Rectal Spacer for Prostate Cancer Radiation Image Guidance: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Rectal Spacer for Prostate Cancer Radiation Image Guidance (NCT05650021), a na na study. The primary outcome is Difference in 3-dimensional Shifts Between Fiducial Marker and SpaceOAR Vue Alignment.

2. Clinical Question

The clinical question is about the agreement of two image-guidance alignment approaches. It is not a comparison of separate treatment groups or a test of cancer-treatment outcomes. The registry describes a hypothesis that alignment based on the spacer would be similar to alignment based on fiducial markers, but it does not report the outcome estimates needed to assess the observed degree of similarity.

3. Trial Design

This is a completed, single-group device study sponsored by Milton S. Hershey Medical Center. The registry lists an enrollment of 17 and identifies the condition as prostate cancer. The study began on 2024-03-08, reached primary completion on 2025-09-15, and has a completed status. The phase and allocation are listed as NA, and masking is listed as none.

The intervention is SpaceOAR Vue, a radiopaque hydrogel rectal spacer. Participants were treated according to standard of care, including daily CBCT for image-guided radiotherapy. On the first day of treatment, the investigator reviewed the CBCT and aligned the image so that fiducials matched while also assessing alignment of the radiopaque spacer.

The primary analysis uses CBCT images acquired during the first five radiotherapy fractions. Each image was independently aligned using fiducial markers and the spacer. Two readers performed the alignment measurements independently. The comparison is therefore within participants and within images: both methods are applied to the same imaging observations. This design allows the analysis to focus on differences between alignment methods without treating the methods as separate participant groups.

The registry does not describe random assignment, and the study is explicitly single-group. The two methods are measurement approaches within the study, not randomized treatment arms or a like-for-like control group. The record does not report a protocol rationale for enrollment, a sample-size justification, or details of interim analyses. It describes the primary model and the coordinate-specific difference thresholds, but does not report numerical outcome results.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: Participants had SpaceOAR Vue and fiducial markers placed as part of standard-of-care care before undergoing radiotherapy.

Overall Study

Analysis population for the primary outcome, as defined in the registry: All 17 enrolled participants who completed the study were included in the primary analysis. Each participant contributed five CBCT images, with each image independently assessed by two readers. The two alignment methods were evaluated within the same participants; therefore, results are reported as the differences between SpaceOAR Vue and fiducial-marker alignment rather than as separate intervention groups.

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.

Population: All 17 participants enrolled in the study are included in the baseline analysis population. Each participant contributes one set of baseline characteristics.

6. Primary Endpoint

The registry lists the primary outcome as Difference in 3-dimensional Shifts Between Fiducial Marker and SpaceOAR Vue Alignment, measured over First through fifth radiotherapy fractions, occurring within 14 days after the first fraction..

The primary endpoint is the mean difference in three-dimensional shifts between alignment based on fiducial markers and alignment based on SpaceOAR Vue. Differences in the x, y, and z coordinates were measured for each CBCT image. The endpoint is expressed in millimeters, and the primary analysis estimates a separate mean difference for each coordinate.

Both alignment methods were applied to the same images, making the coordinate differences a within-image comparison. The analysis also accounts for repeated measurements within participants and for measurements made by two readers. The registry specifies a linear mixed-effects model with reader as a fixed effect and participant as a random effect.

The stated thresholds of difference are ±2 mm for the x and z coordinates and ±1 mm for the y coordinate. These give coordinate-specific reference points for interpreting the size of the estimated differences. A mean difference describes the average direction and magnitude of the paired measurements; it does not by itself describe how much individual images or participants vary around that average. Confidence intervals are important because they convey uncertainty in the estimated mean. The registry lists a 95% confidence interval as the dispersion measure, but does not report the interval limits or the mean estimates.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous: it is a mean difference in measured coordinate shifts, reported in millimeters. For each CBCT image, alignment using fiducials and alignment using the spacer produce coordinate positions. The analysis compares these positions by calculating the differences in x, y, and z. The registry specifies an estimated mean difference for each coordinate, with a 95% confidence interval as the measure of uncertainty.

Why compare measurements within the same image?

When both alignment methods are applied to the same image, the comparison is paired by design. Image characteristics shared by both measurements are not treated as if they belonged to separate groups. The question is how far the two methods differ on the same observation. This is distinct from comparing average measurements in unrelated sets of participants, where between-person differences could contribute to the comparison.

