Characterization of Aortic Valve Stenosis by CT Angiography in a Diverse US Cohort

Abstract

Background

Aortic stenosis (AS) involves calcific and fibrotic degeneration of the valve tissue. The only non-invasive method for evaluating both processes is contrast-enhanced computed tomography (CTA). We aimed to explore the differences in aortic valve (AV) tissue composition across sex, race/ethnicity, and AS hemodynamic phenotype in U.S. patients referred for transcatheter AV replacement (TAVR) planning.

Methods

We retrospectively analyzed symptomatic AS patients who underwent CTA for TAVR planning between 2015 and 2022. Using semi-automated software, we quantified the AV tissue composition by fibrotic, calcific, and fibrocalcific volumes, and the fibro-calcific ratio (fibrotic/calcific volume) as a measure of valve phenotype.

Results

The study included 651 patients (mean age 84 years; 55% women) with 38% non-Hispanic (NH)-White, 27% Hispanic, and 13% NH-Black. Women had lower fibro-calcific (230 mm³/cm² vs. 293 mm³/cm², P<0.001) and calcific volumes (85mm³/cm² vs. 149 mm³/cm², P<0.001), and higher fibrocalcific ratio (1.47 vs. 0.83, P<0.001). No differences were observed in the fibrotic volumes (P=0.805). NH-White women had higher fibro-calcific (256mm³/cm², P=0.002) and fibrotic volumes (145 mm³/cm², P<0.001), and fibro-calcific ratio (1.57, P=0.01) compared to Hispanic and NH-Black. No differences were found among men. High-gradient AS had higher fibro-calcific (295 vs 219 mm³/cm², P<0.001) and calcific volumes (148 vs 88 mm³/cm², P<0.001), and a lower fibro-calcific ratio (0.90 vs 1.45, P<0.001), although no difference in fibrotic volume (P=0.099) compared to low-gradient.

Conclusions

Phenotypical differences in CTA valve tissue composition exist in AS patients referred for TAVR, with females and low-gradient AS showing a proportional more fibrotic phenotype. NH-White women have the highest fibrotic tissue composition, and no differences are evident among men.

Graphical Abstract

INTRODUCTION

Degenerative aortic stenosis (AS) is the most common valvular heart disease requiring intervention in Europe and North America¹. Its prevalence rises with age, peaking at 12.4% among the elderly, and due to the current aging population, it is projected to more than double by 2050². The pathophysiology is complex and involves a progressive fibro-calcific remodeling of the leaflets that leads to restrictive opening and ultimately myocardial remodeling³. Echocardiography remains the primary guideline-recommended method to assess AS.^1,2. However, accurately determining the severity of AS using echocardiography alone can be challenging in real-world settings, with up to 40% of patients showing discordant findings. In such cases, assessing aortic valve calcification using a calcium score derived from non-contrast computed tomography (CT) has proven to be valuable^4–6. Notwithstanding, while the AV calcium score serves as a hemodynamically independent marker, is not without limitations, and exclusively evaluates the calcific process of AS^5,7. Even though mineralization/calcification of the leaflets has been long considered the main driver of the development and progression of AS, it is also apparent that women have lower AV calcium scores and indexed calcium density values than men for the same hemodynamic severity of AS^8–10. Currently, there are no established medical therapies for AS. While most trials have focused on calcium metabolism, the fibrotic process has been largely overlooked, in part due to its limited visibility with conventional imaging techniques. Recent widespread use of contrast-enhanced ECG-synchronized computed tomography angiography (CTA), together with the development of dedicated semiautomatic software platforms, has provided a unique opportunity to further characterize the phenotype of AS by assessing simultaneously the calcific and fibrotic remodeling of the AV leaflets^11,12.

Initial investigations using this CTA-based approach have identified varied patterns of AV fibro-calcific remodeling, attributed to a complex interplay of factors including age, sex, valve morphology (tricuspid vs bicuspid), and hemodynamic phenotype (high versus low gradient-AS)^13–15. However, most of the studies were relatively small and performed in exclusively European cohorts with no mention of race/ethnicity. In addition, no effort was made to explore the influence of race/ethnicity on degenerative AS. Hence, there is currently no data available that sheds light on the composition of the AV among different racial/ethnic subgroups with AS¹⁶.

We aimed to investigate the differences in CTA-AV phenotype in a diverse US population with severe AS referred for CTA-TAVR evaluation.

METHODS

Study Population

All consecutive adults with native AS (aged >18 years) who underwent a cardiac CTA in the setting of pre-TAVR planning at Montefiore Healthcare System from 2015 to 2022 were identified from the institutional cardiac CT registry. We excluded studies with suboptimal CTA images for valve quantification or those with concordant moderate AS by echocardiographic criteria. Demographic data and cardiovascular risk factors such as body mass index (BMI), hyperlipidemia, hypertension, diabetes mellitus, coronary artery disease, and lipid levels, were collected through retrospective chart review. Race and ethnicity were self-identified at the time of registration and recorded in our medical records. Ethnicity was categorized as Hispanic or non-Hispanic (NH). Within the NH group, race was categorized as NH White, NH Black, NH Asian, and other/unknown. The study was approved by our Office of Human Research Affairs at Albert Einstein College of Medicine and consent was waived due to study design. Data are available upon request and restricted to institutional and IRB approvals.

