Dose Ranging of GSK2336805 in Combination Therapy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Dose Ranging of GSK2336805 in Combination Therapy (NCT01648140), a randomized phase 2 study. The primary outcome is Number of Participants Achieving eRVR.

2. Clinical Question

The clinical question is therefore about observed early viral-response counts across dose groups and treatment regimens, rather than a single comparison in one uniform population. The randomized genotype 1 groups can inform comparisons between the assigned regimens, while the separate genotype 4 cohort has no randomized comparator in this record. The eRVR result is a response measure at specified assessments; it does not, by itself, summarize every virologic outcome listed for the study. The record also lists sustained virologic response outcomes among the secondary measures, but the primary endpoint is eRVR.

3. Trial Design

This was a phase II, multicenter, parallel-group study described as randomized and double-masked. The registry records enrollment of 286 participants and a completed status. The study start date is 2012-08-01, and the primary completion date is 2014-07-01. The official title describes 12 weeks of oral GSK2336805 administration with pegylated interferon and ribavirin in treatment-naïve participants with chronic genotype 1 or genotype 4 hepatitis C.

The randomized genotype 1 groups received GSK2336805 40 mg or 60 mg, or telaprevir, with pegylated interferon alpha-2a and ribavirin. The registry separately describes a nonrandomized, single-arm genotype 4 cohort receiving GSK2336805 60 mg with pegylated interferon and ribavirin. Thus, randomization applies to the randomized study groups; it should not be assumed for the genotype 4 cohort. The registry record does not report the allocation ratio or the specific randomization and masking procedures.

The primary endpoint was the number of participants achieving eRVR at Week 4 and Week 12. Participants who discontinued before the Week 12 assessments or had missing HCV RNA values at Week 4 and Week 12 were treated as nonresponders. The stated intent-to-treat population required randomized assignment, receipt of at least 1 dose, and at least 1 post-Baseline HCV RNA measurement. These criteria describe the population used for the primary endpoint; they do not mean that every enrolled participant necessarily contributed to it.

The registry lists safety outcomes and several virologic and pharmacokinetic secondary outcomes. The original sample-size justification, planned allocation details, and other design assumptions 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.

Before assignment: Participants with treatment-naïve(TN) chronic genotype1(G1) hepatitis C virus (HCV) infection were randomly allocated on 2:2:1 basis to 2 dose levels of GSK2336805 or telaprevir. In a nonrandomized single-arm cohort, participants with TN genotype4(G4) chronic HCV infection were enrolled in parallel at dose level of 60 milligrams (mg) of GSK2336805.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent-To-Treat (ITT) Population: comprised of all participants who met the study criteria and were randomly assigned to treatment in the study with documented evidence of having received at least 1 dose of randomized treatment and at least 1 post Baseline HCV RNA measurement.

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 Number of Participants Achieving eRVR, measured over Week 4 and Week 12.

The primary endpoint counts participants achieving extended rapid virologic response, or eRVR. A participant met the definition only if plasma HCV RNA was below the lower limit of quantification and the target was not detected at both Week 4 and Week 12. The endpoint therefore combines findings at two assessment times rather than recording a response at only one visit.

The record specifies how certain incomplete assessments were handled: participants who discontinued before Week 12 or had missing HCV RNA values at Week 4 and Week 12 were classified as nonresponders. This rule makes missingness consequential to the counted response outcome. It does not establish how every other missing-data pattern was handled.

As a binary endpoint, eRVR can be summarized as the number of participants meeting the response definition out of the participants in each group. Such counts are straightforward to interpret, but their precision depends in part on the number of participants observed in each group. The endpoint was used to compare early response across the study groups; it is not a complete measure of safety, tolerability, or longer-term virologic response. Sustained virologic response at 12 and 24 weeks appears among the secondary outcomes, but it is distinct from the primary eRVR definition.

7. Results Posted in the Registry

• GSK2336805 60 mg, Genotype 1 HCV: 52.5% (21/40) versus 56.1% (23/41) on control; risk difference -3.6 percentage points (95% CI -25.3 to 18.1), risk ratio 0.94 (0.63–1.40); unadjusted two-proportion z-test p = 0.745.
• Telaprevir, Genotype 1 HCV: 52.9% (9/17) versus 56.1% (23/41) on control; risk difference -3.2 percentage points (95% CI -31.3 to 25.0), risk ratio 0.94 (0.56–1.59); unadjusted two-proportion z-test p = 0.826.
• GSK2336805 60 mg, Genotype 4 HCV: 69.2% (9/13) versus 56.1% (23/41) on control; risk difference 13.1 percentage points (95% CI -16.2 to 42.5), risk ratio 1.23 (0.79–1.94); unadjusted two-proportion z-test p = 0.401.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary: each participant was classified as achieving eRVR or not achieving it. The reported group results are counts of participants with the event, alongside the group sizes. This representation preserves the scale of the observed data and makes the response definition central to interpretation. Because missing values at the specified assessments were classified as nonresponse under the endpoint rule, the event count depends on both the virologic measurements and that rule.

