The volume of tracer distribution, V_T, is measurable with PET (Perrier and Mayersohn, 1982). Assuming a compartmental tissue model for a PET voxel (Perrier and Mayersohn, 1982; Alpert et al., 2018), and the volume of distribution can be written as

where the experimental endpoints are CPET¯, the steady-state tissue concentration of ¹⁸F-TPP⁺ measured by PET, Cp¯, the steady-state plasma concentration, f_ecs, the tissue volume fraction occupied by the ECS, and f_mito the tissue volume fraction occupied by mitochondria. β=zFRT is a ratio of known physical constants – F is Faraday’s constant, z denotes the valence, R denotes the universal gas constant, and T symbolizes the temperature in degrees Kelvin. f_ecs is measured using ECG-triggered CT scans before and after administration of iodinated contrast media (see below). f_mito was set to 0.2568 mL/mL (Barth et al., 1992). See Alpert et al. (2018) for more details. There are several ways to measure V_T. One way is to use a bolus injection of ¹⁸F-TPP⁺ and a dynamic PET acquisition protocol, followed by kinetic analysis of the measured activity concentration data (Alpert et al., 2018). Administering the tracer with intravenous bolus injection followed immediately by prolonged constant infusion allows an alternative approach requiring establishment of a secular equilibrium during which V_T can be measured as a simple tissue-to-plasma concentration ratio (Endres and Carson, 1998; Carson, 2000).

Animals used in this study were housed and maintained under the supervision of the Massachusetts General Hospital (MGH), Institutional Animal Care and Use Committee (IACUC), and our study was conducted under a protocol approved by the MGH IACUC.

Adult Yucatan minipigs were anesthetized with intramuscular telazol (4.4 mg/kg) and xylazine (2 mg/kg), tracheally intubated, and mechanically ventilated. Anesthesia was maintained with a mixture of isoflurane (1.5%) and oxygen. Once intubated and under general anesthesia, swine were positioned supine on the procedure table with a heating pad underneath. All procedures were performed using aseptic technique. Bilateral auricular dorsolateral veins were cannulated with a 22G angiocatheter for administration of IV fluids and medications. Heparin (150 units/kg) was administered intravenously prior to introducer sheath placement and additional heparin administered as needed to maintain activated clotting time (ACT) > 200 s. The femoral veins were cannulated percutaneously with 8F introducer sheaths for administration of TPP, CT, contrast, and venous blood sampling. The left femoral artery was cannulated percutaneously with a 6F introducer to monitor arterial blood pressure and blood gasses. A 7F introducer sheath was placed in the right femoral artery for introduction of a guide catheter. The 7F guide catheter was inserted through the introducer and advanced into the ostium of the main left coronary artery under fluoroscopic guidance. Angiographic images of the coronary artery were obtained from the standard 25 degree left anterior oblique view with a 10 degree cranial angle. Intracoronary BAM15 drug delivery was accomplished with a 0.92 mm over the wire Maverick Angioplastic PTCA catheter (Boston Scientific). The drug delivery catheter was advanced over a 0.36 mm guide wire through the 7F guide catheter. Under fluoroscopy, radio-opaque marker bands located at the tip of the catheter were used to position the catheter tip in the mid-left anterior descending (LAD) artery. Once drug delivery catheter position was confirmed by fluoroscopy, the 0.36 mm guide wire was withdrawn and the lumen used for intracoronary BAM15 delivery. Both catheters were anchored and fixed externally with sutures to avoid displacement during imaging.

Vascular access for cardiac hemodynamic monitoring was achieved via a 2 cm midline neck incision over the right side of the trachea. The right common carotid artery was identified and an 8F introducer sheath inserted via a guide wire into the lumen of the artery. Two surgical ties were positioned on either side of the introducer sheath for bleeding control. Under fluoroscopic observation, the 5F Transonic Scisense catheter was inserted through the 8FR introducer, advanced across the aortic valve via the aortic root, and placed into the left ventricular chamber.

