A Study In Patients With Non-Small Cell Lung Cancer Testing If Erlotinib Plus SU011248 (Sunitinib) Is Better Than Erlotinib Alone: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study In Patients With Non-Small Cell Lung Cancer Testing If Erlotinib Plus SU011248 (Sunitinib) Is Better Than Erlotinib Alone (NCT00265317), a randomized phase 2 study. The primary outcome is Progression-Free Survival (PFS).

2. Clinical Question

The trial was designed to compare the combination with erlotinib alone in previously treated patients with advanced or metastatic disease. Its primary endpoint reflects the time until disease progression or death, rather than response alone or survival alone. The registry reports medians and a hazard ratio for the comparison. These measures address different aspects of the time-to-event data: the median describes the time by which half of the analyzed group has experienced an event, while the hazard ratio summarizes the relative event rate over follow-up under the analysis method.

The registry describes a Full Analysis Set based on randomized assignment. This means the primary population was not restricted to participants who actually received the medication assigned to them. The clinical question is therefore framed around randomized groups and the endpoint definition recorded for the trial. The registry record does not report further detail about eligibility criteria beyond the description of advanced or metastatic non-small-cell lung cancer and previous platinum-based treatment.

3. Trial Design

This was a randomized, double-blind, parallel-group phase 2 study sponsored by Pfizer. The registry lists an enrollment of 162 and a completed status. Participants were assigned to sunitinib plus erlotinib or erlotinib plus placebo. The registry describes the study purpose as testing whether the combination was better than erlotinib alone in patients with advanced or metastatic lung cancer previously treated with a platinum-based regimen.

The primary endpoint was progression-free survival, measured in weeks. The primary analysis population was the Full Analysis Set, defined as randomized participants in the randomized phase, kept in the group assigned at initial randomization regardless of whether they received study medication or a different medication. The posted group sizes for this endpoint are 65 for sunitinib plus erlotinib and 67 for erlotinib plus placebo. Those group counts describe the primary endpoint results and should not be confused with total enrollment.

The primary endpoint was assessed from randomization at weeks 8 and 12, then every 8 weeks until disease progression or death, with follow-up described up to Month 17. The registry lists objective response, time to tumor progression, duration of response, overall survival, survival at 1 year, and pharmacokinetic measures among secondary outcomes. It reports 67 outcome measures and 212 analyses, while 62 secondary outcomes are posted.

The registry identifies randomization and double masking but does not report details of the randomization procedure, the masking arrangements, treatment duration, or the protocol's original sample-size assumptions. The arm descriptions are not populated in the registry record. No further design detail should be inferred from the intervention names alone.

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.

Lead-In

Reasons for not completing this period:

Phase 2

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Full Analysis Set (FA):all participants in randomized phase randomized with study medication assignment designated according to initial randomization, regardless of whether participants actually received study medication or received different medication from what they were randomized.

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.

6. Primary Endpoint

The registry lists the primary outcome as Progression-Free Survival (PFS), measured over From randomization to Weeks 8 and 12, then every 8 weeks until disease progression or death (up to Month 17).

Progression-free survival (PFS) measures the time from randomization until the first recorded progression of disease or death on study from any cause, whichever occurs first. In this trial, progression was defined using imaging criteria: at least a 20% increase in the sum of the longest dimensions of target lesions, using the smallest recorded sum since the first dose as the reference, or the appearance of at least one new lesion. The registry says the endpoint was based on third-party independent imaging review.

The reported time was calculated from the event date and randomization date, with the elapsed time expressed in weeks using 7.02 days per week. Assessments were scheduled at weeks 8 and 12, then every 8 weeks until progression or death, up to Month 17. The registry's primary population was the Full Analysis Set, based on initial randomized assignment.

In general, a time-to-event endpoint retains information about when an event occurs, not only whether it occurred by a particular visit. PFS combines progression and death into one endpoint, so it captures either event rather than treating them as separate primary outcomes. The median is the time by which half of the analyzed group has had an event; it does not describe the experience of every participant. The hazard ratio provides a separate comparison of event rates over follow-up. Neither measure alone gives a complete account of individual outcomes or the reasons an event occurred.

