PET Measurements of Organ Metabolism: The Devil Is in the Details Open Access

Abnormalities in myocardial fatty acid (FA) metabolism, which are primarily driven by an increase in plasma FAs delivered to the heart, play a central role in the pathogenesis of the cardiomyopathic processes associated with obesity, insulin resistance, and diabetes (1). Moreover, the prevalence and magnitude of the cardiovascular complications of these diseases, such as left ventricular hypertrophy, heart failure, and ischemic heart disease, are typically more pronounced in women compared with men (2,3). Thus, it is tempting to speculate that the sex differences in myocardial FA metabolism may be contributory to this interaction. Evidence to support this notion has come from studies using positron emission tomography (PET), which because of its inherent quantitative capabilities and diversity of metabolic radiotracers is the most commonly used imaging method to measure organ metabolism. PET studies using the FA radiotracer 1-¹¹C-palmitate have recently shown that myocardial FA uptake and oxidation are higher in obese women compared with men and that this sex disparity appears to become more pronounced when one transitions to obesity with concurrent type 2 diabetes (4,5).

However, myocardial FA metabolism is dependent on the plasma delivery of FA from nonesterified FA (NEFA) bound to albumin as well as those incorporated into triglyceride (TG), either as chylomicrons (CM-TG) or VLDLs (VLDL-TG). Fatty acids bound to TGs are made available to the myocardium via their release by lipoprotein lipase located on the capillary endothelium. Measuring this contribution of TG-extracted FAs destined for metabolism is important because it represents a major source of FAs from dietary intake (DFA) and evidence suggesting that the source of extracted FAs may impact their metabolic fate in tissue (6,7). To date, the PET method described above to measure myocardial FA metabolism relies on intravenous administration of FA radiotracers and thus only evaluates the uptake of NEFAs produced primarily by adipose tissues in communication with the blood, ignoring DFA as a source of input to peripheral organs, including the heart. Consequently, the question remains whether similar sex differences exist in myocardial DFA uptake.

In this issue, Kunach et al. (8) attempt to answer this question in patients with impaired glucose tolerance (IGT). They use a novel method developed by their group whereby PET is performed after the FA radiotracer 14(R,S)-[¹⁸F]-fluoro-6-thia-heptadecanoic acid (¹⁸F-FTHA) is incorporated into a gelatin capsule and administered orally. ¹⁸F-FTHA is an FA analog whose tissue retention reflects, in general, the combined effects of FA uptake and oxidation. Oral administration of the ¹⁸F-FTHA capsule results in the incorporation of ¹⁸F-FTHA initially into CM-TGs, which are then delivered to the heart and other tissues that take up fat (9). In the current study, patients with IGT (IGT⁺: 12 women and 8 men) and without IGT (IGT⁻: 9 women and 12 men) ingested a standard liquid meal with the ¹⁸F-FTHA capsule and then underwent PET and computed tomography (CT) imaging at 90–120 min and 360 min. Serial blood sampling was performed to assess changes in the plasma substrate environment and the source of ¹⁸F-blood activity (e.g., ¹⁸F-CM-TGs, ¹⁸F-VLDL-TGs, or ¹⁸F-NEFAs). From the PET data obtained between 90–120 min, the rate of fractional uptake of DFA by the myocardium (K_i) was measured using Patlak graphical analysis. Moreover, it was assumed all of the ¹⁸F-FTHA in the blood is incorporated into CM-TGs, and the authors calculated the net DFA uptake rate into the myocardium by multiplying the K_i by the plasma concentration of CM-TGs. Whole-body imaging was performed at 360 min to determine the relative uptake or partitioning of DFA in various tissues based on the standard uptake value, a semiquantitative measure of radiotracer uptake, in various organs of interest. The authors observed that both the K_i at 90–120 min and myocardial DFA partitioning at 6 h were higher in IGT⁺ subjects compared with their IGT⁻ counterparts, although no sex differences were noted. Similarly, net myocardial DFA uptake was higher in IGT⁺ subjects compared with IGT⁻ subjects, but men exhibited higher net DFA uptake compared with women.

The potential implications of the study are significant. If correct, they suggest that, at least in the presence of IGT, the male heart may be more dependent on dietary sources of FAs, particularly CM-TGs, whereas the female heart is likely more dependent FAs arising from adipose stores. In addition, and consistent with the PET studies that use intravenously administered FA radiotracers, the results suggest the plasma levels of these FAs are the major determinants of the observed sex differences as opposed to differences in intracellular processes responsible for FA uptake (4,5).

