Genetically Modified Neural Stem Cells, Flucytosine, and Leucovorin for Treating Patients With Recurrent High-Grade Gliomas: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Genetically Modified Neural Stem Cells, Flucytosine, and Leucovorin for Treating Patients With Recurrent High-Grade Gliomas (NCT02015819), a na phase 1 study. The primary outcome is Number of Participants With Dose Limiting Toxicities (DLTs) and Maximum Tolerated Dose (MTD).

2. Clinical Question

3. Trial Design

This was a Phase I study sponsored by City of Hope Medical Center. The registry record gives the study as completed, with enrollment of 16. The recorded start date is 2014-10-28 and the primary completion date is 2017-10-07. The study used a single-group model, with allocation listed as NA and masking listed as none. It was not a randomized comparison against a control group.

The intervention was organized into dose levels. The listed treatment included intracranial CD-expressing neural stem cells on days 1 and 15, oral flucytosine every 6 hours on days 4-10 and 18-24, and, for some participants depending on entry timing, oral leucovorin every 6 hours on days 4-10 and 18-24. Courses were described as repeating every 28 day. Dose level 4 also included leucovorin and microdialysis. The registry record lists the neural stem cell and flucytosine doses for each level; it does not establish that the groups should be treated as comparable treatment arms.

The primary endpoint was assessed through Day 28 of course 1 and concerned DLTs and the MTD or MFD. Secondary outcomes included antibodies against neural stem cells, steady-state 5-FC and 5-FU levels in brain, 5-FC concentrations in plasma, a brain-to-plasma comparison of 5-FC, and tumor response using RANO criteria. The registry record does not report numerical findings for these outcomes. It also does not report the protocol’s detailed dose-escalation rules, sample-size justification, interim analyses, or statistical analysis plan.

4. Participant Flow and Accounting

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

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: 1 participant in dose level 4 was unevaluable and had to be replaced after receiving the first dose of CD-NSCs due to the tip of the Rickham catheter migrating into the lateral ventricle.

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 With Dose Limiting Toxicities (DLTs) and Maximum Tolerated Dose (MTD), measured over Day 28 of course 1.

The primary endpoint counted participants with dose-limiting toxicities and addressed determination of the maximum tolerated dose (MTD), with the endpoint description also referring to MFD. The stated assessment period was Day 28 of course 1. The record describes reporting DLT counts per dose level and determining the MTD or MFD, but does not report the observed counts or dose determination.

In a phase I dose-finding study, DLTs are used to characterize treatment-limiting toxicity within a defined assessment period and to inform dose selection. The endpoint is a count of participants with an event, not a continuous measure of symptom severity or a time-to-event outcome. Interpreting such counts requires knowing how DLTs were defined, how participants were assigned to dose levels, and how unevaluable participants were handled. The detailed DLT criteria and dose-selection rules are not reported in the registry record.

The registry record notes that one participant at dose level 4 was unevaluable and had to be replaced after receiving the first dose of CD-NSCs because the tip of the Rickham catheter migrated into the lateral ventricle. That note is relevant to the endpoint population and dose-level accounting, but it does not provide the DLT result for that participant or the other participants. Without the primary endpoint counts and dose determination, the endpoint’s planned role can be explained, but the study’s dose-finding conclusion cannot be read from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record identifies the primary outcome as a count of participants with DLTs and an MTD determination, with assessment through Day 28 of course 1. The outcome type is binary at the participant level: a participant either meets the endpoint definition or does not. Such participant-level results may then be summarized as counts within dose levels. The record does not give the operational definition of a DLT, the observed endpoint counts, or the method used to select the MTD or MFD.

What does a participant count tell the reader?

A count conveys how many participants experienced an event under a specified definition and period. It is useful for describing observed events in an early-phase safety study, especially when the purpose includes dose characterization. A count alone does not express the precision of an underlying event rate. To assess precision, readers would usually need a denominator and an uncertainty interval, as well as a clear account of which participants were included. The registry’s adverse-event summaries include numbers at risk and affected for serious and other events, but the primary DLT counts are not reported.

General statistical explanation: a confidence interval describes a range of values compatible with the observed data under the assumptions of the chosen method. It can show how uncertain an estimate is, rather than converting a small observed count into a definitive risk estimate. No confidence interval for the primary endpoint is reported in the registry record. Neither a point estimate nor an interval should be reconstructed from the available summaries when the endpoint definition and analysis details are absent.

Why does the dose-level structure matter?

Dose-level summaries are descriptive groupings, not necessarily treatment comparisons. This study is listed as single group, with no randomized allocation and no masking. The dose levels differ in the listed neural stem cell doses, and dose level 4 also includes leucovorin and microdialysis. As a result, apparent differences across levels cannot be assigned to a single component or interpreted as if participants had been randomized between otherwise equivalent groups. The registry record does not provide the dose-escalation algorithm or explain the exact decision rule used to move between levels.

General statistical explanation: dose-finding decisions depend on both observed toxicity and the prespecified rules for evaluating each dose. Those rules define what counts as a DLT, which participants contribute to the decision, and how uncertainty is balanced against the aim of selecting a dose. Without the protocol details and endpoint results, readers cannot reconstruct those decisions from adverse-event summaries alone.

