ONC201 in Adults With Recurrent H3 K27M-mutant Glioma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for ONC201 in Adults With Recurrent H3 K27M-mutant Glioma (NCT03295396), a non-randomized phase 2 study. The primary outcome is Number of Patients With Overall Response.

2. Clinical Question

The primary endpoint was the number of participants with an overall response. The registry defines overall response as achieving either complete response or partial response. Complete response requires disappearance of all target lesions; partial response requires at least a 50% decrease in the sum of products of perpendicular diameters of target lesions from baseline. RANO also considers non-target lesions, corticosteroid use, and performance status.

The question is descriptive as well as clinical: it concerns the observed response experience among participants who received study treatment. Since allocation was not randomized and there was no control group, the study design does not support a direct randomized comparison against another treatment or no treatment. The registry record does not report the number or proportion of participants who achieved response, so the primary result cannot be quantified from the posted record.

3. Trial Design

This was a Phase II, open-label study of dordaviprone (ONC201) in adults with recurrent H3 K27M-mutant glioma. The registry lists a single-group model and nonrandomized allocation, while the study summary describes two arms. Both arms received the same listed weekly dose, but their eligibility descriptions differed. The arms therefore represent distinct cohorts rather than randomized treatment alternatives.

Arm A included patients with recurrent H3 K27M-mutant glioma, including patients with DIPG, primary spinal tumors, and some atypical histologies. Arm B excluded patients whose primary malignant lesion was in the pons or spinal cord, as well as patients with atypical non-astrocytic histologies such as ependymoma and ganglioma. The Arm B description ends mid-list in the registry record, so the full set of exclusions is not reported in the registry record.

The registry lists enrollment of 73. The primary endpoint description says that all patients who received at least one dose were included in the response analysis, including patients with DIPG or primary spinal tumors who were historically not assessable by RANO-HGG because of anatomical or imaging limitations. The record also notes that 9 patients in Arm A, including 7 with primary spinal tumors and 2 with DIPG, were included in the overall analysis despite not being assessable for response per RANO-HGG.

The study is listed as terminated. The reason for termination is not reported in the registry record. The listed start date is 2017-10-30, and the primary completion date is 2022-12-31. Those dates describe the registry timeline; they do not establish why the study ended or whether the planned analyses were completed.

Because the study was open-label and nonrandomized, participants and investigators were not masked to treatment, and assignment to the two arms was not random. The registry record does not report a separate control intervention. These design features matter when interpreting response assessments and any differences between cohorts: neither arm is a randomized control for the other.

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

Analysis population for the primary outcome, as defined in the registry: Note: A total of 9 (21%) patients in Arm A, including 7 with primary spinal tumors and 2 with DIPG, were included in the overall analysis (i.e. Overall Response Rate) despite not being assessable for response per RANO-HGG.

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 Patients With Overall Response, measured over From first dose of study treatment through study completion, an average of 1 year.

The primary endpoint is the number of participants with overall response. It is based on radiographic tumor assessment under RANO criteria for high-grade glioma. Complete response means that all target lesions disappear. Partial response requires at least a 50% decrease from baseline in the sum of products of perpendicular diameters of target lesions. The criteria also take account of non-target lesions, corticosteroids, and performance status. Overall response combines participants meeting either the complete-response or partial-response definition.

This endpoint is used to capture tumor changes that meet prespecified response criteria, rather than relying on an informal description of imaging. Repeated assessments allow response to be evaluated over the course of treatment. In this study, tumor assessments were scheduled at 8 weeks, with a 1-week allowance, after therapy began and every 8 weeks, with a 1-week allowance, thereafter. The stated timeframe runs from first dose through study completion, described as an average of 1 year.

The endpoint has an important scope qualification. All treated participants were included in the overall response analysis, including some with DIPG or primary spinal tumors that were not assessable under RANO-HGG because of anatomical or imaging limitations. The registry notes that 9 Arm A patients were included despite this limitation. The number with observed response, the response proportion, and the handling of participants without an assessable response are not reported in the registry record. Thus, the endpoint definition is available, but its result cannot be evaluated from the posted registry information.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a response count measure?

