Skip main navigation

Preoperative Computed Tomography Angiography Reveals Leaflet-Specific Calcification and Excursion Patterns in Aortic Stenosis

Originally published Cardiovascular Imaging. 2021;14:1122–1132



Computed tomography–based evaluation of aortic stenosis (AS) by calcium scoring does not consider interleaflet differences in leaflet characteristics. Here, we sought to examine the functional implications of these differences.


We retrospectively reviewed the computed tomography angiograms of 200 male patients with degenerative calcific AS undergoing transcatheter aortic valve replacement and 20 male patients with normal aortic valves. We compared the computed tomography angiography (CTA)-derived aortic valve leaflet calcification load (AVLCCTA), appearance, and systolic leaflet excursion (LEsys) of individual leaflets. We performed computer simulations of normal valves to investigate how interleaflet differences in LEsys affect aortic valve area. We used linear regression to identify predictors of leaflet-specific calcification in patients with AS.


In patients with AS, the noncoronary cusp (NCC) carried the greatest AVLCCTA (365.9 [237.3–595.4] Agatston unit), compared to the left coronary cusp (LCC, 278.5 [169.2–478.8] Agatston unit) and the right coronary cusp (RCC, 240.6 [137.3–439.0] Agatston unit; both P<0.001). However, LCC conferred the least LEsys (42.8° [38.8°–49.0°]) compared to NCC (44.8° [41.1°–49.78°], P=0.001) and RCC (47.7° [42.0°–52.3°], P<0.001) and was more often characterized as predominantly thickened (23.5%) compared to NCC (12.5%) and RCC (16.5%). Computer simulations of normal valves revealed greater reductions in aortic valve area following closures of NCC (−32.2 [−38.4 to −25.8]%) and RCC (−35.7 [−40.2 to −32.9]%) than LCC (−24.5 [−28.5 to −18.3]%; both P<0.001). By linear regression, the AVLCCTA of NCC and RCC, but not LCC, predicted LEsys (both P<0.001) in patients with AS. Both ostial occlusion and ostial height of the right coronary artery predicted AVLCCTA, RCC (P=0.005 and P=0.001).


In male patients, the AVLCCTA of NCC and RCC contribute more to AS than that of LCC. LCC’s propensity for noncalcific leaflet thickening and worse LEsys, however, should not be underestimated when using calcium scores to assess AS severity.


There has been a strong push to use computed tomography to adjudicate the severity of aortic stenosis (AS) based on calcium scoring of the aortic valve. Increasing evidence suggests that patients with clinically significant AS not only have varying levels of aortic valve calcification load but also significant interleaflet differences in calcification load, the functional implication of which remains unclear. The results of our study suggest that the calcification loads of the noncoronary and right coronary cusps contribute significantly more to AS than that of the left coronary cusp due to their greater values, predictability for reduced leaflet motion, and potential impact on aortic valve area. However, compared to other leaflets, the left coronary cusp is associated with the least systolic excursion and is more often characterized by noncalcific leaflet thickening as the predominant disease feature. These results raise the possibility for subthreshold levels of aortic valve calcification load to underestimate AS severity, especially in cases where severe noncalcific thickening of the left coronary cusp is strongly suspected. Further development of techniques to quantify both valvular calcification and noncalcific leaflet thickening/fibrosis will be helpful to overcome this concern and provide a more comprehensive assessment of AS than calcium scoring alone.

Computed tomography (CT)–based scoring of aortic valve calcification (AVC) is gaining popularity as an alternative to echocardiography for adjudicating the severity of aortic stenosis (AS) and guiding the timing of valve replacement.1 The utility of assessing AVC is well established by previous studies demonstrating strong correlations between CT-determined AVC load (AVCCT) and AS patient survival.2,3 Although AVCCT load as a single measurement confers simplicity for clinical use, it masks distinct contributions from individual valve leaflets. The importance of examining valvular calcification in a leaflet-specific manner is further supported by the association of regional leaflet calcification with transcatheter aortic valve replacement (TAVR)–related complications (eg, advanced heart block).4

Although previous studies have noted significant differences in the calcification loads of individual valve leaflets, little is known about their functional contributions to AS.4,5 In this study, we aimed to (1) retrospectively analyze pre-TAVR CT angiograms of patients with AS to characterize the interleaflet differences in calcification load, appearance, systolic excursion, and systolic excursion-calcification load relationships, (2) conduct computer simulations to examine how interleaflet differences in systolic excursion affect aortic valve area (AVA), and (3) identify factors that may affect calcification in a leaflet-specific manner.


All supporting data are available upon reasonable request to the corresponding author.

Study Population

This cross-sectional study involving patients from the Veterans Affairs Palo Alto Health Care System was conducted with prior approval from the Stanford-Veterans Affairs Institutional Review Board. Informed consent was waived by the Institutional Review Board. Two hundred consecutive male patients with degenerative (tricuspid) calcific AS who underwent retrospective electrocardiography–gated pre-TAVR computed tomography angiography (CTA) from July 2011 to October 2019 were enrolled (Figure S1 for study cohort flow chart). The only female patient who underwent TAVR during this period was not included due to known sex differences in AVCCT,6 which could confound data analysis. Twenty consecutive male patients with normal aortic valves on echocardiography who underwent clinically indicated retrospective electrocardiography-gated coronary CTA with maximum tube current during both systole and diastole between December 2017 and October 2019 were enrolled as controls. Patients with bicuspid aortic valves, prosthetic aortic valves, inadequate image quality, insufficient image data, and history of endocarditis, rheumatic valve disease, and mediastinal radiation were excluded.


