Clinical Course of Acute Canine Polyradiculoneuritis Following Treatment with Human IV Immunoglobulin

Introduction

Acute canine polyradiculoneuritis (ACP) is an acquired peripheral neuropathy primarily involving the ventral nerve roots characterized by the rapid development of a nonambulatory, lower motor neuron (LMN) tetraparesis/tetraplegia. Dogs are often unable to lift their head, cannot vocalize, and some react hyperesthetically to pressure on the distal limbs. In severe cases, the respiratory muscles may also be affected. Cranial nerves (other than the facial nerve), tail wagging, and bladder and bowel function are usually not impaired.^1–3 However, trigeminal neuropathy and Horner´s syndrome were the prominent clinical features in one dog with polyradiculoneuritis and ganglionitis.⁴ ACP is also commonly called coonhound paralysis because it is believed to be caused by an immune-mediated reaction to an antigen present in the saliva of a raccoon, but a raccoon bite is not a requirement for this diagnosis.¹

Guillain-Barré syndrome (GBS) is considered the human counterpart of ACP.^1,5,6 Although the exact pathogenesis remains unknown, some authors have postulated that this syndrome may be caused by an autoimmune response against peripheral nerve antigens. Potential triggers are thought to include exposure to bacteria such as Campylobacter jejuni, Mycoplasma pneumonia, and Haemophilus influenzae and viruses such as cytomegalovirus and Epstein-Barr virus, as a kind of molecular mimicry.^1,5,7,8

Dogs will often recover within 3–5 wk with only physical rehabilitation and supportive care. However, prolonged courses (up to 3 mo), incomplete recovery and lack of improvement have been observed. Thus, some animals may not improve within the period that supportive care is tenable for the owners.^1,2,5,9,10

In humans with GBS, treatment with either plasmapheresis or high-dose IV immunoglobulin (IVIg) has been evaluated in controlled clinical trials.^11–14 Treated patients had a significantly more rapid motor recovery, a shortened time to recover walking without aid, and less frequently required artificial ventilation than untreated patients.¹⁵ Because IVIg is much simpler to administer and less likely to cause complications than plasmapheresis, IVIg has been adopted as the favored treatment in humans.^11,15,16 Its mechanisms of action have not yet been completely elucidated, but are mainly attributed to sialylated Fc-linked glycans.¹⁷

In veterinary medicine, IVIg is considered effective for the treatment of auto-immune hemolytic anemia (AIHA) and, more recently, has been used for immune-mediated thrombocytopenia, severe adverse cutaneous drug reactions, Stevens-Johnson syndrome, and pemphigus foliaceus.^18–24 The purpose of this clinical pilot study was to describe the clinical course of ACP following treatment with IVIg and compare this treatment with dogs with ACP that were treated with supportive care only.

Materials and Methods

Sixteen client-owned dogs diagnosed with ACP and treated with IVIg were included. Five dogs were identified retrospectively through a search of the medical records database of the Clinic of Small Animal Medicine, Ludwig-Maximilians University, Munich, Germany (2000–2004), and the remaining 11 dogs were included prospectively in an ongoing study (2005–2010). Informed consent was obtained from all the owners. A control group including 14 client-owned dogs with ACP that did not receive IVIg was identified retrospectively through a medical record search (2000–2010, Animal Health Trust, Newmarket).

The inclusion criteria for both groups were rapid development of nonambulatory LMN tetraparesis or tetraplegia within 2 wk of onset, no historical evidence of possible exposure to botulinum neurotoxin, hematology, serum biochemistry (including creatine kinase), and at least 6 mo of follow-up information. When available, serology for Toxoplasma gondii immunoglobulin (Ig)M/IgG and Neospora caninum IgG, serum total thyroxine, thyroid-stimulating hormone, electrodiagnostic examination, cerebrospinal fluid (CSF) analysis, MRI, computed tomography, and muscle/nerve biopsy were performed to exclude other causes of pelvic limb weakness on initial presentation.

Information retrieved from the medical records was verified and completed during telephone conversations with clients and veterinarians. The dogs’ disabilities were evaluated before and after IVIg treatment using a modified, simple 6-point scale based on the functional GBS-grading scale used in humans.^25,26 In that scale, 6 was normal strength; 5 was minor symptoms or signs of neuropathy, but capable of running (weakness); 4 was ambulating without assistance at least five steps, but incapable of running; 3 was ambulating with assistance; 2 was nonambulatory tetraparesis; 1 was tetraplegia; and 0 was ventilatory support.

