A Study Evaluating the Efficacy and Safety of Multiple Treatment Combinations in Patients With Melanoma (Morpheus-Melanoma): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study Evaluating the Efficacy and Safety of Multiple Treatment Combinations in Patients With Melanoma (Morpheus-Melanoma) (NCT05116202), a randomized phase 1, phase 2 study. The primary outcome is Pathologic Response Rate (pRR) for Cohort 1 as Determined by Independent Pathologic Review.

2. Clinical Question

3. Trial Design

Morpheus-Melanoma was a Phase Ib/II, open-label, multicenter, randomized umbrella study with a parallel-group model and no masking. The registry record lists an enrollment of 110 and a completed study status. Its broad objective was to evaluate efficacy, safety, and pharmacokinetics of treatment combinations in cancer immunotherapy-naive participants with resectable Stage III melanoma in Cohort 1 and participants with Stage IV melanoma in Cohort 2.

The umbrella structure matters for interpreting the posted comparisons. The registry describes the study as flexible: new treatment arms could be opened as treatments became available, existing arms could be closed for minimal clinical activity or unacceptable toxicity, and the participant population could be modified. The posted primary response results are for Cohort 1, and compare the nivolumab-plus-ipilimumab control with three experimental arms. The results table gives different participant counts by arm; it does not show an equal-sized comparison.

For Cohort 1, the control and experimental regimens were described as treatment before surgery, generally for 2 cycles (6 weeks) or until surgery, unacceptable toxicity, or loss of clinical benefit. The Cohort 2 description posted for RO7247669 plus tiragolumab instead specifies treatment until unacceptable toxicity or loss of clinical benefit, with an investigator assessment drawing on radiographic and biochemical data, local biopsy results if available, and clinical status. Cohort 2 does not contribute to the primary response comparison.

The registry lists study start as 2022-02-02 and primary completion as 2023-09-22. These are record dates, not evidence about the timing or conduct of particular analyses. The registry record does not report the original sample-size justification or the full statistical analysis plan.

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: A total of 110 participants with resectable Stage III melanoma and Stage IV melanoma took part in Cohort 1 and Cohort 2, respectively. A total of 6 treatment arms were planned in Cohort 1 and 102 participants were enrolled in only 4 arms. The study was closed before the Cohort 1 arms tobemstomig 600 milligrams (mg), and tobemstomig 600 mg + tiragolumab were opened. 8 participants were enrolled in 1 arm in Cohort 2.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Efficacy-evaluable population included all participants in Cohort 1 who received at least one dose of each drug for their assigned treatment regimen.

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: Intent-to-treat (ITT) population included all participants who were enrolled in the study.

6. Primary Endpoint

The registry lists the primary outcome as Pathologic Response Rate (pRR) for Cohort 1 as Determined by Independent Pathologic Review, measured over Time of surgery (scheduled at Week 7).

The primary endpoint is the pathologic response rate, or pRR, in Cohort 1, assessed at the time of surgery scheduled at Week 7. It is the percentage of participants whose dissected lymph node showed a pathologic complete response, near complete response, or partial response on independent pathologic review. A complete response meant a complete absence of viable tumor cells; a near complete response meant more than 0 to at most 10% viable tumor cells; and a partial response meant more than 10 to at most 50% viable tumor cells.

This endpoint captures the extent of viable tumor remaining in the surgical specimen after the pre-surgery treatment period. Using a defined pathologic assessment can make the response classification more reproducible than relying on an unspecified clinical impression. The registry specifies independent review for the primary endpoint and reports the response proportion for each arm with a 95% confidence interval.

The analysis rule for missing assessments is important: participants with missing or no pathologic response assessment, including participants who did not proceed to complete lymph node dissection, were classified as non-responders. The efficacy-evaluable population consisted of Cohort 1 participants who received at least one dose of each drug for their assigned treatment regimen. Thus, the reported response proportions should be read in light of both that population definition and the non-responder rule. Pathologic response is the primary endpoint; it is not itself a measure of event-free survival, relapse-free survival, overall survival, or safety.

