This page separates reported trial information from statistical interpretation. The ClinicalTrials.gov record identifies a completed, single-arm study with results posted, but no formal statistical analyses are included in the ClinicalTrials.gov record.
1. Trial at a Glance
GioTag was a completed industry-sponsored, single-arm study of afatinib followed by osimertinib in non-small-cell lung cancer. The study enrolled 204 participants and registered one primary endpoint focused on time on treatment across the afatinib-to-osimertinib sequence.
| Feature | GioTag |
|---|---|
| Trial name | GioTag |
| NCT ID | NCT03370770 |
| Brief title | Afatinib Osimertinib Sequencing NIS |
| Condition | Carcinoma, Non-Small-Cell Lung |
| Status | COMPLETED |
| Enrollment | 204 |
| Arms | 1 |
| Interventions | Afatinib; Osimertinib |
| Lead sponsor | Boehringer Ingelheim |
| Sponsor type | INDUSTRY |
| Start date | 2017-12-28 |
| Primary completion | 2019-11-28 | Results posted | Yes |
2. Clinical Question
The study evaluates a treatment sequence rather than a randomized comparison. The registered primary endpoint asks how long patients remain on treatment across an afatinib-first strategy followed by osimertinib.
Population
Patients represented in the GioTag study of carcinoma, non-small-cell lung. The ClinicalTrials.gov record identifies the serious-adverse-event denominator as patients with EGFR mutation-positive NSC.
Intervention sequence
Afatinib followed by osimertinib. The registered endpoint specifically measures time across the two treatment stages.
Comparator
There is no comparator arm. The study contains 1 arm, so the registry data do not provide a randomized control group against which the treatment sequence can be compared.
Primary question
What is the observed time on treatment with afatinib followed by osimertinib, according to the registry's prespecified definition?
3. Trial Design
What the single-arm structure changes statistically
A one-arm study can describe treatment exposure, response, treatment duration, and safety within the enrolled population. What it cannot establish on its own is how the observed outcome differs from what would have happened under a contemporaneous alternative treatment.
Randomization is the mechanism that makes treatment groups comparable, on average, with respect to both measured and unmeasured baseline characteristics. GioTag does not have that comparison because there is only one study arm. Consequently, a statistic calculated from the cohort is primarily descriptive unless an external comparator or prespecified benchmark is introduced. No such comparator is reported in the ClinicalTrials.gov record.
4. Endpoints
| Endpoint | Registry definition / time frame | Registry status |
|---|---|---|
| Time on Treatment With Afatinib (Gi(l)Otrif®) Followed by Osimertinib | Time on treatment, defined as time in months from the start date of Afatinib (Gi[l]otrif®) treatment ("start date of initial dose" for First-Line Treatment) to the end date of Osimertinib treatment (maximum between "end date of initial dose" and the last "end date of dose modification" for Second-Line Treatment) or death date due to any cause ("date of death"). | Primary endpoint; results posted |
The registered time frame states that data were collected from start of treatment until data entry completion, up to 96.8 months for the first analysis and up to 114.1 for the subsequent stated time horizon.
5. Understanding the Primary Endpoint
The GioTag endpoint combines the duration of two treatment stages into a single treatment-pathway measure. The starting point is the start of afatinib treatment. The endpoint continues through the subsequent osimertinib treatment stage and ends at the registry-defined end of osimertinib treatment or death.
The endpoint is defined using treatment dates recorded during the afatinib and osimertinib stages, with death due to any cause serving as an endpoint under the registered definition.
Why the definition is more than a simple duration
A treatment-duration endpoint can reflect several processes simultaneously: remaining on the first treatment, transitioning to the second treatment, continuing the second treatment, discontinuing treatment, and death. It therefore summarizes a clinical treatment pathway rather than isolating a single pharmacologic effect.
This distinction is important when interpreting the statistic. A longer observed treatment duration can arise from prolonged treatment continuation, successful sequencing, different discontinuation patterns, or other characteristics of the enrolled population. Without a comparator group, the data do not identify which portion of the observed duration is attributable specifically to the treatment strategy rather than to patient or disease characteristics.
