multiple Randomized Comparisons
TAPUR contains 12 registered arms, but the study is still classified as non-randomized with a single-group design model. This distinction is important when thinking about multiplicity.
In a conventional multi-arm randomized trial, multiple treatment comparisons can create a formal family of hypothesis tests requiring prespecified control of type I error. Here, the registry describes a collection of cohort-specific treatment evaluations rather than randomized comparisons among the 12 arms.
12 arms
The registry lists 12 intervention groups, including individual drugs and drug combinations.
Non-randomized
Participants are not described as being randomized among those 12 interventions.
Single-group model
The registered design model is single group rather than parallel-group randomized.
Cohort-level question
The primary endpoint is defined within cohorts sharing tumor type, genomic variant, and study drug.
14. Endpoint Measurement Across Disease Types
| Disease category | Registry-specified response framework | Primary assessment |
|---|---|---|
| Advanced solid tumors | Response Evaluation Criteria for Solid Tumors (RECIST) | At 16 weeks of treatment |
| Non-Hodgkin lymphoma | Lugano Criteria | At 16 weeks of treatment |
| Multiple myeloma | International Uniform Response Criteria for Multiple Myeloma | At 16 weeks of treatment |
The statistical summary may have the same mathematical form across these disease categories—a percentage—but the underlying clinical classification is disease-specific. A response percentage is therefore meaningful only when the corresponding response framework and cohort definition are also understood.
15. Missing Data and Denominator Considerations
Fixed-time response endpoints require a clear definition of the analysis population and denominator, particularly when assessment occurs at a specified treatment week. The registry information provided for this record does not specify a missing-data or imputation method.
That distinction matters because several different denominators can produce different percentages from the same observed patients. A proportion among all enrolled participants, a proportion among participants evaluable at 16 weeks, and a proportion among participants with an available disease assessment are not automatically interchangeable.
16. Interim Analysis, Multiplicity, and Other Design Topics
The registry information for TAPUR identifies a phase 2, non-randomized, single-group design with 12 arms. It does not report a non-inferiority margin, crossover design, factorial design, Bayesian analysis, interim-analysis boundary, or formal multiplicity-adjustment strategy for the primary endpoint.
| Methodological topic | Registry information |
|---|---|
| Non-inferiority margin | Not reported |
| Crossover | Not reported |
| Factorial design | Not reported |
| Interim analysis | Not reported |
| Missing-data / imputation method | Not reported |
| Bayesian methods | Not reported |
| Formal multiplicity strategy | Not reported |
| Randomization stratification | Not applicable to the registered non-randomized allocation |
These omissions are statistically meaningful because each method addresses a different source of uncertainty or design complexity. They should not be inferred simply from the presence of multiple cohorts or interventions.
17. Enrollment and Study Timeline
Study start
The registry lists 2016-03-14 as the study start date.
Recruiting
The ClinicalTrials.gov record lists the study as Recruiting.
Primary completion date
The registry lists 2027-12-31 as the primary completion date.
The planned enrollment is 4200 participants across the study, with 12 registered arms. The primary endpoint is assessed at 16 weeks of treatment within the defined cohorts.
18. Statistical Interpretation of the TAPUR Design
A cohort-specific percentage can quantify how frequently the registered endpoint definition is observed at 16 weeks among participants in a defined tumor type, genomic variant, and study-drug cohort.
A cohort response percentage does not by itself establish that the study drug caused the observed outcome or that it performs better than an alternative treatment. Those conclusions require an appropriate comparator or additional assumptions and evidence.
The cohort is not defined only by clinical disease. The registry explicitly incorporates genomic variant and study drug. That creates a more specific estimand: observed activity within a molecularly and therapeutically defined population.
A response measure at 16 weeks is a landmark-type summary. It differs from a time-to-event endpoint in which the exact timing of events and censoring contributes to the analysis throughout follow-up.
19. Limitations
- Non-randomized allocation: the study does not use randomization to create comparable treatment groups.
- Single-group design: the registered model does not provide a concurrent randomized comparator for the primary endpoint.
- Cohort heterogeneity across the study: the registry includes different tumor types, genomic variants, and study drugs, making cohort-specific interpretation essential.
- Fixed-time endpoint: the primary endpoint is assessed at 16 weeks, so it does not by itself summarize the entire subsequent course of disease.
- Disease-specific assessment: response is evaluated using different criteria for solid tumors, non-Hodgkin lymphoma, and multiple myeloma.
- Unreported analysis details: the registry does not post formal statistical analyses or a missing-data/imputation strategy for the primary endpoint.
- Denominator sensitivity: interpretation of a fixed-time percentage depends on the population included in the denominator.
- Multiple intervention groups: the existence of 12 arms does not create randomized pairwise treatment comparisons.
- Generalizability: applicability of a cohort estimate depends on how closely a future patient resembles the specific tumor type, genomic variant, and clinical population represented by that cohort.
20. Why This Trial Matters Statistically
TAPUR is a useful statistical teaching case because it illustrates a different problem from the conventional randomized phase 3 trial. Instead of beginning with two randomized groups and estimating a treatment contrast, the analysis begins with a molecularly defined cohort and asks how often a prespecified clinical outcome occurs at a fixed treatment time.
| Concept | How it appears in TAPUR |
|---|---|
| Phase 2 design | The study is registered as Phase 2. |
| Non-randomized allocation | Allocation is registered as non-randomized. |
| Single-group design | The registered design model is single group. |
| Multiple intervention groups | 12 arms are registered. |
| Molecularly defined cohorts | Each cohort includes the same tumor type, genomic variant, and study drug. |
| Fixed-time endpoint | The primary endpoint is assessed at 16 weeks of treatment. |
| Binomial proportion | The endpoint is expressed as a percentage of participants meeting the specified response categories. |
| Disease-specific response criteria | RECIST, Lugano Criteria, and International Uniform Response Criteria for Multiple Myeloma are specified. |
| Causal interpretation | Non-randomized single-group data require caution when separating treatment activity from patient and disease characteristics. |
| Denominator definition | The interpretation of a fixed-time percentage depends on the cohort and analysis population used. |
21. Sources
- ClinicalTrials.gov: TAPUR, NCT02693535.
- PubMed: PubMed record 42456088.
- PubMed: PubMed record 42060861.
- PubMed: PubMed record 42018965.
- PubMed: PubMed record 41538762.
- PubMed: PubMed record 41100773.
Continue through the Clinical Biostats knowledge graph
Explore additional biostatistics tutorials, statistical calculators, and clinical trial analyses through Clinical Biostats.
22. Record Summary
TAPUR is a phase 2, non-randomized, single-group study with 12 registered arms and planned enrollment of 4200 participants. Its primary endpoint is an objective response rate defined by the registry as the percentage of participants in a cohort with complete or partial response or stable disease according to standard response criteria, assessed at 16 weeks of treatment. The statistical structure is therefore centered on cohort-specific fixed-time proportions rather than randomized treatment contrasts. The cohort definition—same tumor type, genomic variant, and study drug—and the disease-specific response criteria are essential to interpreting any future numerical result.