A Study To Evaluate Hypofractionated Proton Therapy Or IMRT For Recurrent, Oligometastatic Prostate Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study To Evaluate Hypofractionated Proton Therapy Or IMRT For Recurrent, Oligometastatic Prostate Cancer (NCT04190446), a randomized phase 2 study. The primary outcome is Incidence of Late Grade 3 or Higher GI and/or GU Toxicity.

2. Clinical Question

3. Trial Design

The registry identifies this study as a randomized, parallel phase II trial. It lists no masking. Mayo Clinic is the sponsor, and the listed enrollment is 84. The record describes the study as evaluating side effects of hypofractionated proton therapy or IMRT in recurrent or oligometastatic prostate cancer after primary localized treatment.

The two arms differ in fractionation schedule. Arm I, moderate hypofractionation, involves IMRT or PBRT 5 days a week over 3 weeks. Arm II, conventional fractionation, involves IMRT or PBRT 5 days a week over 5 weeks. Because both arms permit either radiation modality, the posted primary results are best read as a comparison by treatment schedule, not as a direct comparison of proton therapy with IMRT. The registry record does not report how the radiation modality was selected within each arm.

The study description also lists quality-of-life assessment, questionnaires, imaging, and biospecimen collection. PET scans, CT scans, MRI, and blood sample collection are described as occurring throughout the study. Those procedures provide context for the planned assessments, but their results are not included in the primary outcome figures.

The primary outcome is late grade 3 or higher gastrointestinal and/or genitourinary toxicity, assessed at 24 months. The record lists acute and late adverse events, grade 3 or higher GI or GU adverse events by treatment schedule, and adverse events among its secondary outcomes. It reports enrollment of 84, while the primary outcome groups contain 53 participants in Arm I and 25 in Arm II. The reasons that those group counts differ from the listed enrollment are not reported in the registry record. The study status is listed as active, not recruiting.

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 Incidence of Late Grade 3 or Higher GI and/or GU Toxicity, measured over 24 months.

The primary endpoint measures the percent of participants who experience a late grade 3 or higher gastrointestinal (GI) and/or genitourinary (GU) adverse event judged possibly, probably, or definitely related to radiation therapy. Late is defined as at least 90 days after radiation therapy starts, and the assessment time frame is 24 months. Events are graded using Common Terminology Criteria for Adverse Events (CTCAE) version 4.0.

This endpoint focuses on serious toxicity in the GI or GU systems and distinguishes events by both severity and timing. Reporting a percentage makes the result interpretable as the observed proportion of participants with a qualifying event in each treatment schedule. The combined GI and/or GU definition means an event in either system can meet the endpoint; it does not require an event in both.

The registry describes this outcome as a safety measure. It does not report the endpoint population beyond the arm-level participant counts, nor does it explain how participants without a recorded assessment were handled. Those details matter when interpreting percentages because the denominator and completeness of follow-up determine which participants contribute to an estimate. The registry record also does not report whether the 24-month assessment was based on a particular visit window or another timing rule.

7. Results Posted in the Registry

• Arm II (Conventional Fractionation): 0.0% (0/25) versus 5.7% (3/53) on control; risk difference -5.7 percentage points (95% CI -11.9 to 0.6); unadjusted two-proportion z-test p = 0.225.

8. Statistical Interpretation

9. Statistical Methods Explained

General: A binary outcome records whether each participant experienced a defined event or did not. The primary endpoint is reported as a percentage with an event, so its arm-level value summarizes the observed frequency of qualifying late toxicity. The event definition is specific: grade 3 or higher GI and/or GU toxicity, at least 90 days after radiation therapy starts, and assessed as possibly, probably, or definitely related to radiation therapy.

The registry reports 3 events among 53 participants in Arm I and 0 among 25 participants in Arm II. It reports the corresponding percentages as 5.7% and 0.0%. The denominators attached to these results are important: they describe the participants counted for the primary outcome and are not the same as the total enrollment listed for the trial. Why the endpoint denominators differ from enrollment is not reported in the registry record.