Scan Protocol and Image Reconstruction

All scans were performed according to the Society of Cardiovascular CT guidelines¹⁷. The CTA acquisition was performed using three scanners: 64-slice GE Lightspeed VCT, Optima CT660, and 64-slice Philips IQon. The scans utilized retrospective gating without dose modulation and 80–120 mL of Isovue-370 contrast, adjusted based on the patient’s renal function and weight. All studies were acquired using 120 kV, and the tube current (mA) was set according to the patient’s body mass index as per scanner protocol. No medication was administered before the scan. Images were reconstructed from 5–95% of the cardiac cycle with a slice thickness of 0.625 mm and a slice increment of 0.625 mm using a standard kernel.

CT Analysis of Aortic Valve

Quantitative analyses of AV tissue composition were performed using semi-automated software (Autoplaque version 2.5, Cedars-Sinai Medical Center, Los Angeles, CA, USA) on the best diastolic phase (50–95%) of the R-R interval when the aortic valve was closed. The detailed process has been described previously, and the intra and interobserver reproducibility has been demonstrated^14,15,18. Image quality was graded on a four-tier scale with 1=suboptimal and 4=excellent image quality, considering the presence of cardiac motion, inadequate contrast opacification (<150 HU), and overall image noise. Patients with suboptimal image quality were excluded from the analysis as mentioned before.

The region of interest involving the aortic valve was defined between the lower coronary ostium and the CTA-defined virtual basal ring¹⁷ using multiplanar reconstructions. Serial cross-sectional images orthogonal to the longitudinal axis of the ascending aorta were used to contour the valve around its perimeter with 15–20 adjustable points, excluding the aortic wall. Adaptive scan-specific Hounsfield unit (HU) thresholds for fibrotic and calcific tissue components were automatically identified for each patient using Gaussian mixture modeling, which fits Gaussian functions to the HU histograms for the blood pool, non-calcific, and calcific AV tissue. Thresholds for fibrotic (lower 99.7th percentile of blood pool HU) and calcific tissue (upper 99.7th percentile blood pool HU) were automatically generated. The volumes of fibrotic and calcific tissue in the AV leaflets were measured and then indexed to the AV annulus area, to account for differences in valve size as previously reported^11,18. The sum of the fibrotic and calcific volume provided the fibro-calcific volume, while the fibrotic volume divided by the calcific volume provided the fibro-calcific ratio. The fibrocalcific ratio reflects the relative composition of valve tissue, with values >1 indicating a predominantly fibrotic phenotype and values <1 signifying a predominantly calcific phenotype, where calcification comprises at least 50% of the measured volume.

CT analyses and measurements were conducted by a Level III board-certified cardiac CT cardiologist (D.L.) and two experienced CT readers (A.F. and P.P.).

Echocardiography

Transthoracic echocardiography was performed and reported according to the American Society of Echocardiography guidelines. We included patients with high-gradient severe AS (HG-AS) as defined by an AV peak velocity (Vmax) ≥4 m/s or mean pressure gradient (ΔP) ≥40 mm Hg and AV area (AVA) ≤1.0 cm²; and low-gradient severe AS (LG-AS) was defined as AVA ≤1.0 cm² with resting Vmax <4 m/s or mean ΔP <40 mm Hg regardless of LVEF or left ventricular stroke volume index as recently published^19,20. Left ventricular systolic function was assessed using LVEF as measured by Simpson’s biplane method.

Statistical Analysis

Continuous variables are presented as mean ± standard deviation (SD) when normally distributed, and as median and interquartile range (IQR) when not normally distributed. Normal distribution was evaluated with the Shapiro-Wilk test. Categorical data are presented as counts and percentages. Continuous variables were compared with one-way ANOVA and Student’s unpaired t-test (parametric test) or the Kruskal-Wallis test and Wilcoxon rank-sum test (non-parametric test) between three and two groups, respectively. Categorical variables (as frequencies or percentages) were compared with χ² test or the Fisher exact test, as appropriate. Multivariable linear regression analyses were performed to identify factors associated with different components of aortic valve volume. The analysis included clinically and statistically relevant variables: age, sex, BMI, hypertension, diabetes, chronic kidney disease, dyslipidemia, and race/ethnicity. All statistical tests were two-sided, with P-values <0.05 considered statistically significant. Statistical analyses were performed using R, version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

Study population

After the exclusion of patients with suboptimal CTA image quality (n=172) and concordant moderate AS (n=83), the final cohort consisted of 651 patients, Figure 1, Table 1. Overall, 55% were female and the median age was 81 (73–86) years. The racial/ethnic distribution was 38% NH-White (n=245), 27% Hispanic (n=176), and 13% NH-Black (n=84). Due to the small sample size of NH-Asian patients (n=8), they were excluded from the analysis, along with those in the Unknown subgroup (n=138, 21%). There was a higher prevalence of NH-Black women compared to men (17.7% vs. 7.1%, P<0.001). Furthermore, women exhibited higher low-density lipoprotein-cholesterol (LDL-C) and total cholesterol levels (P=0.024 and P<0.001, respectively) and were less frequently affected by coronary artery disease (CAD; 12.1% vs 25.4%, P<0.001) when compared to men. Detailed demographic characteristics stratified by sex are detailed in Table 1.

