Study of ITCA 650 (Exenatide in DUROS) in Subjects With Type 2 Diabetes Mellitus: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of ITCA 650 (Exenatide in DUROS) in Subjects With Type 2 Diabetes Mellitus (NCT00943917), a randomized phase 2 study. The primary outcome is Mean Change in HbA1c (Per Protocol).

2. Clinical Question

The registry describes a study of ITCA 650 in subjects with type 2 diabetes, with safety, efficacy and tolerability as its stated purpose. The primary posted outcome is mean change in HbA1c from Day 0 to Week 12 in the Stage I per-protocol population. The posted results report that mean separately for ITCA 650 20 mcg/day, ITCA 650 40 mcg/day and exenatide injection.

The posted results answer this question descriptively through group means and standard deviations and comparatively through estimated mean differences, intervals and unadjusted p-values. They do not, by themselves, establish that the groups are equivalent or that a difference is clinically important. The registry record does not report a clinical threshold for interpreting HbA1c differences, or a prespecified hypothesis for these comparisons.

3. Trial Design

This is a phase 2, randomized, parallel-group study. The registry lists masking as none, so it was open-label according to the record. The sponsor is Intarcia Therapeutics. The study status is completed, and the registered enrollment is 155. The condition listed is type 2 diabetes.

The registry lists a broad set of interventions, including ITCA 650, exenatide injection and combination-labelled intervention entries. Its arm titles include several dose or sequence labels. The primary endpoint results, however, are posted for three Stage I groups: ITCA 650 20 mcg/Day, ITCA 650 40 mcg/Day and exenatide injection. The outcome values for those groups are based on the per-protocol population described for the endpoint, rather than a stated analysis of every enrolled participant.

The study start is recorded as 2009-08 and primary completion as 2010-08. The primary outcome timeframe is Day 0 and Week 12, and its description identifies the mean change over the first 12 weeks as Stage I. The registry lists six outcome measures but zero analyses. These record entries do not describe all study procedures or establish how every arm in the full study relates to the three groups with posted primary outcome results.

Randomization can help make treatment groups comparable at the start of a trial, but the registry entry alone does not supply details about sequence generation, allocation concealment or implementation. Those details are not reported in the registry record. Likewise, the original sample-size justification, planned enrollment rationale and full protocol are not reported in the registry record. The reported trial enrollment should not be conflated with the per-protocol group sizes used for the posted endpoint.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: At Stage II subjects divided into 6 total arms: ITCA 650 20 mcg/day assigned to either ITCA 650 20 mcg/day or ITCA 650 40 mcg/day; ITCA 650 40 mcg/day assigned to ITCA 650 40 mcg/day or ITCA 650 80 mcg/day; Ex Inj assigned to either ITCA 650 40 mcg/day or ITCA 650 60 mcg/day

Only the first 6 of 15 groups are shown.

Stage I

Reasons for not completing this period:

Stage II

Reasons for not completing this period:

Stage II Continuation

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Per protocol analysis- Only subjects with values at baseline and Week 12 endpoint are included.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Only the first 6 of 16 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Mean Change in HbA1c (Per Protocol), measured over Day 0 and Week 12.

The primary endpoint is titled Mean Change in HbA1c (Per Protocol). HbA1c is the continuous measure named in the registry; the registered unit is percent change. The outcome description specifies mean change from Day 0 to Week 12, over the first 12 weeks of Stage I.

A mean describes the central value of a continuous outcome within a group. The associated standard deviation describes variation among individual measurements around that group mean. Reporting both helps distinguish the group’s average change from the spread of participant-level outcomes. Neither statistic alone shows how much participants’ outcomes overlap across groups.

The registry defines the primary endpoint population as subjects with values at baseline and the Week 12 endpoint who are included in the per-protocol analysis. This is narrower than a general statement about everyone enrolled: a participant without both specified values would not meet the endpoint’s stated inclusion description. The reason any individual may be absent from that analysis is not reported in the registry record.

