A Study of ABT-414 in Participants With Newly Diagnosed Glioblastoma (GBM) With Epidermal Growth Factor Receptor (EGFR) Amplification: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of ABT-414 in Participants With Newly Diagnosed Glioblastoma (GBM) With Epidermal Growth Factor Receptor (EGFR) Amplification (NCT02573324), a randomized phase 3 study. The primary outcome is Overall Survival (OS).

2. Clinical Question

Population

Participants had newly diagnosed glioblastoma with epidermal growth factor receptor (EGFR) amplification. The primary analysis population was the Full Analysis Set: randomized participants enrolled in the main study by 31-Mar-2018, excluding the open-label hepatic-impairment sub-study. The outcome definition included deaths due to any cause.

Intervention

The experimental regimen combined depatuxizumab mafodotin, radiation and temozolomide (TMZ). During chemoradiation, depatuxizumab mafodotin was scheduled on Day 1 of Week 1, 3 and 5 with standard therapy. During adjuvant treatment it was scheduled on Day 1 and 15 of each cycle with TMZ on Days 1-5 of each cycle.

Comparator

The comparison group received placebo for ABT-414 alongside radiation and TMZ. Placebo followed the corresponding schedule during chemoradiation and the adjuvant phase. The trial was randomized, triple-masked and parallel-group.

Primary question

The central question was whether adding ABT-414, also called depatuxizumab mafodotin, to concurrent chemoradiation and subsequent adjuvant TMZ prolonged overall survival in this population. Overall survival (OS) was defined as time from randomization to death from any cause. The posted comparison reports median OS of 18.7 months for placebo plus radiation and TMZ and 18.9 months for depatuxizumab mafodotin plus radiation and TMZ. The registry-reported hazard ratio for placebo versus depatuxizumab mafodotin was 1.02, with a 95% confidence interval of 0.82 to 1.26 and p=0.633. These estimates describe the observed comparison and its uncertainty; they do not by themselves establish a clinically meaningful benefit or harm.

The record also includes an open-label Phase 1 sub-study assessing pharmacokinetics, safety and tolerability in participants with mild or moderate hepatic impairment. That sub-study is distinct from the randomized primary comparison.

3. Trial Design

This was a randomized, placebo-controlled Phase 3 study with a parallel-group model and triple masking. Participants in the main study had newly diagnosed glioblastoma with EGFR amplification. The overall registry enrollment was 691. The primary outcome population was narrower than the overall enrollment: it included randomized main-study participants enrolled on or prior to 31-Mar-2018 and excluded those in the open-label hepatic sub-study. The registry record does not report the trial's original sample-size justification or the full protocol-level analysis plan.

The experimental group received depatuxizumab mafodotin with radiation and TMZ during chemoradiation, followed by depatuxizumab mafodotin with adjuvant TMZ. The comparator received placebo with the same background radiation and TMZ approach. According to the arm descriptions, study drug or placebo was scheduled on Day 1 of Week 1, 3 and 5 in chemoradiation, and on Day 1 and 15 of each cycle during adjuvant treatment. TMZ was given on Days 1-5 of each adjuvant cycle. The record identifies the interventions as depatuxizumab mafodotin, placebo for ABT-414, radiation and temozolomide.

The posted primary OS results give arm sizes of 316 for placebo, radiation and TMZ and 323 for depatuxizumab mafodotin, radiation and TMZ. These are the reported outcome-analysis arm sizes, not a recalculation of total enrollment. The overall median duration of follow-up was 15.5 months, with a range of 0.1 to 35.6 months.

A separate open-label Phase 1 sub-study concerned participants with mild or moderate hepatic impairment. It was not part of the primary Full Analysis Set. Its regimen and purpose differ from the masked randomized comparison, so its observations should not be blended into interpretation of the randomized OS estimate. The record marks the study as completed and lists its start as 2015-01-04 and primary completion as 2022-04-04.

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: An efficacy futility analysis, conducted in accordance with the SAP, generated a conclusive result of futility, meeting the main study's objective of determining whether the addition of Depatux-M to standard therapy improves overall survival. The study was modified at that point, remaining open until after all participants had discontinued study treatment. AbbVie therefore considered this study completed (according to the rules set out for the study) and not terminated.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Full Analysis Set (FAS): all randomized participants (regardless of whether they received study treatment) satisfying the following criteria: * Enrollment date was on or prior to 31-Mar-2018 * Enrolled in the main study (not in the open-label hepatic sub-study)

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 Overall Survival (OS), measured over Overall median duration of follow-up was 15.5 months (range: 0.1, 35.6)..

