CAR T Cell Receptor Immunotherapy Targeting EGFRvIII for Patients With Malignant Gliomas Expressing EGFRvIII: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for CAR T Cell Receptor Immunotherapy Targeting EGFRvIII for Patients With Malignant Gliomas Expressing EGFRvIII (NCT01454596), a non-randomized phase 1, phase 2 study. The primary outcome is Number of Treatment Related Adverse Events.

2. Clinical Question

The study asks whether anti-EGFRvIII chimeric antigen receptor (CAR) T cells can be administered to patients with malignant gliomas expressing EGFRvIII, and what dose can be given safely. The registry describes a therapy made by collecting a patient’s white blood cells, modifying them to express a CAR targeting EGFRvIII, and returning the cells to the patient. The treatment regimen also includes cyclophosphamide and fludarabine before cell infusion, and aldesleukin.

The registry also lists objective response, circulating CAR-positive cells in peripheral blood at 1 month post-treatment, and serious and non-serious adverse events among the secondary outcomes. Their numerical results are not reported in the registry record. Thus, the clinical question encompasses both safety and feasibility and possible antitumor activity, but the posted information does not provide a numerical estimate of response or a basis for comparing treatment with an untreated or alternative-treatment group.

3. Trial Design

This is a completed Phase I/II study with non-randomized allocation, no masking, and a sequential model. The registry lists enrollment of 18. The study began on 2012-05-16, and primary completion was on 2018-11-01. The record identifies the National Cancer Institute (NCI) as sponsor.

The Phase I arm was designed around escalating doses of anti-EGFRvIII CAR-transduced peripheral blood lymphocytes (PBL), given after a non-myeloablative lymphodepleting preparative regimen of cyclophosphamide and fludarabine, with aldesleukin. The Phase II arm used the maximum tolerated dose established in Phase I with the same accompanying regimen. This sequence makes the dose-finding stage part of the study’s design; the Phase II regimen is not an independent randomized comparison.

The adverse-event results list cohorts by dose and by steroid status. Labels include Group A (Steroids), Group B (No Steroids), and combined steroids/no steroids cohorts. The registry’s overall design classification remains non-randomized. Differences between cohort labels therefore should not be interpreted as treatment effects of steroids or dose: the record does not describe random assignment or a controlled comparison. Cohort membership, timing, and other characteristics that would support such a comparison are not reported in the registry record.

The registry describes eligibility as adults age 18-70 with malignant glioma expressing EGFRvIII. Its brief summary outlines screening, leukapheresis to collect white blood cells, cell preparation, and return of the modified cells. Further details about the conduct of treatment are not fully available in the registry record. In particular, a statistical analysis plan, sample-size justification, interim analyses, and details of analysis populations are not reported in the registry record. The study status is completed, but completion status alone does not establish that every planned outcome has a posted result.

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 patients were treated on the Ph I portion of the study. After encountering a patient mortality at dose level 8 and a grade 3 pulmonary toxicity at dose level 9 without seeing any clinical responses, the principal investigator and senior staff re-evaluated the protocol and elected to close the protocol rather than proceed into the Ph II portion.

Only the first 6 of 12 groups are shown.

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.

Only the first 6 of 13 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Number of Treatment Related Adverse Events, measured over From 4 weeks after cell infusion up to 77 days.

The primary endpoint is the number of treatment-related adverse events that were Grade 3 or higher. The registry defines the measure as the aggregate of adverse events judged possibly, probably, or definitely related to treatment. Events were assessed using the Common Terminology Criteria in Adverse Events (CTCAE v4.0). The observation period is from 4 weeks after cell infusion up to 77 days.

This is a count of adverse events, not a rate or a percentage of participants. A participant could potentially experience more than one event, so an event count does not automatically tell the reader how many people experienced an event. The registry’s endpoint description does not specify whether the posted count, if available, would also be accompanied by participant-level results. The numerical primary endpoint result is not reported in the registry record.

