A Study of DSP-7888 Dosing Emulsion in Combination With Bevacizumab in Patients With Recurrent or Progressive Glioblastoma Following Initial Therapy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of DSP-7888 Dosing Emulsion in Combination With Bevacizumab in Patients With Recurrent or Progressive Glioblastoma Following Initial Therapy (NCT03149003), a randomized phase 3 study. The primary outcome is Number of Participants Who Experienced a Dose-limiting Toxicity.

2. Clinical Question

The broader trial question was whether adding DSP-7888 Dosing Emulsion to bevacizumab could be evaluated against bevacizumab alone in patients with recurrent or progressive glioblastoma after initial therapy. The listed secondary outcomes include overall survival at 12 months, progression-free survival, progression-free survival rate at 6 months, response rate, duration of response, and adverse events. Those outcomes describe the scope of the study, but the registry record does not include results for them. The central distinction is therefore between the trial’s randomized comparison and the primary result that is actually posted: the latter is a small Part 1 safety count, not an estimate of the treatment difference in Part 2.

3. Trial Design

WIZARD 201G, registered as NCT03149003, is described as a randomized, multicenter, adaptive Phase 3 study with a parallel-group design and no masking. Its official title identifies the comparison as DSP-7888 Dosing Emulsion plus bevacizumab versus bevacizumab alone in patients with recurrent or progressive glioblastoma following initial therapy. The study is also described as event driven and open-label. Overall enrollment is listed as 221, and the registry status is completed. The listed start date is 2017-12-08 and the primary completion date is 2021-08-30.

The record distinguishes Part 1 from Part 2. The primary outcome is defined for patients enrolled into Part 1, the safety set, and counts participants experiencing a dose-limiting toxicity. Its evaluation period runs from Day 1 through Day 29. The posted Part 1 combination group has 4.0 participants and a reported outcome count of 0. The Part 2 combination arm and bevacizumab-alone arm have 0.0 participants listed in the primary outcome results, with no outcome values. These entries do not provide a usable randomized comparison for the primary outcome.

Although the overall study is randomized, the primary outcome shown is not an arm-to-arm estimate across the Phase 3 groups. The registry describes the Part 2 treatment arms but does not post primary outcome values for either one. The record lists 8 outcome measures and 6 secondary outcomes, while the number of posted analyses is 0. The listed secondary outcomes include survival, progression, response, and safety measures, but results for those outcomes are not included in the record.

The design description supplies broad structural features, but not all operational details. The randomization ratio, adaptive decision rules, event target, and statistical plan are not reported in the registry record. The term “adaptive” alone does not establish which features could change or how decisions were made.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Analysis limited to patients enrolled into Part 1 - the safety set.

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 Participants Who Experienced a Dose-limiting Toxicity, measured over Dose-limiting toxicity will be evaluated and applied from Day 1 through Day 29.

The primary endpoint counts participants who experienced a dose-limiting toxicity (DLT). The registry defines the population as patients enrolled into Part 1, the safety set, and specifies that DLT is evaluated from Day 1 through Day 29. The posted result is a count rather than a rate, risk difference, or time-to-event estimate. The record lists 4.0 participants in Part 1 Arm 1 and an outcome count of 0.

In general, a DLT endpoint is used to summarize specified toxicities during a defined evaluation period. The count can communicate how many participants met the endpoint, but by itself it does not describe the full range, severity, timing, or duration of adverse events. The registry record does not provide a detailed DLT definition or criteria for deciding which events qualify. That information is not reported in the registry record.

The Part 1 endpoint should be kept distinct from the listed Phase 3 efficacy outcomes, such as overall survival, progression-free survival, and response. A safety count from the Part 1 set does not estimate the effect of the randomized Part 2 comparison on those outcomes. Nor does a zero count in this small posted group establish that no DLTs would occur in a larger or different population. The endpoint answers a limited descriptive question about the participants and evaluation window recorded for Part 1.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Several statistical ideas help distinguish what the registry describes from what the posted results can establish. The design is randomized and parallel-group: in general, random allocation aims to make treatment groups comparable on average, while a parallel design assigns participants to one group rather than crossing them between treatments. Randomization does not itself provide a treatment-effect estimate. That requires outcome data for the relevant groups and an analysis that compares them. The primary results recorded give a Part 1 safety count but no estimable Part 2 comparison.

