Different calcification patterns of tricuspid and bicuspid aortic valves and their clinical impact

Abstract

OBJECTIVES

Mechanical strain plays a major role in the development of aortic calcification. We hypothesized that (i) valvular calcifications are most pronounced at the localizations subjected to the highest mechanical strain and (ii) calcification patterns are different in patients with bicuspid and tricuspid aortic valves.

METHODS

Multislice computed tomography scans of 101 patients with severe aortic stenosis were analysed using a 3-dimensional post-processing software to quantify calcification of tricuspid aortic valves (n = 51) and bicuspid aortic valves (n = 50) after matching.

RESULTS

Bicuspid aortic valves exhibited higher calcification volumes and increased calcification of the non-coronary cusp with significantly higher calcification of the free leaflet edge. The non-coronary cusp showed the highest calcium load compared to the other leaflets. Patients with annular calcification above the median had an impaired survival compared to patients with low annular calcification, whereas patients with calcification of the free leaflet edge above the median did not (P = 0.53).

CONCLUSIONS

Calcification patterns are different in patients with aortic stenosis with bicuspid and tricuspid aortic valves. Patients with high annular calcification might have an impaired prognosis.

Aortic valve stenosis represents a major socioeconomic burden in the Western world with a prevalence of 2.8% in >75-year-old individuals and 9.8% in octogenerians [1].

INTRODUCTION

Aortic valve stenosis represents a major socioeconomic burden in the Western world with a prevalence of 2.8% in >75-year-old individuals and 9.8% in octogenerians [1]. Due to the ageing of society, its prevalence is continuously increasing [2].

The aortic valve comprises 2 cell types: (i) valvular endothelial cells covering the valvular surface and preventing embolism by production of nitric oxide and (ii) valvular interstitial cells (VIC) responsible for the production and maintenance of valvular extracellular matrix proteins including collagen, elastin and glycosaminogylcans [3]. These proteins are necessary for the structural resistance of the valve to mechanical strain [3, 4]. Excessive mechanical strain causes a phenotypic switch of VIC towards osteoblasts that initiaties valvular calcification and finally results in aortic valve stenosis [5]. During the cardiac cycle, different mechanical forces impact the aortic valve [6]. During systole, there is laminar shear stress on the ventricular side of the aortic valve, sensed mainly by endothelial cells and resulting in the release of nitric oxide. In contrast, during diastole, significant axial pressure is present upon the aortic side of the aortic valve leaflets with tensile strain on the extracellular matrix and on the VIC. The highest stress during the cardiac cycle is present at the valvular commissures [7]. At the same time, laminar shear stress occurs over the right coronary cusp (RCC) and left coronary cusp (LCC) as blood flows into the coronary ostia. On the contrary, oscillatory shear stress is observed over the non-coronary cusp (NCC) because there is no runoff for the blood column (Fig. 1). Thus, the NCC is subjected to the highest tensile strain and oscillatory shear stress, both of which promote an osteoblastic phenotype switch of the VIC [8].

Figure 1:

Flow patterns and mechanical strain on tricuspid aortic valves and bicuspid aortic valves. During systole, laminar shear stress occurs on the ventricular side of the aortic valve. During diastole, significant axial pressure occurs on the aortic side of the aortic valve leaflets. The pressure is highest on the valvular commissures in tricuspid aortic valves. At the same time, oscillatory shear stress occurs over the non-coronary cusp because there is no runoff for the blood column (modified from [8]).

Bicuspid aortic valves (BAV) are classified using the Sievers classification. Type 0 valves do not exhibit a raphe, whereas type 1 valves have a single raphe. The raphe is formed by the malformation and obliteration of 2 of the 3 separate cusps during development. The 2 malformed cusps that merge into 1 cusp are separated by a prominent raphe [9]. Basic research experiments using computed or porcine models show changes in haemodynamics in BAV leading to abnormal stress relative to TAV leaflets [10, 11].

Additionally, due to their unnatural geometry, BAV undergo excessive strain on their free leaflet edges (FLE) during opening [12]. The abnormal stress leads to an inflammatory response and to endothelial dysfunction in the VICs, extracellular matrix remodelling and a VIC phenotype switch. All these components finally lead to calcification [13].

