Vaccine Therapy for the Treatment of Newly Diagnosed Glioblastoma Multiforme: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Vaccine Therapy for the Treatment of Newly Diagnosed Glioblastoma Multiforme (NCT02465268), a randomized phase 2 study. The primary outcome is Comparison of Overall Survival (OS) Between the Active Treatment Group (Arms 1 and 2) and the Control Group (Arm 3).

2. Clinical Question

The clinical question is therefore whether the posted survival data distinguish either vaccine formulation from the control in this trial population. The results describe the observed comparisons; they do not by themselves establish a causal explanation for differences or similarities between groups.

3. Trial Design

The registry identifies ATTAC-II as a phase II randomized, single-masked, parallel, placebo-controlled trial. It lists 175 as the enrollment and records the study as completed. The sponsor was the University of Florida. The listed start date was 2016-08-09, and the primary completion date was 2023-11-30.

There were three arms. Arm 1 received pp65-shLAMP DC with GM-CSF and Td; Arm 2 received pp65-flLAMP DC with GM-CSF and Td; and Arm 3 received unpulsed PBMC and saline. The descriptions state that dosing began on day 22-24 after the first temozolomide cycle, followed by doses at 2 week intervals. Doses 4-10 were scheduled on day 22-24 of each temozolomide cycle. Dosing could continue until a total of 10 doses, disease progression, or unacceptable toxicity.

The registry describes overall survival as the primary outcome, measured from first vaccination until death, with follow-up up to 48 months. The outcome description says survival would be censored at last follow-up if death had not occurred. The stated primary comparison describes Arms 1 and 2 together against Arm 3, but the posted hazard ratios and arm-level medians are reported separately for each active arm versus the control.

Progression-free survival and immune assays were listed as secondary outcomes. Arm-level outcome counts differ across measures, so the number contributing to one analysis should not be assumed to apply to another. The trial's original sample-size justification, detailed allocation procedures, and full statistical analysis plan are not reported in the registry record.

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: Participants were evaluated for inclusion. 17 were determined not eligible and 4 participants withdrew. 154 participants were randomized with an allocation ratio of 1:1:1 into one of three arms: 1) Arm 1: pp65-shLAMP mRNA DCs with GM-CSF 150 µg (Td skin prep); 2) Arm 2: pp65-flLAMP mRNA DCs with GM-CSF 150 µg (Td skin prep); or 3) Arm 3: Unpulsed PBMCs (saline skin prep).

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Survival in Patients receiving vaccine vs placebo.

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 subjects who received at least one DC vaccine were included in the modified intent to treat analysis.

6. Primary Endpoint

The registry lists the primary outcome as Comparison of Overall Survival (OS) Between the Active Treatment Group (Arms 1 and 2) and the Control Group (Arm 3), measured over From date of first vaccine until the date of death, up to 48 months.

Overall survival is the time from first vaccination to death. The registry specifies that participants who had not died by last follow-up would be censored at that time, and gives a follow-up timeframe of up to 48 months. This endpoint focuses on time until death rather than a change in a laboratory measure or a tumor assessment at a single visit.

Survival is commonly analyzed as a time-to-event outcome because participants may have different lengths of follow-up and not all will have experienced the event by the time results are analyzed. A median survival time describes the point at which the estimated survival curve reaches the halfway mark. It is not an average lifetime and does not describe every participant's experience. The registry reports median overall survival in months for each arm, with 95% confidence intervals.

The primary outcome description specifies Kaplan-Meier curves and a log-rank test to characterize and compare overall survival. The posted results also include hazard ratios for comparisons of each active arm with the control. A hazard ratio summarizes the relative event rate over follow-up under the assumptions of the survival analysis; it is not a difference in median survival. These measures answer related but distinct questions. The registry record does not report an overall-survival estimate combining the two active arms, although the primary outcome description refers to that combined group.

