Chemotherapy, Surgery, and Radiation Therapy in Treating Patients With Gastric Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Chemotherapy, Surgery, and Radiation Therapy in Treating Patients With Gastric Cancer (NCT00003298), a na phase 2 study. The primary outcome is Grade 3 or Higher Toxicity Incidence on Step 1.

2. Clinical Question

3. Trial Design

This is a completed Phase II, single-group trial of a treatment sequence for gastric cancer. The registry lists an enrollment of 39, identifies no masking, and records allocation as not applicable. The study is not a randomized comparison, so there is no concurrent control group in the registry record against which to estimate a treatment difference.

The official title describes neoadjuvant paclitaxel-cisplatin chemotherapy, surgery, and adjuvant radiation therapy with 5-FU/leucovorin. The brief summary likewise describes preoperative chemotherapy followed by surgery and postoperative chemoradiation. The listed interventions include cisplatin, paclitaxel, fluorouracil, leucovorin calcium, surgery, and radiation therapy. The registry's experimental-arm description is incomplete, so details beyond those stated in the title, summary, and listed interventions are not reported in the registry record.

The planned preoperative treatment is described as three courses, given on day 1 every 21 days, with cisplatin and paclitaxel infusions. The primary endpoint timeframe says toxicity was assessed at the end of every cycle during treatment. The brief summary says the study investigated whether the approach would be well tolerated and have a high curative resection rate; these are stated aims, not results.

The registry lists overall survival, progression-free survival, and best confirmed response to neoadjuvant therapy as secondary outcomes. It records four outcome measures, but the results for these measures are not reported in the registry record. The study start is listed as 1999-06-01, primary completion as 2010-02, and status as completed.

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.

Step 1 (Neoadjuvant Therapy)

Reasons for not completing this period:

Surgery

Reasons for not completing this period:

Re-register to Step 2 After Surgery

Reasons for not completing this period:

Step 2 (Adjuvant Therapy)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: eligible and treated patients on step 1

5. Baseline Characteristics

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

Population: Eligible and treated in step 1

6. Primary Endpoint

The registry lists the primary outcome as Grade 3 or Higher Toxicity Incidence on Step 1, measured over assessed at the end of every cycle (cycle=21 days) during treatment (3 cycles in total).

The primary endpoint is the incidence of grade 3 or higher treatment-related toxicity on step 1. The registry defines incidence as the proportion of patients with any such toxicity among all treated patients, and identifies the relevant population as eligible and treated patients on step 1. Assessment was planned at the end of every cycle during the three-cycle treatment course.

This endpoint summarizes the occurrence of severe treatment-related toxicities during the neoadjuvant stage. The “any” definition means that the endpoint concerns whether a patient experienced at least one qualifying toxicity, rather than the total number of separate toxicity events. That distinction matters: a patient with multiple qualifying toxicities still contributes to a proportion of patients with an event, rather than necessarily contributing multiple times to its numerator.

The registry specifies a 95% confidence interval as the endpoint's dispersion measure. A confidence interval communicates uncertainty around an estimated proportion; it does not eliminate uncertainty or establish that the estimate applies identically to other patient groups. Neither the primary endpoint estimate nor its interval is reported in the registry record. The serious adverse event figures listed for treatment steps are separate safety summaries and should not be substituted for the registered primary endpoint.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes the primary outcome as a percentage of participants with an event. Its event is any treatment-related toxicity of grade 3 or higher, and its population is eligible and treated patients on step 1. The measure is therefore a proportion of patients meeting the endpoint definition, not a count of all toxicity episodes. The registry gives the assessment schedule as the end of every cycle during the three cycles of treatment.

What does a confidence interval convey?

General statistical explanation: a confidence interval expresses uncertainty around an estimated quantity. For a proportion, the estimate describes the observed share of the defined population meeting the endpoint, while the interval conveys how imprecisely that share is estimated. A narrower interval indicates greater precision than a wider interval, but does not address whether the endpoint is clinically important or whether the study design supports a causal comparison. The registry identifies a 95% confidence interval for the primary endpoint, but does not report the estimate or its interval.

In a single-group study, an interval around an event proportion describes uncertainty about that group's observed outcome under the endpoint definition. It is not an interval for a difference against a control group, because the registry does not identify a separate control group. A confidence interval should also be interpreted with attention to the endpoint population and follow-up schedule: an estimate for eligible and treated patients on step 1 is not automatically an estimate for all enrolled patients or for patients receiving the later treatment step.

Why distinguish patients from events?

General statistical explanation: patient-level incidence and event counts answer different questions. An incidence measure asks whether a patient experienced an event that satisfies the definition. A count of events can include repeated occurrences in the same patient. When the endpoint is defined as patients with any qualifying toxicity, the relevant unit is the patient, even if a patient experiences more than one qualifying toxicity. The registry definition explicitly uses any grade 3 or higher treatment-related toxicity among treated patients.

