Authors: Jared A Jaffey, Andrew S Hanzlicek, Laura H Rayhel, Eric T Hostnik, Chase Irwin, Lauren Chittick
Categories: Original Research, antibody serology, Coccidioides, cough, fungal, Valley Fever
Source: Journal of Veterinary Internal Medicine
Pulmonary coccidioidomycosis (PC) is a common cause of respiratory tract disease in dogs from endemic regions, but evidence-based guidelines for monitoring treatment and determining remission are lacking.
Describe temporal changes in clinical signs, antibody serology, and thoracic radiographs in dogs with PC, and evaluate the performance of a treatment monitoring protocol to determine clinical remission.
Thirty-one client-owned dogs with newly diagnosed PC.
A prospective, cohort study. Dogs were examined at diagnosis (T0) and once every 3 months (T1-T4) until remission or for a maximum of 12 months. Clinical status was assessed using an owner-completed visual analog scale (VAS) representative of the 24 hours and 7 days that preceded the visit.
The 24-hour and 7-day VAS scores increased (ie, clinical signs improved) from T0 (24-hour: mean, 95% confidence interval [CI]; 42.3 mm, 35.0-50.2; 7-day: 39.6 mm, 31.9-47.3) to T1 (24-hour, 85.7 mm, 78.1-93.3, 7-day, 84.4 mm, 76.8-92.1, P < .001) with minimal change thereafter. Immunoglobulin G titers decreased from T0 (median, interquartile range [IQR]; 16, 4-1:32) to T1 (1:2, 1-1:8, P < .001) without significant changes thereafter. Eighty-four percent (26/31) of dogs achieved remission in a mean of 258.9 days (SD, 108.9), which occurred at T1 (13%, 4/31), T2 (23%, 7/31), T3 (32%, 10/31), and T4 (16%, 5/31). Relapse occurred in 15% (4/26) of dogs.
Most dogs with PC are expected to achieve remission within 6-9 months of starting antifungal treatment, and relapse is not uncommon.
Keywords: antibody serology, Coccidioides, cough, fungal, Valley Fever
Coccidioides spp. are soil-dwelling fungi endemic to the southwestern United States, northern Mexico, and parts of Washington State. Transmission occurs predominantly by inhalation of arthroconidia that are aerosolized from soil.^1^ Arthroconidia undergo transformation into spherules within the terminal airways and are either eliminated or cause clinical disease.^1^ As in human patients, it is suspected that most dogs infected with coccidioidomycosis remain subclinical or develop mild, transient clinical signs.^2–4^
Clinical infection in dogs with coccidioidomycosis can be confined to the respiratory tract and regional intrathoracic lymph nodes (primary pulmonary) or disseminate to extrathoracic sites. Subclinical coccidioidomycosis is estimated to occur in approximately 60% of dogs in endemic areas.^2^ Among dogs that develop clinical disease, pulmonary coccidioidomycosis (PC) is suspected to account for approximately 75% of cases, but epidemiologic studies systematically exploring the distribution of disease forms in dogs are lacking.^5^ The most common clinical presentation for dogs with PC consists of an acute onset of clinical signs associated with respiratory tract disease, often in addition to ≥1 non-specific sign of illness.^5–9^
Currently, evidence-based recommendations for treatment monitoring and criteria to determine clinical remission in dogs with PC are lacking. Published recommendations are based on expert opinion and limited to generalized guidelines for typical estimated treatment duration and some proposed remission criteria, such as resolution of clinical signs and radiographic abnormalities, and low immunoglobulin (Ig)G antibody titers.^5,9^ The absence of a structured treatment monitoring protocol and clearly defined criteria for remission may result in unnecessary, prolonged administration of antifungal drugs in some dogs. This can cause an undue burden on dogs and owners through an increased risk of treatment-related complications and increased cost of treatment and monitoring.
Our objectives (1) to describe temporal changes in respiratory tract clinical signs, anti-Coccidioides spp. antibody serology, and thoracic radiographs in dogs with PC from the time of diagnosis until discontinuation of treatment, and (2) to investigate the performance of a treatment monitoring protocol to determine clinical remission. We hypothesized that clinical signs, antibody serology, and thoracic radiographs would improve after the initiation of antifungal treatment, and clinical remission would be achieved within 6 months of treatment in >50% of dogs with PC.
