A Study to Determine the Activity of Robatumumab (SCH 717454, MK-7454) in Participants With Relapsed or Recurrent Colorectal Cancer (P04721, MK-7454-003): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Determine the Activity of Robatumumab (SCH 717454, MK-7454) in Participants With Relapsed or Recurrent Colorectal Cancer (P04721, MK-7454-003) (NCT00551213), a randomized phase 2 study. The primary outcome is Number of Participants With a >20% Decrease in Positron Emission Tomography (PET)-Assessed Tumor Glucose Metabolism: Fluorodeoxyglucose (FDG) Standardized Uptake Value (SUV) in the Target Lesion.

2. Clinical Question

3. Trial Design

The registry describes a randomized, open-label, parallel phase 2 study of robatumumab in participants with relapsed or recurrent colorectal cancer. Enrollment is listed as 67, and the study status is completed. The registry dates the study start to 2007-11-21 and primary completion to 2009-06-04. Merck Sharp & Dohme LLC is named as sponsor.

The study used two treatment sequences. In the Robatumumab→Robatumumab group, participants received one dose of robatumumab 0.3 mg/kg intravenously and then robatumumab 10 mg/kg intravenously once every 2 weeks until disease progression. The registry defines a robatumumab cycle as 2 weeks of treatment, with no recovery period between cycles.

In the Chemotherapy→Robatumumab group, participants first received one cycle of standard colorectal cancer chemotherapy selected by the investigator based on prior treatment, followed by robatumumab 10 mg/kg intravenously once every 2 weeks until disease progression. Listed chemotherapy choices include irinotecan, capecitabine, FOLFOX, CAPEOX/XELOX, FOLFIRI, and cetuximab-containing options, as described in the registry record.

The study was open-label, so masking is recorded as none. The randomization procedure, allocation concealment, and any further details about how participants entered the treatment sequences are not reported in the registry record. The primary endpoint was a PET-based measure rather than a conventional tumor response endpoint. Secondary outcomes included adverse events, investigator and central review of best overall tumor response, discontinuation due to adverse events, and change from baseline in tumor growth rate.

4. Participant Flow and Accounting

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

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All participants who received ≥1 dose of robatumumab in Periods 1 and 2 and had SUV data before and after the first dose of robatumumab in Period 2 were included in the analysis. No participants in the Chemotherapy→Robatumumab group received robatumumab in Period 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.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants With a >20% Decrease in Positron Emission Tomography (PET)-Assessed Tumor Glucose Metabolism: Fluorodeoxyglucose (FDG) Standardized Uptake Value (SUV) in the Target Lesion, measured over After the first robatumumab dose in Period 2 (Up to approximately 4 weeks after first robatumumab dose in Period 1).

The primary endpoint counted participants with a greater than 20% decrease in FDG SUV in the target lesion, measured using positron emission tomography. FDG is a glucose analogue, and the endpoint was intended to assess tumor glucose metabolism as an indicator of biological activity. It is a measurement of a metabolic imaging change, not the same outcome as tumor shrinkage assessed using RECIST or a clinical outcome such as survival.

The registry states that RECIST version 1.0 was used to select target lesions. Up to 5 lesions per organ and 10 lesions in total could be identified and measured at baseline. For each participant, the change in SUVmax was calculated as endpoint SUVmax minus baseline SUVmax, divided by baseline SUVmax, and expressed as a percentage. When multiple target lesions were measured at a visit, their percentage changes were averaged for that participant. A participant was classified as an FDG SUVmax responder if SUVmax decreased by greater than 20% after the first robatumumab cycle in Period 2.

Using a defined threshold turns a continuous imaging change into a count of participants meeting a responder definition. That makes the endpoint straightforward to summarize, but it also discards information about the size of changes within each category. The primary endpoint timeframe is described as after the first robatumumab dose in Period 2, up to approximately 4 weeks after the first robatumumab dose in Period 1. It does not, by itself, establish how a metabolic imaging change relates to longer-term tumor control or patient outcomes.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

A binary endpoint classifies each participant into a category, such as meeting or not meeting a responder definition. Here, the responder category was based on a greater than 20% decrease in SUVmax after the first robatumumab cycle in Period 2. The posted result is a count of participants who met that rule. For a binary outcome, an analysis might compare the proportion meeting the rule across groups, but such a comparison requires outcome data for the groups being contrasted and a clearly defined analysis population.

