The Dominant Anterior Thoracic Artery of the Spinal Cord

Variations of the arterial supply to the anterior thoracic cord are important due to their implications for ischemic events as well as surgical and endovascular procedures. Radiculomedullary arteries (RMAs) represent the dominant arterial supply to the spinal cord and are the sole segmental supply to the anterior spinal artery (ASA) along the ventral cord.¹ The ASA ranges from 0.2 to 0.8 mm and has an inconstant caliber along its length.^2⇓-4 The fluctuating size is related to the variable number of RMAs specifically providing arterial supply to the ventral cord.^4,5 The number, caliber, and entrance level of the anterior RMAs have been previously studied^6⇓-8 and ranges in number from 2–14.²

Dominant radiculomedullary and radiculopial arteries such as the artery of lumbar enlargement (ALE),^2⇓-4 the artery of cervical enlargement (ACE),¹ and the great posterior radiculomedullary artery⁸ are well-documented. Historically, 2 RMAs have received the most consideration. In the lower thoracic or upper lumbar region, the dominant and most important radiculomedullary supply is the ALE, otherwise known as the arteria radiculomedullaris magna or, more commonly, the artery of Adamkiewicz.^2⇓-4 The ALE arises between T8 and L2 in 75% of individuals with a diameter ranging between 0.5 and 1.2 mm.^2,3 The largest anterior cervical radiculomedullary branch is generally in the region of C5 through C6 and has been termed the ACE.¹ Although dominant RMAs such as the ALE and ACE are well-documented, variability exists as to the size, number, and location of other RMAs contributing supply to the ASA. Because RMAs other than the ALE have a considerably smaller diameter, they are difficult to identify on imaging, including angiography. Traditionally, RMAs arising from upper thoracic segmental arteries are generally regarded as branches only rarely present and therefore of little clinical significance. This notion has led to the commonly taught and widely held belief that the anterior thoracic arterial system represents a watershed territory.

A spinal angiography study conducted in 50 individuals identified a significant upper thoracic anterior RMA distinct from the ALE in 86% of patients.⁹ Due to the relatively constant nature of this artery, the author suggested naming this vessel the artery of von Haller. This was proposed in honor of Albrecht von Haller, who documented the arterial supply to the ASA including this vessel as well as the ALE in 1754,^5,10 more than a century before Albert Adamkiewicz^11,12 and Kadyi.¹³ However, instead of the artery of von Haller, we opted to refer to this vessel as the dominant anterior thoracic artery (DATA) so that the name could provide easy recognition and support more reliable terminology. The DATA has attracted little attention in the literature, and its importance is not widely known or taught. In addition, its angiographic identification is quite difficult due its caliber as well as obscuration by overlying structures, including osseous and muscular arterial branches as well as the trachea.⁹ The aim of this anatomic study was to analyze the arterial supply to the ventral spinal cord and determine the prevalence of the DATA along with its morphometric characteristics in cadaveric specimens.

RESULTS

No vascular malformations, fistulas, or other vascular abnormalities were identified in any of the specimens.

In only 3 specimens (3/50) was the DATA not identified. At least 1 DATA was evident in 94% of spinal cord specimens (47/50). A single DATA was present in 66% of the specimens (33/50) (Fig 1). Two DATAs were identified in 26% (13/50) (Fig 2). Three DATAs were identified in a single (1/50) specimen (2%), all arising on the left at the level of T3, T7, and T9. The laterality and spinal level of the DATA are summarized in Fig 3. A DATA was most frequently identified at the level of T4, with 15 left- and no right-sided specimens identified at this level. The average DATA size was 0.446 mm, with a range of 0.300–0.759 mm on the left and 0.270–0.569 mm on the right.

• Download figure
• Open in new tab
• Download powerpoint

FIG 1.

Single DATA. The anterior view of the spinal cord during arterial injection (A) clearly shows a T9 ALE (black arrow) as well as a single T4 DATA (white arrow), the most common level in our analysis. Three other specimens after fixation (B–D) clearly demonstrate a single dominant upper thoracic anterior RMA supplying the ASA, consistent with a DATA (white arrows).

• Download figure
• Open in new tab
• Download powerpoint

FIG 2.

Two DATAs. The anterior view of the spinal cord clearly reveals evidence of the ALE (black arrow) as well as 2 left-sided DATAs consistent with 2 separate DATAs (white arrows) supplying the ASA.

