Histopathological Confirmation of Polyneuropathy in 11 Dogs With Laryngeal Paralysis

Introduction

Laryngeal paralysis (LP) is an important cause of upper airway obstruction in dogs¹ where dyspnea occurs secondary to the lack of abduction of the arytenoid cartilages during inspiration. Clinical signs of LP include stridor, exercise intolerance, dysphonia, cyanosis, respiratory distress, cough or gag, hyperthermia, and collapse.^2–9 In addition, dogs with LP may present with clinical signs of generalized muscle weakness including difficulty rising, paresis, and dysphagia, or they may develop these clinical signs several months after the recognition of respiratory disease.^5,10–13

Although LP can affect either young or old dogs of any breed as an acquired condition, a breed predisposition in the Bouvier des Flandres, Siberian husky, bull terrier, Dalmatian, leonberger, Pyrenean mountain dog, and rottweiler is apparent, suggesting a hereditary basis for LP.^10,14–18 To our knowledge, neither an inheritance pattern nor a specific mutation in a disease-causing gene has been identified for any of the known, breed-specific LP polyneuropathy syndromes. Breed-specific LP may be a single entity or part of a generalized polyneuropathy complex.^10,14 Some dogs diagnosed with breed-specific LP may also have gait abnormalities, hyporeflexia, weakness, exercise intolerance, proprioceptive deficits, and muscle atrophy.^14,15

Acquired LP is most often diagnosed in large-breed dogs, with Labrador retrievers and rottweilers being the most common.¹ Old dogs are diagnosed with LP more often than young dogs, with the majority of affected dogs being over 6 years of age.^1,13 The acquired form of LP is most often considered idiopathic but can be caused by trauma, neoplasia, or an infection. It may also be iatrogenic, occur as part of a paraneoplastic syndrome, or be associated with endocrine or metabolic disorders (such as hypothyroidism or diabetes mellitus).^3,9,13 Myasthenia gravis (MG) has also been implicated as a cause of LP, although in such cases, the LP usually occurs with other clinical signs of MG and not as an isolated respiratory presentation.^19,20

Idiopathic LP was once thought to be an isolated clinical problem, but recent studies suggest that LP may actually be one component of a generalized neuromuscular disorder.^3,5,11,12 In one study, confirmed neuromuscular disease was reported in 22% of dogs surgically treated for LP.¹³ Electrophysiological changes, such as decreased amplitude of the compound muscle action potential and decreased motor nerve conduction velocities, support degeneration of the distal part of the longest peripheral nerve fibers as an underlying cause of the LP.¹² While Braund et al¹² describe their histopathological findings in muscle and nerve biopsies obtained from two dogs with acquired LP, the full report only describes electrodiagnostic findings and muscle and nerve biopsy results in one mature dog with acquired LP.¹²

The purposes of this study were to investigate a group of mature dogs presented to our hospital with acquired LP and to histologically confirm the presence of generalized polyneuropathy in those dogs. In this study, electrodiagnostic testing and histopathology were evaluated in laryngeal and pelvic limb muscles and nerves to test the hypothesis that dogs with acquired LP have a generalized polyneuropathy complex. To our knowledge, no previous study has described the histopathological findings from biopsies of the muscular branch of the recurrent laryngeal nerve and cricoarytenoideus dorsalis muscle, distant peroneal nerve, and cranial tibial muscle in the same dogs. Additionally, no previous studies have reported histopathological results from biopsies of the recurrent laryngeal nerve, peroneal nerve, cricoarytenoideus dorsalis muscle, and cranial tibial muscle in combination with electrodiagnostic results for more than one dog.

Materials and Methods

Enrolled in this study were 11 mature, client-owned dogs that were presented to the University of Tennessee Veterinary Teaching Hospital between April 2004 and March 2006 for surgical treatment of LP. This study was approved by our institution’s laboratory animal care and use committee, and consent for the procedures performed in this study was obtained from each dog owner.

