Linsitinib or Topotecan Hydrochloride in Treating Patients With Relapsed Small Cell Lung Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Linsitinib or Topotecan Hydrochloride in Treating Patients With Relapsed Small Cell Lung Cancer (NCT01533181), a randomized phase 2 study. The primary outcome is Median Progression Free Survival (PFS).

2. Clinical Question

This was a randomized, double-masked phase II study of treatments for relapsed small cell lung cancer. Its stated purpose was to evaluate how OS-906 compared with topotecan in slowing tumor growth or progression. The study also planned to assess side effects and explore whether proteins measured in blood or tumor, and imaging features from computed tomography scans, could help predict response. Results for those biomarker and imaging aims are not reported in the registry record.

The central statistical comparison is between the time-to-event experience in the two arms, not simply between two isolated median values. The primary result describes the distribution of time until progression or death among participants who received treatment. The posted hazard ratio gives a relative comparison over the observed follow-up, while the arm-specific medians describe the time at which the estimated progression-free survival curve reaches its midpoint.

3. Trial Design

The registry describes a randomized, double-masked phase II study with a crossover model. It was sponsored by the National Cancer Institute (NCI), began in 2012-02, reached primary completion in 2014-11, and is listed as completed. The total enrollment is recorded as 44. The record does not report the method used to generate the random allocation sequence, the allocation concealment procedures, or the practical details of masking.

Participants were assigned to topotecan or OS-906, also called linsitinib. The topotecan arm allowed intravenous administration over 30 minutes or oral dosing once daily on days 1–5. The OS-906 arm used daily continuous dosing in three-week cycles. Participants receiving topotecan could cross over to Arm I at progressive disease. The registry does not report how many participants crossed over, or provide enough detail to characterize post-crossover treatment exposure.

The primary endpoint was median progression-free survival, measured from randomization to progression or death, and assessed with Kaplan–Meier methods. The endpoint description gives a timeframe of up to 6 months and defines the analysis population as all participants who received treatment. The posted primary arm results list 14 participants for topotecan and 28 for OS-906. The registry does not explain the difference between total enrollment and the arm-specific result counts, or report participant flow and reasons for any exclusions.

Overall survival was listed as a secondary outcome, with a timeframe of up to 2 years. Other listed secondary outcomes included disease control rate and serious adverse events possibly, probably, or definitely related to study drugs. The registry also lists laboratory biomarker analysis and pharmacological study interventions. Detailed statistical methods for the secondary outcomes, biomarker objectives, and safety summaries are not reported in the registry record.

The original protocol’s sample-size justification, intended allocation ratio, planned interim analyses, and prespecified statistical analysis plan are not reported in the registry record. Randomization and masking are important design features, but the available description does not establish how well those procedures were implemented or whether the planned analyses were followed in full.

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: All participants who received treatment

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: All participants

6. Primary Endpoint

The registry lists the primary outcome as Median Progression Free Survival (PFS), measured over Up to 6 months.

Progression-free survival is the time from randomization until disease progression or death. It combines these events into a time-to-event outcome, so participants who have not experienced either event by the end of their observed follow-up can still contribute information up to that point. The registry specifies that the primary endpoint was summarized using the Kaplan–Meier method, with median progression-free survival and confidence intervals for the median and rates at different time points constructed when appropriate.

The median is the time at which the estimated proportion remaining progression-free reaches one-half. It is a compact description of the survival curve, but it does not describe every feature of that curve. Two arms can have similar medians and different patterns at other times, or different medians while their curves overlap during part of follow-up. Reading the medians alongside the hazard ratio and confidence intervals therefore gives a more complete account than treating either median as the full result.

In general, a time-to-event endpoint can use information from participants with different follow-up durations, rather than reducing each participant’s experience to a single fixed-time status. That advantage depends on appropriate definitions of progression, death, censoring, and follow-up. The registry gives the endpoint definition and method, but does not report the assessment schedule, censoring rules, or details of progression ascertainment.

The primary endpoint concerns disease progression or death during the analyzed period. It does not by itself answer questions about overall survival, quality of life, or the effects of treatment after progression. Those questions require separate outcomes and analyses.

