This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Irinotecan and Cisplatin in Treating Patients Who Are Undergoing Surgery For Locally Advanced Cancer of the Stomach or Gastroesophageal Junction (NCT00062374), a na phase 2 study. The primary outcome is Histological Response Determined by FDG Uptake Correlates.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT00062374 |
| Phase | Phase 2 |
| Condition | Gastric Adenocarcinoma, Stage II Gastric Cancer, Stage III Gastric Cancer |
| Allocation | Not applicable (single group) |
| Design model | Single Group |
| Masking | None (open label) |
| Enrollment | 55 |
| Arms in the primary outcome | n/a |
| Lead sponsor | National Cancer Institute (NCI) |
| Results posted | Yes |
2. Clinical Question
Population
The study concerned patients undergoing surgical resection for locally advanced cancer of the stomach or gastroesophageal junction. The registry lists gastric adenocarcinoma and stage II, stage III, and stage IV gastric cancer among the conditions. Enrollment was 55. The registry does not report the participant characteristics needed to describe who contributed usable imaging or pathology data.
Intervention
The treatment was preoperative chemotherapy with cisplatin and irinotecan. The regimen description specifies intravenous doses of both drugs on days 1, 8, 22, and 29, with treatment repeating every 6 weeks for 2 courses if there was no disease progression or unacceptable toxicity. Surgery was planned within 4 weeks after completion of neoadjuvant chemotherapy.
Comparator
This was a single-group study, not a randomized comparison. The registry does not describe a separate control group receiving another treatment or no preoperative chemotherapy. The planned statistical comparison was instead between participants classified as pathological responders and nonresponders, using early change in FDG-SUV.
Primary question
The primary objective was to determine whether a decrease in FDG-SUV early during treatment discriminated pathological response from nonresponse. Pathological response was based on microscopic assessment of residual cancer cells and fibrosis. The registry says the FDG-SUV measurements were unreliable or uninterpretable and that response adjudication was invalid and needed to be repeated; therefore, only histological response is reported, without a numerical result in the registry record.
3. Trial Design
This was a phase II, single-group study with no masking. Allocation is listed as not applicable, consistent with the absence of a randomized treatment comparison. The registry records 55 participants enrolled. The study is marked completed; its recorded start was 2003-06 and its primary completion was 2011-06. The sponsor was the National Cancer Institute.
The treatment plan combined cisplatin and irinotecan before surgery. The regimen description specifies each drug intravenously over 30 minutes on days 1, 8, 22, and 29, with the treatment repeated every 6 weeks for 2 courses in the absence of disease progression or unacceptable toxicity. Surgery was planned within 4 weeks after completion of neoadjuvant chemotherapy. The registry description does not give further surgical details in the record represented.
The study also lists computed tomography, positron emission tomography, fludeoxyglucose F-18, fluorothymidine F-18, and therapeutic conventional surgery among its interventions. The primary objective focused on whether an early decrease in FDG-SUV could distinguish pathological responders from nonresponders. The registry describes a two-sample t-test for that comparison, but says technical problems made FDG-SUV data unreliable or uninterpretable and that response adjudication was invalid and needed re-adjudication.
These design features limit the kind of inference available. A single-group study can describe outcomes among enrolled participants, but without a concurrent comparison group it cannot isolate a treatment contrast against another strategy. In addition, the central planned relationship between imaging change and pathological response cannot be evaluated from the record because the relevant data were not considered valid. Numerical histological response results 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
| Stage | Arm I |
|---|---|
| STARTED | 55 |
| COMPLETED | 43 |
| NOT COMPLETED | 12 |
Reasons for not completing this period:
| Reason | Arm I |
|---|---|
| Not treated | 12 |
Analysis population for the primary outcome, as defined in the registry: Technical problems with measurement of FDG-SUV data leading to unreliable or uninterpretable data. Data adjudication was invalid, and needs to be re-adjudicated. Therefore, only histological response is reported.
