This page separates reported trial results from statistical interpretation. The numerical trial results presented here are restricted to the statistical analyses posted in the ClinicalTrials.gov data posted on ClinicalTrials.gov for EURTAC.
1. Trial at a Glance
EURTAC was a randomized, open-label, parallel phase 3 treatment trial comparing an erlotinib group with a standard chemotherapy group in non-small cell lung cancer. The registry reports 174 enrolled participants and two treatment arms. Its registered primary endpoint was progression-free survival, with overall survival posted as a secondary endpoint analysis.
| Feature | EURTAC |
|---|---|
| Trial name | EURTAC |
| Phase | Phase 3 |
| Condition | Non-Small Cell Lung Cancer |
| Design | Randomized, parallel, unmasked |
| Primary purpose | Treatment |
| Enrollment | 174 |
| Primary endpoint | Progression Free-survival |
| Primary endpoint type | Time-to-event |
| Primary analysis | Log-rank test |
| Primary effect measure | Hazard ratio |
| Hypothesis type | Superiority |
| Trial status | Completed |
| ClinicalTrials.gov | NCT00446225 |
2. Clinical Question
The statistical question is whether randomized assignment to erlotinib produces a different time-to-event outcome than assignment to standard chemotherapy in the registered trial population. The primary question is expressed through progression-free survival, defined in the registry as the time from enrollment to tumor progression or death from any cause, whichever occurs first.
Population
Patients enrolled in the EURTAC phase 3 trial for non-small cell lung cancer, with the study described as treating patients with mutations in the TK domain of EGFR.
Intervention
Erlotinib group. Erlotinib is the drug intervention identified for this treatment arm in the registry data.
Comparator
Standard chemotherapy group. The registered chemotherapy interventions include carboplatin, gemcitabine, docetaxel, and cisplatin.
Primary question
Does erlotinib improve progression-free survival relative to standard chemotherapy under a superiority hypothesis?
3. Trial Design
Erlotinib
- Erlotinib is the registered drug intervention for this group.
- The posted primary and secondary analyses compare this group with standard chemotherapy.
Standard chemotherapy
- The registry identifies carboplatin, gemcitabine, docetaxel, and cisplatin among the chemotherapy interventions.
- The statistical analyses posted by the registry use this group as the comparator.
4. Endpoints
| Endpoint | Registry definition / time frame | Role |
|---|---|---|
| Progression Free-survival | The time from enrollment in the study to tumor progression or death from any cause, whichever occurs first. From the date of randomization to the date of last follow up, assessed up to 24 months. | Primary |
| Overall Survival | From the date of randomization to the date of last follow up, assessed up to 24 months. | Secondary |
The primary endpoint is explicitly time-to-event. This distinction is important because progression-free survival is not simply a proportion of patients who progress: each participant contributes a follow-up time, and participants who have not experienced progression or death by their last follow-up can contribute censored information.
5. Statistical Methodology
Primary analysis: log-rank test
The registry reports a log-rank analysis for the primary progression-free survival endpoint. The log-rank test is designed for comparing survival or time-to-event distributions between groups while accounting for the timing of events and censoring.
For EURTAC, the registry identifies the hypothesis as superiority. The posted analysis therefore evaluates whether the randomized treatment groups differ on the time-to-event endpoint rather than testing a non-inferiority margin.
Hazard ratio
The effect measure reported for both posted analyses is the hazard ratio. A hazard ratio compares the estimated instantaneous event rates between the two groups within a time-to-event framework.
An HR below 1 indicates a lower estimated instantaneous event rate in the erlotinib group relative to the standard chemotherapy group, under the model and follow-up represented by the analysis.
Confidence intervals
Each posted hazard ratio is accompanied by a two-sided 95% confidence interval. The interval describes statistical uncertainty around the estimated relative treatment effect. It does not describe the range of outcomes that individual patients will experience.
P-values
The posted analyses also include p-values. A p-value addresses compatibility of the observed data with the specified null hypothesis; it is not a measure of how large or clinically important an effect is. The hazard ratio and its confidence interval are therefore essential parts of the result.
Time-to-event analysis and censoring
For a time-to-event endpoint, not every participant necessarily experiences the event during the observation period. Kaplan-Meier estimation is the standard descriptive framework for displaying such data, while the log-rank test compares the groups and a hazard ratio provides a relative effect measure. The EURTAC ClinicalTrials.gov record report the log-rank method and hazard ratio but do not provide the underlying event-by-event data needed to reconstruct a Kaplan-Meier curve.
