A Study of Tarceva (Erlotinib) to Compare Two Different Doses in in Currently Smoking Patients With Advanced or Metastatic Non-Small Cell Lung Cancer (CURRENTS): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Tarceva (Erlotinib) to Compare Two Different Doses in in Currently Smoking Patients With Advanced or Metastatic Non-Small Cell Lung Cancer (CURRENTS) (NCT01183858), a randomized phase 3 study. The primary outcome is Progression-Free Survival (PFS).

2. Clinical Question

The research question was a dose comparison within a defined clinical setting, not a comparison of erlotinib with no treatment or with another drug. The registry describes a prospective, double-blind, randomized phase III study with a parallel-group design. Participants were assigned to one of the two daily doses, and treatment was to continue until disease progression. The trial also assessed response, disease control, overall survival, time to progression, and adverse events. Those measures broaden the description of outcomes, but the primary question is specifically about progression-free survival.

3. Trial Design

CURRENTS was a randomized, double-blind, parallel-group phase III study sponsored by Hoffmann-La Roche. Its title describes a prospective comparison of erlotinib 300 mg with erlotinib 150 mg in current smokers with locally advanced or metastatic non-small cell lung cancer in the second-line setting after chemotherapy failure. The registry lists an enrollment of 315. The study status is completed; the recorded start was 2010-10 and primary completion was 2013-10.

Participants were assigned to receive one of two single daily oral doses of erlotinib. The arm descriptions specify continuation until disease progression. The registry summary says participants had advanced or metastatic disease after failure of first-line platinum-based chemotherapy and were current smokers who did not intend to stop smoking during the study. The randomized, double-blind structure was intended to compare the two dose groups while limiting bias in assignment and in the conduct or assessment of the study. Further operational detail about how masking was maintained is not reported in the registry record.

The primary endpoint was progression-free survival, measured from randomization to disease progression or death. The posted analysis gives median values in weeks, confidence intervals, and a log-rank hazard-ratio comparison. The primary timeframe runs from randomization to the clinical cutoff of 28 October 2013, up to 36.5 months. Secondary measures included overall survival, overall response rate, disease control rate, time to progression, and adverse-event outcomes. The registry lists 8 outcome measures and 6 posted secondary outcomes.

The primary analysis description calls the population intent-to-treat and says 2 participants were excluded: 1 randomized in error and 1 with missing source data. The group-level analysis counts and this stated exclusion should be read as the registry reports them; additional detail about analysis-set construction is 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.

Before assignment: 315 participants were randomized. 313 participants were included in the Intent-to -treat (ITT) population. The ITT population excluded 2 randomized participants: 1 participant randomized in error and 1 participant with missing source data.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent-to-treat Population included all randomized participants. 2 participants were excluded from analysis: 1 participant randomized in error and 1 participant with missing source data.

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: Intent-to-treat Population included all randomized participants. 2 participants were excluded from analysis: 1 participant randomized in error and 1 participant with missing source data.

6. Primary Endpoint

The registry lists the primary outcome as Progression-Free Survival (PFS), measured over Randomization to Clinical Cutoff: 28 October 2013 (Up to 36.5 Months).

Progression-free survival (PFS) measures the time from randomization until the first occurrence of disease progression or death. It combines these events into a time-to-event outcome, so it reflects how long participants remain alive without recorded progression. It is distinct from overall survival, which concerns time to death, and from time to progression, which does not necessarily include death as an event.

The registry defines progression using lesion assessments. For target lesions, progressive disease was defined as at least a 20% increase in the sum of the longest diameter, using the smallest recorded sum since treatment started as the reference, or the appearance of one or more new lesions. For non-target lesions, progression meant the appearance of one or more new lesions and/or unequivocal progression of existing non-target lesions. These criteria make the endpoint depend on both the timing and the recorded assessment of disease status.

The registry reports median PFS in weeks for each dose group. A median is a summary of the event-time distribution: it identifies the point by which half of the relevant distribution has had an event, subject to the analysis of censored observations. It does not describe every participant's experience or the shape of the full survival curve. PFS is often used in trials of advanced cancer because progression can be observed before death for some participants, allowing an assessment of disease control over follow-up. That general rationale does not establish how much weight participants or clinicians should give the endpoint in a particular setting.

