Aortic Valve Calcification as a Marker for Aortic Stenosis Severity: Assessment on 16-MDCT

Ralf Koos 1 Andreas Horst Mahnken 2 Anil Martin Sinha 1 Joachim Ernst Wildberger 2 Rainer Hoffmann 1 Harald Peter Kühl 1 Received March 3, 2004; accepted after revision May 18, 2004. 1 Department of Cardiology, University Hospital Aachen, Pauwelsstraße 30, Aachen 52072, Germany. Address correspondence to R. Koos. 2 Department of Diagnostic Radiology, University Hospital Aachen, Aachen 52072, Germany. AJR 2004;183:1813 1818 0361 803X/04/1836 1813 American Roentgen Ray Society Aortic Valve Calcification as a Marker for Aortic Stenosis Severity: Assessment on 16-MDCT OBJECTIVE. The degree of valvular calcification in patients with aortic stenosis was determined with retrospectively ECG-gated 16-MDCT and correlated with the severity of stenosis assessed at cardiac catheterization. SUBJECTS AND METHODS. We conducted a prospective study of 72 patients (38 men and 34 women; mean age ± SD, 69.5 ± 8.8 years) with aortic stenosis who underwent 16-MDCT and cardiac catheterization. Aortic valve calcification was assessed using the aortic Agatston score, aortic mass score, and aortic volume score. Severity of aortic stenosis was classified at cardiac catheterization. Aortic valve area and peak-to-peak and mean transvalvular gradients were correlated with the degree of calcification determined on MDCT. RESULTS. All measured aortic valve calcification scores were significantly higher in patients with severe aortic stenosis (n = 46) than in patients with moderate (n = 15) or mild (n = 11, p < 0.001) aortic stenosis. Aortic valve calcification scores were inversely related to aortic valve area (r = 0.67, p for aortic mass score) and correlated significantly with peak-to-peak (r = 0.70, p ) and mean transvalvular (r = 0.72, p ) gradients. No correlation between the aortic valve calcification and the total coronary calcium scores was observed. CONCLUSION. Aortic valve calcification assessed on 16-MDCT is associated with severity of aortic stenosis. Thus, aortic valve calcification scores should be calculated routinely in all patients undergoing MDCT for assessment of coronary calcification. High aortic valve calcification scores indicate possibly severe aortic stenosis and should prompt a further functional evaluation. N onrheumatic calcific aortic stenosis is the most common valvular heart disease in the population older than 65 years, with a prevalence of 2 7% [1]. The grade of calcification appears to correlate with the rate of disease progression [2]. Moreover, moderate or severe aortic valve calcification has been shown to be a strong and independent predictor for an adverse clinical outcome, including an increased risk of death and a need for aortic valve replacement [3]. Thus, interest is growing in the detection and accurate quantification of aortic valve calcification. Electron beam CT has shown a good reproducibility for quantification of valvular calcium in aortic stenosis. Previous studies have compared electron beam tomography and echocardiography for assessment of aortic stenosis [4 6]. A correlation between the degree of valvular calcification and the severity of aortic stenosis has been shown [7]. Retrospectively ECG-gated MDCT has been evaluated primarily for noninvasive coronary angiography and for the detection and quantification of calcified and noncalcified coronary plaques [8, 9]. Because CT is a sensitive method for the detection of calcification, it is potentially useful for the assessment of aortic valve morphology and quantification of the degree of calcification [10, 11]. Preliminary data have shown a good correlation between the degree of aortic valve calcification assessed on 4-MDCT and transvalvular gradients and aortic valve area measured on Doppler echocardiography [12]. Nevertheless, the applicability of these findings to 16- MDCT scanners has yet to be proven. The aim of this prospective study was to correlate the degree of valvular calcification assessed on 16-MDCT using three aortic valve calcification scores with the severity of aortic stenosis evaluated at cardiac catheterization in patients with aortic stenosis. AJR:183, December 2004 1813

