This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Metformin Hydrochloride in Treating Patients With Prostate Cancer Undergoing Surgery (NCT01433913), a randomized phase 2 study. The primary outcome is Cell Proliferation in the Prostatectomy Tissue as Assessed by Ki67 Expression Using Immunohistochemistry (IHC).
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT01433913 |
| Phase | Phase 2 |
| Condition | Adenocarcinoma of the Prostate, Recurrent Prostate Cancer, Stage I Prostate Cancer |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Double |
| Enrollment | 20 |
| Arms in the primary outcome | 2 |
| Lead sponsor | National Cancer Institute (NCI) |
| Results posted | Yes |
2. Clinical Question
Population
Patients with prostate cancer undergoing surgery were eligible for the pre-prostatectomy study. The primary outcome population was narrower: participants with tissue sections available from prostatectomy. The registry lists adenocarcinoma, recurrent prostate cancer, and stage I, IIA, and IIB prostate cancer among the conditions studied. Enrollment was 20.
Intervention
One group received extended-release metformin hydrochloride by mouth once daily for 4-12 weeks before surgery. The registry describes the study as a phase II trial. Its brief summary gives a proposed biological rationale involving enzyme activity and tumor-cell growth; this rationale is not itself evidence of an effect on the primary outcome.
Comparator
The comparison group received placebo by mouth once daily for 4-12 weeks. The study used randomized allocation, double masking, and a parallel-group model. The registry does not report further details of the allocation procedure or masking implementation.
Primary question
The primary question was whether Ki67 expression in prostatectomy tissue differed between the metformin and placebo groups. Ki67 was assessed by immunohistochemistry and expressed as the percentage of positively stained nuclei. The outcome was specified at 12 weeks. The registry reports group medians and inter-quartile ranges and a Wilcoxon (Mann-Whitney) comparison, with p = 0.23. The record does not report a confidence interval for the between-group effect, so the precision of that effect cannot be read directly from a confidence interval.
3. Trial Design
This completed phase II study randomized participants to metformin hydrochloride or placebo in a parallel-group, double-blind design. Enrollment was 20. The trial focused on patients with prostate cancer undergoing surgery, and the primary outcome population was participants with prostatectomy tissue sections available. The registry does not report the number randomized to each group, or explain why tissue was unavailable for some participants.
Participants in the metformin group received extended-release metformin hydrochloride by mouth once daily; participants in the comparator group received placebo by mouth once daily. Both regimens are described as lasting 4-12 weeks. The primary outcome timeframe is 12 weeks. The record does not give more detail about how the treatment duration within that range was determined.
The primary endpoint was cell proliferation in prostatectomy tissue, assessed using Ki67 expression by immunohistochemistry. The measure was the percentage of positively stained nuclei. Group results were reported as medians with inter-quartile ranges. The registry describes a two-group t-test at a two-sided significance level of 0.05, with a non-parametric rank-sum test to be used if the data were not normally distributed. The analysis actually reported for the primary comparison was Wilcoxon (Mann-Whitney), with p = 0.23.
Other registered outcomes included tissue drug concentration, several tissue biomarkers, and changes in serum PSA. The registry lists 14 outcome measures and 13 posted secondary outcomes. It does not report additional design information such as the randomization sequence method, allocation concealment, or detailed procedures for maintaining masking.
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
| Stage | Arm I (Metformin Hydrochloride) | Arm II (Placebo) |
|---|---|---|
| STARTED | 10 | 10 |
| COMPLETED | 9 | 10 |
| NOT COMPLETED | 1 | 0 |
Analysis population for the primary outcome, as defined in the registry: Participants with tissue sections available from prostatectomy
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.
| Characteristic | Unit | Arm I (Metformin Hydrochloride) | Arm II (Placebo) | Total |
|---|---|---|---|---|
| Participants analysed | 10 | 10 | 20 | |
| Age, Categorical: <=18 years | Participants | 0 | 0 | 0 |
| Age, Categorical: Between 18 and 65 years | Participants | 4 | 6 | 10 |
| Age, Categorical: >=65 years | Participants | 6 | 4 | 10 |
| Age, Continuous | years | 65 (SD 10) | 61 (SD 6) | 63 (SD 8) |
| Sex: Female, Male: Female | Participants | 0 | 0 | 0 |
| Sex: Female, Male: Male | Participants | 10 | 10 | 20 |
| Region of Enrollment: United States | participants | 10 | 10 | 20 |
6. Primary Endpoint
The registry lists the primary outcome as Cell Proliferation in the Prostatectomy Tissue as Assessed by Ki67 Expression Using Immunohistochemistry (IHC), measured over 12 weeks.
