Paravalvular leakage in patients with prosthetic heart valves: cardiac computed tomography findings and clinical features

Paravalvular leakage in patients with prosthetic heart valves: cardiac computed tomography... Abstract Aims Morphological characteristics of paravalvular leakage (PVL) have not yet been well characterized by computed tomography (CT). The purpose of this study was to demonstrate the morphological characteristics of PVLs using cardiac CT in patients diagnosed with PVL. Methods and results Between May 2011 and December 2013, 46 patients who had been diagnosed with PVL and underwent cardiac CT were included in this study. On CT, the characteristics of PVLs including number, size, location, and shape are described. Inter- and intra-observer agreement of CT were assessed. The extent of PVL on CT and the degree of regurgitant grade on echocardiography are compared. The size of PVLs were compared between patients who underwent surgical correction and who treated with percutaneous device closure or observed without treatment. All PVLs detected on surgical filed were the same with the number and locations of PVLs demonstrated on CT. Interobserver agreement for CT measurements of the PVL was good, ranging from 0.78 to 0.99. The sizes of PVL were smaller in patients who treated with device closure or observed without treatment compared to those in patients who underwent surgical correction (median areas of aortic PVL, 64.5 vs. 15.5 mm2 and mitral PVL, 40.0 vs. 27.0 mm2). PVL sizes were larger in patients with higher regurgitant grades on echocardiography (P < 0.05). Conclusion Cardiac CT can demonstrate the location and size of PVL. The size of PVLs are larger in patients who managed by surgical correction. PVL size measured on CT is correlated with the regurgitant grade on echocardiography. paravalvular leakage, computed tomography, prosthetic valve Introduction Paravalvular leakage (PVL) is defined as abnormal communication between the sewing ring and valve annulus.1,2 The prevalence of PVL in patients who underwent valve replacement surgery has been reported to vary from 1% to 10% around the aortic valve1,3 and from 3% to 15% around the mitral valve.1,4 Although small PVLs are common and usually benign, they have been associated with poorer survival.1,3 Notably, the incidence of major leaks ranges from 2% to 10%.5–7 As a treatment for severe PVL, surgical corrections have been recommended to perform either repair of the leak or re-replacement of the valve. However, both procedures have failure rates that range from 12% to 35%, and such cases increase the need for re-intervention.8 Moreover, recurrence of a PVL can also reduce survival rates among surgical candidates.9 Recently, percutaneous device closure has been introduced as an alternative option to treat PVL.5,6,10–14 To reduce the failure rate associated with these treatment methods, a thorough review of the morphological characteristics of the PVLs should be performed before conducting any of these procedures. Standards of care dictate that imaging studies for an initial diagnosis of PVL are usually made by echocardiography. Recently, several case series have evaluated the utilization of cardiac computed tomography (CT) to evaluate PVLs.2,5,14,15 These two different imaging modalities each have their own benefits as flow velocity helps clinicians to interpret disease severity, whereas cardiac CT may be better suited for preoperative planning by yielding more precise anatomical details. Moreover, cardiac CT can help a physician select an optimal treatment method by providing a more accurate localization and visualization of these leaks. Although several anatomical reports have been published based on surgical findings,5,14,15 CT-based morphological details have not yet been well characterized. We hypothesized that the morphological characteristics of PVL could be described thoroughly using cardiac CT. Herein, we aimed to demonstrate the morphological characteristics of PVLs, and its potential clinical impact using cardiac CT in patients who had detected PVL on echocardiography. Materials and methods Study population The institutional review board of our institution approved this retrospective study and waived the requirement to obtain informed consent. Between May 2011 and December 2013, 131 patients who showed PVL by echocardiography were retrospectively reviewed. They underwent echocardiography for routine postoperative care (n = 82) or a symptomatic evaluation (n = 49). Symptomatic patients were defined based on the presence of congestive heart failure, systemic emboli, haemolysis, and new-onset atrioventricular block. Congestive heart failure was defined based on criteria developed by the Framingham Heart Study.16 Haemolysis was defined as a serum lactate dehydrogenase level >460 U/L along with any two of the following criteria: (i) level of blood haemoglobin <13.8 g/dL for males or <12.4 g/dL for females; (ii) level of serum haptoglobin <50 mg/dL; or (iii) reticulocyte count >2%. After excluding two patients with suboptimal CT image quality due to severe motion artefacts, 46 patients (mean age 62.0 years, 24 males) subsequently underwent electrocardiogram-gated cardiac CT and were ultimately enrolled. Reasons for the examination by cardiac CT were most commonly for demonstration of paravalvular anatomy to plan treatment of a PVL (n = 44), followed by evaluation of coronary artery disease (n = 2).17–19 All CT scans were obtained within 7 days of the initial transoesophageal echocardiography (3 ± 2 days). Patient demographics, clinical findings, echocardiography findings, and surgical records were reviewed based on electronic medical records. Overall mean follow-up periods were 410 ± 253 days. CT technique Retrospective electrocardiogram (ECG)-gated cardiac CT was performed using a second-generation dual-source CT (Somatom Definition Flash, Siemens, Erlangen, Germany). The scan delay was determined using a bolus tracking method with a threshold of 100 HU, and the region of interest was placed in the ascending aorta. Images were obtained after injecting 60–80 mL iomeprol-400 (Iomeron; Bracco Imaging, Milan, Italy), followed by 40 mL saline chaser at 4 mL/s. The body size radiation dose-adaptive protocol, in which the tube voltage and current time product were adjusted according to the patient body habitus, was used to reduce the radiation dose as follows: 80–120 kVp tube voltage, 185–380 mA tube current per rotation with a 280 ms gantry rotation time, and 128 × 0.6 mm collimation. ECG-based tube current modulation was applied (R–R interval, 30–80%). The mean dose length product was 967.1 [standard deviation (SD) 450.3] mGy⋅cm and the mean effective dose was 14.5 (SD 8.6) mSv. CT image analysis Image reconstruction CT images were reconstructed in a 5% R–R interval using a 1-mm slice and a B26 kernel, which were transferred to an external workstation (Aquarius, TeraRecon, Foster City, CA, USA) for post-processing. Before interpreting the cardiac CTs, details of the echocardiography regarding the size, location, lesion number, and shape of the PVL were blinded, except the presence of PVL. Evaluation of the PVL was independently performed by two radiologists for the following steps (see Supplementary data online, Videos S1 and S2). First, cardiac phases with the least motion and beam-hardening artefacts were selected. The least amount of motion around the prosthesis on CT images could be distinguished by reviewing different intervals of the cardiac phases. The largest size of PVL during whole obtained cardiac cycles is measured. Because of cardiac motion artefacts and beam hardening artefacts from prosthetic valve, not all phases could be used for evaluation of PVLs. We performed the comparison of PVL sizes between systolic and diastolic phases in a subset of patients. In consequence, aortic PVLs are measured during the systolic phase (20–40% R–R) and mitral PVLs are drawn in the diastolic phase (65–85% R–R). Second, multiplanar reformatted (MPR) images were generated by sagittal view (Figure 1A and D) and en face views of the prosthetic valves (Figure 1B and E). Third, the morphological characteristics of the PVL were determined. A contrast media-filled lesion adjacent to the prosthetic valve that was continuous between the two cardiac chambers was defined as a PVL. The PVLs are frequently small and the peri-dehiscence tissues may be thin, so to avoid overestimation of the true area of the dehiscence, further image reconstruction was carried out by either rotating the imaging plane to centre the dehiscence or using a 360° rotational technique following the ‘double-checking’ method. Fourth, en face views of the volume rendering images were provided as surgeon-friendly images (surgeon’s view; Figure 1C and F). The raysum technique was used to generate fluoroscopy-like images (fluoroscopic projection view; Figure 1G). The entrance site to the PVL was marked on fluoroscopic views with diverse angles (see Supplementary data online, Video S3). The raysum images with fluoroscopic views for an interventional cardiologist was generated using a separate workstation (Advantage Server, General Electric, MI, USA). Figure 1 View largeDownload slide A 67-year-old woman who presented with paravalvular leakage (PVL) at the prosthetic aortic and mitral valves. Sagittal (A), en face (B), and surgeon’s views (C) of mitral plane images showing a PVL (arrow) in the antero-lateral aspect (11 o’clock) of the prosthetic valve. The left circumflex artery indicates the lateral direction of the mitral annulus. Sagittal (D), en face (E), and surgeon’s views (F) of aortic valve (asterisk) images showing PVL (arrow) near the left coronary artery orifice. (G) Mitral (right arrow) and aortic (left arrow) PVLs are noted in fluoroscopic view which was obtained using the ray sum technique. (H) The two lesions were successfully occluded using vascular plugs (arrows) by a percutaneous approach. Ao, aorta; CAU, caudal; LA, left atrium; LAA, left atrial appendage; LAO, left anterior oblique; LV, left ventricle; PAV, prosthetic aortic valve; PMV, prosthetic mitral valve; RA, right atrium. Figure 1 View largeDownload slide A 67-year-old woman who presented with paravalvular leakage (PVL) at the prosthetic aortic and mitral valves. Sagittal (A), en face (B), and surgeon’s views (C) of mitral plane images showing a PVL (arrow) in the antero-lateral aspect (11 o’clock) of the prosthetic valve. The left circumflex artery indicates the lateral direction of the mitral annulus. Sagittal (D), en face (E), and surgeon’s views (F) of aortic valve (asterisk) images showing PVL (arrow) near the left coronary artery orifice. (G) Mitral (right arrow) and aortic (left arrow) PVLs are noted in fluoroscopic view which was obtained using the ray sum technique. (H) The two lesions were successfully occluded using vascular plugs (arrows) by a percutaneous approach. Ao, aorta; CAU, caudal; LA, left atrium; LAA, left atrial appendage; LAO, left anterior oblique; LV, left ventricle; PAV, prosthetic aortic valve; PMV, prosthetic mitral valve; RA, right atrium. Morphological characteristics of PVLs on CT The location and extent of each PVL was marked based on a clock-wise format on en face views of CT images (Figure 2). Twelve o’clock was assigned to the commissure between the left- and right-coronary sinuses, 4 o’clock to the commissure between the right- and non-coronary sinuses, and 8 o’clock to the commissure between the left- and non-coronary sinuses. A reference line was drawn from the commissure of the left- and right-coronary cusp to the mid-portion of the non-coronary sinus. Similar to an aortic valve assessment, the mitral PVL location could also be reported. The mid-points of the anterior and posterior annulus were represented as 12 and 6 o’clock, respectively. The postero-medial commissure and interatrial septum were at 3 o’clock, while the antero-lateral commissure and atrial appendage were at 9 o’clock.20 Figure 2 View largeDownload slide Distribution of paravalvular leakages in prosthetic (A) aortic and (B) mitral valves. AO, aorta; LA, left atrium; LAA, left atrial appendage; LCC, left coronary cusp; NCC, non-coronary cusp; RA, right atrium; RCC, right coronary cusp; 6, 6 o’clock direction; 12, 12 o’clock direction. Figure 2 View largeDownload slide Distribution of paravalvular leakages in prosthetic (A) aortic and (B) mitral valves. AO, aorta; LA, left atrium; LAA, left atrial appendage; LCC, left coronary cusp; NCC, non-coronary cusp; RA, right atrium; RCC, right coronary cusp; 6, 6 o’clock direction; 12, 12 o’clock direction. Based on the centre point of the dehiscence, the location of the PVL was classified as follows: (i) right-coronary cusp (12–4 o’clock), non-coronary cusp (4–8 o’clock), and left-coronary cusp (8–12 o’clock) in the aortic prosthetic valve and (ii) anterior (directions of the hour hand of a clock at 10:30–1:30), medial (1:30–4:30), posterior (4:30–7:30), and lateral (7:30–10:30) in the mitral prosthetic valve. For anatomical descriptions of the PVL, the size, shape, location, and number of the PVLs were determined based on MPR images. On an en face view of the prosthetic valve, the long- and short-diameters, perimeter, area, and involved angle of the PVL were measured to describe the extent of the PVL (see Supplementary data online, Figure S1). The length of the dehiscence tract was measured on a profile view of the prosthetic valve that was centred at the dehiscence. The shape of the PVL was categorized as round, ovoid, or crescent using a ratio of the long diameter to short diameter (L/S ratio) as follows: (i) round shape, L/S