Study of Pembrolizumab in Combination With Chemotherapy for Patients With Advanced Colorectal Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of Pembrolizumab in Combination With Chemotherapy for Patients With Advanced Colorectal Cancer (NCT02375672), a na phase 2 study. The primary outcome is Median Progression Free Survival (mPFS).

2. Clinical Question

The central statistical issue is the distinction between a question and evidence answering that question. A single-arm study can describe outcomes among people who received the combination, but it does not create a randomized estimate of the difference between the combination and another treatment. Any comparison with historical standards depends on information outside the concurrent study group, and the record does not identify the standards or provide a comparative estimate.

The registry lists disease assessment for objective response rate, disease control rate, overall survival, and grade 3 or higher treatment-related adverse events as secondary outcomes. These measures address different aspects of outcomes and safety; they are not substitutes for the primary progression-free survival question. The registry does not report numerical results for these secondary outcomes.

3. Trial Design

This is a multicenter, single-arm, open-label phase II study, registered as completed. The registry record reports enrollment of 30. Its brief description notes a safety run-in cohort. Allocation is listed as not applicable, and masking as none; the study description likewise states that randomization and blinding were not involved.

The intervention group is named pembrolizumab (MK-3475) plus mFOLFOX6. Following the safety run-in cohort, mFOLFOX6 was scheduled on D1 and D15 every 2 weeks, while pembrolizumab was administered intravenously over 30 minutes every 3 weeks. The registry does not report further details about the safety run-in or how the schedule was modified, if at all, for individual participants.

The registry lists the study start as 2015-05-28 and primary completion as 2020-10-08. These dates describe the record’s study timeline; they do not, by themselves, specify how long each participant was followed. The primary endpoint timeframe is described as beginning at registration and ending at documented progression per RECIST 1.1 or death, with an estimate of 14 months.

Because all participants received the named combination and no concurrent control arm is listed, treatment effects cannot be separated from other differences between this study population and a historical group. Open-label conduct also means that treatment assignment was known. The registry does not report whether endpoint assessment was performed by independent reviewers or whether any other safeguards were used to limit assessment bias. Those details should not be inferred from the phase or study title.

The record marks the trial completed but lists 0 analyses. It does not report the primary endpoint estimate, confidence interval, or comparative analysis. Thus, the available registry entries describe the design and intended question, not a numerical answer to that question.

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:

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 Median Progression Free Survival (mPFS), measured over From time of registration to the time of documented progression per RECIST 1.1 or subject death (estimate 14 months).

The primary endpoint is median progression-free survival, abbreviated mPFS. The registry defines the time interval as beginning at registration and ending at documented progression under RECIST 1.1 or death. In general, progression-free survival combines these events into a time-to-event outcome: it describes how long participants remain alive without documented disease progression. The median is the time by which half of the observed event-time distribution has had an event, subject to the handling of censoring in the analysis.

Time-to-event endpoints are useful when participants may have different lengths of follow-up or when some have not experienced the event by the time data are analyzed. Standard analysis methods can use information from participants who are followed for different periods, while accounting for observations that end before progression or death is documented. The registry record does not report the censoring rules or the analysis method used for this endpoint.

The record gives an 80% confidence interval as the endpoint’s dispersion measure, but does not report the median estimate or its interval limits. A confidence interval generally represents uncertainty around an estimated quantity under the assumptions of the analysis; it is not a range containing a specified proportion of individual participants’ survival times. Without the estimate and interval, neither the observed mPFS nor its precision can be assessed.

The timeframe description includes an estimate of 14 months. This is a stated estimate in the endpoint description, not a reported median result or a participant-level follow-up summary. The registry describes the purpose as determining whether the combination improves mPFS compared with historical standards, but it does not report the historical benchmark or a formal comparison.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods help turn participant follow-up into estimates, but the registry record does not report the analysis method for the primary endpoint. The explanations below are general descriptions of methods relevant to the named endpoint; they are not claims about what this trial’s investigators used.

What does a median time-to-event endpoint represent?

For progression-free survival, each participant contributes a time from registration until documented progression or death, if one of those events is observed. A participant without a documented event by the end of their observed follow-up contributes censored information. Censoring means the event time is not fully observed; it does not mean that progression or death occurred at the end of follow-up. The method used to define censoring and the analysis cutoff can affect an estimate, and those details are not reported in the registry record.

The median is one summary of the time-to-event distribution. It indicates the point at which the estimated event-free proportion reaches its midpoint, under the analysis method. It does not describe every participant’s experience, and it does not give the full shape of the survival curve. The registry names median progression-free survival, but provides no estimate from which to interpret its magnitude.

Why is a confidence interval useful?

A confidence interval expresses uncertainty around an estimate, accounting for sampling variability under the statistical model. A narrower interval generally indicates greater precision than a wider interval, although the interval’s validity depends on the analysis assumptions and data. The registry specifies an 80% confidence interval for the primary endpoint, but does not give its bounds or the estimate it accompanies. It is therefore not possible to judge precision from the record.

