Durvalumab for MSI-H or POLE Mutated Metastatic Colorectal Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Durvalumab for MSI-H or POLE Mutated Metastatic Colorectal Cancer (NCT03435107), a na phase 2 study. The primary outcome is Objective Response Rates (RECIST 1.1).

2. Clinical Question

3. Trial Design

This was a phase II, single-group study of durvalumab in metastatic colorectal cancer. The registry lists Asan Medical Center as the sponsor, records the study as completed, and gives enrollment as 33. The recorded study dates are 2018-04-12 for start and 2023-04-26 for primary completion. The record identifies no allocation to different treatment groups and no masking. Because there was one treatment group, the design does not create a randomized, concurrent comparison.

The intended population was molecularly selected: mismatch repair deficient or microsatellite instable disease, or POLE-mutated disease. The registry summary describes participants as previously treated and refractory to fluoropyrimidines, irinotecan, and oxaliplatin, with or without targeted agents. These eligibility features matter when considering how well the results, if reported, might apply to other colorectal cancer populations. The registry record does not report details of recruitment, eligibility screening, or the characteristics of those who entered the analysis.

The primary outcome was objective response rate under RECIST 1.1. The stated assessment schedule begins with a first measurement at 8weeks after first administration, followed by assessments every 8weeks until disease progression or death from any cause, with assessment up to 46months. The registry record does not provide the actual follow-up duration for participants or explain how missing assessments were handled.

Disease control rate is listed as a secondary outcome. The record reports two outcome measures but does not give an analysis count. No objective-response result, disease-control result, confidence interval, or statistical test is reported in the registry record. The study's phase and single-group structure describe its design; they do not by themselves establish the precision or strength of any treatment effect.

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

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 Objective Response Rates (RECIST 1.1), measured over First measurement should be at 8weeks from first administration.After first measurement, it should be followed up at every 8weeks until date of progression disease or date of death from any cause, whichever came first, assessed up to 46months...

The primary endpoint is objective response rate (ORR) according to RECIST 1.1. It is a binary outcome at the participant level: a participant either meets the stated response definition or does not. The endpoint counts complete and partial responses. Complete response is defined as disappearance of all target lesions; partial response is defined as at least a 30% decrease in the sum of the longest diameters of target lesions. Assessments are based on CT or MRI, according to the registry description.

ORR is used to describe the proportion of participants whose measured tumors meet a prespecified response criterion. In a single-group study, it provides a summary of response observed among participants receiving the study intervention. It does not, by itself, estimate the difference from another treatment, and it does not capture every possible dimension of clinical outcome. For example, response rate is distinct from the time until progression or death.

The registry specifies the assessment timing: a first measurement at 8weeks after first administration, then every 8weeks until progression or death, with assessment up to 46months. The registry record does not report the number of participants who had a complete or partial response, the resulting response percentage, or its 95% confidence interval. Thus, the endpoint's definition is available, but its observed value is not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods concern how outcomes are defined, summarized, compared, and interpreted. The registry record provides the primary endpoint definition and describes a single-group design, but it does not report the statistical analysis plan, a primary-endpoint estimate, or any analysis results. The explanations below are general guidance for reading a binary response outcome; they do not imply that any particular unreported method was used in this study.

What does a binary response measure?

General: A binary endpoint classifies each participant into one of two outcome states, such as meeting or not meeting a response criterion. An outcome percentage expresses the share of participants classified as having the event of interest within a defined analysis population. To interpret that percentage, a reader needs both the number meeting the definition and the denominator used. It also matters whether participants without an evaluable assessment were counted as nonresponders, excluded, or treated in another way. Those analysis details are not reported in the registry record.

For this study, the response definition is based on RECIST 1.1 and combines complete and partial response. The registry gives the imaging assessment schedule, but does not report a response count or percentage. It also does not state whether response assessments were reviewed by an independent assessor, how disagreements were resolved, or what rules governed incomplete imaging data. Those details cannot be inferred from the endpoint name alone.

What does a confidence interval add?

General: A confidence interval describes the uncertainty around an estimate under a specified statistical procedure. For a binary response percentage, it helps show how much the estimated response frequency could vary because the study observed a limited group of participants. A narrower interval generally indicates greater precision, while a wider interval indicates less precision. The interval does not eliminate limitations from selection, measurement, missing data, or the lack of a comparator.

The primary outcome is described with a 95% confidence interval, but the registry record does not report the actual interval or the underlying response estimate. It also does not identify the method used to calculate that interval. As a result, the precision of the primary result cannot be evaluated from the registry record.

How does the single-group design affect inference?

General: A single-group study estimates outcomes within the treated group. It does not provide a concurrent group to account for what might have happened to similar participants under another intervention. A comparison with an external benchmark, if one is made, can be affected by differences in eligibility, disease severity, assessment timing, and other features. The registry record does not report a formal benchmark comparison or a statistical test against a reference value.

