Table 17: Study characteristics of prognostic studies.

Author, Year	Study Design & No. Patients	Patient Population	Mean Follow-up	Technique (tracer)	Outcomes	Viability Criteria
Desideri et al., 2005 (52)	Prospective cohort n=261	Patients with chronic CAD with LVEF ≤40% being evaluated for surgical revasc.	2.1y (median)	13N- ammonia / FDG	Cardiac death	Mismatch: reduced perfusion (ammonia uptake < 70% of maximum) with relatively preserved FDG uptake (FDG ≥ 70%)
Feola et al., 2008 (53)	Prospective cohort n=93*	Patients with previous MI and LVEF <40%	342 ± 78 days Range: 110- 434	13N-ammonia / FDG	Cardiac death, hospital re-admission for MI or worsening heart failure	Mismatch: FDG uptake scores ranging from 0 to 2 with reduced (< 2.5 standard deviations) with respect for normal ammonia myocardial blood flow
Sawada et al., 2005 (54)	Prospective cohort n=61	Patients with diabetes and ischemic LV dysfunction	4.3 ± 8.8 years	13N-ammonia / FDG	Cardiac death	Mismatch: FDG minus ammonia > 2 standard deviations above the mean FDG minus ammonia difference of the normal database
Rohatgi et al., 2001 (55)	Retrospective cohort n=99†	Patients with known CAD being considered for revasc. with no viable tissue based on thallium SPECT imaging	25±9 months Range: 9-50	13N-ammonia / FDG	Sudden cardiac death (from MI or arrhythmia), MI, or admission to hospital for dyspnea or other signs of congestive heart failure	Mismatch: areas with diminished perfusion (<70% of peak counts) with a >12.5 increase in FDG activity relative to flow involving at least 20% of contiguous myocardium
Zhang et al., 2001 (56)	Cohort‡ n=123	Patients with previous MI and LV ≤ 45% (41 patients had LV aneurysms)	26±10 months Range: 1-36 Median: 28	Tc-MIBI SPECT / FDG PET	Cardiac events: cardiac death, acute MI, unstable angina requiring revasc. and late revasc (>3 mo after imaging)	Mismatch: regions with perfusion defect but preserved FDG uptake
Di Carli et al., 1998 (57)	Cohort‡ n=93	Patients undergoing myocardial viability assessment for CAD with LVEF <40% who are candidates for revasc.	3.8 years Range: 0-6.2	13N-ammonia / FDG	Cardiac death (deaths caused by acute MI, ventricular arrhythmia, and congestive heart failure)	Mismatch: segmental FDG uptake was increased relative to perfusion. Patients were classified as viable if > 5% of the LV had a mismatch pattern (based on ROC curve analysis)
vom Dahl et al., 1997 (58)	Prospective cohort n=161*	Patients with CAD and ischemic regional wall motion abnormalities being considered for revasc.	29±6 months Range: 22-44	99Tc-MIBI SPECT / FDG PET	Cardiac events: cardiac death; MI; unstable angina requiring revasc.; heart transplantation; or survived cardiopulmonary resuscitation Improvement in NYHA or CCS status	Mismatch: Tc uptake < 70%, FDG minus Tc uptake > 20%, and FDG uptake >70%. One territory for each patient was selected for analysis and the patient was defined as viable or not based on whether match or mismatch was present in most segments within the territory.
Lee et al., 1994 (59)	Retrospective cohort n=129§	Patients with LV dysfunction related to a previous MI and known coronary anatomy	17±9 months	Rb-82 / FDG	Cardiac events: unstable angina; MI; cardiac death	Mismatch: presence of FDG activity > 2 SD above the reference normal standard in segments with reduced perfusion. Patients were classified as viable if they had ≥ 1 mismatch segment.
Eitzman et al., 1992 (60)	Retrospective cohort n=82║	Patients with known CAD, decreased LVEF, and ≥1 major vessel with significant stenosis who underwent PET viability scanning between 1988 and 1990	12 months	13N-ammonia / FDG or Rb-82 / FDG	NYHA classification for heart failure, Canadian Cardiovascular Society classification for angina, MI, death, or urgent revasc. due to symptom development	Mismatch: an area with decreased blood flow and relatively increased FDG uptake

CAD refers to coronary artery disease; CCS, Canadian Cardiovascular Society; d, days; FDG, F-18-flurodeoxyglucose; LV, left ventricular; LVEF, left ventricular ejection fraction; mo, months; MI, myocardial infarction; NR, not reported; NYHA, New York Heart Association; revasc, revascularization; Rb-82, rubidium-82; yr, years

^*Consecutive patients

^†144 patients were eligible for inclusion, but only 99 were included as 46 were lost to follow-up.

^‡Prospective or retrospective not reported in paper.

^§137 patients were enrolled in the study, but the analysis only includes 129 patients. Reasons for exclusion were: technically inadequate studies (2 patients); cardiac transplantation (4 patients); loss to follow-up (2 patients).

^║110 patients were eligible for the study, but complete data acquisition was not possible in 23 patients, 3 received heart transplants, and 2 had poor quality PET scans which provided inadequate data for inclusion in the study.