A Study Evaluating the Safety and Efficacy of Glofitamab or Mosunetuzumab in Combination With Gemcitabine Plus Oxaliplatin in Participants With Relapsed or Refractory Diffuse Large B-Cell Lymphoma and High-Grade Large B-Cell Lymphoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study Evaluating the Safety and Efficacy of Glofitamab or Mosunetuzumab in Combination With Gemcitabine Plus Oxaliplatin in Participants With Relapsed or Refractory Diffuse Large B-Cell Lymphoma and High-Grade Large B-Cell Lymphoma (NCT04313608), a non-randomized phase 1 study. The primary outcome is Number of Deaths Due to Adverse Events (AEs).

2. Clinical Question

The broader study aim was to evaluate safety and efficacy for the two combinations in participants with relapsed or refractory diffuse large B-cell lymphoma or high-grade large B-cell lymphoma. Secondary outcomes included treatment tolerability, complete response and objective response according to the 2014 Lugano Response Criteria, and maximum serum concentration of glofitamab. These registered aims identify what the study intended to examine; they do not, by themselves, establish the results for those outcomes. Efficacy estimates are not reported in the registry record.

The clinical question is therefore best read as an early-phase assessment of two regimens, with safety as the primary registered outcome. Because treatment allocation was nonrandomized, the arm results should not be treated as a randomized comparison of the regimens. The registry record does not report how participants were assigned to the arms or provide an adjustment strategy for differences between them.

3. Trial Design

This was a Phase Ib, open-label, multicenter study with nonrandomized allocation and a sequential model. The record gives an enrollment of 23 and lists the study as completed. The study start date was 2020-06-04, and the primary completion date was 2021-10-26. Those dates describe the registry timeline; they do not provide information about individual participants’ follow-up duration beyond the primary outcome’s stated timeframe.

Arm A evaluated glofitamab with gemcitabine and oxaliplatin, administered in 21-day cycles for up to 8 cycles, followed by up to 4 cycles of glofitamab monotherapy. A single dose of obinutuzumab was to be given 7 days before the first glofitamab dose. Arm B evaluated mosunetuzumab with gemcitabine and oxaliplatin, administered in 21-day cycles for up to 8 cycles. The listed interventions also include tocilizumab, but the registry information does not describe its use in either arm.

Open-label means that the study was not masked, as also indicated by the record. In general, lack of masking can affect treatment delivery, reporting, or assessment, depending on the outcome and study procedures. The record does not explain whether particular assessments used independent review or other safeguards. The meaning of “sequential” in the study’s implementation, including the order or timing of any cohorts, is not reported in the registry record.

Nonrandomized allocation matters when interpreting differences between the arms. Random assignment is one way to balance measured and unmeasured factors on average; this design did not use that approach. The record itself cautions that the arms are not a like-for-like control comparison. It does not report a protocol rationale for the allocation method, a sample-size justification, interim analyses, or a detailed 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.

Overall Study

Reasons for not completing this period:

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.

6. Primary Endpoint

The registry lists the primary outcome as Number of Deaths Due to Adverse Events (AEs), measured over Baseline - 90 days after last dose of study treatment.

The primary endpoint was the number of deaths due to adverse events, in participants, from baseline through 90 days after the last dose of study treatment. The posted counts are 1 in Arm A and 0 in Arm B. These are event counts, not estimates of treatment effect, and the registry record does not provide a confidence interval or a statistical test for this endpoint.

The record defines an adverse event broadly as an untoward medical occurrence in a participant given a pharmaceutical product. An event does not have to be causally related to the treatment to meet that definition. The definition includes unfavorable signs, symptoms, disease, and abnormal laboratory findings temporally associated with product use; worsening pre-existing conditions are also considered adverse events. The endpoint’s wording specifies deaths due to adverse events, but the registry record does not give further details about attribution procedures or adjudication.

