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 Capecitabine With or Without Sunitinib Malate as First-Line Therapy in Treating Patients With Metastatic Cancer of the Esophagus or Gastroesophageal Junction (NCT00891878), a randomized phase 2 study. The primary outcome is Comparison of Progression-free Survival.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT00891878 |
| Phase | Phase 2 |
| Condition | Adenocarcinoma of the Gastroesophageal Junction, Esophageal Cancer |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None (open label) |
| Enrollment | 12 |
| Arms in the primary outcome | 2 |
| Lead sponsor | Alliance for Clinical Trials in Oncology |
| Results posted | Yes |
2. Clinical Question
Population
The trial concerned patients with metastatic adenocarcinoma of the esophagus or gastroesophageal junction. Its official title describes elderly and/or poor performance status patients, but detailed eligibility criteria and the age or performance-status characteristics of participants are not reported in the registry record. The registry lists 12 participants enrolled.
Intervention
Arm B received capecitabine with sunitinib malate. The registry describes oral capecitabine on days 1–14 and oral sunitinib malate once daily on days 1–21. It does not state the doses in the posted arm description.
Comparator
Arm A received oral capecitabine on days 1–14. The registry says that patients whose disease progressed could cross over to Arm B at the physician’s discretion. The number who crossed over is not reported in the registry record.
Primary question
The primary question was whether progression-free survival differed between the capecitabine-alone and combination arms. Progression-free survival was summarized by median time in months, with a 95% confidence interval for each arm. The registry describes a one-sided log-rank comparison of the Kaplan–Meier curves; the posted analysis gives a log-rank p-value of 0.43.
3. Trial Design
This was a randomized, open-label, parallel-group phase II trial. The registry lists 12 participants enrolled and identifies the trial as completed. Its stated purpose was to compare capecitabine alone with capecitabine plus sunitinib malate as first-line therapy for metastatic cancer of the esophagus or gastroesophageal junction. The official title specifies metastatic adenocarcinoma and describes the intended population as elderly and/or having poor performance status. The enrolled participants’ characteristics and the detailed eligibility criteria are not reported in the registry record.
Arm A received oral capecitabine twice daily on days 1–14. Patients experiencing progression could cross over to Arm B at the physician’s discretion. Arm B received capecitabine as in Arm A and oral sunitinib malate once daily on days 1–21. Specific drug doses and treatment-cycle details beyond these schedules are not reported in the registry record. The posted primary-outcome entries list six participants in each arm; this does not by itself establish how many met the endpoint’s evaluability criteria.
The primary endpoint was progression-free survival, summarized as median time in months with a 95% confidence interval. The planned primary comparison was a one-sided log-rank test between Kaplan–Meier curves. The registry defines progression using RECIST v1.0: a 20% increase in the sum of the longest diameters of target lesions, a measurable increase in a non-target lesion, or the appearance of new lesions. Patients meeting eligibility criteria who started treatment were described as evaluable for the primary endpoint. Follow-up was specified as up to 3 years, with no further follow-up required if a patient remained alive 3 years after registration.
The registry does not report the randomization procedure, allocation concealment, the number of participants who crossed over, or details of treatment adherence. A protocol-based sample-size justification and interim-monitoring plan 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 A (Capecitabine) | Arm B (Capecitabine+Sunitinib) |
|---|---|---|
| STARTED | 6 | 6 |
| COMPLETED | 6 | 6 |
| NOT COMPLETED | 0 | 0 |
Analysis population for the primary outcome, as defined in the registry: All evaluable 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.
| Characteristic | Unit | Arm A (Capecitabine) | Arm B (Capecitabine+Sunitinib) | Total |
|---|---|---|---|---|
| Participants analysed | 6 | 6 | 12 | |
| Age, Continuous | years | 74.5 (56 to 82) | 73 (47 to 80) | 73 (47 to 82) |
| Sex: Female, Male: Female | Participants | 1 | 0 | 1 |
| Sex: Female, Male: Male | Participants | 5 | 6 | 11 |
| Ethnicity (NIH/OMB): Hispanic or Latino | Participants | 0 | 0 | 0 |
| Ethnicity (NIH/OMB): Not Hispanic or Latino | Participants | 6 | 6 | 12 |
| Ethnicity (NIH/OMB): Unknown or Not Reported | Participants | 0 | 0 | 0 |
6. Primary Endpoint
The registry lists the primary outcome as Comparison of Progression-free Survival, measured over Up to 3 years.
| Element | Registry entry |
|---|---|
| Result type | Median |
| Measure of spread | 95% Confidence Interval |
| Unit | Months |
| Endpoint classification | Time-to-Event Analysis |
Progression-free survival measures the time until disease progression or death, according to the endpoint definition used for the analysis. The registry’s description defines progression by specified radiographic changes under RECIST v1.0, including a 20% increase in the summed longest diameters of target lesions, measurable worsening of a non-target lesion, or new lesions. The registry does not separately detail how deaths were handled in the endpoint calculation.
