A Phase 2 Dose-finding Study of Atacicept in Subjects With Rheumatoid Arthritis (AUGUST I): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Phase 2 Dose-finding Study of Atacicept in Subjects With Rheumatoid Arthritis (AUGUST I) (NCT00430495), a randomized phase 2 study. The primary outcome is Percentage of Participants Achieving American College of Rheumatology 20 Response Based on C-reactive Protein (ACR20-CRP) at Week 26.

2. Clinical Question

3. Trial Design

This was a phase 2, randomized, double-blind, placebo-controlled, multicentre, parallel-arm dose-finding trial. The registry record describes it as prospective and lists atacicept and placebo matched to atacicept as the study interventions. The registered condition was rheumatoid arthritis. The reported enrollment was 256.

Four groups were listed: placebo and atacicept 25 mg, 75 mg, and 150 mg. For the primary Week 26 outcome, the posted analysis groups were 62.0 for placebo, 66.0 for atacicept 25 mg, 62.0 for atacicept 75 mg, and 64.0 for atacicept 150 mg. These are the outcome-group sizes; they need not represent every participant enrolled or every analysis population. The registry record does not explain any difference between enrollment and the group sizes reported for this outcome.

The primary outcome was the percentage achieving ACR20-CRP response at Week 26. Other listed outcomes included more stringent ACR responses, DAS28-CRP thresholds, HAQ-DI improvement, EULAR response, and adverse events. The record marks the trial completed. It lists a start in 2006-12 and primary completion in 2009-09.

The primary population is described as ITT participants who were randomized and received at least 1 treatment dose. The registry record does not report the original statistical analysis plan, allocation concealment procedures, treatment adherence, or how missing Week 26 outcomes were handled. It also does not report the protocol's own sample-size justification or original design assumptions.

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: A total of 456 subjects were screened, of whom 256 were enrolled and randomized of which 254 received the trial medication and included in Intention to Treat (ITT) population.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: ITT population included all randomized participants who received at least 1 treatment dose.

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.

Population: Intention-to-treat (ITT) population included all randomized participants who received at least one treatment dose.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants Achieving American College of Rheumatology 20 Response Based on C-reactive Protein (ACR20-CRP) at Week 26, measured over Week 26.

The primary endpoint was the percentage of participants achieving an ACR20-CRP response at Week 26. It is a binary response measure: each participant either met the response definition or did not. The reported percentage therefore describes the proportion of the analysis group meeting the specified criteria, rather than the average change in a continuous measure.

The definition requires at least 20 percent improvement in both tender joint counts, based on 68 joints, and swollen joint counts, based on 66 joints. It also requires at least 20 percent improvement in at least 3 of 5 measures: participant assessment of pain, participant global assessment of disease activity, physician global assessment of disease activity, participant assessment of physical function, and the acute-phase marker CRP.

As a general statistical concept, a composite response endpoint can capture improvement across multiple aspects of a condition in one outcome. Requiring improvement across joint counts and additional measures makes the response more demanding than relying on a single component alone. At the same time, an overall response percentage does not show which individual components changed, or how much each component changed. The endpoint is useful for comparing the proportion meeting a defined response threshold across randomized groups, but it does not by itself describe every dimension of disease activity or the pattern of change within participants.

The registry record reports percentages at Week 26, but does not report a continuous change-from-baseline estimate for the primary endpoint or a more detailed component-level analysis.