• Download figure
• Open in new tab
• Download powerpoint

FIG 3.

Distribution, number, and side of the ACE, DATA, and ALE.

The ALE was identified between T3 and L2 in all 50 spinal cords (100%). A left-sided ALE was present in 44 specimens, and a right-sided ALE was present in 6 specimens. The most proximal ALE was evident at the level of T3 on the left, and the most distal, at the level of L2 on the left. The side and spinal level of ALEs are summarized in Fig 3. On average, the diameter of the ALE was 0.83 mm, with the smallest caliber ALE measuring 0.49 mm and the largest measuring 1.21 mm.

The ACE was identified in 84% of specimens (42/50). A single ACE was present in 20 specimens; 2 separate ACEs, in 12 specimens; 3 ACEs, in 6 specimens; and 4 ACEs, in 4 specimens. A total of 78 ACEs were, therefore, documented, with 38 on the left and 40 on the right. The most proximal ACE was evident at the level of C2, and the most distal, at the level of C8. The side and spinal level of the ACEs are summarized in Fig 3. On average, the diameter of the ACE was 0.49 mm, with the smallest caliber ACE measuring 0.13 mm and the largest measuring 0.73 mm. In the 8 specimens in which no ACE was identified, the DATA originated at T2 in 2 specimens, T3 in 3 specimens, T4 in 1 specimen, and T8 in 2 specimens.

DISCUSSION

Our results demonstrate that a DATA is present in most individuals (94%). Our findings are similar but more robust than those in the study of Gailloud,⁹ in 2013, in which a total of 43 of 50 patients (86%) had at least 1 RMA branch between T3 and T7, with a single anterior branch in 68% (34/50), 2 anterior branches in 18% (9/50), and none with 3 anterior branches. In contrast, we found a single DATA in 66% (33/50), 2 in 26% (13/50), and 3 in 2% (1/50). Improved detection in our cadaveric findings is likely due to increased sensitivity and accuracy because the upper thoracic RMAs may be obscured by overlying osseous or muscular branches on spinal angiography.

Although the caliber of the DATA did not approximate that of the ALE, the vessels were truly dominant and easily distinguished from the tiny caliber of the radicular arteries that accompany spinal nerves. Of note, no absolute cutoff measurement was used to establish the presence of a DATA because establishing the presence of a dominant arterial feeder is a relative process specific to each individual. For instance, paucisegmental and plurisegmental ASA circulations are associated with markedly differing calibers of the ALE.

Although naming RMAs is necessarily arbitrary in nature, the presence of an ALE, ACE, and DATA in 42 of our specimens validates the naming of each of these vessels. In other words, the dominant upper thoracic RMA in a specimen clearly demonstrating an ALE and an ACE is best regarded as a DATA rather than a high-origin ALE or a low-origin ACE. Furthermore, in the 8 cases in which we did not identify an ACE, the DATA distribution from T2 through T8 did not demonstrate a high thoracic clustering, arguing against naming all these vessels as low-origin ACEs. In particular, it is difficult to justify the classification of the cases with no ACE and a DATA at either T4 or T8 as a low ACE variant instead.

These results are contrary to the notion that in most individuals, there is a thoracic spinal cord watershed zone at risk of ischemia. The presence of a DATA in most individuals counters the erroneous belief that there is a poorly arterialized thoracic segment. Instead, the concept that in most individuals there are at least 3 dominant RMAs such as the ALE, ACE, and DATA should be endorsed and taught during anatomic discussions (Fig 4). Our results have implications for reducing the risk of arterial spinal and aortic surgical and endovascular procedures. In particular, our findings are important for addressing vascular malformations of the spinal cord. The degree of involvement of the DATA in spinal cord pathology has yet to be studied, and we believe this would be a fruitful area for future research.

• Download figure
• Open in new tab
• Download powerpoint

FIG 4.

Dominant RMAs. In most individuals, at least 3 dominant RMAs exist and include the ALE, ACE, and DATA as illustrated and labeled above. The DATA is shown on the left at the level of T4, the most common side and level in our study. a. indicates artery.

The number of specimens we examined supports the reliability of our vessel distribution, laterality, and diameter determinations. However, because most of our anatomic specimens were from male cadavers, we cannot exclude the unlikely possibility that sex differences exist. In addition, the possibility of racial variations cannot be excluded on the basis of this study.

• Received August 31, 2022.
• Accepted after revision October 25, 2022.

• © 2023 by American Journal of Neuroradiology