Immunologic Determinants of Response to Pembrolizumab (MK-3475) in Advanced Melanoma (MK-3475-161/KEYNOTE-161): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Immunologic Determinants of Response to Pembrolizumab (MK-3475) in Advanced Melanoma (MK-3475-161/KEYNOTE-161) (NCT03407170), a na phase 2 study. The primary outcome is Mean Fraction of Cytotoxic T-lymphocytes (FCT) for Participants With Response Versus Participants With Progression.

2. Clinical Question

The study’s broader stated aim was to examine immune features associated with pembrolizumab-mediated tumor regression. The registry description lists three hypotheses: responders would have a higher baseline fraction of cytotoxic tumor-infiltrating T-lymphocytes (FCT), a higher fold-increase in FCT from baseline, and a higher Average Specific Cytotoxic T-lymphocyte Frequency Ratio (ASCTFR) than participants who did not respond. These are biological association questions within a treated group, not questions about whether pembrolizumab performs better than another treatment.

The primary outcome title narrows the comparison to mean FCT for participants with response versus participants with progression. Its stated timeframe is up to approximately 59 weeks. However, the registry record says that no primary-outcome data were collected or analyzed because of early termination. It therefore contains no estimate of the difference between those groups, and the stated biological hypotheses cannot be evaluated from the primary outcome in this record.

3. Trial Design

This was an open-label, phase II study of pembrolizumab in advanced melanoma, identified as KEYNOTE-161. The registry lists a single-group model, no masking, and allocation as not applicable. No randomized treatment allocation or active comparator is described. The study is therefore not a like-for-like comparison of treatments; its primary comparison concerns participants classified by response or progression after treatment.

The listed intervention was pembrolizumab 200 mg by intravenous infusion every 3 weeks, for up to 24 months. The registry gives a start date of 2018-06-28 and a primary completion date of 2019-08-14, and lists the study status as terminated. Enrollment is listed as 1. The registry does not report why the study was terminated or provide further detail about how the enrollment figure relates to the planned or analyzed populations.

The primary outcome was mean FCT for participants with response versus participants with progression, with a timeframe of up to approximately 59 weeks. FCT was defined using tumor-biopsy CD8+ T-cells and a predefined single-cell RNA gene signature. The study description also identifies blood and tumor analyses and lists other outcomes concerning adverse events, response, progression-free survival, overall survival, neoepitope burden, and antigenic determinants of highly functional CD8+ T-cell clones.

Despite those planned outcomes, the registry record explicitly states that no data were collected or analyzed for the primary measure because of early termination. It lists no analyses. Details such as the intended analysis population, the planned sample-size justification, the handling of missing observations, and any protocol-defined interim analyses are not reported in the registry record. The design labels alone do not establish what statistical procedures were intended.

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: No data were collected or analyzed for this outcome measure due to early termination of the 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: Zero participants are reported due to the risk of identification of a person.

6. Primary Endpoint

The registry lists the primary outcome as Mean Fraction of Cytotoxic T-lymphocytes (FCT) for Participants With Response Versus Participants With Progression, measured over Up to approximately 59 weeks.

The primary endpoint is titled as the mean fraction of cytotoxic T-lymphocytes for participants with response versus participants with progression. FCT is defined as the fraction of CD8+ T-cells expressing a predefined single-cell ribonucleic acid (RNA) gene signature, relative to the total tumor-infiltrating CD8+ T-cells isolated from tumor biopsies. The listed timeframe is up to approximately 59 weeks.

In general, a fraction expresses the size of a defined subgroup relative to a specified total. Here, the numerator is the CD8+ T-cell subset with the specified gene signature, and the denominator is the total tumor-infiltrating CD8+ T-cell population isolated from the biopsy. Using a proportion rather than a raw cell count can describe the composition of the sampled immune-cell population, though it does not by itself convey how many cells were present in the tumor.

The endpoint is used to investigate whether this immune-cell feature is associated with clinical response. Comparing the group means could describe whether average FCT differs between response and progression groups, but an association would not by itself establish that FCT causes response or that it predicts response in other patients. Interpretation would depend on the number of evaluable biopsies, measurement procedures, variability, and the analysis plan; these details are not reported in the registry record.

Most importantly, the record says no data were collected or analyzed for this outcome because the study terminated early. Thus, it provides an endpoint definition and a planned comparison, but no observed FCT means, group sizes, effect estimate, or uncertainty interval for that comparison.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record describes a single-group, open-label phase II study. These design labels identify the broad structure but do not specify an analysis method. No analysis is listed, and details such as the statistical model, test, covariate adjustment, or decision criteria are not reported in the registry record. The explanations below describe general statistical concepts relevant to the stated outcome; they are not claims about procedures used in this study.

What does the endpoint measure?