How should precision and planning be distinguished?

The site’s sample-size material is described as a planning illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. For non-time-to-event outcomes, such planning values are expressed as participants per arm for 80% power or other stated power targets. Such values are hypothetical planning illustrations, not observed power and not evidence that the single-group study had a randomized two-arm design. The registry record does not report a sample-size justification for this trial’s phase I dose-finding purpose.

General statistical explanation: power calculations describe the sample size that a hypothetical study might need for a specified effect and assumptions. They do not measure the probability that an observed result is true, and they do not show whether an already completed study detected a real effect. If the illustration uses a two-arm comparison while the actual trial has one group and dose levels, its assumptions do not describe the trial’s design. No recalculated treatment effect or comparison is available for the primary endpoint in the registry record.

What can adverse-event summaries establish?

The serious and other adverse-event entries give affected-participant counts within each dose-level group. The serious-event terms are also presented with percentages. Such summaries can describe what was recorded, but do not by themselves establish causality, severity over time, or a dose-response pattern. General statistical explanation: sparse counts tend to carry substantial uncertainty, and uneven group sizes complicate direct comparisons. The registry record does not report adjusted comparisons, uncertainty intervals for these safety figures, or a causal analysis.

The record lists one analysis, but does not give enough detail to identify a statistical model, test, adjustment strategy, or analysis population for the primary endpoint. There is no reported inferential result such as a p-value for that endpoint. In a single-group phase I study, the principal interpretive emphasis is on the defined safety outcomes and dose-selection process; absent the outcome counts and decision rules, neither can be evaluated from 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("Dose Level 1: (NSC 5x10^7 and 5-FC 37.5 mg/kg)", "Dose Level 2: (NSC 1x10^8 and 5-FC 37.5 mg/kg)", "Dose Level 3 (NSC 1.5x10^8 and 5-FC 37.5 mg/kg)", "Dose Level 4 (NSC 1.5x10^8 and 5-FC 37.5 mg/kg + L")
n      <- c(3, 3, 3, 6)
events <- c(0, 0, 0, 1)
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  Dose Level 1: (NSC 5x10^7 and 5-FC 37.5 mg/kg)  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Dose Level 2: (NSC 1x10^8 and 5-FC 37.5 mg/kg)  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Dose Level 3 (NSC 1.5x10^8 and 5-FC 37.5 mg/kg)  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Dose Level 4 (NSC 1.5x10^8 and 5-FC 37.5 mg/kg + L  6  1  16.7  0.4  64.1  3.0  56.4  3.7  57.9

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists several secondary outcomes, including immune response against neural stem cells, drug levels in brain and plasma, their comparison, and tumor response using RANO criteria. It does not report a multiplicity adjustment or explain whether these outcomes were treated as confirmatory, exploratory, or descriptive. When multiple outcomes are examined, the chance of obtaining at least one apparently notable result can be affected by how many analyses are conducted and how results are selected. General statistical explanation: multiplicity procedures can control the error rate across a family of tests, but their use and interpretation depend on a prespecified analysis plan. No such plan is reported in the registry record.

The record does not describe how missing outcome data were handled, define analysis populations, or report the number of participants included in each primary or secondary analysis. It notes that one participant at dose level 4 was unevaluable and replaced after the first dose because the Rickham catheter tip migrated into the lateral ventricle. The record does not state how this replacement affected the primary endpoint analysis or provide a broader account of exclusions or missingness.

The primary endpoint’s assessment period is specified as Day 28 of course 1, but details about follow-up completion and endpoint ascertainment are not reported in the registry record. It also does not describe interim analyses, sensitivity analyses, or a plan for participants who did not complete the assessment period. These omissions limit the ability to judge whether the analyzed participants were representative of everyone enrolled, how robust the findings were to missing data, or how results across the many outcomes should be interpreted. Because the primary outcome findings themselves are not reported, these issues cannot be resolved through a comparison of planned and observed analyses.

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 (33.3%), skin infection (33.3%), death (33.3%), muscle weakness left-sided (33.3%), hypertension (14.3%).

The serious adverse-event summary reports affected participants and participants at risk by dose level: 2 among 3 at dose level 1, 1 among 3 at dose level 2, 1 among 3 at dose level 3, and 4 among 7 at dose level 4. For other adverse events, the listed counts are 3 among 3 at each of the first three levels and 7 among 7 at dose level 4. These are descriptive participant counts, not counts of all individual events, and they do not establish that the treatment caused the events.

The record separately lists deaths by group and gives percentages for selected serious-event terms, including seizure, skin infection, death, muscle weakness left-sided, hypertension, and brain abscess. The displayed percentages should not be treated as comparative treatment effects. The registry record does not report causal attribution for each event, detailed event timing, or a complete account of how events were collected and classified. The presence of deaths in the safety record is important to recognize, but the available summaries do not explain their causes or relationship to study treatment.

