Nivolumab Plus Relatlimab or Ipilimumab in Metastatic Melanoma Stratified by MHC-II Expression: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Nivolumab Plus Relatlimab or Ipilimumab in Metastatic Melanoma Stratified by MHC-II Expression (NCT03724968), a non-randomized phase 2 study. The primary outcome is Change in Activated GZMB+ CD8+T-cell Density Intratumorally, of Two Immunotherapy Regimens.

2. Clinical Question

The clinical question concerns a tissue-based measure of immune-cell density, rather than a reported clinical response or time-to-event result. The record lists response rate, median progression-free survival, median overall survival, and treatment-related adverse events as secondary outcomes. However, it does not report results for those outcomes, and it reports no analyses. As a result, the record cannot establish whether the regimens differed on the primary measure or on the listed secondary outcomes.

The study was open-label and non-randomized. Treatment assignment therefore was not described as random, and the record does not establish that the groups were comparable at baseline. MHC-II expression is part of the study description, but the record does not report the stratification procedure or group-specific MHC-II information. The central question is consequently the one the trial intended to assess, not a question answered by posted comparative outcome data.

3. Trial Design

This was a phase II, open-label, non-randomized, parallel, two-arm study in metastatic melanoma. Arm A was nivolumab plus relatlimab; Arm B was nivolumab plus ipilimumab. The registry record lists enrollment of 2 and gives the study status as terminated. The study start date was 2019-01-17, and the primary completion date was 2020-05-28.

Because treatment allocation was non-randomized, the arms should not be interpreted as randomized groups. In general, random allocation helps balance measured and unmeasured participant characteristics across groups. Without it, differences in outcomes, if available, could reflect differences in who received each regimen as well as treatment. The registry record does not report how treatment assignment was determined, whether the groups were similar at entry, or whether any adjustment was planned to address group differences.

Open-label means the study was not masked. The record does not describe any masking of participants, clinicians, or outcome assessors. The primary endpoint was change in activated GZMB+ CD8+ T-cell density intratumorally, with a timeframe of approximately 16 months. The registry explicitly states that these data were not collected due to loss of funding. Thus, the intended design can be described, but the record does not contain a primary outcome comparison.

The record lists five outcome measures and zero analyses. Its listed secondary outcomes include response rate, median progression-free survival, median overall survival, and treatment-related adverse events assessed by CTCAE Version 5.0. Outcome results for these measures are not reported in the registry record. A sample-size justification and detailed analysis plan are also 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.

Before assignment: Three participants were consented to take part in this study and 1 participant was determined to be ineligible. This study closed early due to low accrual. Both participants were assigned to Arm B. Since only 2 participants were enrolled and both were on Arm B, a statistical analysis cannot be done and some outcome measures can't be reported.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Due to loss of funding data were not collected

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: Study stopped early due to low accrual and no participants were enrolled to Arm A

6. Primary Endpoint

The registry lists the primary outcome as Change in Activated GZMB+ CD8+T-cell Density Intratumorally, of Two Immunotherapy Regimens, measured over Approximately 16 months.

The primary endpoint was change in activated GZMB+ CD8+ T-cell density within tumors, comparing the two immunotherapy regimens. The record gives an approximate timeframe of 16 months. Activated GZMB+ CD8+ T cells are the specific cell population named in the endpoint; density describes how concentrated that population is in tumor tissue. The record does not specify the unit of measurement, the exact calculation of change, the collection schedule, or the method used to assess the tissue.

In general, a tissue-based endpoint can be used to examine a biological feature of a tumor during treatment. It can address a different question from a clinical outcome such as response rate or time to progression: it focuses on a measured feature in tumor tissue rather than directly summarizing a clinical event. The endpoint title indicates that the planned comparison concerned change between two regimens, but it does not establish how the endpoint was to be statistically summarized.

The key interpretive limitation is that the registry record says the primary endpoint data were not collected due to loss of funding. No endpoint estimate, variability measure, confidence interval, or test result is reported. Accordingly, the planned tissue measure cannot be used to describe an observed difference between the arms.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does non-randomized allocation mean?

In a randomized trial, chance determines assignment to treatment groups. In general, randomization is used to reduce systematic differences between groups, including differences in characteristics that might affect an outcome. Here, the registry record labels allocation non-randomized. The two arms therefore should not be treated as a randomized treatment-versus-control comparison. If outcomes had been reported, group differences could have reflected treatment, differences between participants, or both. The record does not report the assignment procedure or a statistical adjustment plan.

What does an estimate and its confidence interval tell a reader?