As general context, a count endpoint can summarize how many participants experienced a specified event during a defined observation period. It is useful for describing an important safety outcome, but a count alone does not show how precise an underlying risk estimate is or explain why an event occurred. Interpretation also depends on the number of participants observed, the completeness and length of follow-up, and consistent event classification. The posted arm counts and at-risk counts provide some context, but other details needed to assess precision and ascertainment are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record describes a binary-outcome study record, with outcomes expressed as participants with an event. For the primary endpoint, the event is death due to an adverse event within the stated observation window. Counts are a direct descriptive way to report how many participants experienced an outcome. They do not, by themselves, quantify the uncertainty around an event rate or test whether arm results differ beyond what might occur by chance.

What does a confidence interval add?

As general statistical context, a confidence interval gives a range of values compatible with the observed data under a specified statistical method and its assumptions. It helps readers judge precision: narrower intervals generally indicate more precise estimates, while wider intervals indicate greater uncertainty. The registry record does not report confidence intervals for the primary deaths or the other safety counts. Without them, readers cannot assess the interval-based precision of an estimated event rate from this record.

A confidence interval would not resolve every interpretive issue. It quantifies sampling uncertainty under the analysis model, but it does not correct nonrandom allocation, missing information, outcome misclassification, or differences in follow-up. In this study, the record explicitly warns that the arms are not a like-for-like control comparison. Even a precise estimate would not, on its own, establish a causal difference between the regimens.

Why does the allocation matter?

As general context, randomization can help make treatment groups comparable at baseline by assigning treatment independently of participant characteristics. In a nonrandomized study, group composition may reflect clinical or operational decisions. If those factors are related to outcomes, observed differences may be affected by confounding. The registry record does not describe participant assignment details, baseline characteristics, or an adjustment analysis. Consequently, it is not possible to determine from the record whether the arms differed in prognostic factors or how any such differences were handled.

The study is also identified as sequential. That label is part of the design description, but the registry record does not explain the implementation or report interim analyses. It would be inappropriate to infer a stopping rule, cohort sequence, or repeated-testing procedure from the design label alone.

What do the sample-size illustrations mean?

The record describes sample-size illustrations as planning calculations based on an observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation. The intended effect is expressed as a stated fraction of the observed effect. For time-to-event outcomes, the described calculation uses a total event count; for other outcomes, it gives participants per arm. The numerical illustration results are not reported in the registry record. These calculations, as characterized in the record, are planning illustrations and not observed power. They are not evidence that the study was designed to detect a particular clinically meaningful effect, and they do not replace a protocol-based sample-size justification. The trial’s own sample-size justification is not reported in the registry record.

In general, such planning calculations depend on their assumptions. Equal allocation may not match actual group sizes, and a normal approximation may be less reliable when event counts are sparse. An assumed effect based on observed data may also be uncertain. Because the illustration values and underlying effect estimates are not reported, their sensitivity or practical implications cannot be assessed.

How should multiple outcomes be considered?

The registry record lists 8 outcome measures, including the primary safety outcome and secondary measures. When a study examines multiple outcomes, the chance of finding at least one apparently notable result can increase if each is tested separately without an adjustment or a prespecified hierarchy. This is a general statistical issue, not a claim that such testing occurred in this study. The record does not report a multiplicity strategy, testing hierarchy, or adjustment. It also reports 0 analysis entries, so inferential analyses cannot be reconstructed from the registry record.

Likewise, methods for missing outcome data and analysis populations are not reported in the registry record. As general context, missingness can affect estimates if participants with unavailable outcomes differ systematically from those with observed outcomes. An analysis population definition clarifies which participants contribute to each result. Without the missing-data procedures, analysis populations, and outcome estimates, readers cannot evaluate those aspects of the study’s statistical analysis.

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("Arm A: Glofit-GemOx", "Arm B: Mosun-GemOx")
n      <- c(17, 6)
events <- c(1, 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  Arm A: Glofit-GemOx  17  1  5.9  0.1  28.7  1.0  27.0  1.4  27.3
ARM  Arm B: Mosun-GemOx  6  0  0.0  0.0  45.9  0.0  39.0  0.4  41.0

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists 8 outcome measures, including the primary endpoint of deaths due to adverse events and secondary measures covering tolerability, complete response, objective response, and maximum serum concentration of glofitamab. It reports 0 analysis entries. No statistical testing hierarchy, multiplicity adjustment, or approach to interpreting results across the outcomes is reported in the registry record.