Small, unequal groups make descriptive comparisons especially limited. The dose levels were not randomized control groups, and dose level 4 included additional components. The serious-event counts therefore should not be read as proof that one level is safer or less safe than another.

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

The record documents a completed Phase I, single-group study of CD-expressing neural stem cells with oral flucytosine, with leucovorin included for some participants. It specifies a primary safety and dose-finding endpoint assessed through Day 28 of course 1, and it gives selected adverse-event summaries. These details describe the study question and some recorded safety information; they do not answer the primary question because the DLT counts and MTD or MFD determination are not reported.

The intervention rationale is described in the registry: CD-expressing neural stem cells were intended to convert flucytosine into 5-fluorouracil near tumor tissue, and leucovorin was described as a possible aid to 5-FU activity. A biological rationale is not a clinical result. The registry record does not report primary results establishing the extent of drug conversion in participants, the effect on tumors, or a treatment effect compared with another strategy. Secondary outcomes concerning drug levels and tumor response are listed, but their findings are not reported in the record.

There is no randomized comparator, and dose levels are not a like-for-like control comparison. Consequently, the listed group counts cannot establish comparative safety or effectiveness. The safety summaries also do not identify whether each event was judged related to the intervention. Any interpretation should remain descriptive and bounded by the information recorded.

General statistical explanation: a completed study can still have results that are insufficiently reported for readers to evaluate its main question. Enrollment and completion status describe study conduct, not the magnitude or precision of an endpoint result. For this trial, the registry record supports understanding the intended design and the posted safety counts, but does not support a conclusion about which dose was selected or whether the intervention changed clinical outcomes.

14. Why This Trial Matters Statistically

This study illustrates a key statistical distinction in early-phase research: a registered dose-finding question is not answered merely by knowing that a study was completed or by reviewing general adverse-event counts. The primary endpoint was specifically tied to DLTs during Day 28 of course 1 and to determining an MTD or MFD. Without the DLT results and the dose-selection rule, the central statistical purpose cannot be assessed.

The trial also highlights why design structure matters when interpreting group summaries. It is listed as a single-group study, with no randomization or masking. The dose levels contain different listed treatment components, and dose level 4 includes leucovorin and microdialysis. Counts across these groups are useful as descriptions of recorded events, but they do not provide randomized comparisons or isolate the effect of one treatment component.

The reported unevaluable participant illustrates a further practical issue: analysis depends on how participants and outcomes are classified. The catheter migration note explains why one participant at dose level 4 was replaced, but the record does not state how this affected the primary analysis. Clear reporting of evaluability, endpoint definitions, and analysis populations is important for interpreting a phase I dose decision.

Finally, this record distinguishes a mechanistic rationale from statistical evidence of clinical outcome. It describes the intended conversion of flucytosine to 5-FU by the modified cells and lists pharmacologic and tumor-response outcomes. Yet their findings are not reported. For statistical readers, the main lesson is to separate the proposed mechanism, the study design, the posted descriptive safety data, and the unreported primary result.

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 for interpreting the trial record is the absence of primary endpoint results. The record does not give the DLT counts by dose level or state which dose, if any, was determined to be the MTD or MFD. It therefore does not allow an assessment of the main registered dose-finding objective.

The design is single group, with allocation listed as NA and masking as none. There is no randomized control group against which to interpret safety or tumor response. The dose levels are not interchangeable comparators: the listed doses differ, and dose level 4 includes leucovorin and microdialysis. Thus, differences in descriptive event counts cannot be attributed to a specific dose or component.

Other limitations in the record include the absence of detailed DLT criteria, dose-escalation rules, a sample-size justification, and a statistical analysis plan. The record does not report analysis populations, missing-data methods, uncertainty intervals, or multiplicity handling for the listed secondary outcomes. It notes one unevaluable participant at dose level 4 who was replaced after catheter migration, but does not describe the effect of that replacement on analysis.

Adverse-event figures are limited to selected group summaries and serious-event terms. The record does not report event-level causal attribution, timing, or enough detail to interpret the listed deaths in relation to treatment. The secondary outcomes are named, but their results are not reported. These gaps constrain interpretation to the study design and the particular descriptive safety figures included in the registry record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Genetically Modified Neural Stem Cells, Flucytosine, and Leucovorin for Treating Patients With Recurrent High-Grade Gliomas — NCT02015819.
• Registry record last updated 2026-11-03; status last verified 2026-10.
• Results first posted 2021-09-29.
• 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

NCT02015819 is a completed Phase I, single-group study sponsored by City of Hope Medical Center, with enrollment of 16. The intervention combined intracranial CD-expressing neural stem cells and oral flucytosine, with leucovorin used for some participants. The primary endpoint concerned participants with DLTs and determination of the MTD or MFD through Day 28 of course 1.

The registry record lists dose-level safety summaries and secondary outcomes concerning antibodies, drug concentrations, and tumor response. It notes one unevaluable participant at dose level 4 who was replaced after catheter migration. The primary DLT counts, dose determination, secondary outcome findings, and detailed statistical analysis plan are not reported in the registry record. Safety counts are descriptive and do not constitute randomized comparisons.