A Study of an Investigational V212/Heat-Treated Varicella-Zoster Virus (VZV) Vaccine in Immunocompromised Adults (V212-002): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of an Investigational V212/Heat-Treated Varicella-Zoster Virus (VZV) Vaccine in Immunocompromised Adults (V212-002) (NCT00535236), a randomized phase 1 study. The primary outcome is Geometric Mean Fold Rise (GMFR) in Varicella-Zoster (VZV) Antibody Responses Measured by Glycoprotein Enzyme-linked Immunosorbent Assay (gpELISA).

2. Clinical Question

This was a safety and immunogenicity study, not a trial with a posted primary clinical outcome comparing V212 with placebo for prevention of herpes zoster or its complications. Its primary endpoint describes antibody change over time among selected V212 recipients. That makes the central statistical question one of the magnitude and precision of the measured immune response, rather than a direct estimate of a treatment effect on disease occurrence. The randomized placebo groups remain relevant to the broader study design and safety assessment, but the posted primary immunogenicity estimates do not give a placebo comparison for the groups named in that endpoint.

3. Trial Design

The registry identifies NCT00535236 as a Phase I, randomized, double-blind, placebo-controlled, multicenter, parallel-group trial. It enrolled 341 participants and is listed as completed. The recorded study dates are 2007-11-02 for the start and 2010-01-25 for primary completion. The sponsor was Merck Sharp & Dohme LLC. The conditions listed are herpes zoster and herpes zoster-related complications.

Participants belonged to clinically distinct groups: autologous and allogeneic hematopoietic cell transplant recipients; people with HIV infection and a baseline CD4 cell count ≤200 cells/mm^3; people receiving chemotherapy for specified solid tumor malignancies; and people with specified hematologic malignancies. V212 and placebo were given in group-specific arms. The intervention descriptions specify a 0.65 mL subcutaneous injection in a four-dose regimen, with doses approximately 30 days apart. For the groups described, injections were preferably given in the deltoid area, with arms alternated between doses.

The primary immunogenicity endpoint was framed for participants in the solid tumor, hematologic malignancy, and HIV groups who received at least one V212 dose and had valid immunogenicity information and post-vaccination gpELISA data. The registry record explicitly describes the primary immunogenicity objective and endpoints for the transplant groups as exploratory. Thus, the trial included several populations, but the primary immune-response question did not apply to every group in the same way.

The registry record does not report the randomization sequence, allocation concealment procedures, or detailed masking procedures. It identifies the design as randomized and double-blind, but further operational details are not reported in the registry record. A sample-size justification, planned interim analyses, and the full statistical analysis plan are also not reported in the registry record. Those omissions limit what can be inferred about the planned precision, monitoring, and analysis decisions beyond the outcome descriptions and analyses recorded.

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.

Only the first 6 of 10 groups are shown.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All randomised participants in STM, HM and HIV arms who had received at least 1 dose of V212, had at least 1 valid immunogenicity evaluation for VZV antibody and had post-vaccination gpELISA data available for endpoint. As per the protocol, immunogenicity results for the autologous and allogeneic HCT groups were considered exploratory.

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.

Only the first 6 of 11 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Geometric Mean Fold Rise (GMFR) in Varicella-Zoster (VZV) Antibody Responses Measured by Glycoprotein Enzyme-linked Immunosorbent Assay (gpELISA), measured over Baseline (Day 1 predose vaccination 1) and 28 days postdose 4 (~Day 118).

The primary endpoint is geometric mean fold rise in VZV antibody response measured by gpELISA. Blood was assessed before vaccination on Day 1 and 28 days after dose 4, at approximately Day 118. The registry describes the fold rise as the post-dose geometric mean titre divided by the prevaccination geometric mean titre. The reported scale is a ratio, and the registry gives a 90% confidence interval for each group estimate.

A geometric mean is commonly used for antibody measurements because such measurements can be skewed: a small number of high values may pull an arithmetic average upward. Taking logarithms before averaging and then transforming back produces a geometric mean, which is often more representative of the typical proportional level. A fold-rise endpoint expresses the post-vaccination level relative to baseline, making it useful for describing changes in antibody response when the underlying values span a wide range.

