Capecitabine, Oxaliplatin, and Bevacizumab in Treating Patients With Metastatic or Recurrent Colorectal Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Capecitabine, Oxaliplatin, and Bevacizumab in Treating Patients With Metastatic or Recurrent Colorectal Cancer (NCT00416494), a non-randomized phase 2 study. The primary outcome is Response Rate (Percentage of Participants With Partial or Complete Response).

2. Clinical Question

3. Trial Design

This was a phase II study of oxaliplatin, capecitabine, and bevacizumab in metastatic colorectal cancer. The registry identifies the design as non-randomized, with no masking and a single-group model. Enrollment is recorded as 50. Although the record labels two cohorts, Initial Cohort and Second Cohort, it gives no cohort descriptions and does not establish that they received different regimens. Their labels therefore do not create a randomized treatment comparison.

The study status is completed. The recorded start is 2003-09 and the primary completion is 2008-01. These dates describe the registry timeline; they do not, by themselves, establish how long any individual participant was followed. For the primary response endpoint, the record states that restaging took place every 9 weeks after study-drug initiation until disease progression, with assessment up to 24 months.

The intervention list names bevacizumab, oxaliplatin, and capecitabine, with capecitabine listed twice. The record does not report dose, route, treatment duration, treatment modifications, or the reason for the duplicate listing. It also does not provide a detailed protocol description for the cohorts. The stated rationale describes using drugs with different ways of acting on tumor growth, but this rationale is not evidence of a clinical outcome.

Several secondary outcomes are named: time to progression, disease-free survival, overall survival, and safety and tolerability. The registry records 7 outcome measures but lists no analysis records. The statistical analysis plan, including the intended methods for the secondary outcomes, is not reported in the registry record. No sample-size justification or interim-analysis plan is reported either.

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

Analysis population for the primary outcome, as defined in the registry: All subjects who had restaging scans were included in the analysis.

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 Response Rate (Percentage of Participants With Partial or Complete Response), measured over After all subjects were evaluated for restaging which occured every 9 weeks from drug initiation until disease progression, assesed up to 24 months..

The primary endpoint is the percentage of participants with a partial or complete response. It is a binary outcome at the participant level: a participant is classified according to whether the disease assessment meets the response criteria. The registry says that disease assessments were recorded according to RECIST v.1.0. A complete response is defined as disappearance of all target lesions. A partial response is defined as at least a 30% decrease in the sum of the longest diameters of target lesions, using the baseline sum as reference.

The record describes stable disease as neither sufficient shrinkage to qualify as partial response nor sufficient increase to qualify as progressive disease, using the smallest sum of target-lesion diameters since treatment started as reference. Progressive disease is described as at least a 20% increase from the smallest recorded sum since treatment started, or the appearance of one or more new lesions. The registry text labels partial response with the abbreviation PD, which it also uses for progressive disease; that abbreviation is therefore ambiguous in the written definitions.

Response is commonly used in clinical studies because it provides a structured way to describe tumor change after treatment. It does not, on its own, describe how long a response lasts, whether disease progression is delayed, or how participants fare overall. The registry lists those as separate secondary outcomes. It does not report the primary response percentage or a confidence interval, so neither the observed frequency nor its statistical precision can be assessed from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

General statistical explanation: the primary outcome is a binary response classification, so each participant included in the response analysis contributes to one of two broad categories: response or no response. The response category combines partial and complete response, as stated in the primary endpoint title. A response percentage is interpreted relative to the participants included in the analysis, not automatically relative to everyone enrolled. Here, the registry defines the population as all subjects who had restaging scans, but does not report how many participants met that definition or how missing assessments were handled.

What does a confidence interval communicate?

The registry identifies a 95% confidence interval as the dispersion measure for the primary endpoint, but no interval limits are reported. In general, a confidence interval gives a range of estimates compatible with the observed data under a specified statistical method. It communicates precision rather than certainty that a particular value is true. The interval’s width depends on the observed data and the method used. Without the reported limits and method details, the precision of this study’s response estimate cannot be evaluated.

A confidence interval does not replace the point estimate: the estimate describes the observed response frequency, while the interval expresses uncertainty around it. Nor does an interval alone establish a causal treatment effect. In this study, there is no randomized comparator, so even a precise response estimate would describe the observed single-group experience rather than a randomized difference between treatments.

Why does the analysis population matter?

The registry states that all subjects who had restaging scans were included in the primary analysis. That definition makes scan availability relevant to the denominator. Participants without restaging scans may not be represented in that analysis population, but the record does not say whether any participants lacked scans, why scans might be unavailable, or how such cases were treated. It also does not report the number of participants analyzed for the endpoint. These omissions prevent assessment of the relationship between enrollment and the response estimate.

General statistical explanation: excluding participants without an outcome assessment can affect an estimate if the chance of having an assessment is related to the participant’s disease status or other factors. This possibility does not establish that bias occurred in this trial. It explains why the number analyzed and the missing-data approach are important details. Neither a missing-data method nor sensitivity analysis is reported in the registry record.

What can be inferred from a single-group design?