Client-owned dogs with a new diagnosis of PC between October 2020 and February 2021 were eligible for inclusion in our prospective cohort study. Data from dogs in this cohort were used in separate published studies that had unrelated objectives.^10–12^ Dogs were eligible for inclusion if the following criteria were (1) dogs exhibited ≥1 clinical signs associated with respiratory tract disease, such as cough, wheeze, increased respiratory effort, tachypnea, exercise intolerance, cyanosis, or syncope, (2) there was a positive anti-Coccidioides spp. antibody serologic test result, and (3) a minimum of 2-view thoracic radiographs were available for review. Dogs with normal thoracic radiographs at baseline were retained in the study if there was a strong clinical suspicion of PC based on baseline diagnostic test results, clinical history, and physical examination findings, in combination with improvement in clinical status after the initiation of antifungal treatment. Exclusion criteria were clinical findings suggestive of concurrent disseminated disease, presence of any comorbidities that may cause immune dysregulation (eg, hyperadrenocorticism, diabetes mellitus, immune-mediated disorders, neoplasia), the administration of immunosuppressive medications or chemotherapy, or a historical diagnosis of coccidioidomycosis. Prednisone administration after diagnosis was permitted, provided the dosage did not exceed 1.5 mg/kg/day and the duration was < 3 weeks. Dogs also were excluded if they had received antifungal treatment for >7 days before enrollment and if compounded antifungal drugs were utilized. Clinical decisions that were not directly related to the study were made by the attending clinician. The study was conducted in accordance with guidelines for clinical studies and was approved by the Midwestern University Animal Care and Use Committee (protocol # 3024) with written owner consent.
Dogs were presented for evaluation once every 3 months after the initiation of a Food and Drug Administration (FDA)-approved generic formulation of fluconazole until remission was achieved or for a maximum of 12 months (Figure 1). Alternative antifungal drugs were allowed in dogs that failed to respond adequately, as determined by the attending clinician. Time points were designated as baseline (ie, time of diagnosis; T0), 3-month (T1), 6-month (T2), 9-month (T3), and 12-month (T4). The following data were collected at evaluations T1-T4: (i) 3-view thoracic radiographs, (ii) coccidioidal antibody serology (agar gel immunodiffusion [AGID] and enzyme immunoassay [EIA]), and (iii) clinical status.
Figure 1 Flow diagram illustrating the number of dogs with newly diagnosed pulmonary coccidioidomycosis enrolled between October 2020 and February 2021 as well as time from diagnosis and the number of dogs that achieved remission and those with unresolved infection at each time point. Time point designations were baseline (T0), 3-month (T1), 6-month (T2), 9-month (T3), and 12-month (T4). ^ represents 1 dog that was not evaluated at T3 for the scheduled examination because of owner availability. Telephone communication with the owner indicated the dog was still clinical, fluconazole administration was continued, and the dog was evaluated at T4.
Clinical remission was defined as the first evaluation at which all 3 of the following criteria were (1) an anti-Coccidioides IgG titer of ≤ 1:8 with no detection of immunoglobulin (Ig) M antibodies using AGID, or an IgG concentration ≤ 20 EIA units (EU) in dogs with negative AGID results throughout the study period, (2) thoracic radiographs that were either normal or had static, mild abnormalities without tracheobronchial lymphadenopathy (TBL) compared to the previous evaluation; and (3) clinical signs resolved. Antifungal administration was discontinued at the time of remission. Owners were then contacted by research investigators by telephone or electronic mail once every 3 months for a 12-month time period after remission to assess clinical status. Research investigators were available to be contacted by owners to report abnormal clinical signs at any time during the 12-month study follow-up period. Diagnostic investigations were carried out in dogs suspected to have a relapse, and testing included coccidioidal antibody serology, a CBC, serum biochemistry, and urinalysis. Additional diagnostic testing was performed as needed on a case-by-case basis. Relapse was defined as the recurrence of abnormal clinical signs (respiratory or extrapulmonary) in conjunction with a ≥2-fold increase in the AGID IgG titer relative to remission results, supported by a review of available clinicopathologic data and imaging findings, with unrelated disorders excluded to the extent possible. Relapse also was considered possible in dogs without serologic recurrence if sufficient evidence supported this determination on a case-by-case basis (Supplementary Material A). All cases were assessed by a single board-certified small animal internist (JAJ). Long-term follow-up for all dogs, regardless of remission status, was defined as the time from clinical remission or T4 (if remission was not achieved) to April 4, 2025 or the date of death.
One reference laboratory (MiraVista Diagnostics) was used for serologic testing for IgM and IgG against Coccidioides spp. using AGID (IgM and IgG) and EIA (IgG) at all evaluations. A positive AGID IgM or IgG result was defined as detectable antibodies in an undiluted serum sample. If positive, serial dilutions were tested (up to 128) and the highest dilution with detectable IgG antibodies was reported as the final serum titer result. A positive EIA IgG was defined as the results that were ≥10 EU and the upper quantifiable range was 80 EU.
Thoracic radiographs were reviewed using commercial software (HorosTM, Horosproject.org, Nimble Co. LLC) by a single board-certified veterinary radiologist (ETH) who was not blinded to the suspected diagnosis of PC at T0 but was blinded to clinical information at all subsequent hospital visits.
