Ertugliflozin and Sitagliptin Co-administration Factorial Study (VERTIS FACTORAL, MK-8835-005): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Ertugliflozin and Sitagliptin Co-administration Factorial Study (VERTIS FACTORAL, MK-8835-005) (NCT02099110), a randomized phase 3 study. The primary outcome is Change From Baseline in A1C at Week 26: Excluding Rescue Approach.

2. Clinical Question

The registry describes a Phase III, randomized, double-blind, multicenter, parallel-group study. Metformin was the background treatment: the clinical question concerns adding the study drugs to metformin, rather than replacing it. The primary hypothesis specifically focused on the 15 mg ertugliflozin and 100 mg sitagliptin combination versus sitagliptin alone. The posted primary comparisons also include the 5 mg combination versus ertugliflozin 5 mg alone and comparisons of the 15 mg combination with ertugliflozin 15 mg and sitagliptin 100 mg.

The outcome is a measure of change in a blood marker of average glucose over time, not a direct measure of symptoms or longer-term clinical events. The registry record does not report a separate primary question about those other outcomes. The results therefore address the specified A1C comparisons at Week 26 under the stated analysis approach; they do not by themselves answer broader questions about effects beyond the measured outcomes and follow-up.

3. Trial Design

NCT02099110, titled Ertugliflozin and Sitagliptin Co-administration Factorial Study (VERTIS FACTORAL, MK-8835-005), is recorded as a completed Phase III trial sponsored by Merck Sharp & Dohme LLC. The registry lists randomized allocation, double masking, and a parallel model. Enrollment is recorded as 1233. The study dates recorded are 2014-04-22 for the start and 2016-05-26 for primary completion.

The treatment regimens included ertugliflozin 5 mg, ertugliflozin 15 mg, sitagliptin 100 mg, and combinations of ertugliflozin 5 mg or 15 mg with sitagliptin 100 mg. The listed study drugs were taken orally once daily for 52 weeks. Participants were receiving metformin at a dose described in the intervention record as at least 1500 mg/day. Insulin glargine is listed as rescue medication. The registry describes placebo-matching interventions, consistent with the recorded double masking, but the detailed masking procedures are not reported in the registry record.

The primary endpoint was change from baseline in A1C at Week 26, using an analysis that excluded data after rescue therapy began. The posted arm summaries include least-squares mean changes and 95% confidence intervals. The trial also listed secondary outcomes covering body weight, fasting plasma glucose, the proportion reaching an A1C threshold, a beta-cell sensitivity index, and systolic blood pressure. The registry record lists 8 outcome measures and 28 analyses. The detailed randomization procedure, allocation concealment, site count, and rationale for the enrollment target are not reported in the registry record.

4. Participant Flow and Accounting

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

Before assignment: Participants on ≥1500 mg/day of metformin for ≥8 weeks with A1C ≥7.5 and ≤11% at screening could directly enter a 2-week, single-blind, placebo run-in period. Participants who did not meet these criteria, received diet/exercise counseling and metformin titration (as necessary) for \~8 weeks before entering the 2-week, placebo run-in period.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The analysis population consisted of all randomized participants who received at least one dose of study treatment, had a baseline measurement or a post-randomization measurement for the A1C change from baseline at Week 26 analysis endpoint subsequent to at least one dose of study treatment.

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: The Baseline Analysis Population was the All-Participants-As-Treated population. One participant randomized to the Ertugliflozin 15 mg + Sitagliptin 100 mg arm did not receive study medication.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in A1C at Week 26: Excluding Rescue Approach, measured over Baseline and Week 26.

The primary endpoint is change from baseline in A1C at Week 26. A1C is a blood marker used to report average blood glucose levels over prolonged periods and is expressed as a percentage. The change is calculated as Week 26 A1C minus Week 0 A1C. A negative change therefore represents a lower measured A1C at Week 26 than at baseline; a more negative mean indicates a larger average decrease in the analyzed group.

A1C is used in this study as a laboratory measure of glycemic control over time. It offers a common continuous outcome for comparing the groups, rather than reducing every participant's result to whether or not they crossed a target. The secondary outcomes include a separate measure based on achieving an A1C below 7% (53 mmol/Mol); that threshold outcome and the primary continuous change outcome answer related but different questions.

