Gadobutrol/Gadavist-enhanced Cardiac Magnetic Resonance Imaging (CMRI) to Detect Coronary Artery Disease (CAD): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Gadobutrol/Gadavist-enhanced Cardiac Magnetic Resonance Imaging (CMRI) to Detect Coronary Artery Disease (CAD) (NCT01890421), a na phase 3 study. The primary outcome is Presence of a Myocardial Perfusion Defect Indicating Significant CAD Per Participant on Gadobutrol-enhanced CMRI (Based on RPS) - Primary Analysis of Sensitivity Based on Blinded Readers' Assessment.

2. Clinical Question

3. Trial Design

This Phase 3 study is described as multicenter and open-label, with a single-group model and no allocation to competing trial arms. The registry records enrollment of 426 and lists the study status as completed. Gadobutrol was the intervention; there was no separate untreated or active-comparator study group. Consequently, the main purpose was diagnostic evaluation against a reference standard, not an estimate of a randomized treatment contrast.

Participants being assessed for suspected or known coronary artery disease were invited to take part. The study description says that participation could last from 2 days to 4-6 weeks and involve one to four study-doctor visits. Coronary angiography or computed tomography angiography could have been performed before enrollment or scheduled after the study, within the timing described in the registry. Cardiac magnetic resonance images and reference-standard images were collected for independent image review.

Image reading was blinded, although the overall study is classified as open-label. These describe different aspects of the design: open-label does not mean that the diagnostic image readers necessarily knew the reference results. The primary assessment was based on blinded readers' interpretation of regional perfusion scores. The registry also lists investigator assessments and comparisons involving unenhanced wall-motion images among secondary outcomes.

The diagnostic reference was routine clinical coronary angiography or computed tomography angiography. The primary definition of significant disease used quantitative coronary angiography stenosis of >=50%. The record does not report how participants were distributed across the reference methods or give enough information to reconstruct the full analysis workflow. It also does not report the protocol's sample-size justification or the detailed statistical analysis plan.

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: Overall, 456 participants signed the informed consent, of them 19 did not finish their baseline visit (6 screening failure, 13 dropped out), 1 discontinued the study due to an adverse event (AE). A total of 436 participants entered the diagnostic imaging phase, of them 426 were treated with gadobutrol and entered the follow-up phase.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Full analysis set (FAS): participants who underwent pharmacologic stress and for whom electronic case report form entries, adequate image sets for unenhanced and gadobutrol-enhanced CMRI, and the complete image set for SoR diagnosis were available. N=FAS with significant CAD (defined as maximum stenosis severity of \>=50% by QCA) as verified by SoR.

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: Safety analysis set (SAF)

6. Primary Endpoint

The registry lists the primary outcome as Presence of a Myocardial Perfusion Defect Indicating Significant CAD Per Participant on Gadobutrol-enhanced CMRI (Based on RPS) - Primary Analysis of Sensitivity Based on Blinded Readers' Assessment, measured over 0 to 30/40 minute (min) post-injection.

The primary endpoint was sensitivity for detecting significant coronary artery disease using gadobutrol-enhanced cardiac magnetic resonance imaging. Blinded readers assessed myocardial regions using a regional perfusion score. A score of 0 represented normal perfusion; scores of 1 or higher represented a perfusion defect. The registry describes the higher scores as distinguishing reversible, mixed, and fixed or permanent defects.

For this primary analysis, significant disease was defined as quantitative coronary angiography stenosis of >=50%. Sensitivity is the proportion of participants who have the disease according to the reference standard and whose imaging assessment is positive. In general, sensitivity focuses on detection among people classified as having the condition; it does not by itself describe how often a negative image correctly excludes disease. Specificity addresses that complementary question and appears among the listed secondary outcomes.

