Safety and Effectiveness of Live Zoster Vaccine in Anti-Tumor Necrosis Factor (TNF) Users (VERVE Trial): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety and Effectiveness of Live Zoster Vaccine in Anti-Tumor Necrosis Factor (TNF) Users (VERVE Trial) (NCT02538341), a randomized phase 2 study. The primary outcome is GMFR (Geometric Mean Fold Rise ) in Varicella Zoster Virus (VZV) Glycoprotein Enzyme-linked Immunosorbent Assay (gpELISA) Immunoglobulin G (IgG) Levels From Baseline at 6 Weeks.

2. Clinical Question

3. Trial Design

VERVE was a Phase 2, randomized, quadruple-masked, parallel-group trial. Its stated purpose was to evaluate the safety and effectiveness of live zoster vaccine in people using anti-TNF therapy. The registry lists University of Alabama at Birmingham as sponsor and the study status as completed. The study started in 2016-05, and primary completion was recorded as 2019-12. The planned enrollment was 617.

The two intervention groups were Zoster Vaccine Live (Zostavax) and Placebo Normal Saline. Each intervention was described as a single 0.65 mL injection under the skin in the deltoid region of the upper arm. The registry says allocation was randomized and masking was quadruple. Specific information about the randomization sequence, allocation concealment, and which participant, clinician, assessor, or other roles were masked is not reported in the registry record.

The primary outcome was geometric mean fold rise in VZV gpELISA IgG levels from baseline at six weeks after vaccination. Secondary outcomes included antibody and cellular immune measures at one year, confirmed varicella samples, vaccine tolerability within 42 days, and rheumatoid arthritis disease activity using CDAI. The registry lists five secondary outcomes and seven outcome measures overall. These outcomes concern more than one domain and follow-up period; their presence does not establish which analyses were prespecified or how their results were handled together.

The posted primary arm results include 303 participants for vaccine and 299 for placebo. These are the arm counts attached to the primary outcome result, not a substitute for the planned enrollment figure. The registry’s description also reports that some participants did not complete the six-week assessment and that some samples were not usable. Further information about the trial protocol, analysis populations, and statistical analysis plan is not reported in the registry record.

4. Participant Flow and Accounting

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

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Zoster vaccine live arm 7 did not complete Week 6; 5 lost to follow-up, 2 withdrew consent. Placebo arm 8 did not complete Week 6; 5 lost to follow-up, 3 withdrew consent. Zoster vaccine live arm 300 immunogenicity samples collected, 294 useable immunoglobulin G (IgG) samples, 287 usable immunoglobulin G (IgG) sample pairs. For the placebo arm 298 immunogenicity samples collected, 289 useable immunoglobulin G (IgG) samples, 259 usable immunoglobulin G (IgG) sample pairs.

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 GMFR (Geometric Mean Fold Rise ) in Varicella Zoster Virus (VZV) Glycoprotein Enzyme-linked Immunosorbent Assay (gpELISA) Immunoglobulin G (IgG) Levels From Baseline at 6 Weeks, measured over 6 weeks post vaccination.

The primary endpoint is the geometric mean fold rise in immunoglobulin G antibodies to VZV glycoprotein, measured with gpELISA. It describes change from baseline at six weeks after vaccination. The record calls this a protocol-defined measure for immunogenicity samples. The posted results give an arm-level estimate of 1.33 for Zoster Vaccine Live and 1.02 for placebo, each with a 95% confidence interval.

In general, a fold-rise measure expresses change relative to a starting measurement rather than reporting an absolute antibody concentration. A geometric mean is suited to summarizing multiplicative measurements, which commonly have a skewed distribution. It gives a central summary on a multiplicative scale, rather than the ordinary arithmetic average. These are general statistical properties, not additional trial findings.

An antibody response is an immunogenicity outcome. It can describe a biological response to vaccination, but it is not itself a direct count of shingles cases, a measure of vaccine effectiveness against clinical disease, or proof of protection. The registry lists confirmed varicella as a separate secondary outcome. The primary endpoint’s six-week timing also differs from the secondary antibody outcomes listed at one year. The registry record does not report a threshold for interpreting the fold rise as clinically meaningful.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is expressed as a geometric mean fold rise, with a 95% confidence interval for each arm. The registry also reports a generalized linear model analysis and a p-value for the comparison. These are different statistical summaries: the arm estimates describe each group, while the model analysis addresses a contrast under its specified assumptions. The details needed to reconstruct that contrast are not reported in the registry record.

