Mitral Valve Prolapse: Evaluation With ECG-Gated Cardiac CT ... - AJR

Cardiopulmonar y Imaging • Original Research Shah et al. Cardiac CT Angiography of Mitral Valve Prolapse

FOCUS ON:

Cardiopulmonary Imaging Original Research

Rajnil G. Shah1 Gian M. Novaro 2 Rodolfo J. Blandon1 Lana Wilkinson1 Craig R. Asher 2 Jacobo Kirsch1 Shah RG, Novaro GM, Blandon RJ, Wilkinson L, Asher CR, Kirsch J

Mitral Valve Prolapse: Evaluation With ECG-Gated Cardiac CT Angiography OBJECTIVE. The purpose of our study was to evaluate the feasibility of detecting mitral valve prolapse with ECG-gated 64-MDCT angiography in comparison with the reference standard, transthoracic echocardiography. MATERIALS AND METHODS. The charts of patients consecutively referred for clinically indicated 64-MDCT angiography were reviewed. The study cohort consisted of patients who had undergone transthoracic echocardiography. Two experienced radiologists performed blinded consensus review of the MDCT angiograms of 20 patients, and the findings were compared with those of transthoracic echocardiography, which was the reference standard. RESULTS. With the findings on each anterior and posterior leaflet as separate data points, sensitivity was calculated to be 69.2–84.6% and specificity, 100%. The positive and negative predictive values were estimated to be 100% and 87.0–93.1%. CONCLUSION. ECG-gated cardiac 64-MDCT angiography can be used reliably to detect mitral valve prolapse.

M

Keywords: CT angiography, echocardiography, MDCT, mitral valve, mitral valve prolapse DOI:10.2214/AJR.09.2545 Received February 7, 2009; accepted after revision September 12, 2009. 1 Department of Radiology, Cleveland Clinic Florida, 2950 Cleveland Clinic Blvd., Weston, FL 33331. Address correspondence to J. Kirsch ([email protected]). 2 Department of Cardiology, Cleveland Clinic Florida, Weston, FL.

AJR 2010; 194:579–584 0361–803X/10/1943–579 © American Roentgen Ray Society

AJR:194, March 2010

itral valve prolapse (MVP) is defined echocardiographically as single or bileaflet prolapse 2 mm or more beyond the long-axis annular plane with or without leaflet thickening [1]. It can be caused by histologic abnormalities of valvular tissue and by geometric disparities between the left ventricle and mitral valve. Physical examination and 2D trans thoracic echocardiography (TTE) historically have been the diagnostic reference standards for MVP. However, physical findings such as a classic mid-to-late systolic click with a high-pitched late systolic murmur are low in specificity, and echocardiography occasionally does not yield sufficient information for a diagnosis [1, 2]. Advances in CT technology, including MDCT, ECG gating, and faster gantry rotation, have improved temporal resolution to the point that CT of the heart throughout the cardiac cycle is possible. Increases in the number of high-resolution detector elements have further improved the quality and robustness of cardiac CT [3]. Studies of cardiac MDCT angiography have focused on imaging of the coronary arteries. However, the same technology that facilitates submillimeter isotropic voxel imaging of coronary anatomic features also facilitates direct imaging

of intracardiac anatomic features. Structures that previously were suboptimally visualized owing to motion artifact at CT, such as the cardiac valve apparatus, can now be imaged as a part of a routine cardiac evaluation. To our knowledge, no studies with published results have been conducted on the use of cardiac CT angiography for the diagnosis of MVP. The purpose of our study was to evaluate whether ECG-gated cardiac CT angiography with a 64-MDCT scanner can be used reliably for detection of MVP with TTE as the reference standard. Materials and Methods In accordance with the policies for exemption set by our internal institutional review board, we performed a retrospective review of the charts of patients consecutively referred for cardiac CT angiography from September 2007 through July 2008. Patients who had undergone a documented TTE examination within 3 months of MDCT were established as the study cohort. All seven patients with a diagnosis of MVP at TTE were included, as were 13 patients who underwent both examinations without visualization of MVP at TTE. Review of these patients’ charts showed that no intercurrent ischemic or other event that might have caused a change in mitral valve status had occurred between TTE and CT. Twelve of the 20

579

Shah et al.

