IMAging With Opto-acoustics to downgradE BI-RADS claSsificaTion Relative tO Other Diagnostic Methodologies (MAESTRO): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for IMAging With Opto-acoustics to downgradE BI-RADS claSsificaTion Relative tO Other Diagnostic Methodologies (MAESTRO) (NCT02364388), a non-randomized na study. The primary outcome is OA/US Specificity (Downgrade (%) for BI-RADS 4A & 4B) of Benign Masses.

2. Clinical Question

BI-RADS is the Breast Imaging Reporting and Data System established by the American College of Radiology. The registry describes scores as ranging from 0 to 6, with higher scores indicating a greater probability of malignancy, and states that a score of 4 or more indicates the need for biopsy. In this study's endpoint description, the relevant question is whether OA/US imaging can support a benign or probably benign classification for masses that had been considered suspicious on the basis of the earlier assessment.

This is a diagnostic classification question, not a randomized comparison of clinical outcomes. The record describes a post-market, non-randomized multi-center surveillance study. It does not establish that the two listed imaging approaches were assigned under comparable conditions, and it does not report a primary outcome estimate that can be used to quantify the observed downgrading percentage.

3. Trial Design

MAESTRO, registered as NCT02364388, is described as a post-market, non-randomized multi-center surveillance study of the Imagio breast imaging system. The system is designed to collect opto-acoustic images together with diagnostic ultrasound images. The listed devices are Imagio IUS gray-scale ultrasound and Imagio OA/US, described as opto-acoustic plus gray-scale ultrasound.

The registry records the study as completed. The recorded start date is 2015-03-27, and the primary completion date is 2016-02-15. Enrollment is listed as 217. The phase is listed as NA. The study model is sequential, allocation is non-randomized, and masking is none.

Those design features matter when interpreting comparisons between the listed imaging approaches. Because allocation was not randomized, differences between arms could reflect how participants or assessments entered the study as well as differences associated with the imaging method. The registry specifically cautions that the arms are not a like-for-like control comparison. Accordingly, the design should not be read as a randomized efficacy trial, and contrasts between arms should not be attributed to the device alone.

The primary outcome description places the initial clinical decision and CDU imaging at screening, OA/US imaging within 10 days of screening, and biopsy within 30 days of screening. The outcome is restricted to benign masses in the intent-to-diagnose population described for specificity. Details such as the recruitment process, site-level enrollment, eligibility criteria, and how sequential participation was implemented are not reported in the registry record.

The registry lists two outcome measures and zero analysis records. It does not report a statistical analysis plan, a sample-size justification for the original study, or details about planned interim reviews. The planning calculations associated with the posted results are illustrative scenarios rather than evidence about the original design or observed study power.

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: Subjects with BI-RADS 4A or 4B by Conventional Diagnostic Ultrasound

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The intent-to-diagnose population of participants with benign masses. This is the ITD population of masses for specificity.

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: Intent-to-diagnose (ITD)

6. Primary Endpoint

The registry lists the primary outcome as OA/US Specificity (Downgrade (%) for BI-RADS 4A & 4B) of Benign Masses, measured over CDU images and decision to biopsy to be done at Screening. OA/US imaging to be done within 10 days of Screening. Biopsy to be done within 30 days of Screening..

The primary endpoint is the percentage of benign masses correctly downgraded by OA/US from a suspicious BI-RADS 4A or 4B classification to BI-RADS 2 or 3. The registry labels this outcome OA/US specificity and describes its unit as a percentage of masses. Its target population is the intent-to-diagnose population of participants with benign masses.

Specificity is useful in this setting because the endpoint asks how often the imaging classification identifies benign masses as benign or probably benign rather than leaving them in a suspicious category. In general, a diagnostic specificity measure describes correct negative classifications among cases that do not have the condition of interest. Here, the registry operationalizes the question through the downgrade of benign masses. That makes the endpoint relevant to the stated classification task, but it does not by itself describe sensitivity, overall diagnostic accuracy, or patient outcomes.

The registry states that specificity is reported with a 96% confidence interval using a normal approximation. A confidence interval expresses uncertainty around an estimated quantity; it does not establish that the estimate is free from bias or that the study design supports causal comparisons. The observed percentage, its confidence limits, and the number of benign masses contributing to this endpoint are not reported in the registry record.

The timing description links the outcome to screening, subsequent OA/US imaging, and biopsy. The registry does not report further detail about how indeterminate or unavailable assessments were handled in the endpoint calculation.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary measure is a percentage, with the mass—not an imaging arm comparison—as the unit described in the endpoint record. The target set is benign masses in the intent-to-diagnose population. In general, an intent-to-diagnose approach aims to evaluate participants or diagnostic cases according to the intended assessment process, rather than limiting interpretation only to ideal or fully evaluable cases. The registry does not give the operational rules used to define evaluability or to handle unavailable assessments.

What does the percentage represent?

