Anti-PD-L1 in Stage IIIA(N2) Non-small Cell Lung Cancer (NSCLC): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Anti-PD-L1 in Stage IIIA(N2) Non-small Cell Lung Cancer (NSCLC) (NCT02572843), a na phase 2 study. The primary outcome is Event-free Survival (EFS) Rate.

2. Clinical Question

3. Trial Design

NCT02572843 is registered as a multicenter, single-arm phase II trial. Its official title describes MEDI4736 added to neoadjuvant chemotherapy for patients with stage IIIA(N2) NSCLC. The registry lists an enrollment of 68 and a completed status. The record gives a start date of 2016-06-16 and a primary completion date of 2020-01-20.

The intended population was patients with primary, resectable disease. In the arm description, patients whose tumors were deemed resectable at diagnosis were to receive cisplatin/docetaxel chemotherapy, followed by neoadjuvant MEDI4736 and surgery. The study objective refers to adding both neoadjuvant and adjuvant immunotherapy. However, the postoperative arm description is incomplete in the registry record, so the planned adjuvant regimen and its eligibility conditions cannot be fully characterized from this information.

Because the design is single-group, outcomes are observed under the study regimen without a concurrent control arm. This can describe the treated group and its follow-up, but it does not isolate the effect of MEDI4736 from chemotherapy, surgery, patient selection, or other influences. The registry lists no masking and no allocation to different groups.

The primary outcome is the EFS rate at 12 months in the FAS, estimated with the Kaplan-Meier method and accompanied by a 90% confidence interval. Secondary outcomes listed in the record include median EFS, overall survival rate, objective response after neoadjuvant chemotherapy and after neoadjuvant immunotherapy, pathological complete response, and adverse events assessed according to NCI CTCAE v4.0. Numerical results for these outcomes 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.

Accrual

Reasons for not completing this period:

Neoadjuvant chemotherapy

Reasons for not completing this period:

Neoadjuvant immunotherapy

Reasons for not completing this period:

Surgery

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: FAS

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: FAS

6. Primary Endpoint

The registry lists the primary outcome as Event-free Survival (EFS) Rate, measured over at 12 months.

Event-free survival (EFS) concerns the time participants remain free of events defined by the trial. The primary outcome is expressed as an EFS rate at 12 months, rather than as a comparison between treatment groups. The registry identifies the population as the FAS and specifies estimation by the Kaplan-Meier method. It does not report the event definition, the rules for censoring, or the estimate and its 90% confidence interval.

In general, a time-to-event endpoint can reflect more than a single type of clinical event and can account for differing lengths of follow-up. Kaplan-Meier estimation uses participants’ observed follow-up and event times to estimate the proportion remaining event-free over time. Participants who have not had a recorded event by their last observation may contribute follow-up up to that point; interpreting the estimate depends on the study’s censoring rules and on follow-up being adequately described. The specific rules used in this trial are not reported in the registry record.

A 12-month rate offers a defined time point for describing EFS in the study population. It does not, on its own, establish what would have happened under another treatment, particularly in a single-arm study. The confidence interval is intended to convey statistical uncertainty around the estimate: narrower intervals indicate greater precision, while wider intervals indicate less precision. Because neither the estimate nor the interval is reported, the precision of the primary result cannot be assessed from this record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The methods can be understood by separating the study design, the outcome definition, and the uncertainty around an estimate. The registry identifies this investigation as a multicenter, single-arm phase II trial. It lists no masking and no allocation to treatment groups. The primary endpoint uses a time-to-event method, while the safety information is presented as counts and percentages. These methods answer different descriptive questions and should not be treated as interchangeable evidence.

What does a single-arm design estimate?

A single-arm study follows a treated group and describes outcomes within that group. It can estimate the observed outcome under the study regimen, subject to eligibility, follow-up, endpoint definitions, and missing-data handling. It does not provide an internal treatment contrast. Without a concurrent comparator, an observed outcome cannot by itself show what would have occurred to a similar group receiving another regimen. This is why the design should not be read as a randomized estimate of the added effect of MEDI4736.

The registry describes treatment with cisplatin/docetaxel followed by neoadjuvant MEDI4736 and surgery for patients whose tumors were considered resectable at diagnosis. The objective includes adjuvant immunotherapy, but the postoperative treatment details are incomplete in the record. The planned course therefore cannot be reconstructed fully from the available arm description.

How does Kaplan-Meier estimation work?

Kaplan-Meier estimation is a general method for summarizing time until an event. It uses event times and follow-up information, including observations that end before an event is recorded. Such observations are commonly treated as censored at the last known follow-up, subject to the study’s rules. The method can estimate the proportion remaining event-free at a specified time even when participants have different follow-up lengths.

For this trial, the primary outcome is the EFS rate at 12 months, measured by Kaplan-Meier in the FAS. The registry does not state which events count as EFS events, how censoring was defined, how participants with incomplete follow-up were handled, or whether assumptions and diagnostics were assessed. These details matter because they determine which observations contribute to the estimate and how the endpoint should be interpreted.

What does a confidence interval communicate?

