Abstract
OBJECTIVES
We compared the effectiveness of virtual 3-dimensional (3D) models with 2-dimensional (2D) transthoracic echocardiography (TTE) for evaluating the anatomy of the interventricular septum (IVS) and abnormal muscle bundles (AMBs) in planning septal myectomy (SM).
METHODS
Between January 2017 and July 2020, 103 consecutive symptomatic patients with hypertrophic cardiomyopathy underwent 2D TTE and cardiovascular magnetic resonance imaging in 49 (47.6%) or computed tomography angiography in 54 (52.4%) patients with 3D IVS modelling for SM planning. We evaluated maximal IVS thickness and location, length and thickness of AMBs.
RESULTS
The mean maximal IVS thickness by 2D TTE was 7.3 [standard deviation (SD) 4.8] mm less than that based on the 3D model analysis: 21.4 (SD 3.7) vs 28.6 (SD 5.5) mm, respectively (P < 0.001, 95% confidence interval 6.4–8.2). The planned volume of ideal SM was larger than that of performed SM: 26.2 (18.4–39.4) vs 10.3 (7.4–12.8) cm3, respectively (P < 0.001). The sensitivity and specificity of 2D TTE in diagnosing AMBs were 36.9% and 95%, and those of cardiovascular magnetic resonance and computed tomography angiography with 3D modelling were 97.1% and 100% for cardiovascular magnetic resonance and 98% and 100% for computed tomography angiography, respectively. AMBs occurred in 84 (81.6%) patients. No patient required mitral valve replacement. The 30-day mortality was 1 patient. There were 4 late non-cardiac deaths (3.9%) within 18.1 (standard error 1.32) months.
CONCLUSIONS
Anatomical analysis of the IVS and AMBs based on their virtual 3D models is highly effective for SM planning.
Keywords: Hypertrophic obstructive cardiomyopathy, Septal myectomy, 3-Dimensional model, Interventricular septum
Septal myectomy (SM) is the key to managing hypertrophic obstructive cardiomyopathy (HOCM) [1, 2].
INTRODUCTION
Septal myectomy (SM) is the key to managing hypertrophic obstructive cardiomyopathy (HOCM) [1, 2]. A precise understanding of the anatomy of the interventricular septum (IVS) is the basis for SM planning. Abnormal muscle bundles (AMBs) are additional morphological markers [3] for hypertrophic cardiomyopathy. AMBs are thought to contribute to the mechanism of left ventricular outflow tract obstruction, but their role has not been confirmed [3]. The mechanisms responsible for latent obstruction are also unclear [4]. Cardiovascular magnetic resonance (CMR) imaging and computed tomography angiography (CTA) can identify segmental left ventricular hypertrophy that cannot be visualized reliably by 2-dimensional (2D) transthoracic (TTE) or transoesophageal echocardiography [5–7]. Computer and 3-dimensonal (3D) heart models based on CMR and CTA are increasingly being used for SM planning [2, 8–11]. However, no algorithm has been described for the analysis of the anatomy of IVS and AMBs using 3D heart models and for extended SM planning.
MATERIALS AND METHODS
Ethical statement
The study was approved by the ethics committee of our centre (No 12/19 December 2016), and all participants signed a written informed consent form.
Patients, examination, surgical planning and surgery
One hundred and three consecutive symptomatic patients admitted to our centre between January 2017 and July 2020 were included in this study. The inclusion criterion was a peak pressure gradient in the area of left ventricle (LV) obstruction ≥30 mmHg at rest (for HOCM) or with exercise (latent obstruction). The exclusion criterion was contraindications for surgery. Table 1 summarizes the patients’ characteristics.
