The Role of Cardiovascular Computed Tomographic Angiography for Coronary Sinus Mitral Annuloplasty

ABSTRACT: Background. The coronary sinus (CS) travels in close proximity to the left circumflex (LCX) artery. Percutaneously placed CS devices used to treat mitral regurgitation (MR) therefore have the potential to impinge upon the LCX arterial distribution and compromise coronary flow. Objectives. In this study, we sought to analyze the anatomic relationship between the CS, LCX and mitral annulus (MA) in patients with right dominant (RCD), left dominant (LCD) and codominant (CCD) arterial systems using a novel systematic approach. Methods. We retrospectively studied 102 normal patients (46 females) and 27 consecutive patients (5 females) with ischemic severe MR. All patients underwent cardiovascular computed tomographic (CCT) angiography with a 64 multidetector scanner for clinical indications. Images were analyzed using a GE Advantage workstation, version 4.4, capable of advanced image processing and manipulation. Results. In patients with a normal mitral valve, the LCX initially crossed under the coronary sinus/great cardiac vein (CS/GCV) in 74% with RCD, 83% with LCD and 97% with CCD. In patients with ischemic severe MR, the LCX initially crossed under the CS/GCV in 96%. Conclusions. The majority of patients, especially those with a CCD, have the LCX initially coursing under the CS/GCV. CCT data analysis using our newly established method is an excellent tool to evaluate the anatomic course of the LCX in patients being evaluated for percutaneous CS device placement.

J INVASIVE CARDIOL 2010;22:67–73

Key words: cardiovascular CT, cardiac CT, tomography, coronary sinus, mitral regurgitation, percutaneous devices, interventions
Patients with mitral regurgitation (MR) have a dilated and outward displacement of the coronary sinus (CS).¹ The CS is currently being evaluated as a potential anatomic site for device deployment, with the goal of displacing the posterior mitral valve leaflet forward in an attempt to change the anterior-posterior dimension of the mitral valve and reduce MR severity. The prevalence of MR is reported to be about 21%.^2–4 Data show that mitral valve repair provides better long-term outcome than valve replacement.^5–10 However, the use of surgical mitral valve repair for advanced heart failure is technically challenging and is therefore limited due to concerns of significant surgical morbidity and mortality.^11,12 A minimally invasive option to treat MR with percutaneous transvenous catheter-based deployment of an annuloplasty device (PCAD) would decrease recovery time and help reduce surgical morbidity and mortality. Animal studies have shown the feasibility of PCAD for MR.^13–16 However, given the close proximity of the CS to the left circumflex artery (LCX), there is a risk of CS-based devices potentially impinging on the LCX. A recent study¹⁷ showed the feasibility of percutaneous reduction in functional MR with a novel CS-based mitral annuloplasty device in patients with heart failure, and was associated with an improvement in quality of life and exercise tolerance. Of the 48 patients enrolled in that trial, 30 received the device. Eighteen patients did not receive a device because of access issues, insufficient acute functional MR reduction, or coronary artery compromise. Coronary arteries were crossed in 36 of the 43 implantation attempts. For the 17% of implants in which a significant compromise of a coronary artery was observed, the device was recaptured without sequelae. Therefore, an accurate preprocedural understanding of the CS anatomy as it relates to the mitral valve, specifically the posterior mitral leaflet (PML) and the adjacent LCX, is vital for this approach to be efficient and successful. Cardiovascular computed tomographic angiography (CCT) has the capability to provide precise characterization of the CS prior to PCAD and to assess the suitability of the patient’s anatomy.1 The available literature currently lacks a detailed analysis of the CS and LCX relationship in different types of coronary arterial dominances. In addition, a systematic CCT evaluation of the CS would be useful and could potentially be widely applied in clinical practice. In this study, we sought to analyze the anatomic relationship between the CS, LCX and mitral annulus (MA) in patients with right dominant (RCD), left dominant (LCD) and codominant (CCD) arterial systems using a novel systematic approach.