The numerator for the endpoint conceptually consists of benign masses correctly classified at the stated lower-risk categories after OA/US imaging: BI-RADS 2 or 3, following an earlier BI-RADS 4A or 4B assessment. The denominator is the set of benign masses in the specified population. This explains the meaning of the measure, but the actual numerator and denominator are not reported in the registry record, and no percentage can be reconstructed from the available endpoint description.

In general, specificity concerns correct negative classifications among cases without the condition being targeted. The record applies that idea to downgrading benign masses. This is not a complete account of diagnostic performance: sensitivity concerns identification of cases with the condition, while overall accuracy combines correct classifications across relevant categories. Neither measure can be inferred from the primary endpoint alone.

How should the confidence interval be read?

The registry specifies a 96% confidence interval calculated using a normal approximation. Generally, a confidence interval is a range of values compatible with the observed estimate under the statistical method and its assumptions. Its width communicates sampling uncertainty, not every kind of uncertainty. It does not account automatically for systematic differences between groups, imperfect reference classifications, or missing data.

A normal approximation uses a normal distribution to approximate the sampling behavior of an estimate. Whether that approximation is suitable depends on features such as the amount of information and the distribution of observations. Those underlying details, including the number of benign masses and the event counts for the endpoint, are not reported in the registry record. The numerical interval is also not reported, so the precision of the estimate cannot be assessed.

Why does the design affect comparisons?

Random allocation can help balance measured and unmeasured characteristics between comparison groups. In this study allocation was non-randomized, and the registry warns that the arms are not a like-for-like control comparison. As a general matter, differences in participant characteristics, timing, or clinical assessment can influence observed outcomes when treatment or device assignment is not randomized. The record does not report enough detail to assess or adjust for such differences. Consequently, a between-arm contrast should not be interpreted as a causal device effect.

The study model is sequential and masking is recorded as none. These terms describe features of the study structure, but the registry record does not provide detailed procedures for sequence, reader blinding, or adjudication. The lack of masking can matter in general when an outcome depends on interpretation or classification, since awareness of an imaging method may influence decisions. Whether and how that issue affected this endpoint cannot be determined from the registry record.

What do the planning calculations mean?

The planning illustrations use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. They describe sample-size scenarios for detecting a stated fraction of the observed effect. Such calculations are prospective planning aids: they show how assumed effect size and design choices relate to a sample-size target. They are not observed power, do not describe the study's actual enrollment rationale, and do not establish what the original investigators planned.

For non-time-to-event outcomes, those illustrations are expressed as participants per arm for 80% power or other stated planning targets. The actual scenario values are not reported in the registry record. The observed effect used as the basis for these illustrations is also not reported in the registry record, so the assumptions cannot be independently assessed from the endpoint results. Planning values should not be read as an estimate of the probability that the completed study detected a real effect.

The registry lists zero analysis records. It does not report the inferential procedure for any additional outcome, adjustments for covariates, handling of multiple assessments, or sensitivity analyses. Those omissions limit the ability to reproduce or fully evaluate the statistical analysis.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists two outcome measures and zero analysis records, but it does not report a detailed statistical analysis plan or explain how the measures relate to any family of hypotheses. A multiplicity strategy—such as adjustment for testing several outcomes or contrasts—is not reported in the registry record. Without that information, the chance of false-positive findings across multiple analyses cannot be assessed. The primary outcome is identified, but that alone does not establish how other outcomes were handled statistically.

The primary endpoint population is described as the intent-to-diagnose population of participants with benign masses. The registry does not give the number of masses included in that analysis, the number with missing imaging or biopsy information, or the rules used to classify incomplete records. Methods for handling missing data, including whether observations were excluded, imputed, or otherwise addressed, are not reported in the registry record.

These omissions matter because missingness can affect both the estimated percentage and its uncertainty, especially if cases without complete assessment differ from cases included in the analysis. The endpoint description establishes the intended population and the imaging-to-biopsy timing, but does not explain how deviations from that sequence or unavailable reference information were treated.

The record also does not report subgroup analyses, covariate adjustment, interim analyses, or sensitivity analyses. No conclusions about their use should be drawn from their absence in the registry record.

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 non-serious adverse event terms

The adverse-event record lists the MAESTRO group with 217 participants at risk for serious adverse events and zero affected. For other adverse events, it lists 217 at risk and one affected. Deaths are listed as zero, and no top serious adverse-event terms are listed.

These are counts in the registry record, not comparative safety estimates. The study is non-randomized, and the registry cautions that the arms are not a like-for-like control comparison. The safety figures therefore do not support a causal comparison between Imagio IUS and Imagio OA/US. They also do not establish that serious events could not occur outside the reported observation or recording process.

The record does not provide event definitions, follow-up duration for safety, severity or relatedness assessments, or details about how events were collected. The absence of listed serious-event terms should be read as no terms reported in this record, not as evidence that all possible harms were assessed and ruled out. The single other adverse event is not described by type in the registry entries.