Random allocation is a method used in comparative trials to help make treatment groups comparable with respect to measured and unmeasured characteristics. It is not applicable to the recorded single-group model. Masking is recorded as none. These design descriptions do not establish that outcome assessment was biased, but they identify features relevant to interpreting a response estimate if one becomes available.

What can planning calculations tell a reader?

General: Sample-size planning calculations describe the number of participants that might be needed under specified assumptions to detect a chosen effect with a chosen statistical power and significance level. A planning illustration is not the same as an observed result, and it is not observed power. When an illustration is based on an effect observed in a study, it depends on that estimated effect and should be treated as a planning aid rather than a guarantee that a future study will reproduce it.

The registry record does not report the trial's sample-size justification or statistical plan. It therefore cannot establish the original target assumptions, whether any planning calculation guided enrollment, or whether the study was designed to detect a particular response rate. No such protocol details should be inferred from the enrollment figure alone.

Why can over-dispersion matter?

General: Over-dispersion occurs when the variability in observed data is greater than a model assumes. In binary outcomes, an unsuitable model or an overly simple assumption about independent observations can lead to uncertainty estimates that are too small or too large. For a basic participant-level response proportion, the choice of interval method and the handling of clustering or repeated assessments can influence the precision estimate. The registry record does not report a model, an over-dispersion assessment, or whether repeated measurements entered the primary analysis. Repeated imaging is used to determine response status over time, but it does not by itself mean that each scan is an independent participant-level outcome.

The primary endpoint is a response proportion rather than a time-to-event endpoint. The registry record gives a follow-up assessment schedule, but does not report a time-to-event estimate or a statistical model for such an outcome. No analysis method, adjustment strategy, or sensitivity analysis is reported in the registry record. These absences limit methodological appraisal; they are not evidence that a particular procedure was or was not carried out outside the registry.

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("Durvalumab")
n      <- c(33)
events <- c(14)
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  Durvalumab  33  14  42.4  25.5  60.8  27.2  59.2  27.2  59.3

11. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when a study makes several statistical comparisons or evaluates multiple outcomes. In general, the more inferences made, the greater the chance that at least one apparently notable result could occur by chance unless the analysis plan accounts for that multiplicity. The registry lists objective response rate as the primary outcome and disease control rate as a secondary outcome. It reports two outcome measures, but does not provide the full set of outcome definitions or any analysis results. The registry record does not report a multiplicity adjustment or explain how the primary and secondary outcomes were prioritized in statistical testing.

The registry record reports no analyses. It does not describe an analysis population, such as whether the primary endpoint was evaluated among all enrolled participants, eligible participants, or participants with an adequate imaging assessment. Nor does it report how missing response assessments, withdrawal, or other incomplete outcome data were handled. These matters can affect a binary response estimate because the denominator and classification of participants without an assessment may change the result.

General: Missing-data methods make assumptions about the relationship between missing outcomes and observed information. Different assumptions can yield different estimates, so transparent reporting of the analysis population and sensitivity analyses is important. The registry record does not describe such assumptions or analyses. This should be read as a limitation of the available record, not as proof that the study had missing data or that no analyses were done outside the record.

There is no reported statistical analysis plan, interim-analysis information, or protocol-level explanation of sample-size justification in the registry record. Accordingly, the planned handling of multiplicity, missing data, and analysis populations cannot be assessed from this record.

12. 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: abdominal pain (6.1%), acute kidney injury (6.1%), dyspnea (3.0%), infection (3.0%), both knee pain (3.0%).

The safety fields list 13 participants affected by serious adverse events among 33 at risk, and 28 affected by other adverse events among 33 at risk. The registry lists 15 deaths in the durvalumab group. These are separate recorded figures; the record does not explain whether categories overlap or provide event timing, duration, severity grading, or causal attribution. A death count should not be treated as a count of deaths caused by treatment.

The registry's listed serious adverse-event terms include abdominal pain and acute kidney injury, each shown as 6.1%, and dyspnea, infection, both knee pain, and anorexia, each shown as 3.0%. These are the terms and percentages presented in the record. The record does not provide the counts behind each percentage, details about coding, or a complete account of all safety events in the available information.

General: Safety percentages describe events under the definitions and collection rules used in a study. They do not, on their own, establish whether an event was caused by an intervention. Interpretation also depends on the duration and completeness of follow-up, the population at risk, event severity, and comparison with an appropriate control group. This study has no concurrent comparator, and the registry record does not report an event-attribution analysis. The figures therefore describe recorded safety experience, not a comparative risk estimate.

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

The record establishes that a completed phase II, single-group study evaluated durvalumab in a selected population with metastatic colorectal cancer described as mismatch repair deficient, microsatellite instable, or POLE mutated. It also specifies objective response rate under RECIST 1.1 as the primary outcome and disease control rate as a secondary outcome. These design and endpoint details clarify what the study set out to assess.

