A Phase 3 Study to Evaluate the Safety, Tolerability, and Immunogenicity of Multiple Production Lots and Dose Levels of BNT162b2 RNA-Based COVID-19 Vaccines Against COVID-19 in Healthy Participants: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Phase 3 Study to Evaluate the Safety, Tolerability, and Immunogenicity of Multiple Production Lots and Dose Levels of BNT162b2 RNA-Based COVID-19 Vaccines Against COVID-19 in Healthy Participants (NCT04713553), a randomized phase 3 study. The primary outcome is Geometric Mean Ratios (GMRs) of Full-Length S-Binding Immunoglobulin G (IgG) Concentrations Between Individual US Lots 1, 2, and 3 at 1 Month After Dose 2: Primary Study.

2. Clinical Question

3. Trial Design

This was a Phase 3, randomized, parallel study with triple masking recorded in the registry. The study description also calls it observer-blind. The registry record does not explain how these masking labels were implemented, who was masked, or how randomization was conducted. Total enrollment was 1574, and the study status is completed. The recorded start date is 2021-02-15 and the primary completion date is 2021-07-22.

The primary study evaluated BNT162b2 in healthy participants 12 through 50 years of age. Its planned vaccine schedule consisted of two doses separated by 21 days. The study description includes 30-microgram doses from four manufacturing lots, including three US lots and an EU lot, and a 20-microgram dose from a US lot corresponding to one of the US 30-microgram lots. The primary endpoint analysis is narrower: it compares the three US 30-microgram lots.

A booster study evaluated BNT162b2 and BNT162b2.B.1.351 as 30-microgram booster doses in healthy participants 18 through 50 years of age. The description says the booster was administered approximately 3 months after dose two. The posted primary endpoint does not compare the booster groups; its outcome is full-length S-binding IgG after the primary two-dose schedule.

The registry reports 52 outcome measures and 5 analyses. It does not report the allocation ratio, the number assigned to each group for the full study, or a sample-size justification. Total enrollment should not be read as the number included in the primary immunogenicity analysis: the analysis population had additional dose, timing, and sample-result criteria.

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: Total number of participants enrolled in study and assigned to study intervention were 1574, however, only 1573 participants received study intervention.

Only the first 6 of 7 groups are shown.

Primary Study (2 Months)

Reasons for not completing this period:

Booster Study (1 Month)

Analysis population for the primary outcome, as defined in the registry: Evaluable immunogenicity population: Participants who received 2 doses of vaccine in primary study, with Dose 2 received within predefined window, had at least 1 valid immunogenicity result from blood sample collected within an appropriate window at 1 month after Dose 2, were negative for both SARS-CoV-2 test (RT-PCR and N-blinding antibody) during study and had no other protocol deviations determined by clinician. Here, "N"= participants evaluable for this outcome measure.

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: Safety Population comprising of all randomized participants who received at least 1 dose of the study intervention were included.

6. Primary Endpoint

The registry lists the primary outcome as Geometric Mean Ratios (GMRs) of Full-Length S-Binding Immunoglobulin G (IgG) Concentrations Between Individual US Lots 1, 2, and 3 at 1 Month After Dose 2: Primary Study, measured over 1 Month after Dose 2.

The primary endpoint is the geometric mean ratio of full-length S-binding immunoglobulin G (IgG) concentrations between individual US lots 1, 2, and 3, measured one month after dose two. The registry describes calculating geometric mean concentrations for each lot and then reporting ratios between those concentrations. The posted pairwise results are the lot comparisons, rather than a single overall measure of immunogenicity.

A geometric mean is a measure of the central level of measurements that accounts for their multiplicative scale. A ratio compares the central levels in two groups: a ratio of 1 represents equal geometric means, while a ratio above or below 1 indicates a higher geometric mean in the numerator or denominator group, respectively. This is a general explanation of the statistic, not an additional trial result.

Antibody concentration is used as an immunogenicity measure: it quantifies an immune response to the vaccine antigen. It is suited to comparing measured antibody levels across manufacturing lots, but it is not itself a clinical endpoint such as COVID-19 incidence. The registry says assay results below the lower limit of quantitation were set to 0.5*LLOQ. The rationale for that convention and any sensitivity analyses are not reported in the registry record. A clinically meaningful difference threshold or an equivalence margin is also not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary analysis uses geometric mean concentrations (GMCs) and geometric mean ratios (GMRs). These statistics are common for measurements such as antibody concentrations, which are naturally interpreted on a multiplicative scale. In general, a geometric mean summarizes the central level after accounting for the scale of the measurements, and a ratio expresses the relative level between groups. This general explanation does not specify the exact computational procedure used in this trial; the registry does not report that procedure in detail.

