**Anatomical Assessment of the Thorax in the Neonatal Foal Using Computed Tomography Angiography, Sectional Anatomy, and Gross Dissections**

#### **Alberto Arencibia 1,\*, Juan Alberto Corbera 2, Gregorio Ramírez 3, María Luisa Díaz-Bertrana 4, Lidia Pitti 4, Manuel Morales <sup>2</sup> and José Raduan Jaber <sup>1</sup>**


Received: 27 May 2020; Accepted: 16 June 2020; Published: 17 June 2020

**Simple Summary:** This research aimed to describe the normal appearance of the thorax in neonatal foals by computed tomography angiography (CTA). The newborn foals were imaged using a 16-slice helical CT scanner after the administration of an iodinated contrast medium. CTA images and three-dimensional cardiac volume-rendered reconstructed images were obtained to enhance cardiovascular structures. In addition, thoracic anatomical sections and gross dissections were used as anatomical references. Clinically relevant anatomical structures were identified on the CTA images, anatomical sections, and gross dissections. These findings could serve as a reference to the CTA image assessment of the thorax of neonatal foals.

**Abstract:** The purpose of this study was to correlate the anatomic features of the normal thorax of neonatal foals identified by CTA, with anatomical sections and gross dissections. Contrast-enhanced transverse CTA images were obtained in three neonatal foals using a helical CT scanner. All sections were imaged with a bone, mediastinal, and lung windows setting. Moreover, cardiac volume-rendered reconstructed images were obtained. After CT imaging, the cadaver foals were sectioned and dissected to facilitate the interpretation of the intrathoracic cardiovascular structures to the corresponding CTA images. Anatomic details of the thorax of neonatal foals were identified according to the characteristics of CT density of the different organic tissues and compared with the corresponding anatomical sections and gross dissections. The information obtained provided a valid anatomic pattern of the thorax of foals, and useful information for CTA studies of this region.

**Keywords:** CT angiography; sections; dissections; thorax; anatomy; neonatal foal

#### **1. Introduction**

CTA is a minimally invasive imaging technique used to assess the organs of the respiratory and cardiovascular system. In humans, medicine has become the imaging modality of choice for diagnosis of abnormalities, injuries, and thoracic disease [1,2]. CTA displays the anatomical detail of specific tissue densities and blood vessels more precisely compared with radiography, ultrasonography, and magnetic resonance imaging. It is due to its high spatial resolution, shortened examination time, and improved

visibility of vascular structures and pulmonary parenchyma [3,4]. CTA can be used to obtain high-quality two-dimensional images and three-dimensional cardiac reformatted images that delineate the morphologic features of the cardiovascular system without the superimposition of other structures [5,6].

In equine medicine, radiology has been the main imaging modality used to image the thorax of foals, but the superimposition of adjacent anatomical structures makes its interpretation difficult [7]. Nonetheless, an advanced imaging technique such as the ultrasound [8,9] has given to the equine practitioners the opportunity to obtain an accurate diagnosis. In addition, helical CTA has also been used in horses, but information is limited to reports on the head, spinal cord, musculoskeletal system, and lungs [10–19]. To the best of our knowledge, a detailed anatomical study using CTA with an iodinated contrast medium for imaging the thorax of neonatal foals has not been reported before.

An accurate anatomical interpretation of helical CTA of the thorax could be useful to aid in the diagnosis of different diseases described in neonatal foals such as thoracic trauma, pulmonary disorders, and congenital heart anomalies [20–25]. Therefore, the objectives of this study were: (1) To describe the normal cross-sectional anatomy of the thorax of neonatal foals using helical CTA images with the use of a vascular contrast medium, anatomic sections, and gross dissections, and (2) to obtain three-dimensional cardiac volume-rendered reconstructed images to assist in the understanding of the anatomy of the heart and the main intrathoracic vessels.

#### **2. Materials and Methods**

#### *2.1. Animals*

Three newborn crossbreed foals of 2, 5, and 6 days with weights ranging between 50–55 kg were selected from equine patients attending to the Veterinary Hospital of Las Palmas de Gran Canaria University between January to December 2019. The animals had neurological signs that included head tilt, seizures, circling, and ataxia. No other physical examination abnormalities were detected. After clinical evaluation, a combination of butorphanol 10 mg/mL at a dose of 0.4 mL (Torbugesic®; Zoetis S.L.U., Madrid, Spain), and dexmedetomidine 10 mg/mL at a dose of 0.3 mL (Dexdormitor®; Lab. Dr. Esteve SAU, Barcelona, Spain) injected IM were employed as a preanesthetic medication. Anesthesia using sevuflorane 98 (0.5 to 2%) (Sevoflo.; Abbot Laboratories SA, Madrid, Spain) was maintained during the procedure. After the head scan, the study of the thorax was performed. The euthanasia was done due to the diagnosis of CNS congenital abnormalities. The Animal Ethical Committee of Veterinary Medicine of Las Palmas de Gran Canaria University authorized the research protocol (MV–2016/04). The owners of these foals were informed of the study and signed a consent for participation in the study.

#### *2.2. CTA Technique*

Contrast-enhanced sequential transverse CT slices were performed using a 16-slice helical CT scanner (Toshiba Astelion, Toshiba Medical System, Madrid, Spain). The animals were positioned symmetrically in dorsal recumbency on the CT couch and a standard clinical protocol (120 kVp, 80 mA, 512 × 512 acquisition matrix, 283 × 283 field of view, a spiral pitch factor of 0.94, and a gantry rotation of 1.5 s) was used to acquire transverse CT thorax images, with 3-mm slice thickness. In addition, all foals received a bolus of iomeprol 300 mg/mL at a dose of 2 mL/kg (Iomeron.; Rovi S.A., Madrid, Spain) via the jugular vein. The transverse original data was stored and transferred to the CT workstation. To better evaluate the CT appearance of the thoracic structures, three CT windows were applied by adjusting window widths (WW) and window levels (WL): A bone window setting (WW = 1500; WL = 300), a mediastinal window setting (WW = 248; WL = 123), and a lung window setting (WW = 1400; WL = −500). The original data were used to generate cardiac volume-rendered reconstructed images from the right and left surfaces, and the base of the heart after manual editing of the transverse CT images to remove bone structures and other soft tissues using a standard dicom 3D format (OsiriX MD, Geneva, Switzerland).

#### *2.3. Anatomic Evaluation*

The interpretation of the CTA images was based on anatomical sections and gross dissections to facilitate the identification of the thoracic structures. At the end of the scanning procedure, the euthanized foals were frozen until solid. Later, two frozen cadavers were sectioned using an electric band saw to obtain sequential transverse anatomical sections. The other cadaver was used to perform the gross anatomical dissections within 24 h of death to minimize post-mortem changes. Thoracic structures studied in the CT images were correlated with those identified in the corresponding anatomical sections and gross dissections, evaluated according to the characteristics of CT density of different tissues and labelled to conform to anatomical texts [26,27].

#### **3. Results**

#### *3.1. Transverse Computed Tomography Angiography Images*

The results of the thoracic CTA images are presented in seven sequential transverse CTA images of the thorax at different levels that best correlated with the macroscopic sections (Figures 1–7). Each figure consists of four images: (a) Bone window, (b) mediastinal window, (c) lung window, and (d) anatomical section. Transverse CT images are presented in a cranial to caudal progression from the level of the brachiocephalic trunk (Figure 1) to the level of the left ventricle and apex of the heart (Figure 7).

The CTA images obtained with the use of the bone window setting (Figures 1A–7A), provided a good differentiation between the bones and the soft tissues of the thoracic cavity. Thus, the thoracic vertebrae (including the vertebral body and arch, and the corresponding articular, transverse, and spinous processes), the ribs (with its head tubercle, body, and costal cartilage), and the sternum were well visualized (Figures 1A–7A). In addition, the vertebral cortical and bone marrow fat were also delineated. However, the costovertebral, costochondral, and sternocostal joints and those muscles associated with the thorax such as epaxial (semispinalis, longissimus, and iliocostalis) and thoracic wall (external and internal intercostal, and pectoral) muscles appeared with an intermediate CT density (Figures 1A–7A). Other anatomical structures such as the thoracic duct (Figures 1A–7A), the right and left vagus nerves (Figures 1A–4A), and the dorsal and ventral vagal trunks (Figure 5A) were also identified.

The CT mediastinal window (Figures 1B–7B) showed good visualization of the bony thoracic wall structures. Moreover, this CT window also provided an excellent visualization of the heart with its chambers (atrium and ventricles) and the main arteries and veins, which appeared with a high attenuation due to the intravenous contrast medium (Figures 1B–7B). Thus, important associated vessels such as the cranial (Figures 1B and 2B), and the caudal vena cava (Figures 5B–7B) were seen leading into the right atrium. Additionally, the course of the right azygos vein (Figure 3B–7B), the pulmonary trunk (Figure 3B), and the left and right pulmonary arteries (Figure 4B) were observed. Other intrathoracic vessels, including the aortic root (Figures 2B–4B), ascending (Figure 2B) and descending aorta (Figures 2B–7B), and the internal thoracic arteries and veins (Figures 1B–6B) were also identified. By contrast, with the use of CT bone window images (Figures 1A–7A), these structures appeared with an intermediate CT density. Other structures such as the myocardial walls showed an intermediate CT density and were best visualized in the CT bone and mediastinal windows settings (Figures 1A–7A and 1B–7B).

Concerning the lungs, the CT bone (Figures 1A–7A) and mediastinal (Figures 1B–7B) window settings showed the bronchi and the vascular formations of the lungs, which were only clearly defined at the level of the hilus because of the deeper lumen and the use of intravenous contrast medium. In contrast, the CT lung window (Figures 1C–7C) allowed a better definition of the lobes and a better tomographical definition of the trachea, tracheal bifurcation, main bronchi, and lobar bronchi due to these structures that presented higher attenuation than the lungs. Moreover, it was possible to visualize the triad that comprises the lobar pulmonary vein, the lobar arterial branch, the lobar bronchus, and the pleural cavity.

#### *3.2. Anatomical Sections*

On transverse anatomical sections (Figures 1D–7D), additional morphologic and topographic information about the thoracic structures could be identified when compared with CTA images. All bones, cartilaginous structures, and associated muscles were identified. The respiratory tract structures, including the trachea (Figures 1D–3D) and its bifurcation (Figure 4D), the principal and lobar bronchi, and pulmonary parenchyma (Figures 5D–7D), were also well observed. Other intrathoracic structures such as the heart with its chambers and associated large vessels were likewise visible in Figures 1D–7D. Other anatomical structures such as the thoracic duct (Figures 1D–7D), the right and left vagus nerves (Figures 1D–7D) were identified.

**Figure 1.** Transverse images at the level of the brachiocephalic trunk. (**A**) Computed tomography (CT) bone window; (**B**) CT mediastinal window; (**C**) CT lung window; and (**D**) anatomical section. These images are displayed so that the right side of the foal is to the viewer's left and the dorsal view is at the top. Brachiocephalic trunk (BT); cranial vena cava (CrVC); epaxial muscles (EM); esophagus (Eso); internal thoracic artery and vein (IAV); left lung (LL); left pulmonary root (LPR); left vagus nerve (LVN); rib: Costal bone (R); rib: Costal cartilage (CC); right lung (RL); right pulmonary root (RPR); right vagus nerve (RVN); scapula (Sca); sternum (St); thoracic duct (TD); thoracic vertebra (TV); and trachea (Tra).

**Figure 2.** Transverse images at the level of the ascending aorta. (**A**) CT bone window; (**B**) CT mediastinal window; (**C**) CT lung window; and (**D**) anatomical section. These images are displayed so that the right side of the foal is to the viewer's left and the dorsal view is at the top. Aortic root (AoR); ascending aorta (AAo); cranial vena cava (CrVC); descending aorta (Dao); epaxial muscles (EM); esophagus (Eso); internal thoracic artery and vein (IAV); left lung (LL); left vagus nerve (LVN); pulmonary trunk (PT); rib: Costal bone (R); rib: Costal cartilage (CC); right atrium (RA); right lung (RL); right pulmonary root (RPR); right vagus nerve (RVN); right ventricle (RV); scapula (Sca); sternum (St); thoracic duct (TD); thoracic vertebra (TV); and trachea (Tra).

**Figure 3.** Transverse images at the level of the pulmonary trunk. (**A**) CT bone window; (**B**) CT mediastinal window; (**C**) CT lung window; and (**D**) anatomical section. These images are displayed so that the right side of the foal is to the viewer's left and the dorsal view is at the top. Aortic root (AoR); descending aorta (Dao); epaxial muscles (EM); esophagus (Eso); internal thoracic artery and vein (IAV); left lung (LL); left pulmonary root (LPR); left vagus nerve (LVN); pulmonary trunk (PT); rib: Costal bone (R); rib: Costal cartilage (CC); right atrium (RA); right azygos vein (RAV); right lung (RL); right pulmonary root (RPR); right vagus nerve (RVN); right ventricle (RV); scapula (Sca); sternum (St); thoracic duct (TD); thoracic vertebra (TV); and trachea (Tra).

**Figure 4.** Transverse images at the level of the pulmonary arteries. (**A**) CT bone window; (**B**) CT mediastinal window; (**C**) CT lung window; and (**D**) anatomical section. These images are displayed so that the right side of the foal is to the viewer's left and the dorsal view is at the top. Aortic root (AoR); descending aorta (Dao); epaxial muscles (EM); internal thoracic artery and vein (IAV); left lung (LL); left pulmonary artery (LPA); left pulmonary root (LPR); left vagus nerve (LVN); rib: Costal bone (R); rib: Costal cartilage (CC); right atrium (RA); right azygos vein (RAV); right lung (RL); right pulmonary artery (RPA); right pulmonary root (RPR); right vagus nerve (RVN); right ventricle (RV); scapula (Sca); sternum (St); thoracic duct (TD); thoracic vertebra (TV); and tracheal bifurcation (TB).

**Figure 5.** Transverse images at the level of the lung roots. (**A**) CT bone window; (**B**) CT mediastinal window; (**C**) CT lung window; and (**D**) anatomic section. These images are displayed so that the right side of the foal is to the viewer's left and the dorsal view is at the top. Caudal cava vein (CaCV); descending aorta (Dao); dorsal vagal trunk (DVT); epaxial muscles (EM); internal thoracic artery and vein (IAV); left atrium (LA); left lung (LL); left pulmonary root (LPR); pulmonary vein (PV); rib: Costal bone (R); rib: Costal cartilage (CC); right atrium (RA); right azygos vein (RAV); right lung (RL); right pulmonary root (RPR); right ventricle (RV); scapula (Sca); sternum (St); thoracic duct (TD); thoracic vertebra (TV); and ventral vagal trunk (VVT).

**Figure 6.** Transverse images at the level of the left ventricle. (**A**) CT bone window; (**B**) CT mediastinal window; (**C**) CT lung window; and (**D**) anatomical section. These images are displayed so that the right side of the foal is to the viewer's left and the dorsal view is at the top. Caudal vena cava (CaVC); descending aorta (Dao); dorsal vagal trunk (DVT); epaxial muscles (EM); internal thoracic artery and vein (IAV); interventricular septum (IS); left lung (LL); left pulmonary root (LPR); left ventricle (LV). rib: Costal bone (R); rib: Costal cartilage (CC); right azygos vein (RAV); right lung (RL); right pulmonary root (RPR); right ventricle (RV); sternum (St); thoracic duct (TD); thoracic vertebra (TV); and ventral vagal trunk.

**Figure 7.** Transverse images at the level of the apex of the heart. (**A**) CT bone window; (**B**) CT mediastinal window; (**C**) CT lung window; and (**D**) anatomical section. These images are displayed so that the right side of the foal is to the viewer's left and the dorsal view is at the top. Caudal vena cava (CaVC); descending aorta (Dao); dorsal vagal trunk (DVT); epaxial muscles (EM); left lung (LL); left pulmonary root (LPR); left ventricle (LV); rib: Costal bone (R); right azygos vein (RAV); right lung (RL); right pulmonary root (RPR); thoracic duct (TD); thoracic vertebra (TV); and ventral vagal trunk (VVT).