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.
| Characteristic | Unit | Arm I |
|---|---|---|
| Participants analysed | 55 | |
| Age, Categorical: <=18 years | Participants | 0 |
| Age, Categorical: Between 18 and 65 years | Participants | 35 |
| Age, Categorical: >=65 years | Participants | 20 |
| Sex: Female, Male: Female | Participants | 23 |
| Sex: Female, Male: Male | Participants | 32 |
| Region of Enrollment: United States | participants | 55 |
6. Primary Endpoint
The registry lists the primary outcome as Histological Response Determined by FDG Uptake Correlates, measured over Day 15.
| Element | Registry entry |
|---|---|
| Result type | Number |
| Measure of spread | not applicable |
| Unit | participants |
| Endpoint classification | Binary Endpoint |
The primary objective concerned whether change in FDG-SUV early in treatment could discriminate between pathological responders and nonresponders. The registry identifies the timeframe as Day 15. Pathological response was defined from microscopic inspection for residual cancer cells and fibrosis: nonresponse was Tumor Regression Grade 3 or higher, while response was Tumor Regression Grade 1 (CR) or Grade 2 (PR).
In general, an imaging measure can be studied as an early indicator of a later tissue-based outcome. FDG uptake provides an imaging signal, while the pathological classification is based on tissue examined under a microscope. The research question was whether the early imaging change differed between the two pathology-defined groups. The registry describes a two-sample t-test for this purpose.
The endpoint cannot be interpreted as a demonstrated relationship in this record. The registry states that technical problems led to unreliable or uninterpretable FDG-SUV data and that adjudication of response was invalid and required re-adjudication. It consequently says that only histological response is reported. However, it does not give the number of participants in each response category or a numerical estimate of the imaging-response relationship. Those results are not reported in the registry record.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The main interpretive fact is that the registry does not present a usable result for the primary imaging question. The objective was to test whether an early decrease in FDG-SUV separated pathological responders from nonresponders. The record says the imaging measurements were unreliable or uninterpretable and the response adjudication was invalid and needed to be repeated. It therefore does not support a numerical estimate of how well early FDG-SUV change discriminated pathological response.
The registry states that only histological response is reported, but no response count, proportion, or estimate is included in the registry record recorded for this analysis. The number of participants with each pathology classification, and the number with usable tissue assessments, are not reported in the registry record. Enrollment of 55 should not be read as the number included in an endpoint analysis, since the analysis population is not described.
The planned two-sample t-test would compare early SUV decrease between the pathological responder and nonresponder groups. In general, that test addresses whether the group means differ, subject to its assumptions and the quality of the underlying measurements. It would not by itself establish that the imaging measure predicts response accurately for an individual, nor would it establish a causal treatment effect. In this case, the registry does not provide a valid test result or enough numerical information to assess the size or precision of any difference.
The single-group design also matters. There is no like-for-like control arm in the registry record, so the study cannot furnish a randomized comparison of preoperative chemotherapy against another treatment approach. The intended responder-versus-nonresponder comparison is a biomarker-analysis question within the treated group, not a comparison of assigned treatments. No adjusted or unadjusted analysis result is reported; the record lists no analyses.
For safety, the record lists 14 participants with serious adverse events among 55 at risk, and 44 participants with other adverse events among 55 at risk. Serious-event terms and their reported percentages are listed separately, but these figures do not supply a comparative safety estimate. The record does not report deaths. Counts and percentages describe what was entered in the registry; they should not be treated as proof of causation or as a comparison with another regimen.
Overall, the record documents the study question, design, treatment plan, and safety counts, but does not provide numerical efficacy results for the primary question. The invalid imaging and pathology adjudication is central to interpretation, rather than a minor uncertainty around an otherwise estimable effect.
9. Statistical Methods Explained
The methods described in the record need to be distinguished from results that can actually be interpreted. The stated objective was to assess whether a decrease in FDG-SUV early in treatment differed between pathological responders and nonresponders. The registry names a two-sample t-test for that comparison. It also states that technical problems affected the imaging measurements and that response adjudication was invalid. No test statistic, estimate, confidence interval, or p-value is reported in the registry record, and the record lists no analyses.