It remains unclear where and how the pathology of aortic valve calcification starts and whether BAVs and tricuspid aortic valves (TAV) show different calcification patterns.

Due to the described mechanical phenomena and our clinical observations, we hypothesized that patients with BAV and TAV exhibit a different calcification pattern and that different calcification patterns influence clinical outcome.

METHODS

Ethical statement

The study complies with the Declaration of Helsinki. Permission to perform this study was obtained from the Innsbruck Medical University institutional review board (No.1004/2019) on 24 November 2020. Written informed consent was waived by the institutional review board (retrospective study).

Patients

We identified 131 patients with severe aortic stenosis(AS) who were indicated for transcatheter aortic valve replacement (TAVR) or surgical aortic valve replacement (SAVR) between 2010 and 2019 at our institution. We formed 2 groups, 1 with TAVs and the other with BAVs. Due to the low prevalence of BAVs, we did not perform a further selection in this group. We included the available multislice computed tomography angiography (CTA) data from patients with BAV and performed the matching with the patients with TAV. Within the study population, patients with a BAV were matched to patients with a TAV by age at the time of preoperative CTA and by sex, if possible. Patients who had not undergone preoperative CTA [14] or who had images of inadequate quality were excluded from our study. Patients with BAV and TAV were matched for age, gender, the major cardiovascular (CV) risk factors and other comorbidities. After matching, we included 101 patients: 51 patients with TAV and 50 patients with BAV. The primary end point was survival. Follow-up was 100% complete. Clinical follow-up for survival was assessed after a median of 2.3 years. The optimal treatment for each patient was discussed in a multidisciplinary Heart Team meeting.

Quantification of calcification from computed tomography angiography

A 3-dimensional (3D) post-processing software (3mensio Medical Imaging, Bilthoven, the Netherlands) was used for the quantification of calcification in TAV and BAV as described previously [24]. CTA was performed using 128-slice dual source CTA (Somatom Definition Flash, Siemens Healthineers, Erlangen, Germany) with a detector collimation of 2 × 64 × 0.6 mm, a rotation time of 0.28 s and prospective electrocardiogram-triggering into the diastolic phase (70% of the RR-interval). Iopromide (Ultravist 370, Bayer HealthCare, Whippany, NJ), an iodine contrast agent, was injected and triggered into the arterial phase using bolus tracking (100 Hounsfield units threshold, ascending aorta). Images were reconstructed at a 0.75-mm slice thickness (increment, 0.4 mm).

The average diameter at the height of the annulus and the length of each cusp were measured (Fig. 2A) with a software tool integrated into the 3D post-processing software that served as a digital caliper. Using this software tool, we quantified the aortic root from the basal plate up to 15 mm above the basal plate and the upper part of the left ventricular outflow tract from the basal plate to 5 mm below as the area of interest for calcification. A measuring cylinder with a height of 20 mm (Fig. 2B) was adjusted automatically to the outer wall of our area of interest. A threshold of 650 Hounsfield units was defined as described previously for measurement of calcifications [15]. The lesions were identified with a digital caliper. An automated 3D volume growing algorithm was applied for 3D segmentation and quantification of calcification volumes. Manual corrections of the outer boarders were made in case of inaccuracies. Calcification of each leaflet was additionally measured using a specific Mercedes-star-like scanner tool in 3mensio (3mensio Medical Imaging) (Fig. 2C+D). In a next step, we measured annular calcification deposits (ANC) and calcifications of the FLE separately. ANC were separated from FLE based on visual evaluation on a cross-sectional axial image: ANC were defined as being located on the circular annular shape, and FLE as being located inside, respectively. Finally, we subtracted the FLE from the total aortic valve calcifications (AVC) to obtain the volume of the ANC. Measurements were performed by an experienced observer and supervised by a consultant with experience in CV medicine and imaging.

