Combination Immunotherapy in Colorectal Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Combination Immunotherapy in Colorectal Cancer (NCT05571293), a non-randomized phase 2 study. The primary outcome is Cohort A and B: Pathological Overall Response (pOR) Rate Determined by Analysis of Tissue Resected During Surgery Reported by Cohort.

2. Clinical Question

The clinical question is exploratory: it concerns tissue response after a pre-surgical combination regimen, rather than a randomized comparison of treatment against a control. The primary endpoint is determined from the resected tumor, and the registry specifies that response is reported by cohort. That structure makes the cohort definitions important to interpretation, particularly because the number of balstilimab doses differs between Cohort A and Cohorts B and C, and Cohort C has a specified tumor subtype.

The registry describes a pilot study of investigational drugs and surgery. It does not report a comparative estimate that would establish whether one cohort’s regimen is more effective than another’s. Nor does the registry record report a numerical primary response result in the information available for this analysis. The question that can be addressed is therefore what endpoint was intended and what design features shape its interpretation, not whether the combination improves outcomes relative to an alternative.

3. Trial Design

This completed phase 2 study enrolled 26 participants and used a non-randomized, parallel design with no masking. The study began on 2023-03-17, and primary completion was recorded as 2024-11-14. The sponsor was Weill Medical College of Cornell University. The condition listed is colorectal neoplasms.

There were three cohorts. Cohort A was assigned two doses of balstilimab and one dose of botensilimab before surgery. Cohorts B and C were assigned four doses of balstilimab and one dose of botensilimab before surgery. The brief summary describes Cohort C as enrolling people with dMMR/MSI-High colorectal cancer. The detailed arm descriptions state that balstilimab was administered approximately 2 weeks apart and that botensilimab was given on the same day as the first balstilimab dose. The study used intravenous administration.

Because allocation was non-randomized, cohort membership should not be treated as an interchangeable assignment to comparable groups. The cohorts differ in regimen duration, and Cohort C has a tumor subtype requirement. Any observed differences could reflect these design features or other differences among participants; the registry record does not report a randomized comparison that would separate such explanations.

The primary endpoint is titled as a pathological overall response rate for Cohorts A and B, determined by analysis of tissue resected during surgery and reported by cohort. Its timeframe is immediately following surgical resection of the primary tumor. The registry also lists secondary outcomes concerning changes in minimal residual disease assessed using ctDNA before surgery and 30 days after surgical resection. The full statistical plan, details of cohort assignment, and rationale for the enrollment target are not reported in the registry record.

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: The analysis population includes all eligible enrolled patients who met protocol-defined criteria and were evaluable for the primary outcome. Two enrolled subjects were excluded from the primary analysis: one subject was determined to be ineligible due to incomplete tumor resection, and one subject was determined to be ineligible after enrollment following identification of metastatic disease. Patients from cohort C were not included in the analysis of this outcome measure per study protocol.

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: 2 patients in cohort A were found to be ineligible due to identification of metastatic disease or incomplete resection.

6. Primary Endpoint

The registry lists the primary outcome as Cohort A and B: Pathological Overall Response (pOR) Rate Determined by Analysis of Tissue Resected During Surgery Reported by Cohort, measured over Immediately following surgical resection of the primary tumor.

The primary endpoint concerns pathological response assessed in tissue removed during surgery. The registry describes examination of the resected tumor in its entirety, estimation of the percentage of residual viable tumor in the macroscopically identifiable tumor bed using routine hematoxylin and eosin staining, and classification of regression with the Mandard tumor regression grading system. The endpoint is titled as a pathological overall response rate for Cohorts A and B and is reported by cohort.

The registry defines major pathologic response as at most 10% residual viable tumor cells, corresponding to at least 90% response and Mandard tumor regression grade 1 (CR) or 2 (near-CR). Partial response is defined as at least 50% tumor regression. Because the registry notes a lack of consensus on partial response after immunotherapy, tumors with >50% and <90% residual viable tumor are labeled as having “10-50% tumor regression,” following the definition described in the record. In the specified pMMR responder analysis, this subgroup is included with nonresponders.