HbA1c is used in this trial as the named measure of change in type 2 diabetes. That explains the endpoint’s role in the study, but the record does not provide a threshold for clinical importance or a rule for deciding what size of change would matter to patients. The posted group summaries therefore support description of average change and statistical comparison, not a stand-alone judgment about clinical importance.

7. Results Posted in the Registry

• ITCA 650 40 mcg/Day - STAGE I: mean -1.04 versus -0.96 on control; difference -0.08 (95% CI -0.41 to 0.25), standardized effect d = -0.10; z-test p = 0.637.
• Exenatide Injection - STAGE I: mean -0.82 versus -0.96 on control; difference 0.14 (95% CI -0.22 to 0.50), standardized effect d = 0.16; z-test p = 0.445.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous, and the registry reports it as a mean with standard deviation. A mean summarizes the average change within a group. A standard deviation describes how dispersed individual values are around that mean. A small standard deviation relative to the mean difference can yield a more precise comparison, while considerable within-group variation can make differences harder to estimate precisely. These are general statistical principles; the trial-specific values are those reported in the outcome summaries.

What does a mean difference represent?

A mean difference compares the average outcome in one group with the average in another, using the stated comparison orientation. In the ITCA 650 dose comparison, the estimated difference is -0.08, with ITCA 650 40 mcg/Day compared with ITCA 650 20 mcg/Day. In the exenatide comparison, the estimated difference is 0.14, with exenatide injection compared with ITCA 650 20 mcg/Day. The estimate is a contrast between group means, not an individual participant’s change and not a percentage reduction computed from the posted means.

How should the intervals be read?

An interval around an estimate expresses uncertainty about the size and direction of a group difference under the calculation’s assumptions. The first comparison’s interval, -0.412 to 0.252, includes zero. The second comparison’s interval, -0.219 to 0.499, also includes zero. When an interval includes zero, the data do not pin down a direction of difference with high precision. An interval is not a list of observed individual outcomes, and it does not mean that every value inside it is equally plausible. The confidence level for these posted intervals is not reported in the registry record.

What does a p-value add?

The unadjusted p-values are 0.637 for the ITCA 650 dose comparison and 0.445 for the comparison with exenatide injection. In general, a p-value describes how unusual data at least as discrepant as those observed would be under a specified null model. It is not the probability that the null hypothesis is true, nor does it quantify clinical importance. Here, the values accompany recalculated unadjusted comparisons. The registry does not report a prespecified testing plan or an adjustment for the number of comparisons, so the p-values should not be treated as multiplicity-adjusted evidence.

Why can the group means differ while the evidence remains uncertain?

A point estimate is one best-fitting summary of a difference, but it has sampling uncertainty. The standard deviations describe outcome variation, and the group sizes used in each comparison also affect precision. Even when observed means differ, an interval may remain wide enough to include zero and values in either direction. This distinction between the estimated effect and its precision is central to interpreting the posted results. The standard deviations are not standard errors: a standard error concerns uncertainty in an estimate, whereas a standard deviation describes variation among observations.

What does recalculated mean?

The comparisons are unadjusted recalculations from the posted summary statistics. The 40 mcg/day result is compared with the 20 mcg/day group; the injection result is also compared with the 20 mcg/day group. These are not adjusted model estimates reported by the sponsor. The discrepancy flag is false, but the registry record does not report sponsor inferential analyses for these contrasts. Recalculation cannot reproduce undocumented choices about covariate adjustment, missing values, testing hierarchy or other protocol-specific methods.

Why can over-dispersion matter?

Over-dispersion is a general concern when data vary more than a statistical model expects. It can make standard errors too small if the model understates the actual variability, leading to intervals that are too narrow and p-values that are too optimistic. The primary outcome is continuous and summarized by means and standard deviations; the registry does not identify an over-dispersion model or report that over-dispersion was assessed. This concept should not be mistaken for a finding about this trial.

