Omarigliptin (MK-3102) Clinical Trial - Add-on to Oral Antihyperglycemic Agent Study in Japanese Participants With Type 2 Diabetes Mellitus (MK-3102-015): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Omarigliptin (MK-3102) Clinical Trial - Add-on to Oral Antihyperglycemic Agent Study in Japanese Participants With Type 2 Diabetes Mellitus (MK-3102-015) (NCT01697592), a randomized phase 3 study. The primary outcome is Percentage of Participants Who Experienced at Least One Adverse Event Excluding Data After Glycemic Rescue During Phase A.

2. Clinical Question

3. Trial Design

This was a Phase III, randomized, parallel-group, placebo-controlled trial with a subsequent open-label extension. The registry record reports triple masking, while the official title describes the study as double-blinded. The record does not identify which specific parties were masked, so the masking description cannot be expanded beyond those labels.

The study enrolled 585 participants and was conducted in Japanese participants with type 2 diabetes mellitus and inadequate glycemic control on diet and exercise therapy plus oral antihyperglycemic agent monotherapy. Omarigliptin 25 mg or matching placebo was given orally once weekly. Participants continued their pre-study basal medication, which could be a sulfonylurea, glinide, biguanide, thiazolidinedione, or α-glucosidase inhibitor. The different background medicines define distinct treatment groups in the posted Phase A results.

Phase A lasted 24 weeks and was placebo-controlled. Phase B lasted 28 weeks, bringing the described treatment period to 52 weeks. Participants assigned to omarigliptin continued it in Phase B; those assigned to placebo switched to omarigliptin for the extension. Because the extension was open-label and did not continue the original placebo comparison, safety counts from Phase A+B should not be interpreted as a randomized placebo comparison over the full period.

The primary endpoint concerned adverse events up to 24 weeks, with data after glycemic rescue excluded. The registry lists change from baseline in hemoglobin A1c at Week 24 as a secondary outcome, but the registry entries described for this analysis do not include its results. The study start date was 2012-10-24, the primary completion date was 2014-05-08, and the status is completed.

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: In Phase A, participants were randomized to receive either omarigliptin (MK-3102) 25 mg once weekly or placebo for 24 weeks in a blinded manner. In Phase B, all participants received open-label omarigliptin 25 mg once weekly for 28 weeks.

Only the first 6 of 10 groups are shown.

Phase A (Up to 24 Weeks)

Reasons for not completing this period:

Phase B (Week 25 to Week 52)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All Subjects as Treated (ASaT) population, defined as all randomized participants who received at least one study drug. Participants were included in the treatment group corresponding to the study treatment they actually received for the analysis of safety data using the ASaT population.

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 11 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants Who Experienced at Least One Adverse Event Excluding Data After Glycemic Rescue During Phase A, measured over Up to 24 weeks.

The primary endpoint is the percentage of participants who experienced at least one adverse event during Phase A, up to 24 weeks, excluding data after glycemic rescue. An adverse event was defined in the registry as an unfavorable and unintended sign, including an abnormal laboratory finding, symptom, or disease associated with medical treatment or a procedure, regardless of whether it was considered related to treatment or the procedure.

This endpoint is a participant-level summary: a participant with one or more qualifying events is counted as having experienced an event. It does not convey how many events each participant had, how long events lasted, how severe they were, or whether investigators considered them treatment-related. Those are distinct aspects of safety assessment.

Excluding post-rescue data makes the endpoint specific to the period before or up to the defined rescue handling, rather than a complete account of later follow-up. The record describes glycemic rescue criteria based on fasting plasma glucose and rescue through changes to basal medication or use of metformin or glimepiride. The endpoint is useful for describing the proportion of participants with at least one adverse event during the controlled period; by itself it does not establish causation or summarize every safety outcome.

