Feasibility of multiplane microtransoesophageal echocardiographic guidance in structural heart disease transcatheter interventions in adults

This was a prospective, observational unblinded study to determine the capabilities of microTEE probe imaging to guide the operator during interventional therapy in cardiac disease patients (Fig. 1).

The primary objective of this clinical study was to establish that microTEE imaging is able to visualise required anatomical features with similar image quality compared with TEE and ICE to ensure its feasibility in procedural guidance. Accuracy was tested on standard measurements of the LAA and compared between microTEE and TEE. The secondary objective was to establish that microTEE can be used without the use of general anaesthesia.

The study was performed on patients undergoing either one of the following cardiac interventions between October 2014 and November 2015: (i) LAA closure, (ii) interatrial communication closure, and (iii) atrial transseptal puncture for other purposes.

Patients were included if they fulfilled the following inclusion criteria: (i) subject is 18 years of age or older, and (ii) has given written informed consent to participate in the study. No patients were excluded. This study was approved by a medical ethics committee (NL49744.100.14).

Intraprocedural images of the LAA were recorded with microTEE and subsequently (i) standard 2D/3D TEE in patients undergoing LAA closure and a transseptal puncture for MitraClip implantation, and; (ii) ICE in patients undergoing interatrial communication closure and atrial transseptal puncture during left-sided ablations such as pulmonary vein isolation. The ICE catheter was inserted into the inferior vena cava and positioned within the right atrium. To reduce the bias of prior knowledge of anatomy, microTEE was used prior to TEE or ICE in all patients.

After anaesthesia was induced, the microTEE probe was introduced either transnasally (initial 10 patients) or transorally into the oesophagus (subsequent patients). A full set of images was obtained using standard views. Hereafter, images with standard size 2D/3D TEE probe or ICE were performed. To evaluate patient discomfort during the placement of the microTEE probe, patients undergoing local anaesthesia were asked to evaluate their comfort on a visual analogue scale from 0 (no discomfort or pain) to 10 (unbearable pain).

Two experienced echocardiographers (M.J.S., V.J.N.) independently evaluated each recording in random order. To assess image quality, each echocardiographer assigned an image quality score of 1 to 4 (1 = excellent; 2 = good; 3 = fair; 4 = poor) describing how well structures were seen on the particular recording. Additionally, to assess accuracy, each echocardiographer independently performed measurements of the LAA for appropriate device sizing [1, 9] as depicted in Fig. 2. These measurements included diameters of the ostium, landing zone, depth from the ostium needed for Amplatzer (AGA, St. Jude Medical, Minneapolis, MN, USA) closure, depth from the landing zone needed for the Watchman (Atritech, Boston Scientific, Natick, MA, USA) closure, and the total depth of the LAA. Each structure was measured at the following views: 0°, 45°, 90°, and 135°. Atrial septal defect measurements were not performed.

The S8–3t microTEE probe (Philips Healthcare) has a shaft width of 5.2 mm and length of 85 mm, a transducer tip width of 7.5 mm, height of 5.5 mm, and length of 18.5 mm (Fig. 1). The microTEE transducer is a 32-element phased array transducer that is equipped with 2D B‑mode echocardiography, colour Doppler, pulse-wave and continuous-wave Doppler, M‑mode, and colour flow M‑mode features. It has a centre frequency of 6 MHz on a bandwidth of 3.2 to 7.4 MHz.

The X7–2t 3D TEE-probe (Philips Healthcare), which has an imaging frequency range from 7 to 2 MHz, was used as the standard 2D/2D TEE probe (Fig. 1). Its head has a width, height and length of 16 × 12 × 40 mm. The shaft is 10 mm in diameter and 106 cm long. The probe contains >2,500 elements.

Both the standard 2D/3D TEE and microTEE probe provide a 180° manual image plane control with angular display, anterior and posterior articulation and an articulation brake. Both probe tip surfaces are constantly monitored for temperature to ensure patient safety.

For ICE, the Acuson AcuNav catheter (Siemens Medical Solutions USA, Inc., Mountain View, CA) was used: an 8 French, 2.5 mm, 64-element, linear phased array transducer that is equipped with a single longitudinal plane. Its imaging frequency ranges from 5.5 to 10 MHz.

The image quality scores [1‐4] and the measured LAA dimensions (mm) were presented as median (interquartile range) and mean ± standard deviation (SD), respectively. The measured diameters were provided as the mean of all four standard views by both echocardiographic reviewers. The results of the microTEE were compared with the results of the standard 2D/3D TEE probe and ICE using a paired samples t‑test and dependent Wilcoxon signed-rank test, as appropriate. Accuracy was tested on pre-specified LAA measurements, because these measurements are validated and consider a standardised and reproducible set, by using Pearson’s r correlation test and mean difference according to the Bland-Altman plot. The two-way mixed intraclass correlation coefficient was used to assess interobserver variability (M.J.S. and V.J.N.). Statistical significance was inferred at p < 0.05.

In total, 49 patients were included. There were 20 (49%) women and 29 (51%) men, with a mean age of 64 ± 18 years. Patients underwent the following interventions: LAA closure (n = 18, 37%); interatrial communication closure (n = 14, 29%); and transseptal puncture for MitraClip implantation (n = 12, 24%) or left-sided ablation (n = 5, 10%). General anaesthesia was used in 27 (55%) procedures. In those undergoing local anaesthesia, all patients had no to mild discomfort; 1 patient (5%) had moderate discomfort regarding the microTEE. In the first 10 (20%) patients, the microTEE was introduced transnasally. In one of these cases, minimal epistaxis occurred after which we decided to move to transoral introduction.

Image quality scores are provided in Tab. 1 and 2. There was no qualitative difference in the ease of passage and ability to manipulate the microTEE compared with the standard TEE probe. The standard probe rendered a better overall average quality score than the microTEE probe (1 [1, 2] vs. 2 [1–2]; p = 0.04). Although the overall average scores were significantly different, both scores were between good (2) and excellent (1) in our rating, and good visualisation was achieved using both probes. No anatomical features were missed by the microTEE. Compared with ICE imaging, the microTEE probe offered comparable image quality (2 [1–2] vs. 2 [1–2]; p = 0.13) with a marked increased field of view.

LAA measurements are provided in Tab. 3. Compared with standard 2D/3D TEE imaging, measurements seemed to be underestimated when the microTEE probe was used (Tab. 3). Bland-Altman plotting revealed a mean difference of −1.2 mm (95% CI −2.4 to 1.7 mm) for the ostium, −0.9 mm (95% CI −2.1 to 2.1 mm) for the landing zone, −0.9 mm (95% CI −2.4 to 3.0 mm) for the ostium’s depth, −1.3 mm (95% CI −3.4 to 0.9 mm) for the landing zone’s depth. The total depth of the LAA did not differ. The Pearson correlation coefficients between microTEE and standard TEE were r = 0.95 (p < 0.01) for the ostium, r = 0.94 (p < 0.01) for the landing zone, r = 0.88 (p < 0.01) for the depth of ostium, r = 0.95 (p < 0.01) for the depth of the landing zone, and r = 0.93 (p < 0.01) for the total depth. The intraclass correlation coefficient for interobserver variability was 0.92 (p < 0.01) for the ostium, 0.89 (p < 0.01) for the landing zone, 0.90 (p < 0.01) for the depth of ostium, 0.90 (p < 0.01) for the depth of landing zone, and 0.95 (p < 0.01) for the total depth.