Study With Gefitinib in Combination With Olaparib (AZD2281) Versus Gefitinib Alone: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study With Gefitinib in Combination With Olaparib (AZD2281) Versus Gefitinib Alone (NCT01513174), a randomized phase 1, phase 2 study. The primary outcome is Progression-free Survival (PFS).

2. Clinical Question

The clinical question is about the relative timing of progression or death-related follow-up events as captured by the primary endpoint, not simply whether a participant had a response at a particular assessment. The registry defines progression by RECIST v1.0: a 20% increase in the sum of the longest diameter of target lesions, a measurable increase in a non-target lesion, or the appearance of new lesions. The primary endpoint timeframe runs from randomization until the end of follow-up, up to 30 months.

The comparison is randomized and parallel, but masking is listed as none. This means the record describes an open-label comparison; it does not establish how knowledge of treatment assignment may have affected assessment or care. The record’s primary outcome is progression-free survival. Overall survival and best global response during treatment are also identified as secondary outcomes, although complete results for all listed outcomes are not posted in the available record.

3. Trial Design

The registry describes a randomized, multicenter Phase Ib/IIb study comparing gefitinib plus olaparib with gefitinib alone in patients with EGFR mutation-positive advanced non-small-cell lung cancer. Allocation is listed as randomized, the model as parallel, and masking as none. The record reports enrollment of 186. It lists the study as completed, with a start date of 2011-08 and primary completion in 2016-07.

In the control group, gefitinib was administered continuously once daily at 250 mg/day in 28-day cycles. In the experimental group, gefitinib was administered at 250 mg once daily with olaparib twice daily, continuously in 28-day cycles. The olaparib amount is described as the recommended dose in the previous Phase I study, but the dose amount is not reported in the registry record. The registry identifies the Spanish Lung Cancer Group as sponsor.

The primary endpoint is progression-free survival, summarized as a median with a full range and compared using a log-rank analysis. The primary analysis population is described as per protocol. The arm counts shown for that outcome are 90 for gefitinib and 84 for the combination. Other posted analyses use different arm counts, including 91 in each group for overall survival and the safety figures. The reasons for these differences are not reported in the registry record, so the counts should be understood in the context of each analysis rather than treated as interchangeable.

The record does not report the original sample-size justification, the detailed statistical analysis plan, or whether interim analyses were conducted. The enrollment figure alone does not establish the planned power for the primary comparison. Those details are necessary to distinguish the trial’s original design choices from the separate planning illustrations based on observed effects.

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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The PFS analysis is performed on the PP population

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: The mITT population is defined as the set of patients included in the study who received at least one dose of the study drug. There are 182 patients in the mITT population.

6. Primary Endpoint

The registry lists the primary outcome as Progression-free Survival (PFS), measured over From the date of randomization until end of follow up, up to 30 months..

Progression-free survival measures the time from randomization to the first documented disease progression. In this record, progression is defined using RECIST v1.0: a 20% increase in the sum of the longest diameter of target lesions, a measurable increase in a non-target lesion, or the appearance of new lesions. Follow-up for the endpoint is described as running from randomization to the end of follow-up, up to 30 months. The primary analysis is stated to use the per-protocol population.

Time-to-event endpoints are useful when the timing of an event matters and participants may have different lengths of observation. General explanation: a median time-to-event result is the point by which half of the analyzed population has experienced the event, under the analysis method; it is not an average and does not describe every participant’s course. Participants without a recorded event by the end of their observation can contribute follow-up information through censoring, although the censoring rules used in this trial are not reported in the registry record.

The endpoint is paired with a hazard ratio from a log-rank comparison. General explanation: a hazard ratio summarizes the relative event rate over follow-up under the model or analysis approach, whereas the median describes a particular point on each group’s event-time distribution. These measures answer related but not identical questions. The posted record supplies both median estimates and a sponsor-reported hazard ratio, but does not provide enough detail to reconstruct the underlying event-time data or verify every analytic choice.