An intravenous bolus injection of ∼16 mCi of ¹⁸F-TPP⁺ followed immediately with an infusion of ¹⁸F-TPP⁺ (∼6 mCi in 60 mL, 0.33 mL/min over 180 min) were used for tracer administration. PET list mode data were acquired for 150 min beginning 30 min after the start of ¹⁸F-TPP⁺ bolus/infusion using a hybrid PET/CT system (GE Discovery MI; Pan et al., 2019). List mode data were framed as a dynamic series of 2 min image volumes. PET data were reconstructed using an OSEM algorithm with CT-based attenuation correction to yield radioactivity concentration maps in units of Bq/cc with 89 slices and a voxel size of 2.73 mm × 2.73 mm × 2.8 mm. Reconstructed PET image volumes were reoriented into the standard short axis projection, segmented as standard 17 segment polar maps according to the ACNC guidelines (Dilsizian et al., 2016), and used to form time activity curves (TACs). These TACs were analyzed during the secular equilibrium and inspected for evidence of depolarization during infusion of BAM15.

Venous blood samples were drawn from the femoral vein in pigs 3 and 4 to estimate the ¹⁸F-TPP⁺ concentration during secular equilibrium. Beginning at 90 min after tracer administration and ending 30 min later, samples were drawn every 6 min.

Myocardial f_ECS was measured with a contrast-enhanced cardiac CT protocol consisting of three steps: First, a CT scan was performed to obtain baseline blood and myocardial attenuation in Hounsfield units (HU); second, a bolus of iodine contrast agent was administered intravenously (1.8 mgI/kg; 3 mL/sec) using a power injector followed by a saline flush; third, a repeat CT scan was performed to measure blood and myocardial HU after a standard 15-min delay, required for the contrast agent to reach equilibrium between myocardium and blood pools (Treibel et al., 2015; Lee et al., 2016). All cardiac CTs were acquired with prospective ECG gating (∼65–75% of R-R interval) and the following parameters: helical mode, tube voltage: 120 kVp; tube current: 200 mAs; gantry rotation time: 330 msec, matrix size: 512 × 512, voxel size: 0.70 mm × 0.70 mm × 5mm, and 100 slices. Non-local means filtering was applied to the reconstructed images to increase the image signal to noise ratio (Buades et al., 2005). The filtering parameters were: local patch size 3 × 3 × 3, search window 5 × 5 × 3. Next, the CT images were registered to the PET image (summation of late frames) to account for potential position mismatch between the CT and PET acquisitions. The pre- and post-contrast CT images were first rigidly aligned to the PET image, followed by elastic registration of the myocardium to refine alignment. The resulting deformation field was applied to both pre- and post-contrast CT images. All rigid and non-rigid registrations were performed using the open source software elastix (Klein et al., 2009).

Iodine contrast induces more absorption and scattering of x-ray, which results in increases in CT attenuation. In heart, intravascular contrast rapidly permeates the ECS. The ECS fraction was calculated as fECS=ΔHUMΔHUB×(1-Hct) where ΔHU_M and ΔHU_B represent the change in attenuation before and after contrast in the myocardium and in the blood, respectively, (ΔHU = ΔHU_{post−contrast}−ΔHU_{pre−contrast}), and Hct represents the hematocrit.

BAM15 was prepared with 3% DMSO in 10% P80. BAM15 dose response with respect to PET imaging was assessed in two minipigs: Beginning 50 min post injection/infusion of ¹⁸F-TPP⁺, BAM15 was infused into the LAD coronary artery and the dose was increased in three levels. PET data were reconstructed as 75 frames of 2 min each, reoriented into the standard short axis projection, segmented as standard 17 segment polar maps, and used to form TACs. These TACs were inspected for evidence of depolarization. Visualization of distinct local minima in the TACs following BAM15 infusion was considered evidence of depolarization. The first dose-response experiment began with very conservative BAM15 dose while the second study used higher BAM15 doses to demonstrate that an effect could be observed. Bam15 dosing is summarized in Table 1. Global values of myocardial ¹⁸F-TPP⁺ concentrations were used as the endpoint for assessing if changes could be attributed to BAM15. Blood sampling was not performed in the dose-response studies.