6. Planned Analysis
For an endpoint defined as time from treatment initiation to treatment discontinuation or death, a typical time-to-event analysis would account explicitly for the timing of events and for participants who have not yet reached the endpoint at their last available follow-up. Kaplan-Meier methods are commonly used to estimate the distribution of time-to-event outcomes, while a Cox proportional-hazards model is commonly used when a relative hazard comparison between groups is scientifically appropriate.
Neither approach, however, creates a comparator that does not exist. In a single-arm study, a Kaplan-Meier estimate can describe the observed time-to-event distribution, but a Cox hazard ratio requires a comparison framework. The registry-reported GioTag data do not provide a control group or a formal statistical-analysis record from which such a comparison could be reported.
| Question | What the ClinicalTrials.gov record supports |
|---|---|
| Was the endpoint registered? | Yes. |
| Were results posted? | Yes. |
| Was a formal statistical-analysis record reported? | No; no formal statistical analyses were posted. |
| Is there a comparator arm? | No; the study has 1 arm. |
| Can a treatment-effect hazard ratio be reported? | No, not from the ClinicalTrials.gov record. |
| Can a causal comparison with another treatment be made? | No, not from this single-arm dataset alone. |
7. Statistical Methods Explained
Why would a time-to-event method be appropriate?
The primary endpoint is expressed in months and has a defined endpoint that can occur at different times for different participants. Time-to-event methods are designed for exactly this structure because they preserve information about when an event occurs rather than reducing every participant to a simple event/no-event classification at a fixed calendar date.
Why does censoring matter?
If a participant has not reached the endpoint by the last available observation, the participant's complete event time is unknown. A time-to-event analysis can use the information available up to that observation while recognizing that the final event time has not been observed. Treating every such participant as if the endpoint never occurred would discard information about follow-up duration.
Why can't GioTag produce a treatment hazard ratio from the ClinicalTrials.gov record?
A conventional treatment hazard ratio is a relative comparison between groups. GioTag has 1 arm. A single observed time-to-event distribution can be estimated, but there is no randomized second distribution in the registry-reported study against which to form a treatment hazard ratio.
What would a Kaplan-Meier curve show?
A Kaplan-Meier curve would describe the proportion of the cohort remaining free of the defined endpoint over time, subject to the study's event and censoring rules. It would be descriptive in a single-arm study unless compared with an independently justified reference distribution.
Why isn't a p-value automatically part of a results analysis?
A p-value requires a null hypothesis and a specified probability model for the test statistic. In a single-arm study, a meaningful hypothesis test would need a prespecified benchmark or reference value.
Why is the distinction between treatment duration and treatment efficacy important?
Time on treatment describes how long participants remained within the defined treatment sequence. It does not by itself establish that treatment caused the observed duration, nor does it directly quantify survival benefit, tumor response, or comparative efficacy. Those questions require their own endpoints and appropriate comparison frameworks.
8. Results Availability and Interpretation
The ClinicalTrials.gov record states that results are posted and identify 2 outcome measures posted. They also identify the primary endpoint and its definition. However, no statistical-analysis records are reported with the trial data used for this page.
That distinction is important. "Results posted" means that the registry contains outcome information; it does not necessarily mean that the ClinicalTrials.gov recordset contains a formal inferential comparison. Because the task requires every numerical result to come only from the ClinicalTrials.gov record, this page does not reconstruct or import estimates from the linked publications.
What can be concluded from the ClinicalTrials.gov record?
No formal statistical analyses were posted, and the study has 1 arm.
9. Safety
The registry profile reports serious adverse events by arm as 32/203 among patients identified as "EGFR Mutation-positive NSC." Because the study has only 1 arm, this is a within-study safety count rather than a randomized safety comparison.
| Safety measure | Reported value | Interpretation |
|---|---|---|
| Serious adverse events | 32/203 | Affected / at risk among patients with EGFR Mutation-positive NSC, as reported in the registry in the registry profile. |
Because there is no comparator arm, the reported 32/203 count cannot be converted into a relative safety risk versus another treatment using the ClinicalTrials.gov record. A within-cohort proportion can be calculated mathematically, but doing so would add a derived number that is not necessary to interpret the registry result and would not solve the underlying absence of a comparator.
