Correlation between inertial cavitation dose and endothelial cell damage in vivo

Abstract

Previous in vivo studies have demonstrated that vascular endothelial damage can result when vessels containing gas-based microbubble ultrasound contrast agent (UCA) are exposed to MHz-frequency pulsed ultrasound (US) of sufficient pressure amplitudes, presumably as a result of inertial cavitation (IC). The hypothesis guiding this research was that IC is the primary mechanism by which the vascular endothelium (VE) is damaged when a vessel is exposed to pulsed 1-MHz frequency US in the presence of circulating UCA. The expectation was that a correlation should exist between the magnitude and duration of IC activity and the degree of VE damage. Rabbit auricular vessels were exposed in vivo to 1.17-MHz focused US of variable peak rarefaction pressure amplitude (1, 3, 6.5 or 9 MPa), using low duty factors (0.04% or 0.4%), pulse lengths of 500 or 5000 cycles, with varying treatment durations and with or without infusion of a shelled microbubble contrast agent. A broadband passive cavitation detection system was used to measure IC activity in vivo within the targeted segment of the blood vessel. The magnitude of the detected IC activity was quantified using a previously reported measure of IC dose. Endothelial damage was assessed via scanning electron microscopy image analysis. The results supported the hypothesis and demonstrate that the magnitude of the measured IC dose correlates with the degree of VE damage when UCA is present. These results have implications for therapeutic US-induced vascular occlusion. (E-mail: [email protected])

Introduction

The ability to damage selectively the vascular endothelium (VE) without causing extensive perivascular injury has several potential therapeutic applications, including selective vascular thrombosis for the treatment and prevention of bleeding, tumor therapy and drug and gene delivery (Hynynen et al 1996, Unger et al 2002, Wu et al 2002). In a previous in vivo study, the ability to damage the vascular endothelium by exposing a vessel to pulsed, low duty factor (DF), high spatial-peak, pulse-average intensity (I_SPPA) ultrasound (US) in the presence of circulating ultrasound contrast agent (UCA) was demonstrated (Hwang et al. 2005). The general hypothesis guiding that study was that inertial cavitation (IC), rather than thermal damage, was the dominant mechanism causing endothelial damage. The study demonstrated that increasing endothelial surface damage occurred with increasing peak rarefaction pressure amplitudes (PRPA) and that histology did not provide any evidence of thermal injury to the perivascular tissue. However, the detection or quantification of IC activity in vivo was not attempted because the US transducer was scanned manually along the target vessel with the ear placed on an agar surface (as opposed to being submerged in a water tank). Therefore, direct detection of intravascular IC activity was precluded in those experiments, and the hypothesized correlation between IC activity and endothelial surface damage could not be addressed.

The detection of IC activity can be performed by using a passive cavitation detection (PCD) system that incorporates a broadband US “listening” transducer (Madanshetty et al. 1991). The typical signature of IC activity is that of broadband noise detected by the “listening” transducer. Several studies have demonstrated that IC activity can be quantified using PCD systems (Chen et al 2003a, Everbach et al 1997, Poliachik et al 1999). Chen et al. (2003a) reported a method of calculating an IC dose based on the root-mean-square (RMS) amplitude, in the frequency domain, of the broadband noise within a frequency window that was between harmonic peaks (e.g., between the third and fourth harmonics). They demonstrated excellent correlation between the calculated IC dose and erythrocyte hemolysis.

The general hypothesis guiding the present experimental studies was that IC is the primary mechanism by which the VE is damaged when a vessel is exposed to low DF, high PRPA, 1-MHz frequency US in the presence of circulating UCA. If true, a correlation should exist between the magnitude and duration of IC activity and the degree of VE damage.

Three specific hypotheses were subjected to test in an in vivo rabbit auricular blood vessel model: (1) IC nucleated by gas-based microbubble contrast agent can be detected using a PCD system and a quantitative measurement of IC, given as an IC dose, will correlate with PRPA; (2) for a given PRPA, the IC dose will be greater in vessels with circulating UCA than in those without and (3) the degree of endothelial damage will correlate with the measured IC dose.