Immune Checkpoint Inhibitor Nivolumab in People With Recurrent Select Rare CNS Cancers: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Immune Checkpoint Inhibitor Nivolumab in People With Recurrent Select Rare CNS Cancers (NCT03173950), a na phase 2 study. The primary outcome is Disease Control Rate (Complete Response (CR) + Partial Response (PR) or Durable Stable Disease (SD) for 6 Months Measured in a Variety of Recurrent Refractory Primary Central Nervous System Tumors Reported With Standard Deviation.

2. Clinical Question

3. Trial Design

This was a Phase II study sponsored by the National Cancer Institute (NCI). The registry lists a planned enrollment of 136 and describes a single-group, unmasked design with allocation not applicable. These features mean the study was not designed as a randomized comparison between nivolumab and a control treatment. The adverse-event record separates participants into a heavily pre-treated cohort, a non-heavily pre-treated cohort, and a group enrolled but not treated or completing treatment cycles; the registry does not establish these groups as comparable controls.

The treatment description specifies nivolumab by intravenous infusion. Participants were described as receiving 240 mg every 2 weeks during cycles 1 through 2, then 480 mg every 4 weeks for 14 additional doses. The brief summary describes screening examinations and laboratory tests, MRI and neurologic assessments during the study, and symptom and quality-of-life questionnaires. Follow-up after treatment was planned for up to 5 years. These are registry descriptions of the study procedures; they do not establish how completely the procedures were carried out for each participant.

The registry lists a start date of 2017-07-13, a primary completion date of 2025-06-23, and a status of terminated. The reason for termination is not reported in the registry record. A termination status alone does not explain whether the study stopped early, how much follow-up was completed, or what conclusions the sponsor reached.

The primary endpoint concerns disease control across a variety of recurrent refractory primary CNS tumors. The registry also lists overall survival, symptom burden and interference, changes in symptom scores, and adverse events among secondary outcomes. The record does not report a comparator group or provide outcome estimates that would support a comparative treatment effect.

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: All participants enrolled who have data collected are reported.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: 102/136 participants were analyzed because 19 participants were ineligible, 1 pursued other treatment, 1 withdrew consent and decided not to participate, 9 declined to participate before treatment started (after consent) ,1 withdrew consent and decided not to participate in cohort 2 and 3 did not complete a minimum of one cycle in cohort 1.

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: All participants enrolled who have baseline data are reported.

6. Primary Endpoint

The registry lists the primary outcome as Disease Control Rate (Complete Response (CR) + Partial Response (PR) or Durable Stable Disease (SD) for 6 Months Measured in a Variety of Recurrent Refractory Primary Central Nervous System Tumors Reported With Standard Deviation, measured over From cycle one Day 1 of the treatment through the end of treatment, up to 64 weeks.

The primary endpoint is the disease control rate, defined as a confirmed complete response or partial response, or durable stable disease for 6 months. It was assessed using Immunotherapy Response Assessment in Neuro-Oncology (iRANO) criteria. The registry describes the endpoint as a proportion of participants and gives an assessment period from cycle one Day 1 through the end of treatment, up to 64 weeks.

Under the definition in the record, complete response requires disappearance of all enhancing measurable and non-measurable disease sustained for at least 4 weeks. Partial response requires at least a 50% decrease from baseline in the summed products of perpendicular diameters of measurable enhancing lesions, sustained for at least 4 weeks. Durable stable disease is disease that does not qualify as complete response, partial response, or progression and remains stable for 6 months. Progression is described as an increase greater than 25% in the summed products of perpendicular diameters of enhancing lesions, using best response or baseline as specified in the record.

A disease-control endpoint can capture more than tumor shrinkage: it includes qualifying stable disease as well as complete or partial response. That makes the endpoint relevant to a study asking whether disease can be controlled during treatment. However, the registry record does not report the observed disease-control estimate or its standard deviation, so the extent and precision of the endpoint cannot be assessed from the posted results. The record also labels the endpoint as a mean with standard deviation despite describing it as a proportion; this leaves the intended statistical presentation unclear.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a single-group design estimate?

In a single-group study, participants receive the study intervention without a concurrent control arm. An outcome such as disease control can be summarized as the observed experience of the analyzed participants. It is not, on its own, a comparison of what happened with nivolumab against what would have happened to similar participants under another treatment or no treatment. That distinction matters because disease course, previous therapies, eligibility decisions, and follow-up can all influence observed outcomes.

