Immune-mediated polyarthritis is an important cause of lameness and fever in dogs. In 1 study,1 20% of dogs with fever of unknown origin had polyarthritis. Middle-aged large-breed dogs may be overrepresented.2 The pathogenesis of IMPA is believed to involve a type III hypersensitivity reaction in which immune complexes are deposited in the synovial membrane, initiating an inflammatory cascade. Other clinical signs in affected dogs include those of joint pain, joint swelling, and systemic illness (eg, lethargy, inappetence, and weight loss).3
In dogs, IMPA has been classified on the basis of whether erosive lesions are detected within the joint via radiography and whether an underlying etiology, such as infectious agents or certain drugs (eg, sulfonamides), is identified.2–5 For most affected dogs with nonerosive polyarthritis, an underlying etiology is not identified; this condition has been termed idiopathic (type I) polyarthritis.6 Reactive (type II) polyarthritis may result from infection with Borrelia burgdorferi,7,8 granulocytic ehrlichiae,9,10 Leishmania spp,11 and possibly Bartonella spp12,13 or from underlying bacterial endocarditis.14
Diagnosis of IMPA is based on quantitative and qualitative evaluation of synovial fluid samples obtained via arthrocentesis. In dogs with IMPA, synovial fluid contains an increased proportion of nondegenerate neutrophils, compared with the proportion in healthy dogs.3 Oral administration of immunosuppressive doses of prednisone or prednisolone remains the standard treatment for primary IMPA.6 For affected dogs with refractory disease, azathioprine, leflunomide, cyclophosphamide, levamisole, or mycophenolate mofetil can be used in combination with glucocorticoid drugs for additional immunosuppression and to allow reduction in the dose of glucocorticoid drugs administered.15–18 However, few studies have been conducted to evaluate the efficacy of these treatments. To the authors' knowledge, no prospective clinical studies of treatment efficacy have been reported for dogs with IMPA.
Monitoring of the response to treatment for IMPA can be performed by analysis of synovial fluid samples collected every 3 to 4 weeks.19 Adjustments to treatment are typically made on the basis of clinical and cytologic findings. Adverse effects of glucocorticoid treatment in dogs can be severe and include muscle atrophy, weakness, polyuria, polydipsia, polyphagia, weight gain, and gastrointestinal signs.20 These adverse effects can negatively impact quality of life, particularly when imposed on an existing musculoskeletal disorder.
Cyclosporine is a potent immunosuppressive agent that has been successfully used to treat several immune-mediated or inflammatory conditions in dogs.21 The drug is a calcineurin inhibitor that inhibits T-cell activation and prevents synthesis of several cytokines, particularly interleukin-2.21,22 The most common adverse effects of cyclosporine treatment are gastrointestinal signs (particularly vomiting and diarrhea) and gingival overgrowth. Vomiting and diarrhea reportedly occur in 15% to 30% of treated dogs, with most dogs having < 3 episodes over a period of several months; these episodes typically abate with time.22 The adverse effects of cyclosporine may be less likely than those of glucocorticoids to impact quality of life of dogs with IMPA. Therefore, we hypothesized that treatment with cyclosporine alone for immunosuppression would be as efficacious as prednisone alone for treatment of dogs with primary IMPA, without the adverse effects of polyphagia, polyuria, polydipsia, or muscle atrophy that are associated with prednisone treatment. The purpose of the study reported here was therefore to compare outcomes for a 90-day treatment period involving orally administered cyclosporine or prednisone in dogs with primary IMPA.
Materials and Methods
Animals
Dogs evaluated for IMPA at the veterinary teaching hospital at the University of California-Davis from 2008 to 2011 were eligible for inclusion in the study. Dogs were enrolled when they had a diagnosis of primary (idiopathic) IMPA that had been made on the basis of consistent historical, physical examination, and clinicopathologic findings (ie, results of CBC, serum biochemical analysis, urinalysis, thoracic radiography, and abdominal ultrasonography), coupled with results of synovial fluid analysis that indicated neutrophilic inflammation in 2 or more joints, including a small joint (tarsus or carpus). Criteria for neutrophilic inflammation in joints consistent with IMPA included high cell count (> 3,000 cells/μL), with > 50% nondegenerate neutrophils in the differential nucleated cell count.
Other diagnostic tests performed in an attempt to rule out secondary IMPA (echocardiography, serologic and PCR assays to detect vector-borne disease [including detection of antibodies against Anaplasma spp, Borrelia burgdorferi, and Ehrlichia spp and antigen of Dirofilaria immitisa], serologic testing [immuno-fluorescent antibody assay] for Bartonella infection, spinal or joint radiography, and microbial culture of blood [including for Bartonella], synovial fluid, or urine samples) were performed at the discretion of the attending clinician with consideration of dog demographic and historical factors, physical examination findings, preliminary test results, and client financial resources. Dogs were excluded from the study when they had an established diagnosis of primary IMPA that had not responded to previous immunosuppressive drug treatment, a history of glucocorticoid treatment at dosages ≥ 0.5 mg/kg/d (0.23 mg/lb/d) within the previous month, an established diagnosis of erosive IMPA, or a comorbid disease (eg, concurrent infection) that could have indicated secondary IMPA. Owners signed a written consent form to allow their dogs to participate in the study. The study protocol was approved by the Clinical Trials Review Board of the Veterinary Medical Teaching Hospital, University of California-Davis (protocol No. 09-04-16).