A visual analog scale (VAS) was used to assess owner-perceived severity of clinical signs related to respiratory tract disease, reflecting the 24 hours and 7 days before each evaluation (Supplementary Material B). The VAS used a 100-mm scale to assess the range of observed clinical signs from severe signs (0 mm) to no signs (100 mm). The distance from 0 to the patient score was measured to provide a quantifiable VAS score, with increasing scores associated with less severe signs. Dogs were considered subclinical when both the 24-hour and 7-day VAS scores were 100 mm (ie, complete resolution).
Statistical analyses were performed using proprietary software (SigmaPlot, Systat Software Inc, version 14.5, and SAS, version 9.4). Normality was assessed using the Shapiro-Wilk test. Normally distributed data were presented as mean and SD, whereas data that were not normally distributed were presented as median and interquartile range (IQR), which was expressed as the 25th and 75th percentiles and range when indicated. Categorical data were presented as proportions. The percentage change in VAS scores was calculated using the following example (T1 VAS score – T0 VAS score)/ T0 VAS score × 100. Linear mixed models with individual dogs treated as random effects were used to estimate changes in 24-hour and 7-day VAS scores. Normality and homoscedasticity of residuals were inspected visually. The least-squares mean (LSMean), the corresponding 95% confidence interval (CI), and the sample size were reported for each time point. Custom hypothesis tests were constructed to compare time points.
When the measured IgG antibody was above the upper limit of detection, data were recorded at the upper limit of detection for statistical purposes (ie, AGID at 128 and EIA at 80 EU). Similarly, when the AGID IgG titer was negative, data were recorded as 0 for statistical purposes. The AGID and EIA IgG median and IQR values were reported for each time point. Linear mixed models initially were fitted to analyze AGID and EIA IgG outcomes, but, normality of residuals and homoscedasticity assumptions were violated despite applying logarithmic and square-root transformations. Therefore, the Wilcoxon signed-rank test with the Benjamini and Hochberg’s adjustment for multiple comparisons was used to compare time points. The median and IQR of the difference and the adjusted P-value for each comparison were reported. Time to clinical and radiographic resolution was visualized using Kaplan-Meier curves. We reported the frequency and proportion of dogs that reached resolution, as well as the median time to resolution and its corresponding 95% CI for both outcomes. Four dogs were excluded from the radiographic resolution analysis because they had normal radiographs at baseline and throughout the study. P-values <.05 were considered significant.
Thirty-two dogs were eligible for inclusion, but 1 dog was excluded because of the use of compounded fluconazole, leaving 31 dogs in the study (Figure 1). Median age and mean weight were 4.8 years (IQR, 1.7-8.0) and 22.9 kg (SD, 13.3), respectively. There were 18 purebred dogs and 13 mixed breed dogs. Purebred dogs included Chihuahua (n = 3), Rhodesian ridgeback (n = 3), pit bull terrier (n = 2), and 1 each of the Australian cattle dog, American foxhound, Belgian malinois, English bulldog, English mastiff, German shorthaired pointer, golden retriever, Labrador retriever, miniature poodle, and Queensland heeler. The sex distribution was as male neutered (n = 12; 39%), female spayed (n = 12; 39%), female intact (n = 4; 13%), and male intact (n = 3; 9%).
Dogs were initially treated with an FDA-approved generic formulation of fluconazole. The median PO fluconazole dosage prescribed was 15.9 mg/kg/day (IQR, 13.6-18.6). Eight dogs received prednisone after diagnosis, with a mean PO dosage of 0.7 mg/kg/day (SD, 0.27) for a mean duration of 11.9 days (SD, 3.4). One dog was changed to amphotericin B lipid complex (cumulative dose, 15 mg/kg; Abelcet, Leadiant Biosciences) and itraconazole (dose, 8.8 mg/kg/day; Sporanox, Janssen Pharmaceuticals) after the T1 evaluation because of inadequate clinical improvement and worsening findings on thoracic radiographs. Fluconazole was the only antifungal drug administered to the remaining 30 dogs throughout the study period.
Most dogs exhibited >1 clinical sign (23/31, 74%) at T0, and the median number of signs per dog was 3 (range, 1-6). Cough was the only reported clinical sign in 8 dogs (26%). Overall, cough was the most common clinical sign (30/31, 97%), followed by hyporexia (14/31, 45%), lethargy (14/31, 45%), exercise intolerance (5/31, 16%), tachypnea (4/31, 13%), diarrhea (4/31, 13%), wheeze (3/31, 10%), increased respiratory effort (2/31, 6%), sneeze (2/31, 6%), vomiting (2/31, 6%), and weight loss (1/31, 3%). The median duration of time clinical signs were reported by owners before T0 was 14 days (IQR, 7-37, 3-365 days). Median VAS scores reflective of the 24 hours and 7 days before T0 were 35 mm (IQR, 18-80, 0.0-94 mm) and 35 mm (IQR, 14-65, 0.0-100 mm), respectively. Mean respiratory rate and rectal temperature at the initial presentation were 35.7 breaths/min (SD, 8.5; n = 24) and 103.3°F (SD, 1.1; n = 28), respectively. Rectal temperatures were ≥ 103°F in 64% (18/28) of dogs on presentation.
