Resolution of Polyneuropathy in a Hypothyroid Dog Following Thyroid Supplementation
An 8 yr old male golden retriever was evaluated because of chronic, progressive, multiple neurologic signs. Physical examination showed marked obesity and facial swelling with a “tragic facial expression.” Neurologic evaluation revealed the dog had multiple cranial nerve deficits and lower motor neuron signs in the pelvic limbs. Serum biochemical analysis and thyroid function tests were consistent with hypothyroidism. A biopsy from the common peroneal nerve revealed a loss of myelinated fibers, inappropriately thin myelinated fibers, and resolving subperineurial edema. The diagnosis of polyneuropathy associated with hypothyroidism was made. Levothyroxine therapy was initiated. Response to levothyroxine treatment was slow, with most neurologic abnormalities persisting for >6 wk. However, the dog made a full neurologic recovery within 6 mo. Although the occurrence of polyneuropathy in dogs resulting from hypothyroidism has been controversial, the study authors demonstrated that hypothyroid polyneuropathy can occur in dogs as documented in humans. This is the first report describing long-term follow-up information together with detailed pathological features of hypothyroid polyneuropathy in a dog. In hypothyroid polyneuropathy, the response to thyroid replacement may be slow, but a recovery can be expected if treatment is initiated before peripheral nerve fiber loss becomes severe.
Introduction
Hypothyroidism is the most common hormonal disease in dogs.1 Because thyroid hormone has a wide variety of physiologic effects, this disease can cause a broad range of problems in many organs. Reportedly, various neurologic signs resulting from either central or peripheral nervous system involvement, such as vestibular dysfunction; facial paralysis; and reduced spinal reflexes, ataxia, and paresis, are associated with hypothyroidism. However, a causal relationship with hypothyroidism has been proven in only a limited number of cases.2–7
The occurrence of hypothyroid polyneuropathy in dogs has been controversial, and pathological findings have been mentioned briefly in only rare cases.2,3,5 In a recent study in which hypothyroidism was experimentally induced in dogs, peripheral neuropathy was not evident clinically, electrophysiologically, or histopathologically.8 Based on that study, it was suggested that canine peripheral nerves may be relatively resistant to chronic thyroid hormone deficiency.8
In humans, it is known that hypothyroidism is an uncommon cause of peripheral neuropathy, such as carpal tunnel syndrome and some types of polyneuropathies.9 Although the pathogenesis of peripheral neuropathy due to hypothyroidism remains to be fully elucidated, it is generally understood that hypothyroidism can cause mucous deposits and fluid retention resulting in nerve damage through a compression mechanism.9
In this report, an older golden retriever with chronic, progressive, multiple neurologic signs was diagnosed with hypothyroid polyneuropathy after complete diagnostic investigations. To the authors’ knowledge, this is the first report describing long-term follow-up information together with detailed pathological features of hypothyroid polyneuropathy of a dog.
Case Report
An 8 yr old male golden retriever weighing 47.5 kg was referred to the Azabu University Veterinary Teaching Hospital for evaluation of head tilt, ptosis, pelvic limb ataxia, and exercise intolerance. Clinical signs began 1 mo prior to evaluation and had progressively worsened. The owner also noticed that the dog was less active and appeared somewhat depressed, although appetite and water consumption was normal. The dog’s vaccination status was current. On physical examination, the dog had a rectal temperature of 38.2°C, pulse rate of 96 beats/min, and respiratory rate of 24 breaths/min. The dog was obese (body condition score 5 out of 5) and appeared to be moderately swollen in the face with a “tragic facial expression.” Mild otitis externa of both ears was noted. Neurologic examination revealed ataxia of both pelvic limbs, but conscious proprioception and hopping reactions were normal in all limbs. Patellar and withdrawal reflexes were diminished in the right pelvic limb. There was a right-sided head tilt together with a right-sided ptosis. Pain sensation was considered normal in all limbs. The remaining neurologic examinations did not show any abnormalities. The complete blood cell count and serum biochemical analysis (after fasting for 12 hr) were performed. Results of the complete blood cell count were unremarkable. Serum biochemical analysis revealed hypercholesterolemia (46.1 mmol/L; reference range, 2.77–8.13 mmol/L) and hypertriglyceridemia (13.97 mmol/L; reference range, 0.19–1.15 mmol/L), with no other abnormalities. Hypothyroidism was suspected, and thyroid function tests were performed. They revealed low serum thyroxine (T4) <5.13 nmol/L; (0.3 μg/dL; reference range, 18.8–61.54 nmol/L; 1.1–3.6 μg/dL), low serum free T4 [(3.86 pmol/L; 0.3 ng/dL; reference range, 11.58–33.46 pmol/L; 0.9–2.6 ng/dL)], and an increased serum level of thyroid-stimulating hormone (1.6 ng/mL; reference range, 0.08–0.32 ng/mL). Thoracic radiographs were interpreted as normal.