| Element | Registry entry |
|---|---|
| Result type | Median |
| Measure of spread | Inter-Quartile Range |
| Unit | % positively stained nuclei |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint measured cell proliferation in prostatectomy tissue using Ki67 expression assessed by immunohistochemistry. The registry defines the unit as the percentage of positively stained nuclei and specifies a 12-week timeframe. The outcome population consisted of participants with tissue sections available from prostatectomy.
As a general statistical and biological concept, a tissue biomarker can provide information about a biological process in the sampled tissue. Ki67 staining is used as an indicator related to cell proliferation. It is not, by itself, a direct measure of how a participant feels, functions, or survives, and a difference in a tissue marker would not automatically establish a difference in a clinical outcome.
Studying tissue collected at surgery allows investigators to examine a biological measure in the organ of interest after the pre-surgical treatment period. In this trial, that focus aligns with the registry’s stated interest in laboratory biomarkers. The endpoint should therefore be interpreted as a tissue-level comparison between the randomized groups, rather than as a measure of symptoms, disease progression, or long-term clinical benefit.
The registry reports the endpoint as a median and inter-quartile range for each group. Those summaries describe the observed distributions among participants with available tissue. The record does not provide a confidence interval for the between-group contrast, nor does it report an effect estimate in the primary analysis entry.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
For the primary tissue outcome, the metformin group had a reported median Ki67 value of 6.5, with an inter-quartile range from 3.33 to 9.33. The placebo group had a median of 3.67, with an inter-quartile range from 2.0 to 6.0. These are summaries among participants with available prostatectomy tissue: 8 in the metformin group and 9 in the placebo group.
The registry reports a Wilcoxon (Mann-Whitney) comparison with p = 0.23. This is the reported test result for the comparison; it is not a measure of the size or clinical importance of a difference. The p-value is not below the two-sided 0.05 significance level stated in the registry’s planned analysis description. It should not be read as proof that the groups were identical or that no effect exists. With relatively small analyzed groups and no reported confidence interval for the between-group effect, the result does not show how precisely the effect has been estimated.
The numerical group medians alone do not provide a direct estimate of the difference between groups, and no such estimate is reported in the primary analysis entry. The inter-quartile ranges describe the middle half of the observed values in each group; they are not confidence intervals for the group medians or for a treatment contrast. Their overlap should not be used as a substitute for the reported test or for a confidence interval.
The primary outcome concerns Ki67 staining in tissue collected at surgery. It is a biomarker endpoint, not a clinical endpoint. The registry also lists tissue and serum secondary outcomes, but a collection of biomarker measures does not by itself establish effects on clinical outcomes. Results for additional outcomes would require their own estimates and appropriate attention to the number of comparisons.
Interpretation is also limited by the endpoint analysis population. Enrollment was 20, while the primary outcome is reported for participants with available prostatectomy tissue, with group counts of 8 and 9. Reasons for missing tissue, other missing outcomes, and the handling of missing data are not reported in the registry record. The observed results therefore describe the reported tissue-available groups; the record does not establish how missingness may relate to the outcome.
9. Statistical Methods Explained
The registry describes a two-group t-test at a two-sided 0.05 significance level as the planned primary comparison. It also states that a non-parametric rank-sum test would be used if data were not normally distributed. The primary analysis actually reported is Wilcoxon (Mann-Whitney), a rank-sum procedure, with p = 0.23. The registry does not report whether the test choice followed a formal assessment of the distribution, or give further details of the analysis procedure.
What the planned t-test describes
As a general statistical explanation, a two-group t-test compares group means. Its usual model-based interpretation relies on assumptions about the observations and the distribution of the outcome, with particular attention to normality when samples are small. The registry’s description says the test was planned for the two groups and specifies a two-sided significance level. A two-sided test allows for a difference in either direction; it does not prespecify that only a higher or only a lower value would count as evidence against the null hypothesis.