What does the responder threshold do?

A threshold simplifies a continuous measurement by converting it into a yes-or-no outcome. This can make results easier to communicate, but it loses detail about the magnitude of change. A participant whose change barely passes the cutoff is classified in the same responder category as one with a much larger decrease. Likewise, changes that do not cross the cutoff are grouped together even if they differ in direction or size. The registry describes the calculation of SUVmax change and the lesion-averaging approach, but does not post a distribution of the underlying continuous changes.

Why is the comparison not estimable?

The primary analysis population required at least one robatumumab dose in both Periods 1 and 2 and SUV data before and after the first Period 2 dose. The registry states that no participants in the chemotherapy-first group received robatumumab in Period 1. As a result, that group has no participants meeting this primary analysis definition. Without an outcome count from an eligible comparison group, the posted data do not support an estimate of the difference between sequences for this endpoint. This is a structural mismatch between the analysis-population definition and one treatment sequence, rather than evidence that the two sequences had equal outcomes.

How should a confidence interval be understood?

In general, a confidence interval expresses the uncertainty around an estimated quantity under a specified statistical method. For a group comparison, it would describe a range of values compatible with the observed data and the method’s assumptions. A confidence interval does not remedy the absence of an eligible comparison group. The registry record does not report a confidence interval for a between-sequence contrast, and no such contrast can be interpreted from the posted primary count.

What can randomization contribute?

In general, random allocation can help balance prognostic characteristics between treatment groups, on average, when a valid comparison is performed according to the randomized assignment. It does not guarantee that every characteristic will be balanced in a particular study, and it does not make a comparison possible when a group has no eligible observations for the endpoint analysis. The registry records randomized allocation, but the randomization method and analysis according to assigned group are not reported in the registry record.

What is the relevance of open-label conduct?

Open-label means masking is recorded as none. In general, knowledge of treatment assignment can influence care, assessment, or reporting, although an imaging measurement may involve procedures intended to standardize measurement. The registry does not report whether image readers were blinded to treatment sequence, how readings were adjudicated, or whether the SUV measurements underwent independent review. Those details should not be assumed from the open-label designation alone.

Why can over-dispersion matter?

General statistical explanation: over-dispersion refers to variability in count data that is greater than expected under a particular model. When it is present, a model that assumes less variability can understate uncertainty and produce intervals or tests that are too optimistic. The primary endpoint is a binary participant-level count, not a recurrent-event rate, and the registry record does not report a model or dispersion assessment for this endpoint. Over-dispersion should not be presumed to have been assessed.

What analyses are documented?

The registry record lists the primary outcome and five secondary outcomes, but reports no posted statistical analyses. It does not state a statistical model, a test for the primary endpoint, covariate adjustment, confidence-interval method, or a hierarchy for confirmatory testing. In general, those methods determine how uncertainty and multiplicity are handled; without a documented plan, the descriptive count should not be recast as a formal test of the study hypothesis.

Sample-size calculations are planning tools, not evidence about the result observed in a completed trial. The registry record does not report the study’s sample-size justification or a planned detectable effect. Any planning illustration based on an observed effect would be a separate calculation and should not be interpreted as observed power or as a substitute for an inferential analysis.

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("Robatumumab→Robatumumab")
n      <- c(41)
events <- c(7)
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  Robatumumab→Robatumumab  41  7  17.1  7.2  32.1  8.5  31.3  8.6  31.4

11. Multiplicity, Missing Data and Analysis Populations

The registry lists five secondary outcomes in addition to the primary endpoint. These include adverse events, best overall tumor response under investigator review, discontinuation of study drug due to an adverse event, best overall tumor response under central review, and change from baseline in tumor growth rate. Considering multiple outcomes can raise the chance of finding an apparently notable result by chance when many tests are performed. In general, a statistical plan may address this through a prespecified hierarchy or other multiplicity approach. Such an approach is not reported in the registry record, and no posted statistical analyses are identified.