The primary outcome is expressed as a median time in months, with a 95% confidence interval. A median describes the time by which half of the relevant event-time distribution has occurred, as estimated using the trial’s time-to-event analysis. It is not a statement that every participant experienced progression or death at that time. Participants without a recorded event during observation can contribute information up to their last follow-up rather than being treated as if an event occurred then.
Progression-free survival is commonly used in oncology trials because it captures disease worsening over follow-up and can be analyzed while accounting for differing observation times and censoring. That general rationale does not establish that the endpoint was assessed without bias in this trial. Because the study was open-label, assessment procedures and any safeguards against knowledge of treatment assignment are relevant; such details are not reported in the registry record.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The posted median progression-free survival was 6.1 months in Arm A, capecitabine alone, with a 95% confidence interval from 0.6 to 9.2 months. In Arm B, capecitabine plus sunitinib malate, the median was 2.4 months, with a 95% confidence interval from 0.6 to 8.4 months. The intervals are wide and overlap substantially, indicating that the estimates are imprecise and that the registry results do not sharply distinguish the arms’ underlying event-time distributions.
The posted log-rank analysis reports p = 0.43. This is not strong statistical evidence against the null hypothesis of no difference in the progression-free survival curves under the test. It should not be read as proof that the treatments are equivalent, or as proof that there is no clinically meaningful difference. A non-significant result in a small trial can reflect limited information as well as a genuinely small difference. The registry lists six participants in each arm for the primary outcome, so the event-time estimates and comparison have little data on which to rely. The record does not report a hazard ratio or its confidence interval.
The numerical ordering of the posted medians should be interpreted cautiously. Medians alone do not describe the full time-to-event curves, and they do not show whether the difference between groups is consistent over follow-up. The log-rank test compares the curves using event-time information, rather than testing only the two reported medians. The reported p-value is the sponsor’s posted analysis; no independent recalculation is included in the registry results. Whether the posted p-value is one-sided, as in the planned primary comparison description, is not reported alongside the analysis result.
Cross-over from Arm A to Arm B was permitted after progression at the physician’s discretion. This provision is important to the trial’s treatment structure, but the registry does not report how many participants crossed over or how crossover was handled in analyses other than the primary endpoint description. Because crossover was triggered by progression, it would not undo an already observed progression event for progression-free survival, but it may matter for interpreting later outcomes such as overall survival. No outcome-specific crossover results are reported in the information available.
These results are a small phase II comparison, not a precise estimate of treatment effects for a broad patient population. The posted results describe this trial’s participants and follow-up; they do not establish that either regimen is superior or equivalent. Participant characteristics, event counts, censoring details, and fuller curve estimates are not reported in the registry record, limiting assessment of the stability and generalizability of the comparison.
9. Statistical Methods Explained
The primary endpoint is a time-to-event outcome. In general, time-to-event methods use both the timing of events and the period during which each participant was observed. A participant who has not had a recorded event by the end of observation can be censored: the analysis uses the information available up to that point without assuming that the event happened then. The registry does not provide the event counts, censoring counts, or detailed censoring rules for this comparison.
What a median time represents
In general, the median time-to-event estimate is the time at which the estimated event-free proportion reaches the midpoint of its distribution. It is a summary of an estimated curve, not an average and not a guarantee about an individual participant’s experience. The posted median was 6.1 months for capecitabine alone and 2.4 months for capecitabine plus sunitinib malate. The registry gives a 95% confidence interval of 0.6 to 9.2 months for Arm A and 0.6 to 8.4 months for Arm B. The broad intervals indicate considerable uncertainty around each median.