In general, an estimate gives a summary of the observed effect or difference, while a confidence interval describes uncertainty around that estimate under the statistical model used. A narrower interval indicates greater precision than a wider interval, but a confidence interval does not by itself establish clinical importance. No primary endpoint estimate or confidence interval is reported in the registry record because the endpoint data were not collected. The record also lists zero analyses, so there are no posted confidence intervals or test results to interpret.

How should the planning assumptions be read?

The stated sample-size planning illustration uses a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. It considers effects defined as a stated fraction of an observed effect. These are planning assumptions, not results from this trial and not evidence of the trial's achieved power. The sample-size justification used when designing the trial is not reported in the registry record. The record does not support treating a planning calculation as an observed trial finding.

In general, a sample-size calculation asks how many participants might be needed to detect a chosen effect under specified assumptions. It depends on the outcome, effect size, variability, allocation, and statistical method. A planning illustration cannot compensate for uncollected outcomes, and it cannot establish what the study would have found. No sample-size values or primary endpoint effect estimates are reported in the registry record, so there is no numerical planning result to interpret.

Why does the outcome type matter?

The primary endpoint concerns change in a tissue measure, while the listed secondary outcomes include response rate, medians for progression-free and overall survival, and treatment-related adverse events. Different outcome types generally require different summaries and may rely on different assumptions. For example, a continuous measurement is often summarized with a measure of location and variability, while a time-to-event outcome may be summarized using event-time methods. These are general distinctions, not a description of analyses conducted in this study. The record does not report the endpoint's unit, variability, model, or analysis method.

For event counts, over-dispersion can matter when the observed variation exceeds what a simple statistical model assumes. In general, ignoring over-dispersion can make uncertainty appear smaller than it is. The registry record does not provide outcome distributions or an event-count analysis, so there is no basis to assess over-dispersion in this trial. Similarly, it does not report whether the survival outcomes were analyzed with a particular model or how censoring was handled.

What can the adverse-event summaries establish?

The record gives participant counts for serious and other adverse events by arm and lists adrenal insufficiency as the top serious adverse-event term. These are descriptive safety summaries. In general, counts can help identify what was recorded, but very limited numbers do not precisely characterize event frequency or permit a stable comparison between treatment groups. No comparative statistical analysis of safety is reported. The record's adverse-event counts should therefore be read as recorded events and participants at risk, not as a formal estimate of relative safety.

Finally, the study was open-label and non-randomized, enrolled 2 participants, and was terminated. The record does not report interim analyses, a detailed statistical analysis plan, or a protocol-based sample-size rationale. It also does not report a multiplicity strategy or analysis populations. These omissions prevent reconstruction of a complete inferential analysis. The statistical methods that can be described are limited to the design labels and the general assumptions explicitly stated for the planning illustration.

10. Multiplicity, Missing Data and Analysis Populations

The registry record names a primary endpoint and lists secondary outcomes: response rate, median progression-free survival, median overall survival, and treatment-related adverse events assessed by CTCAE Version 5.0. It does not report whether formal hypothesis tests were planned for these outcomes, how many comparisons were intended, or whether any adjustment for multiple testing was specified. A multiplicity strategy is not reported in the registry record.

In general, multiplicity arises when several outcomes or comparisons are tested. Without an appropriate plan, repeated testing can increase the chance of at least one apparently notable result by chance. This general concern cannot be quantified for this study because the record reports zero analyses and no test results. The endpoint data were not collected, and results for the listed secondary outcomes are not reported in the registry record.

The record also does not describe analysis populations, such as which participants would be included in an efficacy analysis or in a safety analysis. It does not report an intention-to-treat definition, a per-protocol definition, or rules for excluding observations. The analysis population is therefore not reported in the registry record.

For missing data, the record specifically states that primary endpoint data were not collected due to loss of funding. It does not report how missing observations for other outcomes were handled, whether any imputation was planned, or whether sensitivity analyses were intended. No assumptions about missingness or participant follow-up should be inferred from the record. These gaps mean that the intended handling of incomplete data and the robustness of any analysis cannot be assessed.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: adrenal insufficiency (50.0%).

For Arm A, the registry lists 0 participants at risk for serious adverse events, 0 affected by a serious adverse event, 0 at risk for other adverse events, 0 affected by other adverse events, and 0 deaths. For Arm B, it lists 2 participants at risk for serious adverse events and 1 affected; 2 at risk for other adverse events and 2 affected; and 0 deaths. Adrenal insufficiency is listed as the top serious adverse-event term at 50.0%.

These figures are descriptive counts from the registry record. They do not provide a precise estimate of comparative safety, particularly given the total enrollment of 2 and the non-randomized design. Arm A's zero counts should not be read as evidence that no adverse event could occur; the record lists zero participants at risk for those summaries in that arm. The serious-event term and percentage are as recorded, not a basis for broader claims about how frequently the event would occur in other settings.