The primary analysis excluded measurements collected after initiation of rescue therapy at any time point. The registry describes this as an approach intended to avoid the confounding influence of rescue therapy. The posted analysis population was limited to randomized participants who received at least one dose and had a baseline or relevant post-randomization A1C measurement. Thus, the reported mean change describes this analysis population under the exclusion-after-rescue approach, not every randomized participant regardless of treatment or measurement availability.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous: it is a numerical change in A1C rather than a category. The registry reports a least-squares mean for each arm and a difference in least-squares means for the specified comparisons. Least-squares means are model-based group estimates; they are not necessarily the same as a simple arithmetic average of the observed values. The registry identifies constrained longitudinal data analysis as the method for the primary between-group comparisons. The full model specification, covariates, covariance structure, and assumptions used in that model are not reported in the registry record.

What does a confidence interval convey?

A confidence interval describes uncertainty around an estimated effect under the statistical model and its assumptions. A narrower interval indicates greater precision than a wider interval, while an interval spanning a null comparison value would leave more uncertainty about the direction of the contrast. For the primary difference estimates, the posted 95% confidence intervals lie below zero. Because the outcome is change in A1C and the contrasts are expressed as one group minus another, negative estimates indicate a larger decrease in the first-listed combination group relative to its comparator. The interval does not describe the range of individual participant responses, nor does it establish that the effect is clinically important.

Why distinguish group summaries from contrasts?

Each arm's least-squares mean change answers how the modeled outcome changed from baseline within that arm. A between-group contrast answers how the modeled changes differ between two arms. The contrast is the relevant statistic for a claim about relative change. It should not be recreated by subtracting rounded arm means: the reported analysis uses constrained longitudinal data analysis, and its estimates can depend on the underlying model and the data used. The registry's sponsor-reported contrasts are therefore the appropriate reported comparisons for the primary endpoint.

What does excluding data after rescue mean?

The record says that the analysis excluded all data following initiation of rescue therapy at any time point, to avoid the confounding influence of rescue therapy. This approach defines which measurements contribute to the endpoint analysis; it is not the same as estimating what would have happened had rescue therapy never been available. The details of how missing measurements before rescue were handled, and the assumptions used for any missing-data procedure, are not reported in the registry record. The analysis population also required receipt of at least one dose and a baseline or relevant post-randomization measurement.

How should the secondary binary outcome be read?

The secondary A1C threshold outcome is a proportion, not a continuous mean change. The registry reports logistic-regression odds ratios for selected comparisons. An odds ratio compares odds, which are not the same quantity as a risk difference or a ratio of probabilities. The posted secondary results also include recalculated, unadjusted comparisons based on the arm summaries. Those calculations are not the sponsor's logistic-regression results and should be interpreted as a separate, simpler analysis. In general, a confidence interval around a binary comparison indicates uncertainty in the estimated association; it does not account for testing many outcomes unless a multiplicity procedure is applied.

What do the planning illustrations mean?

The sample-size illustrations use an effect based on the observed secondary binary outcome and show planning for 100%, 75%, and 50% of that observed effect. They assume a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. They are prospective planning illustrations, not observed power for this trial and not evidence that a particular effect is certain. Smaller effects generally require more participants for the same intended power, all else held constant. These calculations do not replace the trial's own design justification; that justification is not reported in the registry record.

10. Analyses Reported by the Sponsor

The primary endpoint's arm estimates and pairwise contrasts are reported from the sponsor's constrained longitudinal data analysis. Those contrasts are not labeled as recalculations. The reported analyses compare the combinations with ertugliflozin alone or sitagliptin alone, and include the registry's stated primary hypothesis comparison of ertugliflozin 15 mg plus sitagliptin 100 mg versus sitagliptin 100 mg alone. The registry reports the associated confidence intervals and p-values for these analyses.

For the secondary A1C threshold outcome, the registry also reports sponsor analyses using logistic regression and odds ratios. Separate unadjusted recalculations based on posted summary counts are a different analysis: they use a simpler comparison and do not reproduce the sponsor's regression model. Differences between those estimates can arise because the estimands and methods differ, so they should not be treated as competing versions of the same adjusted result. The registry does not identify a discrepancy for the posted results. The sponsor's reported analysis is the main interpretation of the registered analysis; the simpler recalculations serve as an additional descriptive check, not as a replacement.