The posted timeframe extends from randomization to the clinical cutoff on 28 October 2013, up to 36.5 months. The registry record does not provide further detail about assessment schedules or handling of particular censoring circumstances.

7. Results Posted in the Registry

• Erlotinib 150 mg vs Erlotinib 300 mg: hazard ratio 1.05 (95% CI 0.83–1.33), reported p = 0.671 (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.

Disease Control Rate (DCR)

• 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:
  • Erlotinib 300 mg: 36.5% (58/159) versus 40.3% (62/154) on control; risk difference -3.8 percentage points (95% CI -14.6 to 7.0), risk ratio 0.91 (0.68–1.20); unadjusted two-proportion z-test p = 0.491.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Erlotinib 300 mg and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists secondary outcomes spanning survival, response, disease control, time to progression, and adverse events. This broader set can describe aspects of outcome that PFS alone does not capture. Overall survival concerns death; response and disease control concern tumor status; time to progression concerns the timing of progression; adverse-event measures describe harms recorded during the study. These endpoints differ in definition and interpretation, so their results should not be treated as interchangeable confirmations of one another.

The posted DCR estimates and unadjusted comparison are compatible with uncertainty about both the size and direction of a between-group difference. The remaining secondary endpoints are identified in the registry, but detailed results are not included in the registry record for this explanation; they should not be inferred from the primary PFS result. In general, secondary outcomes can support a fuller account of a trial, but they do not automatically confirm or contradict the primary endpoint.

Unadjusted secondary results carry less inferential weight when several outcomes are considered because multiple separate tests can increase the chance of apparently notable findings. Whether the trial controlled this issue through a hierarchy or adjustment is not reported in the registry record. The registry also does not report enough detail to identify which secondary comparisons, if any, were intended as confirmatory. The DCR recalculation is explicitly unadjusted and has no sponsor-reported analysis listed, so it is best read as a descriptive supplement rather than decisive evidence.

The primary PFS result remains the central reported comparison. A secondary result should be interpreted in relation to its own definition, estimate, uncertainty, and analysis status. A pattern across outcomes can be informative only when the outcomes and their analyses are sufficiently reported; it should not be constructed by assuming results for endpoints whose values are not reported in the registry record.

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

• Overall Survival (OS)
• Overall Response Rate (ORR)
• Disease Control Rate (DCR)
• Time to Progression (TTP)
• Number of Participants With Adverse Events (AEs) at the End of the Study
• Overall Survival (OS) at the End of Study

9. Statistical Interpretation

10. Statistical Methods Explained

The registry reports a log-rank analysis for the primary time-to-event comparison and a hazard ratio with a 95% confidence interval. These methods answer related but distinct questions: the log-rank test assesses evidence against a no-difference null hypothesis across follow-up, while the hazard ratio expresses the relative event rate as summarized by the analysis. The reported estimate should be read with its interval because a point estimate alone does not convey its precision.

What does a hazard ratio describe?

General statistical explanation: A hazard is the instantaneous event rate among participants who have not yet experienced the event, conditional on remaining under observation and event-free up to that time. A hazard ratio compares the hazards between groups under the model or summary used. A hazard ratio near 1 indicates similar estimated hazards; values above or below 1 indicate a relative direction according to the stated group order. It is not a direct measure of the difference in median survival times, nor does it by itself describe the absolute probability of an event by a particular time.

For this trial, the comparison is stated as erlotinib 150 mg versus erlotinib 300 mg. The reported hazard ratio is 1.05, with a 95% confidence interval from 0.83 to 1.33. Because the interval crosses 1, the posted estimate is compatible with no relative hazard difference as well as with relative differences in either direction. The registry does not provide further details about whether proportional hazards were assessed or how departures from that assumption would be handled. Those details are not reported in the registry record.

What does the log-rank test contribute?

General statistical explanation: The log-rank test compares event-time experience between groups over the observed follow-up, accounting for censoring in the survival analysis. It is commonly used when the outcome is time to an event and some participants have not had the event by the end of observation. Its interpretation depends on the event definitions, censoring, and analysis procedures. A p-value describes compatibility with a specified null hypothesis under the test; it does not measure the size or clinical importance of a difference.

