Materials and Methods

Case selection—Medical records were obtained from the following institutions: Cornell University, North Carolina State University, Texas A&M University, University of Georgia, Pet Emergency and Specialty Center of Marin, and the Veterinary Internal Medicine Practice of Northern Virginia. Qualifying records from August 1999 through February 2008 were reviewed to identify dogs with acquired MG. The diagnosis of acquired MG in all dogs was based on a negative cutoff value of 0.6nM/L for the AChR antibody concentration.¹⁵ Case selection was further limited to dogs treated with PYR alone or MMF + PYR.

Medical records review—Data collected from each record included signalment; body weight; clinical form of acquired MG (eg, focal, generalized, acute fulminating, and paraneoplastic); results of an edrophonium chloride response test^b (if performed); whether the dog had megaesophagus, pneumonia, or both; serum T₄ concentration (if measured); remission and time to diagnosis of remission (if applicable); treatment dose administered; length of the follow-up period; and outcome. Megaesophagus was diagnosed by evaluation of thoracic radiographs. Pneumonia was diagnosed by evaluation of thoracic radiographs and clinical signs consistent with pneumonia. Clinical (pharmacological) remission was defined as the resolution of clinical signs of acquired MG while the treatment directed against MG was continued in that dog. Immune remission was defined as the combination of clinical remission with the return of the AChR antibody titer concentration to < 0.6nM/L without concurrent treatment directed against the acquired MG condition. Death, if attributed to the secondary effects of MG, was included for estimates of mortality rates.

Statistical analysis—Mortality and remission rates were compared between treatments by use of the Fisher exact test.^c The AChR antibody titers were compared between treatments by use of a Wilcoxon signed rank test.^c The Kaplan-Meier product limit method^c was used to estimate survival by treatment group. Log-rank analysis^c was used to compare treatment groups for resolution of megaesophagus, pharmacological or immune remission, time to remission (pharmacological or immune), and outcome (number of months from diagnosis until death); additionally, the same analysis was performed to compare remission and outcome in dogs with the generalized form of MG between the 2 treatment groups. Time from diagnosis of MG to initiation of treatment was also compared between treatment groups by use of a Mann-Whitney U test.^c In addition, effects of an AChR antibody titer, megaesophagus, pneumonia, and serum T₄ concentration less than the reference range on time to MG remission and survival time were determined via backward stepwise Cox proportional hazards regression analysis.^c A value of P < 0.05 was considered significant for all analyses.