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

The record defines a primary diagnostic question focused on benign masses initially classified as BI-RADS 4A or 4B and whether OA/US imaging resulted in a downgrade to BI-RADS 2 or 3. It also specifies a 96% confidence interval using a normal approximation. These details make the intended endpoint understandable, but the registry record does not report the observed percentage or its confidence limits. The endpoint's magnitude and statistical precision therefore cannot be stated.

The study is a completed, non-randomized post-market surveillance study. Its design can describe the assessment process and endpoint as registered, but the cautioned non-like-for-like comparison does not support attributing differences between the listed devices to the device alone. The record does not establish a causal treatment effect.

The endpoint is limited to the downgrading of benign masses. It does not, by itself, establish diagnostic performance for malignant masses, patient outcomes, or the consequences of changing a classification. Those results are not reported in the registry record. The record also does not provide the primary analysis denominator or the missing-data procedures, so completeness and representativeness of the endpoint analysis cannot be evaluated.

The adverse-event counts report zero participants affected by serious adverse events, one affected by other adverse events, and zero deaths within the listed group. Those figures describe what was recorded; they do not demonstrate an absence of risk or provide a randomized safety comparison.

Planning calculations are illustrative sample-size scenarios based on assumptions and fractions of an observed effect. They are not observed power and should not be interpreted as evidence that the study had a particular probability of detecting an effect. In short, the registry establishes the study's intended question and broad design, while key numerical efficacy results and analysis details are not reported in the registry record.

13. Why This Trial Matters Statistically

MAESTRO is statistically notable because its primary outcome translates an imaging classification task into a defined percentage: correct downgrading of benign masses from BI-RADS 4A or 4B to BI-RADS 2 or 3. That definition clarifies what the endpoint is intended to measure and keeps attention on a particular diagnostic decision rather than an unspecified notion of imaging performance.

The design also illustrates why statistical interpretation depends on how a study is organized. Non-randomized allocation and the registry's explicit warning that the arms are not like-for-like limit the meaning of between-arm comparisons. An apparent difference, if reported, could not automatically be attributed to the imaging device without addressing the potential for group differences and other sources of bias.

Another important issue is the distinction between an estimate and its precision. The registry specifies a 96% confidence interval and a normal approximation, but does not report the numerical endpoint estimate or limits. Without those values, readers cannot assess the observed percentage or how tightly it is estimated. A methodological label cannot substitute for the result itself.

The record also distinguishes planning from evidence. Sample-size illustrations based on fractions of an observed effect can inform future study planning, but they are not observed power and do not establish the rationale for the completed study's enrollment. That distinction is useful when registry-based calculations appear alongside trial results.

Finally, the safety counts are descriptive and have limited context in the record. Reading them alongside the non-randomized design and unreported event-collection details helps avoid treating a small number of recorded events as a definitive assessment of safety. The study's statistical value therefore lies in its clearly described question and the interpretive lessons of its design, while the missing primary result constrains conclusions about the measured outcome.

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 most direct limitation in the registry record is that the primary endpoint's observed percentage and its 96% confidence interval are not reported. The number of benign masses contributing to the endpoint is also not reported, which prevents assessment of the analysis denominator and statistical precision.

The non-randomized design limits causal comparisons, and the registry explicitly states that the arms are not a like-for-like control comparison. The record does not provide further details about assignment procedures, participant characteristics by arm, or adjustments that might address differences between groups. Masking is recorded as none, while the specific procedures for image reading or outcome adjudication are not reported in the registry record.

The endpoint is restricted to benign masses and a particular downgrade classification. The record does not report corresponding diagnostic results for malignant masses, other classification changes, or patient outcomes. The criteria for the reference diagnosis and the handling of incomplete assessments are not described in the available record.

Although the study is described as multi-center and sequential, details about sites and implementation are not reported. The registry lists two outcome measures and zero analysis records; it does not report multiplicity handling, missing-data methods, interim analyses, or sensitivity analyses. The original sample-size justification is not reported in the registry record. The planning illustrations do not fill that gap because they are not evidence of the study's original design assumptions or observed power.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: IMAging With Opto-acoustics to downgradE BI-RADS claSsificaTion Relative tO Other Diagnostic Methodologies (MAESTRO) — NCT02364388.
• Registry record last updated 2020-04-06; status last verified 2020-03.
• Results first posted 2020-04-06.
• 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

NCT02364388, MAESTRO, is a completed post-market, non-randomized multi-center surveillance study of Imagio breast imaging. The registry lists enrollment of 217, a sequential model, no masking, and phase NA. The recorded study start date is 2015-03-27, with primary completion on 2016-02-15.

The primary outcome is the percentage of benign masses initially classified as BI-RADS 4A or 4B that OA/US correctly downgrades to BI-RADS 2 or 3. The registry specifies a 96% confidence interval using a normal approximation, but the endpoint estimate, confidence limits, and analysis denominator are not reported in the registry record.

Because allocation was non-randomized, the registry warns that the listed arms are not a like-for-like control comparison. Safety records list zero serious adverse events, one other adverse event, and zero deaths. Detailed analysis methods, missing-data handling, and multiplicity procedures are not reported in the registry record.