This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results and trial-specific facts on this page are limited to the ClinicalTrials.gov record.
1. Trial at a Glance
LAURA is a randomized, parallel-group, double-masked phase 3 study evaluating osimertinib 80mg/40mg versus placebo osimertinib 80mg/40mg following chemoradiation in patients with stage III unresectable non-small cell lung cancer.
| Feature | LAURA |
|---|---|
| Trial name | LAURA |
| NCT ID | NCT03521154 |
| Phase | Phase 3 |
| Condition | Non Small Cell Lung Cancer (Stage III) |
| Design | Randomized, parallel |
| Masking | Double |
| Primary purpose | Treatment |
| Enrollment | 216.0 |
| Arms | 2 |
| Lead sponsor | AstraZeneca |
| Sponsor type | Industry |
| Status | Active, not recruiting |
| Start | 2018-07-19 |
| Primary completion | 2024-01-05 |
2. Clinical Question
The central statistical question was whether patients randomized to osimertinib 80mg/40mg would experience a different progression-free survival distribution than patients randomized to placebo osimertinib 80mg/40mg.
Population
Patients with stage III unresectable non-small cell lung cancer participating in the LAURA study.
Intervention
Osimertinib 80mg/40mg.
Comparator
Placebo osimertinib 80mg/40mg.
Primary question
Does osimertinib improve progression-free survival compared with placebo in the randomized study population?
3. Trial Design
LAURA was a phase 3, randomized, parallel-group, double-masked trial with a treatment purpose. The registry reports an enrollment of 216.0 participants and two treatment arms.
Osimertinib
- Osimertinib 80mg/40mg
- Randomized treatment arm
- Compared with placebo in the posted analyses
Placebo
- Placebo osimertinib 80mg/40mg
- Randomized comparator arm
- Reference group for reported treatment-effect estimates
4. Registered Primary Endpoint
| Endpoint | Registry definition | Time frame |
|---|---|---|
| Progression-free Survival (PFS) by Blinded Independent Central Review (BICR) | Time from randomisation until the date of objective disease progression or death (by any cause in the absence of progression) regardless of whether the patient withdraws from randomised therapy or receives another anti-cancer therapy prior to progression, based on BICR assessment according to RECIST v1.1 | Every 8 weeks for first 48 weeks, then every 12 weeks until BICR-confirmed radiological disease progression. Assessed up to date of DCO (05Jan24) to a maximum of approximately 63 months |
This is a classic time-to-event endpoint. The event is the earlier relevant occurrence defined by the registry: objective disease progression or death in the absence of progression. Importantly, the definition explicitly retains events regardless of withdrawal from randomized therapy or receipt of another anti-cancer therapy before progression.
5. Results: Primary Progression-Free Survival
The posted primary analysis compared progression-free survival between osimertinib and placebo in the full analysis set. The registry reports a log-rank analysis, a hazard ratio as the effect measure, a two-sided 95% confidence interval, and a superiority hypothesis.
Primary PFS hazard ratio
95% CI: 0.10–0.24 · P < 0.001
Analysis population: Full analysis set · Method: Log-rank test
A hazard ratio of 0.16 means that, under the time-to-event comparison represented by the reported analysis, the estimated instantaneous rate of progression or death was substantially lower in the osimertinib group than in the placebo group. Expressed as a simple relative interpretation, 0.16 corresponds to an estimated hazard that is 16% of the comparator hazard, or approximately an 84% lower estimated hazard under the model.
The hazard ratio does not mean that 84% of patients were cured, that 84% of patients avoided progression, or that every patient experienced exactly an 84% reduction in individual risk. A hazard ratio is a relative time-to-event measure, not an absolute probability.
The 95% confidence interval of 0.10–0.24 describes statistical uncertainty around the estimated hazard ratio. It does not describe the range of outcomes for individual patients. Its width also matters: the data support a substantially lower estimated hazard, while the interval quantifies uncertainty about the magnitude of that relative effect.
