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 on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for A Study of Erlotinib (Tarceva) Versus Gemcitabine/Cisplatin as First-line Treatment in Patients With Non-small Cell Lung Cancer With EGFR Mutations (NCT01342965), a randomized phase 3 study. The primary outcome is Investigator-assessed Duration of Progression-free Survival.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT01342965 |
| Phase | Phase 3 |
| Condition | Non-Small Cell Lung Cancer |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None (open label) |
| Enrollment | 217 |
| Arms in the primary outcome | 2 |
| Lead sponsor | Hoffmann-La Roche |
| Results posted | Yes |
2. Clinical Question
Population
Patients with stage IIIB/IV non-small cell lung cancer whose tumours had mutations in the tyrosine kinase domain of EGFR. The primary analysis population was the full analysis set, defined as all randomized participants.
Intervention
Erlotinib, also called Tarceva, at 150 mg orally once daily. Treatment continued until disease progression or unacceptable toxicity.
Comparator
Gemcitabine and cisplatin given in three-week cycles. Gemcitabine was administered intravenously on Days 1 and 8, and cisplatin intravenously on Day 1. Treatment continued until progression, unacceptable toxicity, or a total of 4 cycles, whichever came first.
Primary question
How did investigator-assessed progression-free survival compare between the randomized treatment groups? Progression-free survival was measured from randomization until disease progression or death from any cause, whichever occurred first. The registry reports median durations by group but does not report an estimable comparative effect for this time-to-event outcome.
The study addressed a first-line treatment comparison in a defined cancer population, rather than a comparison across all patients with non-small cell lung cancer. Eligibility depended on both disease stage and the specified EGFR tumour mutation. The registry record describes the study as multicenter, randomized, open-label, parallel-arm and phase III, with 217 participants enrolled.
The clinical question has several components. The primary endpoint concerns the time until progression or death, while the secondary outcomes include investigator-assessed response, disease control, duration of response, overall survival, adverse events and quality of life measured with the FACIT-L questionnaire. These outcomes capture different aspects of the trial: tumour response is not the same measure as time to progression, and safety outcomes do not directly quantify tumour control.
The registry describes the comparison and its posted results, but it does not report every detail needed to reconstruct the planned analysis. For example, the statistical plan, sample-size justification and handling of missing data are not reported in the registry record. The available results therefore support a description of the recorded outcomes, but not a complete reconstruction of the trial's design rationale or analysis process.
3. Trial Design
This was a phase III, randomized, open-label, parallel-arm study sponsored by Hoffmann-La Roche. It enrolled 217 participants with stage IIIB/IV non-small cell lung cancer and mutations in the EGFR tyrosine kinase domain. The registry lists the study as completed. Its recorded start was 2011-03, and primary completion was recorded as 2012-07.
Participants were assigned to erlotinib or gemcitabine/cisplatin. The erlotinib group received 150 mg orally once daily until disease progression or unacceptable toxicity. The chemotherapy group received gemcitabine 1250 mg/m^2 intravenously on Days 1 and 8, together with cisplatin 75 mg/m^2 intravenously on Day 1 of each three-week cycle. Chemotherapy stopped at progression, unacceptable toxicity or a total of 4 cycles, whichever came first.
The open-label design means treatment assignment was not masked. The registry identifies the primary outcome as investigator-assessed duration of progression-free survival. Because the endpoint assessment is attributed to investigators and the study was unmasked, interpretation should keep the assessment context in view. The registry does not report additional information about procedures used to reduce possible assessment bias.
The primary endpoint's stated timeframe runs from baseline to the data cut-off date of 20 Jul 2012 (1 year, 4 months). The secondary response and disease-control outcomes have a later stated cut-off date of 19 Nov 2012 (1 year, 8 months). These different cut-off dates matter when describing results: the posted outcome summaries do not all refer to the same data cut-off.
