Durvalumab (MEDI4736) With or Without SBRT in Clinical Stage I, II and IIIA Non-small Cell Lung Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Durvalumab (MEDI4736) With or Without SBRT in Clinical Stage I, II and IIIA Non-small Cell Lung Cancer (NCT02904954), a randomized phase 2 study. The primary outcome is Number of Subjects With Major Pathological Response (MPR).

2. Clinical Question

3. Trial Design

This was a randomized phase 2, parallel-group trial with no masking. The registry record names Weill Medical College of Cornell University as the sponsor and lists carcinoma, non-small-cell lung as the condition. The study status is completed. The recorded enrollment is 60. The trial began on 2016-12-02, and its primary completion date is 2020-09-16.

The two groups received durvalumab, with SBRT added to the pre-operative treatment in one group. In the monotherapy group, durvalumab 1.12 g was administered every 3 weeks for 2 cycles before surgical resection. In the combination group, the same pre-operative durvalumab schedule was paired with radiotherapy in 3 daily fractions, starting concurrently with the first cycle. Both descriptions include post-operative durvalumab monotherapy for 12 months; the monotherapy-group description specifies 1.5 g.

The primary outcome timeframe runs from the durvalumab start date to surgical resection, up to 10 weeks. It is a count of participants meeting a pathological response definition, not a time-to-event measure. The record also lists Kaplan-Meier disease-free survival proportion at 2 years and objective clinical response rate as secondary outcomes. Results for those outcomes are not detailed in the registry record used for this analysis, so their findings are not reported in the registry record as represented.

The posted primary comparison includes 30 participants in each group. The trial record does not report the original sample-size justification, the randomization procedure, allocation concealment, protocol deviations, or detailed eligibility criteria. The absence of masking is recorded; whether outcome assessment was masked is not reported in the registry record. These design details matter when judging how closely the posted comparison can be interpreted as an estimate of the intended randomized treatment contrast.

4. Participant Flow and Accounting

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

Before assignment: Of 64 enrolled participants, 60 met inclusion criteria and were randomized to treatment.

Neoadjuvant (Preoperative) to Surgery

Reasons for not completing this period:

Adjuvant Durvalumab

Reasons for not completing this period:

5. Baseline Characteristics

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

6. Primary Endpoint

The registry lists the primary outcome as Number of Subjects With Major Pathological Response (MPR), measured over Durvalumab start date to surgical resection, up to 10 weeks.

The primary endpoint is the number of participants with a major pathological response (MPR) in the resected specimen. MPR is defined as ≤10% residual viable tumor. The timeframe begins at the durvalumab start date and ends at surgical resection, up to 10 weeks. The endpoint is therefore based on pathology from surgery, rather than on a symptom score, imaging measurement, or survival duration.

A pathological response endpoint can provide an early measure of tumor response in a pre-operative treatment setting. It uses tissue obtained during resection to assess how much viable tumor remains. This makes the endpoint specific to participants with a resected specimen that can be assessed. The registry record does not state how participants without a resection or without an evaluable specimen were handled in the primary analysis, so the analysis population and missing-outcome rules are not reported in the registry record.

The posted counts are 2 participants with MPR in the durvalumab-monotherapy group and 16 in the durvalumab-plus-SBRT group. The endpoint is binary at the participant level: a participant either meets the stated MPR definition or does not. The count comparison summarizes how often that definition was met in each randomized group. It does not capture every dimension of clinical outcome, and it should not be treated as interchangeable with the secondary outcomes of disease-free survival or objective clinical response.

7. Results Posted in the Registry

• Arm 2 (Durvalumab Plus SBRT): 53.3% (16/30) versus 6.7% (2/30) on control; risk difference 46.7 percentage points (95% CI 26.7 to 66.6), risk ratio 8.00 (2.01–31.80); unadjusted two-proportion z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a binary participant-level response: MPR is present when the resected specimen has ≤10% residual viable tumor. The posted counts are 2 participants with MPR among 30 in the durvalumab-monotherapy group and 16 among 30 in the durvalumab-plus-SBRT group. Because this outcome is a count of participants with an event, the comparison can be expressed as group proportions and as a contrast between those proportions.

What does the risk difference describe?