Overall survival is the time from randomization until death from any cause. It is an endpoint that captures survival duration without requiring a cause-specific classification of death. In this trial, the registry reports median OS in months and a hazard ratio comparison between treatment groups. The overall median follow-up was 15.5 months, with a range of 0.1 to 35.6 months.

A median is the point in the observed survival distribution by which half of the group’s participants have experienced the event and half have not, with the exact estimation depending on the survival-analysis method and censoring. In general, time-to-event methods can account for participants whose event time is not observed during follow-up by using their available follow-up information rather than treating them as if the event occurred at a known time.

OS is relevant to the stated study question because the trial sought to determine whether adding depatuxizumab mafodotin to chemoradiation and adjuvant TMZ prolonged survival. The registry defines the event as death from any cause, so the endpoint is not limited to deaths attributed to glioblastoma or treatment. The posted medians describe each arm’s observed survival distribution, while the hazard ratio summarizes a relative comparison over follow-up under the proportional-hazards model used in the registry analysis. Neither the median nor the hazard ratio alone conveys every feature of the survival curves, and the confidence interval is important for understanding uncertainty.

The registry record does not report additional detail about the handling of censoring, endpoint adjudication, or the original statistical analysis plan beyond the stated endpoint definition and listed analysis methods.

7. Results Posted in the Registry

• Placebo, Radiation and TMZ vs Depatuxizumab Mafodotin, Radiation and TMZ: hazard ratio 1.02 (95% CI 0.82–1.26), reported p = 0.633 (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.

OS for the O6-methylguaninemethlytransferese (MGMT) Unmethylated Group

• 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:
  • Placebo, Radiation and TMZ vs Depatuxizumab Mafodotin, Radiation and TMZ: hazard ratio 0.97 (95% CI 0.76–1.24), reported p = 0.504 (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 Placebo, Radiation and TMZ vs Depatuxizumab Mafodotin, Radiation and TMZ and the control. The interval gives the range of effects that remain compatible with the data.

OS for the MGMT Methylated Group

• 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:
  • Placebo, Radiation and TMZ vs Depatuxizumab Mafodotin, Radiation and TMZ: hazard ratio 1.17 (95% CI 0.76–1.80), reported p = 0.773 (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 Placebo, Radiation and TMZ vs Depatuxizumab Mafodotin, Radiation and TMZ and the control. The interval gives the range of effects that remain compatible with the data.

OS for the Epidermal Growth Factor Receptor (EGFR)vIII-Mutated Tumor Subgroup

• 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:
  • Placebo, Radiation and TMZ vs Depatuxizumab Mafodotin, Radiation and TMZ: hazard ratio 0.95 (95% CI 0.71–1.27), reported p = 0.381 (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 Placebo, Radiation and TMZ vs Depatuxizumab Mafodotin, Radiation and TMZ and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary OS analyses address MGMT methylation groups and the EGFRvIII-mutated tumor subgroup. Their point estimates are near the no-relative-hazard reference, and each reported confidence interval includes 1. The subgroup results therefore do not present a clear, precise pattern that would materially alter the reading of the primary OS comparison.

The registry lists additional secondary outcomes involving progression-free survival, symptom severity and interference, and neurocognitive functioning. Results for these outcomes are not reported in the registry record in the information available for interpretation. Their listing indicates outcome domains of interest, not the direction or strength of any findings. The secondary endpoint inventory also includes subgroup analyses; their presence should not be mistaken for evidence of a positive result.

In general, secondary outcomes can help characterize treatment effects beyond the primary question, but they carry less weight when many outcomes or subgroups are assessed without a reported multiplicity strategy. Multiple testing increases the chance that some estimates appear noteworthy by chance. In addition, subgroup analyses often have fewer observations and events than the main analysis, which can make estimates less precise. These considerations are especially relevant when the registry record does not report a prespecified hierarchy, adjusted testing procedure or interaction test.

The primary OS estimate remains the main basis for reading the trial’s stated question. Secondary findings can add context, but they should not be used selectively to replace or overrule the primary result. The registry record does not report a multiplicity adjustment or enough detail to determine whether the listed secondary analyses were confirmatory, exploratory, or both. Accordingly, the secondary findings are best treated as supporting descriptive evidence, with uncertainty and reporting gaps kept in view.