On the basis of thyroid function test results, a diagnosis of hypothyroidism was made. Differential diagnoses for neurologic signs included a brain stem tumor, encephalitis, and polyneuropathy due to variable causes, including hypothyroidism and possibly otitis media and/or interna. Levothyroxine Naa (0.02 mg/kg per os [PO] q 12 hr) was initiated.
Seven days after the first presentation, the dog was reevaluated at the hospital. The dog was more active, but did not show any changes in other clinical signs. Serum T4 levels were measured and were high (130.77 nmol/L). Serum biochemical analysis revealed some improvement of hypercholesterolemia (33.39 mmol/L) and hypertriglyceridemia (2.05 mmol/L). The dose of levothyroxine Na was decreased to 0.01 mg/kg PO q 12 hr.
At that time, the owner decided to have the dog undergo further diagnostic testing rather than adopting a “wait and see” approach for whether levothyroxine treatment would have a positive effect on neurologic signs. The dog was anesthetized and MRI of the head and spine was performed 9 days after initial presentation. T1-weighted, T2-weighted, and fluid-attenuated inversion recovery scans were performed in the transverse and sagittal planes. Additionally, gadoteridolb at 0.2 mL/kg was administered IV, and a postcontrast transverse T1-weighted sequence was obtained. MRI showed no abnormality. Following MRI, cerebrospinal fluid was obtained from the cerebellomedullary cistern. No abnormalities were detected on cytologic analysis including cell counts and protein concentrations. Serologic evaluation for canine distemper virus immunoglobulin G antibodies in the cerebrospinal fluid and serum were negative. Polymerase chain reaction examination of nasal, oral, and conjunctival epithelium, stool, and digital pads was performed and were negative for the canine distemper virus. Toxoplasma gondii and Neospora caninum immunoglobulin G antibodies were both negative in the serum. Serum T4 was measured before anesthesia and was considered still slightly high (94.36 nmol/L). A dose reduction of levothyroxine Na to 0.005 mg/kg PO q 12 hr was instructed. Hypercholesterolemia (19.89 mmol/L) and hypertriglyceridemia (1.36 mmol/L) were further improved.
Five weeks after initial presentation, the dog was reevaluated at the hospital. The body weight decreased to 43.5 kg. Ataxia, exercise intolerance, and facial swelling were resolved, but other neurologic signs including diminished patellar and withdrawal reflexes, head tilt, and ptosis did not show any apparent improvement. Serum creatine kinase was measured and was within normal limits (46.0 IU/L; reference range, 40–151 IU/L). The dog was placed under general anesthesia for evaluation of lower motor neuron dysfunction. Electromyography was within normal limits. Motor nerve conduction velocity in the right tibial nerve was within the reference range, but was polyphasic with a decreased amplitude of the compound muscle action potential (0.9 mV; reference range, 23.3 ± 2.3 mV).10
Biopsy specimens were obtained from the right common peroneal nerve and were fixed in neutral-buffered 10% formalin. Nerve specimens were plastic embedded and evaluated in 1 μm sections (Figure 1). Subjectively, there was a mild (medium-size fascicle) to moderate (smaller fascicle) loss of myelinated fibers without obvious axonal degeneration (Figure 1A). Occasional large and small caliber fibers showed inappropriately thin myelin sheaths for the axon diameter (Figure 1B). Numerous collagen bundles consistent with resolving subperineurial edema were evident in all fascicles (Figure 1A). To determine quantitatively if fiber loss was selective to specific nerve fiber diameters, axonal size frequency distribution was determined for the golden retriever of this report and compared with three large-breed dogs without neurologic abnormalities that were used as controls for another study (Figure 2).11 Compared with the control group, nerve fiber loss was limited to those with axon diameters in the 3–7 µm range.
