Vaccine Hyporesponse in Healthy Elderly Participants (MK-0000-131 AM2): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Vaccine Hyporesponse in Healthy Elderly Participants (MK-0000-131 AM2) (NCT01119703), a registered phase not stated study. The primary outcome is Antibody Titer Responses to Hepatitis B Virus Surface Antigen (HBV sAg) Vaccine, Measured and Predicted Based on Pre-vaccination Biomarkers in Healthy, Elderly Participants..

2. Clinical Question

3. Trial Design

This was a Phase I, open-label observational study of vaccine hyporesponse prediction in healthy participants. The registry record lists an enrollment of 174 and a completed status. The study started in 2010-07, with primary completion in 2011-11. Its stated objective was to develop prospective prediction rules from baseline biomarkers for vaccine response, defined in the summary as the post-baseline absolute serum antibody titer, using the protocol-selected vaccines separately. It also examined rank correlation coefficients for pairs of post-vaccination antibody titers within the same older individuals.

The primary endpoint focused on hepatitis B surface antigen antibody titer in healthy older participants. Participants in this population received hepatitis vaccine, tetanus-diphtheria booster vaccine, and cholera vaccine simultaneously at baseline. The registry also describes a healthy younger group aged 25 to 40 years, while the older group was aged 65 years and older. It explicitly characterizes comparisons between the younger and older cohorts as exploratory and says those comparisons are not reported.

The listed biological interventions were Twinrix, the tetanus and diphtheria booster, and Dukoral. The design is not described as a randomized comparison, and the allocation method is not stated. Accordingly, the arms should not be interpreted as a treatment and control pair. The record does not report the protocol’s sample-size justification, a detailed statistical analysis plan, or whether interim analyses were planned. The record’s description of an observational, open-label study and its stated prediction objective are the appropriate context for interpreting the endpoint.

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: Healthy participants 65 years of age and older who were treated with all three vaccines, and had their post-vaccination titers measured. Results from comparing titers and biomarker responses between younger and older cohorts are exploratory, and therefore not reported.

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 Antibody Titer Responses to Hepatitis B Virus Surface Antigen (HBV sAg) Vaccine, Measured and Predicted Based on Pre-vaccination Biomarkers in Healthy, Elderly Participants., measured over Baseline and 1 month after final vaccination.

The primary endpoint is the antibody titer response to hepatitis B virus surface antigen in healthy older participants. It is expressed as the mean of a continuous measure in ln International Units, with standard deviation identified as the dispersion measure. Titers were measured by enzyme linked immunosorbent assay one month after final vaccination. The assay’s optical-density results were converted to standardized international units of reactivity and transformed to the natural logarithm.

The endpoint was selected to represent the post-vaccination antibody response that the study sought to predict from pre-vaccination biomarkers. Baseline blood samples were used for messenger RNA profiling, biochemical assays, and flow cytometric assays. Those measurements were used in a machine-learning model based on random forests to predict the transformed antibody titers. The endpoint therefore links a measured immune response with a prediction task; it is not itself a direct measure of how accurately the model predicted that response.

The registry record specifies the measurement timing and scale but does not report the observed mean, standard deviation, participant count contributing to the endpoint, or model-performance estimates. Without those results, the endpoint definition explains what was intended to be measured, but not the size or precision of the observed response or the quality of the predictions.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a continuous antibody-titer measure. A continuous outcome can take values across a scale, rather than being limited to a category such as yes or no. The registry describes the primary result as a mean and identifies standard deviation as its measure of dispersion. A mean describes the central level of a set of measurements, while a standard deviation describes how spread out individual measurements are around that mean. Neither quantity is reported numerically in the registry record for this endpoint.

Why use a logarithmic scale?

The antibody titer was converted to natural logarithms, denoted ln, after assay measurements were translated into standardized international units. A logarithmic transformation changes the scale on which the data are analyzed. In general, transformations can make a distribution more suitable for statistical modeling or reduce the influence of very large values. That is a general explanation, not a claim about the distribution observed in this study: the registry record does not report a distributional assessment or a reason for selecting the transformation. Because the endpoint is reported on the ln scale, its mean and spread should be understood on that transformed scale rather than as raw optical-density readings.

What does the assay measure?

The registry describes enzyme linked immunosorbent assay, or ELISA, as the method used to measure HBV sAg antibody titers. Optical densities were converted to standardized international units of reactivity and then transformed to ln values. The measurement time was one month after final vaccination, while biomarker samples used for prediction were collected before vaccination at baseline. This order supports the stated prediction objective: information measured at baseline is used to predict a later antibody-titer outcome. The registry record does not provide assay validation details, laboratory reproducibility estimates, or the observed endpoint values.

How should a prediction model be understood?