7. Results Posted in the Registry

• Arm A: Topotecan vs Arm B: OS-906: hazard ratio 3.90 (95% CI 1.90–8.10), reported p < 0.001 (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 (OS)

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor (Kaplan-Meier); it was not recalculated.
• Result:
  • Arm A: Topotecan vs Arm B: OS-906: hazard ratio 1.10 (95% CI 0.60–2.20), reported p = 0.710 (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 Arm A: Topotecan vs Arm B: OS-906 and the control. The interval gives the range of effects that remain compatible with the data.

Overall survival is the secondary time-to-event outcome with quantitative arm-level results and a reported hazard-ratio comparison. Its estimate is less conclusive than the primary result: the confidence interval includes the no-difference value, and the reported p-value is 0.71. This does not establish that the treatments have equivalent overall survival; it means the posted analysis does not rule out a range of relative effects.

The registry also lists disease control rate and serious adverse events possibly, probably, or definitely related to study drugs as secondary outcomes. Detailed disease-control results and the methods used to analyze that outcome are not reported in the registry record. Serious adverse event counts are available, but their interpretation is descriptive and is addressed separately. The study’s biomarker and imaging aims are also described, but predictive findings are not reported in the registry record.

Secondary outcomes can add context to a primary endpoint, but they answer distinct questions and should not be treated as interchangeable confirmations. Overall survival captures time to death, whereas progression-free survival includes progression or death. Crossover after progressive disease may also complicate the interpretation of later outcomes; crossover counts and its effect on the overall-survival analysis are not reported.

In general, when many outcomes or comparisons are examined, the chance of finding at least one apparently notable result can increase unless the analysis plan accounts for multiplicity. The registry does not report a multiplicity adjustment or a hierarchy for testing secondary outcomes. As a result, secondary findings should be interpreted with more caution than a prespecified primary result, especially when the reporting is incomplete. The registry’s listed outcomes do not reveal how many analyses were planned, which were considered confirmatory, or whether results were selected for posting.

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

• Disease Control Rate (DCR)
• Incidence of Serious Adverse Events (SAEs) Possibly/Probably Definitely Related to Study Drugs
• Overall Survival (OS)

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is a time-to-event measure, so its analysis differs from a comparison of simple averages. Each participant contributes information about how long they remained free of progression or death, and participants may have different lengths of observation. The registry says the Kaplan–Meier method was used to summarize progression-free survival by arm. The method and its assumptions are described as general statistical background; registry-specific details beyond those explicitly reported are not available.

What does Kaplan–Meier estimation do?

In general, Kaplan–Meier estimation constructs a step-shaped estimate of the probability of remaining event-free over time. At each observed event time, the estimate accounts for the participants still being followed and at risk at that time. Participants who have not experienced an event when their observation ends are censored: their known follow-up contributes information up to that point, but their later event time is unknown. This approach can accommodate differing follow-up durations, provided the censoring assumptions and event ascertainment are appropriate.

The registry defines progression-free survival as time from randomization to disease progression or death, and says the method was applied by treatment arm. It does not report censoring rules, how progression was assessed, whether outcome assessment was centrally reviewed, or how deaths and missing assessments were handled. Those details affect how the estimated curves should be understood, and cannot be inferred from the method name alone.

What does the median describe?

In general, the median time-to-event value is the time at which the estimated event-free probability reaches one-half. It condenses an entire estimated curve into one location. The median does not describe the shape of the curve before or after that point, and it is not an estimate of the time every participant will remain event-free. The confidence interval around a median expresses uncertainty in its location.

The posted results give the arm-specific medians and 95% confidence intervals. The registry also says intervals for medians and event-free rates at different time points were to be constructed when appropriate. It does not report which additional time-point rates were available or the specific method used to construct the confidence intervals.

How should the hazard ratio be read?

