Angiographic Anatomy of the Coronary Sinusand Its Trihutaries

MARTINE GILARD, JACQUES MANSOURATI, YVES ETIENNE,JEAN-MARIE LARLET, BERNARD TRUONG, JACQUES BOSCHAT,and JEAN-JACQUES BLANC

From the Department of Cardiology Brest University Hospital, Brest, France

GILARD, M., ET AL.: Angiographic Anatomy of the Coronary Sinus and Its Tributaries. Permanent left ven-tricular pacing has been shown to imporve the hemodynamic and clinical status of patients with severeheart failure. To pace the left ventricle, the electrode is implanted in tributaries of the coronary sinus {CS).However, the anatomy of cardiac veins with this purpose in mind has not been described in detail.Methods: One hundred consecutive patients admitted for coronary angiography had a simultaneous coro-nary venography performed after the injection of 8 to 10 mL of contrast material into the left coronaryartery. Cardiac veins were analyzed in antero-posterior, left anterior oblique 60°, and right anterior oblique30° views by three different observers. The number, dimension, angulation, and position of the coronary si-nus and of its tributaries were studied. Results: Two veins are consistently present: the middle cardiac vein(mean diameter 2.62 ± 1.26 mm) and the great cardiac vein (mean diameter 3.55 ± 1.24 mm). The left pos-terior vein(s) (LPV) (mean diameter 2.25 ±1.2 mm) is (are) variable in number (ranging from 0 to 3), size,and angulation. The absence of LPV limits the ability to pace the left ventricle endovenously. The diame-ter of the vein (< 2 mm) and its angulation may also complicate the insertion of the lead. Conclusion: An-giographic analysis of dimensions, tortuosity, number, and angulation of venous tributaries of the CS seemsto allow the insertion of commercially available pacing leads in approximately 85% of cases. An increasein this percentage hinges on the development of new, dedicated leads. (PACE 1998; 21[Pt. II]: 2280-2284.)

coronary sinus, angiography, left ventricular pacing, anatomy

Introduction

Since its original description the anatomy ofthe cardiac vessels has focused almost exclu-sively on the arterial system. A few authors, how-ever did study the venous system: in 1651,Bartbolinus^ described two coronary veins and,in 1839, Winslow^ named the transverse segmentof the great cardiac vein: the coronary sinus (CS).In 1952, for the first time. Tori described a coro-nary venogram."̂ Since that time, few papers havebeen devoted to the angiographic anatomy of thehuman coronary veins. Despite the scarcity ofdata on this subject, the venous system has beenused increasingly for surgical and electrophysio-logical purposes, including retrograde coronary

Address for reprints: M. Gilard, Department of Cardiology,CHU Cavale Blanche Bd Tanguy Prigent, 29609 BREST C^dex-France. FAX: 33-2-98-34-78-03; Tel 33-02-98-34-73-92;e-mail: Jacques.Boschat@imiv-brest.fr

perfusion for myocardial preservation sensing ofleft atrial activity, and radiofrequency catheterablation. More recently, it has been shown thatleft ventricular-based pacing may markedly im-prove patients with severe heart failure.''•^ Jn thissetting, a permanent pacing electrode is usuallyplaced in a tributary of the CS. This importanttherapeutic method has caused a renewed inter-est in the precise angiographic anatomy of the CSand, more precisely, of its tributaries. The ain:i ofthe present study was to describe the humanangiographic anatomy of the main tributaries ofthe CS potentially available for left ventricularpacing.

Methods

Study Patients

The study group consisted of 100 consecu-tive patients (73 men) with a mean age of 62.2 ±10.7 years (range 36 to 78 years), undergoing for

2280 November 1998, Part II PACE, Vol. 21

ANGIOCRAPHY OF THE CORONARY SINUS AND ITS TRIBUTARIES

coronary angiography. A few patients with thefollowing criteria were excluded: coronary an-giography in the early phase of myocardial in-farction, significant stenosis of the left mainartery, severely impaired renal function, allergicreactions to contrast agents, and lack of coopera-tion.

Coronary Venography

To image the anatomy of the venous system,the left coronary ostium was cannulated with a 6FR Judkins catheter to inject 8 to 10 mL of con-trast material. The venous phase occurred 5 to 10seconds after the injection. Three views were im-aged: right anterior oblique (RAO) 30°, anteropos-terior (AP), and left anterior oblique [LAO) 60°(Fig. 1). Images were recorded on a 35-mm filmand a videotape for subsequent analysis. Allrecordings were analyzed by three different ob-servers for precise radio-anatomic descriptions ofthe number, dimension, angulation (angle be-tween the branching vein and the CS), and tortu-osity ofthe tributaries ofthe CS. Measurements ofthe CS and its trihutaries were performed in theAP view with the ARTREK (Image Comm. Sys-tems, Sunnyvale, CA. USA). The diameter of theCS was calculated as the mean of three measure-ments at the CS ostium, the estimated level oftheMarshall vein, and the mid-point between these

two sites. The diameter of the major tributaryveins was measured at the widest visible site.

