Testing the Addition of a New Immunotherapy Drug, Atezolizumab (MPDL3280A), to the Usual Chemoradiation (CRT) Therapy Treatment for Limited Stage Small Cell Lung Cancer (LS-SCLC): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Testing the Addition of a New Immunotherapy Drug, Atezolizumab (MPDL3280A), to the Usual Chemoradiation (CRT) Therapy Treatment for Limited Stage Small Cell Lung Cancer (LS-SCLC) (NCT03811002), a randomized phase 3 study. The primary outcome is Overall Survival.

2. Clinical Question

3. Trial Design

This is a phase III, randomized, parallel-group trial with no masking reported. It compares chemoradiation alone with chemoradiation plus atezolizumab in people with limited stage small cell lung cancer. The registry lists enrollment of 544. The sponsor is the National Cancer Institute. The recorded start date is 2019-07-26, and the primary completion date is 2024-04-21. The current registry status is active, not recruiting.

Participants in both groups receive etoposide and either cisplatin or carboplatin, along with radiation therapy. The registry lists 3-D conformal radiation therapy and intensity-modulated radiation therapy as radiation options. In the added-atezolizumab group, atezolizumab is given during chemotherapy cycles and may continue for 17 cycles, described as 1 year, in the absence of disease progression or unacceptable toxicity. The comparator group does not receive atezolizumab. Details beyond the listed schedules and conditions are not reported in the registry record.

The primary endpoint is overall survival, assessed from randomization to death or last follow-up. The registry says survival rates and median survival are estimated using the Kaplan-Meier method, with participants alive at analysis censored. Median follow-up at the analysis was 23.8 months. The intent-to-treat population is specified for the primary analysis. The record also lists secondary outcomes related to progression, response, adverse events, quality of life, and fatigue. The protocol’s original sample-size justification, allocation ratio, and planned interim analyses are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Intent-to-treat population

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: Intent-to-treat population

6. Primary Endpoint

The registry lists the primary outcome as Overall Survival, measured over Randomization to date of death due or last follow-up. Median follow-up at the time of analysis was 23.8 months..

Overall survival measures the time from randomization until death from any cause, with follow-up ending at death or the last follow-up. The registry describes the endpoint timeframe as randomization to date of death or last follow-up. Participants who are alive at the analysis are censored in the Kaplan-Meier estimation: their observed follow-up contributes information up to the point they were last known alive, while their later survival time is unknown.

Time-to-event outcomes are useful when the timing of an event matters and participants may have different lengths of follow-up. A median survival estimate identifies the time at which the estimated survival curve reaches the halfway point. It is a summary of the survival distribution, not an average survival time and not a prediction for any one participant. The registry also reports a hazard ratio, which summarizes the relative event rate over follow-up under the analysis model. It should be read alongside its confidence interval and the reported statistical test, rather than in isolation.

Overall survival is a direct endpoint because it concerns death rather than an intermediate measure such as tumor response. That makes it relevant to evaluating the trial question, while also making follow-up duration and censoring important to interpretation. Median follow-up at analysis was 23.8 months; the duration of follow-up varies between participants and does not mean that everyone was observed for that length of time.

7. Results Posted in the Registry

• Arm I (Chemotherapy, Radiation Therapy) vs Arm II (Chemotherapy, Radiation Therapy, Atezolizumab): hazard ratio 1.05 (95% CI 0.82–1.34), reported p = 0.582 (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.

Progression Free Survival (PFS)

• 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:
  • Arm I (Chemotherapy, Radiation Therapy) vs Arm II (Chemotherapy, Radiation Therapy, Atezolizumab): hazard ratio 0.98 (95% CI 0.80–1.21) (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 I (Chemotherapy, Radiation Therapy) vs Arm II (Chemotherapy, Radiation Therapy, Atezolizumab) and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Participants With Confirmed Objective Response (Objective Response Rate (ORR))

• 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:
  • Arm II (Chemotherapy, Radiation Therapy, Atezolizumab): 60.2% (165/274) versus 59.6% (161/270) on control; risk difference 0.6 percentage points (95% CI -7.6 to 8.8), risk ratio 1.01 (0.88–1.16); unadjusted two-proportion z-test p = 0.888.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Arm II (Chemotherapy, Radiation Therapy, Atezolizumab) and the control. The interval gives the range of effects that remain compatible with the data.

