A Phase 3 Lot to Lot Consistency Study of Live Oral Cholera Vaccine, PXVX0200 in Healthy Adults: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Phase 3 Lot to Lot Consistency Study of Live Oral Cholera Vaccine, PXVX0200 in Healthy Adults (NCT02094586), a randomized phase 3 study. The primary outcome is Geometric Mean Ratio (GMR) at Day 11 for Vaccine Lots A and B.

2. Clinical Question

3. Trial Design

This was a Phase 3, randomized, quadruple-masked, parallel-group trial of PXVX0200 and placebo. Its stated primary goal was to compare three vaccine lots for manufacturing consistency. The record describes a single-dose vaccine intervention and a physiological saline placebo. Enrollment was 3146. The trial is marked completed, with a start date of 2014-05 and primary completion in 2015-02.

The randomized groups were Lot A, Lot B, Lot C, and placebo. The primary immunogenicity analysis focused on Day 11 serum vibriocidal antibody titers, comparing the vaccine lots with one another rather than comparing vaccine with placebo. The stated Immunogenicity Evaluable Population included randomized participants with evaluable Day 11 antibody results and no major protocol violations affecting immunogenicity. The number meeting that definition is not reported in the registry record.

Other listed outcomes included Day 11 seroconversion, antibody geometric mean titers and seroconversion at several follow-up days, and adverse events. The record lists four secondary outcomes and seven outcome measures, but does not report the full detail or results for every measure. Specific information about randomization procedures, allocation concealment, masking roles, follow-up completeness, and protocol deviations is not reported in the registry record. Those omissions limit assessment of how the planned design was implemented, even though the trial is identified as randomized and quadruple-masked.

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: Following assignment to either receive PXVX0200 or placebo, subjects in the PXVX0200 arm were randomized to receive Lot A, B or C. Subjects in each of these Arms/Groups add up to the number of subjects in the PXVX0200 (Lot A, B and C) Arm/Group.

Overall Study

Analysis population for the primary outcome, as defined in the registry: The Immunogenicity Evaluable Population (IEP) comprises all randomized subjects who had evaluable vibriocidal antibody results from Day 11 and had no major protocol violations that affected immunogenicity.

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 Geometric Mean Ratio (GMR) at Day 11 for Vaccine Lots A and B, measured over Day 11.

The primary endpoint was the geometric mean titer of serum vibriocidal antibody at Day 11, summarized through ratios between vaccine lots. A titer measures antibody activity in a serum sample under the assay conditions; the geometric mean provides a summary suited to measurements that can span a wide range and are often analyzed on a logarithmic scale. The registry describes taking base-10 logarithms of the titers, calculating the mean on that scale for each lot, and transforming the means back to the original scale.

The lot comparison uses a geometric mean ratio (GMR). A ratio near 1 represents similar geometric mean titers, while values farther from 1 represent a larger relative separation. This is an immunogenicity endpoint, not a direct clinical outcome. It was used to assess whether manufacturing lots produced antibody responses consistent with one another. The registry’s equivalence criterion required each pairwise ratio’s 95% confidence interval to fall within 0.67 to 1.5. Placebo GMT values were excluded from this primary analysis, so the primary question is about consistency among lots, not whether vaccine responses differ from placebo.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Why use a geometric mean?

Antibody titers are often analyzed on a logarithmic scale because their values may vary multiplicatively and span a broad range. On that scale, each observation is transformed before calculating a mean. Transforming the mean back produces a geometric mean titer (GMT), which summarizes the central response while respecting the scale on which the analysis was conducted. The registry describes this process using base-10 logarithms for the Day 11 vibriocidal antibody titers.

A geometric mean is not the same as an arithmetic average of the original titers. It reflects the average on the log scale, returned to the original measurement scale. The GMR compares two such geometric means. A GMR around 1 indicates that the two lot summaries are close relative to one another. The ratio is directional: the stated Lot A versus Lot B ratio expresses the Lot A GMT relative to the Lot B GMT. It is not an absolute difference in antibody units.

