Trigeminal Neuropathy in Dogs: A Retrospective Study of 29 Cases (1991–2000)
The medical records of 29 dogs unable to close their mouths due to flaccid paralysis or paresis of the muscles innervated by the mandibular branch of the trigeminal nerve, were reviewed. Idiopathic trigeminal neuropathy was diagnosed in 26 dogs based on complete resolution of clinical signs and lack of any long-term neurological disease. Of these dogs, golden retrievers were over-represented. No age, sex, or seasonal predispositions were identified. Trigeminal sensory innervation deficits were observed in 35% (9/26), facial nerve deficits were observed in 8% (2/26), and Horner’s syndrome was observed in 8% (2/26) of dogs. Electromyographic examination of the muscles of mastication revealed abnormalities in seven of nine dogs. Results of cerebrospinal fluid analysis were abnormal in seven of eight dogs. Corticosteroid therapy did not affect the clinical course of the disease. Mean time to recovery was 22 days. Lymphosarcoma, Neospora caninum infection, and severe polyneuritis of unknown origin were diagnosed in three of 29 dogs at necropsy.
Introduction
The trigeminal nerve is the largest cranial nerve in dogs and humans and contains both motor and sensory fibers. The cell bodies of the motor neurons are located in the trigeminal motor nucleus in the pons. The sensory cell bodies of the general somatic afferent fibers are located in the trigeminal ganglion within the trigeminal canal. As the nerve emerges from the trigeminal canal, it divides into three branches. The mandibular branch contains both motor and sensory fibers. The motor fibers innervate the masseter and temporalis muscles that close the mouth. The sensory fibers of the mandibular branch innervate the buccal cavity, tongue, teeth of the lower jaw and skin of the mandible, caudal buccal region, and craniolateral pinna. The maxillary and ophthalmic branches contain only sensory fibers. The maxillary branch innervates the lower eyelid, nasal mucosa, upper teeth, upper lip, and nose. The ophthalmic branch is sensory to the eyelids, eyeball, nasal mucosa, and skin of the nose. Ophthalmic (i.e., the lacrimal nerve) and maxillary (i.e., the zygomaticotemporal nerve) branches contain postganglionic parasympathetic fibers that innervate the lacrimal gland. Sympathetic fibers run along the ventral surface of the trigeminal ganglion and then join the ophthalmic branch to provide control of globe position and pupillary dilatation.1
An inability to close the mouth is a relatively common presenting complaint in dogs. Mechanical obstruction due to bilateral luxation of the temporomandibular joints,2 oral foreign bodies or fractures of the mandible,3 or flaccid paralysis of the muscles innervated by the mandibular branch of the trigeminal nerve (i.e., mandibular paralysis) can lead to an inability to close the mouth. Careful examination of the oral cavity will usually distinguish flaccid mandibular paralysis from other causes.
Both benign and malignant disease processes can lead to flaccid paralysis of the muscles of mastication with resultant inability to close the mouth. The most common benign cause is idiopathic trigeminal neuropathy (ITN). Information on only eight cases of ITN in dogs has been published, and in all cases, complete resolution occurred in a mean of 14 days.4–7 Idiopathic trigeminal neuropathy has also been referred to as trigeminal neuropathy,4 trigeminal neuritis,8–11 and trigeminal neurapraxia.6 Flaccid mandibular paralysis in dogs may also be due to neoplastic or infectious disease. Lymphosarcoma,1213 myelomonocytic leukemia,1415 and rabies1617 have all been reported.
The specific goals of this study were to examine clinical findings and clinicopathological data in dogs with ITN as well as with causes of mandibular paralysis other than ITN, and to clinically differentiate the causes of mandibular paralysis in dogs.
Materials and Methods
Medical records of dogs presenting to the Veterinary Hospital of the University of Pennsylvania (VHUP) between 1991 and 2000 with an inability to close the jaw were reviewed. The criteria for inclusion in the study were the presence of flaccid mandibular paralysis resulting in an inability to close the mouth and case follow-up data of time to resolution of clinical signs or necropsy. Idiopathic trigeminal neuropathy was diagnosed when no structural cause for the paralysis was found, and if complete resolution occurred with no long-term neurological disease. Diagnoses other than ITN were made based on the histopathological examination of samples gathered at necropsy.
