Feasibility Study of Metformin Therapy in ADPKD: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Feasibility Study of Metformin Therapy in ADPKD (NCT02903511), a randomized phase 2 study. The primary outcome is Safety and Tolerability of Metformin.

2. Clinical Question

The study’s stated purpose was to assess whether metformin was safe and well tolerated in people with ADPKD who were not diabetic and had slightly decreased kidney function. It also planned to evaluate effects on kidney growth and kidney function. The listed secondary outcomes include change in total kidney volume, change in kidney function and the rate of serious adverse events.

The primary question is about the ability to remain prescribed a planned dose over the study period. That makes it a feasibility and tolerability question, rather than a direct test of whether kidney outcomes improve. The registry describes a randomized comparison with placebo, but does not report the primary endpoint values or an analysis comparing the arms. Consequently, the record does not establish how often participants met either dose-prescription threshold, or how those percentages differed between groups.

3. Trial Design

NCT02903511 is a phase 2 randomized study of metformin therapy in ADPKD. It is listed as completed, with enrollment of 56. The registry lists University of Colorado, Denver as the sponsor. The recorded start is 2016-11 and primary completion is 2020-08-17.

The design is described as randomized, quadruple masked and parallel. In a parallel design, participants are assigned to one study arm and remain in that assigned arm for comparison over the trial period; they do not cross over between treatments. Masking is intended to limit the influence of knowing treatment assignment on study conduct or assessment. The registry identifies quadruple masking but does not specify in the information which roles were masked or how masking was maintained.

The two listed interventions are metformin and placebo. Both arms had the same planned tablet schedule: starting with 500 mg twice a day, increasing by 500 mg every 2 weeks as tolerated, up to 1000 mg by mouth twice a day, for 12 months. This parallel comparison is intended to distinguish the experience under metformin assignment from that under placebo assignment, while accounting for the dose schedule described in the record.

The primary outcome is framed as percentages of participants meeting dose-prescription thresholds at 12 months. The record also lists secondary outcomes related to total kidney volume, kidney function and serious adverse events. It records four outcome measures, but reports no analyses. The primary outcome is flagged as having several categories or time points and therefore is not reduced to a single comparison in the posted data.

The registry does not report the planned sample-size justification, detailed randomization procedures, analysis populations or a statistical analysis plan. The planning assumptions associated with illustrative sample-size calculations are not the same as evidence that the trial was designed using those assumptions. They also do not supply the missing outcome results.

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: 5 participants screen failed or dropped out prior to randomization

Overall Study

Reasons for not completing this period:

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 Safety and Tolerability of Metformin, measured over 12 months.

The primary endpoint describes dose persistence at 12 months. It asks for the percentage of participants who were still prescribed the full randomized dose and the percentage prescribed at least 50% of the randomized dose. These are related but distinct summaries: one concerns remaining at the full planned dose, while the other uses a lower dose threshold.

This kind of endpoint is used in a feasibility study because the practical ability to continue assigned treatment is relevant to whether a larger study can be carried out. Dose reduction or stopping may reflect difficulty tolerating treatment or other circumstances, but the registry record does not report reasons for dose changes or discontinuation. The measure is about what participants were prescribed, not necessarily what they took.

The registry lists the timeframe as 12 months, matching the planned treatment duration. It does not report the percentages for either arm, the numbers contributing to those percentages, confidence intervals, or a comparison between metformin and placebo. The endpoint description also refers to several categories or time points, and the record does not reduce it to one comparison. It is therefore not possible to determine from the posted primary outcome information how dose persistence differed between the randomized groups.

Although kidney growth and kidney function were among the study’s broader aims, neither is the primary endpoint described in the registry entry. The listed secondary outcomes include change in total kidney volume and change in kidney function, but their results are not reported in the information.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical interpretation depends on distinguishing the trial’s design, its endpoint definitions, and its observed results. The registry describes a randomized, parallel comparison of metformin and placebo, with quadruple masking. Random allocation is intended to make treatment groups comparable on average at the point of assignment. It does not guarantee identical groups, particularly in a study of limited size, and it does not substitute for reporting outcome data and an analysis.

What does a percentage endpoint represent?

A percentage is a proportion expressed on a percentage scale: the numerator is the number of participants meeting a stated condition, and the denominator is the number assessed for that condition. For the primary endpoint, the conditions are being prescribed the full randomized dose and being prescribed at least 50% of that dose at 12 months. The registry does not report the numerators or denominators for these primary measures. Without them, the percentages cannot be assessed or reproduced from the registry entry.

These two dose thresholds should not be treated as interchangeable. The full-dose measure is more restrictive, while the threshold of at least 50% includes a broader range of prescribed doses. The record does not say whether a participant could contribute to both categories, how changes between doses were handled, or how discontinuation was classified. Those operational details are not reported in the registry record.

Why report uncertainty with an estimate?

In general, a confidence interval describes the range of values reasonably compatible with an estimate under the analysis assumptions. It communicates precision as well as the estimated effect. A narrow interval indicates greater precision than a wide interval, though the interval does not remove the need to consider study design, missing data or potential bias. The registry does not report confidence intervals for the primary endpoint or for the serious adverse-event counts.