Figure 1. Flowchart of patients’ selection and exclusion.

CTA, computed tomography angiography; TAVR, transcatheter aortic valve replacement.

Table 1.

Demographic and clinical characterization of the overall population, by sex.

	Overall	Women	Men	P value
	N = 651	N = 356	N = 295
Age, years	84 [75, 88]	83 [76, 88]	84 [74, 88]	0.922
BMI	26.4 [23.3, 31.0]	26.8 [23.2, 31.8]	26.2 [23.4, 29.9]	0.168
BSA	1.8 [1.6, 2]	1.7 [1.6, 1.9]	1.9 [1.8, 2.1]	<0.001
Race/Ethnicity
NH-Black	84 (12.9)	63 (17.7)	21 (7.1)	<0.001
NH-White	245 (37.6)	135 (37.9)	110 (37.3)	0.932
Hispanic	176 (27.0)	88 (24.7)	88 (29.8)	0.170
Unknown*	146 (22.4)	70 (19.7)	76 (25.8)	0.078
Hypertension	310 (47.6)	161 (45.2)	149 (50.5)	0.206
Diabetes	152 (23.3)	83 (23.3)	69 (23.4)	1
Dyslipidemia	251 (38.6)	128 (36)	123 (41.7)	0.156
PVD	116 (17.8)	58 (16.3)	58 (19.7)	0.31
CAD	118 (18.1)	43 (12.1)	75 (25.4)	<0.001
CKD	97 (15.8)	43 (13)	54 (19.1)	0.051
Stroke	39 (6.0)	16 (4.5)	23 (7.8)	0.109
Atrial Fibrillation	154 (23.7)	79 (22.2)	75 (25.5)	0.369
Creatinine, mg/dL	1. [0.8, 1.3]	0.9 [0.8, 1.2]	1.1 [1, 1.5]	<0.001
LDL-C, mg/dL	78 [57.8, 105.2]	83 [60, 110]	73 [56, 99]	0.024
Triglycerides, mg/dL	99 [73, 137]	102 [80, 137.2]	96 [67, 136]	0.177
Total Cholesterol, mg/dL	149 [124, 183]	161 [135.2, 192.5]	139.5 [120, 170.8]	<0.001
TAVR	517 (79.4)	276 (77.5)	241 (81.7)	0.226
Death	126 (19.4)	67 (18.8)	59 (20)	0.78

^*The NH-Asian race and individuals with unknown race/ethnicity have been grouped as “Unknown.”

Values are presented as n (%); mean ± SD; median [IQR].

BMI, body mass index; BSA, body surface area; CAD, coronary artery disease; CKD, chronic kidney disease; LDL-C, low-density lipoprotein cholesterol; NH, non-Hispanic; PVD, peripheral vascular disease; TAVR, transcatheter aortic valve replacement.

Aortic valve hemodynamics

The median AVA was 0.70 (0.59–0.83) cm², and the mean AV mean ΔP was 40 (29–47) mmHg. The mean AV Vmax was 4±0.72 m/s. Overall, 51.8% of the patients were classified as HG-AS, and 92.3% had tricuspid AV morphology. The median LVEF was 60 (52–65) %. When stratified by sex, women exhibited a significantly lower absolute AVA, but similar index AVA compared to men [0.7 (0.5–0.8) cm² vs 0.7 (0.6–0.8) cm², P=0.001). They also had smaller left ventricle outflow tract diameter [1.9 (1.8–2) cm vs 2 (2–2.1) cm, P<0.001], and higher LVEF (65% vs 60%, P<0.001). There were no significant differences in terms of AV mean ΔP, Vmax, tricuspid AV morphology, and the prevalence of HG-AS between both sexes. All echocardiographic variables classified by sex are listed in Table 2. There were no major significant differences in echocardiographic measurements among different race/ethnicity subgroups (Table S1).

Table 2.

Echocardiographic and Computed tomography parameters for AV characterization, by sex.