The record does not report a statistical comparison of adverse events between arms. The listed counts therefore do not establish that one regimen is safer or less safe than the other.

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

The record establishes that a phase II, open-label, non-randomized, two-arm study was conducted and later terminated, with enrollment of 2. It identifies the planned regimens and the primary tissue-based endpoint. It also states that the primary endpoint data were not collected due to loss of funding. These registry entries describe the study and its reporting, not a treatment effect.

The record does not show whether either regimen changed intratumoral activated GZMB+ CD8+ T-cell density, or whether changes differed between arms. Nor does it report results for response rate, median progression-free survival, or median overall survival. The secondary outcome names alone are not evidence of benefit, lack of benefit, or a difference between treatments. There are no reported analyses, estimates, or confidence intervals for drawing such conclusions.

The posted serious and other adverse-event counts describe what was recorded for the small enrolled group. They do not provide a dependable comparative safety estimate. Zero deaths are recorded, but this finding in the registry record does not establish that a regimen carries no mortality risk.

In general, an unreported or uncollected endpoint is different from an endpoint that was measured and found to show no difference. Likewise, absence of a statistically significant result cannot be inferred when no analysis is reported. The appropriate interpretation is limited: the trial's intended primary comparison is not answerable from the registry results. Any stronger statement about efficacy, biological activity, equivalence, or comparative safety would exceed the evidence recorded.

13. Why This Trial Matters Statistically

This study illustrates an important distinction between a registered research question and an answered research question. The registry record specifies a primary endpoint and two regimens, but it also states that the primary endpoint data were not collected. Without those measurements, neither the direction nor the magnitude of the planned comparison can be estimated. A clearly named endpoint is not itself an outcome result.

The design also matters. Because allocation was non-randomized, any observed difference would require careful interpretation in light of potential group differences. In general, randomization helps protect a treatment comparison from systematic allocation differences, while open-label conduct can affect some forms of assessment. Here the record reports no outcome analyses that could be evaluated for these issues.

The small enrollment and limited safety counts show why descriptive event information should not be mistaken for a comparative safety conclusion. Counts record events in the study group; they do not by themselves establish stable event rates or differences between regimens. The record also lacks a reported analysis plan, multiplicity approach, and analysis-population definitions, which limits the ability to assess how intended statistical inference would have been carried out.

For clinical researchers and biostatisticians, the record is a reminder to separate design, planned outcomes, collected data, and statistical conclusions. Each answers a different question. In this case, the design and intended endpoint are documented, but the main endpoint data and comparative analyses are not. The distinction prevents missing evidence from being misread as evidence of no effect.

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.

Interpretation is constrained first by the absence of primary endpoint data. The registry record attributes non-collection to loss of funding, so the intended change in intratumoral activated GZMB+ CD8+ T-cell density cannot be evaluated. It does not report a unit, measurement procedure, or a detailed endpoint definition beyond the endpoint title and approximate timeframe.

Enrollment was 2, and the record does not give arm-specific enrollment in the information available for interpretation. The non-randomized, open-label design limits the basis for causal comparisons; details of allocation and baseline comparability are not reported in the registry record. The study is listed as terminated, but a separate reason for termination is not reported in the registry record.

The record lists secondary outcomes but gives no results or analyses for them. It does not report confidence intervals, test statistics, missing-data methods, analysis populations, interim analyses, or a multiplicity plan. A detailed statistical analysis plan and the trial's sample-size justification are not reported in the registry record. These gaps prevent assessment of the intended inferential approach and any sensitivity of conclusions to analysis choices.

Safety interpretation is also limited by the small recorded counts and the absence of a reported comparative analysis. The registry figures describe events in this study; they do not support broad generalization about event frequency or between-regimen differences.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Nivolumab Plus Relatlimab or Ipilimumab in Metastatic Melanoma Stratified by MHC-II Expression — NCT03724968.
• Registry record last updated 2021-04-20; status last verified 2021-03.
• Results first posted 2021-03-05.
• 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.

18. Record Summary

The registry record describes NCT03724968, a phase II, open-label, non-randomized, parallel study in metastatic melanoma. Arm A was nivolumab plus relatlimab, and Arm B was nivolumab plus ipilimumab. The record lists enrollment of 2, a start date of 2019-01-17, a primary completion date of 2020-05-28, and terminated status.

The primary endpoint was change in intratumoral activated GZMB+ CD8+ T-cell density over approximately 16 months. The record states that the primary endpoint data were not collected due to loss of funding. It lists four secondary outcomes and reports zero analyses. Safety summaries record 0 deaths in both arms and list adrenal insufficiency at 50.0% as the top serious adverse-event term. The registry record does not report comparative efficacy or safety conclusions.