7. Results Posted in the Registry

• Atacicept 25 mg: 30.3% (20/66) versus 29.0% (18/62) on control; risk difference 1.3 percentage points (95% CI -14.6 to 17.1), risk ratio 1.04 (0.61–1.78); unadjusted two-proportion z-test p = 0.875.
• Atacicept 75 mg: 27.4% (17/62) versus 29.0% (18/62) on control; risk difference -1.6 percentage points (95% CI -17.5 to 14.2), risk ratio 0.94 (0.54–1.66); unadjusted two-proportion z-test p = 0.842.
• Atacicept 150 mg: 39.1% (25/64) versus 29.0% (18/62) on control; risk difference 10.0 percentage points (95% CI -6.4 to 26.5), risk ratio 1.35 (0.82–2.21); unadjusted two-proportion z-test p = 0.235.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Percentage of Participants Achieving American College of Rheumatology 50 Response Based on CRP (ACR50-CRP) at Week 26

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Atacicept 25 mg: 13.6% (9/66) versus 6.5% (4/62) on control; risk difference 7.2 percentage points (95% CI -3.1 to 17.5), risk ratio 2.11 (0.69–6.51); unadjusted two-proportion z-test p = 0.179.
  • Atacicept 75 mg: 11.3% (7/62) versus 6.5% (4/62) on control; risk difference 4.8 percentage points (95% CI -5.1 to 14.8), risk ratio 1.75 (0.54–5.68); unadjusted two-proportion z-test p = 0.343.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Atacicept 25 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Atacicept 75 mg and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Participants Achieving American College of Rheumatology 70 Response Based on CRP (ACR70-CRP) at Week 26

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Atacicept 25 mg: 6.1% (4/66) versus 0.0% (0/62) on control; risk difference 6.1 percentage points (95% CI 0.3 to 11.8); unadjusted two-proportion z-test p = 0.049.
  • Atacicept 75 mg: 4.8% (3/62) versus 0.0% (0/62) on control; risk difference 4.8 percentage points (95% CI -0.5 to 10.2); unadjusted two-proportion z-test p = 0.080.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Atacicept 25 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Atacicept 75 mg and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Participants Achieving Disease Activity Score in 28 Joints (DAS28) Based on CRP (DAS28-CRP) of Less Than or Equal to (<=) 3.2 at Week 26

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Atacicept 25 mg: 10.6% (7/66) versus 9.7% (6/62) on control; risk difference 0.9 percentage points (95% CI -9.5 to 11.4), risk ratio 1.10 (0.39–3.08); unadjusted two-proportion z-test p = 0.862.
  • Atacicept 75 mg: 9.7% (6/62) versus 9.7% (6/62) on control; risk difference 0.0 percentage points (95% CI -10.4 to 10.4), risk ratio 1.00 (0.34–2.93); unadjusted two-proportion z-test p = 1.000.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Atacicept 25 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Atacicept 75 mg and the control. The interval gives the range of effects that remain compatible with the data.

The listed secondary measures cover several kinds of response: more stringent ACR thresholds, DAS28-CRP cutoffs, HAQ-DI improvement, EULAR response, and adverse events. Together, they offer different views of response and safety rather than repeating the primary endpoint exactly. The registry record provides numerical results for some of these outcomes, but not complete numerical comparisons for every listed measure.

The broad pattern is mixed rather than uniformly aligned across endpoints and treatment groups. Some reported secondary response percentages are numerically higher in an atacicept group than in placebo, while other group percentages are similar. The higher-threshold ACR outcomes are sparse, and the DAS28-CRP outcomes do not show a consistent numerical separation across the reported dose comparisons. These endpoint patterns should be considered alongside the primary result, whose comparative estimates are imprecise and whose confidence intervals include no difference.

Secondary endpoints can support context about whether a primary-endpoint pattern appears across other measures, but they do not automatically confirm it. The endpoints differ in definition and in the degree of response required. In addition, examining multiple outcomes and treatment comparisons increases the chance that some unadjusted results may appear noteworthy by chance. The registry record does not report a multiplicity adjustment, an outcome hierarchy, or a prespecified interpretation plan for the secondary endpoints.