7. Results Posted in the Registry

• pp65-shLAMP DC With GM-CSF and Td (Arm 1) vs Unpulsed PBMC and Saline (Arm 3): hazard ratio 1.29 (95% CI 0.77–2.16), reported p = 0.196 (as reported in the registry).
• pp65-flLAMP DC With GM-CSF and Td (Arm 2) vs Unpulsed PBMC and Saline (Arm 3): hazard ratio 1.63 (95% CI 0.99–2.70), reported p = 0.196 (as reported in the registry).

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Comparison of Progression-free Survival Between the Active Treatment Group (Arms 1 and 2 Combined) and the Control Group (Arm 3)

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor (Log Rank); it was not recalculated.
• Result:
  • pp65-shLAMP DC With GM-CSF and Td (Arm 1) vs Unpulsed PBMC and Saline (Arm 3): hazard ratio 0.98 (95% CI 0.60–1.61), reported p = 0.354 (as reported in the registry).
  • Plus pp65-flLAMP DC With GM-CSF and Td (Arm 2) vs Unpulsed PBMC and Saline (Arm 3): hazard ratio 1.42 (95% CI 0.87–2.33), reported p = 0.354 (as reported in the registry).
• Interpretation:
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between pp65-shLAMP DC With GM-CSF and Td (Arm 1) vs Unpulsed PBMC and Saline (Arm 3) and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Plus pp65-flLAMP DC With GM-CSF and Td (Arm 2) vs Unpulsed PBMC and Saline (Arm 3) and the control. The interval gives the range of effects that remain compatible with the data.

ELISPOT Assay (pp65)

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Arm 2: pp65-flLAMP DC With GM-CSF and Td: mean 48.40 versus 126.30 on control; difference -77.90 (95% CI -441.17 to 285.37), standardized effect d = -0.14; z-test p = 0.674.
  • Arm 3: Unpulsed PBMC and Saline: mean -162.70 versus 126.30 on control; difference -289.00 (95% CI -802.91 to 224.91), standardized effect d = -0.30; z-test p = 0.270.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Arm 2: pp65-flLAMP DC With GM-CSF and Td and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Arm 3: Unpulsed PBMC and Saline and the control. The interval gives the range of effects that remain compatible with the data.

Flow Cytometric Analysis (T Cell)

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • pp65-flLAMP DC With GM-CSF and Td (Arm 2): mean 5.90 versus -0.10 on control; difference 6.00 (95% CI -3.34 to 15.34), standardized effect d = 0.42; z-test p = 0.208.
  • Unpulsed PBMC and Saline (Arm 3): mean 1.70 versus -0.10 on control; difference 1.80 (95% CI -4.73 to 8.33), standardized effect d = 0.20; z-test p = 0.589.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between pp65-flLAMP DC With GM-CSF and Td (Arm 2) and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Unpulsed PBMC and Saline (Arm 3) and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes cover progression-free survival and several immune measures, including ELISPOT assays, flow cytometry, and cytokine analysis. Together, they span clinical time-to-event outcomes and laboratory measures. They can add context about disease progression and measured immune responses, but they do not answer the primary overall-survival question directly.

The posted progression-free survival comparisons do not show a sharply resolved separation: the reported hazard-ratio intervals include the no-difference value, and the listed p-value is 0.354. The immune outcomes are mixed in what they measure and how they are analyzed. The registry reports a sponsor Kruskal-Wallis result for the pp65 ELISPOT assay, while the posted arm summaries and recalculated mean comparisons describe a different aspect of those data. For T-cell flow cytometry, the posted means and unadjusted comparisons do not yield a precise separation. Other listed immune assays do not have outcome results detailed in the record.

Secondary results generally carry less inferential weight than a prespecified primary endpoint because there can be many outcomes, comparisons, time points, and analysis choices. Testing many outcomes increases the chance of apparently notable findings arising by chance if no multiplicity adjustment is used. Whether adjustments were planned or applied is not reported in the registry record. The secondary findings should therefore be treated as supportive or exploratory context, not as independent confirmation of a survival effect.