What does a geometric mean ratio express?

A GMR compares the GMC for one lot with the GMC for another lot. A value of 1 corresponds to equal GMCs. A value above 1 indicates a higher GMC for the numerator group, and a value below 1 indicates a lower GMC for that group relative to the denominator. The reported ratios are 1.01 for US lot 1 versus US lot 2, 0.93 for US lot 1 versus US lot 3, and 0.92 for US lot 2 versus US lot 3. These are relative comparisons of antibody concentrations, not differences in infection risk.

How should a confidence interval be read?

A confidence interval conveys uncertainty about an estimated quantity. In general, a narrower interval indicates greater precision than a wider interval, while the interval endpoints show a range of values compatible with the estimate under the analysis assumptions. The registry reports 95% confidence intervals for the GMRs: 0.91 to 1.13, 0.83 to 1.04, and 0.82 to 1.03. Each includes 1. This observation describes the intervals; it does not by itself establish equivalence or determine whether a result meets a prespecified criterion.

The registry labels these as equivalence comparisons, but does not state the equivalence margin or the method used to construct the intervals. The statistical method is not reported in the registry record. Without the margin and method, readers cannot independently verify the equivalence decision from the label alone. The registry also does not report p-values for the comparisons.

Why does the lower limit of quantitation matter?

Laboratory assays have a lower limit of quantitation (LLOQ), below which a result cannot be quantified with the assay’s stated precision. The registry says results below the LLOQ were set to 0.5*LLOQ. Such a convention permits values below the assay threshold to be included in a summary, but it also affects the resulting estimates. This is a general statistical consideration. The registry does not state how many results were below the LLOQ, whether other substitutions were examined, or whether a sensitivity analysis was conducted.

What does the analysis population represent?

The analysis was based on an evaluable immunogenicity population rather than all enrolled participants. The record describes participants who received two primary-study doses, had dose two within a predefined window, and had at least one valid immunogenicity result from an appropriately timed blood sample one month after dose two. It also begins to specify serostatus criteria, but that portion of the population definition is incomplete in the registry record. The record does not report the full exclusion flow or the number excluded for each reason. Consequently, the results describe the defined evaluable population, not necessarily every randomized or enrolled participant.

Restricting an analysis to participants with valid, appropriately timed measurements can make the outcome definition more consistent. It can also limit interpretation if inclusion depends on factors related to the outcome or differs between groups. Assessing that possibility requires information about missing samples and exclusions, which is not reported in the registry record.

What can the design contribute?

Randomization is generally used to reduce systematic differences between assigned groups, and masking can reduce bias in treatment administration or outcome assessment. These are general features of trial design. The registry records randomization, parallel allocation, and triple masking, and describes the study as observer-blind. It does not give the allocation procedure, concealment details, or the practical masking arrangements. Those details therefore cannot be assessed from the registry record.

The study includes several manufacturing lots, dose levels, and booster groups, while the primary endpoint is restricted to three US lots at the 30-microgram dose. The registry reports 52 outcome measures and 5 analyses, but a multiplicity adjustment or hierarchy of tests is not reported. When multiple outcomes or comparisons are examined, the chance of at least one apparently notable result can increase; methods for managing that issue depend on the analysis plan. No particular adjustment or prespecified hierarchy can be inferred from the registry information.

The registry record does not report a sample-size justification or detailed statistical analysis plan. Sample-size illustrations based on observed effects, if considered alongside these results, are planning exercises rather than evidence about the trial’s original design or its observed power. Their assumptions do not replace a protocol-based justification, which is not reported in the registry record.

10. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary analysis as geometric mean ratios for the three pairwise comparisons among US lots 1, 2, and 3. The posted estimates are 1.01, 0.93, and 0.92, with 95% confidence intervals of 0.91 to 1.13, 0.83 to 1.04, and 0.82 to 1.03, respectively. The registry labels these analyses as equivalence comparisons.

