A Study Of Multiple Immunotherapy-Based Treatment Combinations In Participants With Metastatic Non-Small Cell Lung Cancer (Morpheus- Non-Small Cell Lung Cancer): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study Of Multiple Immunotherapy-Based Treatment Combinations In Participants With Metastatic Non-Small Cell Lung Cancer (Morpheus- Non-Small Cell Lung Cancer) (NCT03337698), a randomized phase 1, phase 2 study. The primary outcome is Stage 1: Percentage of Participants With Objective Response (OR) as Determined by Investigator According to Response Evaluation Criteria in Solid Tumors Version 1.1 (RECIST V1.1).

2. Clinical Question

The study addressed a broad set of treatment comparisons rather than a single two-arm question. Its two cohorts represent different treatment histories and eligibility criteria, so results from one cohort should not be read as directly comparable with those from the other. Stage 2 offered eligible participants a different regimen after progression, loss of clinical benefit, or unacceptable toxicity during Stage 1; Stage 2 is not the population for the primary Stage 1 response outcome.

The trial enrolled 314 participants overall, but the response analyses are reported separately by arm and include much smaller group sizes. The primary outcome’s efficacy population included participants who received at least one dose of each drug in their assigned regimen. The record says a participant in the Stage 1 Cohort 2 Idasanutlin + Docetaxel arm received no study treatment and was excluded from efficacy analyses.

3. Trial Design

This was a Phase Ib/II, randomized, open-label, multicenter, parallel umbrella study. It evaluated multiple immunotherapy-based treatment combinations in participants with metastatic non-small cell lung cancer. The registry record lists an enrollment of 314 and a terminated status. The reason for termination is not reported in the registry record.

Two cohorts were enrolled in parallel. Cohort 1 included participants with tumor PD-L1 expression who had received no prior systemic therapy for metastatic disease. Cohort 2 included participants with disease progression during or after a platinum-containing regimen and a PD-L1/PD-1 checkpoint inhibitor, whether given together as one line or as separate lines, without requiring a particular PD-L1 expression level.

In Stage 1, participants were assigned among several treatment arms. The control was atezolizumab in Cohort 1 and docetaxel in Cohort 2. The posted Cohort 2 arms include docetaxel and a range of atezolizumab combinations. Cohort 1 included atezolizumab, atezolizumab plus cobimetinib, atezolizumab plus CEA-TCB, and atezolizumab plus tiragolumab. The registry describes the original atezolizumab control arm in Cohort 1 as having closed during study conduct and later being reactivated with atezolizumab plus tiragolumab as a new comparator arm.

Participants experiencing progression, loss of clinical benefit, or unacceptable toxicity during Stage 1 could be eligible for a different Stage 2 regimen, subject to the stated eligibility conditions. The study was unmasked, so participants and investigators were not blinded to treatment assignment. The randomization procedure, allocation ratios, and any stratification factors are not reported in the registry record. The record also does not report the rationale for the termination.

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: Eligible participants were assigned to 1 of several treatment arms in Stage 1. Participants with PD, loss of clinical benefit, or unacceptable toxicity in Stage 1 were eligible for a different treatment combination in Stage 2. Participants in Stage 1 who did not enter Stage 2 \& those who completed Stage 2 treatment entered the long-term survival follow-up. Results include only those cohorts that enrolled participants; planned cohorts that were never opened for enrollment are not presented.

Only the first 6 of 19 groups are shown.

Stage 1

Reasons for not completing this period:

Stage 2

Reasons for not completing this period:

Long-term Survival Follow-up

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Efficacy population included all participants who received at least one dose of each drug for their assigned treatment regimen. Cohort 1 with 'Atezolizumab (control arm)' was closed during the study conduct and was reactivated with 'Atezo + Tira' as a new comparator arm. Comparative statistics are not presented for Atezo + Tira arm \& Atezo + CEA-TCB arms (Cohorts 1 \& 2) due to the small sample size (N=1), which precludes any meaningful conclusion.

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: Randomized population included all participants randomized to each treatment arm.

Only the first 6 of 20 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Stage 1: Percentage of Participants With Objective Response (OR) as Determined by Investigator According to Response Evaluation Criteria in Solid Tumors Version 1.1 (RECIST V1.1), measured over Up to 50.4 months.

The primary endpoint was the percentage of participants with an objective response during Stage 1, assessed by the investigator according to RECIST v1.1. An objective response required a complete response or partial response on two consecutive occasions at least four weeks apart. The response rate is the percentage of participants meeting that definition.

A complete response was defined as disappearance of all target and non-target lesions, with any pathological lymph nodes reduced to less than 10 millimeters in short axis. A partial response required at least a 30% decrease in the sum of diameters of target lesions from baseline, without meeting the definition of complete response. The registry record states that 95% confidence intervals for response rates were constructed using the Clopper-Pearson method and that percentages were rounded.