Results

Medical records of 27 dogs were reviewed. Twelve dogs were treated with PYR alone, and 15 were treated with MMF + PYR. Dogs in the PYR group ranged from 1 to 12 years old (mean, 5.5 years; median, 5 years), whereas dogs in the MMF + PYR group were 1 to 15 years old (mean, 6.8 years; median, 6 years). With the exception of 1 dog that was transiently (24 days) and unsuccessfully treated (worsening of clinical condition and maintained an AChR antibody concentration) with azathioprine, no other immunosuppressive treatment was administered to any of the remaining dogs. In the PYR group, there were 8 spayed females, 3 neutered males, and 1 sexually intact male. In the MMF + PYR group, there were 7 spayed females, 7 neutered males, and 1 sexually intact male. Body weight for the PYR group ranged from 2.5 to 46 kg (5.5 to 101.2 lb), with a mean of 27.8 kg (61.2 lb) and a median of 30.1 kg (66.2 lb). Body weight for the MMF + PYR group ranged from 3.6 to 40.2 kg (7.9 to 88.4 lb), with a mean of 24.7 kg (54.3 lb) and a median of 27.4 kg (60.3 lb). Dogs in the PYR group included 2 with focal, 8 with generalized, and 2 with acute fulminating forms of MG; dogs in the MMF + PYR group included 3 with focal, 10 with generalized, and 2 with acute fulminating forms. Three dogs, which included 1 dog with the focal form in the PYR group and 2 dogs with the generalized form in the MMF + PYR group, had radiographic evidence of a mass in the cranial aspect of the mediastinum that was suspected to be a thymoma. At the time of diagnosis, the serum AChR antibody concentrations for dogs in the PYR group ranged from 1.1 to 14.35nM/L (mean, 4.17nM/L; median, 2.51nM/L); for dogs in the MMF + PYR group, these antibody concentrations ranged from 0.96 to 14.4nM/L (mean, 3.70nM/L; median, 2.98nM/L). No significant (P = 0.76; Wilcoxon signed rank test) difference was detected in AChR antibody titers between the treatment groups. Megaesophagus was detected in 11 of 12 dogs treated with PYR and 13 of 15 dogs treated with MMF + PYR. Pneumonia was diagnosed in 8 dogs treated with PYR and 8 dogs treated with MMF + PYR. An edrophonium chloride response test^b was performed in 7 dogs treated with PYR (1 with the acute fulminating and 6 with the generalized forms) and 9 dogs treated with MMF + PYR (7 with the generalized and 2 with the acute fulminating forms); results of this test were considered negative in all dogs with the acute fulminating form of acquired MG. Serum T₄ concentration was determined in 4 dogs treated with PYR and 11 dogs treated with MMF + PYR; T₄ concentrations were less than the reference range (1 to 4 μg/dL) in 2 of the dogs treated with MMF + PYR (0.33 and 0.54 μg/dL, respectively). Except for 1 dog that received a continuous rate infusion, PYR was administered at 8- or 12-hour intervals to dogs in both treatment groups. Dosages in the PYR group ranged from 0.04 to 2.0 mg/kg (0.018 to 0.91 mg/lb), with a mean of 1.1 mg/kg (0.5 mg/lb) and a median of 0.97 mg/kg (0.44 mg/lb). Dosages of PYR in the group treated with MMF + PYR ranged from 0.3 to 2.4 mg/kg (0.14 to 1.09 mg/lb), with a mean of 1.1 mg/kg and a median of 0.8 mg/kg (0.36 mg/lb). Dosages of MMF ranged from 4 to 27 mg/ kg (1.8 to 12.3 mg/lb), with a mean of 16.2 mg/kg (7.36 mg/lb) and a median of 20 mg/kg (9.09 mg/lb). Estimates for the time interval from diagnosis of the acquired MG condition to initiation of treatment were available for 11 dogs treated with PYR and 14 dogs treated with MMF + PYR. The number of days that dogs were treated with PYR ranged from 2 to 16 days (mean, 9.2 days; median, 8 days); the number of days that dogs were treated with MMF + PYR ranged from 1 to 75 days (mean, 12.6 days; median, 5.5 days). No significant (P = 0.31; Mann-Whitney U test) difference was detected between treatment groups for the time interval from diagnosis of the acquired MG condition to initiation of treatment.

The follow-up time for dogs treated with PYR ranged from 0.17 to 26 months (mean, 10.3 months; median, 7.5 months); for dogs treated with MMF + PYR, the follow-up time ranged from 0.1 to 47 months (mean, 16.3 months; median, 9 months). Five dogs treated with PYR achieved pharmacological remission, 2 of which also achieved immune remission. Six dogs treated with MMF + PYR achieved pharmacological remission; 2 of these 6 dogs also achieved immune remission. However, during drug withdrawal, there was a relapse from immune remission in these 2 dogs (ie, serologically diagnosed but clinical signs remained inapparent). The time to remission for dogs treated with PYR ranged from 0.37 to 10 months (mean, 3.5 months; median, 2 months). The time to remission for dogs treated with MMF + PYR ranged from 0.9 to 28.2 months (mean, 6.6 months; median, 1.3 months). Megaesophagus resolved in 3 of 11 dogs treated with PYR and in 7 of 13 dogs treated with MMF + PYR. The serum AChR antibody concentration at the time of reexamination was available for 7 dogs treated with PYR and 9 dogs treated with MMF + PYR. The AChR antibody concentrations for the PYR treatment group at the time of reexamination ranged from 0.02 to 3.85nM/L (mean, 0.84nM/L; median, 0.25nM/L); for the MMF + PYR treatment group, these values at the time of reexamination ranged from 0.16 to 21.6nM/L (mean, 2.90nM/L; median, 0.44nM/L).