ratio of two or less; (ii) ovoid shape, L/S ratio between two and five; and (iii) crescent shape, L/S ratio of more than five. Treatment method Treatment strategies were decided based on patient symptoms, co-morbidity, regurgitant degree with heart function on echocardiography, and the morphological details on cardiac CT. For patients who had undergone cardiac CT, multidisplinary heart team conferences that included the cardiologist, cardiac surgeons, and radiologists were held. The degree of paravalvular regurgitation through the dehiscence was graded as either mild, moderate, or severe based on the color Doppler jet area on echocardiography. In patients with mild PVL and no symptoms, careful follow-up was recommended. In cases of moderate or severe leakage with clinical symptoms, either surgical or interventional treatment was recommended. Before repair of the PVLs, the surgeon’s view or fluoroscopic projection view was provided to the operator. During PVL repair, the morphological details of the PVLs were recorded by the operator regarding the location, number, and shape of the dehiscence. Statistical analysis Continuous variables were analysed using independent samples t-test, and the Fisher’s exact test was used to analyse categorical variables. By the Kruskal–Wallis analysis, sizes of PVLs on CT were compared according to the degree of regurgitant grade on echocardiography. Interobserver agreements on the size, location, and number of lesions on CT were obtained using a two-way random model intra-class correlation coefficient with consistency assumption. A two-sided P-value <0.05 was considered to indicate a statistically significant difference. Statistical analyses were performed using the Statistical Package for the Social Sciences (version 18 for Windows; SPSS Inc., Chicago, IL, USA). Results Patient characteristics Baseline characteristics of patients are summarized in Table 1. Patients who had undergone CT were more symptomatic (74% vs. 18%, P < 0.001) and showed decreased LV systolic function. On CT, a total of 48 PVLs were noted in 46 patients. One patient had two sites of mitral PVL, and another patient presented both aortic and mitral PVLs (Figure 1). CT characteristics and treatment methods for aortic and mitral PVLs are noted in Table 2. Left and right coronary sinuses are common sites of aortic PVLs. Lateral portion (42%) of the mitral annulus was the most common site of mitral PVL. Half of the PVLs had a round shape, which indicated potentially adequate for device occlusion. Inter- and intra-observer agreement for CT measurements of PVLs were good to excellent (range 0.78–0.99) (see Supplementary data online, Figure S2). Table 1 Patient characteristics Patients with CT (n = 46) Patients without CT (n = 85) P-value Age (years) 62.1 ± 11.3 60.6 ± 13.4 0.52 Male gender 24 (52) 45 (53) 1.000 Body weight 58.8 ± 9.8 60.9 ± 11.2 0.28 Height 162.2 ± 8.3 161.3 ± 9.5 0.59 Heart rate 75.0 ± 16.7 76.4 ± 16.4 0.64 Symptom 34 (74)a 15 (18)a <0.001  Asymptomatic 12 (26) 70 (82) <0.001  Congestive heart failure 31 (67) 12 (14) <0.001  Haemolysis 4 (9) 1 (1) 0.05  Systemic emboli 1 (2) 1 (1) 1.00  New-onset AV block 1 (2) 2 (2) 1.00 Laboratory findings  Haemoglobin (g/dL) 10.6 ± 2.1 11.8 ± 2.1 0.002  Creatinine (mg/dL) 1.2 ± 1.3 1.1 ± 1.1 0.89 Involved prosthetic valve  Aortic valve 19 (41) 45 (53)  Mitral valve 24 (52)b 37 (44)  Tricuspid valve 2 (4) 3 (4)  Both aortic and mitral valves 1 (2) 0 (0) Concomitant IE 7 (15) 2 (2) 0.003 Echocardiography  LVEF (%) 52.7 ± 11.4 56.5 ± 10.4 0.06  End diastolic volume (mL) 147.9 ± 69.7 124.8 ± 48.0 0.03  End systolic volume (mL) 74.1 ± 46.6 57.1 ± 34.1 0.02 Number of previous open heart surgery  One 31 (67) 74 (87)  Two (Redo-) 12 (26) 9 (11)  Three (Trido-) 3 (7) 2 (2) Valve type 0.58  Tissue valve 7 (15) 9 (11)  Prosthetic valve 39 (85) 76 (89) Patients with CT (n = 46) Patients without CT (n = 85) P-value Age (years) 62.1 ± 11.3 60.6 ± 13.4 0.52 Male gender 24 (52) 45 (53) 1.000 Body weight 58.8 ± 9.8 60.9 ± 11.2 0.28 Height 162.2 ± 8.3 161.3 ± 9.5 0.59 Heart rate 75.0 ± 16.7 76.4 ± 16.4 0.64 Symptom 34 (74)a 15 (18)a <0.001  Asymptomatic 12 (26) 70 (82) <0.001  Congestive heart failure 31 (67) 12 (14) <0.001  Haemolysis 4 (9) 1 (1) 0.05  Systemic emboli 1 (2) 1 (1) 1.00  New-onset AV block 1 (2) 2 (2) 1.00 Laboratory findings  Haemoglobin (g/dL) 10.6 ± 2.1 11.8 ± 2.1 0.002  Creatinine (mg/dL) 1.2 ± 1.3 1.1 ± 1.1 0.89 Involved prosthetic valve  Aortic valve 19 (41) 45 (53)  Mitral valve 24 (52)b 37 (44)  Tricuspid valve 2 (4) 3 (4)  Both aortic and mitral valves 1 (2) 0 (0) Concomitant IE 7 (15) 2 (2) 0.003 Echocardiography  LVEF (%) 52.7 ± 11.4 56.5 ± 10.4 0.06  End diastolic volume (mL) 147.9 ± 69.7 124.8 ± 48.0 0.03  End systolic volume (mL) 74.1 ± 46.6 57.1 ± 34.1 0.02 Number of previous open heart surgery  One 31 (67) 74 (87)  Two (Redo-) 12 (26) 9 (11)  Three (Trido-) 3 (7) 2 (2) Valve type 0.58  Tissue valve 7 (15) 9 (11)  Prosthetic valve 39 (85) 76 (89) Data represent the mean ± standard deviation or numbers of patients with the percentage indicated in parentheses. a Total number of patients who had one or more symptoms. b One of them had two sites of mitral paravalvular leakages. AV, atrioventricular; CT, computed tomography; IE, infective endocarditis; LVEF, left ventricular ejection fraction. Table 1 Patient characteristics Patients with CT (n = 46) Patients without CT (n = 85) P-value Age (years) 62.1 ± 11.3 60.6 ± 13.4 0.52 Male gender 24 (52) 45 (53) 1.000 Body weight 58.8 ± 9.8 60.9 ± 11.2 0.28 Height 162.2 ± 8.3 161.3 ± 9.5 0.59 Heart rate 75.0 ± 16.7 76.4 ± 16.4 0.64 Symptom 34 (74)a 15 (18)a <0.001  Asymptomatic 12 (26) 70 (82) <0.001  Congestive heart failure 31 (67) 12 (14) <0.001  Haemolysis 4 (9) 1 (1) 0.05  Systemic emboli 1 (2) 1 (1) 1.00  New-onset AV block 1 (2) 2 (2) 1.00 Laboratory findings  Haemoglobin (g/dL) 10.6 ± 2.1 11.8 ± 2.1 0.002  Creatinine (mg/dL) 1.2 ± 1.3 1.1 ± 1.1 0.89 Involved prosthetic valve  Aortic valve 19 (41) 45 (53)  Mitral valve 24 (52)b 37 (44)  Tricuspid valve 2 (4) 3 (4)  Both aortic and mitral valves 1 (2) 0 (0) Concomitant IE 7 (15) 2 (2) 0.003 Echocardiography  LVEF (%) 52.7 ± 11.4 56.5 ± 10.4 0.06  End diastolic volume (mL) 147.9 ± 69.7 124.8 ± 48.0 0.03  End systolic volume (mL) 74.1 ± 46.6 57.1 ± 34.1 0.02 Number of previous open heart surgery  One 31 (67) 74 (87)  Two (Redo-) 12 (26) 9 (11)  Three (Trido-) 3 (7) 2 (2) Valve type 0.58  Tissue valve 7 (15) 9 (11)  Prosthetic valve 39 (85) 76 (89) Patients with CT (n = 46) Patients without CT (n = 85) P-value Age (years) 62.1 ± 11.3 60.6 ± 13.4 0.52 Male gender 24 (52) 45 (53) 1.000 Body weight 58.8 ± 9.8 60.9 ± 11.2 0.28 Height 162.2 ± 8.3 161.3 ± 9.5 0.59 Heart rate 75.0 ± 16.7 76.4 ± 16.4 0.64 Symptom 34 (74)a 15 (18)a <0.001  Asymptomatic 12 (26) 70 (82) <0.001  Congestive heart failure 31 (67) 12 (14) <0.001  Haemolysis 4 (9) 1 (1) 0.05  Systemic emboli 1 (2) 1 (1) 1.00  New-onset AV block 1 (2) 2 (2) 1.00 Laboratory findings  Haemoglobin (g/dL) 10.6 ± 2.1 11.8 ± 2.1 0.002  Creatinine (mg/dL) 1.2 ± 1.3 1.1 ± 1.1 0.89 Involved prosthetic valve  Aortic valve 19 (41) 45 (53)  Mitral valve 24 (52)b 37 (44)  Tricuspid valve 2 (4) 3 (4)  Both aortic and mitral valves 1 (2) 0 (0) Concomitant IE 7 (15) 2 (2) 0.003 Echocardiography  LVEF (%) 52.7 ± 11.4 56.5 ± 10.4 0.06  End diastolic volume (mL) 147.9 ± 69.7 124.8 ± 48.0 0.03  End systolic volume (mL) 74.1 ± 46.6 57.1 ± 34.1 0.02 Number of previous open heart surgery  One 31 (67) 74 (87)  Two (Redo-) 12 (26) 9 (11)  Three (Trido-) 3 (7) 2 (2) Valve type 0.58  Tissue valve 7 (15) 9 (11)  Prosthetic valve 39 (85) 76 (89) Data represent the mean ± standard deviation or numbers of patients with the percentage indicated in parentheses. a Total number of patients who had one or more symptoms. b One of them had two sites of mitral paravalvular leakages. AV, atrioventricular; CT, computed tomography; IE, infective endocarditis; LVEF, left ventricular ejection fraction. Table 2 CT characteristics and treatment of paravalvular leakages (per-lesion based analysis) Prosthetic aortic valve Prosthetic mitral valve P-value Long diameter (mm) 9.5 (5.7–15.4) 12.4 (6.6–15.9) 0.31 Short diameter (mm) 5.0 (2.9–5.5) 4.3 (3.2–5.7) 0.76 Perimeter (mm) 24.0 (14.9–39.0) 29.8 (21.7–45.7) 0.17 Area (mm2) 41.0 (14.3–77.8) 35.3 (17.5–60.0) 0.28 Involved angle (°) 33.0 (28.8–59.5) 31.0 (21.8–58.0) 0.69 Location  Left-coronary 8 (40)  Right-coronary 8 (40)  Non-coronary 4 (20)  Anterior 3 (12)  Posterior 8 (31)  Lateral 11 (42)  Medial 4 (15) Shape 0.80  Round 10 (50) 12 (46)  Ovoid/elongated 9/1 (45/5) 10/4 (38/15) Regurgitant grade 0.62  Mild 5 (25) 5 (19)  Moderate 5 (25) 10 (38)  Severe 10 (50) 11 (42) Concomitant IE 5 (25) 2 (8) Pseudoaneurysm 3 (15) 0 (0) Severe annular calcification 0 (0) 2 (8) Treatment  Surgery 10 (50) 17 (65)  Device closure 1 [1/0] (5)a 5 [3/2] (19)a  Observation 9 (45) 4 (15) Prosthetic aortic valve Prosthetic mitral valve P-value Long diameter (mm) 9.5 (5.7–15.4) 12.4 (6.6–15.9) 0.31 Short diameter (mm) 5.0 (2.9–5.5) 4.3 (3.2–5.7) 0.76 Perimeter (mm) 24.0 (14.9–39.0) 29.8 (21.7–45.7) 0.17 Area (mm2) 41.0 (14.3–77.8) 35.3 (17.5–60.0) 0.28 Involved angle (°) 33.0 (28.8–59.5) 31.0 (21.8–58.0) 0.69 Location  Left-coronary 8 (40)  Right-coronary 8 (40)  Non-coronary 4 (20)  Anterior 3 (12)  Posterior 8 (31)  Lateral 11 (42)  Medial 4 (15) Shape 0.80  Round 10 (50) 12 (46)  Ovoid/elongated 9/1 (45/5) 10/4 (38/15) Regurgitant grade 0.62  Mild 5 (25) 5 (19)  Moderate 5 (25) 10 (38)  Severe 10 (50) 11 (42) Concomitant IE 5 (25) 2 (8) Pseudoaneurysm 3 (15) 0 (0) Severe annular calcification 0 (0) 2 (8) Treatment  Surgery 10 (50) 17 (65)  Device closure 1 [1/0] (5)a 5 [3/2] (19)a  Observation 9 (45) 4 (15) Data represent the median and interquartile ranges or numbers of patients with the percentage indicated in parentheses. a Success/failed cases of percutaneous device closure. IE, infective endocarditis. Table 2 CT characteristics and treatment of paravalvular leakages (per-lesion based analysis) Prosthetic aortic valve Prosthetic mitral valve P-value Long diameter (mm) 9.5 (5.7–15.4) 12.4 (6.6–15.9) 0.31 Short diameter (mm) 5.0 (2.9–5.5) 4.3 (3.2–5.7) 0.76 Perimeter (mm) 24.0 (14.9–39.0) 29.8 (21.7–45.7) 0.17 Area (mm2) 41.0 (14.3–77.8) 35.3 (17.5–60.0) 0.28 Involved angle (°) 33.0 (28.8–59.5) 31.0 (21.8–58.0) 0.69 Location  Left-coronary 8 (40)  Right-coronary 8 (40)  Non-coronary 4 (20)  Anterior 3 (12)  Posterior 8 (31)  Lateral 11 (42)  Medial 4 (15) Shape 0.80  Round 10 (50) 12 (46)  Ovoid/elongated 9/1 (45/5) 10/4 (38/15) Regurgitant grade 0.62  Mild 5 (25) 5 (19)  Moderate 5 (25) 10 (38)  Severe 10 (50) 11 (42) Concomitant IE 5 (25) 2 (8) Pseudoaneurysm 3 (15) 0 (0) Severe annular calcification 0 (0) 2 (8) Treatment  Surgery 10 (50) 17 (65)  Device closure 1 [1/0] (5)a 5 [3/2] (19)a  Observation 9 (45) 4 (15) Prosthetic aortic valve Prosthetic mitral valve P-value Long diameter (mm) 9.5 (5.7–15.4) 12.4 (6.6–15.9) 0.31 Short diameter (mm) 5.0 (2.9–5.5) 4.3 (3.2–5.7) 0.76 Perimeter (mm) 24.0 (14.9–39.0) 29.8 (21.7–45.7) 0.17 Area (mm2) 41.0 (14.3–77.8) 35.3 (17.5–60.0) 0.28 Involved angle (°) 33.0 (28.8–59.5) 31.0 (21.8–58.0) 0.69 Location  Left-coronary 8 (40)  Right-coronary 8 (40)  Non-coronary 4 (20)  Anterior 3 (12)  Posterior 8 (31)  Lateral 11 (42)  Medial 4 (15) Shape 0.80  Round 10 (50) 12 (46)  Ovoid/elongated 9/1 (45/5) 10/4 (38/15) Regurgitant grade 0.62  Mild 5 (25) 5 (19)  Moderate 5 (25) 10 (38)  Severe 10 (50) 11 (42) Concomitant IE 5 (25) 2 (8) Pseudoaneurysm 3 (15) 0 (0) Severe annular calcification 0 (0) 2 (8) Treatment  Surgery 10 (50) 17 (65)  Device closure 1 [1/0] (5)a 5 [3/2] (19)a  Observation 9 (45) 4 (15) Data represent the median and interquartile ranges or numbers of patients with the percentage indicated in parentheses. a Success/failed cases of percutaneous device closure. IE, infective endocarditis. Comparisons between CT and echocardiography The numbers and locations of PVLs detected on CT were the same with echocardiography except one patient. In the patient, two mitral PVLs on transoesophageal echocardiography had shown as a single elongated PVL on CT and finally confirmed to be a single lesion on the surgical inspection (Figure 3). The diameter, perimeter, area, and involved angle of PVLs measured on CT were significantly larger in patients with higher regurgitant grades on echocardiography (Table 3 and Figure 4). However, the length of PVLs were not statistically different according to the degree of regurgitant grades. Table 3 Comparison of the size of paravalvular leakage according to the regurgitant grade using the Kruskal–Wallis test (per-lesion based analysis) Mild Moderate Severe P-value Prosthetic aortic valve Long diameter (mm) 4.4 (3.9–4.9) 7.5 (7.4–12.0) 15.5 (9.8–18.6) 0.02 Short diameter (mm) 2.6 (2.2–2.7) 3.5 (2.9–5.0) 5.6 (5.1–7.1) 0.01 Perimeter (mm) 14.6 (10.0–15.0) 24.0 (22.3–26.0) 39.9 (28.5–42.3) 0.04 Area (mm2) 12.0 (6.0–16.0) 23.8 (15.0–30.9) 86.5 (58.5–134.8) 0.01 Involved angle (°) 27.0 (19.0–32.0) 33.0 (31.7–55.0) 60.0 (36.3–65.8) 0.05 Length (mm) 11.0 (8.3–12.9) 9.0 (8.2–9.1) 7.5 (5.5–12.0) 0.75 Prosthetic mitral valve Long diameter (mm) 8.0 (3.6–9.3) 9.2 (5.6–13.9) 18.0 (12.6–25.5) 0.01 Short diameter (mm) 3.3 (1.6–4.1) 3.6 (2.8–4.3) 5.8 (5.0–7.6) 0.01 Perimeter (mm) 24.0 (8.3–24.3) 25.2 (17.2–29.9) 52.0 (31.8–63.3) 0.01 Area (mm2) 19.0 (5.0–27.0) 22.5 (14.8–39.9) 62.0 (40.0–86.5) 0.005 Involved angle (°) 15.0 (14.0–29.0) 27.0 (19.5–46.3) 50.0 (30.0–98.0) 0.04 Length (mm) 8.0 (6.9–8.0) 9.7 (8.2–11.4) 9.3 (7.2–9.9) 0.20 Mild Moderate Severe P-value Prosthetic aortic valve Long diameter (mm) 4.4 (3.9–4.9) 7.5 (7.4–12.0) 15.5 (9.8–18.6) 0.02 Short diameter (mm) 2.6 (2.2–2.7) 3.5 (2.9–5.0) 5.6 (5.1–7.1) 0.01 Perimeter (mm) 14.6 (10.0–15.0) 24.0 (22.3–26.0) 39.9 (28.5–42.3) 0.04 Area (mm2) 12.0 (6.0–16.0) 23.8 (15.0–30.9) 86.5 (58.5–134.8) 0.01 Involved angle (°) 27.0 (19.0–32.0) 33.0 (31.7–55.0) 60.0 (36.3–65.8) 0.05 Length (mm) 11.0 (8.3–12.9) 9.0 (8.2–9.1) 7.5 (5.5–12.0) 0.75 Prosthetic mitral valve Long diameter (mm) 8.0 (3.6–9.3) 9.2 (5.6–13.9) 18.0 (12.6–25.5) 0.01 Short diameter (mm) 3.3 (1.6–4.1) 3.6 (2.8–4.3) 5.8 (5.0–7.6) 0.01 Perimeter (mm) 24.0 (8.3–24.3) 25.2 (17.2–29.9) 52.0 (31.8–63.3) 0.01 Area (mm2) 19.0 (5.0–27.0) 22.5 (14.8–39.9) 62.0 (40.0–86.5) 0.005 Involved angle (°) 15.0 (14.0–29.0) 27.0 (19.5–46.3) 50.0 (30.0–98.0) 0.04 Length (mm) 8.0 (6.9–8.0) 9.7 (8.2–11.4) 9.3 (7.2–9.9) 0.20 Data represent the median and interquartile ranges. Table 3 Comparison of the size of paravalvular leakage according to the regurgitant grade using the Kruskal–Wallis test (per-lesion based analysis) Mild Moderate Severe P-value Prosthetic aortic valve Long diameter (mm) 4.4 (3.9–4.9) 7.5 (7.4–12.0) 15.5 (9.8–18.6) 0.02 Short diameter (mm) 2.6 (2.2–2.7) 3.5 (2.9–5.0) 5.6 (5.1–7.1) 0.01 Perimeter (mm) 14.6 (10.0–15.0) 24.0 (22.3–26.0) 39.9 (28.5–42.3) 0.04 Area (mm2) 12.0 (6.0–16.0) 23.8 (15.0–30.9) 86.5 (58.5–134.8) 0.01 Involved angle (°) 27.0 (19.0–32.0) 33.0 (31.7–55.0) 60.0 (36.3–65.8) 0.05 Length (mm) 11.0 (8.3–12.9) 9.0 (8.2–9.1) 7.5 (5.5–12.0) 0.75 Prosthetic mitral valve Long diameter (mm) 8.0 (3.6–9.3) 9.2 (5.6–13.9) 18.0 (12.6–25.5) 0.01 Short diameter (mm) 3.3 (1.6–4.1) 3.6 (2.8–4.3) 5.8 (5.0–7.6) 0.01 Perimeter (mm) 24.0 (8.3–24.3) 25.2 (17.2–29.9) 52.0 (31.8–63.3) 0.01 Area (mm2) 19.0 (5.0–27.0) 22.5 (14.8–39.9) 62.0 (40.0–86.5) 0.005 Involved angle (°) 15.0 (14.0–29.0) 27.0 (19.5–46.3) 50.0 (30.0–98.0) 0.04 Length (mm) 8.0 (6.9–8.0) 9.7 (8.2–11.4) 9.3 (7.2–9.9) 0.20 Mild Moderate Severe P-value Prosthetic aortic valve Long diameter (mm) 4.4 (3.9–4.9) 7.5 (7.4–12.0) 15.5 (9.8–18.6) 0.02 Short diameter (mm) 2.6 (2.2–2.7) 3.5 (2.9–5.0) 5.6 (5.1–7.1) 0.01 Perimeter (mm) 14.6 (10.0–15.0) 24.0 (22.3–26.0) 39.9 (28.5–42.3) 0.04 Area (mm2) 12.0 (6.0–16.0) 23.8 (15.0–30.9) 86.5 (58.5–134.8) 0.01 Involved angle (°) 27.0 (19.0–32.0) 33.0 (31.7–55.0) 60.0 (36.3–65.8) 0.05 Length (mm) 11.0 (8.3–12.9) 9.0 (8.2–9.1) 7.5 (5.5–12.0) 0.75 Prosthetic mitral valve Long diameter (mm) 8.0 (3.6–9.3) 9.2 (5.6–13.9) 18.0 (12.6–25.5) 0.01 Short diameter (mm) 3.3 (1.6–4.1) 3.6 (2.8–4.3) 5.8 (5.0–7.6) 0.01 Perimeter (mm) 24.0 (8.3–24.3) 25.2 (17.2–29.9) 52.0 (31.8–63.3) 0.01 Area (mm2) 19.0 (5.0–27.0) 22.5 (14.8–39.9) 62.0 (40.0–86.5) 0.005 Involved angle (°) 15.0 (14.0–29.0) 27.0 (19.5–46.3) 50.0 (30.0–98.0) 0.04 Length (mm) 8.0 (6.9–8.0) 9.7 (8.2–11.4) 9.3 (7.2–9.9) 0.20 Data represent the median and interquartile ranges. Figure 3 View largeDownload slide (A) A 71-year-old male who underwent trido-mitral valve replacement exhibited two para-mitral leakages (arrows) on transoesophageal echocardiography in both the medial and lateral aspect of the mitral annulus. (B, C) On CT, lesions appeared as one large crescent shaped dehiscence (arrows) that involved posterior aspect of the mitral annulus. (D) Eventually, this lesion was confirmed as a single lesion in surgical inspection. Surgical instruments indicate the medial and lateral ends of the paravalvular dehiscence. Asterisk indicates prosthetic mitral valve; arrowhead indicates left circumflex artery; Ao, aorta; CS, coronary sinus; LA, left atrium; LAA, left atrial appendage; LV, left ventricle; RA, right atrium. Figure 3 View largeDownload slide (A) A 71-year-old male who underwent trido-mitral valve replacement exhibited two para-mitral leakages (arrows) on transoesophageal echocardiography in both the medial and lateral aspect of the mitral annulus. (B, C) On CT, lesions appeared as one large crescent shaped dehiscence (arrows) that involved posterior aspect of the mitral annulus. (D) Eventually, this lesion was confirmed as a single lesion in surgical inspection. Surgical instruments indicate the medial and lateral ends of the paravalvular dehiscence. Asterisk indicates prosthetic mitral valve; arrowhead indicates left circumflex artery; Ao, aorta; CS, coronary sinus; LA, left atrium; LAA, left atrial appendage; LV, left ventricle; RA, right atrium. Figure 4 View largeDownload slide The areas and involved angle of paravalvular leakage measured on CT and the degree of regurgitant grades on echocardiography in prosthetic (A, B) aortic and (C, D) mitral valves. PAV, prosthetic aortic valve; PMV, prosthetic mitral valve. Figure 4 View largeDownload slide The areas and involved angle of paravalvular leakage measured on CT and the degree of regurgitant grades on echocardiography in prosthetic (A, B) aortic and (C, D) mitral valves. PAV, prosthetic aortic valve; PMV, prosthetic mitral valve. Size of PVL and treatment methods The size of PVLs were compared between systolic and diastolic phases in a subset of patients (see Supplementary data online, Table S1). Although there was no statistical significant difference, aortic PVLs are slightly larger during the systolic phase (20–40% R–R) and mitral PVLs are slightly larger during the diastolic phase (65–85% R–R). Representative cases are demonstrated in Supplementary data online, Figures S3 and S4. More than half of aortic and mitral PVLs were treated by surgical correction and six sites (five patients including one who had both aortic and mitral PVLs) of PVLs were managed by percutaneous device closure. The numbers and locations of PVLs detected on surgical filed were the same with those demonstrated on CT. The sizes of aortic PVLs are larger in patients who underwent surgical correction compared to those in patients with device closure or who observed without treatment (Table 4). The areas of mitral PVLs in patients with surgery were larger than those in patients without surgery, but marginally significant (P = 0.05). In both aortic and mitral PVLs, surgical correction is preferred to patients who presented severe regurgitant grade on echocardiography (P < 0.05). Among the five patients who were managed by percutaneous interventions, three patients (including the patient who had both aortic and mitral PVLs) were successfully treated, and two patients with small mitral PVLs exhibiting an elongated shape failed because of difficulty in guiding the passage of the catheter through the narrow and long dehiscence tract (Figure 5 and Supplementary data online, Video S1). The remaining 13 PVLs were observed without treatment because of the absence of symptoms (n = 11) or reluctance to undergo invasive treatment (n = 2). Postoperative mortality (range 1–56 days) occurred in three patients due to pneumonia (n = 1), multiorgan failure with a failed prosthetic valve (n = 1), or septic shock secondary to underlying infective endocarditis (n = 1). Table 4 Comparison of size of paravalvular leakage on CT and regurgitant grade on echocardiography according to the treatment methods (per-lesion based analysis) Prosthetic aortic valve Prosthetic mitral valve Surgery (n = 10) Device closure or observation (n = 10) P-value Surgery (n = 17) Device closure or observation (n = 9) P-value Long diameter (mm) 15.2 (8.8–17.6) 7.1 (4.5–10.9) 0.06 13.0 (11.2–18.0) 8.0 (6.5–12.4) 0.56 Short diameter (mm) 5.4 (5.1–6.3) 2.8 (2.3–4.6) 0.01 5.2 (3.6–6.0) 3.6 (3.2–4.4) 0.10 Perimeter (mm) 39.3 (25.3–41.7) 16.2 (13.5–23.9) 0.01 31.0 (23.9–52.0) 24.3 (17.2–31.7) 0.30 Area (mm2) 64.5 (52.8–119.8) 15.5 (10.5–29.1) 0.05 40.0 (22.0–67.7) 27.0 (17.0–40.0) 0.05 Involved angle (°) 58.7 (29.8–64.8) 32.4 (27.8–49.5) 0.11 44.0 (26.0–79.0) 26.0 (15.0–52.0) 0.54 Regurgitant grade 0.001a 0.04a  Mild 1 4 1 4  Moderate 0 5 6 4  Severe 9 1 10 1 Concomitant IE 2 3 2 0 Pseudoaneurysm 2 1 0 0 Severe annular calcification 0 0 1 1 Prosthetic aortic valve Prosthetic mitral valve Surgery (n = 10) Device closure or observation (n = 10) P-value Surgery (n = 17) Device closure or observation (n = 9) P-value Long diameter (mm) 15.2 (8.8–17.6) 7.1 (4.5–10.9) 0.06 13.0 (11.2–18.0) 8.0 (6.5–12.4) 0.56 Short diameter (mm) 5.4 (5.1–6.3) 2.8 (2.3–4.6) 0.01 5.2 (3.6–6.0) 3.6 (3.2–4.4) 0.10 Perimeter (mm) 39.3 (25.3–41.7) 16.2 (13.5–23.9) 0.01 31.0 (23.9–52.0) 24.3 (17.2–31.7) 0.30 Area (mm2) 64.5 (52.8–119.8) 15.5 (10.5–29.1) 0.05 40.0 (22.0–67.7) 27.0 (17.0–40.0) 0.05 Involved angle (°) 58.7 (29.8–64.8) 32.4 (27.8–49.5) 0.11 44.0 (26.0–79.0) 26.0 (15.0–52.0) 0.54 Regurgitant grade 0.001a 0.04a  Mild 1 4 1 4  Moderate 0 5 6 4  Severe 9 1 10 1 Concomitant IE 2 3 2 0 Pseudoaneurysm 2 1 0 0 Severe annular calcification 0 0 1 1 a The P-values are obtained from the comparison between mild or moderate and severe groups using the Fisher’s exact test. IE, infective endocarditis. Table 4 Comparison of size of paravalvular leakage on CT and regurgitant grade on echocardiography according to the treatment methods (per-lesion based analysis) Prosthetic aortic valve Prosthetic mitral valve Surgery (n = 10) Device closure or observation (n = 10) P-value Surgery (n = 17) Device closure or observation (n = 9) P-value Long diameter (mm) 15.2 (8.8–17.6) 7.1 (4.5–10.9) 0.06 13.0 (11.2–18.0) 8.0 (6.5–12.4) 0.56 Short diameter (mm) 5.4 (5.1–6.3) 2.8 (2.3–4.6) 0.01 5.2 (3.6–6.0) 3.6 (3.2–4.4) 0.10 Perimeter (mm) 39.3 (25.3–41.7) 16.2 (13.5–23.9) 0.01 31.0 (23.9–52.0) 24.3 (17.2–31.7) 0.30 Area (mm2) 64.5 (52.8–119.8) 15.5 (10.5–29.1) 0.05 40.0 (22.0–67.7) 27.0 (17.0–40.0) 0.05 Involved angle (°) 58.7 (29.8–64.8) 32.4 (27.8–49.5) 0.11 44.0 (26.0–79.0) 26.0 (15.0–52.0) 0.54 Regurgitant grade 0.001a 0.04a  Mild 1 4 1 4  Moderate 0 5 6 4  Severe 9 1 10 1 Concomitant IE 2 3 2 0 Pseudoaneurysm 2 1 0 0 Severe annular calcification 0 0 1 1 Prosthetic aortic valve Prosthetic mitral valve Surgery (n = 10) Device closure or observation (n = 10) P-value Surgery (n = 17) Device closure or observation (n = 9) P-value Long diameter (mm) 15.2 (8.8–17.6) 7.1 (4.5–10.9) 0.06 13.0 (11.2–18.0) 8.0 (6.5–12.4) 0.56 Short diameter (mm) 5.4 (5.1–6.3) 2.8 (2.3–4.6) 0.01 5.2 (3.6–6.0) 3.6 (3.2–4.4) 0.10 Perimeter (mm) 39.3 (25.3–41.7) 16.2 (13.5–23.9) 0.01 31.0 (23.9–52.0) 24.3 (17.2–31.7) 0.30 Area (mm2) 64.5 (52.8–119.8) 15.5 (10.5–29.1) 0.05 40.0 (22.0–67.7) 27.0 (17.0–40.0) 0.05 Involved angle (°) 58.7 (29.8–64.8) 32.4 (27.8–49.5) 0.11 44.0 (26.0–79.0) 26.0 (15.0–52.0) 0.54 Regurgitant grade 0.001a 0.04a  Mild 1 4 1 4  Moderate 0 5 6 4  Severe 9 1 10 1 Concomitant IE 2 3 2 0 Pseudoaneurysm 2 1 0 0 Severe annular calcification 0 0 1 1 a The P-values are obtained from the comparison between mild or moderate and severe groups using the Fisher’s exact test. IE, infective endocarditis. Figure 5 View largeDownload slide An example of failed device closure in mitral paravalvular leakage because of a long and narrow dehiscence tract. Paravalvular leakage with a thin and long tract (arrows) is located at the posterolateral aspect of the mitral prosthetic valve. LA, left atrium; LV, left ventricle. Figure 5 View largeDownload slide An example of failed device closure in mitral paravalvular leakage because of a long and narrow dehiscence tract. Paravalvular leakage with a thin and long tract (arrows) is located at the posterolateral aspect of the mitral prosthetic valve. LA, left atrium; LV, left ventricle. Discussion The major findings of this study were as follows: (i) cardiac CT could demonstrate the size and location of PVL, and the location of PVLs were matched with those noted on echocardiography and surgical inspection; (ii) the sizes of PVLs were significantly larger in patients with higher regurgitant grade; and (iii) surgical correction is preferred to patients who presented large PVLs on CT and severe regurgitant grade on echocardiography. Making the decision to perform a reoperation in patients with PVL is difficult because of the high surgical morbidity and recurrence rate (12–35%) of PVLs after reoperation.8 Recently, interventional occlusion for PVL has emerged as an alternative treatment strategy.10,21–24 For decision making in PVL treatment, anatomical information about the PVL, including the size, shape, and the three-dimensional relationship between the PVL and cardiac structure, are important. Echocardiography is the initial modality of choice that has excellent temporal resolution and real-time imaging capabilities. Notably, for the detection of PVLs, colour Doppler imaging can easily indicate the location and degree of paravalvular regurgitant jet. However, anatomical demonstration of a PVL using echocardiography often is compromised by metallic artefacts from the prosthetic valve and its limited sonic window. By providing more precise anatomical details, including the exact location and morphology of the PVLs on CT, pre-treatment planning could be better tailored and individualized.25 By simulating the entrance site of the guidewire through the PVL on a fluoroscopic view of CT to guide the percutaneous procedure, both the fluoroscopic time and radiation exposure could be reduced. For PVLs located in the medial aspect of the prosthetic mitral valve, it is challenging to perform a trans-septal access because of the difficulty of manipulating the acute angle with a guidewire.14 Moreover, the shape and length of the dehiscence tract are also determining factors. If the dehiscence tract is long and narrow, it can be difficult to negotiate the guidewire. If the lesion is crescent shaped or large that it cannot be occluded using occlusion devices, surgical management should be considered.14 These conditions can be clearly demonstrated on cardiac CT using reconstructed images.26,27 Detailed anatomic