Mycoplasma spp were first documented in 1934,1 but their role in respiratory disease in dogs remains unclear. Mycoplasma spp are readily isolated from the oropharynges of healthy dogs and are generally accepted as commensal organisms in the oral cavity.2–4 Their presence in the LRT (ie, portion from the trachea to the lungs) is considered to be from endogenous translocation from the oral cavity during normal respiration.2,5 The likelihood of isolating Mycoplasma spp from the LRT is increased in dogs that are < 1 year of age, immunocompromised, or predisposed to aspiration, compared with that of unaffected dogs.3,6 In various studies,2,6,7 Mycoplasma spp have been isolated from the LRT of healthy dogs; however, these organisms have also been associated with airway collapse, pneumonia, and bronchitis in dogs.8 Mycoplasma spp, Mycoplasma cynos in particular, are associated with bronchopneumonia, tracheobronchial septic inflammation, and coinfections with Bordetella spp and Streptococcus spp in dogs.3,9,10
Although some believe that isolating Mycoplasma spp from BALF or transtracheal wash fluid signifies disease,8 the abundance of Mycoplasma spp in the mouth raises the question of whether oral bacterial contamination of BALF impacts bacterial culture or PCR assay results and their interpretation. Hirt et al7 compared bacterial culture results of BALF obtained through sterile protected airways and unprotected airways. Although they did not find a significant difference in bacterial populations between collection methods, oral contamination was not detected with either method. To date, no study has successfully determined whether oral bacterial contamination of BALF influences the detection of Mycoplasma spp from BALF.
The primary purpose of the study reported here was to determine whether an association exists between oral bacterial contamination of BALF and positive PCR assay results for the detection of Mycoplasma spp. The secondary purpose was to determine whether testing all BALF samples for Mycoplasma spp is warranted or if populations categorized under certain signalments or disease diagnoses were more likely to have BALF samples with positive Mycoplasma-specific PCR assay results. Specific aims were to determine the prevalence of positive Mycoplasma-specific PCR assay results in dogs with LRT disease, whether dogs with positive versus negative PCR assay results differed significantly in signalment, and whether disease diagnoses were predictive of Mycoplasma-specific PCR assay results. We hypothesized that a greater percentage of BALF samples contaminated with oral bacteria would have positive Mycoplasma-specific PCR assay results, signalment would be similar between dogs with positive or negative PCR assay results, a greater percentage of dogs with diseases predisposing to aspiration would have BALF samples with positive Mycoplasma-specific PCR assay results, and significantly more dogs with Bordetella spp or Streptococcus spp isolated from BALF samples would have positive Mycoplasma-specific PCR assay results.
Materials and Methods
Case selection—Medical records at the University of Illinois Veterinary Teaching Hospital from the period of January 2005 to April 2012 were reviewed to identify client-owned dogs with LRT disease that had a bronchoalveolar lavage performed (n = 131). For dogs to be included, BALF had to have been obtained and submitted to the University of Illinois Clinical Pathology Laboratory for cytologic evaluation and the University of Illinois Veterinary Diagnostic Laboratory for Mycoplasma-specific PCR assay and aerobic bacterial culture. Dogs were excluded from the study if pulmonary disease was not found. If a patient had undergone bronchoalveolar lavage collection at multiple time points, only the results of the first bronchoalveolar lavage were included. The final study population consisted of 121 dogs.
Medical records review—Patient age, body weight, sex, breed, disease diagnosis, and results of BALF cytologic evaluation, Mycoplasma-specific PCR assay, aerobic bacterial culture, and anaerobic bacterial culture (if performed) were extracted from the medical records. Age was rounded to the nearest whole number and assessed both continuously and categorically as follows: juvenile (< 1 year), young adult (1 to 5 years), middle aged (6 to 10 years), and senior (> 10 years). Body weight was rounded to the nearest whole number (in kilograms) and assessed both continuously and categorically as follows: toy (< 8 kg [< 17.6 lb]), small (8 to 15 kg [17.6 to 33 lb]), medium (16 to 30 kg [35.2 to 66 lb]), large (31 to 45 kg [68.2 to 99 lb]), and giant (> 45 kg [> 99 lb]). Sex was categorized as sexually intact female, spayed female, sexually intact male, and castrated male. Fifty breeds were represented and grouped according to American Kennel Club standards (toy breed dogs, sporting dogs, herding dogs, working dogs, terriers, nonsporting dogs, and hounds), with mixed-breed dogs in a separate category.