CSF was collected by cisternal puncture and processed routinely for total nucleated cell count (reference range, <3/μL), erythrocyte count, protein concentration (reference range, <0.32 g/L), and cytospin differential cell count.

Muscle and nerve biopsies were obtained from either the peroneal nerve or tibial nerve and the gastrocnemius muscle and cranial tibial muscle. Samples were placed on saline-soaked gauze and either immediately submitted to a specialized veterinary neuromuscular laboratory on campus (Institute of Veterinary Pathology/Neuropathology, Ludwig-Maximilians-University Munich) or mailed with cold packs via overnight mail to the Institute for Neuropathology, Heinrich-Heine-University Düsseldorf.

An additional inclusion criterion for the treatment group was a complete electrodiagnostic examination^a. Dogs were anesthetized with diazepam^b, propofol^c and isoflurane^d. Electromyography (EMG) was performed with a concentric needle electrode^e (37 mm in length, 0.46 mm in diameter, with a 0.07 mm² recording area). Routinely, all proximal and distal limb muscles, the tail and paraspinal muscles, and the facial and masticatory muscles on one side of the body were explored. Specific attention was paid to persistent spontaneous electric activity (SPA). Nerve stimulation studies focused on the investigation of the tibial nerve in all but three dogs. Following proximal (trochanter) and distal (hock) tibial nerve stimulation with monopolar Teflon-coated needle electrodes, compound muscle action potentials (CMAPs) were recorded from the plantar interossei muscle using surface recording electrodes [an alligator clamp served as active recording electrode and was placed plantar over the distal aspect of the interossei muscles; a subcutaneous platinum needle served as a reference electrode and was placed plantar at the base of the fourth digit ]. Distal latency and amplitude of the CMAPs were measured. Motor nerve conduction velocity (MNCV) was calculated for the trochanter-hock segment of the tibial nerve. F waves were measured with the same electrode configuration, but with the cathode positioned proximal to the anode. F waves were recorded from the plantar interossei muscle. Pelvic limb length was measured from the greater trochanter of the femur to the tip of the fourth digit. Minimum F wave latencies were measured, and the F ratio was calculated. Reference ranges for all measurements were derived from 22 healthy beagles and Labrador retrievers that were examined using the same technique. Distal CMAP latencies were evaluated with a regression equation.²⁷

During the retrospective phase of the study, dogs were treated with human IVIg^f according to published protocols in veterinary medicine (the dose was 0.5 g/kg).¹⁹ During the prospective phase, higher doses were used as recommended for humans with GBS (1.5 g/kg split into aliquots of 0.5 g/kg, which were administered on three consecutive days).^7,8 Initially, small volumes were infused and the dogs were carefully monitored for any side effects. Clinical observation included mental status, body temperature, capillary refill time, pulse quality, and heart and respiratory rates. Those parameters were monitored every 5 min in the first hour then q 30 min thereafter. Subsequently, the Igs were infused over 6 hr. The primary outcome was the duration (measured in days) from IVIg administration to ambulating without assistance ≥5 steps (i.e., grade 4). In addition, the duration of the initial progressive phase (measured in days from onset to nadir of weakness), the time (measured in days) from onset to first IVIg infusion, the duration of the complete episode (from onset to ambulating without assistance), and long-term follow-up were recorded. Correlations between dose of IVIg and duration from IVIg administration to ambulating without assistance (i.e., the recovery time) and the association between recovery time and body weight (measured in kg) were also calculated.

Statistical Analysis

Age, weight, duration of the initial progressive phase, and duration of the complete episode (onset to ambulation without assistance) were compared between the IVIg-treated and control groups using the Mann-Whitney U-test. Associations between recovery time and IVIg dose or body weight (measured in kg) were assessed by Spearman's rank correlation coefficient. The level of significance was set at P < 0.05 for all calculations. Data were analyzed descriptively (i.e., median, percent and range), and have been displayed graphically. All analyses were performed using SPSS v.17^g.