What does an equivalence interval test?

Equivalence testing asks whether the data rule out differences large enough to be considered unacceptable under prespecified margins. The registry states that the lot ratios had to be within +/- 50% of one another with 95% confidence, expressed as confidence-interval limits within 0.67 to 1.5. Thus, the decision depends on the full confidence interval, not only the point estimate. A point estimate inside the margins would not be enough if its interval extended beyond either bound.

For Lot A versus Lot B, the reported GMR is 0.92 and the 95% confidence interval is 0.78 to 1.08. Both interval limits fall within the stated bounds. This satisfies the stated interval condition for that reported comparison. It does not answer the Lot A versus Lot C or Lot B versus Lot C questions, whose numerical results are not reported in the registry record. A general statistical point is that an equivalence conclusion concerns acceptable similarity within chosen margins; it does not demonstrate exact equality.

Why are confidence intervals useful?

A confidence interval expresses uncertainty around an estimate under the statistical procedure used. It shows a range of values compatible with the data and method, rather than a guarantee that the true value lies inside that particular interval. For an equivalence analysis, the interval has a direct role in the decision: its endpoints are compared with the acceptable bounds. A narrower interval gives a more precise estimate, while a wider interval leaves more uncertainty. The confidence interval does not account for every possible limitation, such as measurement problems, missing outcomes, or bias from departures from the intended analysis population.

The registry gives 95% confidence intervals for the arm-level GMT estimates as well as for the reported GMR. These quantities answer different questions. The arm-level intervals describe uncertainty in each GMT; the GMR interval describes uncertainty in the relative comparison. Equivalence is judged using the ratio interval and the stated margins, not by comparing whether separate arm-level confidence intervals overlap.

Why distinguish the primary and secondary analyses?

The primary endpoint was designed to address manufacturing consistency among lots. Secondary outcomes broaden the description of antibody responses and adverse events, but each additional analysis creates another opportunity to observe an apparently notable result by chance. When many outcomes or comparisons are examined, unadjusted p-values do not control the chance of at least one false positive across the whole set. The registry does not report a multiplicity adjustment strategy, so secondary findings should be treated as less definitive than a prespecified primary analysis.

The primary plan described three pairwise lot comparisons, but the numerical result available in the record covers only Lot A versus Lot B. The record does not report the method used to account for the set of three tests or whether an adjustment was planned. Equivalence decisions for a family of comparisons can depend on how the family is defined and how the analysis is specified; details beyond the stated interval rule are not reported in the registry record.

What can a recalculation establish?

A recalculation from posted summary statistics is an unadjusted check based on the information available, not a replacement for the sponsor’s analysis. Summary statistics may not preserve details of the original analysis population, data handling, or model. In this trial, the registry reports the sponsor’s GMR for Lot A versus Lot B and does not flag a discrepancy. The sponsor’s stated analysis and reported result therefore remain the primary reference for interpreting that comparison.

Sample-size figures based on an observed effect are planning illustrations. They describe hypothetical future-study planning under stated assumptions; they are not the number needed to reproduce this trial and are not observed power. General statistical practice distinguishes such planning calculations from the analysis of the trial that has already taken place.

10. Analyses Reported by the Sponsor

The registry reports the sponsor’s Lot A versus Lot B GMR as 0.92, with a 95% confidence interval of 0.78 to 1.08. This interval falls within the stated equivalence bounds of 0.67 to 1.5 for that pairwise comparison. The posted arm-level GMT estimates are 9220 for Lot A and 10030 for Lot B, but those estimates do not replace the ratio-based equivalence test.

The registry does not flag a discrepancy between the sponsor analysis and the recalculation. No independent numerical result is reported in the record that would alter the sponsor’s GMR or confidence interval. The reported sponsor analysis should therefore be used for the Lot A–Lot B interpretation. This does not resolve the full primary question: the analysis plan described three pairwise lot tests, while numerical results for comparisons involving Lot C are not reported in the registry record.