Table 1:
Baseline clinical characteristics before surgery
| Variables | Value (n = 103) |
|---|---|
| Male gender, n (%) | 56 (54.4) |
| Age (years), mean (SD) | 53.2 (11) |
| Body surface area (m2), mean (SD) | 1.98 (0.17) |
| NYHA functional class II, n (%) | 58 (56.3) |
| NYHA functional class III, n (%) | 40 (38.8) |
| NYHA functional class IV, n (%) | 1 (0.97) |
| NYHA functional class, mean (SD) | 2.4 (0.57) |
| Concomitant coronary disease, n (%) | 16 (15.5) |
| Concomitant organic aortic valve disease, n (%) | 20 (19.4) |
| Concomitant organic mitral valve disease, n (%) | 25 (24.3) |
| Left ventricular aneurysm, n (%) | 3 (2.9) |
| EuroSCORE 2, median (25–75th percentile) | 1.2 (0.91–2.1) |
| Type of HCM, n (%) | |
| Basal | 20 (19.4) |
| Basal–midventricular | 62 (60.2) |
| Midventricular | 4 (3.9) |
| Diffuse | 17 (16.5) |
| Latent obstruction | 7 (6.8) |
EuroSCORE 2: European System for Cardiac Operative Risk Evaluation; HCM: hypertrophic cardiomyopathy; NYHA: New York Heart Association; SD: standard deviation.
Preoperative patient examinations included TTE, CTA (Siemens SOMATOM Force, Siemens Healthcare AG, Erlangen, Germany) or CMR (Siemens MAGNETOM Aera 1.5 T) (see Supplementary Material). Intraoperatively, transoesophageal echocardiography was performed after cardiopulmonary bypass. Table 2 shows the 2D TTE data before SM.
Table 2:
Transthoracic echocardiography data before SM and at the latest follow-up
| Variables | Value, n (%), mean (SD) or median (25–75th percentile) |
P-value | |
|---|---|---|---|
| Pre-op (n = 103) | Follow-up (n = 102) | ||
| Peak systolic gradient in obstruction area at rest (mmHg) | 70 (50–89) | 6 (5–8) | <0.001 |
| Maximal IVS thickness (mm) | 21 (19–23) | 13 (11–15) | <0.001 |
| LVEDD (mm) | 48 (44–50) | 53 (50.3–56) | <0.001 |
| LVESD (mm) | 27 (25–32) | 35 (33–39) | <0.001 |
| LVEDV (ml) (SD) | 112.1 (30.8) | 120 (30.2) | 0.036 |
| LVESV (ml) | 39 (29–48) | 46 (33–58) | <0.001 |
| LVEF (Simpson, %) (SD) | 64.4 (6.6) | 61.1 (8.5) | 0.002 |
| SAM | <0.001 | ||
| Absence | 3 (2.9%) | 100 (98%) | |
| Grade 1/3 | 5 (4.9%) | 2 (1.96%) | |
| Grade 2/3–3/3 | 95 (92.2%) | 0 | |
| Mitral regurgitation: mean | 3 (2.5–3) | 1.5 (1–2) | <0.001 |
| Mild (1/4) | 1 (0.97%) | 32 (31.4%) | |
| Moderate (2/4) | 20 (19.4%) | 62 (60.8%) | |
| Moderately severe (3/4) | 60 (58.3%) | 7 (6.9%) | |
| Severe (4/4) | 22 (21.4%) | 0 | |
IVS: interventricular septum; LVEDD: left ventricular end-diastolic diameter; LVEDV: left ventricular end-diastolic volume; LVEF: left ventricular ejection fraction; LVESD: left ventricular end-systolic diameter; LVESV: left ventricular end-systolic volume; LVOT: left ventricular outflow tract; pre-op: preoperative; SAM: systolic anterior motion; SD: standard deviation; SM: septal myectomy.
Using the OsiriX open source DICOM viewer (https://www.osirix-viewer.com) or InVesalius (https://www.cti.gov.br/pt-br/invesalius), we created virtual 3D models of the IVS and LV AMBs with CMR data for 49 (47.6%) patients and CTA data for 54 (52.4%). Computer-aided mapping of the IVS was performed by cross-sectioning the 3D model in different planes and measuring the thickness and length of the IVS and AMBs. We distinguished 3 varieties of AMBs: anterolateral (ALAMB), i.e. from the basal segments to the base of anterolateral papillary muscle; posteromedial (PMAMB), i.e. to the base of posteromedial papillary muscle; and medial AMB, i.e. between the ALAMB and PMAMB (Fig. 1 and Video 1).