7. Results Posted in the Registry

• Sunitinib + Erlotinib vs Erlotinib + Placebo: hazard ratio 0.90 (95% CI 0.57–1.40), reported p = 0.321 (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.

Percentage of Participants With Objective Response

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Sunitinib + Erlotinib: 4.6% (3/65) versus 3.0% (2/67) on control; risk difference 1.6 percentage points (95% CI -4.9 to 8.2), risk ratio 1.55 (0.27–8.95); unadjusted two-proportion z-test p = 0.624.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Sunitinib + Erlotinib and the control. The interval gives the range of effects that remain compatible with the data.

Time to Tumor Progression (TTP)

• 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:
  • Sunitinib + Erlotinib vs Erlotinib + Placebo: hazard ratio 0.92 (95% CI 0.57–1.49), reported p = 0.373 (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 Sunitinib + Erlotinib vs Erlotinib + Placebo and the control. The interval gives the range of effects that remain compatible with the data.

Overall Survival (OS)

• 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:
  • Sunitinib + Erlotinib vs Erlotinib + Placebo: hazard ratio 1.07 (95% CI 0.70–1.61), reported p = 0.617 (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 Sunitinib + Erlotinib vs Erlotinib + Placebo and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes add context but do not create a clear, consistent pattern across distinct measures. The combination group's posted objective response percentage and median time to tumor progression are numerically higher than those of the comparator group. The overall-survival point estimate does not point in the same direction as those outcomes. These endpoints capture different events and should not be treated as interchangeable confirmations of the primary PFS result.

The secondary confidence intervals are broad. The sponsor-reported objective-response interval includes 1 for the risk ratio, while the hazard-ratio intervals for time to tumor progression and overall survival include 1. In general, an interval that includes the no-difference value means the data are compatible with no difference under that measure, as well as with other values in the interval. The estimates are therefore not precise enough to establish a uniform secondary-outcome pattern.

The registry lists 67 outcome measures and 212 analyses, with 62 secondary outcomes posted. When many outcomes are examined, the chance of finding apparently notable results by chance can increase. In general, a multiplicity strategy helps define how evidence across multiple tests should be interpreted. The registry record does not report such a strategy or a hierarchy for the secondary outcomes. Unadjusted secondary results consequently carry less weight than a clearly specified primary analysis and should be read as exploratory or descriptive evidence, not as independent proof of a treatment difference.

Other listed secondary measures include duration of response, survival at 1 year, and pharmacokinetic measures. Their presence broadens the scope of information collected but does not change the primary endpoint or resolve the uncertainty in the primary comparison. The registry record does not report results for every listed secondary outcome in the posted summary, so conclusions should remain limited to the outcomes with results recorded.

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

• Percentage of Participants With Objective Response
• Time to Tumor Progression (TTP)
• Duration of Response (DR)
• Overall Survival (OS)
• Percentage of Participants Surviving at 1 Year
• Area Under the Curve From Time Zero to 24 Hours [AUC(0-24)] of Erlotinib
• AUC(0-24) of Sunitinib
• AUC(0-24) of SU-012662 (Metabolite of Sunitinib)

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is a time-to-event outcome. The event is the first of documented disease progression or death on study from any cause. Time begins at randomization, and the analysis uses the event timing rather than reducing each participant's experience to a simple yes-or-no result at a single visit. The registry describes scheduled assessments at weeks 8 and 12 and then every 8 weeks until progression or death, up to Month 17.

What does the median describe?

For each randomized group, the median PFS is the estimated time by which half of that group's participants have experienced progression or death. The posted medians are 12.3 weeks for sunitinib plus erlotinib and 8.5 weeks for erlotinib plus placebo. A median is a compact summary of the time-to-event distribution, not an average and not a statement that every participant experienced an event at that time. It does not, by itself, show the shape of the full event-time pattern or the number of participants still followed at every time point.