Pathological response is used because it evaluates tumor tissue directly at resection and can describe the degree of regression after pre-surgical treatment. It is distinct from a longer-term clinical outcome. The registry record does not report the primary response counts or rates in the information available for this analysis, so the endpoint definition can be described but its observed result cannot be assessed.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a response rate describe?

General statistical explanation: a response rate is a proportion of participants meeting a prespecified outcome definition among a stated analysis population. Its meaning depends on both the response criteria and the denominator. Here, the registry defines pathological response through residual viable tumor and tumor regression grading, and identifies an evaluable primary analysis population. However, the response counts, rates, and cohort denominators are not reported in the registry record. Without them, neither the observed proportion nor its precision can be interpreted.

Why does the analysis population matter?

General statistical explanation: excluding participants from an analysis can change the population represented by the result. An evaluable-participant analysis can be informative when the endpoint requires an assessment that cannot be made for everyone, but it should be distinguished from an analysis of all enrolled participants. The registry says that two enrolled subjects were excluded from the primary analysis. It identifies incomplete tumor resection as the reason for one exclusion; the reason for the other is not reported in the registry record. The endpoint denominator and details of any other missing assessments are not reported in the registry record.

Why is this not a treatment comparison?

General statistical explanation: random assignment helps make groups comparable, on average, for measured and unmeasured characteristics. In a non-randomized study, cohort membership may be associated with participant characteristics or treatment decisions. That can make differences in outcomes difficult to attribute to the regimen. This study is non-randomized and has no like-for-like control comparison. Its cohorts also differ in planned balstilimab exposure, and Cohort C has a specified dMMR/MSI-High population. Any cohort contrasts should therefore be read descriptively unless a suitable analysis and its assumptions are reported. An adjusted statistical plan is not reported in the registry record.

How should uncertainty be understood?

General statistical explanation: a confidence interval describes a range of values reasonably compatible with the data under the method used. It conveys precision as well as direction and magnitude; a wide interval indicates substantial uncertainty. No confidence interval for the primary response endpoint is reported in the registry record. Small cohort counts, particularly the single participant listed for Cohort C in the safety information, limit how much a cohort-specific proportion can say about a wider population. No new estimates can be inferred without the underlying outcome data.

What do the safety counts represent?

The safety information distinguishes participants with serious adverse events from participants with other adverse events. These counts describe how many participants in each cohort were affected, not how many separate episodes occurred. General statistical explanation: participant counts and event counts are different measures; one participant can experience more than one event. The registry record does not report enough detail to determine the relationship between the listed adverse-event terms and the cohort-level counts. The listed deaths are counts, and the record lists none in each cohort.

What is the role of the planning illustration?

General statistical explanation: a sample-size calculation is a planning exercise about the amount of information a future study might need to detect a specified effect under stated assumptions. It is not evidence that a completed study had a particular probability of finding an effect. The planning illustration described for this analysis uses a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect defined as a stated fraction of the observed effect. It is not the trial’s own design calculation. The registry record does not report that the study used these assumptions to set enrollment, and the illustration should not be interpreted as observed power.

What can be inferred from the analysis records?

The registry information lists no analysis records. That means no analysis estimates or test results are available in the record for interpretation; it does not establish that no analyses were conducted outside the registry. The statistical analysis plan, including any planned model, covariate adjustment, interim analysis, or multiplicity procedure, is not reported in the registry record. General statistical explanation: when multiple endpoints, cohorts, or subgroup questions are considered, multiplicity can increase the chance of at least one apparently notable result under a null hypothesis. A prespecified plan helps explain which results are primary and how multiple questions are handled. The available record does not permit assessment of such procedures.

