²⁰¹Tl SPECT imaging is the oldest and most common method of assessing myocardial ischemia. It is also well established as a means of measuring myocardial viability. Traditional viability assessment using ²⁰¹Tl may be accomplished by various protocols which universally utilize the favorable kinetics of thallium-201. The initial myocardial uptake early after intravenous injection of thallium is proportional to regional blood flow. However, redistribution of ²⁰¹Tl is related to the rate of ²⁰¹Tl “washout” from the myocardium, dependant on the concentration gradient between myocytes and the thallium concentrations in the blood and the integrity of the membrane-based Na+/K+ ATPase pump [7, 14‐17]. As a potassium analogue ²⁰¹Tl uptake, retention, and washout are partly dependent on this pump and therefore the integrity of the cell membrane. Loss of membrane integrity is indicative of cell death. Therefore, in regions of reduced initial uptake (indicating reduced flow), ²⁰¹Tl washout is slower in regions that are viable because the myocardium is able retain the tracer and take up circulating ²⁰¹Tl via the intact membrane transport. The corollary is that when there is necrosis or scar, there is no mechanism to retain or take up ²⁰¹Tl and hence washout is more rapid. On imaging, the former is seen as redistribution indicating viability; while in the latter, a persistent fixed defect indicates a region with loss of viability.

The most common clinical protocols used include stress-redistribution, rest-redistribution with baseline imaging of rest perfusion and delayed 3 to 4 h redistribution imaging, and stress-redistribution-reinjection. With standard stress-redistribution or rest-redistribution imaging, the 3 to 4 h delayed images may underestimate viability. Subsequent data support improved accuracy for delayed 24-hour redistribution images [18, 19]. Often though, counts are low and image quality is compromised with a single injection. This can be overcome by using a second injection of ²⁰¹Tl immediately after acquisition of delayed images (Fig. 1). This protocol incorporates initial image acquisition with pharmacological or exercise stress, delayed 3–4 h redistribution images, and a reinjection phase with imaging typically 1 h after redistribution [20]. This approach has yielded a high level of accuracy [21, 22] and also concordance with ¹⁸F-FDG PET imaging [21, 22].

Recent efforts have been focused on improving the ability of ²⁰¹Tl SPECT imaging to detect viable myocardium. Adjuvant glucose administration has been proposed. A number of studies using either glucose or glucose-insulin-potassium (GIK) infusion have been published showing that ²⁰¹Tl uptake is enhanced, thus improving detection of viable segments [23‐25]. Since ²⁰¹Tl and potassium possess similar kinetics, glucose administration will increase endogenous insulin release, thus ultimately acting to increase uptake of ²⁰¹Tl in myocardial cells. Hasbek et al. [25] recently demonstrated the benefit of oral glucose administration with ²⁰¹Tl. Using nondiabetic patients, they performed rest-redistribution-24 h late SPECT. When a perfusion defect was observed, 1 mCi of ²⁰¹Tl was given 30 min after a 75-g oral glucose load upon completion of the 24-hour late SPECT. Repeat imaging was then performed. Glucose plus ²⁰¹Tl images yielded the lowest number of perfusion defects, and showed the highest segmental perfusion improvement. The authors also found that ECG-gated in the EF calculations after oral glucose administration were significantly higher compared to EF calculations prior to glucose, attributed to both enhanced ²⁰¹Tl uptake and improvement in wall detection in segments with perfusion defects [25]. In summary, ²⁰¹Tl uptake can be enhanced by oral glucose administration and the detection of viable myocardium may be improved through a simple oral glucose administration protocol that is applicable to routine daily clinical practice.

Heiba et al. [26] recently published a study using a protocol combining rest-redistribution with ²⁰¹Tl and low-dose dobutamine contractility assessment. Data for perfusion were obtained first by a single rest injection of ²⁰¹Tl with gated SPECT image acquisition. Then 4-hour delayed gated SPECT images were acquired during dobutamine infusion. Segments were analyzed for rest-redistribution ²⁰¹Tl perfusion (normal, reversible, and fixed decreased perfusion) and wall motion and/or wall-thickening difference (normal, fixed dysfunctional, improved dysfunctional contractility) with low-dose dobutamine. Furthermore, segments with a fixed decrease in ²⁰¹Tl uptake were subclassified into mild, moderate, and severe fixed segments. In this study, over half of the dysfunctional segments with either a mild to moderate fixed decrease in thallium uptake showed evidence of improved wall motion or wall thickening. Only 30% of the segments with a severe fixed defect on redistribution showed improvement. Using this protocol the authors could obtain complementary data regarding myocardial viability. Using not only perfusion-redistribution data, contractile reserve and regional contractility could simultaneously be assessed adding to the functional characterization of viable myocardium. Perceived nonviable myocardium on ²⁰¹Tl imaging may actually have viable tissue identified with low-dose dobutamine potentially improving the prediction of patient outcomes and impacting management decisions. Another attractive feature is the routine ease of administration of low-dose dobutamine in the clinical setting. There is no additional ²⁰¹Tl injection or extra imaging time, making this protocol particularly attractive.