Twenty-nine percent (9/31) of dogs had positive AGID IgM results. Ninety-four percent (29/31) of dogs had positive AGID IgG titer results. The IgG titers ranged from negative to ≥1:128. The median AGID IgG titer in dogs with positive results was 16 (n = 29). Sixty-five percent (20/31) of dogs had positive EIA IgG results. The mean EIA IgG for the 20 dogs with positive results was 44.1 EU (SD, 21.9). One of the dogs with a negative AGID IgG had a positive IgM and an EIA IgG of 63.3 EU. The other dog with a negative AGID IgG had a negative AGID IgM and an EIA IgG of 25.1 EU.
Pulmonary parenchymal abnormalities were identified on thoracic radiographs performed at T0 in 42% (13/31) of dogs. Most dogs had a solitary pulmonary pattern (77%, 10/13), of which a bronchial pattern was most common (60%, 6/10), followed by an unstructured interstitial pattern (20%, 2/10), and 1 each of a structured interstitial pattern and an alveolar pattern. Twenty-three percent (3/13) of dogs with pulmonary parenchymal abnormalities had concurrent bronchial and unstructured interstitial patterns. One dog had mild pleural effusion, and none had sternal or cranial mediastinal lymphadenopathy.
Tracheobronchial lymphadenopathy (TBL) was identified in 84% (26/31) of dogs at T0, and when present, it was mild (50%, 13/26), moderate (38%, 10/26) or severe (12%, 3/26). A comprehensive investigation of TBL in this cohort has been published elsewhere.^10^ Tracheobronchial lymphadenopathy was identified in 78% (14/18) of the dogs that had no radiographic evidence of pulmonary parenchymal disease at T0.
A significant increase in 24-hour and 7-day VAS scores occurred from T0 to T1 (both P < .001; Figure 2), but no significant changes in VAS scores between successive time points were identified after T1 (Tables S1 and S2). The VAS scores were significantly lower at T0 compared with all other time points (all P < .001; Figure 2). The median percentage change in 24-hour and 7-day VAS scores from T0 to T1 was 165.4% (range, −15.3% to 2350.0%) and 122.2% (range, −10.6 to 1066.7%, n = 28; Table 1), respectively. The percentage change in VAS scores from T0 to T1 was not calculated for 3 dogs because their T0 data were recorded as 0. Minimal deviation occurred in the percentage change in VAS scores between evaluation time points after T1. The median percentage change for VAS scores (24-hour and 7-day) never exceeded 4.2% for the T1-T2, T2-T3, and T3-T4 time intervals (Table 1). Individual dog data for VAS scores are presented in Table S3.
Figure 2 Visual analog scale (VAS) scores (A) 24 hours and (B) 7 days preceding each examination time point. Time point designations were baseline (T0, n = 31), 3-month (T1, n = 31), 6-month (T2, n = 27), 9-month (T3, n = 19), and 12-month (T4, n = 10). Enclosed circle represents the least-squared means, and the bars represent the 95% confidence interval. Light blue lines show individual trajectories. Symbols denote statistically significant differences from immediately preceding time point based on model P < .05 (^^), P < .01 (^^), P < .001 (^^).
Overall, 90% (28/31) of dogs had complete resolution of respiratory tract signs during the 12-month study period, and resolution was first recorded at T1 in 19% (6/31), T2 in 29% (9/31), T3 in 32% (10/31), and T4 in 10% (3/31). The median time to complete resolution of clinical signs associated with respiratory tract disease was 274 days (95% CI, 195-299 days; Figure 3).
Figure 3 Kaplan–Meier curves visualizing probability of clinical and radiographic resolution. Clinical resolution is visualized in red; radiographic resolution is visualized in blue. 28 (90%) out of 31 dogs reached clinical resolution with a median time to resolution of 274 days (95% CI, 195-299 days). 23 (85%) out of 27 dogs reached radiographic resolution with a median time to resolution of 187 days (95% CI, 104-294 days).