In general, a risk difference subtracts one group’s event proportion from the other group’s event proportion. It is an absolute measure, expressed on the same scale as the proportions. A value above zero for the stated comparison means the event proportion is higher in the durvalumab-plus-SBRT group. The unadjusted recalculation is 0.467, with a 95% confidence interval from 0.267 to 0.666. The registry's reported Fisher Exact analysis gives a risk difference of 46.6, with a 95% confidence interval from 26.7 to 66.6.

The risk difference is useful because it describes the absolute separation between the observed group proportions. It is not a relative measure, and it does not identify a biological mechanism. The recalculation uses the posted summary counts without adjustment. The registry record does not report an adjusted primary analysis, covariates used for adjustment, or a prespecified model specification.

Why use Fisher's exact test?

The registry reports Fisher Exact for the risk-difference comparison. In general, Fisher's exact test evaluates the association between group assignment and a binary outcome using the observed cell counts, with a probability calculation that accounts for the fixed margins of the observed table. It is commonly used for categorical outcomes when counts are limited or when a large-sample approximation may be less suitable. The record does not state why this test was selected or whether it was prespecified.

The sponsor's reported p-value is < 0.001. The unadjusted recalculation reports p < 0.001. A p-value summarizes how incompatible the observed data are with a specified null model, assuming the test's conditions hold. It is not the probability that the treatment has no effect, nor does it quantify the magnitude or practical importance of the group difference. The registry's test and the recalculated result are distinct presentations and should be read with their respective estimates.

What do the risk ratio and odds ratio add?

The recalculated risk ratio is 8.0, with a confidence interval from 2.012 to 31.8. A risk ratio compares the event proportions directly, so a value above one indicates a greater observed proportion in the durvalumab-plus-SBRT group. The recalculated odds ratio is 16.0, with a confidence interval from 3.218 to 79.56. An odds ratio compares event odds rather than event proportions. Risk ratios and odds ratios answer related but different questions, and their numerical values should not be treated as interchangeable.

Both relative measures are based on the same posted binary outcome, but they describe the contrast on relative scales. When the outcome proportion in the comparator group is low, a relative measure may be numerically large even though the absolute count difference is a separate and essential part of the interpretation. The broad confidence intervals indicate uncertainty in the size of the relative contrast. The absolute risk difference and the raw counts help keep that scale visible.

How should the confidence interval and test be read?

In general, a confidence interval is a range of values compatible with the data and the method used to calculate the interval. A 95% confidence interval does not mean there is a 95% probability that a fixed true value lies inside this particular interval. Rather, under repeated use of the method and its assumptions, intervals constructed in that way would cover the target value at the stated rate. The interpretation also depends on the study design and on how outcomes and participants were included.

For the recalculated risk difference, the interval extends from 0.267 to 0.666. For the sponsor's reported risk difference, the interval is 26.7 to 66.6. The registry labels the sponsor's confidence interval as 95%. Both intervals are entirely on the positive side of zero in their respective displays, consistent with the reported small p-values for a null difference. This statistical evidence does not settle whether the estimate is clinically meaningful, generalizable, or free of bias.

The primary count endpoint is not a survival outcome, so a time-to-event method is not used for this comparison. The registry separately lists a Kaplan-Meier disease-free survival proportion at 2 years, but the methods and results for that secondary outcome are not reported in the registry record. It would therefore be inappropriate to infer its analysis details from the primary endpoint method.

Why can over-dispersion matter?

In general, over-dispersion means observed variation is greater than a statistical model expects. It is especially relevant when a model assumes a particular distribution for counts or repeated observations. Allowing for over-dispersion can prevent standard errors from being too small and confidence intervals from being too narrow. This primary outcome is reported as a binary participant count, and no over-dispersion model is described in the registry record. There is consequently no basis to claim that over-dispersion was assessed or adjusted for in this analysis.

More broadly, the posted results do not report a detailed statistical analysis plan, the primary analysis population, missing-data procedures, interim analyses, or multiplicity adjustments. Those details can affect how the primary result should be interpreted, especially when a trial reports more than one outcome. Their absence does not invalidate the posted count comparison, but it limits what can be concluded about the full planned analysis and its safeguards against selective reporting.

