Immunogenicity and Safety of 4 Prime-boost Combinations of HIV Vaccine Candidates in Healthy Volunteers: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Immunogenicity and Safety of 4 Prime-boost Combinations of HIV Vaccine Candidates in Healthy Volunteers (NCT02038842), a randomized phase 1, phase 2 study. The primary outcome is To Discard Vaccine Strategies With an Insufficient Level of Immunogenicity, Defined by HIV-specific IFN-γ-ELISPOT Responses, Among 4 HIV Prophylactic Prime-boost Combinations in Healthy Volunteers at Low Risk of HIV Infection.

2. Clinical Question

3. Trial Design

This was a randomized, open-label, multicenter Phase I/II trial with a parallel-group design. The registry record reports 92 enrolled participants and describes the participants as healthy volunteers at low risk of HIV infection. Masking was not used. The trial is recorded as completed.

Four prime-boost schedules were studied. In the MVA HIV-B/LIPO-5 arm, MVA HIV-B was given at Week 0 and Week 8, followed by LIPO-5 at Week 20 and Week 28. In the LIPO-5/MVA HIV-B arm, LIPO-5 was given at Week 0 and Week 8, followed by MVA HIV-B at Week 20 and Week 28. The two GTU-MultiHIV B combinations used priming doses at Week 0, Week 4 and Week 12, followed by either LIPO-5 or MVA HIV-B at Week 20 and Week 28.

The primary endpoint was assessed at Visit Week 30, described as 2 weeks after the last vaccine immunisation. The arm-level endpoint results report 21 participants for MVA HIV-B/LIPO-5, 21 for LIPO-5/MVA HIV-B, 19 for GTU-MultiHIV B/LIPO-5, and 21 for GTU-MultiHIV B/MVA HIV-B. These are the denominators reported alongside the primary outcome; the registry record does not explain their relationship to the total enrollment.

The record lists a start date of 2014-03 and a primary completion date of 2015-10. Further details about randomization procedures, allocation concealment, the original sample-size rationale, or planned interim analyses are 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:

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 To Discard Vaccine Strategies With an Insufficient Level of Immunogenicity, Defined by HIV-specific IFN-γ-ELISPOT Responses, Among 4 HIV Prophylactic Prime-boost Combinations in Healthy Volunteers at Low Risk of HIV Infection, measured over Visit Week 30.

The primary endpoint counted participants in each arm who had an HIV-specific interferon-gamma response measured by an ELISPOT assay at Week 30. A participant was counted as having a response if the test was positive to at least one of the stimulating HIV peptide pools. The pools covered Env, Gag, Pol, and Nef. The assay used stimulated peripheral blood mononuclear cells and was performed 2 weeks after the last vaccine immunisation.

This is a binary endpoint: each participant is classified as having or not having a response under the registry definition. The result is therefore reported as a count of responders within each arm, rather than as a continuous measurement of response magnitude. A positive result to at least one pool captures whether a response was detected under the stated assay definition; it does not, by itself, describe response strength, durability, or protection against infection.

The endpoint was selected to address the trial’s stated immunogenicity question: identifying prime-boost strategies considered to have an insufficient level of response. The registry also lists secondary objectives concerning tolerance and the type of vaccine-induced T-cell response. The specific results for those secondary objectives are not reported in the primary outcome data.

7. Results Posted in the Registry

• LIPO-5/MVA HIV-B: 42.9% (9/21) versus 33.3% (7/21) on control; risk difference 9.5 percentage points (95% CI -19.7 to 38.8), risk ratio 1.29 (0.59–2.81); unadjusted two-proportion z-test p = 0.525.
• GTU-MultiHIV B/LIPO-5: 0.0% (0/19) versus 33.3% (7/21) on control; risk difference -33.3 percentage points (95% CI -53.5 to -13.2); unadjusted two-proportion z-test p = 0.006.
• GTU-MultiHIV B/MVA HIV-B: 66.7% (14/21) versus 33.3% (7/21) on control; risk difference 33.3 percentage points (95% CI 4.8 to 61.8), risk ratio 2.00 (1.02–3.93); unadjusted two-proportion z-test p = 0.031.