In general, a hazard is the instantaneous rate of an event among participants who have not yet experienced it, conditional on remaining at risk up to that time. A hazard ratio compares the hazards between groups. It is not a comparison of the number of events, the event probabilities at a fixed time, or the medians themselves. The reported comparison is labeled topotecan versus OS-906, and the registry gives a hazard ratio of 3.9 with a 95% confidence interval from 1.9 to 8.1.

A single hazard ratio is easiest to interpret when the relative hazards are reasonably stable over time. If the relative hazards change, one summary value may not represent the differences across the full follow-up period well. The registry identifies the Kaplan–Meier method and reports the hazard ratio, but does not state whether proportional hazards were assessed, how the hazard ratio was estimated, or whether alternative models were considered. Those aspects of the statistical plan are not reported in the registry record.

What does a confidence interval contribute?

In general, a confidence interval gives a range of values compatible with the data under the model and procedure used. A 95% confidence interval is produced by a method that, under its assumptions, would cover the true parameter in 95% of repeated samples. It is not a statement that there is a 95% probability that the fixed true value lies inside this particular interval. For this trial, the interval around the hazard ratio shows a range of plausible relative hazard values, not a range of participant outcomes.

A confidence interval also helps distinguish the precision of an estimate from its direction. An interval can be narrow or wide depending on the amount of information and the variability of the data. A statistically notable result can still have uncertainty about its magnitude. Conversely, an interval that includes the no-difference value does not prove equivalence; it indicates that the data do not rule out that value at the interval’s confidence level under the method’s assumptions.

Why might over-dispersion matter?

In general, over-dispersion occurs when the variability in observed data is greater than a statistical model assumes. If the model understates variability, standard errors can be too small and confidence intervals too narrow, making statistical evidence appear stronger than it is. This concern often arises in count outcomes or clustered data, rather than being a feature that can be diagnosed from a reported time-to-event summary alone. The registry does not report an over-dispersion assessment, and the posted primary endpoint is not a simple count outcome.

Other assumptions also matter. In general, informative censoring occurs when the chance that follow-up ends is related to the participant’s underlying event risk in a way not handled by the analysis. Unequal follow-up or missing outcome information can then affect estimates. The registry record does not report enough information to evaluate censoring or missing-data assumptions for this study.

What can randomization and masking add?

In general, randomization is intended to make treatment groups comparable on average with respect to measured and unmeasured baseline characteristics. Masking is intended to reduce the influence of knowledge of treatment assignment on care, assessment, or reporting. The registry labels the study randomized and double-masked, but does not describe sequence generation, concealment, who was masked, or whether masking was maintained. These labels identify design features; they do not by themselves confirm that bias was eliminated.

The registry does not report the original sample-size justification, a prespecified statistical analysis plan, interim analyses, multiplicity procedures, or detailed analysis-population definitions. The planning illustrations are based on the observed effect and are not the trial’s own design assumptions. They should not be treated as evidence that the trial had a particular planned power or that the observed result has a particular probability of replication.

11. Confidence Intervals and Precision

Precision concerns how tightly an estimate is determined by the data and analysis. The primary median estimates are accompanied by 95% confidence intervals: 3.0 months for topotecan, with an interval from 1.5 to 3.6 months, and 1.2 months for OS-906, with an interval from 1.1 to 1.4 months. The hazard ratio is 3.9, with an interval from 1.9 to 8.1. The interval for the hazard ratio is broad enough to leave substantial uncertainty about the exact size of the relative comparison, even though it does not include the no-difference value of 1.

In general, a confidence interval summarizes uncertainty under the statistical method used. Its width reflects the information in the data as well as variability and modeling assumptions. It is not a prediction interval for individual participants, and it does not account for every possible source of bias, such as unreported missing-data procedures or departures from the analysis plan.

The primary arm counts are 14 for topotecan and 28 for OS-906, while total enrollment is 44. The registry does not explain the difference between these counts and total enrollment. The relatively limited arm-specific information is relevant when judging how stable the estimates may be, but the confidence intervals are the direct posted expression of statistical uncertainty and should not be replaced by informal calculations.

