Metformin in Castration-Resistant Prostate Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Metformin in Castration-Resistant Prostate Cancer (NCT01215032), a na phase 2 study. The primary outcome is PSA (Prostate Specific Antigen) Response.

2. Clinical Question

The clinical rationale in the registry centers on metabolic changes associated with ADT and possible links between insulin and prostate cancer. The summary notes that prior studies had suggested a relationship between insulin and tumor-cell growth, and that people taking metformin for diabetes might have different prostate-cancer outcomes. Those statements describe the rationale for studying metformin; they are not findings from this trial.

The question is therefore exploratory: how PSA changed during metformin treatment in this population. With no comparator arm, the study cannot by itself separate a change occurring during treatment from changes that might have occurred over the same period without metformin. The record also does not report a result for the primary endpoint, which prevents even a descriptive account of the observed change.

3. Trial Design

The registry describes a phase 2, open-label, single-group study of metformin in castration-resistant prostate cancer. Massachusetts General Hospital is listed as the sponsor. The study start is recorded as 2010-09, and primary completion as 2015-01. Its status is listed as terminated. The reason for termination is not reported in the registry record.

The enrollment field lists 21. The arm description identifies metformin as the only arm. Allocation is listed as NA, consistent with a design that does not assign participants between treatment groups. There is no separate untreated or active-control group against which to compare the course of PSA or adverse events. The record does not describe randomization, and this is not a randomized comparison.

The brief summary says participants were already receiving ADT, described there as standard care for prostate cancer. The study thus examined metformin in the context of ongoing ADT, rather than metformin as a replacement for it. The registry does not provide details about the timing or handling of ADT during follow-up, metformin dose, adherence, or treatment changes.

The primary outcome is PSA response, defined as percent change from baseline to 12 weeks, with the assessment time described as approximately 12 weeks. The planned summary is the median with full range. Other registered outcomes concern the number of participants with a PSA response, the relationship between baseline metabolomic profile and PSA response, and maintenance of glycemic control. The outcome value for the primary endpoint and the methods used to analyze the secondary outcomes are not reported in the registry record.

Because this is a single-group study, its design can describe outcomes observed among participants receiving metformin, but it cannot estimate a randomized treatment contrast. The posted status and enrollment information characterize the registry record; they do not explain why the study was terminated or establish how many participants contributed data to each endpoint.

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 PSA (Prostate Specific Antigen) Response, measured over Approximately 12 weeks.

The primary endpoint is the percent change in PSA from baseline to approximately 12 weeks. PSA, or prostate-specific antigen, is a biomarker used in prostate cancer research and care. In general, following a biomarker over time can provide a way to describe how a measured disease-related signal changes during a study. It does not, by itself, establish why that change occurred or whether it represents a change in other clinically important outcomes.

The registry specifies the median and full range as the summary for this endpoint. The median describes the middle of the observed values and is less influenced by unusually high or low values than the arithmetic mean. The full range describes the lowest and highest observed values, but it does not show how values are distributed between those endpoints. Neither statistic is a confidence interval: a range of individual observations and an interval quantifying uncertainty about an estimated parameter answer different questions.

The endpoint compares each participant's PSA with that participant's baseline, then summarizes the changes. The registry does not report the observed median or full range, the number of participants with evaluable measurements, or details about how missing measurements were handled. As a result, the posted record does not permit an account of the endpoint's observed magnitude, variability, or precision.

Even if the endpoint result were available, a change over time in a single-group study would remain a descriptive finding. Without a control group, the study cannot determine whether the change was caused by metformin rather than by other influences, including the underlying disease course or concurrent care.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods are useful only when matched to the study question and the data actually reported. In this study, the principal registered measure is a within-participant change in PSA from baseline to approximately 12 weeks. The planned summary is the median and full range. The registry record does not report an analysis plan or a numerical result for that endpoint, so the methods below explain how such summaries are generally interpreted rather than asserting which additional methods were used in this trial.

What does a median summarize?

In general, the median is the central value after observations are ordered. It can be useful when data are skewed or contain unusually extreme observations, because extreme values have less influence on the median than on the mean. A median alone does not describe all individual responses. The same median can occur with very different distributions, so information about spread and the underlying observations matters.

The full range gives the minimum and maximum observed values. It communicates the span of the data, but it is sensitive to extreme observations and does not show whether most values cluster tightly or are widely dispersed. It is also not an uncertainty interval for the median. A confidence interval, in general, represents uncertainty about an estimated quantity under specified assumptions; the range of observed participant values describes the sample itself. The registry specifies a full range, not a confidence interval.

What does a baseline change measure?