8. Statistical Interpretation

9. Statistical Methods Explained

What does a binary outcome mean?

General: A binary outcome places each participant into one of two categories, such as response detected or not detected. The primary result is expressed as a count of participants meeting the response definition, together with the number assessed in each group. In this trial, a participant was counted as a responder after a positive result to at least one of the HIV peptide pools in the IFN-γ ELISPOT assay. The recorded counts are 7 of 21, 9 of 21, 0 of 19, and 14 of 21 across the four arms.

Binary outcomes can be compared using several measures. A risk difference is the difference between the observed response proportions in two groups. A value of no difference corresponds to equal proportions. A risk ratio compares the proportions as a ratio, while an odds ratio compares the odds of response. These measures answer related but distinct questions. The posted pairwise re-calculations focus on the risk difference, so the interpretation emphasizes the absolute separation between response proportions rather than a relative measure.

How should the confidence intervals be read?

General: A confidence interval describes a range of effect estimates compatible with the data and the method used to calculate the interval. It communicates precision as well as direction. For a risk difference, an interval spanning no difference indicates that the data do not exclude equal response proportions under that interval calculation. An interval entirely on one side of no difference indicates that the interval is consistent with a difference in that direction. It does not show that the true effect has a particular probability of lying within the calculated interval.

The interval for LIPO-5/MVA HIV-B versus MVA HIV-B/LIPO-5 includes values below and above no difference. The interval for GTU-MultiHIV B/LIPO-5 is below no difference, while the interval for GTU-MultiHIV B/MVA HIV-B is above no difference. The interval widths also matter: an estimate can have a clear direction while still being imprecise about the size of the difference.

What do the p-values represent?

General: A p-value measures how incompatible the observed data are with a stated null model, under the assumptions of the analysis. It is not the probability that the null hypothesis is true, nor does it measure the probability that a finding is due to chance. It also does not measure the size of an effect or its clinical importance.

The pairwise p-values of 0.525, 0.006, and 0.031 come from unadjusted re-calculations comparing the three other arms with MVA HIV-B/LIPO-5. The registry’s reported analysis lists a Binomial method across all four groups and p = 0.02. These results concern different analysis descriptions and should not be treated as interchangeable. Since the record indicates a discrepancy, the sponsor’s own analysis takes precedence over the simple re-calculations.

Why does the analysis method matter?

General: For a binary response, a binomial model is a natural way to describe counts of responses among participants assessed. The registry labels its overall analysis as Binomial, but does not report additional details about its model specification, test, adjustment variables, or handling of missing observations. A method label alone is not enough to reconstruct all aspects of the analysis.

The posted pairwise calculations are unadjusted comparisons of the observed summary counts. They do not account for covariates or for testing several comparisons. No adjusted pairwise results are reported in the registry record. The pairwise estimates and p-values are therefore best read as descriptive, simple re-calculations and not as a replacement for the sponsor’s analysis.

General: Over-dispersion occurs when observed variation is greater than a model expects. In binary data, differences in participant characteristics or response probabilities can sometimes make a simple model’s uncertainty estimates too optimistic. The registry record does not report an assessment of over-dispersion, so it is not possible to say whether it was evaluated or whether it affected the sponsor’s analysis.

What analysis details are not reported?

The registry record does not report the original statistical analysis plan, the trial’s own sample-size justification, or interim analyses. It does not state how missing or unevaluable Week 30 results were handled, whether the primary analysis used an intention-to-treat or per-protocol population, or whether the pairwise comparisons were prespecified. It also does not report a multiplicity adjustment for the several arm comparisons. These omissions limit how fully the analyses can be reproduced and how strongly the pairwise findings can be interpreted.

10. Confidence Intervals and Precision

The response counts differ across the four arms, but the arm sizes are limited: the reported denominators range from 19 to 21 participants. With binary outcomes, each participant’s classification contributes directly to the count, so estimates from groups of this size can change noticeably when only a few outcomes differ. The confidence intervals are therefore important alongside the observed counts and point estimates.