As general statistical context, examining multiple outcomes can create multiple opportunities for apparently notable results. Prespecifying which outcome is primary and how other outcomes will be interpreted can help readers distinguish confirmatory questions from exploratory ones. The record identifies a primary outcome, but the remaining details of a multiplicity plan are not reported. This does not establish that no such plan existed outside the registry entry.

Missing-data methods and analysis populations are also not reported in the registry record. The posted safety information gives at-risk counts, but does not explain how missing observations, withdrawals, or incomplete follow-up were handled for each outcome. The primary outcome’s timeframe is specified as baseline through 90 days after the last dose; details of ascertainment across that period are not provided.

As general context, missingness can influence estimates when the availability of outcome information is related to participant characteristics or outcomes. An analysis-population definition indicates which participants were included in a particular analysis and can clarify how exclusions or incomplete records were handled. Without these definitions and procedures, the reported counts can be described, but their completeness and the effects of missing information cannot be assessed from the registry record.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: pyrexia (35.3%), fall (33.3%), sepsis (17.6%), diarrhoea (16.7%), asthenia (16.7%).

The registry record reports serious adverse events in 12 participants among the 17 at risk in Arm A and 4 among the 6 at risk in Arm B. It reports other adverse events in 17 of 17 participants in Arm A and 6 of 6 in Arm B. These are participant counts as recorded, not measures of the number of separate events per participant.

The listed serious-event terms and percentages are pyrexia, 35.3%; fall, 33.3%; sepsis, 17.6%; diarrhoea, 16.7%; asthenia, 16.7%; and hepatic cirrhosis, 16.7%. The registry information does not identify the arm or denominator for these percentages, so they should not be assigned to either arm. It also does not report term-level counts in the information available for this interpretation.

As general context, an adverse event is an untoward medical occurrence and need not be caused by treatment. The record’s definition includes events temporally associated with product use and worsening pre-existing conditions. Seriousness and treatment causality are distinct concepts; the posted counts do not establish that an event was caused by a study regimen. The record does not provide details of seriousness criteria or event adjudication. Because the study was nonrandomized and the arm sizes were limited, these descriptive safety figures do not support a causal comparison between regimens.

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

The results reported in the registry record describe safety events observed among participants in two study arms. The primary endpoint count was 1 death due to an adverse event in Arm A and 0 in Arm B. Serious adverse events were recorded in 12 participants in Arm A and 4 in Arm B, with 17 and 6 participants at risk, respectively. Other adverse events were recorded in all participants at risk in each arm. These observations characterize what was recorded; they do not establish why an event occurred.

The study was Phase Ib, open-label, nonrandomized, and sequential. The registry record specifically cautions that the arms are not a like-for-like control comparison. Therefore, the posted counts cannot determine whether one regimen is safer or more effective than the other. Differences in participant characteristics, follow-up, or other factors could affect comparisons, and the record does not report adjustment methods to address such possibilities.

The study was designed to evaluate safety and efficacy, and registered secondary outcomes included complete response and objective response according to the 2014 Lugano Response Criteria. However, estimates for those efficacy outcomes are not reported in the registry record. The safety counts cannot substitute for response results, and the stated study aims should not be mistaken for demonstrated outcomes.

As general statistical context, estimates from small groups are often imprecise, and interpreting them requires attention to uncertainty as well as the observed count. The record does not report confidence intervals or statistical tests for the posted safety outcomes. It also does not provide detailed missing-data methods, analysis populations, or outcome-assessment procedures. Thus, the strongest supported interpretation is descriptive: these are the recorded counts for the named arms and outcomes, subject to important limits in precision, comparability, and reporting.