What the confidence intervals convey
In general, a confidence interval describes the statistical uncertainty in an estimate under the assumptions of the method. A 95% confidence interval is not a range containing 95% of individual participants’ outcomes, nor does it assign a 95% probability to this particular interval containing the true value. Its width depends on the amount and pattern of information available. Here, the intervals extend across a broad range of times, which is consistent with imprecise median estimates in a trial with six participants listed per arm for the primary outcome. The record does not provide enough detail to assess the interval calculation beyond the stated endpoint and confidence level.
How the log-rank test works
In general, the log-rank test compares time-to-event curves across groups by evaluating how observed events are distributed over follow-up relative to what would be expected if the groups’ event patterns were the same. It uses event timing and can account for censored observations under its assumptions. It does not simply compare the medians, and its p-value is not an estimate of the size of a treatment effect. The registry describes a one-sided log-rank test as the planned primary comparison and reports a log-rank p-value of 0.43. Whether that reported p-value is one-sided is not reported alongside the result.
A p-value measures how compatible the data are with a specified null model, assuming that model and the test assumptions apply. A p-value of 0.43 does not mean there is a 43% probability that the null hypothesis is true, and it does not quantify the probability that the observed result happened “by chance.” In general, a large p-value means that the analysis does not provide strong evidence against the null model; it does not demonstrate equivalence. The registry does not report an equivalence margin or a non-inferiority framework.
What the Kaplan–Meier curves add
In general, Kaplan–Meier estimation constructs a stepwise estimate of the probability of remaining event-free over time. At each event time, the estimate reflects the number of participants still under observation and at risk. Censored participants contribute information while observed, then leave the risk set. The registry says that the planned comparison used two Kaplan–Meier curves, but the numerical curve coordinates and risk-set counts are not reported in the registry record. A curve can reveal features that a median does not, such as whether the groups separate early or later, but no specific curve shape should be inferred from the medians alone.
Assumptions and interpretation
In general, the log-rank test is most informative when the relative pattern of event risks is reasonably consistent over follow-up. If the curves cross or the difference changes over time, a single log-rank result may not convey the full pattern. The registry does not report a proportional-hazards assessment, a hazard ratio, or other analyses of changing effects. The log-rank test also relies on suitable handling of censoring and independent observations; the record does not describe checks of these assumptions.
The registry’s primary endpoint description says that all patients meeting eligibility criteria who started treatment were considered evaluable. This is an analysis-population definition, not a full account of the data included: the number meeting those conditions, the treatment-start counts, and handling of any missing endpoint assessments are not reported in the registry record. The follow-up timeframe was up to 3 years, but the record does not give the distribution of follow-up duration.
The posted analysis is a sponsor-reported log-rank comparison. No recalculated treatment-effect estimate is included, and no hazard ratio should be inferred from the medians or p-value. In general, a hazard ratio summarizes the relative instantaneous event rates under its model assumptions; it is distinct from a ratio of medians. The registry record does not report a hazard ratio for this primary comparison.
10. Analyses Reported by the Sponsor
| Method | Estimate | 95% CI | P-value |
|---|---|---|---|
| Log Rank | not reported | 0.430 |
The registry reports the sponsor’s primary comparison as a log-rank test with p = 0.43. It also reports arm-specific median progression-free survival estimates and their 95% confidence intervals. These are the posted analysis results. The record does not include a recalculated hazard ratio, confidence interval for a hazard ratio, or independent recalculated p-value, so there is no separate recalculation to compare with the sponsor’s analysis.
The registry’s endpoint description says that the planned comparison was a one-sided log-rank test. The posted analysis result identifies the method as log rank but does not state whether its p-value is one-sided. This distinction is not resolved in the registry record. No discrepancy between sponsor and recalculated results is identified. Accordingly, the sponsor’s reported analysis is the relevant comparison result, while the arm-specific medians and confidence intervals should be read as estimates with substantial uncertainty rather than as a substitute for the full time-to-event analysis.
11. Multiplicity, Missing Data and Analysis Populations
The registry identifies one primary endpoint, progression-free survival, and lists five secondary outcomes: response rate, overall survival, time to disease progression, time to treatment failure, and duration of response. It reports six outcome measures and two analyses overall, but the listed secondary outcomes do not come with their results in the information presented for this comparison. The analysis plan for those outcomes and the selection rules for reporting them are not reported in the registry record.
When a study evaluates multiple endpoints or makes multiple comparisons, the chance of at least one apparently noteworthy result can increase. General methods for addressing this include prespecifying a hierarchy of outcomes or adjusting significance thresholds. Whether any multiplicity strategy was used in this trial is not reported in the registry record. The primary endpoint p-value should therefore be interpreted as the reported result for that comparison, without assuming that it was adjusted for other tests.