Transthoracic echocardiography was performed per clinical guidelines.7 All examinations involved the inspection of the aortic valve for morphology, leaflet characteristics, and leaflet mobility. For patients with AS, measurements of AVA, mean gradient, peak jet velocity, left ventricular ejection fraction, and stroke volume index were obtained to assess AS disease severity as described.1 For control patients, normal tricuspid aortic valves were confirmed by noting thin leaflets with normal mobility.

Pre-TAVR CT Angiography

Pre-TAVR CTA was performed using either the GE Lightspeed VCT (GE Healthcare, Chicago, IL) or the Siemens SOMATOM Force CT System (Siemens Healthineers AG, Munich, Germany). Retrospective electrocardiography-gated image acquisition was performed through the heart after intravenous administration of 120 to 150 mL of iohexol (Omnipaque 350, GE Healthcare) at 4 to 5 mL/s for 30 seconds followed by 80 to 100 mL of saline at the same rate as the contrast. This was immediately followed by a standard-pitch nongated (GE) or high-pitch electrocardiography-triggered (Siemens) acquisition through the chest, abdomen, and pelvis. Either a fixed x-ray tube energy of 120 kilovoltage peak (kVp; GE) or an automated x-ray tube energy selection tool (70–120 kVp; Siemens, CARE kV) was used. Tube current modulation was performed from 30% to 70% of the R-wave to R-wave interval. Images were reconstructed with a slice thickness of either 0.625 mm (GE) or 0.750 mm (Siemens), with further reconstruction to 2.5 mm for calcium scoring. Retrospective electrocardiography-gated coronary CTA of control patients was similarly performed as the electrocardiography-gated portion of the TAVR protocol except with a triphasic contrast protocol.

CTA Image Analyses

The Aquarius iNtuition Edition Ver.4.4.13.P3A viewer (TeraRecon, Foster City, CA) was used to analyze the CTA images of AS and control cohorts and assess the calcification loads, leaflet scores, and systolic excursions of individual valve leaflets (Figure 1, yellow path).

Figure 1.

Figure 1. Schematic of study workflow to elucidate leaflet-specific contributions to aortic stenosis (AS). Yellow path: Pre–transcatheter aortic valve replacement (TAVR) computed tomography angiography (CTA) images of patients with AS or coronary CTA images of control (CNTRL) patients were assessed for aortic valve calcification load, aortic valve leaflet calcification load, and systolic leaflet excursion. Red path: multivariable linear regression analyses were performed for each leaflet to assess the leaflet excursion-calcification load relationship and the predictors of leaflet calcification. Multinomial logistic regression was used to assess the dependence of visual leaflet score on individual leaflets. Blue path: Coronary CTAs of control patients were processed in sequential steps to enable the construction of properly oriented/aligned computational aortic valve models in different cardiac phases. Displacement fields were mapped for individual leaflets to comprehensively assess systolic motion. Computer simulations were performed to assess leaflet-specific contributions to normal valvular function by manipulating leaflet closure while computing the aortic valve area. LCC indicates left coronary cusp; NCC, noncoronary cusp; and RCC, right coronary cusp.

Calcium scoring of the aortic valve and individual leaflets was performed in midsystole (20%–30% R-wave to R-wave) as previously described.8 Briefly, a 1-cm2 region of interest was first placed in the ascending aorta to obtain the mean Hounsfield unit (HUaorta) and SD. A calcium threshold of HUaorta+2 SD was set before the calcified regions within each leaflet were quantified in terms of Agatston units and summed to obtain the CTA-derived aortic valve leaflet calcification load (AVLCCTA) of that leaflet. The summation of AVLCCTA of the noncoronary (NCC), right coronary (RCC), and left coronary (LCC) cusps then resulted in the CTA-derived aortic valve calcification load (AVCCTA). The fractional AVLCCTA of each leaflet was calculated as the quotient of AVLCCTA of that leaflet divided by AVCCTA.

The leaflet appearance was assessed in diastole in two views using a visual leaflet score (VLS): (1) a short-axis maximal intensity projection of the aortic valve (with a slab thickness large enough to cover all the aortic valve calcium based on the coronal left ventricular outflow tract view) showing the distribution of leaflet calcification and (2) an oblique view orthogonal to the annular plane and cutting through the midline and opposing commissure of each leaflet showing leaflet thickness. A VLS score of one to four (VLS-1 to VLS-4) was assigned for a normal thin leaflet, a predominantly thickened leaflet (<1/3 area with calcification), a leaflet of significant mixed features of calcification (between 1/3 and 2/3 area with calcification) and leaflet thickening, and a predominantly calcified leaflet (>2/3 area with calcification), respectively.

The systolic leaflet excursion (LEsys) for each leaflet was assessed in the same plane as for VLS scoring and obtained by measuring the angle between the annular plane and the longitudinal length of the leaflet body during maximal valve opening (20%–30% R-wave to R-wave).