The broader pattern does not remove uncertainty in the primary result. A laboratory difference, even if detected, would not establish a survival difference, and a secondary result that is inconclusive does not prove that no biological or clinical effect exists. Interpretation depends on the endpoint, analysis method, precision, and completeness of reporting for each outcome.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Comparison of Progression-free Survival Between the Active Treatment Group (Arms 1 and 2 Combined) and the Control Group (Arm 3)
• ELISPOT Assay (pp65)
• Flow Cytometric Analysis (T Cell)
• Cytokine Array Analysis (IFN-g)
• ELISPOT Assay (Actin)
• Flow Cytometric Analysis (NK Cell)
• Flow Cytometric Analysis (CD4.CD25 T Reg)

9. Statistical Interpretation

10. Statistical Methods Explained

How survival analysis handles follow-up. Overall survival is a time-to-event endpoint: each participant contributes information about time from first vaccination until death or last follow-up. If death has not occurred by last follow-up, the registry description says the observation is censored at that point. Censoring means the exact event time is not observed during the recorded follow-up; it does not mean that the participant is treated as having died at the censoring time. Survival methods use the available follow-up while accounting for this incomplete event-time information.

What does a Kaplan-Meier estimate represent?

The Kaplan-Meier method estimates the proportion remaining alive over time while incorporating censored observations. Its curve can show how the estimated survival probability changes through follow-up. The median is the time at which the estimated curve reaches its halfway point, when that point is estimable. It compresses the curve to a single value, so it does not show the full timing or shape of survival across follow-up. The registry reports median times and 95% confidence intervals, but the underlying curve coordinates are not reported in the record.

What does the log-rank test assess?

The log-rank test compares event patterns over follow-up between groups. It is not a test of whether the two medians alone differ, and it does not state the size of an effect. The registry names the log-rank method for the primary survival comparison and reports p-values alongside hazard ratios. A p-value describes how compatible the observed data are with a specified no-difference model under the test's assumptions; it is not the probability that a treatment effect is true. The record does not report further details about the test implementation.

How should a hazard ratio be read?

A hazard ratio summarizes the relative rate of the event between the groups in a stated comparison over follow-up. It is distinct from a ratio of medians and does not directly translate into a fixed difference in survival time. Interpretation also depends on which arm is in the numerator and on whether the relative hazards are reasonably stable over time. The registry reports hazard ratios for Arm 1 versus Arm 3 and Arm 2 versus Arm 3, but does not report an assessment of whether hazards were proportional. The confidence interval conveys uncertainty about the estimated relative hazard; broad intervals indicate that a wide range of effects remains compatible with the data.

What do the continuous-outcome summaries add?

The immune assays include reported means and standard deviations for arm-level measurements at baseline and post-vaccine #3. A mean describes the arithmetic center of recorded values, while a standard deviation describes their spread around that mean. The posted recalculations compare means with unadjusted estimates and confidence intervals. These are not the same analysis as a rank-based test: the sponsor's ELISPOT result is reported using Kruskal-Wallis, which compares distributions using ranks rather than relying on a mean difference alone. The registry does not report a sponsor analysis for the posted T-cell flow-cytometry outcome.

Large standard deviations relative to the means, as seen in the ELISPOT measurements, indicate substantial variability in those measurements. In general, over-dispersion or marked variability can make a mean less representative of a typical observation and can reduce precision. This general statistical point does not establish why the observed variability occurred. The arm-level summaries do not reveal the full distribution, the influence of extreme values, or how missing measurements were handled.

The record's primary survival results rely on sponsor-reported hazard ratios. The recalculated continuous-outcome comparisons are simple, unadjusted summaries based on posted arm-level statistics. They do not reproduce a participant-level analysis, adjust for baseline covariates, or account for any multiplicity unless explicitly stated. No such additional adjustments are reported for these recalculations. They should therefore be treated as descriptive statistical comparisons, not as a replacement for the trial's own analysis.

11. Confidence Intervals and Precision

The 95% confidence intervals communicate how much uncertainty surrounds the estimates under the analysis method. A narrower interval gives a more precise estimate; a wider interval leaves more plausible values unresolved. The intervals are not guarantees that the true value lies within the displayed range, and they do not show the distribution of individual survival times.