7. Results Posted in the Registry

• Omarigliptin 25 mg/SU (Phase A): 59.5% (75/126) versus 60.3% (38/63) on control; risk difference -0.8 percentage points (95% CI -15.6 to 14.0), risk ratio 0.99 (0.77–1.26); unadjusted two-proportion z-test p = 0.916.
• Omarigliptin 25 mg/Gln (Phase A): 53.8% (35/65) versus 60.3% (38/63) on control; risk difference -6.5 percentage points (95% CI -23.6 to 10.6), risk ratio 0.89 (0.66–1.21); unadjusted two-proportion z-test p = 0.460.
• Omarigliptin 25 mg/BG (Phase A): 54.5% (36/66) versus 60.3% (38/63) on control; risk difference -5.8 percentage points (95% CI -22.8 to 11.3), risk ratio 0.90 (0.67–1.22); unadjusted two-proportion z-test p = 0.508.
• Omarigliptin 25 mg/TZD (Phase A): 60.0% (39/65) versus 60.3% (38/63) on control; risk difference -0.3 percentage points (95% CI -17.3 to 16.6), risk ratio 0.99 (0.75–1.32); unadjusted two-proportion z-test p = 0.971.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary at the participant level: each participant either experienced at least one qualifying adverse event in the defined period or did not. The registry expresses the outcome as a percentage of participants. This is a direct way to describe how common the endpoint was in each group, but it compresses multiple aspects of event experience into a single yes-or-no indicator.

What does a risk difference describe?

General statistical explanation: a risk difference subtracts the event proportion in one group from the event proportion in another. A value below zero means the event proportion in the group named for the comparison is lower than the reference proportion; a value above zero means it is higher. The reported risk differences for the omarigliptin comparisons are -0.008, -0.065, -0.058, -0.003, and -0.096. These estimates are unadjusted comparisons, and their uncertainty should be considered rather than reading the point estimates alone.

The record also reports risk ratios and odds ratios with intervals. General statistical explanation: a risk ratio compares event probabilities as a ratio, while an odds ratio compares event odds. Both describe relative rather than absolute differences and may be less intuitive than a risk difference. They are alternative summaries of the same binary outcome, not separate outcomes. The displayed risk-difference results are sufficient to describe the absolute direction of the posted comparisons; the other measures should not be interpreted as additional independent evidence.

How should confidence intervals be read?

General statistical explanation: a confidence interval describes a range of estimates compatible with the data under the method used. It conveys precision as well as direction. An interval that includes zero for a risk difference includes the possibility of no difference under that analysis, but it may also include differences in either direction. The intervals for each of the five omarigliptin risk-difference comparisons span zero. Their breadth matters: it indicates uncertainty about the size and direction of the underlying difference, not proof that the groups have identical event frequencies.

The unadjusted p-values reported for those comparisons are 0.916, 0.46, 0.508, 0.971, and 0.273. General statistical explanation: a p-value assesses how compatible the observed data are with a specified null hypothesis under the model and assumptions used. It is not the probability that the null hypothesis is true, and it does not measure the clinical importance of an estimate. A p-value alone cannot distinguish a small difference from an imprecisely estimated larger difference. Reading it alongside the point estimate, confidence interval, and sample size is more informative.

Why does the comparison structure matter?

The primary endpoint is reported across five background-medication groups. These are separate treatment comparisons, with distinct group sizes and event percentages. The registry lists each omarigliptin background group and placebo background groups; the summaries should not be combined informally into an overall treatment effect. A pooled comparison would require a defined analysis and appropriate handling of background therapy, neither of which is reported in the registry record. The registry’s displayed unadjusted estimates should therefore be interpreted as the comparisons presented, not as a single estimate across all participants.

General statistical explanation: testing multiple comparisons increases the opportunity for a nominally unusual result to appear by chance across the set. A multiplicity procedure can control that issue, but the registry record does not report whether an adjustment was planned or used for the primary comparisons. Consequently, the listed p-values should be understood as unadjusted rather than as a family-wise conclusion.