Figure 1:
Types of abnormal muscle bundles on 3-dimensional model (computed tomography angiography based) of the interventricular septum. ALAMB: anterolateral abnormal muscle bundle; MAMB: medial abnormal muscle bundle; PMAMB: posteromedial abnormal muscle bundle.
The following measurements of the virtual 3D IVS model were taken along the long axis of the perpendicular section through the LV apex and right coronary cusp nadir: thickening length (defines the SM length), depth of thickening from the right coronary cusp nadir (defines the SM depth towards the LV apex) and IVS axis length from right coronary cusp nadir to the apex (Fig. 2). Maximal IVS thickness (MTh) was measured in frontal sections and compared to 2D models. Figure 3 demonstrates the anatomical types of HOCM (basal, basal–midventricular, midventricular and diffuse).
Figure 2:
Measurements of virtual 3-dimensional models (computed tomography angiography based) of the interventricular septum for septal myectomy planning. Longitudinal section through the apex of the left ventricle and right coronary cusp nadir. LAp: interventricular septum axis length; LD: depth of thickening; LTh: length of thickening; LVOT: left ventricular outflow tract; MTh: maximal interventricular septum thickness.
Figure 3:
Anatomical types of hypertrophic obstructive cardiomyopathy (longitudinal section through the apex of the left ventricle and right coronary cusp nadir): (A) basal, (B) midventricular, (C) basal–midventricular and (D) diffuse.
This analysis was followed by performance of a virtual SM (Video 2) to produce an ideal IVS (thickness of 10–11 mm), 3D printing of the IVS model and its fragment in polylactide using a low-cost 3D printer (FlashForge Guider 3D printer; FlashForge Corp., Zhejiang, China). A 3D model build time is ∼2 h.
SM was performed through the ascending aorta (n = 102) and apical (n = 1) approaches as described previously [2, 8] (Video 2). Briefly, we implanted the needles in the 3D model of the ideal IVS to the height of the fragment to accurately mark the depth of myocardial resection. We took a sterile 3D model to the operating room. The excised tissue was sequentially laid out on the bottom of the 3D model until it reached the tips of the needles and weighed [2]. The AMB length and thickness were measured (see Supplementary Material). Given that myocardial density is 1 g/cm3, the excised myocardial mass (g) was considered equal to its volume (cm3).
Complications, 30-day mortality and follow-up
Mortality and major hospital complications were assessed using the Clavien–Dindo classification (grades IIIb–V) within 30 days after surgery. Follow-up examinations included symptom assessment, electrocardiography and TTE. These tests were performed at discharge, 3 months after and every 12 months after the intervention.
Statistical analysis
Continuous variables with a normal distribution assessed by the Kolmogorov–Smirnov test were reported as mean and standard deviation (SD). These values were compared using Student’s t-test. Categorical variables were presented as number and percentage and compared using two-sided Pearson’s χ2 test and Fisher’s exact test. Continuous variables with an asymmetric distribution were presented as median and 25th and 75th percentiles and compared with the Mann–Whitney U-test. Paired t-test, signed-rank test and McNemar’s test were used for the paired comparison between 2D and 3D models and for preoperative versus postoperative comparisons among patients. The relationship between 2 quantitative variables was examined through univariable regression analysis. Follow-up and survival (mean and standard error) analysis was done using Kaplan–Meier method. A statistical hypothesis of equal distributions was rejected at P-value ≤0.05. We used IBM SPSS Statistics software (v. 19.0; IBM Corp., Armonk, NY, USA) for all statistical analyses.
RESULTS
Basal–midventricular type found in 62 patients (60.2%) was predominant (Table 1). Equally common basal and diffuse types were found in 20 patients, (19.4%) and 17 (16.5%), respectively. Midventricular type occurred more rarely: 4 patients (3.9%). Among basal–midventricular types, the equally distributed type was more frequent than basal or midventricular dominant types: 36 (58.1%) vs 23 (37.1%) and 3 (4.8%). Extensive types (basal–midventricular and diffuse) occurred far more frequently than localized types (basal and midventricular): 79 (76.7%) vs 24 (23.3%). HOCM with obstruction at rest was much more common than latent obstruction: 96 (93.2%) vs 7 (6.8%).