#### *3.3. Gross Anatomical Dissections*

Figure 8 is a composition of two anatomical gross dissections at the level of the atrial (Figure 8A) and auricular (Figure 8B) surfaces of the heart. All chambers, grooves, and the main blood vessels

were identified. Thus, the myocardial walls and the coronary groove were well visualised in both images (Figure 8A,B), as well as the subsinuosal interventricular (Figure 8A) and the paracoronary interventricular (Figure 8B) grooves. The location relative to the cranial and caudal vena cava leading into the right atrium could be clearly observed in all views of the heart (Figure 8A,B). In addition, the pulmonary veins were identified in the image corresponding to the atrial surface of the heart (Figure 8A), while the pulmonary trunk arising from the right ventricle was clearly visible in Figure 8B. In addition, the right and left pulmonary arteries were also well identified (Figure 8). The course of the ascending aorta arising from the left ventricle, and the main branches such as the brachiocephalic trunk, the right, and left subclavian arteries, and the descending aorta were also easily identified (Figure 8B).

**Figure 8.** Anatomical dissections of the heart. (**A**) Atrial surface and (**B**) auricular surface. Aortic arch (AoAr); brachiocephalic trunk (BT); caudal vena cava (CaVC); cranial vena cava (CrVC); descending aorta (DA); heart apex (HA); left atrium (LA); left pulmonary artery (LA); left subclavian artery (LSA); left ventricle (LV); paracoronary interventricular groove (PIG); pulmonary trunk (PT); pulmonary vein (PV); right atrium (RA); right pulmonary artery (RPA); right subclavian artery (RSA); subsinuosal interventricular groove (SIG); and right ventricle (RV).

#### *3.4. Cardiac Volume-Rendered Reconstructed CTA Images*

Cardiac three-dimensional volume-rendered reconstructed images corresponding to right (Figure 9A) and left lateral surfaces (Figure 9B) and the base (Figure 10) of the heart are presented. Volume-rendered reconstructed CTA images provided a good visualisation of the heart and the major associated vessels. Thus, the cardiac chambers and the main associated blood vessels were identified in all CT reconstructed images (Figures 9 and 10). The location relative to the cranial and caudal vena cava leading into the right atrium could be clearly observed on all views of the heart (Figures 9 and 10). Other important vessels such as the right azygos and brachiocephalic veins joined to the cranial vena cava were seen (Figures 9 and 10). In addition to these observations, the junction of the pulmonary veins entering into the left atrium was identified in all volume reconstructed CT images (Figures 9 and 10). Figure 9 (panel A) shows the course of the pulmonary trunk originating from the right ventricle. In contrast, the pulmonary artery bifurcation (right and left pulmonary arteries) was clearly identified on the dorsal aspect (Figure 10). The course of the ascending aorta arising from the left ventricle (Figures 9B and 10), and its main branches (such as the descending aorta and brachiocephalic trunk) were easily identified in Figures 9B and 10. The cranial branches of the brachiocephalic trunk such as the left and right subclavian arteries, and bicarotid trunk could be identified on all reconstructed CT images (Figures 9 and 10).

aspect. Aortic arch (AoAr); ascending aorta (AAo); bicarotid trunk (BIT); brachiocephalic trunk (BT); brachiocephalic vein (BV); caudal vena cava (CaVC); cranial vena cava (CrVC); descending aorta (Dao); heart apex (HA); left atrium (LA); left pulmonary artery (LPA); left subclavian artery (LSA); left ventricle (LV); liver (L); pulmonary trunk (PT); pulmonary vein (PV); right atrium (RA); right azygos vein (RAV); right pulmonary artery (RPA); right subclavian artery (RSA); and right ventricle (RV).

**Figure 10.** Three-dimensional volume-rendered reconstructed image of the normal neonatal foal heart and associated blood vessels. Dorsal aspect. Aortic arch (AoAr); ascending aorta (AAo); bicarotid trunk (BIT); brachiocephalic trunk (BT); brachiocephalic vein (BV); caudal vena cava (CaVC); cranial vena cava (CrVC); descending aorta (Dao); left atrium (LA); left pulmonary artery (LPA); left subclavian artery (LSA); left ventricle (LV); liver (L); pulmonary trunk (PT); pulmonary vein (PV); right atrium (RA); right azygos vein (RAV); right pulmonary artery (RPA); and right subclavian artery (RSA).

#### **4. Discussion**

In humans, advanced image-based diagnostic techniques, especially helical computed tomography angiography makes possible the evaluation of the cardiac and vascular thoracic structures due to its fast imaging acquisition, the acquisition of body sections from different tomographic planes, good anatomic resolution without superimposition, high contrast between different vascular structures, and excellent tissue-like differentiation [1–4]. In addition, the use of CTA allows the obtention of three-dimensional volume-rendered reconstructed images that provide excellent detail of the heart, and the arteries and veins of this region [1,5,6].

In veterinary medicine, the use of third or four generations of CT scanners has provided an excellent anatomic resolution of the thoracic structures [28,29]. In the present study, CTA images were obtained using a helical CT scanner that provided a qualitative overview of thoracic morphology, giving adequate information of midline thoracic vascular structures, a good depiction of the four chambers of the heart, as well as serving of a standard reference for the size and positions of the heart and main blood vessels. The use of a 16-slice configuration CT scanner and a similar protocol was reported in other studies performed in humans [1–6], neonatal foal [17–19], dog [30,31], cat [32,33], and goat [34].

Clinical evaluation of the equine thorax is laborious due to its anatomical complexity, which makes it difficult to diagnose diseases by physical examination and conventional diagnostic techniques. Nevertheless, advanced diagnostic techniques as CTA has shown considerable advantages over traditional imaging techniques since it gives an accurate anatomical detail of blood vessels, higher differentiation of tissue densities [30,31]. In addition, CTA is more sensitive in detecting diseases such as congenital abnormalities of the cardiovascular system including the heart, vascular malformations, injuries, tumors, aneurysms, vessels ruptures or tears, and pulmonary embolism [1–5,32].

In this research, an intravascular contrast medium administration was very helpful to identify the heart chambers, the main associated vessels, and the delineation of the adjacent non-vascular structures. In veterinary medicine, only a few studies have applied contrast-enhanced helical CT to perform anatomical or clinical studies of the thoracic cavity in dogs [31,35–37] and cats [32], as well as in other studies performed in the thorax [17–19] and abdomen [38] of foals. Nevertheless, to our knowledge, the use of intravenous contrast agents to describe the normal anatomy of the thorax in neonatal foals has not been reported before. In CT imaging, the use of an appropriate window width is a key to successful diagnosis [30,31]. In the present study, thoracic CTA images were evaluated by the use of bone, mediastinal, and lung window settings. The CT bone window provided some valuable anatomical information of the cortical and medullar marrow fat of the bones, whereas the CT mediastinal window provided an excellent detail of soft tissues, especially the heart and the major associated blood vessels. By contrast, the lung window setting gave a better definition of the respiratory tract and intrapulmonary vascular structures.

The images obtained by volume-rendered tomographic reconstruction are the most flexible 3D visualization tools [1,35–37]. In our study, the contrast CT volume-reconstructed images performed by the post-processing bone removal technique provided an excellent anatomical detail of the lateral and dorsal aspects of the heart and main associated vessels. Lateral CT reconstructed acquisitions were preferred for the evaluation of the anatomic relationships between the heart chambers and the main blood vessels, while the dorsal view was selected for identification of midline thoracic vascular structures because it yielded detailed information about the pulmonary vessels, and the main branches of the brachiocephalic trunk. Usually, motion artifacts make it difficult to identify various parts of the heart or the lungs on CT images [28]. In this study, the use of helical scanning equipment tomography on living foals minimized the artifacts generated by cardiovascular and respiratory movements. However, its use in equine medicine is currently limited because of its expense, availability, and complications of acquiring CT images in older foals and adult horses due to their physical size [17–19].

This CTA anatomic study has confirmed the valid use of cadavers to evaluate different anatomic patterns. The absence of blood flow in dead animals must be taken into account when compared with live specimens. Results from the current study showed that the use of frozen anatomical sections was helpful in the identification of different thoracic structures observed on transverse CTA images and guaranteed the matching accuracy. In addition, the identification of vascular structures of the foal thorax in the volume-rendered reconstructed CTA images were facilitated by gross anatomical dissections of the atrial and auricular surfaces of the heart. Therefore, the three foals used in this study showed cardiovascular anatomy similar to that described in the anatomical literature [26,27]. Thus, the main anatomical differences in the cardiovascular structure of equines compared to dogs such as the subclavian arteries and bicarotid trunk arising from the brachiocephalic trunk could be distinghished.

There are no previously published anatomic identifications as these reported in this study, which could be applied as an initial anatomic approximation to other CTA studies on foals. Therefore, the information provided could be used for the evaluation of CT images of foals with thoracic disease. In humans, new CT scanners have achieved improved diagnostic capabilities to evaluate a wide variety of congenital and acquired heart diseases [1–6]. With improvements in CT protocols and optimized scanners, CT angiography images will become an accurate method for evaluating the foal thorax [17–19], and in the diagnosis of several thoracic diseases described in equine medicine [20–25].

#### **5. Conclusions**

Helical CT images provided adequate detail of the thorax of normal neonatal foals and were a useful imaging modality for anatomical evaluation. This information could serve as an initial anatomic reference aid to clinicians for the diagnosis of suspected thorax-associated diseases in foals.

**Author Contributions:** Conceptualization, A.A. and J.R.J.; methodology, A.A., J.A.C., and M.L.D.-B.; formal analysis, A.A.; investigation, A.A., G.R., L.P., and M.M.; writing—original draft, A.A. and J.R.J.; supervision, A.A. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding and the University of Las Palmas de Gran Canaria funded the study.

**Acknowledgments:** The authors thank foals owners for the cession of the specimens for our study.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Morphometrical Study of the Lumbar Segment of the Internal Vertebral Venous Plexus in Dogs: A Contrast CT-Based Study**

**Valeria Ariete 1, Natalia Barnert 1, Marcelo Gómez 1,\*, Marcelo Mieres 2, Bárbara Pérez <sup>1</sup> and Juan Claudio Gutierrez <sup>3</sup>**


**Simple Summary:** The internal vertebral venous plexus (IVVP) is a valveless venous network running inside the vertebral canal. The objective of this study was to morphometrically describe the IVVP, dural sac, epidural space and vertebral canal of the lumbar segment in dogs with enhanced computerized tomography. Six clinically healthy adult dogs were used for the study. Dorsal reconstructed computed tomography (CT) images showed a continuous rhomboidal morphological pattern for the IVVP. The dural sac was observed as an isodense structure with a rounded shape throughout the vertebral canal. The average percentage area occupied by the IVVP between L1 and L7 vertebrae ranged between 6.3% and 8.9% of the area of the vertebral canal, and the dural sac ranged between 13.8% and 72.2% of the vertebral canal. The epidural space accounted between 27.08% and 86.2% of the lumbar vertebral canal. CT venography is a safe technique that allows adequate visualization and evaluation of the lumbar IVVP and adjacent structures in dogs.

**Abstract:** The internal vertebral venous plexus (IVVP) is a thin-walled, valveless venous network that is located inside the vertebral canal, communicating with the cerebral venous sinuses. The objective of this study was to perform a morphometric analysis of the IVVP, dural sac, epidural space and vertebral canal between the L1 and L7 vertebrae with contrast-enhanced computed tomography (CT). Six clinically healthy adult dogs weighing between 12 kg to 28 kg were used in the study. The CT venographic protocol consisted of a manual injection of 880 mgI/kg of contrast agent (587 mgI/kg in a bolus and 293 mgI/mL by continuous infusion). In all CT images, the dimensions of the IVVP, dural sac, and vertebral canal were collected. Dorsal reconstruction CT images showed a continuous rhomboidal morphological pattern for the IVVP. The dural sac was observed as a rounded isodense structure throughout the vertebral canal. The average area of the IVVP ranged from 0.61 to 0.74 mm2 between L1 and L7 vertebrae (6.3–8.9% of the vertebral canal), and the area of the dural sac was between 1.22 and 7.42 mm<sup>2</sup> (13.8–72.2% of the vertebral canal). The area of the epidural space between L1 and L7 ranged from 2.85 to 7.78 mm2 (27.8–86.2% of the vertebral canal). This CT venography protocol is a safe method that allows adequate visualization and morphometric evaluation of the IVVP and adjacent structures.

**Keywords:** internal vertebral venous plexus; computed tomography; canine

#### **1. Introduction**

The vertebral venous plexus (VVP) is a thin-walled, valveless venous network that surrounds the entire length of the vertebral column, terminating at the cephalad end in the cerebral venous sinuses [1,2]. According to its position inside or outside of the vertebral canal, the vertebral venous plexus can be divided into three intercommunicating divisions:

**Citation:** Ariete, V.; Barnert, N.; Gómez, M.; Mieres, M.; Pérez, B.; Gutierrez, J.C. Morphometrical Study of the Lumbar Segment of the Internal Vertebral Venous Plexus in Dogs: A Contrast CT-Based Study. *Animals* **2021**, *11*, 1502. https:// doi.org/10.3390/ani11061502

Academic Editors: Matilde Lombardero and Mar Yllera Fernández

Received: 18 March 2021 Accepted: 19 May 2021 Published: 22 May 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

the internal vertebral venous plexus, external vertebral venous plexus, and basivertebral veins [1]. The internal vertebral venous plexus (*plexus vertebralis internus ventralis*) (IVVP) is also known as the longitudinal venous sinus, vertebral venous sinuses, epidural venous plexus, paravertebral veins, and meningorachidean plexus [3]. The IVVP lies within the vertebral canal, inside the epidural space, and along the dorsal surface of the vertebral bodies and intervertebral disks [1,2] (Figure 1). Cranially, the IVVP communicates with the basilar sinuses at the level of the foramen magnum, and caudally, it extends to the fourth or sixth caudal vertebra [1]. The VVP drains blood from the vertebral column, the paravertebral musculature, the spinal cord, the meninges, and the nerve roots of the spinal nerves [3]. In addition to its normal drainage function, this vascular network can also be a collateral pathway for blood return towards the heart in cases of occlusion or ligation of the caval venous system [4]. Clinically, in humans and animals, rupture of the IVVP has been associated with the etiology of spontaneous spinal epidural hematomas [5,6]. Doberman pinschers with a deficiency of von Willebrand factor (VIII-related antigen) can develop significant spinal epidural hemorrhage due to laceration of the IVVP, which results in progressive neurological deficits [7]. Following Hansen type I intervertebral disk herniation, the extruded nucleus pulposus can be expelled laterally and rupture the IVVP. This can result in bleeding, hematoma formation, and subsequent extradural spinal cord compression [8]. The continuity of the valveless IVVP throughout the length of the neuroaxis enables bacteria and tumor cells to travel from the thorax, abdomen, and pelvis to the head and vertebral column when intrathoracic or intra-abdominal pressure is increased [9]. This phenomenon has been termed paradoxical embolism, retrograde venous invasion, and Batson's phenomenon [9]. In dogs, metastasis of osteosarcomas and pheochromocytomas into the central nervous system is hypothesized to occur by retrograde venous spread through the VVP. Diskospondylitis, vertebral osteomyelitis, and infectious diskitis in dogs with primary sites of infection elsewhere in the body have been explained by this mechanism [10]. The IVVP may participate in the etiology of fibrocartilaginous embolism of the spinal cord vasculature (also known as embolic myelopathy) in dogs [11]. In humans, the IVVP also participates in the pathophysiology of other spinal cord vascular lesions, such as arteriovenous fistulas and venous malformations [12]. Recently, venous aneurism of the IVVP in a Scottish Deerhound was reported concurrently with severe dilatation of the venous sinuses [13]. Additionally, case reports in dogs have described abnormal enlargement of the IVVP in sighthounds by magnetic resonance imaging (MRI) with clinical signs of radiculopathies and myelopathy [14].

Understanding anatomical vertebral morphometry is an important factor when interpreting the pathological changes associated with several structural spinal diseases. The aims of the present study were to provide an in vivo morphometric description of the normal lumbar IVVP and surrounding vertebral structures using contrast-enhanced computed tomography (CT) in dogs.