The registry lists allocation as not applicable and masking as none. These design features do not create a randomized contrast. Nor does separating participants into cohorts for reporting safety make those cohorts interchangeable control groups. The record does not report a comparative analysis that would support interpreting cohort differences as treatment effects.

What does the disease-control endpoint capture?

The primary endpoint combines confirmed complete response, confirmed partial response, and durable stable disease for 6 months. This composite can count participants whose tumors shrink as well as participants whose disease remains stable for the required duration. It is therefore broader than a response-only endpoint. Its meaning depends on the component definitions, assessment schedule, confirmation requirements, and the population included in the analysis.

The registry refers to iRANO criteria and defines response using changes in enhancing lesions, with response sustained for at least 4 weeks. Stable disease must meet the specified duration and must not qualify as response or progression. As these components represent different clinical states, a single combined rate does not show how many participants met each component unless component-specific results are also reported. Those component counts are not reported in the registry record.

How should a proportion and its uncertainty be read?

A proportion describes the share of an analysis population meeting a defined outcome. Its uncertainty reflects that the observed participants are not the only possible group of similar participants: a different sample could yield a different estimate. A confidence interval expresses a range of values compatible with the data and the interval method, under the method's assumptions. A narrow interval conveys more precision than a wide interval, but neither establishes that the result is free from bias or applies to populations unlike those studied.

The registry calls the primary measure a proportion of participants but also labels it as a mean with standard deviation. A standard deviation describes variation among individual measurements; it is not the same as the uncertainty around an estimated proportion. The record does not explain this labeling or report the primary result, so the intended summary and its uncertainty cannot be evaluated.

What does a standard deviation describe?

For a continuous measure, the standard deviation describes how dispersed participant-level measurements are around their mean. A larger standard deviation indicates more spread in the observed measurements, while a smaller one indicates more clustering. It does not directly tell the reader how precisely the mean has been estimated. Precision of a mean is generally conveyed by a standard error or confidence interval, interpreted with the analysis method and the amount of available data.

The registry lists symptom scores from the MDASI-BT and MDASI-SP, including means, standard deviations, confidence intervals, and change scores at different time points. It also lists comparisons of symptom burden between responders and non-responders. These are distinct summaries with different questions: a time-point mean describes average reported burden, a change score describes movement from baseline, and a group comparison describes differences between the defined responder groups. The registry record does not report those estimates or the methods used to handle missing questionnaire responses.

What is multiplicity?

Multiplicity arises when several outcomes, time points, groups, or analyses are examined. The more questions considered, the greater the chance that at least one result may appear unusual by chance if each is judged separately. Statistical approaches can control or describe this issue, but the registry record does not report a multiplicity adjustment or explain which outcomes were treated as confirmatory and which as exploratory. The presence of 12 outcome measures and 32 analyses indicates a broad set of planned or listed analyses; it does not by itself show that any particular result is false or valid.

How should planning illustrations be interpreted?

Planning calculations based on an observed effect are illustrations of sample size under stated assumptions, not observed power and not evidence that the study achieved a particular probability of detecting an effect. The assumptions described for the planning illustrations include a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the target effect expressed as a fraction of the observed effect. These are planning assumptions and do not match the registry's single-group design. Any such illustration should therefore be read as a hypothetical design exercise, not as a description of the actual trial or its results.

For a time-to-event endpoint, event-based planning and participant-based planning answer different design questions; the record's stated convention distinguishes total events for time-to-event calculations from participants per arm for other calculations. No sample-size justification for the actual single-group study is reported in the registry record. The observed primary outcome and its uncertainty are also not reported, so the posted information does not permit evaluation of whether the study's planned enrollment was adequate for its stated objective.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Cohort 1- Heavily Pre-Treated", "Cohort 2 - Non-Heavily Pre-Treated")
n   <- c(73, 29)
m   <- c(0.219, 0.379)
sp  <- c(0.048, 0.09)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Cohort 1- Heavily Pre-Treated  73  0.219  0.048  0.208  0.230
CARM  Cohort 2 - Non-Heavily Pre-Treated  29  0.379  0.090  0.345  0.413

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 12 outcome measures and 32 analyses, spanning disease control, overall survival, symptom burden and interference, symptom-score changes, and adverse events. Multiple outcomes, time points, and group comparisons can create multiplicity: when many statistical questions are considered, an isolated result may look noteworthy by chance. This does not mean that every analysis is unreliable, but it makes the distinction between primary, secondary, and exploratory questions important.