The ostial heights (OH) of the right coronary artery (RCA) and the left main coronary artery measured from the annular plane and the sinus of Valsalva diameter at each cusp measured by the sinus-to-commissure method were assessed in diastole as previously described.9

Computer Simulation of the Impact of Individual Leaflet Excursions on AVA

Computer simulations were performed to examine the effects of individual leaflet excursions on AVA using a customized workflow (Figure 1, blue path). A more detailed description of the simulation method can be found in Supplemental Methods. Briefly, for each control patient, we used the SimVascular10 and MeshMixer (Autodesk, San Rafael, CA) software to create a three-dimensional valve model in two cardiac phases from the mid-systolic and diastolic CTA images. We used Coherent Point Drift algorithm11 to map the corresponding valve surfaces in the 2 phases to obtain a displacement field, from which the maximal systolic displacement within each leaflet was derived. We then assessed the contribution of each leaflet to AVA by initializing the valve model in midsystole and computing the AVA as we incrementally closed one leaflet at a time.

Statistical Analyses

Stata/SE Version 15 (StataCorp LLC, College Station, TX) was used for all statistical analyses except for sample size calculations which were performed with G*Power This study was powered to (1) detect interleaflet differences in both AVLCCTA and LEsys and identify their predictors in patients with AS and (2) detect interleaflet differences in LEsys in control patients (Supplemental Methods for sample size calculations). Data were presented as mean±SD for normally distributed variables, median (interquartile range) for non-normally distributed variables, and frequency/percentage for nominal variables. All continuous variables were tested for normality using the Shapiro-Wilk test. Student t test, Wilcoxon rank-sum test, and Fisher exact test were used to compare the means of normally distributed variables, the mean ranks of non-normally distributed variables, and the frequencies of nominal variables, respectively, between 2 independent groups. For comparison of multiple related groups, one-way repeated measures ANOVA followed by pairwise paired t tests were used to compare the means of normally distributed variables, whereas the Friedman test followed by pairwise Wilcoxon signed-rank tests were used to compare the mean ranks of non-normally distributed variables.

Multivariable linear regression analyses were performed to assess the LEsys-AVLCCTA relationship of each leaflet and to identify predictors of leaflet-specific calcification (Figure 1, red path). All regression analyses were performed after square root transformation of the dependent variables to satisfy the linear regression model assumptions/conditions (Figures S2 and S3). Method for selecting adjustment variables and handling outliers can be found in Supplemental Methods (also Tables S1 through S3). Multinomial logistic regression was used to assess the dependence of VLS on individual leaflets (Figure 1, red path).

In general, statistical significance was defined as P<0.05 and adjusted as P<0.05/n for n group comparisons (Bonferroni correction).


Clinical Profile of Study Population

Our study population consisted of 200 male patients with degenerative tricuspid AS undergoing TAVR evaluation and 20 male control patients with normal aortic valves (Table 1). The median age of the AS cohort was older than that of the control cohort (78.0 [71.0–84.0] years versus 59.5 [54.3–64.8] years). A large majority (80.5%) of the patients with AS had high-gradient severe AS. Nearly half (40.5%) had ≥3 clinical risk factors for AS. Approximately three-quarters (77.0%) had concurrent coronary artery disease and about one-quarter (22.5%) with left ventricular systolic dysfunction. One-quarter (25.5%) had coronary artery bypass grafting, compared with 15.0% with ostial chronic total/subtotal occlusion (CTO/STO) of RCA, 2.0% with ostial CTO/STO of left main coronary artery, and 11.5% with small nondominant RCA. The AS and control cohorts were similar in terms of risk factors/comorbidities, despite significantly greater AVCCTA in the former (912.4 [590.4–1453.3] Agatston unit versus 17.2 [0.95–41.2] AU; P<0.001).

Table 1. Clinical Characteristics of Study Population

Clinical variableAS cohort (n=200)Control cohort (n=20)P value
Basic demographics
 Age, y78.0 (71.0–84.0)59.5 (54.3–64.8)<0.001*
 Male sex, n (%)200 (100)20 (100.0)N/A
 Body mass index, kg/m227.8 (25.1–31.9)30.4 (25.4–37.0)0.138
Clinical risk factors, n (%)
 Hypertension166 (83.0)16 (80.0)0.757
 Dyslipidemia160 (80.0)13 (60.0)0.149
 Diabetes89 (44.5)7 (35.0)0.483
 Chronic kidney disease38 (19.0)2 (10.0)0.542
 Smoking21 (10.5)3 (15.0)0.465
 ≥3 above risk factors81 (40.5)11 (55.0)0.239
Comorbidities/associated conditions, n (%)
 Coronary artery disease154 (77.0)15 (75.0)0.777
 Ostial RCA CTO/STO30 (15.0)0 (0)0.084
 Ostial LMCA CTO/STO4 (2.0)0 (0)1.000
 sNDRCA23 (11.5)0 (0)0.239
 History of CABG51 (25.5)1 (5.0)0.050
 Left ventricular systolic dysfunction45 (22.5)8 (40.0)0.100
AS severity
 High-gradient severe AS, n (%)161 (80.5)0 (0)<0.001*
 Low-gradient severe AS, n (%)34 (17.0)0 (0)0.049*
 Moderate AS, n (%)5 (2.5)0 (0)1.000
 Aortic valve area, cm20.80 (0.7–0.9)N/AN/A
 Aortic valve mean gradient, mm Hg45.6 (41.0–54.3)N/AN/A
 AVCCTA (AU)912.4 (590.4–1453.3)17.2 (0.95–41.2)<0.001*

AS indicates aortic stenosis; AU, Agatston unit; AVCCTA, computed tomography angiography-derived aortic valve calcification load; CABG, coronary artery bypass grafting; CTO/STO, chronic total/subtotal occlusion; LMCA, left main coronary artery; N/A, not applicable; RCA, right coronary artery; and sNDRCA, small nondominant RCA.