Study protocol
Dogs were randomly assigned by drawing of cards from a hat to receive treatment with prednisone or cyclosporine. Starting on day 0, dogs in the prednisone group received prednisoneb at a dosage of 1 mg/kg (0.45 mg/lb), PO, every 12 hours, with a maximum total dosage of 40 mg, PO, every 12 hours. This dosage was subsequently reduced by 25% every 2 to 3 weeks if physical examination revealed resolution of lameness and joint effusion and cytologic analysis revealed a decrease in the amount of abnormalities detected in synovial fluid samples from a minimum of 3 of 4 joints (estimated total cell count < 3,000 cells/μL). Dogs in the cyclosporine group received cyclosporinec at a dosage of 5 mg/kg (2.3 mg/lb), PO, every 12 hours for the duration of the study period. In that group only, concurrent treatment with carprofend (2.2 mg/kg [1 mg/lb], PO, q 12 h for the first 7 days), tramadole (0.8 to 1.2 mg/kg [0.36 to 0.55 mg/lb], po, q 8 to 12 hours), or both was allowed for analgesia. Owners of dogs in the cyclosporine group were asked to bring their dogs back 7 days after treatment initiation for collection of blood samples, at which point trough whole blood cyclosporine concentrations were determined by means of high-performance liquid chromatography. All blood cyclosporine concentrations were measured at the same laboratory.
Owners were asked to complete a questionnaire on day 0 (just before initiation of treatment) and again on days 14, 45, and 90. This included questions about their dogs with regard to appetite, perceived comfort or gait, and degrees of physical activity, lameness, and lethargy. Responses were provided by use of a 5-point scale (0 = unaffected, 1 = minimally affected, 2 = mildly affected, 3 = moderately affected, and 4 = severely affected). An overall mobility score was derived by adding scores of owner-perceived comfort or gait and lameness. Owners also scored their dogs with respect to perceived quality of life at the same points (0 = excellent, 1 = good, 2 = fair, 3 = poor, and 4 = very poor). On days 14, 45, and 90, owners were also asked to record the frequency with which they observed each of 5 adverse effects (ie, polyuria, polydipsia, panting, vomiting, and diarrhea) by use of a 4 point scale (0 = none, 1 = mild, 2 = moderate, and 3 = severe).
A complete physical examination was performed on each dog on days 0 (just before treatment), 14, 45, and 90, and results were recorded on a standardized scoring sheet at each visit by the same internal medicine clinician that initially evaluated them, although multiple clinicians were involved in the study. Variables documented included body weight, rectal temperature, mentation, hydration status, gait or lameness, and signs of pain, crepitus, and effusion for each of 8 peripheral joints (carpi, tarsi, stifles, and elbows). Subjective findings were scored on a 5-point scale (0 = clinically normal, 1 = minimally affected, 2 = mildly affected, 3 = moderately affected, and 4 = severely affected). An overall locomotion score was calculated by adding the scores assigned for gait, joint effusion, and signs of joint pain.
Arthrocentesis was also performed on days 0 (just before treatment), 14, 45, and 90. Prior to the procedure, food was withheld from dogs for 8 hours. Dogs were sedated with a combination of dexmedetomidine hydrochloridef (2.5 μg/kg [1.1 μg/lb], IV) and butorphanol tartrate8 (0.2 mg/kg [0.1 mg/lb] IV) or butorphanol tartrate (0.2 mg/kg, IV) and midazolam hydrochlorideh (0.2 to 0.3 mg/kg [0.09 to 0.14 mg/lb], IV). A minimum of 4 peripheral joints (both carpi and both tarsi and, for some dogs, the stifle joints or elbows as well) were clipped of hair and aseptically prepared. Arthrocentesis was performed by use of a 3/4-inch, 22-gauge or 5/8-inch, 25-gauge needle attached to a 3-mL syringe. The effects of dexmedetomidine were reversed after the procedure by IM administration of atipamezole hydrochloride.i Synovial fluid samples were submitted to the clinical pathology laboratory at the teaching hospital for analysis. When sufficient synovial fluid was available (typically only for fluid obtained from stifle joints), samples were treated with hyaluronidase and cell counts were performed by use of an automated analyzer.j
Synovial fluid smears on microscope slides were prepared and stained with Wright-Giemsa stain, and a licensed medical technologist examined each to perform differential cell counts of large mononuclear cells (macrophages), small mononuclear cells (lymphocytes), and neutrophils. All stained slides were then examined via light microscopy by 1 board-certified clinical pathologist (WV) who was unaware of treatment group assignment. Cytologic findings were scored by use of a standard scoring system for total and differential cell counts as described elsewhere.23 When identified, inflammation was classified as mononuclear reactivity only (nucleated cell numbers within reference limits but high percentage of large, foamy activated macrophages), mononuclear (only) inflammation, mixed inflammation (neutrophilic and histiocytic), or neutrophilic inflammation. Because in the authors' experience mononuclear inflammation may develop in dogs with long-standing IMPA, treatment effects, or the residuum of prior suppurative inflammation, this classification was included and scored if it was absent at the beginning of the study period (day 0) but was subsequently identified. A final inflammation score was assigned to each joint fluid as follows: 0 = no inflammation, 1 = mononuclear reactivity only, 2 = mild inflammation (regardless of inflammation type), 3 = moderate inflammation (regardless of inflammation type), and 4 = marked inflammation (regardless of inflammation type). For each set of joints from each dog at each assessment point, mean inflammation score and mean neutrophilic inflammation score were calculated by averaging scores across all joints in that set. In addition, maximum neutrophilic inflammation score (neutrophilic inflammation score for the joint in the set that had the highest neutrophilic inflammation score [scored as 0 when no neutrophilic inflammation was identified]) was recorded.
Treatment failure was defined as a need to change to a different drug because of lack of clinical improvement by day 14 (together with adequate trough blood cyclosporine concentrations for dogs in the cyclosporine group), lack of improvement in cytologic scores for synovial fluid samples by day 45 (together with adequate trough blood cyclosporine concentrations for dogs in the cyclosporine group), or need to change to a different drug at any point during the study period because of unacceptable adverse drug effects. For dogs in the cyclosporine group that did not have clinical improvement but had trough blood cyclosporine concentrations < 250 ng/dL, the cyclosporine dosage was increased until adequate trough concentrations (250 to 500 ng/dL) were achieved. When lack of clinical improvement persisted, the affected dog was then considered a treatment failure, and although clinical scoring was continued for monitoring purposes, scores assigned after the point of treatment failure (ie, once the dog was receiving a different treatment) were not included in statistical analysis. Adverse effect scores were no longer recorded when other medications were added to the treatment regimen. However, when a treatment was changed from cyclosporine to prednisone or vice versa for a given dog, adverse effects scoring was continued and used for comparison purposes.