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("Cohort A: Botensilimab and Balstilimab (Bot/Bal)", "Cohort B:")
n      <- c(10, 13)
events <- c(7, 8)
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  Cohort A: Botensilimab and Balstilimab (Bot/Bal)  10  7  70.0  34.8  93.3  39.7  89.2  39.0  89.1
ARM  Cohort B:  13  8  61.5  31.6  86.1  35.5  82.3  35.1  82.3

11. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary pathological response endpoint for Cohorts A and B and secondary outcomes concerning changes in minimal residual disease assessed using ctDNA before surgery and 30 days after resection. It also describes three cohorts, including a tumor-subtype cohort. The statistical plan for prioritizing endpoints, comparing cohorts, or addressing multiple tests is not reported in the registry record. No interim analyses or multiplicity adjustment are reported in the registry record.

Missing-data handling is also not described in enough detail to assess. The primary analysis population is limited to eligible enrolled patients who met protocol-defined criteria and were evaluable for the endpoint. The registry identifies two exclusions from the primary analysis, giving incomplete tumor resection as the reason for one; the other reason is not reported in the registry record. The primary endpoint denominator by cohort, the number of missing assessments, and any imputation or sensitivity analyses are not reported in the registry record.

General statistical explanation: missing outcomes can affect an estimated response rate if the likelihood of being evaluable is related to prognosis or treatment experience. Reporting the reasons for missingness and comparing analysis populations can help readers judge that risk. In a non-randomized study, clear cohort definitions and transparent handling of exclusions are also important because cohort composition may differ. The information available does not establish how these issues were handled beyond the stated evaluability requirement and exclusions.

The registry lists no analysis records. It therefore does not provide estimates or inferential results with which to assess multiplicity, missing-data assumptions, or the prespecified analysis strategy. These details are not reported in the registry record and should not be inferred from the endpoint description alone.

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 serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: diarrhea (16.7%), fatigue (8.3%), atrial fibrillation (7.7%), colitis (7.7%), dehydration (7.7%).

The registry reports serious adverse events and other adverse events by cohort, alongside the number of participants at risk. Serious adverse events affected 3 of 12 participants in Cohort A, 5 of 13 in Cohort B, and 0 of 1 in Cohort C. Other adverse events affected 4 of 12, 10 of 13, and 1 of 1, respectively. Deaths are listed as 0 for each cohort.

These figures are descriptive participant counts. They do not show that one regimen caused more or fewer events than another: the study was non-randomized, and the cohorts are not a like-for-like control comparison. The single participant listed for Cohort C also means that its observed count should not be read as a stable estimate of event frequency in a broader population.

The registry lists serious adverse-event terms with percentages, including diarrhea at 16.7%, fatigue at 8.3%, atrial fibrillation at 7.7%, colitis at 7.7%, dehydration at 7.7%, and edema in limbs at 7.7%. The record does not specify in the available information whether these terms can overlap within the same participant, or how the term-level percentages relate to the cohort-level counts. No causal conclusion should be drawn from these descriptive figures alone.

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

The record describes a completed phase 2 pilot study of botensilimab and balstilimab given before surgery for colorectal cancer. It specifies a tissue-based primary endpoint for Cohorts A and B, a separate tumor-subtype requirement for Cohort C, and safety counts by cohort. These details explain what the study was designed to examine and how the endpoint was defined.

The primary pathological response result is not reported in the registry record. There is no response count or rate, confidence interval, or test result in the available information. Consequently, the record does not establish the magnitude or precision of pathological response in any cohort. The absence of a numerical result in the record is not evidence that the observed response was either favorable or unfavorable.

The non-randomized design does not support a causal comparison between the cohorts or against an unlisted control treatment. The differing number of balstilimab doses and the Cohort C tumor subtype requirement further limit direct comparisons. Safety counts can be described as recorded, but they do not by themselves establish comparative risk or causation.