	Overall	Women	Men	P value
	N = 651	N = 356	N = 295
LVEF, %	60 [52, 65]	65 [55, 68]	60 [50, 65]	<0.001
EF>50%, n	521 (80)	300 (84.3)	221 (74.9)	0.004
AVA, cm2	0.7 [0.6, 0.8]	0.7 [0.5, 0.8]	0.7 [0.6, 0.8]	0.001
AVAi,	0.4 [0.3, 0.5]	0.4 [0.3, 0.5]	0.4 [0.3, 0.4]	0.216
AV mean PG, mmHg	40 [29.4, 47]	40.4 [30, 48]	37.7 [29, 45.2]	0.175
AV max PG, mmHg	64 [50.2, 81.3]	66.9 [51, 82]	63.3 [50, 80]	0.248
AV max velocity, cm/s	4 ± 0.7	4.1 ± 0.7	3.9 ± 0.7	0.069
LVOT diameter	2 [1.9, 2.1]	1.9 [1.8, 2]	2 [2, 2.1]	<0.001
Stroke volume, mL	62.5 [51.3, 74.3]	61 [49, 73]	65.6 [54.4, 75.9]	0.008
Stroke volume indexed, mL/m2	31.6 [22.2, 40.4]	31.9 [22.3, 41.6]	31.3 [22, 38.9]	0.284
IVS diameter	1.2 [1, 1.3]	1.1 [1, 1.3]	1.2 [1.1, 1.3]	0.008
AV Morphology
Tricuspid	605 (92.9)	337 (94.7)	268 (90.8)	0.082
AV Hemodynamics
High-gradient AS	337 (51.8)	195 (54.8)	142 (48.1)	0.108
Fibrotic-calcific volume, mm3/cm2	258 [188, 340]	230 [166, 311]	293 [217, 382]	<0.001
Fibrotic volume, mm3/cm2	126 [82, 193]	127 [79, 204]	125 [87, 184]	0.805
Calcific volume, mm3/cm2	111 [67, 177]	85[49, 132]	149 [101, 217]	<0.001
Fibro-calcific ratio	1.1 [0.6, 2.1]	1.5 [0.8, 2.8]	0.8 [0.5, 1.5]	<0.001
AV Annulus area, cm2	4.3 (1)	3.8 (0.8)	4.9 (1)	<0.001

Values are presented as n (%); mean ± SD; median [IQR].

AS, aortic stenosis; AV, aortic valve; AVA, aortic valve area; EF, ejection Fraction; IVS, interventricular septum; LVOT, left ventricle, outflow tract; PG, pressure gradient.

Advanced CTA-AV tissue composition by sex

Women exhibited significantly lower fibro-calcific volumes [230 (166–311) mm³/cm² vs. 293 (217–382) mm³/cm², P<0.001] and lower calcific volumes [85(49–132) mm³/cm² vs. 149 (101–217) mm³/cm², P<0.001] compared to men. Despite similar fibrotic tissue volume (P=0.805), women had a significantly higher fibro-calcific ratio than men (1.5 [IQR 0.8–2.8] vs. 0.83 [IQR 0.5–1.5], P<0.001). The fibrocalcific ratio reflects the relative tissue composition of the aortic valve. The higher values (>1) observed in women indicate a predominance of fibrotic over calcific tissue in the aortic stenosis phenotype in this population (Figure 2). Detailed CTA-AV characterization by sex is provided in Table 2. These differences remained statistically significant across each racial/ethnic subgroup (NH-Black, NH-White, Hispanic) when analyzed by sex (Table S2). At multivariable linear regression analysis, after adjustment, male sex was independently associated with higher fibro-calcific (β = 70.51, P < 0.0001) and calcific volumes (β = 72.91, P<0.0001), while a lower fibro-calcific ratio (β = −1.69, P<0.001) (Table S3).

Figure 2. Aortic Valve composition, by sex.

Values are presented are fibro-calcific volume (mm³/cm²), fibrotic volume(mm³/cm²), and calcific volume (mm³/cm²), (Panel A). Fibro-calcific ratio is also presented (Panel B).

Advanced CTA-AV tissue composition by race/ethnicity

Overall, NH-White patients had significantly higher fibro-calcific volumes than Hispanic and NH-Black patients [NH-White 266 (203–359) mm³/cm² vs Hispanic 231 (181–319) mm³/cm² vs NH-Black 246 [150–335, P=0.011] mm³/cm². NH-White also exhibited the highest fibrotic volume, while NH-Black had the lowest [138 (97–222) mm³/cm² vs. 111 (71–172) mm³/cm², respectively, P=0.001]. No significant differences were observed in terms of calcific volume (P=0.623) and fibro-calcific ratio (P=0.084) between the different race/ethnicity groups (Figure 3, Table S2). In the multivariable regression analysis, Hispanic ethnicity was independently associated with reduced fibro-calcific (β = −31.09, P=0.002) and fibrotic volumes (β = −23.56, P = 0.014), while non-Hispanic Black individuals demonstrated increased calcific volume (β = 21.45, P=0.046) and a lower fibro-calcific ratio (β = −1.41, P=0.034). No significant interactions between sex and race/ethnicity were observed. (Table S3).

Figure 3. Aortic Valve Tissue Composition by Race/Ethnicity*.

Values are presented are fibro-calcific volume (mm³/cm²), fibrotic volume(mm³/cm²), and calcific volume (mm³/cm²), (Panel A). Fibro-calcific ratio is also presented (Panel B).