Statistical analysis
Proportions of dogs with treatment success and drug adverse effects were compared between treatment groups with the Fisher exact test. The Wilcoxon-Mann-Whitney test was used to compare distributions of owner-perceived overall mobility scores, owner-perceived quality-of-life scores, clinician-assessed overall locomotion scores, rectal temperatures, body weights, mean synovial fluid inflammation scores, mean synovial fluid neutrophilic inflammation scores, and maximum synovial fluid neutrophilic inflammation scores at baseline between treatment groups. This test was also used to compare differences in distributions of scores at each assessment point relative to baseline between groups. The Spearman correlation coefficient (ρ) was calculated to determine whether a correlation existed between subjectively measured clinical variables (overall mobility and locomotion scores) and synovial fluid inflammation scores across all assessment points. For all statistical analyses, values of P < 0.05 were considered significant. Analyses were performed by use of statistical software.k
Results
Animals
Twenty dogs with primary IMPA met all inclusion criteria and were enrolled. Enrolled dogs included 10 males (9 neutered and 1 sexually intact) and 10 females (all spayed). Twelve dogs were considered large breed (> 15 kg [33 lb]), and 8 were considered small breed (≤ 15 kg). Breeds included mixed breed (n = 5), Dachshund (2), and German Shepherd Dog, Australian Shepherd, Staffordshire Bull Terrier, Maltese Terrier, Chihuahua, Vizsla, English Setter, Gordon Setter, Greater Swiss Mountain Dog, Chinese Crested, Italian Greyhound, Rottweiler, and German Shorthaired Pointer (1 each).
Clinical signs of primary IMPA on day 0 (just prior to treatment) consisted of lethargy (n = 19 dogs), decrease in physical activity (19), lameness (16), inappetence (16), and fever (rectal temperature > 39.2°C [102.5°F]; 9). Serologic testing for antibodies against Anaplasma spp, Borrelia burgdorferi, and Ehrlichia spp and for antigen of Dirofilaria immitis was performed at least 10 days after the onset of clinical signs for 16 dogs. Other diagnostic tests performed were aerobic microbial culture of urine samples (n = 11 dogs), cytologic analysis of lymph node or splenic aspirates (10), microbial culture of synovial fluid samples (7), echocardiography for detection of endocarditis (5), microbial culture of blood samples (4), radiography of peripheral joints (4), spinal radiography or spinal MRI (3), and serologic testing for antibodies against Bartonella henselae, Bartonella clarridgeiae, and Bartonella vinsonii subsp berkhoffii (2). One dog had a urinary tract infection caused by a gram-negative bacterium that was successfully treated with antimicrobials immediately before immunosuppressive drug treatment was initiated. Results of all remaining diagnostic assays were unremarkable, and results of all tests for concurrent infectious diseases that could result in secondary IMPA were negative.
Three of the 10 dogs in the prednisone group also received tramadol at the point of enrollment. Carprofen (n = 5), tramadol (2), or both (2) were prescribed to provide analgesia for 9 of the 10 dogs in the cyclosporine group.
Treatment outcome
Clinical treatment success (ie, lack of defined treatment failure) was achieved for 7 dogs in each treatment group of 10 dogs. Of the dogs with treatment failure in the prednisone group, 2 dogs had a lack of clinical improvement by day 14 of treatment and the remaining dog had a lack of improvement in cytologic scores for synovial fluid samples by day 45. Primary IMPA in 1 dog with lack of clinical improvement by day 14 was subsequently managed with a combination of prednisone and cyclosporine, although resolution of synovial fluid cytologic abnormalities was never obtained during the study period. The owner of the second dog with lack of clinical improvement by day 14 declined alternative immunosuppressive drug treatment, and the dog died suddenly on day 45; cause of death was not ascertained. The dog with lack of cytologic improvement by day 45 had azathioprine added to the treatment regimen, and the synovial fluid abnormalities resolved.
Of the 3 dogs with treatment failure in the cyclosporine group, 1 dog developed diarrhea that was believed to be an adverse effect of cyclosporine treatment. That dog was subsequently treated with prednisone, after which diarrhea resolved and a good clinical and cytologic response to treatment was observed. The second dog failed to have clinical improvement by day 14, and it was subsequently successfully treated with prednisone alone. The third dog failed to have cytologic improvement by day 45, and its owners declined additional immunosuppressive treatment because the dog no longer had clinical signs of polyarthritis. The cyclosporine dosage was subsequently tapered, and the dog remained in clinical remission at the end of the study period.
Only 1 of the 7 dogs in the cyclosporine group that was considered to have had treatment success by the definition used had a persistence of clinical signs by day 14. Because that dog had a low trough blood cyclosporine concentration (182 ng/mL) on day 7, the cyclosporine dosage was subsequently increased and the dog remained in the study. By day 75, complete resolution of clinical signs and cytologic abnormalities in synovial fluid samples was achieved. Trough blood cyclosporine concentration was reevaluated after the end of the 90-day study period (on day 104), at which point it was 645 ng/mL. This was the only dog in the cyclosporine group requiring an increase in cyclosporine dosage.