A

B

D

C

E

Fig. 1— 51-year-old woman. Step-by-step method of assessment of mitral valve with cardiac CT angiography. A, After phase of CT closest to end-systole has been selected in axial plane, two-chamber multiplanar reconstruction is made by alignment of plane (dashed line) to cross left ventricular apex and point of coaptation of mitral valve. B, Short-axis multiplanar reconstruction image is obtained by aligning plane (dashed line) perpendicular to long axis of left ventricle at its base, near tip of mitral valve leaflets. C, Plane (dashed line) centered through left ventricular outflow tract is made from short-axis image in B. D, Three-chamber view is equivalent to parasternal long-axis view on echocardiogram. E, Three-chamber view is used to draw imaginary annular line (dotted line) and assess relative position of mitral valve leaflets. LA = left atrium, LV = left ventricle.

patients underwent TTE before CT angiography; four underwent TTE after CT angiography; and four underwent both studies on the same day. No exclusion criteria were implemented. Two experienced cardiologists in consensus reviewed the TTE images for the presence or absence of anterior and posterior leaflet MVP. Leaflet prolapse 2 mm or more beyond the parasternal longaxis annular plane with or without leaflet thickening was used as diagnostic confirmation of MVP. The CT examinations were performed with a 64-MDCT scanner (Sensation 64, Siemens Healthcare). Patients were supine with ECG leads paced. The acquisition parameters were as follows: 650 mAs at 120 kVp; tube rotation time, 0.33 second; detector configuration, 32 × 0.6 mm; reconstructed width, 0.75 mm; reconstructed interval, 0.4 mm; pitch, 0.23; field of view, 25 cm2.

580

Contrast timing was determined with a test bolus consisting of 16–20 mL of iodixanol (Visipaque, GE Healthcare) with an iodine concentration of 320 mg I/mL injected at 4–5 mL/s. Cardiac CT angiography was performed with 80–105 mL iodixanol 320. The total volume of contrast material used for angiography was based on scanning duration according to the following formula: 5 mL/s × (scanning duration + 5 seconds). The saline flush was performed with 40 mL of 0.9 normal saline solution to minimize perivenous artifacts in the superior vena cava and to reduce opacification of the right-sided cardiac chambers. Patients with a heart rate greater than 67 beats/ min (bpm) received metoprolol 5–60 mg IV in increments of 5 mg until reaching a heart rate less than 67 bpm before scanning. Patients with a heart rate less than 67 bpm did not receive rate-control

medications. The mean heart rate of the selected population while undergoing CT was 58 ± 10 (SD) bpm. At the time of CT, patients without contraindications to sublingual nitroglycerin received a one-time sublingual dose of 0.4 mg nitroglycerin. For each patient, the retrospective ECG-gated CT scans were reconstructed into 10 image series, each at a different phase in the cardiac R-R cycle, starting at 0% and increasing in 10% increments until reaching the 90% phase. Each examination was given a separate anonymous identification number, and the order was randomized for the observers. The images contained no patient-identifying information. For evaluation of the CT scans, all 10 phases were loaded into an advanced imaging postprocessing workstation (InSpace, Siemens Healthcare). A three-chamber view (equivalent to a parasternal

AJR:194, March 2010

Cardiac CT Angiography of Mitral Valve Prolapse long-axis view at echocardiography) was methodically obtained by each observer starting in the axial plane at or near end-systole. Figure 1 shows the step-by-step process used by the observers to obtain the desired final plane. Imaging criteria for the presence of MVP were developed. In the three-chamber plane, a line was drawn to join the mitral annulus, which was defined as the insertion site of the base of the mitral leaflets. If the leaflets extended proximally into the left atrium, crossing the drawn imaginary annular line, a measurement was made from this line to the midlevel of each leaflet. If this distance was greater than 2.0 mm, prolapse of that leaflet was considered present (Fig. 2). All CT scans were reviewed independently by two experienced radiologists (observer 1, 3 years of cardiac imaging experience; observer 2, 9 years of cardiovascular imaging experience). For measurement of intraobserver variability, observer 1 performed two readings 1 week apart, the randomization differing between the two sessions. For assessment of interobserver variability, observer 2 evaluated the images once in a different randomized presentation from the two used by observer 1. Observer 1 also measured the maximum thickness of the leaflets in each patient. During the study, the radiologists were blinded to

A

B

Fig. 2—Comparative three-chamber cardiac CT angiographic views show position of mitral valve leaflets in relation to imaginary annular line (dotted line). LV = left ventricle, LA = left atrium, Ao = aorta. A, 35-year-old woman with normal-appearing mitral valve. B, 68-year-old man with mitral valve prolapse confirmed at echocardiography.

the results of TTE. The cardiologists were blinded to the results of CT. Interobserver and intra observer agreement on categoric measures was determined with kappa statistics. The diagnostic characteristics of CT analyzed were sensitivity, specificity, positive predictive value, and negative predictive value.