**Figure 1.** Lumbar internal vertebral venous plexus (IVVP) in a dog. (**A**) Dorsal anatomic view in which the vertebral arch and spinal cord was removed between L1 and L4. The IVVP is observed over vertebral bodies (arrow), intervertebral disk (arrow heads), and dorsolateral to the dorsal longitudinal ligament (asterix), with a typical rhomboidal appearance (arrow). (**B**) Transverse anatomic view of fourth lumbar vertebra where the IVVP (arrow) is located ventrally in the spinal epidural space and dorsal to the L4 vertebral body. (**C**) Schematic representation of some components of the vertebral venous plexus in the dog (Photographs correspond to previous dissections made by the author that do not correspond to the animals examined in this study).

#### **2. Materials and Methods**

#### *2.1. Biological Material*

For the study, 6 healthy adult dogs were used. The age range was between 2 and 6 years old, and the weight range between 12 and 28 kg, with an average weight of 19.5 kg. The dog breed included 1 Boxer, and 5 mixed-breed, with 3 males and 3 females. Owners provided written consent for research purposes via an authorization form. The experimental study was carried out according to the protocols proposed by the Bioethics Committee of the Austral University of Chile and all procedures were monitored throughout the study by a veterinarian

#### *2.2. Anaesthesia Protocol*

All CT studies were performed under general anesthesia with the patient positioned in dorsal recumbency. For the study, dogs were food-deprived for 12 h prior to the CT scan. Each dog was premedicated with 0.5 mg/kg of intramuscular xylazine (2% Xylazine, Xylavet 2% Lab. Alfasan, Worden, Holland). Then, animals were cannulated with a 20 G cannula in the cephalic vein to administer a contrast medium in a continuous infusion of saline (0.9% NaCl) in a volume of 10 mg/kg/h. Anesthesia was induced by intravenous administration of propofol (Propofol 1%, PropoFlo; Zoetis, NJ, USA) at a dose of 4 μg/kg and maintained with 2% isoflurane in oxygen.

#### *2.3. CT Venography Protocol*

For the study, a four-generation CT scanner (Picker PQ 6000, Picker International, Cleveland, OH, USA) was used. For imaging, dogs were positioned in dorsal recumbency with flexed hindlimbs over the CT table. Contiguous CT slices were obtained before and after the application of a nonionic iodinated contrast medium (Iohexol, Omnipaque® 300 mg of iodine/mL, Nycomed Inc, Princeton, NJ, USA) in a total dose of 880 mg/kg IV, applying 587 mg I/kg by bolus, and flow rate of 4 mL/s before the scan. Infusion of an additional 293 mg I/kg as continuous infusion during the scan was administered by means of a constant-rate infusion pump with a flow rate of 0.1 mL/min. Technical parameters included: Slice thickness of 2 mm and an interval of 2 mm and a scan field of view of 180 cm. Settings of the CT scanner included standard resolution, 100 mA and 130 kV per slice. The study area included from L1 to L7 vertebrae. The CT gantry was tilted such that the sections were parallel to intervertebral spaces. All dogs recovered routinely from anesthesia and were clinically normal at 24 and 48 h post CT examination.

#### *2.4. Morphometric Analysis*

The CT images were transferred to a computer for morphometric assessment. Measurements were obtained from CT transverse images using standardized soft tissue window settings (window width WW: 250–450; window level WL: 30–50). All measurements were performed between L1 to L7 at mid-vertebral body level. This segment was chosen due to the scarce morphometric information of components of the vertebral canal at this level in dogs. All image data were imported and recorded into the Osirix® DICOM viewer (Pixmeo Inc., Version 3.9.4., 32 Bit, Bernex, Switzerland), a software for morphometric analysis. Anatomical parameters were measured at the mid-level of the vertebral body of the lumbar segment (L1–L7). Measurements were taken in millimeters in the axial planes and were collected three times by the same observer. The following anatomical parameters were determined considering the anatomical structures shown in Figure 2:


#### *2.5. Statistical Analysis*

Data are presented as the mean ± standard deviation and were analyzed using STATGRAPHICS (Centurion XVI version 16.1.12, 32 bits, StatPoint Inc., Rockville, MD, USA) software.

**Figure 2.** (**1**) Illustration and (**2**) transverse post-contrast CT image at (**3**) the level of L2 vertebra in an adult dog illustrating the measured anatomical parameters. The broken line in the photograph indicates the level of the transverse section where the measurements were performed in the L1-L7 vertebral segment. (**a**) Epidural space, (**b**) dural sac, (**c**) internal vertebral venous plexus.

#### **3. Results**

In all six dogs, appropriate opacification of the vertebral venous component was achieved in all of them. Nonselective CT venography developed in this study allowed adequate visualization and measurement of the IVVP and associated vertebral structures. The contrast injection protocol based on nonionic iodinated contrast medium in a total dose of 880 mg I/kg, 2/3 of which was delivered as an initial bolus before CT image acquisition and 1/3 of which was delivered in the form of a constant-rate infusion pump, resulted in an IVVP average density of 215 UH. Image acquisition time lasted between 25 and 35 min.

#### *3.1. Internal Vertebral Venous Plexus*

The internal vertebral venous plexus (IVVP) was observed as two symmetric longitudinal hyperdense channels, oval in shape, located in the epidural space, dorsal to the vertebral bodies, and ventral to the dural sac (Figure 2). In dorsal reconstruction CT images, it was possible to appreciate their rhomboidal distribution within the vertebral canal (Figure 3). Additionally, CT transverse images obtained at the level of the middle portion of the vertebral body allowed visualization of the basivertebral veins and tributaries of the IVVP (Figure 4). The IVVP between the L1 and L7 vertebrae had an average area of 0.66 mm2, with a minimum value recorded of 0.43 mm2, and the maximum of 1.01 mm2. In vertebra L7, the IVVP presented the lowest average with a 0.61 mm2 (±0.1) area, and L5 vertebrae presented the highest IVVP average area with a 0.74 mm2 (±0.2) area (Table 1, Figure 5). The IVVP occupies an average of 7.67% of the L1–L7 vertebral canal. In the L1 vertebra, there was the highest percentage of the IVVP in relation to the vertebral canal, with 8.9%, while in the L4 vertebra, IVVP occupied the lower percentage, with 6.3% of the vertebral canal (Table 2).

**Figure 3.** Dorsal reconstruction post-contrast CT images at the L3–L6 lumbar level in an adult medium size dog. (**a**) Intervertebral venous plexus, (**cr**) cranial, (**ca**) caudal.

**Figure 4.** Transverse post-contrast CT image of the mid-portion of the L3 vertebral body from an adult medium size dog. (**b**) basivertebral veins, (**L**) left, (**R**) right.

**Table 1.** Average area (mm2) and standard deviation of the IVVP, vertebral canal, and dural sac and epidural space between the L1 and L7 vertebral segments in six adult medium size dogs.


**Figure 5.** Area dimension of the internal vertebral venous plexus (IVVP) between L1 and L7 vertebral segment in six adult medium size dogs.

**Table 2.** Average percentage (%), minimum and maximum of the vertebral canal corresponding to the dural sac, internal vertebral venous plexus (IVVP), and epidural space between vertebrae L1 and L7 in six adult dogs.


#### *3.2. Vertebral Canal*

The vertebral canal measured an average of 8.94 mm2 (±1.7) between the L1 and L7 vertebral segments, the minimum value recorded was 6.14 mm2, and the maximum was 12.92 mm2. Vertebra L1 had the lowest recorded value with a 7.87 mm2 (±1.3) area, while the L4 vertebra had the highest value with a 10.27 mm2 (±1.9) area (Table 1, Figure 6).

**Figure 6.** Area dimension of the vertebral canal between L1 and L7 vertebral segment in six adult medium size dogs.

#### *3.3. Dural Sac*

The dural sac had an average area of 4.83 mm<sup>2</sup> (±2.1) between vertebral segments L1 and L7. The minimum recorded value was 1.11 mm2, and the maximum value was 10.05 mm2. The dural sac in the L7 vertebra had the lowest value with a 1.22 mm2 (±0.1) area, and the L4 vertebra had the highest average with a 7.42 mm2 (±1.6) area (Table 1, Figure 7). The dural sac occupied between 13.8% and 72.2% of the vertebral canal between the L1 and L7 vertebral segments in these dogs (Table 2). It is also possible to appreciate that there is an increase in the occupancy rate toward L4, where it reaches its maximum value, and then it decreases toward L7, where the lowest average value is recorded.

**Figure 7.** Area dimension of the dural sac between L1 and L7 vertebral segment in six adult medium size dogs.

#### *3.4. Epidural Space*

The average area of the epidural space between segments L1 and L7 ranged from 2.85 (±0.8) to 7.78 (±1.6) mm2, corresponding to 27.8% to 86.2% of the vertebral canal (Tables 1 and 2, Figure 8).

**Figure 8.** Area dimension of the epidural space between L1 and L7 vertebral segment in six adult medium size dogs.

#### **4. Discussion**

The nonselective CT venography technique developed in this study allowed adequate visualization and measurement of the lumbar IVVP and associated vertebral structures in a group of dogs. The contrast injection protocol based on nonionic iodinated contrast medium in a total dose of 880 mg I/kg, 2/3 of which was delivered as an initial bolus before CT image acquisition and 1/3 of which was delivered in the form of constant-rate infusion during CT acquisition, resulted in an IVVP density of 215 UH. According to different studies, for optimal vasculature viewing in CT examinations, it is necessary to use a window level ranging between 200 and 400 UH and a window width ranging between 200 and 2000 UH, depending on the vascular concentration of contrast medium [15]. There are various doses and administration protocols for contrast administration published for CT venography in dogs. There have been protocols with values ranging from 720 mg I/kg to 814 mg I/kg in different studies, including the study of IVVP cervical level, diagnosis of portosystemic shunts, multiple vena cava anomalies, pancreatic TC analysis, and evaluation of pulmonary embolisms [16–21]. Although high doses of contrast medium were used in this CT venographic study, no adverse effects were observed in all animals examined. However, it is necessary to consider that large doses of contrast medium in angiographic studies may be associated with major risk of contrast agent hypersensitivity and renal insufficiency [22].

The IVVP appearance on the CT images consisted of two symmetrical oval hyperatenuated structures with defined margins located on the floor of the vertebral canal (Figure 2). The description observed in this study evidenced a rhomboidal distribution with a metameric organization of the venous system consistent with previous publications in dogs [1,17,18]. Regarding morphometry of the IVVP, the average area measurements for the lumbar segment (from 0.61 to 0.74 mm2) were lower than values published previously for the cervical segment in dogs, where an average area of 1.33 mm<sup>2</sup> was recorded in similarly-sized dogs [18]. In addition, the progressive cranio-caudal decrease in the diameter of the IVVP described previously was also observed in previous studies [3,18]. The lumbar IVVP constituted, on average, 7.67% of the area of the vertebral canal and 19.69% of the epidural space, lower values than those reported for the cervical segment values, where these vessels occupy 12.4% of the vertebral canal and 30.61% of the epidural space [18]. These venous trunks tend to decrease in size gradually from vertebrae T4 to L7, away from each other at intervertebral spaces and converging in the center of the vertebral bodies. The size of the IVVP in the L5–L7 segment was smaller and variable, therefore values of the structures in this area should be taken with caution.

In humans, the IVVP is smaller in the cervical area, being bulkier and pronounced as it descends through the vertebral canal, reaching a maximum size at the level of the L4–L5 vertebrae [5]. This difference is most likely related to the postural difference between humans and canids [5]. Dissection studies in humans determined that the anterior IVVP, equivalent to the ventral IVVP in dogs, has a mean length of 103 mm (range 43–153 mm) and a maximum width of 5.8 mm. The maximum caliber of the anterior IVVP in humans is at the thoracolumbar segments (mean maximum width 7.2 ± 2.0 mm) [22].

Previous anatomical descriptions of the IVVP have been performed for humans, dogs, domestic cats, rodents, and even reptiles using various imaging modalities ranging from conventional radiography to more modern techniques, such as magnetic resonance imaging (MRI) and CT [17,18,23–26]. In vivo imaging studies are more appropriate to analyze the vascular anatomy in animals because of vein collapse in cadaver specimens.

In humans, epidural venous engorgement in the lumbar spine, including epidural varices, may cause pain in the lower back and radiculopathy secondary to nerve root impingement [27]. The diagnosis of IVVP engorgement is often mistaken for a herniated disc on radiologic interpretation, and the true diagnosis is finally made in the surgical intervention. In dogs, vertebral venous system abnormalities have been identified on MRI in 12% of all sighthounds, but the underlying cause is unknown [14]. Recently, severe dilation of the right IVVP causing significant compression of the spinal cord and nerve roots was identified in an adult Scottish Deerhound [13]. Bilateral vertebral venous sinus thrombosis and dilation causing cervical spinal cord compression were also recently detected in a 10-year-old male mixed dog [28]. In another study, the visibility of the IVVP was significantly (*p* < 0.001) different between Great Danes with and without clinical signs of cervical spondylomyelopathy [29]. Variation in the size of the IVVP in the central nervous system (CNS) can be explained by the Monro-Kellie doctrine, which establishes an inverse relationship between cerebrospinal fluid (CSF) volume and intracranial blood volume [30]. Hence, as CSF is removed from the intracranial compartment, more blood enters the intracranial compartment. Because the intracranial compartment and vertebral canal are nearly a closed system, the principles of the Monro-Kellie doctrine can be extended to the vertebral canal. As expected, the size of the dural sac increased with increased intracranial

CSF or blood volume and decreased with decreased intracranial CSF or blood volume [30]. In the brain, the dura mater is closely opposed by bone, whereas in the vertebral canal, the dura mater has the IVVP and fat separating it from the fixed bony vertebral canal. As the epidural fat is constant and likely has little movement in and out of the neural foramen, the IVVP is most likely to change in size with changes in dural sac volume [30].

In relation to the dural sac, the increase in the average area recorded at the L4 vertebra level is associated with medullary expansion of the L4-S3 segment, known as lumbar enlargement (*intumescentia lumbalis*) [1]. From this lumbar enlargement, the nervous roots of the lumbosacral plexus emerge, which can be seen in the transverse images from the L5 vertebra caudal. The dural sac area values recorded for each vertebral segment, with the exception of L4, are smaller than those published by Gómez et al. (2005) for the cervical segment in dogs, where an average area of 6.40 + 1.0 mm<sup>2</sup> was recorded. A CT study of the dural sac in humans indicated that between the L4 and L5 segments, the average area of the dural sac was >100 mm2, which was higher than that recorded in the present study for the same vertebral segment in dogs [31]. In humans, a dural sac cross-sectional area < 100 mm2 and dural sac anteroposterior diameter <10 mm are frequently considered to assess the severity of spinal canal stenosis [31,32]. The percentage of the cross-sectional area of the dural sac found in our study was between 50 and 72% of the vertebral canal between L1 and L4. These values were similar to those reported by Lim et al. (2018) [33], with values of 64% in a group of 12 normal Beagle dogs. In this study, a significant reduction in the ratio cross-sectional area of the dural sac (40%) in patients with different spinal disorders (IVDD, spinal tumors, hematomas, etc.) to that of the vertebral canal was observed [33].

The vertebral canal area measured 8.94 mm<sup>2</sup> between the L1 and L7 vertebral segments on average. Similar results were observed in a morphometric CT study of the thoracic spine performed in 13 German Shepherd dogs for the T2–T13 vertebral level (8.36 ± 4.03 mm2) [34]. A decrease in vertebral canal diameter and/or area that results in compression of spinal cord and/or nerve roots is termed absolute stenosis, whereas a diameter that is less than normal but does not cause compression of neural elements is termed relative stenosis [35,36]. Relative vertebral canal stenosis results in decreased available space for the spinal cord to compensate for extradural space-occupying conditions. Relative vertebral canal stenosis therefore predisposes animals to develop clinical signs when relatively mild space-occupying pathologies, such as age-related intervertebral disc protrusion or ligamentous hypertrophy, occur [35,36].