Distant Metastases-free Survival (DMFS)

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor; it was not recalculated.
• Result:
  • Arm I (Chemotherapy, Radiation Therapy) vs Arm II (Chemotherapy, Radiation Therapy, Atezolizumab): hazard ratio 0.95 (95% CI 0.76–1.20) (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 I (Chemotherapy, Radiation Therapy) vs Arm II (Chemotherapy, Radiation Therapy, Atezolizumab) and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes cover several kinds of evidence: progression-free survival and distant metastases-free survival are time-to-event measures; objective response is a binary tumor-response measure; and other outcomes include adverse events, quality of life, and fatigue. The posted time-to-event estimates and the response percentages provide additional context around the overall-survival result, but they do not replace the primary endpoint. Their pattern is mixed rather than a single consistent signal across every measure.

In broad terms, the secondary time-to-event and response comparisons do not provide an unequivocal resolution of the primary result. The progression-free survival hazard-ratio interval includes its null value, as does the distant metastases-free survival hazard-ratio interval. The response percentages are close, and the unadjusted recalculation does not show a statistically clear difference. These measures have distinct definitions and follow-up requirements, so agreement or disagreement among them should not be treated as proof about overall survival.

Multiple secondary outcomes increase the number of opportunities to observe a result that appears notable by chance. Unless a plan controls that multiplicity, individual unadjusted tests are less persuasive than a prespecified primary comparison. The registry record does not report a multiplicity adjustment plan. Nor does it report a complete hierarchy or confirmatory testing strategy for the secondary outcomes. Their results are therefore best understood as supportive or exploratory context, rather than independent proof of a treatment effect.

The registry lists additional outcomes concerning adverse events and patient-reported measures, including quality of life and fatigue. Their presence broadens the description of treatment experience, but outcome definitions, missing-data handling, and analysis details are not reported for all measures in the record. A cautious reading keeps each secondary result tied to its own endpoint and avoids selecting only the most favorable-looking result.

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

• Progression Free Survival (PFS)
• Number of Participants by Highest Grade Adverse Event Reported
• Percentage of Participants With Confirmed Objective Response (Objective Response Rate (ORR))
• Percentage of Participants With Local Progression
• Distant Metastases-free Survival (DMFS)
• Percent of Participants With Deterioration in Functional Assessment of Cancer Therapy-Trial Outcome Index (FACT-TOI)
• Quality-adjusted Survival Months
• Change From Baseline in the 7 Item Patient Reported Outcomes Measurement Information System (PROMIS) Fatigue Short Form Score

9. Statistical Interpretation

10. Statistical Methods Explained

How does Kaplan-Meier estimation work?

Kaplan-Meier estimation is a general method for describing time-to-event outcomes when follow-up times differ and some people have not experienced the event by the time of analysis. Each observed event contributes information to the estimated survival curve. A participant who is alive at last follow-up is censored at that time: the analysis uses the period during which their status is known without assuming what happened afterward. The registry states that the primary survival rates and median are estimated this way, censoring participants alive at analysis.

The median is read from the estimated curve at the point where the estimated proportion surviving reaches one half. It condenses a curve into a single time summary, but it does not show the curve’s full shape. Two groups can have different median estimates even when the uncertainty around the estimates is substantial. Conversely, similar medians do not guarantee that the curves are alike at every point in follow-up.

What does the hazard ratio summarize?

A hazard is the instantaneous event rate among participants who have not yet had the event, conditional on remaining at risk. A hazard ratio compares those rates between groups under the analysis definition. In general, a hazard ratio of 1 corresponds to equal hazards, while values on either side indicate a relative difference in the comparison’s defined direction. The registry reports a log-rank analysis and a hazard ratio for overall survival. The specific model details and direction coding are not reported in the registry record, so interpretation should follow the comparison order and label as posted.

A hazard ratio is not the ratio of median survival times, nor does it directly state how many months an individual may gain or lose. Its summary is most straightforward when the relative hazards are reasonably stable over time. The registry does not report an assessment of that assumption. The log-rank test compares the observed event experience across groups over follow-up; its p-value is interpreted with the effect estimate and confidence interval, not as a measure of the size or importance of an effect.

What does a confidence interval add?

In general, a confidence interval describes the statistical uncertainty around an estimate under the method used. A 95% confidence interval is produced by a procedure that, under its assumptions, would cover the underlying 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. Wider intervals indicate less precision than narrower intervals, though precision alone does not establish clinical importance.

For a hazard ratio, the value 1 is the null comparison. An interval that includes 1 is compatible with equal hazards as well as with values on either side, at the interval’s stated confidence level. For medians, intervals describe uncertainty in each group’s estimated median; overlapping intervals are not themselves a formal test of the between-group contrast. The reported log-rank test is the registry’s stated test for the primary comparison.