The P < 0.001 value addresses the evidence against the relevant null hypothesis under the reported statistical test. It is not a measure of effect size, clinical importance, or the probability that the treatment works. The size of the estimated effect is communicated by the HR, while its statistical uncertainty is communicated by the confidence interval.
Because this is a hazard-ratio analysis, interpretation also depends on the assumptions and censoring structure underlying time-to-event analysis. The ClinicalTrials.gov record does not report a proportional-hazards diagnostic, so the HR should not be treated as a literal constant risk reduction at every point in time.
6. Secondary Time-to-Event Results
The registry also reports several secondary time-to-event analyses. These analyses use the full analysis set and compare osimertinib with placebo. Most are reported using the log-rank test; the EGFR mutation-specific PFS analyses use a Cox proportional-hazards model.
| Secondary endpoint | Method | Effect | 95% CI | P-value |
|---|---|---|---|---|
| Number of Participants With PFS Events in Patients With EGFR Ex19del or L858R Mutation | Cox proportional-hazards model | HR 0.17 | 0.10–0.29 | Not reported |
| Number of Participants With PFS Events in Patients With EGFR Ex19del or L858R Mutation | Cox proportional-hazards model | HR 0.32 | 0.19–0.56 | Not reported |
| Number of Participants With PFS Events in Patients With EGFR Mutations Ex19del or L858R Detectable in Plasma-derived ctDNA | Cox proportional-hazards model | HR 0.22 | 0.15–0.34 | Not reported |
| Central Nervous System (CNS) PFS by BICR | Log-rank test | HR 0.17 | 0.09–0.32 | <0.001 |
| Overall Survival (Duration) | Log-rank test | HR 0.81 | 0.42–1.56 | 0.530 |
| Time to Death or Distant Metastases by BICR | Log-rank test | HR 0.21 | 0.11–0.38 | <0.001 |
| Time to Study Treatment Discontinuation | Log-rank test | HR 0.21 | 0.14–0.32 | <0.001 |
| Time to First Subsequent Treatment | Log-rank test | HR 0.13 | 0.08–0.21 | <0.001 |
| Time to Second Subsequent Treatment | Log-rank test | HR 0.51 | 0.28–0.91 | 0.022 |
| Second Progression-free Survival (PFS2) | Log-rank test | HR 0.62 | 0.35–1.08 | 0.088 |
The registry uses the same broad superiority framework for these reported comparisons. The absence of a posted P-value for the three mutation-specific PFS analyses should not be filled in by calculation; the ClinicalTrials.gov record reports the hazard ratio and confidence interval but no P-value.
CNS Progression-Free Survival
CNS PFS hazard ratio
95% CI: 0.09–0.32 · P < 0.001
Method: Log-rank test · Analysis population: Full analysis set
The CNS PFS estimate is another time-to-event comparison, but the endpoint specifically concerns central nervous system progression-free survival by BICR. The HR below 1 indicates a lower estimated event hazard in the osimertinib group under the reported analysis.
Overall Survival
Overall survival hazard ratio
95% CI: 0.42–1.56 · P = 0.530
Time frame: Date of randomisation to date of death by any cause. Assessed up to a maximum of approximately 63 months.
The reported OS estimate is less precise than the primary PFS estimate, as reflected by the 95% confidence interval extending from 0.42 to 1.56. The ClinicalTrials.gov record does not report a median OS or absolute survival probabilities, so those quantities are not used here to characterize the result.