The AGID IgG titers were significantly higher at T0 compared with all other time points (Figure 4). No significant change in the AGID IgG titer was found between successive time points after T1 (Table S4). Sixty-five percent (20/31) of dogs experienced a ≥2-dilution decrease (ie, improvement) in AGID IgG from T0 to T1, whereas 19% (6/31) had mild increases (1 dilution, n = 4; 2 dilutions, n = 1; and 3 dilutions, n = 1). The dog with the 3-dilution increase had a negative AGID result at T0 and a titer of 4 at T1.
Figure 4 Agar gel immunodiffusion immunoglobulin G inverse titer results for each examination time point. Time point designations were baseline (T0, n = 31), 3-month (T1, n = 31), 6-month (T2, n = 27), 9-month (T3, n = 19), and 12-month (T4, n = 10). Y-axis values are visualized in log-scale. Enclosed circles represent the median and bars represent the interquartile range. Light blue lines show individual trajectories. Symbols denote statistically significant differences from immediately preceding time point based on Wilcoxon signed-rank test with the Benjamini and Hochberg’s adjustment for multiple P < .05 (^^), P < .01 (^^), P < .001 (^^).
Next, we summarized changes in AGID IgG titers over the entire study period. Fifty-two percent (16/31) of dogs had a ≥1-fold dilution decrease (ie, improvement) in AGID IgG titer at T1 that either continued to decrease or remained static through the remainder of the study period. The AGID IgG titer results fluctuated between examination visits in the remaining 48% (15/31) of dogs.
No significant change occurred in IgG EU between time points (all P > .05; Table S5). Likewise, no significant difference was found in IgG EU at T0 compared with all other time points (Figure S1). Forty-five percent (5/11) of dogs that did not have IgG antibodies detected at T0 using EIA had positive detection at T1 with 4 dogs having positive results (ie, ≥ 10 EU) and 1 dog having an intermediate result (8.8 EU). The remaining 6 dogs failed to have IgG antibodies identified using EIA at any point in the study. Information related to IgM results during treatment can be found in Supplementary Material C.
Thoracic radiographic abnormalities resolved in 48% (13/27) or improved in 44% (12/27) of dogs by T1. Two dogs (8%) had worsened pulmonary parenchymal abnormalities on radiographs performed at T1 that subsequently improved at T2 and were resolved by T3. Two different dogs had worsened radiographs at T2 compared with T1, after initially improving from T0 to T1. Temporal changes in radiographic TBL for this cohort have been published previously.^10^ Overall, 85% (23/27) of dogs had complete resolution of radiographic abnormalities during the 12-month study period, and resolution was first recorded at T1 in 44% (12/27), T2 in 19% (5/27), and T3 in 22% (6/27). Three dogs maintained static mild pulmonary parenchymal abnormalities (bronchial, n = 1; unstructured interstitial, n = 1; and small focal alveolar, n = 1) before achieving clinical remission. Thoracic radiographic abnormalities continued to improve beyond T4 in 1 dog. Individual dog data for temporal changes in thoracic radiographs can be found in Table S3. The median time to resolution of radiographic abnormalities was 187 days (95% CI, 104-294 days; Figure 3).
Eighty-four percent (26/31) of dogs achieved remission in a mean of 258.9 days (SD, 108.9). Remission was achieved at T1 in 13% (4/31), T2 in 23% (7/31), T3 in 32% (10/31), and T4 in 16% (5/31; Figure 1). Thirty-four percent (9/26) of dogs had negative AGID IgG titers at the time of remission with the remaining distribution of IgG titer results being 1 (23%, 6/26), 2 (19%, 5/26), 4 (12%, 3/26), and 8 (12%, 3/26). Descriptive details explaining the reasons remission was not achieved can be found in Supplementary Material D. Individual dog data regarding clinical remission can be found in Table S3.
Most dogs (85%, 22/26) that achieved remission did not relapse in the 12-month follow-up period. However, 4 dogs (15%) relapsed at least once with the first occurrence at a median of 113.5 days (IQR, range; 87.3-163.8 days, 85-174 days) after remission. These 4 dogs originally attained remission at T1 (n = 1), T2 (n = 1), and T3 (n = 2). The suspected sites of relapse infection in 3 dogs were 1 each of the skin, central nervous system, and bone. One dog had uncharacterized disease with the acute onset of non-specific pain. Antifungal treatment was restarted, and all dogs had complete resolution of clinical signs. Details related to relapsed infection can be found in Supplementary Material E.
Long-term follow-up information was available for all dogs. The median follow-up was 1316 days (IQR, range; 1206-1393 days, 258-1546 days). Fluconazole was discontinued shortly after T4 in 60% (3/5) of the dogs that did not attain remission during the study period. These included the 2 dogs that remained clinical and the 1 dog that had positive detection of IgM antibodies at T4. None of these dogs exhibited clinical signs concerning for a relapse at a median of 1183 days (range, 652-1240 days) after the discontinuation of fluconazole after T4. The remaining 2 dogs continued fluconazole treatment after T4 and eventually achieved remission 1090 days and 427 days after T4, respectively. The dog that achieved remission 1090 days after T4 remained subclinical for 121 days after fluconazole discontinuation. Fluconazole administration was restarted by the primary care veterinarian in the other dog 106 days after remission, based solely on antibody serology results (IgM, negative; IgG, 4).