General: A risk difference expresses the difference between two event proportions, with the direction depending on which group is subtracted from which. The posted comparison is Arm II compared with Arm I and gives a risk difference of -0.057. In this direction, a negative estimate corresponds to a lower observed proportion in Arm II. Its confidence interval is -0.119 to 0.006. Because the interval includes zero, the data do not precisely distinguish a difference from no difference under this measure. No additional risk-difference calculation is needed to interpret the posted estimate.

General: A confidence interval represents uncertainty around an estimate under the assumptions of the method used. It is not a range that guarantees the true value lies inside it, nor is it a prediction interval for individual outcomes. The registry reports 90% confidence intervals for the arm percentages and for the risk difference. The intervals communicate that these estimates are not exact measures of the underlying event frequency or contrast.

General: A p-value describes how unusual data at least as discrepant as those observed would be under a specified null model and its assumptions. It is not the probability that the null model is true, and it does not measure the size or clinical importance of a difference. The registry gives an unadjusted p-value of 0.225 for the comparison. The statistical analysis plan and the exact test or model used for that comparison are not reported in the registry record, so the p-value should be read as posted rather than attributed to an assumed method.

The trial is randomized and parallel, but the registry does not provide details of allocation procedures, stratification, or an adjusted analysis. The record lists no analysis records. That absence means there is no posted account of additional modeled comparisons or covariate adjustment; it does not justify inferring which methods were or were not used outside the record. The analysis population for the primary outcome is also not specified beyond the group counts.

The registry lists planning assumptions for sample-size illustrations: a two-sided significance level of 0.05, equal allocation, a normal approximation, and an effect to detect defined as a stated proportion of the observed effect. Such planning illustrations are hypothetical calculations based on assumptions, not observed power and not evidence about the trial's realized ability to detect an effect. Numerical sample-size illustration results are not reported in the registry record. The trial's own sample-size justification is not reported in the registry record either.

General: Over-dispersion occurs when observed variability exceeds what a statistical model assumes. It can make uncertainty estimates too narrow if the model does not account for that extra variation. The posted primary outcome is binary and the record does not describe an over-dispersion assessment or a model for repeated observations. It would be inappropriate to assert that over-dispersion was present or that a correction was used.

For a binary endpoint, small event counts can make estimates uncertain, especially when one arm has no observed events. The Arm II point estimate is 0.0%, but its reported 90% confidence interval reaches 10.1%. This interval is a reminder that a zero count in the observed group does not establish zero underlying risk. The record does not state which confidence-interval method was used, so the method should not be guessed.

The registry describes adverse events as graded with CTCAE version 4.0 for the primary endpoint. It does not report details of outcome adjudication, assessor masking, missing-data imputation, or sensitivity analyses. Those methodological details would help readers judge how event classification and incomplete follow-up may affect the estimates, but they cannot be reconstructed from the posted record.

10. Confidence Intervals and Precision

The arm-level estimates are 5.7% for moderate hypofractionation and 0.0% for conventional fractionation. Their reported 90% confidence intervals are 1.6% to 14.0% and 0.0% to 10.1%, respectively. The risk difference for Arm II compared with Arm I is -0.057, with a 90% confidence interval from -0.119 to 0.006. These intervals show that the estimates have meaningful uncertainty; the point estimates alone do not capture that uncertainty.

General: A confidence interval gives a range of values compatible with the data and the interval procedure, under its assumptions. Wider intervals indicate less precision than narrower intervals on the same scale. The risk-difference interval includes zero, so the posted results do not clearly separate a difference from no difference. The interval also extends to values on either side of zero, making the direction uncertain.

The observed event counts are 3 of 53 in Arm I and 0 of 25 in Arm II. The smaller number of participants in Arm II and its zero event count are relevant context for its interval: a zero observed percentage can coexist with uncertainty about the underlying event frequency. The registry provides no precision target or minimum clinically important difference, so the intervals cannot be judged against a registry-defined threshold for adequate precision.