10. Treatment Sequencing as a Statistical Question
GioTag is useful statistically because the intervention is a sequence rather than a single isolated treatment exposure. The endpoint begins with afatinib and continues through subsequent osimertinib treatment. This creates a treatment-pathway estimand rather than a simple "drug A versus drug B" comparison.
Sequence-level question
How long do participants remain on the defined afatinib-to-osimertinib treatment pathway before the registered endpoint occurs?
Comparative question
Would the observed treatment duration differ from another treatment strategy under otherwise comparable conditions?
GioTag answers
The registry-reported study design addresses the first question descriptively but does not provide the randomized comparison required for the second.
Interpretive consequence
Observed duration should not automatically be interpreted as a causal treatment effect.
11. The Role of the Missing Comparator
The absence of a comparator is not a minor reporting detail. It determines what kinds of statistical claims can be supported.
| Statistical question | Single-arm GioTag design |
|---|---|
| What happened in the enrolled cohort? | Can be described. |
| How long did participants remain on the defined treatment sequence? | Can be described using the registered endpoint. |
| What proportion experienced a serious adverse event? | Can be described within the registry-reported safety denominator. |
| Was the sequence better than a concurrent control? | Not answerable from the registry-reported design. |
| What is the randomized treatment hazard ratio? | Not available because there is no comparator arm. |
| Is a p-value for a treatment comparison available? | No formal comparison is reported. |
This is why the registry's own caveat is statistically consequential: without a comparator group, the observed results cannot be placed into the perspective of a contemporaneous alternative treatment using randomized evidence.
12. Missing Data, Censoring, and Endpoint Construction
The registered endpoint is constructed from treatment dates and death information. That means the quality of the endpoint depends on accurate recording of treatment initiation, treatment discontinuation, dose-modification dates, and death dates.
The ClinicalTrials.gov record does not specify a missing-data or imputation strategy, nor do they provide a statistical-analysis plan describing censoring rules beyond the endpoint definition itself. Accordingly, no particular imputation method or censoring convention should be attributed to GioTag beyond the registry-reported definition.
For time-to-event analysis, participants can contribute information even when the endpoint has not occurred by the last available observation. The exact statistical treatment depends on the prespecified analysis rules, which are not reported here.
13. Stratification, Interim Analysis, and Multiplicity
The ClinicalTrials.gov record does not report a stratification scheme, a formal interim-analysis procedure, an alpha-spending method, a multiplicity hierarchy, a non-inferiority margin, or a Bayesian analysis.
| Design topic | What is supported by the ClinicalTrials.gov record |
|---|---|
| Stratification | Not reported in the ClinicalTrials.gov record. |
| Interim analysis | Not reported in the ClinicalTrials.gov record. |
| Alpha spending | Not reported in the ClinicalTrials.gov record. |
| Multiplicity adjustment | Not reported in the ClinicalTrials.gov record. |
| Non-inferiority margin | Not applicable to the registry-reported single-arm description; no non-inferiority comparison is reported. |
| Bayesian methods | Not reported in the ClinicalTrials.gov record. |
| Missing-data / imputation method | Not reported in the ClinicalTrials.gov record. |
These omissions should not be interpreted as evidence that no such methods were used in the underlying study documentation. They mean only that the ClinicalTrials.gov record does not support attributing those methods to the analysis.
14. Statistical Interpretation of a Single-Arm Time-on-Treatment Endpoint
The registered endpoint provides a framework for describing the duration from initiation of afatinib through the defined endpoint following osimertinib treatment or death. It is therefore a measure of persistence within a prespecified treatment sequence.
It cannot establish that the treatment sequence produces a longer duration than another strategy because no randomized comparator is present in the registry-reported design.