Methods

Patient population, data acquisition and workstation image analyses. This was a retrospective study using Digital Imaging and Communications in Medicine standard (DICOM) data from 102 normal patients (46 females) and 27 patients (5 females) with ischemic severe MR. DICOM images were obtained from the CCT database of 3,000 patients. In patients with a normal MV, those with RCD, LCD and CCD systems were identified. Studies had been performed to evaluate coronary arteries using a 64-multislice CT scanner (General Electric Light Speed VCT, GE Healthcare, Waukesha, Wisconsin) for various clinical indications. Clinical indications included chest pain, shortness of breath, abnormal or equivocal stress test, cardiomyopathy, congestive heart failure and syncope. CCT angiography was not performed for patients with an irregular heart rate, allergy to contrast agents or impaired renal function. All patients provided informed consent. Data were acquired using a standard protocol to visualize coronary arteries and were not acquired using a special coronary venous imaging protocol. In all patients, the images were adequate to perform the analysis. We included all of the patient DICOM images for analysis and did not exclude studies based on image quality. The DICOM images were analyzed using a GE AW 4.4 (General Electric Advantage Workstation Version 4.4) capable of advanced image processing and manipulation. The data were analyzed using the following: 3-dimensional volume rendering (3D-VR), axial 4-chamber views (axial-4C), 2-chamber multiplanar reformat views (MPR-2C), 3-chamber multiplanar reformat views (MPR-3C) and curved planar reformat (CPR) views. The slice thickness in the primary axial-4C was 0.625 mm. Data analysis, descriptors and measurements. As a part of our systematic analysis, we chose to introduce some new descriptive terms in order to optimally characterize the CS-LCX-MA relationship: 1. The mitral-left atrial border (MLAB) was determined by the path taken by CS/great cardiac vein (GCV) in the left atrioventricular (AV) groove (Figures 1 and 3D). Since the success of a CS-based device depends upon the device’s actual force trajectory, which could not only be exactly on the MA, but also over the lower left atrium, the mitral left atrial border was felt to be a more encompassing term. 2. LCX-to-CS crossing was used to describe whether the proximal LCX crossed over or under the GCV as both structures prepared to enter and leave the left AV groove (Figure 2). 3. CS was referred to as CS/GCV, as its relationship with the LCX and MA could be better represented given the fact that a CS-based device traverses the length of the CS and GCV (Figure 3). 4. CS os to LCX-CS proximal crossing was defined as the distance from the CS os to the first intersection of proximal LCX with the GCV (Figure 1). 5. CS os to LCX-CS distal crossing was defined as the distance from CS os to the point where the LCX in its distal course leaves the CS/GCV (Figure 1). 6. LCX-CS proximity zone was defined as the distance between the first LCX-CS/GCV crossing point and where the LCX leaves the CS in the left AV groove. 7. Number of crossings was defined as the number of times the LCX leaves and joins the CS/GCV spanning the LCX-CS proximity zone. 8. Posterior lateral branch (PLB) compromise, first obtuse marginal compromise, second obtuse marginal compromise were used when these branches were close enough and crossing under the CS/GCV to potentially be compromised in the event of a CS device deployment. 9. Ramus/diagonal crossover and LAD crossover were used when the anterior interventricular vein (AIV) cut across and crossed over these vessels. 10. CS os diameter was the diameter of the CS ostium in an axial slice as the sinus connects with the right atrium. 11. Proximal CS diameter was defined as the diameter of that segment of the coronary sinus 1 cm from the coronary ostium in an axial slice. The diameter of the GCV and anterior interventricular vein junction (GCV-AIV junction) was also measured in axial slices. 12. Coronary sinus-mitral valve interplanar measure (CS-MV IP) was defined as the perpendicular distance in an MPR slice between two parallel lines; the first line was drawn through the mitral valve plane and the second line was drawn parallel to the first line, through the center of the CS/GCV (Figure 4).⁵ The “straightened” curved planar reformat view (more commonly known as the “pencil view” or “lumen view”) is a software rendering of the CS/GCV in a single plane as a straight line. Using the straightened CPR, the distance from the coronary sinus ostium to the GCV-AIV junction was measured. In both MPR-2C and MPR-3C, a simple multiplanar reformat measurement (sMPR) was obtained by measuring the distance between the center of the CS/GCV and the MA border to show the CS-MA separation distance in a single submillimeter (0.3 mm) plane. Both sMPR and CS-MV IP were also obtained in axial-4C views. It was noted whether the LCX was present between the GCV/CS and the MA in all three of these views. The MA diameter was measured in these three views. Statistical analysis. Statistical analyses were performed using MedCalc for Windows, version 9.6.4.0 (MedCalc Software, Mariakerke, Belgium). Continuous variables are presented as mean ± standard deviation (SD), and categorical variables as number of patients (n) and frequency of patients (%). Continuous variables were analyzed using an unpaired two-tailed t-test. Statistical significance was accepted for two-sided p-values