6. Primary Result: Progression Free-survival
Progression Free-survival is the registered primary endpoint. The registry defines it as the time from enrollment to tumor progression or death from any cause, whichever occurs first, with the time frame specified as from randomization to last follow-up and assessed up to 24 months.
Hazard ratio for progression or death
95% CI: 0.27–0.64 · P = 0.0001
Comparison: Erlotinib Group vs Standard Chemotherapy Group
An HR of 0.42 means that the estimated instantaneous rate of progression or death in the erlotinib group was 42% of the estimated rate in the standard chemotherapy group under the time-to-event analysis. Expressed as a relative difference, this corresponds to an estimated 58% lower hazard of progression or death.
The HR does not mean that 58% of patients avoided progression, that progression-free survival increased by 58%, or that an individual patient's probability of progression or death was reduced by exactly 58%. Hazard is an instantaneous event-rate concept, not a direct probability.
The two-sided 95% CI of 0.27–0.64 describes uncertainty around the estimated HR. It remains below 1 throughout the interval, so the null value of 1 is not contained in the reported confidence interval. The interval is nevertheless not a guarantee that the true effect is any particular value within it.
The p-value of 0.0001 provides evidence against the specified null hypothesis in the reported log-rank analysis. It does not measure effect size, does not indicate the probability that the treatment is effective, and does not replace the HR or confidence interval.
Because this is a time-to-event analysis, interpretation also depends on censoring and on the assumptions underlying the hazard-ratio representation. The ClinicalTrials.gov record does not provide enough information to assess the proportional-hazards assumption directly.
7. Secondary Result: Overall Survival
Overall Survival was posted as a secondary time-to-event analysis. The registry supplies the same broad follow-up frame: from randomization to last follow-up, assessed up to 24 months.
Hazard ratio for overall survival
95% CI: 0.36–0.99 · P = 0.043
Comparison: Erlotinib Group vs Standard Chemotherapy Group
| Endpoint | Method | Effect measure | Estimate | 95% CI | P-value |
|---|---|---|---|---|---|
| Progression Free-survival | Log Rank | Hazard ratio | 0.42 | 0.27–0.64 | 0.0001 |
| Overall Survival | Log Rank | Hazard ratio | 0.59 | 0.36–0.99 | 0.043 |
An OS HR of 0.59 means that the estimated instantaneous rate of death in the erlotinib group was 59% of that in the standard chemotherapy group under the reported time-to-event analysis. In relative terms, the estimate corresponds to approximately a 41% lower estimated hazard of death.
This does not mean that 41% of patients survived, that 41% of patients were cured, or that each patient experienced the same reduction in mortality risk. It is a relative time-to-event effect measure.
The 95% CI of 0.36–0.99 is substantially wider than the primary PFS interval, reflecting greater uncertainty around the posted OS estimate. Its upper limit is close to 1, so the estimate is compatible with a relatively broad range of effects within the reported interval.
The p-value of 0.043 is a test result, not an effect-size metric. It should be interpreted alongside the HR, confidence interval, analysis method, endpoint role, and trial design rather than in isolation.
Because overall survival is a time-to-event endpoint, censoring and the assumptions used to summarize relative hazards remain relevant. The ClinicalTrials.gov record does not provide enough detail to independently evaluate those assumptions or reconstruct the survival curve.
8. Comparing the Two Time-to-Event Results
The two posted analyses tell related but distinct statistical stories. Progression Free-survival is the registered primary endpoint and has an estimated HR of 0.42, whereas Overall Survival is a secondary endpoint with an estimated HR of 0.59.
| Feature | Progression Free-survival | Overall Survival |
|---|---|---|
| Endpoint role | Primary | Secondary |
| Endpoint type | Time-to-event | Time-to-event |
| Method | Log Rank | Log Rank |
| Effect measure | Hazard ratio | Hazard ratio |
| Estimate | 0.42 | 0.59 |
| 95% CI | 0.27–0.64 | 0.36–0.99 |
| P-value | 0.0001 | 0.043 |
| Hypothesis | Superiority | Superiority |
The HRs should not be compared as though they were two measurements of exactly the same quantity. PFS and OS have different event definitions. For PFS, the event is tumor progression or death, whichever occurs first. For OS, the event is death. Consequently, the two hazard ratios answer different questions.