A count of participants with overall response records how many people met the study’s complete-response or partial-response criteria. It is a count outcome, not by itself a comparison of treatments. To interpret it, readers usually need the denominator analyzed, the response count, and a clearly described analysis population. A response proportion can communicate the share of analyzed participants meeting the criteria, but the registry record does not report that proportion or the response count for this study.

In general, an estimate should be read together with its uncertainty. A confidence interval describes a range of values compatible with the data and the method used to calculate it; it does not guarantee that the true value lies in that range. For a response proportion, interval width is influenced by the number of participants and the observed response frequency. No confidence interval for this primary endpoint is reported in the registry record, so its precision cannot be assessed.

Why does the study design matter?

Random assignment helps balance measured and unmeasured participant characteristics across treatment groups on average. This study is listed as nonrandomized, and its arms have different eligibility descriptions. Differences between the arms could reflect those eligibility differences or other cohort characteristics rather than treatment effects. Since both arms received the same listed intervention and there is no randomized control group, the arm structure does not provide a controlled estimate of dordaviprone’s effect.

Open-label conduct means treatment was not masked. In general, knowledge of treatment assignment can influence care, reporting, or assessment. The registry describes radiographic response criteria and assessment timing, but does not report whether independent central review was used or what procedures were used to reduce assessment bias. Those details are not reported in the registry record.

What does an endpoint definition contribute?

Prespecified criteria make outcome classification more systematic. Here, complete response is based on disappearance of all target lesions, and partial response requires at least a 50% decrease in the specified measurement from baseline. The registry also notes that non-target lesions, corticosteroids, and performance status enter RANO response considerations. A clear definition tells readers what qualifies as an event, but does not reveal how many participants met the definition. That result is not reported in the registry record.

The stated assessment schedule—at 8 weeks, with a 1-week allowance, after therapy initiation and every 8 weeks, with a 1-week allowance, thereafter—describes planned observation points. Repeated assessment can capture changes over time, but it also raises methodological questions about how confirmation, missing scans, and changes in clinical status are handled. Those analysis details are not reported in the registry record.

How should missing information be considered?

In general, missing outcome data can affect both the estimated response frequency and its precision. The effect depends on why data are missing and how missingness is handled. The registry says all participants who received at least one dose were included in the analysis, including participants with DIPG or primary spinal tumors that were not assessable by RANO-HGG. It does not report the number of participants with missing assessments, the rules used to classify unevaluable participants, or a sensitivity analysis for missing outcomes. No particular missing-data method can be inferred.

The inclusion of participants described as not assessable by RANO-HGG is a distinct issue from ordinary missing data. Their inclusion in the overall analysis is stated, but the registry record does not explain how they contributed to the response outcome. The number of Arm A participants noted as not assessable is 9, with 7 primary spinal tumors and 2 DIPG. The record does not provide corresponding participant-level outcome classifications.

What can be inferred from the posted analyses?

The registry lists 0 analyses. It therefore reports no primary endpoint estimate, hypothesis test, confidence interval, adjusted analysis, or arm comparison. A nonsignificant result cannot be inferred, because no test result is reported; likewise, statistical significance cannot be inferred. The absence of an analysis in the registry record does not establish that no analysis was conducted elsewhere.

In general, a p-value describes how incompatible observed data are with a specified null model, under assumptions of the analysis. It is not the probability that a treatment works, nor a measure of clinical importance. A confidence interval adds information about the range and precision of an estimate. Neither type of result is reported for the primary endpoint in this registry record.

Planning calculations are also distinct from observed study results. In general, a sample-size illustration based on an assumed effect describes the number of participants that a specified design might require under its assumptions. It is not observed power and does not establish the probability that this completed study detected an effect. The registry record does not report the trial’s sample-size justification, so the design rationale and any planned power target are not reported in the registry record.