When analyzing the AV tissue composition stratified by sex and race/ethnicity, NH-White women (n=145) had significantly higher fibro-calcific volume and higher fibrotic volume [256 (181–328)mm³/cm², P=0.002; and 145(101–236)mm³/cm², P<0.001; respectively] when compared to the other subgroups (Hispanic, n=88 and NH-Black, n=63). In contrast, NH-Black women had the lowest fibrotic volume [108 (71–173) mm³/cm², P<0.001] and fibro-calcific ratio [1.2 (0.6–1.8) mm³/cm², P=0.01], compared to NH-White and Hispanic women, Table 3. There were non-significant differences regarding the calcific valve tissue volume among women of diverse race/ethnicity. In addition, no differences in AV tissue composition were found when analyzing the male population by race/ethnicity, Table 3, Figure 4.

Table 3.

Advanced characterization of AV tissue composition by Race/Ethnicity in women and men.^*

Women
	Hispanic N = 88	NH-White N = 135	NH-Black N = 63	P value
Fibro-calcific volume, mm3/cm2	204 [156, 265]	256[181, 328]	216 [142, 323]	0.002
Fibrotic volume, mm3/cm2	114[63, 176]	145 [101, 236]	108[71, 173]	<0.001
Calcific volume, mm3/cm2	75 [41, 114]	88 [49, 129]	91 [65, 137]	0.083
Fibro-calcific ratio	1.4 [0.7, 3]	1.7 [1, 3.2]	1.2 [0.6, 1.8]	0.01
AV annulus area, cm2	3.5 [3.1, 4.1]	3.8 [3.3, 4.2]	3.7 [3.4, 4.1]	0.015
Men
	Hispanic N = 88	NH-White N = 110	NH-Black N = 21	P value
Fibro-calcific volume, mm3/cm2	287[207, 381]	295[221, 397]	314[223, 361]	0.698
Fibrotic volume, mm3/cm2	124[88, 187]	133[87, 203]	118[74, 170]	0.619
Calcific volume, mm3/cm2	139[98, 204]	146 [97, 220]	183 [133, 239]	0.264
Fibro-calcific ratio	0.9 [0.5, 1.7]	0.9 [0.5, 1.5]	0.7 [0.4, 0.9]	0.212
AV annulus area, cm2	4.6 [4.1, 5.1]	5.0 [4.3, 5.7]	5 [4.2, 6.1]	0.005

^*Individuals of NH-Asian race and those with unknown race/ethnicity were excluded from the analysis.

Values are presented as median [IQR].

AV, aortic valve

Figure 4. Aortic Valve Tissue characterization in women and men by Race/Ethnicity*.

AV tissue volumes and fibro-calcific ratio in women (Top) and men (Bottom).

Aortic Valve Characterization in High vs Low-gradient AS

Patients with HG-AS had significantly higher aortic fibro-calcific volume [295 (222–389) mm³/cm² vs. 219 (171–295) mm³/cm², P<0.001], and a higher calcific volume [111 (72–168) mm³/cm² vs. 62 (33–91) mm³/cm², P<0.001] compared to LG-AS. Despite no significant difference in fibrotic volume (P=0.099), LG-AS patients demonstrated a higher relative fibrotic composition compared to HG-AS, as expressed by a fibrocalcific ratio >1 vs <1 in HG-AS (1.4 vs 0.9, P<0.001) (Table 4, Figure 5). Advanced tissue quantification stratified by sex and hemodynamic phenotype is detailed in Tables S4 and S5. In women with HG-AS, both the fibro-calcific volume [260 (186, 341) mm³/cm² vs. 203(148, 266) mm³/cm², P<0.001] and the calcific volume [111(72, 168) mm³/cm² vs. 62 (33, 91) mm³/cm², P<0.001] were significantly higher compared to women with LG-AS. Although there was no significant difference in fibrotic volumes (P=0.715), the fibro-calcific ratio was higher in women with LG-AS (1.9 vs. 1.2, P<0.001)(Table S4).

Table 4.

Echocardiography and advanced cardiac computed tomography characteristics, stratified by AS hemodynamic phenotype.