7. Results Posted in the Registry

• Cohort 1: Tobemstomig 2100 mg: 80.0% (32/40) versus 77.3% (17/22) on control; risk difference 2.7 percentage points (95% CI -18.7 to 24.2), risk ratio 1.04 (0.79–1.36); unadjusted two-proportion z-test p = 0.801.
• Cohort 1: Atezolizumab + Tiragolumab: 45.0% (9/20) versus 77.3% (17/22) on control; risk difference -32.3 percentage points (95% CI -60.2 to -4.3), risk ratio 0.58 (0.34–0.99); unadjusted two-proportion z-test p = 0.031.
• Cohort 1: Tobemstomig + Tiragolumab: 60.0% (12/20) versus 77.3% (17/22) on control; risk difference -17.3 percentage points (95% CI -45.0 to 10.4), risk ratio 0.78 (0.51–1.19); unadjusted two-proportion z-test p = 0.227.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

pRR for Cohort 1 as Determined by Local Pathologic Assessment

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Cohort 1: Tobemstomig 2100 mg: 75.0% (30/40) versus 81.8% (18/22) on control; risk difference -6.8 percentage points (95% CI -27.8 to 14.2), risk ratio 0.92 (0.70–1.20); unadjusted two-proportion z-test p = 0.539.
  • Cohort 1: Atezolizumab + Tiragolumab: 50.0% (10/20) versus 81.8% (18/22) on control; risk difference -31.8 percentage points (95% CI -59.0 to -4.6), risk ratio 0.61 (0.38–0.99); unadjusted two-proportion z-test p = 0.029.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Cohort 1: Tobemstomig 2100 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Cohort 1: Atezolizumab + Tiragolumab than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Event-free Survival (EFS) for Cohort 1

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor; it was not recalculated.
• Result:
  • Cohort 1: Nivolumab + Ipilimumab (Control) vs Cohort 1: Tobemstomig 2100 mg: hazard ratio 2.09 (95% CI 0.42–10.42) (as reported in the registry).
  • Cohort 1: Nivolumab + Ipilimumab (Control) vs Cohort 1: Atezolizumab + Tiragolumab: hazard ratio 3.95 (95% CI 0.79–19.70) (as reported in the registry).
• Interpretation:
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Cohort 1: Nivolumab + Ipilimumab (Control) vs Cohort 1: Tobemstomig 2100 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Cohort 1: Nivolumab + Ipilimumab (Control) vs Cohort 1: Atezolizumab + Tiragolumab and the control. The interval gives the range of effects that remain compatible with the data.

Relapse-free Survival (RFS) for Cohort 1

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor; it was not recalculated.
• Result:
  • Cohort 1: Nivolumab + Ipilimumab (Control) vs Cohort 1: Tobemstomig 2100 mg: hazard ratio 1.41 (95% CI 0.25–7.75) (as reported in the registry).
  • Cohort 1: Nivolumab + Ipilimumab (Control) vs Cohort 1: Atezolizumab + Tiragolumab: hazard ratio 2.09 (95% CI 0.35–12.62) (as reported in the registry).
• Interpretation:
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Cohort 1: Nivolumab + Ipilimumab (Control) vs Cohort 1: Tobemstomig 2100 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Cohort 1: Nivolumab + Ipilimumab (Control) vs Cohort 1: Atezolizumab + Tiragolumab and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes provide a broader view than the primary pathologic response assessment, spanning local pathologic review, event-free and relapse-free survival, overall survival, response, and safety. The posted results include local pathologic response and selected survival outcomes. Those data offer context about whether the general pattern seen at surgery is accompanied by a clearly precise time-to-event pattern in the available follow-up; the posted survival estimates do not provide that level of precision.

The secondary results should be interpreted alongside, rather than in place of, the primary endpoint. Local pathologic assessment is related to the primary independent pathologic assessment, but it is a distinct assessment. Event-free and relapse-free survival address later events over time, so they answer different questions from whether viable tumor remained in the surgical specimen. The registry lists overall survival and other outcomes, but not all listed secondary endpoints have results summarized in the posted analyses.

There are multiple secondary outcomes and comparisons, and the registry record does not report a prespecified hierarchy or multiplicity adjustment for them. General explanation: examining many results increases the opportunity for an apparently unusual result to arise by chance, particularly when each is interpreted using an unadjusted threshold. For that reason, secondary results without a stated adjustment carry less weight as confirmatory evidence. They can help characterize the study and suggest questions for further investigation, but they do not independently settle whether treatment arms differ.