The percentage of the vertebral canal of the spinal epidural space was minimal between L1 and L4 (27–41%), indicating that these segments are more subject to clinically significant epidural compressive lesions and that L5–L7 had a greater amount of vertebral canal available for the dural sac (47–87%). The large size of the epidural space between L5 and L7 also provides a secure space for epidural instruments or devices (i.e., catheters) in this region. The spinal epidural space surrounds the dural sac and is limited dorsally by the epidural fat, ligamentum flavum and periosteum, ventrally by the dorsal longitudinal ligament, IVVP and vertebral bodies, and laterally bordered by the vertebral pedicles and intervertebral foramina [1]. In MRI and CT imaging studies, severe stenosis is usually associated with subjective signs of the absence of epidural fat [37–39]. There are no quantitative studies using CT images that evaluated the dimensions or proportion of the lumbar epidural space in dogs.

Limitations of the present study were associated with the small number of dogs included and the variety of dogs of different sizes evaluated. Further studies to increase the study population and standardize various breeds to allow analysis by breed (small, medium, large, and giant) size are necessary. These data should thus be interpreted with caution given that measurements were only acquired at the mid-vertebral level and did not include the intervertebral cross-sectional area parameters.

#### **5. Conclusions**

This preliminary study provides reference values of the lumbar IVVP and adjacent structures in a group of dogs. The IVVP constitutes a complex vascular network of the vertebral column that has recently received attention in veterinary medicine. The understanding of its anatomy and morphometry is necessary to adequately diagnose new pathological entities that involve this venous plexus and in preoperative imaging evaluation of the venous morphometry may be useful to avoid complications related with vascular structures.

**Author Contributions:** Conceptualization, V.A., N.B., M.G., and M.M.; methodology, V.A., N.B., M.G., and M.M.; investigation, V.A., M.G., B.P., and J.C.G.; resources, M.G., M.M., B.P., and J.C.G.; writing—original draft preparation, M.G., B.P., and J.C.G.; and writing—review and editing, V.A., N.B., M.G., B.P., and J.C.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Comité de Bioética "Uso de Animales en la Investigación" from Universidad Austral de Chile (UACh), code 22-2011.

**Informed Consent Statement:** Informed consent was obtained from all dog owners.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Review* **The Cat Mandible (I): Anatomical Basis to Avoid Iatrogenic Damage in Veterinary Clinical Practice**

**Matilde Lombardero 1,\*, Diana Alonso-Peñarando 2,3 and María del Mar Yllera <sup>1</sup>**


**Simple Summary:** Nowadays, cats are one of the most common companion animals. They differ from dogs in some important aspects. However, most of the veterinary clinics are oriented towards the care and treatment of dogs, where the cat patient is clinically treated like a small dog. The cat mandible and related structures have some particularities that should be taken into account, when treating a cat, to avoid any unintended medical (iatrogenic) damage. The feline mandible has fewer teeth than a dog's one, but tooth roots and the neurovascular supply account for up 70% of the volume of the mandibular body. This fact makes mandibular fracture repair challenging. In addition, the cat mandible has a prominent angular process that, when the cat is under anesthesia and his mouth is wide open (during oral or transoral manipulation), compresses the maxillary artery (that supplies blood to the brain) inducing temporal or permanent blindness and/or deafness. Other particularities of the cat jaw are also addressed to get a comprehensive knowledge of its functional anatomy, essential to an effective feline clinical practice.

**Abstract:** Cats are one of our favourite pets in the home. They differ considerably from dogs but are usually treated clinically as small dogs, despite some anatomical and physiological dissimilarities. Their mandible is small and has some peculiarities relative to the dentition (only three incisors, a prominent canine, two premolars and one molar); a conical and horizontally oriented condyle, and a protudent angular process in its ventrocaudal part. Most of the body of the mandible is occupied by the mandibular dental roots and the mandibular canal that protects the neurovascular supply: the inferior alveolar artery and vein, and the inferior alveolar nerve that exits the mandible rostrally as the mental nerves. They irrigate and innervate all the teeth and associated structures such as the lips and gingiva. Tooth roots and the mandibular canal account for up to 70% of the volume of the mandibular body. Consequently, when fractured it is difficult to repair without invading the dental roots or vascular structures. Gaining a comprehensive anatomical knowledge and good clinical practice (such as image diagnosis before and post-surgery) will help in the awareness and avoidance of iatrogenic complications in day-to-day feline clinical practice.

**Keywords:** anatomy; feline; lower jaw; neurovascular supply; temporomandibular joint; tooth

#### **1. Introduction**

Cats have been fully appreciated animals for a long time. Ancient Egyptians mummified cats, as pets were buried with their owner, or they were used as votive offerings that depicted the gods [1]. Animals were revered by associating them with deities. In fact, the goddess Bastet was symbolised as a cat or even a woman with a feline head [2].

Nowadays, cats are one of the most numerous pets at home, sharing our company and friendship. According to the American Pet Products Association's (APPA) 2019–2020

**Citation:** Lombardero, M.; Alonso-Peñarando, D.; Yllera, M.d.M. The Cat Mandible (I): Anatomical Basis to Avoid Iatrogenic Damage in Veterinary Clinical Practice. *Animals* **2021**, *11*, 405. https://doi.org/ 10.3390/ani11020405

Academic Editor: Ralf Einspanier Received: 21 December 2020 Accepted: 29 January 2021 Published: 5 February 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

National Pets Owners Survey [3], it is estimated that in the USA, 67% of households own a pet; that is 84.9 million homes with an estimated 63.4 million dogs and 42.7 million cats, among other pets (such as freshwater/saltwater fish, birds, small animals and reptiles, accounting all together for 137 million animals being kept at home). However, compared with 2019 figures, in the USA, an estimated 89.7 million dogs and 94.2 million cats is reported (made in base of multiplying the number of households that own the pet by the average number of pets owned by household), which is a large decrease in both species, but no explanation was given, unless maybe it was a poor rough calculation. In Europe, there were at least 85 million households (38% of all households) owning a minimum of one pet animal [4]: 106.4 million cats and 77.4 million dogs. Hence, in Europe, the numbers of cats largely overtake those of dogs as pets; as they do not take much space, they do not need to go outdoors during the day, and their ownership is more affordable, among other reasons. However, the majority of pet clinics are more oriented towards dog management and treatment. Concerning cat size, they are usually small and consequently their surgery is more difficult. In addition, their lifespan is longer than the dog's, with an average life span of 12 years and ages of 20 years or more not uncommon. The current longevity record cited by Beaver in 2003 [5] was 36 years. Regarding clinical surgery solutions, up to now, little has been is done by companies to provide surgical prostheses and other products adapted to such a small animal, as they are more oriented towards different dog breeds. Hence, unfortunately, and in accordance with Little [6], cats are still the 'poor stepchild' in companion animal medicine, receiving less attention in research into common medical problems or improved diagnostic and treatment approaches than is given to their canine counterparts.

Cats should not be considered or treated as small dogs as there are many differences. This review will be focused specifically on the cat mandible and other related aspects. Taking into account the key relevance of anatomy in high-quality medicine, the treatment of the feline mandible has many important features, most of them due to the singularity of its anatomy when compared to that of the dog. Knowledge of the functional anatomy of the cat mandible will give us clues as to appropriate treatment of the feline patient, minimising the occurrence of potential iatrogenic injuries. Mandibular injuries are varied: from those related to tooth extraction complications to mandibular fractures, or from temporomandibular joint pathologies to neural complications due to an excessive opening of the mouth (usually during oral or transoral manipulation). Consequently, we propose to review the cat mandible and the anatomically related structures from the point of view of functional anatomy with a clinical orientation, as it is of crucial importance in face morphology and for feeding and grooming. Thus, any alteration of the mandible would have important repercussions on the general appearance of the feline patient, affecting both food intake and grooming, as house cats can spend up to 50% of their waking time grooming their coat or performing related behaviour [5]. The various grooming behaviours are important to a normal healthy cat. In the absence of grooming, excess debris can tangle fur, causing painful tugging of the skin and even infection.

#### **2. The Mandible**

Starting with the anatomical description of the feline mandible, and according to the Veterinary Anatomical Nomenclature [7,8], the cat jaw has two halves (*mandibula*) joined rostrally by an *articulatio intermandibularis,* known as the mandibular symphysis (in carnivores, partially conformed by a *synchondrosis* and a *sutura intermandibulares*). Each mandibular half has a horizontal part (body, or *corpus mandibulae*) and a vertical part placed caudally (*ramus mandibulae*; Figure 1). The mandibular ramus has a dorsal part, oriented slightly caudally, with a coronoid process (*processus coronoideus*) on the top (which attaches the temporal muscle), and two faces with both *fossae* for muscle fixation: a masseteric fossa (very deep, laterally—*fossa masseterica*) and a pterigoid fossa (medially—*fossa pterygoidea*). More or less at the same level as the dorsal margin of the body, completely caudally, there is the condylar process (*processus condylaris*), a cylinder oriented horizontally that articulates

with the mandibular fossa of the temporal bone and participates in the temporomandibular joint (TMJ) (*articulatio temporomandibularis*). On the caudal part and ventrally, there is an angular process (*processus angularis*), oriented caudally.

**Figure 1.** Rostro-lateral view of the cat mandible. The most relevant details are shown. A: posterior mental foramen; B: main or middle mental foramen; and C: rostral mental foramen. The prominence of the *processus angularis* is more manifest in a medial view (white arrow).

The mandibular body has two faces—medial or lingual (*facies lingualis*), and lateral or buccal and labial (*facies buccalis* and *labialis*)—separated by two margins - the dorsal or alveolar (*margo alveolaris*), and the ventral—(*margo ventralis*). The ventral margin is smooth, whereas the alveolar margin is very irregular with deep pits where the root teeth are fitted (*alveoli dentales*). The body has two parts, a rostral or incisive part (*pars incisiva*) and the molar part (*pars molaris*). The incisive part supports the incisive and canine teeth, whereas the molar part contains the premolar and molar teeth. All the lower teeth form the *arcus mandibularis inferior*. The buccal face of the mandible is uneventful, except for the two mental foramina (*foramina mentalia)*: the main mental foramen and the posterior mental foramen (or three, the third, minute, one below the incisors, similar to a *foramen nutricium*, but through which the anterior mental nerves exit to innervate the area below the incisors) [9]. Lobprise & Dodd [10] also mention three foramina: the rostral foramen in the incisive part, the middle mental foramen at the level of the labial frenulum, and the caudal mental foramen between the two roots of the third premolar. In contrast, the lingual face has only a mandibular foramen (*foramen mandibulae*) in the rostral part of the ramus for the neurovascular supply of the mandible, which enters into the mandibular canal and provides the sole blood supply (the mandibular artery (*A. alveolaris inferior*-ramus of the maxillary artery (*A. maxillaris*)—to the alveolar bone and teeth, and the mandibular vein (*V. alveolaris inferior*)) and the inferior alveolar (sensory) nerve (*N. alveolaris inferior).* According to Starkie & Stewart [11], a tough fibrous sheath surrounds the main nerve and its larger branches, and only the cat (in contrast to the rabbit and sheep) presents a sheath of compact bone surrounding the neurovascular bundle [11], a fact that in theory would help to localise the feline mandibular canal by radiography.

Pitakarnnop et al. [12] analysed 44 parameters on dried feline bones and demonstrated that there are only three hallmarks for sex identification in cats. One of them is the coronoid process of the mandibular ramus (with an accuracy rate up to 88.2%): 'looking at a lateral view of the mandible, the coronoid process in females was more curved than in males'. The other two were described in the *os coxae* [12]. Hence, the mandible could give us additional and useful information for some scientific domains, such as forensic, developmental and evolutionary sciences, and also for zooarchaeological studies.

To the outer surface of the mandible is attached the Mm. *masseter* and *buccinator* (*partes buccalis* and *molaris*); the inner surface to the Mm. *pterygoidei* (*lateralis* and *medialis*), *mylohyoideus*, *geniohyoideus,* and *geniohyoglossus*; both surfaces at the lower border to the M. *digastricus*, and the upper process (*processus coronoideus*) to the M. *temporalis* [7,8].

When the jaw is closed, the lower incisors normally strike immediately caudal to the upper incisors. The lower canine occludes between the lateral upper incisor and the upper canine. Hence, this arrangement provides a shearing action, particularly between the cheek teeth [13]. The upper and lower teeth do not touch when the jaws move in a sagittal plane. However, when the cat chews on one side of the mouth, the lower jaw must be brought to that side, so the buccal surface of the lower teeth may shear upwards and forwards against the occlusal surface of the upper teeth [13].

Bite force is generated by the interaction of the masticatory muscles, the mandibles and maxillae, the TMJs, and the teeth. The main factors affecting the bite forces in dogs and cats are body weight and the skull's morphology and size [14]. In cats, biting forces of 20–23.25 kg at the canine and up to 28 kg at the carnassial teeth have been reported [15] or, according to Kim et al. [14], an average of 73.3 Newtons (N) and 118.1 N, respectively. However, if converted to kg-force (kgf), it is obvious that these measurements do not match with [15], as they result in 7.47 and 12.04 kgf, respectively.

#### **3. Temporomandibular Joint**

It is worth pointing out that the temporomandibular joint (TMJ) is a cardinal feature that defines the class Mammalia and separates mammals from other vertebrates. Despite its status as a mammalian identifier, the TMJ shows remarkable morphological and functional variations in different species, reflecting the great adaptive diversification of mammals in feeding mechanisms [16]. During evolution, the common features of the TMJ (such as modified hinge joint, fibrocartilaginous articular surfaces, and two synovial joint compartments separated by an articular disc) persisted mostly invariable, except for a few species [17]. The simple components of the TMJ present adaptations, both in form and function, to satisfy the needs of the species, such as feeding and communication [17]. The evolutionary variants include adaptations in the orientation of the joint cavity from parasagittal (many rodents) to transverse (many carnivores), among other features [16]. However, function still remains a problem, because muscles, movements, and joint loads are to a great extent species-dependent [16].

The feline TMJ, which works as a hinge, is a synovial condylar joint formed between the condyloid process of the *ramus mandibulae* and the *fossa mandibularis* of the *pars squamosa* of the temporal bone. Caudally, the deep gutter of the temporomandibular articulation is bounded by a prominent retroarticular process placed behind the mandibular fossa [18]. Medial dissection shows a close relationship of the medial aspect of the articular capsule with the mandibular nerve, the tympanic cord, and maxillary artery. This area is particularly sensitive due to the exaggerated opening of the mouth when using different instruments or clinical exploration or dentistry [19].

It is the unique joint with a whole articular disc (*discus articularis*) that has developed owing to a slight articular surface incongruity [20]. However, in the cat, the temporomandibular joint is highly congruent [17] and the articular disc is very thin and poorly developed. According to Arredondo et al. [21], it is attached around its entire periphery to the capsule, dividing the synovial cavity into two separate spaces—dorsal and ventral. The periphery of the disc is irrigated by small branches from the articular temporomandibular artery [21]. In the cat, the condyloid process formed by the *caput mandibulae* consists of a slender, transverse roller 13–15 mm wide with a diameter of 2–3 mm (Figure 2). The axis of the articulated roller is oriented transversely at the line of the occlusal plane [17]. Caudally oriented, it presents a highly curved convexity medial to the mandibular body that narrows laterally and the outer end is often pointed [22].

**Figure 2.** Caudal view of the cat mandible. Both cone-shaped condylar processes are horizontally oriented.

The mandibular fossa, which lies under the base of the zygomatic arch and is 12–15 mm wide, has a concave *facies articularis* placed between the retroarticular process (*processus retoarticularis*), which is a caudoventral extension (medially placed) of the mandibular fossa, and a rounded and pronounced articular eminence (*tuberculum articularis*) rostrolateral to the mandibular fossa [22] (Figure 3). Therefore, in cats, the mandibular head is completely surrounded by bony structures of the temporal bone, with both eminences acting like two stops to limiting anteroposterior movements, but allowing mandibular motion to the sagittal plane of the cutting edge of the molar/premolar border P4/M1 (just opening and closing the mouth), with very limited lateral [23,24]. According to Crompton et al. [25], the retoarticular process resists the posteriorly directed force of the temporal muscle and the articular tubercle resists the anteriorly directed force of the superficial masseter muscle. When the mouth is closed, the mandibular dental arch fits into the dental arch of the maxilla and leaves no gaps to allow lateral or transverse movement. However, according to Knospe [22], the sideways movement of the lower jaw in the transverse plane occurs only when the oral cavity is slightly to strongly open, providing 2–3 mm to the right or left, with a total of approximately 5 mm, allowing crushing shears in the cat's P4/M1, splitting the jaw pressure into a vertical cutting and a transverse pressure component.