Initial measurements of trough blood cyclosporine concentration were made 7 to 17 days (median, 10 days) after treatment began in the cyclosporine group (Figure 1). Clinical improvement was apparent even with trough cyclosporine concentrations as low as 70 ng/mL
No significant differences in disease severity scores were identified between the 2 treatment groups for any of the variables assessed at baseline, defined as day 0 (prior to initiation of treatment; Table 1). No significant differences were identified between groups in changes from baseline in owner-perceived overall mobility and quality-of-life scores, clinician-assessed locomotion scores, or rectal temperature as assessed on day 14 (P ≥ 0.57 for all), day 45 (P ≥ 0.90 for all), or day 90 (P ≥ 0.42 for all; Figure 2). No significant (P = 0.16) difference between groups was identified for change in body weight between day 0 and day 45.
Whole blood cyclosporine concentrations in 9 dogs with primary IMPA treated with cyclosporine at a dosage of 5 mg/kg (2.3 mg/lb), PO, every 12 hours. Concentrations were measured 7 to 17 days (mean and median, 10 days) after treatment began. Horizontal line represents the median.
Citation: Journal of the American Veterinary Medical Association 248, 4; 10.2460/javma.248.4.395
Whole blood cyclosporine concentrations in 9 dogs with primary IMPA treated with cyclosporine at a dosage of 5 mg/kg (2.3 mg/lb), PO, every 12 hours. Concentrations were measured 7 to 17 days (mean and median, 10 days) after treatment began. Horizontal line represents the median.
Citation: Journal of the American Veterinary Medical Association 248, 4; 10.2460/javma.248.4.395
Whole blood cyclosporine concentrations in 9 dogs with primary IMPA treated with cyclosporine at a dosage of 5 mg/kg (2.3 mg/lb), PO, every 12 hours. Concentrations were measured 7 to 17 days (mean and median, 10 days) after treatment began. Horizontal line represents the median.
Citation: Journal of the American Veterinary Medical Association 248, 4; 10.2460/javma.248.4.395
Mean scores for owner-perceived overall mobility (A), owner-perceived quality of life (B), and clinician-assessed overall locomotion (C), mean rectal temperature (D), and mean body weight (E) for dogs with primary IMPA at various points during treatment with prednisone (1 mg/kg [0.45 mg/lb], PO, q 12 h; solid line; n = 10) or cyclosporine (5 mg/kg, PO, q 12 h; dashed line; 10). Assessments on day 0 were made prior to treatment initiation that day. Overall mobility scores were derived by adding scores for owner-perceived comfort or gait and lameness (for each, 0 = unaffected, 1 = minimally affected, 2 = mildly affected, 3 = moderately affected, and 4 = severely affected). Quality of life was scored on a 5-point scale (0 = excellent, 1 = good, 2 = fair, 3 = poor, 4 = very poor). Overall locomotion score was calculated by adding clinician-assigned scores for gait, joint effusion, and signs of joint pain (for each, 0 = clinically normal, 1 = minimally affected, 2 = mildly affected, 3 = moderately affected, 4 = severely affected). Bars represent SD.
Citation: Journal of the American Veterinary Medical Association 248, 4; 10.2460/javma.248.4.395
Mean scores for owner-perceived overall mobility (A), owner-perceived quality of life (B), and clinician-assessed overall locomotion (C), mean rectal temperature (D), and mean body weight (E) for dogs with primary IMPA at various points during treatment with prednisone (1 mg/kg [0.45 mg/lb], PO, q 12 h; solid line; n = 10) or cyclosporine (5 mg/kg, PO, q 12 h; dashed line; 10). Assessments on day 0 were made prior to treatment initiation that day. Overall mobility scores were derived by adding scores for owner-perceived comfort or gait and lameness (for each, 0 = unaffected, 1 = minimally affected, 2 = mildly affected, 3 = moderately affected, and 4 = severely affected). Quality of life was scored on a 5-point scale (0 = excellent, 1 = good, 2 = fair, 3 = poor, 4 = very poor). Overall locomotion score was calculated by adding clinician-assigned scores for gait, joint effusion, and signs of joint pain (for each, 0 = clinically normal, 1 = minimally affected, 2 = mildly affected, 3 = moderately affected, 4 = severely affected). Bars represent SD.
Citation: Journal of the American Veterinary Medical Association 248, 4; 10.2460/javma.248.4.395
Mean scores for owner-perceived overall mobility (A), owner-perceived quality of life (B), and clinician-assessed overall locomotion (C), mean rectal temperature (D), and mean body weight (E) for dogs with primary IMPA at various points during treatment with prednisone (1 mg/kg [0.45 mg/lb], PO, q 12 h; solid line; n = 10) or cyclosporine (5 mg/kg, PO, q 12 h; dashed line; 10). Assessments on day 0 were made prior to treatment initiation that day. Overall mobility scores were derived by adding scores for owner-perceived comfort or gait and lameness (for each, 0 = unaffected, 1 = minimally affected, 2 = mildly affected, 3 = moderately affected, and 4 = severely affected). Quality of life was scored on a 5-point scale (0 = excellent, 1 = good, 2 = fair, 3 = poor, 4 = very poor). Overall locomotion score was calculated by adding clinician-assigned scores for gait, joint effusion, and signs of joint pain (for each, 0 = clinically normal, 1 = minimally affected, 2 = mildly affected, 3 = moderately affected, 4 = severely affected). Bars represent SD.