A sensitivity endpoint is relevant to a diagnostic study because it evaluates whether the imaging interpretation identifies disease present according to the chosen reference. Its meaning depends on the reference standard, the threshold used to classify disease, the imaging-reading rules, and which participants are included in the analysis. The registry record states a 95% confidence interval as the endpoint's measure of uncertainty and gives the assessment window as 0 to 30/40 minute post-injection. However, neither the primary sensitivity estimate nor its confidence interval is reported in the registry record, so the endpoint's observed performance cannot be described from the recorded results.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

In general, diagnostic accuracy is assessed by comparing an index test with a reference standard. Here, the index test was gadobutrol-enhanced cardiac magnetic resonance imaging, and the reference was routine clinical coronary angiography or computed tomography angiography. The registry defines significant disease for the primary analysis using quantitative coronary angiography stenosis of >=50%. The reference method and disease definition matter because a diagnostic accuracy estimate describes performance relative to the chosen standard, rather than an error-free account of disease in every participant.

What does sensitivity mean?

In general, sensitivity is the proportion of participants with the condition according to the reference standard who receive a positive index-test classification. The registry describes it as true positives divided by true positives plus false negatives. A false negative in this setting means that the reference standard classifies significant disease as present while the imaging assessment does not identify a qualifying perfusion defect. Sensitivity does not tell the reader how often a positive result is correct among all positive images; that is a different quantity. Nor does sensitivity alone describe the ability to rule out disease.

The primary endpoint classified a myocardial region as having a perfusion defect when its regional perfusion score was at least 1, and as normal when the score was 0. The score distinguished reversible, mixed, and fixed or permanent defects. The primary analysis focused on participant-level presence of a defect indicating significant disease, rather than treating every region as an independent participant. This distinction is important in general because multiple measurements from one person are related and should not automatically be interpreted as independent observations.

Why report a confidence interval?

In general, a confidence interval communicates uncertainty around an estimate. A narrower interval indicates greater statistical precision under the method used; a wider interval indicates less precision. It does not eliminate uncertainty about study conduct, the quality of the reference standard, or how well the enrolled participants represent other settings. The registry specifies a 95% confidence interval for the primary endpoint, but the interval and point estimate are not reported in the registry record. Precision therefore cannot be evaluated from the available result description.

The endpoint's timeframe is recorded as 0 to 30/40 minute post-injection. The record also describes assessments during stress and at rest after separate bolus injections. These details identify the intended timing and acquisition context; they do not provide the numerical sensitivity result. The record does not report the specific confidence-interval construction or enough information to reproduce it.

How should the reference standard be considered?

In general, an accuracy estimate depends on the way the reference standard classifies disease. If different reference procedures are used, their measurement properties and the rules for applying them can affect interpretation. This study permitted coronary angiography or computed tomography angiography as the reference, while the primary disease definition mentions quantitative coronary angiography. The record does not report how the methods were combined, how many participants had each method, or whether analyses were stratified by reference method. Those details are relevant when judging how broadly a single diagnostic estimate would apply.

Blinded image review is relevant because knowledge of the reference classification could influence interpretation of the index images. The primary outcome is described as based on blinded readers' assessment, and the image sets were collected for independent review. The overall trial, however, is classified as open-label. These statements are not necessarily contradictory: masking of study participants or personnel and blinding of image readers are separate design features. The registry does not report further operational details about the blinding process.

Why do the analysis population and missing images matter?

The full analysis set required pharmacologic stress, electronic case-report-form entries, adequate unenhanced and gadobutrol-enhanced image sets, and a complete image set for the reference diagnosis. In general, restricting an accuracy analysis to participants with complete test and reference information can affect how representative the analyzed group is. If missingness is associated with participant characteristics or diagnostic difficulty, the estimate among evaluable participants may differ from what would have been observed with complete information for everyone enrolled. The registry does not report the number excluded for each reason or the method used to address missing information.

The primary sensitivity population is further described as the full analysis set with significant disease. The complete disease-population definition and analysis count are not reported in the registry record. The registry's enrollment figure should therefore not be read as the endpoint denominator. Without the number of participants contributing to the sensitivity analysis and the positive and negative classification counts, the calculation cannot be reconstructed.

What do the planning illustrations represent?

In general, sample-size planning asks how many observations might be needed for a proposed design and target effect under specified assumptions. The registry's planning assumptions describe a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and effects framed as fractions of the observed effect. For non-time-to-event outcomes, the illustration is expressed as participants per arm; time-to-event illustrations use total events. These are planning conventions, not observed outcomes, and do not make this single-group diagnostic study a two-arm trial.