**Figure 3.** Lateral view of the cat head. The temporomandibular joint has two 'bumpers' (one rostrally and other caudally) to keep in place the condylar process in the *fossa mandibularis* of the temporal bone.

Problems or perception of pain in opening or closing the mouth should always include a complete evaluation of the bilateral TMJ. Imaging with radiographs can be challenging due to superimposition of maxillofacial/cranial structures, requiring some rotation in either the lateral (10–30◦) or long axis (10–30◦) to isolate the individual structures [10].

In addition, Gracis and Zini [23] stated that the evaluation of the vertical mandibular range of motion or range of mandibular abduction (the distance between the maxillary and mandibular incisor teeth at maximum mandibular extension) should be incorporated into every diagnostic examination, as it may be valuable in showing changes over time for a single patient (concurrently with the patient under general anaesthesia and the musculature is relaxed). Consequently, early detection of a reduction in joint mobility allows a prompt diagnosis of these limiting pathologies or conditions. Those conditions may affect intraarticular or extra-articular TMJ structures, such as ankylosis secondary to fracture, joint luxation, dysplasia and osteoarthritis, which are relatively common TMJ lesions in cats [21], or fracture, osteomyelitis, bone neoplasia, retrobulbar masses, neuromuscular diseases, and trismus [23]. It should be taken into account that in cats there is a positive correlation between the vertical mandibular range of motion and (1) body weight and (2) age. In addition, male cats can open their mouth wider than can females [23].

#### **4. Mandibular Teeth**

The cat has double dentition: deciduous and later permanent teeth. It is edentulous at birth but develops a set of deciduous teeth that start erupting between 2 and 8 weeks after birth [13]. At 60 days, the deciduous dentition is complete [14]. Between 3 and 6 months of age, the deciduous teeth are shed as the permanent teeth erupt, and their full crown height should be achieved by 10–12 months of age [13]. There are four types of teeth depending on their shape and function: incisors (I), canines (C), premolars (PM), and molars (M). The permanent teeth number 30 in all. The formula of the permanent teeth includes I 3/3, C 1/1, PM 3/2, M1/1. The deciduous formula is similar but lacks molar teeth (26 teeth in total). The incisors and canine teeth of the cat are all single-rooted, as those of the dog. The mandibular first and second premolars are normally absent. The mandibular third and fourth premolars and the single mandibular molar each have two roots. The roots of the premolar are nearly equal in size, but the mesial root of the mandibular molar is approximately three times the width of the distal root [10]. The mandibular first molar tooth is considered a carnassial tooth (as well as the maxillary fourth premolar tooth).

Regarding the simple or brachydont tooth structure, as a general overview, the crown of the tooth is covered with enamel, whereas the roots are covered with cementum. Both hard-tissue layers meet at the cement–enamel junction, near the cervical portion of the tooth. Dentine constitutes the major part of the mature tooth. The difference between enamel and dentine microhardness is a result of the percentage of mineralisation they present. Dentine is synthesised by the odontoblasts lying at the pulp's periphery. Since dentine is produced throughout life in a vital tooth, the permanent teeth of old cats have thicker dentineal walls and narrower pulp cavity compared to those of young cats [26]. The enamel thickness is reported to be 0.1–0.3 mm in cats and 0.1–0.6 mm in dogs [10] and, additionally, cat enamel is less hard than that of dogs, according to Hayashi & Hideo [27]. They also reported that enamel microhardness is higher in the outer layer than in the central or inner layer (since calcium ions in saliva infiltrate from the tooth surface), and there is an age-related increase in the microhardness of enamel [27]. In the premolars, the enamel hardness is higher at the top and middle of the crown, and it decreases in the cervical portion. Similarly, the dentine is harder in the cusp than in the rest of the tooth and the dentine microhardness decreases from the outer to the inner part. They also reported that the comparative microhardness enamel/dentine ratio varied from 3–9/1 [27]. Hence, data support that cat teeth are far more fragile compared to dog teeth; that being so, special care should be taken when manipulating.

The dental root is inserted in the dental alveolus and kept in place by the periodontium, which is made up of the gingiva, periodontal ligament, cementum, and alveolar bone. The space between the tooth and the free gingiva is the gingival sulcus, which should be no deeper than 0.5 mm in cats. The periodontal ligament attaches the root to the alveolar bone [26]. The alveolar bone appears with tooth eruption and disappears with tooth loss [14]. It surrounds the alveolar socket with an extension of cortical bone into

the alveolus that outlines a radiopaque lamina dura in radiographic images [26]. The dental sac contacting with the cementum forms fibroblasts that produce collagen fibres while the other components of the periodontal ligament are developing. These are blood vessels, lymphatics, nerves, and various types of connective tissue cells. The nerves of the periodontal ligament are important as they provide additional senses to the tooth. They harbour pain fibres (similar to the pulp), but also pressure, heat, and cold fibres (not present in the pulp) [10].

However, as the aim of this review is not feline dentistry, specialised textbooks should be consulted for further information regarding teeth. Nonetheless, the more frequent dental pathologies seen in cats should be mentioned: periodontal disease and tooth resorption. Periodontal disease is perhaps the most common oral disease seen in dogs and cats and involves the periodontium and is the major cause of tooth loss. In general terms, periodontal disease is caused essentially by the accumulation of plaque on the tooth surfaces, and the severity correlates directly with the quantity of such deposits, producing gingivitis and forming a periodontal pocket. These changes lead to destruction of the gingival tissue with recession of the gums and retraction of the periodontal membrane. Ultimately there is infection of the dental root, destruction of alveolar bone, and dental loss. It is clear that a soft diet correlates positively with periodontal disease [13]. Cats suffer severely with progression of periodontitis and can be seriously anorexic. Gingival recession is common with marked tooth loss as a result of external tooth resorption, where the inflammation at the cement–enamel junction leads to collapse of the tooth crown with retention of the roots. These roots may become chronically inflamed or cystic and be very painful with periodontal abscessation. Treatment consists of removing degenerated teeth and any retained roots [13].

Dental disease is more common in older cats and can lead to other health problems, so maintaining oral health is important. To prevent dental disease, the single most effective method is to brush the cat's teeth daily with a pet-specific toothpaste or powder, although it can sometimes take some time to train them to allow their teeth to be brushed.

It must be taken into account that the morphology of the mandible is conditioned by food habits [28]. As a carnivore, the cat has a mastication pattern consisting of an up and down or hinge movement of the mandible. Their TMJ lies on or close to the same plane as that of the lower dentition, with molars designed for crushing and slicing [25]. In contrast, in the herbivore group, which includes the Ungulates, the main action consists of a grinding movement of the mandible [28], as their TMJ lies above the occlusal plane of the maxillary dentition. In carnivores, the combination of a tall coronoid process and extension of the skull posterior to the TMJ permits both a large gape and a powerful bite. Another characteristic feature of this group is that the temporal muscle is considerably stronger than the masseter muscle [25].

#### **5. Mandibular Neurovascular Supply**

The first vessel to leave the maxillary artery in carnivores is the a. *articularis temporomandibularis* destined to the mandibular joint. The a. *alveolaris mandibularis* arises in a rostrolateral direction from the first part of the maxillary artery and runs towards the mandibular foramen through which it enters into the mandibular canal and provides blood supply to the mandible. Within this canal, it gives off the *rami dentales* to the molar and premolar teeth. Other branches pass through the alveolar canal to the canine and incisor teeth [29]. It exits at the caudal, middle, and rostral mental foramina to supply the lower lips [6]. In carnivores, the *rami mentales* leave the mandibular canal trough the mental foramina and ramify in the region of the *margo interalveolaris* in the gingiva of the incisor teeth and in the lower lip [29]. Veins often exist concurrently with arteries and empty by way of the maxillary and linguofacial veins into the external jugular vein [6].

In the mandibular canal, when examined in cross-section, the mandibular nerve is located in the dorsolateral portion of the canal with the vein in the ventromedial portion and the artery in the middle [10], although according to Davis & Story [30] the a. *alveolaris inferior* is situated lateral to the inferior alveolar nerve as the two enter the foramen.

The maxillary and mandibular branches (*nervi maxillaris* and *mandibularis*) of the trigeminal nerve (nervus trigeminus) are sensory, but the mandibular branch also supplies motor function to the masticatory musculature (temporal muscle, masseter muscle and the lateral and medial pterigoid muscles, which close the mandible) and other muscles. The digastric muscle is the only one responsible for opening the jaws; its rostral belly is stimulated by the mandibular branch of the trigeminal nerve, as the caudal belly is innervated by the facial (VII) nerve [10,31]. The facial nerve also provides motor function to many cutaneous facial muscles and is responsible for taste in the rostral two-thirds of the tongue [6]. The inferior alveolar nerve contains afferent fibres from the ipsilateral lower lip, areas of oral mucous membrane and mandibular teeth. It supplies sensory innervation of the lower teeth and, after exiting the mental foramina as the mental nerves, it innervates the lower lip. According to Robinson [9], fibres supplying the teeth are found in all branches, except the mental ones. At the mandibular foramen, the inferior alveolar nerve is a single bundle with the mandibular artery and vein lying inferior to it. Within the mandible, the nerve divides into several branches, which conform to a basic pattern with some individual variation. Three branches, splitting dorsally the main trunk, supply the alveolar processes (the posterior, middle, and anterior alveolar branches); another supplies the canine and incisor regions (canine/incisor branch), and there are four mental branches (posterior, main and two anterior) that leave through the various mental foramina. Interconnecting fibres are often seen between these principal branches [9]. There is no apparent bilateral symmetry and all the branches contain fibres from at least two adjacent tissues, including afferences from the pulp, periodontal ligament, mucous membrane, and skin. In addition, the nerves supplying one tooth come from different branches because they do not travel in a unique branch of the main trunk (Figure 4). The most proximal branch splitting from the superior aspect of the main trunk is the posterior alveolar branch. It contains afferents from the molar, fourth premolar, and occasionally third premolar teeth [9]. The next alveolar branch is the middle alveolar branch, leaving dorsally the main trunk beneath the distal root of the molar tooth to supply the third premolar, although in 50% of specimens it also supplies the canine teeth and the third incisor. The last alveolar branch is the anterior alveolar nerve, which supplies the canine and third incisor teeth in addition to mechanoreceptor afferents from the mucous membrane and skin adjacent to these teeth. Occasionally it also carries fibres from the second and first incisors and the third premolar teeth. Other branches splitting laterally from the main trunk, such as the posterior and main mental, leave the mandible through the posterior and main mental foramina, respectively, to supply the buccal gingival margin, the mucous membrane on the labial side of the alveolar process and the skin of the chin and lip from the anterior to caudal part (just up to the rostral edge of the molar tooth). The anterior mental branch divides in two terminal branches: the canine/incisor nerve and the anterior mental nerve. The latter does not contain any pulpal or periodontal afference. Lastly, the canine/incisor branch carries fibres from the canines and all three incisors.

Relative to the existence of transmedian innervation, Robinson [9] said that there is no evidence of transmedian innervation of tooth pulps; nonetheless, the cutaneous innervation in the anterior mental nerve crosses the midline for 1–2 mm. In contrast, Anderson & Pearl [32] reported the existence of an extensive transmedian innervation of the teeth in the cat; an innervation which is particularly dense in the canine teeth and extends at least as far laterally as the third premolar teeth. Wilson et al. [33], using the horseradish peroxidase technique, reported that, in addition to the inferior alveolar nerve, the nerve to mylohyoid and possibly other accessory neural pathways is involved in incisor innervation in cats.

**Figure 4.** Innervation of the cat mandible modified from Robinson [8]. Diagram showing the inferior alveolar nerve through the mandibular canal and its division in *rami* to innervate different structures. Dashed lines indicate inconstant afferences. The thickness of the line when detaching from the inferior alveolar nerve is proportional to its real thickness, according to Robinson [8].

According to Izumi et al. [34], ligation or cutting of the inferior alveolar nerve always elicits an increase in gingival blood flow. They also reported that blood flow in the cat gingiva and periodontal ligament is controlled by sympathetic α-adrenergic fibres for vasoconstriction; regarding vasodilatation, sensory fibres are involved besides the mast cells of the gingiva. According to Skerrit [31], a unilateral deficit of the mandibular nerve results in weakness of the chewing muscles when biting and atrophy of the temporal and masseter muscles. When the deficit is bilateral, it leads to drooping of the mandible and inability to close the mouth [31].

Mental nerve block, when needed, should be done at the main or middle mental foramen and anaesthetises the buccal soft tissues and the mandibular incisors and canine on the side injected [35]. The mandibular nerve can be anaesthetised by intraoral or extraoral techniques. Anaesthesia of the nerve results in desensitisation of the mandibular body, the lower portion of the mandibular ramus, all mandibular teeth on the same side, the labial/buccal surfaces of the mandible, and the mucosa and skin of the lower lip and chin [35].

The mandibular canal is not a medullary canal, and treating fractures of the body via an intramedullary pin through this canal will damage the associated neurovascular bundle. In many fractures or tumoural reparative surgery of the mandible, the inferior alveolar nerve is damaged and then resected. However, some experiments provide evidence that peripheral nerve fibres are important, not only in normal bone homeostasis and skeletal growth, but also in their influence on the repair mechanism of bone fracture. Many experiments suggest that sensory and sympathetic nerve fibres do have a role in bone remodelling and osteogenic differentiation of precursor cells during skeletal growth. Hence, the loss of sensory nerves could result in a decrease in the quality of new bone, as reported by Cao et al. [36] on the mandibular distraction osteogenesis in rabbits, stating that the peripheral sensory nervous system plays an important role in bone regeneration. Sensory nerves also play an important role in regulating bone resorptive activity, as shown by Yamashiro et al. [37] during experimental tooth movement in rats. In bone, the areas with the highest metabolic activity receive the richest sensory and sympathetic innervation, which has an effect on the activity of both osteoblasts and osteoclasts [37]. As seen in many articles, the inferior alveolar nerve could also be affected by reparative osteosynthesis, and as in the feline mandible there is very little space to safely place the screws required to

fix the metal plates to resolve the fracture. Regarding the integrity of the inferior alveolar nerve, we consider a conservative option should be prevalent over inferior alveolar nerve resection so as to not impair the outcome of bone repair and sensitivity. Hence, we propose that surgeons should be as conservative as possible and try to leave the nerve intact, with no attempt to pull the nerve as this causes a lesion, and being sure where to place the screw in order to not affect the neurovascular bundle or any dental root.

#### **6. Mandibular Radiographic Images**

Dental radiography requires general anaesthesia to get accurate projections and avoid any trauma or damage to the equipment.

Cats have two mandibular halves that connect rostrally through a mandibular symphysis. The symphysis is represented radiographically as a radiolucent border between the two mandibles. The portions of the mandibles associated with the symphysis are roughly parallel [38].

The lateral projection of the feline mandible (Figure 5) is bordered ventrally by the ventral cortex and dorsally by the cusps of the premolars and molar. The area corresponding to the location of the mandibular canal, containing the neurovascular bundle (the alveolar mandibular nerve, artery and vein) appears as a radiolucent area just dorsal to the ventral cortex and ventral to the dental roots. When evaluating any tooth, the following are assessed: crown (and enamel) and pulp chamber, root and root canal, periodontal ligament space and alveolar bone. The relative radiolucent line outlining the roots (*lamina lucida*) is the periodontal ligament space. It is wider early in age and is typically widest at the coronal and apical one-third of the root [38]. Adjacent to this ligament space is a radiodense line (*lamina dura*), which is the cortical bone of the alveolus. Contiguous to it is the trabecular bone of the alveolus. The crown is covered by a more radiodense margin, which is the enamel, and the bulk is dentine, which is not as radiodense as enamel but is radiodense compared to bone. Because the cementum has nearly the same radiodensity as bone, it is not obvious radiographically. In multirooted teeth, bone should be present up to the apex of the furcation. The centre of the root is the radiolucent root canal, which houses the radicular portion of the pulp.

