DC Migration Study for Newly-Diagnosed GBM: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for DC Migration Study for Newly-Diagnosed GBM (NCT02366728), a randomized phase 2 study. The primary outcome is Median Overall Survival.

2. Clinical Question

3. Trial Design

This was a randomized phase II study with a parallel-group model. The registry lists the allocation as randomized and masking as triple. The study summary says that assignment to Groups I and II was double-blinded; it does not explain how the triple-masking designation relates to that description or identify which roles were masked. Randomization was described as a permuted-block algorithm with a 1:1:1 allocation ratio, stratified by cytomegalovirus status. The planned enrollment was 64.

Participants were assigned after completing standard radiotherapy with concurrent temozolomide. Group I received unpulsed autologous dendritic cells on one side of the groin and saline on the other, before the fourth cytomegalovirus-specific dendritic-cell vaccine. Group II received Td pre-conditioning in place of the unpulsed dendritic cells, with saline on the contralateral side. Group III received basiliximab infusions before the first and second vaccines, and Td pre-conditioning before the fourth vaccine. The fourth vaccine was labeled with indium-111 for migration assessment.

The registry says randomization to Group III was terminated effective March 2017. It does not state in the record why that randomization ended or provide further details about how the change affected enrollment or analysis. For the posted overall-survival outcome, the reported group sizes are 25 for Group I, 27 for Group II, and 8 for Group III. These are the outcome group counts, not a description of who was eligible, randomized, treated, or included in each analysis; those analysis-population details are not reported in the registry record.

The study status is completed. The record gives a start date of 2015-10-12 and primary completion date of 2020-10-31. Its primary outcome timeframe is described as up to 72 months, from randomization of the first patient until approximately 31 months after randomization of the last patient. The record does not give a complete account of follow-up duration for each participant.

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:

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: Participants who received at least one ppp65 vaccine.

6. Primary Endpoint

The registry lists the primary outcome as Median Overall Survival, measured over Up to 72 months (from the time of randomization of the first patient until approximately 31 months after randomization of the last patient).

The primary endpoint is median overall survival in months. Overall survival is defined in the registry as the time from randomization until death, or until last follow-up for a participant who is alive. Kaplan-Meier methods were specified for estimating overall survival. The reported outcome is a median with a 95% confidence interval for each group.

In general, the median survival is the time by which the estimated survival curve reaches the point where half of the group is estimated to have experienced the event. The Kaplan-Meier approach uses participants’ observed follow-up times while accounting for censoring: for example, a participant who is alive at last follow-up contributes information through that time, even though their eventual survival time is unknown. Censoring does not mean that the participant experienced the event at the time follow-up ended.

Overall survival is a time-to-event endpoint. It incorporates the timing of deaths and follow-up, rather than treating each participant’s outcome as a simple yes-or-no status at one fixed date. The confidence interval expresses uncertainty around the estimated median. A wider interval indicates less precision in locating the median, while an interval’s width can also reflect the amount and pattern of follow-up information. The record reports medians of 16.0 months for Group I, 20.0 months for Group II, and 19.0 months for Group III. Group III’s reported interval has a lower bound of 10.26 months, but no upper bound is reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is time to death, with time measured from randomization. For participants alive at last follow-up, the endpoint definition uses that last-follow-up time. This creates a time-to-event dataset in which some participants have an observed event time and others are censored. In general, censoring records that the event had not been observed by the participant’s last known follow-up; it does not imply that the participant remained event-free indefinitely.

How does Kaplan-Meier estimation work?

Kaplan-Meier estimation constructs a survival curve from the sequence of observed event times. At each event time, the estimated probability of remaining alive is updated using the participants still under observation immediately before that time. Censored participants contribute information until censoring and then leave the later risk sets. This method is useful when follow-up times differ and some participants’ event times are not observed, as is common in survival studies. Its interpretation depends on the adequacy of the observed follow-up and the assumptions about censoring; the registry does not describe diagnostic checks or further implementation details.

The median is taken from the estimated survival curve at the point where the estimated probability of survival reaches one half. In general, a median is a useful summary because it is expressed in the same time units as follow-up and does not require a constant event rate over time. It nevertheless compresses a whole survival curve into one number, so two groups with similar medians could have different patterns at other times. Conversely, different median estimates alone do not show how the full curves differ.

