This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record.
1. Trial at a Glance
FIRE-3 was a randomized, parallel-group, phase 3 treatment trial enrolling 568 participants with colorectal cancer and neoplasm metastasis. It compared two first-line regimens built around FOLFIRI: one containing cetuximab and the other containing bevacizumab.
| Feature | FIRE-3 |
|---|---|
| Phase | Phase 3 |
| Condition | Neoplasm Metastasis; Colorectal Cancer |
| Trial title | 5-FU, Folinic Acid and Irinotecan (FOLFIRI) Plus Cetuximab Versus FOLFIRI Plus Bevacizumab in First Line Treatment Colorectal Cancer (CRC) |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None |
| Primary purpose | Treatment |
| Enrollment | 568 |
| Number of arms | 2 |
| Status | Unknown |
| Start date | 2007-01 |
| Primary completion date | 2014-04 |
| Lead sponsor | PD Dr. med. Volker Heinemann |
| Sponsor type | Other |
| ClinicalTrials.gov | NCT00433927 |
2. Clinical Question
The central statistical question is whether first-line treatment with FOLFIRI plus cetuximab differs from FOLFIRI plus bevacizumab with respect to objective response rate, assessed approximately 6 months after randomisation.
Population
Participants with colorectal cancer and neoplasm metastasis enrolled in a phase 3 first-line treatment trial.
Intervention
5-FU, folinic acid and irinotecan (FOLFIRI) plus cetuximab.
Comparator
FOLFIRI plus bevacizumab.
Primary question
What is the objective response rate approximately 6 months after randomisation in the two randomized treatment groups?
3. Trial Design
FOLFIRI plus cetuximab
- 5-FU
- Folinic acid
- Irinotecan
- Cetuximab
FOLFIRI plus bevacizumab
- 5-FU
- Folinic acid
- Irinotecan
- Bevacizumab
The parallel design is important statistically because each participant contributes to one randomized treatment group. The treatment contrast is therefore defined at the level of the randomized groups rather than through a within-patient comparison.
4. Endpoints
| Endpoint | Registry definition / time frame | Statistical role |
|---|---|---|
| Objective response rate | Objective response rate; approximate 6 months after randomisation | Registered primary endpoint |
Objective response rate
Objective response rate is a proportion: it represents the fraction of evaluable participants who meet the prespecified criteria for an objective response during the defined assessment period. Unlike a time-to-event endpoint, it focuses on whether the response criterion has been achieved by the specified assessment time.
The registry gives the time frame as approximately 6 months after randomisation. That timing is part of the endpoint definition and matters when interpreting an observed response proportion: the same treatment comparison could produce a different response rate if response were assessed at a substantially different time.
5. Statistical Methodology
The registered primary endpoint is a binary clinical outcome assessed at a defined post-randomisation time frame. For a randomized two-arm trial, the principal statistical comparison is naturally framed as a comparison of response proportions between the two treatment groups.
The resulting proportion can be reported separately for each randomized treatment group. A between-group analysis can then quantify the difference or ratio between those proportions and provide an appropriate confidence interval.
Randomization and the treatment comparison
Randomization is the central design feature supporting the comparison between FOLFIRI plus cetuximab and FOLFIRI plus bevacizumab. Because treatment assignment is randomized, differences in objective response between the groups can be interpreted as treatment-group differences within the randomized trial framework, subject to the usual issues of endpoint assessment, missing observations, and adherence.
Binary endpoint analysis
Objective response rate differs fundamentally from overall survival or progression-free survival. A response endpoint is summarized as a proportion rather than as a survival curve or hazard. For this reason, the natural statistical quantities are response proportions and their between-group contrasts, rather than Kaplan-Meier estimates or hazard ratios.
Confidence intervals
An interval estimate around each response proportion or around the between-group treatment contrast quantifies statistical uncertainty. The width of that interval depends on the amount of information available and on the variability of the estimated proportion.
Analysis population
The registry record identifies enrollment of 568 participants but does not report a detailed analysis-population definition for the objective response endpoint. The distinction between all randomized participants, treated participants, and participants evaluable for response can be important for a binary endpoint, particularly when some participants do not have an assessable response at the specified time.