A single-group study can estimate outcomes observed among participants receiving the study regimen. Randomization is a method for creating comparable groups, on average, so that differences in outcomes can be interpreted against a concurrent control. This record specifies non-randomized allocation and a single-group model, with no masking. It does not describe a separate comparator. Accordingly, the design does not support a randomized between-treatment contrast. The lack of masking is recorded, but the registry does not explain how outcome assessment was protected from knowledge of treatment.

RECIST v.1.0 supplies stated rules for classifying changes in target lesions. The record defines complete response as disappearance of all target lesions, partial response as at least a 30% decrease in the sum of target-lesion diameters from baseline, and progressive disease as at least a 20% increase from the smallest recorded sum or the appearance of new lesions. The assessment schedule is stated as every 9 weeks from treatment initiation until progression, with assessment up to 24 months. These details define how response was intended to be assessed, but do not provide the actual response classifications.

What do the planning assumptions mean?

General statistical explanation: sample-size planning asks how many participants or events might be needed to detect a specified effect under chosen assumptions. It is a design exercise, not a description of the observed trial result. The planning illustration assumptions described for this record use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The target effect is stated as a fraction of the observed effect. Those assumptions do not make this single-group, non-randomized trial a two-arm randomized study, and the resulting planning values should not be read as observed power.

The illustration is also distinct from the trial’s original sample-size justification. The record does not report the original rationale for enrollment of 50, the effect used in the original design, or any interim-analysis plan. No numerical result from a planning illustration is described in the registry record. Because the primary response result itself is not reported, its interpretation cannot be reconstructed from the enrollment number alone.

How should secondary outcomes be understood?

The registry names time to progression, disease-free survival, overall survival, and safety and tolerability as secondary outcomes. Time-to-event outcomes require details about when the event is counted, how follow-up ends, and how participants without an observed event are treated in the analysis. General statistical methods for such outcomes may account for differing follow-up, but the registry does not state which methods were used. It does not report secondary-outcome results, event counts for those outcomes, or analysis records. No specific model, adjustment strategy, or censoring method should therefore be attributed to the study.

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("Initial Cohort", "Second Cohort")
n      <- c(19, 31)
events <- c(12, 13)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Initial Cohort  19  12  63.2  38.4  83.7  41.0  80.9  40.8  80.9
ARM  Second Cohort  31  13  41.9  24.5  60.9  26.4  59.2  26.4  59.4

11. Multiplicity, Missing Data and Analysis Populations

The registry names several secondary outcomes in addition to the primary response endpoint, but does not report an analysis plan or analysis records. It does not state whether the primary endpoint was tested against a formal null hypothesis, whether secondary outcomes were tested, or whether any adjustment for multiple testing was planned. Multiplicity refers generally to the increased chance of a false-positive finding when multiple statistical tests are considered. Whether that issue was addressed in this study is not reported in the registry record.

The primary analysis population is described as all subjects who had restaging scans. The record does not state how many participants had scans, how many were excluded from the response analysis, or how missing assessments were handled. It gives no reasons for missing scans and no imputation or sensitivity-analysis method. As a result, the effect of incomplete assessment on the endpoint estimate cannot be evaluated.

The registry lists 50 enrolled participants, but enrollment is not necessarily the same as the analysis denominator. The study also lists Initial Cohort and Second Cohort, without cohort descriptions. It does not explain how participants were assigned to these cohorts or how the cohorts relate to the endpoint analysis. The statistical analysis plan, including analysis populations for secondary outcomes, is not reported in the registry record.

General statistical explanation: missing outcome data can limit interpretation when the probability of having an observed outcome is related to the outcome itself or to participant characteristics. That concern is a reason to seek transparent reporting; it is not evidence that missingness caused bias in this study. The available record does not allow that question to be resolved.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: bowel perforation (5.3%), acute calculus cholecystitis (5.3%), wound dehisence (5.3%), congestive heart failure (5.3%), altered mental status (5.3%).

The safety record separates participants into Initial Cohort and Second Cohort. For serious adverse events, 19 participants were listed as at risk and 6 as affected in the Initial Cohort; 31 were listed as at risk and 9 as affected in the Second Cohort. For other adverse events, the Initial Cohort lists 19 at risk and 19 affected, while the Second Cohort lists 31 at risk and 29 affected. These entries are counts, and they should be read within the cohort descriptions rather than as a randomized comparison.

The registry also lists Bowel Perforation, Acute Calculus Cholecystitis, Wound Dehisence, Congestive Heart Failure, Altered mental status, and Diarrhea among serious adverse event terms, each with 5.3%. The record does not identify the cohort for each term or provide further event-level details in this information. Death counts are not reported in the registry record.

General statistical explanation: adverse-event counts describe recorded events among participants considered at risk; they do not by themselves establish that treatment caused an event. Cohort counts should not be compared as if treatment groups had been randomized, particularly because cohort descriptions and a control group are absent. The safety entries provide limited descriptive information and do not report a complete account of event timing, severity, duration, or attribution.