Thallium imaging remains the most utilized method for assessing myocardial viability. Important strides in research have been made to improve its specificity as highlighted above. However, other radionuclide modalities as well have also made important advances in the assessment of myocardial viability.

^99mTc-based radiopharmaceuticals currently available and used in daily clinical practice include sestamibi and tetrofosmin. To briefly review, they are lipid-soluble compounds that cross membranes through passive diffusion driven by the transmembrane electrochemical gradient ultimately being retained within the mitochondria of the myocyte [7]. In comparison to ²⁰¹Tl, ^99mTc-based radiopharmaceuticals have a higher energy spectrum (140 vs 80 keV) and shorter half-life (enabling a higher dose) which can serve to improve image quality; but have a worse relationship to flow and minimal redistributing properties. However, mitochondrial uptake of these compounds requires an intact mitochondrial membrane and oxidative metabolism, thus the basis for sensitive viability detection [27‐29]. A complete review of the properties and pharmacokinetics of ^99mTc-based radiopharmaceuticals is available in any major textbook [30].

In prior studies, extent of myocardial scar has been consistently overestimated using ^99mTc compared to ²⁰¹Tl [31‐33]. A recent study by Crean et al. [33] comparing ¹⁸F-FDG, cardiac magnetic resonance imaging (CMR), and ^99mTc-sestamibi (MIBI) found that a significantly higher number of segments were identified as scar by MIBI than by either ¹⁸F-FDG or CMR. The greatest discrepancy between MIBI and ¹⁸F-FDG and MRI was in detection of scar of the inferior and lateral wall. By the same token, nitrate (sublingual or intravenous) administration prior to MIBI imaging has been proposed to increase ability to detect myocardial viability while performing perfusion imaging. The exact mechanism is not well understood; however, it is theorized that nitrates improve myocardial blood flow to hypoperfused myocardial segments by dilating stenotic epicardial arteries. This may enable increased blood flow through collateral vessels to ischemic segments [34].

However, more important is the clinical decisions to be made from data provided by nitrate-augmented ^99mTc imaging. Yang et al. [34] recently reviewed data from eight ^99mTc-based studies utilizing nitrate imaging for viability detection (Table 1) [34‐42]. There was a consistent agreement between the occurrence of cardiac events and nitrate-augmented myocardial perfusion imaging. Furthermore, Sorrentino et al. [41] compared patient prognosis and outcome between nitrate ^99mTc-tetrofosmin SPECT and ¹⁸F-FDG-PET. This study evaluated 89 patients with LV dysfunction due to CAD with both nitrate-assisted ^99mTc-tetrofosmin SPECT and ¹⁸F-FDG-PET imaging. There was no significant difference in reported event-free survival in the viable myocardium group (and nonviable myocardium group) imaged with either SPECT or PET. The authors concluded that nitrate-assisted SPECT provided similar prognostic information to PET, thus not affecting patient management or outcome.

Administration of nitrates is easy, affordable, and effective in SPECT-assisted imaging for myocardial viability. Although scar-based quantitative and qualitative assessment may differ, patient outcomes and management decisions may not significantly be altered.

Cardiac PET utilizing ¹⁸F-FDG is considered the most sensitive modality for detecting hibernating viable myocardium. The premise behind ¹⁸F-FDG viability imaging relies on basic cardiac physiology during ischemia/injury. Under aerobic conditions and the fasting state, the preferential energy substrate of the heart is fatty acids. However, under conditions of ischemia or injury when oxidative metabolism is compromised, free fatty acid utilization is impaired, while glucose utilization is increased. Increased cardiac glucose utilization is also seen following an oral carbohydrate load. ¹⁸F-FDG is a glucose analog, which is transported into cells though glucose membrane transporters GLUT1 and GLUT4, and subsequently phosphorylated by hexokinase to ¹⁸F-FDG-6-phosphate. FDG does not undergo any further metabolism (unlike glucose) and essentially is trapped within the cell as ¹⁸F-FDG-6-phosphate. Trapped ¹⁸F-FDG accumulates and becomes an index of cellular glucose utilization and hence ongoing cellular metabolism [43].

There are two key patterns of clinical significance identifying either scar or hibernating, viable myocardium. A match pattern is an equivalent reduction in perfusion and FDG uptake (reduced or absent glucose metabolism), thus signifying myocardial scar. Conversely, a mismatch pattern is a visible reduction in perfusion with preserved uptake of FDG, signifying glucose metabolism and thus viability. There is another pattern of preserved perfusion, yet reduced glucose metabolism on PET viability imaging described as “reverse mismatch” and likely represents viable myocardium. This pattern may be seen in left bundle branch block, nonischemic cardiomyopathy, or diabetes; and may be relevant in patients being considered for resynchronization therapy [44, 45].