**Figure 5.** Postmortem radiography of a cat head, laterolateral projection (left lateral recumbency). The yellow outline corresponds to the left mandible and the orange one to the right jaw. The respective coloured empty arrows point the mandibular foramina, and the small arrows indicate the main mental foramina. The orange circle shows signs of tooth root resorption and the white arrow points evidences of a dental root fracture. Note that a plastic needle cap was used to keep the mouth open. As plastic is radiolucent, it does not interfere with the radiologic image.

According to Milella & Smithson [39] ten radiographic films should be taken to assess accurately each tooth in the cat's mouth (upper incisors, upper left canine (anteriorposterior oblique and lateral), upper right canine (anterior-posterior oblique and lateral), upper left maxillary premolars and molar, upper right premolars and molar, lower canines and incisors, lower right mandibular teeth, lower left mandibular premolars and molar). However, in older cats, full mouth radiographs are more to be recommended.

Hoffman & Ridinger [40] pointed out the value of obtaining annual dental radiographies, as well as the importance of interpreting dental radiographic findings in the light of the patient's systemic health and oral examination findings. This information will be essential when planning treatment in differentiating between primary oral disease and those secondary to systemic disease.

#### **7. Topographical Considerations**

The arteries of significance in clinical dentistry and oral surgery have the same origin in the common carotid artery (*arteria carotis comunis*), which branches into the internal and external carotid arteries (*arteriae carotis externa* and *interna*) [10]. The internal carotid artery blood supply is insignificant in the cat. Consequently, the external carotid artery, continuing as the maxillary artery, provides the majority of cerebral blood flow in the cat. The maxillary artery lies medial to the angular process of the mandible and branches into the maxillary rete before entering the skull through the orbital fissure [10]. Exclusively in the cat, the maxillary artery forms a network known as *rete mirabile a. maxillaris*, which is extracranial, near the *foramen ovale*. This network extends dorsally and laterally to the apex of the periorbital region. In the cat, the arteries for the eye and accessory structures arise from this network as do also the *rami retis*, which pass through the *fissura orbitalis* to connect with the *circulus arteriosus cerebri*. At the same time, the maxillary artery itself, being traceable through this network as a stronger vessel, leads the *a. infraorbitalis* [28].

According to Skerritt [41], there is an inverse relationship between the degree of development of the internal carotid and that of the anastomosing ramus of the maxillary artery: when one is large, the other is small. In no species are both of these channels fully developed. In cat and sheep, the lumen of the internal carotid artery becomes obliterated in the weeks or months after birth (although at birth it is fully functional). As a result, the whole of the adult brain is supplied by maxillary blood via the anastomosing ramus of the maxillary artery. In addition, a *rete mirabile a. maxillaris* occurs on the anastomosing ramus of the maxillary artery in all species in which the supply from the maxillary artery is well developed [41].

The significance of the *rete mirabile* has long been discussed. It was thought that it might eliminate pulsation before the blood reaches the brain itself. However, more recent observations indicate that the rete is involved in thermoregulation, as reported by Baker & Hayward [42]; using different anatomical nomenclature, they stated that the plexus of arteries that make up the carotid rete in the cat seems to be able to modify the temperature of central arterial blood as it enters the cranial cavity. In this short paper (only three pages) published in Nature, they described that in the cat, the carotid rete lies outside the cranial cavity, near the apex of the orbit, and is termed an 'extracranial rete', while in artiodactyls the rete is intracranial, lying in the cavernous sinus at the base of the skull. The extracranial rete lies within a venous lake, and the large surface area of the interlacing network of vessels may permit a 'countercurrent' exchange of heat between the arterial blood of the *rete* and the venous return from various regions of the head, having a profound effect on brain temperature that may have significant thermoregulatory consequences, allowing large, rapid changes in the temperature throughout the cranium (0.1–0.7 ◦C). The presence of the extracranial carotid rete is the most important cerebrovascular difference between the cat and the other species studied by Baker & Hayward [42].

Consequently, in the cat, maxillary arterial blood is distributed to all of the brain and any disturbance of the blood flow may have dramatic consequences. Regarding this aspect, it is currently proven that overextension of the mandibles of the cat (during oral and transoral procedures, such as intubation) can lead to compression of the rete and/or compression of the maxillary artery by the angular process of the mandible (Figure 6), leading to cerebral ischaemia and resulting in temporary or permanent cortical blindness, loss of hearing, or possibly death [43,44].

**Figure 6.** Caudolateral view of the cat head with the mouth open. The white arrow points to the channel narrowed by the *processus angularis*, through which the *a. maxillaris* runs to lead the *rete mirabile.*

A comprehensive understanding of the functional anatomy of all structures associated with the caudal angle of the cat mandible may explain why keeping a cat's mouth wide open for a prolonged period of time may result in temporary or permanent neurological deficits, unilaterally or bilaterally, post-anaesthesia [26]. Cats and dogs have the maxillary artery running through this area, but what gives only cats an increased risk of cerebral ischaemia when the mouth is wide open is that the mandibular angular process, presses against the area through which the maxillary artery passes. This reduces to some extent the maxillary artery blood flow. However, an additional feature is decisive to this occurrence: the internal carotid artery (leading the main blood supply to the brain, retina and inner ear) is functionally absent in cats, so all the blood to the brain is supplied exclusively by the bilateral maxillary arteries. Consequently, the longer the duration of the pressure, the higher the risk of onset of cerebral ischaemia and/or blindness and/or deafness. De Miguel García et al. [44] suggested (where the use of a mouth gag is essential for surgery) to reduce the size of the gag or to close the cat's mouth every few minutes (throughout the procedure) to enable restoration of the blood supply. However, the reason why cats are unequally affected is still unknown, although it could be due to collateral or altered blood flow through the basilar arteries (although, according to Skerritt [41], they only carry blood away from the arterial circle).

Thus, the use of spring-loaded mouth gags is no longer recommended in feline patients [10,26] as they apply continual force to keep the mouth open to such an abnormal degree [45]. It is also reported that the vascular flow is more compromised on the side ipsilateral to the mouth gag, perhaps because the distance between the angular process of the mandible and tympanic bulla is smaller on the ipsilateral side [46]. Another drawback would be that the tighter the lips and cheek become when the mouth is wide open, the more difficult it is to retract them, to perform surgery or a thorough oral examination [26]. In contrast, custom-made plastic mouth props (made from a syringe cap) have many benefits: (1) they are gentler on the jaws and appear to induce fewer alterations in blood flow; and (2) they are radiolucent, hence they do not interfere with diagnostic radiological imaging [26], among others. However, if a mouth gag must be used in a feline patient, the smallest possible gag (ideally between 20 and 30 mm), should be chosen to retract the upper and lower lips without difficulty and the duration of use minimised [26,47].

The Nomina Anatomica Veterinaria [7] does not clarify which species present molar salivary glands (*glandulae molares).* Okuda et al. [48] described the bulge lingual to the lower molar tooth to be a small salivary gland corresponding to the lingual molar gland. Interestingly, there is no equivalent salivary gland in the dog. Previously, Orsini & Hennet [14] stated in their review entitled 'Anatomy of the mouth and teeth of the cat' that 'just medial to the lower first molar on the floor of the oral cavity is a mass-like flap of oral mucosa with no known function; its prominent appearance leads to an incorrect identification as an abnormal finding'. This major salivary gland in the cat has two parts (Figure 7): the buccal and lingual molar gland. The secretory portion of the buccal molar gland is close to the commissure (between the M. *orbicularis oris* and the mucous membrane of the lower lip at the angle of the mouth) and it empties into the buccal cavity by several small ducts; and the lingual molar gland located within a membranous bulge caudolingual to the mandibular molar tooth (constituting the molar pad) [10]. Sometimes, the membranous molar pad enlarges and may be traumatised (when chewing), being important not to be mistaken for a tumour or polyp [49].

**Figure 7.** The cat molar salivary glands. (**a**) Ventrolateral view of the buccal molar gland. (**b**) Frontal view of the lingual molar gland, caudomedially to the first molar tooth.

#### **8. Conclusions**

Clinicians must have a deep knowledge of the functional anatomy of cats (taking into account that they present some important differences with respect to the dog) to achieve an effective and high-quality cat medicine.

A good knowledge of the anatomy of the mandible and the TMJ, and their relation to other important structures, such as blood vessels and nerves, is essential for an accurate interpretation of radiographic images and tomographic diagnostic techniques, in order to make a diagnosis and achieve good results in the management of different conditions.

Clinicians and surgeons are increasingly aware of animal welfare to avoid or, at least, minimise any suffering due to iatrogenic complications. Such as those related to a temporary or permanent blindness and/or deafness following general anesthesia when the mouth is held open with a gag (to the compression of the angular process on the maxillary artery, disrupting the blood flow to the maxillar *rete mirabile* and the brain); the resection of the lingual molar pad (that includes the lingual molar salivary gland), or those complications secondary to dental extraction.

**Author Contributions:** Conceptualization, M.L. and M.d.M.Y.; Writing—Original Draft Preparation, M.L., D.A.-P. and M.d.M.Y.; Writing—Review & Editing, M.L., D.A.-P. and M.d.M.Y. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The Institutional Review Board statement was not required in this study as it has been carried out with cat cadavers, previously euthanised due to other pathologies. An oral consent was given by the owners allowing to use their dead pets for research purposes.

**Data Availability Statement:** Data sharing is not applicable to this article as no new data were created or analyzed in this study.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Review* **The Cat Mandible (II): Manipulation of the Jaw, with a New Prosthesis Proposal, to Avoid Iatrogenic Complications**

**Matilde Lombardero 1,\*,†, Mario López-Lombardero 2,†, Diana Alonso-Peñarando 3,4 and María del Mar Yllera <sup>1</sup>**


**Simple Summary:** The small size of the feline mandible makes its manipulation difficult when fixing dislocations of the temporomandibular joint or mandibular fractures. In both cases, non-invasive techniques should be considered first. When not possible, fracture repair with internal fixation using bone plates would be the best option. Simple jaw fractures should be repaired first, and caudal to rostral. In addition, a ventral approach makes the bone fragments exposure and its manipulation easier. However, the cat mandible has little space to safely place the bone plate screws without damaging the tooth roots and/or the mandibular blood and nervous supply. As a consequence, we propose a conceptual model of a mandibular prosthesis that would provide biomechanical stabilization, avoiding any unintended (iatrogenic) damage to those structures. The improvement of imaging techniques and a patient-specific prosthesis made of full biocompatible material are part of the future trends to improve patients' recovery.

**Abstract:** The cat mandible is relatively small, and its manipulation implies the use of fixing methods and different repair techniques according to its small size to keep its biomechanical functionality intact. Attempts to fix dislocations of the temporomandibular joint should be primarily performed by non-invasive techniques (repositioning the bones and immobilisation), although when this is not possible, a surgical method should be used. Regarding mandibular fractures, these are usually concurrent with other traumatic injuries that, if serious, should be treated first. A non-invasive approach should also first be considered to fix mandibular fractures. When this is impractical, internal rigid fixation methods, such as osteosynthesis plates, should be used. However, it should be taken into account that in the cat mandible, dental roots and the mandibular canal structures occupy most of the volume of the mandibular body, a fact that makes it challenging to apply a plate with fixed screw positions without invading dental roots or neurovascular structures. Therefore, we propose a new prosthesis design that will provide acceptable rigid biomechanical stabilisation, but avoid dental root and neurovascular damage, when fixing simple mandibular body fractures. Future trends will include the use of better diagnostic imaging techniques, a patient-specific prosthesis design and the use of more biocompatible materials to minimise the patient's recovery period and suffering.

**Keywords:** anatomy; feline; lower jaw; mandibular fracture; neurovascular supply; temporomandibular joint; tooth

#### **1. Introduction**

As seen in the first part (The cat mandible (I) [1]), a deep knowledge of the functional anatomy of the feline mandible will be the basis for interpreting the diagnostic images that,

**Citation:** Lombardero, M.; López-Lombardero, M.; Alonso-Peñarando, D.; Yllera, M.d.M. The Cat Mandible (II): Manipulation of the Jaw, with a New Prosthesis Proposal, to Avoid Iatrogenic Complications. *Animals* **2021**, *11*, 683. https://doi.org/10.3390/ani11030683

Academic Editor: Ralf Einspanier

Received: 21 December 2020 Accepted: 26 February 2021 Published: 4 March 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

in combination with symptoms, will help to achieve an accurate diagnosis about the current pathology. Afterwards, the clinician should come to a decision and a treatment has to be prescribed. A meticulous therapeutic decision-making process is essential to choose among alternatives and analyse consequences, as some treatment procedures include surgery.

Hence, the present review will be focused on how to solve different pathological situations of the feline mandible and temporomandibular joint, such as luxation or fractures. In addition, we will insist on the importance of keeping the integrity of the tooth roots and the mandibular canal neurovascular bundle, as they could be damaged during fracture repair.

#### **2. Temporomandibular Joint Luxation**

The temporomandibular joint (TMJ) is characterised by the presence of a whole intraarticular disc, although is very thin in the cat and in the dog. Many elements are related to this synovial condylar joint and gaining a comprehensive knowledge of this joint is required to correctly interpret its diagnostic images. This ability is essential in order to make a diagnosis and to achieve good results in the management of different conditions, minimising the incidence of surgical iatrogenic lesions. While TMJ caudal luxation might occur with fractures in the retroarticular process, mandibular fossa or mandibular head, rostral mandible movement is more common [2]. Unilateral rostrodorsal luxation of the mandibular condyle causes the lower jaw to shift laterorostrally to the luxation-opposite side and the inability of the cat to close its mouth fully due to tooth-to-tooth contact. This distinguishes it from open-mouth jaw locking, where the mouth is held wide open with no contact between mandibular and maxillary teeth [3]. The unilateral rostrodorsal luxation of the mandibular condyle could easily be reduced by placing a pencil between the maxillary fourth premolar (PM4) and the mandibular first molar (M1) on the affected side only and closing the mandible against the pencil while simultaneously relieving the jaw caudally (easing the condyle from the articular eminence). However, this treatment is contraindicated in animals with open-mouth jaw locking [3].

Open-mouth jaw locking is characterised by an inability to close the mouth that usually results from fixed mandibular coronoid process displacement lateral to the ipsilateral zygomatic arch and abnormal contact pressure between these two structures [2]. The history is important in the diagnosis, as it is usually observed after animals have yawned, groomed or vocalised [2]. Physical findings of a wide-open mouth and palpable coronoid process superficial to the zygomatic arch help to distinguish open-mouth jaw locking from TMJ dislocation [2]. Manual reduction is the first-line treatment method in open-mouth jaw locking, secondary to coronoid process zygomatic arch interlocking and temporomandibular dislocation. Resolution may be spontaneous or require manual correction, but recurrence is possible [2]. However, and according to Reiter and Lewis [3], the treatment consists of two phases: an acute treatment (under sedation) consisting of opening the jaw even further to release the coronoid process from the lateral aspect of the zygomatic arch, and then closing the mouth. A tape muzzle should be placed until definitive surgery is performed. This consists of a partial resection of the coronoid process, partial resection of the zygomatic arch, or a combination of both [3].

#### **3. Mandibular Fractures**

#### *3.1. General Considerations*

Mandibular fractures are commonly seen in practice, comprising up to 6% of all fractures in dogs, and 11–23% of all fractures in cats, according to Glyde and Lidbetter [4]. They occur more frequently than maxillary fractures [3]. The majority of mandibular fractures in cats may result from road traffic accidents, fighting injuries or falls from heights [4–6] and, unfortunately, also due to human abuse [7]. A traumatic aetiology commonly involves serious concurrent injuries requiring prompt clinical attention, mainly to the brain, maxilla and chest [8]. Management of life-threatening injuries and normalisation of patient physiology is required before surgical stabilisation of mandibular fractures [8]. Less commonly, pathological mandibular fractures may occur secondary to periodontal, oral neoplasia or metabolic disease, and iatrogenic fractures can also occur during dental treatment [3,4]. Although various types of injuries and trauma are typically responsible for fractures of the upper (maxilla) and lower jaw, certain risk factors may predispose a cat to fractures, including oral infections (e.g., periodontal disease, osteomyelitis), certain metabolic diseases (e.g., hypoparathyroidism) and congenital or hereditary factors resulting in a weakened or deformed jaw [9].