A change from baseline expresses how an outcome differs from an earlier measurement in the same participant. For this endpoint, the registered measure is percent change in PSA from baseline. Such a measure can be informative for describing the course of a biomarker during follow-up. It does not automatically estimate what would have happened under a different treatment or without the intervention.

Repeated measurements within a person are related rather than independent. In general, analyses of repeated observations need to account for that relationship if they use multiple time points. The registry identifies a baseline and an assessment at approximately 12 weeks, but it does not report whether additional measurements were collected or how any repeated measurements were analyzed.

Why does the lack of a comparator matter?

In a controlled comparison, the outcome under one treatment can be contrasted with the outcome under another group, subject to the design and analysis assumptions. A single-group study lacks that concurrent reference. A before-and-after difference may combine effects of treatment with time trends, natural fluctuation, changes in other care, and measurement variability. This is a general limitation of interpreting uncontrolled changes, not a claim about what caused any change in this study.

Open-label status means masking was not used. In general, knowledge of treatment assignment can influence behavior, care, reporting, and some outcome assessments. PSA is a laboratory measurement, but the record does not describe laboratory procedures or whether outcome assessment was otherwise protected from bias. No particular bias can be established from the open-label designation alone.

How should uncertainty and sample-size illustrations be read?

Uncertainty intervals are useful because a point estimate alone does not show how compatible the data are with a range of possible values. Their interpretation depends on the estimator, the analysis, and assumptions about the data. No interval for the primary endpoint is reported in the registry record, so the precision of its estimate cannot be evaluated from the posted information.

Planning illustrations based on an observed effect are hypothetical design calculations, not observed trial outcomes and not observed power. The stated assumptions for the illustrations are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is specified as a fraction of the observed effect. These assumptions describe an illustrative planning exercise, not the single-group design of this study. In particular, an assumed equal allocation between arms does not create a control group in the trial.

Such calculations depend on the effect size chosen and the assumptions used. When based on an observed effect, they can vary substantially with that estimate and should not be treated as proof that a study was adequately powered or as evidence that an effect was detected. The registry record does not report the trial's own sample-size justification, so the planning illustration should not be mistaken for the original design rationale.

What can be said about multiplicity and analysis choices?

When a study examines several outcomes or makes several statistical comparisons, the chance of obtaining an apparently notable result by chance can increase. General approaches to multiplicity include prespecifying a primary outcome, adjusting inference across selected comparisons, or treating secondary findings as exploratory. The record identifies a primary outcome and secondary outcome topics, but it does not report a multiplicity strategy, statistical tests, or analysis details. No adjustment should be assumed.

The registry lists zero analysis entries. This indicates that no analyses are listed in the posted results record; it does not establish that no analysis was conducted outside the registry. The primary endpoint's numerical result and its uncertainty remain unavailable in the registry record. Consequently, no recalculated estimate or comparison can be interpreted from the information recorded.

10. Multiplicity, Missing Data and Analysis Populations

The registry identifies PSA response as the primary outcome and also lists secondary outcomes concerning PSA response counts, baseline metabolomic profile in relation to PSA response, and maintenance of glycemic control. Considering several outcomes can raise multiplicity concerns: in general, the more statistical questions examined, the greater the chance that some result appears notable by chance alone. A prespecified hierarchy or adjustment can help address this issue, but the registry record does not report a multiplicity procedure or the tests used.

The registry lists zero analysis entries. This means the posted record contains no analysis entries; it does not establish that analyses were not performed elsewhere. No primary endpoint estimate, uncertainty interval, or secondary outcome result is reported in the registry record. The statistical analysis plan, model specifications, and decision rules are also not reported.

The record does not describe an analysis population, such as whether analyses included everyone enrolled or only participants with an available follow-up measurement. It does not state the number with evaluable PSA data, explain missing measurements, or report whether missing data were imputed or otherwise handled. These omissions matter because an analysis based on available measurements can differ from one that includes all participants under a stated missing-data method.

Nor does the registry report whether analyses were adjusted for baseline characteristics, how response was defined for the secondary response-count outcome, or how the metabolomic relationship was assessed. Those methods should not be inferred from the outcome names. The absence of detail prevents assessment of whether the analyses followed prespecified methods and how sensitive findings might be to analytic choices.

11. 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 non-serious adverse event terms

The registry safety entry lists 21 participants at risk for serious adverse events and 0 affected in the metformin group. For other adverse events, it lists 21 at risk and 16 affected. These figures should be read as registry counts, not as a comparison of treatment groups: the study has a single group and no control arm.

The record does not identify the serious or other adverse-event terms in the listed results, describe event timing, or provide further detail about severity or relatedness. A count of participants affected does not, by itself, describe how many events occurred or their clinical importance. The record also does not report a comparative risk estimate.

