The value of a head turn in neurolocalization

1 INTRODUCTION

Head turn and pleurothotonus have been 2 interchangeable terms that are used frequently to describe postural abnormalities as part of a neurological examination of cats and dogs. Historically, a head turn is thought to be defined as when the head and neck only are deviated on the horizontal plane and pleurothotonus is defined as the head, neck, and trunk twisted to 1 side.¹ Other terms that have been used synonymously to describe such deviations in head and neck posture include torticollis, laterocollis, and scoliosis. Given the discrepancy between quadrupedal and bipedal head posture in relation to the body, there is no direct equivalent term used in the human medical literature for a head turn. However, the presentation of laterocollis, with the head angling toward the shoulder, visually would be the closest equivalent in humans.²

Presence of a head turn historically has been attributable to lesions affecting the forebrain, either the telencephalon or diencephalon, namely, the rostral thalamus.¹ The understanding as to why a head turn manifests itself in cats and dogs, remains elusive. Limited clinical data supports the localization of head turns, with the remainder of the evidence supported by experimental studies.^{3, 4} Our aims were to identify the regions of the central nervous system that could be associated with the presence of head turn with or without body turn in patients presenting with spontaneous neurological dysfunction. We also aimed to identify whether it was possible to differentiate between head turn and pleurothotonus based on lesion location, whether certain lesion locations were more associated with 1 over the other, and, if any other postural deviations of the head (such as head tilt) were identified in association with head turn.

2 MATERIALS AND METHODS

A multicenter prospective observational study was performed. Cases were collected from University of Liverpool, University of Glasgow, University of Edinburgh, Wear Referrals, North Carolina State University, and Hamilton Specialist Referrals. Dogs were prospectively included in the study if they exhibited a head turn with photographic or video evidence on initial presentation. Based on the photographs or videos, the side of the head turn, the degree of the head turn (<45°, 45°-90°, 90°, and 90°-180°), and the concurrence of body turn or head tilt were assessed. Each dog required a full neurological examination performed by a board-certified neurologist or a neurologist-in-training. Magnetic resonance imaging (MRI) or computed tomography (CT) were required for lesion localization and determining which anatomical structures were affected. Based on imaging, the cases were divided into broad categories of sub-regions, namely forebrain, brainstem, cerebellum or a particular spinal cord segment. Signalment and suspected or confirmed diagnoses also were recorded. Dogs were excluded if no structural changes were identified, when multifocal areas were seen on advanced imaging, or if 1 lesion was large enough to affect more than one sub-region, thus preventing determination of the likely cause for the head turn.

Statistical analysis was performed using the software SPSS 27.0 (SPSS Inc, Chicago, Illinois). Fisher's exact test was used to determine if the degree of head turn affected the presence of body turn. Statistical significance was set at P ≤ .05.

3 RESULTS

Forty-nine dogs met the inclusion criteria, with 24/49 having an imaging localization to the forebrain, 9/49 to the brainstem, 9/49 to the cerebellum, and 7/49 to the cervical spinal cord. The range of lesions included intra-axial solitary mass (n = 22; 12/22 suspected intra-axial neoplasia and 10/22 suspected meningoencephalitis of unknown origin), ischemic infarct (n = 10), extra-axial solitary mass (n = 7), nontraumatic hemorrhage (n = 2), atlantoaxial subluxation (n = 2), syringo(hydro)myelia, (n = 1), acute hydrated non-compressive nucleus pulposus extrusion (n = 1), suspected seizure-induced change (n = 1), intervertebral disc protrusion (n = 1), and traumatic spinal cord injury because of dog bite (n = 1). Figure 1 shows the distribution of disease processes in relation to imaging location. Neurological examination localization was suggestive of a neurolocalization compatible with the lesion found on advanced imaging in 38/49 cases. Magnetic resonance imaging was performed in 46/49 and CT performed in 3/49. Figure 2 shows the distribution of combinations of head and body postures based on imaging location. Of the dogs with head turn only (n = 15), 13/15 had forebrain disease. Of the head turn and head tilt only cases (n = 9), 5/9 were found to have brainstem disease. Of the head turn and body turn only cases (n = 12), 9/12 had forebrain disease. Lastly, of the dogs with a head turn, head tilt and body turn (n = 13), 6/13 had cranial cervical disease and 4/13 had cerebellar disease. The brainstem cases mainly presented with head tilts in concurrence with a head turn (7/9). The majority of the cervical cases had a head tilt and body turn in combination with head turn (6/7). The side of the turn or tilt relating to the imaging localization is presented in Table 1. Of the cerebellar cases, 7/9 had contralateral head tilts in coincidence with a head turn ipsilateral to the lesion. In all 7 cervical cases, a contralateral head tilt and head turn were identified, with 6/7 having a contralateral body turn. The degree of head turn did not appear to be associated with the presence of a body turn. A graph depicting the presence of body turn in relation to the degree of head turn is presented in Figure 3.