All 11 dogs had clinical signs consistent with respiratory difficulty, and all were evaluated for LP by direct laryngeal examination under either sedation or light anesthesia. All dogs were evaluated when anesthesia was light enough to permit gagging and swallowing. When the diagnosis was in doubt, expert opinions of multiple observers were solicited, and doxapram HCL^a was used to promote maximal respiratory effort.²¹

A thorough physical examination, neurological examination, complete blood count (CBC), and serum biochemical profile were performed on each of the 11 dogs. All electrodiagnostic recordings were conducted on anesthetized dogs using a multimodal electrodiagnostic computer.^b Electromyography (EMG) was performed for six of the 11 dogs using a monopolar needle as the active electrode referenced to a subdermal needle electrode. The analyzed muscles included the cranial tibial, interosseous, and gastrocnemius. These muscles were sampled by moving the active electrode to several depths and positions. In addition, sciatic/tibial and/or ulnar motor nerve conduction studies were performed for six and five of the 11 dogs, respectively. The amplitudes of proximal and distal-evoked compound muscle action potentials (CMAP) were also measured.

Nine of the 11 dogs diagnosed with LP were treated by a unilateral cricoarytenoid cartilage lateralization procedure.²² Two of the 11 were not treated surgically and were instead euthanized. Cricoarytenoideus dorsalis and cranial tibial muscle biopsies, as well as recurrent laryngeal and peroneal nerve biopsies were collected from each patient (either at the time of surgery or immediately after euthanasia) by the same surgeon. Biopsies from the cricoarytenoideus dorsalis muscle and the muscular branch of the recurrent laryngeal nerve were obtained through the incision created for the cricoarytenoid cartilage lateralization procedure. Biopsies of the cranial tibial muscle and peroneal nerve were harvested from the lateral surface of the pelvic limb, distal to the stifle.²³ All samples were immersion fixed in 10% neutral-buffered formalin. The muscles were embedded in paraffin and evaluated in standard hematoxylin and eosin-stained sections. Following fixation, the peripheral nerves were rinsed, processed, and embedded in araldite resin as previously described.²⁴ Nerve biopsy sections (1 μm) were stained with toluidine blue and evaluated by light microscopy. All biopsies were examined by the same pathologist.

All nine dogs that underwent surgery for acquired LP were followed until either the time of death or for a minimum of 16 months. Follow-up information on eight of the nine survivors was obtained by physical examination or by telephone interview. Only one dog was lost to follow-up.

Results

Eight of the 11 dogs included in this study were Labrador retrievers. Two were mixed-breed dogs, and one was a Brittany spaniel. Seven of the dogs were spayed females, three were neutered males, and one was an intact male. Average age of the dogs was 12.5 years (range 10 to 16 years), and the average weight of the dogs was 30.1 kg (range 15 to 42.3 kg).

All 11 dogs were diagnosed with bilateral LP on direct laryngeal examination. Physical and neurological examinations were performed on each dog. Stridor and/or dyspnea were found on physical examination in eight of the 11 dogs. Nonrespiratory abnormalities were identified on neurological examination in four of the dogs and included generalized weakness (n=1), decreased conscious proprioception of the pelvic limbs (n=1), pelvic limb hyporeflexia (n=2), and difficulty rising (n=1). Difficulty rising was noted in the histories of five other dogs but was not noted on physical examinations. Based on these examinations, a generalized neuromuscular disorder was suspected in at least four of the dogs. Serum biochemical profiles and CBCs were performed for every dog; no consistent abnormalities were present.

Electromyography was performed for six dogs. For two dogs, EMG showed mild to moderate numbers of fibrillation potentials mixed with scattered positive sharp waves in the cranial tibial, interosseous, and gastrocnemius muscles. No EMG abnormalities were found in the other four dogs. Tibial nerve conduction velocity (NCV) was determined in six dogs, and ulnar NCV was determined in five dogs [see Table]. The mean±standard deviation (SD) sciatic/tibial NCV was 40.8±5.5 meters per second (reference range 59 to 69 meters per second).^25,26 One dog’s tibial NCV was not recordable due to technical difficulties and was therefore not included in the calculation of means. In the thoracic limbs, mean±SD of the ulnar NCV was 37.4±16.2 meters per second (reference range 58 to 60 meters per second).^25,26 The CMAP amplitudes were reduced, and CMAP durations were prolonged as a result of stimulating the sciatic/tibial and ulnar nerves at both proximal and distal sites. The mean±SD CMAP amplitude for the proximal site on the sciatic/tibial nerves was 8.4±4.8 mV, and for the distal site it was 10.4±6.5 mV (reference ranges 20.7±2.2 mV and 24.5±3.8 mV for proximal and distal sites, respectively).²⁶ The mean±SD CMAP amplitude for the proximal site on the ulnar nerve was 12.2±8.4 mV, and for the distal site it was 19.8±10 mV (reference ranges 25.1±1.9 mV and 28.5±1.7 mV for proximal and distal sites, respectively).²⁶ Abnormalities were detected in each of the dogs undergoing electrodiagnostic evaluation.

