Address correspondence and reprint requests to Stephen J. Baines, at the above address. E-mail: email@example.com
The anticlinal vertebra is a point in the caudal thoracic vertebral column at which vertebral anatomic features change. It may be used as a point of reference in diagnostic imaging studies. However, its exact definition and anatomic location are not consistent between anatomic texts. The position of the anticlinal vertebra was evaluated radiographically in 100 dogs, using three different definitions of the anticlinal vertebra. Definition P referred to perpendicularity of the spinous proceses, definition V referred to verticality of the spinous processes, and definition J referred to the orientation of the intervertebral joints. The anticlinal vertebra was assessed to be T11 in 85%, 75%, and 89% of dogs for definitions P, V, and J, respectively. Combining definitions improved the likelihood that the vertebra assessed to be the anticlinal vertebra was T11, if the combination included definition J, but at the expense of loss of sensitivity. The degree of agreement between the definitions ranged from fair to poor. The level of agreement between definitions P and J was improved for small and large dogs and the level of agreement between definitions V and J was improved for medium and large dogs. Compared with large breed dogs, small breed dogs were more likely to have T10 described as the anticlinal vertebra. The presence of transitional vertebra did not affect the position of the anticlinal vertebra. It should not be assumed that the anticlinal vertebra is T11 in all dogs. Care should be taken to define this anatomic feature accurately when using it as an anatomic landmark.
The localization of lesions in diagnostic images is often dependent on the assessment of their relationship with other structures. These points of reference should be clearly defined, easily recognized, present in all animals, and have minimal intersubject variation.
Progressing caudally in the thoracic spine, there are two main morphologic vertebral alterations: the orientation of the spinous processes of the caudal thoracic vertebrae changes from a caudal inclination to a cranial inclination; and there is a change in the orientation of the intervertebral synovial joints from the dorsal plane to the sagittal plane.
The term diaphragmatic vertebra has been used to describe the thoracic vertebra whose cranial intervertebral synovial joints lie in the dorsal plane and whose caudal intervertebral synovial joints lie in the sagittal plane.1 The term anticlinal vertebra has been applied to a particular vertebra with certain anatomic features, which pertain primarily to the slope of the spinous processes. However, the exact definition of the anticlinal vertebra varies. Some authors consider the anticlinal vertebra to be the thoracic vertebra in which the spinous process is perpendicular to the vertebral body or the long axis of the spine.2–4 Others describe the anticlinal vertebra as the thoracic vertebra with a vertical spinous process.5–9 The intervertebral synovial joints have also been used to localize the anticlinal vertebra, and it is suggested that the anticlinal vertebra is the one immediately caudal to the diaphragmatic vertebra, as defined above.1,7 In addition to the inconsistency in the anatomic definition, reports of the exact anatomic location of the anticlinal vertebra are inconsistent, and it has been described as both the tenth8 and the eleventh1,3,5,6,8 thoracic vertebra. The aim of this study was to review the anatomic features and the location of the canine anticlinal vertebra according to the definitions provided by anatomic texts.
Materials and Methods
Lateral and dorsoventral/ventrodorsal radiographs of the thorax and abdomen—on which the cervical and thoracolumbar vertebrae were visible—of 100 consecutive dogs were examined by a single board-certified radiologist (E.A.B.). Criteria for inclusion were as follows: visualization of the sixth cervical vertebra on the lateral radiograph to act as a localizer for the rest of the vertebral column, presence of the entire spinous processes of all thoracic vertebrae, and absence of clinical signs of disease of the vertebral column or spine. The radiographic appearance and location of the anticlinal vertebra according to three separate definitions (Table 1 and Fig. 1) were recorded. For definition P (perpendicular), the anticlinal vertebra was that thoracic vertebra whose spinous process most closely approximated a line perpendicular to the long axis of the vertebral column at that site. For definition V (vertical), the anticlinal vertebra was that vertebra where there was a change in the slope of the spinous processes, such that the caudal border of the spinous process sloped cranially and the cranial border of the spinous process sloped caudally, or was vertical. The spinous processes of the vertebrae cranial to this sloped caudally, and those of vertebrae caudal to this sloped cranially. For definition J (joint), the anticlinal vertebra was the first thoracic vertebra in which both intervertebral synovial joints were in the sagittal plane. The vertebra cranial to this had cranial intervertebral joints in the dorsal plane and caudal intervertebral joints in the sagittal plane.