The registry does not report a statistical analysis plan, a prespecified primary comparison method, or whether any interim analyses were planned or conducted. These details are not reported in the registry record and should not be inferred from the randomized designation or the listed outcomes.

The primary endpoint analysis required robatumumab exposure in Periods 1 and 2, as well as SUV data before and after the first dose in Period 2. This eligibility rule limits the analysis to participants meeting those conditions. The registry explicitly states that no participants in the Chemotherapy→Robatumumab group received robatumumab in Period 1, leaving no participants in that group who could satisfy the stated primary analysis criteria. The missing comparator result therefore prevents a two-arm analysis as posted.

The registry does not describe the amount or reasons for missing PET measurements, how missing endpoint data were handled, or whether sensitivity analyses were conducted. It also does not report a separate intention-to-treat primary analysis. Consequently, the posted count should be understood as applying to the defined analysis population, not automatically to every randomized participant. The extent to which missing measurements or eligibility restrictions could affect interpretation cannot be determined from the registry record.

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: dehydration (13.3%), abdominal distension (6.7%), diarrhoea (6.7%), small intestinal obstruction (6.7%), bacteraemia (6.7%).

The registry reports serious adverse event figures separately for the Chemotherapy→Robatumumab and Robatumumab→Robatumumab groups. For the chemotherapy-first group, 6 participants were reported with serious adverse events among 15 at risk. For the robatumumab-first group, 14 were reported with serious adverse events among 49 at risk. These are counts, not a comparison adjusted for differences in exposure or follow-up. The registry also reports other adverse events for these groups: 15 among 15 at risk and 48 among 49 at risk, respectively.

Terms listed among the serious adverse events include DEHYDRATION at 13.3%, ABDOMINAL DISTENSION at 6.7%, DIARRHOEA at 6.7%, SMALL INTESTINAL OBSTRUCTION at 6.7%, BACTERAEMIA at 6.7%, and MUSCULOSKELETAL CHEST PAIN at 6.7%. The registry record does not specify in this listing how those terms are distributed between treatment groups. Death counts are not reported in the registry record.

These safety summaries describe events recorded during the study; they do not establish that a treatment caused an event. The registry information does not provide a comparative statistical analysis, detailed exposure-adjusted rates, or enough context to interpret differences in follow-up. Counts should therefore be read as descriptive safety reporting rather than as a causal or comparative conclusion.

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

The posted primary result documents 7 participants meeting the FDG SUVmax responder definition among 41 participants in the Robatumumab→Robatumumab primary analysis group. It describes a PET-based metabolic response threshold after robatumumab exposure. It is not a direct estimate of tumor response by RECIST, a survival outcome, or a measure of symptom change.

The result does not provide an estimable primary endpoint comparison with the Chemotherapy→Robatumumab group. The primary population required robatumumab in both treatment periods, and the registry states that no participants in the chemotherapy-first group received robatumumab in Period 1. Accordingly, this count cannot establish whether the response frequency differed between the treatment sequences or whether the stated comparative hypothesis was supported.

The threshold-based count also leaves out information about the magnitude and distribution of SUV changes. The registry explains how changes were calculated and averaged across target lesions, but does not report the individual values or a group summary of continuous change. The count alone cannot show whether changes were clustered near the cutoff or substantially beyond it.

More generally, an imaging signal of altered glucose metabolism is a biological measurement. It should not be equated with longer-term benefit or harm. The registry record does not report a link between this primary endpoint and later clinical outcomes, nor does it supply a comparative inferential result for the primary hypothesis. No treatment recommendation or conclusion about clinical effectiveness follows from the posted count alone.

The endpoint is informative as a description of the specified imaging response in the participants who met the analysis criteria. Its interpretation is limited by the restricted analysis population, the lack of an eligible comparator for the primary endpoint, and the absence of reported statistical analyses and missing-data methods. Those limits are central to reading the result, rather than reasons to reinterpret the count as a complete group comparison.