information with a familiar view for either surgeon or interventionist may be helpful for planning of treatment strategy. CT can also be used to evaluate PVL after transcatheter aortic valve implantation (TAVI). Conventional parasternal short-axis analysis using echocardiography can underestimate the extent of PVL with false negative studies up to 14%.28 Moreover, in contrast to angiography, echocardiography did not show good correlation with magnetic resonance imaging in the assessment of PVL after TAVI.29 Similar to the methods used to evaluate PVLs occurred after surgical valve replacement in this study, CT 3-dimensional multiplanar reconstruction images can demonstrate sagittal view and short-axis cuts (en face views) of the prosthetic frame geometry of TAVI device.30,31 To avoid overestimation of the dehiscence, further image reconstruction can be carried out by either rotating the imaging plane to centre the dehiscence or using a 360° rotational technique following the ‘double-checking’ method. The raysum technique can also be used to generate fluoroscopy-like images. Previously, aortic PVLs have been reported to be more commonly located in the right- and non-coronary cusps.32 However, in our present study, right- and left-coronary cusps are common sites of PVLs. Mitral PVLs were found to be evenly distributed around the sewing ring, which was in accord with the findings of a previous study.33 Considering the small sample size of our study, further studies with a larger population will be necessary. Limitations Our study had several limitations. First, there is a selection bias, only 46 patients with CT out of 131 with PVL detected on echocardiography were studied. The patients with CT performed were more symptomatic, had bigger ventricles and lower ejection fraction. Patients with severe regurgitant flow with large sizes of leakages might be included, and the results of this study may not be applicable for small lesions. However, the smallest sizes of aortic and mitral PVLs in our study were 4.4 and 1.5 mm, respectively. These values suggest the potential use of CT for evaluation of PVLs even for small lesions. Also, it may not be easy to distinguish between small leaks and pledgets or stitches because the density is similar during the angiographic phase. A way to avoid mistakes is to perform a pre-contrast gated scan. Second, observer bias should be considered because the readers had known about the presence of the PVL when they interpreted the CT images. Because patients in this study were collected at a tertiary referral centre and cardiac CT imaging was used as a preoperative planning method, thorough evaluations were performed for optimal patient care. To overcome this bias, all cardiac source images were interpreted by two readers who were blind to the size, location, shape, and number of PVLs and interobserver agreement was good. Third, CT interpretation may be highly dependent upon the data processing quality.34 In our study, we exclude two patients with suboptimal CT image quality because of motion artefacts. To measure the largest size of PVL during obtained cardiac cycles, we used systolic phase for aortic PVLs, and diastolic phase for mitral PVLs. However, because of cardiac motion artefacts and beam hardening artefacts from prosthetic valve, not all phases could be used for evaluation of PVLs, thus we could performed the comparison between systolic/diastolic phases in a subset of patients. Further studies to evaluate the changes of PVLs during different cardiac cycle phases would be of value. Nevertheless, our current findings may demonstrate a practical role for cardiac CT in the planning of PVL treatment by providing detailed anatomic data. Finally, because not all patients underwent operations and the limits of surgical records, we could not compare the morphological characteristics of PVLs detected on CT to surgical anatomy. However, we have shown that the morphology data on CT, long/short diameter, angle and length of PVLs, could be used to demonstrate PVLs and the information may aid in the planning of a treatment strategy. Conclusion In conclusion, cardiac CT can be used to detect and localize aortic and mitral PVLs. The visualization of anatomic information using a familiar view for either a surgeon or an interventionist may aid in the planning of a treatment strategy. Supplementary data Supplementary data are available at European Heart Journal - Cardiovascular Imaging online. Funding The work was supported by the Industrial Strategic technology development program (10072064) funded by the Ministry of Trade Industry and Energy (MI, Korea), the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (NRF-2016R1A1A1A05921207) and a grant (2017-7208) from the Asan Institute for Life Sciences, Asan Medical Centre, Seoul, Korea. Conflict of interest: none declared. References 1 Ionescu A , Fraser A , Butchart E. Prevalence and clinical significance of incidental paraprosthetic valvar regurgitation: a prospective study using transoesophageal echocardiography . Heart 2003 ; 89 : 1316 – 21 . http://dx.doi.org/10.1136/heart.89.11.1316 Google Scholar CrossRef Search ADS PubMed 2 Kliger C , Eiros R , Isasti G , Einhorn B , Jelnin V , Cohen H et al. Review of surgical prosthetic paravalvular leaks: diagnosis and catheter-based closure . Eur Heart J 2013 ; 34 : 638 – 49 . Google Scholar CrossRef Search ADS PubMed 3 Rallidis LS , Moyssakis IE , Ikonomidis I , Nihoyannopoulos P. Natural history of early aortic paraprosthetic regurgitation: a five-year follow-up . Am Heart J 1999 ; 138 : 351 – 7 . Google Scholar CrossRef Search ADS PubMed 4 Hammermeister K , Sethi GK , Henderson WG , Grover FL , Oprian C , Rahimtoola SH. Outcomes 15 years after valve replacement with a mechanical versus a bioprosthetic valve: final report of the Veterans Affairs randomized trial . J Am Coll Cardiol 2000 ; 36 : 1152 – 8 . Google Scholar CrossRef Search ADS PubMed 5 Ruiz CE , Jelnin V , Kronzon I , Dudiy Y , Del Valle-Fernandez R , Einhorn BN et al. Clinical outcomes in patients undergoing percutaneous closure of periprosthetic paravalvular leaks . J Am Coll Cardiol 2011 ; 58 : 2210 – 7 . http://dx.doi.org/10.1016/j.jacc.2011.03.074 Google Scholar CrossRef Search ADS PubMed 6 Bhindi R , Bull S , Schrale RG , Wilson N , Ormerod OJ. Surgery insight: percutaneous treatment of prosthetic paravalvular leaks . Nat Clin Pract Cardiovasc Med 2008 ; 5 : 140 – 7 . Google Scholar CrossRef Search ADS PubMed 7 Lesser JR , Han BK , Newell M , Schwartz RS , Pedersen W , Sorajja P. Use of cardiac CT angiography to assist in the diagnosis and treatment of aortic prosthetic paravalvular leak: a practical guide . J Cardiovasc Comput Tomogr 2015 ; 9 : 159 – 64 . Google Scholar CrossRef Search ADS PubMed 8 Echevarria JR , Bernal JM , Rabasa JM , Morales D , Revilla Y , Revuelta JM. Reoperation for bioprosthetic valve dysfunction: a decade of clinical experience . Eur J Cardiothorac Surg 1991 ; 5 : 523 – 6 . Google Scholar CrossRef Search ADS PubMed 9 Safi AM , Kwan T , Afflu E , Al Kamme A , Salciccioli L. Paravalvular regurgitation: a rare complication following valve replacement surgery . Angiology 2000 ; 51 : 479 – 87 . http://dx.doi.org/10.1177/000331970005100605 Google Scholar CrossRef Search ADS PubMed 10 Hein R. Percutaneous closure of paravalvular leaks . J Interv Cardiol 2006 ; 19 : S73 – 7 . Google Scholar CrossRef Search ADS 11 Pate GE , Al Zubaidi A , Chandavimol M , Thompson CR , Munt BI , Webb JG. Percutaneous closure of prosthetic paravalvular leaks: case series and review . Catheter Cardiovasc Interv 2006 ; 68 : 528 – 33 . Google Scholar CrossRef Search ADS PubMed 12 Hildick-Smith D , Behan MWH , De Giovanni J. Percutaneous closure of an aortic paravalvular leak via the transradial approach . Catheter Cardiovasc Interv 2007 ; 69 : 708 – 10 . http://dx.doi.org/10.1002/ccd.21043 Google Scholar CrossRef Search ADS PubMed 13 Momplaisir T , Matthews RV. Paravalvular mitral regurgitation treated with an amplatzer septal occluder device: a case report and review of the literature . J Invasive Cardiol 2007 ; 19 : E46 – 50 . Google Scholar PubMed 14 Ruiz CE , Cohen H , Del Valle-Fernandez R , Jelnin V , Perk G , Kronzon I. Closure of prosthetic paravalvular leaks: a long way to go . Eur Heart J Suppl 2010 ; 12 : E52 – 62 . Google Scholar CrossRef Search ADS 15 Koo HJ , Yang DH , Kang J-W , Han K , Chung CH , Song J-K et al. Demonstration of prosthetic aortic valve dehiscence in a patient with noninfectious aortitis by multimodality imaging: findings of echocardiography and computed tomography . Circulation 2013 ; 128 : 759 – 61 . Google Scholar CrossRef Search ADS PubMed 16 McKee PA , Castelli WP , McNamara PM , Kannel WB. The natural history of congestive heart failure: the Framingham study . N Engl J Med 1971 ; 285 : 1441 – 6 . http://dx.doi.org/10.1056/NEJM197112232852601 Google Scholar CrossRef Search ADS PubMed 17 Kim YJ , Yong HS , Kim SM , Kim JA , Yang DH , Hong YJ. Korean guidelines for the appropriate use of cardiac CT . Korean J Radiol 2015 ; 16 : 251 – 85 . http://dx.doi.org/10.3348/kjr.2015.16.2.251 Google Scholar CrossRef Search ADS PubMed 18 Tsai IC , Choi BW , Chan C , Jinzaki M , Kitagawa K , Yong HS. ASCI 2010 appropriateness criteria for cardiac computed tomography: a report of the Asian Society of Cardiovascular Imaging Cardiac Computed Tomography and Cardiac Magnetic Resonance Imaging Guideline Working Group . Int J Cardiovasc Imaging 2010 ; 26(Suppl. 1) : 1 – 15 . Google Scholar CrossRef Search ADS PubMed 19 Nishimura RA , Otto CM , Bonow RO , Carabello BA , Erwin JP 3rd , Fleisher LA. 2017 AHA/ACC focused update of the 2014 AHA/ACC guideline for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines . Circulation 2017 ; 135 : e1159 – 95 . Google Scholar CrossRef Search ADS PubMed 20 De Cicco G , Russo C , Moreo A , Beghi C , Fucci C , Gerometta P et al. Mitral valve periprosthetic leakage: anatomical observations in 135 patients from a multicentre study . Eur J Cardiothorac Surg 2006 ; 30 : 887 – 91 . Google Scholar CrossRef Search ADS PubMed 21 Boudjemline Y. Percutaneous closure of a paravalvular mitral regurgitation with Amplatzer and coil prostheses . Arch Mal Coeur Vaiss 2002 ; 95 : 483 – 6 . Google Scholar PubMed 22 Piechaud JF. Percutaneous closure of mitral paravalvular leak . J Interv Cardiol 2003 ; 16 : 153 – 5 . http://dx.doi.org/10.1046/j.1540-8183.2003.08028.x Google Scholar CrossRef Search ADS PubMed 23 Kort HW , Sharkey AM , Balzer DT. Novel use of the Amplatzer duct occluder to close perivalvar leak involving a prosthetic mitral valve . Catheter Cardiovasc Interv 2004 ; 61 : 548 – 51 . http://dx.doi.org/10.1002/ccd.10785 Google Scholar CrossRef Search ADS PubMed 24 Webb JG , Pate GE , Munt BI. Percutaneous closure of an aortic prosthetic paravalvular leak with an Amplatzer duct occluder . Catheter Cardiovasc Interv 2005 ; 65 : 69 – 72 . http://dx.doi.org/10.1002/ccd.20337 Google Scholar CrossRef Search ADS PubMed 25 Pate GE , Thompson CR , Munt BI , Webb JG. Techniques for percutaneous closure of prosthetic paravalvular leaks . Catheter Cardiovasc Interv 2006 ; 67 : 158 – 66 . http://dx.doi.org/10.1002/ccd.20560 Google Scholar CrossRef Search ADS PubMed 26 Odell JA , Orszulak TA. Surgical repair and reconstruction of valvular lesions . Curr Opin Cardiol 1995 ; 10 : 135 – 43 . http://dx.doi.org/10.1097/00001573-199503000-00007 Google Scholar CrossRef Search ADS PubMed 27 Dion R. Ischemic mitral regurgitation: when and how should it be corrected? J Heart Valve Dis 1993 ; 2 : 536 – 43 . Google Scholar PubMed 28 Geleijnse ML , Di Martino LFM , Vletter WB , Ren B , Galema TW , Van Mieghem NM et al. Limitations and difficulties of echocardiographic short-axis assessment of paravalvular leakage after corevalve transcatheter aortic valve implantation . Cardiovasc ultrasound 2015 ; 14 : 37. Google Scholar CrossRef Search ADS 29 Sherif M , Abdel-Wahab M , Beurich H-W , Stöcker B , Zachow D , Geist V et al. Haemodynamic evaluation of aortic regurgitation after transcatheter aortic valve implantation using cardiovascular magnetic resonance . EuroIntervention 2011 ; 7 : 57 – 63 . Google Scholar CrossRef Search ADS PubMed 30 Abdelghani M , Soliman OI , Schultz C , Vahanian A , Serruys PW. Adjudicating paravalvular leaks of transcatheter aortic valves: a critical appraisal . Eur Heart J 2016 ; 37 : 2627 – 44 . Google Scholar CrossRef Search ADS PubMed 31 Schultz CJ , Lauritsch G , Van Mieghem N , Rohkohl C , Serruys PW , van Geuns RJ et al. Rotational angiography with motion compensation: first-in-man use for the 3D evaluation of transcatheter valve prostheses . EuroIntervention 2015 ; 11 : 442 – 9 . Google Scholar CrossRef Search ADS PubMed 32 De Cicco G , Lorusso R , Colli A , Nicolini F , Fragnito C , Grimaldi T et al. Aortic valve periprosthetic leakage: anatomic observations and surgical results . Ann Thorac Surg 2005 ; 79 : 1480. Google Scholar CrossRef Search ADS PubMed 33 Foster GP , Isselbacher EM , Rose GA , Torchiana DF , Akins CW , Picard MH. Accurate localization of mitral regurgitant defects using multiplane transesophageal echocardiography . Ann Thorac Surg 1998 ; 65 : 1025 – 31 . Google Scholar CrossRef Search ADS PubMed Published on behalf of the European Society of Cardiology. All rights reserved. © The Author(s) 2018. For permissions, please email: journals.permissions@oup.com. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png European Heart Journal – Cardiovascular Imaging Oxford University Press

Paravalvular leakage in patients with prosthetic heart valves: cardiac computed tomography findings and clinical features

Loading next page...