In this umbrella study, the scope of possible arms and populations could also change over time, according to the registry description. That flexibility is part of the study context, but the registry record does not report how such changes were incorporated into a statistical multiplicity plan. The safest reading is to treat secondary outcomes as descriptive evidence whose precision and analysis context need to be considered, not as a collection of separate definitive tests.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• pRR for Cohort 1 as Determined by Local Pathologic Assessment
• Event-free Survival (EFS) for Cohort 1
• Relapse-free Survival (RFS) for Cohort 1
• Overall Survival (OS) for Cohort 1
• ORR for Cohort 1
• Number of Participants With Adverse Events (AEs) and Severity of AEs Determined According to National Cancer Institute Common Terminology Criteria for Adverse Events, Version 5.0 (NCI CTCAE v5.0) for Cohort 1
• Number of Participants With Immune-related AEs Grade ≥ 3 for Cohort 1
• Rate of Delayed Surgery Due to Treatment-related AEs

9. Statistical Interpretation

10. Statistical Methods Explained

How was the primary response rate defined?

The primary endpoint is binary at the participant level: a participant is counted as having a pathologic response if the independent review identifies a complete, near complete, or partial response under the registry's viable-tumor thresholds. The response rate is the percentage of participants who meet that definition. Participants without an assessment, including those who did not proceed to complete lymph node dissection, are classified as non-responders. This rule makes missing outcome assessments part of the endpoint classification rather than removing them from the response-rate denominator, although the stated efficacy-evaluable population also requires receipt of at least one dose of each assigned drug.

What do the confidence intervals describe?

The registry says each arm's pRR confidence interval was calculated using the Clopper-Pearson method. General explanation: a confidence interval describes uncertainty around an estimate. A wider interval indicates less precision; it is not a range containing a known fixed fraction of individual participant outcomes. For a proportion, uncertainty depends in part on the number of participants and the observed response pattern. The intervals displayed with the arm rates help show how much uncertainty remains around each estimate.

For the difference in pRR between experimental arms and control, the registry specifies a Wald method with continuity correction for the 95% confidence intervals. General explanation: a difference expresses the contrast between two proportions on the same scale, so a value of zero corresponds to no difference. An interval that includes zero does not distinguish the estimated contrast from no difference at the confidence level used. The registry record does not name the method used to calculate the sponsor's posted difference intervals, and the reported comparisons do not include p-values.

How should the recalculations be read?

The simple recalculations are unadjusted comparisons reconstructed from the posted summary counts. They are useful as a check and as an additional way to describe the observed contrast, but they are not a replacement for the analysis as reported by the sponsor. The recalculated p-values are unadjusted; they do not account for the multiple arms and outcomes in this study. No multiplicity adjustment is reported in the registry record. The recalculations also do not add information beyond the posted participant counts and response results.

General explanation: a p-value summarizes how compatible data are with a specified null model under the assumptions of a test. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. A p-value should be interpreted together with the effect estimate, its confidence interval, study design, endpoint definition, and analysis context. Here, the sponsor's posted confidence intervals are central because they display the uncertainty in the estimated differences directly.

What is reported for time-to-event outcomes?

The registry posts median event-free survival and relapse-free survival, along with hazard ratios for selected comparisons. Event-free survival is measured from randomization to disease progression, disease recurrence or death, or the last tumor assessment. Relapse-free survival is measured from surgery to first documented disease recurrence or death, or the last tumor assessment. Some arm medians are not estimable in the posted results, and several confidence limits are open-ended or absent. The hazard ratios are reported by the sponsor, but the method used to estimate them is not reported in the registry record.

General explanation: a median time-to-event estimate describes the time by which half the analyzed group has experienced the event, when that point can be estimated from available follow-up. A hazard ratio summarizes a relative event rate over follow-up under the model used; it is not a difference in median times. Its interpretation depends on the analysis and assumptions, which cannot be fully assessed when the model and method are not reported. The displayed intervals for these hazard ratios are broad, so the estimates have limited precision.

The registry lists 18 outcome measures and 18 analyses, and reports 16 secondary outcomes. It does not describe a statistical hierarchy for the secondary analyses or explain whether any were adjusted for multiple testing. General explanation: when several outcomes and comparisons are examined, the chance of seeing an apparently notable result by chance can increase. This is one reason secondary results, especially unadjusted ones, are generally interpreted as supportive or exploratory rather than as independent confirmation.