7. Results Posted in the Registry

• Control Gefitinib vs Experimental: Gefitinib in Combination With Olaparib: hazard ratio 1.38 (95% CI 1.00–1.92), reported p = 0.124 (as reported in the registry).

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Overall Survival

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor (Log Rank); it was not recalculated.
• Result:
  • Control Gefitinib vs Experimental: Gefitinib in Combination With Olaparib: hazard ratio 1.22 (95% CI 0.81–1.84), reported p = 0.345 (as reported in the registry).
• Interpretation:
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Control Gefitinib vs Experimental: Gefitinib in Combination With Olaparib and the control. The interval gives the range of effects that remain compatible with the data.

The registry identifies overall survival and best global response during treatment as secondary outcomes, and it posts analyses for two secondary outcomes. The overall-survival results show a broad confidence interval around the hazard-ratio estimate, with the interval including 1. The posted best-global-response results are not reported in the registry record. Thus the available secondary evidence is incomplete and does not provide a full set of results for every named secondary outcome.

In broad terms, the secondary overall-survival comparison does not provide a clear, precise separation between the groups, and it does not resolve the uncertainty in the primary progression-free survival comparison. The endpoints address different questions: progression-free survival concerns documented progression, whereas overall survival concerns time to death. A pattern across both can add context, but the endpoints are not interchangeable and one cannot simply substitute for the other.

Unadjusted secondary results generally carry less confirmatory weight when a study evaluates multiple outcomes. General explanation: repeated testing across several outcomes can increase the chance that at least one result appears noteworthy by chance, unless the analysis plan specifies a multiplicity strategy. The registry record does not report whether multiplicity adjustment was used, how outcomes were prioritized, or whether secondary analyses were intended to be confirmatory or exploratory. Without those details, secondary estimates should be treated as supportive or descriptive rather than as independent proof of a treatment effect.

The registry lists three outcome measures but posts only two secondary analyses. The remaining response outcome’s numerical results and inferential details are not reported in the registry record. This limits the ability to judge whether the secondary findings form a coherent pattern across all outcomes that were planned. The secondary results therefore add context but do not overturn or settle the primary comparison.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Overall Survival
• Best Global Response During Treatment Period

9. Statistical Interpretation

10. Statistical Methods Explained

The primary comparison is reported using a log-rank method and a hazard ratio. The endpoint is time from randomization to documented progression, and the primary analysis is described as per protocol. The results are presented as group medians with full ranges, together with a hazard ratio, a 95% confidence interval, and a p-value.

What does the median represent?

General explanation: a median time-to-event estimate marks the time at which the estimated event-time distribution reaches its midpoint. It describes the central point of the distribution, rather than the experience of every participant or the average follow-up time. Two groups can have the same median while their event patterns differ at other points during follow-up. For that reason, the medians are best read alongside the hazard ratio and its uncertainty, not as a complete summary of the time-to-event comparison.

The registry reports the primary medians as 10.9 months in both groups. The listed full ranges are 9.3 to 13.3 months for gefitinib and 8.6 to 13.4 months for the combination. The record labels the dispersion as full range. It does not describe whether these ranges are confidence limits for the median or another kind of range, so they should not be treated as confidence intervals without further documentation.

What does a hazard ratio describe?

General explanation: a hazard ratio compares the event rate between groups over observed follow-up, according to the analysis definition. A value of 1 corresponds to equal hazards under that comparison; values on either side indicate differing relative hazards, with direction determined by which group is in the numerator or reference. The registry labels its reported comparison as control versus experimental and gives a hazard ratio of 1.38. Keeping that comparison label attached to the estimate is important: a hazard ratio is not an absolute difference in months, and it does not state the probability that an individual participant will progress.

The hazard ratio can reflect event-time information across follow-up, while medians focus on the distribution midpoint. They may therefore provide different-looking summaries without being contradictory. General explanation: a single hazard ratio is easiest to interpret when the relative hazards are reasonably stable over time. The registry record does not report checks of that assumption or time-varying estimates, so no conclusion about the pattern of relative hazards over time can be drawn from the record alone.

