Author + information
- Received April 27, 2010
- Revision received July 28, 2010
- Accepted August 6, 2010
- Published online January 1, 2011.
- Haïfa Mahjoub, MD⁎,
- Stéphane Noble, MD⁎,
- Réda Ibrahim, MD⁎,
- Jeannot Potvin, MD⁎,
- Eileen O'Meara, MD⁎,
- Annie Dore, MD⁎,
- François Marcotte, MD⁎,
- Jacques Crépeau, MD⁎,
- Raoul Bonan, MD⁎,
- Asmaa Mansour, MSc‡,
- Denis Bouchard, MD†,
- Anique Ducharme, MD, MSc⁎ and
- Arsène J. Basmadjian, MD, MSc⁎,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Arsène J. Basmadjian, Department of Medicine, Montreal Heart Institute, 5000 Belanger Street, Montreal, Quebec H1T 1C8, Canada
Objectives This study sought to describe and compare a novel fluoroscopic method and a 2-dimensional transesophageal echocardiographic (TEE) method to localize mitral periprosthetic leaks (PPLs) for transcatheter reduction.
Background Transcatheter reduction of significant regurgitation represents a modern and attractive alternative to surgery for the treatment of mitral PPL in high-risk patients. Accurate localization and precise communication between the echocardiographer and the interventional cardiologist are essential for procedural success.
Methods We analyzed TEE and fluoroscopic studies of patients with mitral PPL who underwent multiplane 2-dimensional TEE–guided transcatheter reduction in our institution. Periprosthetic leaks were routinely localized using the “surgeon's-view” time-clock method during periprocedural TEE assessments. The 2-dimensional TEE examinations were later retrospectively reviewed by an echocardiographer blinded to procedural TEE findings. A corresponding surgeon's-view time-clock method was plotted for fluoroscopic PPL localization. Using this fluoroscopic method, offline fluoroscopic images were reviewed by an independent interventional cardiologist blinded to TEE results. Agreement between methods was evaluated.
Results Complete imaging data were available for analysis in 20 patients who, between 2002 and 2009, underwent transcatheter reduction in which the defect was successfully crossed. There was excellent agreement between procedural TEE and retrospective TEE review for PPL localization (100%; p < 0.0001) and between fluoroscopic and procedural TEE localization (90%; 95% confidence interval [CI]: 77% to 100%; p = 0.0003). In the 2 cases where there was disagreement, fluoroscopic PPL localization was adjacent to TEE localization.
Conclusions The surgeon's-view time-clock method of localizing PPL using 2-dimensional TEE is highly reproducible and allows fluoroscopic localization using the same reference system with very good agreement.
- mitral valve
- percutaneous approach
- periprosthetic leak/paravalvular leak
- prosthetic valve
- transcatheter reduction/closure
- transesophageal echocardiography
The incidence of significant periprosthetic leak (PPL) after mitral valve replacement has been shown to vary from 2.5% to 22%, depending on patient population and regurgitation severity (1–3). Although many of these leaks are small and insignificant, some are responsible for serious clinical complications (4,5), which are difficult to treat medically. Indeed, some patients with PPL exhibit sustained hemolysis, with severe anemia requiring repeated blood transfusions (6). When the leak is large, it may be associated with volume overload leading to refractory cardiac failure. Improved imaging techniques with transesophageal echocardiography (TEE) have lead to increased detection of these leaks (7,8). Surgical repair of PPL has traditionally been the standard of care (9), and in recent years, percutaneous leak reduction procedures have been reported with the off-label use of various types of occluder devices. These new transcatheter procedures provide an attractive alternative for high surgical risk patients, with initial encouraging results (10–16).
One of the main limitations encountered during these types of interventional procedures remains the capacity to precisely locate and cross the leak-causing defect; hence, there is an important role for both fluoroscopic and TEE guidance. To allow adequate localization of PPL and successful transcatheter reduction, a common reference system for these 2 imaging modalities is necessary for precise communication between the echocardiographer and the interventional cardiologist.
A method using multiplane 2-dimensional (2D) TEE has been described to accurately localize mitral periprosthetic defects based on a ”surgeon's-view” time-clock method (SVTCM) (17,18). This method, used for pre-operative assessment of PPL and guidance of surgical repair, has shown excellent correlations with surgical findings. Whether a similar system could be used in conjunction with fluoroscopic evaluation for transcatheter reduction has never been tested, to our knowledge. Hence, the objective of this study is to describe and compare a novel fluoroscopic SVTCM with the 2D-TEE method to localize mitral PPLs for transcatheter leak reduction.