11. Confidence Intervals and Precision

The primary response estimates are based on groups of 22, 40, 20, and 20 participants. Their 95% confidence intervals are broad: for example, the control estimate of 77.3% has an interval from 54.63 to 92.18, while the 80.0% estimate for Tobemstomig 2100 mg has an interval from 64.35 to 90.95. The overlap of arm-level intervals is not itself a formal test of the difference between arms; the posted comparison intervals address that question directly.

For the difference comparisons, the sponsor's 95% confidence intervals extend from -22.25 to 27.70 for Tobemstomig 2100 mg versus control, from -65.01 to 0.46 for atezolizumab plus tiragolumab versus control, and from -49.75 to 15.21 for tobemstomig plus tiragolumab versus control. Each interval includes zero. This means that the posted data do not precisely distinguish the corresponding contrast from no difference, and the estimates remain compatible with materially different response rates in either direction. The interval for the atezolizumab-plus-tiragolumab comparison comes close to zero but still includes it.

The survival results also show substantial uncertainty. The reported hazard-ratio intervals for event-free and relapse-free survival are wide, and some upper confidence limits for median estimates are not reported. Such results should not be read as precise rankings of the arms. General explanation: precision is about how narrowly the data constrain an estimate; it is different from whether a point estimate looks large or small. Smaller groups and limited event information can leave broad uncertainty even when a point estimate is available.

The registry does not report the number of events underlying each survival comparison or the details of its survival-analysis methods. That limits assessment of how stable those estimates are. The tables and plots display the reported estimates and intervals; they should be read as summaries of uncertainty, not as proof that the true effects lie at the point estimates.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size table is a planning illustration derived from the observed effect, not a description of the study's original design calculation. For the binary pRR comparisons, the figures are participants per arm. For time-to-event comparisons, the figures are total events, not participants. The table uses a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It considers planning for 100%, 75%, or 50% of the observed effect.

The 80% power column gives the illustrated number of participants per arm, or total events for time-to-event outcomes, associated with an 80% chance of detecting the selected effect under those planning assumptions if that effect is real. The 90% power column describes the corresponding illustration for a 90% chance. For example, the primary comparison with atezolizumab plus tiragolumab uses 35 participants per arm for 80% power and 46 per arm for 90% power when planning for 100% of the observed effect. For 50% of that observed effect, the illustration rises to 128 per arm for 80% power and 170 per arm for 90% power.

Smaller effects require more information to detect under the same assumptions, so the planning figures rise as the assumed effect becomes a smaller fraction of the observed one. For survival outcomes, the analogous table gives total event targets; it does not state a participant target because the number of participants needed depends on how many events accrue. The original trial's own sample-size assumptions and justification are not reported in the registry record. These figures are not observed power; they are hypothetical planning illustrations based on the observed effect and stated assumptions.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.773; treatment proportion p₁ = 0.800; average p̄ = 0.786.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5796 = 1.1361.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5793 = 0.4876.
4. n = (1.1361 + 0.4876)² / (0.0273)² = 3,544.3, rounded up to 3,545 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.772727272727273                          # control proportion
p1 <- 0.8                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(5116202)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 3,545 (80% power), 4,745 (90% power)
 75% of observed effect: n per arm = 6,375 (80% power), 8,534 (90% power)
 50% of observed effect: n per arm = 14,507 (80% power), 19,420 (90% power)
power.prop.test: 3544.3 
Simulated power at n = 3,545 per arm: 0.804 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The sponsor's posted primary analysis reports differences in pRR and 95% confidence intervals for each experimental arm compared with nivolumab plus ipilimumab. The posted estimate is 2.73 for Tobemstomig 2100 mg, -32.27 for atezolizumab plus tiragolumab, and -17.27 for tobemstomig plus tiragolumab. The associated confidence intervals are shown in the results tables. The sponsor's analyses do not include p-values in the registry record, and the method field for these comparisons is not specified there.

