THE ANATOMICAL RECORD 292:23–28 (2009) Microanatomy of Human Left Ventricular Coronary Veins 1 SARA E. ANDERSON,1,2 ALEXANDER J. HILL,3 AND PAUL A. IAIZZO1,2* Department of Biomedical Engineering, University of Minnesota, Minneapolis, Minnesota 2 Department of Surgery, University of Minnesota, Minneapolis, Minnesota 3 Medtronic, Inc., Minneapolis, Minnesota ABSTRACT We describe analyses of the microanatomy of major left ventricular veins, including their relationship to the myocardium. Immediately following fixation of six fresh human hearts, anterior interventricular veins (AIV), left marginal veins (LMV), posterior veins of the left ventricle (PVLV), and posterior interventricular veins (PIV) were sectioned in 5 mm intervals perpendicular to the veins’ length from base to apex. Slides were prepared, digitized, and analyzed; measurements were made of each vein’s wall thickness, circumference, distance between vein wall and myocardium, and distance between vein wall and closest artery. For analyses, based on the length of each vein, slides were grouped into three regions: basal (top third), mid (middle third), and apical (bottom third). Vein wall thicknesses and circumferences were significantly smaller (P < 0.05) in apical than basal regions in all veins. Vein wall thicknesses were significantly larger in the AIV and PIV than in the LMV and PVLV (P < 0.05). The AIV was significantly farther away (1.81–2.99 mm) from the myocardium than the other three veins (P < 0.05). Left ventricular venous microanatomy was quantified and analyzed. Variation in venous microanatomy, including distance between vein walls and excitable myocardium, could impact therapies involving the coronary venous system. Anat Rec, 292:23–28, 2009. Ó 2008 Wiley-Liss, Inc. Key words: veins; coronary; microanatomy; left ventricle; human INTRODUCTION MATERIALS AND METHODS The coronary venous system has been employed in a variety of methods to enhance cardiac therapies because of the venous system’s ‘‘dense meshwork with numerous interconnections’’ and absence of atherosclerotic disease effects (Mohl, 1994). Most recently, the coronary venous system has been targeted as a pacing lead implant site for cardiac resynchronization therapy via biventricular or left ventricular (LV) pacing. Specifically, because of increased clinical indications for cardiac resynchronization therapy relative anatomy of the LV coronary veins have become more clinically relevant because of their close proximity to LV myocardium. However, the specific microanatomy of these veins remains largely unexplored. Therefore, the purpose of this study was to perform detailed analyses of the microanatomy of the major LV veins in the human heart. Fresh human hearts (n 5 6) were obtained via the generous donation of individuals and their surviving family members, bequeathed through the University of Minnesota Anatomy Bequest Program. Each fresh heart (24–48 hr postmortem) was perfusion fixed in a pressurized perfusion fixation chamber (10% formalin). Ó 2008 WILEY-LISS, INC. *Correspondence to: Paul Iaizzo, Ph.D., Department of Surgery, University of Minnesota, B172 Mayo, MMC 107, 420 Delaware Street SE, Minneapolis, MN 55455. Fax: 612-6242002. E-mail: firstname.lastname@example.org Received 30 January 2008; Accepted 20 April 2008 DOI 10.1002/ar.20766 Published online 24 October 2008 in Wiley InterScience (www. interscience.wiley.com). 24 ANDERSON ET AL. Fig. 1. Heart model diagrams show the typical locations of the left ventricular veins: anterior interventricular vein along the anterior interventricular sulcus, the left marginal vein along the left margin of the heart, the posterior vein of the left ventricle along the posterolateral wall, and the posterior interventricular vein along the posterior interventricular sulcus. Perfusion fixation involved cannulating the superior vena cava, pulmonary trunk, aorta, and one pulmonary vein. The heart was submerged in a container of formalin, and the great vessels were connected to a formalinfilled upper chamber. Formalin was continuously pumped to the upper chamber and was forced by gravity into the heart through the cannulated vessels. The height of the upper chamber corresponded to physiologic pressures observed by the heart and thus fixed the hearts in an end-diastolic shape. After the hearts were perfusion fixed for at least 48 hr, the anterior interventricular veins (AIV), left marginal veins (LMV), posterior veins of the left ventricle (PVLV), and posterior interventricular veins (PIV) were dissected into blocks and then sectioned