The incremental prognostic value of myocardial perfusion parameters as measured by positron emission tomography
Williams, Brent Allen
MetadataShow full item record
Background: Cardiac imaging modalities are an important component of the armamentarium in the clinical evaluation of individuals with known or suspected coronary artery disease (CAD), especially for guiding individualized therapeutic decision-making for risk minimization and maximal symptomatic relief. Positron emission tomography myocardial perfusion imaging (PET MPI) is regarded as providing better characterization of myocardial perfusion defects than other MPI modalities, yet widespread implementation of PET MPI has not been realized. Determining the utility of PET MPI in guiding therapeutic management first requires establishing the incremental prognostic value of PET MPI for important clinical outcomes. Accordingly, the specific aims of this study were to determine whether PET MPI provides incremental prognostic value with respect to cardiac and all-cause mortality, and to determine the extent PET MPI improves quantitative risk assessment. Methods: This retrospective cohort study included 3739 individuals who underwent rest-stress rubidium-82 PET MPI for evaluation of known or suspected CAD at the Heart Center of Niagara between April 2000 and June 2006. Automated software calculated the size of rest, stress, and stress-induced perfusion defects measured in percentage of the left ventricular (LV) myocardium, grouped into 0%, 0-5%, 5-10%, 10-20%, and ≥20% of the LV. Mortality information was obtained through the National Death Index with follow-up through December 31, 2006. Cox proportional hazards models were developed to determine the incremental prognostic value of PET MPI beyond important clinical variables. Risk reclassification tables were developed to determine the extent PET MPI permitted improved allocation of individuals into clinically relevant risk strata. Results: Larger perfusion defects were associated with a higher prevalence of traditional cardiovascular disease risk factors, known coronary artery disease, and arterial disease in other vascular beds. Annual cardiac mortality rates across stress perfusion defect size groups were 0.4%, 0.9%, 1.0%, 2.2%, and 3.2% (p for trend < 0.001). Adjusted hazard ratios (HR) across stress perfusion defect size groups were 1.0, 1.8, 1.5, 2.8, and 3.9 (p for trend < 0.001). When separating stress perfusion defects into rest and stress-induced components, a strong association with cardiac death was observed for rest perfusion defects (HRs across groups: 1.0, 2.2, 1.9, 4.0, 6.3; p for trend < 0.001) but not for stress-induced perfusion defects (HRs 1.0, 1.1, 1.0, 1.0, 1.1; p for trend = 0.902). These patterns generally held for all-cause mortality. PET MPI reclassified 15% of individuals into more appropriate cardiac death risk strata; among individuals at intermediate and high pre-PET risk, 36% and 30% were reclassified by PET MPI respectively. Conclusions: PET MPI provided strong incremental prognostic value for mortality endpoints which were limited to resting perfusion defects. Furthermore, PET MPI permitted improved allocation of individuals into clinically relevant risk strata, particularly among individuals at intermediate and high pre-PET risk. These findings suggest PET MPI may be valuable in guiding therapeutic management in a cost-effective manner. Further research incorporating non-fatal endpoints and longer follow-up time is needed to determine the proper incorporation of PET MPI in the clinical evaluation of individuals with known or suspected CAD.