What does the confidence interval convey?

A confidence interval represents the uncertainty associated with an estimate under the method used to construct it. The record gives 95% confidence intervals for the arm-specific medians in months. Group I’s interval is 12.8 to 25.5 months, and Group II’s is 16.7 to 25.6 months. Group III has a lower bound of 10.26 months, with no upper bound reported in the registry record. The available endpoints should not be treated as a complete interval for Group III. The registry does not describe the interval-construction method or explain why the upper bound is absent.

Confidence intervals for separate group medians are not the same as a confidence interval for a contrast between groups. A comparison estimate would quantify a contrast directly and would have its own uncertainty. The registry reports no such contrast interval for these overall-survival comparisons. It also reports no hazard ratio. Thus, the arm-specific intervals cannot be used as a substitute for an unreported comparison-specific interval.

What does the log-rank test assess?

The log-rank test is a general method for comparing time-to-event experience between groups. It compares observed event counts with the counts expected under a null hypothesis of no difference in survival experience, using information across event times. Unlike a simple comparison of medians, it uses the timing of events throughout follow-up. The registry identifies log-rank as the method for the Group I versus Group II and Group I versus Group III comparisons. It reports p-values of 0.072 and 0.089, respectively.

A p-value is a measure of how unusual the data, or data more extreme, would be under the null hypothesis and test assumptions. It is not a measure of the probability that a treatment strategy is effective, and it does not give the magnitude of a difference. Interpretation also depends on the analysis plan and any multiplicity adjustments. Those details are not reported in the registry record. The p-values should therefore be read as the reported results of the stated log-rank comparisons, not as a complete account of the relative survival effects.

Why does the size and follow-up of each group matter?

Survival estimates rely on participants remaining under observation and at risk as time passes. Fewer participants generally mean less information, particularly at later times, and may lead to less precise estimates. The posted overall-survival group counts are 25, 27, and 8. The comparatively small Group III count is relevant when judging its median and its incomplete confidence interval, but the registry does not provide enough detail to quantify the influence of the count on precision. Censoring patterns and numbers at risk over time are also not reported in the registry record for this outcome.

In general, a treatment effect on survival may not be constant over time. The log-rank test and a hazard ratio have different interpretations and assumptions; in particular, a single hazard ratio can be difficult to interpret if relative hazards change over time. No hazard ratios are reported for this endpoint, and no assessment of proportional hazards is reported. The posted medians and log-rank p-values should therefore be understood according to their own definitions rather than as interchangeable effect measures.

What is not documented?

The registry record does not report a sample-size justification, a detailed statistical analysis plan, the handling of missing outcome data, or the populations used for the primary analysis. It does not state whether interim analyses were planned or conducted. The record also does not explain the method for constructing the confidence intervals or provide a comparison-specific effect estimate. Without these details, readers cannot determine the prespecified analysis hierarchy or assess whether the posted analysis followed a fuller protocol plan. These items are not evidence that a particular method was or was not used; they are simply not reported in the registry record.

In general, interpretation of a randomized trial’s estimates also considers randomization, adherence, follow-up, and the analysis population. Random assignment is described in the record, along with stratification by cytomegalovirus status and the termination of randomization to Group III effective March 2017. The consequences of that change, the balance of baseline characteristics, treatment received, and participant flow are not described in the registry record for these results. The posted survival statistics can be interpreted as registry-reported estimates and tests, while these unanswered design and analysis questions remain open.

10. Analyses Reported by the Sponsor

The sponsor-reported overall-survival comparisons use the log-rank method. The reported p-value is 0.072 for Group I versus Group II and 0.089 for Group I versus Group III. These values are the sponsor’s stated results for those comparisons. The registry does not report a hazard ratio, confidence interval for either comparison, or a recalculated comparison estimate for overall survival.

The record includes arm-specific Kaplan-Meier median estimates and confidence intervals, but the comparison results are not presented as differences in medians. A recalculation from summary medians alone would not reproduce a log-rank test, because the test depends on event timing and risk sets over follow-up. The record’s note that values are missing or not estimable for a two-arm comparison for a median-time endpoint is consistent with not presenting a recalculated two-group effect measure from those summaries.