3.1 Dogs with forebrain disease (n = 24)

Twenty-four dogs had a noticeable lesion affecting the forebrain' with 11/24 located in the telencephalon only, 3/24 located in the diencephalon only and 10/24 affecting both telencephalon and diencephalon. The underlying causes were intra-axial in 21/24 and extra-axial in 3/24. Of the telencephalic only lesions, 3/11 affected the basal nuclei, 5/11 affected the frontal cortex, 4/11 affected the temporal cortex, 6/11 affected the parietal cortex, 2/11 affected the occipital cortex, and 2/11 affected the piriform lobe. In the diencephalic only cases, 2/3 affected the dorsal paramedian area and 1/3 was situated in the rostral ventromedial aspect of the thalamus.

An ipsilateral head turn was identified 23/24 cases. In the 1 dog that had a contralateral head turn, it was the only postural finding in the case a seizure-induced change of the piriform lobe was suspected. A concurrent body turn was identified in 10/24 cases, all being ipsilateral to the side of the imaging location. A concurrent head tilt was identified in 1 telencephalic-only case and 1 case with both telencephalic and diencephalic involvement. In the former telencephalic case, the location was specific to the parieto-temporal cortices, and the head tilt was ipsilateral to the imaging localization alongside an ipsilateral head turn. In the case with telencephalic and diencephalic involvement, the location was mostly in the dorsal paramedian area of the thalamus, and the head tilt was contralateral to the imaging localization alongside an ipsilateral head turn.

Neurolocalization based on examination differed from imaging location findings in 2/24 cases; 1 of these neurological deficits suggestive of multifocal localization on examination, and in the other case, an incorrect lateralization was suspected. Compulsive circling was identified in 20/24 cases. The 4 dogs that were not identified as circling were non-ambulatory. Other findings attributable to a forebrain localization included altered mentation (22/24), altered behavior (14/24), aversion turning (8/24), hemi-neglect (reduced awareness of stimuli on 1 side) eating (7/24), proprioceptive deficits (24/24), and menace response deficits (20/24). Images of each posture with forebrain pathology can be found in Figure 4. Postural findings of each dog with forebrain disease can be found in the Supporting Information.

3.2 Dogs with brainstem disease (n = 9)

Nine dogs had noticeable lesions affecting the brainstem, with 5/9 affecting the medulla oblongata (MO) only, 3/9 affecting the midbrain (MI) only, and 1/9 had extensive pathology affecting the midbrain, pons, and medulla oblongata (MPMO). Six of the dogs had intra-axial lesions and the remaining 3 had extra-axial masses. Computed tomography was performed in 1/9 dogs with the remaining dogs undergoing MRI. Of the MO cases, the MO had dorsal-only involvement in 2 cases, in the middle in 1 case and occupying both dorsal and ventral components in 2 cases. Of the MI cases, the midbrain was affected in the middle in 2 cases and dorsally in 1 case. The MPMO case had an extensive lesion causing dorsal and ventral compression of all 3 compartments of the brainstem.