The p-value is reported as < 0.001 for the primary hazard-ratio comparison. A small p-value is not a measure of precision or effect size. The confidence interval is more useful for seeing the range of estimates compatible with the data, while the p-value addresses compatibility with a particular null hypothesis under the test assumptions. Neither resolves limitations in the registry’s reporting of follow-up, censoring, or participant flow.

For overall survival, the reported hazard-ratio interval extends from 0.6 to 2.2. In general, an interval that includes 1 is compatible with both lower and higher hazards in the first-named group relative to the second under the reported contrast. The interval therefore does not establish that the groups are equivalent. Its width also indicates that the OS estimate is not tightly determined by the posted data.

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 scenarios based on the observed effect, not the original design calculation. The assumptions listed for these illustrations are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; the effect to detect is a stated fraction of the observed effect. The original trial’s design assumptions and sample-size justification are not reported in the registry record.

For time-to-event outcomes, the figures are the total number of events needed across the study, not participants per arm. For other outcome types, the figures are participants per arm. The primary and overall-survival illustrations are time-to-event calculations, so their event totals should not be mistaken for enrollment targets or for numbers assigned to each treatment group.

The 80% power column represents a planning scenario in which a study with the stated assumptions would have an 80% chance of detecting the specified effect at the chosen significance level, if the planning model and effect were correct. The 90% power column uses a higher target chance of detection under the same kind of planning framework. These are design concepts, not probabilities that the observed trial result is correct.

The rows for 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect show how the planning totals change when the target effect is made smaller. Smaller effects generally require more information to detect, so the illustrations call for more events as the assumed effect moves farther below the observed estimate. This is not observed power: it does not describe the probability that the completed trial detected its actual effect, and it does not establish that the original study was planned to enroll the illustrated numbers.

These scenarios are particularly sensitive to the observed effect used as their starting point. An estimate from a small study can be uncertain, and planning directly from an observed effect does not remove that uncertainty. The event totals are therefore best read as conditional illustrations under the stated assumptions, rather than as recommendations or as evidence about what sample size the investigators originally intended.

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 = 3.900, so ln(HR) = 1.3610.
2. d = 4 × (1.96 + 0.842)² / [ln(3.900)]² = 16.9, rounded up to 17 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 <- 3.9                              # 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(1533181)
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 = 17 (80% power), 23 (90% power)
 75% of observed effect: events = 31 (80% power), 41 (90% power)
 50% of observed effect: events = 68 (80% power), 91 (90% power)
Simulated power with 17 events: 0.747 (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 registry reports the sponsor’s Kaplan–Meier hazard-ratio analysis for the primary endpoint as 3.9, with a 95% confidence interval from 1.9 to 8.1 and p < 0.001. The posted primary comparison carries the same estimate, confidence interval, and p-value. The record identifies no discrepancy between the reported analysis and the displayed result.

For the secondary overall-survival comparison, the registry reports a hazard ratio of 1.1, with a 95% confidence interval from 0.6 to 2.2 and p = 0.71. The posted secondary result likewise gives those values. The sponsor’s analysis is the relevant reported inferential result; a simple recalculation from summary statistics would not replace it, particularly for a time-to-event analysis whose underlying event and censoring data are not available in the registry record.

In general, a recalculation from summary statistics can be useful as a consistency check, but it may not reproduce the sponsor’s model or analysis choices. A hazard ratio depends on time-to-event information and the method used to estimate it, not only on the arm-specific medians. Since the registry reports the sponsor’s results without a discrepancy, interpretation should stay anchored to those reported analyses rather than imply that a separate calculation provides a superior estimate.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when several outcomes, time points, subgroups, or statistical comparisons are tested. In general, conducting many tests can increase the chance of at least one apparently unusual result even when there is no true effect for any of them. Statistical plans can address this through prespecified testing order or adjustments to significance thresholds, among other approaches. The registry record does not report a multiplicity strategy, a testing hierarchy, or which analyses were confirmatory beyond identifying the primary outcome.

The registry lists five outcome measures and reports two analyses, but it does not describe the full set of planned comparisons or whether any outcomes were omitted from the posted results. The primary endpoint is identified as median progression-free survival. The stated secondary outcomes include disease control rate, serious adverse events, and overall survival, while biomarker and imaging objectives are described in the study summary. The detailed analysis plans for these outcomes are not reported in the registry record.