The registry reports 95% confidence intervals for the medians: 8.1 to 19.5 weeks for the combination group and 8.0 to 12.3 weeks for the comparator group. In general, a confidence interval communicates the uncertainty associated with an estimate under the method used. It is not a range that contains every participant's event time.

What does the hazard ratio compare?

The primary comparison uses a hazard ratio reported from a log-rank analysis. The sponsor-reported estimate is 0.898. In general, the hazard ratio summarizes the relative event rates between groups across follow-up, with 1 representing no relative difference. A value below 1 points toward a lower event rate in the first named group; it is not a ratio of the two medians and should not be interpreted as a direct percentage change in survival time. The sponsor-reported interval is an 80% confidence interval, from 0.575 to 1.404, and the reported p-value is 0.321.

The log-rank method compares the timing of events between groups over follow-up. It is a time-to-event comparison, not a simple comparison of the proportions with an event at the final visit. The registry does not report additional detail on the assumptions or implementation of the log-rank analysis. General statistical interpretation of a hazard ratio also depends on the event-rate contrast being a useful summary over the follow-up period; the registry record does not report a separate assessment of that feature.

How do response measures differ?

Objective response is a binary outcome: participants are classified according to whether they met the response definition. The registry reports percentages and an analysis using a risk ratio. A risk ratio compares the probability of response between groups; a value of 1 represents equal probabilities. The posted sponsor analysis reports a risk ratio of 1.546 with a 95% confidence interval from 0.267 to 8.954 and p = 0.625. Such a broad interval conveys substantial uncertainty in the comparison.

A risk difference is another way to express a binary comparison, describing the difference between the groups' response proportions. In general, these two measures answer related but distinct questions: one is relative and the other is absolute. The registry includes an unadjusted recalculation for the response comparison as well as the sponsor's reported analysis. The sponsor's estimate and p-value are the registry's own reported inferential result; the recalculation is a simple comparison based on posted summary statistics and should not replace the sponsor's method.

Why are endpoint methods not interchangeable?

Time to tumor progression and overall survival are also time-to-event outcomes, but they use different event definitions. Time to tumor progression concerns progression, whereas overall survival concerns death. Objective response is a binary measure. A finding on one endpoint cannot be assumed to describe another: participants can have different response, progression, and survival experiences, and the endpoints capture different aspects of follow-up.

In general, over-dispersion means that variation in observed data is greater than a model assumes. Accounting for it can affect uncertainty estimates for some models. The registry does not identify an over-dispersion adjustment for these outcomes, so no such method should be attributed to the analysis.

What is not specified?

The registry does not report the trial's original sample-size justification, the planned statistical analysis details beyond the methods identified with posted results, interim analyses, or a multiplicity strategy. It also does not report how missing endpoint data were handled. These omissions limit the ability to assess whether the posted analysis followed a pre-specified hierarchy or how unobserved outcomes may have influenced estimates. The sample-size illustrations are separate planning calculations based on observed effects and are not evidence of the trial's original design.

11. Confidence Intervals and Precision

Precision concerns how narrowly an estimate is determined by the observed data and its analysis. For the primary endpoint, the posted median PFS estimates are accompanied by 95% confidence intervals. The interval for sunitinib plus erlotinib runs from 8.1 to 19.5 weeks; for erlotinib plus placebo it runs from 8.0 to 12.3 weeks. These intervals show meaningful uncertainty around each median, and their overlap does not itself serve as a formal test of the difference between groups.

The primary hazard ratio is 0.898, with an 80% confidence interval from 0.575 to 1.404. The interval includes 1, the value corresponding to no relative difference. It also spans values on either side of 1, so the estimate does not pin down the direction or size of a difference with high precision. The associated log-rank p-value is 0.321. The p-value and interval describe uncertainty in different forms; neither is a measure of the probability that a treatment effect is true.

Confidence levels matter when reading intervals. In general, a 95% confidence interval and an 80% confidence interval are constructed to have different long-run coverage properties under their respective methods. The primary median intervals are labeled 95%, while the sponsor-reported hazard-ratio interval is labeled 80%. Those levels should not be conflated or compared as if they represented the same degree of coverage.