Figure 2:

Quantification of calcification using 3mensio Medical Imaging and distinct calcification patterns in tricuspid and bicuspid aortic valves. (A) In a first step, a measuring cylinder within the aortic root was defined in transverse sections. (B) Subsequently, the cranial (15 mm above the annulus) and caudal (5 mm below the basal plate) areas of interest were defined (red lines). The 2 grey lines indicate the height of the annulus. (C) The volume of valvular calcification was measured on each cusp using a specific Mercedes-star-like tool. (D) Finally, valvular calcifications were analysed in in 3 dimensions for every cusp.

The TAV cusps as well as the BAV cusps were treated the same during measuring. For a better comparison of the data, the BAV cusps were also separated into 3 individual cusps. This is possible because in Sievers type I BAV, the merged cusps could be identified [9].

Statistical analyses

The distribution of the data was evaluated using the Shapiro-Wilks test. Non-parametric continuous variables are expressed as median [lower quartile—upper quartile]. Parametrical continuous variables are given as the arithmetic mean [standard deviation (SD)]. Non-parametric data were compared using an unpaired Mann-Whitney U-test. Parametric data were compared by performing an unpaired Student t-test. The χ² test was used to determine differences in categorical variables between the 2 groups. Categorical variables are presented as numbers with corresponding percentages. Outcomes were analysed using time-to-event methods and presented using Kaplan–Meier estimates. The curves were compared using log-rank statistics. Results were considered statistically significant at P-values <0.05. Data management was performed with the 26th version of SPSS Statistics (IBM Corp, Armonk, NY, USA) and Microsoft Excel 365 (Microsoft Corp., Redmond, WA, USA). Statistical analyses were performed using RStudio (RStudio PBC, Boston, MA, USA). Data comparison was facilitated with the compareGroups package for R (R Foundation for Statistical Computing, Institute for Statistics and Mathematics, Vienna Austria). Kaplan–Meier estimates and the corresponding log-rank statistics were calculated and plotted using the R survival ggpubr package.

RESULTS

Quantification of calcification was performed using 101 age and sex matched patients with AS. Fifty-one patients had TAVs, whereas 50 displayed BAVs. A total of 39% of the patients were female, and the median age was 74 [67 – 81] years. Fifty percent were older than 75 years of age when computed tomography imaging was performed. There were no differences in distribution of age or sex after matching, and no difference in the major CV factors and kidney dysfunction. A detailed patient description is displayed in Table 1.

Table 1:

Characteristics of patients with tricuspid or bicuspid aortic valves

	Total N = 101	BAV N = 50	TAV N = 51	P-value
Age, years	74.0 [67.0; 81.0]	74.5 [67.2; 81.0]	74.0 [67.0; 81.0]	0.854
Age over 75, n (%)	50 (49.5)	25 (50.0)	25 (49.0)	1.000
Female, n (%)	39 (38.6)	19 (38.0)	20 (39.2)	1.000
Body mass index, kg/m²	26.0 [23.0; 29.0]	25.0 [23.0; 28.0]	27.0 [24.0; 29.5]	0.068
Weight, kg	76.0 [65.0; 83.0]	73.5 [62.0; 81.8]	76.0 [67.5; 83.5]	0.335
Height, cm	169 (8.85)	169 (9.12)	168 (8.63)	0.511
Hypertension, n (%)	89 (89.9)	44 (91.7)	45 (88.2)	0.742
Diabetes mellitus, n (%)	27 (26.7)	11 (22.0)	16 (31.4)	0.401
Smoking, n (%)	42 (44.7)	20 (43.5)	22 (45.8)	0.982
COPD, n (%)	20 (23.0)	11 (27.5)	9 (19.1)	0.505
CAD, n (%)	60 (59.4)	27 (54.0)	33 (64.7)	0.372
Stroke/TIA, n (%)	3 (3.37)	2 (4.65)	1 (2.17)	0.608
PAD, n (%)	7 (6.93)	2 (4.00)	5 (9.80)	0.436
PAH, n (%)	42 (45.7)	18 (40.9)	24 (50.0)	0.506
Previous revascularization, n (%)	25 (25.5)	11 (22.9)	14 (28.0)	0.730
Preop creatinine level, mg/dl	1.00 [0.90; 1.35]	1.00 [0.90; 1.30]	1.01 [0.90; 1.42]	0.443
Preop GFR, ml/min	60.0 [47.0; 60.0]	60.0 [52.0; 60.0]	60.0 [42.5; 60.0]	0.302
Preop dialysis, n (%)	24 (24.0)	11 (22.4)	13 (25.5)	0.903
Preop atrial fibrillation, n (%)	43 (43.0)	19 (38.0)	24 (48.0)	0.419
NYHA classification, n (%)				0.343
<= I	7 (7.07)	5 (10.0)	2 (4.08)
II	17 (17.2)	10 (20.0)	7 (14.3)
III	62 (62.6)	27 (54.0)	35 (71.4)
IV	13 (13.1)	8 (16.0)	5 (10.2)
Hs–cTnT, ng/l	18.9 [11.7; 40.4]	21.3 [11.6; 43.4]	17.1 [11.8; 36.2]	0.770
NT–proBNP, pg/ml	1319 [588; 3584]	1288 [626; 6014]	1351 [584; 3032]	0.480
CK-MB, U/l	83.0 [55.5; 111]	81.5 [54.0; 108]	83.0 [56.5; 118]	0.553
EuroSCORE II	4.13 [2.16; 7.61]	4.12 [2.59; 7.99]	4.14 [2.09; 7.61]	0.929
STS score	0.03 [0.01; 0.04]	0.03 [0.02; 0.04]	0.03 [0.01; 0.04]	0.504
Sievers classification				<0.001
Type 0		3 (6.00)
Type 1		47 (94.0)
Cusp fusion, n (%)				.
LC/NC		2 (4.26)
LC/RC		35 (74.5)
NC/RC		10 (21.3)
EF, %	51.7 (14.5)	50.4 (13.9)	53.0 (15.1)	0.365
EF < 30%, n (%)	91 (90.1)	45 (90.0)	46 (90.2)	1.000
LVEDD, mm	51.2 [46.1; 57.8]	51.5 [45.1; 57.2]	51.0 [46.4; 58.0]	0.835
LVESD, mm	34.4 [29.3; 41.4]	34.3 [30.0; 42.6]	34.4 [28.8; 40.3]	0.453
AVA, cm²	0.62 (0.17)	0.59 (0.16)	0.65 (0.18)	0.058
Mean AO gradient, mmHg	43.0 [33.0; 52.0]	43.5 [34.5; 51.2]	42.0 [33.0; 53.0]	0.891
Max AO gradient, mmHg	64.5 [54.5; 80.0]	63.0 [56.0; 80.0]	66.0 [54.0; 80.0]	0.913
SVI	29.1 [22.3; 34.8]	29.1 [20.4; 34.0]	29.5 [22.6; 35.2]	0.524
sPAP, mmHg	45.0 [36.5; 55.0]	50.0 [40.0; 60.0]	45.0 [35.0; 50.0]	0.108
sPAP > 36 mmHg, n (%)	46 (74.2)	25 (75.8)	21 (72.4)	0.993
sPAP > 50 mmHg, n (%)	26 (41.9)	17 (51.5)	9 (31.0)	0.170
AI grade, n (%)				0.220
0	27 (27.8)	16 (34.0)	11 (22.0)
I	44 (45.4)	18 (38.3)	26 (52.0)
II	24 (24.7)	11 (23.4)	13 (26.0)
III	2 (2.06)	2 (4.26)	0 (0.00)
Procedure, n (%)				0.249
Conservative	12 (11.9)	9 (18.0)	3 (5.88)
SAVR	36 (35.6)	17 (34.0)	19 (37.3)
SAVR + CABG	15 (14.9)	8 (16.0)	7 (13.7)
TAVR	38 (37.6)	16 (32.0)	22 (43.1)
Prosthesis size, mm	23.0 [23.0; 26.0]	25.0 [23.0; 26.0]	23.0 [22.0; 26.0]	0.252
Prosthesis type, n (%)				0.473
Biological	90 (98.9)	42 (97.7)	48 (100)
Mechanical	1 (1.10)	1 (2.33)	0 (0.00)
TAVR approach, n(%)				0.042
Transapical	13 (34.2)	9 (50.0)	4 (20.0)
Transaxillary	1 (2.63)	1 (5.56)	0 (0.00)
Transfemoral	24 (63.2)	8 (44.4)	16 (80.0)
Cross-clamp time, min	56.5 [0.00; 88.0]	50.0 [0.00; 81.5]	69.0 [0.00; 90.0]	0.404
CPB time, min	87.0 [0.00; 129]	76.0 [0.00; 117]	103 [0.00; 134]	0.253