Deaths are entered as null. A null entry is not a reported zero and should not be interpreted as confirmation that no deaths occurred. More broadly, the posted safety counts are limited to the entries recorded in the registry and cannot establish how complete the safety information is. The design also cannot distinguish event rates attributable to metformin from those that might occur in a comparable group without metformin.

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

The study addresses a biologically motivated question about metformin in people with castration-resistant prostate cancer receiving ADT. Its registered primary endpoint is PSA change from baseline to approximately 12 weeks, summarized by the median and full range. The record also identifies secondary questions involving PSA response, metabolomic profile, and glycemic control.

What the posted record does not establish is the observed primary result. The median percent change and full range are not reported, and no confidence interval or other uncertainty measure is given. The absence of a reported endpoint estimate means the registry information cannot show whether PSA increased or decreased, how large any change was, or how much participants' responses varied.

The design also limits causal interpretation. With a single metformin group and no control arm, a change observed over time could not be attributed to metformin alone. Other influences, including disease course and ongoing care, could contribute. Open-label conduct adds another consideration for outcomes or care that may be affected by awareness of treatment, although the record does not establish that such bias occurred.

The safety entries report zero participants affected by serious adverse events and 16 affected by other adverse events among 21 listed at risk in each category. They do not provide event descriptions or a comparison group. The death field is null, so a death count is not reported. These limited entries should not be taken as a complete comparative safety assessment.

Accordingly, the registry record documents a study question, design features, and some safety counts, but not the primary outcome result needed to describe the PSA response. It neither demonstrates benefit nor demonstrates lack of benefit. It also does not establish that metformin caused any change or provide comparative evidence about safety.

13. Why This Trial Matters Statistically

This trial is statistically informative as an example of the difference between a registered clinical question and an interpretable treatment comparison. Its primary endpoint is a within-group change in PSA, while the single-group design has no concurrent control. Descriptive follow-up can characterize what was observed among participants, but a treatment effect requires a comparison that addresses what might have happened under an alternative condition.

The endpoint's specified summary also highlights why a statistic must be interpreted in context. A median can describe the center of a distribution and the full range can describe its extremes, but neither alone quantifies uncertainty about the estimated response. The registry does not report the endpoint values or an uncertainty interval, so even this descriptive interpretation is unavailable from the posted results.

The study's status and limited posted results make transparent reporting especially important. The record lists the study as terminated, but the reason is not reported. It lists safety counts, but not event terms or comparative rates. It identifies secondary outcomes, but does not report their results or explain the analysis methods. These gaps prevent readers from judging the observed outcome, its precision, and the extent to which analytic choices shaped any conclusions.

Finally, planning calculations should be kept distinct from trial findings. Illustrations based on an observed effect and hypothetical two-arm assumptions are not observed power and do not alter the actual single-group design. This distinction matters when translating a statistical planning exercise into claims about what a clinical study showed.

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 principal limitation for interpreting results is that the primary endpoint value is not reported in the registry record. The record therefore does not provide the median percent change in PSA or its full range. It also does not report an uncertainty interval, the number with an evaluable endpoint, or how missing measurements were addressed.

The single-group, open-label design has no concurrent control. As a result, the study cannot distinguish a treatment-associated change from changes related to time, disease course, ADT, other care, or measurement variability. The registry does not report blinding or independent outcome assessment beyond listing masking as none.

Additional design and conduct details are not reported in the registry record, including metformin dose and schedule, adherence, eligibility criteria, participant characteristics, and the reason for termination. These omissions restrict assessment of who the results might describe and how treatment was implemented. The record also does not specify the analysis population, statistical model, missing-data method, or multiplicity approach.

Safety interpretation is limited by the absence of adverse-event terms and details, the lack of a comparison group, and the null entry for deaths. A null death field is not a reported zero. The registry status and enrollment field do not explain the flow of participants through follow-up or how many contributed to each outcome.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Metformin in Castration-Resistant Prostate Cancer — NCT01215032.
• Registry record last updated 2017-05-15; status last verified 2017-04.
• Results first posted 2017-02-27.
• 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

The registry describes a phase 2, open-label, single-group study of metformin for castration-resistant prostate cancer. Participants were described as already receiving ADT. The sponsor is Massachusetts General Hospital; the study start is recorded as 2010-09, primary completion as 2015-01, and status as terminated. Enrollment is listed as 21, and the reason for termination is not reported in the registry record.

The primary outcome is percent change in PSA from baseline to approximately 12 weeks, summarized by median and full range. The primary outcome result, analysis details, and results for the listed secondary outcomes are not reported in the registry record. Safety entries list 0 participants affected by serious adverse events and 16 affected by other adverse events, with 21 at risk in each category. The death entry is null. The single-group design provides no control comparison.