The primary analysis population is per protocol and requires baseline and Week 12 values, according to the registry definition. A complete-case or per-protocol restriction can affect who contributes to an estimate and may affect how well the result represents all randomized participants. The registry does not report the handling of missing outcome data, reasons for missingness, imputation methods or sensitivity analyses. It also does not report covariate adjustment, interim analyses, the original statistical plan or the assumptions used to select the trial’s enrollment. These omissions limit reconstruction of the sponsor’s intended inferential approach.

10. Confidence Intervals and Precision

The posted comparisons are estimates with substantial uncertainty about the exact mean difference. For ITCA 650 40 mcg/Day compared with ITCA 650 20 mcg/Day, the estimated mean difference is -0.08, with an interval from -0.412 to 0.252. For exenatide injection compared with ITCA 650 20 mcg/Day, the estimated mean difference is 0.14, with an interval from -0.219 to 0.499. Both intervals include zero and extend into values on either side of it.

In general, an interval communicates the range of values reasonably compatible with the estimate under the calculation’s assumptions. Wider intervals indicate less precision; narrower intervals indicate more precision. These intervals show that the point estimates alone should not be read as settled evidence of a difference in either direction. They also do not establish equivalence, because no equivalence margin is reported.

The group standard deviations and the number of observations contributing to each comparison affect the precision of a mean difference. The endpoint results are per protocol, with groups of 47.0, 45.0 and 47.0 for the three posted arms. These analysis counts differ from the registered enrollment of 155; the registry does not explain participant flow from enrollment to endpoint analysis. The confidence level for the posted intervals is not reported in the registry record, so it should not be assumed from the intervals alone.

Precision is distinct from clinical importance. The record does not state a threshold that would distinguish a clinically important difference from a small one. Nor do the intervals provide a guarantee about the result of a future study. They quantify uncertainty in these particular unadjusted recalculations, not certainty about outcomes beyond the defined Stage I endpoint.

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 sample-size figures are planning illustrations, not the original trial’s design justification. They estimate participants per arm for 80% power and participants per arm for 90% power, under a two-sided significance level of 0.05, equal allocation between the two arms and a normal approximation. The assumed effect is a stated fraction of the observed effect. The original trial’s own design assumptions are not reported in the registry record. These figures are not observed power.

For the ITCA 650 40 mcg/Day comparison with ITCA 650 20 mcg/Day, planning for 100% of the observed effect gives 1,637 participants per arm for 80% power and 2,191 for 90% power. Planning for 75% gives 2,910 and 3,895; planning for 50% gives 6,546 and 8,763. For the exenatide injection comparison with ITCA 650 20 mcg/Day, the corresponding figures are 632 and 845 for 100%, 1,122 and 1,502 for 75%, and 2,525 and 3,380 for 50%.

The 80% and 90% columns describe different planning targets: under the stated assumptions, the higher target requires more participants per arm. Likewise, planning for a smaller effect requires more participants, because a smaller difference is harder to distinguish from random variation. The curve should be read as a planning aid reflecting the assumed effect and design conditions, not as a prediction of what the completed study would show. The calculations are based on observed estimates, which themselves are uncertain, so the illustrations should not be treated as validated enrollment recommendations.

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, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -0.080; pooled standard deviation σ = 0.817.
2. n = 2 × (1.96 + 0.842)² × 0.817² / -0.080² = 1,636.4, rounded up to 1,637 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
delta <- -0.0800000000000001                      # observed difference in means
sigma <- 0.816791731920282                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^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 (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(943917)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 1,637 (80% power), 2,191 (90% power)
 75% of observed effect: n per arm = 2,910 (80% power), 3,895 (90% power)
 50% of observed effect: n per arm = 6,546 (80% power), 8,763 (90% power)
power.t.test: 1637.3 
Simulated power at n = 1,637 per arm: 0.798 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-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 posted primary outcome includes two recalculated comparisons, each using ITCA 650 20 mcg/Day as the reference: one contrasts ITCA 650 40 mcg/Day and the other contrasts exenatide injection. The associated p-values are unadjusted. The registry record does not report a multiplicity strategy, a hierarchy of hypotheses or an explanation of whether one contrast was intended to be primary and the other exploratory.

