📉 Preserved MFR alone is no longer enough on PET
📉 Preserved MFR alone is no longer enough on PET
In a multicentre REFINE PET registry study of 16,650 patients undergoing regadenoson 82Rb-PET MPI, elevated AI-derived splenic ratio (SR) was independently linked to higher major adverse cardiovascular event risk—even among patients with preserved myocardial flow reserve (MFR ≥2). The study suggests preserved MFR alone is no longer enough for risk reassurance on PET when stress adequacy, captured by automated SR, is abnormal.
Why It Matters To Your Practice
Inadequate pharmacologic stress can blunt the diagnostic and prognostic value of PET MPI.
An AI-derived SR offers an automated check on stress adequacy using routine stress and rest splenic uptake.
Risk may be underestimated if clinicians rely on preserved MFR alone without accounting for whether vasodilator stress was truly adequate.
Clinical Implications
Consider abnormal SR a potential warning flag when interpreting regadenoson PET studies, especially if perfusion or flow findings appear reassuring.
Patients with preserved MFR but elevated SR may warrant closer clinical follow-up and more cautious risk assessment.
Automated SR extraction could be integrated into PET workflows at scale because it does not depend on manual measurement.
Insights
The analysis included 16,650 patients from six sites with clinically indicated MPI and linked outcomes.
SR was calculated fully automatically as the ratio of splenic uptake at stress versus rest, then stratified by deciles.
The primary endpoint was MACE, assessed with Kaplan-Meier and adjusted Cox models including clinical and imaging covariates such as MFR.
The Bottom Line
AI-derived SR adds prognostic information beyond standard PET metrics and may identify higher-risk patients whose preserved MFR would otherwise seem reassuring.
For clinicians interested in AI, this is a practical example of automation improving interpretation quality and risk stratification rather than replacing clinical judgment.