Comparison of Fasiglifam (TAK-875) to Placebo and Sitagliptin in Combination With Metformin in Participants With Type 2 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Comparison of Fasiglifam (TAK-875) to Placebo and Sitagliptin in Combination With Metformin in Participants With Type 2 Diabetes (NCT01549964), a randomized phase 3 study. The primary outcome is Change From Baseline in Glycosylated Hemoglobin (HbA1c).

2. Clinical Question

3. Trial Design

This was a randomized, parallel-group, phase 3 study with quadruple masking. Its official title describes a multicenter, double-blind, placebo- and active-controlled evaluation of fasiglifam 25 mg and 50 mg against placebo and sitagliptin 100 mg, all used with metformin. The registry lists enrollment as 916. The posted primary outcome analysis includes group-specific sample counts that differ from the registry enrollment figure; those counts should be read as the analysis counts for that outcome, not as a replacement for the enrollment entry.

The four treatment groups were placebo, sitagliptin 100 mg, fasiglifam 25 mg, and fasiglifam 50 mg. The arm descriptions specify oral once-daily study tablets during a 24-week treatment period and describe an optional extension for up to 104 weeks of treatment. The primary endpoint was assessed at Week 24 relative to baseline. The registry record lists a study start of 2012-04, primary completion of 2014-03, and status as terminated. It does not report the reason for termination.

The brief summary frames the objective as evaluating blood-sugar lowering for the two fasiglifam doses plus metformin versus placebo plus metformin and sitagliptin plus metformin. The registry record does not report the planned sample-size justification, the reason for choosing the posted enrollment, or the original trial design assumptions. It also does not report the reason for termination. These absences mean the observed enrollment and analysis counts cannot establish what the investigators intended to recruit or why the study stopped.

4. Participant Flow and Accounting

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

Before assignment: Participants with a diagnosis of Type 2 Diabetes Mellitus were randomly enrolled in 1 of 4 treatment groups in a 1:2:2:2 ratio, once a day placebo, 100 mg sitagliptin, 25 mg fasiglifam or 50 mg fasiglifam in combination with metformin.

24-Week Treatment Period

Reasons for not completing this period:

80-Week Extension Period

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Full Analysis Set included all randomized participants who received at least 1 dose of study drug analyzed according to the treatment group to which they were randomized. A participant was included in the analyses when there was both a Baseline and at least 1 Post-baseline value at Week 24.

5. Baseline Characteristics

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

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in Glycosylated Hemoglobin (HbA1c), measured over Baseline and Week 24.

The primary endpoint is change from baseline in glycosylated hemoglobin (HbA1c), expressed as a percent. The record describes HbA1c as the concentration of glucose bound to hemoglobin relative to the maximum that can be bound. The outcome compares the Week 24 value with baseline; a negative change represents a lower measured HbA1c at the later assessment than at baseline.

HbA1c is used as a summary measure of glucose exposure over time, which makes it relevant to a trial evaluating glycemic control. The registry also reports the incidence of HbA1c below 7% as a secondary outcome, giving a threshold-based view alongside the continuous primary measure. Those outcomes answer related but distinct questions: change in HbA1c preserves the magnitude of observed changes, while a threshold outcome counts how many participants meet a defined cutoff.

The primary result is reported as a least-squares mean with standard error for each group. A least-squares mean is a model-based group estimate, rather than simply the arithmetic average of the observed values. The registry describes a mixed model for repeated measurements that incorporates treatment, country, schedule, visit, and visit-by-treatment interaction as fixed factors, with baseline value and baseline-by-visit interaction as covariates. This endpoint and model concern the specified Week 24 assessment; the primary result alone does not describe longer-term outcomes during the optional extension.

7. Results Posted in the Registry

• Sitagliptin 100 mg: mean -1.07 versus -0.19 on control; difference -0.88 (95% CI -1.12 to -0.64), standardized effect d = -1.02; z-test p = <0.001.
• Fasiglifam 25 mg: mean -0.75 versus -0.19 on control; difference -0.56 (95% CI -0.80 to -0.32), standardized effect d = -0.66; z-test p = <0.001.
• Fasiglifam 50 mg: mean -1.01 versus -0.19 on control; difference -0.82 (95% CI -1.06 to -0.58), standardized effect d = -0.97; z-test p = <0.001.

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.