Time to Death or Distant Metastases
Hazard ratio
95% CI: 0.11–0.38 · P < 0.001
Time frame: Time from randomisation to the date of distant metastases or death. Assessed up to date of DCO (05Jan24) to a maximum of approximately 63 months
Time to Study Treatment Discontinuation
Hazard ratio
95% CI: 0.14–0.32 · P < 0.001
Time frame: Time from randomisation to the earlier of the date of study treatment discontinuation or death. Assessed up to date of DCO (05Jan24) to a maximum of approximately 63 months
Time to First Subsequent Treatment
Hazard ratio
95% CI: 0.08–0.21 · P < 0.001
Time frame: Time from randomisation to the start of first subsequent anti-cancer therapy or death. Assessed up to date of DCO (05Jan24) to a maximum of approximately 63 months
Time to Second Subsequent Treatment
Hazard ratio
95% CI: 0.28–0.91 · P = 0.022
Second Progression-Free Survival
PFS2 hazard ratio
95% CI: 0.35–1.08 · P = 0.088
Time frame: Every 8 weeks for first 48 weeks, then every 12 weeks. Assessed up to date of DCO (05Jan24) to a maximum of approximately 63 months
7. Secondary Binary Endpoint Results
LAURA also reports binary response outcomes. These use logistic regression and odds ratios rather than time-to-event methods. This distinction matters: an odds ratio compares odds, not probabilities directly.
| Endpoint | Method | Effect measure | 95% CI | P-value |
|---|---|---|---|---|
| Objective Response Rate by BICR | Logistic regression | OR 2.77 | 1.54–5.08 | <0.001 |
| Disease Control Rate by BICR | Logistic regression | OR 2.06 | 0.94–4.47 | 0.069 |
Objective Response Rate
The reported odds ratio was 2.77, with a two-sided 95% CI of 1.54–5.08 and P < 0.001. The registry notes that an odds ratio > 1 favors osimertinib.
Disease Control Rate
The reported odds ratio was 2.06, with a two-sided 95% CI of 0.94–4.47 and P = 0.069. The registry notes that an odds ratio > 1 favors osimertinib.
8. Statistical Methodology
Log-rank test
The registry reports the log-rank test for the primary PFS analysis and several secondary time-to-event endpoints. The log-rank test compares survival or event-time distributions between groups while accounting for the timing of events and censoring.
The test uses information from observed event times and the numbers of participants at risk. It is therefore fundamentally different from a simple comparison of proportions at a single time point.
Cox proportional-hazards model
The registry reports a Cox proportional-hazards model for the three mutation-specific PFS analyses. The model produces a hazard ratio that summarizes the relative event hazard associated with treatment under the model.
For example, an HR of 0.22 indicates an estimated hazard that is 22% of the comparator hazard under the fitted model. It is not equivalent to a 78-percentage-point reduction in the probability of an event.
Logistic regression
The registry reports logistic regression for objective response rate and disease control rate. These are binary outcomes, so the analysis works with the odds of the specified response or disease-control state.
Intention-to-treat analysis
The registry's analysis text identifies intention-to-treat analysis as an associated concept for the posted efficacy analyses, while the analysis population is described as the full analysis set. The central principle is that randomized treatment assignment remains the basis of comparison rather than redefining groups according to subsequent treatment experience.
Confidence intervals
The reported effect estimates use two-sided 95% confidence intervals. A confidence interval describes statistical uncertainty around the estimated effect under the analysis framework; it is not a range containing 95% of individual patient outcomes.
Statistical significance and effect size
A small P-value and a large treatment effect are different pieces of information. In LAURA, the primary PFS analysis has both a hazard ratio of 0.16 and P < 0.001. The HR communicates the estimated relative effect, the confidence interval communicates precision, and the P-value addresses the evidence against the null hypothesis under the specified test.
9. Statistical Methods Explained
Why is PFS analyzed as a time-to-event endpoint?
PFS records not only whether progression or death occurs, but also when it occurs. Participants who have not experienced the event by their last evaluable assessment can contribute information up to that point through censoring. This makes ordinary binary comparisons inadequate as the sole analysis framework for the endpoint.
What does an HR of 0.16 mean?
An HR of 0.16 means the estimated instantaneous event hazard for progression or death was 16% of the comparator hazard under the reported time-to-event analysis. A useful relative interpretation is therefore an approximately 84% lower estimated hazard. This is not the same as saying that 84% of patients avoided progression.
Why is the confidence interval important?