Results

A total of 129 patient datasets were evaluated (Table 1). Among the 102 patients with a normal mitral valve, there were 31 patients (30.4%) with a RCD system, 35 patients (34.3%) with an LCD system and 36 patients (35.3%) with a CCD system. All 27 patients with ischemic severe MR had a RCD system. In patients with a normal mitral valve, the LCX initially crossed under the CS/GCV in 74% with RCD, 83% with LCD and 97% with CCD (Table 2). In patients with ischemic severe MR, the LCX initially crossed under the CS/GCV in 96%. Normal mitral valve. The LCD group, in contrast to the RCD group, showed significant differences. The CS os-to-GCV-AIV length was greater in the LCD group by approximately 1 cm (p = 0.0023). There was a significant difference in CS os-to-sistal LCX-CS crossing (81.5 ± 22.7 vs. 20.6 ± 10.7; p Ischemic severe mitral regurgitation. Patients with ischemic severe MR in contrast to patients with a normal mitral valve showed the following differences (Table 3): the CS os-to-GCV-AIV length increased by approximately 43 mm (p Discussion A review of the existing literature shows that in approximately 68% of patients, the LCX courses between the CS and the MA.18 Choure et al found that in those with a structurally normal mitral valve and in patients with severe MR due to mitral valve prolapse, the LCX crossed between the CS and MA in 80% of patients.⁵ In another study, the distance from the CS to the MA and the relationship between the CS and surrounding structures was studied in 61 excised cadaveric human hearts. A diagonal or ramus branch, main circumflex artery, or the obtuse marginal branch was located between anterior interventricular vein/CS and the mitral valve annulus in 16.4% and 63.9% of cases, respectively.¹⁹ The available literature currently lacks a detailed analysis of the CS and LCX relationship in different types of coronary arterial dominances using CCT. In our study, we found that the LCX initially crossed under the CS/GCV in 74% of patients with RCD, 83% of patients with LCD and 97% of patients with CCD. In addition, obtuse marginal branches and posterolateral branches were also in a position to potentially be compressed by a device placed within the CS. This raises concern for potential coronary ischemia induced by a CS-based device. However, it must be emphasized that in some patients, despite the LCX going under GCV in its proximal course, the LCX promptly comes out of this unsuitable position later in its course and then travels in a “safe” position, or might just taper off and be absent distally. In this setting, since the distal end of a CS-based device (which depends upon the device type) is expected not to travel far within the GCV, a safe deployment could be anticipated. Since there is significant anatomic variability from patient to patient, an understanding of the relationship between the CS/GCV and LCX is a critical factor in determining the safety of CS-based devices. In patients with a normal mitral valve, among both LCD and CCD groups, although the initial crossing of the proximal LCX with CS/GCV did not differ very much from those with RCD, the LCX had a lengthier relationship with CS/GCV in the left AV groove, with the artery leaving the vein much more proximally to the CS os than that observed with those with RCD. In patients with severe ischemic MR, the length of the CS/GCV was found to be increased compared to those with a normal mitral valve. Although the actual LCX-CS/GCV proximity zone did not change, the entry and exit zones of the LCX in the left AV groove had significantly moved in a distal fashion from the CS os. This results in an “axial tomographic anti-clockwise rotation” of the entry and exit points of the LCX in relation to the CS-GCV, and is most likely due to the preferential distortion and lengthening of those segments of the CS closer to the CS os. In these patients with severe ischemic MR, the maximum increase in MA diameter was noted in axial-4C, followed by both MPR-2C and MPR-3C. There was a significant increase in the sMPR measurements. In contrast, CS-MV IP measurements were not significantly different, suggesting a more outward, rather than upward displacement of the CS/GCV with the accompanying MA dilation. In addition, a comparison of measurement methods in assessing CS/GCV distance from the MA showed that in normal patients, sMPR values were significantly greater than CS-MV IP values only in MPR-2C and MPR-3C. In patients with severe ischemic MR, sMPR measurements were significantly greater than CS-MV IP values in all three views. These changes most likely reflect the unique changes and distortion of the normal CS/GCV geometry in that region in the setting of established ischemic MR. In ischemic MR, there is eccentric regurgitation due to asymmetric dilation of the posteromedial portion of the MA from the posterior papillary muscle traction. Though further validation is required, both sMPR and CS-MV IP measures may be used in concert with 3D-VR to predict the success of a CS-based device by utilizing a device’s projected force trajectory. That force could be mapped along the MLAB using CPR and cross-referenced with 3D-VR to create a band of compression, which could very well include a sizable portion of the left atrium. CCT provides anatomic information that could be rotated and manipulated in any given plane without loss of resolution. This imaging modality has the ability to provide guidance and essential anatomic information for the use of CS-based devices. CCT could also potentially provide a method to develop customized CS-based devices that would be chosen for patient-specific anatomic issues.