	Low-gradient N = 314	High-gradient N = 337	P value
Ejection Fraction, %	60 [45, 65]	65 [55, 65]	<0.001
EF>50%	222 (70.7)	299 (88.7)	<0.001
AVA, cm2	0.8 [0.6, 0.9]	0.7 [0.5, 0.8]	<0.001
AVAi,	0.4 [0.4, 0.5]	0.4 [0.3, 0.4]	<0.001
AV mean PG, mmHg	29 [24, 33]	47 [43, 54.9]	<0.001
AV max PG, mmHg	50 [42, 57.8]	80 [71, 92]	<0.001
AV max velocity, m/s	3.4 ± 0.4	4.5 ± 0.6	<0.001
LVOT diameter, mm	2 [1.8, 2.1]	2 [1.9, 2]	0.553
Stroke Volume, mL	58 [47.3, 69]	67 [57.5, 80]	<0.001
Stroke Volume indexed, mL/m2	30.2 [22, 36.8]	34.0 [22.3, 42.1]	0.005
IVS diameter, mm	1.1 [1, 1.3]	1.2 [1.1, 1.4]	<0.001
Valve Morphology
Tricuspid	295 (93.9)	310 (92)	0.411
Bicuspid	19 (6.1)	27 (8)
TAVR	225 (71.7)	268 (79.5)	0.025
Death	73 (23.2)	53 (15.7)	0.02
Advanced CT AV characterization
Fibro-calcific volume, mm/cm2	218.9 [171, 295]	294.9 [222.1, 388.8]	<0.001
Fibrotic volume, mm/cm2	117.4 [80.2, 189.4]	136.6 [84, 197]	0.099
Calcific volume, mm/cm2	87.6 [52.9, 124.2]	147.7 [86.5, 211.7]	<0.001
Fibro-calcific ratio	1.4 [0.8, 2.6]	0.9 [0.5, 1.6]	<0.001
Annulus area, cm2	4.3 [3.6, 5.1]	4.1 [3.5, 4.8]	0.044

Values are presented as n (%); mean ± SD; median [IQR].

AV, aortic valve; AVA, aortic valve area; CT, computed tomography; LVOT, left ventricular outflow tract; PG, pressure gradient; TAVR, transcatheter aortic valve replacement.

Figure 5. Characteristics of aortic valve tissue composition in low-gradient and high-gradient aortic valve stenosis in overall population.

Values are presented are fibro-calcific volume (mm³/cm²), fibrotic volume(mm³/cm²), and calcific volume (mm³/cm²), (Panel A). Fibro-calcific ratio is also presented (Panel B).

LG, low-gradient; HG-high-gradient.

Men with HG-AS exhibited higher fibro-calcific [335 (266, 429) mm³/cm² vs 245 (189, 322) mm³/cm², P<0.001], fibrotic [142 (97, 190)mm³/cm² vs 117 (83, 176) mm³/cm², P=0.031)], and calcific volumes [189 (140, 248) mm³/cm² vs 114 (82, 164) mm³/cm², (P<0.001)] compared to men with LG-AS. The fibro-calcific ratio was lower in men with HG-AS vs LG-AS (0.7 vs. 1, P<0.001; Table S4) but both subgroups exhibited a ratio <1 in indicating the relative predominance of calcific over fibrotic tissue.

DISCUSSION

This study represents one of the largest cohorts to date evaluating the AV phenotype by CTA, using a racially and ethnically diverse US cohort of patients with symptomatic AS referred for TAVR evaluation. The main findings are that: 1) sex appears to be the primary driver of the AV phenotype with women having lower calcific volumes and higher fibrocalcific ratio than men for similar hemodynamic severity; 2) there are subtle race-ethnicity differences in AV tissue composition among women (but not in men), with NH-Black women having the lowest fibrotic volume and fibrocalcific ratio; and 3) patients with LG-AS have a more relative fibrotic AS phenotype with higher fibrocalcific ratio and lower calcific and fibrocalcific volumes than those with HG-AS.

From the landmark studies of AV calcium score using non-contrast CT^4,10, we know that women present with lower calcium volumes than men for the same hemodynamic severity of AS. Accordingly, sex-specific cutoffs are currently recommended (1,300 AU in women and 2,000 AU in men)⁵. The ability to quantify AV leaflet fibrosis using dedicated valve segmentation software, which cannot be done with non-contrast CT, has recently been shown feasible and provides important information beyond AV calcium score, especially in women, bicuspid AV, and LG-AS^11,14,15,18. Cartlidge et al studied 164 patients (78% male) with diverse AS severity and found that fibrocalcific volume index (by CTA) increased with increasing hemodynamic severity of AS and, in a small subset of 26 patients, correlated strongly with ex-vivo valve weight and semiquantitative histological assessment. Fibrocalcific volume had a better correlation with AV Vmax than the Agatston volume, particularly in female individuals (r=0.72 and r=0.38, respectively)¹¹. This was the first proof-of-concept study with an advanced semi-automatic AV tissue quantification methodology. Recently, Lembo et al validated this approach in 136 patients with diverse AS severity including 41 valves with ex-vivo histological correlation. They also showed that AS progression could also be accurately tracked with CTA quantification as the increase in fibrocalcific volumes correlated with mean gradient progression at 1 year¹⁸. Beyond sex differences, the AV morphology has been also implicated with the valve tissue phenotype. Patel et al used the same advanced AV segmentation software to evaluate 140 (62% male) propensity-matched patients undergoing TAVR evaluation from a single UK center with tricuspid (n=70) and bicuspid AS (n=70). They found that women had greater fibrotic volumes than men in bicuspid but not on tricuspid valves [224 (181–307) vs. 169 (109–247) mm3/cm2, P=0.042]. Men had greater calcific volumes, and women had a greater fibrocalcific ratio (tricuspid 1.86 vs. 0.86, P=0.001 and bicuspid 1.78 vs. 0.74, P=0.001) regardless of the valve morphology¹⁵. We expand those findings to a larger, more diverse, and US-based cohort of patients with severe AS including those who did not ultimately undergo TAVR (Table 1) in order to avoid selection bias. In this predominantly female (55%) and tricuspid morphology (92%) population, we also found significantly lower fibrocalcific and calcific volumes in women compared to men, together with a higher fibrocalcific ratio suggesting a more relative fibrotic phenotype of AS. Similarly, to Patel et al, we did not find any differences in the indexed fibrotic volume between sexes without considering morphology since our cohort was mostly comprised of tricuspid valves. At present, there is no effective medical therapy to prevent AS or to slow its progression, largely because the understanding of the disease process is incomplete. Inflammation and fibrosis appear to play a major role in the pathophysiology of the disease but ongoing trials are focusing solely on the calcific pathway^21,22.