 
/lp/ou_press/paravalvular-leakage-in-patients-with-prosthetic-heart-valves-cardiac-cN9Tz70IkP
Publisher
Oxford University Press
Copyright
Published on behalf of the European Society of Cardiology. All rights reserved. © The Author(s) 2018. For permissions, please email: journals.permissions@oup.com.
ISSN
2047-2404
D.O.I.
10.1093/ehjci/jex341
Publisher site
See Article on Publisher Site

Abstract

Abstract Aims Morphological characteristics of paravalvular leakage (PVL) have not yet been well characterized by computed tomography (CT). The purpose of this study was to demonstrate the morphological characteristics of PVLs using cardiac CT in patients diagnosed with PVL. Methods and results Between May 2011 and December 2013, 46 patients who had been diagnosed with PVL and underwent cardiac CT were included in this study. On CT, the characteristics of PVLs including number, size, location, and shape are described. Inter- and intra-observer agreement of CT were assessed. The extent of PVL on CT and the degree of regurgitant grade on echocardiography are compared. The size of PVLs were compared between patients who underwent surgical correction and who treated with percutaneous device closure or observed without treatment. All PVLs detected on surgical filed were the same with the number and locations of PVLs demonstrated on CT. Interobserver agreement for CT measurements of the PVL was good, ranging from 0.78 to 0.99. The sizes of PVL were smaller in patients who treated with device closure or observed without treatment compared to those in patients who underwent surgical correction (median areas of aortic PVL, 64.5 vs. 15.5 mm2 and mitral PVL, 40.0 vs. 27.0 mm2). PVL sizes were larger in patients with higher regurgitant grades on echocardiography (P < 0.05). Conclusion Cardiac CT can demonstrate the location and size of PVL. The size of PVLs are larger in patients who managed by surgical correction. PVL size measured on CT is correlated with the regurgitant grade on echocardiography. paravalvular leakage, computed tomography, prosthetic valve Introduction Paravalvular leakage (PVL) is defined as abnormal communication between the sewing ring and valve annulus.1,2 The prevalence of PVL in patients who underwent valve replacement surgery has been reported to vary from 1% to 10% around the aortic valve1,3 and from 3% to 15% around the mitral valve.1,4 Although small PVLs are common and usually benign, they have been associated with poorer survival.1,3 Notably, the incidence of major leaks ranges from 2% to 10%.5–7 As a treatment for severe PVL, surgical corrections have been recommended to perform either repair of the leak or re-replacement of the valve. However, both procedures have failure rates that range from 12% to 35%, and such cases increase the need for re-intervention.8 Moreover, recurrence of a PVL can also reduce survival rates among surgical candidates.9 Recently, percutaneous device closure has been introduced as an alternative option to treat PVL.5,6,10–14 To reduce the failure rate associated with these treatment methods, a thorough review of the morphological characteristics of the PVLs should be performed before conducting any of these procedures. Standards of care dictate that imaging studies for an initial diagnosis of PVL are usually made by echocardiography. Recently, several case series have evaluated the utilization of cardiac computed tomography (CT) to evaluate PVLs.2,5,14,15 These two different imaging modalities each have their own benefits as flow velocity helps clinicians to interpret disease severity, whereas cardiac CT may be better suited for preoperative planning by yielding more precise anatomical details. Moreover, cardiac CT can help a physician select an optimal treatment method by providing a more accurate localization and visualization of these leaks. Although several anatomical reports have been published based on surgical findings,5,14,15 CT-based morphological details have not yet been well characterized. We hypothesized that the morphological characteristics of PVL could be described thoroughly using cardiac CT. Herein, we aimed to demonstrate the morphological characteristics of PVLs, and its potential clinical impact using cardiac CT in patients who had detected PVL on echocardiography. Materials and methods Study population The institutional review board of our institution approved this retrospective study and waived the requirement to obtain informed consent. Between May 2011 and December 2013, 131 patients who showed PVL by echocardiography were retrospectively reviewed. They underwent echocardiography for routine postoperative care (n = 82) or a symptomatic evaluation (n = 49). Symptomatic patients were defined based on the presence of congestive heart failure, systemic emboli, haemolysis, and new-onset atrioventricular block. Congestive heart failure was defined based on criteria developed by the Framingham Heart Study.16 Haemolysis was defined as a serum lactate dehydrogenase level >460 U/L along with any two of the following criteria: (i) level of blood haemoglobin <13.8 g/dL for males or <12.4 g/dL for females; (ii) level of serum haptoglobin <50 mg/dL; or (iii) reticulocyte count >2%. After excluding two patients with suboptimal CT image quality due to severe motion artefacts, 46 patients (mean age 62.0 years, 24 males) subsequently underwent electrocardiogram-gated cardiac CT and were ultimately enrolled. Reasons for the examination by cardiac CT were most commonly for demonstration of paravalvular anatomy to plan treatment of a PVL (n = 44), followed by evaluation of coronary artery disease (n = 2).17–19 All CT scans were obtained within 7 days of the initial transoesophageal echocardiography (3 ± 2 days). Patient demographics, clinical findings, echocardiography findings, and surgical records were reviewed based on electronic medical records. Overall mean follow-up periods were 410 ± 253 days. CT technique Retrospective electrocardiogram (ECG)-gated cardiac CT was performed using a second-generation dual-source CT (Somatom Definition Flash, Siemens, Erlangen, Germany). The scan delay was determined using a bolus tracking method with a threshold of 100 HU, and the region of interest was placed in the ascending aorta. Images were obtained after injecting 60–80 mL iomeprol-400 (Iomeron; Bracco Imaging, Milan, Italy), followed by 40 mL saline chaser at 4 mL/s. The body size radiation dose-adaptive protocol, in which the tube voltage and current time product were adjusted according to the patient body habitus, was used to reduce the radiation dose as follows: 80–120 kVp tube voltage, 185–380 mA tube current per rotation with a 280 ms gantry rotation time, and 128 × 0.6 mm collimation. ECG-based tube current modulation was applied (R–R interval, 30–80%). The mean dose length product was 967.1 [standard deviation (SD) 450.3] mGy⋅cm and the mean effective dose was 14.5 (SD 8.6) mSv. CT image analysis Image reconstruction CT images were reconstructed in a 5% R–R interval using a 1-mm slice and a B26 kernel, which were transferred to an external workstation (Aquarius, TeraRecon, Foster City, CA, USA) for post-processing. Before interpreting the cardiac CTs, details of the echocardiography regarding the size, location, lesion number, and shape of the PVL were blinded, except the presence of PVL. Evaluation of the PVL was independently performed by two radiologists for the following steps (see Supplementary data online, Videos S1 and S2). First, cardiac phases with the least motion and beam-hardening artefacts were selected. The least amount of motion around the prosthesis on CT images could be distinguished by reviewing different intervals of the cardiac phases. The largest size of PVL during whole obtained cardiac cycles is measured. Because of cardiac motion artefacts and beam hardening artefacts from prosthetic valve, not all phases could be used for evaluation of PVLs. We performed the comparison of PVL sizes between systolic and diastolic phases in a subset of patients. In consequence, aortic PVLs are measured during the systolic phase (20–40% R–R) and mitral PVLs are drawn in the diastolic phase (65–85% R–R). Second, multiplanar reformatted (MPR) images were generated by sagittal view (Figure 1A and D) and en face views of the prosthetic valves (Figure 1B and E). Third, the morphological characteristics of the PVL were determined. A contrast media-filled lesion adjacent to the prosthetic valve that was continuous between the two cardiac chambers was defined as a PVL. The PVLs are frequently small and the peri-dehiscence tissues may be thin, so to avoid overestimation of the true area of the dehiscence, further image reconstruction was carried out by either rotating the imaging plane to centre the dehiscence or using a 360° rotational technique following the ‘double-checking’ method. Fourth, en face views of the volume rendering images were provided as surgeon-friendly images (surgeon’s view; Figure 1C and F). The raysum technique was used to generate fluoroscopy-like images (fluoroscopic projection view; Figure 1G). The entrance site to the PVL was marked on fluoroscopic views with diverse angles (see Supplementary data online, Video S3). The raysum images with fluoroscopic views for an interventional cardiologist was generated using a separate workstation (Advantage Server, General Electric, MI, USA). Figure 1 View largeDownload slide A 67-year-old woman who presented with paravalvular leakage (PVL) at the prosthetic aortic and mitral valves. Sagittal (A), en face (B), and surgeon’s views (C) of mitral plane images showing a PVL (arrow) in the antero-lateral aspect (11 o’clock) of the prosthetic valve. The left circumflex artery indicates the lateral direction of the mitral annulus. Sagittal (D), en face (E), and surgeon’s views (F) of aortic valve (asterisk) images showing PVL (arrow) near the left coronary artery orifice. (G) Mitral (right arrow) and aortic (left arrow) PVLs are noted in fluoroscopic view which was obtained using the ray sum technique. (H) The two lesions were successfully occluded using vascular plugs (arrows) by a percutaneous approach. Ao, aorta; CAU, caudal; LA, left atrium; LAA, left atrial appendage; LAO, left anterior oblique; LV, left ventricle; PAV, prosthetic aortic valve; PMV, prosthetic mitral valve; RA, right atrium. Figure 1 View largeDownload slide A 67-year-old woman who presented with paravalvular leakage (PVL) at the prosthetic aortic and mitral valves. Sagittal (A), en face (B), and surgeon’s views (C) of mitral plane images showing a PVL (arrow) in the antero-lateral aspect (11 o’clock) of the prosthetic valve. The left circumflex artery indicates the lateral direction of the mitral annulus. Sagittal (D), en face (E), and surgeon’s views (F) of aortic valve (asterisk) images showing PVL (arrow) near the left coronary artery orifice. (G) Mitral (right arrow) and aortic (left arrow) PVLs are noted in fluoroscopic view which was obtained using the ray sum technique. (H) The two lesions were successfully occluded using vascular plugs (arrows) by a percutaneous approach. Ao, aorta; CAU, caudal; LA, left atrium; LAA, left atrial appendage; LAO, left anterior oblique; LV, left ventricle; PAV, prosthetic aortic valve; PMV, prosthetic mitral valve; RA, right atrium. Morphological characteristics of PVLs on CT The location and extent of each PVL was marked based on a clock-wise format on en face views of CT images (Figure 2). Twelve o’clock was assigned to the commissure between the left- and right-coronary sinuses, 4 o’clock to the commissure between the right- and non-coronary sinuses, and 8 o’clock to the commissure between the left- and non-coronary sinuses. A reference line was drawn from the commissure of the left- and right-coronary cusp to the mid-portion of the non-coronary sinus. Similar to an aortic valve assessment, the mitral PVL location could also be reported. The mid-points of the anterior and posterior annulus were represented as 12 and 6 o’clock, respectively. The postero-medial commissure and interatrial septum were at 3 o’clock, while the antero-lateral commissure and atrial appendage were at 9 o’clock.20 Figure 2 View largeDownload slide Distribution of paravalvular leakages in prosthetic (A) aortic and (B) mitral valves. AO, aorta; LA, left atrium; LAA, left atrial appendage; LCC, left coronary cusp; NCC, non-coronary cusp; RA, right atrium; RCC, right coronary cusp; 6, 6 o’clock direction; 12, 12 o’clock direction. Figure 2 View largeDownload slide Distribution of paravalvular leakages in prosthetic (A) aortic and (B) mitral valves. AO, aorta; LA, left atrium; LAA, left atrial appendage; LCC, left coronary cusp; NCC, non-coronary cusp; RA, right atrium; RCC, right coronary cusp; 6, 6 o’clock direction; 12, 12 o’clock direction. Based on the centre point of the dehiscence, the location of the PVL was classified as follows: (i) right-coronary cusp (12–4 o’clock), non-coronary cusp (4–8 o’clock), and left-coronary cusp (8–12 o’clock) in the aortic prosthetic valve and (ii) anterior (directions of the hour hand of a clock at 10:30–1:30), medial (1:30–4:30), posterior (4:30–7:30), and lateral (7:30–10:30) in the mitral prosthetic valve. For anatomical descriptions of the PVL, the size, shape, location, and number of the PVLs were determined based on MPR images. On an en face view of the prosthetic valve, the long- and short-diameters, perimeter, area, and involved angle of the PVL were measured to describe the extent of the PVL (see Supplementary data online, Figure S1). The length of the dehiscence tract was measured on a profile view of the prosthetic valve that was centred at the