The reported log-rank p-value for PFS is 0.671. It does not show that the treatments are identical, and it does not quantify the probability that the null hypothesis is true. The interval around the hazard ratio is useful alongside this test because it shows a range of relative effects compatible with the data and analysis, rather than a yes-or-no conclusion based only on a significance threshold.

How should the median and its interval be read?

The posted group summaries give median PFS values and 95% confidence intervals. General statistical explanation: A confidence interval is a range produced by a method that, under its assumptions and repeated use, has a stated coverage property. It reflects sampling uncertainty in the estimated quantity. It is not a range containing a specified proportion of individual participant outcomes, and it does not mean there is a 95% probability that the fixed true value lies in the interval. The intervals for the group medians are not themselves a direct test of the between-group hazard ratio.

The registry describes PFS as time from randomization to progression or death, and it specifies lesion-based progression criteria. Time-to-event methods can use information from participants who have not had the event by the cutoff, provided their observation is censored appropriately. The record does not report detailed censoring rules, assessment schedules, or sensitivity analyses. Those omissions limit how fully a reader can reconstruct the analysis from the registry summary.

How do binary secondary outcomes differ?

General statistical explanation: A binary outcome records whether an event or state occurred by a defined assessment or period, rather than the time until it occurred. Disease control rate is posted as a percentage of participants, with group-level 95% confidence intervals. The unadjusted recalculation compares the group proportions using a risk difference, defined as the proportion in the 300 mg group minus the proportion in the 150 mg group. A risk difference of zero represents equal proportions; a negative value indicates a lower observed proportion in the 300 mg group under that subtraction order.

The recalculated DCR interval is from -0.145 to 0.070 and its unadjusted p-value is 0.491. As a general matter, an interval spanning zero does not provide clear evidence for a difference under the stated comparison, while its width indicates uncertainty about the size and direction of the difference. This is a simple unadjusted recalculation from the posted summary statistics. The registry contains no sponsor-reported DCR comparison in its posted analyses, and no adjustment for other outcomes is represented in this calculation.

Why does multiplicity matter?

General statistical explanation: When several outcomes or comparisons are examined, the chance of seeing at least one apparently unusual result can increase if each is judged separately without adjustment. Prespecified outcome hierarchies, multiplicity adjustments, or other analysis strategies can address this issue, but their details need to be reported to assess how confirmatory a secondary result is. The registry does not report a multiplicity strategy for the posted outcomes. Consequently, unadjusted secondary findings should be regarded as less conclusive than a single prespecified primary comparison, particularly when they are considered alongside multiple other measures.

The registry reports the sponsor's PFS hazard ratio and log-rank result. The DCR comparison is a recalculation from posted summaries, not a sponsor-reported inferential analysis. Keeping those sources distinct matters: the sponsor's own reported primary analysis is the relevant reported result, while the recalculated secondary comparison is descriptive and unadjusted. The registry does not report the original statistical analysis plan or enough detail to determine whether additional planned analyses were performed but not posted.

11. Confidence Intervals and Precision

The primary PFS medians were 6.86 weeks for erlotinib 150 mg, with a 95% confidence interval from 6.29 to 12.0, and 7.0 weeks for erlotinib 300 mg, with a 95% confidence interval from 6.29 to 11.0. These intervals describe uncertainty around each group's median estimate. They are not individual outcome ranges and should not be treated as a direct interval for the difference between groups.

The more direct relative comparison is the reported hazard ratio of 1.05 for erlotinib 150 mg versus erlotinib 300 mg, with a 95% confidence interval from 0.83 to 1.33. The interval includes 1, so the data are compatible with no relative hazard difference under this analysis, but also with effects in either direction. Its breadth indicates that the estimate is not precise enough to rule out a range of potentially relevant relative differences. The p-value of 0.671 is consistent with the absence of strong statistical evidence against the null comparison; it does not prove equivalence.