14. Why This Trial Matters Statistically

This study illustrates how a trial’s question, treatment sequence, and endpoint population need to align for a comparative analysis. The stated primary hypothesis concerned changes after robatumumab 10 mg/kg compared with changes after 0.3 mg/kg. Yet the primary endpoint population required robatumumab exposure in both Periods 1 and 2, and no participants in the chemotherapy-first sequence received robatumumab in Period 1. The resulting posted primary outcome is a count in the Robatumumab→Robatumumab group rather than a comparison across the randomized sequences.

The example also highlights the distinction between a randomized trial and an estimable randomized comparison. Random allocation is part of the design, but the endpoint analysis still depends on which participants have the required treatment exposure and measurements. An analysis population defined by post-randomization events or data availability may not retain all the features of the original randomized comparison. The registry record does not report an alternative primary analysis that would address this issue.

The endpoint itself demonstrates a common trade-off in clinical research: a continuous measurement can be reduced to a threshold-based responder count for ease of interpretation, but information about the size of change is lost. Reporting both the threshold result and the underlying continuous change can help readers understand the pattern more fully; the latter is not reported for this primary endpoint in the registry record.

Finally, this record shows why statistical interpretation depends on transparent reporting of analysis methods, missing-data handling, and the relationship between the planned question and available results. Without a reported comparative estimate or analysis plan, the posted result remains descriptive. That distinction is important when registry entries are used to understand what a study measured and what its results can support.

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 immediate limitation is that the registry reports a primary endpoint result for only the Robatumumab→Robatumumab group. The primary analysis population required robatumumab exposure in both periods, while no participants in the Chemotherapy→Robatumumab group received robatumumab in Period 1. A two-arm comparison is therefore not estimable from the posted primary outcome.

The primary analysis also required SUV data before and after the first robatumumab dose in Period 2. The registry does not report how many participants lacked the required data, why measurements were unavailable, or how missing values were handled. It does not report whether the analysis population was specified before study conduct or whether any sensitivity analyses were performed.

The endpoint is a thresholded imaging measure. Although the registry describes the SUVmax calculation and lesion averaging, it does not report the continuous change distribution, measurement variability, or details of image reading and adjudication. The record also does not state whether imaging readers were masked to treatment sequence.

Although allocation is recorded as randomized, the randomization method and allocation concealment are not reported in the registry record. The statistical model, inferential test, confidence-interval method, covariate adjustment, and multiplicity strategy are also not reported. No posted statistical analyses are identified. These omissions limit evaluation of uncertainty and of how the primary hypothesis was tested.

The chemotherapy-first sequence allowed investigator selection from several standard treatment options based on prior treatment. The registry record does not provide the distribution of those choices or a comparative analysis accounting for them. The safety figures are descriptive and do not include reported death counts; details needed to compare event risk in relation to exposure and follow-up are not reported in the registry record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study to Determine the Activity of Robatumumab (SCH 717454, MK-7454) in Participants With Relapsed or Recurrent Colorectal Cancer (P04721, MK-7454-003) — NCT00551213.
• Registry record last updated 2018-08-24; status last verified 2018-07.
• Results first posted 2015-12-15.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

The registry describes a completed randomized, open-label, parallel phase 2 study of robatumumab in participants with relapsed or recurrent colorectal cancer. Enrollment is listed as 67. The study began on 2007-11-21, with primary completion on 2009-06-04.

The primary endpoint counted participants with a greater than 20% decrease in PET-assessed FDG SUVmax in a target lesion. The posted primary result is 7 participants among 41 in the Robatumumab→Robatumumab group. No primary endpoint result is available for the Chemotherapy→Robatumumab group, and a two-arm comparison is not estimable from the registry record because the primary analysis criteria required robatumumab in both periods.

The registry lists five secondary outcomes covering adverse events, tumor response, treatment discontinuation due to adverse events, and tumor growth rate. It does not report posted statistical analyses, a statistical analysis plan, or methods for handling missing primary endpoint data.