What does a risk difference express?

A risk difference compares the response proportions in two groups by subtracting one proportion from the other. It is expressed on the probability scale, so zero corresponds to equal observed response proportions. A negative estimate means the first group named in the comparison has a lower observed response proportion than the reference group, while a positive estimate means a higher proportion. This is a comparison of proportions, not a difference in the number of participants, which can be misleading when group sizes differ.

The reported comparisons use the GSK2336805 40 mg genotype 1 group as the reference for the genotype 1 comparisons. The genotype 4 result is presented against that reference in the calculations, but genotype and randomization status differ. A numerical contrast between those groups cannot be interpreted as an isolated treatment effect because the comparison does not hold population and assignment method constant.

How should confidence intervals be read?

A confidence interval gives a range of values compatible with the observed data under the method used to calculate it. In general statistical terms, a wider interval indicates less precision than a narrower interval. For a risk difference, an interval that crosses zero includes values consistent with a negative difference, no difference, and a positive difference. All three reported risk-difference intervals cross zero. This does not demonstrate equivalence: an equivalence conclusion requires an appropriate design and a prespecified margin, neither of which is reported in the registry record.

What do risk ratios and odds ratios add?

The registry-derived comparisons also include risk-ratio and odds-ratio measures. A risk ratio compares the response proportions as a ratio; a value of one corresponds to equal proportions. An odds ratio compares the odds of response, where odds are the probability of response relative to the probability of nonresponse. These measures describe the same binary outcome on different scales and should not be treated as interchangeable. General statistical explanation: when outcomes are not rare, an odds ratio can look farther from one than a risk ratio even when both describe the same data. The risk difference is often the most direct scale for describing the absolute separation between response proportions.

What does a p-value tell us?

The listed p-values are unadjusted. In general statistical terms, a p-value describes how unusual data at least as incompatible with a specified null hypothesis would be if that null model and the test assumptions were correct. It does not measure the size or clinical importance of a difference, and it does not give the probability that a treatment works or that the null hypothesis is true. The registry record does not report the sponsor's statistical test, model, or adjustment strategy for these comparisons. The p-values should therefore be read alongside the estimates, intervals, group sizes, and endpoint definition, rather than as stand-alone conclusions.

What analysis population and missing-data rule apply?

The primary endpoint population is described as participants who met study criteria, were randomized, received at least 1 dose of randomized treatment, and had at least 1 post-Baseline HCV RNA measurement. This definition resembles an intent-to-treat population but includes explicit treatment-exposure and measurement requirements. The endpoint also classifies specified discontinuations and missing HCV RNA assessments as nonresponse. General statistical explanation: prespecified missing-data rules matter because different assumptions can alter the estimated response proportion, especially when missingness is frequent or differs between groups. The registry record does not report additional missing-data methods beyond the stated endpoint rule.

The registry lists no analysis entries. It does list 45 outcome measures, with 28 secondary outcomes posted. The statistical model, covariate adjustment, handling of any other incomplete data, and interim-analysis plan are not reported in the registry record. No assumptions about those procedures can be inferred from the fact that the study was randomized or double-masked.

Why can over-dispersion matter?

General statistical explanation: over-dispersion occurs when variability in observed data is greater than a statistical model expects. If a model assumes less variability than is present, standard errors may be too small and intervals or tests may appear more certain than warranted. This is a general issue in some count and clustered-data analyses. The primary endpoint is a participant-level binary count, and the registry record does not report an over-dispersion assessment or a model that would make such an assessment relevant. It would be inappropriate to claim that over-dispersion was tested or found in this trial.

10. Confidence Intervals and Precision

Precision is constrained by the number of participants in each group: 41 in the GSK2336805 40 mg genotype 1 group, 40 in the GSK2336805 60 mg genotype 1 group, 17 in the telaprevir genotype 1 group, and 13 in the genotype 4 group. The last two groups are especially small, leaving less information to estimate their response proportions and comparisons precisely.

The confidence intervals for all three reported risk differences cross zero and extend to values on both sides of it. This means that the data do not tightly locate the direction of the underlying difference. The intervals are broad relative to the estimates, so the results remain compatible with materially different response patterns. The estimates should not be read as precise forecasts for a new study or population.

Confidence intervals communicate uncertainty in the estimated comparison, not the range in which every participant's outcome falls. They also do not account for every source of uncertainty, such as unreported analysis choices or multiplicity. The registry record does not report a multiplicity adjustment or the original statistical analysis plan, which limits assessment of how these intervals and tests fit into the study's full analysis strategy.