An ipsilateral head turn, in relation to the imaging findings, was identified in all 9 cases. A head tilt was identified in 7/9 cases; 5 were ipsilateral and 2 contralateral to imaging findings. The 4 dogs with ipsilateral head tilt were identified in the MO group and the 2 contralateral head tilts were identified in the MI group.

A correct lesion localization based on the neurological examination was identified in 6/9 cases. Of the 3 cases mis-localized, all were thought to have a multifocal localization on neurological examination. Other signs of brainstem disease included proprioceptive deficits (9/9), change in mentation (2/9), positional nystagmus (2/9), and ipsilateral circling (7/9). An example of the most common presentation with brainstem pathology can be found in Figure 5. Postural findings of each patient with brainstem disease can be found in the Supporting Information.

3.3 Dogs with cerebellar disease (n = 9)

Of the 9 dogs identified with noticeable cerebellar lesions, 7/9 were isolated to the rostral portion of the cerebellar vermis; culmen (4/6), declive (2/6), lingula (2/6), and lobulus (1/6). The 2 remaining dogs had lesions that occupied a cerebellar hemisphere. Magnetic resonance imaging was performed in all dogs and suggested intra-axial lesions. All head turns were ipsilateral to the side of the lesion with 6/8 dogs having a head turn >90°. A head tilt was identified in 7/9 dogs with the tilt appearing contralateral to the side of the imaging findings. A body turn was identified in 4/9 dogs.

A correct localization on neurological examination was identified in all cases. Other aspects of the neurological examination included ataxia in all dogs with ipsilateral hypermetria (4/9), postural response deficits (5/9), positional ventrolateral strabismus (8/9), and vertical spontaneous nystagmus in the eye ipsilateral to the lesion (5/9). An example of the typical posture seen in cerebellar cases can be found in Figure 6. Postural findings of dogs with cerebellar disease can be found in the Supporting Information.

3.4 Dogs with cervical spinal cord disease (n = 7)

Seven dogs had noticeable lesions affecting the cervical spinal cord; with 5/7 situated cranially (at the level of C1-C3 vertebral bodies), 1/7 situated caudally (at the level of C4-C7 vertebral bodies) and 1/7 being an extensive process throughout the entire cervical spinal cord. Magnetic resonance imaging was performed in 6/7 dogs with CT being performed in the remaining dog. Of the cranial cervical cases, 1/5 had an intradural extramedullary dorsal lesion, 3/5 resulted in ventral extradural compression, and, 1/5 resulted in extradural dorsal and ventral compression. The caudal cervical and the extensive cervical cases were limited to an intramedullary dorsal component.

A contralateral head turn, when compared to imaging findings, was identified in all cervical cases. A body turn was identified in all cases and appeared contralateral to the imaging localization. A head tilt was identified in 6/7 dogs and was contralateral to the imaging localization. All of the dogs with lesions involving the cranial cervical area and the 1 dog with extensive cervical disease had head tilts present.

Localization based on neurological examination was correct in 5/7 cases. The 2 cases mis-localized had suspected neurolocalization to the brainstem and forebrain separately. Both dogs had dull mentation with proprioceptive deficits. The dog with suspected brainstem disease had positional ventrolateral strabismus and the dog with suspected forebrain disease had intermittent menace response deficits. Both dogs had lesions in the cranial cervical area. Other clinical signs identified in dogs with cervical lesions included proprioceptive deficits (7/7), tetraparesis (5/7) hemiparesis (3/7), and Horner syndrome (2/7). Circling, hemi-neglect (reduced awareness of stimuli on 1 side) behavior, and aversion turning were not identified in any dog. An example of the most common presentation seen with cranial cervical pathology can be found in Figure 7. Postural findings of dogs with cervical spinal cord disease can be found in Supporting Information.