A confidence interval conveys uncertainty in an estimated quantity. In general, an interval around an EFS rate indicates the range of values compatible with the data under the statistical procedure used. Its width reflects precision: more uncertainty is generally associated with a wider interval. A confidence interval is not a guarantee that the true value lies within the reported limits, and it does not substitute for examining the study design or possible sources of bias.

The registry specifies a 90% confidence interval for the primary rate, but the interval and point estimate are not reported in the registry record. Consequently, neither the magnitude nor the precision of the primary estimate can be evaluated from the posted information. The confidence level alone does not reveal the result.

Why are analysis populations important?

An analysis population defines which participants contribute to an endpoint analysis. The primary endpoint is assigned to the FAS, but the registry does not define that population or explain how participants with protocol deviations, incomplete treatment, or missing outcome data were handled. The secondary response outcomes refer to FAS and FAS II, but those labels are not defined in the record. Population definitions are important because changing who is included can change the interpretation of an estimate.

How should the planning illustrations be read?

The planning assumptions describe illustrations computed from an observed effect, not the trial’s own design. They specify a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and effects defined as fractions of the observed effect. For time-to-event illustrations, the stated convention is a total event count; for other outcomes, it is participants per arm.

These are hypothetical planning assumptions and do not describe the actual single-arm trial allocation. The registry does not report the effect estimate underlying the illustrations or the trial’s own sample-size justification. Such illustrations are not observed power and cannot establish that the trial was adequately powered. Sample-size calculations concern the performance of a proposed design under assumptions; they do not measure the probability that a result is true or prove an effect.

What can be inferred from the posted analysis information?

The record lists no analysis results. It also does not report a statistical analysis plan, interim-analysis approach, adjustment for multiple outcomes, or detailed missing-data procedures. Those absences limit a full assessment of how analyses were specified and carried out. The most defensible reading is therefore descriptive: the registry states the planned endpoint and method, but does not provide the primary estimate needed to assess the result.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("MEDI4736")
n      <- c(67)
events <- c(49)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  MEDI4736  67  49  73.1  60.9  83.2  61.5  82.3  61.4  82.3

11. Multiplicity, Missing Data and Analysis Populations

The registry lists seven outcome measures. The primary outcome is the EFS rate at 12 months; the named secondary outcomes include median EFS, overall survival rate, objective response after neoadjuvant chemotherapy and after neoadjuvant immunotherapy, pathological response, and adverse events. When a study assesses several outcomes, the chance of obtaining an apparently notable finding among them can differ from the chance associated with a single prespecified test. Statistical plans may address this through a hierarchy, adjustment, or an explicitly exploratory interpretation. The approach used for multiplicity is not reported in the registry record.

The record does not provide a statistical analysis plan, analysis-specific estimates, p-values, or details of any interim analyses. It therefore does not show which outcomes were confirmatory, how comparisons were prioritized, or whether any adjustment was applied. Because the study is single-arm, the main interpretive limitation is not only multiplicity but also the lack of a concurrent randomized comparator.

The FAS is identified for the primary endpoint, but its definition is not reported. The record also labels response populations as FAS and FAS II without defining them. Missing-data methods, follow-up completeness, censoring rules, and sensitivity analyses are not reported in the registry record. For a time-to-event endpoint, loss of follow-up and the handling of censored observations can affect the estimate; for response outcomes, missing assessments can affect who is counted in the analysis. The posted information does not allow those issues to be evaluated.

The absence of posted estimates should not be mistaken for evidence that an outcome was zero or that no events occurred. It means the numerical result is not reported in the registry record. Likewise, a completed status does not reveal whether analyses were conducted according to a prespecified plan.

12. 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: lung infection (9.0%), pleural effusion (4.5%), atrial fibrillation (3.0%), nausea (3.0%), dehydration (3.0%).

The registry lists safety counts for the MEDI4736 group. It records 67 participants at risk for serious adverse events, with 36 affected, and 67 at risk for other adverse events, with 66 affected. It also lists 25 deaths. These are recorded counts; the registry information does not establish that each event or death was caused by study treatment.

The listed serious adverse event terms and percentages are lung infection at 9.0%, pleural effusion at 4.5%, atrial fibrillation at 3.0%, nausea at 3.0%, dehydration at 3.0%, and prostate cancer at 3.0%. The record does not explain the denominator or observation window for these term-specific percentages, nor whether participants could contribute more than one event or term. The percentages should therefore be read as the registry’s reported values, not as values recalculated from the counts.

Adverse-event summaries describe observed safety information in the study group. In general, serious adverse events are a category distinct from other adverse events, and event counts do not necessarily equal the number of affected participants. The registry identifies adverse events as assessed according to NCI CTCAE v4.0, but does not provide further detail on grading, attribution, timing, or the relationship of the listed events to treatment. With no concurrent control group, these figures do not estimate comparative risk.

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

The record documents a phase II investigation of MEDI4736 added to neoadjuvant chemotherapy in primary resectable stage IIIA(N2) NSCLC. It describes a single-arm, multicenter design, an enrollment of 68, the primary EFS endpoint at 12 months, and selected safety counts. It also names secondary outcomes. These details clarify what the trial set out to study and what kinds of outcomes were collected.