The BALF cytologic evaluation results were examined for oral bacterial contamination, which was defined as the presence of organisms in the family Simonsiellaceae or squamous epithelial cells and categorized as contaminated or not. Bronchoalveolar lavage fluid Mycoplasma-specific PCR assay status was categorized as negative or positive, with weak positives being classified as positive. Bronchoalveolar lavage fluid aerobic and anaerobic bacterial culture results were categorized by the bacterial genus, amount of bacterial growth, number of different bacterial genera that were isolated from the BALF sample from a single patient, and pathogenicity of isolates (eg, Bordetella spp and Pasteurella spp). The amount of growth detected was classified as no quantity listed, no growth, growth in enrichment broth only, rare growth (1 to 20 colony-forming units), light growth (21 to 60 colony-forming units), moderate growth (61 to 300 colony-forming units), and heavy growth (> 300 colony-forming units). If more than 1 species within a bacterial genus was isolated from an individual BALF sample, growth for that sample was scored according to the bacterial species having the heaviest growth. For example, if bacterial culture of a BALF sample resulted in moderate growth of Staphylococcus pseudintermedius and rare growth of Staphylococcus aureus, that patient would receive a value of moderate growth for Staphylococcus spp. Because anaerobic cultures of BALF samples were not performed in all patients, this category had the additional classification of not performed.
Disease diagnoses were grouped into 10 categories, with several dogs belonging to multiple categories as follows: tracheal or bronchial collapse, bronchitis, laryngeal paralysis, fungal pneumonia, bacterial pneumonia, aspiration pneumonia, eosinophilic bronchopneumopathy, intraluminal respiratory neoplasia, extraluminal respiratory neoplasia, and brachycephalic airway syndrome. Tracheal or bronchial collapse was defined by bronchoscopic or radiographic evidence of airway collapse. Bronchitis was defined by the presence of a chronic cough with cytologic evidence of chronic inflammation of the respiratory tract in the absence of identifiable neoplastic or other primary disease. Laryngeal paralysis was defined by reduced or absent movement of the arytenoid cartilages despite maximal respiratory effort during sedated laryngeal examination. Fungal pneumonia was defined by cytologic evidence of the presence of fungal organisms in the respiratory tract or detection of fungal antigen within urine samples. Bacterial pneumonia was defined by acute cough and BALF samples that had evidence of acute inflammation on cytologic evaluation, growth on aerobic or anaerobic bacterial cultures, or positive Mycoplasma-specific PCR assay results. Aspiration pneumonia was defined by radiographic evidence of pulmonary-dependent alveolar consolidation with factors predisposing to aspiration. Eosinophilic bronchopneumopathy was defined by cytologic evidence of severe eosinophilic inflammation of the airways or lung parenchyma in the absence of other identifiable causes of eosinophilic infiltration. Intraluminal neoplasia and extraluminal neoplasia were defined by masses being observed or cytologic findings suggestive or confirmatory of neoplasia. Brachycephalic airway syndrome was defined by brachycephalic breed status or the presence of > 1 criterion of brachycephalic airway syndrome (eg, stenotic nares, elongated or thickened soft palate, everted laryngeal saccules, hypoplastic trachea, or laryngeal collapse).
Bronchoalveolar lavage—Because of the retrospective nature of this study, it was impossible to ensure that all BALF samples were collected identically. However, the standard bronchoalveolar lavage protocol that is used at the authors’ veterinary hospital was followed. The dog was anesthetized and maintained with injectable anesthetics while in sternal recumbency. A mouth gag was positioned between a mandibular and maxillary canine tooth. A flexible videoendoscope was passed through the larynx into the trachea. The LRT (ie, portion from the trachea to the lungs) was systematically examined for structural abnormalities. Bronchoalveolar lavage fluid was obtained from abnormal bronchi by flushing sterile saline (0.9% NaCl) solution through the endoscope into the airways in aliquots appropriate for patient size and retrieving fluid by immediate suction. The BALF sample was placed in a sterile serum tube. At least 1 mL of BALF was sent to the University of Illinois Clinical Pathology Laboratory for cytologic evaluation, and at least 2 mL were sent to the University of Illinois Veterinary Diagnostic Laboratory for Mycoplasma-specific PCR assay and aerobic bacterial culture. For 34 patients, BALF was also submitted for anaerobic bacterial culture.
Cytologic evaluation—The standard protocol used by the University of Illinois Clinical Pathology Laboratory involved placing 200 μL of BALF into a cytocentrifuge funnel and dispersing the cells onto glass slides. The sample was centrifuged for 3 minutes at 112.9 × g. Direct smears were prepared for all samples, and if visible mucous aggregates were identified, squash smears for cytologic examination were also prepared. The slides were allowed to dry and then stained with Wright-Giemsa stain. The stained slides were reviewed by a board-certified veterinary clinical pathologist.