6. Planned Analysis
The registry says that objective response rate is the primary endpoint, measured approximately 6 months after randomisation. Because this is a binary response endpoint, the usual analysis would begin by estimating the response proportion in each randomized treatment group.
| Analysis component | How an endpoint of this type is typically evaluated |
|---|---|
| Within-arm response | Calculate the proportion of participants meeting the objective-response criterion in each treatment group. |
| Between-arm comparison | Compare the response proportions between FOLFIRI plus cetuximab and FOLFIRI plus bevacizumab. |
| Effect estimate | Report a prespecified contrast such as a risk difference or risk ratio, together with a confidence interval. |
| Statistical uncertainty | Use an appropriate confidence interval for the response proportion and the treatment-group contrast. |
| Assessment time | Interpret the response comparison in relation to the approximate 6-month post-randomisation assessment window. |
For a binary endpoint, a p-value addresses whether the observed treatment-group difference is inconsistent with a specified null hypothesis. It does not describe the magnitude or clinical importance of the difference. The effect estimate and its confidence interval are therefore important companions to any hypothesis test.
If the response proportion were higher in one randomized group, that would describe a difference in the probability of objective response at the specified assessment time. It would not by itself establish a difference in survival, duration of response, quality of life, or other outcomes that are not part of the registered primary endpoint.
7. Statistical Methods Explained
Why is objective response rate analyzed as a proportion?
Objective response is a categorical outcome: a participant either meets the objective-response definition or does not at the relevant assessment. Once those participant-level outcomes are summarized, the primary quantity is a proportion. This differs from a time-to-event endpoint, where the timing of progression or death is central to the analysis.
Why does the approximately 6-month time frame matter?
A response rate is dependent on when response is assessed. An assessment approximately 6 months after randomisation establishes a common time frame for the randomized comparison. Without a defined assessment period, response proportions could be difficult to compare because participants would have different opportunities to achieve a response.
Why is randomization important for a response-rate comparison?
Randomization creates the treatment groups before the response outcome is observed. This reduces systematic differences in treatment assignment and provides the basis for interpreting a difference in response between groups as a randomized treatment comparison rather than simply an observational association.
Why would a hazard ratio not be the natural primary measure here?
A hazard ratio is designed for a time-to-event endpoint, such as time until death or progression. The registered primary endpoint for FIRE-3 is objective response rate at an approximate assessment time, so the central statistical quantity is a response proportion and its between-group contrast rather than an event hazard.
What does a confidence interval add to an objective response rate?
A response proportion calculated from a finite number of participants is an estimate rather than an exact population quantity. A confidence interval expresses the statistical uncertainty surrounding that estimate. For a treatment comparison, an interval around the between-group contrast also shows how precisely the magnitude of the difference has been estimated.
Why should the analysis population be stated clearly?
The denominator is fundamental to a response rate. If participants who lack an evaluable response are handled differently, the resulting proportion can change. A rigorous analysis therefore needs a clearly defined response-evaluable population and a prespecified approach to participants without an assessment.
8. Interpreting the Primary Endpoint
What the endpoint measures
Objective response rate measures the proportion of participants achieving an objective response at approximately 6 months after randomisation.
What it does not measure
It is not a direct measure of overall survival, progression-free survival, duration of response, or long-term disease control.
Relative versus absolute effects
A difference in response proportions describes an absolute contrast, while a ratio of response proportions describes a relative contrast. They answer different questions.
Statistical versus clinical importance
A statistical test addresses compatibility with a null hypothesis. Clinical importance depends on the magnitude, precision, context, and consequences of the observed treatment difference.
For this trial, the primary endpoint should therefore be understood as a prespecified comparison of tumor response between two randomized first-line treatment strategies. The endpoint provides evidence about response, but response alone does not summarize every possible effect of treatment.
9. Trial Timeline
Trial start
The registry records a study start date of 2007-01.
Randomized treatment comparison
The trial uses randomized allocation, a parallel design, no masking, and a treatment primary purpose.
Primary endpoint assessment
Objective response rate is registered for assessment at approximately this time point.
Primary completion
The registry records a primary completion date of 2014-04.
10. Limitations
- No posted statistical analyses: the ClinicalTrials.gov record does not contain a posted statistical analysis for the registered primary endpoint.