There is no stated discrepancy between the sponsor-reported and recalculated results in the registry record. In any case, the posted log-rank p-values should be understood as the sponsor-reported comparisons, not as evidence of an unreported hazard ratio or effect size. The registry does not report further information about the analysis plan or adjustment for multiple comparisons.

11. Multiplicity, Missing Data and Analysis Populations

The record reports two log-rank comparisons for overall survival: Group I versus Group II and Group I versus Group III. It does not report whether these were the only planned comparisons, whether they were prespecified as primary or secondary contrasts, or whether the p-values were adjusted for multiple testing. It also does not describe a testing hierarchy or identify a single group comparison as the sole confirmatory test. Because these details are not reported in the registry record, the two p-values should be read as the results of the stated comparisons without assuming a particular multiplicity strategy.

The primary endpoint definition says overall survival is measured from randomization until death, or last follow-up if alive, and specifies Kaplan-Meier estimation. This identifies how event time is defined but does not describe how missing or incomplete follow-up was handled beyond the definition of last follow-up. The registry does not report the number of participants with missing outcome information, reasons for loss to follow-up, or sensitivity analyses for missingness. It also does not report whether censoring assumptions were assessed.

The record does not state whether the primary analysis used an intention-to-treat population, a treated population, or another analysis set. It does not provide a participant disposition account or clarify how the analysis population relates to the assigned, treated, and followed participants. The overall-survival outcome groups contain 25, 27, and 8 participants, while the planned enrollment was 64; the registry does not explain the relationship between these counts or any exclusions.

Randomization to Group III was terminated effective March 2017, but the record does not give the reason, the timing relative to analysis decisions, or the impact on the statistical plan. It is not reported whether interim analyses were planned or conducted, whether any stopping rule was used, or whether the analysis plan changed. These are uncertainties about what is documented, not proof that specific procedures were absent. They matter because multiplicity, missing data, and analysis populations affect the interpretation of reported statistical evidence.

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: colonic perforation (11.1%), generalized muscle weakness (11.1%), muscle weakness left-sided (11.1%), pyramidal tract syndrome (11.1%), seizure (7.1%).

The registry reports serious adverse event counts by group: 1 of 27 at risk in Group I, 4 of 28 in Group II, and 2 of 9 in Group III were recorded as affected. It also reports other adverse event counts of 25 of 27, 27 of 28, and 9 of 9, respectively. Death counts are listed as 24, 22, and 6. These are the registry’s event counts and denominators; the record does not provide further detail about event timing, attribution, duration, or severity grading.

The listed serious adverse event terms include colonic perforation, generalized muscle weakness, muscle weakness left-sided, pyramidal tract syndrome, seizure, and lung infection, with percentages reported for each term. The record does not assign these listed term percentages to particular trial groups in the registry record for the outcome, so they should not be interpreted as group-specific estimates. Nor does the listing establish whether an event was caused by a vaccine or another intervention.

The safety denominators differ from the posted overall-survival outcome group counts. The registry does not explain the differing counts or how safety populations were defined. These event figures are descriptive registry records, not adjusted comparisons of risk between groups. They do not on their own establish whether one strategy is safer or less safe than another.

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

The record describes a randomized phase II comparison of dendritic-cell vaccine pre-conditioning strategies in newly diagnosed glioblastoma. For the primary outcome, it reports median overall survival of 16.0 months in Group I, 20.0 months in Group II, and 19.0 months in Group III. It also reports two log-rank p-values: 0.072 for Group I versus Group II and 0.089 for Group I versus Group III. The available statistics describe the posted results; they do not establish a treatment effect or explain why the observed group estimates differ.

The medians are estimates of group-level survival distributions, not predictions of an individual participant’s survival. Their confidence intervals show uncertainty around the arm-specific medians. Group III’s interval is incomplete in the registry record, and no comparison-specific confidence intervals or hazard ratios are reported. The p-values do not quantify the size of a difference, and they cannot substitute for an effect estimate with uncertainty.

Interpretation is also constrained by the small Group III outcome count of 8, the termination of randomization to that group effective March 2017, and the absence of details on analysis populations, missing data, multiplicity, and the impact of the change in randomization. The registry reports planned enrollment of 64, but does not explain how that number relates to the posted outcome groups. The outcome counts alone do not reveal whether participants were excluded, had incomplete follow-up, or were analyzed under another rule.