In general, a severe-event count is used to characterize the burden of important harms during a defined observation period. The CTCAE grades help distinguish levels of severity: Grade 3 events are severe, Grade 4 events are life threatening, and Grade 5 is death. The endpoint’s relatedness categories indicate how investigators attributed events to treatment; they do not by themselves establish that treatment caused an event. Because this is a safety-focused count, it does not measure tumor response or duration of disease control. Objective response is listed separately as a secondary outcome, but its numerical result is not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The study’s design is non-randomized and sequential, with Phase I dose escalation followed by a Phase II arm using the maximum tolerated dose established in Phase I. That structure is important for interpreting the numbers: observations from successive dose cohorts are not equivalent to observations from groups assigned at random to competing treatments. The record lists no masking. It does not report a detailed statistical analysis plan, so the methods below explain general statistical principles rather than asserting which procedures investigators used.

What does a count endpoint represent?

In general, a count endpoint records how many events occurred, rather than how many participants experienced at least one event. When one participant can have multiple events, event counts and participant counts answer different questions. A count can describe the total recorded burden, while a participant-level proportion describes how widespread an event was among people at risk. The primary endpoint is specified as an aggregate of adverse events meeting its severity and relatedness criteria. The numerical primary result is not reported in the registry record, and it is not reported whether the endpoint was additionally summarized at participant level.

Count data can have a distribution unlike continuous measurements. In general, a count model may account for the expected event count and, where appropriate, the time each participant was observed. Over-dispersion means the observed variation in counts is greater than a simple model would expect; ignoring it can make uncertainty estimates too small. These are general considerations, not methods documented for this study. The registry record does not report a count model, an exposure-time adjustment, or a method for handling over-dispersion.

How should cohort counts be compared?

In general, a comparative estimate needs a defined contrast, a suitable denominator, and an approach that accounts for how participants entered each group. Here, the listed dose and steroid-status cohorts are not randomized arms. A difference between their observed counts could reflect factors other than dose or steroid status, including the sequential enrollment and dose-escalation structure. The registry does not report an adjusted comparison or an analysis designed to address these possible differences. The cohort counts are therefore best treated descriptively rather than as estimates of causal effects.

Randomization can, in general, help make groups comparable in expectation, while masking can reduce some forms of assessment bias. Neither feature is listed for this study: allocation is non-randomized and masking is none. Those design features do not make the observed data unusable, but they limit the interpretation of comparisons. The registry does not report any alternative method that would create a randomized contrast.

What does a confidence interval add?

In general, a confidence interval expresses the range of values reasonably compatible with data under a specified statistical method. It helps distinguish an estimate’s size from its precision. A narrow interval indicates greater precision under the model; a wide interval indicates more uncertainty. A confidence interval would not remove bias from non-randomized cohort comparisons or resolve uncertainty about outcome definitions. No confidence interval for the primary endpoint or a cohort comparison is reported in the registry record.

A p-value, in general, assesses how compatible data are with a specified null hypothesis under assumptions of the analysis. It is not a measure of the size or clinical importance of an effect. The registry record does not report p-values for the outcomes, and no hypothesis test should be inferred from the cohort listing alone.

How should the planning illustrations be understood?

The planning illustrations use a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect defined as a stated fraction of the observed effect. For a time-to-event setting, the stated convention is a total event count; for other settings, it is participants per arm. These are planning assumptions, not features of the trial’s actual allocation. In particular, equal-allocation planning does not turn this non-randomized sequential study into an equally allocated randomized comparison.

In general, a planning sample size estimates how many participants an imagined future study might need under selected assumptions. It is not “observed power” and does not quantify the probability that the completed study found a true effect. Planning values also depend on the effect chosen for detection and on the model assumptions. The registry record does not report the study’s own sample-size justification. The numerical results for the planning illustrations are not described in the record summary available for interpretation.