That endpoint should not be conflated with the serious adverse event summary. Seriousness and toxicity grade are distinct classifications, and the record does not say that every serious adverse event meets the primary endpoint definition or that every primary endpoint event is a serious adverse event. Nor does the record provide enough detail to map the serious-event terms to the primary endpoint's treatment-related and grade criteria.

What does the study design permit?

General statistical explanation: a single-group design can describe outcomes within the treated group, but it does not provide a randomized control group for estimating a comparative treatment effect. Without a like-for-like control, differences from other studies or populations can reflect differences in eligibility, care, assessment, or follow-up as well as treatment. The registry explicitly identifies the study as single group and notes that it is not a randomized comparison.

The trial is Phase II and the registry describes a treatment strategy that includes neoadjuvant therapy and later surgery and postoperative treatment. The stages have separate safety counts in the record. The step 2 figures concern the patients listed at risk at that stage, and should not be read as a direct comparison with step 1: the record does not describe step assignment as randomized, and the risk sets differ. The registry does not report why the numbers at risk differ between treatment steps.

What can planning calculations tell readers?

General statistical explanation: a sample-size calculation is a planning exercise that describes what a future design might require under specified assumptions. It is not an observed study result and does not measure the probability that a completed trial found a true effect. The record's sample-size assumptions describe an illustration based on an observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; it says that the effect to detect is a stated fraction of the observed effect. It also distinguishes total events for time-to-event illustrations from participants per arm for other illustrations.

Those assumptions describe planning illustrations, not the registered single-group design. In particular, an equal-allocation, two-arm planning framework does not create a comparator in this trial. The observed effect used for any illustration, the resulting planning values, and a trial-specific sample-size justification are not reported in the registry record. These illustrations must not be interpreted as observed power or as evidence about the trial's results.

What is not available?

The registry records zero posted analyses. It does not report the primary endpoint estimate, its confidence interval, or the statistical procedure used to calculate that interval. The record also does not report an analysis plan, handling of missing observations, or details of any interim analyses. Without those details and the outcome results, readers cannot assess the implementation of the planned analysis or reproduce it from this record.

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("Experimental Arm")
n      <- c(38)
events <- c(25)
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  Experimental Arm  38  25  65.8  48.6  80.4  49.9  78.8  49.8  78.8

11. Multiplicity, Missing Data and Analysis Populations

The registry names a primary toxicity endpoint and lists secondary outcomes of best confirmed response to neoadjuvant therapy, overall survival, and progression-free survival. The record also identifies four outcome measures. It does not report how the primary and secondary outcomes were ordered for inferential testing, whether any adjustment for multiple testing was planned, or how statistical conclusions across several outcomes were to be controlled. That multiplicity plan is not reported in the registry record.

The registry records zero analyses, and results for the listed outcome measures are not reported in the registry record. It therefore is not possible to determine which analyses were conducted, whether they followed a prespecified plan, or how any outcome-level estimates were interpreted. The record's sample-size assumptions describe a planning illustration, not a reported analysis plan for this single-group trial.

Missing-data methods are also not reported in the registry record. There is no description of how patients with incomplete toxicity assessments, early discontinuation, or missing follow-up would enter the primary or secondary analyses. The record does not state whether the primary analysis included every eligible and treated patient with available information or used another approach; the endpoint population is described, but the handling of incomplete observations is not.

The registered primary population is eligible and treated patients on step 1. The registry enrollment is 39, while the serious adverse event summary lists 38 at risk for step 1 and 7 at risk for step 2. The record does not explain the relationship among these counts or specify whether the safety risk sets match the primary endpoint analysis population. These gaps limit interpretation of the endpoint denominator and the extent to which results would apply to all enrolled patients.

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: neutrophil count decreased (71.4%), white blood cell decreased (42.9%), vomiting (42.9%), anorexia (28.6%), dehydration (28.6%).

The registry lists serious adverse event figures separately for step 1 and step 2. For step 1, 38 are listed at risk and 25 as affected; for step 2, 7 are listed at risk and 6 as affected. These are the recorded serious-event counts for each treatment stage. They should not be treated as a head-to-head comparison: the study is single group, and the record does not describe random assignment to treatment steps.

The listed serious-event terms are neutrophil count decreased at 71.4%, white blood cell decreased at 42.9%, vomiting at 42.9%, anorexia at 28.6%, dehydration at 28.6%, and esophagitis at 28.6%. The registry does not report the denominators for these term percentages or identify the treatment step they describe. They cannot be used to calculate the primary toxicity endpoint, which has its own definition and population.

The death field is null for both steps. This means the record does not report a death count; it should not be interpreted as a count of zero. The record also does not provide further detail about event timing, severity distribution, or how serious events relate to the primary endpoint's grade and treatment-relatedness criteria.