dehiscence. The shape of the PVL was categorized as round, ovoid, or crescent using a ratio of the long diameter to short diameter (L/S ratio) as follows: (i) round shape, L/S ratio of two or less; (ii) ovoid shape, L/S ratio between two and five; and (iii) crescent shape, L/S ratio of more than five. Treatment method Treatment strategies were decided based on patient symptoms, co-morbidity, regurgitant degree with heart function on echocardiography, and the morphological details on cardiac CT. For patients who had undergone cardiac CT, multidisplinary heart team conferences that included the cardiologist, cardiac surgeons, and radiologists were held. The degree of paravalvular regurgitation through the dehiscence was graded as either mild, moderate, or severe based on the color Doppler jet area on echocardiography. In patients with mild PVL and no symptoms, careful follow-up was recommended. In cases of moderate or severe leakage with clinical symptoms, either surgical or interventional treatment was recommended. Before repair of the PVLs, the surgeon’s view or fluoroscopic projection view was provided to the operator. During PVL repair, the morphological details of the PVLs were recorded by the operator regarding the location, number, and shape of the dehiscence. Statistical analysis Continuous variables were analysed using independent samples t-test, and the Fisher’s exact test was used to analyse categorical variables. By the Kruskal–Wallis analysis, sizes of PVLs on CT were compared according to the degree of regurgitant grade on echocardiography. Interobserver agreements on the size, location, and number of lesions on CT were obtained using a two-way random model intra-class correlation coefficient with consistency assumption. A two-sided P-value <0.05 was considered to indicate a statistically significant difference. Statistical analyses were performed using the Statistical Package for the Social Sciences (version 18 for Windows; SPSS Inc., Chicago, IL, USA). Results Patient characteristics Baseline characteristics of patients are summarized in Table 1. Patients who had undergone CT were more symptomatic (74% vs. 18%, P < 0.001) and showed decreased LV systolic function. On CT, a total of 48 PVLs were noted in 46 patients. One patient had two sites of mitral PVL, and another patient presented both aortic and mitral PVLs (Figure 1). CT characteristics and treatment methods for aortic and mitral PVLs are noted in Table 2. Left and right coronary sinuses are common sites of aortic PVLs. Lateral portion (42%) of the mitral annulus was the most common site of mitral PVL. Half of the PVLs had a round shape, which indicated potentially adequate for device occlusion. Inter- and intra-observer agreement for CT measurements of PVLs were good to excellent (range 0.78–0.99) (see Supplementary data online, Figure S2). Table 1 Patient characteristics Patients with CT (n = 46) Patients without CT (n = 85) P-value Age (years) 62.1 ± 11.3 60.6 ± 13.4 0.52 Male gender 24 (52) 45 (53) 1.000 Body weight 58.8 ± 9.8 60.9 ± 11.2 0.28 Height 162.2 ± 8.3 161.3 ± 9.5 0.59 Heart rate 75.0 ± 16.7 76.4 ± 16.4 0.64 Symptom 34 (74)a 15 (18)a <0.001  Asymptomatic 12 (26) 70 (82) <0.001  Congestive heart failure 31 (67) 12 (14) <0.001  Haemolysis 4 (9) 1 (1) 0.05  Systemic emboli 1 (2) 1 (1) 1.00  New-onset AV block 1 (2) 2 (2) 1.00 Laboratory findings  Haemoglobin (g/dL) 10.6 ± 2.1 11.8 ± 2.1 0.002  Creatinine (mg/dL) 1.2 ± 1.3 1.1 ± 1.1 0.89 Involved prosthetic valve  Aortic valve 19 (41) 45 (53)  Mitral valve 24 (52)b 37 (44)  Tricuspid valve 2 (4) 3 (4)  Both aortic and mitral valves 1 (2) 0 (0) Concomitant IE 7 (15) 2 (2) 0.003 Echocardiography  LVEF (%) 52.7 ± 11.4 56.5 ± 10.4 0.06  End diastolic volume (mL) 147.9 ± 69.7 124.8 ± 48.0 0.03  End systolic volume (mL) 74.1 ± 46.6 57.1 ± 34.1 0.02 Number of previous open heart surgery  One 31 (67) 74 (87)  Two (Redo-) 12 (26) 9 (11)  Three (Trido-) 3 (7) 2 (2) Valve type 0.58  Tissue valve 7 (15) 9 (11)  Prosthetic valve 39 (85) 76 (89) Patients with CT (n = 46) Patients without CT (n = 85) P-value Age (years) 62.1 ± 11.3 60.6 ± 13.4 0.52 Male gender 24 (52) 45 (53) 1.000 Body weight 58.8 ± 9.8 60.9 ± 11.2 0.28 Height 162.2 ± 8.3 161.3 ± 9.5 0.59 Heart rate 75.0 ± 16.7 76.4 ± 16.4 0.64 Symptom 34 (74)a 15 (18)a <0.001  Asymptomatic 12 (26) 70 (82) <0.001  Congestive heart failure 31 (67) 12 (14) <0.001  Haemolysis 4 (9) 1 (1) 0.05  Systemic emboli 1 (2) 1 (1) 1.00  New-onset AV block 1 (2) 2 (2) 1.00 Laboratory findings  Haemoglobin (g/dL) 10.6 ± 2.1 11.8 ± 2.1 0.002  Creatinine (mg/dL) 1.2 ± 1.3 1.1 ± 1.1 0.89 Involved prosthetic valve  Aortic valve 19 (41) 45 (53)  Mitral valve 24 (52)b 37 (44)  Tricuspid valve 2 (4) 3 (4)  Both aortic and mitral valves 1 (2) 0 (0) Concomitant IE 7 (15) 2 (2) 0.003 Echocardiography  LVEF (%) 52.7 ± 11.4 56.5 ± 10.4 0.06  End diastolic volume (mL) 147.9 ± 69.7 124.8 ± 48.0 0.03  End systolic volume (mL) 74.1 ± 46.6 57.1 ± 34.1 0.02 Number of previous open heart surgery  One 31 (67) 74 (87)  Two (Redo-) 12 (26) 9 (11)  Three (Trido-) 3 (7) 2 (2) Valve type 0.58  Tissue valve 7 (15) 9 (11)  Prosthetic valve 39 (85) 76 (89) Data represent the mean ± standard deviation or numbers of patients with the percentage indicated in parentheses. a Total number of patients who had one or more symptoms. b One of them had two sites of mitral paravalvular leakages. AV, atrioventricular; CT, computed tomography; IE, infective endocarditis; LVEF, left ventricular ejection fraction. Table 1 Patient characteristics Patients with CT (n = 46) Patients without CT (n = 85) P-value Age (years) 62.1 ± 11.3 60.6 ± 13.4 0.52 Male gender 24 (52) 45 (53) 1.000 Body weight 58.8 ± 9.8 60.9 ± 11.2 0.28 Height 162.2 ± 8.3 161.3 ± 9.5 0.59 Heart rate 75.0 ± 16.7 76.4 ± 16.4 0.64 Symptom 34 (74)a 15 (18)a <0.001  Asymptomatic 12 (26) 70 (82) <0.001  Congestive heart failure 31 (67) 12 (14) <0.001  Haemolysis 4 (9) 1 (1) 0.05  Systemic emboli 1 (2) 1 (1) 1.00  New-onset AV block 1 (2) 2 (2) 1.00 Laboratory findings  Haemoglobin (g/dL) 10.6 ± 2.1 11.8 ± 2.1 0.002  Creatinine (mg/dL) 1.2 ± 1.3 1.1 ± 1.1 0.89 Involved prosthetic valve  Aortic valve 19 (41) 45 (53)  Mitral valve 24 (52)b 37 (44)  Tricuspid valve 2 (4) 3 (4)  Both aortic and mitral valves 1 (2) 0 (0) Concomitant IE 7 (15) 2 (2) 0.003 Echocardiography  LVEF (%) 52.7 ± 11.4 56.5 ± 10.4 0.06  End diastolic volume (mL) 147.9 ± 69.7 124.8 ± 48.0 0.03  End systolic volume (mL) 74.1 ± 46.6 57.1 ± 34.1 0.02 Number of previous open heart surgery  One 31 (67) 74 (87)  Two (Redo-) 12 (26) 9 (11)  Three (Trido-) 3 (7) 2 (2) Valve type 0.58  Tissue valve 7 (15) 9 (11)  Prosthetic valve 39 (85) 76 (89) Patients with CT (n = 46) Patients without CT (n = 85) P-value Age (years) 62.1 ± 11.3 60.6 ± 13.4 0.52 Male gender 24 (52) 45 (53) 1.000 Body weight 58.8 ± 9.8 60.9 ± 11.2 0.28 Height 162.2 ± 8.3 161.3 ± 9.5 0.59 Heart rate 75.0 ± 16.7 76.4 ± 16.4 0.64 Symptom 34 (74)a 15 (18)a <0.001  Asymptomatic 12 (26) 70 (82) <0.001  Congestive heart failure 31 (67) 12 (14) <0.001  Haemolysis 4 (9) 1 (1) 0.05  Systemic emboli 1 (2) 1 (1) 1.00  New-onset AV block 1 (2) 2 (2) 1.00 Laboratory findings  Haemoglobin (g/dL) 10.6 ± 2.1 11.8 ± 2.1 0.002  Creatinine (mg/dL) 1.2 ± 1.3 1.1 ± 1.1 0.89 Involved prosthetic valve  Aortic valve 19 (41) 45 (53)  Mitral valve 24 (52)b 37 (44)  Tricuspid valve 2 (4) 3 (4)  Both aortic and mitral valves 1 (2) 0 (0) Concomitant IE 7 (15) 2 (2) 0.003 Echocardiography  LVEF (%) 52.7 ± 11.4 56.5 ± 10.4 0.06  End diastolic volume (mL) 147.9 ± 69.7 124.8 ± 48.0 0.03  End systolic volume (mL) 74.1 ± 46.6 57.1 ± 34.1 0.02 Number of previous open heart surgery  One 31 (67) 74 (87)  Two (Redo-) 12 (26) 9 (11)  Three (Trido-) 3 (7) 2 (2) Valve type 0.58  Tissue valve 7 (15) 9 (11)  Prosthetic valve 39 (85) 76 (89) Data represent the mean ± standard deviation or numbers of patients with the percentage indicated in parentheses. a Total number of patients who had one or more symptoms. b One of them had two sites of mitral paravalvular leakages. AV, atrioventricular; CT, computed tomography; IE, infective endocarditis; LVEF, left ventricular ejection fraction. Table 2 CT characteristics and treatment of paravalvular leakages (per-lesion based analysis) Prosthetic aortic valve Prosthetic mitral valve P-value Long diameter (mm) 9.5 (5.7–15.4) 12.4 (6.6–15.9) 0.31 Short diameter (mm) 5.0 (2.9–5.5) 4.3 (3.2–5.7) 0.76 Perimeter (mm) 24.0 (14.9–39.0) 29.8 (21.7–45.7) 0.17 Area (mm2) 41.0 (14.3–77.8) 35.3 (17.5–60.0) 0.28 Involved angle (°) 33.0 (28.8–59.5) 31.0 (21.8–58.0) 0.69 Location  Left-coronary 8 (40)  Right-coronary 8 (40)  Non-coronary 4 (20)  Anterior 3 (12)  Posterior 8 (31)  Lateral 11 (42)  Medial 4 (15) Shape 0.80  Round 10 (50) 12 (46)  Ovoid/elongated 9/1 (45/5) 10/4 (38/15) Regurgitant grade 0.62  Mild 5 (25) 5 (19)  Moderate 5 (25) 10 (38)  Severe 10 (50) 11 (42) Concomitant IE 5 (25) 2 (8) Pseudoaneurysm 3 (15) 0 (0) Severe annular calcification 0 (0) 2 (8) Treatment  Surgery 10 (50) 17 (65)  Device closure 1 [1/0] (5)a 5 [3/2] (19)a  Observation 9 (45) 4 (15) Prosthetic aortic valve Prosthetic mitral valve P-value Long diameter (mm) 9.5 (5.7–15.4) 12.4 (6.6–15.9) 0.31 Short diameter (mm) 5.0 (2.9–5.5) 4.3 (3.2–5.7) 0.76 Perimeter (mm) 24.0 (14.9–39.0) 29.8 (21.7–45.7) 0.17 Area (mm2) 41.0 (14.3–77.8) 35.3 (17.5–60.0) 0.28 Involved angle (°) 33.0 (28.8–59.5) 31.0 (21.8–58.0) 0.69 Location  Left-coronary 8 (40)  Right-coronary 8 (40)  Non-coronary 4 (20)  Anterior 3 (12)  Posterior 8 (31)  Lateral 11 (42)  Medial 4 (15) Shape 0.80  Round 10 (50) 12 (46)  Ovoid/elongated 9/1 (45/5) 10/4 (38/15) Regurgitant grade 0.62  Mild 5 (25) 5 (19)  Moderate 5 (25) 10 (38)  Severe 10 (50) 11 (42) Concomitant IE 5 (25) 2 (8) Pseudoaneurysm 3 (15) 0 (0) Severe annular calcification 0 (0) 2 (8) Treatment  Surgery 10 (50) 17 (65)  Device closure 1 [1/0] (5)a 5 [3/2] (19)a  Observation 9 (45) 4 (15) Data represent the median and interquartile ranges or numbers of patients with the percentage indicated in parentheses. a Success/failed cases of percutaneous device closure. IE, infective endocarditis. Table 2 CT characteristics and treatment of paravalvular leakages (per-lesion based analysis) Prosthetic aortic valve Prosthetic mitral valve P-value Long diameter (mm) 9.5 (5.7–15.4) 12.4 (6.6–15.9) 0.31 Short diameter (mm) 5.0 (2.9–5.5) 4.3 (3.2–5.7) 0.76 Perimeter (mm) 24.0 (14.9–39.0) 29.8 (21.7–45.7) 0.17 Area (mm2) 41.0 (14.3–77.8) 35.3 (17.5–60.0) 0.28 Involved angle (°) 33.0 (28.8–59.5) 31.0 (21.8–58.0) 0.69 Location  Left-coronary 8 (40)  Right-coronary 8 (40)  Non-coronary 4 (20)  Anterior 3 (12)  Posterior 8 (31)  Lateral 11 (42)  Medial 4 (15) Shape 0.80  Round 10 (50) 12 (46)  Ovoid/elongated 9/1 (45/5) 10/4 (38/15) Regurgitant grade 0.62  Mild 5 (25) 5 (19)  Moderate 5 (25) 10 (38)  Severe 10 (50) 11 (42) Concomitant IE 5 (25) 2 (8) Pseudoaneurysm 3 (15) 0 (0) Severe annular calcification 0 (0) 2 (8) Treatment  Surgery 10 (50) 17 (65)  Device closure 1 [1/0] (5)a 5 [3/2] (19)a  Observation 9 (45) 4 (15) Prosthetic aortic valve Prosthetic mitral valve P-value Long diameter (mm) 9.5 (5.7–15.4) 12.4 (6.6–15.9) 0.31 Short diameter (mm) 5.0 (2.9–5.5) 4.3 (3.2–5.7) 0.76 Perimeter (mm) 24.0 (14.9–39.0) 29.8 (21.7–45.7) 0.17 Area (mm2) 41.0 (14.3–77.8) 35.3 (17.5–60.0) 0.28 Involved angle (°) 33.0 (28.8–59.5) 31.0 (21.8–58.0) 0.69 Location  Left-coronary 8 (40)  Right-coronary 8 (40)  Non-coronary 4 (20)  Anterior 3 (12)  Posterior 8 (31)  Lateral 11 (42)  Medial 4 (15) Shape 0.80  Round 10 (50) 12 (46)  Ovoid/elongated 9/1 (45/5) 10/4 (38/15) Regurgitant grade 0.62  Mild 5 (25) 5 (19)  Moderate 5 (25) 10 (38)  Severe 10 (50) 11 (42) Concomitant IE 5 (25) 2 (8) Pseudoaneurysm 3 (15) 0 (0) Severe annular calcification 0 (0) 2 (8) Treatment  Surgery 10 (50) 17 (65)  Device closure 1 [1/0] (5)a 5 [3/2] (19)a  Observation 9 (45) 4 (15) Data represent the median and interquartile ranges or numbers of patients with the percentage indicated in parentheses. a Success/failed cases of percutaneous device closure. IE, infective endocarditis. Comparisons between CT and echocardiography The numbers and locations of PVLs detected on CT were the same with echocardiography except one patient. In the patient, two mitral PVLs on transoesophageal echocardiography had shown as a single elongated PVL on CT and finally confirmed to be a single lesion on the surgical inspection (Figure 3). The diameter, perimeter, area, and involved angle of PVLs measured on CT were significantly larger in patients with higher regurgitant grades on echocardiography (Table 3 and Figure 4). However, the length of PVLs were not statistically different according to the degree of regurgitant grades. Table 3 Comparison of the size of paravalvular leakage according to the regurgitant grade using the Kruskal–Wallis test (per-lesion based analysis) Mild Moderate Severe P-value Prosthetic aortic valve Long diameter (mm) 4.4 (3.9–4.9) 7.5 (7.4–12.0) 15.5 (9.8–18.6) 0.02 Short diameter (mm) 2.6 (2.2–2.7) 3.5 (2.9–5.0) 5.6 (5.1–7.1) 0.01 Perimeter (mm) 14.6 (10.0–15.0) 24.0 (22.3–26.0) 39.9 (28.5–42.3) 0.04 Area (mm2) 12.0 (6.0–16.0) 23.8 (15.0–30.9) 86.5 (58.5–134.8) 0.01 Involved angle (°) 27.0 (19.0–32.0) 33.0 (31.7–55.0) 60.0 (36.3–65.8) 0.05 Length (mm) 11.0 (8.3–12.9) 9.0 (8.2–9.1) 7.5 (5.5–12.0) 0.75 Prosthetic mitral valve Long diameter (mm) 8.0 (3.6–9.3) 9.2 (5.6–13.9) 18.0 (12.6–25.5) 0.01 Short diameter (mm) 3.3 (1.6–4.1) 3.6 (2.8–4.3) 5.8 (5.0–7.6) 0.01 Perimeter (mm) 24.0 (8.3–24.3) 25.2 (17.2–29.9) 52.0 (31.8–63.3) 0.01 Area (mm2) 19.0 (5.0–27.0) 22.5 (14.8–39.9) 62.0 (40.0–86.5) 0.005 Involved angle (°) 15.0 (14.0–29.0) 27.0 (19.5–46.3) 50.0 (30.0–98.0) 0.04 