What does a confidence interval describe?

In general, a confidence interval is a range produced by a statistical procedure intended to have a stated coverage rate over repeated applications under its assumptions. It communicates the precision of an estimate in a way that a point estimate alone cannot. The posted 95% confidence intervals belong to the vaccine and placebo arm estimates separately. They do not provide the confidence interval for the between-arm contrast. To assess the comparative effect directly, readers would need a contrast estimate and its uncertainty, which are not reported with these arm results.

The reported intervals are 1.17 to 1.51 for vaccine and 0.91 to 1.14 for placebo. Comparing separate intervals by eye is not equivalent to testing or estimating the difference between groups. A formal comparison uses the uncertainty of the contrast, which depends on the statistical model and the data contributing to it. The generalized linear model result is the registry’s reported comparison; the record does not supply its coefficient, contrast scale, or confidence interval.

What does the generalized linear model tell the reader?

The registry identifies the sponsor’s method as a generalized linear model and reports p = 0.002. In general, generalized linear models allow the outcome’s distribution and the relationship between predictors and the expected outcome to be specified in ways that extend ordinary linear regression. The method label alone does not reveal the particular distribution, link function, covariates, handling of repeated or clustered observations, or adjustments used in this analysis. Those details are not reported in the registry record, so the model cannot be independently reconstructed from the available description.

A p-value is evaluated relative to a null hypothesis and a model. In general, it measures how unusual the observed result, or a more extreme result, would be if that null model were true. It does not measure the probability that a treatment effect is real, and it does not show whether the effect is large enough to matter clinically. Here, the p-value should be read as the reported model-based test result for the immunogenicity comparison, not as a measure of protection against shingles.

Why use a geometric mean?

Fold changes are multiplicative: a given fold rise is interpreted relative to the baseline value. In general, a geometric mean summarizes such values on a multiplicative scale and can be more suitable than an arithmetic mean when values are skewed. This summary does not show the full distribution of individual responses. The record does not provide individual-level data, a spread measure for the arm estimates, or details of any transformation used in model fitting.

The posted results use arm-level estimates and confidence intervals. The record’s primary-outcome description also notes incomplete Week 6 assessments and samples described as usable or unusable. Missingness can affect an estimate if the people or samples not represented differ systematically from those analyzed. In general, analysis choices such as excluding incomplete observations, imputing values, or using a model that incorporates incomplete data can yield different interpretations. The registry record does not report which approach was used, the assumptions about missingness, or a sensitivity analysis.

How should the planning illustrations be understood?

The planning illustrations use the observed effect as their starting point, with the effect to detect expressed as a stated fraction of that observed effect. Their stated assumptions are a two-sided significance level of 0.05, equal allocation between the arms, and a normal approximation. They are planning illustrations rather than observed power, and they are not evidence that the original trial was designed around those assumptions. The trial’s sample-size justification is not reported in the registry record.

In general, planning calculations depend on assumptions about the target effect, variability, significance level, allocation, and analysis method. An illustration based on an observed effect should not be treated as independent confirmation of that effect: the observed estimate is itself uncertain. The planning method specified for the posted illustrations does not establish that the trial’s original design used the same assumptions. The registry record does not report the original calculation or any alternative design assumptions.

The record lists multiple outcome measures and secondary outcomes. In general, testing many outcomes can increase the chance of obtaining at least one apparently notable result by chance, unless the analysis plan addresses multiplicity. The registry record does not report a multiplicity adjustment or a hierarchy for interpreting the outcomes. The primary result can therefore be described on its own terms, while the broader pattern across outcomes cannot be evaluated from the primary endpoint alone.

10. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary comparison as a generalized linear model, with p = 0.002. The arm-level results are 1.33 for vaccine and 1.02 for placebo, with their respective 95% confidence intervals. The model-based test is the reported inferential comparison; it should not be confused with comparing the two arm-specific intervals as if they were an interval for the treatment contrast.

A recalculated two-arm estimate is not available for this endpoint in the record, because the values needed for that recalculation are missing or not estimable. The registry does not report the model coefficient, comparative confidence interval, or enough model detail to reconstruct the analysis. No discrepancy is flagged between the sponsor analysis and a recalculation. Accordingly, the sponsor’s reported analysis is the available comparison, while its assumptions and precision beyond the p-value remain incompletely described.