In general, randomized allocation can support a comparison between assigned groups when the design and analysis are carried out as intended. The registry’s summary does not provide enough information to assess all aspects of that process or the comparability of groups at baseline. The findings should therefore be read as the reported median estimates and log-rank tests in this record, with their limitations. They do not establish that any approach improves or worsens survival, and they do not determine what should be done for an individual patient.

14. Why This Trial Matters Statistically

This trial is statistically informative because it places a time-to-event outcome alongside a randomized comparison of several pre-conditioning strategies. The endpoint is overall survival from randomization, and the record identifies Kaplan-Meier estimation and log-rank testing. These methods address the timing of events and incomplete follow-up differently from a simple comparison of proportions at a fixed time. Understanding which estimate or test answers which question is essential when interpreting registry results.

The record also illustrates why arm-specific medians and comparison tests are not interchangeable. A median gives a summary of one group’s estimated survival curve, while a log-rank p-value compares survival experience across groups. Neither alone gives the magnitude and precision of a relative effect. The absence of a reported hazard ratio and comparison-specific confidence interval limits the detail available for understanding the relative survival contrasts.

Design changes and group size are important context. The planned enrollment was 64, the posted outcome groups contain 25, 27, and 8 participants, and randomization to Group III ended effective March 2017. The record does not explain the relationship among these registry entries or describe the effect of the change on analysis. This makes transparent reporting of participant flow, analysis populations, and prespecified methods particularly relevant to interpretation.

Finally, the record shows the value of distinguishing what is documented from what remains unknown. It reports group estimates, confidence intervals where available, log-rank p-values, and safety counts, while leaving several statistical-plan details unreported. A careful reading can describe these findings without treating a p-value as an effect size, an arm-level median as a comparative effect, or descriptive safety counts as a causal assessment.

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 registry reports the arm-specific overall-survival medians, but does not provide comparison-specific effect estimates or confidence intervals. A hazard ratio is not reported. This limits assessment of the relative size and precision of the Group I versus Group II and Group I versus Group III contrasts. Group III’s median confidence interval has no reported upper bound, leaving its uncertainty only partially described.

The Group III outcome group contains 8 participants, and the registry says randomization to this group was terminated effective March 2017. The record does not explain why enrollment to this group ended, how the change affected the randomization process or analysis, or whether it was anticipated in the statistical plan. It also does not state how the outcome group counts relate to the planned enrollment of 64.

Details needed to assess the analysis are not reported in the registry record. These include the sample-size justification, analysis populations, missing-data approach, censoring assessment, multiplicity strategy, and whether interim analyses were planned or conducted. The record does not provide participant flow or explain the different group counts in the survival and safety results. It also does not describe baseline characteristics in the registry record for these results, so the extent of group comparability cannot be assessed from this account.

The registry lists masking as triple but describes assignment to Groups I and II as double-blinded. It does not explain this difference or specify who was masked. Finally, the listed serious adverse event terms are not linked to specific groups in the outcome details, and event attribution is not reported. These limitations concern what can be assessed from the registry record; they do not establish that a particular procedure was not performed.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: DC Migration Study for Newly-Diagnosed GBM — NCT02366728.
• Registry record last updated 2023-06-08; status last verified 2023-06.
• Results first posted 2022-02-01.
• 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

The registry describes a completed randomized phase II, parallel-group study of cytomegalovirus-specific dendritic-cell vaccine pre-conditioning in newly diagnosed glioblastoma. Planned enrollment was 64. The primary outcome was median overall survival in months, defined from randomization to death or last follow-up if alive, with Kaplan-Meier estimation specified.

The reported outcome groups were Group I, unpulsed dendritic-cell pre-conditioning (25 participants); Group II, Td pre-conditioning (27 participants); and Group III, basiliximab plus Td pre-conditioning (8 participants). Their median overall-survival estimates were 16.0, 20.0, and 19.0 months, respectively. The registry reports log-rank p-values of 0.072 for Group I versus Group II and 0.089 for Group I versus Group III.

Randomization to Group III was terminated effective March 2017. A comparison-specific hazard ratio, confidence interval, and several details of the statistical plan are not reported in the registry record. The study status is completed, with primary completion dated 2020-10-31.