Koos et al. Subjects and Methods Our study population was 72 patients (38 men and 34 women; mean age ± SD, 69.5 ± 8.8 years; mean body mass index, 26.5 ± 3.5) with aortic valve stenosis who had been referred for further diagnostic evaluation. MDCT was performed before invasive coronary angiography. with an unstable clinical condition or a heart rate exceeding 75 beats per minute and women of childbearing age were excluded from the study. Informed consent was obtained from each patient before the investigation in accordance with local ethics committee requirements. MDCT All examinations were performed with a 16- MDCT scanner (SOMATOM Sensation 16, Siemens) using a standardized imaging protocol with retrospective ECG-gating. were scanned in the supine position during a single inspiratory breath-hold (mean duration, 21.7 ± 3.4 sec). The mean heart rate of the patients was 69.3 ± 7.4 beats per minute. No medication was administered before the examination. The scanning parameters for the unenhanced MDCT examinations were as follows: collimation, 12 0.75 mm; tube voltage, 120 kv with an effective tube current time product of 133 mas; table feed per rotation, 2.8 mm; and tube rotation time, 420 msec. Axial images were reconstructed at 60% of the R-R interval with an effective slice thickness of 3 mm and a reconstruction increment of 2 mm using a dedicated convolution kernel (B35f, Siemens Medical Solutions). The field of view was 180 180 mm with a 512 512 matrix. Scanner quality assurance was performed by calibration using a standard cardiac phantom (CT cardiac phantom, QRM). The software program (WinDose 2.1, Scanditrox Wellhöfer) was used to calculate the effective radiation dose of unenhanced MDCT examinations [13]. MDCT Image Evaluation Images were assessed in a consensus interpretation by an experienced radiologist and an experienced cardiologist. Both interpreters were blinded to all patient data, including findings at cardiac catheterization. Image analysis was performed on a separate computer workstation (Leonardo, Siemens) equipped with a dedicated software tool for calcium scoring (Calcium Scoring CT, Siemens). For quantitative assessment of aortic valve calcification, three aortic valve calcification scores were calculated with a detection threshold of 130 H: the aortic Agatston score, the aortic mass score, and the aortic volume score [14 16]. The Agatston score was calculated by multiplying the lesion area by an attenuation factor derived from the maximal Hounsfield units within the area, as described by Agatston et al. [10]. We used the volume of calcium measured in cubic millimeters as the aortic volume score. The aortic mass score was measured in milligrams of calcium hydroxyapatite. Calcification was attributed to the aortic valve if it was clearly part of the valve cusps. Supravalvular calcifications and calcifications of the coronary arteries including the ostia were removed by manual segmentation. In addition, the amount of coronary calcification was assessed using the Agatston score, which was referred to as coronary calcium score in our study. Cardiac Catheterization All patients underwent cardiac catheterization 1 or 2 days after MDCT. Peak-to-peak and mean transvalvular gradients were determined, and the aortic valve area was calculated using the Gorlin formula [17]. The severity of aortic stenosis was classified as mild (aortic valve area 1.5 cm 2 ), moderate (aortic valve area between < 1.5 and 1.0 cm 2 ), and severe (aortic valve area < 1.0 cm 2 ), according to the American College of Cardiology American Heart Association practice guidelines [18]. Statistical Analysis Continuous variables are expressed as mean values ± SD. Continuous variables were compared with Student s t test for unpaired samples. Kruskal-Wallis H test was used for comparisons of the different grades of severity of aortic stenosis assuming a nonnormal distribution. Categoric variables were compared with a chi-square analysis. Spearman s rank correlation was used to assess the relationship between the aortic valve TABLE 1 calcification identified on MDCT images and the severity of aortic stenosis identified on angiograms. A p value of less than 0.05 was considered statistically significant. Statistical analysis was performed with use of statistical software (SPSS version 10.0 for Windows [Microsoft], Statistical Package for the Social Sciences). Results Note. Values are mean ± SD or absolute values (%). NS = not significant. Patient Characteristics Baseline