Accordingly, unadjusted secondary findings carry less inferential weight than a prespecified primary analysis, particularly when there are several outcomes and the record does not document how they were prioritized. The registry record also does not report numerical results for all listed secondary measures, so a complete cross-endpoint assessment is not possible from the posted information. The secondary results are best read as complementary descriptive evidence, not as a replacement for the uncertain primary comparisons.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Percentage of Participants Achieving American College of Rheumatology 50 Response Based on CRP (ACR50-CRP) at Week 26
• Percentage of Participants Achieving American College of Rheumatology 70 Response Based on CRP (ACR70-CRP) at Week 26
• Percentage of Participants Achieving Disease Activity Score in 28 Joints (DAS28) Based on CRP (DAS28-CRP) of Less Than or Equal to (<=) 3.2 at Week 26
• Percentage of Participants Achieving Disease Activity Score in 28 Joints (DAS28) Based on CRP (DAS28-CRP) of <=2.6 at Week 26
• Percentage of Participants Achieving Improvement in Health Assessment Questionnaire Disability Index (HAQ-DI) of at Least 0.3 From Baseline at Week 26
• Percentage of Participants With Good or Moderate European League Against Rheumatism (EULAR) Response at Week 26
• Number of Participants With Adverse Events (AEs) and Serious Adverse Events (SAEs)

9. Statistical Interpretation

10. Statistical Methods Explained

What does the response percentage represent?

The primary outcome is binary at the participant level: the ACR20-CRP definition is either met or not met by Week 26. The group percentage expresses how common a qualifying response was among the participants included in that outcome group. The posted primary counts were 18 responses among 62.0 placebo participants, 20 among 66.0 participants receiving 25 mg, 17 among 62.0 receiving 75 mg, and 25 among 64.0 receiving 150 mg.

As a general statistical concept, a binary outcome is often summarized both as a proportion within each group and as a contrast between groups. The group percentages are descriptive. A risk difference expresses the difference in response probability between an atacicept dose and placebo; a positive value means the estimated response probability is higher in the dose group, while a negative value means it is lower. The recalculated primary contrasts are expressed as proportions, not percentage-point values.

How should the confidence intervals and p-values be understood?

A confidence interval gives a range of effects compatible with the observed data under the method used to calculate it. A wide interval indicates limited precision. For each primary dose comparison, the 95% interval includes no difference and includes values on both sides of it. This means the observed estimate is uncertain; it does not prove that the groups are identical.

The reported p-values for the primary comparisons are unadjusted recalculations from posted counts. A p-value describes how compatible the data are with a specified no-difference model under the assumptions of the test. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of an effect. Interpretation should consider the effect estimate and interval as well as the p-value. The registry record does not report the sponsor's inferential analysis for these comparisons.

Why are the recalculations unadjusted?

The contrasts are recalculated from the summary counts in each group without covariate adjustment. Randomization supports comparisons between groups, but a recalculation based only on group counts does not incorporate any adjustment that may have been used in a sponsor analysis. It also does not account for a possible prespecified testing hierarchy or multiplicity correction. Because the registry record reports no sponsor analysis for these comparisons, the recalculated results are descriptive inferential summaries rather than a reconstruction of the full analysis plan.

As a general statistical concept, adjustment can account for baseline variables or design features that were specified for analysis. Whether adjustment is appropriate depends on the trial protocol and analysis plan; those details are not reported in the registry record. The posted counts also do not provide individual participant data, which would be needed for many adjusted analyses.

What does randomization contribute?

Random assignment is intended to make treatment groups comparable, on average, with respect to characteristics that could affect outcomes. Double-blinding can help limit some differences in care and outcome assessment. The record also describes quadruple masking. It does not specify the roles masked, the allocation sequence method, or the concealment procedures. These design features support a comparative interpretation, but they do not eliminate random uncertainty, missing-data concerns, or the need to evaluate how outcomes were analyzed.

The study was parallel-arm: participants were assigned to one listed group rather than receiving a sequence of treatments as part of a crossover design. The dose-finding purpose makes comparison across several atacicept doses relevant, but the record does not report a formal test of trend or identify a particular dose comparison as the only planned inferential test.

What limits the available analysis?