Citation: Journal of the American Animal Hospital Association 50, 5; 10.5326/JAAHA-MS-6035
Citation: Journal of the American Animal Hospital Association 50, 5; 10.5326/JAAHA-MS-6035
Serum T4 was again reassessed and was within the normal range (40.85 nmol/L). Serum biochemical analysis revealed further improvement in hypercholesterolemia (10.46 mmol/L) and hypertriglyceridemia (0.59 mmol/L). Although the neurologic signs remained, biochemical abnormalities showed marked improvement and the levothyroxine Na dosage was not changed (i.e., 0.005 mg/kg PO q 12 hr).
Six weeks after the initial presentation, the withdrawal reflex normalized but other neurologic abnormalities including diminished patellar reflex, head tilt, and ptosis persisted. The dog was again reevaluated at 6 mo following the initial presentation. The owner reported that the dog showed steady improvement and became completely normal. The body weight increased to 45.9 kg. The head tilt and ptosis resolved, and the patellar reflexes normalized. There were no abnormalities noted on neurologic examination at this time.
Discussion
The polyneuropathy documented in this report is most likely a result of hypothyroidism because the dog had clinical and biochemical abnormalities consistent with hypothyroidism, other causes of neurologic signs such as intracranial disorders were ruled out, and lastly, complete recovery of neurologic abnormalities was achieved following supplementation with levothyroxine Na alone.
In a recent study in which hypothyroidism was experimentally induced in young adult beagles, peripheral neuropathy was not evident 18 mo following induction.8 However, older dogs and dogs with a larger body size (like the dog described herein) were not studied. Peripheral neuropathy associated with hypothyroidism may target only some dogs, and the authors speculate that age and body size may play a role in disease development.
The pathophysiology of hypothyroidism-induced peripheral neuropathy in adults has not been clarified; however, mucinous deposits around peripheral nerves (i.e., mechanical compression and entrapment of nerves by localized myxedema) is considered to be the main cause of hypothyroidism-associated mononeuropathy and multiple mononeuropathy.9 In diffuse polyneuropathy associated with hypothyroidism, direct metabolic changes due to a decreased thyroid hormone level in either Schwann cells or neurons may be considered.9,12 For example, it has been demonstrated in rat studies that thyroidectomy results in slowed axonal transport.13 Because thyroid hormone acts on mitochondria directly and increases adenosine triphosphate (ATP) production and ATPase activity by promoting cellular respiration, reduced thyroid hormone levels induce ATP deficiency and reduce ATPase activity levels in nerve cell bodies. That results in reduced Na+/potassium pump activity and impairment of axonal transport.13
Motor nerve conduction velocity was not decreased in the dog described herein. That finding is not unexpected given that the largest caliber fastest conducting nerve fibers were not decreased (Figure 2). However, a markedly decreased amplitude of the compound muscle action potential and temporal dispersion were noted. Those findings may reflect loss of 3–7 µm nerve fibers and the presence of the inappropriately thin myelinated fibers identified by peripheral nerve biopsy in the dog of this report. Regarding electrophysiological abnormalities of spontaneously occurring hypothyroidism in dogs, variable changes (including reduced motor-nerve conduction velocity and electromyographic abnormalities, such as fibrillation potentials, positive sharp waves, and complex repetitive discharges) have been reported.2,3 Abnormalities on electromyography were not noted in the dog in this report. Although muscle biopsies were not performed, the absence of both muscle atrophy and hypertrophy clinically together with the depressed spinal reflexes suggest that hypothyroid myopathy was unlikely.
There is only limited information concerning pathological changes in peripheral nerves in dogs with hypothyroidism. Peripheral nerve biopsy abnormalities included myelin irregularities, intercalated internodes, internodal globules, and axonal necrosis.2,3,5 Neurogenic atrophy was often seen on muscle biopsy of dogs showing clinical signs of polyneuropathy.2,3 Segmental demyelination and marked myelinated fiber loss accompanied by axonal degeneration were noted on histopathological examination of peripheral nerves in two dogs with signs of polyneuropathy and reduced T4 levels.14 However, both dogs were diagnosed with neoplasia and thyroid-stimulating hormone concentration was not measured. Therefore, it cannot be ruled out that polyneuropathy in those cases was a paraneoplastic syndrome or the cause of reduced T4 levels was euthyroid sickness.
