S0356 Oxaliplatin, 5-FU, Radiation Therapy (RT), Surgery for Pts With Stage II or III Cancer of Esophagus or Gastroesophageal (GE) Junction: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for S0356 Oxaliplatin, 5-FU, Radiation Therapy (RT), Surgery for Pts With Stage II or III Cancer of Esophagus or Gastroesophageal (GE) Junction (NCT00086996), a na phase 2 study. The primary outcome is Pathological Complete Response.

2. Clinical Question

The study asked whether a treatment sequence combining fluorouracil, oxaliplatin, radiation therapy, and surgery could be evaluated in patients with stage II or III cancer of the esophagus or gastroesophageal junction. Its primary endpoint was pathological complete response after chemoradiotherapy and surgery. The registry defines this as no evidence of residual disease on pathology review. It also specifies that patients who did not receive surgery were counted as not having responded.

This question is about a pathological finding after treatment, rather than a comparison of outcomes between assigned groups. The study also listed overall survival, progression-free survival, and adverse events as secondary outcomes. Their results are not reported in the registry record. Accordingly, the available record does not establish how frequently the primary outcome occurred or describe the numerical results for those secondary outcomes.

3. Trial Design

This was a phase II, single-group study sponsored by SWOG Cancer Research Network. The registry records enrollment of 98 and lists esophageal cancer as the condition. The study began in 2004-09, reached primary completion in 2009-08, and is marked completed. Allocation is listed as not applicable, and masking is listed as none.

Participants in the listed group received neoadjuvant fluorouracil, oxaliplatin, and radiation therapy, followed by conventional surgery and adjuvant fluorouracil and oxaliplatin. The registry describes the primary endpoint as pathological complete response, assessed 10-16 weeks after beginning study treatment. The outcome definition counts a patient who did not undergo surgery as not having responded.

A single-group design describes outcomes in the treated group but does not supply a concurrent control group. It therefore does not provide a like-for-like randomized comparison against another treatment or usual care. Allocation is not reported as random, and the registry explicitly identifies the study as single-group. The interpretation should stay within that design: a response result, if reported, would describe this group under the study’s treatment sequence, not establish what would have happened to the same patients under an alternative.

The registry lists four outcome measures, including the primary response outcome, grade 3 through 5 adverse events related to study drug, overall survival, and progression-free survival. It reports no analyses. The statistical analysis plan, including any prespecified model or approach to uncertainty, is not reported in the registry record. The official title refers to molecular correlates, but molecular results and their analysis are not reported in the registry record.

4. Participant Flow and Accounting

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

Overall Study

Reasons for not completing this period:

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

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, measured over 10-16 weeks after beginning study treatment.

Pathological complete response is a binary endpoint: each eligible patient is classified as having met the response definition or not. The registry defines response as no evidence of residual disease on pathology review after chemoradiotherapy and surgery. Assessment was scheduled for 10-16 weeks after beginning study treatment. Patients who did not receive surgery are explicitly classified as not having responded, which makes failure to reach surgery part of the endpoint’s outcome definition rather than an unclassified response.

As a general statistical concept, a pathological endpoint can provide evidence about tumor findings at a specified point in a treatment course. It is distinct from outcomes measured over follow-up, such as overall survival or progression-free survival. A response count or proportion would describe how often the endpoint occurred in the analyzed population; an uncertainty interval would indicate how precisely that proportion is estimated. Neither the response count nor a response proportion is reported in the registry record, and no uncertainty interval is reported there.

The endpoint definition alone does not reveal how many patients had residual disease, how many underwent surgery, or how the pathology assessments were conducted. Those details are not reported in the registry record. The listed eligible-patient population and the rule for patients without surgery give important context, but they do not substitute for the endpoint result.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies the study as a single-group phase II trial and lists pathological complete response as its primary outcome. The statistical concepts below explain how such outcomes are generally interpreted; they do not describe analyses confirmed to have been conducted for this study. The registry reports no analyses, and the statistical plan is not reported in the registry record.

What does a binary endpoint represent?

For a binary outcome, each participant is placed in one of two categories, such as meeting or not meeting a response definition. The registry’s pathological complete response endpoint follows this structure: response requires no evidence of residual disease on pathology review, and a patient who does not receive surgery is counted as not responding. In general, an outcome frequency is based on the number meeting the definition relative to the relevant analysis population. The registry does not report the response count or proportion, so neither can be interpreted for this trial.

The endpoint definition also affects what the result would mean. Counting patients without surgery as nonresponders includes treatment-course completion in the outcome classification. It does not, by itself, explain why surgery was not performed or indicate how many patients were affected. Reasons for not receiving surgery and counts by treatment stage are not reported in the registry record.

What would an uncertainty interval tell a reader?

As a general concept, a confidence interval represents a range of values compatible with the observed data under a specified statistical method. Its width reflects precision: more uncertainty produces a wider interval, while more precise estimation produces a narrower one. An interval does not guarantee that the true value lies within its particular limits, and it is not a measure of the probability that a treatment works. The registry record does not report a response estimate or confidence interval for this study, so the precision of the primary outcome cannot be assessed from the reported information.