Objective response is used to describe whether measurable disease shrank enough to meet prespecified response criteria. It is a binary outcome at the participant level: each participant either met the response definition or did not. A response percentage is therefore a summary of the observed participant outcomes, not a direct measure of how long a response lasts or how long participants live without progression. Those are distinct outcomes and are reported separately in the registry.

The primary outcome’s timeframe was up to 50.4 months. Its efficacy population consisted of participants who received at least one dose of each drug in their assigned regimen. That population definition matters because it is narrower than the overall enrollment count and because the treatment arms had unequal and often small analysis groups.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is binary at the participant level: a participant either met the objective-response definition or did not. The reported percentage summarizes the number meeting that definition relative to the efficacy analysis group. This makes the percentage intuitive, but its precision depends strongly on the size of that group. In small arms, a single participant’s outcome can materially affect the estimate.

The registry reports 95% confidence intervals for response rates using the Clopper-Pearson method. A confidence interval communicates the range of values reasonably compatible with the observed data under the method’s assumptions. Wider intervals mean less precision. The very wide intervals in the smallest arms reflect the limited information in those groups; they do not imply that the true response rate is known to lie at the observed percentage. Percentages in the registry were rounded.

For Cohort 2 comparisons, the registry reports differences in overall response rates and 95% confidence intervals. A difference of zero would mean equal observed response rates on the scale used for the comparison. An interval containing zero is compatible with a range of possible differences that includes no difference. An interval that excludes zero indicates that the reported interval does not cover zero, but interpretation still depends on the comparison method, analysis plan, and multiplicity handling. Those methods are not reported in the registry record, and p-values are not reported for these comparisons.

For time-to-event outcomes such as progression-free survival and overall survival, medians describe the time by which half of the analysis group experienced the event, subject to the endpoint definition and censoring. A hazard ratio summarizes the relative event rates over follow-up under the model used for its estimation. A hazard ratio confidence interval that includes 1 is compatible with no relative difference under that measure, as well as with effects in either direction covered by the interval. The registry reports hazard ratios and 95% confidence intervals for selected comparisons, but the analysis method and p-values are not reported. The intervals are broad, so the estimates are imprecise.

Some registry outcomes have values that are not estimable or are missing for particular arms. A missing median or an absent confidence limit should not be read as evidence of no event or as a favorable outcome. It indicates that the corresponding summary is not available in the posted results. Small event counts can also leave survival estimates and their intervals unstable. For duration of response, only participants with a documented response contribute to that outcome, so analysis groups can be smaller still.

The registry includes planning illustrations based on observed effects. These illustrations assume a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; for time-to-event outcomes, the planning figures are total events. Rows considering 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect express hypothetical planning scenarios, not the trial’s observed power and not the original trial’s sample-size justification. The design’s sample-size rationale is not reported in the registry record. Such calculations are sensitive to the observed effect, which can be unstable in small groups, and should not be interpreted as a prediction that a future study will reproduce it.

Over-dispersion is a general concern when outcome variability exceeds what a simple statistical model assumes. In a multi-arm study with heterogeneous populations and small groups, model assumptions and the chosen analysis method can affect uncertainty estimates. The registry record does not state whether any over-dispersion assessment or adjustment was used. More generally, confidence intervals and point estimates are only as informative as the underlying analysis population, endpoint definitions, follow-up, and statistical model.

The primary response confidence intervals have a stated method, but the registry does not report the method for the response-rate difference comparisons. It also does not report an adjustment for the many treatment arms and outcomes. If multiple comparisons are made, the chance of at least one apparently unusual result can increase; this is a general statistical issue, not proof that any particular result is spurious. Without a reported multiplicity strategy, isolated comparisons should be treated cautiously.

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("Stage 1 Cohort 1: Control (Atezolizumab)", "Stage 1 Cohort 1: Atezolizumab + Cobimetinib", "Stage 1 Cohort 1: Atezolizumab + CEA-TCB", "Stage 1 Cohort 1: Atezolizumab + Tiragolumab", "Stage 1 Cohort 2: Control (Docetaxel)", "Stage 1 Cohort 2: Atezolizumab + Cobimetinib")
n      <- c(3, 5, 1, 1, 61, 15)
events <- c(1, 0, 1, 0, 8, 0)
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  Stage 1 Cohort 1: Control (Atezolizumab)  3  1  33.3  0.8  90.6  6.1  79.2  6.8  80.6
ARM  Stage 1 Cohort 1: Atezolizumab + Cobimetinib  5  0  0.0  0.0  52.2  0.0  43.4  0.4  45.9
ARM  Stage 1 Cohort 1: Atezolizumab + CEA-TCB  1  1  100.0  2.5  100.0  20.7  100.0  15.8  98.7
ARM  Stage 1 Cohort 1: Atezolizumab + Tiragolumab  1  0  0.0  0.0  97.5  0.0  79.3  1.3  84.2
ARM  Stage 1 Cohort 2: Control (Docetaxel)  61  8  13.1  5.8  24.2  6.8  23.8  6.9  23.9
ARM  Stage 1 Cohort 2: Atezolizumab + Cobimetinib  15  0  0.0  0.0  21.8  0.0  20.4  0.2  20.6