The secondary estimates are also imprecise. For objective response, the sponsor-reported risk-ratio interval extends from 0.267 to 8.954. For time to tumor progression and overall survival, the reported hazard-ratio intervals include 1. These intervals leave room for effects in more than one direction. The registry's summaries therefore support cautious descriptions of point estimates, but provide limited precision for identifying the size of a between-group difference.

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 sample-size illustrations are planning calculations based on the observed effect, not the trial's original design assumptions. For the primary time-to-event endpoint, the figures are total events, not participants per arm. Under the row labeled 100% of the observed effect, the illustration gives 2,713 total events for 80% power and 3,632 total events for 90% power. The rows for 75% and 50% of the observed effect give larger event requirements.

For outcomes that are not time-to-event measures, the figures are participants per arm. The 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 the stated fraction of the observed effect. The original trial's own sample-size assumptions are not reported in the registry record.

In plain terms, the 80% power column gives the planned number of events or participants for a design that would have an 80% chance of detecting the assumed effect under the calculation's conditions, if that effect were present. The 90% power column represents a higher chance of detection under those same conditions and therefore requires a larger planned sample or event count. Smaller effects are harder to distinguish from random variation, so the rows using smaller fractions of the observed effect require more participants or events.

These calculations are not observed power. They are planning illustrations: they do not describe the probability that the completed trial's result is correct, nor do they change the uncertainty in the reported treatment comparison. They also do not imply that the trial was originally designed around the observed effect.

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 = 0.898, so ln(HR) = -0.1076.
2. d = 4 × (1.96 + 0.842)² / [ln(0.898)]² = 2,712.5, rounded up to 2,713 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 <- 0.898                              # 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(265317)
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 = 2,713 (80% power), 3,632 (90% power)
 75% of observed effect: events = 4,823 (80% power), 6,456 (90% power)
 50% of observed effect: events = 10,850 (80% power), 14,525 (90% power)
Simulated power with 2,713 events: 0.803 (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. Group-Sequential Design Analysis

Many time-to-event trials, especially in oncology, stop or report early when interim data are overwhelming, so a single fixed-sample number understates the real design problem. The calculation below is an illustrative group-sequential design built from this trial's posted control median and hazard ratio. The registry record does not state the planned number of looks or the spending function, so the design is a planning illustration, not a reconstruction of the protocol. Assumed: one-sided α = 0.025, 90% power, three equally spaced looks on information, Lan-DeMets O'Brien-Fleming spending, 24 months of accrual, 60 months in total and a 1% monthly dropout hazard.

Under these assumptions the design needs about 3,784 participants in total, with the boundaries below.

Fixed design versus group-sequential alternatives

Group-sequential designs cost a small increase in maximum sample size and events compared with the fixed design, in exchange for the chance to stop early for overwhelming efficacy. O'Brien-Fleming spending keeps early boundaries very conservative and the final boundary close to the fixed-design value, whereas Pocock spending spends more alpha early.

Sensitivity to a smaller true effect

Planning for exactly the observed effect is optimistic, because observed effects in published trials tend to be somewhat larger than the true effect. The table repeats the main design (three looks, O'Brien-Fleming spending, 90% power) for smaller true hazard ratios.

For context, this trial enrolled 162 participants.

Cross-check, upper boundaries: gsDesign 3.71, 2.511, 1.993 | rpact 3.71, 2.511, 1.993. Agreement between the two packages confirms the boundary calculation.

R code and output

suppressWarnings(suppressMessages({ library(gsDesign); library(rpact) }))
med_control <- 1.96153846153846          # control-arm median, months (converted from the posted unit)
hr          <- 0.898          # hazard ratio used for planning
lambdaC     <- log(2) / med_control
eta         <- 0.01               # dropout hazard per month (assumption)
acc_months  <- 24                 # accrual duration (assumption)
study_months<- 60                 # total study duration (assumption)

design <- function(k, sfu, beta, h = hr) gsSurv(k = k, test.type = 1, alpha = 0.025, beta = beta, sfu = sfu,
  lambdaC = lambdaC, hr = h, hr0 = 1, eta = eta, gamma = 1, R = acc_months, T = study_months, minfup = study_months - acc_months)