No discrepancy is flagged between the sponsor-reported result and the available re-calculation. The registry record does not report a separate recalculated GMR, so there is no distinct recalculated estimate to compare numerically with the sponsor’s estimate. The descriptive GMCs are reported as 6300, 6232, and 6775 for the three groups; those summaries should not be mistaken for a recalculation of the pairwise analysis.

The statistical method, p-values, equivalence margin, and decision rule are not reported in the registry record. As a result, the posted estimates and confidence intervals can be described, but the equivalence conclusion cannot be independently checked against a specified margin from the registry information alone.

11. Multiplicity, Missing Data and Analysis Populations

The study covers multiple outcomes and comparisons. The registry reports 52 outcome measures and 5 analyses, and lists secondary immunogenicity outcomes involving full-length S-binding IgG and SARS-CoV-2 neutralizing titers. The primary endpoint itself involves three pairwise lot comparisons. A multiplicity procedure, such as an adjustment to account for multiple testing, is not reported in the registry record. A prespecified hierarchy that would indicate which outcomes or comparisons were confirmatory is also not reported.

Without those details, the reported confidence intervals should be read as belonging to their stated comparisons, rather than as evidence that a family-wide error rate was controlled. This is a general consideration when a study examines multiple outcomes or comparisons; it does not establish that any particular result is spurious. The registry does not report interim analyses or stopping rules for the immunogenicity comparisons.

The primary analysis population is described as evaluable for immunogenicity, with requirements concerning receipt and timing of doses and availability of valid blood-sample results. The record’s population description also begins a serostatus condition but is incomplete. The full criteria, missing-sample counts, reasons for exclusion, and handling of missing results beyond the LLOQ convention are not reported in the registry record. The number analyzed for the primary endpoint is reported by lot, but the record does not explain the flow from enrollment to that population.

These gaps matter because analysis populations and missing-data rules define what the estimates represent. The reported endpoint summarizes participants meeting the evaluability criteria; it does not automatically generalize to all enrolled participants. The registry does not give enough detail to assess whether exclusions or missing results were balanced across groups or related to antibody levels.

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: migrainosus (0.6%), abortion spontaneous (0.3%).

The registry lists serious adverse events and other adverse events by study group. For serious adverse events, the counts affected were 0 of 351 for US lot 1 at 30 micrograms, 0 of 352 for US lot 2 at 30 micrograms, 1 of 346 for US lot 3 at 30 micrograms, 1 of 173 for the EU lot at 30 micrograms, and 0 of 351 for the US lot at 20 micrograms. For the booster groups, the counts were 0 of 31 for BNT162b2 and 0 of 31 for BNT162b2.B.1.351. Deaths were recorded as 0 in each group.

The registry lists other adverse events affecting 341 of 351, 336 of 352, 335 of 347, 165 of 173, and 342 of 351 participants in the primary-study groups, respectively. In the booster groups, the corresponding counts are 30 of 31 and 31 of 31. These counts are descriptive; they do not establish that events were caused by the vaccine or show comparative risk after adjustment. The registry also lists serious-event terms Migrainosus at 0.6% and Abortion spontaneous at 0.3%. The record does not provide enough context in this summary to infer event causality or compare event rates statistically across groups.

Safety groups and the primary immunogenicity analysis population are not interchangeable. The safety counts have their own denominators, and the registry does not explain all analysis-population details in this summary. Rare serious events and zero recorded deaths should be interpreted as counts within the reported groups, not as proof that an event cannot occur.

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

The results describe measured full-length S-binding IgG concentrations one month after dose two among evaluable participants assigned to three US manufacturing lots of 30-microgram BNT162b2. The reported geometric mean ratios are close to 1, and their 95% confidence intervals all include 1. This indicates that the estimates are compatible with equal geometric mean concentrations as well as with some differences between lots, within the ranges represented by the intervals.

The registry labels the analyses as equivalence comparisons, but does not report the equivalence margin or decision rule. A conclusion of equivalence depends on demonstrating that uncertainty falls within an acceptable prespecified range, not simply on observing a ratio near 1 or a confidence interval that includes 1. The registry record therefore does not allow the equivalence criterion to be independently evaluated.