Currently, ¹⁸F-FDG PET is considered the most sensitive means of assessing viable myocardium and hence predicting LV functional recovery post–coronary revascularization [2••, 4, 5, 46••, 47‐54]. Prior observational outcome studies have suggested that ¹⁸F-FDG PET can help identify patients at high risk for cardiac events [55]. The PET and Recovery Following Revascularization (PARR 2) randomized controlled trial of 430 patients investigated patients with an EF < 35% and suspected or confirmed CAD, who were considered for revascularization, transplant, or heart failure work-up. A standardized clinical report was produced by the interpreting physician detailing the amount of scar, viable myocardium, and degree of mismatch. The results along with the perceived likelihood of recovery were conveyed to the treating physician and a decision was made on an individual basis. PET-assisted decision-making compared to standard care showed a trend but did not reach statistical significance with respect to the primary composite end point (cardiac death, myocardial infarction, or recurrent hospital admission for cardiac causes). However, a post hoc analysis of the data was performed investigating adherence to PET recommendations. When management adhered to the imaging recommendations, a significant reduction in adverse outcomes was observed in the PET arm [3].

In a recent substudy of the PARR-2 trial, D’Egidio et al. [2••] set out to determine the predictors of outcome. A total of 182 patients randomized to the PET arm of the PARR-2 trial were included. A key finding was the interaction of mismatch and revascularization such that as mismatch extent increased so too did the benefit of revascularization (Figs. 2, 3 and 4). In those patients with a mismatch score of <7%, there was no significant difference in the primary outcome regardless of whether revascularization was performed or not. However, patients with a mismatch score of ≥7% undergoing revascularization showed a decreased occurrence of the primary outcome. In addition, this study also identified plasma creatinine levels as an independent predictor of outcome, with baseline LV EF demonstrating a trend toward increased risk regardless of whether the patient had revascularization. Despite the post hoc nature of the analysis, this information may be helpful in identifying patients who will likely benefit from revascularization versus medical therapy and identify when the risk of revascularization is not justified.

Interestingly, Inaba et al. [56••], in a recent meta-analysis including seven ¹⁸F-FDG PET studies, found the optimal cutoff value for the amount of viable myocardium to be 25.8% (CI = 16.6–35.0) that would lead to a survival benefit with revascularization in comparison to standard medical therapy. This was the lowest of any viability imaging method (Table 2). However, all of the data used in this meta-analysis represented observational data in which the patient and physician contributed to the decision to treat, which may have contributed to the findings. As well the recent data from D’Egidio et al. were not considered.

Cardiac PET is not yet as widely available as SPECT imaging. Also, experience in interpretation and operation may vary widely. Established institutions routinely performing PET viability imaging may have higher accuracy, standardization protocols and reporting, and confidence of interpretation rates than centers with newly or part-time instituted PET programs, thus ultimately influencing management decisions. The Ottawa-FIVE substudy of the PARR-2 trial investigated the outcomes of a site with ready accessibility to ¹⁸F-FDG supply in addition to a comprehensive and integrative clinical management approach in patients with LV dysfunction and suspected CAD [57]. Patients enrolled in the PARR-2 trial in Ottawa, Canada were randomized into two groups: ¹⁸F-FDG PET assisted management group and a standard care group. Patients with unsuitable anatomy for revascularization were excluded from the analysis. The primary event of interest was the composite of cardiac death, myocardial infarction, or cardiac hospitalization as in the main PARR 2 trial. In the ¹⁸F-FDG PET group, 10 patients of 56 (19%) experienced the primary end point compared to 22 of 55 patients (41%) in the standard care group. Adjusted survival curves are shown in Fig. 5. However, when patients in the rest of the PARR-2 (non-Ottawa sites) were analyzed, there was no significant difference in occurrence of the composite end point between the PET arm and standard treatment arm. The findings of this substudy strongly suggest that clinical outcomes may be significantly improved in centers with ready access to ¹⁸F-FDG PET imaging and experience with the methodology, in addition to an integrative approach incorporating heart failure, imaging, and revascularization teams. In Ottawa, cardiac PET has been an established method of perfusion and viability imaging with a dedicated PET scanner and internationally renowned PET research department. However, in the PARR-2 trial there were eight other recruiting sites, five of which did not have ready access to ¹⁸F-FDG PET and three of which had more limited experience in terms of prior volume. As such, these sites and their respective clinical teams may have been less accustomed to interpreting and translating PET data into practice, potentially contributing to the nonsignificant trend toward benefit of the main PARR-2 study [57]. In light of these data, quality assurance programs have been established in regions, such as the Cardiac FDG PET Registry (CADRE) in Ontario, Canada, whereby sites are trained and clinical scans are regularly reviewed to provide technical and clinical feedback to sites that are establishing new FDG PET imaging programs [58, 59].

With more data and ongoing investigation, ¹⁸F-FDG PET imaging will likely continue to be a key modality into the investigation of myocardial viability. Image quality and confidence in interpretation is high with PET imaging, and this can be attributed to many factors, such as camera technology, attenuation correction techniques, and the pathophysiology of hibernation and its effect on FDG/glucose metabolism among others. However, a very important question is how PET data, imaging, and interpretation can impact clinical outcomes in comparison to SPECT imaging. Outcome comparison studies are limited to date but a multicenter randomized trial is now underway (IMAGE-HF), which will evaluate and compare heart failure patients using ¹⁸F-FDG PET, SPECT, and CMR, and how relevant clinical decisions and outcomes are impacted [60] by the different approaches.