The available result supports the stated equivalence criterion for the comparison it covers, not a broader conclusion about every planned lot comparison. The record’s numerical result and the scope of the primary analysis need to be kept distinct when describing the trial’s findings.

11. Multiplicity, Missing Data and Analysis Populations

The registry’s primary analysis description specifies three pairwise equivalence tests among the vaccine lots. Only the Lot A versus Lot B GMR and confidence interval are reported numerically. Results for Lot A versus Lot C and Lot B versus Lot C are not reported in the registry record. Without those results, the full set of stated primary comparisons cannot be assessed from the record.

The registry does not report whether the three primary tests used a multiplicity procedure, how any such procedure was defined, or whether the confidence intervals were adjusted. It also does not report an interim-analysis plan or a sample-size justification. These omissions should not be filled by assuming a particular protocol or analytical strategy. The stated equivalence margins and the reported interval can be interpreted for the reported pair, but details about family-wise error control or other testing choices remain unavailable.

Several secondary outcomes and time points are listed. When many outcomes are examined, the chance of at least one apparently unusual result can rise. An unadjusted p-value describes evidence for an individual comparison under its test assumptions; it does not, by itself, account for the full collection of analyses. The registry does not report a secondary multiplicity strategy, so those findings warrant less confirmatory weight than a clearly specified primary analysis.

The primary analysis population is defined as randomized participants with evaluable Day 11 vibriocidal antibody results and no major protocol violations affecting immunogenicity. The number included, the number excluded, and the reasons for exclusions are not reported in the registry record. Missing-data handling and sensitivity analyses are also not reported in the registry record. These details matter because a population limited to evaluable participants may differ from the full randomized population, and missing outcomes can affect both precision and interpretation. The record does not provide enough information to assess whether missingness or exclusions could have influenced the reported estimate.

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: colitis (0.3%), kidney infection (0.3%), tonsillitis (0.3%), post procedural haemorrhage (0.3%), appendicitis (0.2%).

The serious adverse event figures are counts among the participants listed as at risk in each arm. The record reports 7 serious adverse events among 926 participants at risk in Lot A, 9 among 928 in Lot B, and 4 among 935 in Lot C. For all PXVX0200 lots combined, it reports 20 among 2789 at risk; for placebo, it reports 3 among 350 at risk. These are registry counts, not evidence on their own that an event was caused by the assigned intervention.

The record lists one death for Lot A and one death across all PXVX0200 lots, and none for the other listed individual lots or placebo. The relationship of the death to treatment is not reported in the registry record. It also lists serious-event terms and percentages, including colitis, kidney infection, tonsillitis, post procedural haemorrhage, appendicitis, and pyelonephritis. The arm-specific distribution of those terms is not reported in the registry record.

Other adverse events are reported for the combined vaccine-lot group and placebo, but the record does not provide event-level detail for each vaccine lot in those figures. Interpretation is limited by the absence of further information about event definitions, follow-up duration, severity grading, causality assessment, and missing safety observations. Counts should be read descriptively, not as a complete comparison of safety or as evidence about causation.

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

The reported primary result indicates that the Day 11 vibriocidal antibody response ratio for Lot A versus Lot B met the registry’s stated equivalence interval criterion. The GMR is 0.92, and its 95% confidence interval, 0.78 to 1.08, lies inside the prespecified bounds of 0.67 to 1.5. This is evidence about the relative antibody response for that pair of lots under the reported analysis.

It does not establish that the two lots have identical responses. Equivalence means the observed uncertainty interval falls within the trial’s chosen range for acceptable similarity. Nor does this pairwise result establish equivalence between Lot A and Lot C or between Lot B and Lot C. Those numerical comparisons are not reported in the registry record, despite the stated plan for three pairwise tests.

The endpoint is immunogenicity, not a direct measure of clinical protection or disease prevention. The registry result does not show how the vaccine performs against cholera in routine use, and it does not establish the duration of protection. It also does not provide a basis for medical advice. Secondary antibody and seroconversion outcomes may add context, but they do not change the scope of the reported primary comparison.