The mean MTh by 2D TTE data was 7.3 (SD 4.8) mm less than that based on the 3D virtual model analysis (Tables 2 and 3): 21.4 (SD 3.7) vs 28.6 (SD 5.5) mm [P < 0.001, 95% confidence interval (CI) 6.4–8.2]. Table 3 summarizes the main anatomical characteristics of the IVS based on the 3D virtual model analysis and implementation of the excision plan.
Table 3:
Main anatomical characteristics of the IVS based on the 3D virtual model analysis and implementation of the excision plan
| Variables | MTh (mm) (SD) | LTh (mm) (SD) | LD (mm) (SD) | LAp (mm) (SD) | Peak gradient (mmHg) (SD) | Planned ideal excision volume (cm3) (SD) | Real excised mass (g) (SD) | %Real versus planned excision (SD) |
|---|---|---|---|---|---|---|---|---|
| Basal (n = 20) | 26.3 (3.6) | 36 (7.1) | 44.6 (7.6) | 88 (13.2) | 75.3 (25.2) | 23.4 (11) | 8.1 (3.1) | 39.4 (18.1) |
| Basal-midventricular (n = 62) | 28.2 (5.2) | 50.1 (8.9) | 57.5 (8.8) | 88.1 (10.7) | 69.2 (31.7) | 29.8 (14.5) | 11.1 (4.6) | 39.4 (11.2) |
| Midventricular (n = 4) | 31.7 (2.9) | 41.4 (6.6) | 67.6 (10.5) | 99.2 (12.3) | 53.3 (17.7) | 39.4 (1.4) | 9.9 (2.7) | 25.1 (6.5) |
| Diffuse (n = 17) | 32.5 (6.8) | 71.2 (9.6) | 78.6 (9.8) | 90.2 (12.9) | 73.3 (20.4) | 45.7 (31.7) | 13.2 (6) | 35.9 (18.8) |
| Obstructive at rest (n = 96) | 28.8 (5.4) | 50.7 (13.9) | 59 (13.8) | 89 (11.6) | 74.1 (26) | 27.7 (19; 39.6) | 10.9 (4.8) | 38 (14.4) |
| Latent (n = 7) | 26.8 (6.6) | 47.2 (10.6) | 57.2 (10.6) | 87 (12.9) | 44.7a (17.6) | 22.2 (11.7) | 9.2 (4.5) | 42.4 (9.8) |
| Total (n = 103) | 28.6 (5.5) | 50.5 (13.7) | 58.8 (13.6) | 88.8 (11.7) | 70.5 (28.5) | 26.2 (18.4; 39.4) | 10.8 (4.8) | 38.3 (14.2) |
Provocative.
3D: 3-dimensional; IVS: interventricular septum; LAp: axis length; LD: depth of thickening; LTh: thickening length; MTh: maximal thickness; SD: standard deviation.
MTh was greater for diffuse than for basal types (P = 0.003, 95% CI 2.4–10). No significant differences in MTh were observed between other types (midventricular, basal–midventricular, latent).
The planned excision volume (PEV) at ideal SM was larger for diffuse and midventricular types: midventricular versus basal–midventricular (P < 0.001, 95% CI 5.6–13.5); midventricular versus basal (P < 0.001, 95% CI 10.7–21.3); diffuse versus basal (P = 0.012, 95% CI 5.4–39.2); and diffuse versus basal–midventricular (P = 0.004, 95% CI 5.3–26.5). The real excision volume (REV) obtained during SM was larger for diffuse than for basal type (P = 0.004, 95% CI 1.8–8.4) . Overall, PEV exceeded REV: 26.2 (18.4–39.4) vs 10.3 (7.4–12.8) ml (P < 0.001).
The percentage for REV from PEV (%REV) did not differ significantly in all types except for midventricular one: midventricular versus basal–midventricular (P = 0.014, 95% CI −0.3 to −0.05) and midventricular versus basal (P = 0.015, 95% CI −0.3 to −0.03). PEV, REV and %REV did not differ between latent and obstructive forms.
