Preoperative Anti-PD-1 Antibody Combined With Chemoradiotherapy for Locally Advanced Squamous Cell Carcinoma of Esophagus: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Preoperative Anti-PD-1 Antibody Combined With Chemoradiotherapy for Locally Advanced Squamous Cell Carcinoma of Esophagus (NCT03792347), a na phase 1 study. The primary outcome is Safety (Number of Participants With Grade 3 and Higher-grade Treatment-related Adverse Events).

2. Clinical Question

3. Trial Design

This was a phase Ib, single-group study of preoperative pembrolizumab combined with chemoradiotherapy in locally advanced esophageal squamous cell carcinoma. The registry lists enrollment of 20 and identifies the status as completed. The start date is 2019-01-21, and the primary completion date is 2020-03-17. The sponsor is Hecheng Li M.D., Ph.D.

Because the study had a single group, it was not designed as a randomized comparison between treatment strategies. Allocation is recorded as not applicable, and masking is recorded as none. The study therefore describes outcomes observed among participants receiving the combined preoperative approach; it cannot isolate the contribution of pembrolizumab from the effects of chemoradiotherapy or other aspects of care.

The arm description states that participants received carboplatin and paclitaxel, with radiotherapy starting on the first day of chemotherapy. It specifies carboplatin at AUC=2 and paclitaxel at 50mg/m² on days 1, 8, 15, 22, and 29, and describes a radiotherapy course of 41.4 Gy in 23 fractions of 1.8 Gy, 5 fractions a week. The description is incomplete in the registry record, so additional treatment details are not reported in the registry record.

The primary outcome concerns grade 3 and higher treatment-related adverse events. Secondary outcome labels cover feasibility, pathologic complete response, and radiographic response. The record lists four outcome measures, but outcome results and analysis details are not reported in the registry record. The study’s stated purpose was to evaluate safety and feasibility and inform further clinical trials; that purpose does not itself establish a comparative treatment effect.

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 Safety (Number of Participants With Grade 3 and Higher-grade Treatment-related Adverse Events), measured over From date of treatment allocation until surgery was applied during study period or up to at least 90 days after last dose.

The primary endpoint is the number of participants with grade 3 and higher-grade treatment-related adverse events. The registry says adverse events were evaluated using CTCAE 4.0 and defines the observation period as running from treatment allocation until surgery was applied during the study period or up to at least 90 days after the last dose. The endpoint is a count of participants experiencing qualifying events, rather than a measure of the number of separate events or the duration of an event.

Counting participants with serious treatment-related toxicity is one way to describe the safety experience of a regimen. It can indicate how many participants crossed a specified severity threshold, but it does not convey the full distribution of event types, timing, duration, or clinical consequences. Nor does a count alone establish whether events were caused by one component of a combined regimen.

The registry record does not report the primary endpoint count, its denominator for analysis, or an uncertainty interval. It also does not report a statistical comparison for this endpoint. The serious adverse event summary is a separate set of registry figures and should not be treated as the primary endpoint result: serious-event status and grade 3 or higher treatment-related status are not interchangeable definitions.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods determine how observations are summarized and how much inferential weight can be placed on them. For this study, the key interpretive fact is the single-group design: there is no concurrent randomized control group in the registry record. Descriptive counts can report what was observed in the treated group, but a treatment contrast requires a comparator. No statistical analysis entries are listed in the registry record.

What does a participant count measure?

The primary outcome is defined as the number of participants with grade 3 and higher-grade treatment-related adverse events. A participant count records whether participants met the endpoint definition; it is not the same as a count of all adverse event episodes. One participant could experience more than one event, but the endpoint wording counts participants, not event occurrences. The registry record does not provide the primary endpoint count or an analysis denominator.

The serious adverse event summary has its own at-risk and affected counts, as does the summary of other adverse events. Those are descriptive safety figures. The endpoint’s wording additionally specifies a treatment-related assessment and a grade threshold. Seriousness, severity grade, and attribution to treatment are distinct concepts, so the available serious-event figures cannot be assumed to equal the primary endpoint count.

Why does the absence of a comparator matter?

A randomized comparison is designed to make groups comparable in expectation, allowing a difference between groups to be interpreted as an estimate of a treatment contrast, subject to the study’s assumptions and conduct. This study has one treatment group. Its observed outcomes may reflect the combination of pembrolizumab, chemotherapy, radiotherapy, surgery, patient characteristics, and chance. Without a comparator, the registry record cannot quantify what would have happened to a similar group receiving another strategy.

General statistical explanation: random variation affects observed counts, especially when a study has limited information. An interval estimate can express uncertainty around an estimate, while a hypothesis test addresses compatibility with a specified null hypothesis under a model. Neither an interval nor a test can repair the absence of a relevant comparator for a comparative question. The registry record does not report confidence intervals or hypothesis tests for the outcomes.

How should event proportions and uncertainty be considered?

A proportion relates the number of participants with an event to a defined population at risk or analyzed population. The registry gives at-risk and affected counts for its serious and other adverse event summaries, but it does not report a corresponding denominator for the primary endpoint analysis. No new proportion should be inferred from those separate summaries as if it were the primary endpoint result.