The 95% CI of 0.10–0.24 places the estimated primary PFS HR in the context of sampling uncertainty. A point estimate alone can give a false impression of precision; the interval shows how much uncertainty surrounds that estimate within the statistical framework.
Why does the P-value not measure effect size?
The P-value is a measure of how incompatible the observed data are with a specified null hypothesis under the test procedure. It does not tell us how large the treatment effect is. The HR of 0.16 is the effect estimate; the 95% CI describes its uncertainty; P < 0.001 describes the hypothesis-test result.
Why use logistic regression for objective response rate?
Objective response is a binary outcome: a participant either meets the response definition or does not. Logistic regression models the probability through the odds scale, allowing an odds ratio to summarize the treatment comparison. An OR of 2.77 therefore means the estimated odds of the specified response were 2.77 times those in the comparator group under the reported model; it is not itself a 2.77-fold probability.
Why are the PFS and response analyses statistically different?
PFS incorporates the timing of progression or death and censoring, so survival-analysis methods are appropriate. Objective response rate is a binary outcome assessed over the specified response-assessment period, so logistic regression is appropriate for the reported comparison. Using different methods for different endpoint structures is a feature of the analysis rather than an inconsistency.
Why should subgroup or biomarker-specific estimates be read cautiously?
The ClinicalTrials.gov record includes mutation-specific and plasma-derived ctDNA-defined analyses. These are narrower analysis populations or endpoint definitions than the overall primary comparison. Their estimates should be interpreted in the context of the corresponding analysis population and confidence interval rather than assuming that every secondary estimate has the same evidentiary role as the primary endpoint.
10. Endpoint Time Frames
| Endpoint | Registry time frame |
|---|---|
| Primary PFS by BICR | Every 8 weeks for first 48 weeks, then every 12 weeks until BICR-confirmed radiological disease progression. Assessed up to date of DCO (05Jan24) to a maximum of approximately 63 months |
| Mutation-specific PFS analyses | Every 8 weeks for first 48 weeks, then every 12 weeks until BICR-confirmed radiological disease progression. Assessed up to date of DCO (05Jan24) to a maximum of approximately 63 months |
| CNS PFS by BICR | Every 8 weeks for first 48 weeks, then every 12 weeks until BICR-confirmed radiological disease progression. Assessed up to date of DCO (05Jan24) to a maximum of approximately 63 months |
| Overall Survival | Date of randomisation to date of death by any cause. Assessed up to a maximum of approximately 63 months |
| Objective Response Rate by BICR | Every 8 weeks for first 48 weeks, then every 12 weeks until radiological disease progression. Assessed up to date of DCO (05Jan24) to a maximum of approximately 63 months |
| Time to Death or Distant Metastases by BICR | Time from randomisation to the date of distant metastases or death. Assessed up to date of DCO (05Jan24) to a maximum of approximately 63 months |
| Time to Study Treatment Discontinuation | Time from randomisation to the earlier of the date of study treatment discontinuation or death. Assessed up to date of DCO (05Jan24) to a maximum of approximately 63 months |
| Time to First Subsequent Treatment | Time from randomisation to the start of first subsequent anti-cancer therapy or death. Assessed up to date of DCO (05Jan24) to a maximum of approximately 63 months |
| Time to Second Subsequent Treatment | Time from randomisation to the start of second subsequent anti-cancer therapy or death. Assessed up to date of DCO (05Jan24) to a maximum of approximately 63 months |
| PFS2 | Every 8 weeks for first 48 weeks, then every 12 weeks. Assessed up to date of DCO (05Jan24) to a maximum of approximately 63 months |
11. Analysis Population and Interpretation
| Feature | Reported information |
|---|---|
| Primary PFS analysis population | Full analysis set |
| Secondary time-to-event analysis population | Full analysis set |
| Binary endpoint analysis population | Full analysis set |
| Groups compared | Osimertinib vs Placebo |
| Hypothesis type | Superiority |
| Confidence interval | 95%, two-sided |
The use of the full analysis set means the posted efficacy estimates are not presented as a comparison restricted only to participants who completed treatment. This distinction is important in randomized trials because restricting analysis according to post-randomization behavior can alter the treatment comparison that randomization was designed to establish.