The registry identifies disease control as the primary endpoint, but a multiplicity strategy is not reported in the registry record. It is not reported whether the secondary outcomes were adjusted for multiple testing, whether selected outcomes or time points were prioritized, or whether analyses were intended to be confirmatory or exploratory. The number of analyses alone does not reveal the impact of multiplicity on any particular estimate.

The primary population description states that 102/136 participants were analyzed and gives several reasons for non-analysis, including ineligibility, other treatment, and withdrawal or declining participation. Its final reason is incomplete, and a full accounting of participant flow is not reported in the registry record. The record does not say how missing primary endpoint assessments were handled, whether participants with incomplete follow-up were counted as having no disease control, or whether sensitivity analyses were performed.

Missing-data handling is also relevant to symptom questionnaires, repeated imaging, survival follow-up, and adverse-event observation. The registry lists these assessments but does not report completion rates, missingness patterns, imputation methods, or analyses testing how missing observations could affect results. Without those details, the reader cannot determine whether the analyzed observations represent all participants who were intended to contribute to each outcome.

The registry record does not report the statistical analysis plan, interim analyses, or a sample-size justification for the actual trial. These omissions limit assessment of prespecification and the intended handling of multiple outcomes and incomplete data.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: general disorders and administration site conditions - other, specify: disease progression (13.2%), colitis (6.7%), surgical and medical procedures - other, specify: laminectomy (6.7%), thromboembolic event (6.7%), headache (5.3%).

The adverse-event record separates a heavily pre-treated cohort, a non-heavily pre-treated cohort, and participants enrolled but not treated or completing treatment cycles. For serious adverse events, the record gives 43 of 76 participants in the heavily pre-treated cohort, 13 of 30 in the non-heavily pre-treated cohort, and 0 of 30 in the group enrolled but not treated or completing treatment cycles. For other adverse events, it gives 70 of 76, 29 of 30, and 0 of 30, respectively. The record also lists deaths of 42, 10, and 1 across those groups.

These are descriptive counts in the registry's safety groups. They should not be read as a randomized comparison or as proof that nivolumab caused an event or death. The group described as not treated or not completing treatment cycles is not a matched control group, and the record does not provide information that would establish comparable follow-up or risk between groups. The death counts are listed in the safety record; their relationship to treatment is not specified by those counts alone.

Among the listed serious adverse-event terms are disease progression, colitis, laminectomy, thromboembolic event, headache, and hydrocephalus, with reported percentages of 13.2%, 6.7%, 6.7%, 6.7%, 5.3%, and 5.3%, respectively. The registry does not provide enough context alongside this list to infer causation from the term or percentage. Counts and percentages describe reported events; they do not by themselves establish comparative safety or the likelihood of an event for a different patient.

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

The registry record documents a study of nivolumab in adults with recurrent rare CNS cancers and defines disease control as the primary endpoint. It describes response and stable-disease criteria, an assessment period through the end of treatment, and a single-group design. These details clarify what the study intended to measure and the population to which the study was directed.

However, the primary endpoint estimate is not reported in the registry record. There is no reported disease-control proportion or measure of its precision to interpret. The record also does not report the secondary outcome estimates for overall survival, symptom burden, symptom changes, or the comparisons between responders and non-responders. The treatment schedule and study design cannot substitute for outcome results.

Even if an observed disease-control rate were available, a single-group result would describe outcomes among analyzed participants rather than establish what would have happened under a different treatment. The absence of a concurrent control limits causal interpretation. The range of tumor types and the separately described treatment-history cohorts also mean that a pooled outcome could conceal variation, but subgroup outcome estimates are not reported.

The registry status is terminated, but the reason for termination is not reported in the registry record. That status does not, by itself, establish why the study ended or whether termination was related to efficacy, safety, recruitment, or another consideration. The posted safety counts are descriptive and do not prove treatment causation or establish a comparative risk.

Accordingly, the record supports a description of the trial's question, design, endpoint definitions, and listed safety counts. It does not support a conclusion about the observed primary endpoint result, the precision of that result, or a causal effect of nivolumab. Those conclusions would require outcome estimates and fuller analytic information not reported in the registry record.