The full analysis set for the primary outcome is defined as all randomized participants. Group sizes for that outcome were 110 for erlotinib and 107 for chemotherapy. For the safety summaries, the recorded at-risk counts are 110 and 104, respectively. The registry does not explain the difference between the chemotherapy safety count and its primary-outcome group size. Further details about randomization procedures, allocation concealment, stratification and protocol deviations are not reported in the registry record.
4. Participant Flow and Accounting
The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.
Overall Study
| Stage | Erlotinib | Chemotherapy |
|---|---|---|
| STARTED | 110 | 107 |
| Received Treatment | 110 | 104 |
| COMPLETED | 1 | 0 |
| NOT COMPLETED | 109 | 107 |
Reasons for not completing this period:
| Reason | Erlotinib | Chemotherapy |
|---|---|---|
| Death | 58 | 57 |
| Lost to Follow-up | 5 | 3 |
| Other Reasons- Unspecified | 44 | 38 |
| Withdrawal by Subject | 2 | 9 |
Analysis population for the primary outcome, as defined in the registry: Full analysis set: All randomized participants.
5. Baseline Characteristics
The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.
Population: Full analysis set: All randomized participants.
| Characteristic | Unit | Erlotinib | Chemotherapy | Total |
|---|---|---|---|---|
| Participants analysed | 110 | 107 | 217 | |
| Age, Continuous | years | 56.7 (SD 10.37) | 55.8 (SD 10.41) | 56.3 (SD 10.37) |
| Sex: Female, Male: Female | Participants | 68 | 65 | 133 |
| Sex: Female, Male: Male | Participants | 42 | 42 | 84 |
6. Primary Endpoint
The registry lists the primary outcome as Investigator-assessed Duration of Progression-free Survival, measured over Baseline to the data cut-off date of 20 Jul 2012 (1 year, 4 months).
| Element | Registry entry |
|---|---|
| Result type | Median |
| Measure of spread | 95% Confidence Interval |
| Unit | Months |
| Endpoint classification | Time-to-Event Analysis |
Progression-free survival measures the time from randomization until the first of disease progression or death from any cause. It combines two clinically important events into one time-to-event outcome: a participant is counted as having reached the endpoint when either event occurs. The registry defines progression through changes in target lesions, unequivocal worsening of non-target lesions, or the appearance of new malignant lesions.
For target lesions, the definition requires an increase of at least 20% in the sum of diameters from the smallest sum recorded on study, with an absolute increase of at least 5 mm. The registry also describes how progression is judged when there are non-target lesions, including cases where measurable disease is absent. These rules specify what event marks the end of progression-free survival; they do not by themselves describe how investigators handled uncertain assessments or missed visits.
The registry reports the median progression-free survival in each group. A median is the time by which half of the analysis population has experienced the event, under the time-to-event analysis. It gives a compact description of the event-time distribution, but it does not show every participant's experience or the shape of the full survival curve. It also is not itself a measure of the relative treatment effect.
Progression-free survival is used in trials because it captures time until progression or death, rather than only whether a response occurred at a particular assessment. General statistical context: time-to-event methods can account for participants whose event has not been observed by the end of follow-up, provided their follow-up and censoring are handled appropriately. The registry record does not report the censoring details or the statistical method used for the primary analysis.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The posted primary outcome gives a median investigator-assessed progression-free survival of 11.0 months for erlotinib, with a reported 95% confidence interval from 8.3 months and no reported upper limit. For chemotherapy, the median is 5.5 months, with a 95% confidence interval from 4.2 to 7.1 months. These are group-level median estimates, not a direct estimate of the relative effect between treatments.
The registry does not report an estimable two-arm comparison for the primary median time-to-event outcome. In particular, no hazard ratio or other comparative primary-effect estimate is reported. The two medians alone cannot establish the size or precision of a treatment contrast across the event-time distributions. The confidence interval for the erlotinib median is also incomplete in the record, limiting interpretation of the uncertainty around that estimate.