For a binary outcome, estimates and intervals depend on the observed number of events, the denominator, and the method used. Different methods can behave differently, especially when event counts are small. The analysis method and outcome counts are not reported, so no method-specific estimate can be reconstructed from the registry record.

Why does the single-group design matter?

A single-group study can describe the outcomes observed among participants who received the study regimen. It does not directly estimate a contrast against a concurrent control group. General statistical reasoning distinguishes description from comparison: observing an outcome in one group cannot isolate the effect of treatment from other explanations, such as participant characteristics, other care, or the process used to assess outcomes. This is especially important when comparing a single-group result with results from another study, because populations and study procedures may differ.

The registry lists allocation as not applicable and masking as none. It does not describe randomized allocation or a blinded assessment process. These design details do not invalidate a descriptive outcome, but they limit what can be concluded about comparative effects and leave the possibility of assessment influences. Further details about pathology review procedures are not reported in the registry record.

How should planning illustrations be distinguished from results?

The planning assumptions describe an illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. They specify that the effect to detect is a stated fraction of the observed effect. These assumptions are not described as the trial’s own design. The actual illustration values and the fraction used are not reported in the registry record.

Such sample-size calculations are planning illustrations, not observed power. They describe sample sizes under chosen assumptions; they do not establish the probability that a completed study detected an effect, and they cannot replace the outcome result. The illustration’s two-arm allocation assumption also does not change the registry’s description of this trial as single-group. No comparison arm or randomized allocation is listed for the trial itself.

What applies to survival outcomes?

Overall survival and progression-free survival are time-to-event outcomes. In general, these outcomes use both whether an event occurred and the time until it occurred, while accounting for participants whose event time is not observed during follow-up. Censoring methods are commonly used for that purpose. The registry lists these outcomes but reports no analyses, event counts, follow-up summaries, survival estimates, or handling of censoring. None of those results can be inferred from the endpoint names alone.

Interpretation of a time-to-event analysis also depends on the precise event definition and follow-up period. Those details and the statistical model, if any, are not reported in the registry record. The listed outcomes should therefore be treated as outcome categories, not as evidence of a particular survival or progression 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("Chemo Plus RT, Surgery, Chemo")
n      <- c(93)
events <- c(26)
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  Chemo Plus RT, Surgery, Chemo  93  26  28.0  19.1  38.2  19.9  37.8  19.9  37.8

11. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary outcome and several secondary outcomes, but a plan for managing multiple statistical tests is not reported in the registry record. As a general statistical concept, examining several outcomes can increase the chance of obtaining at least one apparently notable result by chance if each is tested separately. A prespecified hierarchy or adjustment can address this issue, but the record does not describe such a procedure. It also reports no analyses or outcome-specific statistical results.

The primary endpoint definition includes a specific rule for patients who did not undergo surgery: they are counted as not having responded. This is an explicit part of the registry’s response classification. Other missing-data procedures are not reported in the registry record, including how unavailable pathology assessments, incomplete follow-up, or missing covariates were handled. The number of participants with missing endpoint information is not reported.

The record names eligible patients as the primary endpoint population but does not describe analysis populations such as intention-to-treat or per-protocol populations. It does not report how many enrolled participants were included in each outcome analysis or whether participants were excluded. The enrollment figure alone does not provide those analysis denominators.

For overall survival and progression-free survival, the registry does not report follow-up duration, censoring rules, or the number of events. These details are important for interpreting time-to-event estimates. Without analysis methods and results, neither missing-data assumptions nor multiplicity handling can be evaluated. The absence of these details in the registry record does not demonstrate that a particular method was or was not used; it means the methods are not reported there.

12. Safety Results

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

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: adult respiratory distress syndrome (ards) (2.2%), hypoxia (2.2%), cardiac-ischemia/infarction (1.1%), pericardial effusion (non-malignant) (1.1%), svt and nodal arrhythmia - atrial fibrillation (1.1%).

The registry safety summary records 93 participants at risk for serious adverse events, with 7 affected. For other adverse events, it records 93 at risk and 93 affected. These are separate summary categories; the record does not describe overlap between them or provide further details about how event counts were classified. The listed primary safety outcome refers to grade 3 through 5 adverse events related to study drug, but the summary figures alone do not show which reported events met that grade range or were judged related to study drug.

Selected serious adverse-event terms are reported with percentages: adult respiratory distress syndrome (ARDS), 2.2%; hypoxia, 2.2%; cardiac-ischemia/infarction, 1.1%; pericardial effusion (non-malignant), 1.1%; SVT and nodal arrhythmia - atrial fibrillation, 1.1%; and typhlitis (cecal inflammation), 1.1%. These are reported term-specific percentages. The registry does not give further details on whether a participant could contribute to more than one term or on the calculation basis for each percentage beyond the group safety summary.

The safety entries are descriptive. They do not establish causation, and the term list is not presented as an exhaustive account of all events. The death count is not reported in the registry record; a missing death figure should not be interpreted as zero.