The primary population includes randomized participants who received at least 1 treatment dose. That definition is narrower than all randomized participants if any randomized participant received no dose. The registry record does not report whether this occurred, how many participants had missing Week 26 responses, or the method used to handle missing outcomes. Consequently, the posted response percentages cannot be assessed for sensitivity to different missing-data assumptions from the information in the record.

The primary outcome is assessed at a single reported time point, Week 26. The registry record does not report repeated-measures modeling, interim analyses, or the original analysis plan. As a general principle, a clearly prespecified primary endpoint and analysis help distinguish confirmatory comparisons from exploratory ones. Without those protocol details, the posted results can be described, but some aspects of the inferential framework remain unknown.

11. Confidence Intervals and Precision

Precision is limited for the primary treatment contrasts. The recalculated risk difference for atacicept 25 mg versus placebo is 0.013, with a 95% confidence interval from -0.146 to 0.171. For 75 mg, it is -0.016, with a 95% confidence interval from -0.174 to 0.142. For 150 mg, it is 0.100, with a 95% confidence interval from -0.064 to 0.265. All three intervals cross no difference and extend to effects in both directions.

These intervals are wide relative to the point estimates. For the 25 mg and 75 mg comparisons, values compatible with the data include differences that favor either placebo or atacicept. The interval for 150 mg also includes no difference and values in either direction, despite its positive point estimate. The observed ranking of percentages is therefore more definite than the evidence about the underlying differences: the data identify the observed percentages, but do not estimate the comparative effects narrowly.

As a general statistical concept, a confidence interval is not a range containing a known proportion of individual outcomes, and it does not assign a probability to the true effect after observing the data. It describes uncertainty from sampling under the method's assumptions. Here, the intervals come from unadjusted recalculations using the posted summary statistics. The registry record does not report sponsor confidence intervals for the primary contrasts, and the recalculations should be read with that limitation in mind.

The group sizes reported for the primary outcome are 62.0, 66.0, 62.0, and 64.0. With response proportions based on groups of this scale, a small number of participants changing response status could affect the observed percentages. The exact impact cannot be quantified without a separate analysis, and no new estimate is implied by this observation. The key message is that the point estimates are uncertain, rather than that one treatment group is established as superior or equivalent.

12. 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 sample-size illustrations describe participants per arm for comparisons between one atacicept dose and placebo. They are not total sample sizes across all four groups. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The assumed effect to detect is the stated fraction of the observed effect.

The 80% power column gives an illustrative number of participants per arm under those assumptions for an 80% chance of detecting the specified effect if that effect were the true one. The 90% power column gives the corresponding illustration for a 90% chance. Power is a property of a planned design under specified assumptions; it is not a measure of how likely the observed result is to be correct.

The illustrations are based on the observed risk differences, so their size can vary greatly by comparison. When the effect used for planning is reduced to a smaller fraction of the observed effect, the required number of participants per arm increases. This pattern reflects the general statistical principle that smaller differences are harder to distinguish from random variation. For some comparisons, the observed difference is close to zero, producing very large planning illustrations when the target effect is a fraction of that small observed difference.

These figures are planning illustrations, not observed power. The registry record does not report the trial's original sample-size justification or its own design assumptions. The displayed assumptions describe these retrospective illustrations only; they should not be attributed to the trial protocol or treated as evidence that the study was designed around the observed results.

13. 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 = 0.290; treatment proportion p₁ = 0.303; average p̄ = 0.297.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6460 = 1.2661.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6459 = 0.5436.
4. n = (1.2661 + 0.5436)² / (0.0127)² = 20,282.2, rounded up to 20,283 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 <- 0.290322580645161                          # control proportion
p1 <- 0.303030303030303                          # 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(430495)
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 = 20,283 (80% power), 27,152 (90% power)
 75% of observed effect: n per arm = 35,947 (80% power), 48,122 (90% power)
 50% of observed effect: n per arm = 80,627 (80% power), 107,935 (90% power)
power.prop.test: 20282.2 
Simulated power at n = 20,283 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.