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

• OS for the O6-methylguaninemethlytransferese (MGMT) Unmethylated Group
• OS for the MGMT Methylated Group
• OS for the Epidermal Growth Factor Receptor (EGFR)vIII-Mutated Tumor Subgroup
• Progression-Free Survival (PFS)
• PFS for EGFRvIII-Mutated Tumor Subgroup
• Deterioration Free Survival in M.D. Anderson Symptom Inventory Brain Tumor Module (MDASI-BT) Symptom Severity Score
• Deterioration Free Survival in MDASI-BT Symptom Interference Score
• Deterioration Free Survival in Neurocognitive Functioning on the Hopkins Verbal Learning Test Revised (HVLT-R) Total Recall Score

9. Statistical Interpretation

10. Statistical Methods Explained

The primary comparison is a time-to-event analysis. Participants enter follow-up at randomization, and the event is death from any cause. The endpoint is expressed as median survival time in months, while the treatment comparison is expressed as a hazard ratio from a Cox proportional-hazards analysis. The registry identifies the comparison method as log rank and reports a Cox proportional-hazards estimate.

What does the hazard ratio describe?

In general, a hazard is the event rate at a particular point in follow-up among participants who have not yet experienced the event. A hazard ratio compares those rates between groups over time. It is not the ratio of the medians and should not be described as a direct percentage change in survival time. The posted comparison is labeled placebo, radiation and TMZ versus depatuxizumab mafodotin, radiation and TMZ, with a reported hazard ratio of 1.02. The direction of interpretation follows that registry comparison order.

The Cox proportional-hazards model estimates a relative hazard while accounting for the timing of events and available follow-up. Its proportional-hazards assumption is that the relative hazard is reasonably stable over time. The registry record names this model but does not report diagnostics or other information that would show how the assumption was assessed. If hazards vary over time, a single hazard ratio may not fully represent the changing pattern.

What does the log-rank test contribute?

In general, the log-rank test compares survival experience across groups over follow-up, using event-time information rather than comparing only the medians. The registry reports a log-rank p-value of 0.633 with the Cox estimate and also lists a separate log-rank p-value of 0.704 without an effect estimate. The record does not explain why both are listed or specify whether they correspond to distinct test formulations. A p-value measures how unusual the test statistic would be under its null model; it is not a measure of effect size or the probability that the null hypothesis is true.

Why report both a median and a hazard ratio?

The median is an absolute time summary that is comparatively direct to read: it gives a point in the estimated survival distribution. The hazard ratio is a relative comparison that uses the event timing across follow-up. The two answer related but different questions. The median does not fully describe the survival curve, and a hazard ratio does not indicate the number of months an individual might live. In general, a full interpretation benefits from looking at survival curves and confidence intervals as well as the single-number summaries.

The arm medians are 18.7 months and 18.9 months, with 95% confidence intervals of 17.0 to 20.3 and 17.4 to 20.8, respectively. The hazard-ratio confidence interval is 0.82 to 1.26. These intervals describe statistical uncertainty around different estimands: the medians and the relative hazard. Their overlap is not itself a formal test of the treatment comparison.

How should subgroup analyses be read?

The posted subgroup analyses estimate OS separately for MGMT unmethylated, MGMT methylated and EGFRvIII-mutated tumor groups. In general, subgroup results can help describe whether patterns appear different across clinically characterized groups, but they often have less precision than the main comparison. A difference between subgroup estimates is not, by itself, evidence that the treatment effect differs between subgroups. A formal interaction analysis would be needed to assess effect modification; such a result is not reported in the registry record.

The record lists secondary outcomes including progression-free survival and symptom, interference and neurocognitive deterioration-free survival. Results for several of those outcomes are not reported in the registry record. The presence of an outcome in the registry does not establish that a result was posted or that it was statistically significant.

Statistical methods and assumptions for the original trial are not fully described. The record does not report the original sample-size justification, the full covariate adjustment strategy, missing-data methods, interim analyses, or diagnostics for the proportional-hazards assumption. General descriptions of these methods should not be mistaken for confirmation that any unreported procedure was used in this trial.

11. Confidence Intervals and Precision

The primary hazard-ratio estimate is 1.02, with a 95% confidence interval from 0.82 to 1.26. In general, a confidence interval represents a range of estimates reasonably compatible with the observed data and the statistical model, under repeated-sampling assumptions. It conveys uncertainty more directly than a p-value alone. It is not a range that contains a fixed proportion of individual participants’ outcomes, nor does it assign a probability to the true effect after the data have been observed.