The main anatomical AMBs characteristics based on the 3D virtual model analysis and confirmed intraoperatively are presented in the Supplementary Material, Table. According to the intraoperative data, AMBs were not visualized on 2D TTE in 53 (51.5%) patients, 49 with ALAMB (92.5%), while overdiagnosis occurred in 1 (1%) patient. CMR and CTA (with virtual 3D modelling) did not show AMBs in 1 (1%) patient using each method, and no patient received an overdiagnosis. AMBs were correctly diagnosed using 2D and 3D models in 31 (30.1%) and 82 (79.6%) patients, respectively, P < 0.001. All 3D-modelled AMBs were found intraoperatively and excised. The respective sensitivity and specificity for 2D TTE in diagnosing AMBs were 36.9% and 95%, and those for CMR and CTA with virtual 3D modelling were 97.1% and 100% for CMR and 98% and 100% for CTA.
ALAMBs were found in 76 (73.7%) patients. ALAMBs were longer than MAMBs [46.8 (SD 10.6) vs 37 (SD 11.3) mm (P < 0.001, 95% CI 4.9–14.6)] and PMAMBs [46.8 (SD 10.6) vs 38.5 (SD 17.7) mm (P < 0.04, 95% CI 0.6–16)]. The thickness did not differ between AMBs. Doubled AMBs (only ALAMBs) were found in 12 (11.7%) patients and tripled AMBs in 1 (1%) patient. One type of bundle was more frequent than the combination of 2 different AMBs: 57 (55.3%) vs 25 (24.3%). All 3 types of AMBs were noted in 2 (1.9%) patients.
Fibrous and short secondary chordae were resected; an average of 2.8 (SD 1.5, range 1–7) were resected in 31 patients (30.1%). The head of an anterolateral papillary muscle with direct insertion into the anterior mitral leaflet was resected in 1 patient.
Since August 2018 to prevent long-term aortic valve regurgitation we have performed its resuspension in most patients—61/72, 84.7% (the effectiveness of this procedure will be assessed later).
Table 4 demonstrates patients’ main perioperative characteristics. No patient had an iatrogenic ventricular septal defect or needed a mitral valve replacement. The mean cross-clamp time in a relatively small (mainly due to frequently performed aortic valve resuspensions) number of patients with an isolated SM (n = 20, 19.4%) was 89.6 (SD: 18.0) min. Thirteen patients (12.6%) required cardiopulmonary bypass restart, 9 (8.7%) patients because of mitral regurgitation grade 3–4 out of 4 that was managed with mitral valve repair and 1 patient for aortic valve replacement for organic valve disease. Three patients (2.9%) had the residual peak gradient >15 mmHg reduced by additional SM. Sixteen patients (15.5%) required pacemaker implantation for complete atrioventricular block. In addition, 3 (2.9%) cardiac complications were observed: 1 (1%) involved rupture of the LV posterolateral wall in an elderly female patient with mitral valve stenosis and its annular calcification and 2 (2%) involved pericarditis. Six (5.9%) non-cardiac events were recorded: 2 (2%) strokes, 1 (1%) perforated peptic ulcer with peritonitis, 2 (2%) occurrences of postoperative bleeding and 1 (1%) deep wound infection. One patient (1%) died from peritonitis on day 24.
Table 4:
Main perioperative characteristics
| Variables | Value (n = 103) (%) |
|---|---|
| AV replacement | 6 (5.8) |
| AV repair (total/resuspension) | 59 (57.3)/54 (52.4) |
| Aorta replacement | 6 (5.8) |
| CABG | 16 (15.5) |
| Maze procedure | 3 (2.9) |
| MV repair | 13 (12.6) |
| Concomitant with SM | 4 (3.9) |
| Secondary (CPB restart) | 9 (8.7) |
| Type of MV repair | |
| Alfieri stitch | 3 (2.9) |
| Posterior semi-annulus suture plasty | 8 (7.8) |
| Carpentier ring annuloplasty | 1 (1) |
| Commissurotomy | 1 (1) |
| Concomitant procedure (mean) | 1 (1; 2) |
| Cross-clamp time (min) | 115.2 (SD: 28.7) |
| Weight of resected muscle (g) | 10.8 (SD: 4.8), min. 4.1, max. 32.6 |
| Major perioperative complications | 9 (8.7) |
| 30-Day mortality | 1 (1) |
| Permanent pacemaker implantation | 16 (15.5) |
AV: aortic valve; CABG: coronary aortic bypass graft; CPB: cardiopulmonary bypass; MV: mitral valve; SD: standard deviation; SM: septal myectomy.
