Clinical Anatomy 8:347-351 (1995)

The Size of the Anterior Spinal Artery in Relation to the Arteria Medullaris Magna Anterior in Humans

TAO ZHANG, LYNN HARSTAD, JOSEPH E. PARISI, AND MICHAEL J. MURRAY The Departments of Anesthesiology (T.Z., L.H., M.J.M.) and Laboratory Medicine and Pathology (J.E.P.), Mayo Clinic

and Mayo Foundation, Rochester, Minnesota

We studied the anterior spinal artery (ASA) in 25 cadaveric human spinal cords to determine the cross-sectional area of the ASA cephalad and caudal to the entry of the arteria medullaris magna anterior (AMMA). Spinal cords were removed en bloc and latex was injected into the AMMA. The preparations were then fixed in formalin, embedded in paraffin, sectioned, mounted, and stained with hematoxylin and eosin. The diameter and cross-sectional area of the ASA 1 cm above and 1 cm below the entry of the AMMA were measured. The mean radius of the ASA above the entry of AMMA was 0.14 +- 0.03 mm compared to 0.28 2 0.05 mm below the entry of the AMMA ( P < 0.001). According to Poiseuille's equation, the resistance to blood flow in the MA cephalad to the AMMA would be 14.8 times greater than the resistance to blood flow caudal to the AMMA. This resistance could affect the distribution of blood flow in the distal spinal cord whenever flow in the AMMA or distal ASA is altered. 0 1995 Wiley-Liss, Inc.

Key words: spinal cord, vascular resistance, postoperative paraplegia, anatomy

INTRODUCTION Paraplegia, an infrequent but serious complication of

surgical procedures on the thoracoabdominal aorta, de- velops because of ischemic reperfusion injury to the distal spinal cord. An understanding of the vascular anatomy of the distal spinal cord, therefore, is impor- tant in the understanding and prevention of paraplegia. Few functional anastomoses exist between the three main vascular territories along the vertical axis of the spinal cord: the cervicothoracic, the midthoracic (com- prising the fourth to eighth thoracic segments), and the thoracolumbar (comprising the remainder of the cord) regions (Aminnoff, 1976). In the thoracolumbar-vascu- lar bed, one of the anterior medullary arteries is domi- nant in caliber and was, therefore, termed the arteria medullaris magna anterior (AMMA) by Adamkiewicz (1882). T h e AMMA usually enters along a nerve root on the left side of the body, with its origin most often between the T 9 and T12 spinal-cord segments (62%), but occasionally lower in the lumber region (26%) or more cephalad, between T 6 and T8 (12%) (Jellinger,

1966; Thron, 1988). Ligature of the AMMA without aortic occlusion or alteration in hemodynamics results in a 72% rate of paraplegia in pigs (Wadouh et al., 1984), a result that underscores the importance of this artery in supplying the distal spinal cord. We believe, based on animal studies (Svensson et al., 1986b; Mur- ray et al., 1992), that in addition to the AMMA, the size of the anterior spinal artery (ASA) caudal and cephalad to the entry of the AMMA may also affect the distribu- tion of blood flow within the spinal cord. In this study, we measured the ASA diameter, radius, and cross- sectional area 1 cm above and 1 cm below the entry of the AMMA.

METHODS We studied 25 spinal cords from adult cadavers of

both sexes and ascertained the cause of death for each

Received for publication August 2, 1994; revised December 28, 1994. Address reprint requests to Dr. Michael J . Murray, Mavo Clinic, Rochester, MN 55905.

0 1995 Wiley-Liss, Inc.

348 Zhang et al.

cadaver. Cases with known spinal pathology had been previously excluded.