For DCR, the group confidence intervals were 32.4% to 48.5% for erlotinib 150 mg and 29.0% to 44.5% for erlotinib 300 mg. The unadjusted recalculated risk-difference interval, from -0.145 to 0.070, includes zero and leaves uncertainty about direction. The group intervals and the difference interval answer different questions, so they should not be compared as though they were interchangeable.

General statistical explanation: Confidence intervals become narrower when estimates are more precise and wider when the available information leaves greater uncertainty. Precision depends on factors such as the number of observed events, variability, follow-up, and the analysis method. The registry does not report all details needed to assess these factors fully. Neither a confidence interval nor a p-value alone determines clinical importance; that requires context about the outcome and the size of effects that matter, which is not reported in the registry record.

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 figures are illustrations based on the observed effects, not the trial's original design calculations. For PFS, the planning quantities are total numbers of events, not participants per arm. For the DCR comparison, the quantities are participants per arm. The two types of figures therefore have different units and should not be compared as though they represented the same resource requirement.

The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The calculation uses the stated fraction of the observed effect as the effect to detect. For time-to-event outcomes, the event totals use the Schoenfeld method. The trial's own original sample-size assumptions and justification are not reported in the registry record.

In plain language, the 80% power column gives the planned number of events or participants required under those assumptions for an 80% chance of detecting the specified effect if that effect is the true one. The 90% power column gives the corresponding planning quantity for a 90% chance. These are long-run properties of a design under its assumptions, not probabilities that the finding in this completed trial is correct. The figures are not observed power.

The planning rows consider 100%, 75%, and 50% of the observed effect. A smaller effect is harder to distinguish from random variation, so the planning quantity rises as the effect to detect becomes smaller. For the PFS illustration, the values are total event counts; for DCR, they are participants per arm. The associated curve should be interpreted in the same planning context: it relates the assumed effect size to the amount of information needed, rather than describing a new trial result or changing the observed outcome.

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 = 1.050, so ln(HR) = 0.0488.
2. d = 4 × (1.96 + 0.842)² / [ln(1.050)]² = 13,188.7, rounded up to 13,189 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 <- 1.05                              # 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(1183858)
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 = 13,189 (80% power), 17,656 (90% power)
 75% of observed effect: events = 23,447 (80% power), 31,389 (90% power)
 50% of observed effect: events = 52,755 (80% power), 70,624 (90% power)
Simulated power with 13,189 events: 0.799 (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 primary PFS result is the sponsor's reported analysis: a log-rank hazard ratio of 1.05 for erlotinib 150 mg versus erlotinib 300 mg, with a 95% confidence interval from 0.83 to 1.33 and p = 0.671. The primary comparison is not replaced by a newly calculated hazard ratio from the posted medians. Time-to-event hazard ratios cannot be reconstructed reliably from the two median values alone, and the registry's reported analysis is the applicable result.

The registry does not flag a discrepancy between the reported primary analysis and the results summarized for the comparison. For the DCR endpoint, there is no sponsor-reported inferential analysis listed. Its risk difference, confidence interval, and p-value are an unadjusted recalculation from the posted group summaries and should be distinguished from the sponsor's PFS analysis. The DCR calculation is useful as a descriptive comparison but does not carry the same status as a reported primary analysis.

More broadly, a recalculation from summary data can check or express a simple comparison, but it cannot reproduce analyses that depend on participant-level data, censoring, covariate adjustment, or other modeling details. Those details are not reported in the registry record for the additional comparisons. The sponsor's own reported analysis therefore remains the source for the primary PFS inferential result.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists multiple outcomes, including PFS, overall survival, response, disease control, time to progression, and adverse-event measures. Considering multiple outcomes raises the general statistical issue of multiplicity: when many hypotheses are tested, some results may appear unusual by chance even if there are no underlying differences. A planned hierarchy or adjustment can help control the interpretation of multiple tests, but neither should be assumed without documentation.

The multiplicity strategy is not reported in the registry record. The original statistical analysis plan, including any hierarchy among secondary endpoints or adjustment of secondary comparisons, is also not reported in the registry record. This means the posted secondary results cannot be classified confidently as confirmatory or exploratory on the basis of the registry summary alone. The DCR comparison is an unadjusted recalculation and should be interpreted with that limitation in mind. The reported primary PFS analysis is distinct from that secondary recalculation.