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

The registry record documents a phase II study of oxaliplatin, capecitabine, and bevacizumab in metastatic colorectal cancer, with response rate as the primary endpoint. It specifies RECIST v.1.0 assessment and a schedule of restaging scans every 9 weeks from treatment initiation until disease progression, with assessment up to 24 months. These details describe the intended endpoint and assessment approach, not the study’s observed response finding.

The primary response percentage and its confidence interval are not reported in the registry record. The record therefore does not establish the observed proportion of participants who had a partial or complete response, nor how precisely that proportion was estimated. It also does not report the number of participants who had restaging scans, so the analysis denominator cannot be identified from the endpoint description alone.

The study’s single-group, non-randomized design has no stated control treatment. Any outcomes observed in the treated group would not, by themselves, provide a randomized comparison with another treatment. The two cohort names do not supply such a comparison: their descriptions are blank, and the record does not state that they represent different randomized regimens.

The record lists time to progression, disease-free survival, overall survival, and safety and tolerability as secondary outcomes, but their results are not reported in the information available. Safety counts and serious adverse-event terms are present, but these are descriptive and should not be interpreted as causal or comparative findings. No advice about treatment choice follows from these registry entries.

General statistical interpretation: a planned measure, assessment schedule, or named outcome should not be mistaken for an observed result. The distinction matters because the registry record documents what was to be assessed while leaving the primary numerical outcome and analysis details unreported. Conclusions about effect size, precision, comparative benefit, or the robustness of the findings cannot be drawn from the record alone.

14. Why This Trial Matters Statistically

This study is statistically informative as an example of the distinction between an endpoint definition and an endpoint result. The registry describes a response-rate outcome, identifies RECIST v.1.0, and gives a restaging schedule. Yet it does not report the observed response percentage or its confidence interval. A clear outcome definition is important, but without the estimate and its denominator, readers cannot assess the result’s magnitude or precision.

The design also illustrates why a cohort label is not equivalent to a control group. The registry identifies a non-randomized, single-group study and names Initial Cohort and Second Cohort, but supplies no cohort descriptions or randomized allocation. Safety counts organized by cohort therefore remain descriptive; they do not create a fair treatment comparison.

Several secondary outcomes are named, while analysis records and statistical methods are absent. This gap illustrates the importance of distinguishing outcomes listed in a registry from outcomes with reported estimates and documented analyses. It also leaves multiplicity, missing-data handling, and analysis-population choices unresolved.

Finally, planning illustrations answer a different question from trial results. Their stated assumptions—two-sided significance level 0.05, equal allocation between two arms, and normal approximation—describe a hypothetical planning exercise based on a fraction of the observed effect. They do not explain the original enrollment of 50 and do not convert the study into a two-arm randomized comparison. The trial’s own sample-size justification is not reported in the registry record.

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 most consequential limitation for interpreting the primary endpoint is that the registry record does not report the response percentage, confidence interval, or number of participants included in the analysis. Although 50 participants were enrolled, the analysis population is defined as those with restaging scans, and its size is not reported. The record also does not explain how participants without scans, if any, were handled.

The design is non-randomized and single-group, with no described control treatment. The Initial Cohort and Second Cohort have blank descriptions, leaving their relationship to each other and to the treatment regimen unclear. The available information does not support treating them as comparable treatment arms.

Details needed to interpret the endpoint implementation are incomplete. The registry specifies RECIST v.1.0 and scan timing, but not the statistical method used to calculate the confidence interval, the assessment procedures beyond the stated schedule, or handling of incomplete or unevaluable scans. Its response definitions use PD for both partial response and progressive disease, creating an abbreviation ambiguity in the written criteria.

The registry names secondary outcomes but does not report their results or methods. It does not report a multiplicity plan, missing-data strategy, interim-analysis plan, or sample-size justification. Safety information is also limited: death counts are not reported, serious-event terms are not assigned to a cohort, and details such as event timing and attribution are absent. These are limitations of the information reported in the registry record; they do not establish that an unreported procedure or analysis was not conducted.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Capecitabine, Oxaliplatin, and Bevacizumab in Treating Patients With Metastatic or Recurrent Colorectal Cancer — NCT00416494.
• Registry record last updated 2014-09-03; status last verified 2014-08.
• Results first posted 2013-03-14.
• 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 describes a completed phase II study of oxaliplatin, capecitabine, and bevacizumab for metastatic or recurrent colorectal cancer. It reports enrollment of 50, non-randomized allocation, no masking, and a single-group model. The recorded study start is 2003-09, and primary completion is 2008-01.

The primary endpoint is the percentage of participants with partial or complete response under RECIST v.1.0. Restaging scans were scheduled every 9 weeks from treatment initiation until disease progression, with assessment up to 24 months. The primary analysis population is described as participants who had restaging scans, but its size and the response result are not reported in the registry record.

Time to progression, disease-free survival, overall survival, and safety and tolerability are listed as secondary outcomes. Safety counts are reported for two named cohorts, whose descriptions are blank. No analysis records are listed, and the statistical plan, sample-size justification, and missing-data approach are not reported in the registry record.