AI: aortic valve insufficiency; AO gradient: aortic valve gradient; AVA: aortic valve area; BAV: bicuspid aortic valve; CAD: coronary artery disease; CK-MB: creatine kinase MB; CABG: coronary artery bypass grafting; COPD: chronic obstructive pulmonary disease; CPB: cardiopulmonary bypass; EF: ejection fraction; EuroSCORE II: The European System for Cardiac Operative Risk Evaluation II; GFR: glomerular filtration rate; Hs-cTnT: high-sensitivity cardiac troponin T; LCC: left coronary cusp; LVEDD: left ventricular diastolic dysfunction; LVESD: left ventricular systolic dysfunction; NCC: non-coronary cusp; NT–proBNP: N-terminal prohormone of brain natriuretic peptide; NYHA: new York Heart Association;PAD: peripheral artery disease; Preop: preoperative; SAVR: surgical aortic valve replacement; RCC: right coronary cusp; sPAP: systolic pulmonal arterial pressure; STS score: Society of Thoracic Surgeons score; SVI: stroke volume index; TAV: tricuspid aortic valve; TAVR: transcatheter aortic valve replacement.

A total of 94% of patients with BAV showed type 1 morphology according to the Sievers classification. Twenty-one percentexhibited a fusion of the NCC and the RCC, whereas 75% had a fusion of the LCC and the RCC, and 4% had fusion of the LCC and NCC (Table 1). The mean ejection fraction was 53% (15.1) in the TAV and 50% (13.9) in the BAV group. The aortic valve area was 0.65 cm² in the TAV and 0.6 cm² in the BAV group, and the mean pressure gradient showed no significant differences in the 2 groups. There was no significant difference in the stroke volume index or in the rates of aortic regurgitation (Table 1).

A total of 36% of the patients underwent SAVR, 15% had combined SAVR and CABG, 38% received TAVR and 12% were treated conservatively. There was no difference in treatment in the BAV and TAV groups. There was no significant differences in surgical approach for the TAVR procedure between the groups. Neither the cardiopulmonary nor the aortic cross-clamp times were different between patients with BAV and TAV (Table 1).

The aortic valves showed a mean annulus diameter of 29 (2.6) mm in the TAV group and 30 (3.1) mm in the BAV group. The leaflet length of the RCC was increased in the BAV group. BAV and TAV showed significant differences in calcification patterns (Fig. 3; Fig. 4). BAV had a higher AVC volume (Table 2). Moreover, BAV showed increased calcification of the NCC and a significantly more calcification of the FLE (Table 2; Fig. 3). The NCC showed the highest calcium load (Table 2; Fig. 3).

Figure 3:

Calcification patterns in tricuspid and bicuspid aortic valves. (A) The non-coronary cusp showed the highest calcium load compared to the other leaflets irrespective of the biscuspid aortic valve . The bicuspid and tricuspid aortic valves showed significant differences in calcification patterns. (B) Bicuspid aortic valves were similar, with a greater total amount of aortic valvular calcification. (C) There was no difference in annular calcification between patients with tricuspid and bicuspid aortic valves. (D) Bicuspid aortic valves showed significantly higher calcification of the free leaflet edge. AN: annular; BAV: bicuspid aortic valve; FLE: free leaflet edges; LCC: left coronary cusp; NCC: non-coronary cusp; RCC: richt coronary cusp; TAV: tricuspid aortic valve.

