Metformin as a Novel Therapy for Autosomal Dominant Polycystic Kidney Disease: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Metformin as a Novel Therapy for Autosomal Dominant Polycystic Kidney Disease (NCT02656017), a randomized phase 2 study. The primary outcome is Change in the Gastrointestinal Symptoms Rating Scale (GSRS) to 24 Months.

2. Clinical Question

3. Trial Design

The registry identifies NCT02656017 as a completed, randomized, parallel-group phase 2 study. Enrollment was 97. The record lists 49 participants in the metformin arm and 48 in the placebo arm for the primary outcome. The study began on 2016-06-27, and primary completion was recorded as 2020-12-07. These dates describe the registry record; they do not establish the duration of follow-up for every participant.

Masking is listed as quadruple. The registry does not report the specific roles masked, how masking was maintained, or whether masking was assessed. The two arms used matching scheduled dose titrations as described in the record. The intended metformin regimen reached 1000 mg twice daily at week 6 and continued for 26 months; the placebo arm followed the corresponding schedule.

The study focused on adults with ADPKD and beginning stages of chronic kidney disease. The brief description says investigators would assess safety and tolerability, kidney disease progression reflected in kidney size and kidney function, kidney pain, and quality of life. Secondary outcomes listed include physical and mental quality-of-life components, back-pain frequency, estimated glomerular filtration rate, and MRI measures of kidney and liver volumes, including cyst volumes.

The registry lists 14 outcome measures and 14 analyses, with 11 secondary outcomes posted. It does not report further details about the recruitment process, eligibility thresholds, treatment adherence, participant flow, or reasons for withdrawal. Those omissions limit how fully the conduct and completeness of follow-up can be assessed from the 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.

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 Change in the Gastrointestinal Symptoms Rating Scale (GSRS) to 24 Months, measured over Baseline, 2 weeks, 6 weeks, 1 month, 3 months and every 3 months thereafter to 24 months.

The primary endpoint is change in the Gastrointestinal Symptoms Rating Scale (GSRS) to 24 months. The GSRS is described as a validated, widely used 15-item questionnaire assessing gastrointestinal symptom burden. Its score ranges from 1 to 7, with higher mean scores indicating worse outcomes. The endpoint therefore summarizes reported gastrointestinal symptoms rather than a laboratory measure or an imaging result.

The registry describes mean change estimated using repeated measurements beginning at baseline and continuing at specified visits through 24 months. Those visits include 2 weeks, 6 weeks, 1 month, 3 months and every 3 months thereafter to 24 months. Using measurements across visits can characterize how scores change over follow-up and can use the repeated observations within the analysis, rather than treating each visit as an unrelated endpoint.

In general, a symptom questionnaire can capture experiences that are important to participants but may vary between people and across time. Reporting a mean change gives a group-level summary; it does not show the distribution of individual experiences or establish how many participants had a particular symptom. The registry does not report a threshold for a clinically important GSRS change, so the clinical importance of a given score difference cannot be determined from the posted result alone.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the reported mean represent?

The primary endpoint is continuous: it is a score on a scale, and the registry reports a mean change for each arm. A mean is a group-level summary that combines the observed scores into an average. It can be informative about the overall direction of change, but it does not show the range of scores, the distribution of changes, or whether every participant changed in the same direction.

The reported values are mean changes to 24 months. The metformin group mean was -0.04 and the placebo group mean was -0.11. Because higher GSRS scores indicate worse outcomes, a negative change represents a lower score relative to baseline. These values should be read as estimates, not as exact changes experienced by all participants.

Why use repeated measurements?

The registry describes observations at baseline, 2 weeks, 6 weeks, 1 month, 3 months and every 3 months thereafter to 24 months. Measurements from the same participant are related: a person’s score at one visit may resemble their score at another visit. In general, repeated-measures methods account for this within-person structure when estimating patterns or average change over time. Treating every observation as independent could understate or misrepresent uncertainty because observations from one person are not necessarily independent.

The stated method is a linear mixed model. In general, mixed models can represent repeated outcomes while accounting for participant-level clustering and can estimate average patterns across visits. The registry says the mean change to 24 months was estimated with this approach. It does not report the model’s exact specification, the covariance structure, the handling of baseline scores, or which covariates, if any, were included. Those details should not be inferred from the method label alone.

What does the confidence interval say?

The reported between-group estimate is a net mean difference of 0.07, with a 95% confidence interval from -0.13 to 0.26. In general, a confidence interval describes a range of values compatible with the data and the statistical model at the stated confidence level. It expresses uncertainty about the estimated group difference; it is not a range containing a specified proportion of individual participants’ changes.

Here the interval includes zero and extends to values on both sides of zero. This means that the estimate is imprecise enough to remain compatible with a range of possible mean differences, including no difference. It does not prove that there is no difference. Nor does the interval alone identify which values would be important to patients, because a clinically important threshold is not reported in the registry record.

How should the p-value be read?