The different widths of the confidence intervals also matter. The PFS interval, 0.27–0.64, gives a more precise estimate than the OS interval, 0.36–0.99, in the statistical sense of interval width. Precision and statistical significance are related but are not interchangeable: a narrower interval can coexist with a larger p-value if the estimated effect is closer to the null.
9. What a Hazard Ratio of 0.42 Means
The PFS HR of 0.42 is a ratio of estimated event rates within a time-to-event framework. It indicates a lower estimated instantaneous rate of progression or death for the erlotinib group than for the standard chemotherapy group.
It is not a direct probability, risk ratio, odds ratio, median difference, or absolute difference in survival. It also does not imply that every patient experiences the same relative reduction in event risk.
The 95% CI of 0.27–0.64 expresses uncertainty around the estimated HR. It is more informative than the point estimate alone because it shows both the direction of the estimated effect and the range of parameter values compatible with the statistical model and data at the stated confidence level.
The p-value of 0.0001 addresses the null hypothesis tested by the reported analysis. It does not quantify the magnitude of benefit. Two studies can have identical hazard ratios but different p-values if their information levels differ.
10. Statistical Methods Explained
Why was a log-rank test used?
The registered primary endpoint is time-to-event, not a simple binary outcome. A log-rank test is appropriate for comparing the timing of events between randomized groups while incorporating follow-up and censoring. In EURTAC, the registry specifically identifies the log-rank method for both posted analyses.
What does the PFS hazard ratio of 0.42 mean?
It means the estimated instantaneous rate of progression or death in the erlotinib group was 42% of the corresponding estimated rate in the standard chemotherapy group. The complementary interpretation is an estimated 58% lower hazard. This is a relative rate interpretation, not a statement that 58% of patients avoided the event.
Why is the confidence interval important?
The point estimate alone gives only one estimate of the relative treatment effect. The 95% CI of 0.27–0.64 communicates uncertainty around that estimate. Its width provides information about precision, while its location relative to 1 helps describe compatibility with the null value for a hazard ratio.
Why is the p-value not the effect size?
The p-value depends on both the observed effect and the amount of statistical information available. A very small p-value can accompany a modest effect in a large information set, while a substantial effect can have a less extreme p-value when uncertainty is high. For that reason, the HR and CI should remain central to interpretation.
Why are PFS and OS not interchangeable?
PFS counts tumor progression or death as the first event, whereas OS concerns death. A patient can therefore contribute a PFS event without having an OS event at the same time. The two endpoints measure different aspects of the clinical time course and should be interpreted separately.
What does the OS HR of 0.59 tell us?
It estimates the relative rate of death between the randomized groups. An HR of 0.59 corresponds to an estimated 41% lower hazard of death in the erlotinib group relative to standard chemotherapy. It does not specify an absolute survival probability or tell us how long an individual patient will live.
Why does the proportional-hazards assumption matter?
A single hazard ratio is especially straightforward to interpret when the relative hazards are reasonably stable over time. If hazards vary substantially in a non-proportional way, one HR can compress a more complicated time-varying pattern into a single number. The registry-reported EURTAC registry data do not contain enough information to determine whether that assumption is adequately satisfied.
11. Statistical Interpretation of the Primary Analysis
The primary PFS analysis combines three complementary pieces of information: the estimated hazard ratio, its confidence interval, and the p-value from the log-rank analysis.
Effect size
HR 0.42 describes the estimated relative difference in the instantaneous rate of progression or death between randomized groups.
Precision
The 95% CI of 0.27–0.64 describes uncertainty around the HR estimate and remains below the null value of 1.
Hypothesis test
P = 0.0001 is the reported log-rank test result under the superiority framework.
Clinical meaning
The registry result supports a difference in the time-to-event experience between randomized groups, but the HR alone does not provide an absolute treatment benefit.
This is an important distinction in clinical-trial interpretation. A statistically strong comparison does not automatically answer every clinical question. The result establishes the statistical evidence represented by the endpoint and analysis; understanding absolute survival, patient-level benefit, toxicity, quality of life, and other considerations would require additional data.
12. Interpretation of the Overall Survival Analysis
The OS analysis has an HR of 0.59 with a two-sided 95% CI of 0.36–0.99 and a p-value of 0.043. The estimate is below the null value of 1, and the upper confidence limit is close to 1.
That combination is statistically informative because the point estimate and interval should be considered together. The point estimate suggests a lower estimated rate of death in the erlotinib group, while the interval indicates meaningful uncertainty about the exact magnitude of the relative effect.