Why can event counts matter?

In general, time-to-event analyses gain information from observed events, not simply from enrolling more participants who have not yet had the event. Participants with shorter follow-up contribute less information about later event patterns than participants followed longer, though their observed time is still used. Censoring is therefore part of the analysis rather than a reason to discard every participant who is alive at analysis. The registry reports enrollment and median follow-up, but does not provide a detailed accounting of event counts or numbers at risk over time in the registry record for this analysis.

Over-dispersion is a general issue for count outcomes when the observed variability exceeds what a simple Poisson model assumes; it can make standard errors and confidence intervals too small if ignored. The primary endpoint is time-to-event, not a count outcome, and the registry does not report a Poisson model for it. The relevant general concerns for this survival result are censoring, follow-up, the hazard-ratio interpretation, and uncertainty in the estimates.

How should the objective-response comparison be read?

Objective response is reported as a percentage of participants with confirmed response. It is a binary outcome at the participant level: each participant either meets the response definition or does not. A risk difference expresses the difference between the two groups’ response proportions, while a risk ratio compares the proportions as a ratio. The registry record posts group response estimates and confidence intervals, but does not report a sponsor analysis for this endpoint. The unadjusted recalculation is a simple comparison based on the posted summary statistics and should not be mistaken for a protocol-specified adjusted model.

The reported p-value for that recalculation is unadjusted. In general, an unadjusted test does not account for examining multiple outcomes or multiple comparisons. A confidence interval around a binary measure conveys uncertainty in the estimated proportion or contrast, but does not resolve concerns about selection, missing outcome assessments, or multiplicity. The registry does not report enough detail to assess all of those issues.

Which details are not documented?

The registry record identifies the primary population, the Kaplan-Meier method, censoring of participants alive at analysis, and the log-rank comparison. A full statistical analysis plan, the original sample-size justification, the handling of missing data beyond censoring, and the assessment of model assumptions are not reported in the registry record. Those absences limit the ability to evaluate whether every analysis choice was prespecified or how alternative methods might affect the estimates.

11. Confidence Intervals and Precision

The primary hazard-ratio estimate is 1.05, with a 95% confidence interval from 0.82 to 1.34. This interval includes 1, the null value for a hazard ratio, and spans values on both sides of it. The estimate is therefore not precise enough to distinguish clearly between equal hazards and a range of relative hazard differences under the reported analysis. The p-value of 0.582 is consistent with that uncertainty; it should not be read as a measure of how large or important the effect is.

The median overall survival estimates also have uncertainty. The chemoradiation group’s estimate is 36.1 months, with a 95% confidence interval from 28.1 to 42.5 months. The group receiving atezolizumab has an estimate of 31.1 months, with a 95% confidence interval from 28.5 to 44.7 months. These are group-specific intervals. Their overlap does not substitute for a confidence interval or test of the difference between medians.

Precision depends on the number and timing of observed events, how long participants are followed, and how much information remains at later follow-up. The median follow-up was 23.8 months, and participants alive at analysis were censored. The registry does not report numbers at risk over time or a separate confidence interval for the difference in medians. The result should therefore be described through the reported estimates and their uncertainty rather than by treating a single point estimate as definitive.

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 effect, not a reconstruction of the trial’s original design. The registry’s original sample-size justification and design assumptions are not reported in the registry record. The illustrations use a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. For each scenario, the effect to detect is the stated fraction of the observed effect.

For time-to-event outcomes, the figures represent the total number of events, calculated using the Schoenfeld approach. They are not participants per arm. For other outcomes, the figures represent participants per arm. The distinction matters because time-to-event information depends substantially on events occurring during follow-up, whereas a binary outcome planning illustration is expressed as the number of participants in each arm.

The 80% power column describes the event or participant count that, under the stated planning assumptions and effect scenario, would give an 80% chance of detecting an effect at the specified significance level if that effect were present. The 90% power column describes the corresponding count for a 90% chance. These are planning concepts, not guarantees: actual power and results depend on the true effect, follow-up, event occurrence, missingness, and analysis conditions.

Rows planning for a smaller fraction of the observed effect require more events or participants than rows planning for the full observed effect. This is a general feature of power calculations: smaller effects are harder to distinguish from random variation and require more information. These figures are not observed power.

