Human stenotic aortic valves were obtained from a leftover of valve surgery replacement after approval from the Ethical Committee at Centro Cardiologico Monzino for the purpose of the present study. Porcine pericardia were collected from euthanized adult pigs in the experimental piggery facility and the slaughterhouse of the Department of Veterinary Medical Sciences of the University of Bologna. The pig in vivo trial was performed at an external facility authorized for the execution of cardiothoracic surgery trials. Animals were treated according to the ethical guidelines approved by the Italian Ministry of Health.

The TDD used in the present study was an evolution of the device we preliminarily tested (19). The theoretical considerations underlying the use of shockwaves is provided for interested readers in the supplementary information. Briefly, the TDD comprised a pulse wave generator that was designed to provide two narrow impulsive electrical signals ranging from 50 to 75V. For the present study, one electrical signal was emitted pulses at 100 kHz and the other emitted at 3MHz to the ablation unit in an alternating pattern, with a time interval of 6 s. The combination of these different frequencies improved the disruptive effects of calcium deposits in the aortic valve cusps, avoiding thermal injury and the breaking of the transducers. The ablation unit was based on two mechanically bound piezoceramic transducers produced by PI Ceramics that had dimensions of 2.7 mm × 8 mm ×.7 mm. Each transducer was electrically connected to a Kapton flexible printed circuit and housed in metal support. The metal support was engineered to create a backing effect on the ultrasounds emitted by the transducer as the waves are conveyed in the direction of treatment. To obtain this effect, the thickness of the wall where the transducer was anchored is 34λ. The transducer must be electrically isolated to work in a biological environment. The insulating material was a variant of parylene. It is deposited with a thickness of λ4 that facilitates the passage of ultrasonic waves as the effects of refraction and reflection are limited. The debridement device used for the execution of the tests that is reported in this article consisted of a clamp whose faces are constituted by two piezoceramic transducers, but the final version of the device also included an artificial temporary valve with a Nitinol support structure intended to be positioned within the native valve to keep it open during the positioning of the transducers. Figure 1 illustrates the main constructive elements of the TDD and its experimental setup for the removal of calcium deposits in the soft valve tissue. The calcific leaflets were continuously treated for 30 min at alternating frequencies of 3 MHz and 100 KHz, with a switch between the two frequencies every 6 s. The combination of two-frequency ultrasound fields, as indicated in the literature (REF), increases the cavitation effects compared to the use of a single frequency. Hence, it is likely to preserve the integrity of the transducer and reduce the heat emitted.

Immediately after gentle isolation of the parietal pericardial membrane, the tissue was immersed in a solution of Phosphate-buffered saline (PBS)-containing antibiotics (penicillin/streptomycin) and stored or shipped at 4°C until the beginning of the experiments. Before the treatment with shockwaves, the remaining fat tissue was removed from the pericardial flap under sterile conditions. Subsequently, it was divided into stripes with a sterile knife and further cut into three samples/stripes. TDD treatment was localized at the center of the samples after marking with a surgical pen for discriminating the treated vs. the untreated portions in subsequent analyses.

Candidate valves for TDD testing were selected on the basis of the degree of calcification as detected by echocardiographic characteristics before the valve substitution intervention. In general, valves with low or moderate stenosis levels were selected due to the inability of our experimental TDD to ablate nodules with a volume smaller than 100 mm³. The system was mounted with the two piezoelectric transducers clamping the portion of the tissue to be treated before connecting to the system amplifier (Figure 1).

The experimental treatment was carried out in two adult pigs in a fully equipped operating room with a portable angiographic C-arm. Under general anesthesia, the animals were fully heparinized, and the femoral artery was exposed to perform the treatment with TDD using a minimally invasive procedure. The animals were kept anesthetized for the whole duration of the treatments and were immediately euthanized after the procedure to recover the heart and dissect the treated valves. The mean arterial pressure was stable during the entire procedure (mean arterial pressure of 80 mmHg with heart rate of ~130 beats per min). No moderate to severe aortic regurgitation was ever detected by echocardiographic monitoring during and after TDD treatment. No pacing or rapid pacing was needed for the treatment.

Human and pig valve samples were fixed in 4% paraformaldehyde (4°C overnight), dehydrated in alcoholic scale, and embedded in paraffin. Histological sections (5 μm in thickness) were cut, dewaxed, and stained. Von Kossa staining was performed to evaluate calcium deposits in untreated vs. treated portions of the specimens. Conventional staining (hematoxylin/eosin, Masson's trichrome) was performed to observe in greater detail the structure of the tissue. Images of the tissue sections were acquired using an AxioPlane optical microscope (Carl Zeiss) and analyzed with ZenBlue software.

To preliminarily test the efficiency of calcium deposits debridement in human aortic valves, experiments were performed on explanted and formaldehyde-fixed aortic valve leaflets. Under these conditions, we obtained an average volume reduction of 13.87% with a single piezo and a supply voltage of 55V, which allowed to obtain a peak of acoustic pressure of .2Mpa. The emitted energy is 150 mj/mm² in the center of the transducer and 80 mj/mm² at the edges (the shape of the emitted field is pyramidal). Volume reduction was calculated from the post-treatment vs. pre-treatment leaflet by CT scan (Supplementary Figure S1).