These findings address an immunogenicity measure, not clinical protection. The endpoint is antibody concentration, and the registry does not report a COVID-19 incidence comparison for this primary analysis. The results do not establish that the lots are identical in all immune responses, that their clinical effects are the same, or that the findings apply to groups outside the stated study populations.

The results also depend on the evaluable immunogenicity population and the assay convention for results below the LLOQ. The population description is incomplete in the registry record, and details on missing measurements and exclusions are not reported. The results should therefore be understood as summaries for the reported analysis, with limits on what can be inferred about participants who were not included.

14. Why This Trial Matters Statistically

This trial’s primary statistical question concerns consistency of an immune-response measure across manufacturing lots. That is distinct from asking whether a vaccine reduces clinical events. The distinction matters because the endpoint, comparison, and interpretation must match the question being asked. Here, the reported GMRs compare concentrations of full-length S-binding IgG after the primary schedule; they are not estimates of relative COVID-19 risk.

Geometric mean ratios are useful for comparing measurements on a multiplicative scale, while confidence intervals communicate the precision of those comparisons. The reported intervals include 1, but the registry’s equivalence label cannot be fully assessed without the margin and decision rule. This illustrates why a statistical label, estimate, and interval need to be read together with the prespecified criterion that defines a meaningful conclusion.

The analysis also highlights how population definitions shape a result. The endpoint is calculated in an evaluable immunogenicity population with dose and sample-timing requirements, not simply in everyone enrolled. Understanding those requirements, and how missing data were handled, is essential when deciding how far the estimate can be generalized. The registry record does not give all the information needed to examine those issues.

Finally, the study has several lots, dose levels, booster groups, and outcome measures, while the posted primary endpoint focuses on a specific subset of comparisons. The absence of a reported multiplicity plan or detailed statistical analysis plan limits assessment of how the primary endpoint fits within the full set of analyses. These are practical lessons in interpreting a complex trial: the statistical question, analysis population, interval, and decision criterion all matter.

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 gives the primary endpoint definition and reported GMRs, but several details needed for a fuller statistical assessment are not reported. These include the statistical method used to estimate the ratios and confidence intervals, the equivalence margin, the decision rule, and p-values. The record labels the comparisons as equivalence analyses, but that label does not reveal the criterion used to reach the conclusion.

The evaluable immunogenicity population description is incomplete. It specifies dose receipt, timing, and valid-result requirements, and begins a serostatus condition, but the remainder of that condition is not reported in the registry record. Counts and reasons for exclusion, missing sample handling, and sensitivity analyses are also not reported. The assay convention for below-LLOQ values is stated, but the number of affected results is not.

The study includes multiple outcome measures, lots, dose levels, and booster groups. A multiplicity plan, outcome hierarchy, interim analysis plan, and detailed statistical analysis plan are not reported in the registry record. The record also does not report a sample-size justification, so the basis for the planned enrollment cannot be assessed. The total enrollment does not substitute for the endpoint-specific analysis counts.

The primary endpoint is an antibody concentration measure at one post-dose time point. It does not directly measure clinical COVID-19 outcomes, and the registry does not report a clinical efficacy comparison for this endpoint. The posted group summaries and ratios also do not show individual-level distributions. These limitations restrict independent evaluation of the analysis and the scope of conclusions that can be drawn from the registry results.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Phase 3 Study to Evaluate the Safety, Tolerability, and Immunogenicity of Multiple Production Lots and Dose Levels of BNT162b2 RNA-Based COVID-19 Vaccines Against COVID-19 in Healthy Participants — NCT04713553.
• Registry record last updated 2022-12-22; status last verified 2022-11.
• Results first posted 2022-12-22.
• 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

NCT04713553 is a completed Phase 3 randomized, parallel study with triple masking recorded in the registry and an observer-blind study description. It enrolled 1574 healthy participants and evaluated primary-study vaccine lots and dose levels, as well as booster candidates in separate groups. The primary endpoint compares full-length S-binding IgG geometric mean concentrations one month after dose two across three US lots of 30-microgram BNT162b2.

The reported GMRs are 1.01, 0.93, and 0.92, with 95% confidence intervals of 0.91 to 1.13, 0.83 to 1.04, and 0.82 to 1.03. The registry labels the comparisons as equivalence analyses, but the margin and decision rule are not reported in the registry record. The endpoint is an immunogenicity measure, and the registry does not report a clinical efficacy result for this comparison.