Finally, statistical interpretation depends on specifying the analysis population, outcome window, event definitions, handling of missing observations, and any multiplicity adjustments. These details are not reported in the registry record. Without them and without numerical outcome estimates, readers can describe the design and endpoint definitions but cannot reconstruct a complete inferential analysis.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists several outcomes: the primary count of treatment-related adverse events, objective response, circulating CAR-positive cells in peripheral blood at 1 month post-treatment, and serious and non-serious adverse events. Considering multiple outcomes can raise multiplicity concerns. In general, when several statistical tests are performed, the chance of at least one apparently unusual result may increase unless the analysis plan addresses the issue. The registry record does not report whether any multiplicity adjustment or prespecified hierarchy among outcomes was used.

Numerical results for the listed secondary outcomes are not reported in the registry record. The primary endpoint’s numerical result is also not reported. No inferential analysis or p-values are reported, and the record does not provide enough information to determine whether outcome analyses were performed but not posted, were not performed, or were handled in another way.

Missing-data methods are not reported in the registry record. The available information does not describe whether participants who did not complete follow-up were included, how incomplete outcome observations were handled, or whether any imputation or sensitivity analysis was used. The analysis populations for the primary and secondary outcomes are also not reported. Enrollment is listed as 18, while the adverse-event listing gives cohort-specific at-risk counts; the record does not explain how these counts relate to each outcome’s analysis population.

These omissions matter because outcome availability and analysis choices can affect both the estimated result and its uncertainty. They do not establish that data were missing or that a particular method was inappropriate. They mean the registry record does not provide the information needed to assess those questions.

11. 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: death not associated with ctcae term: multi-organ failure (100.0%), dyspnea (shortness of breath) (100.0%), hypoxia (100.0%).

The adverse-event listing is organized by dose and steroid-status cohort. For cohorts 1 through 7, the serious-event entries show 0 affected, with 1 or 3 participants at risk as listed for each cohort. Cohort 8, labeled combined steroids/no steroids at 3-6x10(10), shows 1 serious adverse event among 1 at risk and 1 death. Cohort 9, labeled combined steroids/no steroids at 3x10(10), shows 1 serious adverse event among 1 at risk and no deaths. These are cohort-level entries and should not be combined into an overall rate without further information.

The “other” adverse-event entries show an affected count matching the at-risk count for every listed cohort. Those events are not necessarily serious and should not be confused with the primary endpoint, which is specifically an aggregate count of treatment-related events Grade 3 or higher. The adverse-event listing also names “Death not associated with CTCAE term: Multi-organ failure,” “Dyspnea (shortness of breath),” and “Hypoxia,” each with a displayed value of 100.0%. The denominator for those term percentages is not reported in the registry record, so their scope should not be assumed.

These figures are descriptive. They do not establish that the treatment caused an event, and they do not supply the missing numerical result for the primary endpoint.

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

The record describes a sequential, non-randomized Phase I/II study testing the safety and feasibility of anti-EGFRvIII CAR-transduced cells in adults with malignant glioma expressing EGFRvIII. It defines a primary safety outcome based on treatment-related adverse events of Grade 3 or higher and specifies the period over which that outcome was assessed. It also describes the Phase I dose-escalation approach and the Phase II use of the maximum tolerated dose established in Phase I.

The record does not report the numerical primary endpoint result or numerical results for the listed secondary outcomes. Therefore, it does not provide a numerical estimate of the frequency of the primary outcome, a measure of uncertainty, or an estimate of response. The listed adverse-event cohort entries allow a limited descriptive account of the serious-event and other-event counts, but they are not a substitute for the primary endpoint result and do not provide a randomized safety comparison.

Because allocation was non-randomized and the model was sequential, differences among dose or steroid-status cohorts cannot be interpreted as the causal effect of dose or steroid use. There is no randomized comparator described. The posted information also does not provide an adjusted analysis, analysis populations, missing-data procedures, or multiplicity methods. These limits prevent a complete assessment of comparative effects and statistical precision.

The absence of numerical results in the registry record should not be interpreted as a result of zero events, no response, or no treatment effect. It means the record does not give the numerical information needed to evaluate those outcomes. Likewise, a completed study status indicates that the trial is recorded as completed; it does not fill in outcome values that are not reported.