The registry record does not provide enough detail to reproduce the primary model independently, including its complete specification and missing-data handling. Consequently, the primary results should be read as reported rather than reconstructed from rounded arm summaries.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary hypothesis and multiple primary comparisons, as well as secondary outcomes and analyses. It does not report a multiplicity adjustment, a testing hierarchy, or a distinction between confirmatory and exploratory secondary analyses. Without that information, the nominal p-values should be understood as results for the reported comparisons, not as a demonstrated family-wise error-controlled set of tests. This matters because repeated testing can make apparently strong findings more likely even when no true effect exists for some comparisons. The registry record also does not report whether interim analyses were planned or conducted.

Missing-data methods are not fully described. The primary analysis excluded measurements after rescue therapy began, but that statement does not specify how other missing Week 26 measurements were treated, the assumptions made about missingness, or whether sensitivity analyses were conducted. Those details are not reported in the registry record. The exclusion-after-rescue approach also means that the estimate is conditional on the analysis strategy; it does not directly describe outcomes incorporating measurements after rescue therapy.

The primary analysis population was described as randomized participants who received at least one dose of study treatment and had a baseline measurement or a post-randomization measurement for the A1C change analysis subsequent to at least one dose. This definition is narrower than a description of every randomized participant, because it includes treatment exposure and measurement requirements. The record does not provide a separate account of exclusions from this analysis population or the reasons for missing measurements. Enrollment is recorded as 1233, while the arm-level primary endpoint summaries report their own analysis counts. These figures describe different aspects of the record and should not be combined into a new total or treated as interchangeable denominators.

Randomized allocation and double masking are recorded, but the procedures used to preserve allocation concealment and masking are not reported in the registry record. A sample-size rationale and any planned assumptions about attrition are also not reported. These omissions limit assessment of the full prespecified statistical plan and the extent to which the posted analyses follow it.

12. 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: acute myocardial infarction (0.8%), vomiting (0.8%), blood potassium decreased (0.4%), blood sodium decreased (0.4%), acute coronary syndrome (0.4%).

The safety record reports serious adverse events and other adverse events separately for each treatment group. The serious adverse event counts range from 5 to 12 across the listed groups, with the group-specific denominators shown in the registry record. The other adverse event counts range from 28 to 43 across those groups, also with group-specific denominators. These are counts, not comparative risk estimates. The registry does not provide a statistical comparison of adverse-event rates in the registry record, so differences in counts alone should not be interpreted as evidence of a treatment-related difference.

Listed serious adverse event terms include acute myocardial infarction and vomiting at 0.8% each, and blood potassium decreased, blood sodium decreased, acute coronary syndrome, and myocardial infarction at 0.4% each. The registry record does not assign these term-level percentages to a specific arm in the information available, so they should not be read as arm-specific rates. Death counts are not reported in the registry record.

Safety counts require context about exposure, observation time, event definitions, and how events were collected. The record describes treatment for 52 weeks, but the figures alone do not establish causality or provide a complete comparative safety assessment. They are descriptive postings from this trial, not a basis for individual medical decisions.

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

The primary results show that the combination groups had more negative modeled mean A1C changes at Week 26 than the listed single-drug comparator groups. The sponsor-reported differences in least-squares means are also negative, and their 95% confidence intervals are below zero for the comparisons posted. This describes a difference in a measured laboratory outcome under the registry's analysis approach, which excludes data after rescue therapy began.

The result does not, by itself, establish the size of benefit an individual participant would experience. Group averages do not describe the full distribution of responses, and confidence intervals around group contrasts do not predict an individual's outcome. Statistical evidence against a no-difference comparison is also distinct from whether a change is clinically meaningful; the registry record does not state a threshold for clinical importance.

The finding is limited to the population and endpoint described: participants with type 2 diabetes and inadequate glycemic control on metformin, the specified study regimens, and the Week 26 A1C outcome. It does not establish effects on every secondary outcome, long-term outcomes beyond the reported follow-up, or results for populations not described in the trial. Rescue therapy and the exclusion of subsequent measurements are part of the endpoint definition and matter when interpreting what the estimate represents.