Follow-up
The median follow-up was 18.1 (standard error 1.32, min 2.9, max 46.3) months, 168.8 person-years of 100% complete follow-up. Two-year survival was 95% (standard error 2.5). There were 4 late deaths (3.9%) from non-cardiac causes: 2 (2%) because of malignant tumours, 1 (1%) alcoholism and 1 (1%) mediastinitis. Nine (8.8%) major adverse events were observed: myocardial infarction in 4 (3.9%) patients and complete atrioventricular block with pacemaker implantation, sternum osteomyelitis, mediastinitis, pericarditis and ischaemic stroke in 1 (1%) patient each. The sustained sinus rhythm after pacemaker implantation was restored in 4 (3.9%) patients. Table 2 shows the main TTE data at follow-up. Only 1 patient (1%) had a suboptimal (20–30 mmHg) residual gradient between the mid- and apical LV segments with diffuse type and the largest length from right coronary cusp nadir to the apex (121 mm). Both volumetric and dimensional TTE parameters of the LV increased compared with preoperative values (Table 2). Two (2%) patients progressed from grade 2 aortic regurgitation (preoperative) to grade 3 with regurgitation volume >30 ml. All patients showed improved New York Heart Association (NYHA) functional class compared with preoperative values (Table 1): 74 (72.5%) patients were in NYHA I, 25 (24.5%) in NYHA II and 3 (2.9%) in NYHA III (P < 0.001). Repeat CTA in 3 (2.9%) patients showed a 35–50% MTh reduction.
DISCUSSION
To our knowledge, our study is the largest single-centre study of cardiovascular surgery based on the 3D analysis of heart models created with 3D printing [11].
The underestimation of MTh by 2D TTE is associated with poor visualization of some IVS segments [6, 12]. We found a difference of >7 mm. The 3D visualization of the IVS is important for the accurate assessment of the anatomical substrate for LV cavity obstruction and SM planning. The absence of IVS hypertrophy associated with LV cavity obstruction can be confirmed only after 3D visualization and 2D TTE data are not sufficient. We observed no HOCM without IVS hypertrophy and SM was effective in all patients. Thus, we consider 3D visualization to be key to HOCM surgery.
The measurements obtained using our 3D IVS model (i.e. thickening length, length from right coronary cusp nadir to the apex, depth of thickening, MTh) allowed us to get a precise view of the IVS anatomy and optimal excision planning. It is desirable to take measurements along the axis towards the LV apex as well as the upper and lower borders of the excision close to the base of anterolateral and posteromedial papillary muscles. The myocardial excision can vary in depth. The major IVS thickening is more frequently observed in the direction from the basal segments to the apex at an angle to the LV longitudinal axis forming the septal band [13].
The lack of difference in PEV, REV and %REV between the latent and obstructive forms of hypertrophic cardiomyopathy argues for the involvement of a pronounced functional factor in the pathophysiology of the latent form. However, we observed hypertrophied IVS segments >13 mm in all patients with the latent form, and SM with AMBs excision was effective in all of them.
We propose that the large discrepancy between PEV and REV (%REV of ∼40%) reflects technical difficulties in reaching the distal IVS portions, an excision risk when approaching the base of the papillary muscle, and the intention to leave the IVS thicker than the ideal 10 mm because of the presence of crypts, the irregular IVS surface, possible 3D printing inaccuracies, and the layer-by-layer excision method. Therefore, we perform a 3D-modelled extended SM rather than ideal SM, guided by the original IVS thickness, anatomical type of HOCM and technical options. Our experience shows that the %REV of 30–40% is sufficient to achieve good SM outcomes.
The lack of effective visualization of the IVS using 2D TTE contributed to underestimation of the IVS thickness and the low TTE sensitivity for diagnosing AMBs. This may be explained by the high prevalence of ALAMBs (73.7%) in HOCM patients, which localized in the zone of poor 2D TTE visualization. By contrast, CMR and CTA with 3D modelling provided higher sensitivity and specificity in diagnosing the location, length and thickness of AMBs.