What the planned two-sample test addresses
In general, a two-sample t-test compares the average value of a continuous measure in two groups. Here, the groups were defined by pathological response status, and the continuous measure of interest was early decrease in FDG-SUV. A test of group means asks whether the observed difference is compatible with a null hypothesis of no difference, under the test assumptions. The registry describes the proposed test but gives no result from it. Without valid measurements and valid group adjudication, the intended comparison cannot be interpreted.
In general, a p-value measures how compatible the data are with a specified null model; it is not the probability that the null hypothesis is true, and it does not describe the size or clinical importance of an effect. A confidence interval, when reported, gives a range of estimates compatible with the data and model at a stated confidence level. Neither a p-value nor a confidence interval for the imaging comparison is reported in the registry record. There is therefore no basis to characterize the magnitude or precision of the planned difference.
Why measurement quality is essential
A statistical test cannot repair unreliable input data. If an imaging value is unreliable or uninterpretable, the calculated group summaries may not represent the underlying imaging signal. Likewise, if pathology classifications are invalid, participants may be placed in the wrong response group or the endpoint may not be consistently defined. General statistical methods assume that variables and groups have been measured and classified adequately. The registry specifically flags problems with both the FDG-SUV data and adjudication, so the intended association should not be treated as established.
General concepts such as missingness and measurement error are relevant, but the record does not quantify how many observations were affected, why individual observations were unusable, or how any such cases were handled. It does not report a sensitivity analysis or an alternative analysis plan. Those details are not reported in the registry record.
Single-group design and the target of inference
In general, a single-group study can describe outcomes in the enrolled group but lacks a concurrent randomized control for estimating the effect of assigning one treatment rather than another. In this study, the planned comparison was between pathological response categories among participants receiving the preoperative regimen. That comparison is not a randomized treatment contrast. Even if a difference in early imaging change between response groups had been validly estimated, the result would concern association with the pathology classification, not a causal effect of treatment or a direct comparison with another regimen.
The record lists allocation as not applicable, masking as none, and the model as single group. It does not report further procedures to reduce bias in image assessment, pathology assessment, or analysis. No assumptions about such procedures can be made from the record.
Planning illustrations and power
General sample-size planning asks how many participants would be needed to detect an assumed effect under specified statistical assumptions. Any planning illustration associated with this study is not an observed-power calculation and should not be read as evidence that the study did or did not detect an effect. The stated illustration assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect to detect expressed as a fraction of the observed effect. For time-to-event outcomes, the stated convention is a total number of events; for other outcomes, the convention is participants per arm.
Those assumptions describe a hypothetical planning exercise, not the design of this single-group trial. In particular, equal allocation between two arms does not change the registry’s single-group design. The record does not report the trial’s own sample-size justification, and no observed effect estimate for the primary imaging comparison is available to interpret. Planning numbers cannot replace valid endpoint data, and they are not observed power.
More generally, precision depends on the amount and quality of usable data, as well as variation in the measure. A confidence interval would communicate uncertainty around an estimate, while an analysis based on invalid measurements could give misleading precision even if calculated. Since the registry does not report numerical endpoint analyses, neither the statistical significance nor the precision of the intended primary comparison can be assessed.
10. Multiplicity, Missing Data and Analysis Populations
The registry describes one primary objective: to assess whether early decrease in FDG-SUV discriminated between pathological responders and nonresponders. It describes a two-sample t-test for that comparison. It does not report whether additional confirmatory hypotheses were prespecified, whether any multiplicity adjustment was planned, or how multiple outcomes or analyses would be handled. Those details are not reported in the registry record.
The record lists 2 outcome measures and no analyses. The primary description states that only histological response is reported after technical problems with FDG-SUV measurement and invalid response adjudication. It gives no numerical histological response results, analysis estimates, or uncertainty intervals. The criteria for response and nonresponse are described, but the number classified in each group is not reported.
Missing-data procedures are also not described. The record does not state how many participants had unusable imaging or pathology data, whether missing measurements were excluded, or whether any imputation or sensitivity analysis was used. The stated data problems make these details important: the usable analysis population may differ from the enrollment count, but its size is not reported.