Length (mm) 8.0 (6.9–8.0) 9.7 (8.2–11.4) 9.3 (7.2–9.9) 0.20 Mild Moderate Severe P-value Prosthetic aortic valve Long diameter (mm) 4.4 (3.9–4.9) 7.5 (7.4–12.0) 15.5 (9.8–18.6) 0.02 Short diameter (mm) 2.6 (2.2–2.7) 3.5 (2.9–5.0) 5.6 (5.1–7.1) 0.01 Perimeter (mm) 14.6 (10.0–15.0) 24.0 (22.3–26.0) 39.9 (28.5–42.3) 0.04 Area (mm2) 12.0 (6.0–16.0) 23.8 (15.0–30.9) 86.5 (58.5–134.8) 0.01 Involved angle (°) 27.0 (19.0–32.0) 33.0 (31.7–55.0) 60.0 (36.3–65.8) 0.05 Length (mm) 11.0 (8.3–12.9) 9.0 (8.2–9.1) 7.5 (5.5–12.0) 0.75 Prosthetic mitral valve Long diameter (mm) 8.0 (3.6–9.3) 9.2 (5.6–13.9) 18.0 (12.6–25.5) 0.01 Short diameter (mm) 3.3 (1.6–4.1) 3.6 (2.8–4.3) 5.8 (5.0–7.6) 0.01 Perimeter (mm) 24.0 (8.3–24.3) 25.2 (17.2–29.9) 52.0 (31.8–63.3) 0.01 Area (mm2) 19.0 (5.0–27.0) 22.5 (14.8–39.9) 62.0 (40.0–86.5) 0.005 Involved angle (°) 15.0 (14.0–29.0) 27.0 (19.5–46.3) 50.0 (30.0–98.0) 0.04 Length (mm) 8.0 (6.9–8.0) 9.7 (8.2–11.4) 9.3 (7.2–9.9) 0.20 Data represent the median and interquartile ranges. Table 3 Comparison of the size of paravalvular leakage according to the regurgitant grade using the Kruskal–Wallis test (per-lesion based analysis) Mild Moderate Severe P-value Prosthetic aortic valve Long diameter (mm) 4.4 (3.9–4.9) 7.5 (7.4–12.0) 15.5 (9.8–18.6) 0.02 Short diameter (mm) 2.6 (2.2–2.7) 3.5 (2.9–5.0) 5.6 (5.1–7.1) 0.01 Perimeter (mm) 14.6 (10.0–15.0) 24.0 (22.3–26.0) 39.9 (28.5–42.3) 0.04 Area (mm2) 12.0 (6.0–16.0) 23.8 (15.0–30.9) 86.5 (58.5–134.8) 0.01 Involved angle (°) 27.0 (19.0–32.0) 33.0 (31.7–55.0) 60.0 (36.3–65.8) 0.05 Length (mm) 11.0 (8.3–12.9) 9.0 (8.2–9.1) 7.5 (5.5–12.0) 0.75 Prosthetic mitral valve Long diameter (mm) 8.0 (3.6–9.3) 9.2 (5.6–13.9) 18.0 (12.6–25.5) 0.01 Short diameter (mm) 3.3 (1.6–4.1) 3.6 (2.8–4.3) 5.8 (5.0–7.6) 0.01 Perimeter (mm) 24.0 (8.3–24.3) 25.2 (17.2–29.9) 52.0 (31.8–63.3) 0.01 Area (mm2) 19.0 (5.0–27.0) 22.5 (14.8–39.9) 62.0 (40.0–86.5) 0.005 Involved angle (°) 15.0 (14.0–29.0) 27.0 (19.5–46.3) 50.0 (30.0–98.0) 0.04 Length (mm) 8.0 (6.9–8.0) 9.7 (8.2–11.4) 9.3 (7.2–9.9) 0.20 Mild Moderate Severe P-value Prosthetic aortic valve Long diameter (mm) 4.4 (3.9–4.9) 7.5 (7.4–12.0) 15.5 (9.8–18.6) 0.02 Short diameter (mm) 2.6 (2.2–2.7) 3.5 (2.9–5.0) 5.6 (5.1–7.1) 0.01 Perimeter (mm) 14.6 (10.0–15.0) 24.0 (22.3–26.0) 39.9 (28.5–42.3) 0.04 Area (mm2) 12.0 (6.0–16.0) 23.8 (15.0–30.9) 86.5 (58.5–134.8) 0.01 Involved angle (°) 27.0 (19.0–32.0) 33.0 (31.7–55.0) 60.0 (36.3–65.8) 0.05 Length (mm) 11.0 (8.3–12.9) 9.0 (8.2–9.1) 7.5 (5.5–12.0) 0.75 Prosthetic mitral valve Long diameter (mm) 8.0 (3.6–9.3) 9.2 (5.6–13.9) 18.0 (12.6–25.5) 0.01 Short diameter (mm) 3.3 (1.6–4.1) 3.6 (2.8–4.3) 5.8 (5.0–7.6) 0.01 Perimeter (mm) 24.0 (8.3–24.3) 25.2 (17.2–29.9) 52.0 (31.8–63.3) 0.01 Area (mm2) 19.0 (5.0–27.0) 22.5 (14.8–39.9) 62.0 (40.0–86.5) 0.005 Involved angle (°) 15.0 (14.0–29.0) 27.0 (19.5–46.3) 50.0 (30.0–98.0) 0.04 Length (mm) 8.0 (6.9–8.0) 9.7 (8.2–11.4) 9.3 (7.2–9.9) 0.20 Data represent the median and interquartile ranges. Figure 3 View largeDownload slide (A) A 71-year-old male who underwent trido-mitral valve replacement exhibited two para-mitral leakages (arrows) on transoesophageal echocardiography in both the medial and lateral aspect of the mitral annulus. (B, C) On CT, lesions appeared as one large crescent shaped dehiscence (arrows) that involved posterior aspect of the mitral annulus. (D) Eventually, this lesion was confirmed as a single lesion in surgical inspection. Surgical instruments indicate the medial and lateral ends of the paravalvular dehiscence. Asterisk indicates prosthetic mitral valve; arrowhead indicates left circumflex artery; Ao, aorta; CS, coronary sinus; LA, left atrium; LAA, left atrial appendage; LV, left ventricle; RA, right atrium. Figure 3 View largeDownload slide (A) A 71-year-old male who underwent trido-mitral valve replacement exhibited two para-mitral leakages (arrows) on transoesophageal echocardiography in both the medial and lateral aspect of the mitral annulus. (B, C) On CT, lesions appeared as one large crescent shaped dehiscence (arrows) that involved posterior aspect of the mitral annulus. (D) Eventually, this lesion was confirmed as a single lesion in surgical inspection. Surgical instruments indicate the medial and lateral ends of the paravalvular dehiscence. Asterisk indicates prosthetic mitral valve; arrowhead indicates left circumflex artery; Ao, aorta; CS, coronary sinus; LA, left atrium; LAA, left atrial appendage; LV, left ventricle; RA, right atrium. Figure 4 View largeDownload slide The areas and involved angle of paravalvular leakage measured on CT and the degree of regurgitant grades on echocardiography in prosthetic (A, B) aortic and (C, D) mitral valves. PAV, prosthetic aortic valve; PMV, prosthetic mitral valve. Figure 4 View largeDownload slide The areas and involved angle of paravalvular leakage measured on CT and the degree of regurgitant grades on echocardiography in prosthetic (A, B) aortic and (C, D) mitral valves. PAV, prosthetic aortic valve; PMV, prosthetic mitral valve. Size of PVL and treatment methods The size of PVLs were compared between systolic and diastolic phases in a subset of patients (see Supplementary data online, Table S1). Although there was no statistical significant difference, aortic PVLs are slightly larger during the systolic phase (20–40% R–R) and mitral PVLs are slightly larger during the diastolic phase (65–85% R–R). Representative cases are demonstrated in Supplementary data online, Figures S3 and S4. More than half of aortic and mitral PVLs were treated by surgical correction and six sites (five patients including one who had both aortic and mitral PVLs) of PVLs were managed by percutaneous device closure. The numbers and locations of PVLs detected on surgical filed were the same with those demonstrated on CT. The sizes of aortic PVLs are larger in patients who underwent surgical correction compared to those in patients with device closure or who observed without treatment (Table 4). The areas of mitral PVLs in patients with surgery were larger than those in patients without surgery, but marginally significant (P = 0.05). In both aortic and mitral PVLs, surgical correction is preferred to patients who presented severe regurgitant grade on echocardiography (P < 0.05). Among the five patients who were managed by percutaneous interventions, three patients (including the patient who had both aortic and mitral PVLs) were successfully treated, and two patients with small mitral PVLs exhibiting an elongated shape failed because of difficulty in guiding the passage of the catheter through the narrow and long dehiscence tract (Figure 5 and Supplementary data online, Video S1). The remaining 13 PVLs were observed without treatment because of the absence of symptoms (n = 11) or reluctance to undergo invasive treatment (n = 2). Postoperative mortality (range 1–56 days) occurred in three patients due to pneumonia (n = 1), multiorgan failure with a failed prosthetic valve (n = 1), or septic shock secondary to underlying infective endocarditis (n = 1). Table 4 Comparison of size of paravalvular leakage on CT and regurgitant grade on echocardiography according to the treatment methods (per-lesion based analysis) Prosthetic aortic valve Prosthetic mitral valve Surgery (n = 10) Device closure or observation (n = 10) P-value Surgery (n = 17) Device closure or observation (n = 9) P-value Long diameter (mm) 15.2 (8.8–17.6) 7.1 (4.5–10.9) 0.06 13.0 (11.2–18.0) 8.0 (6.5–12.4) 0.56 Short diameter (mm) 5.4 (5.1–6.3) 2.8 (2.3–4.6) 0.01 5.2 (3.6–6.0) 3.6 (3.2–4.4) 0.10 Perimeter (mm) 39.3 (25.3–41.7) 16.2 (13.5–23.9) 0.01 31.0 (23.9–52.0) 24.3 (17.2–31.7) 0.30 Area (mm2) 64.5 (52.8–119.8) 15.5 (10.5–29.1) 0.05 40.0 (22.0–67.7) 27.0 (17.0–40.0) 0.05 Involved angle (°) 58.7 (29.8–64.8) 32.4 (27.8–49.5) 0.11 44.0 (26.0–79.0) 26.0 (15.0–52.0) 0.54 Regurgitant grade 0.001a 0.04a  Mild 1 4 1 4  Moderate 0 5 6 4  Severe 9 1 10 1 Concomitant IE 2 3 2 0 Pseudoaneurysm 2 1 0 0 Severe annular calcification 0 0 1 1 Prosthetic aortic valve Prosthetic mitral valve Surgery (n = 10) Device closure or observation (n = 10) P-value Surgery (n = 17) Device closure or observation (n = 9) P-value Long diameter (mm) 15.2 (8.8–17.6) 7.1 (4.5–10.9) 0.06 13.0 (11.2–18.0) 8.0 (6.5–12.4) 0.56 Short diameter (mm) 5.4 (5.1–6.3) 2.8 (2.3–4.6) 0.01 5.2 (3.6–6.0) 3.6 (3.2–4.4) 0.10 Perimeter (mm) 39.3 (25.3–41.7) 16.2 (13.5–23.9) 0.01 31.0 (23.9–52.0) 24.3 (17.2–31.7) 0.30 Area (mm2) 64.5 (52.8–119.8) 15.5 (10.5–29.1) 0.05 40.0 (22.0–67.7) 27.0 (17.0–40.0) 0.05 Involved angle (°) 58.7 (29.8–64.8) 32.4 (27.8–49.5) 0.11 44.0 (26.0–79.0) 26.0 (15.0–52.0) 0.54 Regurgitant grade 0.001a 0.04a  Mild 1 4 1 4  Moderate 0 5 6 4  Severe 9 1 10 1 Concomitant IE 2 3 2 0 Pseudoaneurysm 2 1 0 0 Severe annular calcification 0 0 1 1 a The P-values are obtained from the comparison between mild or moderate and severe groups using the Fisher’s exact test. IE, infective endocarditis. Table 4 Comparison of size of paravalvular leakage on CT and regurgitant grade on echocardiography according to the treatment methods (per-lesion based analysis) Prosthetic aortic valve Prosthetic mitral valve Surgery (n = 10) Device closure or observation (n = 10) P-value Surgery (n = 17) Device closure or observation (n = 9) P-value Long diameter (mm) 15.2 (8.8–17.6) 7.1 (4.5–10.9) 0.06 13.0 (11.2–18.0) 8.0 (6.5–12.4) 0.56 Short diameter (mm) 5.4 (5.1–6.3) 2.8 (2.3–4.6) 0.01 5.2 (3.6–6.0) 3.6 (3.2–4.4) 0.10 Perimeter (mm) 39.3 (25.3–41.7) 16.2 (13.5–23.9) 0.01 31.0 (23.9–52.0) 24.3 (17.2–31.7) 0.30 Area (mm2) 64.5 (52.8–119.8) 15.5 (10.5–29.1) 0.05 40.0 (22.0–67.7) 27.0 (17.0–40.0) 0.05 Involved angle (°) 58.7 (29.8–64.8) 32.4 (27.8–49.5) 0.11 44.0 (26.0–79.0) 26.0 (15.0–52.0) 0.54 Regurgitant grade 0.001a 0.04a  Mild 1 4 1 4  Moderate 0 5 6 4  Severe 9 1 10 1 Concomitant IE 2 3 2 0 Pseudoaneurysm 2 1 0 0 Severe annular calcification 0 0 1 1 Prosthetic aortic valve Prosthetic mitral valve Surgery (n = 10) Device closure or observation (n = 10) P-value Surgery (n = 17) Device closure or observation (n = 9) P-value Long diameter (mm) 15.2 (8.8–17.6) 7.1 (4.5–10.9) 0.06 13.0 (11.2–18.0) 8.0 (6.5–12.4) 0.56 Short diameter (mm) 5.4 (5.1–6.3) 2.8 (2.3–4.6) 0.01 5.2 (3.6–6.0) 3.6 (3.2–4.4) 0.10 Perimeter (mm) 39.3 (25.3–41.7) 16.2 (13.5–23.9) 0.01 31.0 (23.9–52.0) 24.3 (17.2–31.7) 0.30 Area (mm2) 64.5 (52.8–119.8) 15.5 (10.5–29.1) 0.05 40.0 (22.0–67.7) 27.0 (17.0–40.0) 0.05 Involved angle (°) 58.7 (29.8–64.8) 32.4 (27.8–49.5) 0.11 44.0 (26.0–79.0) 26.0 (15.0–52.0) 0.54 Regurgitant grade 0.001a 0.04a  Mild 1 4 1 4  Moderate 0 5 6 4  Severe 9 1 10 1 Concomitant IE 2 3 2 0 Pseudoaneurysm 2 1 0 0 Severe annular calcification 0 0 1 1 a The P-values are obtained from the comparison between mild or moderate and severe groups using the Fisher’s exact test. IE, infective endocarditis. Figure 5 View largeDownload slide An example of failed device closure in mitral paravalvular leakage because of a long and narrow dehiscence tract. Paravalvular leakage with a thin and long tract (arrows) is located at the posterolateral aspect of the mitral prosthetic valve. LA, left atrium; LV, left ventricle. Figure 5 View largeDownload slide An example of failed device closure in mitral paravalvular leakage because of a long and narrow dehiscence tract. Paravalvular leakage with a thin and long tract (arrows) is located at the posterolateral aspect of the mitral prosthetic valve. LA, left atrium; LV, left ventricle. Discussion The major findings of this study were as follows: (i) cardiac CT could demonstrate the size and location of PVL, and the location of PVLs were matched with those noted on echocardiography and surgical inspection; (ii) the sizes of PVLs were significantly larger in patients with higher regurgitant grade; and (iii) surgical correction is preferred to patients who presented large PVLs on CT and severe regurgitant grade on echocardiography. Making the decision to perform a reoperation in patients with PVL is difficult because of the high surgical morbidity and recurrence rate (12–35%) of PVLs after reoperation.8 Recently, interventional occlusion for PVL has emerged as an alternative treatment strategy.10,21–24 For decision making in PVL treatment, anatomical information about the PVL, including the size, shape, and the three-dimensional relationship between the PVL and cardiac structure, are important. Echocardiography is the initial modality of choice that has excellent temporal resolution and real-time imaging capabilities. Notably, for the detection of PVLs, colour Doppler imaging can easily indicate the location and degree of paravalvular regurgitant jet. However, anatomical demonstration of a PVL using echocardiography often is compromised by metallic artefacts from the prosthetic valve and its limited sonic window. By providing more precise anatomical details, including the exact location and morphology of the PVLs on CT, pre-treatment planning could be better tailored and individualized.25 By simulating the entrance site of the guidewire through the PVL on a fluoroscopic view of CT to guide the percutaneous procedure, both the fluoroscopic time and radiation exposure could be reduced. For PVLs located in the medial aspect of the prosthetic mitral valve, it is challenging to perform a trans-septal access because of the difficulty of manipulating the acute angle with a guidewire.14 Moreover, the shape and length of the dehiscence tract are also determining factors. If the dehiscence tract is long and narrow, it can be difficult to negotiate the guidewire. If the lesion is crescent shaped or large that it cannot be occluded using occlusion devices, surgical management