PCR assay—The standard protocol used by the University of Illinois Veterinary Diagnostic Laboratory for Mycoplasma-specific PCR assay is based on the published protocol by Lauerman.11 Samples of BALF were inoculated into enrichment brotha (incubated at 37°C; 9% CO2 for 48 hours). Deoxyribonucleic acid was extracted from a 200-μL aliquot of the broth with a kitb per manufacturer's instructions.
Fifty microliters of general Mycoplasma reaction mix contained 25 μL of master mix,c 0.5 μL of 10μM forward primer (5′-ACA CCA TGG GAG CTG GTA AT-3′), and 0.5 μL of 10μM reverse primer (5′-CCT CAT CGA CTT TCA GAC CCA AGG CAT-3′) with or without 2 μL of internal control plasmid, 5 μL of DNA template, and double-distilled water to 50 μL. The genus specific primers span the intergenic region between the 16S and 23S rRNA genes.
The PCR conditions specified 94°C for 2 minutes, followed by 40 cycles of 94°C for 30 seconds, 55°C for 30 seconds, and 72°C for 30 seconds. After the final cycle, the reaction was held at 72°C for 5 minutes and then held at 4°C. Ten microliters of the amplicon was mixed with 5 μL of gel loading dye and electrophoresed on a 2% to 3% precast agarose geld with ethidium bromide at 120 V for 60 to 65 minutes. Gels were placed under UV light and interpreted. Mycoplasma amplicons were usually 400 to 800 base pairs. Detected Mycoplasma organisms were not speciated.
Interpretation of PCR assay results was based on the size and intensity of the stained band resulting from the amplification product. Positive samples, negative samples, and internal PCR control samples were used. Internal control and positive samples competed for primers when run in the same reaction mix. To be considered a negative result, there needed to be amplification of the plasmid (internal control) resulting in a low molecular weight band (327 base pairs) with the Mycoplasma general primers and no amplification without the template control. A positive PCR assay result had an amplicon within 400 to 800 base pairs. A weak positive PCR assay result was a discernible band, whereas a positive PCR assay result was an intense band.
Bacterial culture—The standard protocol used by the University of Illinois Veterinary Diagnostic Laboratory for aerobic culture includes plating the BALF on Columbia blood agar, Columbia blood agar with colistin and nalidixic acid, Chocolate agar II, and MacConkey agare at 37°C in an atmosphere of air with 5% CO2. Plates were inspected for growth at 24 and 48 hours incubation. Bacteria were identified with the aid of traditional tube tests,12 standard identification plates,f or identification microplates.g Specimens for anaerobic culture were plated on Brucella blood agar, laked blood agar with kanamycin and vancomycin, and phenylethyl alcohol agar.e Organisms determined to be obligate anaerobes by aerotolerance testing were identified on the basis of the scheme outlined in the Wadsworth Anaerobic Bacteriology Manual.13
Statistical analysis—Continuous data were evaluated on the basis of the Shapiro-Wilk test, skewness, kurtosis, and q-q plots. Data that were normally distributed were reported as mean, SD, and range values, whereas nonnormally distributed data were reported as the median, 10% to 90%, and range values. The 95% CIs were calculated for binomial proportions. Fisher exact tests or χ2 tests were used to determine whether Mycoplasma-specific PCR assay status (ie, positive or negative result) was associated with potential sample contamination, age, breed, sex, bacteria, or disease diagnosis. Associations found to have a value of P < 0.2 were included in a logistic regression model. Main effect variables were removed individually from full models to assess the effects on the model likelihood ratio statistics, magnitude of the coefficients for other variables, and Hosmer-Lemeshow goodness-of-fit statistics. Interactions between the main effects variables were also evaluated in the models. Statistical softwareh was used for the analysis. Values of P ≤ 0.05 were considered significant.
Results
A total of 121 dogs were included in this study. Testing of BALF samples for Mycoplasma spp revealed that 83 (68.6%; 95% CI, 60.3% to 76.9%) dogs had negative PCR assay results and 38 (31.4%; 95% CI, 23.1% to 39.7%) dogs had positive PCR assay results. Positive PCR assay results were considered only weakly positive for 3 (2.5%) dogs. Contamination of BALF samples with oral bacteria was not detected for most dogs (no contamination, 95/121 [78.5%; 95% CI, 71.2% to 85.8%]; contamination, 26/121 [21.5%; 95% CI, 14.2% to 28.8%]).