Table 1. Definitions of the Anticlinal Vertebra Used
Definition P: Spinous process perpendicular to spine
The first three or four spinous processes are about equal in length. Caudal to this, they gradually become shorter to the 10th and then remain equal. The caudal slope is most marked in the ninth and 10th. The 11th is practically vertical, and the last two slope slightly cranial
Beginning at T11, the plane of articulation is sagittal, that is, the cranial processes clasp the lateral surfaces of the caudal processes of the preceding vertebra, and the joint line is clear on the dorsoventral view
The 10th thoracic vertebra is known as the diaphragmatic vertebra, and possesses horizontally placed articular surfaces on the cranial aspect and vertically placed articular surfaces on the caudal aspect
The breed and presence of any other vertebral anomalies were also recorded. The purebred dogs were classified, according to The Kennel Club definitions for breed groups10 as small, medium, or large; giant breeds were grouped with large breeds. Radiographs from 100 dogs were examined comprising a range of breeds. There were 86 purebred dogs and 13 crossbred dogs. The purebred dogs comprised 23 small breed dogs, 16 medium breed dogs, and 47 large and giant breed dogs.
The measure of agreement between the three definitions was assessed using the kappa statistic and was interpreted according to a modification of Landis and Koch11 as described by Altman.12 Differences between categorical variables were investigated using the χ2-statistic and Fisher's exact. significance was set at P<0.05.
For definition P, the anticlinal vertebra was T10 in five dogs, T11 in 85 dogs, and T12 in 10 dogs. For definition V, the anticlinal vertebra was T10 in 19 dogs, T11 in 75 dogs, and T12 in six dogs. For definition J, the anticlinal vertebra was T10 in one dog, T11 in 89 dogs, and T12 in 10 dogs (Table 2). The level of agreement among the various definitions assessed by the kappa statistic was moderate for P and V, κ=0.424 [0.198–0.650: 95% confidence interval (CI)]; fair for P and J, κ=0.399 (0.107–0.691: 95% CI); and poor for V and J, κ=0.137 (−0.134 to 0.409: 95% CI).
Table 2. Position of the Anticlinal Vertebra According to the Definitions
Group of Animals
All Dogs (n)
Normal Number (n)
Normal Morphology (n)
Normal Number and Morphology (n)
The percentage of animals whose anticlinal vertebra was assessed to be at T10, T11, or T12. The data are presented for all animals and for animals without various anatomic abnormalities.
The usefulness of the three definitions, either singly or in combination, in predicting whether the vertebra assessed to be the anticlinal vertebra was T11, was determined (Table 3). For single definitions, the vertebra identified as the anticlinal vertebra was T11 for definition P in 85% of dogs, for definition V in 75% of dogs, and for definition J in 89% of dogs. Where two definitions identified the same vertebra, this vertebra was T11 for definitions P and V in 70/80 dogs (87.5%), for definitions P and J in 80/86 dogs (93.0%), and for definitions V and J in 68/72 dogs (94.4%). Where all three definitions identified the same vertebra, this was T11 in 65/69 (94.2%) dogs.