As a general statistical point, a confidence interval conveys uncertainty around an estimate as well as its central value. A 90% confidence interval does not mean that a particular participant’s response lies within the interval. It describes the uncertainty associated with estimating the group-level quantity under the analysis method. The endpoint is an immunogenicity measure, not a direct measure of protection from herpes zoster or its complications. The registry record does not report a validated antibody threshold that would translate these fold rises into a clinical level of protection.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical interpretation depends on both the type of outcome and the contrast being estimated. The primary endpoint is a continuous, positive-valued ratio: the geometric mean fold rise in VZV antibody response. Secondary temperature outcomes are binary, recording whether participants met the registry’s elevated oral temperature criterion. These outcome types call for different summaries and different interpretations. A ratio on a geometric scale should not be read like a percentage, while a binary outcome comparison concerns the difference in the proportion of participants experiencing an event.

What does a geometric mean fold rise describe?

For the primary endpoint, antibody measurements before and after vaccination were used to form a ratio, with the post-dose titre relative to the prevaccination titre. A value greater than 1 describes a higher post-vaccination geometric mean than the baseline geometric mean, while a value of 1 would describe no change in those group-level geometric means. This is a general interpretation of a ratio; it does not establish why a change occurred or whether it matters clinically. The registry reports the observed group estimates as 2.35 for the solid tumor group, 1.28 for the hematologic malignancy group, and 1.37 for the HIV group.

Geometric means are useful when measurements are positive and vary multiplicatively. For example, differences in antibody levels may be better described as proportional changes than as fixed additive changes. A common analytical approach is to work with logarithms of the measurements, calculate averages on that scale, and transform the results back to the original ratio scale. The registry identifies the endpoint as a geometric mean and states that a single longitudinal regression model was used for the reported analyses. It does not report whether the model used a logarithmic transformation, how repeated measurements were represented, or which covariates were included. Those details are not reported in the registry record and should not be assumed.

How should confidence intervals be read?

A confidence interval communicates uncertainty in an estimate. The primary endpoint uses 90% confidence intervals. For the solid tumor group, the interval runs from 1.84 to 3.01; for the hematologic malignancy group, from 1.12 to 1.45; and for the HIV group, from 1.11 to 1.7. An interval’s width reflects precision under the method used, with wider intervals indicating more uncertainty about the group-level estimate. It does not capture every possible source of bias, and it is not a range containing most individual responses.

Confidence intervals can also help distinguish estimation from hypothesis testing. A p-value addresses compatibility between data and a specified null model, conditional on the analysis assumptions. A confidence interval displays a range of values compatible with the data under its own assumptions. The registry reports p-values from a single longitudinal regression model but does not specify the tested null contrasts. As a result, the estimates and intervals can be described directly, but the precise hypothesis represented by each p-value cannot be reconstructed from the registry record.

What do binary comparisons measure?

For a binary outcome, the risk difference is the difference between the event proportions in two groups. A risk difference of zero corresponds to equal event proportions. A confidence interval that includes zero indicates that the data are compatible with no difference as well as with differences in either direction under that method. This is a general statistical interpretation, not a conclusion about whether an intervention is beneficial or harmful. The registry also reports sponsor analyses as differences in percentages with 95% confidence intervals and p-values for selected comparisons. For the transplant-group fever comparisons, those intervals include zero.

The simple unadjusted re-calculations use the posted group counts and event counts. They are not the sponsor’s fitted model and may use a different interval or test procedure. For example, the sponsor’s asymptotic method and an unadjusted risk-difference calculation can yield slightly different p-values or interval endpoints without one necessarily indicating a transcription error. The analyses answer related but not perfectly identical statistical questions because their methods differ. The sponsor’s registered analysis remains the relevant reported analysis, while re-calculations can serve as a transparent check based on summary counts.

Why does over-dispersion matter?