Information obtained from the medical record included signalment, vaccination history, presenting complaint, time of onset, progression, results of diagnostic tests, therapeutic intervention, time to recovery, and evidence of intercurrent disease. All dogs received a general physical examination as well as a neurological examination on presentation. Results of hematology; serum biochemical analysis; antinuclear antibody, thyroid hormone level, or function testing; titers for Borrelia burgdorferi (B. burgdorferi), Rickettsia rickettsii (R. rickettsii), Ehrlichia canis (E. canis), Neospora caninum (N. caninum), and Toxoplasma gondii (T. gondii); Cryptococcus neoformans (C. neoformans) agglutination tests; cerebrospinal fluid (CSF) analysis; bone-marrow aspirate; electromyography (EMG); and radiographic or ultrasound examinations were evaluated. Long-term follow-up information was obtained from the record, by telephone interview, or by questionnaire.
Statistical Methods
Only dogs for which a diagnosis of ITN was made (n=26) were included in the statistical analyses. To determine if there was a breed predilection, Fisher’s exact test was used. Data was presented as odds ratios with 95% confidence intervals. To evaluate if ITN was seasonal, a Kolmogorov-Smirnov test was used. Sex predilection was evaluated using the chi-square test. Student’s t-tests were used to determine if dogs with ITN differ in age from the hospital population and to evaluate whether time to resolution of mandibular paralysis is influenced by the use of corticosteroids or the presence of trigeminal sensory deficit.
Results
Twenty-nine dogs met the inclusion criteria. Twenty-six dogs were diagnosed with ITN [Table 1]. Three dogs had a diagnosis other than ITN [Table 2]. Eight more dogs were excluded due to lack of follow-up information.
Idiopathic Trigeminal Neuropathy (n=26)
The mean age at presentation was 63 months, with a range of 13 to 156 months. Twelve dogs were male and 14 were female. Statistical analysis revealed no sex (P=0.6) or age (P=0.24) predilections. Golden retrievers were over-represented relative to the hospital population during the study period (P=0.01). Four of five golden retrievers were spayed females; one was an intact male. The mean weight of all dogs with ITN at presentation was 26.6 kg, with a range of 8.7 to 73 kg. Mean time to last vaccination was 8.7 months, with only three dogs receiving a vaccination within the previous month. Dogs were admitted to the hospital with approximately equal frequency during all seasons of the year, and no statistically significant seasonal predilection was identified relative to overall hospital admissions during the study period (P=0.74).
The most common presenting complaints and physical examination findings were hypersalivation (14/26, 54%), difficulty eating/drinking (12/26, 46%), anorexia (6/26, 23%), scleral injection (5/26, 19%), weight loss (3/26, 12%), lethargy (3/26, 12%), oral ulceration (3/26, 12%), dyspnea (2/26, 8%), vomiting (2/26, 8%), diarrhea (2/26, 8%), and dehydration (2/26, 8%). The results of neurological examination are summarized in Table 1.
Twenty of 26 animals had a complete blood count and serum biochemical analysis performed. Only one dog had a mature neutrophilia (leukocytes, 24.5 × 103/μL; reference range, 5.3 to 19.8 × 103/μL; neutrophils, 23.6 × 103/μL; reference range, 3.1 to 14.4 × 103/μL). Two dogs had increased serum alkaline phosphatase (SAP; 361 U/L and 861 U/L; reference range, 24 to 174 U/L) levels measured, of which one was receiving phenobarbital for epilepsy. Two of 26 animals were severely thrombocytopenic; one case was diagnosed with immune-mediated thrombocytopenia and hemolytic anemia (packed cell volume [PCV], 22%; platelets, 11 × 103/μL; reference range, 177 to 398 × 103/μL; total bilirubin, 1.3 mg/dL; reference range, 0.3 to 0.9 mg/dL), and in one dog the etiology was not discovered (platelets, 20 × 103/μL). Otherwise, no significant findings were noted. Serum creatine kinase level was measured in five of 26 dogs and was found to be normal in all cases. Eight of 26 dogs had thyroid hormone levels or thyroid function assessed, and all were normal. Eight of 26 dogs had CSF collected at the cerebellomedullary cistern and had their CSF analyzed. In the authors’ reference laboratory, a CSF total protein content of <25 mg/dL is considered normal with a nucleated cell count of <5/μL. One of eight dogs with ITN showed no abnormalities of either protein content or nucleated cell count. In six of eight dogs, there was lymphocytic pleocytosis; in one of eight dogs, there was a mild neutrophilic pleocytosis. Mean total protein content was 29 mg/dL, with a range of 9 to 51 mg/dL. The nucleated cell count ranged from 0 to 10/μL, with a mean of 4/μL [Table 1]. Electromyography of the masseter and temporalis muscles was carried out in nine of 26 dogs. Positive sharp waves, fibrillation potentials, or both were recorded in the masseter or temporalis muscles in seven of nine dogs and were absent in two cases. In these two latter cases, time of onset to EMG examination was 3 and 7 days respectively. Titers for antibodies against B. burgdorferi (i.e., Lyme disease) were measured in five of 26 dogs; three were negative and two were positive. Four of 26 dogs had T. gondii titers measured; three of 26 dogs had N. caninum titers measured, one of 26 dogs had E. canis and R. rickettsii titers measured; and three of 26 dogs had C. neoformans agglutination tests. All were within the reference range. One dog had temporalis muscle biopsies, which were considered normal on histopathological examination, and one dog had an antitype 2M myofiber antibody test that was negative.