In general, a safety-focused early-phase study can describe tolerability and inform subsequent study planning, while an uncontrolled design cannot by itself establish comparative benefit. For this record, interpretation should remain anchored to the outcomes and cohort counts actually reported, without extending them to conclusions about efficacy or comparative safety.

13. Why This Trial Matters Statistically

This trial illustrates several statistical issues that arise when early-phase studies combine dose escalation with later evaluation. The sequential structure is part of the clinical design: Phase I escalated cell doses, and Phase II used the maximum tolerated dose established in Phase I. That structure is useful context for understanding why dose cohorts should not be treated as if they were randomized treatment groups.

The primary outcome is a count of severe treatment-related adverse events, not a simple yes-or-no participant outcome. Keeping event counts distinct from the number of participants affected is important because one person may experience more than one event. The registry’s adverse-event listing separately reports counts at risk and affected, but those cohort-level entries do not supply the missing numerical primary endpoint result.

The record also shows why outcome definitions and observation periods matter. The primary endpoint requires Grade 3 or higher severity, a specified assessment of treatment relatedness, and a defined post-infusion window. Other adverse events, serious adverse events, and deaths are related but not interchangeable measures. Careful interpretation depends on respecting those distinctions rather than treating all safety entries as one outcome.

Finally, the study is non-randomized and unmasked, and the registry does not report a statistical analysis plan or numerical efficacy results. Those features illustrate the difference between describing clinical observations and estimating a comparative treatment effect. A reader can identify what was intended to be measured and how cohorts were structured, while recognizing that the posted record does not support a precise estimate of benefit or a controlled comparison of safety.

14. 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.

A central limitation is that the numerical primary endpoint result is not reported in the registry record. Numerical results for objective response, circulating CAR-positive cells, and the listed serious and non-serious adverse events are also not reported as outcome summaries. The cohort-level adverse-event entries do not resolve these omissions because they are not identified as the primary endpoint analysis.

The study is non-randomized, unmasked, and sequential. The dose and steroid-status cohorts therefore do not constitute randomized comparison groups. The record does not report enough detail to assess whether cohort differences in patient characteristics, timing, or other factors were addressed analytically. No adjusted comparison is reported.

The primary endpoint concerns events during a specified interval after cell infusion, while cohort adverse-event listings may have different scope or criteria; their equivalence is not reported. The endpoint is described as an aggregate event count, but the record does not explain whether recurrent events in one participant were counted separately in the primary result or how the event count would be summarized with participant-level outcomes.

Analysis populations, missing-data handling, multiplicity procedures, and a detailed statistical analysis plan are not reported in the registry record. The record also does not report a sample-size justification, interim analyses, or uncertainty estimates for the outcomes. These limitations restrict assessment of precision, completeness, and inferential methods. They do not, on their own, show that the trial’s conduct or analysis was deficient; they define what cannot be evaluated from the registry record.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: CAR T Cell Receptor Immunotherapy Targeting EGFRvIII for Patients With Malignant Gliomas Expressing EGFRvIII — NCT01454596.
• Registry record last updated 2019-08-21; status last verified 2019-08.
• Results first posted 2019-08-21.
• 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.

18. Record Summary

The registry describes a completed, non-randomized, unmasked, sequential Phase I/II study of anti-EGFRvIII CAR-transduced PBL for malignant gliomas expressing EGFRvIII. Enrollment is listed as 18. Phase I used escalating cell doses with cyclophosphamide, fludarabine, and aldesleukin; Phase II used the maximum tolerated dose established in Phase I with the accompanying regimen.

The primary endpoint is the aggregate number of treatment-related Grade 3 or higher adverse events from 4 weeks after cell infusion up to 77 days. Its numerical result is not reported in the registry record. Secondary outcomes include objective response, circulating CAR-positive cells at 1 month post-treatment, and serious and non-serious adverse events. The registry provides cohort-level adverse-event entries, but these do not create a randomized comparison or replace the primary endpoint result. Numerical analyses, uncertainty estimates, analysis populations, and missing-data methods are not reported in the registry record.