Independently of the aetiology, in cats, and according to Umphlet and Johnson [10], mandibular fractures accounted for 14.5% of all fractures seen in a total of more than 500 cat specimens (*n* = 517). Symphyseal fractures were the most common (73.3%), followed by fractures of the body (16%), condyle (6.7%) and coronoid process (4%). Complications developed more commonly in cats with multiple or open fractures. Clinical union occurred after an average of 6 weeks (range 3–12 weeks) for symphyseal fractures, 10 weeks (range 8–16 weeks) for body fractures, 6 weeks for coronoid fractures and 6 weeks (range 4–8 weeks) for condylar fractures [10]. In contrast, in dogs, fractures in the premolar region are significantly more frequent than in other regions [11]. Umphlet and Johnson [11] also reported that fractures in the rostral portion of the mandible had shorter average time to clinical union than did other mandibular fractures. However, the average time to clinical union for fractures in the caudal portions of the mandible was longer than that currently reported [11]. Nonetheless, overall prognosis depends on type, extent, location of trauma, quality of home care and selection of treatment modality [9].

Accordingly, Little [12] reported that mandibular fractures in cats are typically located in the area of mandibular symphysis or the mandibular ramus (fractures of the condylar process or coronoid process). The midportion of the mandibular body is less frequently fractured in cats.

A non-invasive approach should receive primary consideration, and an invasive option is employed only if non-invasive treatment is insufficient or impractical [13]. Tape muzzling is a non-invasive and inexpensive treatment option for mid-body and caudal mandibular fractures and for TMJ luxation and open-mouth jaw locking after manual correction [13]. Tape muzzles could be used as temporary, definitive or adjunctive therapy. This is a good method in cases of pathological fractures or where the bone is very porous and will not support a fixative device. Where fractures are stable, this is also a good technique [14]. It may be curative for mandibular fractures in immature, adolescent and young adult animals with good bone healing capacities. In addition, muzzling allows some TMJ movement, thus reducing the risk of ankylosis between fractured bones in that area [13].

In the cat, when the jaw is immobilised to allow healing of the fracture, the mouth should be kept open no less than 5 mm and no more than 10 mm (as measured between the incisal edges of the maxillary and mandibular incisors) to allow for the tongue to protrude and lap water and a slurry diet. If the mouth is open too far, it will result in difficulty in swallowing [13]. This immobilisation of the mandible, to limit oral opening, could be done with a tape muzzle, fixation with composite spanning the ipsilateral canine teeth, or through labial buttons placed with suture material [15].

Southerden et al. [16] reported that there is a low level of asymmetry between contralateral mandibles in cats, allowing the use of a mirror image of an intact mandible for planning and evaluating the accuracy of fixation of a contralateral mandible. The most consistent measurement among 27 specimens was the lateral ramus inclination angle. However, the least consistent measurements were ramus height and jaw width at the mental foramen [16]. This type of study may facilitate the development of standardised pre-contoured locking plates for cat mandible repair.

Regarding mandibular biomechanics, as reported by Spodnick and Boudrieau [17], a continuity of tensile to compressive stresses exists from one side of the bone to the other during bending stress. Maximal tensile stresses are present at the oral (alveolar) surface and maximal compressive stresses at the aboral (ventral) surface; therefore, distraction is created

at the oral margin. These bending moments increase from caudal to rostral; furthermore, shear forces are maximal at the ramus, while rotational forces are most prominent rostral to the canine teeth and maximal at the mandibular symphysis [17]. Taking these biomechanics into account would be of great help when fixing a mandibular fracture. Consequently, invasive jaw fracture repair techniques (osseous wiring, external skeletal fixation and bone plating) should be carefully planned and be accompanied by dental radiography, both intraand post-operatively [16].

#### *3.2. Symphyseal Fractures*

Mandibular symphysis fractures are the simplest and are best treated with cerclage wire [4]. Palpable instability of the symphysis is not pathognomonic for traumatic symphyseal separation, as instability may result from periodontal disease, laxity of the ligamentous attachment, neoplasia, or fracture of the mandible [18]. Relative to the placement of a circumferential wire for mandibular symphyseal fracture repair, and according to Glyde and Lidbetter [4], it is necessary to place a hypodermic needle to act as support, making a hole from the oral cavity (from the caudolateral edge of the canine tooth) and leave the wire end at the level of a skin incision in the intermandibular space; however, on the other side, the needle is introduced from the bottom to the oral cavity to allow removing the needle once the wire is in place. This is a logical sequence, taking into account that the wire should make a loop to press the two mandibles together and reassure the mandibular symphysis. Make sure that the incisor teeth remain in alignment; otherwise, step defects can be generated [15]. The wire should be removed once union has been achieved [4].

Parasymphyseal fracture is a common iatrogenic injury during extraction of the mandibular canine tooth in cats [15]. The fracture may occur due to pre-existing periodontal or endodontic disease, insufficient preparation prior to extraction, or excessive force used by the operator, or a combination [15,19]. It turns into an important pathology that remains undetected if postoperative radiographs are not obtained, as the fracture often is nondisplaced [19]. As a recommendation of good practice, Hoffman [20], a diplomate of the American Veterinary Dental College and board-certified in veterinary dentistry, advised that two dental X-rays should be always taken: (1) before extractions (this will allow the veterinary dentist to assess the health of the bone and the anatomy of each tooth, including its roots, taking into account that advanced dental disease contributes to bone loss and increased risk of iatrogenic trauma), and (2) after an extraction to ensure that the entire dental root has indeed been removed. In addition, the client should be informed upfront that a jaw fracture is a possibility, to avoid difficulties regarding iatrogenic jaw fractures secondary to tooth extraction. Furthermore, her advice is that once you are faced with an iatrogenic mandibular fracture, the case should be always referred to a specialist [20]. Parasymphyseal fractures could be treated with circumferential wire [15].

#### *3.3. Body Fractures*

When assessing the mandibular body for fractures, the direction of the fracture and the location of the fracture in relation to the dental roots should be evaluated. From a biomechanical point of view, fractures could be simple (in which the fracture line is perpendicular to the long axis of the mandible), or oblique fractures, that may be described as favourable or unfavourable according to the difficulty of immobilisation. This distinction results from the forces that the muscles of mastication place on the mandible as they either compress (favourable) or distract (unfavourable) the fracture segments [15,21] (Figure 1). Hence, a fracture that travels dorsocaudal to ventrorostral is considered favourable, whereas a fracture that travels dorsorostral to ventrocaudal and distracts the fracture fragments is considered unfavourable [15,21,22]. Favourable fractures compress because the upward pulling of the masseter and temporalis muscles and the downward and caudal pulling of the digastricus will hold the fracture segments in apposition, to a large extent. They are relatively stable, and stabilisation of the tension surface may be all that is required for bone healing [21]. In contrast, in unfavourable fractures, the alveolar crestal bone is considered

the tension surface, while the ventral cortex is considered the compression surface [21], and the muscular forces will lead to considerable displacement of the fracture segments [3].

**Figure 1.** Drawings depicting the lateral view of the feline left mandible with a favourable fracture (**a**), as it compresses the fracture fragments, and an unfavourable fracture (**b**) in which the fracture segments are distracted.

As far as fracture repair is concerned, it is extremely important to take notice of a large percentage of the mandibular body being occupied by deep dental roots—between 50% and 70% of its dorsomedial depth [4]—and the neurovascular structures (along the mandibular canal), facts that largely limit the areas for safe implant/screw placement (i.e., for plate fixation). This is particularly true in cats, where only a small amount of 'free' bone exists rostrally to the first premolar and caudally to the molar tooth [4]. Radiologically, the mandibular canal can be placed parallel and just coronal to the ventral cortex (the radiopaque bone in the ventral mandibular body margin) as a thick, horizontal radiolucent line in close apposition to the mandibular premolar and molar dental root apexes [18,23]. In that sense, Bellows [21] advises "Unless you've had advanced training, avoid plating jaw fractures for fear of compromising dental roots. Also avoid placing intramedullary pins into the mandibular canal. The mandibular canal carries the neurovascular structures—it is not an intramedullary canal". Accordingly, one myth stated by Hoffman et al. [24], to dispel common misconceptions relative to "intramedullary pins in the mandibular canal are an option for treatment of fractures of the mandibular body", said that experimentally, inserting pins into the body of the mandible results in delayed healing and considerable damage to many dental roots, whereas clinically, malalignment is a common complication.

As stated by Lantz [25], the principles of facial repair include (1) restoration of occlusion and anatomic reduction, (2) stable fixation to neutralise detrimental forces on the fracture line(s), (3) preservation of blood supply, (4) preservation of soft tissue attachment to bone fragments by gentle tissue manipulations and minimal tissue elevation, (5) avoidance of iatrogenic dental trauma, not injuring the dental roots, and (6) extraction of diseased teeth at the line(s). The selected method of repair should provide occlusion and, ideally, rigid stability of all major fragments. In addition, the device should allow immediate return to oral function, be lightweight, economical and not cumbersome for the patient. However, Spodnick and Boudrieau [17] reported that removal of teeth may increase complications due to disruption of the blood supply and iatrogenic trauma to the adjacent tissues, including further displacement of the bone fragments, elimination of occlusal landmarks useful in realigning bone segments to allow functional occlusion, elimination of available structures for use in the fixation of bone fragments and creation of a large bony defect adding to the difficulty of reduction and stabilisation. Preservation of teeth involved within a line in a mandibular fracture has also been reported to have a favourable prognosis if optimal reduction and stabilisation of the jaw has been achieved. Therefore, removal of teeth is not advised unless the teeth involved are fractured (even here universal removal is not recommended if the tooth contributes to stabilisation and if the fracture of the tooth does not involve the root). Nonetheless, advanced periodontitis or periapical abscessation are situations in which in-line teeth should be removed since they have contributed to pathological fracture of the mandible [17].

Mandibular body fractures may be treated in various ways. Favourable body fractures could heal using conservative methods to get stabilization of the tension surface [15,21]. Internal fixation with interfragmentary wires is indicated for simple mandibular body fractures, such as unfavourable fractures. The primary function of osseous wiring is to reduce fractures and prevent their displacement by the passive function of the muscles of mastication. The first rule of osseous wiring is to do no harm or to attempt to avoid injury to other structures within the jaws, such as the mandibular canal, roots of teeth and the periodontal ligament. The second rule is always to attempt to re-establish normal functional occlusion [15], as the mouth must be able to shut after surgery. To repair an unfavourable fracture, two intra-osseus wires are needed (one dorsally and the other ventrally) between the two fragments [15], or a triangular method could be used instead, consisting in one hole rostrally to fix two wires in a perpendicular arrangement coming from two holes placed caudally to the fracture. The dorsal wire should have a horizontal disposition and the second wire provides additional stability and prevents shear or rotation of the fragments around the primary wire [8]. Additionally, interdental wiring could also be used to increase the fracture stability [15].

Another option to treat unfavourable fractures is by using internal fixation with conventional bone miniplates and screws. The advantages of internal rigid fixation in the treatment of mandibular fractures are the accurate restoration of normal anatomy and occlusion and rapid return to normal function [16]. Nevertheless, every effort should be made to avoid injury to the roots and periodontal ligament of the teeth and the mandibular canal during placement of the anchorage holes through the bone [15].

The mandible is not an easy bone in which to use plating techniques for stabilisation of fractures. According to Higgins [8], at least two plates are required in the mandible. One is placed along the alveolar surface to resist bending and act as a tension band. A secondary plate is required on the ventral surface to resist shear and rotation. However, the anatomy of the mandible in the cat precludes placement in a biomechanically advantageous position of the minimum two pins in each fragment without the risk of compromising either the dental roots or the mandibular canal [4,26]. It should be noted that dental damage may result in pain, infection, tooth death, periapical lesions and ultimately failure of fixation because of persistent infection and inflammation at the fixation site [4]. The application of other metal implants should be undertaken noting the following: large teeth occupy 70% of the depth of the bone, damage to vessels and nerves within the mandibular canal should be avoided and feeding tubes should be considered [27]. Thus, bone plates can be placed in the caudal toothless part of the mandible at the junction of the body and the ramus [4].

The article by Greiner et al. [28] biomechanically evaluates two plating configurations (using one or two internal fixation plates) to fix a simple transverse caudal mandibular fracture, in particular, shown in Figure 1 (and onwards), in which the legend states that the ideal region for miniplate application (where the bone is able to support internal fixation) in the cat mandible is depicted in blue. However, the blue colour coincides all along with the path of the mandibular neurovascular bundle, the main blood and sensory supply. They considered the nonideal region for miniplate application to be the area occupied by the dental roots, even though other figures show superimposition of the screw holes with the dental roots. Biomechanically, it could be an acceptable model, but it must be extremely painful for the cat patient. This happens because, usually, the mandible is considered like any other long bone, without taking into account its particularities: (1) the jaw has specialized structures, the teeth, whose roots are inserted in the alveoli and kept in position thanks to the periodontal ligaments, and (2) it does not have a medullary canal, but an inner canal that runs along the jaw (from caudal to rostral) providing blood supply and sensitive innervation, not only to the bone, but also the teeth and other related structures, such as mucosa and gingiva. Thus, if any of these structures become damaged, it does not matter that the prosthesis would be acceptable from a biomechanical point of view, because it is functionally unsatisfactory. Hence, it would be advisable that researchers, clinicians

and surgeons be aware of the importance of maintaining the integrity of the neurovascular supply and dental roots in order to achieve a favourable outcome.

Plate placement on the ventral (compression) border of the mandibular body, which avoids the neurovascular structures of the mandibular canal, increases the load on the plate. Mono-cortical application of bone plates to the mid-buccal surface of the mandible has been recommended to reduce the risk of iatrogenic damage [4]. Consequently, standard bone plates in accordance with the feline mandibular size must be used. Nowadays, there are many companies producing osteosynthesis systems with titanium microplates (up to a minimum thickness of 0.6 mm to fix with 1 mm diameter screws). These sizes, and similar, are recommended for the repair of feline mandibular fractures. The plate should therefore be placed on the ventral surface to avoid these structures, although the tension side is the oral side. Plating should only be considered for simple fractures that can be very accurately reduced. Any malalignment may lead to dental malocclusion, which would be difficult to correct. Locking plates may be more useful than non-locking implants as, if plate contouring is not perfect, they should not distort the fragments. The most common complications of surgical repair are malocclusion and osteomyelitis [15,21].

In the case of comminuted or open fractures, in which soft tissue wounds prevent the use of internal fixation, an external fixation with pins and a mandibular bumper bar is appropriate for their reduction [26].

According to Spodnick and Boudrieau [17], the management of simple fractures (large fragments without comminution) could be done with routine induction and endotracheal intubation per os for anaesthetic maintenance and surgery. In these instances, anatomical re-alignment and reduction of the fragments, rather than dental occlusion, is used to determine the accuracy of surgical reduction. Alternatively, in cases of severely comminuted fractures or those with bone loss, dental occlusion must be used to access the accuracy of the surgical reduction. In these cases, the endotracheal tube impedes the assessment of occlusion by preventing full closure of the mouth and must be replaced to bypass the mouth (endotracheal intubation via pharyngotomy). Hence, occlusion is used to determine the accuracy of the reduction when comminution or gaps in the bone are present. Simpler fractures may be reconstructed anatomically. As usually one side of the head/face is more severely injured, a reasonable approach is to repair the side with the simpler fractures first. It is highly recommended to repair the mandible from caudal to rostral, with symphyseal separations secured as the final step [17]. Interestingly, the ventral approach to each mandible facilitates exposure and bone fragment manipulation, including the ability to perform an accurate reduction and stabilisation. Consequently, the patient should be placed in dorsal recumbency to get the head in the ventral approach, fixing it by taping the maxilla to the table [17].