The secondary outcome results show a similar direction in the two posted binary summaries: the recorded response percentage and disease-control percentage are higher in the erlotinib group than in the chemotherapy group. The unadjusted comparison for response has p < 0.001, and the unadjusted comparison for disease control has p = 0.035. These secondary comparisons should not be treated as substitutes for the primary time-to-event comparison. They address different outcomes, and the registry does not report adjusted analyses for them.
A p-value describes how compatible the observed data are with a specified null hypothesis under the assumptions of a particular test. It does not measure the probability that the null hypothesis is true, the size of a clinical effect, or the chance that the result will replicate. The unadjusted secondary p-values also do not account for the fact that the registry lists multiple secondary outcomes. The registry record does not report a multiplicity adjustment or a prespecified hierarchy for interpreting these outcomes.
Overall, the record contains descriptive primary medians and posted secondary percentages, but lacks an estimable comparative primary time-to-event result. The data can therefore be described by outcome and group, while the relative primary effect and its precision remain unquantified in the registry record. This distinction is important: numerical differences in group summaries do not, on their own, supply a complete inferential analysis of the primary endpoint.
9. Statistical Methods Explained
The primary outcome is a time-to-event endpoint. Each participant's event time begins at randomization and ends at disease progression or death, whichever comes first. The registry reports group medians and confidence intervals, but does not report a hazard ratio, a comparative test, or the analysis method used to estimate the medians. The record therefore supports interpretation of the posted group summaries, not reconstruction of the primary inferential analysis.
What does a median time-to-event estimate describe?
A median is the point in the event-time distribution by which half of the participants have experienced the event. It does not describe the event times of every participant, and it does not tell whether the treatment contrast is constant over follow-up. General statistical context: a survival curve displays the estimated probability of remaining event-free over time, while a median reduces that curve to one time point. Two medians can be useful summaries, but a comparison of medians is not interchangeable with a hazard ratio or a test of the entire curves.
A confidence interval describes uncertainty in an estimated quantity under the statistical model and method used. General statistical context: a wider interval indicates less precision than a narrower interval, and the interval is not a range that contains every participant's event time. The record gives a 95% confidence interval for each median only partially: the erlotinib interval has a lower limit but no upper limit, while both limits are recorded for chemotherapy. The reason the erlotinib upper limit is absent is not reported in the registry record.
Why is there no primary between-group estimate?
The posted primary outcome is a median-time endpoint, and the record explicitly indicates that values are missing or not estimable for a two-arm comparison. It therefore does not provide a comparative primary estimate that can be interpreted as a hazard ratio, risk difference or other measure. A comparison of the medians by subtraction would not replace the missing time-to-event analysis and is not performed. The number and timing of progression or death events, censoring information, and the method for the primary comparison are not reported in the registry record.
How should the secondary percentages be read?
The registry posts group percentages for investigator-assessed response and disease control, along with unadjusted comparisons. A risk difference compares event proportions between groups on the absolute scale. General statistical context: its confidence interval communicates uncertainty in that difference, while a p-value assesses compatibility with a null hypothesis under the test assumptions. These measures answer different questions: the estimate describes a contrast, its interval describes precision, and the p-value does not express effect size.
The unadjusted response comparison has p < 0.001, and the unadjusted disease-control comparison has p = 0.035. Those values are attached to secondary outcomes, not to the primary progression-free survival endpoint. The registry lists several secondary outcomes but reports no analyses for the other listed outcomes in the record. It also does not report whether the secondary comparisons were prespecified as confirmatory or how multiplicity was handled.
General statistical context: when several hypotheses are tested, the chance of at least one apparently unusual result can increase if the tests are interpreted independently without adjustment. This does not mean a posted result is invalid; it means that its evidential weight depends partly on the planned testing strategy. Because the strategy is not reported, the unadjusted secondary p-values should be read as individual, unadjusted comparisons rather than as a family-wise confirmatory conclusion.
What do the planning illustrations mean?