12. Safety Results
The ClinicalTrials.gov record provides serious adverse-event counts by treatment arm as affected/at risk. These should be read as reported counts rather than converted into additional comparative estimates.
| Arm | Serious adverse events | Reported format |
|---|---|---|
| Osimertinib | 55/143 | Affected / at risk |
| Placebo | 11/73 | Affected / at risk |
The registry caveat states that, as per CSP Section 7.2, patient monitoring during treatment with open-label osimertinib is monitored by the investigator per local clinical guidance. It also states that serious adverse events and AESIs are reported in the eCRF but were not analyzed or tabulated at the time of primary analysis, and that safety assessments are to be performed as per local clinical practice and not collected as part of study data capture in line with the pro...
13. Biomarker-Specific PFS Analyses
Three secondary analyses focus on PFS events in biomarker-defined populations. The registry reports Cox proportional-hazards models and hazard ratios for these analyses.
| Analysis | HR | 95% CI | Analysis note |
|---|---|---|---|
| Patients with EGFR Ex19del or L858R mutation | 0.17 | 0.10–0.29 | PFS in Ex19Del positive patients; a hazard ratio < 1 favours osimertinib |
| Patients with EGFR Ex19del or L858R mutation | 0.32 | 0.19–0.56 | PFS in L858R positive patients; a hazard ratio < 1 favours osimertinib |
| EGFR Ex19del or L858R detectable in plasma-derived ctDNA | 0.22 | 0.15–0.34 | Negative at screening; a hazard ratio < 1 favours osimertinib |
The two mutation-specific entries concern distinct mutation categories identified in the analysis notes, while the third concerns detectability in plasma-derived ctDNA. The registry extract does not provide P-values for these three analyses, so the HR and confidence interval are the appropriate reported statistical quantities here.
14. Understanding the Hazard Ratios Across Endpoints
| Endpoint | HR | 95% CI | P-value |
|---|---|---|---|
| Primary PFS by BICR | 0.16 | 0.10–0.24 | <0.001 |
| CNS PFS by BICR | 0.17 | 0.09–0.32 | <0.001 |
| Overall Survival | 0.81 | 0.42–1.56 | 0.530 |
| Time to Death or Distant Metastases | 0.21 | 0.11–0.38 | <0.001 |
| Time to Study Treatment Discontinuation | 0.21 | 0.14–0.32 | <0.001 |
| Time to First Subsequent Treatment | 0.13 | 0.08–0.21 | <0.001 |
| Time to Second Subsequent Treatment | 0.51 | 0.28–0.91 | 0.022 |
| PFS2 | 0.62 | 0.35–1.08 | 0.088 |
The estimates vary across endpoints because the endpoints measure different clinical events or event combinations. A hazard ratio should therefore be interpreted within its endpoint definition rather than treating all HRs as interchangeable measures of one underlying outcome.
The primary PFS HR of 0.16 and the OS HR of 0.81 are not contradictory simply because they differ. PFS measures progression or death according to its registered definition, while OS measures death from any cause. Different endpoints can have different event timing, censoring patterns, and relationships to subsequent treatment.
15. Odds Ratios for Objective Response and Disease Control
Objective response rate
95% CI: 1.54–5.08 · P < 0.001
Method: Logistic regression · Full analysis set
The odds ratio above 1 favors osimertinib according to the registry's analysis note. An OR of 2.77 means the estimated odds of objective response were 2.77 times those in the placebo group under the reported logistic model.
Disease control rate
95% CI: 0.94–4.47 · P = 0.069
Method: Logistic regression · Full analysis set
The disease-control estimate is also above 1, but its confidence interval includes 1 and its reported P-value is 0.069. The appropriate statistical description is therefore the reported estimate, uncertainty interval, and hypothesis-test result rather than converting the result into a probability of treatment benefit.