Figure 4:

Bicuspid aortic valves calcify at the free leaflet edge. Quantification of valvular calcifications reveals distinct calcification patterns in transverse (A) and coronary sections (B) as well as in 3-dimensional reconstruction images (C). (D) In analogy to the valvular flow patterns, bicuspid aortic valves show calcifications on the free leaflet edge, whereas tricuspid aortic valves are mainly calcified on the commissures and the valvular annulus. BAV: bicuspid aortic valve; LCC: left coronary cusp; NCC: non-coronary cusp; RCC: richt coronary cusp; TAV: tricuspid aortic valve.

Table 2:

Quantification of calcification

	Total N = 101	BAV N = 50	TAV N = 51	P-value
Avg. diameter basal plate, mm	29.3 (2.87)	29.9 (3.05)	28.7 (2.58)	0.046
Leaflet length, mm
NCC	17.5 (2.73)	18.0 (3.06)	17.1 (2.30)	0.092
RCC	15.3 [13.4; 16.7]	15.9 [14.1; 17.7]	14.0 [13.1; 15.8]	0.006
LCC	16.3 (2.79)	16.6 (2.89)	16.0 (2.69)	0.248
AVC, mm³
Total	718 [390; 1073]	782 [480; 1270]	568 [307; 846]	0.014
NCC	267 [140; 430]	325 [165; 592]	242 [109; 368]	0.018
RCC	185 [96.2; 312]	212 [94.5; 368]	165 [96.6; 261]	0.143
LCC	180 [82.2; 298]	190 [91.0; 334]	172 [73.9; 290]	0.366
FLE calcification, mm³
Total	271 [140; 532]	407 [254; 666]	192 [116; 316]	0.001
NCC	123 [32.2; 224]	157 [85.2; 343]	82.4 [20.8; 189]	0.004
RCC	85.3 [28.9; 141]	115 [46.2; 200]	52.5 [18.3; 103]	0.001
LCC	64.0 [16.9; 144]	88.2 [28.7; 191]	43.7 [13.5; 92.9]	0.035
AN calcification, mm³
Total	341 [189; 546]	397 [193; 568]	333 [185; 525]	0.625
NCC	123 [73.1; 229]	137 [88.6; 249]	117 [63.2; 192]	0.130
RCC	96.6 [42.1; 178]	87.2 [42.2; 176]	97.9 [48.3; 175]	0.726
LCC	98.4 [44.5; 176]	93.4 [41.8; 173]	99.9 [48.0; 190]	0.575

AN: annular; AVC: aortic valve calcification; Avg.: average; BAV: bicuspid aortic valve; FLE: free leaflet edge; LCC: left coronary cusp; NCC: non-coronary cusp; RCC: right coronary cusp; TAV: tricuspid aortic valve.

Table 3:

Calcification characteristics

TAV (n = 51)	NCC	RCC	LCC	P-value
Leaflet length, mm	17.1 (2.30)	14.5 (2.00)	16.0 (2.69)	<0.001
AVC, mm³	242 [109; 368]	165 [96.6; 261]	172 [73.9; 290]	0.132
FLE calcification, mm³	82.4 [20.8; 189]	52.5 [18.3; 103]	43.7 [13.5; 92.9]	0.102
AN calcification, mm³	117 [63.2; 192]	97.9 [48.3; 175]	99.9 [48.0; 190]	0.669

BAV (n = 50)	NCC	RCC	LCC	P-value
Leaflet length, mm	18.0 (3.06)	16.0 (2.80)	16.6 (2.89)	0.003
AVC, mm³	325 [165; 592]	212 [94.5; 368]	190 [91.0; 334]	0.013
FLE calcification, mm³	157 [85.2; 343]	115 [46.2; 200]	88.2 [28.7; 191]	0.033
AN calcification, mm³	137 [88.6; 249]	87.2 [42.2; 176]	93.4 [41.8; 173]	0.022

AN: annular; AVC: aortic valve calcification; BAV: bicuspid aortic valve; FLE: free leaflet edge; LCC: left coronary cusp; NCC: non coronary cusp; RCC: right coronary cusp; TAV tricuspid aortic valve;

The total amount of calcification was not associated with survival (Fig. 5). However, patients with increased amounts of calcification were generally sicker with increased EuroSCORE II, Society of Thoracic Surgeons score and increased nt-proBNP levels (Supplementary Tables 1–4).