As a general statistical concept, over-dispersion occurs when observed variation is greater than a model’s assumed variation. If a model underestimates variability, standard errors may be too small and confidence intervals too narrow; p-values can then overstate how incompatible the data are with a null hypothesis. This concern can arise in clustered, repeated, or heterogeneous data, but it cannot be concluded that over-dispersion occurred in this trial. The registry record does not report a dispersion assessment or a method for adjusting for it. Likewise, the record does not describe how dependence among repeated antibody measurements was handled beyond naming a single longitudinal regression model.

What can planning illustrations tell us?

Sample-size calculations based on observed effects are planning illustrations, not observed power and not evidence that a study was adequately powered. Their values depend on assumptions about the effect to detect, allocation, significance level, and calculation method. The posted planning illustration states that it uses a two-sided significance level of 0.05, equal allocation, a normal approximation, and fractions of the observed effect. It is not described as the trial’s own design calculation. The trial’s sample-size justification is not reported in the registry record, so those illustrations cannot establish what effect the trial was originally planned to detect.

In this study, binary-outcome planning estimates are particularly sensitive to the observed event proportions and to the small placebo-group counts in the reported transplant comparisons. A planning calculation that takes an observed difference as its target assumes that the observed estimate is a useful stand-in for a future effect. That assumption can be uncertain, especially when estimates have wide confidence intervals. The appropriate use of such calculations is to illustrate how planning requirements change under stated assumptions, not to claim that the trial had a particular probability of finding a result.

10. Analyses Reported by the Sponsor

The sponsor’s primary analyses are reported as single longitudinal regression model analyses. The registry lists p < 0.001 for the solid tumor group, p = 0.003 for the hematologic malignancy group, and p = 0.017 for the HIV group. The record does not give the model coefficients, standard errors, exact tested contrasts, or full model specification. The associated group-level fold-rise estimates and 90% confidence intervals are therefore the most directly interpretable measures of the primary result in the registry entry.

For selected secondary comparisons, the sponsor reports an asymptotic method for differences in percentages, with 95% confidence intervals and p-values. The simple unadjusted re-calculations use the posted counts and risk differences. For the autologous transplant comparison, the recalculated risk difference is 0.105, with a confidence interval from -0.241 to 0.452 and p = 0.548; the sponsor reports a difference in percentages of 10.5, with a confidence interval from -21.3 to 41.8 and p = 0.552. For the allogeneic transplant comparison, the recalculated risk difference is 0.167, with a confidence interval from -0.177 to 0.510 and p = 0.329; the sponsor reports a difference in percentages of 6.7, with a confidence interval from -22.5 to 39.4 and p = 0.696.

These are related comparisons but use different procedures, and their results should not be substituted for one another. The registry’s sponsor-reported analysis is the primary record of the sponsor’s method. The recalculations are unadjusted checks, not adjusted or model-based reanalyses, and they do not override the sponsor analysis.

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists multiple study populations, several outcomes, and more than one analysis. It does not report a statistical testing hierarchy, a multiplicity adjustment, or a plan specifying which outcomes would be treated as confirmatory. Without that information, the nominal p-values should be interpreted as results for their respective reported analyses rather than as familywise error-controlled evidence across all endpoints and groups. This is especially relevant to secondary and subgroup results, where examining multiple comparisons can increase the chance of at least one apparently unusual result.

The primary hypothesis in the brief summary concerns a significant VZV-specific immune response measured by gpELISA or VZV IFN-ELISPOT at 28 days after dose 4 in the HIV, solid tumor, and hematologic malignancy groups. The posted primary endpoint described in detail is gpELISA geometric mean fold rise. The record also says that the primary immunogenicity objective and endpoints for the autologous and allogeneic transplant groups were considered exploratory. It does not report a complete accounting of how the two assay approaches or the different populations were combined in hypothesis testing, or how missing assay results were handled.

The primary endpoint population required participants to have received at least one dose of V212, have at least one valid immunogenicity evaluation, and have post-vaccination gpELISA data available. This is not the same as an analysis of all randomized participants. The number of participants excluded for not meeting these criteria, the reasons for missing measurements, and any imputation or sensitivity analyses are not reported in the registry record. If participants with unavailable measurements differed from those with observed measurements, the analyzed group could differ from the randomized group in ways that affect interpretation. The record does not provide enough information to assess that possibility.

