Study of Durvalumab and Tremelimumab After Radiation for Microsatellite Stable Metastatic Colorectal Cancer Progressing on Chemotherapy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of Durvalumab and Tremelimumab After Radiation for Microsatellite Stable Metastatic Colorectal Cancer Progressing on Chemotherapy (NCT03007407), a na phase 2 study. The primary outcome is Overall Objective Response Rate (ORR) of Dual Immune Checkpoint Blockade by RECIST 1.1.

2. Clinical Question

3. Trial Design

This is a phase II, single-group study of durvalumab plus tremelimumab after palliative hypofractionated radiation in patients with microsatellite stable metastatic colorectal cancer progressing on chemotherapy. The registry lists enrollment of 33. It names NSABP Foundation Inc as the sponsor and records the study as completed. The recorded start date is 2017-07-31, and the primary completion date is 2019-04-18.

The design has no randomized comparator: allocation is listed as not applicable, the model is single-group, and masking is none. Thus, the study describes outcomes in participants receiving the study regimen, rather than a contrast between assigned treatment groups. The contribution of the drugs cannot be separated from the preceding radiation or from other influences on outcomes using a concurrent control comparison in this design.

The primary outcome is overall objective response rate by RECIST 1.1, assessed through treatment, up to 1.3 years. The registry also lists clinical benefit, duration of response, and adverse-event frequency as secondary outcomes. Although enrollment is recorded as 33, the adverse-event summary reports 29 participants at risk for the listed event categories. The registry record does not explain the relationship between these figures or specify the analysis populations for the response outcomes.

The summary says the study included research tests using diagnostic tumor samples, fresh tumor samples from a metastatic site, and blood samples. Their analytic aims and results are not reported in the registry record. The record also does not describe the treatment schedule, dose levels, eligibility details, or a statistical analysis plan. These omissions limit how fully the design and analysis can be reconstructed from the registry information.

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.

Population: Participant could identify as more than one race/ethnicity and may be counted in more than one race/ethnicity category.

6. Primary Endpoint

The registry lists the primary outcome as Overall Objective Response Rate (ORR) of Dual Immune Checkpoint Blockade by RECIST 1.1, measured over Through treatment, up to 1.3 years.

The primary endpoint is the overall objective response rate, defined as the percentage of participants with a complete response or partial response by RECIST 1.1. The assessment period is through treatment, up to 1.3 years. RECIST provides standardized rules for classifying changes in measurable disease, making it possible to describe response using prespecified categories rather than an informal impression.

For this study, complete response requires disappearance of all non-target lesions and normalization of the tumor marker level; lymph nodes must be <10mm on the short axis. Partial response requires a decrease of 30% or more in the sum of target-lesion diameters relative to baseline. The criteria also define progressive disease as an increase of 20% or more from the smallest sum on study, with an absolute increase of 5mm or more; a new lesion also counts as progression. Stable disease is neither sufficient shrinkage for partial response nor sufficient increase for progression.

A response rate is a proportion, and its confidence interval communicates uncertainty around the estimated proportion. The registry names a 95% confidence interval as the measure of dispersion. However, it does not report the observed response rate, confidence interval, or response count. The endpoint definition is available, but the magnitude and precision of the primary outcome cannot be assessed from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a binary response classification at the participant level: a participant either meets the RECIST definition of complete or partial response, or does not. A response rate summarizes the share of participants classified as responders. A confidence interval around that rate expresses the uncertainty associated with estimating a proportion from a finite group; it is not a range containing a known individual treatment effect. The registry specifies a 95% confidence interval, but does not report the estimate or interval itself.

How does RECIST define response?

RECIST 1.1 applies rules to target lesions, non-target lesions, lymph nodes, and new lesions. In this record, complete response requires disappearance of all non-target lesions and normalization of the tumor marker level, with all lymph nodes <10mm on the short axis. Partial response requires a decrease of 30% or more in the sum of target-lesion diameters from baseline. These definitions make the endpoint more reproducible than a judgment without shared criteria, although the registry does not report details of assessment procedures or review.

The definition also distinguishes stable disease from progression. Progressive disease requires an increase of 20% or more in the sum of target-lesion diameters from the smallest sum on study, with an absolute increase of 5mm or more; the appearance of a new lesion also counts as progression. Stable disease is the category that meets neither the response threshold nor the progression threshold. A participant with stable disease is not counted as an objective responder under the stated primary endpoint.

What does the single-group design allow?

Because the study has a single-group model, there is no randomized comparison to estimate a treatment contrast. A response proportion in the treated group is descriptive. It does not show how outcomes would differ under a comparator, and it cannot distinguish effects of the study drugs from the preceding radiation or other influences. Randomization, when used, can help balance measured and unmeasured prognostic factors between groups; this record does not describe a randomized design.

