Pembrolizumab, Combination Chemotherapy, and Radiation Therapy Before Surgery in Treating Adult Patients With Locally Advanced Gastroesophageal Junction or Gastric Cardia Cancer That Can Be Removed by Surgery: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Pembrolizumab, Combination Chemotherapy, and Radiation Therapy Before Surgery in Treating Adult Patients With Locally Advanced Gastroesophageal Junction or Gastric Cardia Cancer That Can Be Removed by Surgery (NCT02730546), a na phase 1, phase 2 study. The primary outcome is Pathological Complete Response (PathCR) Rate.

2. Clinical Question

The trial addresses a treatment strategy delivered before surgery, rather than a comparison between randomized treatment groups. Its primary outcome is pathological: it concerns whether cancer is recognized in the resected esophageal specimen and accompanying lymph nodes, and whether margins are free of tumor. This makes the endpoint distinct from outcomes based only on imaging or clinical assessment.

The registry also names complete resection with no tumor within 1 mm of the resection margins, disease-free survival, time to relapse, overall survival at two years, progression-free survival rate at two years, and dose-limiting toxicities as secondary outcomes. The record does not report numerical results for these outcomes. Consequently, it does not establish how often response occurred or how outcomes compare with another treatment approach.

3. Trial Design

NCT02730546 is listed as a phase 1b/2 trial sponsored by Mayo Clinic. The registry describes one treatment group receiving pembrolizumab, chemotherapy, radiation, and surgery. Allocation is listed as not applicable, masking as none, and the model as single group. These features mean the trial is not a randomized comparison and does not provide a concurrent control group for estimating a treatment contrast.

The trial concerns adults with resectable, locally advanced adenocarcinoma of the gastroesophageal junction or gastric cardia. The conditions include gastric cardia adenocarcinoma, gastroesophageal junction adenocarcinoma, and gastric cancer stages IB, II, IIA, IIB, IIIA, and IIIB under AJCC version 7. The registry lists an enrollment of 31. The arm description refers readers to a detailed description, so the complete treatment schedule and eligibility details are not reported in the registry record.

The official title describes neoadjuvant pembrolizumab plus concurrent chemoradiotherapy with weekly carboplatin and paclitaxel. The intervention list also includes fluorouracil, leucovorin calcium, and oxaliplatin, as well as computed tomography and laboratory biomarker analysis. The registry listing alone does not establish which listed drugs were given to every participant or explain their roles in the protocol.

The record gives a start date of 2016-06-24, a primary completion date of 2021-07-01, and a status of completed. These are registry milestones, not outcome findings. The stated primary endpoint is pathological complete response rate, defined using the pathology findings following resection. The record specifies a timeframe of up to 3 years. It does not report the actual response estimate, the number evaluable for that endpoint, or a statistical analysis plan.

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

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 Pathological Complete Response (PathCR) Rate, measured over Up to 3 years.

The primary endpoint is the pathological complete response rate, expressed as a percentage of patients. The registry defines it as the number of patients with a pathological complete response divided by the total evaluable patients. A response requires no recognized cancer and margins free of tumor, as determined by a pathologist following resection of the esophageal specimen and accompanying lymph nodes. The specified timeframe is up to 3 years.

This endpoint is used to describe the pathological findings after treatment before surgery. Unlike an assessment based only on scans or symptoms, it is based on examination of tissue removed at resection. It therefore measures a specific feature of the surgical pathology outcome; it does not, by itself, describe later recurrence, survival, or the overall balance of benefits and harms.

The denominator is evaluable patients, not necessarily everyone enrolled. The registry does not report how many participants were evaluable, the response count, or the resulting percentage. It lists a 95% confidence interval as the measure of dispersion, but does not provide the interval limits. Without those results, the endpoint definition can be interpreted, but the observed rate and its precision cannot be assessed from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a binary event summarized as a percentage: each evaluable patient either meets the pathological complete response definition or does not. A percentage is a way to express the frequency of an event relative to its stated denominator. In this trial, the denominator is explicitly the total evaluable patients, not automatically the enrollment figure. Because the record does not give the count of evaluable patients or the count meeting the response definition, it does not permit the rate to be checked or interpreted.

What does a confidence interval add?

In general, a confidence interval represents uncertainty around an estimated rate. It indicates a range of values compatible with the data under the method used; it is not a guarantee that the true value lies within that particular interval. The registry specifies a 95% confidence interval for the primary endpoint, but the interval itself is not reported. As a result, neither the estimated response rate nor the precision of that estimate can be assessed. The interval method is also not reported in the registry record.

Why does the endpoint denominator matter?

The endpoint definition uses evaluable patients. In general, an analysis restricted to evaluable participants can differ from an analysis using everyone enrolled if some participants lack the information needed to classify the endpoint. The registry does not say how evaluability was determined, how missing pathology information was handled, or whether participants who did not undergo resection were included in a particular way. These details matter because the denominator affects the interpretation of a percentage. Enrollment of 31 does not resolve these questions.