The interval includes 1, the no-relative-hazard reference for a hazard ratio. It also extends below and above that reference. Thus, the primary estimate is not very precise about the direction of the relative hazard. The estimate is near 1, but the interval does not demonstrate equivalence; an equivalence conclusion would require a prespecified equivalence margin and an analysis designed for that purpose, neither of which is reported in the registry record.

The medians also have uncertainty intervals. Placebo, radiation and TMZ had a median OS of 18.7 months (95% CI 17.0 to 20.3), and depatuxizumab mafodotin, radiation and TMZ had a median OS of 18.9 months (95% CI 17.4 to 20.8). Those intervals give context for the median estimates but do not substitute for the reported hazard-ratio comparison. Overlap or non-overlap of separate confidence intervals is not, by itself, the appropriate test of a between-group difference.

Precision is also limited in the subgroup estimates. Their reported hazard-ratio confidence intervals extend across 1, including the MGMT unmethylated, MGMT methylated and EGFRvIII-mutated groups. Subgroup intervals are important because a point estimate can look more or less favorable while remaining uncertain. The record does not report a prespecified hierarchy for these subgroup comparisons or a multiplicity adjustment, so they should be treated as descriptive rather than independent confirmatory evidence.

The follow-up information matters for interpretation: the overall median duration was 15.5 months, with a range of 0.1 to 35.6 months. The registry record does not provide further detail about the distribution of follow-up by arm or the completeness of outcome ascertainment.

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 illustration uses time-to-event planning expressed as the total number of events, not participants per arm. It is based on the observed effect and uses a two-sided significance level of 0.05, equal allocation between the two arms and a normal approximation. The effect to detect is set at the stated fraction of the observed effect. These are planning assumptions for an illustration, not the trial’s original design assumptions; the original design assumptions are not reported in the registry record.

The 80% power column describes the event count associated with an 80% chance of detecting the specified effect under the planning assumptions, if that effect is true. The 90% power column describes the corresponding event count for a 90% chance. These are target operating characteristics for a hypothetical design, not probabilities that the posted trial result is correct. This is not observed power.

The rows scale the effect used for planning: the observed effect, 75% of the observed effect and 50% of the observed effect. Smaller effects are harder to distinguish from random variation, so they require more events to achieve the same planned power under otherwise unchanged assumptions. The total-event figures should not be mistaken for enrollment targets or for numbers of participants in either arm. In a time-to-event trial, event accumulation, follow-up and censoring all affect how many participants would be needed to reach an event target; the registry illustration does not give a participant total for these time-to-event scenarios.

The figures are calculated from the observed effect and are not a reconstruction of the sponsor’s original planning. They should be read as conditional illustrations of how event requirements change with the assumed effect and target power, not as evidence that the completed trial had a particular power or that any planned target was met.

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 = 1.020, so ln(HR) = 0.0198.
2. d = 4 × (1.96 + 0.842)² / [ln(1.020)]² = 80,061.2, rounded up to 80,062 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 <- 1.02                              # 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(2573324)
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 = 80,062 (80% power), 107,180 (90% power)
 75% of observed effect: events = 142,332 (80% power), 190,541 (90% power)
 50% of observed effect: events = 320,245 (80% power), 428,717 (90% power)
Simulated power with 80,062 events: 0.805 (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 sponsor-reported primary analysis gives a Cox proportional-hazards estimate of 1.02 for placebo, radiation and TMZ versus depatuxizumab mafodotin, radiation and TMZ, with a 95% confidence interval of 0.82 to 1.26 and p=0.633 using the log-rank method. The registry also lists a log-rank p-value of 0.704 without an accompanying effect estimate or confidence interval. The record does not explain the relationship between these two entries.

The sponsor’s reported hazard ratio is the primary source for interpreting the trial comparison. The additional planning illustrations are not a replacement for that analysis: they are constructed from the observed effect under stated assumptions and address hypothetical event requirements. No discrepancy between the posted comparison and the sponsor analysis is identified in the registry record, so there is no conflict requiring a recalculated estimate to take precedence.

Any simple re-analysis based on posted summary statistics would be unadjusted and would not reproduce the participant-level survival analysis. It cannot account for the exact event times, censoring, stratification or covariate handling unless those data and procedures are available. The registry’s sponsor-reported analysis therefore remains the appropriate result to cite, while the planning figures serve a separate illustrative purpose. The record does not report enough detail to independently verify all elements of the sponsor’s analysis.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when several endpoints, subgroups or statistical tests are examined. In general, the more comparisons made, the greater the chance that at least one result may appear unusual even when there is no corresponding effect. A prespecified testing hierarchy or adjustment can help control this issue, but the registry record does not report whether a multiplicity procedure was used for the secondary outcomes or subgroup analyses.