Table 3. Prediction of T11: Likelihood of Whether the Vertebra Assessed by the Definition Given will be T11, Either as a Single Definition, or when Two or Three Definitions Agree
Number of Vertebra Assessed
Number of Vertebra which are T11
Vertebrae that are T11 (%)
P and V
P and J
V and J
P, V, and J
The agreement among the three definitions was assessed for each of the three breed size groups. For small dogs, the level of agreement among the various definitions was fair for P and V, κ=0.369 (−0.021 to 0.759: 95% CI); moderate for P and J, κ=0.436 (−0.067 to 0.938: 95% CI); and poor for V and J, κ=0.023 (−0.394 to 0.440: 95% CI). For medium dogs, the level of agreement among the various definitions was fair for P and V, κ=0.310 (−0.191 to 0.812: 95% CI); fair for P and J, κ=0.333 (−0.347 to 1.013: 95% CI); and fair for V and J, κ=0.281 (−0.329 to 0.891: 95% CI). For large dogs, the level of agreement among the various definitions was fair for P and V, κ=0.359 (−0.120 to 0.838: 95% CI); good for P and J, κ=0.636 (0.237–1.035: 95% CI); and fair for V and J, κ=0.232 (−0.219 to 0.683: 95% CI).
When subdivided according to breed size, the level of agreement between definitions P and J was improved for small and large dogs, and the level of agreement between definitions V and J was improved for medium and large dogs. The agreement between definitions P and V was reduced for all breed sizes.
There was a significant difference in the position of the anticlinal vertebra among the breed sizes for definition P (P=0.002) and definition V (P=0.008), but not for definition J (P=0.50). Based on post hoc comparisons, large breed dogs had a higher proportion of dogs with T11 identified as the anticlinal vertebra and small breed dogs had a higher proportion of dogs with T10 identified as the anticlinal vertebra, both for definition P (P=0.01) and definition V (P=0.003), but not definition J.
Abnormalities in shape and number of vertebrae were identified in 13 animals. These consisted of single transitional vertebrae in nine dogs: at T13 (six), L7 (two), and L1 (one). Three of the dogs with a transitional vertebra at T13 were assessed to have 12 thoracic vertebrae and eight lumbar vertebrae (i.e., bilateral transitional vertebra). One dog had two transitional vertebrae: one at T13 and one at L7. In addition, three dogs had variation in the number of vertebrae in the thoracolumbar segment: two dogs had 19 thoracolumbar vertebrae, with a transitional vertebra at the junction (13th vertebra), and one dog had 21 thoracolumbar vertebrae, which were assessed as 13 thoracic and eight lumbar vertebrae. For those animals with a normal number of vertebrae in the thoracolumbar segment, there was no significant difference in the proportion of anticlinal vertebrae assessed to be T11, as opposed to T10 or T12, between those animals with and without a transitional vertebra, for definition P (P>0.99), definition V (P=0.22), and definition J (P=0.59). In the two dogs with 19 thoracolumbar vertebrae, the anticlinal vertebra was assessed to be T10 by definitions P and V and T11 by definition J in one and T11 by all definitions in the other. In the one dog with 21 thoracolumbar vertebrae, the anticlinal vertebra was assessed to be T12 by all definitions. There were too few animals with an abnormal number of vertebrae to test the hypothesis that changes in the total number of thoracolumbar vertebrae has an effect on the position of the anticlinal vertebra.
The term anticlinal refers to either (1) sloping downward in opposite directions, as in the geological term, or (2) alignment of a linear structure perpendicular to another, such as the plane of a cell division perpendicular to the surface of a plant organ, as in the botanical term.13 These two definitions encompass the anatomic definitions V and P, respectively. When used in anatomy, the former definition is principally used and indeed, in the case of the anticlinal vertebra, it is the sloping of the spinous processes away from a central point that is considered the hallmark of this feature. Definition J was used in this study because it appears in the literature,1,7 and although it does not refer to an angle of inclination, it has objective morphological features as its basis.