Results A total of 413 patients underwent cardiac CT angiography from September 2007 through July 2008. One hundred twenty-two of these patients had previously undergone TTE. A total of 20 patients (11 men, nine women; mean age, 57 years; range of 26–85

TABLE 1: Indications for Transthoracic Echocardiography and Cardiac CT Angiography Patient No.

Sex

Age (y)

Indication for Transthoracic Echocardiography

Indication for CT

1

M

58

Coronary atherosclerosis

Coronary bypass grafting, evaluation of ventricular function

2

F

48

Mitral valve disorder, hypertension

History of hypertension, chest discomfort

3

M

26

Cardiac murmur

Pulmonary stenosis, tetralogy of Fallot

4

F

35

Chest pain

Chest pain

5

F

54

Chest pain

Chest pain

6

M

68

Chest pain

Chest pain, indeterminate findings at nuclear perfusion imaging

7

M

36

Aortic valve replacement

Aftermath of aortic valve replacement; hypercholesterolemia

8

M

63

Chest pain

Acute chest pain, hypertension, hyperlipidemia

9

F

58

Hypertension

Hypertension, chest pain

10

F

37

Congenital coronary anomaly

Congenital coronary anomaly, chest pain

11

M

70

Mitral valve disorder

History of murmur, preoperative evaluation

12

F

45


Acute chest pain

13

F

72

Ventricular septal defect repair

Preoperative evaluation

14

M

63

Atherosclerosis

Chest pain

15

M

62

Coronary artery disease

Dyspnea on exertion, coronary bypass grafting

16

F

51

Atherosclerosis

Coronary bypass grafting, evaluation of ventricular function

17

M

79



18

F

85

Angina, stable

Chest pain, dyspnea on exertion

19

M

56

Hypertension

Cardiac murmur, abnormal ECG findings

20

M

68



AJR:194, March 2010

581

Shah et al. TABLE 2: Findings at Transthoracic Echocardiography and Cardiac CT Angiography Patient No. 2

Mitral Valve Prolapse at Echocardiography

Mitral Valve Regurgitation

Interobserver Agreement Observer 1 determined the presence of MVP of the anterior leaflet in five cases during the first review of images and in six cases during the second review. Observer 2 determined the presence of MVP of the anterior leaflet in five cases. The agreement between the first set of readings by observer 1 and the readings by observer 2 was perfect (κ = 1.00, p < 0.001). The agreement between the second set of readings by observer 1 and the set of readings by observer 2 also was very high (κ = 0.87, p < 0.001). Observer 1 determined the presence of MVP of the posterior leaflet in five cases in both the first and second reviews. Observer 2 determined the presence of MVP of the posterior leaflet in four cases. The agreement between the first set of readings by observer 1 and the set of readings by observer 2 was strong (κ = 0.87, p < 0.001). The agreement between the second set of readings by observer 1 and the set of readings by observer 2 was identical to the comparison between the first set by observer 1 and the set of readings by observer 2 (κ = 0.87, p < 0.001) (Tables 4 and 5).

Mitral Valve Prolapse at CT

Severe

Moderate

Prolapse of both leaflets

10

Mild

Trivial

Anterior leaflet prolapse

11

Moderate

Moderately severe


12

Mild

Mild

Not detected

13

Mild

Severe


17

Severe

Severe


20

Severe

Severe


A

Diagnostic Measures Seven of the 20 patients who underwent TTE and cardiac MDCT angiography within a 3-month period were found to have MVP at TTE. According to the echocardiographic criteria of at least 2-mm single or bileaflet prolapse beyond the long-axis annular plane, 11 of the 13 leaflet prolapses were detected with MDCT in the second set of readings by observer 1 (sensitivity, 84.6%; overall sensitivity, 69.2–84.6%). In the 13 patients with no evidence of MVP at TTE, all 26 leaflets plus the one normal leaflet in patient 10 were correctly identified as normal at MDCT (specificity, 100%) (Table 6).