What does the log-rank test contribute?

General explanation: the log-rank test compares time-to-event experience across groups while accounting for the timing of observed events and censored observations. Its p-value assesses how compatible the observed separation is with the null comparison under the test’s assumptions. It does not give the probability that a treatment has no effect, and it does not measure the clinical importance of a difference. The posted p-value for the primary comparison is 0.124. The registry does not report whether this test was the only prespecified inferential analysis or how the testing plan handled multiple outcomes.

Why do confidence intervals matter?

General explanation: a confidence interval communicates uncertainty around an estimate under the statistical procedure used. It shows a range of values compatible with the data and method at the stated confidence level; it is not a range containing the true value with a probability assigned after the analysis. The primary 95% interval extends from 1.00 to 1.92. Its width makes clear that the estimate is not precise enough to rule out a range of possible relative effects. The p-value and interval are related summaries of the same comparison, not independent confirmations.

What is the role of the analysis population?

The primary outcome description says that the analysis uses the per-protocol population, and the arm counts for that analysis are 90 and 84. General explanation: a per-protocol analysis typically focuses on participants who meet defined protocol requirements, but the exact criteria vary and must be specified. The registry record does not give the criteria used to define this population or explain how participants were handled if they did not meet them. A per-protocol result should therefore not automatically be interpreted as an analysis of every randomized participant.

Random allocation is listed, which is intended to create comparable groups at assignment. However, a registry description does not provide enough detail to assess the randomization sequence, concealment, or implementation. Masking is listed as none. General explanation: in an unmasked study, participants and study personnel know the assigned group, which can matter for outcomes affected by assessment or decisions. The registry defines progression using RECIST, but does not report whether imaging review was independent or blinded. That detail is not reported in the registry record.

How should safety counts be understood?

The safety record gives numbers at risk and numbers affected separately for serious adverse events and other adverse events, as well as deaths. General explanation: these counts describe reported events or participants affected according to the registry’s categories, but they do not by themselves establish that an event was caused by study treatment. Interpretation also depends on definitions, observation periods, exposure, and rules for counting repeated events. Those details are not reported in the registry record, so counts should be read as descriptive safety information rather than as causal comparisons.

The record’s planning illustrations use observed hazard-ratio effects and report total events for time-to-event outcomes. These are separate from the sponsor’s original trial-design calculations, which are not reported in the registry record. General explanation: planning calculations depend on assumptions about the effect to detect, allocation, significance level, power, and analysis method. Changing the effect assumption changes the amount of information needed. The illustrations cannot establish what the trial was originally designed to detect or whether the observed analysis met a prespecified target.

11. Confidence Intervals and Precision

The primary hazard-ratio estimate is 1.38, with a 95% confidence interval from 1.00 to 1.92. This interval is broad relative to the point estimate and begins at 1.00, the value corresponding to equal hazards under the stated comparison. The data therefore leave substantial uncertainty about the relative event experience. The reported p-value is 0.124. Neither the estimate nor the p-value should be interpreted without the comparison label, which is control versus experimental.

The medians are both reported as 10.9 months, but equality at the displayed precision does not establish that the full event-time distributions are identical. The full ranges overlap, and the registry does not label them as confidence intervals for the medians. They should be read as the reported range information, not as a formal measure of precision around each median.

The secondary overall-survival estimate has a 95% confidence interval from 0.806 to 1.845 around the reported hazard ratio of 1.220, with p = 0.345. This interval includes 1 and covers a range of possible relative hazards. The median estimates also have broad reported ranges. As with the primary result, these statistics are compatible with considerable uncertainty and do not establish equivalence.