The reported p-value is 0.50. In general, a p-value describes how unusual the observed data, or data more extreme, would be under a specified null hypothesis and the assumptions of the analysis. It is not the probability that the null hypothesis is true, and it is not a measure of the probability that the result occurred by chance. The value should be read alongside the estimated difference and its confidence interval. In this case, the confidence interval includes zero and the p-value does not indicate a statistically significant difference under conventional thresholds.

Why do model assumptions matter?

In general, a linear mixed model relies on assumptions about the relationship between the outcome and predictors, the distribution of model errors, and the structure of correlation among repeated observations. Those assumptions affect estimates and uncertainty. The record does not report diagnostic checks, model fit information, or sensitivity analyses, so their adequacy cannot be evaluated from the posted result.

Missing observations also matter in repeated-measures analyses. In general, a mixed model can use available observations under particular assumptions about the missing-data process, but it does not automatically remove the risk of bias from missingness. The registry does not report the amount or pattern of missing GSRS data, reasons for missing visits, or the specific missing-data assumptions used.

What does the analysis not establish?

The reported comparison addresses the average GSRS change under the stated analysis. It does not establish equivalence, rule out clinically important differences, or describe individual responses. It also does not substitute for results on other outcomes. The record does not report whether the analysis followed an intention-to-treat population, how protocol deviations were handled, or whether any additional analyses were prespecified. Those aspects of the analysis plan are not reported in the registry record.

10. Analyses Reported by the Sponsor

The registry reports a mixed models analysis with a net mean difference of 0.07, a 95% confidence interval from -0.13 to 0.26, and p = 0.50. This is the sponsor’s reported comparison for the primary GSRS endpoint. The record does not flag a discrepancy between that result and a recalculation. A separate simple, unadjusted recalculation of the comparison is not reported in the registry record, so there is no second estimate to set alongside the model-based result.

In general, a recalculation from posted summary statistics may differ from a trial’s own analysis because it may not use the same repeated observations, model structure, covariates, or handling of missing data. Such a calculation would be a simplified check rather than a replacement for the planned analysis. In this case, the registry’s reported mixed models estimate is the available inferential result; the record does not provide enough model detail to reproduce it independently.

The reported confidence interval includes zero and the p-value is 0.50. These figures describe uncertainty in the sponsor’s reported comparison, not proof of equivalence or proof that no difference exists. The registry does not report additional model specifications or sensitivity results that would allow assessment of how robust the estimate is to alternative analysis choices.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 14 outcome measures and 14 analyses, and 11 secondary outcomes are posted. The study description includes multiple domains, such as gastrointestinal symptoms, kidney function and volume, pain, and quality of life. When several outcomes or time points are analyzed, the chance of observing at least one apparently notable result can increase. In general, multiplicity procedures can address that issue, but the record does not report whether an adjustment was planned or applied, or how the primary and secondary results were prioritized.

The primary endpoint is identified as GSRS change to 24 months, and its reported comparison has a 95% confidence interval and p-value. The registry does not report whether there were interim analyses, a hierarchy for testing outcomes, or a prespecified statistical analysis plan. These details are not reported in the registry record and should not be inferred from the number of posted outcomes.

Missing data are another important consideration for an endpoint collected repeatedly over time. The record describes a repeated-measures linear mixed model, but does not report how many GSRS observations were missing, when they were missing, why they were missing, or whether participants discontinued treatment or follow-up. It also does not describe imputation, sensitivity analyses, or the assumptions made about missing outcomes. A model’s use of repeated observations does not by itself show that missing data had no effect on the result.

The analysis population is not reported in the registry record. In particular, the record does not say whether the primary analysis included everyone as randomized, only participants who received treatment, or another defined group. Information about protocol deviations and adherence is also not reported. Without those details, the estimate cannot be fully connected to a specified analysis population or assessed for possible effects of exclusions.

These omissions do not change the posted estimate, but they limit independent appraisal of how the analysis was conducted and how broadly it represents the randomized groups. Interpretation should remain focused on the reported result and acknowledge that multiplicity, missing-data handling, and analysis-population details are not reported.

12. 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: other - lightheadedness (2.1%), appendicitis perforated (2.1%), other - urinary tract infection and renal cyst hemorrhage (mild) (2.1%), other - renal cyst (2.1%), other - renal cyst rupture (2.1%).

The safety record lists 4 participants with serious adverse events in the metformin group and 4 in the placebo group. The groups at risk were 49 and 48, respectively. The record lists 21 participants with other adverse events in the metformin group and 13 in the placebo group. Deaths are recorded as 0 in each group. These are counts as posted; they do not establish whether the events were caused by the assigned intervention.

The serious-event terms listed include lightheadedness, perforated appendicitis, urinary tract infection and renal cyst hemorrhage (mild), renal cyst, renal cyst rupture, and diarrhea (c. difficile infection). The registry gives percentages alongside these terms, including 2.1% for several terms and 2.0% for diarrhea (c. difficile infection). These listed terms do not necessarily describe every serious event in the trial, and the record does not provide fuller event narratives or participant-level details.

In general, safety counts need context: event definitions, severity, timing, exposure, and the way events were grouped can affect their interpretation. The registry record does not report those details here, nor does it provide a statistical comparison of adverse-event rates. The figures should therefore be read as descriptive posted counts, not as evidence that the groups had the same or different safety profile.

