High-intensity focused ultrasound (HIFU) has been applied clinically as a noninvasive therapeutic tool. Atrial septostomy is a palliative treatment for pulmonary artery hypertension. The purpose of this study was to assess the feasibility of atrial septal ablation in vitro using HIFU.
Fourteen sections of atrial septum from pig hearts were treated. Focused ultrasound energy was applied with an operating frequency of 5.25 MHz at the nominal focal point intensity of 4.0 kW/cm 2 for 0.4 sec in 1-sec intervals.
Lesions were created with ultrasonic exposures ranging from 40 to 120 pulses. There were significant relationships between HIFU exposure time and lesion area on the exposed site ( R 2 = 0.3389, P < .0001) and lesion volume ( R 2 = 0.6161, P < .0001).
HIFU has the potential to create focal perforations without direct tissue contact. This method may prove useful for noninvasive atrial septostomy.
High-intensity focused ultrasound (HIFU) is a noninvasive extracorporeal therapeutic technique for thermally ablating without injuring intervening tissues. Ultrasound energy can be applied in a target volume to induce molecular agitation, absorptive heating, and thermal coagulative tissue necrosis. Multiple studies have examined the histologic changes related to HIFU ablations in the liver, kidney, prostate, breast, and brain. HIFU is being explored as a therapeutic modality in almost every tissue that is accessible by ultrasound. Previously, a novel system of delivering HIFU energy in vitro and in vivo to canine myocardial samples without direct contact with the target tissue was reported. HIFU offers several potential advantages over other therapy modalities as a technique to create focal lesions because HIFU does not apply ionizing radiation and does not require direct contact with target tissues. This thermal ablative technique has been applied to various ventricular myocardium and cardiac valve settings and as a noninvasive analog to the Cox-Maze procedure for treating atrial fibrillation. However, the use of HIFU for creating holes in atrial tissue has not been completely studied.
Pulmonary artery hypertension (PAH) is characterized by the existence of hypertensive arteriopathy in the pulmonary circulation. PAH includes primary pulmonary hypertension and secondary pulmonary hypertension due mainly to chronic pulmonary thromboembolism, collagen vascular diseases, and Eisenmenger syndrome associated with congenital heart diseases. Current medical treatment includes the administration of calcium channel blockers and long-term continuous epoprostenol (prostacyclin) therapy, although a beneficial response to vasodilators occurs in only 25% to 30% of the population, and <20% of patients respond to calcium channel antagonists. Lung transplantation will ultimately be required for many patients with idiopathic PAH. Atrial septostomy (AS) is a palliative treatment for refractory PAH. AS involves the creation of a right-to-left shunt, producing a pathway that increases left ventricular preload, and thus systemic cardiac output, while pulmonary blood pressure can be significantly decreased. It is used as a bridge to lung transplantation ; however, the immediate death rate associated with this interventional procedure is still high (5.4%–13.0%). This procedural death rate is associated with profound and refractory hypoxemia due to the massive right-to-left shunt across the interatrial septum. Also, spontaneous closure has been noted in 26% of patients receiving this treatment.
We hypothesized that HIFU would have the potential to ablate and create a therapeutic shunt on an atrial septum using thermal technology. The purpose of this study was to assess the feasibility and characteristics of in vitro atrial septum lesions using HIFU for the potential application.
Fourteen pig hearts (American Duroc breeds; mean age, 3 years) were examined. Pig hearts, vacuum packed soon after slaughter, were obtained. We dissected atrial septum specimens from 14 pig hearts and manually stretched them with the minimum force required to keep each specimen flat. The edges of the atrial septum were affixed to a rubber pad with steel pins. The rubber pad with the attached atrial septum was placed in a polyethylene container filled with normal phosphate-buffered saline solution, which was degassed using a deaerator and a dry vacuum pump. The container was immersed in a water bath maintained at 37°C with an electric heater.
HIFU energy was supplied with a signal generator (model 33250A; Agilent Technologies, Santa Clara, CA), a power amplifier (model 2100L; ENI, Rochester, NY), and a therapeutic HIFU transducer made from piezoelectric ceramic (Sonic Concepts, Bothell, WA). The transducer had an outer radius of 16.5 mm, an inner radius of 6.9 mm, a focal length of 35 mm, and a central hole that housed a 7.5-MHz diagnostic B-mode transducer (model 8663; B-K Medical, Herlev, Denmark). The diagnostic transducer was aligned to be coaxial and confocal with the HIFU transducer ( Figure 1 ). The operating frequency of the HIFU transducer was 5.25 MHz. The acoustic HIFU output power was 66 W, as determined by the radiation force on an absorber. This corresponds to a nominal spatial average intensity of 4.0 kW/cm 2 . The focal zone beam shape was measured using a pulse-echo reciprocity technique with a point target. At the half-power points, the focal zone was approximately 3.3 mm axially and 0.37 mm transversely ( Figure 1 ). The ultrasound beam focus was positioned at the midpoint of each atrial septum specimen with guidance from the diagnostic B-mode transducer. A custom software package was used to control the HIFU system.
We planned the timing of the exposures on the basis of the consideration that this method would eventually be used for a beating heart. The HIFU transducer was operated at a pulse duration of 0.4 sec with a 1.7-Hz pulse repetition frequency.
After creating holes in the atrial septum, the lesion area and depth were measured macroscopically using high-frequency diagnostic ultrasound (Vevo 770; VisualSonics, Inc., Toronto, ON, Canada). Macroscopic and echocardiographic B-mode measurements were compared by two observers. The area of the lesion on the exposed site ( A entrance ) and the reverse site ( A exit ) of the specimen was calculated as 0.785 × length × width, because of the ellipsoid shape. The lesion volume ( V ), approximating a conic trapezoid, was calculated as follows:
V = d ( A entrance + A exit + A entrance + A exit ) / 3 ,