For an arm comparison, the analysis would need to define the effect measure and its direction. For example, a comparison could describe a difference between percentages, but no such contrast is reported for the primary outcome. No effect estimate should be inferred from the endpoint description or from the adverse-event counts alone.

What can the adverse-event counts show?

The serious adverse-event entries are descriptive counts with an at-risk denominator for each group. The metformin entry records 2 affected participants among 26 at risk; the placebo entry records 0 among 25 at risk. These figures indicate what was recorded in the registry, but they do not provide an adjusted or unadjusted comparative estimate, an uncertainty interval, or a test of a group difference. For uncommon outcomes, counts can be especially imprecise, and a count of zero does not imply that the true event probability is zero.

General statistical methods for comparing proportions may include a confidence interval for each arm or for a contrast between arms. The method should reflect the design, event frequency and analysis plan. The registry does not report which inferential method, if any, was used for these adverse-event data. No method should be attributed to the investigators without that information.

What do the planning assumptions mean?

The registry includes assumptions for planning illustrations: a two-sided significance level of 0.05, equal allocation between arms, a normal approximation, and an effect to detect defined as a stated fraction of the observed effect. These are assumptions for illustrative sample-size calculations, not evidence of the trial’s original design rationale, and the record explicitly distinguishes them from the trial’s own design. The associated calculations are not observed power and do not establish that the study had adequate power for its primary endpoint.

In general, sample-size planning asks how many participants would be needed to detect a specified effect under stated assumptions. Such a calculation depends on the outcome, anticipated effect, variability or event rate, allocation and statistical procedure. A planning illustration based on an observed effect is retrospective in its reference to that effect; it does not replace a prespecified design justification. The registry does not report the protocol’s sample-size justification or its original statistical plan.

Why does the endpoint structure matter?

The record flags the primary outcome as having several categories or time points and does not reduce it to one comparison. That prevents a single summary estimate from representing all aspects of the stated endpoint. Separate thresholds can answer different questions, and multiple time points can create multiple opportunities to observe a result. The record does not specify how those outcomes were prioritized or whether any formal adjustment was planned.

In general, over-dispersion is relevant when observed variability exceeds what a model assumes. It can make standard errors too small and confidence intervals too narrow if ignored. The registry does not report a model, a variance estimate or evidence of over-dispersion for the outcomes in this trial, so no conclusion about its presence or handling is warranted.

10. Multiplicity, Missing Data and Analysis Populations

The primary outcome description contains more than one dose threshold and refers to several categories or time points. When a study evaluates multiple outcomes, thresholds or time points, the chance of finding at least one apparently notable result can increase if each is treated as a separate test. General approaches include defining one primary contrast in advance, specifying a hierarchy among outcomes, or adjusting inference for multiple comparisons. The registry record does not report whether any such approach was planned or used.

The registry lists four outcome measures and no analyses. It does not report a statistical analysis plan, a prespecified hierarchy, or adjustments for multiplicity. Without those details and the corresponding estimates, the strength of evidence across multiple outcome measures cannot be assessed.

Missing-data handling is also not reported in the registry record. The primary endpoint is defined at 12 months, but the record does not state how participants without a 12-month assessment, those who stopped treatment, or those with incomplete dose information were handled. It also does not report whether outcomes were analyzed according to randomized assignment or in another analysis population.

These omissions do not show that missing observations occurred or that a particular analysis was inappropriate. They mean the registry entry does not explain how missingness, exclusions or analysis populations were addressed. Those details are important because estimates can depend on which participants are included and how unavailable measurements are treated.

The record does not report interim analyses or stopping rules. Their presence or absence should not be inferred from the completion status. The registry therefore supports no conclusion about interim monitoring, multiplicity control, or missing-data assumptions.

11. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

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: infection (7.7%).

The registry lists serious adverse events by group. In the metformin group, 2 participants were affected among 26 at risk. In the placebo group, 0 participants were affected among 25 at risk. No deaths are listed in either group. These are descriptive registry counts, not a formal estimate of the difference in serious adverse-event risk.

A count of zero in one group should not be read as proof that serious events cannot occur under that assignment. With finite observations, uncommon outcomes may not appear even when their underlying probability is not zero. The record does not provide confidence intervals or a statistical comparison, so the precision of any difference is not reported.

The leading serious-event term is Infection, reported as 7.7%. The available entry does not identify a group-specific rate for that term or provide enough detail to interpret it as a comparison. It also does not report whether events were judged related to study treatment. The serious-event counts alone cannot establish causation.

The reported at-risk denominators are outcome-specific figures. The registry does not explain how they relate to enrollment or describe any missing safety follow-up. These limitations warrant reading the counts as the posted descriptive record, without extending them to an unreported causal or comparative conclusion.

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

This registry record establishes that a randomized, quadruple-masked, parallel phase 2 study compared a planned metformin regimen with placebo for 12 months in people with ADPKD. It defines dose persistence as the primary outcome and lists kidney volume, kidney function and serious adverse events among the secondary outcomes. It also reports serious adverse-event counts and deaths by group.