The registry lists nine outcome measures and thirteen analyses. It reports primary OS and several subgroup OS results, while results for other listed outcomes are not reported in the registry record. The presence of multiple analyses means that secondary p-values should not be interpreted as if each were the only planned test. The record does not describe which analyses were prespecified, whether any were exploratory, or how conclusions were to be protected against multiple testing.

The primary analysis population is defined as the Full Analysis Set: all randomized participants satisfying the main-study and enrollment-date criteria, whether or not they received study treatment. The open-label hepatic sub-study is excluded. This definition supports an analysis based on randomized assignment rather than treatment received, but the registry record does not provide further details about exclusions after randomization, missing outcome data, or how censoring was handled.

In general, time-to-event analyses can retain information from participants whose event has not been observed by the end of their available follow-up through censoring. The validity of that approach depends on assumptions about the censoring process. The record does not describe censoring rules or provide enough information to evaluate whether missingness or loss to follow-up could have affected the estimates. It also does not report interim analyses or stopping rules. These omissions do not establish that such procedures were absent; they mean that their details are not reported in the registry record.

Without the full statistical analysis plan, it is not possible to determine the planned testing hierarchy, analysis populations for every secondary outcome, handling of missing data, or any interim monitoring strategy. The reported estimates should therefore be interpreted as registry results with these documentation limits, not as a complete account of every analytical decision.

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 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 (16.7%), malignant neoplasm progression (3.5%), pneumonia (2.9%), thrombocytopenia (2.9%), brain oedema (2.6%).

The registry reports serious adverse events by treatment group and study phase, alongside the number at risk and the number affected. These are counts within the listed phase-specific groups; the phases should not be combined or treated as independent cohorts. For example, the placebo arm reports 56 serious adverse events among 335 at risk during chemoradiation and 85 among 285 at risk during the adjuvant phase. In follow-up, 0 were reported among 326 at risk. The depatuxizumab mafodotin arm reports 68 among 342 at risk during chemoradiation, 107 among 273 at risk during adjuvant treatment, and 2 among 326 at risk in follow-up.

For the open-label sub-study, the registry lists 1 serious adverse event among 6 at risk in each of the chemoradiation and adjuvant phases, and 0 among 5 at risk in follow-up. These small phase-specific counts should be read in the context of the separate open-label sub-study. The record also lists deaths by phase, but these should not be interpreted as treatment-caused deaths without additional attribution information.

The most frequently listed serious-event terms include seizure at 16.7%, malignant neoplasm progression at 3.5%, pneumonia and thrombocytopenia at 2.9% each, brain oedema at 2.6%, and pyrexia at 1.8%. The record does not specify in this summary the denominator or arm attribution for those term percentages. Serious-event counts describe reported safety observations; they do not by themselves establish causality or comparative risk.

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

The posted primary result describes OS in randomized participants with newly diagnosed EGFR-amplified glioblastoma who met the Full Analysis Set criteria. The median estimates were 18.7 months with placebo, radiation and TMZ and 18.9 months with depatuxizumab mafodotin, radiation and TMZ. The reported hazard ratio was 1.02, with a 95% confidence interval of 0.82 to 1.26 and p=0.633. Taken together, these statistics show a point estimate close to the no-relative-hazard reference, with uncertainty that includes values on either side.

That result does not establish that the regimens are equivalent. A non-significant superiority test is not an equivalence or non-inferiority test. Nor does the result prove that no effect exists; the interval allows a range of possible relative hazards. The registry record does not report a clinical threshold defining what size of OS difference would be meaningful, so statistical estimates alone cannot determine whether a difference would matter to patients or clinicians.

The result also does not describe every aspect of benefit and harm. The primary endpoint is survival, while other listed endpoints include progression-free survival, symptoms and neurocognitive outcomes. Some of those results are not reported in the registry record. Safety entries summarize serious adverse events by phase, but the counts do not establish that treatment caused an event or provide a complete comparative safety assessment.

General interpretation of a randomized comparison depends on the design, analysis population, endpoint definition, follow-up, uncertainty and completeness of reporting. Here, the record identifies randomization and triple masking, defines the primary population and outcome, and posts the main estimate. It does not provide the full protocol or statistical analysis plan, and it does not report all details needed to assess missing data, model assumptions, multiplicity or subgroup hypothesis testing.