Bacteria were not isolated from BALF samples by aerobic or anaerobic culture from 45 of 121 (37.2%) dogs. In these 45 dogs, 8 (17.8%) had BALF samples with positive Mycoplasma-specific PCR assay results and 37 (82.2%) had negative PCR assay results (these dogs represented 6.6% [8/121] and 30.6% [37/121] of the entire population, respectively; Table 1). Seven of 8 dogs with negative bacterial culture results, but positive Mycoplasma-specific PCR assay results, also did not have evidence of oral bacterial contamination of BALF samples on the basis of cytologic examination; the remaining dog, however, did have cytologic evidence of oral bacterial contamination.
Comparison of the number of different bacterial genera that were isolated from bacterial culture of BALF samples of dogs (n = 121) with LRT disease to the number of BALF samples with cytologic evidence (present or absent) of oral bacterial contamination and PCR assay results (negative or positive) for Mycoplasma spp in BALF samples.
| Evidence of bacterial contamination | Mycoplasma-specific PCR assay results | ||||
|---|---|---|---|---|---|
| No. of bacterial genera | Absent | Present | Negative | Positive | Total |
| 0 | 41 | 4 | 37 | 8 | 45 |
| 1 | 28 | 1 | 21 | 8 | 29 |
| 2 | 13 | 2 | 10 | 5 | 15 |
| 3 | 3 | 3 | 2 | 4 | 6 |
| 4 | 5 | 5 | 7 | 3 | 10 |
| 5 | 4 | 1 | 3 | 2 | 5 |
| 6 | 0 | 7 | 2 | 5 | 7 |
| 7 | 1 | 1 | 0 | 2 | 2 |
| 8 | 0 | 2 | 1 | 1 | 2 |
| Total | 95 | 26 | 83 | 38 | |
The majority of dogs tested were considered middle aged (48/121 [39.7%; 95% CI, 30.9% to 48.4%]) to senior (32/121 [26.4%; 95% CI, 18.5% to 34.2%]), with lesser numbers of young adults (32/121 [26.4%; 95% CI, 18.5% to 34.2%]) or juveniles (9/121 [7.4%; 95% CI, 2.7% to 12.1%]). Toy breed dogs (31/121 [25.6%; 95% CI, 17.8% to 33.3%]) represented the most common breed group in this sample population, followed by sporting dogs (26/121 [21.5%; 95% CI, 14.2% to 28.8%]), mixed-breed dogs (22/121 [18.2%; 95% CI, 11.3% to 25.1%]), herding dogs (12/121 [9.9%; 95% CI, 4.6% to 15.2%]), working dogs (10/121 [8.3%; 95% CI, 3.3% to 13.3%]), terriers (7/121 [5.8%; 95% CI, 1.6% to 10.0%]), nonsporting dogs (7/121 [5.8%; 95% CI, 1.6% to 10.0%]), and hounds (6/121 [5.0%; 95% CI, 1.2% to 8.8%]). Females (72/121 [59.5%, 95% CI, 50.7% to 68.2%]), both spayed (60/121 [49.6%]) and sexually intact (12/121 [9.9%]), were more common in this sample population than males (total, 49/121 [40.5%; 95% CI, 31.7% to 49.3%]; castrated, 36/121 [29.8%]; sexually intact, 13/121 [10.7%]). The median body weight for this population of dogs was 14.0 kg (30.8 lb), with 10% to 90% weighing 3.2 to 38.4 kg (7.0 to 84.5 lb; range, 1.0 to 65.0 kg [2.2 lb to 143.0 lb]). Final diagnoses for this sample population were summarized (Table 2).
Number (%) of dogs (n = 121) with each disease diagnosis for which cytologic evidence of oral bacterial contamination was present or absent.
| Disease* | Present | Absent |
|---|---|---|
| Tracheal or bronchial collapse | 46 (38.0) | 75 (62.0) |
| Bronchitis | 44 (36.4) | 77 (63.6) |
| Laryngeal paralysis | 21 (17.4) | 100 (82.6) |
| Fungal pneumonia | 9 (7.4) | 112 (92.6) |
| Bacterial pneumonia | 17 (14.0) | 104 (86.0) |
| Aspiration pneumonia | 10 (8.3) | 111 (91.7) |
| Eosinophilic bronchopneumopathy | 6 (5.0) | 115 (95.0) |
| Intraluminal neoplasia | 2 (1.7) | 119 (98.3) |
| Extraluminal neoplasia | 6 (5.0) | 115 (95.0) |
| Brachycephalic airway syndrome | 14 (11.6) | 107 (88.4) |
Disease diagnoses were grouped into 10 categories, with several dogs belonging to multiple categories.