Terminology

The terminology used to describe the varioustributary veins is the one recommended by MacAlpine^:

The great cardiac vein: It originates, in themajority of cases, from the lower or middle thirdof the interventricular sulcus, courses with in it,and it then turns toward the left side of the atri-oventricular groove and enters the CS at an ap-proximately 180° angle. Its ostium coincides, inthe majority of cases, with the valve of Vieussens.

The middle cardiac vein: It originates near thecardiac apex, courses within the posterior inter-ventricular groove, and drains into the CS just be-fore it arrives in the right atrium. It may also draindirectly into the right atrium.

The left posterior vein: Left ventricular ve-nous branches arise from the lateral and posterioraspects ofthe left ventricle and drain into the greatcardiac vein or into the CS.

The left atrial oblique vein of Marshall: Itcourses diagonally on the posterior surface of theleft atrium and joins the great cardiac vein at thepoint where the latter becomes the CS. This veinis small in humans and was difficult to visualizein our study.

Figure 1. Angiographic RAO 30°, AP, and LAO 60° views of the venous system of the heart inthe same patient. GCV = great cardiac vein; LPV - left posterior vein; MCV = middle cardiacvein.

PACE, Vol. 21 November 1998, Part II 2281

GILARD, ET AL

Statistical Analysis

All values are expressed as mean ± SD. Com-parisons of data among different groups were per-formed using Student's t-test. A P value < 0.05was considered statistically significant.

Results

Coronary Arteriography and LeftVentriculography

The coronary arteries were distributed as usu-ally described: right dominant in 68, left dominantin 17, and mixed in 15 patients.

The mean end-diastolic volume (EDV) was112.6 ± 52.3mL/mMbetween30and304mL/M^).

Coronary Venography

CS, great cardiac vein, and middle cardiacvein were visualized in all patients.

Coronary Sinus

In the AP view, the orifice of the CS was su-perimposed over the left portion of a thoracic ver-tebra at the level of the diaphragm. However, theselandmarks are too variable to be useful in individ-ual cases. For anatomists, the CS ends at the levelof the Thebesan valve and tbe great cardiac veinends in the CS at the level of tbe Vieussensvalve.^"^ Since these valves were not visible onangiography, the precise length of the CS couldnot be measured. Its mean diameter was 6.80 ±1.9mm [range 3.1 to 15.1 mm).

The Middle cardiac vein and great cardiacvein were consistently present on all venograms.

Middle Cardiac Vein

It consistently coursed in all patients as de-scribed by anatomists, as described earlier and re-joined the CS perpendicularly near its ostium af-ter a consistent 60° to 90° angle situated 1 cmbefore its anastomosis. Its mean diameter was 2.6± 1.3 (from 0.7 to 6 mm).

Great Cardiac Vein

It originated in the anterior interventricularsulcus coursed toward the atrioventricular groove.At that point, in the AP view, it appeared to have

a variable curve (when this curve made a U-turn,the vein was considered as tortuous). Past thatcurve, the great cardiac vein remained in tbe sameaxis as the CS and joined it at approximately thelevel of the Marshall vein (when the latter was vis-ible). This anatomy was concordant in all patients.Variations were observed in tortuosity and diame-ter: the latter measured 3.51 ± 1.2 mm (range 1.3to 6.7 mm) at its distal segment. Before the curvewhen the vein is still on the epicardial surface ofthe left ventricle, the diameter is considerablysmaller and difficult to measure, generally be-tween 0.5 and 3 mm. Tortuosity, as defined ear-lier, was observed in 13 patients (Fig. 2).

Left Posterior Veins

These veins originate fTom the posterior andlateral aspects of the left ventricle and join the CS(posterior veins) or the great cardiac vein (lateralveins). As opposed to the presence of great andmiddle cardiac veins, these posterior veins arehighly variable in number, size and angulation.The number of visible left posterior veins on an-giography ranged between 0 in 1% of our patientsand 3 in 3%, with intermediate numbers of 1 in51% and 2 in 45% of patients. The mean diameterofthelargest vein was 2.25 ± 1.2 mm (range 0.3 to6 mm). It is noteworthy that in patients fonnd tohave a single posterior vein, it was very posterior,parallel, and close to tbe middle cardiac vein,probably not suitable for optimal left ventricularpacing in six cases. It should also be emphasizedthat in some patients the diameter of the largestvein was < 1 mm. Tbe angle of the junction be-tween the left posterior vein and the CS or greatcardiac vein was acute (< 90°) in 43% of cases(Fig. 3). This acute angulation might create tech-nical difficulties in the insertion of a permanentpacing lead.

Discussion

The venous phase of coronary angiograpbywas used to examine the anatomy of tbe venoussystem of the heart. The main advantages of thismethod, compared to retrograde venography,^'^° isits simplicity; cannulation of the CS is not neces-sary and, more importantly, the diameters and po-sitions of the tributaries of tbe CS remain morenatural. The limitation of this method is the less

2282 November 1998, Part II PACE, Vol. 21

ANGJOGRAPHY OF THE CORONARY SINUS AND ITS TRIBUTARIES

Figure 2. RAO 30° view of the coronary venous system. Note the tortuosity (a U-turn) of the greatcardiac vein (GCV) before it joins the CS.

accurate visualization of the small vessels, requir-ing the acute occlusion of the coronary sinus, aprocedure that clearly modifies the dimensions ofthe vessels and probably their angulation at pointsof anastomosis.