When several hypotheses are tested, the chance of at least one apparently unusual result can increase. Statistical procedures can account for multiple testing, but the record does not identify such a procedure for these results. The unadjusted p-values should therefore be understood as individual comparison statistics rather than as evidence adjusted across a family of tests. No conclusion about formal statistical significance under a prespecified family-wise plan can be reconstructed from the registry entries alone.

The primary population is per protocol and includes only subjects with values at baseline and Week 12. This definition indicates that the endpoint analysis is not described as including everyone enrolled regardless of endpoint availability. The registry does not report the number or reasons for missing outcome values beyond the group counts in the posted results, nor does it describe imputation, last-observation methods, sensitivity analyses or an intention-to-treat analysis. The effect of missingness on the estimates is therefore not assessable from the record.

The registry also does not report how protocol deviations were defined or how they affected inclusion in the per-protocol population. No statistical analysis plan, interim-analysis plan, covariate adjustment strategy or sponsor inferential analysis is listed. These omissions matter because decisions about population, missing data and multiple testing can affect estimates and uncertainty. They do not show that a particular method was or was not used; those details are simply not reported in the registry record.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: rectal cancer (7.7%), labyrinthitis (6.7%), myocardial infarction (6.2%), subclavian artery stenosis (6.2%), leukopenia (5.3%).

The registry lists serious and other adverse-event counts for multiple groups, including the three Stage I groups and additional dose, sequence and continuation-labelled groups. For the three initial groups, serious adverse events are reported as 1 among 51 at risk for ITCA 650 20 mcg/Day, 0 among 51 for ITCA 650 40 mcg/Day, and 0 among 53 for exenatide injection. Other adverse events are reported as 39 among 51, 43 among 51 and 38 among 53, respectively.

These are event summaries, not comparisons adjusted for differences in follow-up or participant characteristics. The record does not provide enough detail in these entries to infer the duration of risk or to determine whether a person could contribute to more than one event category. Death counts are not reported in the registry record. Serious adverse event terms listed with percentages include rectal cancer, labyrinthitis, myocardial infarction, subclavian artery stenosis, leukopenia and meningitis cryptococcal. The record does not establish from these entries that a particular event was caused by a study intervention. Safety counts should be read as descriptive registry reports, not as a causal comparison or a basis for treatment decisions.

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

The posted results describe average Week 12 HbA1c change in three Stage I per-protocol groups and give unadjusted estimates comparing two groups with the ITCA 650 20 mcg/Day group. The means are -0.96 for ITCA 650 20 mcg/Day, -1.04 for ITCA 650 40 mcg/Day and -0.82 for exenatide injection. The corresponding estimated differences are -0.08 for the higher-dose ITCA 650 comparison and 0.14 for the injection comparison.

Both intervals include zero and include values in either direction. The results therefore leave uncertainty about the direction and size of the differences. They do not show that the groups are equivalent, because the registry record does not state an equivalence margin. Nor do the posted means and comparisons, on their own, establish whether any difference is clinically important: no clinical-importance threshold is reported.

The results apply specifically to the endpoint and analysis population described in the registry: mean change over the first 12 weeks, among subjects with baseline and Week 12 values in the per-protocol analysis. They do not describe every intervention or continuation group listed in the record, and they do not provide a longer-term comparison. The findings should not be generalized beyond the stated outcome and population without additional evidence.