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

The registry record establishes that this was a phase II, single-group study of a treatment sequence for stage II or III esophageal or gastroesophageal junction cancer. It defines pathological complete response and describes the timing of assessment. It also records selected adverse-event counts and terms. These details provide context for understanding what the study was designed to assess and how some outcomes were classified.

The primary response result is not reported in the registry record. Consequently, the record does not show how many patients met the pathological complete response definition, what proportion did so, or how precisely that result was estimated. The listed secondary outcomes of overall survival, progression-free survival, and grade 3 through 5 adverse events related to study drug also lack reported analyses and numerical results.

Even if an outcome frequency were available, the single-group design would support a description of outcomes in the study group, not a randomized comparison of treatment strategies. A response observed after a treatment sequence is not, on its own, proof that the sequence caused the response or changed outcomes relative to an alternative. The registry does not provide a concurrent comparator for such a contrast.

The safety figures should likewise be read as reported event summaries, not as proof that treatment caused every event. The record does not give enough detail to determine the relationship between all reported serious events and the listed drug-related safety outcome, nor does it report a death count.

In short, the registry describes the trial’s design, treatment sequence, endpoint definition, and selected safety information. It does not provide the efficacy results or statistical analyses needed to quantify the primary endpoint or assess survival outcomes. The available record therefore supports interpretation of what was measured and how the study was structured, but not a conclusion about the magnitude or precision of treatment outcomes.

14. Why This Trial Matters Statistically

This trial illustrates why a clear endpoint definition and a clear account of study design are essential to statistical interpretation. Pathological complete response is defined by pathology review after chemoradiotherapy and surgery, and the registry specifies that patients who did not receive surgery count as nonresponders. That rule makes the outcome classification explicit, while also linking response status to whether participants reached surgery.

The single-group design is equally important. It can describe outcomes in the enrolled group, but it does not create a randomized comparison. Recognizing this distinction prevents an observed outcome frequency from being mistaken for a treatment effect relative to another strategy. The registry’s allocation and masking entries also help clarify what can and cannot be inferred from the study structure.

The record demonstrates another practical issue: an outcome may be named and defined without its numerical result or analysis being reported. Here, the primary endpoint and secondary outcomes are listed, but the response result and analyses are not reported in the registry record. Without those results, readers cannot assess the response frequency, its precision, or the reported survival outcomes. A design description is useful, but it cannot substitute for outcome reporting.

Finally, the safety summary shows why event categories and attribution must be kept distinct. Counts for serious and other adverse events, selected event terms, and a listed drug-related safety outcome are not interchangeable. The available information supports a careful description of the recorded safety entries, but does not establish causation or provide every detail needed to reconcile those categories.

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 for statistical interpretation is that the registry record does not report the primary endpoint result. It also reports no analyses. Without response counts, denominators for the endpoint analysis, or an uncertainty estimate, the frequency and precision of pathological complete response cannot be evaluated.

The design is single-group, with no listed concurrent comparator. This limits the study’s ability to estimate a comparative treatment effect from the registry information. The record lists masking as none and does not report details of pathology review procedures, so the extent of any blinding or safeguards in outcome assessment is not described.

The registry records enrollment of 98, but does not report how many participants underwent surgery, how many were included in each outcome analysis, or how many had missing outcome information. Although patients who did not receive surgery are counted as nonresponders by definition, the number and reasons for not receiving surgery are not reported. Other missing-data rules and analysis populations are also not reported.

The secondary outcomes have no reported numerical results or analyses. For overall survival and progression-free survival, the record does not report event counts, follow-up summaries, or censoring methods. For safety, selected counts and serious adverse-event terms are available, but the record does not report a death count or enough detail to establish how the summary entries relate to the drug-related grade 3 through 5 outcome.

The statistical analysis plan, multiplicity procedures, and molecular-correlate results are not reported in the registry record. These limitations concern what can be assessed from the registry information; they do not establish what analyses or procedures may have been conducted outside the reported record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: S0356 Oxaliplatin, 5-FU, Radiation Therapy (RT), Surgery for Pts With Stage II or III Cancer of Esophagus or Gastroesophageal (GE) Junction — NCT00086996.
• Registry record last updated 2013-10-11; status last verified 2013-09.
• Results first posted 2012-04-03.
• 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.

19. Record Summary

The registry describes a completed phase II, single-group trial sponsored by SWOG Cancer Research Network, with enrollment recorded as 98. The treatment sequence was neoadjuvant fluorouracil, oxaliplatin, and radiation therapy, followed by conventional surgery and adjuvant fluorouracil and oxaliplatin. The primary endpoint was pathological complete response, assessed 10-16 weeks after beginning treatment; patients who did not receive surgery were classified as not having responded.

The registry lists grade 3 through 5 adverse events related to study drug, overall survival, and progression-free survival as secondary outcomes. It reports selected safety figures, including 7 participants affected by serious adverse events among 93 at risk. The primary response result, secondary outcome analyses, and statistical analysis plan are not reported in the registry record. The single-group design does not provide a randomized comparator.