4 DISCUSSION

We sought to determine whether the presence of a head turn in combination with other postural abnormalities could facilitate neurolocalization. This goal, however, was not achieved, because the concurrence of head tilts and body turns was identified in each broad location of the brain and cervical spinal cord. An isolated head turn in the absence of other neurological deficits was not identified in our study, although this has been reported previously.⁵ The presence of other abnormalities in the neurological examination assisted in neurolocalization in most cases although mis-localizations were still frequent.

When a head turn, body turn, and head tilt are seen in combination, it has commonly been called torticollis or scoliosis. However, similar terms have been implied in cases of head turn and head tilt. Also, as identified in our study, the side of the head tilt was not always the same side as the head turn and body turn (difference seen between cervical spine cases and cerebellar cases), making specific documentation of posture necessary.

The most common imaging-based location in our study was the forebrain with all but 1 dog, having an ipsilateral head turn to the site of the lesion. One possible explanation for this outlier is that the piriform lobe and amygdala, in this case, was stimulated rather than had a loss of function. The head turn abnormality was identified when it was suspected the patient was undergoing partial seizure activity. This resultant contralateral head turn has been documented previously in experimentally induced stimulation of the amygdaloid nuclear complex in cats.^{4, 6} Aside from this case, a correct neurolocalization based on examination was made. However, the presence of a head turn could have influenced neurolocalization as because it historically has been associated with lateralized forebrain disease. Of the telencephalic cases, the main areas affected were the frontal, parietal, and temporal cortices. These cortices are important in sensory processing, with the components of the frontal and parietal cortices giving rise to the sensorimotor cortex. It has been hypothesized in humans that a lesion affecting the somatosensory cortex could result in a loss of normal suppressive input leading to motor overflow resulting in co-contraction of muscles, although this finding has been thought to represent a theoretical cause for idiopathic cervical dystonia.^{7, 8} Some dogs had isolated lesions in the basal nuclei. The caudate nuclei, along with the striatum, may play a role in the manifestation of abnormal ipsilateral head turning in humans as a result of serotonergic depletion.^9-11 This finding has been identified in dogs with striate artery occlusion.¹² An uncommon accompaniment to the altered head posture was a head tilt, identified in 2 dogs. A head tilt in association with head turn has been reported previously in association with the paramedian aspect of the thalamus in dogs with caudal perforating artery occlusion.^{12, 13} Hypotheses behind this particular head deviation include dysfunction of the mediodorsal nucleus of the thalamus, which has been purported to have connections to the interstitial nucleus of Cajal (INC) and the cerebellum.^{14, 15} A head tilt and head turn in association with telencephalic lesions have not been reported in dogs. The parieto-insular cortex is an area, identified in humans and macaques, that provides sensory vestibular input, originating from receptors of the head and neck, ears, and vision. Dysfunction of this particular region theoretically could lead to alterations in head posture by the individual losing a sense of verticality on 1 side. Further investigations are required to determine whether this finding is similar in quadrupedal mammals. The presence of a body turn in concurrence with a head turn was an inconsistent finding in the forebrain cases, thus not making it a reliable indicator for forebrain disease.

In the dogs with brainstem disease, a head turn was never seen alone but always accompanied with either a head tilt only, body turn only, or both head tilt and body turn. Many components of the brainstem may contribute to posture of the head, neck, and body. The interstitial nucleus of Cajal (located in the midbrain), and the descending interstitiospinal fibers within the medial longitudinal fasciculus have direct excitatory synaptic connections to the ipsilateral cervical muscles, including the biventer, splenius, and sternocleidomastoid muscles.^16-18 Therefore, disruption to these areas of the brainstem theoretically could deviate the head in both the horizontal and vertical planes. A ventrolateral lesion of the midbrain has been associated with a contralateral head turn in cats that had electrolytic lesions affecting the substantia nigra reticulata.¹⁹ This finding however was not identified in any of our dogs. The dorsal raphe nucleus plays a role in serotonergic innervation of the substantia nigra and rostral cortex. Loss of this specific function in rats has resulted in an ipsilateral head turn.²⁰ A body turn identified in 2 of our dogs could have been a result of disruption to the reticular formation, which plays an essential role in facilitating muscle tone.