The reported intervals are 90% intervals, not 95% intervals. No alternative interval level or method is reported. The registry record does not report follow-up completeness, missing-outcome handling, or an analysis population definition that could further clarify how precisely the target population is represented by these estimates.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 9 outcome measures and 4 secondary outcomes, including acute and late adverse events, grade 3 or higher GI or GU adverse events by treatment schedule, and adverse events. The record lists no analysis records. It does not report a multiplicity adjustment, a hierarchy among outcomes, or a plan identifying which comparisons would be treated as confirmatory. Without that information, the primary outcome can be described as the designated primary measure, but the approach to interpreting multiple outcomes and comparisons is not reported in the registry record.

The primary endpoint description specifies a 24-month time frame and the event criteria, but it does not state how incomplete follow-up or missing endpoint assessments were handled. The primary outcome population field is blank. The posted arm counts are 53 for Arm I and 25 for Arm II, while trial enrollment is listed as 84. The registry record does not explain the difference between those counts, and no reason for exclusion or non-assessment should be inferred.

The registry does not report whether the primary comparison used an intention-to-treat population, a per-protocol population, or another analysis set. It also does not report imputation, censoring rules, sensitivity analyses, or procedures for participants who withdrew or lacked a 24-month assessment. These omissions limit assessment of whether the estimates could be affected by missing outcome data or by choices about which participants were included.

Planning illustrations are described as using a two-sided significance level of 0.05, equal allocation, a normal approximation, and a stated proportion of the observed effect. The registry does not report the trial's own sample-size justification or numerical illustration results. The planning assumptions should not be mistaken for a documented analysis plan, and they do not establish observed power.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: small intestinal obstruction (3.3%).

The serious adverse event summary reports 0 affected participants among 54 at risk in Arm I and 1 among 30 at risk in Arm II. It lists 1 death in Arm I and 0 in Arm II. These are serious adverse event summary counts, not the primary endpoint counts: the primary endpoint concerns late grade 3 or higher GI and/or GU toxicity and uses different arm denominators.

The registry also reports other adverse events in 53 of 54 participants in Arm I and 28 of 30 in Arm II. Those figures describe the adverse-event summary and should not be interpreted as rates of the primary endpoint. The record lists Small intestinal obstruction at 3.3% as a leading serious adverse event term. The registry does not provide further context for that percentage in the listed term summary, so its denominator and relationship to the arm-specific counts should not be assumed.

General: Serious adverse event counts and event percentages answer different questions from a severity-specific, treatment-related primary toxicity endpoint. The serious event summary does not by itself establish whether an event met the primary endpoint definition, whether it was late, or whether it was judged related to radiation therapy. The posted counts are descriptive; the registry record does not report a statistical comparison for these serious adverse event figures.

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

The posted data show 3 qualifying late grade 3 or higher GI and/or GU toxicity events among 53 participants in Arm I and 0 among 25 in Arm II. The reported percentages are 5.7% and 0.0%. The comparison estimate is -0.057 for Arm II relative to Arm I, with a 90% confidence interval from -0.119 to 0.006 and an unadjusted p-value of 0.225. These results describe the observed primary endpoint in the posted groups.

The results do not establish that conventional fractionation reduces toxicity, nor do they establish that the schedules are equivalent. The confidence interval includes zero and the estimate remains uncertain. In particular, no observed events in Arm II is not evidence that the underlying event rate must be zero; the reported interval for that arm extends to 10.1%.

The comparison is by fractionation schedule, not by radiation modality. Both arms include IMRT or PBRT, and the registry record does not report modality-specific primary results. The findings therefore cannot isolate a difference between proton therapy and IMRT.

The primary endpoint is about late toxicity and does not answer questions about cancer control, quality of life, or other outcomes. The record lists secondary outcomes but does not provide results for them in the registry record reported for this analysis. No conclusion about those outcomes should be drawn from the primary toxicity comparison.

Finally, enrollment and the primary outcome group counts are not identical, and the reason is not reported in the registry record. The record also does not describe missing-data handling, the analysis population, or detailed statistical methods. Those limitations matter when considering how well the posted estimates represent all enrolled participants and how robust they are to analysis choices.