The sample-size illustrations use the observed effect as a planning reference and show hypothetical participants per arm for specified power levels and fractions of that effect. They are not the trial's observed power and are not evidence that the trial was designed using those assumptions. General statistical context: planning calculations estimate the sample size needed under stated assumptions about the target effect, significance level, allocation and analysis. If the target effect is smaller, a larger sample is generally needed to distinguish it from the null with the same planned power.
The record states the illustrations assume a two-sided significance level of 0.05, equal allocation, a normal approximation, and an effect set to a stated fraction of the observed effect. It separately notes that time-to-event planning uses a total event count. These are illustrations computed from observed results, not the trial's own design calculation. The actual sample-size justification, target effect, event target and power assumptions for the trial are not reported in the registry record.
What do the analysis-population labels establish?
The primary endpoint's analysis population is the full analysis set, defined as all randomized participants. This identifies who is included according to the registry description, but it does not specify how missing endpoint assessments, treatment discontinuation, crossover or protocol deviations were treated. General statistical context: analysis-population definitions matter because changing who contributes to an analysis can affect the estimand and the interpretation of comparisons. Details beyond the full-analysis-set definition are not reported in the registry record.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists multiple secondary outcomes alongside the primary progression-free survival endpoint. These include response, disease control, duration of response, overall survival, adverse events and quality of life. It posts unadjusted comparisons for response and disease control, but does not report analyses for the remaining listed secondary outcomes in the record.
When several outcomes are tested, multiplicity becomes relevant because repeated testing can increase the chance of at least one small p-value under a set of null hypotheses. General statistical context: investigators may address this through a prespecified hierarchy, adjustment procedure or other strategy, but the choice depends on the trial's aims and analysis plan. The multiplicity strategy, including whether the secondary outcomes were treated as confirmatory or exploratory, is not reported in the registry record.
The primary outcome is described as using the full analysis set of all randomized participants. That definition gives an analysis-population label, but it does not explain how missing assessments, withdrawal, treatment discontinuation, or censoring were handled. The statistical methods for these issues are not reported in the registry record. In particular, the posted medians do not disclose how participants without an observed progression or death by the data cut-off contributed to the analysis.
The primary endpoint is investigator-assessed in an open-label study. The registry does not report details of independent review, adjudication, or safeguards for assessment. This makes the assessment context relevant when interpreting a progression endpoint, without establishing that any particular assessment bias occurred.
The record also does not report interim analyses, stopping rules, a statistical analysis plan, or the trial's sample-size justification. These omissions do not establish that such procedures were absent; they mean their details are not reported in the registry record. Consequently, the posted results cannot be evaluated against an available description of the complete planned analysis strategy.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Erlotinib | 22 | 110 | 20.0% |
| Chemotherapy | 16 | 104 | 15.4% |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 5% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Erlotinib | 99 of 110 |
| Chemotherapy | 96 of 104 |
Most frequent serious adverse event terms
| Event term | Organ system | Erlotinib | Chemotherapy |
|---|---|---|---|
| Anaemia | Blood and lymphatic system disorders | 1 of 110 | 4 of 104 |
| Dyspnoea | Respiratory, thoracic and mediastinal disorders | 3 of 110 | 0 of 104 |
| Thrombocytopenia | Blood and lymphatic system disorders | 0 of 110 | 3 of 104 |
| Gastroenteritis | Infections and infestations | 2 of 110 | 1 of 104 |
| Pneumonia | Infections and infestations | 3 of 110 | 0 of 104 |
| Diarrhoea | Gastrointestinal disorders | 3 of 110 | 0 of 104 |