There was a significant (χ2 = 7.5; P = 0.02) association between Mycoplasma-specific PCR assay status (ie, positive or negative result) and the likelihood of oral bacterial contamination of BALF samples, with 13 of 26 (50%; 95% CI, 30.7% to 69.2%) contaminated BALF samples having positive Mycoplasma-specific PCR assay results and only 22 of 95 (23.2%; 95% CI, 14.7% to 31.6%) noncontaminated BALF samples having positive PCR assay results. There was no significant (P = 0.15) difference in the likelihood of oral bacterial contamination of BALF samples in dogs with positive Mycoplasma-specific PCR assay results when taking into account the presence or absence of bacterial growth on culture. When individual bacterial genera or known respiratory pathogens (ie, Bordetella spp and Pasteurella spp) were compared in terms of the likelihood of a positive Mycoplasma-specific PCR assay result, no significant association was found for any bacteria, individually (χ2 ≥ 0.45; P ≥ 0.1) or when grouped (χ2 = 0.15; P = 0.69) by known pathogenicity. When the presence or absence of isolation of bacteria from BALF samples by aerobic or anaerobic culture was compared with the likelihood of a positive Mycoplasma-specific PCR assay result (χ2 = 1.6; P = 0.23) or oral bacterial contamination of BALF (P = 0.66), no significant association was found. There was no significant (χ2 = 10.9; P = 0.09) association between Mycoplasma-specific PCR assay status (ie, positive or negative result) and age groups, although some age groups had a higher proportion of dogs with BALF samples that had positive Mycoplasma-specific PCR assay results than did others: juveniles (5/9 [55.6%]), middle aged (18/48 [37.5%]), seniors (7/32 [21.9%]), and young adults (7/32 [21.9%]). There was no significant difference in Mycoplasma-specific PCR assay status by body weight (χ2 = 12.3; P = 0.14), although dogs with low body weights had BALF samples with positive Mycoplasma-specific PCR assay results more often than dogs with high body weights (< 8 kg, 16/39 [41%]; 8 to 15 kg: 10/26 [38.5%]; 16 to 30 kg, 7/30 [23.3%]; 31 to 45 kg, 4/22 [18.2%]; and > 45 kg, 1/4). There was a significant (χ2 = 26.7; P = 0.02) difference in Mycoplasma-specific PCR assay status by breed, with terriers being the most likely to have a positive PCR assay result (5/7 [71.4%]), followed by nonsporting breeds (4/7 [57.1%]), toy breeds (11/31 [35.5%]), mixed-breed dogs (7/22 [31.8%]), sporting breeds (8/26 [30.8%]), working dogs (2/10 [20.0%]), herding dogs (1/12 [8.3%]), and hounds (0/6 [0%]). There was no significant (χ2 = 4.0; P = 0.67) difference in Mycoplasma-specific PCR assay status by sex. Data on PCR assay results and disease diagnoses are summarized (Table 3); significant associations were only found between PCR assay results and bronchitis (χ2 = 3.8; P = 0.05), brachycephalic airway syndrome (χ2 = 4.9; P = 0.03), and isolation of Oligella spp from BALF samples (P = 0.05).
Associations between positive and negative Mycoplasma-specific PCR assay results of BALF samples and final disease diagnosis in dogs (n = 121) with LRT disease that had (present) or did not have (absent) cytologic evidence of oral bacterial contamination of BALF samples.
| Present | Absent | |||||
|---|---|---|---|---|---|---|
| Disease* | PCR positive | PCR negative | PCR positive | PCR negative | χ2 | P value |
| Tracheal or bronchial collapse | 19 | 27 | 19 | 56 | 3.4 | 0.07 |
| Bronchitis | 9 | 35 | 29 | 48 | 3.8† | 0.05† |
| Laryngeal paralysis | 5 | 16 | 33 | 67 | 0.61 | 0.41 |
| Fungal pneumonia | 3 | 6 | 35 | 77 | — | 0.58 |
| Bacterial pneumonia | 4 | 13 | 34 | 70 | — | 0.58 |
| Aspiration pneumonia | 2 | 8 | 33 | 75 | — | 0.50 |
| Eosinophilic bronchopneumopathy | 1 | 5 | 35 | 77 | — | 0.66 |
| Intraluminal neoplasia | 1 | 1 | 34 | 82 | — | 0.53 |
| Extraluminal neoplasia | 3 | 3 | 32 | 80 | — | 0.38 |
| Brachycephalic airway syndrome | 8 | 6 | 27 | 77 | 4.9† | 0.03† |
— = Not applicable.
Significant (P ≤ 0.05) effect on Mycoplasma-specific PCR assay results.