Except for the 4 dogs that relapsed in the initial 12-month follow-up period, none of the remaining 22 dogs that achieved remission went on to relapse during long-term follow-up. Seventy-five percent (3/4) of the dogs that relapsed went on to relapse at least 1 more time (range, 2-3 relapses). Data related to additional relapses in these 3 dogs can be found in Supplementary Material F.
Nineteen percent (6/31) of dogs died or were euthanized at a median of 730 days (IQR, range; 441-1095.8 days, 258-1206 days) after remission or T4. Two dogs died at home unexpectedly without a known cause. Two dogs were euthanized because of multicentric lymphoma and complications associated with a large splenic mass, respectively. One dog was euthanized because of the development of clinical signs associated with respiratory tract disease after initially achieving remission at T3. One dog was euthanized because of refractory seizures that were suspected to be related to coccidioidomycosis. Necropsy examinations were not performed in any dog.
Pulmonary coccidioidomycosis remains one of the most common infectious respiratory tract disorders of dogs in the southwestern United States, but limited information exists related to structured plans for assessing treatment response, criteria to determine remission, and the timeline for expected clinical response. Therefore, our primary objectives were to critically evaluate a treatment monitoring regimen and criteria to identify clinical remission in dogs with PC. Substantial improvement in clinical signs associated with respiratory tract disease and IgG titers occurred within the first 3 months (ie, T0-T1) of antifungal treatment, with minimal change thereafter. Nearly the entire cohort had improved or resolved thoracic radiographic changes by T1. Most dogs (84%, 26/31) achieved remission within the study period, which occurred in two-thirds of dogs by T3. Relapse occurred in 4 dogs (15%) within the 12-month and long-term follow-up periods.
Clinical signs associated with respiratory tract disease, as depicted by VAS scores, significantly improved from T0 to T1, with minimal change thereafter. The data also were analyzed on the basis of percentage change, and results were similar to the mixed linear effects model with a rapid and substantial improvement in the severity of respiratory tract signs that proved to be equivocal at successive time intervals. Specifically, the median percentage change in 24-hour and 7-day VAS scores from T0 to T1 was 165.4% and 122.2% respectively, with the median percentage change never exceeding 4.2% after T1. One study highlighted that the prevalence of clinical signs in a cohort of patients with mild to moderate PC rapidly decreased, with peak improvement identified within 12 weeks of antifungal treatment.^13^ The same study also reported that the median time to a 50% reduction in symptom score was 9.9 weeks and the time to complete symptom resolution was 18.7 weeks.^13^ Complete clinical resolution occurred in 90% of the dogs in our study and was identified in 48% of dogs by T2 and 80% by T3.
Temporally accelerated clinical improvement rates early in the course of treatment, with minimal proportional change after T1, were an unexpected observation in our study. The specific reason for this trend is unknown, and it is likely multifactorial. One possible explanation is that 3 months of PO fluconazole administration is sufficient to abrogate a large proportion of spherule propagation, with residual replication effectively neutralized thereafter in most dogs with PC. Repeated cycles of spherule growth and rupture trigger a marked host inflammatory response that ultimately manifests as clinical signs associated with systemic illness and the destruction of contiguous tissues.^14,15^ Conversely, dormant spherules do not elicit the same inflammatory response.^14,15^ This hypothesis was illustrated in a study that utilized this same cohort of 31 dogs to investigate longitudinal changes in positive acute phase proteins, such as C-reactive protein (CRP) and haptoglobin, from the time of diagnosis to remission or 12 months.^11^ Serum concentrations of CRP and haptoglobin significantly decreased from T0 to T1, whereas only mild, non-significant changes occurred thereafter.^11^
Collectively, our data indicate that attending clinicians can expect the most substantial clinical improvement in dogs with newly diagnosed PC within 3 months of antifungal treatment. This observation is important because it provides clinicians with reasonable expectations for clinical responses that—in conjunction with antibody serology and thoracic radiographs—can be used to determine whether an adjustment in antifungal treatment is warranted. For example, in our study, most dogs had 24-hour or 7-day VAS scores that were >70 mm at T1. Only 1 dog had both VAS scores < 50 mm at T1, and that dog transitioned from fluconazole to amphotericin B lipid complex and long-term treatment with itraconazole. Clinical signs nearly completely resolved in this dog after 6 months of appropriate antifungal treatment with 24-hour and 7-day VAS scores of 96 mm and 97 mm, respectively.