* Significant difference (P<0.05) between patients with AS and control patients.

Leaflet-Specific Calcification Loads

The AVCCTA of our patients with AS was highly heterogeneous and for the most part continuous from 116.0 AU to 3219 AU (Figure 2A) in a right-skewed distribution (Figure 2B). The AVLCCTA varied greatly among the leaflets such that AVLCCTA, NCC (365.9 [237.3–595.4] AU) was significantly greater than AVLCCTA, RCC (278.5 [169.2–478.8] AU) and AVLCCTA, LCC (240.6 [137.3–439.0] AU; P<0.001 comparing all leaflets; P<0.001 comparing NCC and RCC, P=0.055 comparing RCC and LCC, P<0.001 comparing NCC and LCC; Figure 2C). More than two-thirds (71.0%) of the patients with AS had fractional AVLCCTA, NCC >0.333 (ie, NCC-dominant), the expected fraction if the entire AVCCTA was evenly distributed among the 3 leaflets (Figure 2D). The AVLCCTA of all leaflets in control patients were trivial, with no significant differences among them (NCC 7.9 [0.15–12.2] AU, RCC 3.5 [0–13.0] AU, LCC 2.8 [0–13.6] AU; P=0.270).

Figure 2.

Figure 2. Pre–transcatheter aortic valve replacement computed tomography angiography (CTA) assessment of valvular and leaflet calcification in patients with aortic stenosis (AS). A, The CTA–derived aortic valve calcification load (AVCCTA) of patients with AS is displayed for patients ranked from low to high AVCCTA. B, The histogram of AVCCTA is skewed to the right, with median (red line) and mean (blue line) values indicated. C, Tukey box plots of the CTA-derived aortic valve leaflet calcification load (AVLCCTA; not transformed) of different leaflets are shown. Statistical significance is P<0.05 and P<0.0167 for multiple and pairwise comparisons, respectively. D, A stacked column plot of the fractional AVLCCTA (fAVLCCTA) for noncoronary cusp (NCC; black), right coronary cusp (RCC; light gray), and left coronary cusp (LCC; dark gray) is displayed for patients with AS ranked from low to high fAVLCCTA, NCC. AU indicates Agatston unit.

Interleaflet Differences in VLSs

All leaflets in AS and control patients had abnormal (>VLS-1) and normal (VLS-1) appearances by VLS scores, respectively (Figure 3A and 3B for conventions). In patients with AS, compared with NCC, LCC was descriptively associated with greater proportions of VLS-2 (thickened; 23.5% versus 12.5%) and VLS-3 (mixed; 25.0% versus 18.5%), and a lesser proportion of VLS-4 (calcified; 51.5% versus 69%; Figure 3C). The proportions of visual scores for RCC were intermediate to those of NCC and LCC (16.5%, 20.5%, and 63% with VLS-2, VLS-3, and VLS-4, respectively). Multinomial logistic regression further found LCC to differ significantly from NCC in association with VLS-2 (thickened; relative risk ratio=3.071, P<0.001) and VLS-3 (mixed; relative risk ratio=1.964, P=0.014) relative to VLS-4 (calcified; baseline outcome; Table S4). LCC differed significantly from RCC in association with only VLS-2 (relative risk ratio=1.983, P=0.016) but not VLS-3 (relative risk ratio=1.575, P=0.068) relative to VLS-4. There was no significant difference between RCC and NCC in association with either VLS-2 (P=0.169) or VLS-3 (P=0.420) relative to VLS-4.

Figure 3.

Figure 3. Computed tomography angiography–based visual assessment of leaflet calcification and thickening. A, A schematic shows how the visual leaflet score (VLS) is evaluated for each leaflet based on the area coverage of calcification (white) in the short-axis view and the presence of leaflet thickening (black) in the oblique view. B, Representative diastolic multiplanar reformation images (MPR; at the level of leaflet apposition) and maximal intensity projection (MIP) images of the aortic valve are displayed for 2 patients with aortic stenosis (AS) and one control (CNTRL) patient. The corresponding oblique cuts through the midlines of individual leaflets (noncoronary cusp [NCC]/right coronary cusp [RCC]/left coronary cusp [LCC]; arrows) are displayed with VLS numerically indicated. C, Stacked column plots of VLS proportions are displayed for individual leaflets.

Leaflet-Specific Excursion Patterns

We observed small interleaflet differences in the LEsys of our patients with AS (P<0.001) such that LEsys, RCC was the greatest (47.7° [42.0°–52.3°]), followed by LEsys, NCC (44.8° [41.1°–49.78°]), and LEsys, LCC (42.8° [38.8°–49.0°]; P=0.027 comparing NCC and RCC, P<0.001 comparing RCC and LCC, P=0.001 comparing NCC and LCC; Figure 4A, 4B). This pattern of interleaflet differences was similar but even more pronounced in control patients: LEsys, RCC 84.3° (80.6°–89.9°), LEsys, NCC 79.4° (°71.9–83.0°), LEsys, LCC 72.6° (69.9°–76.0°; P<0.001 comparing all leaflets; P=0.002 comparing NCC and RCC, P<0.001 comparing RCC and LCC, P<0.001 comparing NCC and LCC; Figure 4C and 4D).

Figure 4.

