VELOCITY: An Anthrax Vaccine Clinical Study: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for VELOCITY: An Anthrax Vaccine Clinical Study (NCT03877926), a randomized phase 3 study. The primary outcome is Geometric Mean Titer (GMT) of Toxin Neutralizing Antibody (TNA) 50% Neutralization Factor (NF50) at Day 64.

2. Clinical Question

3. Trial Design

VELOCITY was a Phase 3, randomized, double-blind, parallel-group trial of AV7909 and BioThrax in healthy adults. The registry lists quadruple masking. Which groups of people were masked is not reported in the registry record. The study was designed to assess lot consistency, immunogenicity, and safety, and the registry lists anthrax as the condition.

Participants were randomized among four groups: AV7909 Lot 1, AV7909 Lot 2, AV7909 Lot 3, and BioThrax. The arm descriptions state an allocation scheme of 2:2:2:1. The first three groups received different consecutively manufactured AV7909 lots; the BioThrax group received one lot. The registry reports enrollment of 3689. The primary outcome population consists of participants meeting protocol-defined immunogenicity criteria, with group counts reported separately from overall enrollment.

The trial began on 2019-03-11, reached primary completion on 2020-08-06, and is listed as completed. The registry describes the primary endpoint as Day 64 TNA NF50 GMT, with the timeframe seven weeks after the second AV7909 vaccination. It also lists secondary outcomes for a Day 29 antibody threshold, adverse events, adverse events of special interest, and solicited systemic and injection-site reactogenicity events.

The registry record does not report the full protocol, the randomization sequence-generation or concealment procedures, the details of masking implementation, or the planned statistical methods beyond the stated primary equivalence criterion. It also does not report a sample-size justification or planned interim analyses. Those details should not be inferred from the completed status or the outcome counts.

4. Participant Flow and Accounting

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

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Participants that met protocol-defined immunogenicity criteria from AV7909 Lot 1 group (n=835), AV7909 Lot 2 group (n=854), AV7909 Lot 3 group (n=854) and BioThrax group (n=430) are included.

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: Intent-to-Treat Population (i.e., number of randomized participants) is used for baseline analysis.

6. Primary Endpoint

The registry lists the primary outcome as Geometric Mean Titer (GMT) of Toxin Neutralizing Antibody (TNA) 50% Neutralization Factor (NF50) at Day 64, measured over Day 64 (seven weeks after second AV7909 vaccination).

The primary endpoint is the geometric mean titer of toxin-neutralizing antibody, measured as TNA NF50 at Day 64. The registry defines the assessment time as seven weeks after the second AV7909 vaccination. A titer summarizes the antibody response on a scale for which the geometric mean is commonly used; unlike an arithmetic mean, it is based on averaging on a logarithmic scale and returning to the original scale. This can be useful when measurements are positive and vary multiplicatively.

The registry reports Day 64 GMT estimates and 95% confidence intervals for the three AV7909 lots and BioThrax. Its primary lot-consistency question is not whether each lot has an identical estimate. Instead, it evaluates the ratios between the three AV7909 lot GMTs. The stated equivalence requirement is that the 95% confidence interval for each lot-to-lot ratio falls within the margins of 0.5 and 2.0.

Equivalence margins define a range of differences considered compatible with the trial’s stated consistency criterion. A confidence interval fully inside the margins supports that criterion for the comparison being assessed; an interval extending beyond a margin would not meet it. The registry reports three lot-to-lot ratios and their confidence intervals, each of which can be judged against the prespecified margins as described in the record.

The endpoint is an immune-response measure, not a direct measure of anthrax cases or prevention. The registry record does not report a clinical disease endpoint in these results. Thus the antibody endpoint addresses immunogenicity and consistency, rather than demonstrating clinical effectiveness against anthrax.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary measure is a geometric mean titer. In general, a geometric mean is calculated by transforming measurements to a logarithmic scale, averaging there, and transforming back. This approach is useful for measurements that behave multiplicatively and may be strongly skewed on their original scale. The registry reports the GMTs and their 95% confidence intervals, but does not report the precise estimation procedure, any transformations beyond what the endpoint implies, or whether the analysis adjusted for covariates.

What does a confidence interval communicate?

In general, a confidence interval describes the uncertainty associated with an estimated quantity under the model and analysis assumptions. A narrower interval indicates greater precision than a wider interval, while an interval’s location describes the values compatible with the data and those assumptions. A confidence interval is not a range containing a fixed proportion of individual participants’ measurements, and it does not guarantee that the true value lies within its particular bounds.