Conclusion

In summary, we found that a large proportion of patients have a LCX arterial distribution that travels under the CS/GCV. CCT data analysis using our newly established method is an excellent tool to evaluate the anatomic course of the LCX in patients being evaluated for percutaneous CS device placement. Study limitations. Our study does not prove the clinical utility of this methodology and does not demonstrate if it works in a clinical situation for guiding patient care. Further studies are needed to help validate its usefulness in guiding percutaneous transvenous catheter-based deployment of mitral valve annuloplasty devices. This study is a retrospective analysis of data and only included patients with severe ischemic MR. We did not include those with mild-to-moderate cases of ischemic MR and nonischemic MR, and are therefore unable to provide data on how the CS/GCV geometry changes or evolves in those conditions. In clinical practice, many patients with significant MR who are not considered surgical candidates may not be ideal candidates for CCT. Examples of this include those with a cardiac resynchronization lead in place or those with atrial fibrillation. Our study was a retrospective analysis of CT-based datasets and did not have such a population. Beam-hardening artifacts from the CRT lead and gating issues due to atrial fibrillation may compromise the accurate interpretation of data. Finally, the data in this study were post-processed, manipulated, analyzed and reported by experienced readers in the field of CCT. It may take some time and training to apply these methods in general clinical practice. Nevertheless, a standardized reporting approach in CCT for CS-based devices would help select appropriate patients and help to effectively share anatomic data between practitioners and institutions.