The authors should be complemented for implementing a novel method to measure myocardial FA metabolism from FAs derived from dietary sources to address a fundamental and important question. However, as with any new method, some caution is always warranted before one can draw definitive conclusions based on its application. In PET, the underlying biological process is determined by analyzing the input(s), representing the delivery of material from blood, and measuring the tissue response, representing, for example, a metabolic process. With that in mind, the general scheme in quantification of PET data is to derive the radiolabeled metabolite-corrected input function, which is then used in conjunction with a mathematical model, or variations thereof, to quantify biological/physiological processes by optimizing the model against PET-measured tissue response data, ideally corrected for geometric considerations such as partial volume and spillover effects. The measured tissue response in PET imaging reflects the sum total of radioactivity from all radiolabeled species in the tissue of interest. As such, the contribution of individual species, and metabolic rates thereof, must be determined by using complex mathematical models of radiotracer kinetics. This is especially critical when, in this case, the parent radiotracer ¹⁸F-FTHA is metabolized to produce radiolabeled species ¹⁸F-CM-TGs and ¹⁸F-VLDL-TGs, each of which is a source of input to peripheral tissues.

Thus, in the context of the work by Kunach et al. (8), there are three potential sources of input to the heart reflecting contribution from ¹⁸F-FTHA (representing ¹⁸F-NEFAs), ¹⁸F-CM-TGs, and ¹⁸F-VLDL-TGs in a complex dynamic kinetics between blood and heart tissue (Fig. 1). One way to mitigate this complexity is to assume that one source of input is dominant and constant over the time course of the imaging study. In the current study, the authors assume that ¹⁸F-CM-TGs account for the majority of the ¹⁸F radioactivity in blood from 90–120 min. However, this may not have been the case. For example, during the 90–120 min interval, it appears plasma ¹⁸F activity reflected not just the ¹⁸F-CM-TG concentration but a significant contribution from ¹⁸F-NEFAs with a lesser contribution from ¹⁸F-VLDL-TGs and with the relative proportions depending upon sex and IGT status (as shown in Supplementary Fig. 1 [8]). As a consequence, the myocardial tissue response cannot be solely attributed to ¹⁸F-CM-TG uptake and may, in fact, represent total lipid uptake.

Importantly, the authors resort to a simplified Patlak analysis approach to derive measures of K_i constant. In Patlak analysis, the linear segment of plot depicting the ratio of the radiolabeled metabolite concentration in tissue to the plasma concentration at the respective times versus the ratio of the plasma concentration–time integral of the metabolite species to plasma concentration at a given time yields the K_i (mL/g/min) constant (10). The product of K_i with the concentration of a metabolite of interest in plasma yields the net uptake constant (K_m; nmol/g/min). In light of the arguments for potentially three sources of input to the myocardium, the reported myocardial K_i may represent an integration of the individual K_i values for ¹⁸F-FTHA provided from each of the three lipid species and not just chylomicrons. Moreover, as ¹⁸F-NEFAs, ¹⁸F-CM-TGs, and ¹⁸F-VLDL-TGs likely exhibit different uptake and metabolism kinetics, the timing of steady state or equilibrium of each species relative to the concentration of a given species in blood may not be the same, suggesting that individual K_i values may not be additive unless their measurement and timing for determination have been validated a priori. Accordingly, given that we do not know individual K_i values, one cannot determine the myocardial net DFA uptake value (K_m) for a single lipid species, such as CM-TGs, and compare values between groups. Finally, ascertaining if the net myocardial DFA flux differs between men and women with IGT requires multiplying the K_i by total lipid levels representing all three species ideally measured during the 90–120-min time interval for determination of K_i. Thus, the multiple sources of input to the heart coupled with the complexity of the kinetics underscore the need for additional studies to facilitate the interpretation of metabolic imaging studies based on DFA.

Overall, the methodology developed by Kunach et al. (8) for determining the contribution of DFA to tissue uptake must be applauded. It opens a new dimension in quantification of FA metabolism and sources of FA uptake and metabolism beyond traditional methods of intravenous administration of radiotracers. However, there remain technical challenges in quantifying the response from the resultant multiple sources of radiolabeled metabolites contributing the signal in PET imaging. The devil is always in the details.