For the genotype 4 cohort, uncertainty is compounded by its nonrandomized, single-arm design and its different participant genotype. Its response count can describe that cohort, but the comparison with the genotype 1 reference cannot isolate a treatment contrast. The registry-derived interval does not remove this design limitation.

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 a total across both arms and not total events. Each comparison has a set of illustrative participant counts for detecting the stated fraction of its observed effect. The curve presents the same planning relationship visually: as the effect assumed for planning becomes smaller, the required number of participants rises.

The calculation assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is the stated fraction of the observed effect. The 80% power column gives the number of participants per arm that the illustration estimates would be needed for an 80% chance of detecting that assumed effect under those assumptions. The 90% power column gives the corresponding per-arm count for a 90% chance. Higher target power generally requires more participants under otherwise identical assumptions.

The rows for 100%, 75%, and 50% of the observed effect are planning scenarios, not separate observed results. A smaller assumed effect requires more participants because subtle differences are harder to distinguish from sampling variability. The calculations are planning illustrations based on the observed comparison estimates, not the trial's original design justification. The original sample-size assumptions are not reported in the registry record. These figures are not observed power and should not be interpreted as evidence that the completed study had a particular probability of detecting its actual effect.

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 = 0.561; treatment proportion p₁ = 0.525; average p̄ = 0.543.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7045 = 1.3808.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.7040 = 0.5925.
4. n = (1.3808 + 0.5925)² / (-0.0360)² = 3,008.6, rounded up to 3,009 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 <- 0.560975609756098                          # control proportion
p1 <- 0.525                          # 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(1648140)
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 = 3,009 (80% power), 4,028 (90% power)
 75% of observed effect: n per arm = 5,341 (80% power), 7,150 (90% power)
 50% of observed effect: n per arm = 11,997 (80% power), 16,060 (90% power)
power.prop.test: 3008.6 
Simulated power at n = 3,009 per arm: 0.802 (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. Multiplicity, Missing Data and Analysis Populations

The primary endpoint has a specific missing-data rule: participants who discontinued before Week 12 assessments or had missing HCV RNA values at Week 4 and Week 12 were classified as nonresponders. This provides an explicit rule for those cases, but the registry record does not report how other missing-data patterns were handled or whether sensitivity analyses were conducted.

The primary endpoint population is described as participants who met study criteria, were randomized, received at least 1 dose, and had at least 1 post-Baseline HCV RNA measurement. This is not simply every person enrolled. The registry record does not report further details on exclusions, missing outcome counts, or whether additional analysis populations were used.

The record lists 45 outcome measures and posts 28 secondary outcomes. The secondary measures include adverse events, virologic response measures, and pharmacokinetic summaries. Testing or estimating many outcomes and making multiple comparisons can increase the chance that some results appear notable by chance. General statistical explanation: multiplicity procedures can control or describe this issue, but the appropriate approach depends on which outcomes and comparisons were prespecified and how they were prioritized.

A multiplicity strategy, adjusted p-values, and the full analysis plan are not reported in the registry record. The comparisons discussed are unadjusted, so they should not be treated as if they account for testing across all outcomes or contrasts. Interim analyses and any associated decision rules are also not reported. The record does not give enough information to determine whether the posted secondary results reflect every planned analysis or how any selection among analyses was made.

14. 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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: erysipelas (7.7%), dermatitis allergic (5.9%), myalgia (5.9%), anal abscess (5.9%), haemorrhoids (2.5%).

The registry reports participants with serious adverse events and participants with other adverse events in each group. Serious adverse events were reported for 2 of 41 participants receiving GSK2336805 40 mg in genotype 1, 2 of 40 receiving GSK2336805 60 mg in genotype 1, 3 of 17 receiving telaprevir in genotype 1, and 1 of 13 receiving GSK2336805 60 mg in genotype 4. The record also lists other adverse events for 41 of 41, 38 of 40, 17 of 17, and 13 of 13 participants, respectively.

These are participant counts, not a measure of event severity beyond the separate serious-event classification, and they do not establish that treatment caused an event. The registry lists serious-event terms including erysipelas, dermatitis allergic, myalgia, anal abscess, haemorrhoids, and hyperglycaemia, with percentages attached to those terms. The record does not establish from these listings whether a particular event was attributed to treatment. Death counts are not reported in the registry record.

Differences between groups should be interpreted cautiously because group sizes differ and the genotype 4 cohort was not randomized. The available counts alone do not characterize timing, duration, recurrence, or the full clinical context of each event.

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

The posted results describe eRVR counts in four study groups and give unadjusted comparisons against the GSK2336805 40 mg genotype 1 reference group. The genotype 1 randomized comparisons provide the clearest basis for comparing assigned regimens within the same stated genotype. Even there, the confidence intervals are broad and cross zero, so the estimates do not establish a clear direction or precise size of difference.