Pairing can make a method comparison more directly focused on the difference between methods, but it does not guarantee a precise estimate. Precision still depends on the pattern and variability of the paired differences, the amount of usable data, and the analysis model. The registry does not report the observed differences or their variability, so the precision cannot be assessed from the posted results.

Why use a linear mixed-effects model?

The registry describes a linear mixed-effects model, with reader included as a fixed effect and participant included as a random effect. A mixed-effects model can represent the fact that observations are grouped or repeated rather than independent. In this study, repeated images come from participants, and each image is assessed by two readers. Measurements from the same participant may be more alike than measurements from different participants. Accounting for participant-level grouping helps the model reflect that dependence when estimating mean differences and their uncertainty.

A fixed effect estimates a modeled contribution associated with a specified factor, while a random effect represents variation associated with sampled units or groups. Here, reader is treated as a fixed effect, and participant is treated as a random effect. The registry does not describe additional model terms, estimation details, or diagnostic checks. The exact model specification beyond the stated effects is not reported in the registry record.

Repeated observations create a statistical issue because the number of measurements is not the same as the number of independent participants. General statistical reasoning: ignoring within-participant correlation can make standard errors and confidence intervals unreliable. A mixed model is one way to account for this structure, provided its assumptions and specification are suitable for the data. The registry identifies the model structure but does not report diagnostics or sensitivity analyses.

What do the coordinate thresholds contribute?

The registry specifies thresholds of difference of ±2 mm for the x and z coordinates and ±1 mm for the y coordinate. These are stated as thresholds for the differences, not as estimates of the observed differences. In general, a threshold can provide a reference for deciding whether an estimated difference is small enough for a particular purpose. To evaluate a threshold meaningfully, the estimates and their uncertainty need to be considered together. The record does not state a formal decision rule, a named equivalence or non-inferiority framework, or a specific conclusion based on these thresholds.

A mean difference summarizes the average across measurements. General statistical reasoning: averages can mask variation, so the mean alone does not show whether individual paired measurements cluster closely or include larger discrepancies. Confidence intervals address uncertainty about the estimated mean, but do not describe the full distribution of individual differences. The registry does not report additional summaries of that distribution.

What does a 95% confidence interval convey?

In general, a confidence interval gives a range of values compatible with the data and model under the interval procedure. Its width communicates precision: a wider interval leaves more uncertainty about the mean difference, while a narrower interval gives a more precise estimate. A confidence interval is not a guarantee that the true value lies within the particular reported interval, and it does not establish practical similarity without reference to the relevant thresholds and analysis framework. Although the registry specifies 95% confidence intervals, the interval values are not reported in the registry record.

The registry record also does not report the numerical mean differences or p-values. Without estimates and intervals, the reader cannot assess direction, magnitude, or precision, and cannot judge how the observations relate to the coordinate thresholds. No recalculated results are described in the registry record, and no sponsor analysis results are reported there.

10. Multiplicity, Missing Data and Analysis Populations

The primary endpoint involves three coordinate-specific mean differences: x, y, and z. The registry does not report whether or how multiplicity across these coordinate analyses was addressed. It also does not report a hierarchy of primary and secondary tests, an adjustment procedure, or an overall decision rule combining the coordinate results. The coordinate-specific thresholds are stated, but the record does not describe how they were used in formal inference.

The registry identifies all 17 enrolled participants who completed the study as included in the primary analysis. It describes five CBCT images per participant and independent assessments by two readers. These statements identify the intended analysis population and measurement structure, but do not establish that every image and every reader assessment was available. The number of missing images or measurements, reasons for missingness, and any imputation or exclusion procedures are not reported in the registry record.

The record does not report an interim analysis plan, a missing-data sensitivity analysis, or the handling of incomplete coordinate measurements. It also does not describe a broader analysis-population framework beyond the statement that participants who completed the study were included. The available information is therefore insufficient to determine whether all repeated measurements contributed to every coordinate estimate or whether the analysis population varied by outcome.

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

The study addresses a measurement question: whether alignment based on a radiopaque rectal spacer produces coordinate shifts similar to alignment based on fiducial markers when both are applied to the same CBCT images. The registry describes a within-participant comparison and a model intended to account for repeated participant measurements and reader effects.

The registry record does not report the primary mean differences or their 95% confidence intervals. Consequently, it does not show the observed degree of agreement, the precision of the estimates, or whether the coordinate differences were within the stated thresholds of ±2 mm for x and z and ±1 mm for y. The stated hypothesis of similarity is a study rationale, not a reported finding.