In pigs 3 and 4, the distribution of ¹⁸F-TPP⁺ was allowed to evolve for 100 min under constant infusion, leading to a secular equilibrium of ¹⁸F-TPP⁺ in the myocardium. No BAM15 was administered during this period. Data obtained in the epoch 95–100 min was used to measure VT=CT¯Cp¯ and to estimate ΔΨ_T. Infusion of BAM15 commenced at 100 min and continued until 125 min. In those cardiac segments showing a clear decrease in TPP⁺ concentration, we used V_T ≈ f_m⋅(1−f_ECS)e^−βΔΨ_T as an approximation of Eq. 1 to estimate the value of V_T and ΔΨ_T in the period of lowest excursion, as well as for computation of parametric maps of ΔΨ_T.

As shown in Figure 1A, during constant intravenous infusion of ¹⁸F-TPP⁺, the myocardial ¹⁸F-TPP⁺ concentration rose nearly linearly between 40 and 95 min while the BAM15 infusion rate was increased ten-fold, from 0.01 mg/min (total 0.15 mg) to 0.1 mg/min (total 1.5 mg), with no measurable effects of BAM15. At 95 min the BAM15 dose was increased to 0.5 mg/min (total 7.5 mg) and the PET concentration decreased about 20 min after the beginning of BAM15 infusion.

Figure 1B shows the average myocardial TAC of the second dosing study. The BAM15 infusion rate was set at 1 mg/min and infused into the LAD artery over a 25 min period beginning at 100 min after the PET infusion began. The global ¹⁸F-TPP⁺ myocardial concentration immediately decreased and a new secular equilibrium was observed in the period 110–120 min. At the end of the first BAM15 infusion the concentration of ¹⁸F-TPP⁺ resumed a roughly linear increase until 155 min when the second BAM15 infusion started. The second infusion of 1.2 mg/min was terminated after 10 min due to cardiac dysfunction. The cumulative BAM15 dose before end of study was 37 mg.

Figure 2 demonstrates that BAM15 infusion into the mid LAD artery causes a decrease in the concentration of ¹⁸F-TPP⁺ in the corresponding vascular territory. The top panel presents TACs for apical anterior and mid anteroseptal myocardium, segments 8 and 14 of the standard 17 segment polar model. Shortly after start of BAM15 infusion into the LAD, concentration of ¹⁸F-TPP⁺ decreased, and remained depressed for about 10 min before beginning an upward trajectory. A second BAM15 infusion, begun 25 min later, demonstrated a much weaker response in these segments. Considering the whole myocardium, the response to BAM15 varied from segment to segment; greatest in apical segment 14 and least in segments 10 and 11 which are not parts of the LAD territory. These results are consistent with the placement of the catheter midway in the LAD artery and support the idea that it is possible to monitor, in real time, changes in ¹⁸F-TPP⁺ concentration, as an indicator of mitochondrial membrane depolarization.

Figure 3 provides a more detailed quantitative analysis, comparing data obtained at secular equilibrium with and without BAM15. Parametric images of V_T and ΔΨ_T, are shown in the vertical long axis projection, allowing examination of effects in apex and septum. Color/intensity bars on the right side of Figure 3 provide a quantitative index of the parameter values. Before the infusion of BAM15 the parametric images of V_T and ΔΨ_T are similar to those obtained in our earlier study (Alpert et al., 2018), with values of V_T about 23 mL/mL and ΔΨ_T near -133 mV. Comparison with the “after BAM15” images shows that V_T decreased to about 18 mL/mL and ΔΨ_T fell to about -119 mV. In other word, assuming that BAM15 did not affect the cellular membrane, there was about 10 mV depolarization of mitochondrial membrane potential attributable to BAM15 uncoupling of the proton current.