A potential problem for the treatment

A potential problem for the treatment, at least in rats, is the substantial variation in treatment results for seemingly identical conditions. One source of variation is the variation in attenuation and scattering by the intervening tissue. When the echogenicity of the tissue (skin, intercostal space) between the chest surface and the heart was measured, substantial variation was found (Fig. 6). The variation was modestly correlated with a decreasing readout of LVE, which suggests that the ultrasonic properties of the intervening tissue are important.
Another source of variation in treatment outcome was thought to be variation in the nicotinic receptor agonist agent dose actually reaching the circulation, which was estimated by scattering measurements for ROIs within the left ventricle (Figs. 2 and 3). This pre-treatment measure of LVE, expressed in acoustical units, was proportional to the infusion rate in sham-treatment tests (Fig. 4). The LVE was reduced substantially by the therapy, with a typical agent destruction and refill pattern (Fig. 5). This identified another important agent dose parameter, which was the LVE at refill, immediately before the pulse-burst trigger. The LVE results exhibited relatively large variation, with standard deviations often 30% of the means (Fig. 7), not unlike the substantial variation in SCS results (Fig. 7). However, comparison of the LVE results with the pre-treatment and refill echogenicities to SCS results indicated that the variation in LVE was not predictive of treatment outcome (Figs. 8 and 9). Therefore, adjustment of the contrast AU readout of the LVE either before or during treatment cannot be used to reduce the variation in treatment results.
This finding was initially puzzling, because the cavitational microlesions produced by cavitation nucleation from the microbubbles should be related to their infusion. However, an important factor is the strong dependence of ultrasound scattering on microbubble size to the sixth power (Forsberg and Shi 2001), such that large bubbles would generate most of the LVE. We hypothesize that the outcome truly is related to the infusion of the optimum cavitation nuclei, but that the measured echogenicity does not indicate the concentration of these nuclei. The results of the experiment using microbubble suspensions with enhanced small- or large-microbubble populations supported this hypothesis (Figs. 10 and 11). Because the normal suspension is dominated by the small-size-range microbubbles, most of the efficacy for normal suspensions must arise from the small microbubbles, which were nearly undetectable by the LVE measurements. The large-microbubble fraction of the normal suspension, which must generate most of the LVE, therefore has only a small role in treatment efficacy. Variation in the relative population of easily destroyed large bubbles caused by mixing in a syringe or the vial, flotation, catheter constrictions, animal-dependent removal in the lungs and possibly other perturbations would cause LVE to vary, yielding the very weak relationship to efficacy, as illustrated in Figures 8 and 9.
A better approach than LVE for real-time control of MCET impact may be passive cavitation monitoring. This has been reported in vitro and in tissues (Haworth et al. 2012; Jensen et al. 2012; Salgaonkar et al. 2009). Imaging and mapping of cavitation nucleation sites can use harmonic, subharmonic, ultraharmonic and broadband emission from transient cavitation. The use of imaging probes has been reported in heart, liver and brain (Vignon et al. 2013), and transcranial 3-D imaging has been found to be feasible for monitoring blood–brain barrier disruption (O\’Reilly et al. 2014). Spatial mapping will be limited, as the left ventricular wall thickness measures only a few millimeters in rats. Moreover, cavitational emissions from the microbubble-laden blood in the ventricles could overwhelm the small signals from the myocardium. Adequate time gating and limited expectations with respect to spatial resolution would make passive cavitation mapping/monitoring an orthogonal therapy feedback in addition to the ECG. This approach to treatment management would be investigated for also providing spatial feedback when using a larger animal model for MCET.