Figure 4. Computed tomography angiography (CTA)–based assessment of interleaflet differences in systolic leaflet excursion (LEsys). A, Representative oblique CTA images of a patient with aortic stenosis (AS) show reduced opening of the valve leaflets (red arrows) at their midlines in midsystole. B, Tukey box plots of LEsys for individual leaflets are displayed for patients with AS. C, Representative oblique CTA images of a control (CNTRL) patient with normal aortic valve show maximal opening of each leaflet in midsystole. LEsys is indicated as the angle between the annular plane (white dashed line) and the longitudinal body of the leaflet (yellow dashed line). D, Tukey box plots of LEsys for individual leaflets are displayed for control patients. Statistical significance is P<0.05 and P<0.0167 for multiple and pairwise comparisons, respectively. LCC indicates left coronary cusp; NCC, noncoronary cusp; and RCC, right coronary cusp.

Leaflet-Specific Excursion-Calcification Load Relationships

Conceptually, LEsys should correlate well with AVLCCTA if valvular calcification were to be the predominant mechanism of restricted leaflet motion in AS. Because this relationship seemed to depend heavily on AVCCTA (Figure S4), we included AVCCTA as a covariate in our subsequent regression analysis to examine the LEsys-AVLCCTA relationship for each leaflet. After adjusting for AVCCTA and other variables that could influence either LEsys (left ventricular systolic dysfunction [left ventricular ejection fraction<50%]) or AVLCCTA (age, body mass index, hypertension, dyslipidemia, diabetes, chronic kidney disease, smoking) or its measurement (scanner type), we found the relationship between LEsys (transformed) and AVLCCTA to be significant for both NCC (β=−0.0012, P<0.001) and RCC (β=−0.0013, P<0.001), but not LCC (β=-0.0006, P=0.084; Table 2).

Table 2. Linear Regression Analysis of the Relationship Between Leaflet Excursion and Calcification Load

Model by leafletDependent variableIndependent variableβ (95% CI)P value
NCC*LEsys, NCCAVLCCTA, NCC−0.0012 (−0.0017, −0.0006)<0.001
RCC*LEsys, RCCAVLCCTA, RCC−0.0013 (−0.0020, −0.0007)<0.001
LCC*LEsys, LCCAVLCCTA, LCC−0.0006 (−0.0012, 0.0001)0.084

Summary of linear regression models of LEsys (square root transformed dependent variable) with AVLCCTA as the main independent variable. β indicates unstandardized coefficient; AVLCCTA, CTA-derived aortic valve leaflet calcification load; CTA, computed tomography angiography; LCC, left coronary cusp; LEsys, systolic leaflet excursion; NCC, noncoronary cusp; and RCC, right coronary cusp.

* Model adjusted for AVLCCTA, age, body mass index, left ventricular systolic dysfunction, hypertension, dyslipidemia, diabetes, chronic kidney disease, smoking, and scanner type.

† Statistical significance (P<0.05).

Impact of Leaflet-Specific Excursion on AVA

To examine the functional implications of asymmetrical LEsys observed in control patients, we performed computer simulation by initializing the aortic valve model of each patient in maximally opened configuration and incrementally closing one leaflet at a time while assessing AVA (Figure 5A). With the valve fully opened, the maximal displacement value within each leaflet from its diastolic configuration was significantly greater for RCC (10.4 [9.5–11.7] mm) and NCC (9.1 [8.5–10.6] mm) than LCC (7.9 [6.3–10.1] mm; P<0.001 comparing all leaflets; P=0.349 comparing NCC and RCC, P<0.001 comparing RCC and LCC, P=0.008 comparing NCC and LCC; Figure 5B). With incremental closure of each leaflet, the degree of reduction in percentage AVA was greater for NCC (–32.2% [–38.4% to –25.8%]) and RCC (−35.7% [−40.2% to −32.9%]) than LCC (−24.5% [−28.5% to −18.3%]; P<0.001 comparing all leaflets; P<0.001 comparing RCC and LCC, P<0.004 comparing NCC and LCC, P=0.055 comparing NCC and RCC; Figure 5C).

Figure 5.

Figure 5. Computer simulation of leaflet-specific excursion and its effect on aortic valve area (AVA) in control (CNTRL) patients. A, Computer simulated images of a normal aortic valve are shown for all leaflets fully closed (top left), all leaflets fully open (top middle), only noncoronary cusp (NCC) fully closed (bottom left), only right coronary cusp (RCC) fully closed (bottom middle), and only left coronary cusp (LCC) fully closed (bottom right). A representative displacement map (top right) for full valve opening shows greater overall systolic displacement of NCC and RCC than LCC. B, Tukey box plots of maximal systolic leaflet displacement are shown for individual leaflets. C, Tukey box plots of AVA as a percentage of maximal AVA are shown for incremental closures of NCC (orange), RCC (red), and LCC (blue). Statistical significance is P<0.05 and P<0.0167 for multiple and pairwise comparisons, respectively.