It does not establish the primary outcome result. The percentages meeting the full-dose and at-least-50% dose thresholds are not reported, and there is no primary endpoint comparison or uncertainty interval. As a result, the record cannot show from these data how tolerable the assigned dose schedule was, how often participants remained prescribed those doses, or whether the groups differed on dose persistence.

The reported serious adverse-event counts are useful as descriptive information, but they are not enough to establish a treatment effect on safety. The record gives no comparative analysis, confidence intervals or causal assessment. The listed Infection term does not include a group-specific result in the available entry.

The study’s stated aims also included kidney growth and kidney function. The listed outcomes alone do not show whether either changed, and the registry does not report those results in the information. No conclusion about effects on kidney outcomes follows from the design description or the adverse-event counts.

In general, a feasibility study can inform whether a treatment schedule and outcome collection are practical to study. The registry’s stated focus on safety and tolerability is consistent with that kind of question. However, without the primary endpoint results and supporting analysis details, the posted record cannot quantify feasibility or resolve how the groups compared. It should not be interpreted as evidence for or against clinical benefit.

13. Why This Trial Matters Statistically

This trial is statistically notable because its primary question concerns dose persistence rather than a direct clinical kidney outcome. Whether participants remain prescribed a full or partial randomized dose can help characterize the practicality of a regimen for study. It is a different estimand from change in total kidney volume or kidney function, which the registry lists as secondary outcomes.

The randomized, masked, parallel design provides a structure for comparing assigned treatment with placebo. But a sound design description is only one part of a statistical interpretation. Readers also need outcome counts or estimates, denominators, measures of uncertainty, and information about the analysis population. Those primary endpoint results and analysis details are not reported in the registry entry.

The record’s serious adverse-event counts illustrate why raw event figures require context. Counts can describe what was recorded, but without an analysis and uncertainty they do not establish a precise group difference. The listed leading serious-event term likewise cannot be interpreted as a group-specific comparison from the available information.

The trial’s listed aims span tolerability, kidney growth and kidney function. These are distinct questions, and evaluating several outcomes or categories raises issues of prioritization and multiplicity. The registry does not report how those issues were addressed. The primary endpoint’s multiple categories or time points are specifically not reduced to a single comparison.

For biostatisticians and clinical researchers, the record thus illustrates the distinction between trial registration, descriptive results and inferential evidence. The design and outcome definitions provide useful context, while the absent primary outcome values, uncertainty estimates and analysis plan limit what can be concluded. The limitation is not a finding of no effect; it is that the posted record does not report enough information to estimate the primary comparison.

14. 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 main trial-specific limitation for interpretation is that the registry record does not report the primary endpoint results. It defines two dose-prescription thresholds at 12 months but gives no percentages, numerator or denominator for either threshold, arm comparison, or uncertainty interval. The primary outcome is also flagged as having several categories or time points rather than one reduced comparison.

Although the record lists four outcome measures and secondary outcomes involving total kidney volume, kidney function and serious adverse events, their analyses are not reported. The registry lists no analyses. It also does not provide the statistical analysis plan, sample-size justification, analysis populations, multiplicity approach, missing-data procedures or any interim-analysis information.

For safety, group-specific serious-event counts and deaths are reported, but the entry provides no inferential comparison, uncertainty estimates or treatment-relatedness assessment. The leading serious-event term, Infection, is reported as 7.7%, without enough detail in the available information to establish its group-specific attribution. The at-risk denominators are recorded, but the registry does not explain how they relate to enrollment or how incomplete safety follow-up was handled.

The registry identifies quadruple masking but does not specify which roles were masked or describe how masking was maintained. It gives a treatment schedule but does not report reasons for dose changes, discontinuation, or whether prescription status reflected medication actually taken. These details matter for interpreting a dose-persistence endpoint, but are not reported in the registry record.

These are limitations of the information available for interpretation, not proof that the underlying study lacked procedures or analyses. The posted record does not permit a quantitative conclusion about the primary endpoint or kidney outcomes.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Feasibility Study of Metformin Therapy in ADPKD — NCT02903511.
• Registry record last updated 2021-09-05; status last verified 2021-08.
• Results first posted 2021-09-05.
• 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.

18. Record Summary

NCT02903511 is a completed randomized, quadruple-masked, parallel phase 2 study sponsored by University of Colorado, Denver. It enrolled 56 participants and compared metformin with placebo in autosomal dominant polycystic kidney disease. Both arms had a planned dose-escalation schedule for 12 months.

The primary outcome concerned the percentages prescribed the full randomized dose and at least 50% of that dose at 12 months. The registry record does not report those results or any analyses. Listed secondary outcomes include changes in total kidney volume and kidney function, and the rate of serious adverse events.

The serious adverse-event entries report 2 affected participants among 26 at risk for metformin and 0 among 25 at risk for placebo. No deaths are listed in either group. These counts are descriptive; no comparative analysis or uncertainty estimates are reported.