Serum IgG titers using AGID significantly decreased in the first 3 months of antifungal treatment without additional significant changes after T1. This finding aligns with our hypothesis, and similar trends are found in people with uncomplicated PC.^16^ One study found that the median serologic improvement rate of IgG titers in people with uncomplicated PC was 91 days/dilution, and a more accelerated rate of improvement occurred immediately after the serologic maxima, early in the course of antifungal treatment.^16^ It is common practice to review coccidioidal antibody serology during treatment, because these titers usually decrease over time with patient improvement. However, there are several inherent limitations of utilizing antibody serology that should preclude it being used as the sole determinant of treatment response in dogs with PC. For example, 1 dog had improvement in the AGID IgG titer (T0: 16, T1: 8) while thoracic radiographic abnormalities worsened. Continuing fluconazole solely on the basis of improved serology would likely have been inappropriate, because marked and rapid clinical and radiographic improvement occurred after the transition to amphotericin B and itraconazole. Three months later (T2), despite the near resolution of clinical signs (VAS 24-hr: 96 mm; VAS 7-day: 97 mm) and substantially improved thoracic radiographs, the AGID IgG increased to 64. Treatment was maintained, and the dog subsequently achieved remission at the next evaluation (T3). In humans with uncomplicated PC, antibody serology is used alongside clinical response and physical examination findings to guide treatment decisions, but it is rarely used as the sole factor determining whether to change, discontinue, or re-initiate antifungal treatment.^17^
One limitation of antibody serology is a lack of information related to serologic kinetics in dogs with PC. One study in people with coccidioidomycosis found that maximal IgG titers developed a mean of 31 days after initial serologic positivity, and 10% of patients had maximal titers > 50 days after diagnosis.^16^ Therefore, without knowing when a dog was initially infected, it is reasonable to suspect that clinical improvement can occur alongside an increasing IgG titer at T1. Indeed, 6 dogs (19%) in our study had higher IgG titers at T1 compared with T0 despite clinical improvement. These results, if overinterpreted by clinicians, may have prompted an erroneous adjustment to antifungal treatment. Another limitation is that many dogs and people^16^ with PC maintain persistently positive IgG titers, despite complete clinical resolution, for many years after the discontinuation of antifungal treatment. Serologic resolution of IgG titers occurred in 56.2% of people with uncomplicated PC in 1 study.^16^ Likewise, 57% (15/26) of dogs in our study had IgG titers <1:2 at the end of the study period. Therefore, it is important to recognize that IgG titers do not need to be negative to safely discontinue antifungal treatment in dogs with PC. In fact, results from our study emphasize that antifungal treatment can be safely discontinued in dogs with PC that have IgG titers as high as 8, as long as the other criteria for remission are met. Lastly, antibody serology results can fluctuate during treatment for reasons independent of clinical status, as was observed in our study.^16,17^
Eighty-four percent (26/31) of dogs achieved remission in a mean of 258.9 days, which translated to remission at T1 in 13% (4/31), T2 in 23% (7/31), T3 in 32% (10/31), and T4 in 16% (5/31). The optimal duration of treatment and criteria to determine remission in people with mild to moderate PC have not been defined. The most recent practice guidelines from the Infectious Disease Society of America simply advise that PO azole treatment be administered for 3-6 months or longer, depending on clinical response.^18^ The guidelines outline that treatment can be discontinued when the patient’s symptoms and inflammatory markers have resolved, and serology and radiographs have stabilized.^18^ One expert in coccidioidomycosis in humans suggests that the decision to discontinue treatment be made on a case-by-case basis, but it would be uncommon for patients with otherwise uncomplicated PC to require > 6 months of antifungal treatment.^17^ Collectively, criteria for the cessation of antifungal treatment in people with uncomplicated PC are more flexible than those used in our study. Specifically, we applied a strict IgG titer cutoff to aid in defining remission, an approach that is not used in human patients. If overly stringent, our remission criteria may have resulted in unnecessarily prolonged treatment in some dogs. Additional studies using criteria such as those used in humans with PC, along with CRP concentrations and long-term follow-up, are needed to optimize treatment monitoring and the duration of treatment in dogs with PC.
Relapse occurred in 15% (4/26) of dogs that achieved remission and was first identified a median of 113.5 days (range, 85-174 days) after remission. Three of these dogs went on to relapse ≥1 more time. No other dog that attained remission relapsed in the 12-month or long-term follow-up period (median, 1316 days). One study in people with uncomplicated PC treated with fluconazole reported a 15% (8/54) relapse rate, which aligns with our results.^19^ Our results suggest that a relapse of infection in dogs with PC is not uncommon and, when it does occur, it typically happens within 6 months after discontinuing antifungal treatment. Therefore, a reasonable post-treatment monitoring schedule would include at least 2 visits at 3-month intervals, each consisting of a physical examination, antibody serology, and any additional diagnostic testing as indicated.