FCT is a proportion formed from tumor-biopsy cells. Its numerator is CD8+ T-cells expressing a predefined single-cell RNA gene signature, and its denominator is the total tumor-infiltrating CD8+ T-cells isolated from the biopsy. The endpoint title specifies a comparison of mean FCT between participants with response and participants with progression. A mean summarizes the average across observations, but it can be influenced by unusually high or low measurements. The record does not report the observed means, the number of usable biopsies, or the distribution of FCT values.

Biological measurements derived from biopsies can vary across participants and may also depend on tissue sampling and measurement procedures. In general, that variability matters because the observed difference between groups is only an estimate of the underlying contrast. The registry description defines the signature concept but does not report assay details, reproducibility information, or rules for selecting or processing specimens.

How would a group comparison be interpreted?

In general, a difference in group means describes the direction and magnitude of the average contrast in the analyzed sample. It does not, on its own, show that the immune measure caused the clinical outcome. Because response groups are defined by what happened after treatment rather than assigned at random, a comparison may reflect other differences among participants as well as any relationship between FCT and response. The registry record does not report an adjustment strategy or any other method for addressing potential confounding.

A confidence interval is a statistical summary of uncertainty around an estimate under the assumptions of the analysis. A narrower interval indicates greater precision, while a wider interval indicates less precision. A p-value is a measure tied to a specified null hypothesis and model; it does not express the probability that the null hypothesis is true, and it does not measure the size or importance of an effect. No confidence interval, p-value, or primary effect estimate is reported for this outcome because the record states no data were collected or analyzed.

What does over-dispersion mean in general?

For outcomes analyzed as counts or proportions, a simple statistical model may assume a particular amount of variation. Over-dispersion means that the observed variation is greater than that assumption allows. If it is ignored, standard errors may be too small and intervals or tests may convey more certainty than warranted. Whether over-dispersion is relevant depends on the actual data and analysis model. The registry does not report a model or data for the primary outcome, so it is not possible to say whether this issue arose in this study.

What do the planning illustrations represent?

Planning sample sizes are not observations from a trial and are not a measure of achieved power. The described planning illustration uses a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect to detect expressed as a stated fraction of the observed effect. For time-to-event outcomes, the illustration is based on a total number of events; for other outcomes, it is expressed as participants per arm. These are planning assumptions, not a record of this single-group study’s enrollment plan or an analysis of its participants.

Sample-size planning generally asks how many observations would be needed to detect a chosen effect under specified assumptions. The answer depends on the effect size, variability, statistical threshold, allocation, and analysis method. An assumed effect is not guaranteed to recur in a new study, and an illustration based on an observed effect should not be read as evidence that the trial itself was designed or powered that way. The registry does not report the study’s own sample-size justification.

Here, no primary outcome data were analyzed, and the registry lists no analyses. Therefore, no trial-specific estimate, standard error, interval, or hypothesis test can be interpreted. The stated endpoint and planning assumptions provide context for the intended statistical question, but they do not substitute for observed outcome data or a documented analysis plan.

10. Multiplicity, Missing Data and Analysis Populations

The study description names several immune hypotheses and the registry lists multiple additional outcomes. When many hypotheses are tested, the chance of at least one apparently notable finding can increase. In general, multiplicity procedures may control the overall error rate or otherwise prioritize a limited set of confirmatory questions. The registry record does not report a multiplicity strategy, a hierarchy among endpoints, or whether analyses were intended to be confirmatory or exploratory.

The record lists 10 outcome measures and no analyses. It does not report whether all measures were collected, how outcomes were selected for reporting, or whether any outcomes were modified after study initiation. Because the primary outcome has no analyzed data, there are no reported primary results for which a multiplicity adjustment could be assessed.

The registry explicitly states that no data were collected or analyzed for the primary endpoint because of early termination. It does not describe how missing data would have been handled, whether any biopsies or other measurements were available, or what analysis populations were intended. These details are not reported in the registry record. The enrollment listing of 1 should not be treated as the number with an evaluable primary measurement.

In general, missing observations can affect both the precision and the validity of an estimate, particularly if the likelihood of missingness is related to participant characteristics or outcomes. Methods such as complete-case analysis or imputation require assumptions and are not interchangeable. No missing-data method can be attributed to this study from the registry entry. With no primary-outcome analysis, the extent and consequences of missingness cannot be evaluated.

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

The record establishes that investigators intended to study whether tumor immune features were associated with response to pembrolizumab in advanced melanoma. It defines FCT and gives the planned comparison between response and progression groups. It also records early termination and states that no data were collected or analyzed for the primary outcome. The statistical conclusion for that endpoint is therefore that no comparison is reported, not that the groups were equal or different.

The design does not include a randomized control treatment. Even if an association between FCT and response had been reported, that type of comparison would not alone show that pembrolizumab caused the immune feature or that the feature caused response. It would also not establish how pembrolizumab compares with another treatment. Those are distinct questions requiring appropriate designs and analyses.