For LIPO-5/MVA HIV-B compared with MVA HIV-B/LIPO-5, the risk-difference interval includes no difference and extends in both directions. This indicates uncertainty not only about the magnitude but also about the direction of the difference. For GTU-MultiHIV B/LIPO-5, the interval is entirely below no difference; for GTU-MultiHIV B/MVA HIV-B, it is entirely above no difference. Those directions are supported by the respective unadjusted intervals, but the intervals still allow a range of possible effect sizes.

General: A confidence interval is not a guarantee that the underlying effect lies within its endpoints. It is a method for expressing uncertainty under a statistical procedure and its assumptions. The registry does not report the interval method used for the sponsor analysis, and the posted intervals accompany simple unadjusted re-calculations. They should not be read as adjusted intervals or as accounting for all sources of uncertainty, including multiple comparisons or missing outcomes.

The small arm sizes and the discrepancy between the sponsor’s reported analysis and the re-calculations make precision and analysis provenance especially relevant. The counts and intervals describe the observed endpoint data; they do not establish how consistent the response differences would be in another group of volunteers.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm, not total events. They are illustrations based on the observed effect for the pairwise comparisons where planning calculations are reported. The assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect used for each planning scenario is the stated fraction of the observed effect. The registry record does not report the assumptions used for the trial’s original design or its own sample-size justification.

The 80% power column gives the estimated participants per arm for a hypothetical study to have an 80% chance of detecting the specified effect if that effect and the planning assumptions were correct. The 90% power column gives the corresponding number for a higher chance of detection. These are planning targets, not descriptions of the number actually enrolled or assessed. This is not observed power.

The scenarios consider planning for 100%, 75%, and 50% of the observed effect. When the target effect is smaller, more participants are needed to distinguish it from random variation under the same assumptions. This pattern is especially important when the observed effect comes from small groups, because it may not be a stable estimate of the effect a future study should target.

For the comparison of LIPO-5/MVA HIV-B with MVA HIV-B/LIPO-5, the planning figures are 407 participants per arm for 80% power and 545 for 90% power at 100% of the observed effect; for 75%, they are 716 and 958; and for 50%, they are 1,589 and 2,126. For GTU-MultiHIV B/MVA HIV-B compared with MVA HIV-B/LIPO-5, the corresponding figures are 35 and 46 at 100%, 62 and 82 at 75%, and 137 and 182 at 50%. No planning figures are reported for the GTU-MultiHIV B/LIPO-5 comparison.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.333; treatment proportion p₁ = 0.429; average p̄ = 0.381.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6868 = 1.3460.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6835 = 0.5752.
4. n = (1.3460 + 0.5752)² / (0.0952)² = 407.0, rounded up to 407 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.333333333333333                          # control proportion
p1 <- 0.428571428571429                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(2038842)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 407 (80% power), 545 (90% power)
 75% of observed effect: n per arm = 716 (80% power), 958 (90% power)
 50% of observed effect: n per arm = 1,589 (80% power), 2,126 (90% power)
power.prop.test: 407 
Simulated power at n = 407 per arm: 0.804 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Analyses Reported by the Sponsor

The registry record reports a Binomial analysis across all four groups with p = 0.02. It does not provide the full specification of that analysis, such as the exact test procedure or any adjustments. The posted pairwise estimates and p-values are simple unadjusted re-calculations from the arm-level response counts and compare each non-reference arm with MVA HIV-B/LIPO-5.

These analyses do not ask precisely the same question. An overall analysis across four groups assesses evidence of variation among the groups as a set, whereas each pairwise re-calculation focuses on one comparison with the reference arm. The pairwise re-calculations also do not include adjustments for multiple comparisons. Their results should not be substituted for the sponsor’s reported analysis.

The record indicates a discrepancy between the sponsor analysis and the re-calculations. In that circumstance, the sponsor’s own analysis takes precedence. The re-calculations can still help describe how the posted group counts compare, but they should be presented as unadjusted analyses of summary statistics rather than as the trial’s definitive inferential analysis. The available record does not explain the discrepancy or provide enough detail to reconcile the analyses.