For overall survival, the median estimates are accompanied by intervals that extend across a substantial range of months. The hazard-ratio intervals are also broad: Arm 1 versus Arm 3 has a 95% confidence interval from 0.77 to 2.16, and Arm 2 versus Arm 3 has a 95% confidence interval from 0.99 to 2.70. Each includes the no-relative-difference value or comes very close to it, while also allowing for effects in either direction. This means the point estimates alone should not be read as precise evidence about the relative survival experience.

Confidence intervals for the secondary progression-free survival comparisons are likewise broad. The immune-assay mean-difference intervals also span values on both sides of no difference. Precision varies by outcome, and the number contributing to one measure cannot be assumed to match the number contributing to another. The registry record does not report the reasons for these differences in outcome counts.

Intervals describe statistical uncertainty conditional on the recorded data and analysis. They do not account for every possible limitation, such as unreported missing-data procedures or the selection of outcomes for emphasis. A confidence interval that includes no difference does not prove equality; it indicates that the data do not pin down the effect with enough precision to exclude that possibility under the stated analysis.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size illustrations are planning calculations based on the observed effect, not the trial's original design assumptions. For time-to-event outcomes, the displayed quantities are total events required. For the other outcomes, they are participants per arm. Thus the survival illustrations should not be read as a per-arm enrollment target, while the continuous-outcome illustrations should not be read as a total across arms.

The calculations assume a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. They consider planning to detect 100%, 75%, or 50% of the observed effect. The trial's own original sample-size assumptions and justification are not reported in the registry record.

The 80% power column describes the planned event count or per-arm participant count associated with an 80% chance of detecting the specified effect under the calculation assumptions, if that effect is present and the assumptions hold. The 90% power column gives the corresponding planning quantity for a 90% chance. These are design illustrations, not guarantees about what a future study would observe.

Smaller effects are harder to distinguish from random variation, so planning to detect a smaller fraction of the observed effect requires more events or participants. This is not observed power. It does not measure the probability that the trial's result is correct, and it should not be used to reinterpret the observed p-values as evidence for or against an effect.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

d is the total number of events needed (Schoenfeld's formula) with 1:1 allocation and proportional hazards.

Worked calculation for this trial

1. Hazard ratio HR = 1.290, so ln(HR) = 0.2546.
2. d = 4 × (1.96 + 0.842)² / [ln(1.290)]² = 484.2, rounded up to 485 events in total for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
hr <- 1.29                              # observed hazard ratio
events_needed <- function(power, frac = 1) ceiling(round(4 * (z_a + qnorm(power))^2 / log(hr^frac)^2, 6))   # Schoenfeld, 1:1 allocation
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: events = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(events_needed(0.80, f), big.mark = ","), format(events_needed(0.90, f), big.mark = ",")))

# Simulation: given d total events under proportional hazards, events in the treatment arm are Binomial(d, hr / (1 + hr)).
# The score (log-rank type) statistic is (D1 - d / 2) / sqrt(d / 4).
set.seed(2465268)
d    <- events_needed(0.80)
reps <- 20000
D1 <- rbinom(reps, d, hr / (1 + hr))
z  <- (D1 - d / 2) / sqrt(d / 4)
cat(sprintf("Simulated power with %s events: %.3f (%d replicates)\n", format(d, big.mark = ","), mean(abs(z) > z_a), reps))

Output from running the code

100% of observed effect: events = 485 (80% power), 649 (90% power)
 75% of observed effect: events = 861 (80% power), 1,153 (90% power)
 50% of observed effect: events = 1,937 (80% power), 2,593 (90% power)
Simulated power with 485 events: 0.789 (20000 replicates)

The simulation uses the fact that, with d events in total and proportional hazards, the number of events in the treatment arm is binomial. It is an approximation that conditions on the total number of events, and it should land close to the target power.

14. Analyses Reported by the Sponsor

The primary survival comparisons are reported by the sponsor using hazard ratios from log-rank analyses. The registry lists a hazard ratio of 1.29 for Arm 1 versus Arm 3 and 1.63 for Arm 2 versus Arm 3, with 95% confidence intervals and p-values for each comparison. These are the sponsor's analyses and are the relevant results for interpreting the trial's reported survival comparisons.