patient characteristics are shown in Table 1. According to the aortic valve area as determined at cardiac catheterization, 11 patients had mild, 15 patients had moderate, and 46 patients had severe aortic stenosis. The mean effective radiation dose was 2.08 ± 0.55 msv for men and 2.84 ± 0.73 msv for women. Quantitative Assessment of Aortic Valve Calcification and Its Relationship to Cardiac Catheterization The ranges for the aortic valve calcification scores for the entire group of patients were 49 10,227 by the Agatston method, 45 8,149 by the volumetric method, and 7 2,962 by the mass method. All measured aortic valve calcification scores were different among the catheterization-determined severity groups (Table 2). with severe aortic stenosis showed significantly higher Characteristics of, Classified by Aortic Valve Area Value (n = 11) Aortic Stenosis (n = 15) (n = 46) Age (yr) 67.7 ± 11.6 70.9 ± 7.3 69.4 ± 8.7 NS Male sex (%) 6 (55) 9 (60) 23 (50) NS Coronary calcium score 383 ± 369 708 ± 1,139 739 ± 1,133 NS Cardiac catheterization Peak systolic gradient (mm Hg) 43.6 ± 12.8 52.4 ± 8.3 94.4 ± 28.5 Mean systolic gradient (mm Hg) 21.0 ± 5.2 31.4 ± 6.5 60.9 ± 18.3 Aortic valve area (cm 2 ) 1.75 ± 0.20 1.24 ± 0.16 0.73 ± 0.17 TABLE 2 Relation of Aortic Valve Calcification Assessed on MDCT and Severity of Aortic Stenosis Aortic Valve Calcification Score Note. Values are mean ± SD. (n = 11) Aortic Stenosis (n = 15) (n = 46) Aortic Agatston score 1,066.2 ± 779.2 2,236.5 ± 575.4 4,560.5 ± 2,252.2 Aortic volume score 863.1 ± 620.9 1,801.3 ± 476.5 3,621.3 ± 1,778.7 Aortic mass score 211.6 ± 170.0 452.3 ± 138.6 1,070.7 ± 614.9 p p 1814 AJR:183, December 2004

Assessing Aortic Valve Calcification on 16-MDCT 12,000 3,000 10,000 Aortic Agatston Score 8,000 6,000 4,000 2,000 0 No. of 11 aortic valve calcification scores than patients with moderate or mild aortic stenosis (Figs. 1 6). In addition, patients with moderate aortic stenosis showed higher aortic valve calcification scores than those with mild aortic stenosis (p = 0.001 for all aortic valve calcification scores). Aortic stenosis severity was inversely correlated to the three different aortic valve calcification scores (p ). A significant correlation was found between the invasively determined aortic valve area and all measured aortic valve calcification scores (Table 3). In addition, the aortic valve calcification scores showed significant correlation with the mean and the peak-to-peak transvalvular gradients (Table 3). To further characterize the usefulness of aortic valve calcification scores in identifying patients with severe aortic stenosis, we constructed receiver operating characteristic (ROC) curves. The best ROC curve for the aortic mass score had an area under the curve (A z ) of 0.91 (Fig. 7). The A z values for the aortic volume score and the aortic Agatston score were similar (Table 4). As a tool for distinguishing patients with severe aortic stenosis from those with moderate or mild aortic stenosis, an aortic valve calcification score of 563 derived with the mass method had a sensitivity of 85%, a specificity of 92%, a positive predictive value of 95%, and a negative predictive value of 77%. Table 4 lists the sensitivities and specificities for different aortic valve calcification cutoff scores. 15 Stenosis Severity 46 Fig. 1. Graph shows box plots indicating median (horizontal line within box) and 25th 75th percentiles for aortic Agatston score in relation to stenosis severity determined at cardiac catheterization. Bars above and below boxes indicate minimum and maximum scores. Aortic Mass Score 2,000 1,000 0 No. of Coronary Calcification The coronary calcium score by the Agatston method was not significantly different among the aortic stenosis severity groups (p = 0.87). We found no correlation between the aortic valve calcification (aortic Agatston score) and the total coronary calcium scores (r = 0.17, p = 0.20). Discussion Our study investigated the calcification of the aortic valve by quantitative analysis using 16-MDCT and correlated the results Fig. 3. Graph shows box plots including median (horizontal line within box) and 25th 75th percentiles for aortic volume score in relation to stenosis severity determined at cardiac catheterization. Bars above and below boxes indicate minimum and maximum scores. 