What does the safety population mean?

The primary safety analysis used the All Subjects as Treated population. The registry defines it as all randomized participants who received at least one study drug, with each participant assigned for safety analysis according to the treatment actually received. This is a treatment-received approach to safety description. The registry does not report the full participant flow or the number excluded from the safety population, so the relation between total enrollment and each outcome denominator cannot be reconstructed from these registry entries alone.

The endpoint excludes data after glycemic rescue. This defines a restricted observation period for the primary endpoint. General statistical explanation: rescue treatment can alter subsequent exposure or outcome experience, so a rule for handling post-rescue data affects what the endpoint estimates. The registry specifies the exclusion and outlines rescue criteria, but the full missing-data approach and any sensitivity analyses are not reported in the registry record. The results therefore describe the endpoint under the stated rescue handling, not every possible approach to post-rescue observations.

The record identifies randomization, parallel groups, placebo control, and masking, but it does not report sequence generation, allocation concealment, the exact statistical model, variance method, or a detailed analysis plan. Those details should not be inferred from the trial’s design labels or from the posted outcome summaries. The trial’s posted comparison statistics are useful summaries, but their limitations include the absence of further details on adjustment, missing outcomes, and the original prespecified analysis.

10. Confidence Intervals and Precision

Precision is about how narrowly the data estimate the difference between groups. For the five posted omarigliptin comparisons, every risk-difference confidence interval includes zero and extends to values on either side of it. This means each interval is compatible with no difference as well as with differences in both directions, under the method used. The intervals are not a demonstration that event frequencies are equal.

The posted group sizes are modest and differ by background medicine: the omarigliptin groups contain 126, 65, 66, 65, and 67 participants, while the placebo groups contain 63, 34, 33, 34, and 32 participants. With this structure, the estimated difference for an individual comparison may be imprecise. The interval is the relevant description of that uncertainty; a point estimate alone can give a misleading impression of certainty or importance.

General statistical explanation: confidence intervals are affected by both the observed event information and the analysis method. More information can narrow an interval, while less information tends to leave a wider range of plausible values. A wide interval does not establish that a difference exists, and an interval spanning zero does not establish that a difference is absent. The registry record does not report a prespecified equivalence margin, so these results cannot be interpreted as an equivalence finding.

The numerical endpoints of the intervals are not needed to see their central feature: all span zero. The unadjusted p-values also do not replace the intervals. Together, the estimates and intervals indicate that the posted results do not precisely identify the direction or size of the event-rate differences in these comparisons.

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 illustrations express required sample size as participants per arm, not as total enrollment or total events. They use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect. These are planning assumptions for the illustrations; the trial’s original design assumptions are not reported in the registry record.

The 80% power column gives the planned number of participants per arm under those assumptions for an 80% chance of detecting the selected effect if that effect were the true one. The 90% power column uses a higher target chance of detection and therefore generally requires more participants per arm. For example, for the omarigliptin-versus-placebo comparison with sulfonylurea background therapy, the illustration at 100% of the observed effect gives 59,851 participants per arm for 80% power and 80,123 participants per arm for 90% power.

Rows for 75% and 50% of the observed effect ask what sample size would be needed to detect a smaller effect than the observed estimate. Smaller effects are harder to distinguish from random variation, so the illustrations require more participants per arm. For the same comparison, planning for 75% of the observed effect gives 106,315 participants per arm for 80% power and 142,324 participants per arm for 90% power; planning for 50% gives 239,009 and 319,965 participants per arm, respectively.

These calculations are planning illustrations based on observed estimates, not the original trial’s design justification. They are not observed power and do not indicate the probability that the trial’s result is correct. A curve, where shown, conveys how the planning requirement changes as the assumed effect changes; it should be read under the same assumptions as the corresponding illustrations.