The record’s listed secondary outcomes and safety categories do not supply evidence of effects when no results or analyses are reported. In particular, zeros in the adverse-event group are accompanied by zero participants at risk. That registry entry cannot be interpreted as proof of an absence of adverse events in a treated population. Likewise, the enrollment figure does not demonstrate that any participant contributed to the primary endpoint.

In general, a statistically estimated association is subject to uncertainty, measurement limitations, and the assumptions of its analysis. No effect estimate or uncertainty interval is available for this primary outcome, so the magnitude and precision of a possible FCT difference remain unknown from this record. The proposed hypotheses remain untested in the reported primary analysis.

These statements are limited to the registry record. They do not establish what may have been collected or reported elsewhere, and no further results or explanation for termination are reported in the registry record. The appropriate interpretation is narrow: the registry documents the planned immune-correlate question and the absence of analyzed primary-outcome data.

12. Why This Trial Matters Statistically

This study is statistically notable because its primary question concerns a biomarker comparison within a treated group rather than a randomized comparison of treatments. That distinction changes what an estimate could mean. A difference in FCT between response and progression groups would describe an association between an immune measure and clinical outcome; it would not, by itself, identify a causal effect of pembrolizumab or establish the measure’s predictive value in another population.

The endpoint also illustrates why a biomarker needs a clear operational definition. FCT is defined as a signature-positive fraction among tumor-infiltrating CD8+ T-cells isolated from biopsies. The numerator and denominator determine what the proportion represents, while biopsy collection and assay procedures can affect measurement quality. Statistical interpretation depends on both the biological definition and the quality and completeness of the measurements.

Most importantly, a well-defined endpoint cannot answer a scientific question without analyzed observations. The record states that early termination meant no primary-outcome data were collected or analyzed. Thus, the intended comparison has no estimate or precision measure in the registry record. This is a useful distinction for readers: a registered hypothesis, an endpoint definition, an outcome listing, and an analyzed result are not interchangeable.

The record also shows why safety counts require a denominator. A listing of zero events alongside zero participants at risk cannot establish the frequency of adverse events among exposed participants. Statistical interpretation depends on knowing who was observed and for how long, not merely on reading an event count in isolation.

Finally, the planning assumptions described for sample-size illustrations should be kept separate from the study’s actual design and findings. Such calculations are planning tools under stated assumptions, not evidence of observed power or evidence that a trial met its scientific aims. In this case, the registry does not report a study-specific sample-size justification or a primary analysis.

13. 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 central limitation is that the registry record reports no data collected or analyzed for the primary endpoint because of early termination. Without analyzed FCT observations, there is no group mean, comparison estimate, or measure of uncertainty to interpret. The record lists enrollment of 1 but does not explain whether any participant posted an evaluable biopsy or how that figure relates to the primary-outcome statement.

The study status is terminated, but the reason for termination is not reported in the registry record. The registry gives start and primary-completion dates but does not explain the timing or circumstances of the decision to terminate. It also does not report recruitment progress, protocol deviations, or whether any outcomes other than the primary outcome were collected.

The record does not describe the intended statistical model, test, covariates, multiplicity approach, missing-data procedure, or analysis population. It does not report the sample-size justification for the trial. These omissions prevent assessment of whether a planned analysis would have been appropriate for the endpoint or how its uncertainty would have been quantified.

The primary comparison is between groups defined by response or progression, not between randomized treatment arms. Any hypothetical group difference would therefore be an association and could be influenced by participant or disease characteristics. No adjustment plan is reported. The endpoint also relies on tumor-biopsy measurements, but assay performance, specimen handling, and measurement reproducibility are not described in the registry record.

Finally, the safety listing gives zero participants at risk and zero events, which does not provide an evaluable denominator for assessing safety. Overall, the registry record documents the intended scientific question and key design features, but the absence of analyzed outcome data sharply limits what can be concluded.

14. Related Tutorials

15. Related Statistical Calculators

16. Sources

• ClinicalTrials.gov: Immunologic Determinants of Response to Pembrolizumab (MK-3475) in Advanced Melanoma (MK-3475-161/KEYNOTE-161) — NCT03407170.
• Registry record last updated 2020-07-15; status last verified 2020-06.
• Results first posted 2020-07-15.
• 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.

17. Record Summary

NCT03407170 is a terminated, open-label, single-group phase II study of pembrolizumab in advanced melanoma, identified as KEYNOTE-161. The listed regimen was 200 mg by intravenous infusion every 3 weeks for up to 24 months. Enrollment is listed as 1.

The primary outcome was mean FCT in participants with response versus participants with progression, measured from tumor-biopsy CD8+ T-cells using a predefined single-cell RNA gene signature. The registry states that no data were collected or analyzed for this outcome because the study ended early. It lists no analyses.

The record also lists safety, response, survival, and immune-related outcomes, but does not provide analyzed results for them. It does not report the reason for termination, the trial’s sample-size justification, or key statistical procedures such as missing-data and multiplicity methods. The registry therefore describes the study’s intended question and design, but offers no primary-outcome estimate.