The OS result should also be kept in its registered role as a secondary endpoint. The ClinicalTrials.gov record does not identify an endpoint hierarchy beyond the designation of PFS as primary and OS as secondary, nor do they provide a detailed multiplicity procedure. Consequently, this page does not assign an additional confirmatory interpretation to the OS p-value beyond what the registry reports.
13. Trial Timeline
Trial start
The EURTAC trial is recorded as starting on February 15, 2007.
Randomized comparative treatment study
The trial is recorded as a phase 3, randomized, parallel study with no masking and a primary purpose of treatment.
Primary completion
The registry records April 11, 2012 as the primary completion date.
Results posted
The registry record is marked completed and indicates that results and statistical analyses have been posted.
14. Safety Results
The ClinicalTrials.gov record provides serious adverse event counts by randomized treatment arm. They do not provide a more detailed breakdown of individual adverse-event types or grades.
| Safety measure | Erlotinib Group | Standard Chemotherapy Group |
|---|---|---|
| Serious adverse events, affected / at risk | 27 / 86 | 25 / 87 |
The denominator is important. The ClinicalTrials.gov record reports 27 of 86 patients in the erlotinib group and 25 of 87 patients in the standard chemotherapy group as affected by serious adverse events. These counts should not be confused with the efficacy analysis population of 174 enrolled participants because the registry-reported safety field identifies its own affected/at-risk denominators.
The registry data do not provide enough detail here to attribute particular serious adverse events to specific drugs, characterize severity beyond the serious-adverse-event designation, or perform a formal comparative safety hypothesis test. The appropriate interpretation is therefore descriptive.
15. Analysis Populations and What Is Reported
The registry-reported EURTAC data identify enrollment of 174 participants and provide serious-adverse-event denominators of 86 and 87. They do not provide explicit definitions of an intention-to-treat population, per-protocol population, or modified analysis population.
| Population / feature | Information reported |
|---|---|
| Total enrollment | 174 |
| Erlotinib safety denominator | 86 |
| Standard chemotherapy safety denominator | 87 |
| Formal ITT definition | Not reported in the ClinicalTrials.gov record |
| Per-protocol definition | Not reported in the ClinicalTrials.gov record |
| Missing-data method | Not reported in the ClinicalTrials.gov record |
| Imputation method | Not reported in the ClinicalTrials.gov record |
For a randomized time-to-event analysis, the treatment assignment is normally retained as the basis of the primary efficacy comparison. However, because the ClinicalTrials.gov record does not explicitly define the analysis population, this page does not assign a formal ITT label to the posted EURTAC analysis.
16. Randomization and Causal Interpretation
Randomization is one of the central statistical features of EURTAC. The registry identifies the allocation as randomized, which means treatment assignment rather than post-treatment outcome was used to create the comparison.
Randomization matters because, in expectation, it balances both measured and unmeasured prognostic factors between treatment groups. This is the foundation for interpreting a difference in outcomes as a treatment-assignment comparison rather than simply an observational association.
The fact that the trial is unmasked is also important. Without masking, knowledge of treatment assignment can potentially influence aspects of treatment administration, assessment, reporting, or other processes. The magnitude and relevance of such effects depend on the endpoint and trial procedures. For a time-to-event endpoint such as PFS, the exact mechanisms by which assessment knowledge could matter cannot be evaluated from the ClinicalTrials.gov record alone.
17. Missing Data, Censoring, and Follow-Up
Progression-free survival and overall survival are time-to-event outcomes, so censoring is an intrinsic part of their statistical analysis. A patient who has not experienced the relevant event by the last observed follow-up does not simply disappear from the analysis; their available follow-up contributes information up to the censoring time.
The registry-reported EURTAC data specify the primary PFS time frame as from randomization to last follow-up, assessed up to 24 months. They do not provide the individual censoring rules, reasons for censoring, missing-assessment rules, or imputation methods.
Censoring
The registry defines follow-up through a last-follow-up framework, consistent with the need to account for incomplete event observation in time-to-event analysis.
Imputation
No missing-data or imputation procedure is reported in the ClinicalTrials.gov record.
Follow-up limit
The registered endpoint time frame states that assessment was up to 24 months.
Curve reconstruction
The registry-reported summary statistics are insufficient to recreate a valid Kaplan-Meier curve.
This limitation matters because survival methods rely on assumptions about censoring. In standard analyses, censoring should not systematically depend on the unobserved event process in a way that invalidates the estimator. The ClinicalTrials.gov record does not allow that assumption to be examined directly.