14. Multiplicity, Missing Data and Analysis Populations

The trial has a primary endpoint and several listed secondary endpoints, with comparisons across three atacicept doses and placebo. Considering multiple outcomes and treatment contrasts raises a multiplicity issue: even when each individual test uses a conventional threshold, the chance of at least one apparently unusual result can increase when many tests are examined. As a general statistical concept, multiplicity procedures or a prespecified testing hierarchy can control how confirmatory claims are made across a family of tests.

The registry record does not report a multiplicity adjustment, a hierarchy among secondary outcomes, or whether the primary dose comparisons were tested as a defined family. The posted recalculated p-values are unadjusted. They should not be interpreted as though they account for all outcome and dose comparisons in the trial. The record also does not report interim analyses or stopping rules, so their role in the analysis cannot be assessed.

The primary population is defined as all randomized participants who received at least 1 treatment dose. This definition is stated in the registry record, but the record does not describe an analysis including every randomized participant regardless of treatment received. It does not report the number of missing Week 26 observations, reasons for missingness, the handling of withdrawals, or any sensitivity analyses for missing data. These omissions limit assessment of whether the reported response percentages could be affected by missing outcomes.

As a general statistical concept, missing-data methods rely on assumptions about why outcomes are unavailable. Different assumptions can lead to different estimates, especially when the outcome is measured at a fixed follow-up time. Without the missingness and analysis details, readers cannot evaluate the robustness of the primary result under alternative assumptions. The registry record also does not report a full statistical analysis plan or explain whether the endpoint population definition changed from the protocol.

15. 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 serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: rheumatoid arthritis (3.2%), hip fracture (3.2%), back pain (1.6%), osteonecrosis (1.6%), femur fracture (1.6%).

The posted serious adverse event counts were 3 in the placebo group, 9 with atacicept 25 mg, 8 with atacicept 75 mg, and 5 with atacicept 150 mg. The corresponding groups at risk were 62, 66, 62, and 64. These counts describe participants with serious adverse events as recorded; they are not a measure of causation and do not by themselves establish a comparative safety effect.

Other adverse event counts were 15 with placebo, 25 with atacicept 25 mg, 16 with atacicept 75 mg, and 26 with atacicept 150 mg. The registry record lists serious adverse event terms including rheumatoid arthritis, hip fracture, back pain, osteonecrosis, femur fracture, and acute myocardial infarction, with the displayed percentages attached to those terms. The posted information does not give a complete narrative context for each event or establish whether an event was considered treatment-related.

Deaths are not reported in the registry record. The adverse event counts should be interpreted with the group sizes, follow-up information, and definitions in mind; details of exposure duration and event-collection procedures are not reported in the registry record. These figures describe events recorded during the trial and are not, alone, a comprehensive comparison of safety.

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

The primary result answers a narrow question: what proportion in each registered group met the ACR20-CRP response definition at Week 26, and what are the unadjusted recalculated differences versus placebo? The observed percentages range from 27.4% to 39.1% across the four groups. The largest observed percentage is in the 150 mg group, while the 25 mg and 75 mg observations are close to the placebo value.

The comparison estimates are uncertain. Their confidence intervals include no difference and extend in both directions. Thus, the posted results do not precisely establish the size or direction of the underlying treatment differences. A point estimate above placebo is not sufficient to establish a reliable difference, and an interval that includes no difference does not prove equivalence or absence of an effect.

The endpoint also has a particular meaning: a participant must meet the joint-count and additional-measure criteria to be classified as a responder. The percentage does not show the degree of improvement among those who responded, the changes among nonresponders, or the separate behavior of each component. Nor does one Week 26 response endpoint describe all possible outcomes relevant to rheumatoid arthritis.

The registry record is a limited basis for judging the full analysis. It does not report the original statistical plan, missing-data methods, multiplicity approach, or sponsor analysis of the primary comparisons. Secondary information is incomplete for some listed outcomes. These gaps restrict how far the posted numbers can support broader conclusions.