14. Multiplicity, Missing Data and Analysis Populations

Four prime-boost combinations were evaluated, and the posted re-calculations compare three of them with MVA HIV-B/LIPO-5. Testing several comparisons can increase the chance that at least one result appears statistically unusual even when there is no underlying difference. Statistical methods can address this issue through prespecified testing strategies or adjustments, but the registry record does not report a multiplicity adjustment or a planned hierarchy of comparisons. The unadjusted pairwise p-values should therefore be interpreted with this limitation in mind.

The record does report an overall Binomial analysis across the four groups with p = 0.02. It does not give enough detail to establish the test specification or whether any additional analyses were planned. The sponsor’s analysis takes precedence over the re-calculated pairwise results because the registry indicates a discrepancy, but the available details do not allow a full assessment of how multiplicity was handled in the sponsor analysis.

Missing-data handling is also not reported. The endpoint counts include the arm denominators listed for Week 30, but the registry does not explain whether any randomized participants lacked a Week 30 result, how unevaluable assays were treated, or whether missing responses were counted as nonresponses. Nor does it describe imputation or sensitivity analyses.

The analysis population is not specified in the primary endpoint description. The record does not state whether the primary analysis was intention-to-treat, per-protocol, or based on participants with an available assay result. The relationship between the total enrollment and the reported endpoint denominators is not explained. Without these details, it is difficult to assess how participant exclusions, protocol deviations, or missing measurements might affect the comparison.

The registry also does not report interim analyses, the original statistical analysis plan, or the trial’s own sample-size justification. These details would help clarify which comparisons were planned, which populations were intended for analysis, and how the sponsor’s reported result should be interpreted.

15. 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: blood creatine phosphokinase increased (8.7%), blood urine present (8.7%), anaemia (4.3%), gilbert's syndrome (4.3%), asthenia (4.3%).

The registry reports serious adverse events for groups with 23 participants at risk in each arm. The numbers affected were 3 for MVA HIV-B/LIPO-5, 5 for LIPO-5/MVA HIV-B, 7 for GTU-MultiHIV B/LIPO-5, and 3 for GTU-MultiHIV B/MVA HIV-B. The reported number of deaths was 0 in each arm. These are counts of participants reported as affected, not counts of all events.

The listed serious event terms include Blood creatine phosphokinase increased and Blood urine present, each reported as 8.7%. Anaemia, Gilbert's syndrome, Asthenia, and Vitiligo are each reported as 4.3%. These terms describe events recorded in the registry; the record does not provide enough detail to determine their timing, severity beyond the serious-event classification, relationship to vaccination, or whether any participant experienced more than one listed term.

Serious adverse event counts should not be confused with counts of all adverse events or with event severity ratings. The registry separately reports other adverse events, but these serious-event figures alone do not summarize the full safety profile. The counts are descriptive and should be interpreted in the context of the small reported at-risk groups.

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

The primary endpoint indicates whether an HIV-specific IFN-γ ELISPOT response was detected at Week 30 under the registry’s definition. The observed counts varied among the four combinations: 7 of 21, 9 of 21, 0 of 19, and 14 of 21. The unadjusted pairwise re-calculations show different directions and levels of uncertainty relative to MVA HIV-B/LIPO-5. The sponsor’s reported overall analysis is p = 0.02, and the registry record indicates that it differs from the re-calculations.

These results concern an assay-based immune response in healthy volunteers at low risk of HIV infection. They do not establish whether a combination prevents HIV infection, how long any response lasts, or whether the response translates into a clinical benefit. The primary endpoint is not a measure of vaccine effectiveness against infection.

The observed response counts are also not proof that the underlying response rates for the combinations will be the same in other participants. The groups were small, and the reported re-calculated confidence intervals reflect uncertainty in the pairwise estimates. The record does not report multiplicity handling, missing-data methods, or the analysis population, which further limits interpretation of the pairwise findings.

The record includes secondary objectives concerning tolerance and the type of vaccine-induced T-cell response, but the corresponding outcome results are not reported in the primary data. Serious adverse event counts are reported, including no deaths in the four groups, but they are descriptive and do not by themselves establish causation or provide a complete account of safety.

In short, the registry reports immune-response counts and a sponsor analysis across four combinations. It does not support conclusions about prevention of HIV infection, and the difference between the sponsor analysis and the simple re-calculations should remain explicit when interpreting the posted results.