Relationships Between Leaflet Calcification Load and Local Coronary Factors

To identify local factors that may influence calcification load in a leaflet-specific manner, we considered factors that could influence coronary flow because the lack thereof had been thought to predispose NCC to disease.13 We performed multivariable linear regression to examine the relationship between the AVLCCTA of each leaflet (dependent variable) and the different coronary factors, including ostial RCA CTO/STO, RCA OH, sinus of Valsalva diameter at RCC (SOVdRCC). small nondominant RCA, left main coronary artery OH, and sinus of Valsalva diameter at LCC (SOVdLCC). Ostial left main coronary artery CTO/STO was not included due to its rare prevalence in our patients with AS (2%). For the corresponding regression model of each leaflet, we adjusted for AVCCTA to account for its significant interindividual differences and factors that could influence either valvular calcification (age, body mass index, hypertension, dyslipidemia, diabetes, chronic kidney disease, smoking) or its measurement (scanner type). We found both ostial RCA CTO/STO and RCA OH to be significant positive predictors of AVLCCTA, RCC (β=1.8792 and P=0.005 for ostial RCA CTO/STO; β=0.3437 and P=0.001 for RCA OH; Table 3). Additionally, ostial RCA CTO/STO and RCA OH were found to be negative predictors of AVLCCTA, NCC (β=−2.1196, P=0.002) and AVLCCTA, LCC (β=−0.2223, P=0.032), respectively.

Table 3. Linear Regression Analysis of Coronary-Related Predictors of Leaflet-Specific Calcification Load

Model by leafletDependent variableIndependent variableβ (95% CI)P value
NCC*AVLCCTA, NCCOstial RCA CTO/STO−2.1196 (−3.4235 to −0.8158)0.002
RCA OH0.0263 (−0.1609 to 0.2134)0.782
SOVdRCC−0.1222 (−0.4314 to 0.1870)0.437
sNDRCA0.3756 (−1.1810 to 1.9321)0.635
LMCA OH−0.1069 (−0.3095 to 0.0956)0.299
SOVdLCC−0.1007 (−0.3985 to 0.1971)0.506
RCC*AVLCCTA, RCCOstial RCA CTO/STO1.8792 (0.5615 to 3.1970)0.005
RCA OH0.3437 (0.1523 to 0.5352)0.001
SOVdRCC−0.0491 (−0.3668 to 0.2687)0.761
sNDRCA−0.1805 (−1.7195 to 1.3585)0.817
LMCA OH−0.0558 (−0.2625 to 0.1526)0.598
SOVdLCC0.1700 (−0.1359 to 0.4759)0.274
LCC*AVLCCTA, LCCOstial RCA CTO/STO−0.1018 (−1.4989 to 1.2952)0.886
RCA OH−0.2223 (−0.4253 to −0.0193)0.032
SOVdRCC0.0524 (−0.2845 to 0.3893)0.759
sNDRCA1.0010 (−0.6306 to 2.6326)0.228
LMCA OH0.2183 (−0.0027 to 0.4392)0.053
SOVdLCC0.1233 (−0.2010 to 0.4476)0.454

Summary of linear regression models by leaflet type with AVLCCTA as the square root transformed dependent variable and the coronary factors listed as independent variables. β indicates unstandardized coefficient; AVLCCTA, CTA-derived aortic valve leaflet calcification load; CTA, computed tomography angiography; CTO/STO, chronic total/subtotal occlusion; LCC, left coronary cusp; LMCA, left main coronary artery; NCC, noncoronary cusp; OH, ostial height; RCA, right coronary artery; RCC, right coronary cusp; sNDRCA, small nondominant RCA; SOVdLCC, sinus of Valsalva diameter at LCC; and SOVdRCC, sinus of Valsalva diameter at RCC.

* Model adjusted for AVLCCTA, age, body mass index, hypertension, dyslipidemia, diabetes, chronic kidney disease, smoking, and scanner type.

† Statistical significance (P<0.05).


This study addresses a significant knowledge gap regarding interleaflet differences in contributions to AS. The main findings were: (1) In male patients with AS being considered for TAVR, significant heterogeneity was observed for their AVCCTA and AVLCCTA; (2) AVLCCTA was the greatest for NCC compared to other leaflets, whereas leaflet thickening, as the predominant disease feature, was more often visualized in LCC than other leaflets; (3) LEsys was the least for LCC compared to other leaflets, and the same pattern was seen in control patients; (4) The LEsys-AVLCCTA relationship was significant for both NCC and RCC but not LCC; (5) In male control patients, closures of both NCC and RCC were associated with greater reductions in AVA than that of LCC, suggesting their greater functional contributions to AVA; (6) Both ostial CTO/STO and OH of RCA were significant positive predictors of AVLCCTA, RCC. Collectively, these findings support the notion that there are significant interleaflet differences in contribution to AS, some of which may be related to local coronary flow.

This work was motivated by the clinical observation that patients with AS often have heterogeneous distribution of leaflet calcification that does not always correlate with reduced leaflet motion. Here, we sought to address these discrepancies by examining the interleaflet differences in AVLCCTA, VLS, LEsys, and LEsys-AVLCCTA relationships. We first observed significant interleaflet differences in AVLCCTA in patients with AS, with NCC carrying the most load as previously described.4 However, it was LCC, rather than NCC, that had the least LEsys. Although noncalcific leaflet thickening/fibrosis could have significantly affected LCC at least in part to underlie these findings, the fact that a similar pattern of interleaflet differences in LEsys was also observed in control patients suggests baseline functional asymmetry among the leaflets. As the flow profile across the left ventricular outflow tract could be skewed, with maximal velocity biasing towards the anteromedial sector leading to NCC/RCC,14 the asymmetrical fluid forces impinging on the leaflets could have contributed to their differential LEsys under normal (and diseased) conditions. The implication of asymmetrical leaflet opening is that NCC and RCC contribute more to AVA than LCC and their protection from disease would be paramount for retarding AS progression.