This study was approved by the Research and Ethics Committees of our institution. Complete TEE and fluoroscopic data were available for comparison in 20 patients who underwent transcatheter mitral PPL reduction procedures at the Montreal Heart Institute from June 2002 to October 2009. All procedures were performed under general anesthesia. Fluoroscopic data were considered suitable for analysis if a left anterior oblique with or without caudal view showed a guidewire, balloon, or device crossing the paravalvular leak. In this study, TEE was performed by experienced echocardiographers, using GE Vivid 7 (GE Healthcare, Milwaukee, Wisconsin) or Philips Sonos 5500 (Philips Medical Systems, Andover, Massachusetts) imaging systems, in all patients before and during procedures. During these TEE examinations, PPLs were prospectively localized using the SVTCM. Images were stored in digital format or on VHS videotapes.
All TEE images were reviewed by an independent echocardiographer, blinded to clinical, echocardiographic, and fluoroscopic findings, in order to assess interobserver reproducibility of PPL localization. If there was more than 1 leak, only the largest defect was reported for analysis because it corresponds to the defect that was addressed for transcatheter reduction. Fluoroscopic images showing a guidewire, a catheter, or a device crossing the PPL defect were reviewed by an interventional cardiologist, blinded to TEE and fluoroscopic findings, in order to assess the fluoroscopic localization of the PPL. Agreement with TEE using the fluoroscopic SVTCM was evaluated.
TEE localization method
Localization of PPL using multiplane 2D-TEE can be reported in reference to neighboring anatomical landmarks, such as the atrial septum, the aortic valve, or the left atrial appendage (LAA). The method used in this study provides accurate anatomic localization, based on the polar coordinate system of a surgeon's-view time clock (Fig. 1). By convention, the LAA was set at 9:00, because this structure or its remnant can almost always be identified surgically when looking through the left atrium. Using anatomic landmarks, the entire prosthetic ring could be screened and PPL localized according to the TEE degrees at which the leak is found, depending on whether the defect is on the same side of the landmark or on the opposite region. For example, at 0° (4-chamber view) using the TEE anatomical view display (left atrium up), the medial or left side of the screen, near the atrial septum, corresponds to 0:00/12:00 on the surgeon's view and the lateral or right side on the screen corresponds to 6:00. At 90° (2-chamber view), the left side of the screen is at 3:00, whereas the right side corresponds to 9:00, adjacent to the LAA. At 150° (3-chamber view), the aortic valve corresponds to 11:00, whereas a defect occurring on the left side of the display, opposite to the aortic valve, would correspond to 5:00 (Figs. 1 and 2). Periprosthetic leak was reported as a single hour if the defect was very localized or, more commonly, a range of hours if the defect was larger.
Fluoroscopic localization method
The projection views to perform the procedure were typically the right anterior oblique, and the left anterior oblique with or without caudal angulations, the latter being the most helpful to localize the paravalvular leak. As no angiographic contrast was used during the procedures, fluoroscopic offline localization of PPL was possible only if a wire, a catheter, a sizing balloon, or the occluder device was positioned across the leak. From the available images for offline analysis, an independent interventional cardiologist, blinded to TEE results, visually assessed PPL localization based on sizing balloon or device position, over a range of hours, using a fluoroscopic SVTCM (Figs. 1 and 2).
In the left anterior oblique view (at 30°) with or without caudal view, the mitral prosthetic ring is visualized en face as a mirror image of the surgeon's view with 0:00/12:00 and 6:00 in the vertical axis (12:00 at the upper position), and 3:00 and 9:00 in the transverse axis with 3:00 on the left of the display screen or septal side, and 9:00 on the right or LAA side. In general, most patients with valvular disease have landmarks such as aortic valve calcifications or a prosthetic aortic valve, a tricuspid annuloplasty, or a pacemaker wire, which give supplemental indices in localizing PPL. Figures 3 and 4 show examples of PPL localization by TEE and fluoroscopy in Patient #3 and Patient #20, respectively.
Baseline data are presented as mean ± SD, median (interquartile range), or number of patients (percentage). The TEE and fluoroscopic examinations were compared on a 1-h section-by-section basis (12 sections for each patient for each imaging modality) and on a patient-by-patient basis. It was ruled that there was agreement between methods when there was overlap of 1 h or more for PPL localization on the SVTCM between the original prospective TEE and the offline TEE review as well as between original prospective TEE and offline fluoroscopic localizations. It was ruled that there was disagreement when there was no overlap for PPL localization between methods. For instance, if the TEE localization was 7:00 to 11:00 and the fluoroscopic localization was 10:00 to 12:00, there was agreement (for patient-by-patient analysis) between methods because of an overlap of 1 h from 10:00 to 11:00 (Patient #1). Proportion of agreement between echocardiographic interpretations for the TEE method and proportion of agreement between TEE and fluoroscopic examinations were provided with a 2-sided 95% confidence interval (CI), when applicable. The 2 proportions of agreement were also tested against the null hypothesis of proportion of agreement of 50% due to chance, using a binomial test statistic.