Missing-data and censoring procedures are also important for time-to-event outcomes. The registry defines progression-free survival and states that Kaplan–Meier methods were used, but does not report the censoring rules, assessment schedule, or how missing progression assessments were handled. The overall-survival record likewise does not give detailed rules for follow-up and censoring. These omissions do not show that the analyses were incorrect, but they prevent a reader from fully evaluating their assumptions.

The primary endpoint description refers to all participants who received treatment. The registry does not report a complete analysis-population definition, participant flow, or reasons why the primary results list 14 participants for topotecan and 28 for OS-906 while overall enrollment is 44. It is therefore not possible to determine from the record whether an intention-to-treat analysis was performed or how randomized participants who did not receive treatment were handled.

Interim analyses, stopping rules, and the original statistical analysis plan are not reported in the registry record. This limits assessment of whether the posted analyses match prespecified methods and whether decisions during the trial could have affected the interpretation of the p-values or confidence intervals. These are uncertainties in reporting, not evidence that a particular procedure was used or omitted in practice.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: death nos (50.0%), neoplasms benign, malignant and unspecified - other, death (21.4%), anemia (14.3%), white blood cell decreased (14.3%), dehydration (14.3%).

The registry reports serious adverse events affecting 7 of 14 participants at risk in the topotecan arm and 18 of 28 at risk in the OS-906 arm. It also lists other adverse events affecting 12 of 14 participants at risk in the topotecan arm and 28 of 28 at risk in the OS-906 arm. These are descriptive counts from the posted safety record. The record does not provide enough detail to determine event timing, duration, recurrence, or how adverse events were attributed beyond the listed relatedness language for the secondary outcome.

Several serious-event terms are listed with percentages, including “Death NOS” at 50.0%, “Neoplasms benign, malignant and unspecified - Other, Death” at 21.4%, anemia at 14.3%, white blood cell decreased at 14.3%, dehydration at 14.3%, and lung infection at 10.7%. The registry does not clarify the denominator or whether these percentages refer to participants, events, or another reporting basis. They should therefore be read as registry-listed term percentages, not converted into participant-level risks.

Safety counts from a small trial can describe what was recorded in these groups, but they do not by themselves establish comparative safety or causation. The registry’s adverse-event definitions, collection procedures, and detailed exposure-adjusted analyses are not reported. These figures should be considered alongside the limits of the posted safety information rather than interpreted as a complete account of risk.

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

The primary result is a registry-reported comparison of progression-free survival in participants with relapsed or recurrent small cell lung cancer who received treatment in this randomized phase II study. The reported medians and hazard ratio indicate a difference in the observed time-to-event summaries between the arms. The confidence interval around the hazard ratio gives a range of relative effects compatible with the analysis, and the small reported p-value indicates that the result is not readily compatible with a no-difference null under the test assumptions.

That finding is not a guarantee about an individual participant’s outcome and does not, on its own, answer every question about treatment benefit or harm. Progression-free survival includes progression or death; it is not the same outcome as overall survival. The secondary overall-survival estimate is uncertain, and its confidence interval includes the no-difference value. The record also does not report full results for disease control rate or the biomarker and imaging objectives.

Interpretation is bounded by the trial’s scale and reporting. The registry lists 44 participants enrolled, with primary arm results for 14 and 28 participants. The reason for the difference is not reported in the registry record. The record does not provide the full analysis plan, participant flow, censoring rules, details of missing data, or assessment procedures. These details are relevant to judging how robustly the posted estimates represent the randomized comparison.

The crossover option allows participants in the topotecan arm to receive OS-906 after progressive disease. This is part of the design, but the registry does not say how many participants crossed over or detail its implications for the secondary outcomes. Since progression is itself part of the primary endpoint, crossover after progression does not change the definition of the primary event; it may be relevant to interpreting later outcomes.