Figure 3. An RAO 30° view of the coronary venoussystem. Note the sharp angle of the left posterior vein(LPV) when it joins the great cardiac vein (GCV).

It has been reported that the diameter of theCS increases with the dimensions of the left ven-tricle.'^ Hovi'ever, to the best of our knowledge,the dependence of the diameter of the great car-diac vein on the dimensions of tbe left ventricle,two measurements statistically correlated in ourseries, has not been reported.

We have confirmed in vivo tbe anatomic find-ings of Aho^: the dimensions of the great and mid-dle cardiac veins are negatively correlated witbthe number of posterior veins. This may havepractical, though undesirable implications: themore tributaries, the less accessible they are witbavailable pacing leads.

Tbe main finding of this study was tbe widevariations in number and size of tbe so-called leftposterior veins (in fact, sometimes situated in alateral position). This finding is of interest with re-spect to insertion of permanent pacing leads. Tbeabsence of left posterior or lateral veins may prac-tically limit the ability to pace the left ventricleendovenously. However, this situation was foundin < 5% of patients. Two otber factors may com-plicate tbe insertion of tbe lead, namely tbe diam-eter of tbe vein and its angulation. It sbould, how-ever, be emphasized tbat diameters measuredangiographically in vivo are smaller than the onesmeasured by anatomists. Furthermore, the greatcardiac vein was always present and, in most

PACE. Vol. 21 November 1998, Part II 2283

GILARD, ET AL

cases, had a diameter > 2 mm, allowing the place-ment of a lead. The main obstacle in the catheter-ization of this vein np to the epicardial surface ofthe left ventricle is its tortuosity, with a completeU-turn visible in 13% of cases. The other reliablevein was the middle cardiac vein. Though its di-ameter is usually wide, its catheterization may notbe easy because of its 90° angulation at its junctionwith the CS and its proximity to the coronary os-tium, causing the lead to pull back before enteringthe middle cardiac vein. Furthermore, pacing theleft ventricle from this vein (or the posterior veinsclose to it) is probably not optimal since it maypace the diaphragm, in the absence of demon-strated bemodynamic improvement at this veryposterior site.

Limitations of the Study

The main limitations of this series is its ap-plication to "normal" or slightly dilated hearts.Whether these findings apply to markedly dilated

References

7.

Bartholinus T. Anatomia reformata. Ludg, Batav.,1651.Tandler J. Anatomie des Herzens, Jena., 1913.Tori G. Radiological visualization of the coronarysinus and coronary veins. Acta Radiol 1952; 36:405-415.Blanc JJ, Etienne Y, Gilard M, et al. Evaluation ofdifferent pacing sites in patients with severe car-diac failure. Results of an acute hemodynamicstudy. Circulation 1997; 96:3273-3277,Cazeau S. Ritter P, Lazarus A, et al. Multisite pac-ing for end-stage heart failure; Early experience.PAGE 1996; 19:1748-1757.McAlpine WA. Heart and coronary arteries.Berlin-Heildelberg-New York, Springer-Verlag,1975, pp. 188-191.Aho A. On the venous network of the human heart

hearts, in which left ventricular pacing is oftenconsidered, remains to be demonstrated.

Because of the technique used, the results ob-tained should not viewed as definitive anatomicdata. Only "the major" veins were visualized, notthe small ones. However, this is of little impor-tance since our study was designed with the in-sertion of a permanent electrode in mind.

Conclusion

Tbe coronary venous system presents muchwider interindividual variations than the arterialsystem, particularly as it pertains to the left poste-rior veins. The main conclusion of this angio-graphic in vivo study is that diameters and angula-tions of the CS tributaries may allow the insertionof commercially available leads [6 Fr, 2 mm) inabout 85% of cases inat least one vein, though per-haps not the more suitable one. To increase thissuccess rate, electrodes sbould be more adapted tothe anatomy of the coronary venous system.

and its arteriovenous anastomoses, Ann Med Ex-per Biol Fenniae 1950; 28:9-51.Rouviere H, Delmas A. Anatomie Humaine De-scriptive, topographique et fonctionnelle. Tome 2.Paris, Masson, 1985, pp, 155-156.Gensini GG, Giorgi S, Coskun O, et al. Anatomy ofthe coronary circulation in living man: Coronaryvenography. Circulation 1965; 31:778-784.Schumacher B. Tebbenjohanns J, Pfeiffer D, et al.Prospective study of retrograde coronary venogra-phy in patients with posteroseptal and left-sidedaccessory atrioventricular pathways. Am Heart J1995; 130:1031-1039.Potkin BN, Roberts WC. Size of coronary sinus atnecropsy in subjects without cardiac disease andin patients with various cardiac conditions. Am JGardiol 1987; 60:1418-1421.

10.

11.

2284 November 1998, Part II PACE, Vol. 21