The primary endpoint description defines the evaluable population as patients who met eligibility criteria and started treatment. The registry does not give the count meeting those conditions or explain whether any participants were excluded from the primary analysis. The posted primary-outcome entries list six participants per arm. Details of missing endpoint assessments, censoring, reasons for loss to follow-up, and imputation are not reported in the registry record. These omissions make it difficult to assess whether missing or incomplete follow-up could have influenced the estimates.
The protocol allowed crossover from Arm A to Arm B after progression at the physician’s discretion. The number crossing over and the handling of crossover for secondary outcomes are not reported. The record also does not report interim analyses, stopping rules, or a statistical plan for the secondary endpoints. Those details should not be inferred from the primary comparison alone.
12. 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 A (Capecitabine) | 1 | 6 | 16.7% |
| Arm B (Capecitabine+Sunitinib Malate) | 0 | 6 | 0.0% |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Arm A (Capecitabine) | 6 of 6 |
| Arm B (Capecitabine+Sunitinib Malate) | 6 of 6 |
Most frequent serious adverse event terms
| Event term | Organ system | Arm A (Capecitabine) | Arm B (Capecitabine+Sunitinib Malate) |
|---|---|---|---|
| Anorexia | Metabolism and nutrition disorders | 1 of 6 | 0 of 6 |
| Dehydration | Metabolism and nutrition disorders | 1 of 6 | 0 of 6 |
| Muscle weakness | Musculoskeletal and connective tissue disorders | 1 of 6 | 0 of 6 |
| Thrombosis | Vascular disorders | 1 of 6 | 0 of 6 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Arm A (Capecitabine) | Arm B (Capecitabine+Sunitinib Malate) |
|---|---|---|---|
| Hemoglobin decreased | Blood and lymphatic system disorders | 6 of 6 | 6 of 6 |
| Nausea | Gastrointestinal disorders | 6 of 6 | 4 of 6 |
| Fatigue | General disorders | 4 of 6 | 3 of 6 |
| Hand-and-foot syndrome | Skin and subcutaneous tissue disorders | 3 of 6 | 4 of 6 |
| Diarrhea | Gastrointestinal disorders | 4 of 6 | 2 of 6 |
| Vomiting | Gastrointestinal disorders | 3 of 6 | 3 of 6 |
| Leukocyte count decreased | Investigations | 3 of 6 | 3 of 6 |
| Cough | Respiratory, thoracic and mediastinal disorders | 4 of 6 | 2 of 6 |
| Ear, nose and throat examination abnormal | Gastrointestinal disorders | 2 of 6 | 3 of 6 |
| Serum sodium decreased | Metabolism and nutrition disorders | 2 of 6 | 3 of 6 |
Serious events with the highest arm-level rate: anorexia (16.7%), dehydration (16.7%), muscle weakness (16.7%), thrombosis (16.7%).
The adverse-event record lists one participant with a serious adverse event in Arm A and none in Arm B; six participants are listed at risk in each arm. It also lists other adverse events affecting six participants in each arm. These counts describe the registry’s posted safety entries and should not be treated as a complete account of event type, severity, duration, or attribution. The record does not provide the number of events per affected participant or detailed criteria for event classification in the information presented.
The registry lists anorexia, dehydration, muscle weakness, and thrombosis among the serious adverse-event terms, each with a reported percentage of 16.7%. These terms are not assigned to a specific arm in the posted top-term list, so an arm-specific interpretation is not supported. The death counts are not reported in the registry record. With six participants listed at risk in each arm, these safety observations are highly limited and cannot establish that one regimen has a different safety profile from the other.
13. What the Result Does, and Does Not, Mean
The registry results describe a randomized phase II trial comparing capecitabine alone with capecitabine plus sunitinib malate in metastatic esophageal or gastroesophageal junction cancer. The primary endpoint was progression-free survival. The reported median was 6.1 months in Arm A and 2.4 months in Arm B, with 95% confidence intervals of 0.6 to 9.2 months and 0.6 to 8.4 months, respectively. The sponsor-reported log-rank p-value was 0.43.
These estimates are uncertain, and the posted comparison does not provide strong statistical evidence against the null model tested by the log-rank analysis. It does not establish that the treatments are equivalent, that a difference is absent, or that either treatment is superior. The reported medians are not a complete description of the event-time curves, and the registry does not provide a hazard ratio or event counts to quantify the contrast in another way.