Our study had some limitations. Dogs were diagnosed with PC based on the interpretation of clinical criteria and treatment response by a single board-certified small animal internist with extensive experience managing coccidioidomycosis. We utilized a VAS to provide a semiquantitative assessment of respiratory tract clinical signs. Although this survey was adequate to determine when clinical signs were completely resolved, a more comprehensive questionnaire including a variety of individual clinical signs may have provided a more thorough understanding of longitudinal changes in clinical status. Recent studies have shown that not all FDA-approved generic formulations of fluconazole are bioequivalent in dogs, despite being considered bioequivalent in people.^20,21^ In our study, dogs did not necessarily receive fluconazole from the same manufacturer, which may have introduced variability in treatment response. Prednisone was administered at anti-inflammatory doses to 8 dogs for short periods to temporarily control clinical signs in moderately to severely affected cases. Although our study was not designed to evaluate the impact of prednisone on treatment response or the risk of dissemination, this approach remains an important area for future investigation. Notably, prolonged prednisone administration, regardless of dose, is contraindicated in patients with coccidioidomycosis. Ours was a single-arm study without a control group. Inclusion of an untreated control cohort was not feasible, because withholding antifungal treatment from dogs with PC would be unethical. Positivity for IgM is often thought to be associated with active replication of Coccidioides spp. and can aid in the diagnosis or detection of relapse in clinically ill patients. Its utility for monitoring treatment response, however, remains unclear, particularly in patients showing clinical improvement. This limitation was reflected in our study, in which several dogs exhibited intermittent IgM positivity despite sustained clinical and radiographic resolution, and this was the sole reason for a lack of remission in 1 dog. The specific reasons for intermittent IgM positivity are unknown but speculated to be related to nonspecific background reactivity or laboratory variability. Accordingly, requiring IgM negativity at the time of presumptive clinical remission may have been overly conservative and merits further evaluation. Lastly, the study design lacked clearly defined criteria for when and how antifungal treatment should be adjusted. Had such criteria been implemented, some dogs might have achieved remission sooner, and others that failed to achieve remission may have done so.
Our study provided data on the temporal patterns of respiratory tract clinical signs, thoracic radiographic findings, and antibody serology in dogs with PC that received antifungal treatment. Substantial improvement in clinical signs, thoracic radiographic abnormalities, and antibody serology typically occurred within the first 3 months of antifungal treatment, followed by more gradual progress. Based on the criteria used in our study, most dogs with PC are expected to achieve remission within 12 months of starting treatment, most commonly between 6 and 9 months. Lastly, a relapse of infection after treatment discontinuation is not uncommon and, when it does occur, it typically arises within 6 months.
The authors thank Sasha Willis for technical assistance.
Jared A Jaffey, Department of Specialty Medicine, Midwestern University, College of Veterinary Medicine, Glendale, AZ 85308, United States.
Andrew S Hanzlicek, MiraVista Diagnostics, Indianapolis, IN 46241, United States.
Laura H Rayhel, Department of Specialty Medicine, Midwestern University, College of Veterinary Medicine, Glendale, AZ 85308, United States.
Eric T Hostnik, Department of Veterinary Clinical Sciences, Ohio State University, Veterinary Clinical Sciences, Columbus, OH 43210, United States.
Chase Irwin, Office of Research and Sponsored Programs, Midwestern University, Glendale, AZ 85308, United States.
Lauren Chittick, Department of Specialty Medicine, Midwestern University, College of Veterinary Medicine, Glendale, AZ 85308, United States.
Jared A. Jaffey (Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Writing—original draft, Writing—review & editing), Andrew S. Hanzlicek (Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Writing—original draft, Writing—review & editing), Laura H. Rayhel (Conceptualization, Data curation, Formal analysis, Investigation, Resources, Writing—original draft, Writing—review & editing), Eric T. Hostnik (Conceptualization, Data curation, Formal analysis, Investigation, Resources, Writing—original draft, Writing—review & editing), Chase Irwin (Data curation, Formal analysis, Investigation, Methodology, Writing—original draft, Writing—review & editing), and Lauren Chittick (Data curation, Investigation, Software, Writing—original draft, Writing—review & editing).
Andrew Hanzlicek is employed by MiraVista Diagnostics and Jared Jaffey is a consultant for the company. No other authors have a conflict of interest.
This study was partially funded by MiraVista Diagnostics.
The authors declare no off-label use of antimicrobials.
This study was conducted in accordance with guidelines for clinical studies and approved by the Midwestern University Animal Care and Use Committee (protocol # 3024).
The authors declare no human ethics approval was required.