The influence of race/ethnicity in AS has not been properly studied with CTA since the above-mentioned reports were done on predominantly White European cohorts and the authors did not report race/ethnicity of the patient cohorts. Among the few studies published to date reporting on the race/ethnicity implications it has been shown that severe AS is less prevalent among NH-Black, as well as the frequency of bicuspid morphology^23,24. A recent study examined the rates of AV calcification using non-contrast CT in 2,283 elderly NH-Black and NH-White patients from the Atherosclerosis Risk in Communities Study. The authors found that the prevalence of AV calcification was highest among White males, followed by Black males, White females, and Black females individuals²⁵. As far as we know, our cohort is the first of its kind allowing analysis of racial/ethnic differences on AS phenotype due to its unique composition including commonly underrepresented race/ethnicities in TAVR trials as NH-Black and Hispanics patients²⁶. We found that the overall cohort sex differences in AV tissue composition persisted when stratified by race/ethnicity subgroups, suggesting that sex could be a stronger driver of the phenotype. Furthermore, when male patients were stratified by race/ethnicity, we did not find any differences in terms of tissue composition. However, subtle differences were found within the female subgroup. Despite all female patients presenting with a predominantly relative fibrotic valve phenotype (fibro-calcific ratio >1), NH-White patients had the highest fibrotic volume and fibro-calcific ratio, followed by Hispanic and NH-Black patients. Although the descriptive analysis showed no significant differences in calcific volume or fibro-calcific ratio by race/ethnicity, multivariable analysis revealed a significant association between non-Hispanic Black race and higher calcific volume (β = 21.45, P=0.046) after adjusting for established risk factors. This discrepancy suggests that underlying race/ethnic-specific pathophysiological mechanisms may influence aortic valve calcification independent of traditional cardiovascular risk factors. The pathophysiological mechanisms underlying race/ethnicity differences in the fibrotic component remain poorly understood. Further research is needed to confirm these findings, elucidate the underlying mechanisms, and determine their clinical implications for risk stratification and therapeutic approaches. Notably, the widely accepted AV calcium score cutoffs were derived from exclusively Caucasian cohorts, underscoring the need for validation in diverse cohorts. When examining the association of the hemodynamic phenotype with the valve composition, we found that patients with HG-AS had higher fibrocalcific and calcific volumes with a lower fibrocalcific ratio compared to those with LG-AS. This suggests that HG-AS is characterized by a more calcific phenotype (fibrocalcific ratio<1) whereas LG-AS exhibits a predominantly fibrotic tissue composition. Furthermore, when analyzing the sex implications, women with LG-AS exhibited the highest relative fibrotic tissue composition (fibrocalcific ratio=1.89) while men with HG-AS showed the highest calcific phenotype (fibrocalcific ratio= 0.74). Our results are in line with the previous report from Grodecki et al, who analyzed a multicenter Polish TAVR cohort of 447 patients (73% HG-AS) and compared it to 224 matched controls with normal AV¹⁴. They also found a higher proportion of non-calcific (fibrotic) tissue composition in patients with LF-AS compared to HG-AS independently of the LVEF (reduced vs preserved). However, the authors did not report on sex differences concerning the hemodynamic phenotype. Interestingly, they demonstrated that incorporating CTA tissue measurements enabled more accurate reclassification of patients, surpassing the accuracy of traditional AV calcium scoring. Furthermore, the accuracy of AV calcium score to distinguish between true severe and pseudo-severe/moderate AS has been challenged in a recent international multicenter cohort, which found an area under the receiver operating curve of 0.508 for men and 0.524 for women.⁷.

Our work emphasizes the important role of valve tissue fibrosis in LG-AS in a large and diverse US cohort. This could potentially have direct clinical implications since the AV calcium score is currently recommended along with dobutamine stress-echocardiography on both American and European guidelines in the setting of LG-AS^1,2.