What does a count measure?

A count records the number of participants meeting a stated endpoint. Here, the endpoint is participants experiencing a dose-limiting toxicity in the Part 1 safety set during the period from Day 1 through Day 29. A count is direct and easy to interpret, but its meaning depends on the population, event definition, and observation window. Without a detailed DLT definition or additional information about event timing and severity, the posted count cannot describe those features. The registry record does not report the detailed criteria.

Counts also need their denominators to convey how much information they represent. The listed Part 1 group has 4.0 participants and an outcome count of 0. That is a small set of observations, so the count should not be generalized as though it precisely characterized a broader patient population. In general, a confidence interval expresses a range of values compatible with an estimate under specified statistical assumptions. No confidence interval is posted for this primary result, and one should not infer a precise underlying event frequency from the count alone.

Why is a two-arm estimate different?

A comparative analysis asks how outcomes differ between groups, rather than describing one group in isolation. In general, an unadjusted comparison of binary outcomes can be based on event counts and denominators in each group. But the registry’s Part 2 primary outcome entries have 0.0 participants and no values in both arms. Without posted outcome information for those groups, there is no comparison to interpret. The record explicitly identifies the two-arm comparison as not estimable for the binary endpoint. This absence does not mean the groups had identical results; it means the posted data do not yield the comparison.

The record lists no analyses. Accordingly, there are no posted primary-endpoint test results, confidence intervals, or analysis details to explain. The registry record does not report whether an adjusted model was planned or used, which covariates would be included, or how a treatment effect would be summarized. These details cannot be inferred from the trial being randomized or from its Phase 3 designation.

What does “event driven” or “adaptive” establish?

In general, an event-driven study bases some aspect of timing or analysis on accumulating events, and an adaptive design may allow prespecified changes as information accrues. Those general labels do not identify the actual event target, the timing of any interim looks, the criteria for adaptation, or whether any adjustment to statistical error rates was planned. The record calls the study event driven and adaptive, but the operational rules and interim-analysis details are not reported in the registry record. It would therefore be inappropriate to infer a particular adaptive procedure.

Likewise, “open-label” means masking is listed as none. In general, knowledge of assigned treatment can affect care, reporting, or assessment, depending on the outcome and how it is measured. The registry record does not describe additional safeguards for outcome assessment, so none should be presumed.

How should planning illustrations be read?

In general, a sample-size calculation describes the number of participants that would be needed under stated assumptions to detect a specified effect with a chosen significance level and power. Such a calculation is a planning illustration, not an observed result and not “observed power.” The sample-size assumptions associated with the record describe a two-sided significance level of 0.05, equal allocation, a normal approximation, and effects defined as fractions of an observed effect; time-to-event calculations are described in terms of total events, while other calculations are described per arm. However, no planning estimates or observed treatment-effect estimate are included in the registry record. No numerical sample-size requirement can be interpreted from these details alone.

In general, over-dispersion matters when outcome variation exceeds what a chosen statistical model expects; ignoring it can make uncertainty estimates too narrow. That concept is relevant to some count outcomes, but the posted primary measure is a participant count and the record does not report a model for it. There is no basis for claiming that over-dispersion was assessed or that a particular model was used.