# Main design: 3 looks, Lan-DeMets O'Brien-Fleming spending, one-sided alpha 0.025, 90% power
x <- design(3, sfLDOF, 0.10)
cat(sprintf("N_TOTAL|%s\n", format(ceiling(sum(x$gamma * x$R)), big.mark = ",")))
for (i in 1:3) cat(sprintf("GS|%d|%.1f|%.0f|%.3f|%.5f|%.5f\n", i, x$T[i], x$n.I[i], x$upper$bound[i], pnorm(-x$upper$bound[i]), cumsum(x$upper$spend)[i]))
cat("\n"); print(gsBoundSummary(x))

# Comparison of designs
cmp <- list(list("Fixed design, one analysis", 1, sfLDOF), list("Two looks, O'Brien-Fleming", 2, sfLDOF),
            list("Three looks, O'Brien-Fleming", 3, sfLDOF), list("Three looks, Pocock", 3, sfLDPocock))
cat("\n")
for (b in c(0.20, 0.10)) for (d in cmp) {
  y <- design(d[[2]], d[[3]], b)
  cat(sprintf("DES|%s|%d|%.0f|%s|%.1f\n", d[[1]], round(100 * (1 - b)), tail(y$n.I, 1), format(ceiling(sum(y$gamma * y$R)), big.mark = ","), tail(y$T, 1)))
}

# Sensitivity: what if the true effect is smaller than the observed one? (3 looks, O'Brien-Fleming, 90% power)
cat("\n")
for (fr in c(1, 0.75, 0.5)) {
  y <- design(3, sfLDOF, 0.10, exp(log(hr) * fr))
  cat(sprintf("SENS|%.2f|%.3f|%.0f|%s\n", fr, exp(log(hr) * fr), tail(y$n.I, 1), format(ceiling(sum(y$gamma * y$R)), big.mark = ",")))
}

# Independent cross-check of the boundaries with rpact
r <- getDesignGroupSequential(kMax = 3, alpha = 0.025, beta = 0.10, sided = 1, typeOfDesign = "asOF", informationRates = x$timing)
cat(sprintf("\nCross-check, upper boundaries: gsDesign %s | rpact %s\n", paste(round(x$upper$bound, 3), collapse = ", "), paste(round(r$criticalValues, 3), collapse = ", ")))

Method: LachinFoulkes 
     Analysis              Value Efficacy
    IA 1: 33%                  Z   3.7103
      N: 1708        p (1-sided)   0.0001
 Events: 1225       ~HR at bound   0.8089
    Month: 11   P(Cross) if HR=1   0.0001
              P(Cross) if HR=0.9   0.0338
    IA 2: 67%                  Z   2.5114
      N: 2980        p (1-sided)   0.0060
 Events: 2450       ~HR at bound   0.9035
    Month: 19   P(Cross) if HR=1   0.0060
              P(Cross) if HR=0.9   0.5603
        Final                  Z   1.9930
      N: 3784        p (1-sided)   0.0231
 Events: 3674       ~HR at bound   0.9364
    Month: 60   P(Cross) if HR=1   0.0250
              P(Cross) if HR=0.9   0.9000



Cross-check, upper boundaries: gsDesign 3.71, 2.511, 1.993 | rpact 3.71, 2.511, 1.993

15. Analyses Reported by the Sponsor

The registry reports the sponsor's primary PFS comparison as a log-rank hazard ratio of 0.898, with an 80% confidence interval from 0.575 to 1.404 and p = 0.321. The posted median estimates and their 95% confidence intervals describe each group separately; the hazard ratio and its interval describe the between-group time-to-event comparison. They are complementary summaries, not interchangeable calculations.

The record does not flag a discrepancy between the sponsor's reported analysis and the simple recalculated summaries. The sponsor's own inferential analysis remains the relevant reported result for the primary comparison. Recalculations based on posted summary statistics are unadjusted and do not reproduce the full participant-level analysis or necessarily use the sponsor's analytic details.