See Table 2 for remainder of key.
Independent variables associated with positive Mycoplasma-specific PCR assay results were entered into the initial logistic regression model as follows: oral bacterial contamination of BALF; dog age, body weight, and breed; isolation of Oligella spp, Bacteroides spp, Enterococcus spp, Moraxella spp, Ochrobactrum spp, or Propionibacteria spp from bacterial culture of BALF; and whether a dog had brachycephalic airway syndrome or bronchitis. Values were χ2 = 27.0 and P = 0.001 for the final logistic regression model. Variables found to have a significant influence in the final model were dog breed (P = 0.01), oral bacterial contamination of BALF (P = 0.003), and bronchitis (P = 0.05). Bronchoalveolar lavage fluid samples that had cytologic evidence of oral bacterial contamination were 5.1 times (95% CI, 1.6 to 15.9 times) as likely to have positive Mycoplasma-specific PCR assay results as noncontaminated BALF samples. Breeds other than those considered hound or herding dogs were found to be 13.6 times (95% CI, 1.6 to 113.5 times) as likely to have a positive PCR assay result, and dogs with bronchitis were less likely (OR, 0.38; 95% CI, 0.14 to 1.0) to have BALF samples with positive Mycoplasma-specific PCR assay results.
Discussion
This study is the first to confirm that BALF samples with oral bacterial contamination are significantly more likely to have positive Mycoplasma-specific PCR assay results than noncontaminated BALF samples. The finding that contaminated BALF samples were 5.1 times as likely to have positive Mycoplasma-specific PCR assay results strongly suggests that the Mycoplasma spp isolated from the respiratory tract originated from the mouth and were actually commensals instead of LRT pathogens. With this in mind, antimicrobial treatment of patients on the basis of positive Mycoplasma-specific PCR assay results may be unnecessary. Nevertheless, we cannot rule out LRT colonization occurring concurrently with oral bacterial contamination of BALF. Of the 8 bronchoalveolar lavage samples that had positive Mycoplasma-specific PCR assay results and negative bacterial culture results, 7 did not have cytologic evidence of bacterial contamination. Bronchoalveolar lavage fluid from which only Mycoplasma spp were detected was not significantly more likely to lack oral bacterial contamination than BALF from which Mycoplasma spp were detected and other bacterial cultured. Nevertheless, a PCR-positive and bacterial culture-negative BALF sample is still strongly suggestive of LRT colonization over simple contamination with Mycoplasma spp. Given that Mycoplasma spp were detected with and without concurrent bacteria, our positive Mycoplasma-specific PCR assay results probably include those that are translocated and LRT residents. Because of the retrospective nature of this study and the lack of evidence that oral Mycoplasma spp can be differentiated from those in the LRT, we were unable to confirm the origin of the Mycoplasma spp detected by PCR assay. Prospective studies concurrently collecting paired samples from the oral cavity and LRT would be necessary to further investigate this. Although our results do not discourage performing Mycoplasma-specific PCR assays on BALF samples that have been contaminated with oral bacteria or from which other bacteria are isolated, they do emphasize the importance of considering potential misclassification of positive PCR assay results in these BALF samples.
All BALF samples were submitted for aerobic bacterial culture, and 28% of samples were submitted for anaerobic bacterial culture. No significant association was found between BALF samples with positive Mycoplasma-specific PCR assay results and isolation of any bacteria, individually or grouped by known pathogenicity. Surprisingly, not even dogs with Streptococcus spp or Bordetella spp isolated from BALF samples were significantly more likely to have positive Mycoplasma-specific PCR assay results, as Randolph et al3 have reported; however, heavy growth of Bordetella spp or Streptococcus spp could have selectively excluded Mycoplasma spp from the LRT. Although Jameson et al6 have demonstrated that dogs with positive Mycoplasma-specific PCR assay results are significantly more likely to have > 1 bacterial species isolated from transtracheal wash fluid, our results were not similar. Differing culture media, techniques, or vulnerability of organisms may account for these discrepancies. We were unable to demonstrate that BALF from which ≥ 1 bacterium were cultured was more likely to have oral bacterial contamination, although isolation of Oligella spp, bacteria normally residing in the oral cavity, did significantly increase the likelihood of oral bacterial contamination of BALF. Jameson et al6 stated that oropharyngeal contamination could have affected their findings, yet their samples were not evaluated for contamination. Considering that oral bacterial contamination introduces multiple bacteria, perhaps the tendency for Mycoplasma spp to be isolated with other bacteria is partially the result of oral bacterial contamination and not LRT colonization.