#### *3.4. Fractures of the Ramus*

Fractures of the mandibular ramus are relatively stable because the surrounding muscles usually prevent gross displacement of the fracture segments [3]. Condylar process fractures often heal as pain-free and functional non-union without surgery, but comminuted fractures could result in a TMJ ankylosis in immature and young cats [3]. Ramus fractures in dogs and humans are frequently stabilised using internal rigid fixation with plates and screws, providing accurate reduction and good construct stability [16]. But in cats, this is complicated because of their small size, the need for greater contouring of implants, difficulty in positioning small fragile fragments of bone during the application of implants and the very small cross-sectional surface area of bone at the fracture sites, making accurate anatomical reduction challenging [16]. However, Southerden et al. [16] proposed the development of a small range of standard pre-contoured locking plates for the fixation of caudal mandibular fractures in cats due to the small variation in shape and size of mandibles between animals (among 38 specimens).

#### *3.5. Impairment of the Nervous Supply*

It should be taken into account that more than the upper two-thirds of the body of the mandible is occupied by dental roots. The ventral third includes the mandibular canal, containing the inferior alveolar nerve and associated blood vessels, and the inferior alveolar artery and vein. The inferior alveolar nerve provides sensory innervation for the teeth and leaves the bone through three mental foramina as the mental nerves. These nerves are sensory to the soft tissues of the rostral part of the mandible. The blood vessels in the mandibular canal supply all the teeth by the way of small dental branches (*rami dentales*) entering the apical foramina and the bone itself [17].

In humans, according to Misch and Resnik [29], and regarding nerve injuries after dental implant procedures, traumatic and iatrogenic nerve complications may involve total or partial nerve resection, crushing, stretching or entrapment injuries. As a consequence, the resulting sensory deficits may range from nonpainful minor loss of sensation to a permanent and severe debilitating pain dysfunction. Regarding oral and maxillofacial tumours in cats, Little [12] shows an example of mandibulectomy (removing left total and right partial mandibles), in which "the inferior alveolar artery and vein entering and exiting the mandibular canal through the mandibular foramen at the medial aspect of the mandible are ligated and transected". But nothing is said about the alveolar mandibular nerve, as it is supposed that its resection might become very painful (even post-resection) as it has a sensory component. The point is that humans can express and describe their loss of sensation or pain, but what about cat patients after mandibular surgery? It is reasonable to assume that they can also suffer from the same type of impairment, but they cannot describe their sensations and could be suffering from severe debilitating pain dysfunctions or neuropathic pain and not even be willing to eat normally by themselves again.

In mandibular fractures, besides mandibular reconstruction, the integrity of the mandibular nerve is a fundamental aspect to take into consideration, since it has a sensitive, in addition to the motor, component. Consequently, it is not recommended to proceed to cut the mandibular nerve just to remove it, given that it can induce trigeminal ganglion degeneration, as reported by Gobel and Binck [30] following pulp removal in cats, inducing degenerative changes in primary trigeminal axons and in neurons in the *nucleus caudalis*.

Mandibular fracture types, their incidence and treatment methods are summarized in Table 1.


**Table 1.** Mandibular fracture types. Their incidence, possible treatment methods and some recommendations are compiled.

#### **4. Prosthesis Proposal to Fix a Simple Fracture of the Mandibular Body**

Taking into account all the concerns previously revealed, we propose a fixation method to repair simple fractures of the mandibular body that will provide acceptable rigid biomechanical stabilisation yet avoiding dental root and neurovascular damage.

The current prosthesis design proposal was made by using the design program Solidworks®. This computer-aided design (CAD) model was built starting from the threedimensional (3D) model of a real jaw. A cloud of points, attained after laser scanning a real jaw, has been post-processed to obtain the 3D model of the cat mandible used. Meshlab and Cloud Compare software were used to post-process the data obtained by laser scanning.

The proposed prosthesis (Figures 2–5) is a conceptual design that, if referred to the technological maturity level scale known as Technology Readiness Levels (TRL), would be equivalent to a TRL2, which corresponds to the formulation of a conceptual solution without yet testing experimentally. Hence, seven other TRL levels would lay ahead, including research in the laboratory environment (up to TRL4), then, the simulation environment (TRL5–6), and finally, the real environment (TRL7–9).

**Figure 2.** Lateral view of the mandibular prosthesis conceptual model in place, as the upper screws should be fixed where there are no dental roots. This model would be useful to repair body fractures between the third premolar and the first molar.

We think it could achieve good results in further TRL levels as it has three fixation points with small screws and the fourth is like a folded tab for fastening the ventral edge, therefore avoiding drilling the mandibular canal. The design with a two-sided 'Y' was chosen in a way that the screw positions do not imply any risk of perforating dental roots or damaging the neurovascular support. This shape also allows keeping safe the mental nerve that goes through the main mental foramen and also avoiding the caudal mental foramen.

The suitable material to implement this prosthesis, given its high requirements of resistance and biocompatibility, is titanium. The optimal manufacturing method would be 'additive manufacturing' (a transformative approach to industrial production that enables the creation of lighter, stronger parts and systems), i.e., metal 3D printing, since the prosthesis is a unique piece of small size, and up to now, this is the only method that makes production economically viable. In addition, given that its geometric complexity does not pose problems for its manufacture, this method is considered the best option.

**Figure 3.** Magnification of Figure 2. The asymmetric shape of the anterior horizontal "Y" avoids damaging the nerves that come out through the main and mental foramina.

**Figure 4.** A ventrolateral view of the left jaw displaying the prosthesis conceptual model to show the fourth fixing point with no screw holes. This part consists of a flat hook-like device that surrounds and embraces the mandibular ventral margin to avoid damaging the neurovascular supply when drilling the mandibular canal. As this prosthesis is custom-designed, the flat hook size (thickness, width and length) will be variable, depending on the patient.

**Figure 5.** A frontolateral view of the left mandible with the proposed prosthesis. Note the prosthesis thickness is variable depending on the mandibular area in order to fully adapt to its contour. Prosthesis thickness does not exceed 1.2 mm at any point.

In order to promote good adhesion of the prosthesis to the bone, and trying to avoid damage to the bone, the surface of the prosthesis should have a high roughness (undulations, geometric patterns with protrusions or some geometry with an equivalent effect). Thus, the effective area of contact with the bone (i.e., the actual contact area) would be minimised while increasing the friction force of the prosthesis with respect to the bone, which greatly reduces the possible relative slipping that may exist between the elements.

#### *4.1. Calculations*

Considering that at the time of application of the prosthesis, the jaw is immobilised, and that once the bone has healed the stresses of the prosthesis are greatly reduced, the function of the prosthesis at the beginning consists fundamentally in ensuring that the bone fragments are correctly and firmly positioned. A calculation of resistance based on a bite force of 10 kg is considered conservative.

#### *4.2. Flexural Strength*

Considering this force and that the central part of the prosthesis is the weakest part (its most critical section), which is placed 2 cm away from the canines, it is calculated that it supports a moment (*M*) of:

$$M = F \times d = 50 \text{ N} \times 25 \text{ mm} = 1250 \text{ N} \cdot \text{mm}$$

where:


Note that the force considered is 5 kg, since it is assumed that half of the effort is carried by the other jaw.

Taking into account the moment calculated previously, a first approximation of flexure resistance could be obtained dividing the moment by the bending moment resistance (Wb), as is shown in the next equation [31], with a considered thickness of 1.5 mm and width of about 3 mm, there is a maximum bending stress of:

$$
\sigma^{\text{max}} = \frac{M}{\frac{b \cdot h^2}{6}} = \frac{1250 \text{ N} \cdot \text{mm}}{\frac{1.5 \text{ mm} \times (3 \text{ mm})^2}{6}} = 555.56 \text{ MPa} \tag{1}
$$

Equation (1): Maximum traction tension considering only flexion efforts. where:


The value of the bending moment resistance is obtained considering a squared section of flexure; if the design of the prosthesis changes, as a result of an irregular section chosen, the finite element method (FEM) would be needed to calculate the maximum bending tension.

As it can be seen, the stress obtained is considerably lower than the elastic limit of titanium. Thus, it could be considered a valid measurement. However, a balance should be struck between material resistance and physiology. Regarding the field of material resistance, it is advisable to make the prosthesis with a width as large as possible to achieve greater rigidity. However, this is inadvisable from the physiological point of view. Therefore, it is a question of achieving a balance between good rigidity and comfort.

#### *4.3. Shear Resistance of Screws*

Screws with a diameter of 1 mm have been considered, and taking into account the previous data and the conservative hypothesis that all the force is supported by one of them, the following shear stress is obtained [31]:

$$
\tau^{\text{max}} = \frac{4}{3} \times \frac{F}{\pi r^2} = 85 \text{ MPa} \tag{2}
$$

Equation (2): Maximum shearing tension considering bite force where:


The resulting stress is considerably lower than the shear strength of titanium, which means that the diameter of the screws can be even smaller.

Depending on the shape or position of the screws, a deeper calculation based in the FEM would be needed.

As previously indicated, the prosthesis shown is purely conceptual, and although an idea of the thickness and width that might be needed has been given, the calculations must be reviewed in practical clinical cases. Moreover, it would be convenient to undertake a more detailed study based on calculation software using the finite element method to obtain a more precise notion of the stresses to which the prosthesis is subjected. A topological optimisation of the geometry using suitable software could also be useful.

The current proposal would be equivalent to a TRL2 prototype according to the well-known TRL scale applied to technology, whose value (from 1–9) gives an indication of the level of maturity of the product, which implies that it is the formulation of a concept that has not yet been validated in the laboratory. Therefore, our proposal considers the mandibular prosthesis concept like an embryo, without practical validity, although its correct evolution and development could mean important advances in the field of intervention for mandibular fractures, not only in cats, but also in other species, by avoiding complications caused by damaging the integrity of the tooth roots and neurovascular supply when placing the screws to fix the jaw prosthesis.

This prosthesis model aims to visually reveal the proposals and ideas expressed throughout this review, without delving into the details about main dimensions, materials, manufacturing process and other aspects that may affect its implementation. Prostheses of these characteristics should be designed specifically for each specimen, adapted in a 3D model of the cat patient's jaw. In this way, not only will the size be adjusted, but the surfaces will also adapt much better, and thus a better grip will be ensured with all the advantages that this entails for fracture healing.

#### **5. Future Trends**

On one hand, the use of virtual surgical planning (VSP) and CAD/CAM (Computer-Aided Design/Computer-Aided Manufacturing) technology has added a new dimension to surgical planning, especially in the areas of craniomaxillofacial trauma, orthognatic surgery and reconstructive maxillofacial surgery [32]. This technology allows increased accuracy of reconstruction, decreased operative time, decreased flap ischaemic times, ease of use, improved predictability of outcomes, improved patient satisfaction and decreased complications [30]. In addition, with the aid of the surgical cutting guide, the positioning and fixation of the prosthesis would be accurate and almost flawless, thereby greatly reducing the operating time [32].

On the other hand, a comparison of three cone-beam computed tomography (CBCT) methods with dental radiography was analysed by Heney et al. [33]. They found that CBCT methods were better suited than dental radiography to the identification of anatomical structures in the full mouth of cats [15,21]. Cone-beam CT may prove to be the next major advancement in veterinary dentoalveolar and maxillofacial imaging because of its ability to provide 3D imaging at a lower cost than conventional CT, and with a lower radiation risk. The use of rapid scan technology, which allows faster image acquisition than conventional CT, and the ability to post-process the volumetric data into various two-dimensional (2D) and 3D reconstructions, makes CBCT an attractive imaging modality [33].

Liptak et al. [34] were one of the first to report a partial reconstruction of the mandibular body with a customised 3D-printed titanium prosthesis in a cat after removal of a mandibular osteosarcoma, as mandibulectomy in cats is associated with a high complication rate [35], including short-term (<4 weeks) and long-term (>4 weeks) adverse effects such as dysphagia or inappetence, mandibular drift or malocclusion. To avoid anorexia after surgery due to pain, supplemental tube feeding is recommended following mandibulectomy in cats. However, almost one in eight of the cats in the study never returned to voluntary eating after mandibulectomy [35]. Liptak et al. [34] pointed out that, in cats and dogs, dental roots and neurovascular structures comprise most of the bone volume in the mandibles and maxilla, and avoiding these structures is important during mandibular and maxillary repair. This is a further challenge in the rostral portion of the mandible because the canine dental root fills most of the mandible. The rationale for avoiding the dental roots is the high likelihood of tooth death and consequent periapical periodontitis, resulting in infection and potential implant failure. In general, this case is a good example of the future trends in cat mandibular osteosynthesis. However, we consider that surgeons would have been more concerned about the following issues: (1) neurovascular bundle and (2) the screw size used in this cat patient. Regarding the neurovascular supply, the authors did not report the fate of the neurovascular bundle in the mandibular canal nor what they did with these structures before/during/after cutting the mandibular body [34]. Also, in this case report, a bone fragment was kept in the rostral part of the mandible to fix the prosthesis (after a mandibulectomy of 40 mm in the mandibular body). It could be expected that lacking ipsilateral blood supply and innervation, this fragment would not survive. On the contrary, it showed no evidence of failure 14 months postoperatively. It is reasonable to think that this bone fragment might have achieved collateral

blood supply from the mandibular symphysis vessels due to tissue hypoxia that may have promoted angiogenesis. However, in our opinion, surgeons must be more attentive to the importance of avoiding damage to the neurovascular bundle because new blood vessels may grow from pre-existing ones (angiogenesis), but the sensitive (and motor) neural components do not recuperate and could cause additional neural damage. This is important to the well-being of the cat. Regarding point (2), the 2.0 mm-diameter screws used to fix the custom-made prosthesis seemed too large and long for the cat mandible, and after traversing both bone cortexes, their tips would become embedded into other structures, such as muscles or damage any intermandibular structure. Hence, using a screw size suitable to the cat patient is highly recommended.

On one hand, the mandibular prosthesis size should be adapted to the small size of feline patients. It seems that most pet prostheses are designed for bigger specimens (except those designed for a specific cat patient, printed in titanium), they must have no sharp edges, and finally, they should have plenty of holes to place the screws and for muscle or tissue attachment. However, the screws used are usually too long, far surpassing the contralateral cortical bone.

On the other hand, materials used in osteosynthesis must be more biologically acceptable because soft tissue is not going to attach to metal [36], so a deeper understanding is needed, and further research will be required to produce more biocompatible prostheses.

To close this section, we cannot agree more with Vaughan [36], who stated "customshaped implants will likely be part of the future, but the process needs to be refined from a biomechanical and biological perspective".

#### **6. Conclusions**

In the body of the cat mandible, dental roots and the mandibular canal (with the vascular supply and the inferior alveolar nerve) occupy most of the volume. Therefore, in mandibular fractures (due to a variety of causes, such as periodontitis, tooth resorption, trauma, or secondary to tooth extraction), it makes it challenging to apply a plate with fixed screw positions without invading dental roots or neurovascular structures. Otherwise, it would be a very painful process, including the failure of fixation due to chronic infection and inflammation at the fixation site. In the face of all these difficulties, we proposed a suitable prosthesis design, produced by additive manufacturing, that would provide acceptably rigid biomechanical stabilisation and avoid damage to any of those structures when fixing a mandibular body fracture. The future depends on the improvement of diagnostic images and Computer-Aided Design/Computer-Aided Manufacturing (CAD/CAM) technology to manufacture custom-designed prostheses made of highly biocompatible material.

**Author Contributions:** Conceptualisation, M.L. and M.d.M.Y.; Writing—Original Draft Preparation, M.L., D.A.-P., M.L.-L. and M.d.M.Y.; Prosthesis conceptualisation M.L.-L. and M.L.; Writing—Review and Editing, M.L., D.A.-P., M.L.-L. and M.d.M.Y. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The Institutional Review Board statement was not required as no animals were used to carry out this study.

**Data Availability Statement:** Data sharing is not applicable to this article as all data associated is available in the text.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


*Article*