Pathological changes in muscle biopsies from all 11 cases included combinations of large- and small-grouped atrophy in the cricoarytenoideus dorsalis muscle and cranial tibial muscle specimens with varying degrees of severity [Figures 1A–1D]. Concurrent with this pattern of muscle fiber atrophy were large nerve fiber loss, axonal degeneration, and endoneurial fibrosis, which were consistently identified within the resin-embedded nerve sections [Figures 2A–2D]. Mixed axonal degeneration and demyelination were found in two cases. Regenerative changes were not observed. Together, these changes confirmed the diagnosis of polyneuropathy due to chronic axonal degeneration and nerve fiber loss. A specific cause for neuropathy could not be identified from the muscle and nerve specimens.

In the nine surviving dogs, progression of clinical signs between the time of surgery and the time of follow-up or death included dysphagia (n=2), megaesophagus (n=1), lower motor neuron tetraparesis (n=1), pelvic limb hyporeflexia (n=1), decreased conscious proprioception of the thoracic limbs (n=1), and decreased conscious proprioception in the pelvic limbs (n=2). Six of the 11 dogs with acquired LP were euthanized or died within 15 months of diagnosis.

Discussion

All 11 dogs diagnosed with LP had pathological changes in muscle and nerve biopsies from both the laryngeal region and the pelvic limb that were consistent with denervation. In light of the histopathological findings, decreased CMAP amplitudes, prolonged CMAP durations, and decreased NCV, we suggest that LP in the examined dogs was an early clinical sign of a generalized polyneuropathy.¹¹ A specific cause for denervation could not be determined from the biopsy sections. As none of the dogs were hypothyroid, diabetic, or had a detectable neoplasia, a degenerative neuropathy was considered the most likely cause of the generalized polyneuropathy.

Muscle specimens were evaluated in paraffin-embedded sections, which is not the optimal method of evaluation. Frozen sections would have provided an opportunity for fiber typing and determination of fiber-type grouping, which might have further supported a diagnosis of neurogenic atrophy. Even so, pathological findings in paraffin sections of both the pelvic limb and laryngeal muscles, as well as the peroneal and laryngeal nerve biopsies, were consistent with active denervation. The pattern of muscle fiber atrophy with small and large groups of atrophic fibers in the absence of specific myopathic changes is typical of denervation. Unfortunately, the most distal intramuscular nerve branches could not be adequately evaluated in these preparations, so we could not determine if these dogs had a “dying-back” type of neuropathy that typically shows fiber loss distally.

Consistent with the pathological changes observed in the muscle biopsies, loss of large-caliber nerve fibers, axonal degeneration, and endoneurial fibrosis were identified within the resin-embedded sections of both the peroneal and recurrent laryngeal nerves in all cases. In a previous report, distal nerve biopsies were performed on only two dogs with acquired LP, and electrodiagnostic findings were reported for only one of these cases.¹² In these previously reported cases, histopathological findings were similar to those described in the present study. Additionally, the present study included biopsies of the muscular branch of the recurrent laryngeal nerve, cricoarytenoideus dorsalis muscle, peroneal nerve, and cranial tibial muscle in 11 dogs, with electrodiagnostic findings in six of these dogs. To our knowledge, a study this extensive has not been previously reported.