14. Why This Trial Matters Statistically

This trial illustrates why the endpoint definition and the comparison being made need to be read together. The primary outcome is not a general count of all adverse events: it is restricted to late grade 3 or higher GI and/or GU events considered related to radiation therapy. The reported percentages therefore answer a narrower safety question than the overall adverse-event summaries.

The arms compare schedules, with moderate hypofractionation over 3 weeks and conventional fractionation over 5 weeks. Both schedules can use IMRT or PBRT. This distinction matters statistically because the arm contrast does not isolate the radiation modality. Treating it as a direct proton-versus-IMRT comparison would misstate the trial question.

The event counts are small, particularly the zero observed events in Arm II. The accompanying confidence intervals make clear that a point estimate of 0.0% does not imply that the underlying event risk is known to be zero. The risk-difference interval also includes zero, so the data leave uncertainty about both the size and direction of the contrast.

The registry lists total enrollment of 84 but reports primary outcome groups of 53 and 25. Since the reason for this difference and the missing-data approach are not reported, readers cannot determine from the record alone whether incomplete follow-up or analysis-population choices affected the endpoint estimates.

More broadly, the record distinguishes a randomized phase II design from a complete report of statistical methods. It gives the primary outcome results but does not report the analysis plan, multiplicity approach, or trial-specific sample-size justification. Recognizing both the reported evidence and those unanswered questions supports a measured interpretation without turning a single safety comparison into a broader claim.

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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The primary endpoint results are based on 53 participants in Arm I and 25 in Arm II, while listed enrollment is 84. The registry record does not report why these counts differ, whether all enrolled participants reached the 24-month assessment, or how missing assessments were handled. The primary outcome population is not otherwise specified.

The event counts are limited: 3 in Arm I and 0 in Arm II. The arm-level confidence intervals are correspondingly important to interpretation, and the zero observed events in Arm II should not be mistaken for a precisely established absence of risk. The risk-difference interval includes zero, leaving uncertainty about the direction and size of the schedule contrast.

The trial allows either IMRT or PBRT in both arms. The posted primary outcome is grouped by schedule, and no modality-specific outcome results are reported in the registry record. The schedule comparison therefore cannot be used to separate outcomes by radiation modality.

The registry record does not report the statistical test or model used for the comparison, an adjusted analysis, the analysis population, allocation details beyond randomization, or a missing-data strategy. It also does not report multiplicity procedures, sensitivity analyses, or the trial's own sample-size justification. These gaps prevent a fuller evaluation of analytic choices and the impact of incomplete outcome data.

The record lists adverse event summaries, but these do not replace the primary endpoint definition or provide a full account of event timing and attribution. The serious adverse event term summary identifies Small intestinal obstruction at 3.3%, but further denominator context is not reported in the registry record. As a result, the term percentage should be interpreted cautiously and not conflated with the primary toxicity percentages.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study To Evaluate Hypofractionated Proton Therapy Or IMRT For Recurrent, Oligometastatic Prostate Cancer — NCT04190446.
• Registry record last updated 2026-10-20; status last verified 2026-09.
• Results first posted 2026-10-20.
• 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

The registry describes NCT04190446, titled A Study To Evaluate Hypofractionated Proton Therapy Or IMRT For Recurrent, Oligometastatic Prostate Cancer, as a randomized, parallel phase II trial with no masking. Mayo Clinic is the sponsor, and enrollment is listed as 84. The record status is active, not recruiting.

Arm I uses moderate hypofractionation over 3 weeks; Arm II uses conventional fractionation over 5 weeks. Both arms may receive IMRT or PBRT. The primary outcome is the percent of participants with late grade 3 or higher GI and/or GU toxicity at 24 months. The posted percentages are 5.7% for Arm I and 0.0% for Arm II, with a reported risk difference of -0.057 and an unadjusted p-value of 0.225. The comparison confidence interval includes zero.

The record lists 4 secondary outcomes and 9 outcome measures. Details of the primary analysis population, missing-data handling, multiplicity procedures, and trial-specific sample-size justification are not reported in the registry record.