For this endpoint, the record reports 95% confidence intervals for each group GMT and for the ratios between AV7909 lots. The group intervals convey uncertainty about the GMT estimates. The ratio intervals address the consistency question more directly, because the prespecified criterion concerns the interval for each lot-to-lot ratio rather than whether the separate GMT intervals overlap.

How does an equivalence test differ from a difference test?

In general, a conventional test for a difference asks whether the data are inconsistent with an exact null of no difference. An equivalence analysis instead asks whether the plausible range of the comparison lies within bounds judged acceptable for the purpose of the study. The registry states margins of 0.5 and 2.0 for the AV7909 lot GMT ratios. For this criterion, the relevant evidence is whether each reported 95% confidence interval stays within both boundaries.

The three reported ratio intervals—0.93 to 1.10, 0.96 to 1.14, and 0.95 to 1.13—are each inside the stated margins. This supports the registry’s lot-consistency conclusion according to its stated rule. A ratio near 1 indicates similar estimated GMTs on the ratio scale, but the conclusion relies on both the estimate and its interval. The registry does not report the method used to construct these intervals or the detailed statistical test procedure.

Why is the analysis population important?

The primary endpoint population includes participants who met protocol-defined immunogenicity criteria. The record reports group counts of 835, 854, 854, and 430 for AV7909 Lot 1, AV7909 Lot 2, AV7909 Lot 3, and BioThrax, respectively. These are not the same as the randomized allocation counts described in the safety data. Different outcomes can use different analysis populations because measurements may be available or eligible for analysis for different participants. The registry does not report the full participant flow, the number missing a Day 64 measurement, or the detailed criteria used to define the immunogenicity population.

When an analysis population is selected using criteria set in advance, those criteria can help focus the analysis on participants with relevant measurements. However, exclusions or missing measurements can affect how representative the result is of all randomized participants. Without a detailed flow and missing-data methods, readers cannot assess the extent or direction of any such effect. The posted confidence intervals quantify uncertainty within the analyzed data under the chosen method; they do not automatically account for every uncertainty arising from missing or excluded outcomes.

How should the other outcome comparisons be read?

For binary outcomes, a risk difference compares the proportion with an event in one group with the proportion in another. In general, a confidence interval for a risk difference describes uncertainty in that contrast; if it includes zero, the data are compatible with no difference as well as with values on either side. A relative risk and an odds ratio express contrasts on different scales and should not be confused with absolute differences. The registry’s secondary results include participant counts and percentages, while the recalculated comparisons are simple unadjusted comparisons from posted summary statistics.

Those recalculations are not necessarily the sponsor’s planned analyses. They do not incorporate adjustment for other variables, account for multiple comparisons, or establish causation. The registry does not report methods for these comparisons. Unadjusted p-values, where included in the posted analysis material, should therefore be treated as descriptive evidence rather than as a complete inferential plan.

Why does over-dispersion matter in general?

For count outcomes, over-dispersion means the observed variation exceeds what a simple model assumes. If it is present and ignored, standard errors and confidence intervals can be misleading. That general concern does not establish that over-dispersion occurred in this trial. The primary endpoint is a geometric mean titer, and the registry does not report a count-model analysis or information about over-dispersion for the primary result.

The registry’s stated equivalence bounds are the interpretive framework for the primary lot comparisons. Its statistical analysis plan, detailed model specifications, multiplicity procedures, and missing-data procedures are not reported in the registry record. Those omissions limit independent assessment of how the posted results were generated, but do not change what the listed estimates and intervals show against the stated margins.

10. Analyses Reported by the Sponsor

The registry reports sponsor analyses for the primary AV7909 lot-to-lot comparisons. These are ratios of geometric mean titers with 95% confidence intervals. The intervals are 0.93 to 1.10 for Lots 1 and 2, 0.96 to 1.14 for Lots 1 and 3, and 0.95 to 1.13 for Lots 2 and 3. Each interval falls within the registry’s stated equivalence margins of 0.5 and 2.0, so each comparison meets the stated lot-consistency criterion.

No p-values are listed for these equivalence analyses. A p-value is not needed to read the criterion as stated: the key question is whether the full confidence interval is contained within the equivalence margins. The registry does not identify the statistical method used to derive the intervals or provide a fuller analysis plan.

The discrepancy indicator is false, and the primary comparison results are sponsor-reported rather than recalculated. There is therefore no conflicting recalculation to resolve for the primary endpoint. Where the posted material includes recalculated comparisons for secondary outcomes, those are simple unadjusted calculations from summary statistics and are not substitutes for a sponsor analysis. If a discrepancy between a recalculation and the sponsor’s own analysis were present, the sponsor’s own analysis would take precedence; no such discrepancy is indicated for this trial’s primary result.