What can a single-group design estimate?

In general, a single-group study can describe outcomes observed among participants receiving its treatment strategy. A randomized comparison is designed to estimate differences between allocated groups, whereas this registry lists one group and no randomized comparator. An observed outcome in a single group does not by itself show what would have happened to comparable participants under another strategy. This limits causal interpretation and prevents a direct estimate of a between-group treatment effect from this design.

The registry lists allocation as not applicable, masking as none, and the model as single group. It does not report an adjusted primary analysis, a hypothesis-testing procedure, or a statistical plan. The record also lists no statistical analyses. That means a particular model, test, or adjustment should not be inferred from the endpoint definition alone.

How should the planning assumptions be read?

The record describes sample-size illustrations based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It states that the target effect is a specified fraction of the observed effect. For time-to-event outcomes, the illustration uses the total number of events; for other outcomes, it uses participants per arm. These are planning assumptions, not evidence that the trial used that design. They are not observed power, and they do not establish the original protocol’s rationale or sample-size justification. Numerical illustration results are not reported in the registry record.

In general, over-dispersion means that observed variability exceeds what a chosen statistical model expects. It can matter for some count or clustered-data analyses because underestimated variability can make uncertainty appear too small. The registry does not report an analysis model for which over-dispersion was assessed, so no such adjustment can be attributed to this trial. The primary endpoint is defined as a binary percentage, but the actual estimation method is not stated.

Time-to-event outcomes such as disease-free survival, time to relapse, overall survival, and progression-free survival concern the timing of events as well as whether they occur. In general, analyses of such outcomes must account for participants whose event has not been observed by the end of their follow-up. The registry names these outcomes but does not report their estimates, event counts, censoring details, follow-up summaries, or analysis methods. No assumptions about those features should be inferred.

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("Treatment (Pembrolizumab, Chemotherapy, Radiation,")
n      <- c(31)
events <- c(7)
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  Treatment (Pembrolizumab, Chemotherapy, Radiation,  31  7  22.6  9.6  41.1  11.4  39.8  11.5  40.0

11. Multiplicity, Missing Data and Analysis Populations

The registry lists several secondary outcomes alongside pathological complete response: complete resection with no tumor within 1 mm of the resection margins, disease-free survival, time to relapse, overall survival at two years, progression-free survival rate at two years, and dose-limiting toxicities. These outcomes represent multiple opportunities to describe results. In general, when many outcomes are tested, the chance of at least one apparently notable finding can increase unless the analysis plan addresses multiplicity. The registry does not report a multiplicity strategy, testing hierarchy, or adjustment method.

The record lists no statistical analyses. This does not establish that no analyses were conducted; it means analysis details are not available in the registry record. The statistical plan, including any interim analyses or prespecified subgroup analyses, is not reported in the registry record. No significance tests or adjusted estimates can therefore be attributed to the study from this information.

The primary endpoint denominator is defined as evaluable patients, but the criteria for evaluability are not reported. The record does not state how missing pathology assessments, incomplete follow-up, withdrawals, or participants not undergoing resection were handled. The analysis population for the primary endpoint and for the secondary outcomes is also not reported. These omissions make it difficult to judge the potential effect of missing outcome information or to reproduce the intended analysis.

In general, missing-data methods and analysis-population definitions help readers understand which participants contribute to an estimate and what assumptions support it. Because these details are absent, the reported enrollment cannot be treated as the denominator for every endpoint, and no missing-data adjustment should be assumed.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: sepsis (9.7%), infections and infestations - oth spec (6.5%), investigations - other, specify (6.5%), lymphocyte count decreased (6.5%), febrile neutropenia (3.2%).

For the treatment group, the registry lists 31 participants at risk for serious adverse events and 13 affected. It also lists 31 at risk for other adverse events and 31 affected, along with 11 deaths. These are descriptive figures. They do not establish that the treatment caused an event or death, and the registry record does not provide a comparison group for judging whether the counts differ from another treatment approach.

The listed serious adverse event terms include sepsis at 9.7%, infections and infestations - other specified at 6.5%, investigations - other, specify at 6.5%, lymphocyte count decreased at 6.5%, febrile neutropenia at 3.2%, and hemolysis at 3.2%. The record does not provide enough detail to determine the observation period, whether terms overlap within participants, or the event-level severity and attribution. The percentages should be read as reported, not reinterpreted as treatment-related risks.

Safety comparisons require consistent definitions, follow-up, and a suitable comparator. Those elements are not reported for a comparative analysis in this single-group record. The listed deaths should not be assumed to be caused by an adverse event or by study treatment; cause and attribution are not reported in the registry record.