The anatomic features that form the basis of these definitions used in this paper also have functional relevance. The orientation of the intervertebral synovial joints results in increased movement in a dorsal plane for the cranial thoracic spine and increased movement in a sagittal plane in the caudal thoracic spine. The position of the change in slope of the spinous processes affects the muscular advantage of the longitudinal spinal muscles for different gaits and activities. Mammalian species where these two anatomic features occur in the same or adjacent vertebra have greater flexibility in the spine than those where these anatomic features occur in vertebrae which are further apart.14 The anticlinal vertebra is positioned at the apex of the curve of the back in galloping individuals6 In fact, as a descriptive anatomic landmark, the position of the anticlinal vertebra has a valuable role as a paleobiological tool in mammalian taxonomy.12,15
In some diagnostic imaging modalities, if a small field of view is used, or if the tomographic nature of the imaging excludes anatomic landmarks which might be useful in lesion localization, then identification of additional anatomic features as reference points will aid interpretation.16 For instance, in one study, the usefulness of the coeliac and cranial mesenteric arteries as landmarks for magnetic resonance imaging of the spine was investigated, but it was found that the position of these arteries was too variable to allow them to be used as reference points.16 This author suggested that the anticlinal vertebra, described as T11, might be useful as an anatomic landmark.
Of the three definitions, definition J was the most consistent in identifying a single vertebra as the anticlinal vertebra and this definition resulted in T11 being the anticlinal vertebra in the majority animals, whereas definition V was the least consistent at identifying a single vertebra and identified T11 as the anticlinal vertebra in fewer animals. The reason for this may be that definition J is the most objective, as the orientation of the intervertebral synovial joints is easy to define. In applying definition V, the assessment of the vertebra with the most vertical spinous process was more subjective. In addition, definition J may also be readily applied to an image containing only a few thoracic vertebrae, whereas both definitions P and V required a certain number of vertebra either side of the proposed anticlinal vertebra to assess the long axis of the vertebral column (definition P) or assess that consecutive vertebrae had spinous processes sloping in the same direction (definition V). Furthermore, definition J may also be applied if the tips of the spinous processes are not included in the primary beam, whereas this would make the other definitions more difficult to use.
The use of combinations of definitions only improved the likelihood that the designated vertebra was T11, compared with definition J alone, if the combination included definition J. However, this increased specificity was achieved only at the expense of a loss of sensitivity, that is, the method is not applicable in animals where the definitions were discordant.
The three definitions used in this study are given as fact in anatomic textbooks and papers. Some of the texts use elements of both definition P and V in their description of the anticlinal vertebra, but none considered that these definitions might not be the same. However, it is apparent that these definitions are not in agreement in all dogs. The greatest degree of agreement was between definitions P and V, suggesting that these definitions refer to similar features, that is, the spinous process perpendicular to the long axis of the spine is likely to be the most vertical.
When subdivided according to breed size, the agreement among definitions was different in the size group from the group of all dogs in most categories. An improvement in the degree of agreement for definitions P and J, and definitions V and J, may reflect the relative ease in applying a more subjective definition (P or V) in dogs of different sizes.
A limitation of the study is that the information was taken from radiographs of the thorax and abdomen and not radiographs positioned and centered for the vertebral column. However, this allowed the acquisition of a large number of radiographs of animals with a normal vertebral column made with the same technique. In addition, the landmark of the anticlinal vertebra may still be useful for localizing lesions in close proximity to the vertebral column visualized on radiographs that are not positioned and centered for the vertebral column. Another potential limitation is that the study is based on radiographs rather than dissected anatomic specimens. However, our intention was to provide information that would be useful from a diagnostic imaging point of view.
There is more than one definition of the anticlinal vertebra and these definitions are not always in agreement. It should not be assumed that the anticlinal vertebra is T11 in all dogs. Care should be taken to define this anatomic feature accurately when using it as an anatomic landmark and this study emphasizes the difficulty in using skeletal landmarks for surgical or other interventional approaches. Definition J, which is the least commonly used definition in anatomy texts, was more objective in assessment, easier to use, and least affected by breed.
The authors would like to thank Aviva Petrie for her help and assistance with the statistical analysis and presentation of the data.