The sponsor’s analysis concerns the six-week antibody fold-rise measure. It does not transform the immunogenicity endpoint into an estimate of clinical protection, and the reported p-value does not quantify clinical importance. A fuller account would require the analysis specification and an estimate of the between-arm effect with its uncertainty; these are not reported in the registry record.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary endpoint and five secondary outcomes, including antibody and cellular immune measures at one year, confirmed varicella samples, tolerability within 42 days, and CDAI disease activity. It also records seven outcome measures overall and two analyses. These entries show that the trial measured more than the six-week primary immunogenicity endpoint. They do not, on their own, establish a prespecified testing hierarchy or a multiplicity strategy.

When several outcomes or comparisons are examined, the chance of at least one nominally unusual finding can increase. In general, multiplicity methods may control an error rate across tests, or investigators may specify a primary outcome and treat other analyses as supportive or exploratory. The registry record does not report whether multiplicity was addressed, which outcomes were included in any shared testing family, or whether particular outcomes were prioritized beyond their primary or secondary labels. Therefore the p-value for the primary endpoint should be interpreted as the reported test for that endpoint, not as a complete account of evidence across all outcomes.

The primary-outcome description reports that seven vaccine-arm participants and eight placebo-arm participants did not complete Week 6. It gives loss to follow-up and withdrawal of consent as reasons. It also reports 300 vaccine-arm immunogenicity samples collected, 294 usable IgG samples, and an incomplete further description beginning with 287 usable immunoglobu... The record does not finish that measure’s label in the posted text. It does not report corresponding complete sample information for placebo in this description, nor does it state how missing observations or unusable specimens were addressed in the analysis.

Analysis populations are also not fully described. The primary result gives arm counts of 303 and 299, but the registry record does not explain whether these counts represent all randomized participants, participants with an available assessment, or another analysis population. The analysis population definition, missing-data assumptions, and sensitivity analyses are not reported in the registry record. Without those details, readers cannot assess how robust the result is to alternative handling of missing data or sample usability.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: blocked bowel (1.0%), death (1.0%), syncope and collapse (0.3%), respiratory syncytial virus pneumonia (0.3%), sepsis due to pyelonephritis (0.3%).

The registry reports serious adverse events for groups with 310 participants at risk in the vaccine arm and 307 at risk in the placebo arm. Serious adverse events affected 10 in the vaccine arm and 8 in the placebo arm. Deaths were recorded for 3 in the vaccine arm and 0 in the placebo arm. These are descriptive counts from the registry record. They do not establish that an event was caused by the assigned intervention, and they do not by themselves show whether the groups differed in risk under a prespecified statistical comparison.

The serious-event terms listed most often include Blocked Bowel and Death at 1.0% each, followed by Syncope and collapse, Respiratory Syncytial Virus Pneumonia, Sepsis due to pyelonephritis, and Hip fracture at 0.3% each. The registry does not provide enough detail in the registry record to determine the precise denominator or event-counting rule behind each listed percentage. These terms are event categories, not evidence of causality.

The arm counts reported for serious adverse events differ from the arm counts attached to the primary immunogenicity result. The registry record does not explain the relationship between those populations or denominators. Safety counts should therefore be read separately from the primary antibody analysis. Details such as event timing, severity classification, relatedness assessment, and comparative statistical analysis are not reported in the registry record.

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

The posted result describes an immune response measured by VZV gpELISA IgG six weeks after vaccination. The vaccine-arm geometric mean fold rise was 1.33, and the placebo-arm estimate was 1.02. The registry reports a generalized linear model comparison with p = 0.002. Together, these values describe the recorded immunogenicity outcome and the sponsor’s model-based test for that outcome.

The result does not directly measure shingles prevention. An antibody fold rise is a laboratory endpoint, whereas clinical effectiveness would require an outcome such as herpes zoster occurrence and an appropriate comparative analysis. The registry lists confirmed varicella as a secondary outcome, but a result for that outcome is not reported in the primary details used for this explanation. The record also does not report a clinical threshold that would connect the observed antibody change to protection.

The p-value does not establish the magnitude or clinical importance of the comparative effect. The arm confidence intervals describe each estimate separately; a confidence interval for the between-arm contrast is not reported. The model’s detailed specification and the handling of incomplete assessments are also not reported. These omissions limit how fully readers can evaluate precision, assumptions, and robustness.