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

The primary result is an estimate of the difference in mean GSRS change between the metformin and placebo groups through 24 months. The posted mean difference is 0.07, with a 95% confidence interval from -0.13 to 0.26 and p = 0.50. The interval includes zero, and the result does not provide clear statistical evidence of a difference in average GSRS change under the reported analysis.

That interpretation is narrower than saying the treatments are equivalent. Equivalence requires a prespecified margin and an analysis designed to assess whether the plausible difference lies within that margin. No such margin is reported in the registry record. The interval also allows values in both directions, so it does not rule out a difference that could matter; the record does not identify a threshold for clinical importance.

The result is about a group-average symptom score. It does not show each participant’s experience, the number who improved or worsened, or whether a particular symptom changed. The mean changes in the two arms are close as point estimates, but the uncertainty interval is important: the data are compatible with a range of mean differences rather than a single precise value.

The outcome is gastrointestinal symptom burden, not kidney disease progression. Although the study description and secondary outcomes cover kidney function, kidney and liver imaging measures, pain, and quality of life, their results are not reported among the registry entries for this primary endpoint. The GSRS result cannot be used to draw conclusions about those other outcomes.

Likewise, the safety counts are descriptive and do not establish causality or comparative safety. A complete interpretation of the study would require further information about other outcomes, missing data, analysis populations, and model specifications. The registry result supports a bounded conclusion about the reported primary comparison, not a broad claim about all effects of metformin in ADPKD.

14. Why This Trial Matters Statistically

This trial illustrates why a point estimate should be read together with its uncertainty. The reported net mean difference is 0.07, but the 95% confidence interval extends from -0.13 to 0.26 and includes zero. The estimate therefore does not precisely determine the direction or size of a possible difference. A p-value alone would not convey that range of uncertainty or the values still compatible with the analysis.

The endpoint also shows the role of measurement choice. GSRS is a patient-reported symptom scale, while the study’s broader aims include kidney function, imaging measures, pain, and quality of life. Each captures a different aspect of the clinical question. A result for one endpoint should not be generalized to the others, and a group mean does not describe the range of individual experiences.

The repeated-measures design is statistically relevant because the GSRS was collected over multiple visits. A model that accounts for measurements within participants can use the longitudinal structure, but interpretation depends on its specification and assumptions. The registry names the linear mixed model without reporting enough detail to assess those choices or reproduce the estimate independently.

Finally, this record demonstrates how missing details affect interpretation without changing the reported result. Multiplicity procedures, missing-data handling, analysis populations, and model diagnostics are not reported. These limits matter when assessing precision and robustness. The estimate and interval remain the central evidence for the primary comparison, while unanswered methodological questions define how cautiously that evidence should be generalized.

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.

Several trial-specific details needed for a fuller assessment are not reported in the registry record. The eligibility criteria are not described beyond adults with ADPKD and beginning stages of chronic kidney disease. The record does not specify how that stage was defined, or provide further information about recruitment, baseline characteristics, or the comparability of the randomized groups.

The registry states the planned dose titration and a 26-month duration, but does not report adherence, dose reductions, treatment interruptions, or actual exposure for participants. It also does not provide participant flow, withdrawal reasons, or completeness of follow-up. Consequently, the posted group counts do not establish that all randomized participants contributed complete endpoint data.

The primary result is identified as a repeated-measures linear mixed model, but its detailed specification is not reported. Covariate adjustment, covariance structure, model diagnostics, missing-data assumptions, and sensitivity analyses are not described. The analysis population is also not reported. These omissions prevent a full assessment of model fit, robustness, and the relationship between the reported estimate and the randomized population.

The registry does not report a clinically important GSRS difference, so the practical meaning of the confidence interval cannot be determined from a threshold. The record also does not state whether multiplicity was addressed across the 14 outcome measures and 14 analyses, or whether interim analyses occurred. Those points are relevant to interpretation, but should not be guessed.

Finally, the results discussed for the primary endpoint do not include findings for the listed secondary outcomes. The trial’s broader questions about kidney measures, pain, quality of life, and other safety outcomes therefore cannot be evaluated from this primary result alone.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Metformin as a Novel Therapy for Autosomal Dominant Polycystic Kidney Disease — NCT02656017.
• Registry record last updated 2022-08-09; status last verified 2022-08.
• Results first posted 2022-08-09.
• 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

NCT02656017 is a completed, randomized, quadruple-masked, parallel-group phase 2 trial of metformin versus placebo in adults with ADPKD and beginning stages of chronic kidney disease. Enrollment was 97. The primary endpoint was mean change in GSRS to 24 months, estimated using a repeated-measures linear mixed model.

The reported mean changes were -0.04 for metformin and -0.11 for placebo. The reported net mean difference was 0.07, with a 95% confidence interval from -0.13 to 0.26 and p = 0.50. The interval includes zero, and the registry does not report a clinically important difference threshold. These findings concern the primary gastrointestinal symptom endpoint, not the study’s other kidney, pain, quality-of-life, or safety questions.