The overall prevalence of positive Mycoplasma-specific PCR assay results (31.4%) in this study was similar to results stated in one report3 (34%) but lower than another8 (49.5%), in which dogs with LRT disease had positive PCR assay results. When looking at the prevalence of positive Mycoplasma-specific PCR assay results in dogs with bacterial pneumonia (14%) and aspiration pneumonia (8.3%), our results were much lower than those reported previously by Jameson et al6 (70%) and Tart et al13 (21.3%). Perhaps the disparities arose from differing diagnostic methods, sample populations, or inclusion criteria; both of their studies used bacterial culture, which may be more sensitive than PCR assay on BALF evaluation.1 However, Mycoplasma-specific PCR assay results have been reported to be similar to those obtained by means of bacterial culture.i Additionally, the University of Illinois Veterinary Diagnostic Laboratory PCR assay uses enrichment broth prior to PCR amplification. This enhances the detection of Mycoplasma spp that grow in broth but could represent a 1:20-fold reduction in organism DNA for strains incapable of growth in this broth. The limit of sensitivity of the University of Illinois Veterinary Diagnostic Laboratory assay has been determined to be approximately 1,000 organisms/mL. Regardless, if even one-third of dogs with LRT disease are likely to have BALF samples with positive Mycoplasma-specific PCR assay results, regular testing for Mycoplasma spp is reasonable.
Surprisingly, dogs in herding or hound groups were significantly less likely to have BALF samples with positive Mycoplasma-specific PCR assay results than other groups. Of the 12 herding dogs, 11 had negative PCR assay results, and of the 6 hounds, all had negative PCR assay results. So far, no breed predispositions have been reported for Mycoplasma infections. Perhaps the lifestyle or body weight of these breeds made them less prone to LRT invasion by Mycoplasma spp. Dogs with low body weights had a higher proportion of positive Mycoplasma-specific PCR assay results, compared with dogs with high body weights, even though body weight was not a significant variable. Ten of 12 herding dogs and 1 of 6 hounds were > 15 kg (> 33 lb), so body weight and breed likely have some shared effect, although there was no significant interaction found in the regression model.
Even though there was no significant correlation between age groups and Mycoplasma-specific PCR assay results, our findings (5/9) were similar to those of the study by Randolph et al3 (7/9), in which dogs < 1 year of age were more likely to have positive Mycoplasma-specific PCR assay results than were older dogs. Our results did not agree with the study of Jameson et al,6 in which no age of dogs was significantly more likely to result in positive Mycoplasma-specific PCR assay results, but unlike the present study, their sample population contained mostly (60%) dogs > 5 years of age.
Although it is logical to conclude that dogs aspirating oral contents would be more likely to have Mycoplasma spp isolated from their BALF, our results were not supportive of this. Those with aspiration pneumonia and laryngeal paralysis were not significantly more likely to have BALF samples with positive Mycoplasma-specific PCR assay results. Although dogs with brachycephalic airway syndrome were significantly more likely to have BALF samples with positive Mycoplasma-specific PCR assay results when compared by use of univariate analysis, no significant association was found with the regression model, suggesting that this variable was masked by the breed effect, given that brachycephalic airway syndrome crossed over several breeds. Unexpectedly, we found that dogs with bronchitis were less likely to have BALF samples with positive Mycoplasma-specific PCR assay results. A significant relationship has not been previously described, but another study8 found bronchitis in 30% of animals that had Mycoplasma spp as a sole BALF isolate. Perhaps the chronicities of their diseases, previous exposure to antimicrobials, or concurrent bacteria created environments unfavorable for Mycoplasma colonization. Alternatively, Mycoplasma could have been present but adherence to respiratory cilia, as seen with some Mycoplasma spp,14 could have hindered isolation.
We recognize the inherent shortcomings in this retrospective study, including the inability to standardize sample collection procedures and the absence of control groups (because only dogs with suspected pulmonary disease underwent bronchoalveolar lavage). To minimize some of these potential biases, only dogs seen by our institute were included; this indirectly resulted in inclusion of dogs living in the Midwest, dogs with owners able to afford advanced diagnostic testing, and dogs in which previous medical management was unsuccessful. Although the narrow sample population prevents our results from being applicable to the general population, it is representative of animals having bronchoalveolar lavage performed on a regular basis. Other minor weaknesses included failure to speciate the Mycoplasma spp that were isolated and to include prior antimicrobial treatments and disease chronicities. Speciation after detection of Mycoplasma organisms by use of PCR assay is not routinely performed at the authors’ practice and was not available for inclusion in our study. This may have been useful in estimating the pathogenicity of the Mycoplasma spp detected, and it would have been interesting to note whether M cynos was more prevalent in dogs < 1 year old9 or those from which results of bacterial culture were negative. Although prior treatment and disease chronicity certainly could help explain why some dogs had negative Mycoplasma-specific PCR results, these variables were excluded because of incomplete or inexact histories. For example, owners often could not remember drug names or would give vague time intervals regarding chronicity.