11 15 Stenosis Severity 46 Fig. 2. Graph shows box plots indicating median (horizontal line within box) and 25th 75th percentiles for aortic mass score in relation to invasively determined stenosis severity. Dot above boxes represents single outlier. Bars above and below boxes indicate minimum and maximum scores. Aortic Volume Score 10,000 with the severity of aortic stenosis as classified at cardiac catheterization. The findings of this study show that quantification of aortic valve calcification determined by 16- MDCT has a close correlation to the severity of aortic stenosis. Previous studies comparing electron beam CT and echocardiography in patients with aortic stenosis have shown an inverse correlation between the degree of valvular calcification and noninvasively determined aortic valve area [5, 7, 19]. Preliminary MDCT data showed a correlation between the mean echocardiographic gradient and valvular cal- 8,000 6,000 4,000 2,000 0 No. of 11 15 Stenosis Severity 46 AJR:183, December 2004 1815

Fig. 4. 72-year-old man with mild aortic stenosis assessed at cardiac catheterization. Multiplanar reformation of axial MDCT scans shows mild calcification of aortic valve leaflets (aortic Agatston score 807, aortic volume score 663, aortic mass score 155). Locations of right ventricle (RV), left ventricle (LV), right atrium (RA), and left atrium (LA) are indicated. Fig. 5. 68-year-old man with moderate aortic stenosis assessed at cardiac catheterization. Multiplanar reformation of axial MDCT scans shows moderate calcification of aortic valve leaflets (aortic Agatston score 1,949; aortic volume score 1,597; aortic mass score 467). Locations of right ventricle (RV), left ventricle (LV), right atrium (RA), and left atrium (LA) are indicated. Fig. 6. 78-year-old man with severe aortic stenosis assessed at cardiac catheterization. Multiplanar reformation of axial MDCT scans shows severe calcifications of all aortic valve leaflets (aortic Agatston score 7,750; aortic volume score 5,949; aortic mass score 1,630). Locations of right ventricle (RV), left ventricle (LV), right atrium (RA), and left atrium (LA) are indicated. cification assessed by semiquantitative analysis [20]. In contrast to the findings of a study by Cowell et al. [21] using MDCT without ECG-gating, our data showed a better correlation between aortic valve area or transvalvular gradients and the severity of aortic valve calcification, a finding that emphasizes the importance of ECG-gating. In a recent study using a 4-MDCT scanner, a good correlation was found between the degree of aortic valve calcification as measured on a 3D volume score and the severity of aortic stenosis as assessed on Doppler echocardiography [12]. The results of our study are in agreement with the results of that report. In addition to the factors studied in that report, however, we evaluated all routinely available calcification scores including Agatston, volume, and mass scores, introduced to improve reproducibility, using a 16-MDCT scanner. All aortic valve calcification scores showed a similar correlation between the severity of the aortic stenosis and the degree of aortic valve calcification. Moreover, we used the invasively determined aortic valve area as the reference standard, which is the standard of choice for the evaluation of aortic stenosis, especially in patients whose echocardiograms may be of poor quality. Our results emphasize that a large amount of aortic valve calcification seen on MDCT, especially in patients whose aortic valve calcification scores indicate severe aortic stenosis, should prompt a further cardiologic evaluation, including echocardiography, to assess the hemodynamic burden imposed by the calcified aortic valve. TABLE 3 Correlation of Aortic Valve Calcification Scores with Severity of Aortic Stenosis Score Aortic Agatston area Aortic valve area Mean gradient Peak-to-peak gradient Aortic mass score Aortic valve area Mean gradient Peak-to-peak gradient Aortic volume score Aortic valve area Mean gradient Peak-to-peak gradient 1816 r 1.0 p 0.8 0.67 0.71 0.67 0.67 0.72 0.70 0.67 0.71 0.66 Sensitivity Koos et al. 0.5 0.3 0.0 0.0 0.3 0.5 1 Specificity 0.8 1.0 Fig. 7. Graph shows receiver operating curve (ROC) for aortic mass score. Area under ROC curve for aortic mass scores is 0.91, indicating that test has high degree of accuracy for distinguishing patients with severe stenosis from those with mild or moderate aortic stenosis. AJR:183, December 2004