Since application of shockwaves to tissues can determine mechanical damages to cells, we were prompted to assess the effects on cellular survival and tissue ruptures using our experimental setup. This was done by measuring the vitality of living pericardial membranes from pigs, a tissue that we already employed for engineering valve tissues after recellularization (20) and is the golden standard material for manufacturing valve replacements. As shown in Figures 2A,B, the samples, stained with 3-(4.5-dimethylthiazol-2-yl)2,5-diphenyltetrazolium bromide (MTT) and treated with the device, exhibited no differences in the overall vitality of the treated vs. the non-treated areas on both sides of the tissue. The absence of large ruptures and the presence of cells with normal morphology in transversal sections of the treated and untreated areas showed that the type and intensity of the shockwaves delivered by the TDD did not affect, at least macroscopically, the integrity of the tissue (Figure 2C). Confirming the macroscopic observations by MTT staining, a normal amount of cells in the treated vs. the non-treated areas were counted in high magnifications of transversal tissue sections of the pericardial material (Figures 2C,D). This bona fide consolidates the lack of important structural deterioration of a valve-resembling living tissue treated with the device and warrants the safety of the procedure for employment in vivo.

To validate the double-piezo device for fragmentation and/or disruption of the calcium deposits in living calcified valves, we employed the TDD to treat calcified areas of human stenotic valves leaflets that were freshly explanted from patients with CAVD who were undergoing surgical valve replacement (Figure 3A). Treatments were performed using a 3 MHz ultrasound field combined with a field at a low frequency of 100 kHz, which, according to preliminary tests, were found to ensure the most efficient removal of the nodules of intermediate or low size (up to 100 mm³). To this aim, we selected areas of the leaflets that were not characterized by the largest calcific deposits or affected by sclerosis, where, instead, calcium deposition is not involved (21). Leaflets were treated in areas juxtaposed to the visible calcifications with an orientation, as indicated in Figure 3B. From the morphological observation of the leaflets after treatment, the tissues, in no case, showed signs of burns or obvious damages. Histology staining was performed after transversally cutting the treated areas to analyze possible microscopic ruptures to the extracellular matrix components (e.g., collagen and elastic fibers) and variations in the dimensions of the calcium nodules. Figures 3C–F show representative low or high magnifications of calcifications in control (Figures 3C,E) and treated (Figures 3C,F, Supplementary Figure S2) leaflet sections stained with von Kossa. In treated leaflets, it was evident that a partial fragmentation or disruption of the calcium deposits appeared less compact than in untreated samples, as shown by the reduction of the area covered by the compact bone, and the presence of areas where the calcium appeared pulverized. The elimination of calcium deposits was also macroscopically observed by treating a calcified human valve in a cadaveric heart (Supplementary Figure S2). Together, these data macroscopically and microscopically demonstrate the removal of calcium nodules from the calcific leaflets by TDD.

A pigtail catheter, under fluoroscopic guidance, was advanced through a 6 Fr arterial introducer and over a 0.035″ standard wire until the right aortic valve sinus of Valsalva to mark aortic valve plane and leaflet. The TDD, integrated in a transfemoral 21 Fr delivery system, was advanced over a 0.035″ stiff wire through the thoracic aorta and aortic arch to the aortic root (Figure 4A, left center, Supplementary Video 1). An angiography was performed to highlight the correspondence between the position of the device and the aortic valve plane (Figure 4A, right, Supplementary Video 1). After opening the device, under intravascular ultrasound control (Figure 4B), the active element was correctly positioned by laying on the aortic side of the leaflet. The device only kept in closed position at which the leaflet was lying on, allowing the two other leaflets to normally open and close for the whole duration of the treatment (Figure 4B, center and right) and after retraction of the TDD from the valve (Figure 4C and Supplementary Video 1). It was interesting to note that during the activation of the piezo, the TDD ablator interfered with the echo signal. While this suggests an efficient in vivo functioning of the device, it also shows that echocardiographic assessment of leaflet motion during ultrasound treatment is not possible, thus leaving to angiographic monitoring the elective way to control the positioning of the piezo during the procedure. Every 10 min, the treatment was interrupted to check the status of the electrical insulation of the device. The check was carried out by pinching the electrical cables connected to the generator, and the piezoelectric capacity was measured with a tester.

After the end of the experiments, the valves were explanted and processed for histological sectioning. Given that the entire valves were available, the three leaflets were individually dissected from the Valsalva sinuses and transversally sectioned in a circumferential direction from commissure to commissure to screen for potential tissue damage. Figures 5a–c show the picture of the right, left, and non-coronary leaflets, respectively, of one of the valves treated by the TDD in vivo. To identify possible tissue damages, staining with Masson's and Hematoxylin/Eosin (H&E) solutions were performed. The magnifications in each of the panels indicate areas with possible tissue damage due to the treatment. As expected, evident signs of mild/moderate ruptures on the aortic side of the right leaflet (Figure 5a), the one that received the ultrasounds according to echocardiography data. These ruptures, however, were found only on the most external surface of the tissue and did not affect the collagenous structure of the fibrosa layer, suggesting that the shockwaves were very focused on the area of the tissue localized immediately below the contact point of the piezoelectric transducer to the aortic side of the leaflet. This finding, in keeping with the data obtained in animal pericardium and the human calcific leaflets, shows a localized effect of the TDD that is tailored to fracture the calcific lesions without involving major tissue ruptures.