The primary analysis description says the intent-to-treat population included all randomized participants, while also stating that 2 participants were excluded from analysis: 1 participant randomized in error and 1 participant with missing source data. Further information about the rationale, timing, and effects of these exclusions is not reported in the registry record. The group analysis counts are recorded as 154 for erlotinib 150 mg and 159 for erlotinib 300 mg. The registry does not provide additional participant-level accounting that would clarify the relationship between the stated exclusions and every analysis denominator.

Missing-data handling is important because participants without complete outcome information may differ from those with observed outcomes. For a time-to-event endpoint, incomplete follow-up is commonly handled through censoring, but the registry summary does not give detailed censoring rules or sensitivity analyses. For binary secondary outcomes, the record does not describe how missing assessments were handled. These methods are not reported in the registry record, so their possible influence cannot be assessed from the posted summary.

The registry also does not report details of interim analyses, monitoring-related statistical boundaries, or a complete analysis plan. Those procedures should not be inferred. The available results support interpretation of the reported estimates and their stated uncertainty, but not a reconstruction of all planned and conducted statistical decisions.

16. 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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: pneumonia (3.9%), dyspnoea (1.9%), respiratory tract infection (1.9%), general physical health deterioration (1.9%), myocardial infarction (1.9%).

The posted safety counts describe participants with serious adverse events and participants with other adverse events in each dose group. For erlotinib 150 mg, the registry reports 29 participants with serious adverse events among 154 at risk and 115 with other adverse events among 154 at risk. For erlotinib 300 mg, it reports 35 with serious adverse events among 158 at risk and 128 with other adverse events among 158 at risk. These are counts as posted, not a comparative safety analysis.

The most frequent listed serious adverse-event terms include pneumonia at 3.9%, dyspnoea at 1.9%, respiratory tract infection at 1.9%, general physical health deterioration at 1.9%, myocardial infarction at 1.9%, and dehydration at 1.3%. These term-specific percentages describe the registry's listed events; they should not be interpreted as a complete account of all adverse events or as evidence that a particular dose caused an event.

Different adverse-event categories may have different definitions and reporting rules. The registry does not report enough detail to assess event duration, severity grading, causality judgments, exposure-adjusted rates, or comparative statistical testing. Death counts are not reported in the registry record. The posted safety figures therefore offer a descriptive account, not a definitive estimate of relative safety between doses.

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

The registry reports that median PFS was 6.86 weeks in the erlotinib 150 mg group and 7.0 weeks in the erlotinib 300 mg group. The sponsor's log-rank analysis estimated a hazard ratio of 1.05 for 150 mg versus 300 mg, with a 95% confidence interval from 0.83 to 1.33 and p = 0.671. Taken together, these results do not provide clear statistical evidence of a difference in PFS between the two dose groups in this trial.

That interpretation is not the same as demonstrating that the doses are equivalent or that no difference exists. The confidence interval includes 1 and leaves uncertainty about the relative hazard in either direction. Equivalence or non-inferiority requires a defined margin and an analysis designed to address that question; such a margin and analysis are not reported in the registry record. A non-significant p-value by itself cannot establish equality.

The result applies to the randomized comparison in current smokers with locally advanced or metastatic non-small cell lung cancer in the second-line setting described by the registry, under the study's follow-up and analysis. It does not establish how the doses compare with another drug, with no treatment, or in populations outside that setting. The registry summary alone does not support conclusions about unreported outcomes or about individual participants.

Secondary outcomes may add information about survival, response, disease control, progression, and adverse events, but they have distinct definitions and may be subject to multiplicity and missing-data limitations. The DCR recalculation is unadjusted, and the registry does not report a multiplicity strategy. The posted safety counts are descriptive and do not by themselves establish comparative safety. These limits mean the results should be understood as estimates with uncertainty, rather than as a complete account of benefit and harm.

Statistical evidence is only one part of interpreting a clinical trial. The registry does not report a threshold for a clinically important PFS difference or detailed patient-level information needed to judge the practical importance of the estimates. The findings therefore describe what the posted comparison supports statistically, without making a treatment recommendation.