Pathological changes in peripheral nerves in the dog described herein included myelinated fiber loss accompanied by resolving subperineurial edema. To the authors’ knowledge, edematous changes within nerves in dogs with this disease have not been previously reported, but are consistent with pathological findings in human hypothyroid polyneuropathy.15 Although the cause of edema was not investigated in this case, epi- and perineurial infiltration by edematous protein assumed to be a metachromatic mucoid material, probably mucopolysaccharide-protein complexes, has also been observed in humans.15 Perineurial deposition of that substance has not been reported in hypothyroid dogs, but subcutaneous hyaluronic acid deposition manifesting as skin myxedema has been described.16 Hyaluronic acid is a mucopolysaccharide and its deposition may cause edema because it is markedly hygroscopic. Furthermore, myelinated fiber loss apparent in the current case is a common pathological finding in human hypothyroid neuropathy.9 In this dog, edematous deposition resulting in compression of nerve fibers may have occurred due to hypothyroidism.
In canine hypothyroid polyneuropathy, both spinal and cranial nerves may be impaired.5 However, there are reports in which various cranial nerve abnormalities were induced by cerebral infarction due to the cerebral artery atherosclerosis.17,18 Because a necropsy was not performed in this dog, it cannot be definitively concluded that the cranial nerve abnormalities were part of a polyneuropathy and not a central manifestation. However, the neurologic history of this dog (i.e., chronic onset and progression) was not suggestive of a vascular disorder. Furthermore, abnormalities such as necrotic lesions were not noted on MRI. Cranial nerve signs slowly improved and resolved along with spinal nerve signs following only hormone replacement therapy.
Most dogs with hypothyroid myopathy rapidly respond to thyroid replacement therapy, and clinical normalization is obtained within 2–8 wk.2,3 In contrast, the response of peripheral neuropathy to thyroid replacement therapy is not clear and it is said that responses to treatment may be poor.19 The dog in this report required 6 wk to observe the first improvement of neurologic signs after initiation of thyroid replacement. The remaining signs slowly improved, and the dog was neurologically normal on follow up at 6 mo. It has been pointed out that relatively early impairments caused by hypothyroidism are mainly metabolic functional abnormalities, rather than structural changes in peripheral nerves.20 Hence, early recovery by thyroid replacement can be expected at this stage. In contrast, when either apparent structural changes of nerves occur or there is nerve fiber loss resulting from axonal degeneration, a prolonged time may be required to recover or recovery may be incomplete. Therefore, establishing an early and accurate diagnosis and initiating appropriate treatment are necessary when hypothyroid neuropathy is suspected.
Conclusion
The study authors demonstrated that hypothyroid polyneuropathy can occur in dogs as described in humans. Pathological changes in peripheral nerves included myelinated fiber loss accompanied by subperineurial edema. The dog made a full recovery with only thyroid supplementation. However, response to the treatment was slow and most neurologic abnormalities persisted for >6 wk. When apparent structural changes of nerves occur, a prolonged time may be required to recover or recovery may be incomplete. Therefore, establishing an early and accurate diagnosis and initiating appropriate treatment are both necessary when hypothyroid neuropathy is suspected.
A: Low-power, resin-embedded, toluidine blue-stained 1 µm section from the peroneal nerve biopsy showing two nerve fascicles. Resolving subperineurial edema is present in both fascicles. Although axonal degeneration was not noted, nerve fiber loss was suspected indicating previous degeneration. B: Higher-power image of the larger nerve fascicle highlighting edema (*) and nerve fibers with inappropriately thin myelin sheaths for the axon diameter (arrows). Bar represents 50 µm.
Histogram showing axonal size frequency distribution of myelinated nerve fibers from the peroneal nerve of the dog of this report compared with peroneal nerves from three large-breed control dogs ≥8 yr of age and no neurologic dysfunction. Loss of myelinated fibers in the 3–7 µm range was demonstrated in the hypothyroid dog. Bars, standard error of the mean for control dogs.
Contributor Notes