General explanation: precision is about how narrowly an analysis estimates an effect, whereas statistical significance is a decision made using a testing threshold. The registry record does not report the original threshold or a prespecified margin for equivalence or non-inferiority. A confidence interval can be useful even when a result does not cross a particular threshold because it shows which effect sizes remain compatible with the analysis. Here, the intervals discourage a confident claim of either a precise benefit or a precise absence of effect.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning illustrations for progression-free survival and overall survival report the total number of events needed, not participants per arm. This distinction matters for time-to-event outcomes: the calculation is based on the number of outcome events, while the number of participants required to observe those events also depends on follow-up and event occurrence. The registry record does not report the trial’s original sample-size justification, so these illustrations should not be read as the design target used when the study was planned.

The listed planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect. The event calculations use the Schoenfeld approach. They are planning illustrations computed from the observed effect, not a statement of the sponsor’s original assumptions.

The 80% power column describes the event count that the planning calculation associates with an 80% chance of detecting the assumed effect under those assumptions, if that effect and the calculation assumptions hold. The 90% power column describes the corresponding event count for a 90% chance. These are prospective planning concepts, not probabilities that the completed trial found an effect. The figures are not observed power.

Rows that plan for 75% or 50% of the observed effect use smaller target effects than the row based on 100% of the observed effect. Smaller effects generally require more information to distinguish from no difference, so the event requirements increase as the target effect shrinks. This pattern is a feature of planning calculations, not evidence that a smaller effect occurred. The original design justification and the actual number of events used in each analysis are not reported in the registry record.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

d is the total number of events needed (Schoenfeld's formula) with 1:1 allocation and proportional hazards.

Worked calculation for this trial

1. Hazard ratio HR = 1.380, so ln(HR) = 0.3221.
2. d = 4 × (1.96 + 0.842)² / [ln(1.380)]² = 302.6, rounded up to 303 events in total for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
hr <- 1.38                              # observed hazard ratio
events_needed <- function(power, frac = 1) ceiling(round(4 * (z_a + qnorm(power))^2 / log(hr^frac)^2, 6))   # Schoenfeld, 1:1 allocation
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: events = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(events_needed(0.80, f), big.mark = ","), format(events_needed(0.90, f), big.mark = ",")))

# Simulation: given d total events under proportional hazards, events in the treatment arm are Binomial(d, hr / (1 + hr)).
# The score (log-rank type) statistic is (D1 - d / 2) / sqrt(d / 4).
set.seed(1513174)
d    <- events_needed(0.80)
reps <- 20000
D1 <- rbinom(reps, d, hr / (1 + hr))
z  <- (D1 - d / 2) / sqrt(d / 4)
cat(sprintf("Simulated power with %s events: %.3f (%d replicates)\n", format(d, big.mark = ","), mean(abs(z) > z_a), reps))

Output from running the code

100% of observed effect: events = 303 (80% power), 406 (90% power)
 75% of observed effect: events = 539 (80% power), 721 (90% power)
 50% of observed effect: events = 1,211 (80% power), 1,621 (90% power)
Simulated power with 303 events: 0.800 (20000 replicates)

The simulation uses the fact that, with d events in total and proportional hazards, the number of events in the treatment arm is binomial. It is an approximation that conditions on the total number of events, and it should land close to the target power.

14. Analyses Reported by the Sponsor

The sponsor’s reported primary analysis is a log-rank comparison with a hazard ratio of 1.38, a 95% confidence interval from 1.00 to 1.92, and p = 0.124. The posted statistical presentation uses the sponsor-reported hazard ratio; it does not replace that estimate with a newly calculated hazard ratio from the group medians. A median cannot be used to reconstruct a hazard ratio because the latter depends on time-to-event information across follow-up.

The registry record does not flag a discrepancy between the sponsor’s analysis and the posted presentation. Accordingly, the sponsor’s own analysis is the relevant inferential result to describe. The arm medians and their listed ranges provide descriptive context, but they are not a substitute for the reported log-rank analysis.

For overall survival, the sponsor-reported analysis gives a hazard ratio of 1.220, a 95% confidence interval from 0.806 to 1.845, and p = 0.345. Again, this is a reported time-to-event comparison, not an estimate recalculated from the medians. The available record does not provide participant-level data or enough detail to reproduce either analysis independently. It also does not report the full statistical analysis plan, so the extent to which these analyses follow prespecified procedures cannot be assessed.