Of the dogs identified with cerebellar disease, a large proportion of these localized to the rostral vermis; including the culmen, lobulus, and declive. All of these dogs had an ipsilateral body turn with a contralateral head tilt. Clinical findings of rostral cerebellar infarctions have been described previously in dogs with 1 study identifying dogs with ipsilateral or contralateral head tilts toward the lesion^{12, 21} and a contralateral torticollis in 1 dog.²¹ However, this example is the first documented presentation of a head turn and tilt seen concurrently in dogs with cerebellar disease. The rostral vermis of the cerebellum is known to play a role in co-ordination of gait and postural adjustment of the head, trunk, and legs. Of the 2 remaining cases localizing to the hemisphere, an ipsilateral head turn was identified with a head tilt that was contralateral in 1 and ipsilateral in the other dog. A pleurothotonus associated with this area has been described clinically in a calf,²² and experimentally in rabbits²³ and primates²⁴ after ablation of the fastigial nucleus. The rostral aspect of the fastigial nucleus may play a role in integrating spatial information from the head and body to regulate and adapt an animal's posture.²¹ The head tilt identified possibly could relate to the fact that the fastigial nucleus plays an important role in the vestibular system by sending axons through the caudal cerebellar peduncle to the vestibular nuclei to maintain balance.^{1, 25}

Similar to the dogs with brainstem disease, no dogs with cervical spinal cord disease exhibited a head turn in isolation, with only 1 having only a concurrent body turn and the remainder having concurrent head tilts and body turns. A cervical localization of abnormal head posture has been identified previously in dogs with meningoencephalitis of unknown origin⁵ and syringo(hydro)myelia,²⁶ both identified as cervical scoliosis or torticollis. The current hypothesis on pathogenesis suggests this finding could be as a result of damage to either the dorsal gray matter or the ventral gray and white matter. Ventral horn damage and ventral white matter damage may result in de-innervation and atrophy of the ipsilateral paraspinal muscles, causing an imbalance of muscle tone.²⁷ This imbalance of tone could lead to deviation of the head and neck away from the lesion.²⁸ Dorsal horn damage results in interruption of the sensory proprioceptive fibers sending information from the receptors of the joints, muscles and ligaments. This loss results in unopposed contractions of the contralateral spinal muscles leading to deviation of the head and neck away from the lesion.^{28, 29} Another confounding factor that could have played a role in postural deviation is the presence of pain on 1 side preventing the patient from the turning to the side of the lesion. Both the dorsal- and ventral-located lesions identified in our study reflected this phenomenon with all cases having a contralateral body turn. It was interesting that a head tilt was only identified in dogs with cranial cervical lesions. The central cervical nucleus resides within spinal cord segments C1-C4; receiving input from neck muscle spindle afferents and projecting to the contralateral cerebellum as well as having collateral projections to the vestibular nuclei. Experimental destruction of this nucleus has given rise to changes in neck rotation.³⁰ Of these cases, 2 were atlantoaxial subluxations. In humans, the most common manifestation for atlantoaxial rotatory subluxation is torticollis, described as a “cock-robin” posture of the neck.³¹ The proposed cause is damage of the greater occipital nerve (branch of C2 nerve) or the C2 nerve root, both of which innervate the sternomastoideus and trapezius muscles. These muscles are integral to maintenance of rotatory control of the head and neck³² and thus ipsilateral loss can lead to both turn and tilt.