- No numerical primary results: the registry record identifies objective response rate as the primary endpoint but does not report numerical response estimates, confidence intervals, or p-values.
- Analysis population: the registry record does not provide a detailed definition of the population used for the objective response analysis.
- Missing response assessments: the registry record does not specify how participants without an evaluable response at approximately 6 months are handled statistically.
- Endpoint methodology: the registry record identifies objective response rate and its time frame but does not provide a detailed statistical analysis plan for the endpoint.
- Limited design detail: the record identifies randomization, a parallel design, no masking, phase 3 status, and treatment purpose, but does not report additional statistical design features such as stratification, multiplicity procedures, interim-analysis rules, or missing-data methods.
- Status: the registry status is listed as UNKNOWN, so the current status cannot be characterized further from this record.
11. Why This Trial Matters Statistically
FIRE-3 is a useful teaching example because its registered primary endpoint illustrates a different statistical problem from the familiar time-to-event analyses used in many oncology trials. The core comparison is a binary response outcome assessed at a defined post-randomisation time.
| Concept | How it appears in FIRE-3 |
|---|---|
| Randomization | Participants were assigned using randomized allocation. |
| Parallel design | The trial has two parallel treatment arms. |
| Phase 3 design | The study is registered as phase 3. |
| Binary endpoint | Objective response rate is the registered primary endpoint. |
| Fixed assessment window | The primary endpoint is measured approximately 6 months after randomisation. |
| Between-group comparison | The primary statistical question concerns response under two randomized treatment strategies. |
| Confidence intervals | Precision around response proportions and treatment contrasts is important for interpretation. |
| Analysis population | The denominator used for objective response requires a clearly defined analysis population. |
| Missing assessments | Participants without an evaluable response can affect the estimated response rate and require a prespecified approach. |
The trial also illustrates why the choice of endpoint determines the statistical framework. Objective response rate is fundamentally different from a survival endpoint: the analysis centers on proportions and treatment contrasts rather than the distribution of event times.
12. Statistical Interpretation vs Clinical Interpretation
Statistical interpretation
The registered primary analysis concerns a comparison of objective response rates between two randomized treatment groups at approximately 6 months after randomisation.
Clinical interpretation
The response endpoint addresses whether patients achieved an objective tumor response within the specified assessment period. It does not by itself summarize all dimensions of treatment benefit or harm.
A statistically estimated response difference should therefore be presented with its uncertainty and with a clear definition of the denominator and assessment time. A large numerical response difference can be clinically relevant, but statistical significance and clinical importance are separate concepts.
13. What Would a Complete Response Analysis Show?
A complete statistical presentation of the primary endpoint would ordinarily include the response proportion in each randomized group, an explicitly defined between-group effect measure, a confidence interval, and an appropriate hypothesis test if one was prespecified. The interpretation would then distinguish the magnitude of the response difference from the evidence against the corresponding null hypothesis.
A positive or negative difference would describe the direction and magnitude of the contrast, while its confidence interval would quantify statistical precision. The actual FIRE-3 estimate is not reported in the ClinicalTrials.gov record.
For a proportion, uncertainty is especially important when the number of evaluable participants is smaller than the number originally randomized. The denominator should therefore always accompany the response percentage in a rigorous report.
14. Sources
- ClinicalTrials.gov: FIRE-3, NCT00433927.
Continue through the Clinical Biostats trial library
Explore additional clinical-trial analyses and statistical methods across randomized studies and their endpoints.
15. Record Summary
FIRE-3 is registered as a randomized, parallel-group, phase 3 treatment trial enrolling 568 participants with colorectal cancer and neoplasm metastasis. It compares FOLFIRI plus cetuximab with FOLFIRI plus bevacizumab, with objective response rate registered as the primary endpoint at approximately 6 months after randomisation.
Statistically, the key feature is that the primary endpoint is a binary response measure rather than a time-to-event endpoint. The appropriate framework therefore centers on response proportions, between-group contrasts, confidence intervals, and a clearly defined analysis population. The ClinicalTrials.gov record identifies the endpoint and its assessment time but contains no posted formal statistical analyses or numerical primary results.