These conclusions concern lot consistency for an antibody endpoint in the stated immunogenicity population. They should not be expanded into claims about clinical protection or a formal AV7909-versus-BioThrax treatment effect, neither of which is established by these lot-to-lot analyses.

11. Multiplicity, Missing Data and Analysis Populations

The primary endpoint involves three lot-to-lot comparisons, and the registry states an equivalence criterion for each. It reports all three ratio confidence intervals and each lies within the margins of 0.5 and 2.0. The record does not report whether the statistical plan included a multiplicity procedure for the set of comparisons, how the overall family of hypotheses was defined, or whether any additional primary endpoints were planned. The conclusion is therefore tied to the stated criterion and reported intervals, not to an unreported multiplicity strategy.

There are also several secondary outcomes, and some include comparisons across multiple groups. When many outcomes or comparisons are examined, the likelihood of an apparently unusual result arising by chance can increase. This is why unadjusted secondary p-values, if considered, should not be treated as though they came from a single prespecified test. The registry record does not report an adjustment procedure, a testing hierarchy, or which secondary outcomes were confirmatory versus exploratory.

Missing-data methods are not reported in the registry record. The primary endpoint population is described as participants meeting protocol-defined immunogenicity criteria, but the detailed criteria and participant flow are not stated in the outcome description. The record does not say how missing Day 64 titers were handled, whether any values were imputed, or how protocol deviations affected inclusion. It also does not provide a full account of withdrawals or missing secondary outcome observations.

These omissions matter because the reported confidence intervals describe uncertainty under the analysis used; they do not reveal whether missingness could have changed the estimates or whether exclusions affected the representativeness of the analysis population. The record gives outcome-specific group counts, which should be read as the populations analyzed for those outcomes rather than assumed to equal overall enrollment. The randomized design is important, but randomization alone does not resolve uncertainty introduced by missing outcomes or analysis-population selection.

The registry also does not report a sample-size justification, interim analyses, or the full statistical analysis plan. No conclusions about those design choices should be inferred from the posted enrollment or the completed trial status. A fuller protocol and analysis plan would be needed to assess these aspects in detail.

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 1% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: abortion spontaneous (0.3%), lumbar spinal stenosis (0.2%), completed suicide (0.2%), lymphadenitis (0.2%), pneumonia (0.2%).

The registry’s serious adverse event figures are counts of affected participants and the corresponding participants at risk in each group. It reports 21 serious adverse events among 1050 at risk for AV7909 Lot 1, 16 among 1053 for Lot 2, 21 among 1048 for Lot 3, and 4 among 533 for BioThrax. Deaths are reported separately: 2 in each AV7909 group and 0 in the BioThrax group. These are descriptive counts; the registry record does not provide a comparative statistical analysis for them in the results.

The registry also lists other adverse event counts: 194 among 1050 for AV7909 Lot 1, 199 among 1053 for Lot 2, 204 among 1048 for Lot 3, and 171 among 533 for BioThrax. The record does not provide enough detail in these figures to determine the cause, timing, severity distribution, or relationship to vaccination. Counts alone do not establish that an event was caused by a vaccine or that differences between groups are meaningful.

Serious adverse events are not interchangeable with solicited reactogenicity events, which were collected in specified post-vaccination windows. The registry lists serious event terms and percentages, including Abortion spontaneous at 0.3%, and Lumbar spinal stenosis, Completed suicide, Lymphadenitis, Pneumonia, and Sialoadenitis at 0.2% each. These entries should be read as registry-listed terms, not as evidence of causation. The record does not report enough detail to assess event-level clinical context or a formal safety comparison.

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

The primary result supports consistency among the three AV7909 lots on the Day 64 TNA NF50 geometric mean titer endpoint, according to the equivalence rule stated in the registry. Each reported 95% confidence interval for a lot-to-lot ratio falls within the margins of 0.5 and 2.0. The result is about whether the lots met that specified immunogenicity consistency criterion in the analysis population.

It does not mean that the three lots produced identical responses in every participant. Group GMTs are summaries, and confidence intervals express uncertainty around estimates of group-level quantities. Individual responses can vary, and the posted GMTs do not show the full distribution of antibody measurements. The registry record does not provide individual participant data or detailed distribution summaries in the results.

The result also does not establish that AV7909 prevents anthrax disease or is clinically superior to BioThrax. The primary endpoint is an antibody titer, and the registry’s stated primary analysis compares AV7909 lots with one another. Although the BioThrax GMT is reported, the record does not present a formal primary comparative treatment analysis between AV7909 and BioThrax. The study’s postexposure prophylaxis context should not be confused with a clinical disease outcome in these posted results.