Our study had some limitations. Albeit solitary, the pathology identified could be more extensive or infiltrative than apparent on imaging, or could be exerting some mass effect on adjacent areas, making it difficult to ascertain which particular isolated cortices or areas could be responsible for a head turn. A larger sample size of lesions specific to 1 location would be needed to identify this possibility. In addition, the sample sizes for each region were disproportionate, preventing direct comparison of sites with each another. Another hindrance to determine site of dysfunction was whether direct compression or indirect compression on the opposite side of the brain or spinal cord was giving rise to the neurological deficits identified. Including only discrete intra-axial lesions with minimal mass effect may resolve this uncertainty. It also became apparent, whereas collecting cases, that head turns and head tilts identified could be a static finding only or change in severity when posture is altered. This positioned or positional head turn was not fully captured on photographs alone.

Lastly, we hoped to shed some light on differentiating between the terms used to describe variations and combinations of head and body posture. Such terms in the literature include pleurothotonus, scoliosis and torticollis and often are used indiscriminately. Because our study identified that certain combinations of posture can be identified, it may be valuable to review such terms. Pleurothotonus (otherwise known as Pisa syndrome) in the human medical literature, has been thought to be a consequence of long-term treatment with neuroleptic drugs. It has been defined as >10° tonic lateral flexion in upright spine without any clinically relevant associated vertebral rotation (resembling the leaning tower of Pisa) or involvement of the neck or head.^{33, 34} In the veterinary medical literature we have reclassified this term as a head, neck and body turn in the same direction, which in this study, was only identified in 33% of forebrain cases and in only a few brainstem cases. Such limited findings in each region could imply that if a true pleurothotonus was present, either site could be considered, and that the lack of this sign in the examination does not exclude neurolocalization to either these regions. Torticollis, or “wryneck” is described in human medical literature as the twist and a turn of the neck as a result of a shortened sternocleidomastoid muscle.³⁵ This shortening results in a tipping of the head toward the affected shortened side and rotation of the chin toward the unaffected side. A similar appearance was seen in 2/7 dogs with brainstem lesions with an ipsilateral head turn and a contralateral head tilt. Interestingly, 5/7 dogs with brainstem lesions presented with an ipsilateral head turn and head tilt. All of these cases had extension to the MO, which houses the vestibular nuclei, thus contributing to the loss of extensor tone of the neck muscles and head tilt to the same side.

Scoliosis in humans is thought to be a tri-dimensional deformity of the spine in the coronal, sagittal, and axial plane (mainly coronal) with a Cobb angle >10°.³⁶ It is unclear whether or not true scoliosis requires confirmation by an imaging modality. Scoliosis has been associated mostly with spinal cord, spinal nerve or vertebral column pathology. When the cervical spinal cord is affected, a change in neck (and subsequently) head posture would occur alongside and change body posture. The majority of our dogs with cervical spinal cord lesions presented with head turn, head tilt, and body turn in the same direction (6/7), thus implying that scoliosis may be an appropriate term for this presentation.

We have identified a heterogeneous distribution of anatomical locations that give rise to the presentation of a head turn. These findings suggest that the presence of a head turn upon neurological examination would not exclusively lead to neurolocalization in the forebrain, as historically thought. We do recognize that postural control of the head and neck is a highly complex system, consisting of afferent signals arising from visual, vestibular, and proprioceptive sources and giving rise to conscious and compensatory efferent pathways to associated nerves and muscles of the neck. Therefore, theoretically, any disruption of such pathways could lead to a postural deviation.

CONFLICT OF INTEREST DECLARATION

Authors declare no conflict of interest.

OFF-LABEL ANTIMICROBIAL DECLARATION

Authors declare no off-label use of antimicrobials.

INSTITUTIONAL ANIMAL CARE AND USE COMMITTEE (IACUC) OR OTHER APPROVAL DECLARATION

Approval was obtained from the affiliated institutions. This either was covered by an approval by the veterinary ethics research committee as well as owners signing a consent form to use advanced imaging and pictures of the animals enrolled in the study.

HUMAN ETHICS APPROVAL DECLARATION

Authors declare human ethics approval was not needed for this study.