The registry does not describe an intention-to-treat, per-protocol, or other analysis population for the primary objective. Nor does it give an analysis plan addressing exclusions or repeated measurements. These omissions prevent assessment of how the final analysis population was selected and whether the intended testing approach was followed.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Arm I | 14 | 55 | 25.5% |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 5% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Arm I | 44 of 55 |
Most frequent serious adverse event terms
| Event term | Organ system | Arm I |
|---|---|---|
| Thrombosis | Vascular disorders | 5 of 55 |
| Febrile neutropenia | Blood and lymphatic system disorders | 3 of 55 |
| CNS/cerebrovascular ischemia | Nervous system disorders | 2 of 55 |
| Chest pain | General disorders | 2 of 55 |
| Diarrhea | Gastrointestinal disorders | 2 of 55 |
| Hemorrhage, NOS | Blood and lymphatic system disorders | 2 of 55 |
| Neutrophil count decrease | Investigations | 2 of 55 |
| Abdominal pain/cramping | Gastrointestinal disorders | 1 of 55 |
| Cardiovascular, other | Cardiac disorders | 1 of 55 |
| Creatinine increased | Investigations | 1 of 55 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Arm I |
|---|---|---|
| Fatigue | General disorders | 36 of 55 |
| Hypocalcemia | Metabolism and nutrition disorders | 35 of 55 |
| Nausea | Gastrointestinal disorders | 33 of 55 |
| White blood cell count decrease | Investigations | 32 of 55 |
| Neutrophil count decrease | Investigations | 31 of 55 |
| Hemoglobin decrease | Blood and lymphatic system disorders | 30 of 55 |
| Diarrhea | Gastrointestinal disorders | 25 of 55 |
| Platelet count decreased | Investigations | 21 of 55 |
| Hyponatremia | Metabolism and nutrition disorders | 20 of 55 |
| Hypophosphatemia | Metabolism and nutrition disorders | 17 of 55 |
Serious events with the highest arm-level rate: thrombosis (9.1%), febrile neutropenia (5.5%), cns/cerebrovascular ischemia (3.6%), chest pain (3.6%), diarrhea (3.6%).
The registry lists 55 participants at risk for both serious adverse events and other adverse events. It records 14 participants affected by serious adverse events and 44 affected by other adverse events. These are participant counts, not counts of individual event episodes. A participant may have experienced more than one event, so the figures should not be added or treated as mutually exclusive categories.
The serious-event terms with reported percentages are thrombosis, 9.1%; febrile neutropenia, 5.5%; CNS/cerebrovascular ischemia, 3.6%; chest pain, 3.6%; diarrhea, 3.6%; and hemorrhage, NOS, 3.6%. The registry does not provide the event counts behind each listed percentage in the registry record recorded for this analysis, and it does not indicate whether participants could appear under more than one term. The percentages should therefore be read as term-specific registry entries, not as a breakdown that can be summed into a total.
No deaths are reported in the record. These safety figures are descriptive for the study group; with no control group, they do not estimate a comparative difference in risk or establish that a particular event was caused by treatment.
12. What the Result Does, and Does Not, Mean
The record establishes that a phase II, single-group study enrolled 55 participants to examine preoperative irinotecan and cisplatin in locally advanced stomach or gastroesophageal-junction cancer. Its central statistical question was whether an early decrease in FDG-SUV could distinguish pathological responders from nonresponders. The registry also documents the treatment schedule and adverse-event counts.
What the record does not establish is a numerical answer to that primary question. The FDG-SUV data were described as unreliable or uninterpretable, and response adjudication was invalid and needed re-adjudication. Although the registry says that only histological response is reported, no histological response count or estimate is included. There is no reported result from the intended two-sample t-test, and no analysis is listed. Thus neither the size nor the precision of the proposed imaging-response relationship can be determined from the registry record.
The design also does not support a randomized comparison of this regimen with another treatment. A single-group study has no concurrent control arm for estimating such a contrast. The planned comparison between response categories would have addressed whether imaging change differed by pathology classification within the studied group; it would not, by itself, demonstrate a causal effect of treatment.