However, the registry record does not report the objective-response result. No response percentage, response count, or 95% confidence interval is given. Therefore, the record cannot establish the observed frequency of response or how precisely it was estimated. The secondary disease-control result is also not reported. The absence of these results in the record should not be recast as a finding of no response or no disease control.

The single-group design also limits what a response result could establish. A response frequency among participants receiving durvalumab would not, by itself, show how outcomes compare with another treatment. There is no randomized comparator group in the recorded design, so treatment differences cannot be estimated directly from this study's structure.

The safety fields report serious adverse events, other adverse events, and deaths, but do not establish causation. The record does not report attribution of deaths or adverse events to treatment, nor a comparative safety estimate. These figures are descriptive and should be read with their reporting and follow-up limitations in mind.

In general, a trial result is most informative when the outcome estimate, its uncertainty, the analysis population, missing-data handling, and statistical methods are reported together. Those efficacy and analysis details are not reported in the registry record. The appropriate interpretation is consequently limited: the design and intended questions are documented, while the primary efficacy answer and its precision remain unavailable from this record.

14. Why This Trial Matters Statistically

This trial is statistically notable as a single-group phase II study with a molecularly selected population and a binary imaging-based primary endpoint. The design makes the distinction between describing an observed outcome and estimating a comparative treatment effect especially important. An objective response rate, if reported, would summarize the frequency of RECIST-defined response in the treated study group; it would not be a randomized comparison.

The endpoint definition also illustrates why clear operational criteria matter. Complete and partial response are combined into one primary measure, with imaging assessment at defined intervals. A reader needs the response numerator, analysis denominator, and uncertainty interval to understand both the estimated response frequency and its precision. The registry record gives the definition but not those efficacy results, leaving the central statistical question unanswered in the record.

The molecular eligibility criteria focus the study on mismatch repair deficient, microsatellite instable, or POLE-mutated metastatic colorectal cancer after prior treatment. That focus can make a result relevant to a particular clinical research population, while also limiting its applicability to people who do not meet those criteria. The registry record does not provide participant characteristics needed to evaluate how closely the enrolled group reflects the intended population.

The safety data also demonstrate the importance of distinguishing event description from attribution. Counts of participants affected and listed adverse-event terms can characterize recorded experience, but they do not estimate excess risk without a suitable comparison. In this study's single-group design, the registry record offers no concurrent basis for a controlled safety contrast.

Ultimately, the study's statistical importance cannot be judged from design labels alone. The missing primary response result, uncertainty estimate, and analysis details prevent a full assessment of efficacy evidence. The record still permits readers to identify the intended question, the population, the endpoint rules, and important limits on inference.

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 main limitation for statistical interpretation is that the registry record does not report a primary objective-response result. Without the response count, percentage, and confidence interval, neither the observed rate nor its precision can be evaluated. The disease-control result is also absent, despite being listed as a secondary outcome.

The study is recorded as single group, with no allocation to a comparator and no masking. Its design therefore does not support a direct randomized comparison. Any response or safety description would pertain to the enrolled group and would not, by itself, estimate a difference from another treatment. The selected eligibility criteria also narrow the population to metastatic colorectal cancer with specified molecular features and prior treatment history.

The registry record does not report baseline characteristics, the analyzed population, the number of participants with evaluable imaging, or how incomplete outcome data were handled. It does not specify the statistical methods used to calculate the primary endpoint's stated 95% confidence interval. Nor does it report an analysis plan, multiplicity procedures, interim-analysis information, or a sample-size justification. These gaps prevent assessment of whether the outcome analysis followed prespecified procedures.

For safety, the record provides counts and selected serious adverse-event terms but not detailed event-level information, timing, severity, attribution, or a comparative analysis. The 15 deaths are not accompanied by a cause or attribution in the available information. Thus, the safety fields should be treated as descriptive rather than causal evidence.

These limitations refer to information not reported in the registry record. They do not establish that the underlying study lacked additional analyses, procedures, or reporting outside the record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Durvalumab for MSI-H or POLE Mutated Metastatic Colorectal Cancer — NCT03435107.
• Registry record last updated 2024-10-15; status last verified 2024-10.
• Results first posted 2024-10-15.
• 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 records a completed phase II, single-group study of durvalumab in previously treated metastatic colorectal cancer described as mismatch repair deficient, microsatellite instable, or POLE mutated. Enrollment is listed as 33, with Asan Medical Center as sponsor. There was no concurrent comparator, and masking is recorded as none.

The primary outcome is objective response rate under RECIST 1.1, using complete and partial response definitions. Disease control rate is listed as a secondary outcome. The registry gives an assessment schedule beginning at 8weeks after first administration and continuing every 8weeks until progression or death, with assessment up to 46months.

The record reports safety counts, including 13 participants affected by serious adverse events, 28 affected by other adverse events, and 15 deaths. It does not report the primary response estimate, disease-control result, confidence interval, or statistical analysis. The central efficacy result and its precision are therefore not reported in the registry record.