Results

A precipitating cause was not identified, and there was no historical evidence of possible exposure to botulinum toxin, in any of the 30 included dogs. Except of one dog (case 6), the complete laboratory evaluation and the serum thyroxine (n = 17) and thyroid stimulating hormone (n = 13) levels were unremarkable. In addition, serology for Toxoplasma gondii (n = 19) and Neospora caninum (n = 24), MRI (n = 12), and computed tomography (n = 1) were unremarkable. CSF (n = 24) did not reveal pleocytosis in any of the included dogs. CSF protein concentrations were normal in 19 dogs (median, 0.18 g/L; range, 0.11–0.32 g/L; reference range <0.32 g/L), but was marginally elevated in five dogs (median 0.39 g/L, range, 0.34–0.45 g/L). Muscle biopsies were performed in 17 dogs, which revealed variable amounts of neurogenic muscle atrophy. Peripheral nerve biopsies did not demonstrate pathologic alterations in most (13/17) dogs. Four dogs showed a reduction in myelinated fiber count, several thin myelinated fibers, and slightly increased inflammatory cell infiltrations in the nerve biopsy specimens.

IVIg Treatment Group

IVIg was administered to 16 dogs diagnosed with ACP. Clinical signs at the nadir and signalment are listed in detail in Table 1. Two dogs (cases 3 and 4) had similar episodes of LMN tetraparesis 1 mo and 4 mo prior to the episode included in this report, from which they fully recovered. In case 6, which had been diagnosed with AIHA two years earlier, the laboratory evaluation was unremarkable, except for a mild anemia (hematocrit, 30%; reference range, 35–58%). An adrenocorticotropic hormone stimulation test was abnormal in case 16.

TABLE 1 Clinical Features of 16 dogs with Acute Canine Polyradiculoneuritis Treated with Human IV Immunoglobulin

*

From onset to ambulating without assistance ≥ five steps (i.e., grade 4)

†

For dogs with multiple episodes, only the episodes in which dogs were nonambulatory and treated with IVIg are specified.

F, female; IVIg, intravenous immunoglobulin; M, male.

View Fullscreen

Electromyography was performed between days 6 and 47 (median, 11 days). All 16 dogs showed widespread SPA in the truncal, proximal, and distal limb muscles and small CMAP amplitudes following distal and proximal peripheral nerve stimulation. Polyphasic CMAPs indicative of temporal dispersion were recorded in seven dogs after distal tibial nerve stimulation (cases 3, 4, 8, 10, 11, 14, and 16). MNCV was either normal or mildly to moderately decreased. F waves were measured in 14 dogs. They were detected and recordable in six dogs, but absent in the others. In five of the six dogs with recordable F waves, the F waves were delayed and the F ratio was variable. CMAP latencies following distal tibial nerve stimulation were prolonged in all but two dogs when referring to a regression equation (Table 2).²⁷

TABLE 2 Electrodiagnostic Features of 16 dogs with Acute Canine Polyradiculoneuritis Treated with Human IV Immunoglobulin

*

Expected distal compound muscle action potentials (CMAP) latency for the tibial nerve is <0.03 × distance (mm) – 0.09. Expected distal CMAP for the ulnar nerve is <0.02 × distance (mm) + 0.75 where the distance (mm) is between the negative stimulating electrode and the negative recording electrode.²⁷

†

Expected minimum F wave latency for the tibial nerve is <0.39 × distance (cm) + 3.49. Expected minimum F wave latency for the ulnar nerve is <0.38 × distance (cm) + 1.94, where distance (cm) of the pelvic limb is the length from the greater trochanter of the femur to the tip of the fourth digit, and distance (cm) of the thoracic limb is the length from the cranial end of the scapula to the tip of the 4th digit.²⁷

§

Ulnar nerve

**

In dogs with multiple episodes, only the episodes in which dogs were nonambulatory and treated with IVIg were included.

††

Radial nerve

§§

Peroneal nerve

Unless otherwise indicated, electrodiagnostics were performed on the tibial nerve. SPA is measured using the following scale: 0, none; 1+, rare; 2+, moderate, unsustained; 3+, moderate, sustained; 4+, profuse, filling the whole screen. ↓, below RR; ↑, above RR; –, not done; CMAP, compound muscle action potential; EMG, electromyography; MNCV, motor nerve conduction velocity; NR, unable to record; p, peroneal nerve; r, radial nerve; RR, reference range; SPA, spontaneous electric activity; t, tibial nerve; u, ulnar nerve.