Treatment given prior to referral to VHUP included use of loose tape muzzles to hold the mouth closed, antibiotics, corticosteroids, and nonsteroidal anti-inflammatory drugs in various combinations. Mean follow-up time for dogs with ITN was 35.8 months (range, 1 to 106 months). Mean time to resolution of mandibular paralysis for all 26 dogs was 22 days (range, 4 to 63 days). Eleven of 26 dogs received corticosteroid therapy either orally or by injection. Fifteen of 26 dogs received no corticosteroids. Mean time to resolution for the group that received corticosteroids was 25 days. The mean time to complete resolution for the untreated group was 19.6 days. No statistically significant difference (P=0.57) was found in time to resolution of mandibular paralysis between dogs treated with corticosteroids and those not treated. No statistically significant difference was found in time to resolution of mandibular paralysis between dogs with sensory deficits and those without (P=0.51). Time to resolution of trigeminal sensory deficits, facial nerve deficits, and Horner’s syndrome was not available. No dogs in this case series of ITN had another episode of inability to close the jaw during the follow-up period.
Diagnoses Other Than ITN (n=3)
Signalment and the results of neurological examination are summarized in Table 2. Case no. 1 presented for mandibular paralysis, vomiting, and lethargy. Thoracic radiographs revealed megaesophagus. Results of serum biochemical analysis and hematology were unremarkable. Schirmer tear test revealed significantly decreased tear production in both eyes (right eye, 1 mm per minute; left eye, 2 mm per minute; reference range, 14.5 to 25.1 mm per minute). This dog died after anesthesia for the placement of a gastrostomy tube. Diffuse, subacute, suppurative neuritis and ganglioneuritis with axonal necrosis and Schwann cell hyperplasia were noted on necropsy in the trigeminal ganglion as well as in the hypoglossal, facial, and oculomotor nerves and vagosympathetic trunk. A nonsuppurative encephalitis was also present. Numerous protozoal tachyzoites and cysts consistent with N. caninum infection were found. This dog had a negative serum T. gondii antibody titer. Cryptococcus neoformans agglutination test was also negative. Titers for N. caninum were not performed. Immunohistochemistry confirmed the tachyzoites to be those of N. caninum.
Case no. 2 presented for weight loss, difficulty eating and drinking, and mandibular paralysis. Hematology revealed a mild anemia, thrombocytopenia, and leukopenia. Serum biochemical analysis was unremarkable. Antibody titers for Lyme disease, Rocky Mountain spotted fever, and E. canis were all normal. Hematuria was noted on urinalysis. Thoracic radiography was unremarkable. Bone-marrow aspiration and cytopathological review confirmed the presence of malignant neoplasia. The dog was euthanized 6 weeks after the onset of clinical signs. Necropsy revealed lymphosarcomatous infiltration of both trigeminal ganglia as well as vagal ganglia, multiple lumbar spinal segments, and abdominal organs.
Case no. 3 presented for mandibular paralysis, weakness, anorexia, and lethargy and was euthanized because of clinical deterioration 1 week after onset of signs. Hematology, serum biochemical analysis, and urinalysis showed no significant abnormalities. Titers for Rocky Mountain spotted fever and E. canis were normal. A positive Lyme disease titer at dilutions between 1:64 and 1:128 was present; however, this could not be followed up with a convalescent sample. Thoracic radiographs had revealed megaesophagus, and analysis of cytopathology from a bone-marrow aspirate was unremarkable. Necropsy revealed multifocal, subacute, lymphoplasmacytic-neutrophilic neuritis affecting cranial nerves III, V, IX, X, and XI as well as epicardial nerves and multiple spinal cord segments and meninges. The final diagnosis was severe polyneuritis of unknown origin.
The mean CSF total protein content for these three dogs was 192 mg/dL (range, 54 to 412 mg/dL), and the mean nucleated cell count was 27/μL (range, 5 to 62/μL) [Table 2].