Recently, the diagnostic and prognostic value of CT AV tissue quantification was demonstrated using a multicenter international AS cohort (n=1521)²⁷. The authors were able to derive sex-specific calcific tissue volume threshold for the diagnosis of severe AS using a validation cohort, that were proven as accurate as calcium score were applied to the derivation cohort (C-statistic>0.90 for both sexes). In addition, the CT AV calcific tissue volume was independently associated with aortic valve replacement or all-cause death. Contrast-enhanced CT quantification of the aortic valve has proven to be a rapid and reproducible¹⁸ process, to allow the differentiation of AS phenotypes and to quantify aortic stenosis severity, even in LG-AS patients¹⁴. Fibrocalcific volume measurements offer significant potential for early detection of AS and could facilitate the timely initiation of novel therapeutic interventions aimed at slowing or arresting disease progression, thereby preventing irreversible structural and functional damage to the aortic valve^28,29. Early detection might facilitate the implementation of effective therapeutic strategies to slow or stop the progression of CAVD towards possibly irreversible damage to the aortic valve.

On the other hand, contrast-enhanced CT could play a more prominent role in the comprehensive evaluation of aortic stenosis (AS), extending beyond its current use in pre-TAVR planning. Given the absence of a gold standard for diagnosing LF-LG severe AS, quantification of fibrotic tissue may enhance diagnostic accuracy and improve risk stratification in patients with discordant findings¹⁴_.However, further research is required before this approach can be widely adopted. This includes the need for deriving and validating thresholds and exploring the impact of different CT platforms and image parameters.

LIMITATIONS

Despite the analysis being conducted in a large, balanced, and race/ethnic diverse cohort of patients undergoing CTA-TAVR evaluation for AS, we had to exclude 172 patients due to suboptimal image quality, which represents 20% of the patient population. This highlights the importance of optimal image quality for performing advanced AV quantification in a similar way to coronary CTA plaque quantification³⁰. Although the cohort was diverse, the analysis was constrained by the exceptionally small representation of self-identified NH-Asians, who accounted for only 1.2% of the population. Due to the insufficient sample size, these individuals had to be excluded from the primary analysis precluding any conclusions to be drawn regarding this subgroup. Additionally, a substantial portion of the cohort (21.2%) did not provide self-identified race/ethnicity data due to patient preference and were therefore excluded from direct race/ethnicity comparisons, which limited the overall completeness of the dataset. Despite being one of the largest cohorts to date, direct race and ethnicity comparisons may be limited by the relatively small number of patients in specific subgroups, such as non-Hispanic Black men (n=21).

In addition, all patients came from a single health network based in New York, NY so further data will be needed to validate the generalizability of the current findings. We did not have sufficient numbers of patients with bicuspid AV to explore the implications of valve morphology in our cohort. We decided not to report or analyze the conventional non-contrast CT calcium score as it was not available in a significant proportion of the TAVR cohort, due to not being included in the institutional protocol until the 2019 guideline recommendation¹⁷. In addition, our cohort had a higher-than-expected proportion of patients with LG-AS (mostly preserved LVEF) and the study did not include a systematic evaluation with dobutamine stress echocardiography or calcium score in those patients. Finally, we did not have histological validation for the CTA-derived valve tissue volumes, although this has been successfully demonstrated in previous studies^11,18., and more recently validated in a prospective heart transplant and AV surgical replacement cohort with an excellent correlation between CTA and histology (r=0.92; P<0.001)³¹.

CONCLUSIONS

In a diverse cohort of patients with symptomatic AS undergoing TAVR evaluation, there are important differences in the AV tissue composition by CTA. Women and patients with LG-AS exhibit higher relative fibrotic valve composition compared to men and those with HG-AS, which had more calcific tissue. There were no significant differences among men of different races/ethnicities but subtle differences in the fibrotic composition were found among NH-White compared to NH-Black and Hispanic women. These findings expand our understanding of the biological differences in AS and, may influence diagnostic algorithms as well as enable tracking of disease progression and response to novel therapies in the future.

Supplementary Material

Supplemental Publication Material

Tables S1–S5

CLINICAL PERSPECTIVE.

Contrast-enhanced computed tomography angiography (CTA) is the only non-invasive tool that assesses both calcific and fibrotic components of aortic valve tissue in aortic stenosis (AS). In this study of patients referred for transcatheter aortic valve replacement (TAVR), women—especially non-Hispanic White women—demonstrated a more fibrotic valve phenotype with higher fibrocalcific ratios compared to men and other racial/ethnic groups. Additionally, low-gradient AS was associated with a predominance of fibrotic tissue, while high-gradient AS showed more calcific degeneration. These findings highlight important phenotypic differences in AS tissue composition across sex, race/ethnicity, and hemodynamic profiles, which could help overcome diagnostic challenges, track disease progression, and inform future research on therapeutic strategies.

NON-STANDARD ABBREVIATIONS AND ACRONYMS

AS

aortic stenosis

AV

aortic valve

AVA

aortic valve area

CAD

coronary artery disease

CT

computed tomography

CTA

computed tomography angiography

HG-AS

high-gradient aortic stenosis

LG-AS

low-gradient aortic stenosis

LVEF

left ventricular outflow tract

NH

non-Hispanic

PG

pressure gradient

TAVR

transcatheter aortic valve replacement