18. Superiority Testing
The registry identifies the hypothesis type for both posted analyses as superiority. This means the statistical question is whether the treatment groups differ, rather than whether one treatment is no worse than another within a prespecified non-inferiority margin.
For a hazard-ratio analysis, 1 is the conventional null value. The ClinicalTrials.gov record reports HR 0.42 for PFS and HR 0.59 for OS.
No non-inferiority margin is reported because the posted analyses are identified as superiority analyses. It would therefore be inappropriate to interpret these results using non-inferiority logic or to introduce a margin that is not present in the ClinicalTrials.gov record.
19. Interim Analysis, Bayesian Methods, and Multiplicity
The registry-reported EURTAC registry data do not report an interim-analysis procedure, alpha-spending method, Bayesian analysis, or detailed multiplicity-adjustment strategy.
| Design topic | What the ClinicalTrials.gov record reports |
|---|---|
| Interim analysis | Not reported in the ClinicalTrials.gov record |
| Alpha spending | Not reported in the ClinicalTrials.gov record |
| Bayesian methods | Not reported in the ClinicalTrials.gov record |
| Non-inferiority margin | Not applicable to the reported superiority hypothesis |
| Multiplicity adjustment | Not reported in the ClinicalTrials.gov record |
This is an example of why statistical interpretation should remain tied to documented methods. It is possible to describe how a method would normally be handled, but it would not be appropriate to attribute an unreported interim boundary, Bayesian prior, or multiplicity strategy to EURTAC.
20. What the Confidence Intervals Tell Us
The 95% CI for the PFS HR is 0.27–0.64. The interval lies below 1, consistent with the direction of the point estimate of 0.42. Its width reflects uncertainty around the estimated relative treatment effect.
The 95% CI for the OS HR is 0.36–0.99. It also lies below 1, but its upper boundary is close to the null value. This indicates more uncertainty about the precise magnitude of the OS effect than would be suggested by the PFS point estimate alone.
A confidence interval should not be interpreted as a range containing a fixed percentage of individual patient outcomes. It is an interval estimate for a population-level parameter under the statistical model and sampling framework used to produce it.
21. Why the P-Values Should Be Read Carefully
The PFS p-value is 0.0001, while the OS p-value is 0.043. These values indicate different levels of compatibility with the respective null hypotheses under the reported log-rank analyses, but they should not be converted into probabilities that the treatment works or into measures of clinical importance.
The PFS result combines a relatively low point estimate with a confidence interval that excludes 1. The OS result has a point estimate below 1 but a confidence interval extending to 0.99. The difference illustrates why reporting only p-values can obscure important information about the estimated size and precision of treatment effects.
P = 0.0001
Reported for the primary PFS log-rank analysis. It addresses the statistical test, not the size of the HR.
P = 0.043
Reported for the secondary OS log-rank analysis. It should be interpreted in the context of the endpoint's secondary role and the documented analysis framework.
22. Important Limitations and Interpretation Issues
- Summary-data limitation: the ClinicalTrials.gov record provides hazard ratios, confidence intervals, and p-values but not patient-level event and censoring information.
- No Kaplan-Meier reconstruction: the available summary statistics are insufficient to recreate a valid Kaplan-Meier curve.
- Proportional-hazards assumption: a single hazard ratio is most straightforward to interpret when relative hazards are reasonably stable over time. The ClinicalTrials.gov record does not allow this assumption to be assessed.
- Endpoint distinction: PFS and OS have different event definitions and therefore answer different clinical questions.
- Secondary endpoint: OS is identified as secondary rather than primary in the ClinicalTrials.gov record.
- Multiplicity information: the ClinicalTrials.gov record does not report a multiplicity-adjustment procedure.
- Missing-data methods: no imputation or detailed missing-assessment strategy is reported in the ClinicalTrials.gov record.
- Analysis populations: explicit ITT and per-protocol definitions are not reported.
- Safety interpretation: the serious-adverse-event data are descriptive counts by arm; the ClinicalTrials.gov record does not include a formal comparative safety analysis.
- Unmasked design: treatment assignment was not masked, which is relevant to interpretation of trial conduct and endpoint assessment.
- Generalizability: the registry describes a specific phase 3 trial population with non-small cell lung cancer and EGFR TK-domain mutations; the ClinicalTrials.gov record does not provide a broader population characterization.