15. Multiplicity, Missing Data and Analysis Populations

The registry record does not report how multiplicity was handled. It names a primary endpoint and identifies secondary outcomes, but it does not state whether the testing sequence was hierarchical, whether any adjustment was applied, or which analyses were intended to be confirmatory. When several outcomes are tested, general statistical reasoning says that unadjusted p-values should be interpreted cautiously because multiple opportunities to detect a chance pattern can weaken the evidential meaning of an isolated result.

Missing-data methods are also not reported in the registry record. For progression-free and overall-survival analyses, relevant details include how participants without a recorded event were censored, how loss to follow-up was handled, and what rules were used for incomplete or unevaluable assessments. General explanation: censoring allows observed follow-up to contribute information without treating a participant’s later event time as known, but valid interpretation relies on assumptions about why observation ended. The registry does not provide enough information to assess those assumptions.

The primary endpoint description specifies a per-protocol population. The criteria for that population, the number excluded from it, and the reasons for exclusion are not reported in the registry record. The posted primary arm counts are 90 and 84, while the overall-survival and safety records use different counts. Those differing counts are reported for different analyses; the registry does not explain the differences. They should not be combined or treated as if they refer to an identical analysis population.

Randomized allocation is listed, but the record does not report the sequence-generation or allocation-concealment procedures, nor does it explain whether analyses followed an intention-to-treat principle. Masking is listed as none. It is not reported whether progression outcomes received blinded independent review. These omissions do not establish that a method was absent; they mean the method is not reported in the registry record. Together, these gaps limit assessment of analysis robustness and risk of bias.

16. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: anemia (17.6%), other - respiratory, thoracic and mediastinal disorders (12.1%), diarrhea (3.3%), pain (3.3%), hypertension (2.2%).

The registry reports serious adverse events for 91 participants at risk in each group. Serious adverse events affected 39 in the gefitinib group and 22 in the gefitinib-plus-olaparib group. Other adverse events affected 88 in the gefitinib group and 90 in the combination group, with 91 at risk in each group. Deaths are reported as 10 and 9, respectively. These are descriptive counts, not proof that treatment caused or prevented an event.

The most frequently listed serious-event terms include anemia at 17.6%, other respiratory, thoracic and mediastinal disorders at 12.1%, diarrhea and pain at 3.3% each, and hypertension and pneumonia at 2.2% each. The registry record does not make clear in these summary details whether terms can overlap within participants, nor does it provide exposure-adjusted rates or detailed event timing. No additional denominators or percentages should be inferred from these counts.

General explanation: safety comparisons depend on event definitions, follow-up duration, treatment exposure, and whether counts represent affected participants or event occurrences. The record supplies at-risk and affected counts but not enough detail to assess all these features. The figures therefore describe reported safety observations in the registry, not a causal comparison or a complete characterization of risk.

17. What the Result Does, and Does Not, Mean

The primary result is an estimate of progression-free survival under the registry’s RECIST-based definition and the stated per-protocol analysis. The posted medians are the same at the reported precision, while the hazard-ratio estimate is accompanied by a confidence interval that begins at 1.00 and extends to 1.92. Together, these results do not establish a clear difference between the groups, but they also do not establish that the treatments are equivalent.

A non-significant or uncertain comparison is not evidence that the groups have identical outcomes. Demonstrating equivalence or non-inferiority requires a prespecified design and margin, and those are not reported in the registry record. Likewise, the point estimate should not be presented as a definitive treatment effect without considering its uncertainty, analysis population, and study methods.

The overall-survival analysis provides additional but still uncertain context. Its interval includes 1, and the registry does not report results for the listed best-global-response outcome. The available findings therefore do not constitute a complete account of every outcome named in the record. The safety counts describe reported events but cannot establish causation or support a complete benefit-risk judgment from these summary details alone.