| Pleural effusion | Respiratory, thoracic and mediastinal disorders | 2 of 110 | 0 of 104 |
| Pneumonitis | Respiratory, thoracic and mediastinal disorders | 1 of 110 | 1 of 104 |
| Respiratory failure | Respiratory, thoracic and mediastinal disorders | 0 of 110 | 2 of 104 |
| Upper gastrointestinal haemorrhage | Gastrointestinal disorders | 2 of 110 | 0 of 104 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Erlotinib | Chemotherapy |
|---|---|---|---|
| Rash | Skin and subcutaneous tissue disorders | 78 of 110 | 11 of 104 |
| Nausea | Gastrointestinal disorders | 5 of 110 | 60 of 104 |
| Vomiting | Gastrointestinal disorders | 7 of 110 | 56 of 104 |
| Diarrhoea | Gastrointestinal disorders | 49 of 110 | 9 of 104 |
| Leukopenia | Blood and lymphatic system disorders | 7 of 110 | 51 of 104 |
| Neutropenia | Blood and lymphatic system disorders | 5 of 110 | 53 of 104 |
| Anaemia | Blood and lymphatic system disorders | 8 of 110 | 44 of 104 |
| Decreased appetite | Metabolism and nutrition disorders | 14 of 110 | 30 of 104 |
| Cough | Respiratory, thoracic and mediastinal disorders | 19 of 110 | 9 of 104 |
| Fatigue | General disorders | 6 of 110 | 20 of 104 |
Serious events with the highest arm-level rate: anaemia (3.8%), thrombocytopenia (2.9%), dyspnoea (2.7%), pneumonia (2.7%), diarrhoea (2.7%).
The safety record gives serious adverse events for 22 of 110 participants at risk in the erlotinib group and 16 of 104 at risk in the chemotherapy group. It also records other adverse events for 99 of 110 and 96 of 104 participants, respectively. These are participant counts in the registry's safety summaries; they do not describe event severity beyond the serious-versus-other grouping or the timing and duration of events.
The listed serious adverse-event terms include anaemia at 3.8%, thrombocytopenia at 2.9%, dyspnoea at 2.7%, pneumonia at 2.7%, diarrhoea at 2.7% and respiratory failure at 1.9%. The registry's list does not assign these percentages to a treatment group in the record, so they should not be interpreted as arm-specific comparisons. The record does not report death counts in the safety summary.
General statistical context: safety counts depend on the defined observation period, event coding, and the number of participants contributing safety data. The registry lists different at-risk counts for the chemotherapy safety group and the primary-outcome group, but does not explain that difference. These summaries can describe what was recorded; they do not by themselves establish comparative safety or explain the causes of individual events.
12. What the Result Does, and Does Not, Mean
The primary result is a pair of group-specific median progression-free survival estimates: 11.0 months for erlotinib and 5.5 months for chemotherapy. The registry also gives confidence intervals, although the upper limit for the erlotinib median is not reported. These values describe the recorded median event times in the full analysis set; they do not provide a complete estimate of the relative treatment effect.
The primary two-arm comparison is not estimable in the registry record. Without a reported comparative estimate and its uncertainty, the primary results cannot quantify the relative difference in progression-free survival or indicate how precisely such a difference was estimated. The medians should not be treated as though they were a hazard ratio or as a formal test of treatment groups.
The posted secondary percentages show group-level response and disease-control results, with unadjusted p-values. They add information about those specific outcomes, but do not answer the primary time-to-event question. Other listed outcomes, including overall survival, duration of response and quality of life, have no numerical analyses reported in the registry record. The safety summaries likewise provide counts and listed event terms, not a complete comparative safety analysis.
General statistical context: statistical results are not the same as clinical recommendations. A group-level estimate does not predict an individual participant's outcome, and a p-value does not establish that a difference is important or that it is caused solely by treatment assignment. Interpretation also depends on the endpoint definition, assessment process, follow-up, missing-data handling and analysis plan. Several of those details are not reported in the registry record.
The record therefore supports a limited, precise statement: it reports the primary medians and selected secondary and safety summaries, but does not report an estimable comparative primary effect. The posted statistics should be interpreted within those limits rather than extended to unreported outcomes or to populations beyond the specified patients with stage IIIB/IV disease and EGFR mutations.