10. Confidence Intervals and Precision

Precision is reflected in the range of values that remain compatible with the data under the chosen analysis. The recalculated absolute risk difference is 0.467, with a 95% confidence interval from 0.267 to 0.666. The sponsor's reported risk difference is 46.6, with a 95% confidence interval from 26.7 to 66.6. These are different numerical presentations of the contrast, and the exact reported values should be retained when describing each analysis.

The intervals do not collapse to a single value, so the observed estimate is not an exact measure of the underlying contrast. The lower and upper bounds convey a range of uncertainty about its size. The recalculated interval lies above zero, and the sponsor's interval also lies above zero in its reported scale. This is consistent with the statistical tests reported for the comparison, but the interval does not establish that every value within it is equally plausible, nor does it capture every possible source of uncertainty such as selection, measurement, or missing outcomes.

The relative estimates are less precise in scale: the risk ratio is 8.0 with a confidence interval from 2.012 to 31.8, and the odds ratio is 16.0 with a confidence interval from 3.218 to 79.56. The ranges are wide, particularly at the upper ends. This cautions against treating the point estimates as stable forecasts of the size of a relative difference in a different setting. The point estimates and their intervals describe the posted data under the recalculation's methods, not a guarantee about future results.

The observed group counts, 2 and 16 participants with MPR, provide context for the uncertainty. The registry does not report additional precision measures such as a prespecified clinically important difference or the original target sample-size calculation. Precision should therefore be discussed in relation to the reported intervals and counts, rather than against an unreported design target.

11. 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 figures are planning illustrations based on the observed effect, not a record of the trial's original design calculation. For this binary participant-count outcome, values are participants per arm. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect used for planning is the stated fraction of the observed effect. The original trial's own design assumptions and sample-size justification are not reported in the registry record.

For planning at 100% of the observed effect, the illustration gives 14 participants per arm for 80% power and 19 participants per arm for 90% power. At 75% of the observed effect, it gives 23 participants per arm for 80% power and 30 participants per arm for 90% power. At 50% of the observed effect, it gives 42 participants per arm for 80% power and 56 participants per arm for 90% power.

In general, power is the probability that a planned study will meet its chosen statistical criterion if a particular effect is truly present and the planning assumptions hold. The 80% column means the illustrated design is sized for an 80% chance of detecting the assumed effect under those assumptions; the 90% column uses a higher target chance. These are design targets, not probabilities that the posted finding is correct. The sample-size figures are not observed power.

The smaller the effect a future study aims to detect, the more participants it generally needs, all else held constant. The curve should be read in that planning context and alongside its assumptions, not as a forecast of results or a recommendation for a definitive study. The planning illustration is tied to the observed effect, which may be imprecise, and it does not incorporate design features that are not listed in the stated assumptions.

12. 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

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.067; treatment proportion p₁ = 0.533; average p̄ = 0.300.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6481 = 1.2702.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5578 = 0.4694.
4. n = (1.2702 + 0.4694)² / (0.4667)² = 13.9, rounded up to 14 participants per arm 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
p0 <- 0.0666666666666667                          # control proportion
p1 <- 0.533333333333333                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(2904954)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 14 (80% power), 19 (90% power)
 75% of observed effect: n per arm = 23 (80% power), 30 (90% power)
 50% of observed effect: n per arm = 42 (80% power), 56 (90% power)
power.prop.test: 13.9 
Simulated power at n = 14 per arm: 0.844 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Analyses Reported by the Sponsor

The registry's reported primary analysis uses Fisher Exact and gives a risk difference of 46.6, a 95% confidence interval from 26.7 to 66.6, and p < 0.001. The unadjusted recalculation from the posted counts gives a risk difference of 0.467, a 95% confidence interval from 0.267 to 0.666, and p < 0.001. The recalculation also reports a risk ratio of 8.0 and an odds ratio of 16.0, with their respective confidence intervals.

The sponsor's risk difference is displayed on a different numerical scale from the recalculated risk difference, and some rounded values differ. The record does not mark a discrepancy. The sponsor analysis is the registry's stated trial analysis; the recalculation is a simple unadjusted analysis of the posted summary counts. They are related descriptions of the comparison, but should not be presented as one identical analysis or used to imply undocumented analytic choices.