The main efficacy result cannot be interpreted from the registry record because the primary EFS estimate and its 90% confidence interval are not reported. The record therefore does not show the observed event-free rate at 12 months or how precisely it was estimated. The absence of that result does not mean the outcome was favorable, unfavorable, or unchanged; it means a numerical result is not reported in the registry record.

Even if the primary estimate were available, the single-arm design would support a description of outcomes in the study group rather than a randomized comparison. Without a concurrent control group, the study cannot separate the contribution of MEDI4736 from the other components of treatment or from differences in prognosis and care. A comparison with an external group would require information about comparable eligibility, outcome definitions, and follow-up, which is not established by the registry information.

The safety counts offer a limited description of events recorded in the group. They do not by themselves establish causation or quantify excess risk relative to another treatment. The term-specific percentages are reported, but the registry does not state their denominator or observation window. Likewise, the listed number of deaths does not identify their causes or relationship to treatment.

The study’s completed status and stated objectives should be kept separate from evidence about outcomes. A registration record can identify the trial and its planned measures, but the information is insufficient to judge the primary endpoint result, the precision of that result, or comparative efficacy and safety.

14. Why This Trial Matters Statistically

This trial is statistically informative as an example of why study design and result reporting must be read together. Its primary endpoint is a time-to-event measure, estimated at a specified follow-up point by Kaplan-Meier methods and paired with a confidence interval. That framework can account for differing follow-up, but it depends on clear event definitions, censoring rules, and sufficient reporting of the estimate. Those details and the primary numerical result are not reported in the registry record.

The single-arm structure is central to interpretation. A within-group outcome can describe what happened among enrolled participants, but it cannot provide a direct randomized estimate of the incremental effect of the immunotherapy component. This distinction is particularly important when an intervention is added to a multi-component regimen that includes chemotherapy and surgery. The observed outcome, if reported, would reflect the study context rather than an isolated treatment effect.

The record also illustrates why a sample-size planning illustration should not be mistaken for a trial result. The listed assumptions describe hypothetical planning calculations based on an observed effect, with a two-arm equal-allocation setup. That setup differs from this single-arm design, and the underlying observed effect is not reported in the registry record. Planning calculations do not establish observed power or replace reporting the actual endpoint estimate and its uncertainty.

Finally, the difference between naming an outcome and reporting its analysis is consequential. The registry identifies EFS, response, survival, pathological response, and adverse events, but does not provide numerical efficacy results. A clear statistical appraisal therefore needs to distinguish protocol-level intent, design limitations, posted counts, and missing analysis results rather than treating them as equivalent evidence.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

Several limitations specific to this trial follow from the registry information. First, the study is single-arm, with no randomized comparator. Any observed outcome would describe the treated group and could not by itself identify the effect attributable to MEDI4736. The record explicitly notes that this is not a randomized comparison and that arms are not a like-for-like control comparison.

Second, the primary result is not reported. The registry identifies EFS rate at 12 months, the FAS, the Kaplan-Meier method, and a 90% confidence interval, but gives no estimate or interval. The event definition and censoring rules are also not reported, limiting assessment of endpoint construction and follow-up.

Third, the analysis populations are not fully described. The FAS is named for the primary endpoint, and FAS and FAS II appear in secondary outcome descriptions, but their definitions and membership are not reported. Missing-data procedures, follow-up completeness, and sensitivity analyses are likewise not reported in the registry record.

Fourth, the arm description is incomplete after surgery, even though the study objective refers to adjuvant immunotherapy. The postoperative treatment plan and its conditions cannot be established from the registry information. The record also does not provide detailed eligibility criteria, response definitions, or outcome-specific assessment schedules beyond the primary endpoint’s 12-month time point.

Finally, the selected safety counts lack detail on event attribution, timing, and the denominators or observation windows for term-specific percentages. With no control group, they cannot support comparative safety conclusions. The record lists a completed study but does not report the efficacy analyses needed to interpret the primary question.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Anti-PD-L1 in Stage IIIA(N2) Non-small Cell Lung Cancer (NSCLC) — NCT02572843.
• Registry record last updated 2025-10-28; status last verified 2025-10.
• Results first posted 2025-10-28.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

NCT02572843 is a completed multicenter, single-arm phase II trial of MEDI4736 added to neoadjuvant chemotherapy in primary resectable stage IIIA(N2) NSCLC. The record lists an enrollment of 68 and identifies 12-month EFS rate in the FAS as the primary outcome, measured by Kaplan-Meier with a 90% confidence interval.

The regimen description includes cisplatin/docetaxel followed by neoadjuvant MEDI4736 and surgery; the postoperative details are incomplete. Secondary outcomes include survival, response, pathological response, and adverse events. The registry reports selected adverse-event counts and percentages, but not the primary EFS estimate, its confidence interval, or numerical results for the secondary efficacy outcomes. A single-arm design does not provide a randomized treatment comparison.