Evaluation of Response to the Neoadjuvant Chemotherapy for Advanced Ovarian Cancer by Multimodal Functional Imaging: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Evaluation of Response to the Neoadjuvant Chemotherapy for Advanced Ovarian Cancer by Multimodal Functional Imaging (NCT02792959), a na na study. The primary outcome is Inter-rater Reliability of Positron Emission Tomography (PET) Maximum Standardized Uptake Value (SUVmax) - Individual Intra-class Correlation Coefficients by Location.

2. Clinical Question

The study also sets out questions about whether changes in PET parameters and apparent diffusion coefficient relate to CA125, and whether imaging changes differ by response under RECIST V1.1. These are secondary questions, not randomized comparisons of treatments. The brief summary expresses an interest in early assessment of treatment efficacy, including with and without an antiangiogenic agent, but the registry record does not describe separate treatment groups or report results for such a comparison. The primary reliability question is about consistency between observers, rather than whether an imaging measure predicts a clinical outcome or establishes treatment benefit. Numerical estimates for the primary outcome are not reported in the registry record.

3. Trial Design

The registry record describes a completed, single-group pilot study of multimodal functional imaging in advanced ovarian cancer. The listed enrollment is 11. The study is sponsored by Institut Bergonié. Its phase and allocation are both listed as NA, and masking is listed as none. The record gives a start of 2014-08 and primary completion of 2017-01.

The imaging procedures are described at two points in the treatment course: before the second course of neoadjuvant chemotherapy and before surgery. The listed intervention names refer to functional imaging at these times, and the arm description calls the study a pilot evaluation using fusion MRI and FDG-PET-CT. The two imaging assessments should not be interpreted as two arms: the record identifies a single group, not a like-for-like control comparison.

The primary outcome concerns observer reliability for PET SUVmax at the pre-therapeutic assessment. It specifies individual intra-class correlation coefficients by location and absolute agreement between observers. The description says PET images are interpreted independently by 2 radiologists and 2 nuclear medicine specialists, blinded to one another’s results. Although the record describes blinded interpretation between observers, the overall masking field is listed as none; these are distinct statements about observer procedures and the study-level masking classification.

Secondary outcomes include correlations with CA125, PET parameter changes after one and four cycles of chemotherapy, apparent diffusion coefficient changes, complete surgery resection, and descriptive PET parameter changes by RECIST V1.1 response. The record lists 56 outcome measures and 12 analyses. A detailed protocol, eligibility criteria, chemotherapy regimens, and a rationale for the enrollment target are not reported in the registry 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.

Before assignment: The first steering committee (which met on 30th June, 2015) decided that one patient was not assessable for primary endpoint and had to be replaced. This patient deceased after the first line chemotherapy.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Eligible and evaluable population

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 Inter-rater Reliability of Positron Emission Tomography (PET) Maximum Standardized Uptake Value (SUVmax) - Individual Intra-class Correlation Coefficients by Location, measured over Pre-therapeutic.

The primary endpoint concerns inter-rater reliability of PET maximum standardized uptake value, or SUVmax, measured by location before treatment. The registry record specifies individual intra-class correlation coefficients using absolute agreement. It lists the parameter as a geometric mean and the dispersion as a 95% confidence interval.

Reliability asks whether observers produce sufficiently consistent measurements when assessing the same imaging information. Absolute agreement is relevant when the numerical values themselves are intended to be comparable: observers may rank locations similarly yet still differ systematically in their measurements. An intra-class correlation coefficient is a way to describe how much of the observed variation reflects differences between measured subjects or locations rather than measurement inconsistency. This is a general explanation of the method, not a reported study finding.

Reliability matters because measurement variation can complicate interpretation of imaging changes over time or comparisons with other measures. However, a reliability estimate does not by itself establish that SUVmax accurately measures tumor response, predicts outcome, or changes treatment decisions. The registry record describes the endpoint and its intended interpretation, but numerical estimates by location and their confidence intervals are not reported in the registry record. Consequently, the precision and magnitude of the observed inter-rater reliability cannot be assessed from the record’s posted outcome information.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary method named in the registry record is an intra-class correlation coefficient (ICC) for PET SUVmax, evaluated by location. The endpoint description specifies absolute agreement between observers, and the results are characterized as individual ICCs by location. A geometric mean and a 95% confidence interval are listed for the primary parameter and its dispersion. Numerical ICC estimates and confidence intervals are not reported in the registry record.