The confidence level is not the probability that the true value lies within a particular calculated interval after the data have been observed. In repeated use of a method under its assumptions, intervals at a stated confidence level are intended to cover the target value at that rate. This general interpretation does not supply missing trial results.

What does the single-arm design imply?

A single-arm study estimates outcomes among participants receiving the study regimen. It does not randomize participants between the regimen and a comparator, so the study cannot directly balance measured and unmeasured prognostic factors across treatment groups. A comparison with historical standards requires external data and assumptions about comparability, including differences in eligibility, assessment, follow-up, and background care. The registry states that improvement against historical standards was the question, but does not report the benchmark or a method for making that comparison.

Randomization is a design strategy that, on average, helps make treatment groups comparable with respect to prognostic factors. Since this study was not randomized and lists no control arm, a causal contrast between treatments is not available from the design alone. This does not prevent description of outcomes in the treated group; it limits what can be attributed specifically to treatment relative to another option.

How should progression criteria be understood?

The registry cites RECIST 1.1 and describes target-lesion criteria for complete response, partial response, progressive disease, and stable disease. In general, standardized response criteria aim to make disease assessments more consistent by defining how changes in lesions are categorized. The record describes partial response as at least a 30% decrease in the sum of target-lesion longest diameters from baseline, and progressive disease as at least a 20% increase from the smallest sum recorded since treatment started or the appearance of new lesions. These definitions describe response classification; they do not provide this trial’s response results.

The registry does not report the schedule of disease assessments, handling of unevaluable assessments, confirmation requirements, or review procedures. These details can affect the timing and classification of progression, so their absence limits the methodological reconstruction possible from the record.

What does the record say about the statistical plan?

The registry record lists 0 analyses and does not report a statistical analysis plan, model, censoring rules, interim analyses, or sample-size justification. Those elements should not be inferred. The record also includes a planning-illustration description based on an observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; time-to-event illustrations are described as using a total number of events, while other illustrations use participants per arm. These are planning assumptions, not evidence that the trial used those assumptions in its design or analysis.

In general, sample-size calculations estimate the amount of information a future study may need under specified assumptions. They are not observed power and do not recover the missing outcome. For a single-arm study, any hypothetical two-arm planning illustration is distinct from the actual study design. No numerical planning outputs are reported in the registry record.

Why can over-dispersion matter?

In general, over-dispersion occurs when observed variability is greater than a statistical model expects. If a model assumes less variation than the data contain, standard errors and confidence intervals may be too small, making estimates appear more precise than warranted. Methods that allow extra variation can improve uncertainty estimates when their assumptions suit the data. This concept is general; the registry record does not report whether over-dispersion was assessed or relevant to any analysis in this trial.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists secondary outcomes for objective response rate, disease control rate, overall survival, and the number of patients with grade 3 or higher treatment-related adverse events. It records 5 outcome measures and 0 analyses. The available record does not give numerical results for these secondary outcomes or explain how their analyses were handled.

Multiplicity arises when several outcomes, time points, subgroups, or statistical tests are examined. In general, testing many hypotheses can increase the chance that at least one result appears noteworthy by chance. Prespecified testing sequences or adjustments can address some multiplicity concerns, but the registry record does not report whether such procedures were planned or applied. It is not possible to infer the testing hierarchy from the list of outcomes alone.

The registry does not report missing-data amounts, reasons for missing assessments, endpoint-specific analysis populations, or rules for handling participants without evaluable follow-up. For a time-to-event endpoint, censoring and incomplete follow-up need defined handling; the record does not describe those methods. For response outcomes, missing or unevaluable assessments can also affect interpretation, but no handling rules are reported.

Analysis populations such as intention-to-treat or per-protocol groups are not reported in the registry record. Neither the population included in the primary endpoint analysis nor the number contributing to that analysis can be established from the listed enrollment alone. The record also does not report an interim-analysis plan, stopping rules, or a statistical analysis plan. These omissions leave important questions about multiplicity, missingness, and analysis populations unresolved; they should not be filled by assumptions based on common practice.

11. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

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: febrile neutropenia (10.0%), anaphylaxis (3.3%), colitis (3.3%), diarrhea (3.3%), endocrine disorders (3.3%).

The registry lists one intervention group, pembrolizumab (MK-3475) plus mFOLFOX6. For serious adverse events, it records 30 at risk and 7 affected. For other adverse events, it records 30 at risk and 30 affected, and it lists 10 deaths. These are registry entries and should be read according to their respective categories; the record does not explain further details about event collection or attribution.

Among the listed serious adverse event terms, febrile neutropenia is reported as 10.0%. Anaphylaxis, colitis, diarrhea, endocrine disorders, and fever are each reported as 3.3%. These percentages are attached to the registry’s serious-event terms; they do not describe all possible events or establish a comparative risk.

Seriousness and treatment relatedness are distinct concepts. The secondary outcome list names grade 3 or higher treatment-related adverse events, but the serious-event listing alone does not establish that every serious event was judged related to treatment. The record does not provide a detailed causality breakdown, event timing, duration, resolution, or the relationship of the listed deaths to treatment. The death count should therefore not be interpreted as a treatment-related death count.