Figure 5:

Survival of patients with high and low calcific loads. (A) The survival between TAV and BAV patients was not affected. (B) Patients with valvular calcification above the median did not show impaired long-term survival. (C) However, annular calcification above the median was associated with impaired outcome. (D) Calcification of the free leaflet edge did not have an impact on long-term survival. BAV: bicuspid aortic valve; FLE: free leaflet edges; TAV: tricuspid aortic valve.

Patients with ANC above the median (>341 cm²) exhibited impaired long-term survival compared to patients with low annular calcifications (Fig. 5). This was not the case for patients with FLE calcifications (Fig. 5). To rule out BAV as a confounding variable for survival, we additionally investigated the survival of patients with BAV versus those with TAV. There was no impaired survival between the different valve types (Fig. 5). The results stayed the same even when we excluded patients who underwent the combined AVR and CABG procedure (Supplementary Fig. 1). Additionally, there was no difference between patients above or below the median for annular calcifications in the different procedure groups (Supplementary Table 5).

DISCUSSION

This study, using multislice computed tomography data, provides evidence for different calcification patterns in patients with AS with BAV and TAV and explores possible implications of calcification patterns for patient prognosis. BAV exhibit higher calcification volumes, increased calcification of the NCC and greater calcification of the FLE. The NCC shows the highest calcium load, irrespective of valve type. Patients with annular calcification above the median exhibit impaired survival compared to patients with low annular calcification.

Due to the rapid evolution of catheter-based valve replacement procedures, aortic valve disease is in the spotlight of CV research. Evidence regarding periprocedural aspects of surgically or transcatheter-based valve replacement is constantly increasing [16]. However, underlying calcification mechanisms are still poorly understood. Haemodynamically relevant AVC develops over many years [17].

Tissue subjected to major mechanical strain is a predilection site for the onset of osteoblastic activity [13]. Our study revealed that the NCC is most affected by calcification. The most likely reason is the distinct flow profile in the non-coronary sinus region where a more oscillating pattern leads to higher sheer stress and an increased degree of calcification [18]. This situation might be due to the missing runoff route for the blood column in the non-coronary leaflet. An historic echocardiographic study revealed beginning sclerotic thickening of the aortic valve mainly on the NCC [19], which is strongly supported by our results.

BAV remain the most common congenital cause for the development of aortic valve calcification. The most common type of BAV is type 1 morphology with raphe L/R (>80%), with a fused right and left coronary leaflet, commonly associated with a longer non-coronary leaflet in order to compensate for the fusion.

It is known that the calcification mechanisms of BAV and TAV are different because they have distinctly different genetic patterns [20] and biochemical properties [21]. Due to their geometry, BAV undergo excessive strain, stretch and wall shear stress on their FLE during opening [7, 12, 13]. Cardiac CT is an established and accurate method for the differentiation between BAV and TAV [22]. Our study confirms for the first time that BAV exhibit a different calcification pattern than TAV by using 3D sizing by multidetector computed tomography. Our results underline the hypothesis that mechanical strain results in an increased calcification load in BAV, in particular on the mechanically challenged valvular FLE [23, 24]. However, it remains unclear what exactly triggers the osteoplastic activity and calcification [5].