In general, a statistically significant result, if one were reported, would not automatically establish that a measure is an accurate individual-level predictor or that treatment benefits patients. Conversely, the absence of a usable analysis in this record is not evidence that no relationship exists. It means the stated primary question cannot be answered from the reported results.
The safety entries describe events recorded among study participants, but without a comparator they do not show how those event frequencies compare with another treatment. The registry record is therefore informative about the study’s design and data-quality problem, while leaving the primary numerical efficacy question unresolved.
13. Why This Trial Matters Statistically
This study raises a statistical issue that is central to biomarker research: an intended relationship between an early measurement and a later outcome depends on the validity of both. The objective was to compare early FDG-SUV decrease across pathology-defined responder groups. The registry reports problems with the imaging data and invalid response adjudication. When either side of that comparison is unreliable, an analysis can no longer provide a dependable estimate of the relationship.
The record also illustrates the distinction between a planned method and an interpretable result. Naming a two-sample t-test specifies a type of comparison, but it does not supply evidence that the assumptions were met, that measurements were valid, or that the comparison was completed successfully. No analysis result is reported. Statistical interpretation must therefore remain focused on what the record can support rather than on the intended test alone.
Design affects the question a study can answer. This was a single-group trial, so it was not structured as a randomized treatment comparison. Its proposed responder-versus-nonresponder analysis was a within-study association question. Keeping those targets distinct helps avoid reading a biomarker comparison as evidence of comparative treatment effectiveness.
Finally, enrollment is not interchangeable with an endpoint analysis population. The record gives enrollment but does not say how many participants had valid imaging, valid pathology adjudication, or complete paired data. Such information is essential for assessing representativeness and precision. The registry’s explicit data-quality caveat and lack of numerical endpoint results make transparent reporting of these details especially important.
14. 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 most consequential trial-specific limitation is the stated technical problem with FDG-SUV measurement. The registry describes the data as unreliable or uninterpretable, which prevents a defensible reading of the intended imaging comparison. It also says that response adjudication was invalid and needed to be repeated. The record does not report whether re-adjudication occurred or what it found.
Although the registry says only histological response is reported, it does not give the number of responders or nonresponders, their pathology results, or the number with assessable tissue. The enrollment count of 55 does not identify the analysis population. The number with usable imaging or paired imaging and pathology data is not reported in the registry record.
The single-group design has no randomized control arm, so there is no like-for-like comparison for evaluating a treatment contrast. The registry also lists no analyses and gives no numerical estimate, test result, or precision measure for the primary objective. The statistical analysis plan beyond the described two-sample t-test is not reported.
Further details about missing-data handling, exclusions, protocol deviations, masking of assessments, and multiplicity procedures are not reported in the registry record. The record lists serious and other adverse-event participant counts, but does not provide a comparator or report deaths. These limitations restrict both efficacy interpretation and comparative safety interpretation.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Irinotecan and Cisplatin in Treating Patients Who Are Undergoing Surgery For Locally Advanced Cancer of the Stomach or Gastroesophageal Junction — NCT00062374.
- Registry record last updated 2017-07-06; status last verified 2017-06.
- Results first posted 2017-07-06.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
NCT00062374 is a completed phase II, single-group study sponsored by the National Cancer Institute. The registry records enrollment of 55 and a study start of 2003-06, with primary completion in 2011-06. The study concerned preoperative cisplatin and irinotecan for locally advanced gastric or gastroesophageal-junction cancer, followed by planned surgery.
The primary objective was to assess whether an early decrease in FDG-SUV distinguished pathological responders from nonresponders. The registry describes a two-sample t-test, but reports unreliable or uninterpretable FDG-SUV data and invalid response adjudication requiring re-adjudication. It says only histological response is reported; numerical response results and analyses are not reported in the registry record.
Safety entries record 14 participants with serious adverse events and 44 with other adverse events among 55 at risk. The record does not report deaths. The single-group design does not provide a randomized treatment comparison.