For objective response, the sponsor reports a Cochran-Mantel-Haenszel risk ratio of 1.546, with a 95% confidence interval from 0.267 to 8.954 and p = 0.625. The registry also contains a simple unadjusted recalculation from summary counts. Because that calculation is not the sponsor's analysis, it is best used as a descriptive cross-check rather than a replacement for the reported method. The registry record does not describe enough detail to reconstruct every aspect of the sponsor's analyses independently.

16. Multiplicity, Missing Data and Analysis Populations

The registry lists a large set of outcomes and analyses: 67 outcome measures and 212 analyses, with 62 secondary outcomes posted. Testing or estimating many outcomes raises an interpretive issue: some results may look unusual through chance alone when many questions are examined. In general, multiplicity procedures or a pre-specified testing hierarchy can control how evidence is judged across multiple comparisons. The registry record does not report whether such procedures were planned or used, and it does not identify a hierarchy for the secondary outcomes. This makes the secondary results less conclusive than a single, clearly pre-specified primary comparison.

The registry defines the primary analysis population as the Full Analysis Set: participants randomized in the randomized phase, categorized according to initial randomization regardless of whether they received study medication or received a different medication. This preserves the randomized assignment as the basis for the primary comparison. The primary endpoint group sizes reported are 65 and 67. The registry does not provide further detail about other analysis populations or how participants were handled in every secondary analysis.

Details of missing data and censoring are important for time-to-event analyses because participants may not all have an observed progression or death during the recorded follow-up. The registry describes the event definition and assessment schedule, but handling of missing assessments, withdrawal, censoring, and incomplete follow-up is not reported in the registry record. Without those details, the effect of incomplete observation on the estimates cannot be evaluated from the summary alone.

The registry record also does not report interim analyses, the original statistical analysis plan, or the original sample-size justification. Their absence from the record does not establish that no such plans existed; it means they cannot be described from the recorded information. These gaps limit assessment of analysis choices and the extent to which the posted results reflect pre-specified decisions.

17. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: dyspnoea (15.4%), disease progression (12.5%), diarrhoea (7.8%), anaemia (7.7%), thrombocytopenia (7.7%).

The serious adverse event figures are reported for four groups: sunitinib plus erlotinib in an original lead-in group, sunitinib plus erlotinib in an amended lead-in group, the sunitinib plus erlotinib group, and erlotinib plus placebo. The numbers at risk and affected are, respectively, 13 and 5, 17 and 7, 64 and 29, and 64 and 28. The lead-in groups are listed separately from the main treatment groups and should not be combined with them without further information about their relationship.

The registry lists dyspnoea at 15.4% and disease progression at 12.5% among the top serious adverse event terms. Diarrhoea is listed at 7.8%; anaemia, thrombocytopenia, and tachycardia are each listed at 7.7%. These are event-term percentages as recorded, not a complete safety profile or a between-group statistical comparison. The registry does not report deaths in these safety summaries. The broad counts and event terms should be interpreted alongside the study's group definitions and the limited safety detail recorded.

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

The posted results describe a randomized comparison of sunitinib plus erlotinib with erlotinib plus placebo in previously treated patients with advanced or metastatic non-small-cell lung cancer. The primary outcome was time from randomization to progression or death. The combination group's median PFS was 12.3 weeks and the comparator group's median was 8.5 weeks. The sponsor-reported hazard ratio was 0.898, with an 80% confidence interval that includes 1 and a log-rank p-value of 0.321.

These figures do not establish a clear between-group difference in PFS. The point estimates provide a direction for the observed comparison, but the confidence interval allows effects on both sides of the no-difference value. The result is therefore uncertain, rather than definitive evidence for or against a difference. A non-small p-value is not proof that groups are identical, just as a point estimate in one direction is not proof that a difference exists.

The secondary outcomes do not remove this uncertainty. They measure different aspects of disease and survival, and the reported estimates are imprecise. The registry record does not report a multiplicity strategy for interpreting the many outcomes and analyses. The results should consequently be understood as the recorded estimates for this trial, not as a guarantee of outcomes for any individual or a basis for medical decisions.