If an appropriate statistical estimate were reported, its confidence interval would describe uncertainty around that estimate under the specified statistical model. A confidence interval would not transform the single-arm study into a randomized comparative trial.
Even where a formal hypothesis test is appropriate, a p-value describes compatibility with a specified null hypothesis; it does not measure the magnitude or clinical importance of an effect. In GioTag, the ClinicalTrials.gov record contains no formal comparative p-value.
15. Limitations
- No comparator group: the study has 1 arm, and the registry explicitly identifies the lack of a comparator as a limitation on interpretation.
- No formal statistical analyses reported: no formal statistical analyses were posted, so comparative estimates and inferential statistics cannot be reported from the ClinicalTrials.gov record.
- Endpoint complexity: time on treatment combines multiple treatment stages and uses treatment dates and death to define the endpoint.
- Potential selection effects: in a single-arm study, patient and disease characteristics cannot be balanced against a randomized control group through the study design.
- External comparison risk: comparing the observed outcome with results from another study can be confounded by differences in eligibility, calendar time, follow-up, treatment exposure, assessment methods, and patient characteristics.
- Incomplete methodological detail: the ClinicalTrials.gov record does not specify stratification, interim-analysis procedures, multiplicity control, imputation, or a detailed statistical-analysis plan.
- Safety comparison is unavailable: the reported 32/203 serious-adverse-event count has no randomized comparator in this study.
- Registry classification: the registry-reported endpoint is classified as binary even though its definition is expressed as time in months; this page preserves the registry classification rather than inferring a different official classification.
16. Why This Trial Matters Statistically
GioTag is a useful teaching example because it illustrates an important principle in clinical-trial statistics: the sophistication of an endpoint does not substitute for the absence of a comparator.
| Concept | How it appears in GioTag |
|---|---|
| Single-arm design | 1 study arm with 204 enrolled participants. |
| Treatment sequencing | Afatinib followed by osimertinib. |
| Time-to-event structure | The primary endpoint is defined as time in months from treatment initiation through the specified treatment endpoint or death. |
| Endpoint construction | The definition uses treatment start, treatment-end, dose-modification, and death dates. |
| Censoring considerations | Time-to-event analysis would ordinarily distinguish observed endpoints from incomplete follow-up. |
| Comparative inference | Not supported by a randomized control group in the registry-reported design. |
| Safety analysis | Serious adverse events are reported as 32/203 among the specified EGFR mutation-positive NSC population. |
| Statistical transparency | Results are posted, but the registry-reported statistical-analysis field is empty. |
| Generalizability | The observed cohort cannot automatically represent patients outside the study population. |
The statistical lesson is therefore broader than this particular treatment sequence. A clinical trial can have a clearly defined endpoint and still provide limited causal evidence if the design lacks a suitable comparison. Good statistical interpretation begins by identifying the estimand that the design can actually support.
17. Related Tutorials
Learn more about the methods used to understand this trial:
18. Related Calculators
19. Sources
- ClinicalTrials.gov: GioTag, NCT03370770.
- PubMed: PMID 32854536.
- PubMed: PMID 32757853.
- PubMed: PMID 31863283.
- PubMed: PMID 31370698.
- PubMed: PMID 30336693.
Continue through the Clinical Biostats statistical library
Explore tutorials and statistical calculators that explain survival analysis, clinical-trial design, confidence intervals, and treatment-effect estimation.
20. Record Summary
GioTag is a completed, industry-sponsored, single-arm study enrolling 204 participants and evaluating an afatinib-to-osimertinib treatment sequence in non-small-cell lung cancer. Its registered primary endpoint measures time on treatment from the start of afatinib through the defined end of osimertinib treatment or death, with data collected through the stated follow-up periods.
The central statistical issue is the distinction between descriptive evidence and comparative evidence. The study can describe what occurred within its enrolled cohort, but its 1-arm design does not provide a randomized comparator. The ClinicalTrials.gov record also contain no formal statistical-analysis records, so this page deliberately does not introduce hazard ratios, confidence intervals, p-values, or other inferential results that are not present in the ClinicalTrials.gov recordset.