References

1. Lee MS, Shah AP, Dang N, et al. Coronary sinus is dilated and outwardly displaced in patients with mitral regurgitation: Quantitative angiographic analysis. Catheter Cardiovasc Interv 2006;67:490–494. 2. Enriquez-Sarano M, Schaff HV, Frye RL. Mitral regurgitation: What causes the leakage is fundamental to the outcome of valve repair. Circulation 2003;108:253–256. 3. Singh JP, Evans JC, Levy D, et al. Prevalence and clinical determinants of mitral, tricuspid, and aortic regurgitation (The Framingham Heart Study). Am J Cardiol 1999;83:897–902. Erratum in Am J Cardiol 1999;84:1143. 4. Jones EC, Devereux RB, Roman MJ, et al. Prevalence and correlates of mitral regurgitation in a population-based sample (the Strong Heart Study). Am J Cardiol 2001;87:298–304. 5. Choure AJ, Garcia MJ, Hesse B, et al. In vivo analysis of the anatomical relationship of coronary sinus to mitral annulus and left circumflex coronary artery using cardiac multidetector computed tomography: Implications for percutaneous coronary sinus mitral annuloplasty. J Am Coll Cardiol 2006;48:1938–1945. 6. Thourani VH, Weintraub WS, Guyton RA, et al. Outcomes and long-term survival for patients undergoing mitral valve repair versus replacement: Effect of age and concomitant coronary artery bypass grafting. Circulation 2003;108:298–304. 7. Goldman ME, Mora F, Guarino T, et al. Mitral valvuloplasty is superior to valve replacement for preservation of left ventricular function: An intraoperative two-dimensional echocardiographic study. J Am Coll Cardiol 1987;10:568–575. 8. Jamieson WR, Edwards FH, Schwartz M, et al. Risk stratification for cardiac valve replacement. National Cardiac Surgery Database. Database Committee of The Society of Thoracic Surgeons. Ann Thorac Surg 1999;67:943–951. 9. Gillinov AM, Cosgrove DM, Blackstone EH, et al. Durability of mitral valve repair for degenerative disease. J Thorac Cardiovasc Surg 1998;116:734–743. 10. Grossi EA, Goldberg JD, LaPietra A, et al. Ischemic mitral valve reconstruction and replacement: Comparison of long-term survival and complications. J Thorac Cardiovasc Surg 2001;122:1107–1124. 11. Romano MA, Bolling SF. Update on mitral repair in dilated cardiomyopathy. J Card Surg 2004;19:396–400. 12. Dahlberg PS, Orszulak TA, Mullany CJ, et al. Late outcome of mitral valve surgery for patients with coronary artery disease. Ann Thorac Surg 2003;76:1539–1547; discussion 1547–1548. 13. Maniu CV, Patel JB, Reuter DG, et al. Acute and chronic reduction of functional mitral regurgitation in experimental heart failure by percutaneous mitral annuloplasty. J Am Coll Cardiol 2004;44:1652–1661. 14. Daimon M, Shiota T, Gillinov AM, et al. Percutaneous mitral valve repair for chronic ischemic mitral regurgitation: A real-time three-dimensional echocardiographic study in an ovine model. Circulation 2005;111:2183–2189. 15. Liddicoat JR, Mac Neill BD, Gillinov AM, et al. Percutaneous mitral valve repair: A feasibility study in an ovine model of acute ischemic mitral regurgitation. Catheter Cardiovasc Interv 2003;60:410–416. 16. Kaye DM, Byrne M, Alferness C, Power J. Feasibility and short-term efficacy of percutaneous mitral annular reduction for the therapy of heart failure-induced mitral regurgitation. Circulation 2003;108:1795–1797. Epub 2003 Oct 6. 17. Schofer J, Siminiak T, Haude M, et al. Percutaneous mitral annuloplasty for functional mitral regurgitation: Results of the CARILLON Mitral Annuloplasty Device European Union Study. Circulation 2009;120:326–333. Epub 2009 Jul 13. 18. Tops LF, Van de Veire NR, Schuijf JD, et al. Noninvasive evaluation of coronary sinus anatomy and its relation to the mitral valve annulus: Implications for percutaneous mitral annuloplasty. Circulation 2007;115:1426–1432. 19. Maselli D, Guarracino F, Chiaramonti F, et al. Percutaneous mitral annuloplasty: An anatomic study of human coronary sinus and its relation with mitral valve annulus and coronary arteries. Circulation 2006;114:377–380.

___________________________________________ From the ^*Division of Cardiology, Loma Linda University Medical Center, Loma Linda, California, the §Division of Cardiology, Los Angeles Biomedical Research Institute at Harbor-UCLA Medical Center, Torrance, California, ^£The Heart Hospital Baylor Plano, Plano, Texas, the ^†Division of Cardiology, University of Southern California, Los Angeles, California; and the ^**Division of Cardiology, University of Chicago, Chicago, Illinois. Disclosure: Dr. Budoff dislcoses that he has received speaker honoraria from GE. Manuscript submitted August 26, 2009, provisional acceptance given September 14, 2009, final version accepted October 19, 2009. Address for correspondence: David M. Shavelle, MD, FACC, FSCAI, Division of Cardiovascular Medicine, Los Angeles County/USC Medical Center, Associate Clinical Professor, Keck School of Medicine of the University of Southern California, Los Angeles, CA 90033. E-mail: david.shavelle@usc.edu