In general, adverse-event counts and percentages describe events recorded under particular definitions and observation periods. They do not by themselves show whether the regimen caused an event or how its safety compares with another treatment. No concurrent comparator is listed in this study.

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

The record establishes that a completed, open-label, single-arm phase II study evaluated pembrolizumab with mFOLFOX6 in advanced colorectal cancer and named median progression-free survival as its primary endpoint. It also provides a progression definition, the intended dosing schedule following the safety run-in cohort, a stated endpoint timeframe, and selected adverse-event entries.

Those details do not establish the magnitude of the primary outcome. The registry record does not report the mPFS estimate or its 80% confidence interval. It also does not report numerical results for the listed secondary outcomes. A reader therefore cannot use this record to quantify progression-free survival, response, overall survival, or the comparative effect of the combination against historical standards.

The single-arm design is important to the meaning of any outcome that might be reported elsewhere. An observed outcome in the treated group would describe that group under the study’s follow-up and assessment procedures. Without a concurrent control group, it would not, by itself, show how outcomes would differ under another treatment. Historical comparisons require an identified reference and assumptions about whether populations and methods are sufficiently comparable; these are not described in the record.

Safety entries likewise require care. The registry reports serious and other adverse-event figures, as well as deaths and selected serious event terms. The entries do not establish treatment causality, comparative safety, or the relationship between deaths and treatment. They should not be collapsed into a single judgment about safety.

In short, the registry supports a description of the intended study question, design, endpoint definition, and selected recorded safety information. It does not support a conclusion that the combination improves or fails to improve progression-free survival, nor a conclusion about efficacy or comparative safety. These are limits of the reported information and design, not findings about an unreported outcome.

13. Why This Trial Matters Statistically

This trial illustrates why a clearly named endpoint and a reported result are both needed for statistical interpretation. Median progression-free survival is defined in the registry, including the event criteria and a confidence-interval measure, but the estimate and interval are absent. Without them, readers cannot assess the outcome’s magnitude or precision. A detailed endpoint definition cannot substitute for the numerical result.

It also illustrates the distinction between describing an outcome and estimating a treatment contrast. The single-arm design can describe outcomes in the treated group, but it does not include a randomized concurrent comparator. The stated comparison with historical standards therefore raises a separate methodological question: what reference is used, and how comparable is it? The registry does not report the reference or comparative analysis.

The record is also a reminder that study status and results reporting are different kinds of information. The study is marked completed, with a primary completion date, yet the registry lists 0 analyses. Completion status alone does not communicate the outcome or make the statistical evidence assessable.

Finally, the safety entries demonstrate why event categories and attribution matter. Serious events, other adverse events, deaths, and treatment-related events are not interchangeable categories. Counts and event terms need their definitions, observation periods, and attribution context to support sound interpretation. The record contains some safety figures but does not report all of those details.

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 main trial-specific limitation in the registry record is the absence of reported efficacy results. Although median progression-free survival is named as the primary endpoint and an 80% confidence interval is specified, the estimate and interval limits are not reported. The registry lists 0 analyses, so it does not provide a numerical primary analysis or results for the listed secondary outcomes.

The design is single-arm and open-label, with no randomized or masked comparison. This limits direct comparative inference and makes a historical comparison dependent on external information and assumptions. The record does not name the historical standards, provide their outcome data, or describe the method used to compare them.

Several analysis details needed to assess the endpoint are not reported in the registry record. These include the censoring rules, analysis cutoff, assessment schedule, independent review procedures, endpoint-specific analysis population, and handling of missing or unevaluable data. The statistical analysis plan, sample-size justification, interim analyses, and stopping rules are also not reported.

The safety information is selective rather than a full account of event definitions, timing, seriousness assessments, treatment attribution, or outcomes. The listed deaths cannot be assumed to be treatment related, and serious-event listings do not provide a comparative safety estimate. These limits restrict what can be concluded from the record, without establishing that unreported outcomes were favorable or unfavorable.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Study of Pembrolizumab in Combination With Chemotherapy for Patients With Advanced Colorectal Cancer — NCT02375672.
• Registry record last updated 2022-02-16; status last verified 2022-02.
• Results first posted 2021-10-08.
• 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.

18. Record Summary

The registry record describes a completed multicenter, open-label, single-arm phase II study of pembrolizumab (MK-3475) with mFOLFOX6 for advanced colorectal cancer. Enrollment is listed as 30, and the study description includes a safety run-in cohort. Following that cohort, mFOLFOX6 was scheduled on D1 and D15 every 2 weeks, with pembrolizumab intravenously over 30 minutes every 3 weeks.

Median progression-free survival is the primary endpoint, defined from registration to documented progression per RECIST 1.1 or death. An 80% confidence interval is specified, but the endpoint result and interval are not reported. The registry lists secondary outcomes and selected adverse-event figures, but does not report their analysis results. It records 0 analyses. No concurrent comparator is listed, so the record does not provide a randomized treatment comparison.