The record describes a machine-learning model based on random forests. In general, a random forest combines predictions from multiple tree-based models, which can represent complex relationships between inputs and an outcome. Here, the inputs were baseline biomarkers measured through messenger RNA profiling, biochemical assays, and flow cytometric assays; the target was the transformed post-vaccination antibody titer. A prediction model’s usefulness depends not only on its ability to fit the data used to build it, but also on how well it predicts data not used for model fitting. The registry record does not report how model development and evaluation were separated, what validation approach was used, or any prediction-performance metric. No conclusion about accuracy, calibration, or generalizability can therefore be drawn from the model description alone.

What does rank correlation describe?

The study summary says it examined rank correlation coefficients for pairs of post-vaccination antibody titers within the same older individuals. In general, rank correlation assesses whether higher values of one measure tend to accompany higher or lower values of another, based on the ordering of observations rather than the raw distances between values. It can describe association without requiring that the relationship be linear. It does not by itself establish that one response causes another, and it is distinct from evaluating the accuracy of the biomarker prediction model. The registry record does not report the coefficients or their uncertainty.

What do confidence intervals and significance tests add?

In general, a confidence interval communicates uncertainty around an estimated quantity: narrower intervals indicate greater precision under the assumptions of the analysis, while wider intervals indicate less precision. A significance test evaluates compatibility between data and a specified null hypothesis, but a p-value does not show the size or practical importance of an effect. The primary endpoint results needed to interpret such quantities are not reported in the registry record. The significance level of 0.05 appears only among the stated assumptions for planning illustrations; it should not be treated as confirmation of the study’s analysis threshold or a reported test result.

How should planning calculations be read?

The record describes sample-size assumptions for planning illustrations based on an observed effect. They use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is defined as a stated fraction of the observed effect. For outcomes other than time-to-event outcomes, the illustrations express participants per arm for 80% power or other listed power targets; time-to-event illustrations express total events. These are planning quantities, not observed power, and they do not establish the study’s original sample-size rationale. The registry record does not report that rationale. Planning calculations depend on their assumptions and should not be read as evidence that a study did or did not have adequate power for its actual analyses.

Finally, over-dispersion is a general issue that arises when variability exceeds what a chosen statistical model expects. If ignored, uncertainty estimates can be too optimistic. The registry record does not report an over-dispersion assessment or identify a model for a count outcome as the primary analysis. This general concept therefore does not imply that over-dispersion occurred in this trial. Interpretation should remain anchored to the specific continuous endpoint and the prediction objective described in the registry.

10. Multiplicity, Missing Data and Analysis Populations

The registry record lists nine outcome measures and fourteen analyses. When a study examines multiple outcomes or performs multiple analyses, the chance of at least one apparently notable result can increase if each result is assessed separately. Statistical approaches to multiplicity can address this issue, but the appropriate method depends on the questions and analysis plan. The registry record does not report whether any adjustment for multiple comparisons was planned or used, so no such adjustment should be presumed.

The primary endpoint is described for healthy older participants who received all three vaccines and had post-vaccination titers measured. The record does not report the number contributing data to that endpoint, the amount or pattern of missing endpoint data, or how missing measurements were handled. It also does not describe imputation, exclusions, or sensitivity analyses. These details are important because analyses based only on participants with observed measurements can be affected if missingness is related to participant characteristics or outcomes.

The analysis population for the primary endpoint is defined in the endpoint description, but the registry record does not provide a detailed participant flow or a broader analysis-population definition. The younger cohort’s comparison with the older cohort is identified as exploratory, and those comparative results are not reported. Allocation is not stated, and the study is not described as a randomized comparison. These reporting limits prevent a full assessment of multiplicity handling, missing-data methods, and the precise populations contributing to each analysis.

11. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent non-serious adverse event terms

The registry record reports safety figures for an All Participants group with 174 participants at risk for serious adverse events and 0 affected. It reports 174 participants at risk for other adverse events and 14 affected. These figures describe the counts recorded for the group; they do not establish a causal relationship between vaccination and an event. The record does not list serious adverse-event terms among the top serious terms.

The death field is not reported in the registry record. It should not be interpreted as a count of zero. The registry information also does not specify event severity beyond the serious-versus-other distinction, the timing of events, the number of events per affected participant, or whether events were judged related to a vaccine. These safety counts should therefore be read as registry summaries, not as a complete clinical assessment of risk or as a comparison between randomized groups. The study is described as observational and does not provide a like-for-like control comparison for the primary endpoint.

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

The study’s central statistical question was whether baseline biomarker measurements could predict later antibody titers in healthy older participants. The registry identifies the target as the natural-log-transformed hepatitis B surface antigen antibody titer measured one month after final vaccination. It also describes use of a random-forest-based model and lists other vaccine-response outcomes. These details define the intended prediction task, but the record does not report the numerical primary endpoint result or the model’s predictive performance.