18. Why This Trial Matters Statistically

This trial is statistically informative because it compares two active dose levels in a defined population using random allocation, double masking, and a parallel-group design. Such a design can reduce some sources of systematic bias in a direct comparison. The primary endpoint is time-to-event, which requires interpretation of both event timing and incomplete follow-up rather than a simple comparison of final proportions.

The result also illustrates why an estimate, its confidence interval, and a test result should be considered together. The reported hazard ratio is close to 1, while its confidence interval includes the null value and extends in both directions. The p-value does not establish equivalence, and the point estimate alone does not describe precision. Reading the measures together gives a more restrained interpretation than relying on any one statistic.

The distinction between sponsor-reported analyses and recalculations is important. The PFS comparison is reported by the sponsor using a log-rank method. The DCR comparison is an unadjusted recalculation from posted summaries and is not listed as a sponsor-reported analysis. Keeping those analytic sources separate prevents a simple secondary calculation from being mistaken for the primary inferential result.

The study also highlights limits on what registry summaries can establish. Multiplicity procedures, detailed missing-data handling, and the original sample-size justification are not reported in the registry record. Without those details, readers should avoid treating all secondary outcomes as equally confirmatory or inferring the rationale behind the study's design. Statistical interpretation depends not only on the numerical estimates but also on how outcomes, analysis populations, and decision rules were specified and implemented.

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.

The registry summary does not provide a complete statistical analysis plan. The original sample-size justification, any planned interim analyses, and the approach to multiplicity are not reported in the registry record. As a result, the posted secondary comparisons cannot be placed within a documented hierarchy of confirmatory tests.

The primary analysis description identifies an intent-to-treat population but also notes that 2 participants were excluded: 1 randomized in error and 1 with missing source data. The reasons and analytic consequences of these exclusions are not further detailed in the registry summary. Additional missing-data procedures, censoring rules, and sensitivity analyses are not reported in the registry record. These omissions constrain assessment of how robust the estimates are to alternative analytic choices.

The PFS result is summarized using medians and a hazard ratio. Further information about the survival curves, assessment frequency, proportional-hazards assessment, or the handling of possible departures from model assumptions is not reported in the registry record. The hazard ratio is therefore best interpreted as the reported comparative summary, not as a complete description of how the relative event rate may have varied during follow-up.

The posted secondary results are incomplete in the information described for this analysis. DCR has group estimates and an unadjusted recalculation, but no sponsor-reported comparison is listed. Other secondary endpoints are named without detailed results in the available record details. The registry also does not provide patient-level results, which limits independent reconstruction of analyses and assessment of subgroups.

Safety counts are descriptive. Details about severity grading, causality assessment, duration, exposure-adjusted rates, and comparative testing are not reported in the registry record. These limitations do not invalidate the posted summaries, but they restrict the conclusions that can be drawn from them.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study of Tarceva (Erlotinib) to Compare Two Different Doses in in Currently Smoking Patients With Advanced or Metastatic Non-Small Cell Lung Cancer (CURRENTS) — NCT01183858.
• Registry record last updated 2015-08-19; status last verified 2015-07.
• Results first posted 2015-02-23.
• 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

CURRENTS was a completed, randomized, double-blind, parallel-group phase III study sponsored by Hoffmann-La Roche. It compared daily oral erlotinib 150 mg with erlotinib 300 mg in current smokers with locally advanced or metastatic non-small cell lung cancer after chemotherapy failure. Treatment was described as continuing until disease progression.

The primary endpoint was PFS from randomization to progression or death. Median PFS was 6.86 weeks for the 150 mg group and 7.0 weeks for the 300 mg group. The sponsor-reported log-rank hazard ratio for 150 mg versus 300 mg was 1.05, with a 95% confidence interval from 0.83 to 1.33 and p = 0.671. This result does not establish a clear PFS difference or equivalence between doses.

Secondary measures covered survival, response, disease control, progression, and adverse events. The DCR comparison is an unadjusted recalculation; the registry does not report a multiplicity strategy or detailed missing-data procedures. These results should be read as estimates with uncertainty, within the population and analysis described in the registry record.