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("Part 1 - Arm 1: DSP-7888 Dosing Emulsion Plus Beva")
n      <- c(4)
events <- c(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  Part 1 - Arm 1: DSP-7888 Dosing Emulsion Plus Beva  4  0  0.0  0.0  60.2  0.0  49.0  0.5  52.2

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 8 outcome measures, including 6 secondary outcomes, but lists 0 analyses. The secondary outcomes cover overall survival at 12 months, progression-free survival, progression-free survival rate at 6 months, response rate, duration of response, and adverse events. These multiple outcomes raise a general statistical issue: when many hypotheses are tested, the chance of at least one apparently unusual result can increase unless the analysis plan addresses multiplicity. The record does not report a multiplicity strategy, a hierarchy of outcomes, or which outcomes were intended to support confirmatory conclusions.

The registry record also does not report a statistical analysis plan, analysis populations beyond the stated Part 1 safety set for the primary endpoint, or methods for handling missing outcome data. It does not say how participants who discontinue, have incomplete follow-up, or lack an assessment would be handled in analyses of the listed secondary outcomes. Those methods are important because different missing-data assumptions can lead to different estimates, particularly for outcomes requiring follow-up over time.

The posted Part 2 primary endpoint entries have 0.0 participants and no values. That is a limitation in the available registry results, but it should not be relabeled as missing participant outcomes without further information: the reason for the empty entries is not reported in the registry record. The record also does not explain whether any analyses were conducted but not posted, whether outcomes were changed, or how the adaptive features affected the analysis. Without these details, the secondary outcomes and the randomized comparison cannot be reconstructed 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: seizure (25.0%), asthenia (3.7%), cerebrovascular accident (2.2%), alanine aminotransferase increased (2.2%), encephalopathy (1.9%).

The adverse-event information is listed separately for Part 1 and the two Part 2 groups. For serious adverse events, the record lists 1 of 4 participants in Part 1, 10 of 108 in the Part 2 combination group, and 6 of 90 in the Part 2 bevacizumab-alone group. For other adverse events, it lists 4 of 4, 106 of 108, and 79 of 90, respectively. These figures describe participants with recorded events in the listed groups; they are not adjusted comparisons and do not establish a causal difference between treatments.

The registry lists the following serious adverse-event terms and percentages: Seizure, 25.0%; Asthenia, 3.7%; Cerebrovascular accident, 2.2%; Alanine aminotransferase increased, 2.2%; Encephalopathy, 1.9%; and Hydrocephalus, 1.9%. The record does not specify in these entries how the listed terms relate to each arm, nor does it provide event timing or participant-level detail. These term percentages should therefore be read as entries in the registry’s serious-event listing, not as arm-specific comparative risks.

In general, adverse-event counts do not by themselves show whether events were caused by treatment. They also do not communicate duration, recurrence, severity beyond the serious classification, or exposure time. The registry’s death counts are not equivalent to a survival analysis, which would require information about follow-up and time to event. These safety entries are descriptive; the record does not report inferential comparisons for them.

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

The posted primary result establishes a narrow descriptive fact: the registry records 0 participants with a dose-limiting toxicity among 4.0 participants in Part 1 Arm 1, with DLT evaluated from Day 1 through Day 29. It does not establish the frequency of DLT in a larger population with precision, because the posted set is small and no uncertainty interval is reported. It also does not estimate how the combination compares with bevacizumab alone on the primary endpoint.

The trial’s broader randomized Phase 3 question concerns DSP-7888 Dosing Emulsion plus bevacizumab versus bevacizumab alone. The registry lists several outcomes relevant to survival, progression, response, and safety, but does not report their results in the record. Consequently, these results cannot establish a comparative benefit or harm, or resolve the stated efficacy questions. The empty Part 2 primary outcome entries should not be interpreted as zero events, equal outcomes, or proof that the treatments have the same effect.

In general, absence of a statistically significant result is not the same as evidence of no effect; in this record, even a primary comparison result is not posted. No p-value or confidence interval for the randomized comparison is reported. A conclusion about the size, direction, or precision of a treatment effect would therefore go beyond the available results.