Nor does meeting an equivalence criterion establish that every safety outcome is equivalent. Safety requires assessment of event types, severity, timing, exposure, and uncertainty, and the registry presents descriptive event counts without a reported formal comparison in the details. The absence of a reported difference is not evidence of no difference, especially when outcomes are uncommon or the analysis methods are not stated.

The conclusions are also bounded by the analysis populations and reporting. The primary endpoint uses participants meeting protocol-defined immunogenicity criteria, and the record does not provide detailed missing-data methods or a full participant flow. The posted estimates are informative about the measurements analyzed, but the registry record alone does not resolve how missing observations, protocol deviations, or the choice of analysis population might affect broader interpretation.

14. Why This Trial Matters Statistically

Lot consistency is a distinct statistical question from whether an intervention differs from a comparator. A manufacturing program may need evidence that separate lots produce sufficiently similar immune responses. This trial makes that question explicit by using three AV7909 lots and a ratio-based equivalence criterion with stated margins. The analysis therefore focuses on the precision of the lot comparisons, not simply on whether group estimates appear close.

The confidence intervals are central to that reasoning. The observed ratios alone would not show how precisely the lot differences were estimated. The reported intervals quantify uncertainty, and the fact that each interval lies inside the stated bounds is the evidence that supports the registry’s consistency conclusion. This illustrates why equivalence analyses require attention to both the estimate and its uncertainty.

The design also separates a primary immunogenicity outcome from a wider set of secondary outcomes, including solicited reactogenicity and adverse events. That distinction matters: findings across different endpoints answer different questions, and unadjusted secondary comparisons should not be allowed to displace the prespecified primary question. The registry’s reporting permits interpretation of the primary criterion, while leaving some aspects of the analysis plan and secondary outcome inference unclear.

Finally, the trial is a reminder that statistical conclusions have a defined scope. The primary result concerns antibody GMT consistency among AV7909 lots in a specified analysis population. It does not itself establish clinical protection, nor does it turn descriptive safety counts into comparative safety evidence. Keeping the endpoint, population, comparison, and uncertainty aligned is essential for a sound reading of the registry results.

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 record leaves several trial-specific details unavailable. It does not report the full protocol or statistical analysis plan, the method used to estimate the primary confidence intervals, whether covariates were included, or how multiplicity was addressed. Without these details, the primary results can be evaluated against the stated equivalence margins, but the full analytical choices cannot be independently assessed.

The primary outcome population is defined by protocol-based immunogenicity criteria, but the record does not describe those criteria in detail. It reports group counts for that population, while the safety outcomes use different at-risk counts. A complete participant flow, including the reasons participants were not included in each analysis, is not reported in the registry record. Missing Day 64 values and the handling of missing outcome data are also not described.

The registry reports an antibody endpoint rather than a clinical anthrax outcome. It therefore does not establish how the measured immune responses relate to disease prevention in this trial. The posted primary result assesses consistency among AV7909 lots; the BioThrax GMT is included descriptively, but a formal primary comparison between vaccine groups is not reported.

The safety information is limited by the level of detail available. Serious adverse event and other adverse event counts are reported, but the material does not provide event-level timing, severity, relatedness, or a formal comparative analysis. The listed serious event terms do not establish causality. Results for all listed secondary outcomes are not detailed, and the registry does not report a multiplicity strategy for the many outcomes and potential comparisons.

Finally, the record does not report a sample-size justification, interim analyses, or details of randomization implementation and masking roles. These should be treated as reporting limitations, not evidence that the procedures were absent. Further protocol and analysis documentation would be needed to address them.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: VELOCITY: An Anthrax Vaccine Clinical Study — NCT03877926.
• Registry record last updated 2026-05-06; status last verified 2026-04.
• Results first posted 2021-12-21.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

VELOCITY (NCT03877926) is a completed Phase 3 randomized, double-blind, parallel-group trial of AV7909 and BioThrax in healthy adults. The registry lists quadruple masking and enrollment of 3689. It began on 2019-03-11, with primary completion on 2020-08-06.

The primary outcome was Day 64 TNA NF50 geometric mean titer, assessed in participants meeting protocol-defined immunogenicity criteria. The registry reports three AV7909 lot-to-lot GMT ratios with 95% confidence intervals, all within the stated equivalence margins of 0.5 and 2.0. This supports the stated lot-consistency criterion for the antibody endpoint.

The results do not establish a clinical anthrax outcome or a formal primary comparison between AV7909 and BioThrax. The registry also reports secondary immunogenicity and reactogenicity data and descriptive adverse event counts. Statistical-plan details, missing-data methods, and results for some listed secondary outcomes are not reported in the registry record.