The trial population described in the registry was over 50 years old, had arthritis or related conditions, was using anti-TNF therapy, and had not previously received the vaccine. The posted result should be understood in relation to that population and the measured endpoint rather than generalized automatically to other populations, interventions, or clinical outcomes. The registry record does not provide results sufficient to establish effectiveness for preventing shingles or to resolve the broader clinical benefit-risk question.

14. Why This Trial Matters Statistically

This trial raises statistical questions that are important in studies of immune response: how to summarize fold-change data, how to distinguish an arm estimate from a comparative estimate, and how to interpret a model-based p-value when the model details are limited. The geometric mean fold rise is a multiplicative summary, while the primary clinical question of interest may ultimately concern disease outcomes. Keeping those quantities separate prevents an immunogenicity result from being mistaken for a direct effectiveness result.

The record also illustrates why confidence intervals should be matched to the estimand they describe. The vaccine and placebo confidence intervals quantify uncertainty around their respective arm summaries. They are not a reported interval for the treatment contrast. A p-value is available for the generalized linear model comparison, but the absence of a comparative estimate and interval means the magnitude and precision of that contrast are not fully communicated in the posted result.

Completeness of outcome and sample information matters as well. The primary description records incomplete follow-up and usable-sample counts, but does not explain how those data were handled analytically. The record lists several secondary outcomes, yet does not report a multiplicity plan. These gaps do not invalidate the posted result, but they limit the ability to assess the analysis assumptions and the pattern of evidence across outcomes.

Finally, safety counts and immunogenicity results answer different questions and use different reported group sizes. Reading each outcome in its own context helps avoid combining denominators or treating descriptive adverse-event counts as a causal comparison. VERVE is therefore statistically informative not only for its reported antibody response, but also for the importance of precise endpoint definitions, transparent analysis populations, and complete comparative uncertainty estimates.

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 primary endpoint is an immunogenicity measure rather than a direct clinical endpoint. The result concerns six-week VZV gpELISA IgG fold rise and does not itself establish prevention of shingles. The registry lists confirmed varicella as a secondary outcome, but a result for that outcome is not included in the primary-result information. A clinical threshold linking the antibody measure to protection is not reported in the registry record.

The comparative analysis is incompletely specified. The registry identifies a generalized linear model and gives p = 0.002, but does not report the model’s distribution, link function, covariates, contrast scale, coefficient, or confidence interval for the comparison. This limits independent assessment of the model and the precision of the between-arm effect. The arm-level confidence intervals cannot substitute for a comparative interval.

Information about missing observations and analysis populations is limited. The record reports non-completion of the Week 6 assessment, collected samples, and usable IgG samples, but does not explain how incomplete follow-up or unusable specimens were handled. The primary-result description is truncated after a count of 287 usable immunoglobu..., so the associated assay or measurement cannot be identified from the record. Definitions of the analysis population and missing-data sensitivity analyses are not reported.

The registry lists several secondary outcomes but does not report a multiplicity approach or an outcome hierarchy. The sample-size justification and original statistical analysis plan are not reported in the registry record. Safety results are presented as descriptive counts, with no comparative model or causal attribution details in the information available. These limitations restrict interpretation of precision, the broader outcome pattern, and the relationship between observed events and treatment assignment.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Safety and Effectiveness of Live Zoster Vaccine in Anti-Tumor Necrosis Factor (TNF) Users (VERVE Trial) — NCT02538341.
• Registry record last updated 2021-11-23; status last verified 2021-10.
• Results first posted 2021-11-23.
• 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

VERVE (NCT02538341) was a completed Phase 2 randomized, quadruple-masked, parallel-group trial sponsored by University of Alabama at Birmingham. The registry describes a planned enrollment of 617 and a population of people over 50 years old with arthritis or related diseases who were using anti-TNF therapy and had not previously received zoster vaccine.

The primary endpoint was geometric mean fold rise in VZV gpELISA IgG from baseline at six weeks. The posted estimates were 1.33 for Zoster Vaccine Live and 1.02 for placebo, with 95% confidence intervals of 1.17 to 1.51 and 0.91 to 1.14. The reported generalized linear model comparison had p = 0.002. This is an immunogenicity result, not a direct estimate of clinical effectiveness against shingles.

The registry also lists secondary immune, tolerability, disease-activity, and varicella outcomes. Important details about model specification, analysis populations, missing-data handling, multiplicity, and the original sample-size justification are not reported in the registry record.