The posted group summaries are medians and inter-quartile ranges, rather than means and standard deviations. These summaries do not establish whether the data were normally distributed. The registry’s use of a rank-sum analysis is consistent with its description of a non-parametric alternative, but the reason for selecting that analysis is not reported in the registry record.
What a Wilcoxon rank-sum test compares
As a general explanation, the Wilcoxon rank-sum test, also known as the Mann-Whitney test, uses the ordering of observations across groups rather than directly comparing group means. Its test result reflects how the ranks are distributed between the groups. It is often described informally as a test of medians, but that interpretation requires additional assumptions about the shapes of the group distributions. Without those assumptions, a significant result can reflect differences in distribution beyond a simple shift in medians.
The registry reports p = 0.23 for this test. A p-value is calculated under a null model and describes how compatible the observed test statistic is with that model. It is not the probability that the null hypothesis is true, nor is it the probability that the result occurred by chance in a general sense. It does not quantify the size of a treatment effect. The record does not report the test statistic, a confidence interval for a group contrast, or a separate effect estimate.
Why report medians and inter-quartile ranges
As a general explanation, the median marks the central point of an ordered set of observations, while the inter-quartile range gives the span between the first and third quartiles. These summaries are less sensitive than the mean and standard deviation to extreme values. They are commonly used to describe skewed or otherwise non-normal data. In this trial, the reported values are 6.5 with an inter-quartile range of 3.33 to 9.33 for metformin, and 3.67 with an inter-quartile range of 2.0 to 6.0 for placebo.
A confidence interval would communicate uncertainty around an estimated quantity, such as a median or a between-group effect, under the analysis method used. It would help readers assess the range of effects compatible with the data and model. No confidence interval for the primary between-group effect is reported in the registry record. The inter-quartile ranges are descriptive spreads of individual observations and should not be mistaken for confidence intervals.
Small groups and precision
As a general statistical consideration, small groups often yield imprecise estimates, and distributional assumptions are harder to assess from limited data. This does not make a test result invalid by itself, but it means that estimates and uncertainty intervals are important for understanding what the data can distinguish. Here, the primary tissue outcome is reported for 8 participants in the metformin group and 9 in the placebo group. The reported p-value alone cannot convey the range of plausible effect sizes.
The registry also lists a sample-size planning illustration based on a fraction of the observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation. Such illustrations are planning scenarios, not observed power, and do not establish the trial’s original design rationale. A sample-size justification for this trial is not reported in the registry record. Numeric planning results and further details of their assumptions are not stated in the record information described for this analysis.
10. Analyses Reported by the Sponsor
| Method | Estimate | 95% CI | P-value |
|---|---|---|---|
| Wilcoxon (Mann-Whitney) | not reported | 0.230 |
The registry-reported primary comparison uses the Wilcoxon (Mann-Whitney) method and gives p = 0.23. It reports the group medians and inter-quartile ranges, but no confidence interval or effect estimate for the contrast. These are the results to use when describing the registry’s own analysis.
A separate unadjusted re-calculated comparison estimate is not reported in the registry record, so numerical agreement between such a calculation and the registry analysis cannot be assessed. Recalculations from summary statistics would not necessarily reproduce a rank-based analysis: the underlying observations and their ranks are not available in the registry result. The median and inter-quartile range summaries do not contain enough information to reconstruct the Wilcoxon test statistic.
The reported p-value should not be treated as an effect size, and the medians should not be treated as a formal estimate of the between-group effect. In the absence of a reported confidence interval, the record also does not directly communicate the precision of that effect. The sponsor’s analysis, as posted in the registry, is therefore best described by its stated test, p-value, and group summaries, without substituting an unreported recalculation.
11. Multiplicity, Missing Data and Analysis Populations
The registry lists 14 outcome measures and 13 posted secondary outcomes, in addition to the primary Ki67 tissue endpoint. The secondary outcomes include tissue drug concentration, apoptosis, cell-cycle and mTOR-related biomarkers, angiogenesis, AMPK activation, retinoblastoma protein phosphorylation, and changes in serum PSA. The record also lists 10 analyses. The number and range of outcomes make it important to distinguish the primary comparison from secondary and exploratory findings.