The registry does not provide further detail about whether the sponsor's test and interval were prespecified, the analysis population, adjustments, or missing-data handling. Those points are not reported in the registry record, so the contrast between the reported analysis and the recalculation should remain limited to the methods and values explicitly listed.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome, the number of participants with MPR, and two secondary outcomes: Kaplan-Meier disease-free survival proportion at 2 years and objective clinical response rate. It records 4 outcome measures and 3 analyses. These listings indicate that more than one outcome or analysis exists in the record, but they do not establish how the analyses were ordered, which were confirmatory, or whether the reported primary comparison was selected before results were known.

Multiplicity arises when several statistical hypotheses are tested. In general, testing more hypotheses can increase the chance of at least one apparently notable result by chance if no adjustment or prespecified hierarchy is used. The registry does not report a multiplicity adjustment, a testing hierarchy, or a detailed statistical analysis plan. These procedures are therefore not reported in the registry record; it would be inappropriate to assume either that an adjustment was used or that none was considered in the trial's full analysis materials.

The primary outcome is defined from the resected specimen, but the record does not explain how participants who did not undergo surgery, lacked an assessable specimen, withdrew, or had missing pathology data were treated. The number analyzed and the reason for any difference between the randomized groups and the outcome groups are not reported in the registry record. The posted group counts and denominators describe the displayed result, but they do not reveal the handling of every participant or missing observation.

The registry record also does not report whether the analysis followed intention-to-treat, per-protocol, or another population definition. It does not give a detailed account of protocol deviations or outcome-assessment procedures. As the trial was unmasked, the record states that masking was none; whether pathologists or other outcome assessors were masked is not reported in the registry record.

These omissions constrain assessment of selective reporting, missing-data bias, and robustness. They do not alter the arithmetic of the posted counts, but they limit the ability to reconstruct the planned analysis and understand how all randomized participants were handled. The interpretation should remain tied to the reported endpoint, groups, counts, and stated test.

15. 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 (26.7%), thromboembolic event (16.7%), fatigue (10.0%), anemia (6.7%), bronchopleural fistula (6.7%).

The adverse-event record reports serious adverse events affecting 12 participants in the durvalumab-monotherapy group and 15 in the durvalumab-plus-SBRT group, with 30 at risk in each group. Other adverse events are reported for 29 participants in the monotherapy group and 30 in the combination group. The record lists 9 deaths in the monotherapy group and 10 in the combination group. These are group-level counts as recorded, not adjusted comparisons or estimates of treatment-attributable risk.

The listed serious adverse-event terms and percentages are lung infection, 26.7%; thromboembolic event, 16.7%; fatigue, 10.0%; anemia, 6.7%; bronchopleural fistula, 6.7%; and dyspnea, 6.7%. The registry excerpt does not specify whether these terms overlap within participants, the precise event-counting rules, severity grades, timing, or attribution to treatment. The listed percentages should therefore be read as presented rather than reinterpreted as mutually exclusive categories.

Counts alone do not establish that an event was caused by a treatment, and the registry information does not support a benefit-risk conclusion. Follow-up duration and detailed safety analysis methods are not reported in the registry record.

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

The primary posted result shows a higher observed MPR count and proportion in the group assigned durvalumab plus SBRT than in the group assigned durvalumab monotherapy. MPR is defined as ≤10% residual viable tumor in the resected specimen, with the endpoint assessed from durvalumab start to surgery, up to 10 weeks. The risk difference and its confidence interval quantify the observed group contrast under the stated analysis; they do not eliminate uncertainty about its size.

The randomized, parallel-group design is relevant because random assignment is intended to support a fair comparison of groups. However, the registry record does not detail how randomization was implemented, how participants were analyzed, whether missing specimens affected inclusion, or whether outcome assessors were masked. The absence of these details makes it difficult to assess all potential sources of bias from the registry record alone. The results are not a guarantee of the same outcome in other clinical settings or populations.

The primary endpoint is pathological response, not disease-free survival or overall survival. Although the record lists disease-free survival at 2 years and objective clinical response rate as secondary outcomes, the corresponding results are not reported in the registry record represented. The primary finding cannot be used as a substitute for those outcomes or as evidence about long-term disease control.

Likewise, the adverse-event counts and deaths are descriptive. Without detailed event timing, attribution, follow-up, and analysis procedures, they do not determine comparative safety or explain the cause of any individual event. No medical recommendation follows from the statistical comparison.