The posted recalculations do not replace those analyses. For overall survival and progression-free survival, the record identifies the sponsor-reported hazard ratios as the analyses to use. There is no recalculated hazard ratio presented as an alternative. The record indicates no discrepancy between the sponsor's analysis and the posted extraction.

For some immune outcomes, the posted unadjusted mean comparisons use different summaries and methods from the sponsor's reported tests. For ELISPOT, the sponsor reports a Kruskal-Wallis p-value, whereas the recalculated comparisons describe differences in means. A rank-based test and a mean-difference calculation need not yield the same result because they answer different statistical questions and use the data differently. The sponsor's own analysis takes precedence if a discrepancy is flagged; no such discrepancy is flagged in this record. The recalculated summaries are best read as supplementary, unadjusted descriptions rather than as a substitute for the sponsor's analysis.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when analyses include several outcomes, treatment comparisons, or time points. If many hypotheses are tested, some results may appear noteworthy by chance even when no underlying effect is present. The registry lists overall survival as primary and includes multiple secondary outcomes, but whether the trial used a multiplicity adjustment, a hierarchy among outcomes, or a prespecified testing sequence is not reported in the registry record. The displayed secondary p-values should therefore not be read as if they necessarily account for all the analyses conducted.

The record does not report the statistical analysis population for each endpoint, such as whether analyses followed intention-to-treat assignment or another population definition. The number contributing to the posted outcomes varies, but reasons for noncontribution, exclusions, or missing measurements are not reported. The handling of missing immune-assay values and the details of censoring or follow-up completeness beyond the stated survival censoring rule are also not reported.

The primary outcome description names Kaplan-Meier curves and a log-rank test. It does not describe covariate adjustment, assessment of proportional hazards, subgroup analyses, interim analyses, or additional sensitivity analyses. Those details are not reported in the registry record, so their use should not be presumed.

These reporting gaps do not establish that appropriate procedures were absent; they mean that the record does not permit readers to assess them. The distinction matters because multiplicity, missingness, and analysis-population choices can affect uncertainty and interpretation. The posted results can be described on their own terms, while conclusions about unreported procedures must remain limited.

16. 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 1% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: seizure (11.1%), neurology - other (specify, cerebral edema) (8.3%), infection (5.6%), gastrointestinal - other (specify, appendicitis) (2.9%), hematoma (2.9%).

The registry lists serious and other adverse-event counts separately by arm. For serious adverse events, the at-risk and affected counts are 36 and 8 in Arm 1, 34 and 6 in Arm 2, and 41 and 5 in Arm 3. For other adverse events, the corresponding at-risk and affected counts are 36 and 16, 34 and 11, and 41 and 13. These are arm-level registry counts; they should not be added together or interpreted as counts of distinct participants across the study.

Frequently listed serious-event terms include seizure at 11.1%, neurology - other (specified as cerebral edema) at 8.3%, and infection at 5.6%. The record also lists gastrointestinal - other (specified as appendicitis), hematoma, and syncope (fainting), each at 2.9%. The registry records deaths as 28 in Arm 1, 31 in Arm 2, and 31 in Arm 3. These death counts are reported separately from the adverse-event listings; the record does not establish that they were caused by treatment or classify them as adverse-event deaths.

These figures describe what was recorded, not a causal comparison of safety. The registry information does not provide enough detail to determine event timing, severity beyond the serious-event classification, recurrence, or attribution. Differences in counts should be interpreted with caution and alongside the study's incomplete reporting of analysis populations and follow-up.

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

The posted results establish that the registry reports arm-level overall-survival medians, hazard ratios, confidence intervals, and p-values for the two active-arm comparisons with control. The estimates are uncertain: the hazard-ratio intervals are broad and include values compatible with no relative difference, while also allowing for effects in either direction. The medians are descriptive summaries and do not alone establish that one arm differs from another.

These data do not demonstrate equivalence between the vaccine arms and control. An inconclusive comparison is not the same as evidence that the treatments have identical effects. Nor do the numerical differences in medians or hazard ratios establish that a treatment caused a difference. The registry record does not report a combined active-arm overall-survival estimate, even though the primary endpoint description refers to a combined active group.