The adverse-event entries provide additional descriptive counts for the groups, but without comparative analyses they cannot distinguish treatment effects from chance variation or other differences. Similarly, the registry’s death counts do not replace the listed overall-survival outcome, which would require a time-to-event analysis. The status of the study as completed and its overall enrollment of 221 do not fill these reporting gaps: those details describe the trial, not the outcome estimates available for interpretation.

The appropriate reading is limited but useful. The record documents the trial design and a Part 1 DLT count, along with adverse-event listings. It does not provide an interpretable randomized efficacy estimate or a complete inferential account of the trial. Any broader conclusion about the comparative effects of the study treatments is not supported by these posted results.

14. Why This Trial Matters Statistically

This trial illustrates why the distinction between a study’s overall design and its posted endpoint results matters. WIZARD 201G is described as a randomized, multicenter, adaptive Phase 3 comparison, but the primary result recorded is a Part 1 safety count. The Part 2 randomized arms have no primary outcome values in the registry record. A reader who focuses only on the Phase 3 label or the overall enrollment of 221 could mistakenly assume that a comparative efficacy estimate is available. The posted primary results do not provide one.

The record also illustrates that endpoint definitions and analysis populations shape interpretation. The primary outcome is confined to a Part 1 safety set and a period from Day 1 through Day 29. It should not be conflated with the listed survival, progression, response, and safety outcomes. Keeping populations and endpoints distinct prevents a descriptive safety observation from being misrepresented as evidence about the randomized treatment contrast.

Finally, the trial is described as event driven and adaptive, which makes transparent reporting of decision rules and analysis methods especially important. In general, readers need to know what changes were permitted, what information guided them, and how the statistical analysis accounted for those decisions. Those details are not reported in the registry record. The result is a useful example of how a record can contain substantial design information while still leaving major questions about analysis, precision, and comparative outcomes unanswered.

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 primary endpoint result is based on the Part 1 safety set and lists 4.0 participants in the combination group, with an outcome count of 0. This limited result cannot support a precise estimate of DLT frequency. No confidence interval is reported, and detailed DLT criteria are not reported in the registry record. The defined window is from Day 1 through Day 29, so the primary endpoint count does not describe events outside that period.

The Part 2 entries list 0.0 participants and no primary outcome values for both treatment groups. The reason for these entries is not reported in the registry record. As a result, the primary outcome cannot be interpreted as a randomized comparison. The record lists no analyses and gives no inferential result for the primary endpoint.

Although the study is described as adaptive and event driven, the operational details are not reported in the registry record. The record also does not report the randomization ratio, interim-analysis procedures, analysis plan, multiplicity approach, or methods for missing data. For the listed secondary outcomes, results and analysis details are not reported in the registry record. These limitations prevent evaluation of the planned and executed methods for the broader trial questions.

The adverse-event listings provide participant counts, but not comparative models, uncertainty estimates, exposure-adjusted rates, or the detail needed to interpret the listed serious-event terms by arm. The death counts likewise lack the information needed to interpret time-to-event outcomes. Overall enrollment and completed status do not resolve the absence of comparative results.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study of DSP-7888 Dosing Emulsion in Combination With Bevacizumab in Patients With Recurrent or Progressive Glioblastoma Following Initial Therapy — NCT03149003.
• Registry record last updated 2023-11-15; status last verified 2023-11.
• Results first posted 2023-01-30.
• 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

NCT03149003 is a completed, randomized, open-label, parallel-group, multicenter Phase 3 study of DSP-7888 Dosing Emulsion plus bevacizumab versus bevacizumab alone in recurrent or progressive glioblastoma after initial therapy. Overall enrollment is listed as 221. The study is described as adaptive and event driven.

The posted primary outcome counts participants with dose-limiting toxicity in the Part 1 safety set from Day 1 through Day 29. The combination group lists 4.0 participants and an outcome count of 0. Part 2 has no posted primary outcome values, so the registry record does not provide an estimable two-arm comparison. Adverse-event counts are listed by group, while the results for the listed secondary outcomes and the statistical analysis details are not reported in the registry record.