should be considered.14 These conditions can be clearly demonstrated on cardiac CT using reconstructed images.26,27 Detailed anatomic information with a familiar view for either surgeon or interventionist may be helpful for planning of treatment strategy. CT can also be used to evaluate PVL after transcatheter aortic valve implantation (TAVI). Conventional parasternal short-axis analysis using echocardiography can underestimate the extent of PVL with false negative studies up to 14%.28 Moreover, in contrast to angiography, echocardiography did not show good correlation with magnetic resonance imaging in the assessment of PVL after TAVI.29 Similar to the methods used to evaluate PVLs occurred after surgical valve replacement in this study, CT 3-dimensional multiplanar reconstruction images can demonstrate sagittal view and short-axis cuts (en face views) of the prosthetic frame geometry of TAVI device.30,31 To avoid overestimation of the dehiscence, further image reconstruction can be carried out by either rotating the imaging plane to centre the dehiscence or using a 360° rotational technique following the ‘double-checking’ method. The raysum technique can also be used to generate fluoroscopy-like images. Previously, aortic PVLs have been reported to be more commonly located in the right- and non-coronary cusps.32 However, in our present study, right- and left-coronary cusps are common sites of PVLs. Mitral PVLs were found to be evenly distributed around the sewing ring, which was in accord with the findings of a previous study.33 Considering the small sample size of our study, further studies with a larger population will be necessary. Limitations Our study had several limitations. First, there is a selection bias, only 46 patients with CT out of 131 with PVL detected on echocardiography were studied. The patients with CT performed were more symptomatic, had bigger ventricles and lower ejection fraction. Patients with severe regurgitant flow with large sizes of leakages might be included, and the results of this study may not be applicable for small lesions. However, the smallest sizes of aortic and mitral PVLs in our study were 4.4 and 1.5 mm, respectively. These values suggest the potential use of CT for evaluation of PVLs even for small lesions. Also, it may not be easy to distinguish between small leaks and pledgets or stitches because the density is similar during the angiographic phase. A way to avoid mistakes is to perform a pre-contrast gated scan. Second, observer bias should be considered because the readers had known about the presence of the PVL when they interpreted the CT images. Because patients in this study were collected at a tertiary referral centre and cardiac CT imaging was used as a preoperative planning method, thorough evaluations were performed for optimal patient care. To overcome this bias, all cardiac source images were interpreted by two readers who were blind to the size, location, shape, and number of PVLs and interobserver agreement was good. Third, CT interpretation may be highly dependent upon the data processing quality.34 In our study, we exclude two patients with suboptimal CT image quality because of motion artefacts. To measure the largest size of PVL during obtained cardiac cycles, we used systolic phase for aortic PVLs, and diastolic phase for mitral PVLs. However, because of cardiac motion artefacts and beam hardening artefacts from prosthetic valve, not all phases could be used for evaluation of PVLs, thus we could performed the comparison between systolic/diastolic phases in a subset of patients. Further studies to evaluate the changes of PVLs during different cardiac cycle phases would be of value. Nevertheless, our current findings may demonstrate a practical role for cardiac CT in the planning of PVL treatment by providing detailed anatomic data. Finally, because not all patients underwent operations and the limits of surgical records, we could not compare the morphological characteristics of PVLs detected on CT to surgical anatomy. However, we have shown that the morphology data on CT, long/short diameter, angle and length of PVLs, could be used to demonstrate PVLs and the information may aid in the planning of a treatment strategy. Conclusion In conclusion, cardiac CT can be used to detect and localize aortic and mitral PVLs. The visualization of anatomic information using a familiar view for either a surgeon or an interventionist may aid in the planning of a treatment strategy. Supplementary data Supplementary data are available at European Heart Journal - Cardiovascular Imaging online. Funding The work was supported by the Industrial Strategic technology development program (10072064) funded by the Ministry of Trade Industry and Energy (MI, Korea), the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (NRF-2016R1A1A1A05921207) and a grant (2017-7208) from the Asan Institute for Life Sciences, Asan Medical Centre, Seoul, Korea. Conflict of interest: none declared. References 1 Ionescu A , Fraser A , Butchart E. Prevalence and clinical significance of incidental paraprosthetic valvar regurgitation: a prospective study using transoesophageal echocardiography . Heart 2003 ; 89 : 1316 – 21 . http://dx.doi.org/10.1136/heart.89.11.1316 Google Scholar CrossRef Search ADS PubMed 2 Kliger C , Eiros R , Isasti G , Einhorn B , Jelnin V , Cohen H et al. Review of surgical prosthetic paravalvular leaks: diagnosis and catheter-based closure . Eur Heart J 2013 ; 34 : 638 – 49 . Google Scholar CrossRef Search ADS PubMed 3 Rallidis LS , Moyssakis IE , Ikonomidis I , Nihoyannopoulos P. Natural history of early aortic paraprosthetic regurgitation: a five-year follow-up . Am Heart J 1999 ; 138 : 351 – 7 . Google Scholar CrossRef Search ADS PubMed 4 Hammermeister K , Sethi GK , Henderson WG , Grover FL , Oprian C , Rahimtoola SH. Outcomes 15 years after valve replacement with a mechanical versus a bioprosthetic valve: final report of the Veterans Affairs randomized trial . J Am Coll Cardiol 2000 ; 36 : 1152 – 8 . Google Scholar CrossRef Search ADS PubMed 5 Ruiz CE , Jelnin V , Kronzon I , Dudiy Y , Del Valle-Fernandez R , Einhorn BN et al. Clinical outcomes in patients undergoing percutaneous closure of periprosthetic paravalvular leaks . J Am Coll Cardiol 2011 ; 58 : 2210 – 7 . http://dx.doi.org/10.1016/j.jacc.2011.03.074 Google Scholar CrossRef Search ADS PubMed 6 Bhindi R , Bull S , Schrale RG , Wilson N , Ormerod OJ. Surgery insight: percutaneous treatment of prosthetic paravalvular leaks . Nat Clin Pract Cardiovasc Med 2008 ; 5 : 140 – 7 . Google Scholar CrossRef Search ADS PubMed 7 Lesser JR , Han BK , Newell M , Schwartz RS , Pedersen W , Sorajja P. Use of cardiac CT angiography to assist in the diagnosis and treatment of aortic prosthetic paravalvular leak: a practical guide . J Cardiovasc Comput Tomogr 2015 ; 9 : 159 – 64 . Google Scholar CrossRef Search ADS PubMed 8 Echevarria JR , Bernal JM , Rabasa JM , Morales D , Revilla Y , Revuelta JM. Reoperation for bioprosthetic valve dysfunction: a decade of clinical experience . Eur J Cardiothorac Surg 1991 ; 5 : 523 – 6 . Google Scholar CrossRef Search ADS PubMed 9 Safi AM , Kwan T , Afflu E , Al Kamme A , Salciccioli L. Paravalvular regurgitation: a rare complication following valve replacement surgery . Angiology 2000 ; 51 : 479 – 87 . http://dx.doi.org/10.1177/000331970005100605 Google Scholar CrossRef Search ADS PubMed 10 Hein R. Percutaneous closure of paravalvular leaks . J Interv Cardiol 2006 ; 19 : S73 – 7 . Google Scholar CrossRef Search ADS 11 Pate GE , Al Zubaidi A , Chandavimol M , Thompson CR , Munt BI , Webb JG. Percutaneous closure of prosthetic paravalvular leaks: case series and review . Catheter Cardiovasc Interv 2006 ; 68 : 528 – 33 . Google Scholar CrossRef Search ADS PubMed 12 Hildick-Smith D , Behan MWH , De Giovanni J. Percutaneous closure of an aortic paravalvular leak via the transradial approach . Catheter Cardiovasc Interv 2007 ; 69 : 708 – 10 . http://dx.doi.org/10.1002/ccd.21043 Google Scholar CrossRef Search ADS PubMed 13 Momplaisir T , Matthews RV. Paravalvular mitral regurgitation treated with an amplatzer septal occluder device: a case report and review of the literature . J Invasive Cardiol 2007 ; 19 : E46 – 50 . Google Scholar PubMed 14 Ruiz CE , Cohen H , Del Valle-Fernandez R , Jelnin V , Perk G , Kronzon I. Closure of prosthetic paravalvular leaks: a long way to go . Eur Heart J Suppl 2010 ; 12 : E52 – 62 . Google Scholar CrossRef Search ADS 15 Koo HJ , Yang DH , Kang J-W , Han K , Chung CH , Song J-K et al. Demonstration of prosthetic aortic valve dehiscence in a patient with noninfectious aortitis by multimodality imaging: findings of echocardiography and computed tomography . Circulation 2013 ; 128 : 759 – 61 . Google Scholar CrossRef Search ADS PubMed 16 McKee PA , Castelli WP , McNamara PM , Kannel WB. The natural history of congestive heart failure: the Framingham study . N Engl J Med 1971 ; 285 : 1441 – 6 . http://dx.doi.org/10.1056/NEJM197112232852601 Google Scholar CrossRef Search ADS PubMed 17 Kim YJ , Yong HS , Kim SM , Kim JA , Yang DH , Hong YJ. Korean guidelines for the appropriate use of cardiac CT . Korean J Radiol 2015 ; 16 : 251 – 85 . http://dx.doi.org/10.3348/kjr.2015.16.2.251 Google Scholar CrossRef Search ADS PubMed 18 Tsai IC , Choi BW , Chan C , Jinzaki M , Kitagawa K , Yong HS. ASCI 2010 appropriateness criteria for cardiac computed tomography: a report of the Asian Society of Cardiovascular Imaging Cardiac Computed Tomography and Cardiac Magnetic Resonance Imaging Guideline Working Group . Int J Cardiovasc Imaging 2010 ; 26(Suppl. 1) : 1 – 15 . Google Scholar CrossRef Search ADS PubMed 19 Nishimura RA , Otto CM , Bonow RO , Carabello BA , Erwin JP 3rd , Fleisher LA. 2017 AHA/ACC focused update of the 2014 AHA/ACC guideline for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines . Circulation 2017 ; 135 : e1159 – 95 . Google Scholar CrossRef Search ADS PubMed 20 De Cicco G , Russo C , Moreo A , Beghi C , Fucci C , Gerometta P et al. Mitral valve periprosthetic leakage: anatomical observations in 135 patients from a multicentre study . Eur J Cardiothorac Surg 2006 ; 30 : 887 – 91 . Google Scholar CrossRef Search ADS PubMed 21 Boudjemline Y. Percutaneous closure of a paravalvular mitral regurgitation with Amplatzer and coil prostheses . Arch Mal Coeur Vaiss 2002 ; 95 : 483 – 6 . Google Scholar PubMed 22 Piechaud JF. Percutaneous closure of mitral paravalvular leak . J Interv Cardiol 2003 ; 16 : 153 – 5 . http://dx.doi.org/10.1046/j.1540-8183.2003.08028.x Google Scholar CrossRef Search ADS PubMed 23 Kort HW , Sharkey AM , Balzer DT. Novel use of the Amplatzer duct occluder to close perivalvar leak involving a prosthetic mitral valve . Catheter Cardiovasc Interv 2004 ; 61 : 548 – 51 . http://dx.doi.org/10.1002/ccd.10785 Google Scholar CrossRef Search ADS PubMed 24 Webb JG , Pate GE , Munt BI. Percutaneous closure of an aortic prosthetic paravalvular leak with an Amplatzer duct occluder . Catheter Cardiovasc Interv 2005 ; 65 : 69 – 72 . http://dx.doi.org/10.1002/ccd.20337 Google Scholar CrossRef Search ADS PubMed 25 Pate GE , Thompson CR , Munt BI , Webb JG. Techniques for percutaneous closure of prosthetic paravalvular leaks . Catheter Cardiovasc Interv 2006 ; 67 : 158 – 66 . http://dx.doi.org/10.1002/ccd.20560 Google Scholar CrossRef Search ADS PubMed 26 Odell JA , Orszulak TA. Surgical repair and reconstruction of valvular lesions . Curr Opin Cardiol 1995 ; 10 : 135 – 43 . http://dx.doi.org/10.1097/00001573-199503000-00007 Google Scholar CrossRef Search ADS PubMed 27 Dion R. Ischemic mitral regurgitation: when and how should it be corrected? J Heart Valve Dis 1993 ; 2 : 536 – 43 . Google Scholar PubMed 28 Geleijnse ML , Di Martino LFM , Vletter WB , Ren B , Galema TW , Van Mieghem NM et al. Limitations and difficulties of echocardiographic short-axis assessment of paravalvular leakage after corevalve transcatheter aortic valve implantation . Cardiovasc ultrasound 2015 ; 14 : 37. Google Scholar CrossRef Search ADS 29 Sherif M , Abdel-Wahab M , Beurich H-W , Stöcker B , Zachow D , Geist V et al. Haemodynamic evaluation of aortic regurgitation after transcatheter aortic valve implantation using cardiovascular magnetic resonance . EuroIntervention 2011 ; 7 : 57 – 63 . Google Scholar CrossRef Search ADS PubMed 30 Abdelghani M , Soliman OI , Schultz C , Vahanian A , Serruys PW. Adjudicating paravalvular leaks of transcatheter aortic valves: a critical appraisal . Eur Heart J 2016 ; 37 : 2627 – 44 . Google Scholar CrossRef Search ADS PubMed 31 Schultz CJ , Lauritsch G , Van Mieghem N , Rohkohl C , Serruys PW , van Geuns RJ et al. Rotational angiography with motion compensation: first-in-man use for the 3D evaluation of transcatheter valve prostheses . EuroIntervention 2015 ; 11 : 442 – 9 . Google Scholar CrossRef Search ADS PubMed 32 De Cicco G , Lorusso R , Colli A , Nicolini F , Fragnito C , Grimaldi T et al. Aortic valve periprosthetic leakage: anatomic observations and surgical results . Ann Thorac Surg 2005 ; 79 : 1480. Google Scholar CrossRef Search ADS PubMed 33 Foster GP , Isselbacher EM , Rose GA , Torchiana DF , Akins CW , Picard MH. Accurate localization of mitral regurgitant defects using multiplane transesophageal echocardiography . Ann Thorac Surg 1998 ; 65 : 1025 – 31 . Google Scholar CrossRef Search ADS PubMed Published on behalf of the European Society of Cardiology. All rights reserved. © The Author(s) 2018. For permissions, please email: journals.permissions@oup.com.

Journal

European Heart Journal – Cardiovascular ImagingOxford University Press

Published: Jan 4, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off