The record characterizes the illustration as based on the observed effect, but it does not report the primary observed sensitivity estimate needed to assess that basis. It also does not report the trial's own sample-size justification. A planning calculation is not observed power: it cannot establish what the completed study achieved, and it should not be interpreted as evidence that a nonsignificant finding would imply no diagnostic performance. No recalculated sensitivity, confidence interval, or hypothesis test can be described because the relevant results are not reported in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 20 outcome measures and 8 analyses. The described outcomes cover several related but distinct questions: sensitivity, specificity, investigator assessments, comparisons with unenhanced wall-motion imaging, and defect localization. Testing or estimating multiple outcomes raises a multiplicity issue in general, because the chance of at least one apparently notable finding can increase when many analyses are considered. The registry record does not report whether a multiplicity adjustment, hierarchical testing strategy, or other prespecified approach was used. It also does not report enough analysis-plan detail to identify which secondary analyses were confirmatory and which were exploratory.

The primary endpoint's full analysis set required pharmacologic stress and specified case-report-form and image availability, including complete reference-standard images. This gives some information about eligibility for the analysis, but it is not a complete account of missing-data handling. The numbers of participants with missing or inadequate images, the reasons for missing information, and any imputation or sensitivity analyses are not reported in the registry record. The complete endpoint-specific denominator is also not reported.

The primary population is described as the full analysis set with significant coronary artery disease, but the registry record does not provide its complete definition or count. It therefore cannot be determined from the record how the enrolled group was reduced to the population contributing to the sensitivity estimate, or how participants without evaluable reference information were handled. These omissions limit evaluation of selection effects and reproducibility.

The detailed statistical analysis plan, including any interim analyses and the trial's sample-size justification, is not reported in the registry record. The planning assumptions described for an illustrative calculation should not be treated as the missing trial analysis plan or as evidence of a multiplicity procedure.

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: chest pain (0.2%).

The registry lists 426 participants at risk in the gadobutrol group. It records one participant affected by a serious adverse event, 47 affected by other adverse events, and zero deaths. Chest pain appears among the serious adverse-event terms at 0.2%. These are descriptive safety entries; they do not report the diagnostic sensitivity endpoint.

In general, an affected-participant count is not necessarily the same as a count of individual event episodes: one participant may experience more than one event. The registry record does not provide enough detail to determine event frequency per participant, timing, severity criteria, or the basis for attributing events to gadobutrol. Nor does the serious-event entry by itself establish causation. The record's open-label, single-group design also means these entries do not supply a randomized safety comparison against a control group.

The listed counts should be interpreted in the context of the recorded intervention group and the registry's event categories. The information does not describe all clinical outcomes or establish a comparative safety profile. The registry record does not report additional detail on event definitions, follow-up completeness, or the relationship between the events and the imaging procedure.

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

The study was designed to assess whether blinded readers could identify perfusion defects on gadobutrol-enhanced cardiac magnetic resonance images among participants classified as having significant coronary artery disease. The registry specifies the reference approach, the quantitative coronary angiography threshold used for the primary disease definition, the regional perfusion scoring framework, and a 95% confidence interval as the measure of endpoint uncertainty.

Those design details explain what the primary result would represent, but the numerical sensitivity estimate and its confidence interval are not reported in the registry record. As a result, the posted record does not establish the observed proportion detected, how precise that estimate was, or how it compares with any target. It also does not report the primary analysis denominator or the underlying classification counts. No conclusion about the magnitude of diagnostic performance can be drawn from the endpoint description alone.

Even if a sensitivity estimate were available, it would describe performance in the analyzed population against the selected reference standard. It would not, by itself, establish specificity, positive predictive value, or performance in every population and clinical setting. The listed secondary outcomes address some other diagnostic questions, but their results are not reported in the registry record.

This study should not be read as a randomized comparison of gadobutrol-enhanced imaging against another imaging method or against no contrast. It is a single-group diagnostic evaluation, and the registry record does not report a control arm. It also does not show that the imaging approach changes patient outcomes. A diagnostic accuracy result and a clinical outcome comparison answer different questions.