View Fullscreen

All but two dogs (cases 12 and 15) were treated with IVIg within 3–24 days (median, 11.5 days) of the first evidence of weakness noted by the owner. At the time of IVIg treatment, those 14 dogs had been nonambulatory for 1–23 days (median, 6 days), and there had been no evidence of motor function recovery. Cases 12 and 15 had been nonambulatory for 39 and 45 days, respectively. Prednisolone^h had been administered to all but four dogs (cases 2, 14, 15, and 16) by the primary veterinarian prior to IVIg, but no improvements had been observed. Because prednisolone was administered by the primary veterinarian, the exact dose for each dog was not known. Glucocorticoid treatments were continued throughout the course of the disease in three dogs (cases 3, 6, and 8), which were identified retrospectively. Case 6 was also treated with azathioprineⁱ because of a history of AIHA, which was in remission.

Pronounced improvement was noted in 10 dogs treated with IVIg (62.5%, cases 1, 3, 5, 6, 9–12, 14, and 15), which were ambulating without assistance within 13 days of IVIg (median, 8 days; range, 3–13 days) as shown in Figure 1. In those dogs, the complete episode (i.e., onset to ambulating without assistance) lasted 15–58 days (median, 21.5 days).

View Figure

• Download as Powerpoint
• Download Figure

Less improvement of motor function was noted in six other dogs (cases 2, 4, 7, 8, 13, and 16), which showed increased voluntary movements that were most notable 1–3 days following IVIg infusion, but those dogs did not regain the ability to ambulate without assistance within 2 wk (Figure 2). In those six dogs, the median time to ambulation without support after IVIg was 65.5 days (range, 19–120 days), and the complete episode (i.e., onset to ambulating without assistance) lasted between 29 and 127 days (median, 81.5). Cases 7, 13, and 16 regained the ability to ambulate without assistance 3 mo, 4 mo, and 1.5 mo later, respectively. A second IVIg infusion was administered to cases two and four after they remained nonambulatory after 3 wk and 2 wk, respectively. Following the second IVIg treatment, both dogs were ambulatory without assistance within 5–10 days thereafter. In case 8, IVIg was discontinued after 0.3 g/kg had been infused because of an adverse reaction (i.e., elevated body temperature, tachycardia, and hyperventilation). A mild microscopic hematuria occurred in another dog (case 4).

View Figure

• Download as Powerpoint
• Download Figure

Considering all dogs, cases 7 and 8 had received the lowest IVIg doses (0.5 g/kg and 0.3 g/kg, respectively). They took 90 and 120 days, respectively, to ambulate without assistance following IVIg administration. There was a significant association between IVIg dose and the recovery time (Spearman's rank correlation coefficient -0.577; P = 0.019), but no association between body weight and recovery time.

Three dogs (cases 2–4) had multiple episodes of tetraparesis. With respect to IVIg treatment, case 4 had one prior, case 2 had one following, and case 3 had one prior and two following IVIg treatment. The interval between paretic episodes ranged between 24 days and 1 yr (median, 60 days). The onset to nadir was ≤9 days for each episode. Cases 3 and 4 recovered from their first episode with supportive care and physical rehabilitation alone, but when LMN tetraparesis recurred 1–4 mo later, they were treated with IVIg because of impending euthanasia. Case 3 had two mild relapses after 12 mo and 14 mo, during which he was still ambulatory, but displayed weakness in all limbs, was unable to jump, and had decreased withdrawal reflexes in all limbs. Each time the treatment with IVIg was repeated, the dog recovered completely in 8–10 days (after time of onset). During the last episode, this dog was treated with 2 g/kg IVIg.

All dogs but one (case 12) recovered completely and have been healthy since time of recovery (range, 2 mo–7.5 yr, date of last follow-up Nov 15, 2010). Case 12 died 1 yr after treatment because of a cardiac tamponade caused by hemangiosarcoma in the right atrial appendage (confirmed at necropsy).