Multiplicity arises when several outcomes or comparisons are assessed, because the chance of at least one apparently unusual result can increase across a family of tests. As a general statistical principle, investigators may use prespecified hierarchies or adjustment procedures to address this issue. The registry record does not report a multiplicity adjustment or a hierarchy for interpreting the secondary outcomes. It also does not report how the secondary analyses should be treated relative to the primary endpoint.
For the primary analysis population, the registry specifies participants with tissue sections available from prostatectomy. The primary outcome group counts are 8 for metformin and 9 for placebo; enrollment was 20. The record does not explain the difference between enrollment and the tissue-available analysis groups, report reasons for tissue unavailability, or describe an imputation or other missing-data procedure. It does not report whether missingness was assessed for its potential relationship to treatment assignment or outcome.
The analysis population for each secondary outcome, and the handling of missing data for those measures, are not reported in the registry record. Nor are details reported about adherence, exclusions after randomization, or whether analyses followed an intention-to-treat principle. These omissions limit assessment of how representative the analyzed participants are of the randomized groups and whether the outcome analyses preserve the benefits of randomization. They do not establish that any particular bias occurred; they leave that question unresolved from the registry information.
12. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Arm I (Metformin Hydrochloride) | 1 | 10 | 10.0% |
| Arm II (Placebo) | 0 | 10 | 0.0% |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 5% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Arm I (Metformin Hydrochloride) | 6 of 10 |
| Arm II (Placebo) | 5 of 10 |
Most frequent serious adverse event terms
| Event term | Organ system | Arm I (Metformin Hydrochloride) | Arm II (Placebo) |
|---|---|---|---|
| influenza | Infections and infestations | 1 of 10 | 0 of 10 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Arm I (Metformin Hydrochloride) | Arm II (Placebo) |
|---|---|---|---|
| Diarrhea | Gastrointestinal disorders | 2 of 10 | 0 of 10 |
| Nausea | Gastrointestinal disorders | 1 of 10 | 1 of 10 |
| Headache | Nervous system disorders | 1 of 10 | 1 of 10 |
| Urinary retention | Renal and urinary disorders | 1 of 10 | 1 of 10 |
| Constipation | Gastrointestinal disorders | 1 of 10 | 0 of 10 |
| Bloating | Gastrointestinal disorders | 0 of 10 | 1 of 10 |
| Diverticulosis | Gastrointestinal disorders | 1 of 10 | 0 of 10 |
| Insomnia | Psychiatric disorders | 0 of 10 | 1 of 10 |
| Upper respiratory infection | Infections and infestations | 1 of 10 | 0 of 10 |
| Chest pain | Cardiac disorders | 0 of 10 | 1 of 10 |
Serious events with the highest arm-level rate: influenza (10.0%).
The safety record lists 10 participants at risk for serious adverse events in each group. One participant in the metformin group is recorded as affected by a serious adverse event, and none in the placebo group is recorded as affected. The record lists influenza as the leading serious adverse event term at 10.0%; it does not identify that term by treatment group in the information.
The record also lists other adverse events affecting 6 participants at risk of 10 in the metformin group and 5 participants at risk of 10 in the placebo group. These counts are separate from the serious adverse event figures and should not be combined with them. Deaths are not reported in the registry record.
These counts describe what was recorded in a small study; they are not, by themselves, a precise comparison of safety between groups. The registry information does not give the full event-level detail, timing, duration, severity distribution, or attribution needed to characterize individual events or compare event patterns comprehensively. The serious-event count should therefore be read as a descriptive registry entry, not as a definitive estimate of relative risk.
13. What the Result Does, and Does Not, Mean
The posted primary result describes Ki67 staining among participants with prostatectomy tissue available. The metformin group’s median was 6.5, and the placebo group’s median was 3.67; the registry-reported Wilcoxon (Mann-Whitney) test gave p = 0.23. The analysis is a comparison of a tissue biomarker in this study population at the specified pre-surgical timeframe.
The result does not establish that the groups were identical. A p-value of 0.23 does not prove the absence of a difference, and the registry does not report a confidence interval that would show the range of effects compatible with the data. Conversely, the different observed medians do not by themselves establish a treatment effect. The recorded analysis does not provide a between-group effect estimate, and the small tissue-available groups make precision especially important to assess.