14. Why This Trial Matters Statistically

This study is statistically notable as an early-phase evaluation of two treatment combinations in relapsed or refractory diffuse large B-cell lymphoma or high-grade large B-cell lymphoma. It combines a registered safety endpoint with secondary response, tolerability, and pharmacokinetic outcomes. The record describes a Phase Ib, open-label, multicenter design and reports enrollment of 23 participants. These characteristics define the setting in which the posted counts should be read.

The nonrandomized allocation is especially important to statistical interpretation. The registry record warns that the arms are not a like-for-like control comparison. That limitation prevents straightforward causal interpretation of any difference between their observed outcomes. The counts can inform a description of what was recorded in the study groups, but the design does not isolate treatment assignment as the reason for a difference.

The record also illustrates why outcome definitions and observation windows matter. The primary endpoint focuses on deaths due to adverse events from baseline through 90 days after the last dose. The adverse-event definition includes occurrences that need not be caused by treatment. Reading the endpoint alongside that definition helps distinguish an event classification from a causal conclusion.

Finally, the registry entry does not report inferential analyses, confidence intervals, sample-size justification, multiplicity procedures, or missing-data methods. Their absence limits assessment of precision and of how results across outcomes were handled. As general statistical context, transparent reporting of these elements helps readers judge uncertainty, comparability, and the scope of conclusions. The posted record supports cautious description of observed safety counts, not a definitive comparative claim.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The arm comparison is limited by the nonrandomized allocation. The registry record explicitly says the arms are not a like-for-like control comparison, and it does not report an adjustment strategy for baseline differences or other confounding. Open-label conduct may also matter for some outcomes as a general methodological consideration; masking was absent, and the record does not describe independent outcome review.

Enrollment was 23, with at-risk counts of 17 in Arm A and 6 in Arm B for the reported safety results. These group sizes constrain the precision and stability of descriptive event estimates. The record does not report confidence intervals or statistical tests for the safety counts, so their uncertainty cannot be quantified from the registry information.

The listed serious-event percentages are not accompanied by arm-specific attribution or term-level counts. The registry record also does not describe event adjudication, ascertainment procedures, or how treatment causality was evaluated. No efficacy estimates are reported, despite registered response outcomes. The number and details of participants contributing to each secondary outcome are not reported in the registry record.

Other limitations in the available record include no reported sample-size justification, interim-analysis details, statistical analysis plan, multiplicity strategy, missing-data methods, or analysis-population definitions. The meaning and implementation of the sequential design are not described. These omissions do not prove that the study lacked such procedures; they limit what can be verified and interpreted from the registry record. The stated primary endpoint timeframe gives an observation window, but detailed follow-up completeness is not reported.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study Evaluating the Safety and Efficacy of Glofitamab or Mosunetuzumab in Combination With Gemcitabine Plus Oxaliplatin in Participants With Relapsed or Refractory Diffuse Large B-Cell Lymphoma and High-Grade Large B-Cell Lymphoma — NCT04313608.
• Registry record last updated 2024-03-25; status last verified 2023-09.
• Results first posted 2024-03-25.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

NCT04313608 was a completed Phase Ib, open-label, multicenter, nonrandomized sequential study sponsored by Hoffmann-La Roche. It evaluated glofitamab with gemcitabine and oxaliplatin or mosunetuzumab with gemcitabine and oxaliplatin in participants with relapsed or refractory diffuse large B-cell lymphoma or high-grade large B-cell lymphoma. Enrollment was 23.

The primary outcome was the number of deaths due to adverse events from baseline through 90 days after the last dose. The posted counts were 1 in Arm A and 0 in Arm B. Serious adverse events were reported in 12 of 17 participants at risk in Arm A and 4 of 6 in Arm B. The registry record cautions that the nonrandomized arms are not a like-for-like control comparison.

Registered secondary outcomes included tolerability, complete response, objective response, and maximum serum concentration of glofitamab. Efficacy estimates and inferential analyses are not reported in the registry record. Sample-size justification, missing-data methods, analysis populations, and multiplicity procedures are also not reported.