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", "Arm B")
n      <- c(43, 30)
events <- c(5, 5)
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  43  5  11.6  3.9  25.1  5.1  24.5  5.2  24.6
ARM  Arm B  30  5  16.7  5.6  34.7  7.3  33.6  7.5  33.7

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 2 outcome measures, identifies overall response as the primary outcome, and lists duration of response as a secondary outcome. It reports 0 analyses. The statistical plan for handling multiple outcomes or multiple analyses is not reported in the registry record. Therefore, it is not possible to determine whether any multiplicity adjustment was planned or used, or how the primary and secondary outcomes were prioritized in analysis.

Missing-data rules are not reported in the registry record. The primary endpoint description states that all participants who received at least one dose were included, including participants with DIPG or primary spinal tumors who were not assessable under RANO-HGG. It does not specify how participants with absent, incomplete, or unevaluable assessments were classified, whether any observations were imputed, or whether sensitivity analyses were conducted.

The registry’s statement about inclusion in the analysis describes an analysis population based on receiving at least one dose. It does not report a full participant accounting for the primary endpoint, such as how many participants had evaluable assessments or how many were missing outcome information. The record also does not report whether analyses followed a protocol-defined intention-to-treat or other analysis-population definition. Those details are not reported in the registry record.

These omissions limit interpretation of the primary endpoint independently of the response criteria. A well-defined outcome does not resolve uncertainty about missing observations, outcome assessability, or multiplicity. No direction or magnitude of bias can be inferred without further information about the underlying observations and analysis methods.

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: encephalopathy (9.3%), hydrocephalus (9.3%), seizure (6.7%), vomiting (4.7%), brain oedema (4.7%).

The registry lists serious adverse events by arm: 17 of 43 participants in Arm A and 7 of 30 in Arm B were affected. It also lists other adverse events for 42 of 43 participants in Arm A and 27 of 30 in Arm B. Deaths are listed as 33 in Arm A and 25 in Arm B. These counts describe the registry entries; the record does not report the timing, circumstances, cause, or relationship to treatment for the deaths.

The most frequently listed serious-event terms include encephalopathy and hydrocephalus, each at 9.3%, followed by seizure at 6.7%, vomiting at 4.7%, brain oedema at 4.7%, and headache at 4.7%. The registry record does not provide enough detail to determine whether these event terms overlap within participants or how the percentages were calculated. They should not be added together or interpreted as mutually exclusive categories.

These are descriptive counts in nonrandomized cohorts, not a controlled comparison of safety. The record does not report the definitions used for serious and other adverse events, event-level timing, exposure-adjusted rates, or attribution to dordaviprone. Consequently, the figures alone do not establish that treatment caused an event or explain differences between arms.

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

The registry record establishes that a Phase II, open-label, nonrandomized study evaluated weekly dordaviprone in adults with recurrent H3 K27M-mutant glioma. It defines overall response using RANO criteria and describes repeated tumor assessments. It also reports the enrolled population, arm eligibility differences, and selected safety counts. These registry entries help readers understand what was studied and how the primary outcome was intended to be assessed.

The record does not report the primary response count, response proportion, duration-of-response result, statistical test, or confidence interval. It therefore does not quantify the observed antitumor activity or the precision of an estimate. The lack of a reported result should not be interpreted as evidence that there were no responses or that the treatment had no effect.

The study’s nonrandomized design and differing arm eligibility also limit causal interpretation. The two arms are not like-for-like controls, and both received the same intervention. Any difference between them would not, on this record, isolate a treatment effect. There is no randomized comparator from which to estimate what would have happened under another treatment or without study treatment.

The safety counts likewise describe events recorded by arm but do not establish causation. The registry record does not report event attribution, timing, or sufficient context to interpret deaths or adverse events as treatment-related. The event percentages and counts should be understood as descriptive registry information rather than as a comparative estimate of risk.