The most defensible reading is therefore descriptive: the registry reports similar median OS point estimates and a hazard-ratio estimate near 1, with a confidence interval that is not sufficiently narrow to establish equivalence or precisely determine the direction and size of any difference. No medical advice or treatment recommendation follows from these registry statistics alone.

18. Why This Trial Matters Statistically

This trial is statistically notable because it asks a clear time-to-event question in a randomized, placebo-controlled Phase 3 setting: whether adding depatuxizumab mafodotin to chemoradiation and adjuvant TMZ prolongs OS in newly diagnosed EGFR-amplified glioblastoma. Randomization and masking are important design features because they can reduce certain sources of systematic bias when implemented and maintained. The parallel-group structure also makes the treatment comparison direct.

The result illustrates why a point estimate should not be read without its confidence interval. The reported hazard ratio of 1.02 is close to 1, but its 95% confidence interval of 0.82 to 1.26 includes both lower and higher relative hazards. The medians are close as well, but median comparisons alone do not capture the full survival experience. The example demonstrates the difference between a lack of strong evidence against a null comparison and proof that two regimens are equivalent.

The record also highlights the distinction between a sponsor’s treatment-effect analysis and a planning illustration based on the observed effect. Event targets in the planning display are conditional on assumptions and are not the original trial design or an estimate of observed power. Keeping those questions separate helps prevent a hypothetical planning calculation from being mistaken for evidence about the completed trial’s statistical success.

Finally, the registry includes several secondary outcomes and subgroup analyses, but the available results and multiplicity details are incomplete. This makes disciplined interpretation important: the primary endpoint should anchor the trial-level conclusion, while subgroup and secondary findings should be assessed with attention to precision, multiplicity and reporting completeness. The case is a practical example of reading registry statistics as estimates with limits rather than as simple yes-or-no verdicts.

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.

The primary outcome population is restricted to randomized main-study participants enrolled on or prior to 31-Mar-2018 and excludes the open-label hepatic sub-study. The overall registry enrollment is 691, while the primary OS arm sizes are 316 and 323. The registry record does not explain the relationship between overall enrollment and the primary analysis population in further detail, so these populations should not be assumed to be interchangeable.

The median follow-up was 15.5 months, with a range of 0.1 to 35.6 months. The record does not provide arm-specific follow-up summaries, survival-curve detail in this discussion, or information about completeness of outcome ascertainment. The reported Cox model depends on a proportional-hazards assumption, but diagnostics or assessment of that assumption are not reported in the registry record.

The record lists two primary-comparison log-rank p-values, 0.633 and 0.704, but explains neither their difference nor the analysis associated with the latter entry. The reported hazard ratio and confidence interval are available for one of the entries. The original statistical analysis plan, sample-size justification, interim analyses, missing-data methods, censoring rules and any testing hierarchy are not reported in the registry record.

Several secondary outcomes are listed, but results for all of them are not reported in the registry record. The subgroup estimates have confidence intervals that include 1, and no interaction analysis or multiplicity adjustment is reported. As a result, subgroup patterns should not be treated as confirmatory evidence of different effects. Safety data are reported by phase, with phase-specific counts and serious-event terms; the term percentages do not have arm attribution or denominators specified in the summarized record. These limitations constrain interpretation beyond the posted primary OS comparison.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study of ABT-414 in Participants With Newly Diagnosed Glioblastoma (GBM) With Epidermal Growth Factor Receptor (EGFR) Amplification — NCT02573324.
• Registry record last updated 2023-05-11; status last verified 2023-05.
• Results first posted 2023-05-11.
• 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

NCT02573324 is a completed randomized, placebo-controlled Phase 3 study of depatuxizumab mafodotin with radiation and TMZ versus placebo with radiation and TMZ in newly diagnosed EGFR-amplified glioblastoma. The study used a parallel-group design and triple masking. A separate open-label Phase 1 sub-study assessed depatuxizumab mafodotin in participants with mild or moderate hepatic impairment.

The primary endpoint was OS, defined as time from randomization to death from any cause. In the Full Analysis Set, median OS was 18.7 months in the placebo arm and 18.9 months in the depatuxizumab mafodotin arm. The reported hazard ratio for placebo versus depatuxizumab mafodotin was 1.02 (95% CI 0.82 to 1.26; p=0.633). The confidence interval includes 1 and leaves uncertainty about the direction and magnitude of the relative hazard.

The registry also lists subgroup OS and other secondary outcomes, but not all secondary results or analysis-plan details are reported in the registry record. The primary estimate is therefore best read together with its uncertainty and the stated limits of the registry reporting.