The safety entries are likewise descriptive and lack a control-group comparison in this record. They do not establish whether listed events were caused by the intervention. Taken together, the record allows readers to understand the study's question and some key methods, while leaving the principal efficacy result, its precision, and important analysis details unreported.

13. Why This Trial Matters Statistically

This trial raises several useful statistical issues for diagnostic research. First, diagnostic performance is defined relative to a reference standard. The registry identifies coronary angiography or computed tomography angiography as reference procedures and specifies a quantitative coronary angiography stenosis threshold for the primary disease definition. Understanding that classification is essential: sensitivity describes detection against the reference used, not a comparison with an independent, universally definitive measure of disease.

Second, the study separates image acquisition and interpretation from verification. The registry describes independent, blinded image review, while the overall study remains open-label. This illustrates why masking should be described by role rather than treated as a single all-or-nothing property. For diagnostic studies, reader knowledge of the reference result can be an important source of bias, so the stated blinded reading is relevant to interpreting the intended method.

Third, a diagnostic endpoint often has an analysis population different from the overall enrollment. Here, the full analysis set required pharmacologic stress and specified complete image and case-report-form information, and the sensitivity analysis further concerned participants with significant disease. The missing endpoint denominator and incomplete population details prevent assessment of how those restrictions affected precision or representativeness.

Finally, the registry lists numerous outcomes and analyses but does not report a multiplicity strategy or the numerical primary result. This makes it impossible to assess the estimate's precision or place secondary findings in a prespecified testing hierarchy. The case illustrates that a clearly defined diagnostic question is not enough for full statistical interpretation: readers also need the results, denominators, missing-data account, and analysis plan.

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 central limitation of the registry record for interpreting this study is that it does not report the primary sensitivity estimate or its 95% confidence interval. The primary analysis denominator and the classification counts needed to reconstruct sensitivity are also not reported. Therefore, the magnitude and precision of the diagnostic result cannot be evaluated.

The full analysis set is described through required pharmacologic stress, case-report-form entries, and image availability, but the record does not report how many participants failed to meet each condition or how missing data were handled. The complete definition and count of the primary population with significant disease are not reported in the registry record. This leaves uncertainty about selection into the endpoint analysis and limits independent reproduction.

The study used coronary angiography or computed tomography angiography as a standard of reference, while its primary disease definition specifies quantitative coronary angiography stenosis of >=50%. The registry does not report the numbers assessed by each reference method, the method-specific results, or further details on combining or applying these references. This limits evaluation of how reference-method differences may have influenced the estimate.

The record describes a single-group, open-label study, not a randomized comparison. Thus, it cannot provide a randomized contrast between imaging strategies. Although the primary readers were described as blinded, further details of the blinding procedures are not reported. The record also does not report the trial's own sample-size justification, complete statistical plan, interim analyses, or multiplicity approach. Finally, the recorded safety counts are descriptive and do not include a control-group comparison or detailed causal assessment.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Gadobutrol/Gadavist-enhanced Cardiac Magnetic Resonance Imaging (CMRI) to Detect Coronary Artery Disease (CAD) — NCT01890421.
• Registry record last updated 2018-05-16; status last verified 2018-04.
• Results first posted 2018-05-16.
• 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

NCT01890421 is a completed Phase 3, multicenter, open-label, single-group diagnostic study of gadobutrol-enhanced cardiac magnetic resonance imaging in participants with known or suspected coronary artery disease. The registry records enrollment of 426. The intervention regimen was 0.1 mmol/kg body weight, given as separate stress and rest injections.

The primary outcome was sensitivity for identifying a myocardial perfusion defect among participants with significant disease, based on blinded readers' assessments and a reference standard. The primary disease definition used quantitative coronary angiography stenosis of >=50%. The registry describes a full analysis set requiring pharmacologic stress and specified complete imaging and case-report-form information.

The primary sensitivity estimate, its 95% confidence interval, and the endpoint-specific analysis count are not reported in the registry record. Safety entries list one participant affected by a serious adverse event, 47 affected by other adverse events, and zero deaths; chest pain is listed among serious adverse-event terms at 0.2%. The record does not report a randomized control comparison or the detailed statistical plan.