Control Group

Fourteen client-owned dogs diagnosed with ACP were treated with nursing care and physiotherapy only. Age, weight, and duration of the initial progressive phase did not differ between dogs with IVIg treatment and controls (Table 3; P = 0.08, 0.87, and 0.45, respectively). Due to the retrospective acquisition of this group, no standardized electrodiagnostic protocol was followed in these patients. Electrodiagnostics were performed in 9 of the 14 dogs between days 1 and 14 after onset of disease (median, 5 days). All dogs that were studied electrodiagnostically at least 2 days after disease onset had similar electrodiagnostic findings as the treatment group, with widespread SPA and either normal or mildly decreased MNCV. Eleven dogs recovered completely and one was ambulatory, but still had signs of weakness 8 mo after the initial presentation. The two other dogs died. One died on day 4 after onset of disease due to respiratory arrest and the second was euthanatized because of lack of improvement after 54 days. Duration of the complete episode from onset to ambulation without assistance ranged from 5 days to 220 days (median, 75.5 days). There was a trend toward faster recovery in dogs with ACP treated with IVIg (range, 15–127 days; median, 27.5 days) compared with the recovery of the retrospective controls without IVIg treatment, but this difference did not reach statistical significance (P = 0.32). Dogs that died were not included in that analysis. Nonetheless, inclusion of dogs that died had no effect on the comparison between groups.

TABLE 3 Clinical Features of the 14 dogs with Acute Canine Polyradiculoneuritis Included in the Control Group

*

From onset to ambulating without assistance ≥5 steps (i.e., grade 4).

F, female; M, male.

View Fullscreen

Discussion

Based on the success of IVIg therapy in humans with GBS, 16 dogs with ACP were treated with IVIg to determine the safety of IVIg and describe the clinical course of dogs with ACP following treatment with IVIg. At the time of IVIg treatment, all 16 dogs had been nonambulatory for 1–45 days (median, 6.5 days), and no improvement in motor function had been noted. Following IVIg treatment, 62.5% of the treated dogs rapidly regained their ability to walk (Figure 1), and subtle improvements in motor function were demonstrated in the others dogs treated with IVIg (Figure 2). Notably, two dogs rapidly regained the ability to walk following IVIg infusion after a prolonged nonambulatory period.

Even with the inclusion of a retrospective control group, this study should only be considered a clinical pilot study. The duration of the complete episodes was longer in dogs that were not treated with IVIg (median, 75.5 days) than in dogs treated with IVIg (median, 27.5. days). Because ACP is a self-limiting disease and because the study was not designed as a prospective, controlled, blinded investigation, it is not possible to definitively indicate whether the course of the 16 dogs treated with IVIg would have been the natural course of their disease or not. In dogs with ACP, recovery of motor function may begin as early as 1 wk after nadir of weakness, and the majority of dogs regain function over a period of a few weeks to several months.^1,9

Although ACP is one of the most commonly recognized canine peripheral neuropathies of dogs, it is a rare disease, and there is a lack of description of the clinical course in the recent literature.¹ Retrospective investigations have detailed the course of ACP (also called coonhound paralysis) in a total of 32 dogs, of which 15 survived and 17 died.^2,5,6,10In 4 of the 17 nonsurviving dogs described in the literature, death was attributed to either pneumonia or respiratory failure, and the others were euthanized at the owner’s request because of poor recovery of function.^2,5,6,10 Similarly, two dogs included in the control group in the study described herein either died or were euthanized. In the treatment group, owners considering euthanasia because of the relapsing course in cases 2–4, the lack of improvement of motor function over prolonged periods of time in cases 12 and 15, or because they could not manage a nonambulatory tetraparetic dog >30 kg (cases 1 and 14). In humans with GBS, which mirrors ACP in many aspects, plasmapheresis and IVIg therapy appear to be successful treatment modalities. This special treatment both prevents complications and hastens recovery by its immunomodulatory effects, leading to a reduction in mortality to 5%.^8,13,28 Reported recovery times (from initiation of therapy to walking without aid) were a median of 51 days, 55 days, or 65 days with IVIg treatment and a median of 85 days or 111 days with supportive care in adults in various studies. A placebo-controlled trial demonstrated a hastened recovery with IVIg treatment (median, 15 days) compared with supportive care (median, 24.5 days) in children.^11,29,30