Assessing Aortic Valve Calcification on 16-MDCT TABLE 4 Cutoff Score Accuracy of Aortic Valve Calcification Cutoff Scores for Predicting Aortic Stenosis Sensitivity (%) Specificity (%) Aortic Agatston score a 890 100 23 1,900 91 50 2,500 83 84 2,800 72 78 3,425 61 100 Aortic mass score b 289 100 39 400 89 60 500 85 80 600 77 92 770 57 100 Aortic volume score c 740 100 23 1,550 91 50 1,680 87 69 2,000 83 81 2,965 54 100 a Area under receiver operating characteristic curve (A z = 0.90. b A z = 0.91. c A z = 0.89. Aortic valve calcification scores are quantitative and highly reproducible using electron beam CT [5, 8]. The volume score and the mass score have been introduced to decrease the variability of the coronary calcium measurements [15, 16, 22]. In measurements of aortic valve calcification, a similarly low interscan variability has been shown for Agatston and volume scores determined with electron beam CT [4]. Ohnesorge et al. [23] showed that MDCT-derived measurements of coronary calcification have a better reproducibility for calcium mass and volume scores than for the Agatston score. Preliminary MDCT data of patients with aortic valve calcification showed an overall median interscan reproducibility of 7.9% [12], comparable to that found with electron beam CT. Thus, a good reproducibility of aortic valve calcification scores has been shown in previous studies. Potential of MDCT Technique Evaluation of aortic valve calcification is of increasing importance as the number of patients who undergo MDCT for evaluation of the coronary arteries or unenhanced MDCT for coronary calcium scoring rises. In the future, an accurate determination of aortic valve calcification may be more important still because the degree of calcification has been shown to be a strong and independent risk factor for disease progression and an adverse clinical outcome [3]. Nevertheless, prospective studies are needed to evaluate whether the degree of calcification as determined on MDCT affects treatment. One group of researchers has suggested that patients with severe aortic stenosis and marked calcification belong to a high-risk subgroup that may benefit from early aortic valve replacement, even in the absence of symptoms [3]. Moreover, several ongoing trials are assessing the impact of lipid-lowering therapy on the rate of progression of aortic stenosis. If statin use can be assumed to be associated with a decreased progression of aortic calcification [24], our study results indicate that MDCT is a valuable method for assessing aortic valve calcification and is potentially useful in monitoring progression of this calcification. Study Limitations Reproducibility of different aortic valve calcification scores in the measurement of aortic valve calcification was not evaluated in our study. Nevertheless, preliminary data have already shown a high reproducibility of aortic valve calcification scores [12]. Although a high agreement between different scanners and coronary calcium scores has been shown [25], the reference ranges of the aortic valve calcium scores are dependent on the imaging protocol and CT equipment used. In conclusion, the results of our study show that a high aortic valve calcification score determined with 16-MDCT may indicate severe aortic stenosis, warranting further hemodynamic examination including echocardiography. We suggest routinely assessing aortic valve calcification scores in patients undergoing MDCT for other reasons, such as coronary calcium scoring. Such routine assessment may identify patients with significant aortic stenosis who warrant further cardiologic evaluation. References 1. Stewart BF, Siscovick D, Lind BK, et al. Clinical factors associated with calcific aortic valve disease. J Am Coll Cardiol 1997;29:630 634 2. Davies SW, Gershlick AH, Balcon R. Progression of valvar aortic stenosis: a long-term retrospective study. Eur Heart J 1991;12:10 14 3. Rosenhek R, Binder T, Porenta G, et al. Predictors of outcome in severe, asymptomatic aortic stenosis. N Engl J Med 2000;343:611 617 4. Budoff MJ, Songshou M, Takasu J, Shavelle DM, Zhao XQ, O Brien KD. Reproducibility of electron-beam CT measures of aortic valve calcification. Acad Radiol 2002;9:1122 1127 5. Kizer JR, Gefter WB, delemos AS, Scoll BJ, Wolfe