Consequently, the registry entries do not establish how high or variable the measured titers were, how closely predictions matched measurements, or whether any prediction rule would perform similarly in another group. A description of a modeling approach is not evidence of its accuracy. Likewise, an objective to develop prediction rules does not establish that a useful or validated rule resulted. The record does not report validation results, calibration, or other performance summaries.

The investigation should also not be interpreted as a randomized test of whether a vaccine works. The participants received three vaccines, and the record does not state a randomized allocation method. It explicitly says that age-cohort comparisons were exploratory and are not reported. The primary endpoint concerns a measured antibody response, not a clinical disease outcome. No treatment-versus-control effect is described in the registry record for this endpoint.

The main contribution that can be read from the registry record is the study’s stated design and measurement strategy: baseline biomarker profiling was intended to support prediction of a later antibody response. What cannot be concluded is equally important. The available record does not give the outcome estimates, their precision, the predictive performance, or the handling of missing data and multiplicity. Without those results, the size, reliability, and broader applicability of any association or prediction remain unreported.

13. Why This Trial Matters Statistically

This trial is statistically notable for placing prediction, rather than a randomized treatment contrast, at the center of its primary question. It sought to use information collected before vaccination to predict a later continuous antibody-titer measurement. That structure highlights an important distinction: identifying a potential predictor and demonstrating reliable prediction are separate tasks. A model can be described and developed without the record establishing that it predicts accurately in new participants.

The endpoint also illustrates how measurement choices shape interpretation. Antibody titers were converted to standardized international units and analyzed on a natural-log scale, with the mean and standard deviation specified as summary measures. The model used several types of baseline biomarker measurements, including messenger RNA profiling, biochemical assays, and flow cytometric assays. Together, these features create a complex prediction problem, but the registry record does not report the evaluation details or performance results required to judge it.

The study further describes multiple vaccine-response outcomes and analyses, as well as rank correlations between pairs of post-vaccination titers. That breadth makes clear why readers need to know how outcomes were prioritized and how multiplicity was handled. Those methods are not reported in the registry record. Similarly, the record identifies the primary population but does not report endpoint-specific missingness or the exact analysis populations.

The statistical importance of the study therefore lies in the question and design described, not in an unreported result. It provides a framework for examining whether baseline biological measurements can forecast a later immune-response measure, while leaving crucial evidence about precision, validation, and reproducibility unavailable in the record.

14. 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 main limitation for interpreting this specific trial is that the registry record describes the primary endpoint and prediction approach without reporting numerical endpoint results or model performance. There is no reported mean, standard deviation, endpoint-specific participant count, or estimate of prediction accuracy in the record details. Thus readers cannot assess the response magnitude, its precision, or the predictive value of baseline biomarkers.

The record also does not report how the model was validated, whether predictions were evaluated on data separate from model development, or whether performance was assessed for calibration or generalizability. These are important considerations for a machine-learning prediction objective. The model’s use of random forests is described, but that description does not reveal how the model was built or how well it worked.

Although the record lists nine outcome measures and fourteen analyses, it does not report a multiplicity strategy. Missing-data handling, participant flow, and endpoint-specific analysis populations are also not detailed. The stated primary population includes older participants who received all three vaccines and had titers measured, but the record does not say how many met that endpoint definition.

The study is not described as randomized, allocation is not stated, and the younger-versus-older comparison is exploratory and not reported. The primary endpoint therefore should not be treated as a controlled vaccine comparison. Finally, the registry record does not report the study’s sample-size justification, so planning illustrations should not be mistaken for the original design rationale.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Vaccine Hyporesponse in Healthy Elderly Participants (MK-0000-131 AM2) — NCT01119703.
• Registry record last updated 2015-12-29; status last verified 2015-11.
• Results first posted 2012-11-22.
• 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.

18. Record Summary

NCT01119703, titled Vaccine Hyporesponse in Healthy Elderly Participants (MK-0000-131 AM2), is recorded as a completed Phase I, open-label observational study sponsored by Merck Sharp & Dohme LLC. The registry lists enrollment of 174, a start date of 2010-07, and primary completion in 2011-11.

The study aimed to develop baseline biomarker-based prediction rules for vaccine responses and examine rank correlations between post-vaccination antibody titers. The primary endpoint was the mean natural-log-transformed hepatitis B surface antigen antibody titer in healthy participants 65 years of age and older, measured one month after final vaccination. The record describes vaccination with hepatitis, tetanus-diphtheria, and cholera vaccines and a random-forest-based prediction model.

Numerical primary outcome and prediction-performance results, detailed missing-data methods, multiplicity handling, and sample-size justification are not reported in the registry record.