In short, the posted information supports conclusions about the study’s design and outcome definitions, but not a numerical conclusion about the primary efficacy result. Additional outcome results and analysis details would be needed to evaluate the response estimate, its uncertainty, and the handling of participants who were not assessable under RANO-HGG.

14. Why This Trial Matters Statistically

This trial raises statistical questions that are especially important when studying a heterogeneous population. The registry describes two cohorts receiving the same weekly intervention but with differing eligibility restrictions for tumor location and histology. That structure makes clear why a two-arm study is not necessarily a comparative trial: allocation was nonrandomized, and the arms were not designed as like-for-like controls.

The response endpoint also illustrates the relationship between a formal outcome definition and the population to which it can be applied. RANO-HGG criteria use imaging and other clinical considerations, while the registry notes that some participants with DIPG or primary spinal tumors were not assessable under those criteria because of anatomical or imaging limitations. Including such participants in the overall analysis raises questions about how the endpoint was operationalized for them. The record states that they were included but does not report their outcome classifications.

There is an important distinction between a registered endpoint and a registered result. The primary outcome is clearly named and described, but the record lists no analyses and gives no response estimate. Readers can identify the intended statistical question without being able to assess its answer or precision. That distinction prevents a study description from being mistaken for evidence about the size of an observed effect.

The case also shows why safety summaries need context. Counts of affected participants and named event terms are useful descriptive information, but causal interpretation depends on definitions, timing, attribution, exposure, and appropriate comparison. Those details are not reported in the registry record. Statistical interpretation therefore requires attention not only to the numbers displayed but also to what the record does and does not establish.

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 most consequential limitation of the registry record for statistical interpretation is that it does not report the primary endpoint result. Although the response criteria and assessment schedule are described, the response count, response proportion, uncertainty interval, and analysis result are absent. The secondary outcome of duration of response is named, but its result is also not reported in the registry record.

The nonrandomized, open-label design and different arm eligibility criteria limit the use of arm-to-arm comparisons. The registry identifies the arms as not being a like-for-like control comparison. The record does not describe an external control or a randomized comparator. It also does not report the reason the study was terminated.

Outcome assessability is another specific concern. The registry states that 9 Arm A participants, including 7 with primary spinal tumors and 2 with DIPG, were included in the overall analysis despite not being assessable per RANO-HGG. The method used to classify these participants for the response endpoint is not reported in the registry record. The handling of other missing or unevaluable assessments is also not reported.

The Arm B eligibility description is incomplete in the registry record, ending after examples of excluded atypical non-astrocytic histologies. As a result, the full eligibility distinction between the cohorts cannot be reconstructed from the record. The statistical analysis plan, sample-size justification, missing-data methods, multiplicity procedures, and any adjusted analyses are not reported in the registry record.

Safety interpretation is limited by the absence of event definitions, timing, treatment attribution, and details about deaths. The descriptive safety counts cannot resolve these uncertainties or support a causal comparison between the nonrandomized arms.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: ONC201 in Adults With Recurrent H3 K27M-mutant Glioma — NCT03295396.
• Registry record last updated 2025-07-28; status last verified 2025-07.
• Results first posted 2025-07-28.
• 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

NCT03295396 is a Phase II, open-label, nonrandomized study of dordaviprone (ONC201) in adults with recurrent H3 K27M-mutant glioma. The registry describes two arms receiving 625 mg once weekly, with differing eligibility criteria. Enrollment is listed as 73, and the study status is terminated.

The primary outcome is the number of participants with overall response under RANO criteria for high-grade glioma. The registry describes the response definitions and assessment schedule, but does not report the response result. It lists 2 outcome measures and 0 analyses. Duration of response is identified as a secondary outcome; its result is not reported in the registry record.

Serious and other adverse-event counts, deaths, and selected serious-event terms are reported descriptively by arm. The record does not provide the statistical analysis plan, missing-data methods, reason for termination, or information sufficient for causal safety comparisons.