In the evaluation of BALF samples, Mycoplasma-specific PCR assay results should always be interpreted in light of the oral bacterial contamination status and concurrent bacterial growth. Even though testing for Mycoplasma spp in herding dogs, hounds, and dogs suspected of having bronchitis may be unrewarding, the prevalence of Mycoplasma spp in dogs with LRT disease is high enough to warrant keeping Mycoplasma spp on our differential diagnosis list and continuing to test for it regularly.
ABBREVIATION
| BALF | Bronchoalveolar lavage fluid |
| CI | Confidence interval |
| LRT | Lower respiratory tract |
Mycoplasma broth media, Hardy Diagnostics, Santa Maria, Calif.
QIAmp DNA Mini Kit, Qiagen, Valencia, Calif.
2× Master Mix, Promega Corp, Madison, Wis.
Bio-Rad Laboratories Inc, Hercules, Calif.
Remel, Lenexa, Kan.
GNID or GPID panels, Sensititre, Westlake, Ohio.
MicroID GN2 or GP2 panels, Biolog Inc, Hayward, Calif.
SPSS, version 18.0, SPSS Inc, Chicago, Ill.
Cruse A, Ratterree W, Sanchez S, et al. Comparison of culture and polymerase chain reaction for the detection of Mycoplasma species in canine and feline respiratory tract samples (abstr). J Vet Intern Med 2008;22:706–707.
References
1 Chalker VJ. Canine mycoplasmas. Res Vet Sci 2005; 79: 1–8.
2 Koshimizu K, Ogata M. Characterization and differentiation of mycoplasmas of canine origin. Nippon Juigaku Zasshi 1974; 36: 391–406.
3 Randolph JF, Moise NS & Scarlett JM, et al. Prevalence of mycoplasmal and ureaplasmal recovery from tracheobronchial lavages and prevalence of mycoplasmal recovery from pharyngeal swab specimens in dogs with or without pulmonary disease. Am J Vet Res 1993; 54: 387–391.
4 Rosendal S. Canine mycoplasmas: their ecologic niche and role in disease. J Am Vet Med Assoc 1982; 180: 1212–1214.
5 Bemis DA. Bordetella and Mycoplasma respiratory-infections in dogs and cats. Vet Clin North Am Small Anim Pract 1992; 22: 1173–1186.
6 Jameson PH, King LA & Lappin MR, et al. Comparison of clinical signs, diagnostic findings, organisms isolated, and clinical outcome in dogs with bacterial pneumonia: 93 cases (1986–1991). J Am Vet Med Assoc 1995; 206: 206–209.
7 Hirt RA, Wiederstein I & Denner EBM, et al. Influence of the collection and oxygenation method on quantitative bacterial composition in bronchoalveolar lavage fluid samples from healthy dogs. Vet J 2010; 184: 77–82.
8 Chandler JC, Lappin MR. Mycoplasmal respiratory infections in small animals: 17 cases (1988–1999). J Am Anim Hosp Assoc 2002; 38: 111–119.
9 Chalker VJ, Owen WMA & Paterson C, et al. Mycoplasmas associated with canine infectious respiratory disease. Microbiology 2004; 150: 3491–3497.
10 Zeugswetter F, Weissenbock H & Shibly S, et al. Lethal bronchopneumonia caused by Mycoplasma cynos in a litter of Golden Retriever puppies. Vet Rec 2007; 161: 626–627.
11 Lauerman LH. Nucleic acid amplification assays for diagnosis of animal diseases. Navarre, Fla: American Association of Veterinary Laboratory Diagnosticians, 1998;1–52.
12 Carter GR, Cole JR. Diagnostic procedures in veterinary bacteriology and mycology. 5th ed. San Diego: Academic Press Inc, 1990.
13 Jousimies-Somer H, Summanen P & Citron D, et al. Wadsworth-KTL anaerobic bacteriology manual. 6th ed. Belmont, Calif: Star Publishing, 2002.
14 Tart KM, Babski DM, Lee JA. Potential risks, prognostic indicators, and diagnostic and treatment modalities affecting survival in dogs with presumptive aspiration pneumonia: 125 cases (2005–2008). J Vet Emerg Crit Care (San Antonio) 2010; 20: 319–329.