The small number of participants listed per arm for the primary outcome limits precision. The trial’s open-label design, permitted crossover, and unreported details about participant characteristics, endpoint missingness, and assessment procedures further constrain interpretation. The record’s description of eligibility and evaluability does not establish that all enrolled participants were included in the primary analysis.
These results apply to the trial as recorded and should not be extended automatically to other patient groups or treatment settings. The registry does not report a sample-size justification, detailed eligibility criteria, or enough information to assess how representative the enrolled participants were. The findings are best understood as an imprecise phase II comparison that leaves substantial uncertainty about the relative progression-free survival patterns.
14. Why This Trial Matters Statistically
This trial illustrates how a randomized design can support a direct comparison while still producing uncertain estimates when few participants are available. Randomization is intended, in general, to reduce systematic differences between treatment groups, but it cannot guarantee balance in a small study. The registry does not provide the participant characteristics needed to judge balance between these arms.
It also demonstrates why a p-value should not be interpreted in isolation. The log-rank p-value of 0.43 does not measure the size of a difference, establish equivalence, or indicate the probability that either treatment is effective. The medians and their confidence intervals add information about estimated timing and uncertainty, but they do not replace the full event-time curves or supply a hazard ratio. The broad intervals and small arm counts make precision central to interpretation.
The crossover provision is another design feature with implications for interpretation, especially for outcomes measured after progression. The registry states that crossover could occur at the physician’s discretion but does not report how often it occurred. This emphasizes the value of reporting implementation details alongside analysis methods.
Finally, the record shows why readers need clear reporting of analysis populations, missing data, multiplicity, and safety outcomes. Without these details, it is harder to assess how robust a result is or what it can support. The statistical lesson is not a treatment recommendation: it is that effect estimates, uncertainty, design features, and reporting gaps all matter when interpreting a small clinical trial.
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 primary limitation is the small trial size: the registry lists 12 participants enrolled and six participants per arm for the posted primary outcome. Such a limited amount of information can yield unstable estimates and wide confidence intervals. The record does not report a protocol-based sample-size justification, so the intended precision or design target cannot be assessed.
The registry does not provide participant ages, performance-status values, disease characteristics, or other baseline details. Although the official title describes elderly and/or poor performance status patients, the record does not show how many participants fell into either category. Detailed eligibility criteria are also not reported, limiting assessment of who the results represent.
The trial was open-label. The registry does not describe blinding of progression assessments, independent review, or safeguards against knowledge of treatment assignment. The primary endpoint depends on disease assessments, making these procedural details relevant. The record also does not provide event counts, censoring details, follow-up distribution, or missing-data handling, so the robustness of the time-to-event estimates cannot be fully evaluated.
Crossover was permitted after progression at the physician’s discretion, but the number of crossovers and their handling in analyses are not reported. The registry does not report a hazard ratio or its confidence interval, curve coordinates, or a proportional-hazards assessment. It also does not state whether the posted p-value was one-sided, despite describing a one-sided planned test. Details of multiplicity control, interim analyses, treatment adherence, and the statistical plan beyond the primary test are not reported in the registry record.
16. Related Tutorials
17. Related Statistical Calculators
18. Sources
- ClinicalTrials.gov: Capecitabine With or Without Sunitinib Malate as First-Line Therapy in Treating Patients With Metastatic Cancer of the Esophagus or Gastroesophageal Junction — NCT00891878.
- Registry record last updated 2018-07-24; status last verified 2018-06.
- Results first posted 2018-07-24.
- 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.
19. Record Summary
NCT00891878 is a completed randomized phase II parallel-group trial of capecitabine alone versus capecitabine plus sunitinib malate for first-line treatment of metastatic cancer of the esophagus or gastroesophageal junction. The registry lists 12 participants enrolled. The official title describes elderly and/or poor performance status patients with metastatic adenocarcinoma.
The primary endpoint was progression-free survival, summarized by median time and 95% confidence interval, with a planned one-sided log-rank comparison. The posted medians were 6.1 months for capecitabine alone and 2.4 months for the combination; the reported log-rank p-value was 0.43. The registry also lists five secondary outcomes and limited adverse-event counts. Details about participant characteristics, crossover, missing data, and several aspects of the statistical plan are not reported in the registry record.