What does an intra-class correlation measure?

As a general statistical explanation, an ICC describes reliability by relating the differences between the objects being measured to the variation in repeated measurements. Depending on the ICC formulation, the question can concern consistency in ranking, absolute agreement, or a particular combination of observer and measurement effects. The distinction matters: observers can rank locations in a similar order while assigning values that differ systematically. The registry record explicitly describes absolute agreement, so the intended question concerns agreement in the measured values, not merely similarity of rankings.

An ICC is not a direct measure of accuracy. Agreement among observers does not show that their shared measurements match a biological truth, nor does it show that SUVmax is a valid surrogate for a clinical outcome. Reliability is one property of a measurement process; validity and clinical usefulness are separate questions. The registry record does not report an external reference standard or a clinical decision threshold for this endpoint.

Why report a confidence interval?

In general, a confidence interval expresses uncertainty around an estimated parameter under the assumptions of the statistical method. A narrower interval indicates greater precision than a wider interval, while the interval’s location describes the range of values compatible with the estimate under that method. It does not mean that the parameter itself varies randomly across repeated interpretations. The record specifies 95% confidence intervals, but without the actual estimates and interval limits, neither the degree of precision nor the plausible range of reliability can be evaluated.

The primary endpoint lists a geometric mean. In general, geometric means are useful for summarizing positive quantities when effects or values are naturally considered on a multiplicative scale. Their interpretation differs from an arithmetic average, which summarizes values on an additive scale. The record does not provide the individual location-specific estimates or further details about how locations contribute to the geometric mean. The calculation procedure is therefore not reported in the registry record.

How do the secondary methods fit the questions?

The secondary outcomes include Spearman correlation analyses between CA125 and relative changes in PET parameters after one and four cycles of chemotherapy, and between relative change in apparent diffusion coefficient and CA125. In general, Spearman correlation describes the strength and direction of a monotonic association using ranks; it does not establish causation and need not imply a linear relationship. A correlation result would need its estimate, uncertainty, and analysis count to be interpreted. Those numerical results are not reported in the registry record.

The listed PET measures include SUVmax, MTV, SUVN, SUL, and TLG. The record also names descriptive statistics for relative PET parameter changes by response under RECIST V1.1, as well as complete surgery resection. Descriptive summaries can characterize observed measurements but, without a comparator and a specified inferential analysis, do not by themselves establish that one treatment caused a difference. The record’s single-group design is important when interpreting any within-course change.

More generally, repeated imaging creates measurements at different times, and a change can reflect biological change, measurement variation, or both. Reliability assessment can help characterize one source of variation, but does not remove all uncertainty from longitudinal comparisons. The registry record does not report the detailed statistical models, rules for defining relative change, handling of repeated measurements, or criteria for response-group summaries.

What do planning calculations mean?

The stated assumptions for sample-size illustrations are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the target effect expressed as a stated fraction of the observed effect. For time-to-event calculations, the stated quantity is the total number of events; other calculations are described as participants per arm. These are planning conventions, not observed results or observed power. The study itself is identified as single-group, with allocation listed as NA, so a two-arm illustration should not be read as a description of the actual study design. Numerical illustration values and the effect estimates on which they depend are not reported in the registry record.

The complete statistical analysis plan, including model specifications and any interim analyses, is not reported in the registry record. The methods named for the endpoints indicate the intended types of analyses, but do not supply enough detail to reproduce them or verify how all outcome measures were handled.

10. Multiplicity, Missing Data and Analysis Populations

The registry record lists 56 outcome measures and 12 analyses. The outcomes span observer reliability, several PET parameters, CA125 associations, apparent diffusion coefficient, surgery, and response under RECIST V1.1. When many outcomes are examined, the chance of at least one apparently notable result can increase. This is a general statistical concern, not evidence that any particular reported result is spurious.