Mortality rate was not significantly (P = 0.70; Fisher exact test) different between dogs treated with PYR (mortality rate, 33.3% [4/12 dogs]) and dogs treated with MMF + PYR (mortality rate, 46.7% [7/15 dogs]). There was no significant difference (P = 1.0; Fisher exact test) detected between treatment groups for the likelihood of resolution of clinical signs of acquired MG (PYR, 5/12 [42%] dogs; MMF + PYR, 6/15 [40%] dogs), time to resolution of clinical signs (P = 0.90; log-rank test), or survival time (P = 0.64; log-rank test). Kaplan-Meier curves were constructed for clinical signs and survival time (Figures 1 and 2). For the generalized form of acquired MG in dogs, there was no difference (P = 0.81; log-rank test) between treatment groups for time to resolution of clinical signs (P = 0.74; log-rank test) or survival time (P = 0.81; log-rank test). Dogs that did not achieve pharmacological remission during treatment had greater AChR antibody concentrations at the time of diagnosis than did those that achieved pharmacological remission during treatment (P = 0.03; Kruskal-Wallis test); however, AChR antibody concentrations at the time of diagnosis were not significantly different (P = 0.30; Kruskal-Wallis test) between survivors and nonsurvivors.

View Full Size

Kaplan-Meier curve comparing the proportion of dogs with clinical signs over time in 27 dogs with acquired MG receiving treatment with PYR alone (dashed line; n = 12 dogs) or MMF + PYR (solid line; 15). Resolution of clinical signs was defined as clinical (pharmacological) remission.

Citation: Journal of the American Veterinary Medical Association 236, 6; 10.2460/javma.236.6.664

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Kaplan-Meier curve comparing survival over time in 27 dogs with acquired MG receiving treatment with PYR alone or MMF + PYR. See Figure 1 for remainder of key.

Citation: Journal of the American Veterinary Medical Association 236, 6; 10.2460/javma.236.6.664

• Download Figure
• Download figure as PowerPoint slide

A backward stepwise Cox proportional hazards analysis revealed that for the interaction of treatment with MMF, megaesophagus and pneumonia had no significant effect on the number of months until resolution of clinical signs of disease (MMF, P = 0.92; megaesophagus, P = 0.17; and pneumonia, P = 0.20) or survival times (MMF, P = 0.52; megaesophagus, P = 0.52; and pneumonia, P = 0.28). Similarly, Cox proportional hazards analysis for the subset of dogs in which concentrations for T₄, free T₄, or both were available (PYR, n = 4; MMF + PYR, 11) revealed no effect of subnormal serum T₄ concentration on the number of months until resolution of clinical signs of disease (P > 0.9) or survival times (P > 0.90). Elevated AChR antibody titers at the time of diagnosis were associated with a decrease in the likelihood of pharmacological remission (hazard ratio, 0.53 [95% confidence interval, 0.29 to 0.97]; coefficient, −0.64 [P = 0.04]; goodness-of-fit model, −2; log likelihood, 47.2; χ², 8.08 [P = 0.005]) and a nonsignificant increase in the risk of death (hazard ratio, 1.18 [95% confidence interval, 1.01 to 1.37]; coefficient, 0.16 [P = 0.04]; goodness-of-fit model, −2; log likelihood, 60.5; χ², 3.58 [P = 0.058]). These results suggest that for every increase of 1nM/L in AChR antibody concentration at the time of diagnosis, dogs were 1.18 times as likely to die and 0.53 times as likely to achieve pharmacological remission at any point during treatment.

Adverse effects (eg, drooling, vomiting, or soft feces) as a result of treatment were reported for 6 of 12 dogs in the PYR treatment group; similarly, adverse effects were reported for 10 dogs in the MMF + PYR treatment group. The reduction or resolution of adverse effects on the gastrointestinal tract coincided with a reduction in the MMF dose.