Even with numerical mean differences, a mean would not establish that every image or participant had a small difference. An average describes the center of the measured differences, while individual results can vary around it. General statistical reasoning: a complete assessment of agreement may require attention to both the mean difference and the spread of individual differences. The registry record does not report such additional summaries.

This is not a randomized comparison of treatment groups, and it does not compare cancer outcomes between interventions. The spacer and fiducial markers are alternative image-alignment references measured on the same images. The record therefore supports describing a planned method-comparison analysis, not making claims about clinical effectiveness or treatment benefit.

The adverse-event entries list zero at risk as well as zero affected. That reporting cannot establish that no adverse events occurred in the study population. The registry record does not provide enough safety information to draw a conclusion about harms.

12. Why This Trial Matters Statistically

Method-comparison studies raise questions that differ from ordinary comparisons of separate groups. Here, the same participants and CBCT images contribute measurements for both alignment approaches. This paired structure focuses attention on the difference between methods and requires an analysis that recognizes repeated observations rather than treating every measurement as independent.

The registry’s stated linear mixed-effects model reflects two features of the data: repeated images within participants and measurements by two readers. Explicitly describing the participant random effect and reader fixed effect makes the intended handling of these sources of structure more transparent. The model’s appropriateness and the resulting precision still depend on the observed data and assumptions, which are not assessable without reported estimates and diagnostics.

Coordinate-specific thresholds also illustrate why statistical interpretation is not limited to whether an average difference is near zero. The x and z thresholds are ±2 mm, while the y threshold is ±1 mm. A useful interpretation would consider the estimate and its uncertainty in relation to the relevant threshold for each coordinate. Those numerical estimates and intervals are not reported in the registry record.

The study also highlights the difference between a completed study and a complete report of results. The record lists completion and describes who was included in the primary analysis, but does not provide the primary outcome estimates. Without those results, readers cannot assess the magnitude or precision of the method differences. The measurement plan can be understood, but the empirical answer to the stated comparison remains unavailable in the registry record.

13. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The main limitation of the registry record for interpreting this study is that it does not report numerical primary outcome results. The mean differences and 95% confidence intervals are absent for all three coordinates. This prevents assessment of effect direction, magnitude, precision, and relation to the stated thresholds.

The record describes the study as single-group, with no masking and allocation listed as NA. Both alignment approaches were assessed within participants, which is appropriate to the stated measurement question, but this is not a randomized comparison of separate groups. The registry does not report a formal equivalence or non-inferiority framework, nor does it state how the coordinate thresholds were incorporated into a decision rule.

The analysis population is described as all 17 enrolled participants who completed the study. The record does not report whether any images or reader measurements were missing, whether all planned assessments were included, or how incomplete data were handled. Details of model diagnostics, sensitivity analyses, and multiplicity handling are also not reported in the registry record.

The record states that two readers independently performed alignments, but does not report reader-specific results or a separate measure of inter-reader agreement. It therefore does not allow assessment of how much measurements varied between readers. Nor does it report the distribution of individual paired differences, which would complement mean differences when describing agreement.

Finally, the listed adverse-event counts show zero at risk and zero affected. Without a nonzero at-risk population or further safety details, those entries do not support a conclusion about adverse events among enrolled participants.

14. Related Tutorials

15. Related Statistical Calculators

16. Sources

• ClinicalTrials.gov: Rectal Spacer for Prostate Cancer Radiation Image Guidance — NCT05650021.
• Registry record last updated 2026-10-27; status last verified 2026-10.
• Results first posted 2026-10-27.
• 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.

17. Record Summary

NCT05650021 is a completed, single-group device study sponsored by Milton S. Hershey Medical Center, with an enrollment of 17. It evaluates SpaceOAR Vue alignment against fiducial-marker alignment on CBCT images acquired during the first five radiotherapy fractions. Two readers independently assessed the images, and the primary analysis was described as a linear mixed-effects model with reader as a fixed effect and participant as a random effect.

The primary outcome is the mean difference in x, y, and z coordinate shifts, with 95% confidence intervals. The stated difference thresholds are ±2 mm for x and z and ±1 mm for y. The registry record describes the analysis plan and completed-study population, but does not report numerical endpoint estimates or confidence intervals. Missing-data handling, multiplicity procedures, and a formal decision rule for interpreting the thresholds are not reported in the registry record.