17. Why This Trial Matters Statistically

This trial is statistically useful because it compares four prime-boost sequences rather than assessing only one vaccine schedule. The endpoint definition is explicit: a participant is counted as a responder if at least one of the Env, Gag, Pol, or Nef peptide pools produces a positive HIV-specific IFN-γ ELISPOT response at Week 30. That makes the outcome interpretable as a binary immune-response measure, while also making clear that it does not describe response magnitude or clinical protection.

The arm-level counts vary, and the comparison between the two GTU-MultiHIV B combinations is particularly notable descriptively: the reported responder count is 0 of 19 with LIPO-5 as the boost and 14 of 21 with MVA HIV-B as the boost. The trial’s small groups mean that these counts should be considered with uncertainty rather than treated as precise estimates of future response rates.

The registry also illustrates why statistical provenance matters. It reports an overall Binomial result across the four groups and the posted re-calculations provide unadjusted comparisons with one reference arm. Those are distinct analyses, and the record indicates a discrepancy. Keeping the sponsor’s result separate from re-calculated contrasts helps readers avoid treating an exploratory or simplified comparison as if it were the trial’s full statistical analysis.

Finally, the record makes visible several interpretive questions that arise in randomized trials: how multiple comparisons were handled, which analysis population was used, how missing outcomes were treated, and what assumptions informed the original design. These details are not reported in the registry record. Their absence limits reproducibility, but identifying that uncertainty is itself an important part of statistical interpretation.

18. 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 reported endpoint groups are small, with 19 or 21 participants in each arm. As a result, the response counts and estimated differences may be sensitive to a small number of participants. The confidence intervals should be considered alongside the point estimates, and the observed arm differences should not be treated as precise estimates of responses in a broader population.

The record does not report how the 92 enrolled participants relate to the arm-level denominators used for the Week 30 endpoint. It does not specify whether all randomized participants were included in the primary analysis, whether missing or unevaluable assays occurred, or how such results were handled. This makes it difficult to assess the potential effect of missing outcome data or exclusions.

The trial was open-label and used no masking. The primary endpoint was based on a laboratory assay, but the record does not state whether assay personnel were blinded to group assignment or provide further information on assay quality control. Randomization methods and allocation concealment are also not reported.

The registry reports an overall Binomial analysis and unadjusted pairwise re-calculations, with a stated discrepancy between them. Details needed to reconcile the results, including the full statistical plan, multiplicity handling, and analysis population, are not reported. Nor are the trial’s original sample-size assumptions or planned interim analyses.

These results address an immunogenicity endpoint at Week 30 in healthy volunteers at low risk of HIV infection. They do not directly measure infection prevention, and the reported serious adverse event counts do not establish causality or fully characterize safety. Secondary objectives are listed, but their results are not reported with the primary outcome information.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Immunogenicity and Safety of 4 Prime-boost Combinations of HIV Vaccine Candidates in Healthy Volunteers — NCT02038842.
• Registry record last updated 2026-04-14; status last verified 2025-09.
• Results first posted 2025-10-14.
• 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.

22. Record Summary

NCT02038842 is a completed randomized, open-label, multicenter Phase I/II parallel-group trial of four HIV vaccine prime-boost combinations in healthy volunteers at low risk of HIV infection. The registry reports 92 enrolled participants, with a start date of 2014-03 and a primary completion date of 2015-10.

The primary endpoint counted participants with an HIV-specific IFN-γ ELISPOT response at Week 30. Reported counts were 7 of 21 for MVA HIV-B/LIPO-5, 9 of 21 for LIPO-5/MVA HIV-B, 0 of 19 for GTU-MultiHIV B/LIPO-5, and 14 of 21 for GTU-MultiHIV B/MVA HIV-B. The registry reports an overall Binomial analysis with p = 0.02. The re-calculated pairwise comparisons are unadjusted, and the sponsor’s own analysis takes precedence because the record indicates a discrepancy.

The registry also reports serious adverse event counts and no deaths in the four arms. Details about missing-data handling, analysis populations, multiplicity, and the trial’s original sample-size justification are not reported in the registry record.