Noncalcific leaflet thickening (fibrosis) is increasingly recognized as a significant contributor of AS and was found in this study to be the predominant disease feature more frequently in LCC than other leaflets. This explains why a significant LEsys-AVLCCTA relationship was found for only NCC and RCC but not LCC. These results caution against the sole use of AVCCT to gauge AS severity, as it does not factor in leaflet thickening. A clinical dilemma would arise when an AS patient with discordant echocardiographic findings is found to have a subthreshold AVCCT composed of considerable AVLCCTA, NCC and AVLCCTA, RCC but low AVLCCTA, LCC. The patient could have either truly nonsevere AS to justify continued clinical monitoring if LCC is mildly thickened with mildly restricted systolic motion or severe AS warranting valve replacement if LCC is severely thickened with severely limited excursion. Although careful examination of individual leaflet characteristics/motion can assist in deciding whether to fully rely on AVCCT to adjudicate AS severity, a more comprehensive, quantitative method for assessing noncalcific leaflet thickening is needed to complement calcium scoring. The recent development of an image analysis platform that allows combined assessment of noncalcific and calcific volumes can potentially overcome the shortcomings of AVCCT load and better predict AS severity.15

To date, the reasons for interleaflet differences in calcification have remained elusive, with the complex hemodynamic milieu of the aortic root likely playing a role. Previous studies have implicated flow disruption in the aortic sinuses as the cause of abnormal shear stress on the aortic valve that leads to endothelial dysfunction and other sequelae of AS.16 Although differences in root dimensions and leaflet sizes can also influence the flow patterns of the aortic root to result in differential leaflet calcification, coronary flow (or lack thereof) is more commonly hypothesized as a primary cause of interleaflet differences in calcification.13 Our results support this hypothesis by showing that ostial CTO/STO and greater OH of RCA (factors affecting RCA drainage) were associated with greater AVLCCTA, RCC (Figure S5). These same coronary factors were also found to have negative and distant associations with AVLCCTA, NCC and AVLCCTA, LCC., respectively. Nevertheless, the nature of this study does not lend itself to establishing causality for these interesting relationships between leaflet calcification and coronary factors. Further investigation using computational modeling will be needed to understand how each coronary factor influences leaflet-specific calcification.

Study Limitations

The main limitation of this study concerns the generalizability of our findings to females, as only males were enrolled. Although similar interleaflet patterns of AVLCCTA have been reported in females,5 our results pertaining to the leaflet-specific excursion, thickening, excursion-calcification relationship, and coronary factors may not apply and will need further validation. Another study limitation concerns our assessment of LEsys and leaflet thickening at the midline of each leaflet, which may not capture their heterogeneity elsewhere in the leaflet. To address this limitation, we have developed a robust algorithm to map local leaflet displacement as a part of the computer simulation presented herein. Further refinement of this technique with additional superimposed mapping of regional leaflet calcification and thickness should enable a more thorough evaluation of the interactions among leaflet calcification, thickness, and motion.


This study contributes to a better understanding of the interleaflet differences in calcification load, appearance, leaflet excursion, and their interrelationships. The results presented herein suggest that calcification contributes more to AS via NCC and RCC, rather than LCC. When assessing AS severity, AVLCCTA, NCC and AVLCCTA, RCC should be given unreserved consideration because of their significant values, predictability for reduced LEsys, functional contribution to AVA, and the latter’s potential relation with local coronary alterations. However, because noncalcific leaflet thickening can be a predominant disease feature in LCC more often than in other leaflets, careful interpretation of AVCCT in the setting of low AVLCCTA, LCC is needed to avoid underestimation of AS severity. The development of new techniques to simultaneously assess valvular calcification and leaflet thickening/fibrosis would be highly desirable for the comprehensive evaluation of AS.



We thank Blake Wu and Dr Pei-Yu Lee for their assistance with article preparation.

Supplemental Materials

Supplemental Methods

Figures S1–S5

Tables S1–S4

Nonstandard Abbreviations and Acronyms


aortic stenosis


aortic valve area


aortic valve calcification load


aortic valve leaflet calcification load


computed tomography


computed tomography angiography


chronic total occlusion


fractional AVLC


left coronary cusp


systolic leaflet excursion


noncoronary cusp


ostial height


right coronary artery


right coronary cusp


subtotal occlusion


transcatheter aortic valve replacement


visual leaflet score

Disclosures None.


*I.Y. Chen and V. Vedula contributed equally.

Supplemental Material is available at

For Sources of Funding and Disclosures, see page 1132.

Correspondence to: Joseph C. Wu, MD, PhD, 265 Campus Dr, G1120B, Stanford, CA 94305-5454, Email ; Alison L. Marsden, PhD, Clark Center E100B, 318 Campus Dr, Stanford, CA 94305-5428, Email ; or Ian Y. Chen, MD, PhD, 3801 Miranda Ave, Ste 111C, Palo Alto, CA 94304-1207, Email:


  • 1. Pawade T, Clavel MA, Tribouilloy C, Dreyfus J, Mathieu T, Tastet L, Renard C, Gun M, Jenkins WSA, Macron Let al. Computed tomography aortic valve calcium scoring in patients with aortic stenosis.Circ Cardiovasc Imaging. 2018; 11:e007146. doi: 10.1161/CIRCIMAGING.117.007146LinkGoogle Scholar
  • 2. Clavel MA, Pibarot P, Messika-Zeitoun D, Capoulade R, Malouf J, Aggarval S, Araoz PA, Michelena HI, Cueff C, Larose Eet al. Impact of aortic valve calcification, as measured by MDCT, on survival in patients with aortic stenosis: results of an international registry study.J Am Coll Cardiol. 2014; 64:1202–1213. doi: 10.1016/j.jacc.2014.05.066CrossrefMedlineGoogle Scholar
  • 3. Nchimi A, Dibato JE, Davin L, Schoysman L, Oury C, Lancellotti P. Predicting disease progression and mortality in aortic stenosis: a systematic review of imaging biomarkers and meta-analysis.Front Cardiovasc Med. 2018; 5:112. doi: 10.3389/fcvm.2018.00112CrossrefMedlineGoogle Scholar
  • 4. Maeno Y, Abramowitz Y, Kawamori H, Kazuno Y, Kubo S, Takahashi N, Mangat G, Okuyama K, Kashif M, Chakravarty Tet al. A highly predictive risk model for pacemaker implantation after TAVR.JACC Cardiovasc Imaging. 2017; 10(10 pt A):1139–1147. doi: 10.1016/j.jcmg.2016.11.020CrossrefMedlineGoogle Scholar
  • 5. Veulemans V, Piayda K, Maier O, Bosbach G, Polzin A, Hellhammer K, Afzal S, Klein K, Dannenberg L, Zako Set al. Aortic valve calcification is subject to aortic stenosis severity and the underlying flow pattern.Heart Vessels. 2021; 36:242–251. doi: 10.1007/s00380-020-01688-9CrossrefMedlineGoogle Scholar
  • 6. Simard L, Côté N, Dagenais F, Mathieu P, Couture C, Trahan S, Bossé Y, Mohammadi S, Pagé S, Joubert Pet al. Sex-related discordance between aortic valve calcification and hemodynamic severity of aortic stenosis: is valvular fibrosis the explanation?Circ Res. 2017; 120:681–691. doi: 10.1161/CIRCRESAHA.116.309306LinkGoogle Scholar
  • 7. Baumgartner H, Hung J, Bermejo J, Chambers JB, Edvardsen T, Goldstein S, Lancellotti P, LeFevre M, Miller F, Otto CM. Recommendations on the echocardiographic assessment of aortic valve stenosis: a focused update from the European Association of Cardiovascular Imaging and the American Society of Echocardiography.J Am Soc Echocardiogr. 2017; 30:372–392. doi: 10.1016/j.echo.2017.02.009CrossrefMedlineGoogle Scholar
  • 8. Pandey NN, Sharma S, Jagia P, Gulati GS, Kumar S. Feasibility and accuracy of aortic valve calcium quantification on computed tomographic angiography in aortic stenosis.Ann Thorac Surg. 2020; 110:537–544. doi: 10.1016/j.athoracsur.2019.11.041CrossrefMedlineGoogle Scholar
  • 9. Blanke P, Weir-McCall JR, Achenbach S, Delgado V, Hausleiter J, Jilaihawi H, Marwan M, Norgaard BL, Piazza N, Schoenhagen Pet al. Computed tomography imaging in the context of transcatheter aortic valve implantation (TAVI) / transcatheter aortic valve replacement (TAVR): an expert consensus document of the Society of Cardiovascular Computed Tomography.J Cardiovasc Comput Tomogr. 2019; 13:1–20. doi: 10.1016/j.jcct.2018.11.008CrossrefMedlineGoogle Scholar
  • 10. Updegrove A, Wilson NM, Merkow J, Lan H, Marsden AL, Shadden SC. Simvascular: an open source pipeline for cardiovascular simulation.Ann Biomed Eng. 2017; 45:525–541. doi: 10.1007/s10439-016-1762-8CrossrefMedlineGoogle Scholar
  • 11. Myronenko A, Song X. Point set registration: coherent point drift.IEEE Trans Pattern Anal Mach Intell. 2010; 32:2262–2275. doi: 10.1109/TPAMI.2010.46CrossrefMedlineGoogle Scholar
  • 12. Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences.Behav Res Methods. 2007; 39:175–191. doi: 10.3758/bf03193146CrossrefMedlineGoogle Scholar
  • 13. Dweck MR, Boon NA, Newby DE. Calcific aortic stenosis: a disease of the valve and the myocardium.J Am Coll Cardiol. 2012; 60:1854–1863. doi: 10.1016/j.jacc.2012.02.093CrossrefMedlineGoogle Scholar
  • 14. Kupari M, Hekali P, Poutanen VP. Cross sectional profiles of systolic flow velocities in left ventricular outflow tract of normal subjects.Br Heart J. 1995; 74:34–39. doi: 10.1136/hrt.74.1.34CrossrefMedlineGoogle Scholar
  • 15. Cartlidge TR, Bing R, Kwiecinski J, Guzzetti E, Pawade TA, Doris MK, Adamson PD, Massera D, Lembo M, Peeters Fet al. Contrast-enhanced computed tomography assessment of aortic stenosis.Heart. 2021; 107:1905–1911. doi: 10.1136/heartjnl-2020-318556CrossrefMedlineGoogle Scholar
  • 16. Hatoum H, Dasi LP. Spatiotemporal complexity of the aortic sinus vortex as a function of leaflet calcification.Ann Biomed Eng. 2019; 47:1116–1128. doi: 10.1007/s10439-019-02224-1CrossrefMedlineGoogle Scholar