More broadly, the registry does not report the full statistical analysis plan, the sample-size justification, or planned interim analyses. It does not report whether analyses were adjusted for baseline antibody levels or other participant characteristics. These omissions do not invalidate the posted estimates, but they constrain assessment of how closely the analyses followed an intended plan, whether missingness could have influenced results, and how much weight to place on multiple p-values.

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: pyrexia (17.5%), acute graft versus host disease in intestine (12.5%), anaemia (10.0%), febrile neutropenia (10.0%), diabetes insipidus (10.0%).

The registry reports serious adverse event counts by study arm and the number of participants at risk. In the autologous transplant groups, serious events affected 13 of 40 V212 recipients and 2 of 10 placebo recipients. In the allogeneic transplant groups, the counts were 32 of 40 and 7 of 10, respectively. For solid tumor malignancy, the counts were 7 of 57 and 2 of 19; for hematologic malignancy, 12 of 61 and 1 of 19; and for HIV, 8 of 60 and 2 of 20. These counts describe the reported groups and should be read with their denominators, which are not the same across all arms.

Deaths were also recorded: 1 in the autologous V212 group, 0 in its placebo group, 11 in the allogeneic V212 group, 2 in its placebo group, 1 in the solid tumor V212 group, 0 in its placebo group, 4 in the hematologic malignancy V212 group, 0 in its placebo group, 1 in the HIV V212 group, and 0 in its placebo group. The registry record does not establish that these events or deaths were caused by study treatment. Serious adverse event totals are not, by themselves, evidence of causation, and differences across arms may reflect the underlying illnesses and group characteristics.

The listed serious-event terms include pyrexia, acute graft versus host disease in intestine, anaemia, febrile neutropenia, diabetes insipidus, and diarrhoea. The registry’s summary does not replace detailed clinical review of event timing, severity, relationship assessment, or duration. Such detail is not reported in the registry record.

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

The primary results mean that the registry reports higher post-vaccination geometric mean antibody titres relative to baseline in the analyzed V212 groups. The fold-rise estimates are 2.35 for the solid tumor group, 1.28 for the hematologic malignancy group, and 1.37 for the HIV group. Their confidence intervals describe uncertainty in those group estimates under the reported analysis. These results are evidence about an antibody measurement at a defined follow-up time, not a direct measure of a clinical outcome.

They do not establish that vaccination prevented herpes zoster, reduced herpes zoster-related complications, or improved health outcomes. No primary clinical event comparison for those outcomes is reported in the record. The estimates also do not establish a causal increase in antibody response relative to placebo because the primary endpoint results shown are for V212 recipients, not a posted randomized comparison with placebo. A pre-to-post change can reflect multiple influences, and without the relevant controlled contrast it should not be interpreted as the estimated treatment effect against placebo.

The results do not show that immune responses differed among the three analyzed populations. Each group has its own estimate and confidence interval, but a comparison of estimates is not equivalent to a formal test between groups. Nor do the p-values establish the clinical importance of the observed responses. The exact null hypotheses tested by the longitudinal regression analyses are not reported in the registry record.

The study’s Phase I designation and safety and immunogenicity focus also matter for interpretation. The reported findings can characterize antibody responses and recorded adverse events in the studied participants, but they cannot by themselves establish effectiveness or fully characterize safety in other populations. The transplant groups’ primary immunogenicity objectives are identified as exploratory. Further, the endpoint analysis excluded randomized participants who did not meet its dose and assay-data criteria, and the record does not provide enough detail to determine how missing measurements may have affected the estimates.

In short, the posted evidence describes an immune-response signal in three groups and reports safety events across multiple arms. It does not settle the clinical effect of V212 on herpes zoster or establish a benefit-risk conclusion. That distinction between a biological measurement and a patient-relevant outcome is central to reading the trial accurately.

14. Why This Trial Matters Statistically

This trial is statistically informative because it brings together multiple immunocompromised populations with different underlying conditions and treatment contexts. The registry’s primary immunogenicity estimates apply to the solid tumor, hematologic malignancy, and HIV groups, while the transplant groups are described as exploratory for the primary immune-response objective. That distinction illustrates why a trial’s overall enrollment should not be mistaken for the sample size supporting each endpoint or population-specific estimate.