The baseline characteristics of the 20 patients with complete imaging data who underwent transcatheter procedures between June 2002 and October 2009 and in whom the defect was successfully crossed are listed in Table 1. All patients except 1 (Patient #15) had mechanical mitral valve prostheses: 16 bi-leaflet and 3 single tilting-disc valves (Patients #4, #18, and #20). In all patients, the defect approached for reduction was responsible for significant mitral regurgitation (grade 3 or 4). The indication for PPL closure was heart failure in 12 patients (60%), severe hemolysis in 1 patient (5%), and both in 7 patients (35%). Many of the 20 patients had multiple previous cardiac operations (median: 2, range 1 to 5). Most had significant comorbidities and high reoperative risk, with a mean logistic EuroSCORE (European System for Cardiac Operative Risk Evaluation) of 15.4% (range 4% to 37%).
Localization of PPL reported according to the SVTCM by prospective TEE, blinded TEE review, and fluoroscopy are shown for the 20 patients included in Table 2. As shown, some PPLs were very localized (single hour or 1-h range), whereas the majority were larger and spanned over a range of 2 to 4 h on the SVTCM representing one-sixth to one-third of the mitral annular circumference. All patients had grade 3 or 4 mitral regurgitation with a mean range of hours of 2.67 ± 1 h (prospective TEE assessment), 2.62 ± 0.8 h (retrospective TEE assessment), and 1.8 ± 0.4 h (fluoroscopic assessment). Because many PPLs spanned over more than 1 quadrant, a predominant quadrant was attributed. The distribution of PPL localizations by quadrant was as follows: predominantly septal (12:00 to 3:00) in 5 patients; posterior (3:00 to 6:00) in 4 patients; lateral (6:00 to 9:00) in 5 patients; and anterior (9:00 to 12:00) in 6 patients.
Section-by-section comparisons were performed on a total of 240 single-hour sections (12 single-hour sections in 20 patients) for each localization method (i.e., prospective TEE, offline TEE review, and fluoroscopy; 240 sections × 3) and showed that there was excellent agreement between prospective TEE and offline TEE review for PPL localization (92.08%, 95% CI: 88.67% to 95.50%; p < 0.0001, when compared with the null hypothesis of 50%), and there was very good section-by-section agreement between prospective TEE and fluoroscopic localization (80.00%, 95% CI: 74.94% to 85.06%; p < 0.0001, when compared with the null hypothesis of 50%).
Patient-by-patient analyses showed that there was excellent agreement (overlap of ≥1 h) between prospective TEE and offline TEE review for PPL localization in all 20 patients (100%; p < 0.0001, when compared with the null hypothesis of 50%), and there was also excellent agreement between prospective TEE and fluoroscopic localization (90.00%, 95% CI: 76.85% to 100%; p = 0.0003, when compared with the null hypothesis of 50%;). In 2 cases of 20, where there was disagreement (no overlap), TEE and fluoroscopic localizations were adjacent to one another (i.e., separated only by a 1-h section).
The PPL reductions were attempted with the Amplatzer Duct Occluder, the Amplatzer mVSD Occluder, or more recently, with the Vascular Plug III (AGA Medical Corporation, Plymouth, Minnesota). Among the 20 procedures, 2 were unsuccessful (Patients #3 and #11) for reasons unrelated to localization. In both, the catheter successfully crossed the defect, but the implanted device caused prosthetic disk mobility interference and was immediately removed. In Patient #18, an mVSD Occluder was successfully replaced by a Vascular Plug III because the mono-disk mobility of the Medtronic Hall valve was impaired by the first device, whereas for Patient #20, a Vascular Plug III was considered unstable and was replaced by an mVSD Occluder. No other procedural complications occurred.
This study demonstrates excellent reproducibility of the SVTCM to localize mitral PPL by 2D-TEE and very good agreement with the described novel fluoroscopic method. This multimodality method represents a highly valuable tool for the guidance of transcatheter PPL reduction, which we have been performing in our center since 2002 (14). We developed this common reference system to precisely localize the defect by TEE and fluoroscopy with the goal to improve communication between echocardiographer and interventional cardiologist. We now routinely use this SVTCM, allowing effective and simple communication between caregivers.
The method used in our study is very reproducible and easy to use with a systematic and comprehensive approach, and it provides a detailed and complete evaluation of PPL. Moreover, this reference system is more accurate than the quadrant method (19), especially when the leak overlaps more than a single quadrant. This method has the advantage of representing a standardized nomenclature to localize PPL and has the potential, not only in our center, but also in others, to improve communication between echocardiographers, interventional cardiologists, and cardiac surgeons, because it is a common reference system based on the anatomical representation of the mitral annulus. This is important because patients with PPL represent a “difficult-to-treat” population, and transcatheter PPL reduction is a technically demanding procedure with a steep learning curve.