Discussion
Idiopathic trigeminal neuropathy appears to be the most common neurological cause of inability to close the mouth in dogs. It is a diagnosis of exclusion and cannot be confirmed by any antemortem test. The etiology remains unknown, and because of complete recovery in all cases, necropsy or biopsy material was unavailable from dogs presented in this report. The terms neuritis and neurapraxia have been used to describe this condition.68–11 Neuritis implies an inflammatory cause, whereas neurapraxia indicates a lesion causing a failure of nerve conduction without structural change. Although an inflammatory neuritis has been confirmed in some cases,18 it is unknown whether either of these processes occurs in all cases, and for this reason the term ITN was adopted.
The most common presenting signs in dogs with ITN were mandibular paralysis, difficulty eating and drinking, and hypersalivation. Sensory deficits in the trigeminal distribution were present in 35% of the authors’ patient population. Other reports8101118 state that sensory deficits are not a feature of ITN in dogs. The detection of sensory deficits during neurological examination may vary based on an individual examiner’s interpretation. The cases in this study were not evaluated by the same clinician in each case. Muscle atrophy, Horner’s syndrome, and facial nerve deficits occur less commonly but do not preclude a diagnosis of ITN. A significant proportion of dogs presented with other signs such as depression, vomiting, diarrhea, dehydration, dyspnea, and scleral injection. The only diagnostic tests that showed abnormal results in most cases were CSF analysis and EMG. Mild to moderate elevations in total protein and mild elevations in nucleated cell counts were the most common findings in the CSF. Cell populations were predominantly lymphocytic with smaller numbers of monocytoid cells. Neutrophils and eosinophils were rare findings [Table 1]. In only two of nine dogs, EMGs were normal. These dogs were examined at 3 and 7 days after onset of mandibular paralysis. This may have been too early for positive sharp waves, fibrillation potentials, or both to become apparent.
Mean time to recovery of motor function to the mandible in dogs with ITN has been reported as being in the range of 2 to 3 weeks.8101819 This was the case in many dogs presented in this study, but the clinical course varied from 4 days to 9 weeks, and the mean time to complete recovery was 22 days. It has been suggested that corticosteroids may hasten recovery.10 Corticosteroid therapy did not appear to alter the course of the disease in the dogs of this study even though the preparations used, dose, and length of course varied. Mean time to recovery was shorter for the untreated group, although this was not a statistically significant difference. Resolution of mandibular paralysis took longer in dogs with sensory deficits (mean, 27 days) compared to those without (mean, 21 days), although this was not a statistically significant difference. Time to resolution of the sensory deficits was not available.
It has been stated that ITN can be an episodic condition.5 Of the dogs in this series, only one had a previous tentative diagnosis of an episode of dropped jaw. No dogs had further episodes after discharge from the hospital in a mean follow-up time of 35.8 months. Repeat episodes are probably a rare occurrence. Seasonality of ITN was not detected as suggested elsewhere.20
Although no specific medications appear to be indicated for ITN, the dogs in this report required significant nursing care. Tape muzzles were recommended to improve ingestion of food. As these dogs are unable to grab food but swallow normally, it was recommended to place firm balls of moist food at the back of affected dogs’ mouths. Some dogs required water to be given by syringe, as they were unable to lap significant quantities of water and became clinically dehydrated.
The etiology of ITN remains unknown. In one published report, histopathology revealed an extensive, bilateral, non-suppurative neuritis of all portions of the trigeminal nerve and ganglion without involvement of the brain stem.18 In this case, demyelination was more prominent than axonal damage. The author suggested that in clinical cases that survive, recovery follows remyelination. Allergic19 or immune-mediated10 causes for the condition have been proposed, although no evidence to support or refute these hypotheses is presented here or elsewhere.
None of the dogs tested for T. gondii, C. neoformans, N. caninum, E. canis, or R. ricketsii in this report were positive. Similarly, no animals that were tested for antinuclear antigen or hypothyroidism were positive. It is unlikely that these disease processes are commonly associated with ITN. Two dogs had positive Lyme disease titers. Neither dog had received Lyme disease vaccine, and neither had a convalescent sample taken. Lyme disease can cause neurological disease, but the significance of these two positive antibody titers is unknown.21
It has been postulated that ITN can be caused by stretching of the mandibular nerve from hyperextension of the jaw when prehending large, heavy objects.6 It was shown that after dissection of the mandibular nerve, the nerve was stretched when the mouth was maximally opened. However, it was noted that the less commonly affected sensory portion of the nerve was the most stretched. No owners in this case series reported their dogs commonly carrying large or heavy objects.