along the length of each vein in 5 mm intervals perpendicular to the veins’ length from base to apex (Fig. 1). Standard histological methods were used to prepare slides using Masson’s trichrome stains. Slides were then digitized (Super Coolscan1, Nikon, Melville, NY) and analyzed (Image-Pro1 Plus 4.1.0, Media Cybernetics1, Bethesda, MD). Measurements were made of each vein’s wall thickness and circumference and distances between the vein walls and the adjacent myocardium (Fig. 2). Vein wall thickness was measured at 12 positions around the circumference of the vein and then averaged to yield an average vein wall thickness for each slide. Vein circumference was measured by tracing the inner lumen of the vein walls. Distance between the vein walls and myocardium was measured in eight approximately equal increments along each vein wall to yield an average distance to the myocardium for each slide. Distance between the vein and the nearest artery was measured by choosing the shortest distance between the vein and the nearest artery. A single investigator performed all histological measurements to minimize variability in measurement technique. The total number of slides for a given vein was divided by three to yield basal (top third), mid (middle third), and apical (bottom third) groups for each vein. For instance, in one heart, the PVLV was sectioned into 19 slides, whereas in another heart, the PVLV was divided into 13 slides. For the first heart, the first six slides comprised the basal section, the next seven comprised the mid section, and the final six comprised the apical section, whereas for the second heart, the first four slides comprised the basal section, the next five, the mid section, and the final four, the apical section. Relative regional differences in vein wall thickness, vein circumference, distance to the myocardium, and distance to arteries were compared using analyses of variances (ANOVA); when the ANOVA was significant (P < 0.05), a Bonferroni multiple comparison correction was also performed. Correlation tests (Prism 3.01, GraphPad Software, San Diego, CA) were used to determine if vein wall thickness and vein circumference were correlated. All data are presented as mean 6 1 SD. RESULTS Five hearts had all four commonly described major LV veins; one heart did not have an LMV (Fig. 1). Two of the six hearts were considered normal or free of cardio- HUMAN CORONARY VENOUS MICROANATOMY 25 Fig. 2. Example of measurements made on a digitized slide of a posterior interventricular vein (PIV). Measurements include vein wall thickness, vein circumference, distance to the myocardium, and distance to the closest artery. myopathies (Table 1); however, no significant differences were observed between normal and diseased hearts. Vein length varied from 60 to 215 mm (n 5 23). Table 2 provides a summary of average vein lengths, average vein wall thicknesses, average vein circumferences, and average distances to the myocardium in each region of the four veins; Fig. 3 shows representative examples of each vein in both basal and apical regions. In general, AIV and PIV vein wall thicknesses were significantly larger when compared with LMV and PVLV vein wall thicknesses (P < 0.05). For example, in the basal region of one heart, the average vein wall thicknesses were 0.19 mm for the AIV and 0.15 mm in the PIV, in comparison with 0.09 mm for the LMV and 0.10 mm in the PVLV. Vein wall thicknesses in apical regions of all four veins were significantly smaller than those in basal regions (P < 0.05). Average vein wall thicknesses in apical regions ranged from 0.09 6 0.03 mm to 0.13 6 0.04 mm, in comparison to 0.11 6 0.03 mm to 0.17 6 0.04 mm in basal regions. The overall circumference of the PIV was significantly larger than all other veins (P < 0.05). More specifically, average vein circumferences for the mid region PIV were 11.19 6 3.68 mm (diameter 5 3.56 6 1.17 mm), whereas association circumference measurements were 7.07 6 3.27 mm (diameter 5 2.25 6 1.04 mm) in the AIV, 6.43 6 1.89 mm (diameter 5 2.05 6 0.60 mm) in the LMV, and 5.95 6 1.43 mm (diameter 5 1.89 6 0.46 mm) in the PVLV. Average vein circumferences in apical regions were significantly smaller (ranging from 4.57 6 1.83 mm to 8.13 6 4.27 mm; diameters from 0.73 6 0.58 mm to 2.59 6 1.36 mm) than in both mid and basal regions, which combined, ranged from 5.95 6 1.43 mm to 13.26 6 3.34 mm (diameters from 1.89 6 0.46 to 4.22 6 1.06 mm, P < 0.05, Table 2). The AIV was significantly farther away from the adjacent myocardium than the other three veins (P < 0.05). For example, in basal regions, the average distance between the AIV and the myocardium was 2.90 6 1.22 mm, whereas the average distances to the myocardium were 1.44 6 1.09 mm for the LMV, 1.60 6 2.16 mm for 26 ANDERSON ET AL. TABLE 1. Patient demographic information Age (years) Gender Weight (lbs) Heart weight (g) 53 M N/A n/a 63 F 250 600 64 68 F F 125 165 369 n/a 81 94 F F 105 140 697 655 Disease state Idiopathic cardiomyopathy, atrial fibrillation, hyperthyroidism, pulmonary hypertension, chronic anticoagulation Cardiac arrest, COPD, chronic emphysema (heavy smoker), lung cancer Leukemia CHF, COPD, appendectomy, Type II diabetes, hip replacement, hyperlipidemia, hypertension, hyperthyroidism, kidney removed, renal disease, renal failure Cerebral vascular accident, dehydration CHF, atrial fibrillation, hypertension, mitral valve disorder, osteoporosis COPD, chronic obstructive pulmonary disease; CHF, congestive heart failure. TABLE 2. Summary of left ventricular coronary veins in human hearts (n 5 6) AIV (n 5 6) Average length (mm) Average vein wall thickness (mm) Average vein circumference (mm)* Average vein diameter (mm) [computed from circumferences] Average distance to myocardium (mm)*** Average distance to nearest artery (mm)y Base Mid Apex Base Mid Apex Base Mid Apex Base Mid Apex Base Mid Apex 113 0.17 0.15 0.11 9.26 7.07 4.57 2.95 2.25 1.46 2.90 2.13 1.70 3.22 3.84 3.74 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 54 0.04 0.04 0.06 1.73 3.27 1.83** 0.55 1.04 0.58 1.22 1.54 1.64 1.51 1.90 1.36 LMV (n 5 5) PVLV (n 5 6) 103 0.13 0.12 0.09 6.28 6.43 4.80 2.00 2.05 1.53 1.44 1.23 1.46 3.11 2.37 2.89 88 0.11 0.10 0.11 6.77 5.95 6.39 2.15 1.89 2.04 1.60 1.00 1.42 2.47 2.36 2.67 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 51 0.07 0.04 0.03 1.76 1.89 1.43** 0.56 0.60 0.45 1.09 1.13 1.60 2.17 1.83 2.02 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 20 0.03 0.04 0.04 3.00 1.43 2.38** 0.95 0.46 0.76 2.16 0.96 0.75 1.45 0.72 1.17 PIV* (n 5 6) 89 0.16 0.14 0.13 13.26 11.19 8.13 2.15 3.56 2.59 1.31 0.41 0.56 2.14 1.43 1.53 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 19 0.10 0.05 0.04 3.34 3.68 4.27** 1.06 1.17 1.36 1.20 0.29 0.41 1.25 0.23 0.53 AIV, anterior interventricular vein; LMV, left marginal vein; PVLV, posterior vein of the left ventricle; PIV, posterior interventricular vein. *P < 0.05 PIV vs. AIV, LMV, and PVLV. **P < 0.05 vs. mid and basal regions within the same vein. ***P < 0.05 AIV vs. LMV, PIV, and PVLV. y P < 0.05 AIV vs. PIV. the PVLV, and 1.31 6 1.20 mm for the PIV. Distances between veins and the myocardium did not vary significantly by region (Table 2). The AIV was significantly farther away from the nearest artery than the PIV (P < 0.05). For instance, average distance between the nearest artery and the mid region was 3.84 6 1.90 mm and 1.43 6 0.23 mm for the AIV and PIV, respectively. Distance between the vein and the nearest artery did not significantly vary when compared by region. DISCUSSION To our knowledge, the microanatomy of the coronary venous system has not been extensively examined with respect to needs associated with current clinical practices. Although the sample size in this study was small, obtaining whole fresh human heart specimens is a rare occurrence. Additionally, the small sample size could account for no significant differences in any microanatomical measurements between normal and diseased hearts. This study provides two principal findings regarding LV venous microanatomy that could have important implications for LV coronary venous pacing lead placement: (1) vein circumferences increased from apical to basal regions and (2) the distances between a given vein and the myocardium did not vary significantly among regions. Smaller pacing thresholds are required to consistently capture myocardial excitation in apical positions in comparison to basal positions (Daubert et al., 1998; Huang et al., 2002). Additionally, depending on pacing lead design, implanting a lead more distally in a vein has the potential added benefit of increased long-term stability HUMAN CORONARY VENOUS MICROANATOMY Fig. 3. Representative examples of an anterior interventricular vein (AIV) in basal (A) and apical (B) regions, left marginal vein (LMV) in basal (C) and apical (D) regions, posterior vein of the left ventricle (PVLV) in basal (E) and apical (F) regions, and posterior interventricular vein (PIV) in basal (G) and apical (H) regions. Scale applies to all examples. 