The safety counts are descriptive and do not establish that an event was caused by the vaccine or placebo. The record lacks details needed for a full causal safety assessment. Any sample-size figures based on observed effects are planning illustrations for hypothetical studies, not observed power and not proof that this trial was adequately powered for every outcome.

14. Why This Trial Matters Statistically

This trial illustrates an important distinction in biostatistics: showing that groups differ is not the same task as showing that they are sufficiently similar. A conventional difference test asks whether the data provide evidence against equality. An equivalence analysis instead defines acceptable margins in advance and asks whether the confidence interval is contained within them. The interval, rather than a nonsignificant difference test, is central to the interpretation.

It also demonstrates why a result must be read at the level of the comparison actually reported. The trial’s primary goal concerned three manufacturing lots, and the analysis description specifies three pairwise tests. The available numerical result addresses only Lot A versus Lot B. That result can be interpreted against the equivalence margins, but it cannot stand in for the unreported comparisons involving Lot C.

Another useful statistical lesson is that endpoint choice determines the scope of a conclusion. Day 11 vibriocidal antibody GMTs address an immunogenicity question. Seroconversion, other time points, and adverse-event outcomes provide different kinds of information and have different sources of uncertainty. Their results should not be merged into a single claim or treated as interchangeable measures.

Finally, the record highlights the importance of reporting analysis populations, missing-data handling, and multiplicity plans. These details help readers assess how estimates were obtained and how much confidence to place in a collection of results. When they are not reported, a careful interpretation should stay close to the results and comparisons that are available.

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 principal limitation for interpreting the primary question is incomplete reporting of the planned pairwise results. The registry describes three lot comparisons, but provides a numerical GMR and confidence interval only for Lot A versus Lot B. The other comparisons are not reported in the registry record, so the consistency of all three lots cannot be evaluated from the posted primary results.

The Immunogenicity Evaluable Population is defined, but its size and the number of participants excluded from it are not reported in the registry record. Reasons for missing Day 11 results, handling of missing data, sensitivity analyses, and the extent of protocol deviations are also not reported in the registry record. This makes it difficult to judge how the analysis population relates to all randomized participants and whether exclusions could have affected the result.

The registry does not report the detailed statistical plan for multiplicity across the three primary comparisons or across secondary outcomes. It also does not report the sample-size justification, interim analyses, or other analysis-plan details needed to assess all design assumptions. No conclusions about those procedures should be inferred from the reported result alone.

Safety interpretation has further limits: event definitions, follow-up duration, event-level arm assignments for listed serious terms, causality assessments, and completeness of safety follow-up are not reported in the registry record. The available counts are therefore descriptive. More generally, antibody outcomes are not direct clinical endpoints, and the record does not establish how the measured responses relate to clinical outcomes. These limitations narrow the conclusions that can be drawn without negating the reported Lot A–Lot B equivalence interval.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Phase 3 Lot to Lot Consistency Study of Live Oral Cholera Vaccine, PXVX0200 in Healthy Adults — NCT02094586.
• Registry record last updated 2023-06-28; status last verified 2023-06.
• Results first posted 2021-07-15.
• 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

NCT02094586 is a completed Phase 3 randomized, quadruple-masked, parallel-group study of PXVX0200 in healthy adults. Enrollment was 3146. The primary goal was consistency across three vaccine lots, assessed through Day 11 serum vibriocidal antibody titers and pairwise geometric mean ratios. The registry reports a Lot A versus Lot B GMR of 0.92 with a 95% confidence interval of 0.78 to 1.08, within the stated equivalence bounds of 0.67 to 1.5.

The stated primary analysis included three lot comparisons, but numerical results for comparisons involving Lot C are not reported in the registry record. Secondary outcomes include seroconversion, antibody measures at other time points, and adverse events. Incomplete reporting of the analysis population, missing-data handling, and multiplicity methods limits assessment beyond the reported comparison.