The reported secondary outcomes do not resolve the uncertainty in the primary endpoint. Progression-free survival is another clinical outcome, while immune assays measure laboratory features; neither should be treated as a substitute for overall survival. Unadjusted secondary analyses also require caution when the multiplicity plan is not reported.

Overall, the record supports a careful description of the observed estimates and their uncertainty, not a broad claim about treatment effect. Interpretation is limited by the reporting available for missing data, analysis populations, multiplicity, and the trial's original design assumptions. These statistical results are not medical advice and do not determine what treatment is appropriate for an individual.

18. Why This Trial Matters Statistically

This trial illustrates why a survival result should be read through both its point estimate and its precision. A median alone can make groups appear meaningfully separated or similar, but the confidence interval and the full time-to-event analysis show how uncertain that summary is. The reported hazard ratios add a relative-rate measure, yet their intervals also leave a broad range of effects plausible.

The design has three arms, including two active vaccine formulations and a control. That structure creates an important reporting distinction: the primary outcome description refers to a combined active group, whereas the posted overall-survival results give separate active-arm comparisons. Readers need to know which contrast is actually estimated before drawing conclusions about the primary question.

The trial also highlights the difference between clinical and biological endpoints. Progression-free survival and immune measurements may help describe other aspects of the study, but their analyses cannot be substituted for an overall-survival comparison. Multiple secondary outcomes make transparent reporting of the analysis plan and multiplicity procedures particularly useful.

Finally, the registry record shows how much interpretation depends on details beyond estimates and p-values. Information about missing data, analysis populations, and the original sample-size rationale is not reported. Recognizing those limits is part of sound statistical reading: it keeps the evidence tied to what is documented and prevents uncertainty from being mistaken for proof in either direction.

19. 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.

Several limitations are specific to the way the trial results are reported. The primary outcome description refers to comparing the two active arms combined with control, but the posted overall-survival hazard ratios are separate comparisons of each active arm with Arm 3. A combined active-arm hazard ratio is not reported, so the planned wording and available comparison estimates do not align fully.

The registry reports arm-level median survival and confidence intervals, but the underlying Kaplan-Meier curves and details about the timing and completeness of follow-up are not described in the record. The censoring rule is stated, but additional diagnostics and sensitivity analyses are not reported. The registry also does not report whether the proportional-hazards assumption was assessed for the hazard-ratio analyses.

Outcome counts differ across the clinical and laboratory measures. The record does not explain the differences, identify the reasons for missing measurements, or state how missing data were handled. It also does not define the analysis population for each result. These omissions limit assessment of selection and missingness.

The trial's original sample-size justification, multiplicity strategy, interim-analysis plan, and full statistical analysis plan are not reported. The planning illustrations based on observed effects do not fill those gaps. The listed immune outcomes include multiple measures, while detailed results are not reported for every listed assay in the record. These reporting limits warrant caution without implying that unreported methods were not used.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Vaccine Therapy for the Treatment of Newly Diagnosed Glioblastoma Multiforme — NCT02465268.
• Registry record last updated 2025-01-14; status last verified 2024-12.
• Results first posted 2025-01-14.
• 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.

23. Record Summary

The registry record describes ATTAC-II, a completed randomized, single-masked, parallel, placebo-controlled phase II trial in newly diagnosed glioblastoma. It lists enrollment of 175 and three arms: two pp65 dendritic-cell vaccine formulations with GM-CSF and Td, and unpulsed PBMC with saline as control.

Overall survival was the primary outcome, measured from first vaccination to death, with a stated timeframe of up to 48 months. The posted results report arm-level medians and separate hazard-ratio comparisons of each active arm with control. The confidence intervals are broad, and the record does not report a combined active-arm survival estimate.

Secondary outcomes include progression-free survival and immune assays. Serious and other adverse-event counts are listed by arm. The registry does not report important details such as the original sample-size justification, multiplicity procedures, analysis populations, or missing-data methods.