Treatment of BAV stenosis with TAVR is currently emerging, with promising initial results [25]. However, it still faces technical challenges due to the asymmetry of the sinuses and leaflets, leading to a higher rate of procedural complications such as paravalvular leakage (PVL) and worse outcomes compared to TAV [26]. The increased calcium load in the non-coronary sinus compared to those in the opposite ones could explain the higher rate of PVL—but also other complications such as annulus rupture during TAVR [16]. A high aortic annular calcium load is a well-known risk factor for both [25]. Furthermore, the variations in the BAV calcification patterns lead to local distortion resulting in suboptimal positioning of the stent, higher stress on calcium deposits, which may result in embolization, and high stress on the crown-shaped profile of the native aortic valve leaflets [16]. Additionally, a higher calcium load is positively associated with an increased risk of annular rupture [16]. The calcium load of the aortic valve in general is higher in patients with BAV [27, 28]. Furthermore, in TAVR, increased amounts of annular and leaflet calcium predict PVL [29, 30]. Additionally, in our study, patients with annular calcification above the median had an impaired long-term survival compared to patients with a calcification load below the median. This result was not true for patients with a higher calcification load on the FLE or a higher total calcification load. However, further studies are necessary to document this conclusion, because our study had several limitations. We assume that annular calcification is more relevant for patient prognosis than is free leaflet calcification, because annular calcifications have a direct periprocedural impact on surgical and transcatheter-based results of valve replacement. Moreover, they could be more relevant for the occurrence of PVL. Many calcified aortic valves show major calcifications of the raphe and its adjoining annular region, which is the region of highest mechanical strain in TAV [8]. We assume that this might be the reason for most of the annular calcification and the calcification of the raphes in TAV.

Limitations

The limitations of our study are the sample size, the lack of propensity matching, the heterogeneity of the sample and the baseline differences that could contribute to worse long-term outcomes. Furthermore, our goal was to characterize our patients in detail and present the data in the main manuscript and in the supplementary material to minimize confounders due to the retrospective character of the trial.

Altogether, our study provides evidence for different calcification patterns in patients with bicuspid and tricuspid aortic valves. Patients with a high calcification load in the valvular annulus—but not the FLE—have an impaired prognosis. Further studies are necessary to identify the exact consequences of the different calcification patterns.

Funding

This work was supported by funds from the Austrian Science Fund (FWF) to CGT, JH and IT (P 32821).

Conflict of interest: The authors have nothing to disclose.

Glossary

Abbreviations

ANC

annular calcification deposits

AS

aortic stenosis

AVC

aortic valve calcification

BAV

bicuspid aortic valve

CT

computed tomography

CTA

computed tomography angiography

FLE

free leaflet edge

LCC

left coronary cusp

NCC

non-coronary cusp

PVL

paravalvular leakage

RCC

right coronary cusp

SAVR

surgical aortic valve replacement

TAV

tricuspid aortic valve

TAVR

transcatheter aortic valve replacement

3D

3-dimensional

VIC

valvular interstitial cells

Data availability statement

The data underlying this article will be shared upon reasonable request to the corresponding author.

Author contributions

Can Gollmann-Tepeköylü: Conceptualization; Data Curation; Formal Analysis; Methodology; Investigation; Validation; Writing—original draft; Writing—review & editing. Felix Nägele: Data Curation; Formal Analysis; Methodology; Investigation; Validation; Writing—original draft; Writing—review & editing. Clemens Engler: Formal Analysis; Methodology; Investigation; Validation; Writing—revision; Writing—review & editing. Leon Stoessel: Formal Analysis; Data Curation; Writing—review & editing. Berit Zellmer: Data Curation; Writing—review & editing. Michael Graber: Data Curation; Writing—review & editing. Jakob Hirsch: Data Curation; Writing—review & editing. Leo Pölzl: Data Curation; Writing—review & editing. Elfriede Ruttmann: Data Curation; Writing—review & editing. Ivan Tancevski: Formal Analysis; Methodology; Writing—original draft; Writing—review & editing. Christina Tiller: Data Curation; Writing—review & editing. Fabian Barbieri: Data Curation; Writing—review & editing. Lukas Stastny: Data Curation; Writing—review & editing. Sebastian J Reinstadler: Data Curation; Writing—review & editing. Cenk Özpeker: Data Curation; Writing—review & editing. Severin Semsroth: Data Curation; Writing—review & editing. Nikolaos Bonaros: Writing—review & editing. Michael Grimm: Writing—review & editing. Gudrun Feuchtner: Methodology; Writing—review & editing. Johannes Holfeld: Project administration; Supervision; Methodology; Writing—original draft; Writing—review & editing.