Citation: Journal of the American Veterinary Medical Association 248, 4; 10.2460/javma.248.4.395
Mean ± SD values of clinical variables for dogs with primary IMPA before treatment with prednisone (1 mg/kg [0.45 mg/lb], PO, q 12 h; n = 10) or cyclosporine (5 mg/kg [0.23 mg/lb], PO, q 12 h; 10).
| Variable | Prednisone | Cyclosporine | P value |
|---|---|---|---|
| Body weight (kg) | 18.8 ± 15.1 | 19.5 ± 9.3 | 0.62 |
| Rectal temperature (°C) | 39.3 ± 1.1 | 39.3 ± 0.8 | 0.90 |
| Owner-perceived overall mobility score | 5.5 ± 2.8 | 6.3 ± 2.2 | 0.54 |
| Owner-perceived quality-of-life score | 2.7 ± 1.3 | 2.8 ± 0.9 | 0.97 |
| Clinician-assessed overall locomotion score | 5.0 ± 3.6 | 4.7 ± 2.0 | 1.00 |
| Mean score for synovial fluid inflammation | 3.2 ± 0.7 | 3.1 ± 0.6 | 0.93 |
| Mean score for synovial fluid neutrophilic inflammation | 3.1 ± 0.8 | 3.0 ± 0.7 | 0.78 |
| Maximum score for synovial fluid neutrophilic inflammation | 3.6 ± 0.5 | 3.6 ± 0.7 | 1.00 |
To convert kilograms to pounds, multiply by 2.2. Values of P < 0.05 were considered significant.
Possible drug-related adverse effects reported by dog owners were summarized (Table 2). Dogs treated with prednisone were more likely to have signs of polyuria, polydipsia, and polyphagia on days 14 and 45 than were dogs treated with cyclosporine. No difference in the prevalence of other adverse effects was identified among assessment points. One dog treated with cyclosporine was moved to the prednisone group in the first week of the study because of diarrhea on day 5 (and therefore adverse effect scores for cyclosporine were not available for this dog). Two other dogs in the cyclosporine group had intermittent vomiting and diarrhea throughout the study period. Two dogs treated with prednisone had diarrhea on day 14, and vomiting and diarrhea were reported for 1 dog treated with prednisone on day 90.
Summary of owner-assessed adverse effects on days 14, 45, and 90 of treatment for the dogs with primary IMPA in Table 1.
| Proportion of dogs affected* | Median overall adverse effect score (range of positive scores) | |||||
|---|---|---|---|---|---|---|
| Clinical sign | Day 14 | Day 45 | Day 90 | Day 14 | Day 45 | Day 90 |
| Polydipsia | ||||||
| Prednisone | 10/11† | 11/11‡ | 6/9 | 3‡ (2–3) | 3‡ (1–3) | 1 (1–3) |
| Cyclosporine | 3/8 | 3/8 | 2/5 | 0 (1–2) | 0 (1–2) | 0 (1) |
| Polyuria | ||||||
| Prednisone | 10/11† | 10/11† | 5/9 | 3‡ (2–3) | 2† (1–3) | 1 (1–3) |
| Cyclosporine | 1/8 | 3/8 | 0/5 | 1 (0) | 0 (1–2) | 0 (0) |
| Panting | ||||||
| Prednisone | 6/11 | 9/11 | 4/9 | 1 (1–3) | 1 (1–3) | 0 (1–2) |
| Cyclosporine | 2/8 | 3/8 | 2/5 | 1 (1) | 1 (1–2) | 0 (1) |
| Polyphagia | ||||||
| Prednisone | 10/11† | 11/11‡ | 8/9 | 2† (1–3) | 2‡ (2–3) | I.5 (1–3) |
| Cyclosporine | 3/8 | 3/8 | 2/5 | 0 (1–2) | 0 (1–3) | 0 (1–2) |
| Vomiting | ||||||
| Prednisone | 0/11 | 0/11 | 1/9 | 0 (0) | 0 (0) | 0 (1) |
| Cyclosporine | 2/8 | 1/8 | 1/5 | 0 (1) | 0 (I) | 0 (1) |
| Diarrhea | ||||||
| Prednisone | 2/11 | 0/11 | 1/9 | 9 (2–3) | 0 (0) | 0 (1) |
| Cyclosporine | 2/8 | 1/8 | 1/5 | 0 (1–2) | 0 (3) | 0 (1) |
Scores were assigned by owners on a 4-point scale (0 = none, 1 = mild, 2 = moderate, and 3 = severe).
One dog treated with cyclosporine was moved to the prednisone group in the first week of the study because of diarrhea on day 5, so data pertaining to that dog are not included in the data reported here for the cyclosporine group.
Value differs significantly (P < 0.05) from the cyclosporine value at the same assessment point.
Value differs significantly (P < 0.005) from the cyclosporine value at the same assessment point.
Infections were identified during the study period in 2 dogs in the cyclosporine group and in no dogs in the prednisone group (P = 0.47). One cyclosporine-treated dog developed generalized demodicosis on day 18 that was effectively treated by oral administration of ivermectin. This dog had the lowest trough blood cyclosporine concentration of all dogs in the cyclosporine group. Erysipelothrix rhusiopathiae bacteremia was diagnosed in the other cyclosporine-treated dog by means of microbial culture of a blood sample on day 75. Cyclosporine treatment was immediately discontinued for that dog, and bacteremia resolved with antimicrobial administration. On day 7, the trough blood cyclosporine concentration for that dog was 252 ng/mL
A significant decrease from baseline in mean joint neutrophilic inflammation score was identified by day 90 of the study for both treatment groups. No significant difference between groups was identified in the change from baseline in mean inflammation score over time or between groups for mean and maximum joint neutrophilic inflammation scores on day 14 (P ≥ 0.81), day 45 (P ≥ 0.49), or day 90 (P ≥ 0.85; Figure 3).