ML, Mohler ER 3rd. Electron beam computed tomography for the quantification of aortic valvular calcification. J Heart Valve Dis 2001;10: 361 366 6. Pohle K, Mäffert R, Ropers D, et al. Progression of aortic valve calcification: association with coronary atherosclerosis and cardiovascular risk factors. Circulation 2001;104:1927 1932 7. Kaden JJ, Freyer S, Weisser G, et al. Correlation of degree of aortic valve stenosis by Doppler echocardiogram to quantify of calcium in the valve by electron beam tomography. Am J Cardiol 2002;90:554 557 8. Achenbach S, Giesler T, Ropers D, et al. Detection of coronary artery stenosis by contrast-enhanced, retrospectively electrocardiographicallygated, multislice spiral computed tomography. Circulation 2001:103:2535 2538 9. Knez A, Becker CR, Leber A, et al. Usefulness of multislice spiral computed tomography angiography for determination of coronary artery stenoses. Am J Cardiol 2001;88:1191 1194 10. Agatston AS, Janowitz WR, Hildner J, Zusmer NR, Viamonte M Jr, Detrano R. Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol 1990;15:827 832 11. Becker CR, Kleffel T, Crispin A, et al. Coronary artery calcium measurement: agreement of multirow detector and electron beam CT. AJR 2001;176:1295 1298 12. Morghan-Hughes GJ, Owens PE, Roobottom CA, Marshall AJ. Three dimensional volume quantification of aortic valve calcification using multislice computed tomography. Heart 2003;89: 1191 1194 13. Kalender WA, Schmidt B, Zankl M, Schmidt M. Program for estimating organ dose and effective dose values in computed tomography. Eur Radiol 1999;9:552 562 14. Hong C, Becker CR, Schoepf UJ, Ohnesorge B, Bruening R, Reiser MF. Coronary artery calcium: absolute quantification in nonenhanced and contrast-enhanced multi-detector row CT studies. Radiology 2002;223:474 480 15. Hong C, Bae KT, Pilgram TK. Coronary artery calcium: accuracy and reproducibility of measurements with multi-detector row CT assessment of effects of different thresholds and quantification methods. Radiology 2003;227: 795 801 16. Callister TQ, Cooil B, Raya SP, Lippolis NJ, Russo DJ, Raggi P. Coronary artery disease; improved reproducibility of calcium scoring with an electron-beam CT volumetric method. Radiology 1998;208:807 814 17. Gorlin R, Gorlin SG. Hydraulic formula for calculation of the area of the stenotic mitral valve, other cardiac valves, and central circulatory shunts. Am Heart J 1951;41:1 29 AJR:183, December 2004 1817

Koos et al. 18. Bonow RO, Carabello B, de Leon AC, et al. Guidelines for the management of patients with valvular heart disease: executive summary a report of the American College of Cardiology American Heart Association Task Force on Practice Guidelines (Committee on Management of with Valvular Heart Disease). Circulation 1998;98:1949 1984 19. Shavelle DM, Budoff MJ, Buljubasic N, et al. Usefulness of aortic valve calcium scores by electron beam computed tomography as a marker for aortic stenosis. Am J Cardiol 2003;92:349 353 20. Willmann JK, Weishaupt D, Lachat M, et al. Electrocardiographically gated multi-detector row CT for assessment of valvular morphology and calcification in aortic stenosis. Radiology 2002; 225:120 128 21. Cowell SJ, Newby DE, Burton J, et al. Aortic valve calcification on computed tomography predicts the severity of aortic stenosis. Clin Radiol 2003;58:712 716 22. Ulzheimer S, Kalender WA. Assessment of calcium scoring performance in cardiac computed tomography. Eur Radiol 2003;13:484 497 23. Ohnesorge B, Flohr T, Fischbach R, et al. Reproducibility of coronary calcium quantification in repeat examinations with retrospectively ECG-gated multisection spiral CT. Eur Radiol 2002;12: 1532 1540 24. Shavelle DM, Takasu J, Budoff MJ, Mao S, Zhao XO, O Brien KD. HMG CoA reductase inhibitor (statin) and aortic valve calcium. Lancet 2002;359:1125 1126 25. Carr JJ, Crouse JR 3rd, Goff DC Jr, D Agostino RB Jr, Peterson NP, Burke GL. Evaluation of subsecond gated helical CT for quantification of coronary artery calcium and comparison with electron beam CT. AJR 2000;174;915 921 For the convenience of AJR authors, a standardized form requesting permission to reprint from other publications is now available via the ARRS Web site at www.ajronline.org/misc/ifora.shtml. Your computer system must have version 3.0 or later of Adobe Acrobat Reader. 1818 AJR:183, December 2004