At the postmortem examination, the anterior ver- tebral column was exposed and the attached muscles were divided. T h e vertebral bodies were removed using a Stryker saw, and the spinal cord was exposed and removed from the medullary to the cauda equina levels. T h e AMMA, defined as the largest of the medullary arteries, was identified and its level of entry recorded. Latex (650 gm of Barosperse [Diagnostic Products Division 0991-27, Mallenckrodt, Inc., St. Louis, MO], 15 gm of gelatin [Sigma 6-9382, Sigma Chemical Co., St. Louis, MO], and 3 gm of Thymol [Fisher Scientific #T 185-100, Springfield, NJ] in 500 m L distilled water) was injected into the AMMA so that it filled the ASA in both directions. T h e specimens were then immersed in a 15% formaldehyde solution for 2 15 days. By having the ASA filled with latex, we hoped to avoid significant shrinkage of the artery; any shrinkage, though, would not differentially affect blood vessels and would, therefore, not affect our final anal- ysis and conclusions. T h e spinal cord 1 cm above and 1 cm below the site of entry of the AMMA was sec- tioned and a specimen obtained. These are points at which flow is laminar and without eddy currents (Nichols and O'Rourke, 1990) and, visually, representative of the cephalad and caudal ASA. Specimens were embed- ded in paraffin, sectioned, mounted, and stained with hematoxylin and eosin. T h e inner radius, diameter, and area of the ASA were then measured using an Olympus microscope with Jandel Video Analysis Soft- ware (JAVA'") (Jandel Scientific, Madera, CA). This

system has ?z 1% precision (personal communication, Robert S. Schwartz, M.D., Mayo Clinic, Rochester, MN).

Statistics: Data were expressed as the mean 2 stan- dard deviation when appropriate. Measurements of the radius, diameter, and area of the ASA above and below the AMMA entry site were compared by a paired t-test. We chose a value of P < 0.05 to denote statistical significance.

RESULTS T h e ASA was easily recognizable as an oval structure

lying in the midline on the spinal cord's anterior surface (Figs. 1, 2). T h e most common level of entry for the AMMA was the ninth (n = 8), followed by the eighth (n = 6) thoracic vertebra, with a range from the seventh to the twelfth thoracic vertebra (Fig. 3). T h e major and minor diameter of the ASA (Fig. 4) were 0.34 2 0.09 mm (X ? SD) and 0.23 IT 0.07 mm, respectively, measured 1 cm above the entry of the AMMA, and 0.67 k 0.09 mm and 0.44 ?z 0.12 mm, 1 cm below the entry level ( P < 0.001 when comparing the above and below measurements).

T h e mean area of the ascending ASA was 0.07 k 0.03 mm', 1 cm above, and 0.25 k 0.09 mm', 1 cm below the level of entry of the AMMA ( P < 0.001). We calculated the mean radius from the measured area using the formula: radius = vA/r , where A was the measured area. T h e mean radius was 0.14 2 0.03 mm vs. 0.28 2 0.05 mm, 1 cm above and below the AMMA junction, respectively ( P < 0.001) (Table 1).

Fig. 1. A representative distal segment of a spinal cord with the anterior spinal artery (ASA) in the midline. The ASA is smaller cephalad to and larger caudal to the entry of the arteria medullaris magna anterior (AMMA) (arrow).

The ASA in Humans 349

Fig. 2. The anterior spinal artery was oval both above (left) and below (right) the entry level of the arteria medullaris magna anterior. Pictures were taken with the same focal length and magnifica- tion ( x 25).

DISCUSSION T h e AMMA provides the main blood supply to the

T h e AMMA plays an important role in perfusing the distal spinal cord (Wadouh et al., 1984), but other factors may be equally important. T h e etiology of spinal-cord injury in patients undergoing operations on the thoracic aorta is poorly understood. T h e incidence of paraplegia varies from 0.4% to 40.0% depending on the age of the patient, extent of disease, and length of time that the thoracic aorta is crossclamped (Murray et al., 1992). Although the etiology of paraplegia is most likely multifactorial, anatomical features no doubt play an important role (Dommisse, 1974). Hypoperfusion, or temporary or permanent interruption of flow to the distal cord or to the lumbosacral roots or plexus, may lead to the development of a neurologic deficit (Gloviczki et al., 1991).