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(3811002)
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: 0.796 (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 primary overall-survival analysis as a log-rank comparison with a hazard ratio of 1.05, a 95% confidence interval from 0.82 to 1.34, and p = 0.582. The primary result is therefore available as a reported analysis, rather than being inferred solely from the group medians. Its confidence interval includes the null hazard ratio, and the reported p-value does not indicate a statistically clear difference under the stated test.

The record does not flag a discrepancy between the sponsor result and the simple recalculations. For time-to-event outcomes, the posted hazard ratios are the sponsor-reported estimates; they are not independently recalculated from the displayed medians. A median alone does not contain enough information to reconstruct the log-rank test or a hazard ratio. The sponsor’s stated analysis is the appropriate reference for that comparison.

For objective response, the registry does not list a sponsor-reported inferential analysis. The unadjusted recalculation from the posted summary statistics gives p = 0.888. That calculation is a simple, unadjusted comparison, not a substitute for a prespecified sponsor analysis or an adjustment for multiple outcomes. The distinction between reported analyses and recalculations is important: they may answer related questions, but they do not necessarily use the same methods or carry the same evidential status.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when several outcomes, time points, subgroups, or comparisons are tested. Even if each test uses a conventional significance threshold, conducting many tests increases the chance that at least one result will appear statistically notable by chance. A prespecified strategy can address this by defining which outcome is primary and how testing proceeds, or by adjusting the threshold or inference for a family of comparisons. The registry identifies overall survival as primary and lists multiple secondary outcomes, but a detailed multiplicity strategy is not reported in the registry record.

The record lists 15 outcome measures and 4 analyses. These counts do not by themselves establish how many hypothesis tests were planned, which tests were confirmatory, or whether adjustment was used. The primary log-rank result should be distinguished from unadjusted recalculations and from secondary comparisons. Without the full analysis plan, the status of every secondary finding within a testing hierarchy cannot be determined.

The primary analysis population is identified as intent-to-treat. In general, intent-to-treat analysis retains participants in the groups to which they were randomized, supporting a comparison based on randomized assignment. The registry does not report detailed information about exclusions, protocol deviations, or the handling of missing endpoint data beyond stating that living participants are censored in the survival estimation. It also does not report a missing-data analysis for the patient-reported outcomes.

For survival outcomes, censoring is not the same as treating an unknown event time as if no event occurred. Kaplan-Meier estimation uses known follow-up until censoring, but interpretation relies on assumptions about the relationship between censoring and future event risk. The registry record does not report diagnostics or sensitivity analyses for that assumption. Interim analyses, stopping boundaries, and the complete statistical analysis plan are also not reported in the registry record. Their absence from the record limits assessment of the full prespecification and analysis process.

16. Safety Results

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

Deaths

Other (non-serious) adverse events

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

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: febrile neutropenia (13.3%), neutrophil count decreased (10.9%), white blood cell decreased (10.1%), pneumonitis (9.1%), lung infection (7.2%).

The serious adverse event summaries report 106 affected participants among 257 at risk in the chemoradiation group, and 159 among 264 at risk in the chemoradiation-plus-atezolizumab group. The at-risk counts are the denominators shown for these summaries. These counts describe participants with serious adverse events in the posted safety data; they do not, by themselves, establish that a treatment caused an event or provide a complete comparison of event severity, duration, or exposure-adjusted rates.

The registry lists febrile neutropenia at 13.3%, neutrophil count decreased at 10.9%, white blood cell decreased at 10.1%, pneumonitis at 9.1%, lung infection at 7.2%, and platelet count decreased at 7.0% among the top serious-event terms. The record also reports 120 deaths in the chemoradiation group and 132 in the group receiving atezolizumab. Those death counts should not be assumed to represent deaths caused by adverse events; that attribution is not reported in the registry record.

Safety figures require attention to denominators and definitions. The registry does not report, in the registry record, the full event-level detail, time at risk, or a causal assessment for each event. These summaries are descriptive and do not alone quantify comparative risk.

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

The posted primary result describes overall survival in two randomized groups receiving chemoradiation, with atezolizumab added in one group. The estimated medians differ, while the reported hazard-ratio confidence interval includes 1 and the log-rank p-value is 0.582. Statistically, the analysis does not show a clear difference in overall survival under the reported comparison. The interval also does not establish that the treatment groups are equivalent or that any possible difference is zero.

A non-significant result is not proof of no effect. It can reflect a small effect, uncertainty, or limited information, and the confidence interval helps show which relative values remain compatible with the data under the method. Here the interval spans values on both sides of the null. It is therefore more accurate to describe the result as inconclusive about a clear relative difference than to claim that the groups are identical.