General statistical explanation: uncertainty around a proportion depends on the amount of information and the assumptions used to construct the interval. Small datasets can yield imprecise estimates, even when the observed count appears clear. This matters when interpreting safety summaries: a descriptive count characterizes the study group, but does not precisely establish the frequency expected in other populations or settings.

What is over-dispersion, and is it relevant?

General statistical explanation: over-dispersion occurs when observed variability exceeds what a chosen statistical model expects. It can affect standard errors and confidence intervals if a model assumes less variability than the data exhibit. The registry record does not report a model for the primary endpoint, event-level data, or an over-dispersion assessment. It is therefore not possible to say whether over-dispersion was assessed or would affect an analysis of these results.

What do the planning assumptions mean?

The planning illustration is described as a calculation based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It considers an effect defined as a stated fraction of the observed effect. These are illustrative planning assumptions, not evidence that the trial had two arms: the study itself is single-group. A planning sample size estimates the number of participants per arm under an assumed effect and model; it is not observed power and does not validate the assumed effect.

General statistical explanation: sample-size planning depends on the question, target effect, variability, allocation, and inferential criteria. If the observed effect used as an input is uncertain, calculations based on it inherit that uncertainty. Planning figures should therefore be treated as scenarios rather than guarantees. The registry record does not report the study’s own sample-size justification, target power, or statistical design calculations.

Can the posted outcomes support a treatment-effect estimate?

The registry lists the primary safety outcome and three secondary outcome labels, but reports no outcome analyses. With no analysis entries, the record does not supply an adjusted or unadjusted effect estimate, its precision, or the assumptions used to obtain it. The serious-event counts remain descriptive evidence about the listed group. They do not by themselves estimate a comparative treatment effect, and they do not replace the missing primary endpoint 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("Preoperative Pembrolizumab With Chemoradiotherapy ")
n      <- c(20)
events <- c(13)
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  Preoperative Pembrolizumab With Chemoradiotherapy   20  13  65.0  40.8  84.6  43.3  81.9  43.0  81.9

11. Multiplicity, Missing Data and Analysis Populations

The registry lists safety as the primary outcome and names feasibility, pathologic complete response, and radiographic response as secondary outcomes. It also lists four outcome measures. Multiple outcomes can create a multiplicity issue when several hypotheses are tested: the chance of at least one apparently notable result can rise as the number of tests increases. General statistical practice may address this through prespecified testing hierarchies or adjustments, but the registry record does not report a multiplicity strategy or a statistical analysis plan. No analysis entries are listed, so the posted information does not show whether outcome testing was conducted or adjusted.

Missing data can also affect estimates. Participants may lack outcome measurements, discontinue treatment, not undergo surgery, or have incomplete follow-up; how these situations are handled depends on the outcome and analysis. The registry record does not report missing-data counts, imputation methods, sensitivity analyses, or rules for classifying participants without a recorded outcome. These approaches must not be inferred from the study’s feasibility endpoint or its treatment schedule.

The record does not report analysis populations, such as whether analyses included all enrolled participants, only those who started treatment, or a subset with evaluable outcome data. Enrollment of 20 is reported, but enrollment alone does not establish the denominator for every endpoint. The serious-event summary separately gives at-risk counts for its listed outcomes; those denominators should not be assumed to define the primary endpoint population.

Without reported outcome analyses, missing-data procedures, analysis populations, or multiplicity handling, it is not possible to determine how these issues affected the posted results. This limits reproducibility and makes it difficult to assess whether outcome estimates, if generated, followed prespecified methods.

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: lymphopenia (60.0%), leukopenia (10.0%), decreased neutrophil count (5.0%), esophageal hemorrhage (5.0%).

The serious adverse event summary lists 20 participants at risk and 13 affected in the preoperative pembrolizumab with chemoradiotherapy group. For other adverse events, it lists 20 at risk and 20 affected. The summary also records 1 death. These are registry safety figures for the listed group; they do not constitute a comparison with another treatment group.

The serious adverse event terms most frequently listed are lymphopenia at 60.0%, leukopenia at 10.0%, decreased neutrophil count at 5.0%, and esophageal hemorrhage at 5.0%. These percentages describe the terms as recorded in the registry summary. The registry record does not provide further information about event timing, duration, recurrence, or the criteria behind each term’s listing.

These figures should be distinguished from the primary endpoint, which counts participants with grade 3 and higher-grade treatment-related adverse events. An event’s seriousness is not the same as its severity grade, and the primary endpoint also specifies treatment-related attribution. The serious-event figures therefore cannot be read as the missing primary endpoint count. The registry record does not report that primary count or an uncertainty interval.

General statistical explanation: descriptive safety counts characterize the experience recorded in a study group. Without a control group, they cannot show whether the observed pattern differs from what would occur under another treatment strategy. They also do not establish a rate for populations beyond the participants represented in the registry record.

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

The registry record describes a completed phase Ib, single-group study evaluating preoperative pembrolizumab with chemoradiotherapy in locally advanced esophageal squamous cell carcinoma. It identifies safety as the primary outcome and lists feasibility, pathologic complete response, and radiographic response as secondary outcomes. The serious adverse event summary gives affected and at-risk counts, names several event terms, and records a death.