The result should not be generalized beyond the studied setting without additional evidence. It is an estimate from one completed phase II registry record, with uncertainty in both effect size and reporting detail. Statistical evidence for a difference is not the same as evidence about clinical importance, and the registry does not provide a threshold for what magnitude of difference would be clinically meaningful.

18. Why This Trial Matters Statistically

This trial illustrates why a time-to-event result should be read through several complementary summaries. The medians describe the center of each estimated progression-free survival distribution, while the hazard ratio compares event hazards across follow-up. Confidence intervals show the uncertainty around those estimates. No single statistic fully captures the timing and pattern of events, so the reported measures need to be interpreted together.

The result also illustrates the distinction between statistical evidence and precision. A reported p-value of < 0.001 accompanies a hazard-ratio estimate of 3.9, but the 95% confidence interval from 1.9 to 8.1 shows that the magnitude remains uncertain. The p-value does not convey how large or clinically important the effect is, and the confidence interval does not remove uncertainty about study conduct or unreported analysis decisions.

The randomized and double-masked design labels signal efforts to reduce allocation and assessment biases, while the crossover option highlights how treatment after progression can matter for later outcomes. The registry does not report procedural details sufficient to judge implementation, and the crossover count is not reported. This makes the distinction between a design feature and evidence that the feature worked as intended important.

Finally, the planning illustrations show how required event totals depend on the effect chosen for planning and the target power. They are not the original sample-size calculation and are not observed power. Separating planning scenarios from the trial’s own design assumptions helps prevent an attractive observed estimate from being treated as a validated planning truth. The record’s incomplete reporting on enrollment flow, multiplicity, censoring, and analysis populations reinforces the need to communicate statistical results with their limits.

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 record does not explain why total enrollment of 44 differs from the arm-specific primary result counts of 14 and 28. It does not give a participant-flow account, the number randomized but untreated, reasons for exclusions, or a detailed definition of the analysis population beyond participants who received treatment. This makes it difficult to assess whether the reported analysis preserves the randomized comparison for all enrolled participants.

Methods that affect interpretation of the primary time-to-event analysis are incompletely described. The registry states that Kaplan–Meier methods were used, but does not report censoring rules, the progression-assessment schedule, missing-assessment handling, or procedures for adjudicating progression. It also does not state whether proportional hazards were examined or how the hazard ratio was estimated. These details are not reported in the registry record.

The study allowed crossover from topotecan to OS-906 at progressive disease, but the number of participants who crossed over and the analysis implications are not reported. Crossover is especially relevant when interpreting overall survival and other outcomes measured after progression. The registry also does not report the original sample-size rationale, interim analyses, stopping rules, or a complete prespecified analysis plan.

The available secondary reporting is incomplete. Overall survival has a posted time-to-event result, but disease control rate and the planned biomarker and imaging investigations do not have detailed results in the registry record. The safety percentages for listed serious-event terms lack an explained denominator, and detailed event-collection and attribution methods are not reported. Together, these limits restrict assessment of how comprehensive the posted results are and how they relate to all planned study aims.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Linsitinib or Topotecan Hydrochloride in Treating Patients With Relapsed Small Cell Lung Cancer — NCT01533181.
• Registry record last updated 2016-01-14; status last verified 2015-08.
• Results first posted 2016-01-14.
• 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

NCT01533181 is a completed randomized, double-masked phase II study of OS-906 (linsitinib) versus topotecan hydrochloride for relapsed or recurrent small cell lung cancer. The registry lists 44 enrolled participants and describes crossover from topotecan to OS-906 after progressive disease.

The primary endpoint was progression-free survival, defined from randomization to progression or death and summarized using Kaplan–Meier methods. The posted comparison reports a hazard ratio of 3.9, with a 95% confidence interval from 1.9 to 8.1 and p < 0.001. Overall survival was a secondary endpoint, with a reported hazard ratio of 1.1 and a 95% confidence interval from 0.6 to 2.2.

The registry does not report the original sample-size justification, a full statistical analysis plan, detailed participant flow, or key censoring and missing-data procedures. The planning illustrations are conditional scenarios based on the observed effect, not the trial’s own design calculation.