The results also do not justify a claim that atezolizumab works or does not work for an individual. Trial-level estimates summarize outcomes across participants and rely on the stated endpoint definitions, follow-up, and analysis. They do not predict what will happen to any particular person. The registry record covers a specific population, treatment comparison, and analysis; it does not establish that the result applies to other cancers, disease settings, regimens, or populations.

Secondary measures add context but have their own limitations and do not override the primary endpoint. Multiple outcomes, incomplete reporting of the analysis plan, and uncertainty around estimates all matter when weighing the full record. The appropriate interpretation is bounded: the registry reports no statistically clear overall-survival difference in this analysis, while leaving a range of possible relative effects compatible with the confidence interval.

18. Why This Trial Matters Statistically

This trial illustrates why a clinical result should not be reduced to a comparison of two medians. The primary analysis combines estimated survival over follow-up with a hazard ratio, confidence interval, and log-rank test. Those components answer related but distinct questions: the medians summarize each group’s estimated survival distribution, while the hazard ratio and test compare event experience across groups under the reported analysis.

It also demonstrates the importance of precision. A point estimate is only one location within a range of uncertainty. The primary hazard-ratio interval spans the null value, so the estimate does not sharply distinguish equal hazards from other relative values. The p-value alone cannot communicate that range, and a non-significant result should not be interpreted as proof of equivalence.

The trial also shows why secondary endpoints need disciplined interpretation. Survival, progression, response, safety, and patient-reported outcomes each capture different aspects of the trial. Looking across them can provide context, but multiple unadjusted tests may produce chance findings. A clear distinction between a prespecified primary analysis, sponsor-reported secondary analyses, and simple recalculations helps prevent overstatement.

Finally, the record makes transparent reporting important. It identifies the primary population and key survival methods, while details such as the original sample-size justification, missing-data approach, and multiplicity strategy are not reported in the registry record. Knowing which details are available—and which are not—helps readers assess what the results support without filling gaps through assumptions.

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 main limitation for interpreting the primary result is the uncertainty around the hazard-ratio estimate. Its 95% confidence interval includes 1 and covers relative values on both sides of the null. The registry reports median follow-up of 23.8 months, while survival estimates depend on the distribution of follow-up and censoring. Numbers at risk over time and a fuller accounting of events are not reported in the registry record, limiting assessment of precision late in follow-up.

The registry does not report the original sample-size justification, the complete statistical analysis plan, or whether interim analyses were conducted. It also does not report a detailed multiplicity strategy for the many secondary outcomes. These omissions make it difficult to assess the full prespecification of the analyses and how confirmatory each secondary result was intended to be.

The posted record identifies intent-to-treat as the primary analysis population, but detailed information about missing data and sensitivity analyses is not reported. This is particularly relevant to outcomes requiring tumor assessments or participant questionnaires, for which data availability can affect estimates. The record also does not report diagnostics for survival-model assumptions or a detailed explanation of hazard-ratio coding.

Safety summaries are limited by the detail reported. The serious adverse event counts are descriptive; exposure-adjusted event rates, full event-level data, and causal attribution are not reported in the registry record. Finally, the trial status is active, not recruiting, and the primary completion date is recorded as 2024-04-21. The results discussed are the posted analysis with median follow-up of 23.8 months, not a claim that every participant has completed follow-up.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Testing the Addition of a New Immunotherapy Drug, Atezolizumab (MPDL3280A), to the Usual Chemoradiation (CRT) Therapy Treatment for Limited Stage Small Cell Lung Cancer (LS-SCLC) — NCT03811002.
• Registry record last updated 2026-08-27; status last verified 2026-08.
• Results first posted 2025-06-12.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

The registry record describes a randomized, unmasked, parallel-group phase III study in limited stage small cell lung cancer. It compares chemotherapy and radiation therapy with the same chemoradiation regimen plus atezolizumab. Enrollment is recorded as 544, and the primary endpoint is overall survival in the intent-to-treat population.

The primary analysis reports median overall survival of 36.1 months for chemoradiation and 31.1 months for chemoradiation plus atezolizumab. The reported hazard ratio is 1.05, with a 95% confidence interval from 0.82 to 1.34 and p = 0.582 by the log-rank method. The interval includes the null value, so the analysis does not show a statistically clear overall-survival difference.

The registry also reports secondary outcomes and adverse-event summaries. Details of the original design assumptions, multiplicity strategy, interim analyses, and several aspects of missing-data handling are not reported in the registry record. The findings are best read as estimates with uncertainty within the population and follow-up described.