In this report as well as in others2223 it was noted that Horner’s syndrome and facial nerve involvement can occur with concurrent mandibular paralysis. The sympathetic tract and trigeminal nerve are closely associated in the trigeminal canal, and concomitant deficits may be an anatomic consideration. Postganglionic sympathetic axons course along the ventral surface of the trigeminal ganglion before traveling adjacent to the ophthalmic nerve. Multiple cranial nerve deficits may be due to one multifocal process or from the extension of one anatomical lesion. It is interesting to note that golden retrievers are also predisposed to facial neuropathy22 and commonly suffer from idiopathic Horner’s syndrome.24
In humans, unlike dogs, sensory neuropathy is much more common than motor neuropathy.25 There is only one report of isolated trigeminal sensory deficit in dogs.26 Sensory neuropathy in humans is associated with multiple sclerosis, systemic lupus erythematosus, dermatomyositis, Sjogren’s syndrome, toxin ingestion (such as Trichlorethylene and Stilbamidine25), and various malignancies.27 It appears, however, that a transient, self-resolving sensory neuropathy occurs rarely in humans.28 The most common trigeminal nerve disorder in humans is trigeminal neuralgia, a condition characterized by an acute onset of exclusively unilateral severe facial pain.29 This has never been described in dogs.
Three dogs in this study were confirmed either by biopsy or necropsy to be suffering from conditions other than ITN. Neoplastic infiltration of the trigeminal motor tracts is an important differential diagnosis for mandibular paralysis in dogs. Two case reports of mandibular paralysis caused by lymphosarcoma are published.1213 In one case there was infiltration of epineurium, perineurium, and endoneurium bilaterally in the trigeminal nerve as well as in the vagus, recurrent laryngeal, mesenteric, and spinal nerves.12 There are two reports of myelomonocytic leukemia causing mandibular paralysis.1415 Clinical neosporosis is rare in adult dogs, and to the author’s knowledge, mandibular paralysis of the muscles of mastication has never been recorded as a presenting sign of this disease.30 It must be remembered that rabies is an important differential diagnosis for mandibular paralysis.1617 Especially after intranasal inoculation, the virus can enter the trigeminal nerves and ganglia prior to invasion of the central nervous system. It is equally possible for the virus to spread to the trigeminal nerve from the brain.3132 Rabies, although rare in the United States, must remain the first consideration of the clinician because of the serious zoonotic potential of the condition.
Because of the multiple potential etiologies of mandibular paralysis in dogs, diagnosis of ITN can only be made with caution. Differentiation of those animals with ITN from those with more malignant or infectious disease may be very difficult. Signalment is of limited use. Golden retrievers were over-represented in the authors’ patient population. No other breed, age, or sex predispositions have been identified. Time to complete recovery can be as long as 9 weeks and so is initially unhelpful, as dogs that do not recover within a short period of time cannot be assumed to have malignant disease. Muscle atrophy and sensory deficits were seen in dogs with ITN as well as dogs with malignant disease. Horner’s syndrome was seen in dogs with ITN and has been reported in a dog with myelomonocytic neoplasia.14 However, animals demonstrating multiple cranial nerve deficits, having a long clinical course (>9 weeks) with no improvement, and significant CSF abnormalities may be considered at higher risk for more malignant disease. Dogs presenting with a sudden onset of mandibular paralysis, no other signs of systemic disease, and no other cranial nerve deficits are more likely to be suffering from ITN. Magnetic resonance images were not obtained on any of the dogs in this report but may have helped in distinguishing benign from malignant disease.
Conclusion
The most common neurological cause of mandibular paralysis in dogs is ITN. Dogs presenting with an acute onset of flaccid paralysis of the muscles of mastication with no other signs of systemic disease and no other cranial nerve deficits are most likely to have ITN. The clinical course does not appear to be influenced by corticosteroid therapy. Dogs that present with multiple cranial nerve or other neurological deficits may be at higher risk of having more malignant disease. This latter group of dogs should undergo further diagnostic assessment to rule out neoplastic or infectious disease. It is suggested that this should include hematology, serum biochemical and CSF analysis, bone-marrow aspiration, and infectious disease titers for neosporosis, toxoplasmosis, and cryptococcosis. Magnetic resonance imaging may be helpful. The possibility of rabies should always be considered, and any dogs exhibiting mandibular paralysis must be placed in isolation and handled with caution.
Acknowledgments
The authors thank Drs. S.A. Steinberg, C.H. Vite, and A. deLahunta for their help in the preparation of this manuscript.