Whether the analyses were organized into a prespecified hierarchy, and whether any multiplicity adjustment was used, is not reported in the registry record. The primary outcome is identified, but the complete statistical analysis plan, rules for interpreting secondary outcomes, and any planned control of the overall false-positive rate are not reported in the registry record. Numerical results for the listed outcomes are also not reported in the registry record, so their statistical significance or uncertainty cannot be assessed.

The primary population is described as eligible and evaluable, but its analysis count is not reported in the registry record. The record lists enrollment as 11, without specifying how many participants contributed to each endpoint or analysis. The reasons for any missing imaging, measurements, or outcome data, and the methods used to address missing data, are not reported in the registry record. It is therefore not possible to determine whether the primary and secondary analyses used the same participants or how missing observations could affect the estimates.

In general, missingness can matter when the chance that an observation is absent is related to the participant’s condition, measurement, or study progress. The appropriate response depends on the amount and pattern of missingness and the planned estimand and analysis. No missing-data assumptions or sensitivity analyses are reported in the registry record. These omissions limit assessment of both the analysis populations and the robustness of any estimates.

11. 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: anemia (9.1%), colonic obstruction (9.1%), vomiting (9.1%), sepsis (9.1%).

The registry record lists 11 participants at risk for serious adverse events and 3 affected, and 11 at risk for other adverse events and 11 affected. It also lists 1 death. These counts describe the recorded safety information for the functional imaging group; they do not establish that an event was caused by imaging or chemotherapy. The relationship of events to a particular intervention, the timing of the events, and the criteria used to classify them are not reported in the registry record.

The listed serious-event terms are anemia, colonic obstruction, vomiting, and sepsis, each reported as 9.1%. These are event terms with their recorded percentages, not necessarily mutually exclusive categories. A participant may have experienced more than one event, so the term percentages should not be added to infer a total number of affected participants.

The counts should be read with the registry’s group and at-risk denominators in mind. They are not comparative risks: the study has a single group and no randomized control arm. The record does not report a detailed event listing, severity grades beyond the serious-event classification, exposure time, or causal attribution. The safety counts therefore characterize what is recorded, without supporting a comparative safety conclusion.

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

The registry record identifies a pilot investigation of multimodal functional imaging in advanced ovarian cancer. Its primary statistical question is whether different observers agree when measuring PET SUVmax by location. It also lists questions about imaging changes, CA125, apparent diffusion coefficient, response classification, and surgery. These aims indicate what the study was designed to examine, but they are not themselves results.

The posted outcome information does not include numerical estimates for the primary ICCs or their 95% confidence intervals. It also does not report numerical findings for the listed secondary outcomes. As a result, the magnitude and precision of observer agreement, the strength of any correlation, and the pattern of imaging change cannot be determined from the registry record. A defined endpoint should not be mistaken for a finding on that endpoint.

The single-group design is also important. Measurements made at different points in a treatment course can describe change over time, but without a randomized comparison they cannot isolate the causal effect of an imaging approach or distinguish treatment effects from other changes over time. The record does not describe a treatment comparison with and without an antiangiogenic agent, despite mentioning that interest in the brief summary. No conclusion about comparative treatment efficacy follows from the design details recorded.

The safety counts provide a description of recorded events in the group, but do not show comparative risks or establish event causation. Likewise, a reliability estimate, if available, would address consistency of measurement and not prove that the measure is clinically valid or useful. The listed enrollment is 11, while endpoint-specific analysis counts are not reported in the registry record.

In general, a study’s interpretation depends on its estimates, uncertainty, analysis population, and design as well as its stated aims. Those numerical primary and secondary results and several analysis details are not reported in the registry record. The appropriate reading is therefore limited to the stated study questions, design, and recorded safety counts, rather than a conclusion about imaging performance or treatment efficacy.