General statistical explanation: a study result is most informative when its endpoint, analysis population, estimate, and uncertainty are all available. Here, the endpoint definition and some population details are recorded, but key numerical efficacy results and analysis methods are not. Any interpretation should remain within those boundaries. The registry record supports statements about study structure and listed safety counts, not a conclusion about treatment benefit or comparative effectiveness.

14. Why This Trial Matters Statistically

This trial raises several statistical issues that recur in early clinical studies. First, its primary endpoint is based on pathology after surgery, so the definition of response and the evaluable population are central to the interpretation. The registry gives detailed criteria for classifying tumor regression, but does not report the resulting cohort-specific response estimates in the information.

Second, the non-randomized design places limits on comparisons. The three cohorts are not equivalent control groups: the planned balstilimab exposure differs, and Cohort C is restricted to dMMR/MSI-High colorectal cancer. If outcomes vary across cohorts, the design alone cannot establish that the regimen difference caused the variation. This is a useful distinction between describing outcomes and estimating a causal treatment effect.

Third, the primary analysis excludes two enrolled subjects, with incomplete tumor resection identified for one. The other reason and fuller missing-data handling are not reported in the registry record. Transparent reporting of analysis populations and exclusions is important because the denominator determines what a response proportion represents.

Finally, safety results need to be read in the context of cohort size and design. The record reports cohort-level participant counts and serious adverse-event terms, but does not provide a randomized safety comparison. The planning illustration is also distinct from the trial’s design justification and from observed power. Together, these features make the record a case study in separating endpoint definition, descriptive evidence, uncertainty, and causal claims.

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.

Several trial-specific limitations constrain interpretation. The study was non-randomized and had no like-for-like control comparison. The cohorts differed in planned balstilimab exposure, and Cohort C was defined by dMMR/MSI-High colorectal cancer. Comparisons among cohorts therefore cannot be treated as randomized estimates of regimen effects.

The registry lists an enrollment of 26, but does not report the primary response counts or rates in the information available for this analysis. It also does not report response estimates with confidence intervals or test results. The primary analysis population was restricted to eligible participants evaluable for the endpoint, and two enrolled subjects were excluded. Incomplete tumor resection was the reason for one exclusion; the other reason is not reported in the registry record.

The registry record does not report the sample-size justification, detailed statistical analysis plan, handling of missing primary outcome data, or any interim analysis plan. The ctDNA outcomes are listed, but their results and analysis methods are not reported in the registry record. The available safety information gives cohort-level counts and selected serious adverse-event terms, but does not clarify whether term counts overlap or establish causal attribution.

Cohort C’s safety information covers 1 participant, limiting the interpretability of its descriptive count. More broadly, the lack of reported primary outcome estimates prevents assessment of effect magnitude and precision. These limitations describe what can be evaluated from the registry record; they do not establish that analyses or additional reporting do not exist elsewhere.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Combination Immunotherapy in Colorectal Cancer — NCT05571293.
• Registry record last updated 2026-10-09; status last verified 2026-10.
• Results first posted 2026-05-22.
• 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

NCT05571293 is a completed phase 2, non-randomized, unmasked, parallel study sponsored by Weill Medical College of Cornell University. It enrolled 26 participants and examined intravenous botensilimab plus balstilimab before surgery for colorectal cancer. Cohort A received two balstilimab doses; Cohorts B and C received four. Cohort C was for dMMR/MSI-High colorectal cancer.

The primary endpoint was pathological overall response in Cohorts A and B, assessed from resected tissue immediately following surgery. Two enrolled subjects were excluded from the primary analysis; incomplete tumor resection was recorded for one, and the other reason is not reported in the registry record. Numerical primary response results and the detailed statistical analysis plan are not reported in the registry record.

Safety counts are reported by cohort, with 0 deaths listed in each. The study began on 2023-03-17, and primary completion was recorded as 2024-11-14.