Those details provide a descriptive account of the study’s treatment context and posted safety summaries. They do not provide the primary endpoint count. Nor does the record report results for the listed secondary outcomes, outcome analyses, confidence intervals, or p-values. Thus, the available information does not establish the magnitude or precision of the primary safety endpoint, or the results for the secondary outcomes.

The single-group design also sets a firm limit on interpretation. The study has no randomized comparator, so differences attributable specifically to pembrolizumab cannot be isolated from chemotherapy, radiotherapy, surgery, or other factors. The posted adverse event figures describe participants in the treatment group; they do not show how those figures compare with an alternative approach.

General statistical explanation: a study can be useful for describing feasibility and generating information for future research without providing a comparative estimate. A larger or controlled study may address questions this design cannot answer, but the registry record does not report what future study design or analysis will be used. The study’s stated purpose of informing further trials is a research aim, not a result about treatment benefit.

Accordingly, the findings should be read as limited registry evidence about a particular preoperative treatment combination. They do not establish a causal treatment effect, confirm the frequency of the primary endpoint, or determine how outcomes would differ under another regimen. The missing endpoint results and analysis details are central to judging both the observed safety profile and its uncertainty.

14. Why This Trial Matters Statistically

This study matters statistically because it illustrates the distinction between describing outcomes and estimating a comparative effect. A phase Ib study with one treatment group can record adverse events and report whether participants reached defined milestones. Those descriptions may inform future research questions. But they cannot show, on their own, how outcomes differ from another strategy.

The primary endpoint is a count of participants with grade 3 and higher-grade treatment-related adverse events, a clinically focused safety measure. Yet the registry record does not report that count. This creates a basic reporting limitation: the central endpoint named in the record cannot be evaluated from the posted results. The separate serious adverse event figures offer some safety context, but their definitions do not match the primary endpoint automatically.

The record also highlights why analysis methods and populations matter. A count is interpretable only in relation to a clearly defined outcome, follow-up period, and analysis population. The record describes the endpoint timeframe but does not report the primary analysis denominator, missing-data methods, or statistical plan. Without these details, readers cannot assess precision, reproducibility, or the sensitivity of a result to analysis choices.

General statistical explanation: early-phase research often informs whether a treatment approach merits further study, while later comparative research is needed to estimate differences between approaches. The registry’s stated goal of informing further trials is consistent with a hypothesis-generating role, but no outcome result or inference should be assumed from that goal. Statistical value depends not only on the design but also on transparent outcome reporting.

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 principal limitation is the single-group design. No control arm is listed, so the registry record cannot support a randomized comparison or isolate the effect of pembrolizumab from the concurrent chemotherapy and radiotherapy. Allocation is recorded as not applicable, and masking is recorded as none. These design features limit causal interpretation.

The primary endpoint result is not reported in the registry record. The serious adverse event summary cannot be used as a substitute because the primary endpoint specifies grade 3 and higher-grade treatment-related events, while the safety summary reports serious and other adverse event counts. The record does not provide the primary endpoint denominator, confidence intervals, or statistical tests.

Although the study is marked completed, the record does not report results for feasibility, pathologic complete response, or radiographic response. It lists four outcome measures, but analysis entries are absent. The statistical analysis plan, analysis populations, multiplicity approach, and missing-data procedures are not reported in the registry record.

The arm description is incomplete: it gives details for carboplatin, paclitaxel, and radiotherapy, but ends before all regimen information is provided. The registry record does not report further details in that description. This constrains a full reconstruction of the intervention from the record alone.

The listed adverse event summaries do not provide event-level detail about timing, duration, recurrence, or uncertainty. No comparator is available to contextualize the counts against another treatment approach. These limitations mean the posted information supports only a narrow descriptive reading of the study and does not establish comparative safety or efficacy.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Preoperative Anti-PD-1 Antibody Combined With Chemoradiotherapy for Locally Advanced Squamous Cell Carcinoma of Esophagus — NCT03792347.
• Registry record last updated 2023-11-22; status last verified 2023-11.
• Results first posted 2023-11-22.
• 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

NCT03792347 is a completed phase Ib, single-group study of preoperative pembrolizumab with chemoradiotherapy for locally advanced esophageal squamous cell carcinoma. The registry lists enrollment of 20, a start date of 2019-01-21, and a primary completion date of 2020-03-17. Allocation is not applicable, and masking is none.

The primary outcome is the number of participants with grade 3 and higher-grade treatment-related adverse events, evaluated using CTCAE 4.0 over the stated treatment and follow-up period. The primary endpoint result is not reported in the registry record. Secondary outcomes include feasibility, pathologic complete response, and radiographic response; their results are also not reported in the registry record.

The serious adverse event summary lists 13 affected among 20 at risk, other adverse events affecting 20 among 20 at risk, and 1 death. These descriptive figures are not a randomized comparison and are not equivalent to the primary endpoint. No analysis entries, statistical plan, or outcome uncertainty estimates are reported in the registry record.