Mean of the mean inflammation (A), neutrophilic inflammation (B), and maximum neutrophilic inflammation (C) scores for cytologic assessment of synovial fluid samples collected from a minimum of 3 of 4 joints of the dogs in Figure 2. Scores were assigned on a 5-point scale (0 = no inflammation, 1 = mononuclear reactivity only, 2 = mild inflammation, 3 = moderate inflammation, and 4 = marked inflammation). No significant (ie, P ≥ 0.05) differences were identified between treatment groups. See Figure 2 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 248, 4; 10.2460/javma.248.4.395
Mean of the mean inflammation (A), neutrophilic inflammation (B), and maximum neutrophilic inflammation (C) scores for cytologic assessment of synovial fluid samples collected from a minimum of 3 of 4 joints of the dogs in Figure 2. Scores were assigned on a 5-point scale (0 = no inflammation, 1 = mononuclear reactivity only, 2 = mild inflammation, 3 = moderate inflammation, and 4 = marked inflammation). No significant (ie, P ≥ 0.05) differences were identified between treatment groups. See Figure 2 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 248, 4; 10.2460/javma.248.4.395
Mean of the mean inflammation (A), neutrophilic inflammation (B), and maximum neutrophilic inflammation (C) scores for cytologic assessment of synovial fluid samples collected from a minimum of 3 of 4 joints of the dogs in Figure 2. Scores were assigned on a 5-point scale (0 = no inflammation, 1 = mononuclear reactivity only, 2 = mild inflammation, 3 = moderate inflammation, and 4 = marked inflammation). No significant (ie, P ≥ 0.05) differences were identified between treatment groups. See Figure 2 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 248, 4; 10.2460/javma.248.4.395
Correlations between clinical variables and synovial fluid composition
At day 0, owner-perceived mobility score was significantly and moderately correlated with mean joint neutrophilic inflammation score (P = 0.01; ρ = 0.55) and mean joint inflammation score (P = 0.04; ρ = 0.45). Overall locomotion score was also moderately correlated with maximum joint neutrophilic inflammation score at day 0 (P = 0.01; ρ = 0.58). At day 14, owner-perceived mobility score was correlated with mean joint neutrophilic inflammation score (P = 0.05; ρ = 0.46). No other correlations were detected among owner-perceived mobility, clinician-assessed locomotion, and cytologic scores.
Discussion
Results of the present study indicated that oral administration of cyclosporine may be a suitable alternative to oral administration of prednisone for treatment of primary IMPA in dogs, given the similar clinical and cytologic response rates of the 2 treatment groups over a 90-day period. Alternatives to prednisone for immunosuppressant treatment are needed for dogs that are highly susceptible to adverse effects of glucocorticoids, such as marked polyuria, polydipsia, polyphagia, muscle weakness, weight gain, lethargy, and gastrointestinal ulceration.20 This is often the situation for large-breed dogs, which can be predisposed to primary IMPA.2,3,15
To the authors' knowledge, only a few studies have been conducted to evaluate outcomes of treatment for primary IMPA in dogs. In a retrospective study15 involving 39 dogs with primary IMPA from 3 referral hospitals, 56% of dogs had a complete response to treatment, as defined by clinical and cytologic variables.15 However, the frequency with which repeated arthrocentesis had been performed was not reported. In that study, 26 dogs treated with prednisolone alone achieved a complete response, and treatment was tapered and discontinued for 13 of those dogs, without subsequent disease relapse during the follow-up period. Six dogs failed to respond to prednisone, and additional medications were provided, including levamisole, cyclophosphamide, azathioprine, and cyclosporine. Three dogs in that study15 were treated with cyclosporine alone. In contrast to the results of the present study, none of those dogs responded to cyclosporine within a mean treatment duration of 6 weeks. However, the dosage of cyclosporine administered was 5 mg/kg, PO, once a day, which represents half the dosage used in the study reported here. No significant differences were identified between results of immunosuppressive treatment protocols in the other study.15
Another retrospective study16 involved evaluation of the effect of leflunomide treatment in 14 dogs with IMPA. Complete clinical resolution was achieved for 8 dogs, 5 had partial clinical resolution, and 1 had minimal improvement. Dose and duration of leflunomide treatment differed among dogs in that study,16 and response to treatment was judged on the basis of clinical response alone and not the results of synovial fluid assessment. A more recent study24 revealed major differences in the pharmacokinetics of leflunomide in dogs, compared with the pharmacokinetics in humans, and results suggested that leflunomide administration protocols currently used for dogs may require reevaluation to achieve an optimal degree of immunosuppression.
The clinical scoring system used in the present study was a metric that has not been validated for primary IMPA; however, it was used because no other clinical scoring system has been validated for primary IMPA in dogs. In addition, scoring systems validated for lameness, such as the CBPI,25,26 do not account for systemic signs (eg, lethargy and inappetence) or physical examination findings (eg, multiple swollen joints and fever) associated with primary IMPA. The system that we used provided subjective information, and involvement of multiple clinicians in the study may have led to slight variation in subjective score assignment for physical examination findings. In addition, clinicians could not be blinded to treatment group because administration protocols for prednisone and cyclosporine differ (ie, the dosage of prednisone is typically tapered early in the course of treatment, whereas the dosage of cyclosporine is maintained throughout). We attempted to minimize differences in interpretation of response to treatment through use of the scoring system as well as by assignment of the same clinician to each dog for assessment over time (as was also the situation for owner scoring).
The CBPI has been used in combination with accelerometry to assess pain in dogs with IMPA.27 A combination of accelerometry and CBPI scoring, together with modifications that address systemic effects of primary IMPA, may be useful to provide standardized subjective and objective information on pain associated with primary IMPA in dogs in the future. Force-plate gait analysis could also be considered a method by which objective measurements might be obtained, but such data may be unobtainable from systemically ill dogs that are unable to stand or severely lame.
In contrast to the scoring system used by owners and clinicians in the present study, the scoring system used for cytologic analysis of synovial fluid samples was objective and the clinical pathologist that performed the scoring was blinded to treatment group. The value of serial cytologic assessment of synovial fluid samples for monitoring primary IMPA is unclear, but such assessment can be costly and time-consuming to perform given that dogs must be sedated for the arthrocentesis collection procedure. To the authors' knowledge, no studies have been conducted to examine the correlation between mobility and cytologic findings in dogs with primary IMPA. However, correlations have been identified between plasma C-reactive protein concentrations and median CBPI scores, C-reactive protein concentrations and joint cellularity, and C-reactive protein concentrations and mobility as measured by accelerometry.27
In the study reported here, a correlation was identified between owner-perceived overall mobility scores and cytologic findings for synovial fluid on day 14, but clinician-assessed overall locomotion scores did not correlate with cytologic values during treatment. These data suggested that, after day 14 of treatment for primary IMPA, the only way to determine whether evidence of inflammation in synovial fluid no longer exists is to perform arthrocentesis or possibly to measure plasma C-reactive protein concentrations.27 Additional studies are needed to determine whether persistent evidence of inflammation in synovial fluid samples despite resolution of clinical signs requires continuation of treatment to prevent disease relapse or the development of chronic sequelae of inflammation such as osteoarthritis.