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

The trial record describes a Phase II evaluation of neoadjuvant paclitaxel and cisplatin followed by surgery and planned postoperative therapy for gastric cancer. Its primary endpoint is the incidence of any treatment-related toxicity of grade 3 or higher among eligible and treated patients on step 1, assessed at the end of each cycle during the three-cycle treatment course. Those details define the question the study set out to address.

The record does not report the primary endpoint result. It therefore cannot establish from the recorded information how often the defined toxicity occurred, or how precise an estimate would be. The 95% confidence interval is named, but its bounds and the endpoint estimate are absent. The listed serious adverse event counts and percentages do not fill that gap because they are separate summaries and do not have enough recorded detail to equate them with the primary endpoint.

The brief summary says the trial was conducted to study the treatment sequence and that investigators hypothesized it would be well tolerated and have a high curative resection rate. Those statements describe the study rationale and hypothesis, not evidence that the treatment was well tolerated or that resection outcomes met an expectation. The curative resection result is not reported in the registry record.

Likewise, overall survival, progression-free survival, and best confirmed response are named as secondary outcomes, but their results are not reported in the registry record. No conclusion about the treatment's effects on those outcomes follows from their listing alone. Null death fields in the safety summary do not establish that there were no deaths.

Finally, the single-group design does not offer a randomized control comparison. Even if outcome estimates were available, this design would not by itself establish what would have happened to a comparable group under a different treatment. The defensible reading is limited to the study question, design, and recorded safety figures; treatment benefit, comparative effect, and the missing efficacy outcomes cannot be determined from this record.

14. Why This Trial Matters Statistically

This trial is statistically instructive because its central question is a patient-level toxicity proportion, while the registry also presents serious adverse event summaries by treatment stage. Those measures are related to safety but are not interchangeable. The primary endpoint is defined by treatment relationship and toxicity grade, and counts patients with any qualifying event. The serious-event summaries have their own classification and risk sets. Keeping these measures distinct is essential to avoid treating one safety summary as if it answered the primary endpoint question.

The study also illustrates what a single-group Phase II design can and cannot establish. Such a design can describe outcomes within the treatment pathway under study. It does not provide a randomized control group for an unbiased direct comparison of treatment strategies. The fact that the registered hypothesis concerned tolerability and curative resection does not make those outcomes observed findings, and the record does not report their results.

Precision is another important part of interpretation. The registry identifies a 95% confidence interval for the primary toxicity endpoint, but does not report the estimate or interval. Readers therefore cannot assess how uncertain the endpoint estimate is. A named confidence level is not a substitute for the interval itself.

Finally, the registry separates stages of treatment and lists different numbers at risk for step 1 and step 2. These counts should not be treated as comparable groups. The record does not explain the movement between stages or how missing observations were handled. This makes clear why population definitions, denominators, and analysis details matter alongside an endpoint's label.

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 interpreting the registered question is that the primary endpoint result is not reported in the registry record. Although the endpoint definition, assessment schedule, and 95% confidence interval measure are identified, the estimate and interval are absent. The registry also does not report results for the secondary outcomes of best confirmed response, overall survival, or progression-free survival.

The trial is a single-group study rather than a randomized comparison. As a result, the registry record does not support a direct comparison with a concurrent control group. The treatment sequence includes preoperative therapy, surgery, and planned postoperative therapy, but the arm description is incomplete. Full regimen details beyond the treatment elements in the title, summary, and intervention list are not reported in the registry record.

The registry enrollment is 39, whereas the serious-event summary lists 38 at risk on step 1 and 7 at risk on step 2. The record does not explain the differences or establish that these safety denominators correspond to the primary endpoint population. The listed serious-event term percentages have no reported denominators or specified treatment step, and the death fields are null rather than reported counts.

Further statistical details are not reported in the registry record. These include the primary endpoint analysis procedure, missing-data handling, the multiplicity approach for the listed outcomes, and any interim analyses. The record lists zero analyses. Consequently, the planned analysis cannot be evaluated against posted results, and the available information does not permit an assessment of estimate precision, robustness, or generalizability beyond the endpoint and populations described.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Chemotherapy, Surgery, and Radiation Therapy in Treating Patients With Gastric Cancer — NCT00003298.
• Registry record last updated 2023-07-06; status last verified 2023-06.
• Results first posted 2015-06-23.
• 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 record is for a completed Phase II, single-group study of neoadjuvant paclitaxel and cisplatin followed by surgery and planned postoperative radiation and chemotherapy for gastric cancer. The primary endpoint is the percentage of eligible and treated patients on step 1 with any treatment-related toxicity of grade 3 or higher. It was assessed at the end of each cycle during the three-cycle treatment course, with a 95% confidence interval identified as the dispersion measure.

The registry lists serious adverse event counts for step 1 and step 2 and six serious-event terms with percentages. The primary endpoint estimate, confidence interval, and results for the named secondary outcomes are not reported in the registry record. The study has no randomized control group, and additional analysis, missing-data, and multiplicity details are not reported.