B

Fig. 3—Patient with mitral valve prolapse. Comparative three-chamber views. A and B, Parasternal long-axis-view transthoracic echocardiogram (A) and MDCT image (B) show prolapse. Dotted line indicates imaginary annular line. See also Figure S3C, cine loop, in supplemental data online.

years) who underwent CT and TTE within a 3-month period were included in the study. The average time span between CT and TTE was 23 days (range, 0–65 days). The indications for both cardiac CT angiography and TTE are summarized in Table 1. In seven patients in the cohort, MVP was diagnosed during TTE. MVP was graded categorically as mild, moderate, or severe and was qualified as bileaflet, anterior, or posterior leaflet prolapse. Mitral regurgitation was graded semiquantitatively with color Doppler flow mapping (grade 0, trivial; grade 1, mild; grade 2, moderate; grade 3, moderately severe; and grade 4, severe) and with the proximal flow convergence method. Five of the patients had bileaflet prolapse, and one had single leaflet prolapse. The TTE and CT findings are summarized in Table 2 and Figure 3. (Figure S3C can be seen in the AJR electronic supplement to this article, available at www.ajronline.org). All patients without MVP had a leaflet thickness less than 2

582

mm, but the patients with MVP had an average leaflet thickness of 2.36 ± 1.49 mm. Intraobserver Agreement In two readings of the CT scans, observer 1 determined the presence of MVP of the anterior leaflet in five cases during the first review and in six cases during the second review. Agreement between the two sets of observations was strong (κ = 0.87, p < 0.001). For determination of the presence of posterior leaflet prolapse, agreement was perfect (five cases during each review) (κ = 1.00, p < 0.001) (Table 3).

Discussion Within the limitations of a small cohort, our results show that MDCT can be used for the diagnosis of MVP with high sensitivity and specificity with TTE as the reference

TABLE 3: Intraobserver Agreement (Observer 1) on CT Findings of Mitral Valve Prolapse in Anterior and Posterior Leaflets Image Set 1 Anterior (κ = 0.87) Image Set 2

Posterior (κ = 1.00)

Absent

Present

Absent

Present

Absent

14

0

15

0

Present

1

5

0

5

AJR:194, March 2010

Cardiac CT Angiography of Mitral Valve Prolapse TABLE 4: Interobserver Agreement on CT Findings: First Observations of Mitral Valve Prolapse by Observer 1 Versus Observations of Anterior and Posterior Leaflets by Observer 2 Observer 1, Set 1 Anterior (κ = 1.00) Observer 2


Absent

Present

Absent

Present

Absent

15

0

15

1

Present

0

5

0

4

TABLE 5: Interobserver Agreement on CT Findings: Second Observations of Mitral Valve Prolapse by Observer 1 Versus Observations in Anterior and Posterior Leaflets by Observer 2 Observer 1, Set 2 Anterior (κ = 0.87) Observer 2

Absent


Present

Absent

Present

Absent

14

1

15

1

Present

0

5

0

4

raphy is an alternative, although it is partially invasive [12]. MRI has been proved useful in the evaluation and detection of leaflet abnormalities due to MVP and of jet direction in mitral regurgitation [13]. Limitations on spatial and temporal resolution previously prevented CT from being reliable in the diagnosis of MVP. With improvements in CT scanner technology, particularly the advent of submillimeter isotropic voxel resolution and faster gantry rotation, evaluation of intracardiac structures has become a reality. With CT, however, radiation dose exposure can be a concern. With dose modulation, the estimated effective dose for 64-MDCT angiography of the coronary arteries is approximately 10 mSv. The data on all patients included were not available for accurate assessment of the radiation doses used in our cohort. However, we do not recommend that CT be used for direct assessment of MVP; we recommend use of this technique to add value to an already radiationintensive study. The purpose of this study, the first of its kind to our knowledge, was to determine the diagnostic accuracy of CT for MVP with TTE as the reference standard. Our study showed the feasibility of retrospective ECG-

standard. MVP is a common valvular abnormality that is the most common cause of severe nonischemic mitral regurgitation in the United States [4, 5]. At auscultation, MVP is heard as a mid-to-late systolic click frequently associated with a high-pitched late systolic murmur. The symptoms are dyspnea on exertion and atypical chest pain. MVP is defined echocardiographically as single or bileaflet prolapse 2 mm or more beyond the long-axis annular plane with or without leaflet thickening [1]. When prolapse is accompanied by thickening of the valve leaflets, a 5-mm thickness is considered classic prolapse and is more likely associated with symptoms, mitral regurgitation, and cardiac events [6, 7]. Although it is a multifactorial abnormality, the most common cause of prolapse has been identified as myxomatous degeneration [4, 8–10]. Patients with examination findings or symptoms that suggest MVP traditionally have been referred for 2D TTE to confirm the diagnosis [1]. This method has proved useful in stratifying the severity of MVP because it shows redundancy and the degree of mitral regurgitation [6, 7, 11]. TTE is the study of choice for assessing MVP and mitral regurgitation, but transesophageal echocardiog-

TABLE 6: CT Versus Transthoracic Echocardiographic Detection of Mitral Valve Prolapse (Observer 1, Set 2) Transthoracic Echocardiographic Finding Present

Absent

Present

CT Finding

11

0

Absent

2

27

Note—Sensitivity, 84.6%; positive predictive value, 100%; specificity, 100%; negative predictive value, 93.1%.