In this study, ACP, similar to GBS in humans, was predominantly a clinical diagnosis, but electrophysiological tests, CSF analysis, and muscle and nerve biopsies were used to confirm the diagnosis and exclude other causes of acute LMN tetraparesis.^8,31 Consistent electrodiagnostic findings were rapidly evolving widespread SPA, decreased CMAP amplitudes with or without temporal dispersion, and either delayed or absent F waves, which, in combination with the acute progressive clinical course, strongly suggested a diagnosis of polyradiculoneuritis.³ Similarly, small amplitude CMAPs and widespread SPA have been previously described in cases of ACP and in people with the axonal form of GBS.^3,32 Small amplitude CMAPs may also be a feature of botulism, but intensive questioning failed to reveal any possible source of botulism in any of the dogs, and neither temporal dispersion nor severe widespread SPA is typically found in botulism. In some dogs, failure to observe F waves may have been secondary to the already decreased CMAP amplitude. CSF analysis was performed to exclude inflammatory CNS diseases, such as poliomyelitis. This analysis typically reveals increased protein without an increase in WBCs (i.e., albuminocytologic dissociation) in dogs and humans.^1,6,8,28,33 Only five dogs of our study showed increased CSF protein, but all CSF samples were taken by cisternal puncture, and protein elevation may be more evident in the lumbar CSF than in cisternal CSF of dogs with ACP.^1,6 Because blood contamination occurs more frequently during lumbar puncture, cisternal puncture was preferred because of the necessity to exclude infectious and immune-mediated myelitis.

Diagnostic criteria for GBS in humans include features required for diagnosis (i.e., progressive motor weakness of more than one limb and areflexia) and features that are strongly supportive of the diagnosis (i.e., progression for <4 wk, symmetry of symptoms, albuminocytologic dissociation), but a diagnosis does not require nerve biopsies because nerves are usually biopsied distally, where inflammatory cell infiltrates are not commonly found.^{1,6,10,28,31,33} Consequently, muscle and nerve biopsies may only be necessary to rule out other underlying causes with similar clinical presentation, such as acute inflammatory myopathy and necrotizing myopathy, and to investigate pathogenetic factors.⁷ As expected in ACP, the biopsy results demonstrated mostly normal peripheral nerves and variable neurogenic muscle atrophy.

There is no specific proven effective treatment of ACP, and most clinical studies mention the lack of efficacy of corticosteroids in dogs with ACP.^1,33 Thus, the treatment of ACP is currently limited to supportive care and physical therapy. This study confirmed that steroids are not effective in that neither the dogs pretreated with steroids nor the two dogs in which glucocorticoids were continued throughout the course of the disease showed any improvement in motor function prior to IVIg. Similarly, controlled clinical trials in humans show that steroids are not effective for the treatment of GBS.^7,8,15,16,28 Furthermore, treatment with oral corticosteroids for ≥2 wk may even slow the recovery from GBS.^26,34 The lack of response to corticosteroids is poorly explained in the literature. It is possible that corticosteroids may have harmful effects on denervated muscle or inhibit macrophage repair processes.^7,8,26

The median IVIg dose used in the dogs included in this study (1.3 g/kg) was only marginally lower than the GBS IVIg regimen recommended in people (2 g/kg split into aliquots of 0.4 g/kg administered for 5 consecutive days). The exact dose of IVIg for GBS treatment has not been determined.^7,8,14,35 However, one study involving human patients compared 3 days versus 6 days of 0.4 g/kg IVIg, with a beneficial trend observed in favor of the larger dose.³⁶ Another study in human patients showed that infusion with a standard regimen (2 g/kg) in GBS patients resulted in considerable variability in the increase of serum IgG levels (ΔIgG) between patients, which is related to clinical outcome. In that study, a low ΔIgG was significantly associated with poor outcome.³⁷ In the current study, recovery time following IVIg treatment was inversely associated with the dose of IVIg (P < 0.05). The lowest doses were administered to cases 7 and 8 (who were treated with 0.5 g/kg and 0.3 g/kg, respectively), and only subtle improvements in motor function were noted. The time needed to regain the ability to walk without support after IVIg infusion was very long in those two cases (90 days and 120 days) compared with the other dogs treated with IVIg (mean, 11 days). Thus, it is not possible to exclude a potential dose-response effect of IVIg.