Masking is listed as none. In general, knowledge of treatment assignment can influence some aspects of care, assessment, or reporting. RECIST offers defined classification rules, but the registry record does not say whether assessments were independently reviewed or how potential assessment bias was addressed. Those details are not reported in the registry record.

What do the adverse-event counts mean?

The safety summary gives counts at risk and counts affected for serious events, other events, and deaths. Counts describe what was recorded in the summarized group; they do not establish that treatment caused an event. A zero count of serious events in the reported group does not prove that the risk is zero, particularly without details of event definitions, follow-up completeness, and the analysis population. General statistical principle: when events are uncommon, estimates based on limited observations can be imprecise even when no event is observed.

The summary lists 29 at risk for both serious and other adverse events, with 0 serious events and 29 other adverse events. It records 6 deaths. The registry does not provide event-level timing, causes, attribution, or the serious adverse-event terms. The relationship between the enrollment count and safety denominator is also not explained. These limits prevent a fuller assessment of the safety data from the summary alone.

How should planning calculations be read?

The registry describes sample-size planning illustrations based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The target effect is described as a stated fraction of the observed effect. Such calculations are hypothetical planning illustrations, not the trial’s observed power and not a description of the actual single-group design. The registry’s design has no randomized two-arm allocation, and the record does not report the trial’s own sample-size justification or design assumptions.

General statistical principle: a planning calculation estimates the sample size that might be needed under assumed conditions; it does not indicate the probability that a completed study’s result is true or that a treatment works. A calculation based on an observed effect can also be optimistic if that estimate is unstable. The specific numerical planning outputs and the fractions used are not reported in the registry record, so they cannot be evaluated from the information available.

The record does not report the statistical method used to calculate the response confidence interval, the handling of unevaluable participants, or the rules for missing response assessments. It also does not report an analysis population for the primary outcome. Those methods are essential to reproducing the estimate, but the primary estimate itself is not reported either.

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 and Tremelimumab")
n      <- c(21)
events <- c(2)
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 and Tremelimumab  21  2  9.5  1.2  30.4  2.7  28.9  2.9  29.2

11. Multiplicity, Missing Data and Analysis Populations

The registry identifies overall objective response rate as the primary outcome and lists clinical benefit, duration of response, and adverse-event frequency as secondary outcomes. It records four outcome measures in total. Assessing several outcomes can raise multiplicity concerns: when many hypotheses are tested, the chance of at least one apparently notable result can increase. The registry record does not report a multiplicity adjustment, a hierarchy among secondary outcomes, or a plan for interpreting multiple tests. It is therefore not possible to determine whether or how the analyses accounted for multiple outcomes.

The record does not report how missing response assessments, unevaluable participants, or withdrawals were handled in the primary analysis. It also does not specify whether the response rate used all enrolled participants, participants who received treatment, or another analysis population. Enrollment is listed as 33, while the adverse-event summary lists 29 at risk; the reason for the difference is not reported in the registry record. That safety denominator should not be assumed to be the response denominator.

Details about imputation, sensitivity analyses, and assumptions about missing data are not reported in the registry record. Missing outcomes can affect a proportion if the chance of being observed is related to response. Without the analysis population and missing-data rules, the response estimate would be difficult to reproduce even if it were reported. In this case, the primary response estimate and confidence interval are themselves not reported, further limiting interpretation.

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 5% reporting threshold.

Most frequent non-serious adverse event terms

The adverse-event summary for the durvalumab and tremelimumab group reports 29 participants at risk for serious adverse events, with 0 affected. It reports 29 participants at risk for other adverse events, with 29 affected. The same summary records 6 deaths. These are counts in the registry’s safety summary, not comparisons against a control group.

A count of 0 serious events does not establish absence of risk. It describes the reported observations in the group, and the registry does not give event terms, detailed follow-up, or the criteria used to classify events in this summary. The record also does not indicate whether any death was considered related to treatment or provide causes of death. No causal interpretation should be attached to the death count from these figures alone.

The registry lists enrollment of 33, whereas the safety summary reports 29 at risk. The reason is not reported in the registry record. The two counts should not be treated as interchangeable or used to infer who was included in other analyses. The secondary outcome description refers to adverse events assessed by CTCAE 4.0 from the beginning of treatment to 90 days after the last dose, but the summarized counts do not provide event-level detail.