The posted results do not answer whether the same pattern would occur under other treatment regimens, in other populations, or with different approaches to rescue treatment or missing data. Nor do the A1C findings alone establish a broader clinical effect. The registry supports an interpretation about the measured A1C comparisons; claims beyond those outcomes and analyses are not established by the reported results.

14. Why This Trial Matters Statistically

This study illustrates why an outcome definition and a comparison method are central to interpreting a randomized trial. The endpoint is not simply an A1C value at Week 26: it is change from baseline, analyzed after excluding measurements following rescue therapy. A reader needs to distinguish that estimand from a treatment-policy question that would include later observations regardless of rescue. The same trial can therefore yield different answers to differently framed questions without either calculation being a transcription error.

It also shows why within-group change and between-group effect are not the same statistic. Each arm's mean change describes its modeled trajectory from baseline, whereas a contrast estimates the difference between two groups. The reported constrained longitudinal analysis is the source of the primary contrasts; rounded group means cannot substitute for the complete model-based comparison. Confidence intervals add information about precision that a point estimate alone cannot convey.

The secondary threshold outcome provides a further statistical distinction. Converting a continuous A1C measure into a threshold creates an interpretable proportion, but discards information about how far values lie above or below the cutoff. Logistic-regression odds ratios and unadjusted risk-difference calculations also represent different measures and methods. They should be read on their own terms rather than treated as directly interchangeable.

Finally, multiple outcomes and comparisons raise questions about multiplicity, while rescue-related exclusions raise questions about missingness and the target of inference. The registry record does not report key details of those plans. Recognizing those gaps helps separate what the posted analyses show from conclusions that would require additional design or analysis information.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The primary analysis excludes data after rescue therapy began, so its result pertains to that analysis approach rather than to an estimate incorporating all post-rescue measurements. The registry explains the purpose of the exclusion but does not report how missing values before rescue were handled or whether sensitivity analyses assessed alternative assumptions. This leaves uncertainty about how robust the primary estimate is to missing-data choices.

The full specification for the constrained longitudinal data analysis is not reported in the registry record. Details about covariates, model structure, and handling of repeated measurements are therefore unavailable for independent reproduction. The record also does not report a multiplicity plan, a testing hierarchy, interim analyses, or the sample-size justification. These omissions make it difficult to assess the prespecified status and overall error control of the multiple reported comparisons.

Although the trial is recorded as randomized, double-blind, and multicenter, the randomization procedure, allocation concealment, masking procedures, and number of centers are not reported in the registry record. Participant characteristics beyond the condition and inadequate control on metformin are not described in the registry record, limiting assessment of how closely the enrolled population may match other groups. Detailed numerical results for all listed secondary outcomes are also not reported in the registry record.

The safety postings are counts and selected event terms, without a reported comparative safety analysis in the registry record. Death counts are not reported in the registry record. The results therefore offer a limited description of the posted safety data and do not establish causal attribution or a complete balance of benefits and harms. These limitations concern what can be concluded from the registry record, not necessarily everything collected or analyzed during the trial.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Ertugliflozin and Sitagliptin Co-administration Factorial Study (VERTIS FACTORAL, MK-8835-005) — NCT02099110.
• Registry record last updated 2018-09-12; status last verified 2018-08.
• Results first posted 2017-11-01.
• 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.

19. Record Summary

NCT02099110 is a completed Phase III randomized, double-blind, multicenter, parallel-group study of ertugliflozin and sitagliptin, alone and in combination, alongside metformin for participants with type 2 diabetes and inadequate glycemic control on metformin monotherapy. The registry records enrollment of 1233 and treatment regimens lasting 52 weeks.

The primary outcome was change from baseline in A1C at Week 26, excluding data after rescue therapy began. Sponsor-reported constrained longitudinal analyses found negative differences in least-squares means for the posted combination-versus-single-drug comparisons, with 95% confidence intervals below zero and p < 0.001. The registry also lists secondary metabolic outcomes and safety counts. Details of multiplicity control, missing-data methods, and the complete primary model specification are not reported in the registry record.