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 = 0.603; treatment proportion p₁ = 0.595; average p̄ = 0.599.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6930 = 1.3584.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6930 = 0.5833.
4. n = (1.3584 + 0.5833)² / (-0.0079)² = 59,850.5, rounded up to 59,851 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 <- 0.603174603174603                          # control proportion
p1 <- 0.595238095238095                          # 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(1697592)
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 = 59,851 (80% power), 80,123 (90% power)
 75% of observed effect: n per arm = 106,315 (80% power), 142,324 (90% power)
 50% of observed effect: n per arm = 239,009 (80% power), 319,965 (90% power)
power.prop.test: 59850.5 
Simulated power at n = 59,851 per arm: 0.800 (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 registry record reports five background-medication-specific primary safety comparisons and unadjusted p-values for the posted comparisons. It does not report a multiplicity adjustment, a hierarchy among comparisons, or a prespecified rule for interpreting the family of tests. Therefore the listed p-values should be read as individual unadjusted results, not as a multiplicity-controlled overall conclusion. The record also does not report interim analyses or stopping rules.

The primary analysis population is defined as randomized participants who received at least one study drug, classified according to treatment actually received for safety analyses. This definition is reported, but the registry record does not provide a detailed accounting of missing outcome data, withdrawals, or participants excluded from a particular endpoint analysis. It also does not state how missing adverse-event information was handled. These omissions limit assessment of whether the observed percentages could be affected by incomplete follow-up.

Post-rescue data were excluded from the primary endpoint, and the record describes the glycemic rescue criteria and the rescue approaches. It does not report additional sensitivity analyses for alternative handling of post-rescue observations. Nor does it report the statistical model or a fuller analysis plan for the primary comparisons. The registry lists change from baseline in HbA1c at Week 24 as a secondary outcome, but its results are not reported in the registry record described for this analysis.

These reporting gaps do not invalidate the posted summaries, but they constrain what can be established from them. In particular, neither multiplicity control nor robustness to missingness and rescue handling can be assessed from the registry information available.

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: cataract (1.6%), macular hole (1.5%), appendicitis (1.5%), intervertebral disc protrusion (1.5%), breast cancer (1.5%).

The safety summaries report serious adverse events and other adverse events separately, with the number of participants at risk and the number affected for each listed group. For example, the Phase A omarigliptin-with-sulfonylurea group reports 3 participants with a serious adverse event among 126 at risk and 33 with another adverse event among 126 at risk. These counts describe the registry categories; they do not by themselves establish that treatment caused an event.

Phase A+B entries include the open-label extension and therefore cover a different exposure period from the primary endpoint. They should not be compared directly with Phase A placebo counts as though the groups remained randomized and placebo-controlled throughout. The registry also lists a pooled placebo/ABM group and Phase B treatment groups, so group labels and follow-up period matter when reading counts.

The deaths field is not reported in the registry record. Serious adverse-event terms are listed with percentages, including cataract, macular hole, appendicitis, intervertebral disc protrusion, breast cancer, and colon cancer. These terms describe events reported in the safety record; they do not provide a causal assessment or a comparative risk estimate.

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

The primary results describe the percentage of participants with at least one adverse event during the placebo-controlled Phase A, excluding data after glycemic rescue. The event percentages vary across the five background-medication groups and the corresponding placebo groups. The posted unadjusted risk-difference intervals all include zero, and the unadjusted p-values do not indicate a clear difference in those comparisons. This is a statement about the evidence and precision of the posted comparisons, not proof that the event rates are identical.

The results do not establish that omarigliptin did or did not cause any particular event. The endpoint includes adverse events regardless of whether they were considered related to treatment, and the summary does not distinguish related from unrelated events, severity, duration, or recurrence. The separate serious-event summaries can add descriptive safety information, but they cover varying treatment periods and do not replace the primary comparison.

The results also do not answer whether glycemic control improved. The registry lists a secondary HbA1c outcome, but its result is not reported in the registry record. The primary endpoint is a safety measure, and no efficacy conclusion should be drawn from its percentages.