Despite the contradictory information from some authors [6] about the role of AMBs in obstruction, our experience confirms their presence in most patients (80.2%) with obstruction symptoms at rest and all patients with latent obstruction. We often observed AMBs attached to papillary muscles, which is likely to contribute to SAM development especially in patients with latent obstruction. We consider intraoperative search and excision of AMBs to be mandatory.
The proposed classification of AMBs (i.e. ALAMB, medial AMB, PMAMB) will subsequently lead to standardizing their analysis. These represent longitudinal bundles (along the LV long axis), although there are rare cases of transverse bundles. We observed them in 1 patient (1%).
A relatively high mean cross-clamp time (with no associated complications, however) of an isolated SM can be explained by the fact that intraoperatively we lay out carefully and sequentially without haste the excised tissue of the myocardium on the bottom of the 3D-model with the implanted needles showing the depth of the excision (Video 2).
Even though 4% of patients showed the restoration of sustained sinus rhythm within a month after surgery, the permanent pacemaker implantation rate turned out to be higher than expected (on average 9.8%) [14]. This might result from numerous concomitant procedures or the pursuit of radical SM sometimes exceeding the classic limits of safe resection, in particular, in case of severe or extreme hypertrophies.
Our study was also aimed to demonstrate the possibility, efficacy and applicability of creating 3D IVS models, performing virtual SM and using the 3D-printed results of virtual surgery without considerable extra time (the overall 3D-printed model build time did not exceed 14 h) and money expenses (the average 3D-printed model cost approximated $15) as a standard approach to each patient planned for SM. Such a personalized approach for planning and performing SM allowed us to eliminate the incidence of mitral valve replacements, the development of a non-optimal (≥30 mmHg) residual gradient in the LV obstruction area and LV obstruction recurrence in this study, thus significantly enhancing our previous (before 2017 when we started applying the 3D technology) SM outcomes [2] in HOCM patients.
Limitations
The most important limitations are the single-centre nature of our study and the small number of patients with absent comparative postoperative CTA and CMR data for most patients. Further patient enrolment is required to analyse fully the data for SM based on 3D IVS and AMBs models and to compare the outcomes with those of other surgical treatment options for HOCM.
CONCLUSION
Anatomical analysis of the IVS and AMBs based on CMR and CTA data with virtual 3D modelling is highly effective for SM planning. 2D TTE data is less informative.
SUPPLEMENTARY MATERIAL
Supplementary material is available at ICVTS online.
Conflict of interest: none declared.
Author contributions
Uladzimir Andrushchuk: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Supervision; Writing—original draft. Artsem Niavyhlas: Data curation; Investigation; Methodology; Software; Visualization; Writing—original draft. Vitali Adzintsou: Data curation; Formal analysis; Investigation; Methodology; Writing—original draft. Iryna Haidzel: Data curation; Methodology; Visualization. Hanna Model: Data curation; Methodology; Visualization. Aliaksandr Shket: Data curation; Methodology.
Reviewer information
Interactive CardioVascular and Thoracic Surgery thanks Walter J. Gomes, Leonardo Paim, Akif Turna and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.
Supplementary Material
ABBREVIATIONS
- %REV
Percentage for REV from PEV
- 2D
2-Dimensional
- 3D
3-Dimensional
- ALAMB
Anterolateral abnormal muscle bundle
- AMB
Abnormal muscle bundle
- CI
Confidence interval
- CMR
Cardiovascular magnetic resonance
- CTA
Computed tomography angiography
- HOCM
Hypertrophic obstructive cardiomyopathy
- IVS
Interventricular septum
- LV
Left ventricle
- MTh
Maximal IVS thickness
- NYHA
New York Heart Association
- PEV
Planned excision volume
- PMAMB
Posteromedial abnormal muscle bundle
- REV
Real excision volume
- SD
Standard deviation
- SM
Septal myectomy
- TTE
Transthoracic echocardiography
Presented at the 34th Annual Meeting of the European Association for Cardio-Thoracic Surgery, Barcelona, Spain, 8–10 October 2020.