27 28 ANDERSON ET AL. (Hansky et al., 2002). Lower pacing thresholds and greater lead implant stability in more apical regions of coronary veins have been assumed to be associated with smaller vein circumferences. However, although distance between a given vein and the myocardium did not vary significantly among regions, in the two most commonly targeted veins for cardiac resynchronization therapy, the LMV and PVLV, it was noted that these veins were closer to the myocardium in the mid-ventricular region than in the basal or apical regions. Yet, although the apical distances from the myocardium were greater than the mid-ventricular distances in the PVLV, they were still smaller than the basal distances; this supports the finding of the reported lower thresholds seen in these implant locations (Huang et al., 2002) and would suggest that stable positions from the mid-ventricular to apical regions are preferred because of the proximity to the myocardium. When individual vein characteristics are examined for lead implant stability (e.g., vein circumference) and low pacing threshold (distance to myocardium), it is apparent that implants in the PIV would most likely be more stable in the apical region because of the relatively larger vein circumferences seen in the mid and basal regions. This position would also suggest low thresholds because of its relatively small vein to myocardium distance. In contrast, the data suggest that the preferred implant location in the PVLV would be the mid-ventricular region, because of the smaller vein circumference and smaller vein to myocardium distance, although apical placements would also be acceptable (Table 2). Knowledge of vein dimensions in these apical locations is important to consider in pacing lead design. Although the LV coronary veins are commonly utilized in cardiac resynchronization therapy, an optimal and universal implant site has not been identified and may need to be individually optimized (Gasparini et al., 2003; Albertsen et al., 2005; Gold et al., 2005). The purpose of this study was to explore the microanatomy of the LV coronary veins in an effort to better define anatomical features that could enhance left heart lead placement or other therapy delivery. Because of widely varying anatomy, it is important to continue to investigate the microanatomy of the LV veins to facilitate finding an optimal anatomical site for therapy delivery. Future studies should also focus on relative anatomical changes that occur with aging, disease, and/or body mass indices. Nevertheless, this dataset provides several unique insights on general coronary venous anatomy. ACKNOWLEDGMENTS The authors sincerely thank the University of Minnesota Anatomy Bequest Program for their assistance with this study and the generous individuals and families who, by donating their hearts for research, made this study possible. They also thank Louanne Cheever for preparing the slides, Monica Mahre for assistance with manuscript preparation, and Jason Quill for assistance with the figures. LITERATURE CITED Albertsen AE, Nielsen JC, Pedersen AK, Hansen PS, Jensen HK, Mortensen PT. 2005. Left ventricular lead performance in cardiac resynchronization therapy: impact of lead localization and complications. Pacing Clin Electrophysiol 28:483–488. Daubert JC, Ritter P, Le Breton H, Gras D, Leclercq C, Lazarus A, Mugica J, Mabo P, Cazeau S. 1998. Permanent left ventricular pacing with transvenous leads inserted into the coronary veins. Pacing Clin Electrophysiol 21:239–245. Gasparini M, Mantica M, Galimberti P, Bocciolone M, Genovese L, Mangiavacchi M, Marchesina UL, Faletra F, Klersy C, Coates R, Gronda E. 2003. Is the left ventricular lateral wall the best lead implantation site for cardiac resynchronization therapy? Pacing Clin Electrophysiol 26:162–168. Gold MR, Auricchio A, Hummel JD, Giudici MC, Ding J, Tockman B, Spinelli J. 2005. Comparison of stimulation sites within left ventricular veins on the acute hemodynamic effects of cardiac resynchronization therapy. Heart Rhythm 2:376–381. Hansky B, Guldner H, Vogt J, Minami K, Tenderich G, Horstkotte D, Korfer R. 2002. Coronary vein leads for cardiac pacing in patients with tricuspid valve replacement. Thorac Cardiovasc Surg 50:120–121. Huang J, Walcott GP, Killingsworth CR, Smith WM, Kenknight BH, Ideker RE. 2002. Effect of electrode location in great cardiac vein on the ventricular defibrillation threshold. Pacing Clin Electrophysiol 25:42–48. Mohl W. 1994. Basic considerations and techniques in coronary sinus interventions. In: Mohl W, editor. Coronary sinus interventions in cardiac surgery. Austin: RG Landes Company.