The results should therefore be understood as recorded outcome percentages and imprecise, unadjusted comparisons in this phase 2 dose-finding trial. They do not by themselves establish treatment benefit, lack of benefit, equivalence, or a definitive dose-response relationship. Interpretation of treatment choices requires evidence beyond these registry summaries; no clinical recommendation follows from these statistical descriptions.

17. Why This Trial Matters Statistically

This trial illustrates a central challenge in dose-finding studies: several active-dose groups can produce a visible ordering of observed response percentages, while uncertainty around each comparison remains large. The 150 mg group has the highest posted primary response percentage, but its unadjusted recalculated risk difference has a confidence interval that includes no difference and effects in either direction. The apparent ordering should not be mistaken for a precise estimate of a dose effect.

It also shows why a binary endpoint needs more than a percentage alone. Response counts and group sizes help describe the data, while a contrast and confidence interval show how uncertain the comparative estimate is. A p-value on its own cannot convey the magnitude, direction, or range of plausible effects. The primary comparisons demonstrate the value of reading these measures together.

The trial's several doses and multiple secondary outcomes raise questions about how results were prioritized. Without a reported multiplicity plan, isolated unadjusted findings have a more exploratory interpretation. Missing information about analysis populations and incomplete follow-up reporting also makes transparency important: readers need those details to evaluate whether the posted summary fully represents the planned analysis.

Finally, retrospective sample-size illustrations based on observed effects are not substitutes for an original design rationale. They can show how planning requirements change with the assumed effect size, but they do not reveal what assumptions investigators used before the trial or how much power the completed study had. This distinction is essential when interpreting a small phase 2 study and considering what additional evidence would be needed for a more precise estimate.

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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The registry record leaves several trial-specific uncertainties. Although enrollment is listed as 256, the primary outcome group sizes are 62.0, 66.0, 62.0, and 64.0. The record does not explain the relationship between the enrollment figure and the outcome-analysis groups, or report how many participants were randomized but did not receive treatment.

The primary population is defined as randomized participants who received at least 1 treatment dose, but the record does not state how missing Week 26 outcome values were handled. It also does not provide reasons for missingness, participant flow by treatment group, or sensitivity analyses. These details would help assess whether the observed response percentages represent all participants intended for the primary analysis.

The registry record does not report the sponsor's primary inferential analysis, the original statistical analysis plan, multiplicity procedures, or a formal dose-trend analysis. The risk differences and confidence intervals discussed are unadjusted recalculations from posted summary counts. They may not match an analysis using prespecified methods or participant-level data.

Although the trial is described as quadruple-masked, the record does not state which roles were masked or report allocation concealment details. It does not give dosing frequency, detailed treatment exposure, or full outcome-assessment procedures. Numerical results for all listed secondary endpoints are not available in the registry record, and deaths are not reported there. The safety summary also lacks the detail needed to interpret event timing, attribution, and exposure in depth.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: A Phase 2 Dose-finding Study of Atacicept in Subjects With Rheumatoid Arthritis (AUGUST I) — NCT00430495.
• Registry record last updated 2016-02-17; status last verified 2016-01.
• Results first posted 2016-02-17.
• 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

NCT00430495 is a completed, randomized, double-blind, placebo-controlled, multicentre, phase 2 parallel-arm dose-finding trial of atacicept in rheumatoid arthritis. The study enrolled participants with active disease and an inadequate response to a three month therapeutic trial with a TNFa antagonist due to lack of efficacy. Four groups were registered: placebo and atacicept 25 mg, 75 mg, and 150 mg.

The primary endpoint was ACR20-CRP response at Week 26. Posted response percentages were 29.0% for placebo, 30.3% for atacicept 25 mg, 27.4% for 75 mg, and 39.1% for 150 mg. Unadjusted recalculated comparisons have broad confidence intervals that include no difference. The registry record does not report the sponsor's analysis, missing-data methods, or multiplicity plan, and several listed secondary outcomes lack complete numerical results.