Incidence of HbA1c <7%

• 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:
  • Sitagliptin 100 mg: 42.8% (62/145) versus 14.9% (10/67) on control; risk difference 27.8 percentage points (95% CI 16.1 to 39.6), risk ratio 2.86 (1.57–5.23); unadjusted two-proportion z-test p = <0.001.
  • Fasiglifam 25 mg: 24.3% (35/144) versus 14.9% (10/67) on control; risk difference 9.4 percentage points (95% CI -1.7 to 20.4), risk ratio 1.63 (0.86–3.09); unadjusted two-proportion z-test p = 0.122.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Sitagliptin 100 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 Fasiglifam 25 mg and the control. The interval gives the range of effects that remain compatible with the data.

Change From Baseline in Fasting Plasma Glucose (FPG)

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Sitagliptin 100 mg: mean -21.70 versus -1.60 on control; difference -20.10 (95% CI -30.16 to -10.04), standardized effect d = -0.55; z-test p = <0.001.
  • Fasiglifam 25 mg: mean -26.90 versus -1.60 on control; difference -25.30 (95% CI -35.36 to -15.24), standardized effect d = -0.70; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Sitagliptin 100 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Fasiglifam 25 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry lists two secondary outcomes: incidence of HbA1c below 7% at 24 weeks and change from baseline in fasting plasma glucose. Together, they complement the primary continuous HbA1c endpoint by adding a threshold-based measure and another glycemic measure. The secondary estimates therefore offer related evidence about the trial’s glycemic outcomes, but they are not interchangeable with the primary endpoint.

The reported secondary analyses do not provide a complete set of sponsor comparisons for every group and outcome. In particular, the registry’s posted analyses are limited to the comparisons it lists; an unlisted comparison should not be inferred from the group summaries. The simple recalculations can describe additional contrasts from posted summaries, but the discrepancy flag means the sponsor’s analysis takes precedence where the results differ.

Unadjusted secondary results carry less weight than the prespecified primary analysis for several reasons. Secondary outcomes can be exploratory, and examining multiple outcomes and contrasts increases the chance that some results will appear notable by chance. The record does not report a multiplicity adjustment strategy, so the secondary p-values should not be treated as if they account for all analyses. The registry record also does not report a hierarchy or gatekeeping procedure for interpreting secondary findings.

The overall secondary pattern can be considered in relation to the primary HbA1c findings, but agreement across related outcomes does not by itself remove uncertainty from multiplicity, missing data, or analysis choices. Nor does a difference on a secondary measure establish a clinically important effect. These findings are best interpreted as supporting context for the primary result, with attention to the posted estimates, the sponsor’s analysis, and the limits on formal inference.

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

• Incidence of HbA1c <7%
• Change From Baseline in Fasting Plasma Glucose (FPG)

9. Statistical Interpretation

10. Statistical Methods Explained

The primary analysis uses a mixed model for repeated measurements (MMRM). In general, this model is designed for outcomes measured at multiple visits and estimates treatment-group patterns while accounting for the fact that observations from the same participant are related. The registry says the model included treatment, country, schedule, visit, and visit-by-treatment interaction as fixed factors. It also included baseline HbA1c and a baseline-by-visit interaction as covariates, with an unstructured covariance structure. The record does not give further detail about model diagnostics, the choice of covariance structure, or the handling of particular departures from model assumptions.

What is a least-squares mean?

A least-squares mean is a model-based estimate adjusted for the other terms included in the analysis. It is not necessarily the same as the raw mean calculated directly from all observed values. The registry reports the primary group results as least-squares means and the between-group results as least-squares mean differences. This distinction matters because the sponsor’s reported results come from the stated repeated-measures model, whereas the recalculations use only posted summary statistics and are simple, unadjusted comparisons.

What does the standard error describe?

The primary group estimates are reported with standard errors. In general, a standard error describes the statistical uncertainty around an estimated mean, reflecting how precisely the mean is estimated under the analysis. It is not the spread of individual participant outcomes. The registry lists standard errors of 0.098 for placebo, 0.074 for sitagliptin 100 mg, 0.073 for fasiglifam 25 mg, and 0.074 for fasiglifam 50 mg. The sponsor’s comparison results give confidence intervals for the contrasts, which provide a direct range of values compatible with the model and data under the specified analysis.

How should the recalculations be understood?

The recalculated comparisons use the posted group estimates and their standard errors to form approximate, unadjusted contrasts. The record identifies the standard deviations used in these calculations as approximate. They are not a reproduction of the MMRM, do not apply its adjustment terms, and should not be treated as an alternative sponsor analysis. The presence of a discrepancy is a reason to give the sponsor’s reported analysis priority when describing the trial’s inferential results.