During the procedure, both imaging modalities are essential. Fluoroscopy is important during insertion of guidewires, catheter delivery, and device deployment. Ultrasound is used to localize and assess the defect, guide transseptal puncture and guidewire introduction, and ensure correct device delivery. Transesophageal echocardiography is also extremely useful, not only for localization and procedural guidance, but also to rule out immediate complications such as prosthetic dysfunction associated with the device or pericardial hemorrhage, and TEE gives immediate feedback on the results of the procedure by visualizing and grading severity of residual regurgitation. Moreover, TEE guidance avoids the use of potentially nephrotoxic contrast agents in these fragile patients. Initial reported experiences of transcatheter PPL closure under TEE guidance have shown promising results (10–16,20).
Recently, a new generation of TEE probes with a novel matrix array was introduced, allowing 3-dimensional (3D) representation of cardiac structures in real time with high spatial resolution. Three-dimensional TEE has been used to guide various transcatheter procedures such as atrial or ventricular septal defect closures, valvular procedures, and electrophysiological interventions (21), and very recently, has been described for transcatheter reduction of PPL (22–24). This technique has the potential to increase the ease of localization and, potentially, the efficacy of PPL reductions. However, precise anatomical knowledge remains essential for the interventional cardiologists who perform these procedures and for the echocardiographers involved in PPL evaluation and procedural guidance. The PPL localization should be reported in a standardized fashion, regardless of the imaging technique used, to ensure proper communication between caregivers. The SVTCM can readily be applied to 3D echocardiography, and application of this methodology has the potential to simplify 3D interpretation and reporting. Placing the mitral prosthesis in 3D anatomical or surgeon's view should allow direct and simple interpretation for localization of PPL that could be reported using a single standardized method for echocardiographers, surgeons, and interventional cardiologists, as described in this study.
This is a retrospective analysis of echocardiographic and fluoroscopic data. To allow comparison of fluoroscopic PPL localization with TEE, only patients in whom the defect was actually crossed with a catheter or device could be included. The number of patients is limited, but percutaneous reduction procedures are reserved for high-risk patients after intensive multidisciplinary evaluation, and to date, reported series include small numbers of patients (10–16). To limit bias, an echocardiographer blinded to previous findings independently reviewed the echocardiographic data offline, and an interventional cardiologist who was not the principle operator, also blinded to previous findings, reviewed the fluoroscopic data offline. Comparison of the fluoroscopic method was made with the original, prospectively acquired TEE images and interpretations. The localizations of PPL were, by chance, evenly distributed among all 4 quadrants in our study population.
Localization of PPL with the novel fluoroscopic SVTCM shows very good agreement with the highly reproducible TEE method. In our experience, the use of this method provides a standardized nomenclature for echocardiographers and interventional cardiologists, and it improves communication and image guidance for these complex interventions, allowing better care for patients with PPL.
The authors would like to thank Denis Burelle, MD; Philippe Demers, MD; André Denault, MD, PhD; Lise-Andrée Mercier, MD; Guy B. Pelletier, MD; and Marie-Claude Guertin, PhD, for their contribution to the realization of this study. The authors also thank Dr. André Denault for his contribution in preparing Figure 2 of this paper.
Dr. Marcotte is a lecturer to Actelion Pharma and St. Jude Medical. Dr. Bonan is consultant to Medtronic Transcatheter. All other authors report that they have no relationships to disclose. The first two authors contributed equally to this work.
- Abbreviations and Acronyms
- left atrial appendage
- periprosthetic leak
- surgeon's-view time-clock method
- transesophageal echocardiography
- Received April 27, 2010.
- Revision received July 28, 2010.
- Accepted August 6, 2010.
- American College of Cardiology Foundation
- O'Rourke D.J.,
- Palac R.T.,
- Malenka D.J.,
- Marrin C.A.,
- Arbuckle B.E.,
- Plenh J.F.
- Ionescu A.,
- Fraser A.G.,
- Butchart E.G.
- Genoni M.,
- Franzen D.,
- Vogt P.,
- et al.
- Potvin J.,
- Noble S.,
- Basmadjian A.,
- Crepeau J.,
- Ibrahim R.
- Basmadjian A.,
- Pellerin M.
- Hammerstingl C.,
- Lickfett L.,
- Nickening G.
- Kim M.S.,
- Casserly I.P.,
- Garcia J.A.,
- Klein A.J.,
- Salcedo E.E.,
- Caroll J.D.
- García-Fernández M.A.,
- Cortés M.,
- García-Robles J.A.,
- Gomez de Diego J.J.,
- Perez-David E.,
- García E.