Except for cases 12 and 15, IVIg infusion was initiated within the first 3 wk after the onset of weakness when the dogs were still nonambulatory without any evidence of improvement. Although there are no studies regarding the benefit of early IVIg administration in humans, early treatment is recommended when plasmapheresis is used as a treatment of GBS.^7,16,28,38 Nonetheless, steady improvement of motor function was also observed in cases 12 and 15 on the second day following IVIg infusion, and those two dogs had been nonambulatory for 5.5 wk and 6.5 wk, respectively, so the benefits of IVIg could have been part of the natural course of the disease.

In GBS patients who are treated with IVIg, severe adverse reactions, such as acute renal failure, aseptic meningitis, skin reactions and anaphylaxis, occur in <4% of GBS patients.³⁹ However, 30–40% of GBS patients suffer from mild and transient adverse effects, including headache, nausea, fever, and fatigue. These effects usually disappear when infusion is resumed at a slower rate.^{11,14,17,35,39} In the present study, adverse effects were noted during (case 8) and after (case 4) IVIg infusion in two dogs. Because IVIg therapy can increase the risk of thromboembolism, the development of renal microthrombi is one theory to explain how IVIg may have caused the observed hematuria in case 4.^40,41 The adverse reaction in case 8 was so profound that treatment with IVIg was discontinued. In other reports of IVIg use in dogs, no relevant adverse effects were observed.^18–24,40 Because there is some concern that dogs may develop antibodies to human immunoglobulin, which could precipitate a severe hypersensitivity reaction, the administration of multiple doses of IVIg is not recommended.^19,21,22,24 Despite that concern, two dogs in this study (cases 2 and 4), and a few dogs with AIHA in other studies, were treated a second time. In addition, one dog with pemphigus foliaceus was administered seven IVIg cycles in yet another study, all with no apparent adverse effects.^20,21,40 Nevertheless, the safety of administering multiple IVIg doses needs to be more thoroughly examined.

Recurrent ACP is a rare condition, and there are only a few reports of dogs with multiple episodes of otherwise typical ACP.^2,5 Cases 2–4 had two or more episodes of ACP, and the clinical features of the individual episodes were similar to acute monophasic illness. Acute onset of symptoms, complete recovery, normal MNCV, lack of inflammation on peripheral nerve biopsies, and lack of an apparent response to glucocorticoids were used to distinguish the acute form of polyradiculoneuritis from chronic idiopathic demyelinating polyneuropathy (CIDP), which often has a relapsing course and is generally corticosteroid-responsive.⁴² In humans, relapses and worsening after initial improvement (i.e., treatment-related fluctuation) have been described in up to 16% of GBS patients. Those relapses may be induced by either insufficient dosage or duration of IVIg therapy. Conversely, up to 16% of human patients with CIDP may present acutely (i.e., acute CIDP). It is difficult to distinguish these two forms. The diagnosis can only be confirmed with certainty at follow-up. In humans, acute CIDP should be suspected when either three or more episodes of deterioration occur or progression takes place after 9 wk from onset of disease.^7,43,44 Based on this definition, at least case 3 (classified as ACP) may instead have had CIDP with an acute presentation. As previously mentioned, there is some concern that IVIg treatment affects the immune system, and treated dogs may develop antibodies to human Igs. Because there is strong suspicion that immune stimulation plays an important role in the pathogenesis of ACP, this may be an alternative explanation for the recurrence of ACP in the dogs included in this study. However, two of the three dogs with multiple episodes of paresis developed their first relapse before being treated with IVIg. In addition, there is no current evidence in other reports of IVIg use in dogs and human patients that IVIg treatment results in ACP/GBS or a relapse of these diseases. Therefore, a causal association between IVIg treatment and relapse seems unlikely.

Conclusion

The limitations of this study were the comparison with a retrospective control group and the small number of dogs included. A medical treatment that could alter the clinical course of ACP would be valuable for dogs for which euthanasia is considered due to a lack of improvement within the period of supportive care that is tenable for the owner. Further prospective, placebo-controlled, blinded investigations are warranted to prove and more closely describe the possible beneficial effect, should one exist, of IVIg in ACP.