For a continuous outcome, a mean difference is the contrast between group mean estimates. A negative contrast in these results indicates a lower change estimate for the named treatment group than for the comparator group, using the ordering in the reported analysis. The contrast is distinct from the outcome’s absolute change from baseline: it expresses the difference between groups, not the change experienced by an individual participant.

What does a p-value say?

In general, a p-value measures how compatible the observed result is with a specified null hypothesis under the model and its assumptions. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of an effect. The sponsor reports p < 0.001 for three primary contrasts and p = 0.524 for sitagliptin 100 mg versus fasiglifam 50 mg. Interpretation should consider the estimates and confidence intervals as well as the p-values, and should account for the trial’s multiple comparisons.

The primary population is defined in the registry as the Full Analysis Set: randomized participants who received at least 1 dose, analyzed according to the group to which they were randomized. For inclusion in the analyses, a participant needed both a baseline value and at least 1 post-baseline value at Week 24. This is the analysis population as described; the record does not report a separate detailed account of missing-data imputation or sensitivity analyses.

11. Confidence Intervals and Precision

The sponsor’s primary comparison results include 95% confidence intervals. A confidence interval describes the uncertainty around an estimated contrast under the analysis and its assumptions. In repeated use of the same method, intervals constructed in this way would capture the target quantity at the stated rate over the long run; an individual interval is not a probability statement that the true effect lies within its particular endpoints.

For fasiglifam 25 mg versus placebo, the reported least-squares mean difference is -0.55 and the 95% confidence interval ranges from -0.77 to -0.34. For fasiglifam 50 mg versus placebo, the reported difference is -0.82 and the interval ranges from -1.03 to -0.60. These intervals are wholly on the negative side of zero. The sponsor reports p < 0.001 for both comparisons. For sitagliptin 100 mg versus fasiglifam 50 mg, the reported difference is 0.06 and the interval ranges from -0.12 to 0.23; this interval includes zero, and the reported p-value is 0.524.

Interval width is one way to judge precision: narrower intervals indicate a more precise estimate, while wider intervals leave more uncertainty about the size of the contrast. Precision is not the same as clinical importance. The record does not provide a threshold for a clinically important HbA1c difference, so the statistical intervals alone cannot establish whether a difference would be meaningful to patients.

The simple unadjusted recalculations have their own intervals, but they are not the sponsor’s MMRM intervals. Because the record flags a discrepancy, the sponsor’s own analysis takes precedence. The recalculated intervals are best treated as approximate context rather than as a substitute for the reported model-based results.

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 table and curve are planning illustrations based on the observed effect, not the trial’s original design calculation. For these outcomes, the displayed figures are participants per arm, not total events. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; each scenario uses the stated fraction of the observed effect as the effect to detect. The registry record does not report the trial’s own original sample-size assumptions or justification.

The 80% power column gives the illustrative number of participants per arm for a design that would have an 80% chance of detecting the assumed effect if that effect and the planning assumptions held. The 90% power column gives the corresponding per-arm planning figure for a 90% chance. These are design quantities, not estimates of what happened in this trial. This is not observed power.

Rows planning for 100% of the observed effect use the full observed contrast as the assumed target; rows planning for 75% or 50% use smaller target effects. A smaller effect is harder to distinguish from random variation, so it generally requires more participants per arm under the same planning assumptions. The illustrated counts can therefore change substantially across effect-size scenarios. They are useful for understanding how assumptions affect planning, but should not be read as a recommendation or as evidence that the original trial had those planned group sizes.

For the continuous primary endpoint, the table includes planning illustrations for each comparison against placebo. Other posted continuous outcomes have their own illustrations where recalculations are available. For the binary secondary outcome, the same per-arm convention applies. The figures remain approximate because they depend on the stated model assumptions and the observed effect used as a planning anchor.

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

Worked calculation for this trial

1. Observed difference Δ = -0.880; pooled standard deviation σ = 0.864.
2. n = 2 × (1.96 + 0.842)² × 0.864² / -0.880² = 15.1, rounded up to 16 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 <- -0.88                      # observed difference in means
sigma <- 0.864120163601601                     # 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(1549964)
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 = 16 (80% power), 21 (90% power)
 75% of observed effect: n per arm = 27 (80% power), 37 (90% power)
 50% of observed effect: n per arm = 61 (80% power), 82 (90% power)
power.t.test: 16.2 
Simulated power at n = 16 per arm: 0.794 (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.

14. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary comparisons from a mixed model for repeated measurements, including least-squares mean differences, 95% confidence intervals, and p-values. These model-based estimates are the trial’s reported inferential results. The record also flags a discrepancy between the sponsor’s results and simple recalculations from the posted group summaries.

The recalculations are unadjusted contrasts formed from the posted group estimates and standard errors, with approximate standard deviations. They do not reproduce the sponsor’s model, which incorporates treatment, country, schedule, visit, visit-by-treatment interaction, baseline value, and baseline-by-visit interaction. Differences between a simple summary-based calculation and a model-based analysis can arise because the analyses use different information and adjustments.

Accordingly, the sponsor’s own analysis takes precedence. The recalculated values are useful only as a rough cross-check of direction and scale; they should not be used to replace the sponsor’s reported estimate, interval, or p-value. This distinction is especially important when communicating a result as a formal trial finding, since the sponsor’s analysis is the one specified in the registry for the primary endpoint.

15. Multiplicity, Missing Data and Analysis Populations

The record lists multiple treatment comparisons for the primary outcome and more than one secondary outcome. Testing multiple contrasts increases the opportunity to observe a small p-value by chance even when there is no underlying difference. The registry record does not report whether the primary comparisons were adjusted for multiplicity, whether comparisons followed a prespecified hierarchy, or whether a gatekeeping strategy was used. It also does not report a multiplicity plan for the secondary outcomes. Therefore, the posted p-values should be read as reported, without assuming they control the overall error rate across all comparisons.

The primary analysis population includes randomized participants who received at least 1 dose, analyzed according to their randomized group, and who had both a baseline and at least 1 post-baseline value at Week 24. Requiring an outcome measurement for inclusion means that participants without the required data do not contribute to that analysis as described. The record does not report the number or reasons for missing primary endpoint measurements, the missing-data assumptions, or whether sensitivity analyses were conducted. It is therefore not possible to assess from the registry record how robust the primary result is to alternative missing-data assumptions.

The registry describes an MMRM and an unstructured covariance structure, but does not give details about diagnostics or the treatment of incomplete measurement patterns beyond the stated analysis population. General statistical practice recognizes that repeated-measures models rely on assumptions about the outcome process and missingness; the registry does not report enough information to evaluate those assumptions for this trial.

The record also does not report interim analyses, stopping rules, or the reason for the terminated status. Their absence from the registry record should not be filled in by speculation. These gaps limit interpretation of the analysis plan and the path from the trial’s planned conduct to its posted results.

16. 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 2% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: sepsis (0.8%), phimosis (0.8%), oesophageal carcinoma (0.8%), atrioventricular block first degree (0.4%), cholecystitis (0.4%).

The adverse-event record reports serious adverse events by group using both a count of participants at risk and a count affected. The figures are: placebo, 2 affected among 132 at risk; sitagliptin 100 mg, 10 among 260; fasiglifam 25 mg, 14 among 263; and fasiglifam 50 mg, 4 among 260. These are descriptive registry counts. They do not, on their own, establish that differences between groups were caused by treatment or are statistically reliable.

The record also lists serious event terms with percentages, including sepsis, phimosis, oesophageal carcinoma, atrioventricular block first degree, cholecystitis, and liver disorder. It does not give arm-specific attribution for these listed terms in the registry record for this summary, so they should not be assigned to a particular treatment group. Death counts are not reported in the registry record.

The posted serious-event figures are not a complete assessment of safety. The registry also reports other adverse-event counts by group, but those counts do not establish event severity, duration, causality, or comparative risk. The results should be read as descriptive reporting over the recorded study period, not as a standalone basis for judging treatment safety.

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

The posted results describe changes in HbA1c and other glycemic measures among the trial’s analysis groups over the assessment period. The sponsor’s primary analysis reports lower HbA1c change estimates for fasiglifam groups than placebo, with the reported intervals and p-values detailed for those comparisons. The comparison between sitagliptin 100 mg and fasiglifam 50 mg has a confidence interval that includes zero and a reported p-value of 0.524. These are statistical findings from the registry record, not a complete account of clinical outcomes.

A statistically significant comparison does not by itself show that an effect is important to patients. The registry record does not supply a threshold for a clinically meaningful HbA1c difference, and statistical significance does not address all possible benefits, harms, or treatment choices. Conversely, a non-significant result does not demonstrate that two treatments are equivalent; it indicates that the analysis did not establish a difference at the reported level of precision.