As expected, the prevalence of polyuria, polydipsia, and polyphagia was significantly greater in prednisone-treated dogs than in cyclosporine-treated dogs on days 14 and 45 in the study reported here. A significant difference between groups was not maintained on day 90, likely because of the lower number of dogs eligible for adverse-event scoring as well as the lower dosages of prednisone used at that point. We chose not to score clinical signs that could not be readily distinguished from nonspecific primary IMPA-related clinical signs as adverse effects (eg, weakness, lethargy, or inappetence).
Diarrhea was the most prevalent adverse effect in the cyclosporine group and was identified for 3 of 8 dogs for which scores were available on day 14. In addition, treatment for a fourth dog was changed from cyclosporine to prednisone in the first week of the study because it developed diarrhea. Although the prevalence of opportunistic infections did not differ significantly between treatment groups, the finding that 2 dogs in the cyclosporine-treated group and none of the dogs in the prednisone-treated group developed an infection during the study period was cause for concern. Additional studies are needed to assess whether treatment of dogs with primary IMPA with cyclosporine is associated with a greater risk of opportunistic infections than is treatment with other immunosuppressive medications.
The cost of cyclosporine is considerably higher than that of prednisone, which may also influence choice of drug. In the present study, mean and median cost of 30 days of treatment with cyclosporine was $215.16 and $186.00, respectively (range, $155.40 to $341.40), with the added cost of an assay for measurement of trough whole blood cyclosporine concentration ($98.50). In comparison, mean and median cost for 30 days of treatment with prednisone was $23.94 and $23.40, respectively (range, $10.80 to $43.20).
Interindividual variability exists in the pharmacokinetics of cyclosporine in dogs, and measurement of trough whole blood concentrations has been recommended to achieve optimal immunosuppression.21,28 Trough concentrations that range from 100 to 750 ng/mL have been suggested for treatment of various disorders that require immunosuppression.21,28 However, the ideal whole blood cyclosporine concentration for immunosuppression has not been determined for dogs nor has it been determined whether peak, as opposed to trough, concentrations are preferable for therapeutic monitoring. In the study reported here, satisfactory clinical improvement was achieved for 3 of 4 dogs that had a trough whole blood cyclosporine concentration on day 7 < 200 ng/mL. This finding suggested that trough concentrations may correlate poorly with response to treatment or, alternatively, that those dogs improved clinically by day 14 because of spontaneous disease resolution (ie, despite immunosuppressive treatment). In humans, peak whole blood cyclosporine concentrations at 2 hours after dose administration are now preferred to trough cyclosporine concentrations for therapeutic monitoring of transplant recipients.21 Therapeutic monitoring may be most useful for dogs that fail to respond to treatment and to prevent opportunistic infections associated with excessive immunosuppression.
The primary limitation of the study reported here was the small number of dogs enrolled in each group, which reduced the statistical power. Dog numbers were restricted because of limited study funding, and it was difficult to enroll a large number of dogs in a prospective study over a reasonable period given the strict study inclusion criteria. To confirm whether a difference exists in response between treatments, a large clinical trial would be needed, which would require multi-institutional collaboration over a prolonged period. An additional limitation was that, for financial reasons, not all dogs in this study had diagnostic testing performed to exclude the possibility of erosive polyarthritis, septic polyarthritis, or uncommon secondary causes of IMPA such as endocarditis. Such testing was not routinely performed because of the low prevalence of these diseases in our hospital population and lack of historical, physical examination, or laboratory findings that might suggest the enrolled dogs had such diseases. None of the enrolled dogs had evidence from thorough diagnostic testing that an infectious disease might have been contributing to their IMPA, although reactive polyarthritis can never be completely ruled out even with extensive diagnostic testing. Medications other than the study drugs (eg, antimicrobials and antacids) that were being administered to some dogs at the time of enrollment were not uniform because of patient- and clinician-related factors and may have influenced clinical scores. However, that situation could also be expected in clinical practice.
Results of the study reported here suggested that dogs with primary IMPA had similar clinical outcomes regardless of whether they were treated with cyclosporine or prednisone for immunosuppression. Consequently, cyclosporine may be a suitable alternative to prednisone in situations in which the adverse effects of prednisone are considered undesirable. Additional studies are needed to determine whether complete or partial remission rates are similar between these 2 treatments over a longer period than was evaluated in the present study. Larger prospective studies are also warranted to determine whether differences exist between cyclosporine and prednisone treatment in rates of life-threatening opportunistic infections and to identify optimal protocols for monitoring therapeutic drug concentrations and response to treatment.
Acknowledgments
Supported by the Center for Companion Animal Health, University of California-Davis.
Presented in abstract form at the 2013 American College of Veterinary Internal Medicine Forum, Seattle, June 2013.
ABBREVIATIONS
| CBPI | Canine Brief Pain Inventory |
| IMPA | Immune-mediated polyarthritis |
Footnotes
SNAP 4Dx, IDEXX Laboratories, Portland, Me.
Roxane Laboratories, Columbus, Ohio.
Atopica, Novartis Animal Health, Greensboro, NC.
Rimadyl, Pfizer Animal Health, New York, NY.
Amneal Pharmaceuticals, Hauppauge, NY.
Domitor, Orion Corp, Espoo, Finland.