The registry record does not report the sponsor’s inferential analyses for these contrasts, a multiplicity plan, missing-data methods, a statistical analysis plan or the original sample-size justification. Consequently, the recalculated comparisons are useful for understanding the posted summaries but cannot substitute for a complete protocol-based analysis. These data are a statistical description of one completed trial’s registry results, not medical advice and not a conclusion that a treatment works or does not work.

16. Why This Trial Matters Statistically

This trial illustrates why a difference between sample means is not enough to settle a comparison. The mean change in the ITCA 650 40 mcg/Day group is -1.04, compared with -0.96 for ITCA 650 20 mcg/Day, yet the estimated difference is -0.08 and its interval extends from -0.412 to 0.252. The point estimate and its uncertainty tell different parts of the story: the former gives the central estimate, while the latter shows that the data do not precisely determine the direction or size of the contrast.

The exenatide injection comparison provides a similar lesson. Its estimated difference is 0.14, but the interval from -0.219 to 0.499 also includes zero. A reader who focuses only on the point estimate may overstate what the data resolve; a reader who looks only at a p-value may miss the range of effects compatible with the estimate. Interpreting the estimate and interval together gives a more informative account of precision.

The record also distinguishes the registered enrollment from the per-protocol endpoint analysis. The endpoint definition requires both baseline and Week 12 values, while the rationale for the analyzed group counts and any missing data procedures are not reported. This makes analysis-population details important to understanding what the results represent.

Finally, planning calculations illustrate how the assumed target effect influences a future study’s size. Smaller effects require more participants per arm under the listed assumptions. Such calculations are not observed power and do not recover the rationale for the original trial. Taken together, these features make the study useful for teaching the difference between descriptive outcomes, uncertain comparisons, analysis populations and planning illustrations.

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 endpoint results cover the first 12 weeks of Stage I and are limited to the registry-defined per-protocol population with baseline and Week 12 values. The record does not explain why the endpoint group counts differ from the overall enrollment, how many values were missing, or how protocol deviations were handled. This limits assessment of selection into the analysis.

The study is listed as open-label, with masking none. The registry does not describe procedures to reduce potential effects of unmasked treatment knowledge on measurement, participant behavior or study conduct. This is a design feature, not proof that bias occurred. Randomization is listed, but details of sequence generation and allocation concealment are not reported in the registry record.

The primary comparisons are unadjusted recalculations from summary statistics. The registry does not report sponsor inferential results, covariate-adjusted estimates, a prespecified hypothesis-testing plan or a multiplicity correction. The confidence level for the posted intervals is also not reported in the registry record. These omissions limit assessment of the intended inferential framework and how the comparisons were selected.

The record gives no threshold for clinical importance and no original sample-size justification. Although planning illustrations are posted, they use observed effects and stated calculation assumptions; they are not the trial’s documented design rationale. The registry lists additional arms and adverse-event groups, but detailed outcome results for those groups are not included in the primary endpoint summaries. These limitations mean the available results support a narrow interpretation of the posted Week 12 summaries, rather than a comprehensive account of every study outcome or longer-term effect.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Study of ITCA 650 (Exenatide in DUROS) in Subjects With Type 2 Diabetes Mellitus — NCT00943917.
• Registry record last updated 2015-04-13; status last verified 2015-03.
• Results first posted 2012-05-03.
• 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

The registry describes a completed phase 2 randomized, parallel, open-label study sponsored by Intarcia Therapeutics in subjects with type 2 diabetes. Registered enrollment is 155. The study start is recorded as 2009-08 and primary completion as 2010-08.

The primary posted outcome is mean change in HbA1c from Day 0 to Week 12 in a per-protocol population requiring baseline and Week 12 values. Results are reported for ITCA 650 20 mcg/Day, ITCA 650 40 mcg/Day and exenatide injection. The recalculated unadjusted comparisons have intervals that include zero, leaving uncertainty about the direction and size of group differences. The record does not report a multiplicity plan, missing-data methods or the original sample-size justification.