11. Analyses Reported by the Sponsor

The registry reports the sponsor’s analyses for selected treatment comparisons, including response-rate differences and hazard ratios with 95% confidence intervals. For the primary response endpoint, one reported Cohort 2 comparison is the docetaxel control versus atezolizumab plus cobimetinib; its reported difference is -13.11 with a 95% confidence interval from -25.74 to -0.49. Other posted response comparisons have intervals that include zero. The registry does not give p-values or identify the method used for those difference estimates.

For the selected time-to-event comparisons, the registry reports hazard ratios and confidence intervals, but not the analysis method or p-values. The reported estimates should be read alongside their intervals, which are wide and include 1. The registry record does not establish the statistical procedure used to obtain those estimates.

Any simple unadjusted re-calculation from posted summary statistics is a separate descriptive check, not a replacement for the sponsor’s analysis. No discrepancy is identified in the registry record. When an unadjusted calculation differs from a sponsor-reported analysis, the sponsor’s own analysis takes precedence; such calculations also cannot reproduce adjustments or individual-participant methods that are not reported. Here, the registry’s missing method details limit the extent to which the reported comparisons can be independently assessed.

12. Multiplicity, Missing Data and Analysis Populations

The study includes many treatment arms, two cohorts, and multiple outcomes. This creates a multiplicity concern: as the number of comparisons increases, the chance of finding at least one result that appears unusual by chance can also increase. This general issue makes it important to know which comparisons were prespecified as primary, whether secondary outcomes were tested formally, and whether any adjustment or hierarchical testing plan was used. Those details are not reported in the registry record.

The registry identifies the Stage 1 objective-response outcome as primary and lists secondary outcomes that include survival, response duration, disease control, and adverse events. It reports several pairwise comparisons, but their statistical methods and p-values are not available. A complete statistical analysis plan, including any interim analyses or stopping rules, is not reported in the registry record. The record also does not report the sample-size justification for the trial’s design. Planning illustrations derived from the observed effects do not supply that missing design rationale.

The primary efficacy population included participants who received at least one dose of each drug in the assigned regimen. The registry further notes that a participant in the Stage 1 Cohort 2 Idasanutlin + Docetaxel arm did not receive study treatment and was excluded from efficacy analyses. Details of how other missing outcome data were handled, including any imputation or sensitivity analyses, are not reported in the registry record. Methods for censoring in time-to-event outcomes are also not reported in the registry record.

The registry describes random allocation and an open-label design, but the randomization procedure, allocation ratios, stratification factors, and concealment details are not reported. These omissions make it harder to assess how baseline balance was protected and how the analyses were specified. The reason for termination and the relation between termination and any planned outcome analyses are not reported in the registry record. As a result, the available results support cautious descriptive interpretation rather than a complete evaluation of the prespecified testing strategy.

13. 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: atrial fibrillation (100.0%), colitis (100.0%), covid-19 (100.0%), respiratory tract infection (100.0%), spinal fracture (100.0%).

The safety results report counts for participants at risk and affected by serious adverse events, other adverse events, and deaths in the listed Stage 1 and Stage 2 groups. These are counts, not comparative causal estimates. They should be interpreted in the context of the group sizes, which vary considerably. For example, the Stage 1 Cohort 2 docetaxel group lists 61 participants at risk for serious adverse events and 20 affected; the corresponding atezolizumab plus cobimetinib group lists 15 at risk and 8 affected.

The registry also lists selected serious adverse-event terms with percentages of 100.0%. The term-specific denominators are not reported in the registry record, so those percentages should not be read as the proportion of all participants in an arm who experienced the event. The registry’s displayed counts do not establish that an event was caused by treatment, and the posted summary does not provide enough detail to compare the safety profiles of regimens reliably.

Death counts are presented separately from adverse-event counts. The record does not say in these figures whether deaths were considered treatment-related. Different group sizes, follow-up, treatment exposure, and eligibility for Stage 2 can all affect raw counts; corresponding adjustment methods are not reported. The safety summaries are therefore descriptive and should not be interpreted as a randomized comparison of risk without further analysis.

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

The results describe what was observed among the participants included in the posted analyses. They show that objective responses occurred in some treatment arms and not in others, with estimates that differ in precision. The confidence intervals make that precision visible: large intervals, particularly in small arms, allow a wide range of underlying response rates. A point estimate alone does not capture that uncertainty.