Abnormalities in electrodiagnostic procedures performed on the thoracic and pelvic limbs were consistent with a generalized polyneuropathy. Fibrillation potentials and positive sharp waves are indicative of denervation and occur when lower motor neurons are lost. Generally, the frequency of these potentials is correlated with the number of motor neurons lost and, at least initially, with lapsed time following destruction. In two of the six dogs undergoing electrodiagnostic testing, EMG showed mild to moderate numbers of fibrillation potentials indicating denervation. In all of the cases in which electrodiagnostic procedures were performed, CMAP amplitude was reduced.

The onset latency of the CMAP is the point at which the waveform deviates from baseline and is used to measure the velocity of the fastest-conducting motor axons. The motor units innervated by slower-conducting axons are activated later and participate in the CMAP without distinction. A decrease in motor NCV reveals involvement of at least the largest motor axons in the nerve being tested. The amplitude of a CMAP is a partial reflection of the number of activated motor units, their sizes, and synchrony. The reduction in CMAP amplitudes found in the dogs included in this study indicates a reduction in participating motor units and/or dispersion. Conditions that cause peripheral dispersion of motor axon action potentials will increase the CMAP duration and may change the waveform morphology from a bi- or triphasic waveform to polyphasic waveform. Therefore, based on evidence of denervation, slowed motor NCV, and reduction of CMAP amplitude, we know that lower motor neurons in the dogs of this study showed functional impairment.

The first clinical sign of generalized polyneuropathy appears to be LP. One possible explanation for this supposition could be related to the length of the recurrent laryngeal nerve. The recurrent laryngeal nerve has a long route descending with the vagosympathetic trunk, separating from the vagus nerve and ascending along the trachea to innervate the larynx.²⁷ This theory would be consistent with our findings of slowed NCV in the distal pelvic limb and histopathological abnormalities in the peroneal nerve, as the distal pelvic limb is an area innervated by a branch of the sciatic nerve, which is also a long peripheral nerve.

The electrodiagnostic findings in this study were consistent with a “dying-back” type of neuropathy. Since the intramuscular nerve branches could not be evaluated in muscle biopsies and only one point was examined along a peripheral nerve, this classification can only be assumed pathologically. Nonetheless, the extent of nerve fiber loss and absence of regenerative changes are important in the long-term prognosis for dogs with this condition. In these cases, regeneration was not observed, making the long-term prognosis guarded.

The 11 dogs included in this study exhibited clinical signs of neuropathy that localized to the recurrent laryngeal nerve (resulting in LP in all of the dogs) and to the sciatic nerve (resulting in pelvic limb weakness in some of the dogs). Of the dogs exhibiting conscious proprioception deficits, one dog had deficits in all four limbs, while three dogs had deficits in the pelvic limbs only. Additionally, three dogs exhibited pelvic limb hyporeflexia without hyporeflexia in other locations. Specifically, these three dogs had hyporeflexia localizing to the sciatic nerve, including decreased gastrocnemius reflex and decreased cranial tibial reflex.

A major limitation of this study is the absence of an age-matched control population. Although age-matched unaffected dogs were not available for comparison of histopathological findings and electrodiagnostic procedures, normal values are established for adult dogs to which these samples were compared. One dog in this study had a sciatic/tibial NCV that was not recordable by our equipment. We suspect this was due to human error or equipment malfunction, as the dog was able to walk. Additional limitations include the small population size, subjectivity in client interviews, and lack of complete data in all cases.

Conclusion

The 11 patients included in this study all exhibited generalized neuromuscular dysfunction as evidenced by electrodiagnostic testing and histopathology of muscle and peripheral nerve biopsy specimens. Electrodiagnostic testing should be performed (if available), and muscle and peripheral nerve biopsies should be collected (either as a separate evaluation or during anesthesia for laryngeal surgery) to establish a diagnosis and determine a long-term prognosis. Clinicians should be aware that while LP may be the only presenting clinical sign, the LP could be part of a generalized neuro-muscular disorder. Although dogs diagnosed with LP may eventually develop clinical signs of a generalized polyneuropathy, we do not believe that this possibility should discourage veterinarians from performing corrective surgery for the LP, because the polyneuropathy may be slowly progressive. Laryngeal paralysis can be a life-threatening condition, so surgery is warranted in dogs with LP; however, owners need to be informed of the potential long-term prognosis and the possibility for development of additional, slowly progressive, neuromuscular abnormalities.