14. Why This Trial Matters Statistically

Rare CNS cancers pose statistical challenges because the registry describes conditions that affect relatively small populations and groups several tumor types under a shared study framework. Studying them within one protocol can make systematic assessment possible, but the combined result may not represent every tumor type equally. Interpretation depends on knowing which participants contributed to each analysis and whether outcomes differ across clinically relevant groups. The registry does not report primary endpoint results by tumor type or treatment-history cohort.

The endpoint also illustrates how a clinical question becomes a statistical measure. Disease control includes complete response, partial response, and durable stable disease, each defined through imaging criteria and duration. This broad composite can capture more than tumor shrinkage, but its overall result would not reveal which component drove the rate unless component-specific findings were reported. Such component results are not reported in the registry record.

The trial's single-group design is important for biostatistical interpretation. A carefully defined outcome can describe what occurred in the study, but without a concurrent comparison group it cannot isolate the intervention's effect from other explanations. This distinction is especially important when outcomes are interpreted across different tumor types, prior-treatment histories, and follow-up experiences.

Finally, the registry lists a broad set of secondary outcomes and analyses, while the primary endpoint result and methods for multiplicity and missing data are not reported. The record therefore illustrates why transparent reporting requires more than an endpoint definition: readers also need outcome estimates, uncertainty, analysis-population details, and clear methods for handling multiple questions and incomplete observations. Those pieces are necessary to judge the strength and scope of statistical evidence.

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 limitation for interpretation is that the registry record does not report the primary disease-control result or its uncertainty. Despite a detailed endpoint definition, no estimate is available to assess the observed frequency of disease control or the precision of that estimate. Secondary outcome estimates are also not reported in the information summarized by the registry record.

The population accounting is incomplete. The record identifies 102/136 participants as analyzed and gives several reasons for non-analysis, but the final reason is truncated and a full accounting is not reported. It is also not reported how missing or unevaluable primary endpoint assessments were handled. This makes it difficult to assess the relationship between enrollment, treatment exposure, eligibility, follow-up, and analysis inclusion.

The study includes a variety of rare CNS tumors, but the registry does not report primary results by tumor type or explain how any pooled estimate should be interpreted across those conditions. Cohort definitions beyond the labels heavily pre-treated and non-heavily pre-treated are not reported. Consequently, neither the comparability of cohorts nor the generalizability of a pooled result can be evaluated from the record.

The single-group design does not supply a concurrent control for separating treatment-associated outcomes from other influences. The registry also does not report a multiplicity strategy, missing-data methods, interim analyses, or a statistical analysis plan. The stated primary measure is described as a proportion but labeled as a mean with standard deviation, an ambiguity that further limits understanding of the intended statistical summary.

The status is terminated, but the reason for termination is not reported in the registry record. The record does not explain whether planned follow-up or all assessments were completed. These gaps prevent a fuller appraisal of outcome completeness, analysis decisions, and the scope of conclusions that could be drawn.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Immune Checkpoint Inhibitor Nivolumab in People With Recurrent Select Rare CNS Cancers — NCT03173950.
• Registry record last updated 2026-10-27; status last verified 2026-10.
• Results first posted 2026-08-18.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

This registry record describes a Phase II, single-group study of nivolumab for adults with recurrent select rare CNS cancers, sponsored by the National Cancer Institute (NCI). The listed conditions include medulloblastoma, ependymoma, pineal region tumors, choroid plexus tumors, and atypical or malignant meningioma. The record lists planned enrollment of 136, a start date of 2017-07-13, a primary completion date of 2025-06-23, and a status of terminated.

The primary endpoint is disease control, defined as confirmed complete response or partial response, or durable stable disease for 6 months using iRANO criteria. The registry describes the treatment schedule and the intended assessment period, but the primary endpoint result is not reported in the registry record. The analyzed population is described as 102/136, with an incomplete account of reasons for non-analysis. Safety counts are listed for separate cohorts, but they do not constitute a randomized comparison or establish treatment causation.

The record does not report the reason for termination, the primary outcome estimate and its precision, or the methods used for multiplicity and missing data. Its design and endpoint definitions can be described, but efficacy and comparative safety conclusions cannot be drawn from the reported information.