T7 T8 T9 T10 T 1 1 T12

Level of Spinal Cord

thoracolumbar region of the spinal cord, supplying nearly 70% of the blood to this area (Barth, 1990). Jellinger (1966) reported that 66% of the time, the AMMA joined the ASA with the T9 to T12 nerve roots (most frequently with the T 9 root) and 28% of the time with the L1 to L3 nerve roots. In another study, the AMMA was the most-caudal blood supply to the spinal cord in two thirds of cases, and the AMMA was the only medullary artery below T6 in one third of the cases (Rodriguez-Baeza et al., 1991). In our study, the AMMA originated between T 7 and T12, most often (8 of 25) at T9 .

Dommisse and Enslin (1970) reported four cases of paraplegia in a series of 68 patients undergoing surgical correction of a spinal deformity. Although only 18 of the 68 patients underwent operations involving the T5 through T 9 region of the spinal cord, three of the four cases of paraplegia occurred in this group. In these three patients, the AMMA rose between T5 and T9; in the fourth patient, the AMMA originated below the T10 level. In the remaining 50 patients, only one per- son in this latter group developed paraplegia, and this patient was noted preoperatively to have a partial neu- rologic deficit.

According to Poiseuille’s equation (Q = 7rr4 (Pl-P2)/8Lrl, where Q = flow, r = radius, P = pressure, L = length, and q = viscosity constant), flow is inversely proportional to the fourth power of the radius. Though Poiseuille’s equation applies to New- tonian fluids, one can use the derived information to gain insight into what might occur with blood flow. Ohm’s law states that R = P/Q, where R = resistance, P = pressure, and 0 = flow. Based on Ohm’s law, the

~ i ~ . 3. The level of the spinal cord (n = 25) at which the resisiance in the ASA cephalad to the AMMA would be arteria medullaris magna anterior entered the anterior spinal artery.

14.8 times greater than that in the caudal ASA. Because little of the blood entering the spinal canal in the

350 Zhang et al.

Fig. 4. A diagram of the system used to measure the dimensions of the anterior spinal artery (ASA). O n the ASA shown on the cathode ray tube, we have displayed the “major” and “minor” diameters (axes) and the oval traced on the internal diameter of the vessel that was used to calculate the cross-sectional area

cervical region normally flows the length of the ASA to reach the lower thoracic or lumbar regions (Gillilan, 1962), the lower thoracic cord is mostly dependent on blood flowing cephalad in the ASA. This results in risk to the lower cord when flow through the AMMA is diminished.

Svensson et al. (1986a) reported that a shunt to perfuse the distal aorta in baboons undergoing aortic crossclamping distal to the left subclavian artery and above the diaphragm did not entirely protect against the development of paraplegia, and that little of the shunted blood perfused the lower thoracic spinal cord. T h e paraplegic baboons had an inadequate thoracic spinal-cord blood flow, despite an adequate lumbar

‘IIABLE 1. Dimensions of the anterior spinal artery cephalad and caudal to the junction of the arteria medullaris makna anterior (AMMA)”

Length Measured 1 cm above AMMA 1 cm below AMMA

Major diameter (mm) 0.34 2 0.09’ 0.67 2 0.09 Minor diameter (mm) 0.23 2 0.07’ 0.44 ? 0.12 Mean diameter (mm) 0.28 k 0.07’ 0.54 ? 0.09 Cross-section area (mm‘) 0.07 ? 0.03’ 0.25 2 0.09 Radius from area (mm) 0.14 2 0.03’ 0.28 2 0.05

“Mean ? SD, n = 25. ‘P < 0.01 compared to measurement taken below the AMMA.

spinal-cord blood flow during distal perfusion. Obvi- ously, if the AMMA arises from the occluded segment of the aorta, then perfusion of the distal aorta beyond the lower clamp is unlikely to have a protective effect (Svensson et al., 1986b). In our study, we demon- strated that even if the AMMA were perfused by a shunt, the distal thoracic cord may still be at a higher risk of ischemic injury because the resistance to blood flow cephalad to the AMMA is 14.8 times greater than is blood flow caudal to the AMMA. In order to continue adequate thoracic spinal cord perfusion, one would have to increase ASA perfusion cephalad to the site of entry of the AMMA.