The page also presents simple unadjusted recalculations from the posted summary statistics, including unadjusted p-values. These provide a separate descriptive check, but should not be treated as the sponsor's analysis or as a substitute for its estimates. The registry record does not flag a discrepancy between the sponsor results and the recalculations. Differences in calculation methods can matter, and the sponsor's reported analysis is the primary account of the study results; the recalculations are supplementary context.

Neither the sponsor's confidence intervals nor the recalculated p-values account, in the information posted here, for the study's multiple arms and outcomes. The record does not report the method behind the sponsor's comparison intervals or a multiplicity procedure. Readers should therefore focus on the effect estimates and their uncertainty, and avoid interpreting an isolated unadjusted p-value as a definitive finding.

15. Multiplicity, Missing Data and Analysis Populations

The primary endpoint has a defined response classification and analysis population, but the registry record does not report a complete statistical analysis plan. It states that pRR confidence intervals were calculated for each arm using the Clopper-Pearson method and that the differences in pRR were calculated with 95% confidence intervals using a Wald method with continuity correction. It does not identify the method used for the sponsor's posted difference intervals or report p-values for those comparisons.

Multiplicity is a particular consideration because the study is an umbrella design with several treatment arms and multiple outcomes. The registry lists 18 outcome measures and 18 analyses, and the study description allows arms and participant populations to be modified. The registry record does not report a multiplicity adjustment, a formal hierarchy among endpoints, or a strategy specifying which comparisons are confirmatory. Consequently, the posted comparison results should not be interpreted as if they were the only analyses performed.

The primary endpoint has an explicit missing-assessment rule: participants without pathologic response assessment, including those who did not proceed to complete lymph node dissection, were classified as non-responders. The efficacy-evaluable population included Cohort 1 participants who received at least one dose of each drug in the assigned regimen. This definition is reported, but the record does not provide additional sensitivity analyses showing how alternative handling of missing assessments might affect the results.

The registry record does not report the original sample-size justification, interim-analysis plan, detailed analysis populations for all outcomes, or the statistical methods used for the survival analyses. It also does not explain the methods underlying the sponsor's reported comparisons beyond the primary endpoint description. These omissions limit assessment of prespecification, model assumptions, and the role of secondary analyses. They are reasons to keep interpretations tied closely to the analyses and definitions that are actually described.

16. 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: enteritis (12.5%), myocarditis (5.0%), cellulitis (5.0%), postoperative wound infection (5.0%), infusion related reaction (5.0%).

The safety table reports serious adverse events and other adverse events by group, as well as deaths. In Cohort 1, serious adverse events were reported for 4 of 22 participants in the nivolumab-plus-ipilimumab control, 9 of 40 in the Tobemstomig 2100 mg arm, 3 of 20 in the atezolizumab-plus-tiragolumab arm, and 6 of 20 in the tobemstomig-plus-tiragolumab arm. The table also reports other adverse events for 18 of 22, 34 of 40, 18 of 20, and 18 of 20 participants, respectively. These are counts within the listed groups, not adjusted comparisons.

The Cohort 2 group of 8 participants had 1 participant with a serious adverse event and 7 with other adverse events. The record also lists deaths by group, but those counts alone do not establish whether a death was caused by treatment. The most frequently listed serious-event terms include enteritis, myocarditis, cellulitis, postoperative wound infection, infusion related reaction, and wound dehiscence. The registry lists rates for those terms, but does not provide enough information to infer comparative risk or causality.

These safety counts are descriptive. The groups differ in size and include different cohorts, so direct numerical ranking can be misleading. The record does not provide a detailed causal assessment or a full account of exposure time for these summaries.

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

The results describe what was observed in the posted trial groups: pathologic response proportions at surgery, selected time-to-event estimates, and adverse-event counts. They do not establish that one regimen works or does not work. The response-rate differences have broad confidence intervals, and the sponsor-reported intervals for all three experimental comparisons include zero. These results leave substantial uncertainty about the size and direction of the underlying contrasts.

The primary endpoint addresses pathologic findings at surgery in Cohort 1. It is not a direct measure of survival, long-term disease control, or quality of life. The secondary survival endpoints address later outcomes, but their estimates are also imprecise in the posted record. Safety counts help describe events reported in the groups, but do not alone prove that a treatment caused an event or establish a comparative safety profile.