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

The study asks whether objective responses were observed after durvalumab and tremelimumab were given following palliative hypofractionated radiation to patients with microsatellite stable metastatic colorectal cancer progressing on chemotherapy. Its primary outcome uses RECIST 1.1, with a stated assessment period through treatment, up to 1.3 years. The registry record provides the endpoint definition, but not the observed response percentage or its 95% confidence interval. Thus, the central response result cannot be judged from the record.

The design is single-group and unmasked. It can describe outcomes among participants who received the regimen, but it does not provide a controlled estimate of how outcomes compare with another approach. Without a concurrent comparator, observed responses cannot be attributed to the study drugs alone, especially because radiation preceded treatment. The record does not support a conclusion that the regimen improves outcomes relative to any alternative.

The adverse-event counts provide limited descriptive safety information: 0 serious adverse events among 29 at risk, 29 with other adverse events among 29 at risk, and 6 deaths. They do not provide event details or a basis for causal attribution. The registry also does not explain why the safety denominator differs from recorded enrollment.

Statistical interpretation depends not only on an endpoint definition but also on the observed estimate, its uncertainty, the analysis population, and the handling of missing data. The registry record does not report the primary estimate, analysis population, missing-data methods, or statistical plan. It therefore permits a description of what was intended to be measured and selected safety counts, but not a quantitative judgment of response or a comparative treatment effect.

14. Why This Trial Matters Statistically

This trial raises a useful statistical distinction between measuring an outcome and demonstrating a treatment contrast. RECIST 1.1 gives a structured definition of objective response, and the registry specifies a 95% confidence interval as the measure of precision. Yet the response estimate and interval are not reported in the registry record. Knowing the measurement rules is not enough to assess the magnitude or uncertainty of the result.

The single-group design also shapes the claims that can be supported. A response proportion can characterize outcomes in the treated group, but without a concurrent comparator it cannot establish how outcomes would differ under another strategy. This is especially important when the treatment sequence includes both radiation and the study drugs. The design does not isolate the contribution of each component.

The safety summary illustrates a related point: counts require context. The record reports participants at risk, participants with events, and deaths, but does not provide event-level descriptions or attribution. Enrollment and the adverse-event denominator differ, with no explanation in the record. Clear analysis populations and outcome definitions are essential for readers to understand what a count represents.

Finally, the registry lists several outcomes but does not report a statistical plan for multiplicity, missing data, or the primary analysis population. These details affect reproducibility and interpretation. The trial is statistically informative as an example of why endpoint definitions, design, precision, and analysis transparency all matter; the available registry record does not permit a complete assessment of efficacy results.

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 interpreting efficacy is that the registry record does not report the primary response estimate, confidence interval, or response count. Although the endpoint definition is detailed, it is not possible to assess the observed response proportion or its precision. The listed secondary outcomes likewise have no reported numerical results in the available record.

The single-group, unmasked design has no concurrent comparator. It therefore cannot estimate a comparative treatment effect or separate the contribution of durvalumab and tremelimumab from the preceding radiation. The registry does not describe blinded outcome review, so the extent to which response assessments were protected from knowledge of treatment is not reported.

The record lists enrollment of 33 and an adverse-event denominator of 29, without explaining the difference. It does not report the analysis population for the primary outcome, the handling of missing or unevaluable response assessments, or the use of imputation or sensitivity analyses. These omissions limit reproducibility and make the relationship between enrollment and outcome denominators unclear.

Other details relevant to interpretation are not reported in the registry record, including the full eligibility criteria, treatment doses and schedule, statistical analysis plan, multiplicity strategy, and event-level safety information. The serious adverse-event summary has no listed terms, and the 6 deaths are not described by cause or attribution. These limits mean the available record supports only a partial description of the study design and selected safety counts.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Study of Durvalumab and Tremelimumab After Radiation for Microsatellite Stable Metastatic Colorectal Cancer Progressing on Chemotherapy — NCT03007407.
• Registry record last updated 2022-12-05; status last verified 2022-11.
• Results first posted 2022-12-05.
• 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 record describes a completed phase II, single-group study of durvalumab and tremelimumab after palliative hypofractionated radiation in patients with microsatellite stable metastatic colorectal cancer progressing on chemotherapy. Enrollment is listed as 33. The primary outcome is overall objective response rate by RECIST 1.1 through treatment, up to 1.3 years, with a 95% confidence interval specified.

The record does not report the primary response estimate or confidence interval. It lists clinical benefit, duration of response, and adverse-event frequency as secondary outcomes, without numerical results in the available information. The safety summary reports 0 serious adverse events among 29 at risk, 29 participants with other adverse events among 29 at risk, and 6 deaths. Event terms, causes, and attribution are not reported. The registry does not explain the difference between enrollment and the safety denominator or provide the analysis population and missing-data methods.