13. Why This Trial Matters Statistically
This study is statistically informative because it illustrates the distinction between a descriptive time-to-event summary and a comparative treatment-effect estimate. The registry reports a median progression-free survival for each randomized group, but no estimable two-arm comparison for that endpoint. Having group medians is not equivalent to having a quantified contrast with an uncertainty interval or a reported time-to-event test.
The endpoint also depends on event assessment. Progression-free survival combines progression and death and uses investigator assessment of progression. The study was open-label, so the assessment context matters when considering how endpoint ascertainment might influence results. The registry record does not report further assessment safeguards, which limits evaluation of that aspect of the design.
There is also a useful distinction between the primary and secondary evidence. The response and disease-control outcomes have posted percentages and unadjusted comparisons, but they do not replace the primary endpoint. Their interpretation also depends on the unreported multiplicity strategy. This makes clear why a small unadjusted p-value for one secondary outcome should not automatically be treated as a definitive answer to the primary question or as a summary of every outcome.
Finally, the record shows why analysis details matter alongside numerical results. The primary analysis population is identified, but the handling of missing assessments, censoring and other analysis choices is not reported. The trial's sample-size justification and statistical plan are also not reported. These absences constrain what can be concluded from registry results, even when the posted group summaries are straightforward to read.
14. Limitations
- The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
- Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.
The main limitation for interpreting the primary outcome is that the registry does not report an estimable two-arm comparison. The posted medians can be described separately, but no comparative primary effect or its precision is available. The confidence interval for the erlotinib median is incomplete because its upper limit is not reported in the registry record.
The primary endpoint is investigator-assessed, and the study is open-label. The record does not report whether progression assessments were independently reviewed or adjudicated, nor does it provide further information about assessment procedures. This is a limitation of the available documentation, not proof that assessments were biased.
The registry lists a broad set of secondary outcomes, but numerical analyses are posted for only some of them in the record. The response and disease-control comparisons are unadjusted, and a multiplicity strategy is not reported. Results for overall survival, duration of response and quality of life are not reported in the registry record.
Important analysis details are also unavailable: missing-data handling, censoring rules, interim analyses, the statistical analysis plan and the sample-size justification are not reported in the registry record. The chemotherapy safety group has 104 participants at risk, while its primary-outcome group has 107; the record does not explain this difference. Safety summaries also lack death counts and do not allocate the listed serious adverse-event percentages to treatment groups.
Finally, the primary and secondary outcomes have different stated data cut-off dates. This should be kept in mind when considering the set of results together. The record does not provide enough detail to assess how follow-up completeness or the timing of assessments affected each endpoint.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: A Study of Erlotinib (Tarceva) Versus Gemcitabine/Cisplatin as First-line Treatment in Patients With Non-small Cell Lung Cancer With EGFR Mutations — NCT01342965.
- Registry record last updated 2015-02-24; status last verified 2015-02.
- Results first posted 2015-02-24.
- Data retrieved from ClinicalTrials.gov on 2026-11-09.
- Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
NCT01342965 is a completed, randomized, open-label, parallel-arm phase III study sponsored by Hoffmann-La Roche. It compared first-line erlotinib with gemcitabine/cisplatin in patients with stage IIIB/IV non-small cell lung cancer with EGFR mutations. The registry records 217 participants enrolled.
The primary outcome was investigator-assessed progression-free survival, defined from randomization to progression or death, whichever occurred first. The reported median was 11.0 months for erlotinib and 5.5 months for chemotherapy. A two-arm primary comparison is not estimable in the registry record, and the upper confidence limit for the erlotinib median is not reported.
The registry also reports response and disease-control percentages with unadjusted comparisons, plus safety counts and selected serious adverse-event terms. Numerical results for several other listed secondary outcomes are not reported in the registry record. Details of the statistical plan, multiplicity handling, missing-data methods and sample-size justification are not reported in the registry record.