16. Why This Trial Matters Statistically
LAURA is a useful teaching case because it combines several major clinical-trial statistical structures within one randomized comparison.
| Concept | How it appears in LAURA |
|---|---|
| Randomization | Participants were randomized to two parallel treatment arms. |
| Blinding | The registry reports double masking. |
| Time-to-event analysis | The primary endpoint is progression-free survival. |
| Log-rank test | Used for the primary PFS comparison and multiple secondary time-to-event endpoints. |
| Hazard ratio | Used to quantify relative treatment effects for time-to-event outcomes. |
| Cox model | Reported for the mutation-specific PFS analyses. |
| Logistic regression | Reported for objective response rate and disease control rate. |
| Odds ratio | Used for the two reported binary efficacy endpoints. |
| Confidence intervals | Two-sided 95% CIs accompany the posted effect estimates. |
| Intention-to-treat concept | Identified in the analysis text for the posted efficacy analyses. |
| Biomarker-defined analyses | Separate PFS analyses are reported for EGFR mutation and plasma-derived ctDNA categories. |
| Safety denominator | Serious adverse events are reported as affected/at risk by arm. |
The trial therefore illustrates an important principle in clinical biostatistics: The statistical method should follow the structure of the endpoint. Time-to-event endpoints require methods that account for event timing and censoring, whereas binary endpoints can be modeled using logistic regression.
17. Limitations
- Registry-level scope: this page is restricted to the ClinicalTrials.gov record and the specified linked records. It does not import additional numerical results from publications.
- No median survival values: the ClinicalTrials.gov record reports hazard ratios for OS and other time-to-event endpoints but do not provide median OS, median PFS, or other median event times.
- No absolute survival estimates: the ClinicalTrials.gov record does not report Kaplan-Meier survival probabilities at specified time points.
- No subgroup forest plot: the ClinicalTrials.gov record includes selected biomarker-specific analyses but does not provide a broader subgroup framework or interaction tests.
- No reported multiplicity strategy: the ClinicalTrials.gov record does not specify how the multiple secondary endpoints were incorporated into an overall type I error-control strategy.
- No reported interim strategy: the ClinicalTrials.gov record does not provide an interim-analysis boundary, alpha-spending method, or information fraction.
- No reported imputation strategy: the ClinicalTrials.gov record does not describe missing-data handling or imputation procedures.
- Safety analysis caveat: the registry specifically states that serious adverse events and AESIs were not analyzed or tabulated at the time of primary analysis.
- Hazard-ratio interpretation: HRs are model-based relative measures and should not automatically be interpreted as constant relative risk reductions over the entire follow-up period.
18. What the Primary PFS Result Does — and Does Not — Mean
The primary PFS HR of 0.16 indicates a substantially lower estimated instantaneous hazard of progression or death for osimertinib relative to placebo under the reported analysis.
The HR is not the proportion of patients who remained progression-free, is not an absolute risk difference, and does not mean that every participant experienced the same proportional change in risk.
The 95% CI of 0.10–0.24 quantifies uncertainty around the primary estimate. The interval is therefore essential to interpreting the precision of the reported HR rather than treating 0.16 as an exact population value.
The P-value of <0.001 summarizes the result of the reported hypothesis test. It should not be interpreted as the probability that the null hypothesis is true or as a measure of how large the treatment effect is.
19. Endpoint Definitions and Statistical Interpretation
Progression-free survival
Combines objective disease progression and death in the absence of progression into a time-to-event endpoint.
Overall survival
Measures time from randomisation to death by any cause.
Objective response rate
Uses a binary response framework and is analyzed in the registry using logistic regression.
Disease control rate
Is another binary endpoint analyzed using logistic regression and an odds ratio.
Keeping these definitions separate prevents a common statistical mistake: treating all reported treatment effects as if they measure the same outcome. An HR for PFS, an HR for OS, and an OR for response are mathematically and clinically different quantities.