The endpoint is not a clinical outcome such as symptoms or survival. As a general interpretation principle, evidence about a biological marker is not interchangeable with evidence about how a disease affects patients. The registry’s description of a proposed biological rationale does not demonstrate that the intervention changed the marker, and a marker result alone would not establish a clinical benefit.
The result also cannot answer questions beyond the registered comparison, such as effects over a longer period or in groups not represented by this pre-prostatectomy cohort. Follow-up outcomes and analysis details not included in the registry record cannot be inferred from the primary result. Overall, the posted finding is a non-significant rank-based comparison of the available tissue measurements, with uncertainty about the size and precision of any underlying difference.
14. Why This Trial Matters Statistically
This trial illustrates why endpoint type, analysis method, and analysis population need to be read together. Its primary measure is Ki67 staining in prostatectomy tissue, summarized by medians and inter-quartile ranges. The reported inferential test is a rank-based comparison. Readers should not assume that a test of ranks is a direct test of the difference between the reported medians without additional distributional assumptions.
It also shows why a p-value alone is an incomplete account of a trial result. The registry reports p = 0.23, but no confidence interval or between-group effect estimate. The test result does not communicate the range of effects compatible with the data, while the group summaries do not substitute for that uncertainty information. This is particularly relevant when the primary endpoint is available for 8 and 9 participants in the two groups.
The registry’s planned analysis description and reported analysis also make statistical reporting choices visible: it describes a t-test with a non-parametric rank-sum alternative and posts a Wilcoxon (Mann-Whitney) result. Clear reporting of the reason for the final method, the corresponding effect estimate, and its uncertainty would help readers interpret the finding. The registry does not report that additional rationale.
Finally, the record includes numerous secondary biomarker outcomes alongside the primary endpoint. Distinguishing confirmatory questions from exploratory comparisons matters when interpreting multiple results. The record does not report a multiplicity strategy. These features make the study useful for understanding how a biomarker-focused randomized trial should be assessed, without treating a single statistical result as a complete account of the evidence.
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 endpoint was analyzed among participants with prostatectomy tissue sections available, rather than all enrolled participants. The primary group counts are 8 and 9, while enrollment was 20. The registry does not report why tissue was unavailable for some participants, whether availability differed for reasons related to treatment or outcome, or how missing primary and secondary outcome data were handled.
The primary result is presented as group medians and inter-quartile ranges with a Wilcoxon (Mann-Whitney) p-value. No confidence interval or effect estimate for the between-group contrast is reported. This limits assessment of precision and of the magnitude of any difference. The record also does not explain why the posted rank-based test was selected instead of the t-test described in the analysis plan, or provide information about distributional checks.
The study has several registered outcome measures, including 13 posted secondary outcomes, but a multiplicity strategy is not reported. The record does not state how these outcomes were prioritized or how the secondary analyses should be interpreted relative to the primary test. Detailed analysis populations for the secondary outcomes are also not reported.
Further trial details relevant to assessment, including the randomization sequence method, allocation concealment, masking procedures, adherence, and post-randomization exclusions, are not reported in the registry record. A sample-size justification and details of any interim analyses are also not reported. These omissions limit evaluation of design and analysis choices; they do not demonstrate that a particular design flaw or bias occurred.
16. Related Tutorials
17. Related Statistical Calculators
18. Sources
- ClinicalTrials.gov: Metformin Hydrochloride in Treating Patients With Prostate Cancer Undergoing Surgery — NCT01433913.
- Registry record last updated 2018-01-10; status last verified 2017-12.
- Results first posted 2015-05-15.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
19. Record Summary
This completed randomized, double-blind phase II study compared extended-release metformin hydrochloride with placebo in patients with prostate cancer undergoing surgery. Enrollment was 20. The primary endpoint was Ki67 expression in prostatectomy tissue, expressed as the percentage of positively stained nuclei at 12 weeks.
Among participants with tissue available, the metformin group had a median of 6.5 and an inter-quartile range of 3.33 to 9.33; the placebo group had a median of 3.67 and an inter-quartile range of 2.0 to 6.0. The registry-reported Wilcoxon (Mann-Whitney) comparison gave p = 0.23. No confidence interval or primary effect estimate is reported. The registry lists 13 secondary outcomes, while details of missing-data handling and multiplicity procedures are not reported.