23. Why This Trial Matters Statistically
EURTAC is a useful teaching case because its posted results illustrate the core statistical structure of a randomized time-to-event comparison without requiring a large collection of secondary analyses. The primary endpoint is explicitly time-to-event, the analysis is identified as a log-rank test, and the effect measure is a hazard ratio with a two-sided 95% confidence interval.
| Concept | How it appears in EURTAC |
|---|---|
| Randomization | The registry identifies the allocation as randomized. |
| Parallel design | The design model is recorded as parallel with two arms. |
| Time-to-event endpoint | Progression Free-survival is the registered primary endpoint. |
| Log-rank test | Used for the posted PFS and OS analyses. |
| Hazard ratio | Reported as the effect measure for both analyses. |
| Confidence interval | Both hazard ratios have two-sided 95% confidence intervals. |
| P-value | Reported for both primary and secondary statistical analyses. |
| Superiority | The registry identifies superiority as the hypothesis type. |
| Endpoint hierarchy | PFS is primary; OS is secondary. |
| Censoring | Relevant because both endpoints are time-to-event measures. |
| Unmasked design | The masking field is listed as none. |
| Safety denominators | Serious adverse events are reported as affected patients over patients at risk. |
The most important statistical lesson is that the hazard ratio should never be read in isolation. A complete interpretation requires the endpoint definition, treatment comparison, analysis method, confidence interval, p-value, analysis population, and assumptions of the time-to-event model. EURTAC provides a compact example of how these pieces fit together.
24. A Practical Reading Sequence for EURTAC
A useful way to read the statistical results is to move from design to endpoint to effect estimate rather than starting with the p-value.
This sequence prevents a common statistical mistake: interpreting a small p-value before establishing what was actually tested. In EURTAC, the primary test concerns progression-free survival, not survival in general. The HR of 0.42 describes the relative event rate for progression or death, while the OS HR of 0.59 concerns death as a separate secondary endpoint.
25. Results Summary
| Measure | Erlotinib vs Standard Chemotherapy | Statistical interpretation |
|---|---|---|
| Primary PFS HR | 0.42 | Estimated 58% lower hazard of progression or death |
| PFS 95% CI | 0.27–0.64 | Two-sided uncertainty interval entirely below 1 |
| PFS P-value | 0.0001 | Reported log-rank test result |
| Secondary OS HR | 0.59 | Estimated 41% lower hazard of death |
| OS 95% CI | 0.36–0.99 | Two-sided uncertainty interval with upper limit near 1 |
| OS P-value | 0.043 | Reported log-rank test result |
| Serious AEs | 27/86 vs 25/87 | Descriptive affected/at-risk counts |
These are the statistical results posted on ClinicalTrials.gov for EURTAC. No median PFS, median OS, survival percentages, subgroup estimates, response rates, or additional efficacy estimates are added here because they are not contained in the ClinicalTrials.gov record.
26. Related Tutorials
Learn more about the methods used in this trial:
27. Related Calculators
28. Sources
- ClinicalTrials.gov: EURTAC, NCT00446225.
- PubMed record: PMID 22285168.
- PubMed record: PMID 26181014.
- PubMed record: PMID 25806291.
- PubMed record: PMID 24493829.
Continue with the underlying statistical methods
Explore the survival-analysis concepts that make the EURTAC results interpretable, from Kaplan-Meier estimation and log-rank testing to hazard ratios and confidence intervals.
29. Record Summary
EURTAC is a randomized phase 3, parallel, unmasked clinical trial with 174 enrolled participants and two treatment groups: an erlotinib group and a standard chemotherapy group. Its registered primary endpoint was progression-free survival, analyzed as a time-to-event endpoint using a log-rank test and summarized with a hazard ratio.
The posted primary analysis reports a PFS HR of 0.42 with a two-sided 95% CI of 0.27–0.64 and P = 0.0001. The secondary overall-survival analysis reports an OS HR of 0.59 with a two-sided 95% CI of 0.36–0.99 and P = 0.043. These estimates indicate lower estimated event hazards in the erlotinib group under the reported analyses, while the confidence intervals quantify uncertainty around those relative effects.
The statistical interpretation should remain anchored to the endpoint definitions. PFS concerns tumor progression or death, whereas OS concerns death. Neither hazard ratio is an absolute risk measure, and neither p-value is a measure of treatment-effect magnitude. The available registry data also do not provide enough information to assess proportional hazards, reconstruct Kaplan-Meier curves, evaluate missing-data procedures, or describe an explicit multiplicity strategy.