Discussion

The treatment of acquired MG in dogs can be challenging. Currently, there is a paucity of controlled and uncontrolled studies conducted to clearly determine the immunosuppressive treatment that is of most benefit. A similar situation exists in the treatment of MG in humans. The benefit of MMF in the treatment of acquired MG in humans is controversial. In contrast to the medical management of acquired MG in dogs, prednisone is considered the immunosuppressive treatment of choice for acquired MG in human patients.¹⁶ Typically, evaluation of the efficacy of other immunosuppressive drugs for acquired MG in humans is accomplished by comparing results for prednisone-treated patients with results for patients treated with prednisone and an investigational drug. The investigational drug is typically prescribed as an adjunct to prednisone treatment, which may also reduce the dose of prednisone. Case reports,^17,18 retrospective case series,^19,20 and open-label uncontrolled studies^21,22 suggest that MMF use is beneficial in the treatment of acquired MG in humans. However, 2 randomized, double-blinded, placebo-controlled prospective studies^23,24 did not find a significant benefit from the administration of MMF over prednisone for the treatment of acquired MG in humans. In contrast to humans with acquired MG, there appears to be a substantial percentage of dogs with acquired MG that will undergo disease remission while only being treated with PYR.⁷ Therefore, dogs being treated because of acquired MG are treated with either PYR alone or PYR with 1 or more immunosuppressive drugs (other than glucocorticoids). Because the dogs in the control group (ie, PYR group) of the present study did not receive an immunosuppressive drug, our study differs from comparative studies of acquired MG in humans.^19–24

Two dogs from each treatment group, which were being treated for acute fulminating MG, died shortly after hospital admission. There was no difference between treatment groups for remission or mortality rates. Because most dogs in both treatment groups had the generalized form of MG, dogs with the generalized form were separately compared by treatment group for remission and mortality rates; no difference was detected between treatment groups for dogs that had the generalized form of MG. Analysis of the results of our retrospective case series does not support the routine use of MMF for the treatment of dogs with acquired MG.

Randomized studies^23,24 have detailed a significant lack of efficacy of MMF; however, MMF has been commonly used and generally has been considered a safe and effective treatment option for acquired MG in humans.^25–27 Adverse effects localized to the gastrointestinal system have been reported.^16,19,20,22 In humans treated with MMF; these effects are detected in a minority of treated patients, are not generally considered major effects, and typically are reversible with a reduction in dose of MMF. Although similar adverse effects appeared to be dose related, these effects were detected in 18 of 27 (67%) dogs treated with MMF in the present study. Even though this was a self-limiting phenomenon, the likelihood that MMF treatment would cause a disturbance in the gastrointestinal tract of dogs needs to be considered as well as the apparent overall lack of efficacy of MMF for controlling the myasthenic condition.

Limitations are associated with our study, most of which are disadvantages of a retrospective case series. In addition to the relatively small number of cases, dogs were not randomly assigned to treatment groups. Because veterinarians typically include MMF as a treatment in dogs that exhibit an initial poor response to PYR alone, the MMF + PYR group may have been biased by dogs with acquired MG that were more refractory to treatment. Because 2 dogs treated with PYR achieved sustained immune remission, whereas the acquired MG condition relapsed in 3 dogs treated with MMF + PYR during the withdrawal phase of MMF treatment, the concept that the MMF + PYR group may have consisted of dogs with acquired MG that was more difficult to treat could be supported. Even if we assume this concept is correct, despite not having a comparative group of dogs, a pharmacological remission rate of only 6 of 15 (40%) and a mortality rate of 46.7% do not provide support for including MMF during treatment of dogs with refractory MG. Furthermore, we cannot predict the mortality rate in a group of dogs with acquired MG in which MMF was not administered. The treatment groups compared in the present study were similar with respect to signalment, serum AChR antibody concentrations, and distribution of clinical forms of acquired MG; however, there was a substantial range in the doses administered for both PYR and MMF in each group, which could have affected the outcome of the treatments. Ideally, a prospective, double-blinded, placebo-controlled study should be conducted to evaluate MMF treatment in dogs with acquired MG. Because of the risk of death in dogs with poorly controlled acquired MG and the lack of an accepted standard immunosuppressive drug for treating this disease in dogs, such an investigation poses an ethical dilemma.

An interesting finding of the present study, unrelated to drug regimen, was the inverse relationship between AChR antibody titer at the time of diagnosis and likelihood of attaining disease remission for all dogs. This finding may have some clinical use when planning a treatment strategy, frequency of reevaluation of AChR antibody concentrations, and rate of drug withdrawal.

In this retrospective case series, no significant benefit was detected in the treatment of dogs with acquired MG by administration of MMF + PYR versus PYR alone. Clinicians need to interpret the results of this investigation cautiously for several reasons, which involve the small number of dogs included in this investigation, the lack of randomization to treatment groups, and the variability of disease severity among the dogs and dose of drugs administered.