The most defensible interpretation is narrow: among the posted primary-outcome groups, the observed number meeting the registry's MPR definition was higher with durvalumab plus SBRT than with durvalumab alone, and the reported statistical analyses indicate a difference under their assumptions. The interval estimates show uncertainty around the size of that contrast. Broader conclusions about treatment effectiveness, long-term outcomes, generalizability, or safety require information not reported in the registry record.

17. Why This Trial Matters Statistically

This trial is statistically instructive because its primary outcome is a clear binary pathology measure, while its comparison can be described on both absolute and relative scales. The risk difference communicates the absolute separation between group proportions; the risk ratio and odds ratio express relative contrasts. Reading all of these alongside the raw counts helps prevent a large relative estimate from obscuring the scale and uncertainty of the underlying data.

The reported groups each contain 30 participants, and the counts with MPR are 2 and 16. Those counts are small enough that interval estimates deserve close attention. The recalculated risk difference has a 95% confidence interval from 0.267 to 0.666, while the relative-effect intervals also span substantial ranges. A statistically notable result and uncertainty about the magnitude can coexist. That distinction is central to interpreting small phase 2 trials.

The record also illustrates the difference between a trial's reported analysis and a transparent unadjusted recalculation. The sponsor reports Fisher Exact, a risk difference of 46.6, and p < 0.001. The recalculation from posted counts reports its own risk difference and p-value. Keeping those results distinct makes clear which estimate comes from the registry's stated analysis and which is derived from the posted summary counts.

Finally, sample-size illustrations based on an observed effect can help show how planning depends on the effect a future study seeks to detect. They are not substitutes for the original design rationale and do not show the probability that the current result is true. The trial's missing analysis details and unreported secondary outcome results also demonstrate why statistical interpretation requires attention to both numerical findings and the completeness of the registry record.

18. 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 primary result is based on a pathological response measured in a resected specimen. The registry record does not report how participants without surgery or an evaluable specimen were handled, or whether the outcome counts include all randomized participants. The analysis population, missing-data rules, and the number of participants excluded from the primary analysis are not reported in the registry record.

The trial is described as randomized and unmasked, but the randomization procedure, allocation concealment, and masking of outcome assessors are not reported. The record does not supply a detailed statistical analysis plan, original sample-size justification, interim-analysis information, covariate adjustment plan, or multiplicity approach. These gaps limit assessment of prespecification and the complete risk of bias.

The registry lists secondary outcomes, including disease-free survival proportion at 2 years and objective clinical response rate, but their results and analysis details are not reported in the registry record represented in the registry record. The primary pathological endpoint does not answer those longer-term or distinct outcome questions.

Safety information is presented as group counts and selected serious-event terms. The record does not report detailed event timing, duration of follow-up for safety, severity grading, attribution, participant-level overlap across event terms, or a comparative safety analysis. Death counts are listed, but their causes and relation to treatment are not reported in the registry record.

The posted sample-size illustration uses the observed effect with stated planning assumptions. It is not the trial's original design calculation, and the record does not report the original target rationale. As a phase 2 comparison with posted group counts of 30 per arm, its estimates should be interpreted with their confidence intervals rather than as precise forecasts for another setting.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Durvalumab (MEDI4736) With or Without SBRT in Clinical Stage I, II and IIIA Non-small Cell Lung Cancer — NCT02904954.
• Registry record last updated 2023-10-11; status last verified 2023-01.
• Results first posted 2021-08-03.
• 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.

22. Record Summary

NCT02904954 is a completed randomized phase 2 parallel-group trial of durvalumab with or without SBRT in clinical stage I, II, and IIIA non-small cell lung cancer. The registry record reports no masking and lists Weill Medical College of Cornell University as sponsor. Enrollment is recorded as 60.

The primary endpoint is the number of participants with MPR, defined as ≤10% residual viable tumor in the resected specimen, from durvalumab start to surgical resection, up to 10 weeks. The posted counts are 2 of 30 in the durvalumab-monotherapy group and 16 of 30 in the durvalumab-plus-SBRT group. The registry reports a Fisher Exact risk-difference analysis; an unadjusted recalculation also describes the contrast.

The record lists disease-free survival proportion at 2 years and objective clinical response rate as secondary outcomes. Detailed analysis populations, missing-data handling, multiplicity procedures, and the results of those secondary outcomes are not reported in the registry record.