These results apply to the study’s enrolled population and the treatment regimens recorded for this trial. They do not establish outcomes for all patients with lung cancer, for different treatment schedules, or for settings beyond the study. The record’s open-label design and missing details about censoring, missing outcomes, population definitions, and multiplicity further limit how far the results can be interpreted.

The practical statistical conclusion is modest: the registry records a randomized comparison, but its posted primary estimate is imprecise and does not provide clear evidence of a difference. The result should be read as an estimate with uncertainty, not as proof that the combination is beneficial, ineffective, safer, or interchangeable with gefitinib alone.

18. Why This Trial Matters Statistically

This trial illustrates why a time-to-event result should not be reduced to the median alone. The posted progression-free survival medians are equal at the reported precision, but the analysis also includes a hazard ratio that summarizes the comparison across follow-up. Reading both measures together makes visible the distinction between a distribution midpoint and a relative event-rate estimate.

It also shows why uncertainty is central to interpretation. The primary point estimate is not a complete result without its confidence interval and p-value. The interval is broad, and the p-value does not establish a clear difference. The data therefore leave room for a range of relative effects rather than providing a precise answer. A lack of clear separation should not be converted into a claim of equivalence.

The registry also distinguishes an outcome population from other analysis populations. The primary endpoint is described as per protocol, while other posted analyses use different arm counts. Understanding which participants contribute to each result is essential for interpreting whether the analysis represents randomized assignment or a selected subset. The available record does not explain the population differences, highlighting the importance of complete reporting.

Finally, the planning illustrations demonstrate how power calculations depend on the effect size chosen for planning and on the number of time-to-event events. They are separate from the original design justification, which is not reported. Keeping those planning calculations distinct from observed results helps avoid confusing a hypothetical design requirement with evidence generated by the completed trial.

19. 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.

Several trial-specific reporting gaps constrain interpretation. The registry describes the primary analysis as per protocol but does not give the criteria for that population or explain why its arm counts differ from those reported for overall survival and safety. It also does not report an intention-to-treat analysis, so the effect of the per-protocol choice relative to randomized assignment cannot be assessed from this record.

The study is listed as unmasked. Although progression is defined using RECIST v1.0, the registry does not report whether scans or progression assessments were reviewed independently or blinded to treatment assignment. The potential influence of unmasked assessment therefore cannot be evaluated from the available record.

Important time-to-event details are not reported, including the censoring rules, handling of missing or unevaluable assessments, and checks of the assumptions underlying a single hazard ratio. The record also does not report the original sample-size justification, the detailed statistical plan, or interim-analysis procedures. These omissions limit assessment of the precision intended by design and the extent to which the posted analyses were prespecified.

Multiplicity handling is not reported, and best-global-response results are not posted despite being listed as a secondary outcome. The safety summary lacks details on event definitions, exposure time, timing, and whether event terms can overlap. These limitations do not show that the methods or analyses were inadequate; they identify information not reported in the registry record and therefore unavailable for independent evaluation.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Study With Gefitinib in Combination With Olaparib (AZD2281) Versus Gefitinib Alone — NCT01513174.
• Registry record last updated 2024-06-28; status last verified 2024-01.
• Results first posted 2024-06-28.
• 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.

23. Record Summary

The registry describes a completed randomized, parallel, unmasked Phase Ib/IIb study of gefitinib plus olaparib versus gefitinib alone in EGFR mutation-positive advanced non-small-cell lung cancer. The primary outcome is progression-free survival, analyzed in a per-protocol population and summarized with medians and a sponsor-reported log-rank hazard ratio.

The primary medians are 10.9 months in both groups. The reported hazard ratio is 1.38, with a 95% confidence interval from 1.00 to 1.92 and p = 0.124. This estimate is uncertain and does not establish a clear group difference or equivalence. Overall survival is posted as a secondary analysis; best-global-response results are not reported in the registry record.

Interpretation is limited by unreported details on analysis populations, missing outcomes, censoring, multiplicity, and the original design assumptions. Safety figures are descriptive and should not be treated as evidence of causation.