AJR:194, March 2010

gated cardiac CT angiography with a 64MDCT in the diagnosis of MVP. MDCT had good sensitivity (85%) and specificity (100%) for the diagnosis. The study showed that with MDCT, we can confidently establish the diagnosis of MVP. Although cardiac CT is not expected to replace echocardiography as a first-line test for the evaluation of MVP, the finding that this disease can be detected with ECG-gated CT angiography may add benefit to cardiac CT. With the increased utility of CT in the evaluation of chest pain, the finding of MVP may offer an alternative explanation. This initial feasibility study may serve as a pilot study for future investigations. The limitations of our study included the small number of patients evaluated. In addition, although this factor did not appear to have a negative influence in our study, the CT reconstructed phase images used for assessment of MVP were those acquired during end-systole and by definition were the images expected to be the most affected by motion artifact. Larger-scale studies may be required to assess the effect of this limitation. With current prospective gating technique, images can be acquired during diastole alone. In the assessment of MVP, CT scans have to be specifically acquired in the late systolic phase. In addition, an inherent limitation of MDCT is that direct quantification of the extent of mitral regurgitation is not possible. Our preliminary experience indicates that the detection of MVP with ECG-gated cardiac 64MDCT angiography is possible and has high sensitivity and specificity for this diagnosis. Acknowledgment We thank Benjamin Nutter for help with the mathematical analysis. References 1. Hayek E, Gring CN, Griffin BP. Mitral valve prolapse. Lancet 2005; 365:507–518 2. Weis AJ, Salcedo EE, Stewart WJ, et al. Anatomic explanation of mobile systolic clicks: implications for the clinical and echocardiographic diagnosis of mitral valve prolapse. Am Heart J 1995; 129:314–320 3. Boiselle P, White C. New techniques in cardiothoracic imaging. New York, NY: Informa Healthcare, 2007 4. Olson LJ, Subramanian R, Ackermann DM, et al. Surgical pathology of the mitral valve: a study of 712 cases spanning 21 years. Mayo Clin Proc 1987; 62:22–34 5. Waller BF, Morrow AG, Maron BJ, et al. Etiology

583

Shah et al. of clinically isolated, severe, chronic, pure mitral regurgitation: analysis of 97 patients over 30 years of age having mitral valve replacement. Am Heart J 1982; 104:276–288 6. Marks AR, Choong CY, Sanfilippo AJ, et al. Identification of high-risk and low-risk subgroups of patients with mitral valve prolapse. N Engl J Med 1989; 320:1031–1036 7. Nishimura RA, McGoon MD, Shub C, et al. Echocardiographically documented mitral valve prolapse: long term follow up of 237 patients. N Engl J Med 1985; 313:1305–1309 8. Guthrie RB, Edwards JE. Pathology of the myx-

omatous mitral value: nature, secondary changes and complications. Minn Med 1976; 59:637–647 9. Davies MJ, Moore BP, Braimbridge MV. The floppy mitral valve: study of incidence, pathology, and complications in surgical, necropsy, and forensic material. Br Heart J 1978; 40:468–481 10. Agozzino L, Falco A, de Vivo F, et al. Surgical pathology of the mitral valve: gross and histological study of 1288 surgically excised valves. Int J Cardiol 1992; 37:79–89 11. Agricola E, Oppizzi M, De Bonis M, et al. Multiplane transesophageal echocardiography performed according to the guidelines of the Ameri-

can Society of Echocardiography in patients with flail, and endocarditis: diagnostic accuracy in the identification of mitral regurgitant defects by correlation with surgical findings. J Am Soc Echocardiogr 2003; 16:61–66 12. Brickner ME. Transesophageal echocardiography. J Diagn Med Sonogr 2005; 21:309–317 13. Gabriel RS, Kerr AJ, Raffel OC, Stewart RA, Cowan BR, Occleshaw CJ. Mapping of mitral regurgitant defects by cardiovascular magnetic resonance in moderate or severe mitral regurgitation secondary to mitral valve prolapse. J Cardiovasc Magn Reson 2008; 10:16

F O R YO U R I N F O R M AT I O N

A data supplement for this article can be viewed in the online version of the article at: www.ajronline.org.

584

AJR:194, March 2010