The genotype 4 group is a separate nonrandomized cohort. Its observed response count describes that group, but it cannot establish how the same participants would have responded under another regimen. Differences between it and the genotype 1 group may reflect population and design differences as well as treatment differences.

The results do not establish equivalence, prove that there is no difference, or establish that one regimen is superior. A non-significant p-value is not proof of equality. Likewise, the direction of an estimate is not, by itself, evidence of a reliable or clinically important effect. Interpretation must account for the endpoint definition, the response counts, uncertainty, and the comparison design.

eRVR is an early virologic endpoint defined at Week 4 and Week 12. It should not be treated as interchangeable with sustained virologic response, safety outcomes, tolerability, or pharmacokinetic measures. The registry lists those as separate study outcomes. Their findings would need to be assessed separately to answer questions beyond the early response count.

The registry record does not report the original sample-size justification, statistical analysis plan, multiplicity strategy, or interim analyses. Consequently, the planning illustrations should not be mistaken for the trial's original design target, and the unadjusted comparisons do not answer how the complete set of outcomes was intended to be interpreted.

16. Why This Trial Matters Statistically

This study illustrates why the design and population behind a comparison matter as much as the direction of its point estimate. The genotype 1 treatment groups were part of a randomized, double-masked, parallel-group study, whereas the genotype 4 group was described as a separate nonrandomized cohort. Treating all four groups as if they formed one randomized comparison would obscure that distinction.

It also shows the value of reading a binary endpoint with its denominator and its definition. eRVR required findings at both Week 4 and Week 12, and specified missing assessments were classified as nonresponse. A count of responders therefore reflects a rule-based endpoint, not merely the number of available positive tests.

The small group sizes and broad confidence intervals make precision an important part of the interpretation. A point estimate can look directional while remaining consistent with effects on either side of no difference. The unadjusted p-values alone cannot convey that uncertainty, and they do not establish equivalence or clinical importance.

Finally, the registry includes multiple outcome areas, from virologic response to safety and pharmacokinetics, but does not report the multiplicity strategy or full analysis plan. This is a reminder that posted results should be read in relation to the intended hierarchy of outcomes and the analysis decisions that are documented. Where those details are not reported in the registry record, conclusions should remain limited to the reported counts and comparisons.

17. 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 response counts come from groups of different sizes, with particularly limited numbers in the telaprevir and genotype 4 groups. This limits precision and makes the estimates sensitive to the outcomes of a small number of participants. The broad intervals reflect substantial uncertainty about the size and direction of the comparisons.

The genotype 4 cohort was nonrandomized and single-arm, while the comparison reference was genotype 1. This makes its contrast especially difficult to interpret as a treatment effect. Even though a numerical comparison is available, it does not separate treatment differences from differences in population or assignment.

The endpoint uses an explicit nonresponse rule for certain discontinuations and missing values, but the record does not report other missing-data methods or sensitivity analyses. The intent-to-treat population definition also requires treatment receipt and a post-Baseline HCV RNA measurement, so it is not identical to all enrolled participants.

The registry lists 45 outcome measures and posts 28 secondary outcomes, but the statistical analysis plan and multiplicity approach are not reported. The record also does not report the original sample-size justification, allocation ratio, detailed randomization or masking procedures, interim analyses, or the statistical test and model for the comparisons. These omissions limit the ability to assess whether the posted analyses match the planned analysis strategy.

Safety listings give participant counts and selected serious-event terms, but the registry record does not supply full context for attribution, timing, recurrence, or deaths. The eRVR endpoint is also distinct from longer-term response and cannot alone summarize the range of outcomes the study set out to assess.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Dose Ranging of GSK2336805 in Combination Therapy — NCT01648140.
• Registry record last updated 2017-06-02; status last verified 2016-11.
• Results first posted 2017-06-02.
• 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.

21. Record Summary

NCT01648140 is a completed phase II, multicenter, parallel-group study recorded as randomized and double-masked. It evaluated GSK2336805 with pegylated interferon alpha-2a and ribavirin in treatment-naïve participants with chronic genotype 1 or genotype 4 hepatitis C. The registry records enrollment of 286 participants.

The primary endpoint was the number achieving eRVR at Week 4 and Week 12, with specified missing assessments classified as nonresponse. In the genotype 1 groups, the reported counts were 23 of 41 for GSK2336805 40 mg, 21 of 40 for GSK2336805 60 mg, and 9 of 17 for telaprevir. The separate genotype 4 group had 9 of 13 participants achieving eRVR.

The reported comparison intervals cross zero, and the genotype 4 cohort was nonrandomized. The record does not report the original sample-size justification, multiplicity strategy, or full statistical analysis plan.