ACKNOWLEDGMENTS T h e authors wish to acknowledge and thank Ms.

Robin Williams for her secretarial skills, Dr. M. Kumar for his technical assistance, and the staff of the gross- tissue laboratory for their technical assistance. We also wish to thank Catherine Friederich for editorial assis- tance.

REFERENCES Adamkiewicz, A. 1882 Die Blutgefasse des menschlichen

Ruckenmarkes: 11. Die Gefasse der Ruckenmarksober-

The ASA in Humans 351

flache, Sitzungsb. d. k. Adad. Wissensch. Math.-Naturw.

Aminnoff, M. J. (ed.) 1976 Spinal Angiomas. Oxford: Blackwell,

Barth, J.A. 1990 Feeder arteries, longitudinal arterial trunks and arterial anastomoses of the lower human spinal cord. Zent. bl. Neurochir., 51: 18 1 - 184.

Dommisse, G.F. 1974 T h e blood supply of the spinal cord. J. Bone Joint Surg., .568:225-235.

Dommisse, G.F., and T.B. Enslin 1970 Hodgson’s circum- ferential osteotomy in the correction of spinal deformity. J. Bone Joint Surg., .52B:778.

Gillilan, L.A. 1962 Blood vessels, meninges, cerebrospinal fluid: Blood supply to the central nervous system. In Correla- tive Anatomy of the Nervous System. E.C. Crosby, T. Humphrey, E.W. Lauer (eds.). New York: Macmillan, pp. 550-579.

Gloviczki, P., S. A. Cross, A. W. Stanson, S. W. Carmichael, T. C. Bower, P. C. Pairolero, J.W. Hallett, B.J. Toomey, and K. J. Cherry, Jr. 1991 Ischemic injury to the spinal cord or lumbosacral plexus after aorto-iliac reconstruction. Am. J. Surg., f62:131-136.

Jellinger, K. 1966 Zur Orthologie und Pathologie der Rucken- marksdurchblutung. New York: Springer-Verlag, p. 21.

Murray, M.J., T.C. Bower, and S.W. Carmichael 1992 Anatomy of the anterior spinal artery in pigs. Clin. Anat., .5:452-457.

CL, 85: 101-130.

pp. 5-17.

Nichols, W.W., and M.F. O’Rourke. 1990 T h e nature of flow of a fluid. In McDonald’s Blood Flow in Arteries: Theoretic, Experimental and Clinical Principles. W. W. Nichols, M.F. O’Rourke (eds.). Philadelphia: Lea and Febiger, pp. 12-53.

Rodriguez-Baeza, A., A. Muset-Lara, M. Rodriguez-Pazos, and J.M. Domenech-Mateu 1991 T h e arterial supply of the hu- man spinal cord: A new approach to the arteria radicularis magna of Adamkiewicz. Acta. Neurochir., fO9:57-62.

Svensson, L.G., P. Klepp, and R.A. Hinder 1986a A spinal cord anatomy of the baboon: Comparison with man and implica- tions for spinal cord blood flow during thoracic aortic cross- clamping. Suid-Afrikaanse Tydskrif vir Chirurgie, 24:

Svensson, L.G., E. Rickards, A. Coull, G. Fogers, C.J. Fim- mel, and R.A. Hinder 1986b Relationship of spinal cord blood flow to vascular anatomy during thoracic aortic cross- clamping and shunting. J . Thorac. Cardiovasc. Surg., 9f: 71-78.

Thron, A.K. 1988 Vascular Anatomy of the Spinal Cord. New York: Springer-Verlag, pp. 8-12.

Wadouh, F., E. Lindemann, C.F. Arndt, R. Hetzer, and H.G. Borst 1984 T h e arteria radicularis magna anterior as a decisive factor influencing spinal cord damage during aortic occlusion. J. Thorac. Cardiovasc. Surg., 88: 1-10.

32-34.