The trial applies to the enrolled Japanese participants described in the record, receiving one of several existing oral medicines, over the specified controlled Phase A period for the primary endpoint. The findings should not be generalized automatically to other populations, treatment settings, background therapies, or longer randomized comparisons. The open-label extension has a different design from the placebo-controlled phase. Overall, the posted results are descriptive and comparative evidence with uncertainty; they do not support claims of equivalence, causation, or treatment benefit.

16. Why This Trial Matters Statistically

This trial illustrates why a safety endpoint should be interpreted in relation to its exact definition and follow-up period. The primary measure counts participants with at least one adverse event, excludes data after glycemic rescue, and is limited to Phase A. It is not interchangeable with broader adverse-event counts spanning the open-label extension.

The design also separates participants by background oral medicine. That structure matters statistically: the result is not a single pooled comparison, and each background-specific estimate has its own group sizes and uncertainty. The posted risk differences and intervals provide a way to examine those comparisons while avoiding the mistaken impression that one percentage represents all participants.

The trial is also a practical example of why a non-significant result should not be equated with proof of no difference. Each posted interval spans zero, while also allowing differences in either direction. The evidence is therefore compatible with a range of possibilities rather than demonstrating equal event frequencies. Looking at estimates, intervals, group sizes, and unadjusted p-values together gives a more complete reading than relying on a single test result.

Finally, the record shows the distinction between randomized controlled follow-up and an open-label extension. Phase A provides the placebo-controlled comparison for the primary endpoint; later Phase A+B summaries include treatment after the placebo groups switched to omarigliptin. Recognizing this change in design prevents unlike follow-up periods from being treated as a single randomized contrast.

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

The primary comparisons are unadjusted, and the registry record does not describe multiplicity control. With several background-specific comparisons, the individual p-values do not by themselves provide a family-wide statistical conclusion. The record also does not report the trial’s original statistical analysis plan, sample-size justification, or interim analysis procedures.

The confidence intervals span zero for all five posted omarigliptin risk differences, leaving uncertainty about both direction and size. The groups are not equally sized, and the placebo groups are smaller than their corresponding omarigliptin groups. The registry record does not provide an equivalence margin, so the comparisons cannot establish equivalence.

The endpoint captures whether a participant had at least one adverse event, not the number, severity, duration, or treatment attribution of events. Post-rescue data are excluded, but details about missing outcome data and sensitivity analyses for rescue handling are not reported in the registry record. Safety summaries include different phases and group definitions, limiting direct comparisons across them.

The record lists an HbA1c secondary outcome but does not report its result in the information described for this analysis. The primary completion date and total enrollment are recorded, but the registry information does not provide further participant-flow detail for the endpoint groups. The findings are specific to the reported Phase A comparisons and should not be treated as a general conclusion across different populations or treatment durations.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Omarigliptin (MK-3102) Clinical Trial - Add-on to Oral Antihyperglycemic Agent Study in Japanese Participants With Type 2 Diabetes Mellitus (MK-3102-015) — NCT01697592.
• Registry record last updated 2018-09-10; status last verified 2018-08.
• Results first posted 2015-10-29.
• 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 record describes a completed, randomized, Phase III, parallel-group, placebo-controlled study in Japanese participants with type 2 diabetes mellitus. The study enrolled 585 participants and compared once-weekly omarigliptin 25 mg with matching placebo, added to one of five pre-study oral antihyperglycemic medicines. Phase A was placebo-controlled for 24 weeks, followed by a 28-week open-label Phase B.

The primary endpoint was the percentage of participants with at least one adverse event up to 24 weeks, excluding data after glycemic rescue. Posted percentages differed across background-medication groups, while the five omarigliptin risk-difference confidence intervals spanned zero. Unadjusted p-values were reported. Results for the listed HbA1c secondary outcome are not reported in the registry record described for this analysis.