20. Clinical Interpretation vs Statistical Interpretation
Statistical interpretation
The primary PFS comparison produced an HR of 0.16 with a two-sided 95% CI of 0.10–0.24 and P < 0.001 in the full analysis set. The registry identifies the hypothesis as superiority.
Clinical interpretation
The reported endpoints describe different dimensions of disease control and treatment experience. PFS, OS, CNS PFS, subsequent-treatment timing, response, and safety should therefore be considered as distinct pieces of the trial evidence rather than collapsed into a single statistic.
21. Registry Caveats and What the Data Do Not Establish
The ClinicalTrials.gov record is sufficient to describe the primary statistical result and a substantial set of secondary analyses, but it does not provide every component that would normally appear in a complete statistical analysis plan.
| Topic | What can be stated from the ClinicalTrials.gov record |
|---|---|
| Primary endpoint | One registered time-to-event endpoint: PFS by BICR. |
| Primary method | Log-rank test. |
| Primary effect measure | Hazard ratio. |
| Primary analysis population | Full analysis set. |
| Hypothesis | Superiority. |
| Confidence interval | Two-sided 95% CI. |
| Multiplicity | Not specified in the ClinicalTrials.gov record. |
| Interim analysis | Not specified in the ClinicalTrials.gov record. |
| Missing-data method | Not specified in the ClinicalTrials.gov record. |
| Stratification | Not specified in the ClinicalTrials.gov record. |
| Bayesian methods | Not reported in the ClinicalTrials.gov record. |
| Non-inferiority margin | Not applicable to the reported superiority hypothesis; no non-inferiority margin is reported. |
22. A Compact Statistical Summary
| Question | Reported answer |
|---|---|
| What was the primary endpoint? | PFS by BICR. |
| What population was analyzed? | Full analysis set. |
| What method was used? | Log-rank test. |
| What was the primary effect estimate? | HR 0.16. |
| How precise was the estimate? | 95% CI 0.10–0.24. |
| What was the reported P-value? | <0.001. |
| What was the hypothesis type? | Superiority. |
| What other major methods were reported? | Cox proportional-hazards model and logistic regression. |
| What binary outcomes were analyzed? | Objective response rate and disease control rate. |
| What safety information is posted in the ClinicalTrials.gov record? | Serious adverse events by arm as affected/at risk: 55/143 for osimertinib and 11/73 for placebo. |
23. Related Tutorials
Learn more about the methods used in this trial:
24. Related Calculators
These tools connect directly to the main statistical structures used in the LAURA analysis:
25. Sources
- ClinicalTrials.gov: LAURA, NCT03521154.
- Linked publication record: PubMed PMID 42296617.
- Linked publication record: PubMed PMID 42037107.
- Linked publication record: PubMed PMID 42019226.
- Linked publication record: PubMed PMID 40311309.
- Linked publication record: PubMed PMID 39289145.
Continue through Clinical Biostats
Explore the statistical concepts behind randomized trials, time-to-event endpoints, regression models, confidence intervals, and clinical-trial analysis.
26. Record Summary
LAURA provides a detailed example of a randomized phase 3 clinical-trial analysis centered on a time-to-event primary endpoint. The registry reports PFS by BICR as the primary endpoint, analyzed in the full analysis set with a log-rank test and a hazard ratio of 0.16 (95% CI 0.10–0.24; P < 0.001). Secondary analyses extend the statistical picture to CNS PFS, overall survival, time to distant metastases or death, treatment discontinuation, subsequent treatment, PFS2, biomarker-defined PFS, objective response rate, and disease control rate.
The statistical lesson is broader than the primary P-value. The primary HR communicates the relative time-to-event effect; the confidence interval communicates its precision; the log-rank test provides the reported hypothesis-test framework; Cox models are used for selected biomarker-defined PFS analyses; and logistic regression provides a different framework for binary response outcomes. Reading these quantities together produces a more complete understanding of the trial than any single number can provide.