It also demonstrates the importance of matching the statistical summary to the question. Geometric mean fold rise is a proportional antibody-response measure; a binary fever outcome is a participant-level event proportion; serious adverse event counts are safety summaries. None can stand in for the others. The primary endpoint gives information about antibody levels over time, while the secondary and safety outcomes describe different aspects of the study experience.

The registry record also illustrates why precision and design information matter. Confidence intervals show uncertainty that a point estimate alone would conceal. Smaller placebo groups in some population-specific comparisons make the estimates less precise, while the registry’s missing detail about multiplicity, missing data, and analysis planning limits a full appraisal of the inferential framework. These are not merely technical qualifications: they determine how far a result can be generalized and how strongly it can support a claim.

Finally, sample-size planning based on observed effects must be distinguished from a trial’s original design and from observed power. The posted planning illustrations make their assumptions explicit, but the registry does not report the original sample-size justification. Treating the illustrations as hypothetical planning aids avoids giving them more evidential meaning than they have. A careful reading therefore separates descriptive immune-response findings, controlled treatment comparisons, safety summaries, and assumptions used for future planning.

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.

Several limitations are specific to interpreting this registry record. The primary endpoint estimates are reported for V212 recipients in three groups, without corresponding placebo antibody-response estimates. As a result, the primary findings describe within-group antibody change rather than a direct randomized treatment contrast. The endpoint also requires receipt of at least one dose and available valid assay data. The number and characteristics of participants not included under those criteria, along with reasons for missing data, are not reported in the registry record.

The primary endpoint is a biomarker measure. The record does not report a threshold that links the measured fold rise to protection from herpes zoster or its complications, and it does not provide a primary clinical outcome comparison for those events. The numerical immune response therefore cannot be translated directly into a clinical benefit. The transplant groups’ primary immunogenicity objectives are identified as exploratory, which further limits confirmatory interpretation for those populations.

The trial includes biologically and clinically different groups. The group-specific estimates should not be pooled or compared informally as if the populations were interchangeable. The registry does not report a formal interaction analysis or direct contrasts among the populations. Nor does it provide detailed model specifications, covariate adjustments, diagnostic checks, or the exact null contrasts behind the primary p-values.

Safety interpretation is also constrained. Serious adverse events and deaths are reported as counts, but the record does not provide enough detail about timing, duration, severity, relatedness, or event adjudication to attribute them to V212. Group sizes and underlying conditions differ, and the reported counts do not by themselves separate treatment effects from background risks.

Finally, a full analysis plan, multiplicity approach, missing-data sensitivity analyses, interim-analysis plan, and original sample-size justification are not reported in the registry record. These gaps limit assessment of prespecification, statistical control across the many outcomes and groups, and the robustness of estimates to unobserved data.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study of an Investigational V212/Heat-Treated Varicella-Zoster Virus (VZV) Vaccine in Immunocompromised Adults (V212-002) — NCT00535236.
• Registry record last updated 2019-05-20; status last verified 2019-05.
• Results first posted 2019-04-16.
• 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

NCT00535236 was a randomized, double-blind, placebo-controlled, multicenter Phase I trial of V212 in immunocompromised adults. The study enrolled 341 participants and is listed as completed. Its primary posted immunogenicity outcome was geometric mean fold rise in VZV antibody response by gpELISA, measured from baseline to 28 days after dose 4. Results are reported for V212 recipients in the solid tumor, hematologic malignancy, and HIV groups; the transplant groups’ primary immunogenicity objectives were considered exploratory.

The reported fold-rise estimates were 2.35, 1.28, and 1.37, respectively, with 90% confidence intervals. The registry also reports selected fever comparisons and serious adverse event counts across study arms. These findings describe antibody measurements and recorded safety events, but do not establish a clinical effect on herpes zoster or a primary placebo-controlled antibody-response effect. Detailed information on multiplicity, missing-data handling, the full analysis plan, and the original sample-size justification is not reported in the registry record.