The study also has an umbrella design with multiple arms and the stated flexibility to open or close arms and modify populations. That structure supports evaluation of several combinations, but makes it especially important to know which comparisons were prespecified and how multiplicity was handled. Those details are not reported in the registry record. The sample-size illustrations on this page are not the study's original design justification; they show hypothetical planning needs under stated assumptions.

Accordingly, the most defensible interpretation is descriptive and uncertainty-aware. The findings can indicate the response patterns observed in these groups and the precision of the estimates, but they cannot by themselves establish a definitive treatment ranking or generalize to populations not represented in the posted Cohort 1 results. The registry record's stated primary population was cancer immunotherapy-naive participants with resectable Stage III melanoma; the primary results should be understood within that scope.

18. Why This Trial Matters Statistically

This trial is statistically notable for bringing several treatment combinations into one randomized umbrella study while retaining a defined control arm for the Cohort 1 primary response comparisons. That creates a setting where comparisons can be described side by side, but it also makes the analysis context important: there are multiple arms, multiple outcomes, and the registry describes flexibility to alter arms and populations as the study proceeds.

The primary endpoint illustrates the difference between an observed proportion and a precise estimate. The response proportions range across the posted arms, yet the comparison intervals are wide. The data therefore show patterns, but the precision is not sufficient to treat the point estimates as settled treatment effects. This is a useful example of why estimates and intervals should be considered together rather than focusing only on a single rate or p-value.

The record also makes transparent one important missing-data decision for the primary endpoint: unassessed participants were classified as non-responders. That rule is part of the estimand represented by the reported response rates and should be considered when interpreting them. At the same time, the registry does not report alternative missing-data analyses or a complete multiplicity plan.

Finally, the record distinguishes an observed result from a planning exercise. The sample-size illustrations use the observed effects under explicit assumptions, but they are not the trial's original design calculations. Keeping those purposes separate prevents retrospective calculations from being mistaken for evidence about the design's original power or objectives.

19. 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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

Several features of this specific record limit interpretation. First, the primary response analysis has relatively small arm denominators, and the corresponding confidence intervals are broad. The posted survival comparisons are also imprecise, with wide hazard-ratio intervals and incomplete median confidence limits. The registry does not report the number of events underlying each survival analysis, which would help readers assess the information available for those comparisons.

Second, this is a flexible umbrella study. The record says that treatment arms could be opened or closed and the participant population could be modified, but it does not report the statistical procedures used to address those changes or the overall multiplicity of comparisons. The original sample-size justification, interim-analysis plan, and full statistical analysis plan are not reported in the registry record.

Third, primary pRR results use an efficacy-evaluable population and classify missing or absent pathologic assessments as non-responders. The record does not show sensitivity analyses for alternative handling of missing assessments. It also does not report a method for the sponsor's primary difference intervals, or p-values for those analyses. The recalculated p-values are unadjusted and cannot fill those reporting gaps.

Finally, the results titles use Tobemstomig while intervention and arm descriptions elsewhere in the record use RO7247669. The record does not explain this naming difference, so it is safest to retain the terminology attached to each posted description rather than assume an equivalence. The primary results concern Cohort 1; Cohort 2 has a different population and treatment context and should not be conflated with those comparisons.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study Evaluating the Efficacy and Safety of Multiple Treatment Combinations in Patients With Melanoma (Morpheus-Melanoma) — NCT05116202.
• Registry record last updated 2025-07-18; status last verified 2025-06.
• Results first posted 2025-07-18.
• 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.

23. Record Summary

NCT05116202, Morpheus-Melanoma, is a completed Phase Ib/II, open-label, multicenter, randomized umbrella study with a parallel-group model. The registry lists enrollment of 110 and describes Cohort 1 as cancer immunotherapy-naive participants with resectable Stage III melanoma and Cohort 2 as participants with Stage IV melanoma.

The primary Cohort 1 endpoint was pathologic response rate at surgery scheduled at Week 7, based on independent pathologic review. The posted results compare nivolumab plus ipilimumab with three experimental arms. The sponsor's difference intervals are broad and include zero. The registry also reports selected local pathologic response and survival outcomes, along with adverse-event summaries.

The record does not report the original sample-size justification, a full statistical analysis plan, or a multiplicity strategy. The planning figures are illustrations based on observed effects, not the original trial design assumptions.