Vedoc Inc, St Joseph, Mo.
Versed, Westward, Eatontown, NJ.
Antisedan, Pharmacia Animal Health, Puurs, Belgium.
Advia 120, Siemens USA, Deerfield, Ill.
StatXact, version 9, Cytel Software Corp, Cambridge, Mass.
References
1. Dunn KJ, Dunn JK. Diagnostic investigations in 101 dogs with pyrexia of unknown origin. J Small Anim Pract 1998; 39: 574–580.
2. Rondeau MP, Walton RM, Bissett S, et al. Suppurative, nonseptic polyarthropathy in dogs. J Vet Intern Med 2005; 19: 654–662.
3. Goldstein RE. Swollen joints and lameness. In: Ettinger SJ, Feldman EC, eds. Textbook of veterinary internal medicine. 7th ed. St Louis: Saunders Elsevier, 2010;130–133.
4. Bennett D. Treatment of the immune-based inflammatory arthropathies of the dog and cat. In: Bonagura JD, ed. Kirk's current veterinary therapy XII. Philadelphia: WB Saunders Co, 1995;1188–1195.
5. Giger U, Werner LL, Millichamp NJ, et al. Sulfadiazine-induced allergy in six Doberman Pinschers. J Am Vet Med Assoc 1985; 186: 479–484.
6. Bennett D. Immune-mediated and infective arthritis. In: Ettinger SJ, Feldman EC, eds. Textbook of veterinary internal medicine. 7th ed. St Louis: Saunders Elsevier, 2010;743–749.
7. Appel MJ, Allan S, Jacobson RH, et al. Experimental Lyme disease in dogs produces arthritis and persistent infection. J Infect Dis 1993; 167: 651–664.
8. Kornblatt AN, Urband PH, Steere AC. Arthritis caused by Borrelia burgdorferi in dogs. J Am Vet Med Assoc 1985; 186: 960–964.
9. Foley J, Drazenovich N, Leutenegger CM, et al. Association between polyarthritis and thrombocytopenia and increased prevalence of vectorborne pathogens in Californian dogs. Vet Rec 2007; 160: 159–162.
10. Goodman RA, Hawkins EC, Olby NJ, et al. Molecular identification of Ehrlichia ewingii infection in dogs: 15 cases (1997–2001). J Am Vet Med Assoc 2003; 222: 1102–1107.
11. Santos M, Marcos R, Assuncao M, et al. Polyarthritis associated with visceral leishmaniasis in a juvenile dog. Vet Parasitol 2006; 141: 340–344.
12. Goodman RA, Breitschwerdt EB. Clinicopathologic findings in dogs seroreactive to Bartonella henselae antigens. Am J Vet Res 2005; 66: 2060–2064.
13. Henn JB, Liu CH, Kasten RW, et al. Seroprevalence of antibodies against Bartonella species and evaluation of risk factors and clinical signs associated with seropositivity in dogs. Am J Vet Res 2005; 66: 688–694.
14. Sykes JE, Kittleson MD, Pesavento PA, et al. Evaluation of the relationship between causative organisms and clinical characteristics of infective endocarditis in dogs: 71 cases (1992–2005). J Am Vet Med Assoc 2006; 228: 1723–1734.
15. Clements DN, Gear RN, Tattersall J, et al. Type I immune-mediated polyarthritis in dogs: 39 cases (1997–2002). J Am Vet Med Assoc 2004; 224: 1323–1327.
16. Colopy SA, Baker TA, Muir P. Efficacy of leflunomide for treatment of immune-mediated polyarthritis in dogs: 14 cases (2006–2008). J Am Vet Med Assoc 2010; 236: 312–318.
17. Johnson KC, Mackin A. Canine immune-mediated polyarthritis: part 2: diagnosis and treatment. J Am Anim Hosp Assoc 2012; 48: 71–82.
18. Miller E. Immunosuppressive therapy in the treatment of immune-mediated disease. J Vet Intern Med 1992; 6: 206–213.
19. Johnson KC, Mackin A. Canine immune-mediated polyarthritis: part 1: pathophysiology. J Am Anim Hosp Assoc 2012; 48: 12–17.
20. Cohn LA. Glucocorticoid therapy. In: Ettinger SJ, Feldman EC, eds. Textbook of veterinary internal medicine. 7th ed. St Louis: Saunders Elsevier, 2010;602–608.
21. Archer TM, Boothe DM, Langston VC, et al. Oral cyclosporine treatment in dogs: a review of the literature. J Vet Intern Med 2014; 28: 1–20.
22. Palmeiro BS. Cyclosporine in veterinary dermatology. Vet Clin North Am Small Anim Pract 2013; 43: 153–171.
23. Berg RI, Sykes JE, Kass PH, et al. Effect of repeated arthrocentesis on cytologic analysis of synovial fluid in dogs. J Vet Intern Med 2009; 23: 814–817.
24. Singer LM, Cohn LA, Reinero CR, et al. Leflunomide pharmacokinetics after single oral administration to dogs. J Vet Pharmacol Ther 2011; 34: 609–611.
25. Brown DC, Boston RC, Coyne JC, et al. Development and psychometric testing of an instrument designed to measure chronic pain in dogs with osteoarthritis. Am J Vet Res 2007; 68: 631–637.
26. Brown DC, Boston RC, Coyne JC, et al. Ability of the canine brief pain inventory to detect response to treatment in dogs with osteoarthritis. J Am Vet Med Assoc 2008; 233: 1278–1283.
27. Foster JD, Sample S, Kohler R, et al. Serum biomarkers of clinical and cytologic response in dogs with idiopathic immune-mediated polyarthropathy. J Vet Intern Med 2014; 28: 905–911.
28. Viviano KR. Update on immununosuppressive therapies for dogs and cats. Vet Clin North Am Small Anim Pract 2013; 43: 1149–1170.