13. Why This Trial Matters Statistically

Reliability is a foundational statistical issue in imaging research. If different observers do not measure a feature consistently, apparent differences between locations or across assessment times may partly reflect measurement variation. Evaluating observer agreement can therefore help readers understand whether an imaging measure is reproducible enough to support further investigation. This is a general reason the endpoint is relevant; the registry record does not report the numerical reliability results needed to judge the degree of agreement in this study.

The study’s multimodal focus is also statistically important. It considers PET measurements alongside MRI-related functional imaging and associations with CA125 and apparent diffusion coefficient. Combining or comparing measures raises questions about scale, repeatability, timing, and the interpretation of multiple outcomes. The registry record lists these questions, while the results needed to quantify the relationships are not reported in the registry record.

The pilot, single-group design places the focus on feasibility and measurement questions rather than randomized treatment contrasts. This distinction helps prevent a common interpretive error: observations before and during treatment do not alone show what would have happened under an alternative treatment strategy. The record does not report a like-for-like control comparison, and the treatment comparison mentioned in the brief summary is not described as a separate randomized group.

There is also a useful distinction between describing a statistical target and showing evidence about it. The record names ICCs, confidence intervals, Spearman correlations, and descriptive summaries. Those methods relate to meaningful questions, but their interpretation depends on the actual estimates, uncertainty, denominators, and analysis details. Where these are not reported in the registry record, statistical conclusions should remain limited.

The record is consequently informative about which measurement and association questions were registered, and it documents safety counts, but it does not provide the primary numerical outcome estimates. That separation between study intent and reported evidence is central to reading small exploratory studies carefully.

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 main trial-specific limitation for interpreting the registered primary question is that numerical ICC estimates and 95% confidence intervals are not reported in the registry record. Although the endpoint is specified by location and absolute agreement, the location-specific results and the analysis details needed to interpret the stated geometric mean are absent. The primary outcome’s precision and magnitude therefore cannot be evaluated.

The study is recorded as single-group, with allocation listed as NA and masking as none. There is no randomized control comparison in the registry record. Imaging time points before the second course and before surgery are assessments within the study, not independent treatment groups. Thus, any within-course changes would not by themselves provide a causal comparison between treatment strategies. The brief summary refers to interest in treatment with and without an antiangiogenic agent, but separate arms and comparative results are not reported in the registry record.

The listed enrollment is 11, and the primary population is described as eligible and evaluable. The number actually contributing to each outcome analysis is not reported in the registry record. Eligibility criteria, detailed chemotherapy regimens, analysis denominators, missing-data patterns, and methods for handling missing measurements are not reported in the registry record. These gaps make it difficult to assess the representativeness and completeness of each analysis.

The registry lists 56 outcome measures and 12 analyses, while multiplicity procedures and the full statistical analysis plan are not reported in the registry record. The numerical results for the secondary analyses are also not reported. Finally, details such as how imaging locations were selected and how the individual ICC estimates were combined are not reported in the registry record. These limitations restrict reproducibility and the strength of any broader interpretation.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Evaluation of Response to the Neoadjuvant Chemotherapy for Advanced Ovarian Cancer by Multimodal Functional Imaging — NCT02792959.
• Registry record last updated 2021-04-21; status last verified 2021-03.
• Results first posted 2021-04-21.
• 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 record describes a completed pilot study of multimodal functional imaging for advanced ovarian cancer. It lists enrollment as 11, a single-group design, allocation as NA, and masking as none. Functional imaging is assessed before the second chemotherapy course and before surgery, using fusion MRI and FDG-PET-CT.

The primary outcome concerns observer agreement for PET SUVmax by location at the pre-therapeutic assessment, with absolute agreement, a geometric mean, and 95% confidence intervals specified. The record also lists secondary questions involving CA125, PET parameters, apparent diffusion coefficient, RECIST V1.1 response, and surgery. Numerical estimates for the primary and secondary outcomes, endpoint-specific analysis counts, and detailed statistical procedures are not reported in the registry record.

Safety information lists 3 participants with serious adverse events, 11 with other adverse events, and 1 death among the recorded group. These figures are descriptive and do not establish causation or comparative risk. The record does not report a randomized treatment comparison.