The trial’s primary result concerns the change from baseline to Week 24. Although the arm descriptions mention an optional extension, the primary endpoint results do not establish what happened over that longer period. Nor do the posted results alone support conclusions about outcomes not reported in the record. The study status is listed as terminated, but the reason is not reported in the registry record.

The secondary outcomes offer additional information about glycemic measures but are subject to limitations in multiplicity and reporting. The adverse-event counts provide descriptive safety information, not proof of causation or a full comparative risk assessment. Overall, the record supports a statistical reading of the posted estimates and their uncertainty; it does not establish a treatment recommendation or resolve questions that the reported endpoints and analyses did not address.

18. Why This Trial Matters Statistically

This trial illustrates why a model-based result and a simple calculation from group summaries should not automatically be treated as equivalent. The sponsor reports an MMRM analysis that accounts for specified study factors, visit structure, baseline values, and repeated measurements. The separate recalculations use posted summary statistics and are explicitly approximate and unadjusted. When these approaches show a discrepancy, the sponsor’s own analysis takes precedence for describing the trial’s inferential result.

It also demonstrates the value of reading estimates alongside their uncertainty. The primary comparisons include point estimates, confidence intervals, and p-values, while the group-level results include standard errors. Considering these together helps distinguish the direction of an estimated contrast from its precision. A p-value alone cannot communicate the size of an effect or whether it is clinically important.

The trial also makes clear that a primary endpoint and related secondary measures answer different questions. Continuous HbA1c change, a threshold-based HbA1c outcome, and fasting plasma glucose provide distinct views of glycemic outcomes. Multiple outcomes and comparisons complicate inference when the registry does not report a multiplicity strategy. Finally, the analysis population and the missing-data reporting matter: the stated inclusion criteria describe who contributed to the primary analysis, but the registry does not report enough detail to assess the impact of missing measurements. These are central considerations when interpreting clinical-trial statistics, not merely technical details.

19. 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.

Several trial-specific gaps constrain interpretation. Although the record lists a randomized, quadruple-masked, parallel-group phase 3 design, it does not report the original sample-size justification, the design assumptions used by the sponsor, or the reason for the terminated status. The planning illustrations derived from observed effects cannot fill those gaps or reconstruct the trial’s original design.

The primary analysis population required a baseline and at least 1 post-baseline value at Week 24. The registry does not report how many participants were excluded from that endpoint analysis because of missing measurements, why data were missing, or the results of sensitivity analyses. This leaves uncertainty about the influence of missing outcome data. The record describes the MMRM structure but does not report diagnostics or further details that would allow independent evaluation of model assumptions.

The record flags a discrepancy between simple recalculations and sponsor results. The recalculations are approximate and unadjusted, so they are not a substitute for the sponsor’s model; the sponsor analysis should take precedence. The registry also does not report whether multiplicity was controlled across the primary contrasts and secondary outcomes. Secondary p-values therefore require particular caution.

Finally, the primary endpoint is assessed at Week 24, and the posted analyses do not establish longer-term outcomes despite the arm descriptions’ optional extension. The serious adverse-event summaries are descriptive, with no deaths reported in the registry record and no arm attribution for the listed serious-event terms in this summary. These limits preclude a complete assessment of longer-term effectiveness, safety, and clinical importance.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Comparison of Fasiglifam (TAK-875) to Placebo and Sitagliptin in Combination With Metformin in Participants With Type 2 Diabetes — NCT01549964.
• Registry record last updated 2016-06-02; status last verified 2016-04.
• Results first posted 2016-06-02.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

The registry describes NCT01549964 as a randomized, quadruple-masked, parallel-group phase 3 study of fasiglifam 25 mg and 50 mg, sitagliptin 100 mg, and placebo, each in combination with metformin. Enrollment is listed as 916, and the study status is terminated. The primary endpoint is change in HbA1c from baseline to Week 24, analyzed with a mixed model for repeated measurements in the stated Full Analysis Set.

The sponsor reports model-based group contrasts with confidence intervals and p-values. The record also includes secondary glycemic outcomes and descriptive adverse-event counts. It does not report the original sample-size justification, multiplicity strategy, reason for termination, or detailed missing-data sensitivity analyses. The registry flags a discrepancy with simple unadjusted recalculations, so the sponsor’s own analysis takes precedence.