In general, clinical interpretation also depends on the population enrolled, the treatment regimens, follow-up, outcome definitions, and conduct of the trial. The registry describes enrollment of 162 and a primary analysis based on randomized assignment, but does not provide all details needed to judge generalizability or the impact of missing data. These statistical results alone do not settle questions beyond the recorded trial comparison.

19. Why This Trial Matters Statistically

This trial illustrates why a clinical-trial result should be read through its endpoint definition, analysis population, estimate, and uncertainty together. PFS combines documented progression and death into one time-to-event outcome. The median summarizes when half of a group experienced an event, while the hazard ratio compares event rates across follow-up. Neither summary alone captures every feature of the event-time distributions.

It also demonstrates why the confidence interval matters alongside the point estimate. The primary hazard ratio points in one direction, but its 80% confidence interval includes the no-difference value and extends to values in the opposite direction. The median estimates are accompanied by 95% intervals with uncertainty. Reading only the point estimates would overstate how precisely the comparison is known.

The record offers a useful distinction between sponsor-reported analyses and unadjusted recalculations from posted summaries. A recalculation can help readers understand the data's scale, but it does not reproduce an analysis based on participant-level information or replace the sponsor's reported method. The sample-size illustrations similarly demonstrate the dependence of planning requirements on the effect chosen to detect and on the assumed design conditions. They are not a retrospective measure of the trial's power.

Finally, the registry's numerous outcomes and analyses make multiplicity relevant. Without a reported adjustment or hierarchy, secondary findings need cautious interpretation. These are general statistical lessons illustrated by the recorded trial; they do not substitute for a complete protocol or statistical analysis plan.

20. 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 registry record leaves several trial-specific limitations unresolved. It describes a completed phase 2 study and an enrollment of 162, but does not report the original sample-size justification or the assumptions used when the trial was designed. The planning illustrations based on observed effects cannot fill that gap or establish how the trial was originally powered.

The primary endpoint follow-up is described up to Month 17, with scheduled assessments at weeks 8 and 12 and then every 8 weeks. The record does not report missing-data handling, censoring details, withdrawal patterns, or how incomplete imaging assessments were addressed. These details matter to the interpretation of time-to-event estimates and cannot be recovered from the posted summaries alone.

The registry does not report a multiplicity strategy despite listing many outcome measures and analyses. It also does not report interim analyses or enough detail to assess whether all analysis decisions were pre-specified. The arm descriptions are blank, and further details about treatment administration and randomization are not included in the recorded summary.

Safety information is presented in separate lead-in and main treatment groups, and the registry does not report deaths in those summaries. The available safety terms are not a complete accounting of all possible events. The posted aggregate results also do not permit evaluation of individual participant trajectories or subgroup effects. These limitations counsel against extending the summaries beyond the population, outcomes, and comparisons recorded.

21. Related Tutorials

22. Related Statistical Calculators

23. Sources

• ClinicalTrials.gov: A Study In Patients With Non-Small Cell Lung Cancer Testing If Erlotinib Plus SU011248 (Sunitinib) Is Better Than Erlotinib Alone — NCT00265317.
• Registry record last updated 2013-02-06; status last verified 2013-01.
• Results first posted 2011-06-10.
• 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.

24. Record Summary

NCT00265317 is a completed randomized, double-blind, parallel-group phase 2 study sponsored by Pfizer. It enrolled 162 participants with non-small-cell lung cancer and compared sunitinib plus erlotinib with erlotinib plus placebo. The primary outcome was PFS, defined as time from randomization to documented progression or death, whichever occurred first.

The primary group medians were 12.3 weeks and 8.5 weeks. The sponsor-reported hazard ratio was 0.898, with an 80% confidence interval from 0.575 to 1.404 and p = 0.321. The interval includes the no-difference value, leaving substantial uncertainty about the comparison. Secondary outcomes and safety summaries provide additional descriptive information, but the registry record does not report a multiplicity strategy, missing-data handling, or the original sample-size justification.