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

The registry describes a completed, single-group phase 1b/2 study of pembrolizumab with chemotherapy and radiation before surgery in adults with resectable, locally advanced adenocarcinoma of the gastroesophageal junction or gastric cardia. It specifies pathological complete response as the primary endpoint and gives a pathology-based definition. These details identify the question the trial was designed to address.

The record does not report the primary response count, rate, denominator, or confidence interval. It therefore cannot show how often pathological complete response occurred or how precise an estimate was. The listed secondary outcomes also have no numerical results in the registry record. Completion status should not be interpreted as a substitute for those findings.

Even if an outcome estimate were available, the single-group design would describe the experience of participants in that group rather than establish a randomized difference from another treatment. The record does not provide a concurrent control group or a comparative treatment-effect estimate. Consequently, it does not support a causal conclusion about the effect of the treatment sequence compared with another approach.

The safety summaries document serious and other adverse events and deaths in the treatment group, but they do not establish attribution. Without a comparator and details about the observation period and event assessment, the figures cannot quantify excess risk or distinguish treatment-related events from events that could occur for other reasons.

The most defensible reading is therefore limited: the registry records the study design, the treatment strategy, the endpoint definitions, enrollment, and certain safety counts. It does not provide the efficacy results or statistical analysis details needed to assess response, survival, comparative effects, or the precision of the primary endpoint. Those unavailable results should not be interpreted as either favorable or unfavorable.

14. Why This Trial Matters Statistically

This trial raises a statistical question that is common in early-phase and single-group oncology studies: how to describe an outcome after a treatment strategy when a randomized comparator is not part of the design. A pathological endpoint can be defined precisely, but a precise definition does not supply an estimate. Readers need the response count, the evaluable denominator, and uncertainty around the rate to assess what the study observed.

The registry gives the endpoint definition and specifies a 95% confidence interval, yet reports neither the estimate nor the interval limits. That gap illustrates why outcome definitions and outcome results are not interchangeable. Without the results, readers cannot assess the response frequency or its precision. The evaluable-patient denominator also makes analysis-population reporting essential: enrollment alone does not identify who contributed to the endpoint.

The study design is also central to interpretation. A single-group result can be descriptive, but it cannot provide the same direct treatment contrast as a randomized comparison. This matters when readers distinguish evidence that an outcome occurred after a treatment sequence from evidence that the sequence caused an outcome or performed differently from an alternative.

Finally, the record lists multiple secondary outcomes and safety summaries without accompanying statistical analysis details or numerical efficacy findings. This emphasizes the value of clear reporting of analysis populations, missing-data handling, multiplicity, follow-up, and uncertainty. Those elements are necessary to understand how estimates were obtained and how far conclusions can reasonably extend. In this record, the design and endpoint are identifiable, while the primary result and its precision are not reported.

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.

The main limitation for statistical interpretation is that the registry record does not report the primary endpoint result. The response count, evaluable denominator, percentage, and 95% confidence interval are all absent. Although enrollment is listed as 31, it cannot be used as a substitute for the evaluable population specified in the endpoint definition.

The single-group design has no randomized comparator. It can describe outcomes in the treatment group, but it cannot directly estimate a randomized contrast or rule out explanations related to participant selection, disease features, or other care. The record does not report an adjusted comparative analysis.

Several details needed to assess reproducibility and potential bias are not reported in the registry record: the statistical analysis plan, the criteria for evaluability, missing-data handling, analysis populations, and methods for the secondary outcomes. The record also does not provide the numerical results for those secondary outcomes. Follow-up and event-assessment details for the safety summaries are not reported, and the deaths should not be assigned a cause or treatment relationship from these figures alone.

The official title specifies weekly carboplatin and paclitaxel, while the intervention list includes other drugs. The arm description refers to a detailed description, so the full regimen and the roles of the listed interventions are not reported in the registry record. This limits how specifically the administered treatment can be reconstructed from the record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Pembrolizumab, Combination Chemotherapy, and Radiation Therapy Before Surgery in Treating Adult Patients With Locally Advanced Gastroesophageal Junction or Gastric Cardia Cancer That Can Be Removed by Surgery — NCT02730546.
• Registry record last updated 2025-06-22; status last verified 2024-04.
• Results first posted 2022-08-25.
• 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

NCT02730546 is a completed phase 1b/2, single-group trial sponsored by Mayo Clinic. It studied pembrolizumab with chemotherapy and radiation before surgery in adults with resectable, locally advanced adenocarcinoma of the gastroesophageal junction or gastric cardia. Enrollment is listed as 31; allocation is not applicable, and masking is listed as none.

The primary endpoint is pathological complete response rate, defined by pathology after resection and assessed within up to 3 years. The registry does not report the endpoint estimate, its denominator, or its confidence interval. It lists several secondary outcomes and safety summaries, but does not provide numerical results for the secondary outcomes or a statistical analysis plan. The safety record lists 13 affected by serious adverse events among 31 at risk, 31 affected by other adverse events among 31 at risk, and 11 deaths; these figures do not establish causation.