REFERENCES
- 1.Messmer BJ. Extended myectomy for hypertrophic obstructive cardiomyopathy. Ann Thorac Surg 1994;58:575–7. [DOI] [PubMed] [Google Scholar]
- 2.Andrushchuk U, Adzintsou V, Niavyhlas A, Model H, Ostrovsky Y.. Early results of optimal septal myectomy using 3-dimensional printed models. Kardiochir Torakochirurgia Pol 2019;16:74–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Gruner D, Chan RH, Crean A, Rakowski H, Rowin EJ, Care M. et al. Significance of left ventricular apical–basal muscle bundle identified by cardiovascular magnetic resonance imaging in patients with hypertrophic cardiomyopathy. Eur Heart J 2014;35:2706–13. [DOI] [PubMed] [Google Scholar]
- 4.Schaff HV, Dearani JA, Ommen SR, Sorajja P, Nishimura RA.. Expanding the indications for septal myectomy in patients with hypertrophic cardiomyopathy: results of operation in patients with latent obstruction. J Thorac Cardiovasc Surg 2012;143:303–9. [DOI] [PubMed] [Google Scholar]
- 5.Tang B, Song Y, Cui H, Ji K, Yu Q, Zhu C. et al. Prognosis of adult obstructive hypertrophic cardiomyopathy patients with different morphological types after surgical myectomy. Eur J Cardiothorac Surg 2018;54:310–17. [DOI] [PubMed] [Google Scholar]
- 6.Maron MS. Clinical utility of cardiovascular magnetic resonance in hypertrophic cardiomyopathy. J Cardiovasc Magn Reson 2012;14:13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Ferrazzi P, Spirito P, Iacovoni A, Calabrese A, Migliorati K, Simon C. et al. Transaortic chordal cutting: mitral valve repair for obstructive hypertrophic cardiomyopathy with mild septal hypertrophy. J Am Coll Cardiol 2015;66:1687–96. [DOI] [PubMed] [Google Scholar]
- 8.Andrushchuk U, Adzintsou V, Nevyglas A, Model H.. Virtual and real septal myectomy using 3-dimensional printed models. Interact CardioVasc Thorac Surg 2018;26:881–2. [DOI] [PubMed] [Google Scholar]
- 9.Guo HC, Wang Y, Dai J, Ren CW, Li JH, Lai YQ.. Application of 3D printing in the surgical planning of hypertrophic obstructive cardiomyopathy and physician–patient communication: a preliminary study. J Thorac Dis 2018;10:867–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Hermsen JL, Burke TM, Seslar SP, Owens DS, Ripley BA, Mokadam NA. et al. Scan, plan, print, practice, perform: development and use of a patient-specific 3-dimensional printed model in adult cardiac surgery. J Thorac Cardiovasc Surg 2017;153:132–40. [DOI] [PubMed] [Google Scholar]
- 11.Ferrari E, Gallo M, Wang C, Zhang L, Taramasso M, Maisano F. et al. Three-dimensional printing in adult cardiovascular medicine for surgical and transcatheter procedural planning, teaching and technological innovation. Interact CardioVasc Thorac Surg 2020;30:203–14. [DOI] [PubMed] [Google Scholar]
- 12.Hoey ETD, Elassaly M, Ganeshan A, Watkin RW, Simpson H.. The role of magnetic resonance imaging in hypertrophic cardiomyopathy. Quant Imaging Med Surg 2014;4:397–406. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Takayama H, Yu SN, Sorabella R, Leb J, Pulerwitz TC, Cooper C. et al. Virtual septal myectomy for preoperative planning in hypertrophic cardiomyopathy. J Thorac Cardiovasc Surg 2019;158:455–63. [DOI] [PubMed] [Google Scholar]
- 14.Kim LK, Swaminathan RV, Looser P, Minutello RM, Wong SC, Bergman G. et al. Hospital volume outcomes after septal myectomy and alcohol septal ablation for treatment of obstructive hypertrophic cardiomyopathy: US Nationwide Inpatient Database, 2003-2011. JAMA Cardiol 2016;1:324–32. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.