The reported Cohort 2 comparison of docetaxel with atezolizumab plus cobimetinib has a confidence interval for the response-rate difference that excludes zero. This is a feature of that reported interval, not a basis for a broad conclusion about all combinations or both cohorts. Other response comparisons have intervals including zero, and the statistical methods and multiplicity strategy for the comparisons are not reported. The posted findings do not establish a reliable ranking across the many regimens.

These outcomes do not answer every clinically relevant question. Objective response is not the same as response duration, progression-free survival, overall survival, or safety. The study’s secondary outcomes address some of those dimensions, but their estimates are also limited by uncertainty and missing analysis details. The cohorts differ in treatment history and control regimen, so results should not be carried across cohorts as if they came from one common comparison.

Because the study is open-label and the posted analysis groups are often small, the observed numbers should be read as descriptive evidence from this trial rather than as definitive estimates of treatment effects. The registry does not report the reason for termination, a complete statistical analysis plan, or the basis for the original sample size. None of the posted statistics alone establishes what an individual should receive; the results are a record of trial outcomes, not medical advice.

15. Why This Trial Matters Statistically

This trial is statistically notable for its umbrella structure: multiple combinations were evaluated in parallel across cohorts defined by different treatment histories. That structure can efficiently explore several regimens within one study, but it also creates many arms and comparisons. Interpretation depends on keeping the cohort-specific populations and control groups distinct, and on accounting for the increased multiplicity that comes with numerous analyses.

The posted arm sizes range from very small groups to larger groups, so precision is uneven. The confidence intervals make this imbalance consequential: estimates based on a few participants can be extremely uncertain, while larger arms still have uncertainty. A response percentage without its interval can make small-group findings look more definite than the data support.

The registry also illustrates why endpoint types should not be collapsed into one narrative. Objective response is a proportion, while progression-free and overall survival are time-to-event outcomes, and duration of response applies only to participants who responded. The analysis populations and amount of information can therefore differ across endpoints. Comparing point estimates across these measures without accounting for their definitions and uncertainty would be misleading.

Finally, the record includes selected sponsor-reported comparisons but leaves important statistical details unreported, including the comparison methods, p-values, multiplicity handling, and the original sample-size rationale. Those omissions constrain reproducibility and make the distinction between exploratory signals and confirmatory evidence harder to assess. The results are most informative when read as a multi-arm set of estimates with substantial variation in precision, rather than as a definitive ranking of treatment combinations.

16. 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 limitations are specific to the posted record. The Cohort 1 arms are exceptionally small, including arms with only 1 participant, and the corresponding response estimates have very wide confidence intervals. Some Cohort 2 arms are also much smaller than the docetaxel control group. Unequal group sizes and small analysis groups limit the precision of comparisons.

The two cohorts have different eligibility criteria and different control regimens. Their results are not direct comparisons with one another. In Cohort 1, the atezolizumab control was closed during study conduct and later reactivated with atezolizumab plus tiragolumab as a new comparator arm. This change complicates interpretation of the cohort’s comparator history; the registry does not report further details about the change or its analytical implications.

The study is open-label, and the primary endpoint was investigator-assessed. The record does not report additional assessment procedures that would clarify how potential knowledge of treatment assignment was handled. The reported efficacy population requires receipt of at least one dose of each assigned drug, while details of other missing-data decisions are not reported in the registry record.

The registry status is terminated, but the reason for termination is not reported. It is also not reported whether termination affected follow-up or the planned analyses. Statistical methods for the reported response-rate differences and selected survival comparisons, p-values, multiplicity procedures, and the full analysis plan are not reported in the registry record. These limitations do not invalidate the posted estimates, but they constrain the strength and reproducibility of comparative conclusions.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: A Study Of Multiple Immunotherapy-Based Treatment Combinations In Participants With Metastatic Non-Small Cell Lung Cancer (Morpheus- Non-Small Cell Lung Cancer) — NCT03337698.
• Registry record last updated 2026-05-12; status last verified 2026-04.
• Results first posted 2026-05-12.
• 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.

20. Record Summary

The registry record describes NCT03337698 as a randomized, open-label, Phase Ib/II, multicenter, parallel umbrella study of immunotherapy-based combinations for metastatic non-small cell lung cancer. It lists an enrollment of 314 and a terminated status. The study enrolled two cohorts with different treatment histories and evaluated multiple Stage 1 arms, with Stage 2 options for eligible participants after progression, loss of clinical benefit, or unacceptable toxicity.

The primary Stage 1 outcome was investigator-assessed objective response under RECIST v1.1. The posted estimates vary across arms and are often imprecise, especially in small groups. Selected secondary outcomes include progression-free survival, overall survival, duration of response, disease control, and adverse events. The registry record does not report several important statistical details, including the methods for some comparisons, multiplicity handling, the original sample-size rationale, or the reason for termination.