Histologic and Molecular Correlation in Shelter Cats with Acute Upper Respiratory Infection

JOURNAL OF CLINICAL MICROBIOLOGY, July 2011, p. 2454 2460 Vol. 49, No. 7 0095-1137/11/$12.00 doi:10.1128/jcm.00187-11 Copyright 2011, American Society for Microbiology. All Rights Reserved. Histologic and Molecular Correlation in Shelter Cats with Acute Upper Respiratory Infection Rachel E. Burns, 1 Denae C. Wagner, 2 Christian M. Leutenegger, 3 and Patricia A. Pesavento 4 * Wildlife Disease Laboratories, San Diego Zoo Institute for Conservation Research, P.O. Box 120551, San Diego, California 92112 1 ; Department of Pathology, Microbiology and Immunology 4 and the Koret Shelter Medicine Program, Center for Companion Animal Health, 2 School of Veterinary Medicine, University of California, Davis, 1 Shields Avenue, Davis, California 95616-5270; and IDEXX Laboratories, Inc., West Sacramento, California, 95605 3 Received 29 January 2011/Returned for modification 25 March 2011/Accepted 27 April 2011 This is a descriptive study designed to correlate diagnostic real-time PCR results with histopathologic lesions in cats with clinical signs of upper respiratory infection (URI). The study occurred over a 9-month period in a single open-intake animal shelter. Cats that were selected for euthanasia by the shelter staff and additionally had URI were included in the study, for a total of 22 study cats. Combined conjunctival and oropharyngeal swab specimens were tested by quantitative real-time PCR (qpcr) for feline herpesvirus type 1 (FHV-1), feline calicivirus (FCV), Mycoplasma felis, Chlamydophila felis, and Bordetella bronchiseptica. Necropsy was performed on all cats, and a complete set of respiratory tract tissues was examined by histopathology. Among 22 cats, 20 were qpcr positive for FHV-1, 7 for M. felis, 5 for FCV, 1 for C. felis, and 0 for B. bronchiseptica. Nine cats were positive for two or more pathogens. Histopathologic lesions were present in all cats, with consistent lesions in the nasal cavity, including acute necroulcerative rhinitis in 16 cats. Histologic or antigenic detection of FHV-1 was seen in 18 of 20 cats positive for FHV-1 by qpcr. No lesions that could be specifically attributed to FCV, M. felis, or C. felis were seen, although interpretation in this cohort could be confounded by coinfection with FHV-1. A significant agreement was found between the amount of FHV-1 DNA determined by qpcr and the presence of specific histopathologic lesions for FHV-1 but not for the other respiratory pathogens. Upper respiratory infection (URI) is the most frequent disease reported in the 2 million to 6 million cats estimated to pass through United States shelters each year. Feline URI results from a complex, multifactorial interaction of respiratory pathogens, stress, and animal susceptibility. Intensive housing of cats, such as that found in animal shelters, catteries, and multicat households, can contribute to disease both by compromise to the host (stress, travel, poor health status) and by exposure, dose, and, potentially, evolution of the contributing pathogens. Although mortality is low, clinical signs of URI are discerning criteria in many shelters for euthanasia, so the consequences for affected cats are profound. Moreover, chronic, recurrent rhinosinusitis, considered incurable, may be a sequela of acute rhinitis (10). Many epidemiologic and diagnostic studies have identified five pathogens commonly associated with feline URI, which are feline herpesvirus type 1 (FHV-1), feline calicivirus (FCV), Bordetella bronchiseptica, Mycoplasma felis, and Chlamydophila felis (1, 5, 9, 15, 17, 21). Clinical and diagnostic investigation is complicated for a number of reasons: clinical signs of infection with any one or combination of these five pathogens are overlapping and often nonspecific. Vaccination, common for FHV-1, FCV, and C. felis, can attenuate but is not protective of disease and/or shedding (14, 16, 20), and lastly, all of these microbes can be carried by and detected in clinically normal * Corresponding author. Mailing address: University of California, Davis, Vet Med:PMI, 4206 VM3A, 1 Shields Avenue, Davis, CA 95616-5270. Phone: (530) 752-1166. Fax: (530) 752-3349. E-mail: papesavento@ucdavis.edu. Published ahead of print on 11 May 2011. cats either as latent (13) or persistent (3) infections or as part of the normal flora (6, 8, 4). There is a comprehensive body of epidemiologic literature that explores the potential significance of identification (by culture/isolation or by PCR) of these pathogens in cats with or without clinical URI (1, 5, 11, 16, 17). The collective results are confounding because studies alternately reveal a significant association or no significant association between the presence of pathogens and the clinical state and are valid only on the basis of a single population, time, and situation. It is impossible to extrapolate, by detection alone, which pathogen is actually causing URI in any single cat or group of cats at a given period of time. Studies on pathogen detection have compared sites for sampling (17) and methods of detection, including PCR, culture, virus isolation, and indirect immunofluorescence (1, 2). Enigmatically, choosing a diagnostic method or sample site on the basis of sensitivity alone will increase detection in low-shedding carriers, without regard to true causation. For example, by PCR, vaccine strains cannot routinely be distinguished from wild-type strains of FHV-1 (12). We have designed this study to correlate pathological findings in cases of feline URI with the quantitative detection of pathogens by quantitative realtime PCR (qpcr). Validation of diagnostic results with histopathology and immunohistochemistry is an approach uniquely capable of establishing causation in the case of feline URI. Over a 9-month period of time, detection of common respiratory pathogens by qpcr was compared to necropsy results from 22 cats euthanized with clinical URI. A full necropsy was performed in each case, and a comprehensive set of tissues, including those from the upper and lower respiratory tracts, was examined by histopathology. 2454

VOL. 49, 2011 HISTOLOGIC AND MOLECULAR CORRELATION OF FELINE URI 2455 FIG. 1. Nasal cavity, maxillary turbinates. The histologic grading scheme for rhinitis included type and extent of inflammation as well as changes to the epithelium. Grades I to V in the figure correlate with grades 1 to 5 in the text. MATERIALS AND METHODS Shelter cat population. URI pathogens were monitored in the cat population at the shelter via PCR testing from August 2008 through September 2009. Sampling occurred quarterly for the year period of time, and sample collection occurred over a 1- to 2-week time period each quarter. Approximately 40 cats were sampled at each quarter (20 cats within 24 h of arrival at the shelter [at intake] and 20 cats with clinical signs of URI). Two cotton swab samples were collected from each individual cat, one from each of the conjunctival and deep laryngeal areas, and submitted for qpcr testing. Pathology study group. From August 2008 to May 2009, full necropsies were performed on 22 cats with clinical signs of URI that were selected by the shelter staff for euthanasia. None of these cats overlapped the above-described sampled population. All cats were from a single, open-intake California animal shelter. One of two veterinarians examined each cat immediately prior to euthanasia. Clinical signs, including respiratory effort, ocular discharge, nasal discharge, oral lesions, and lung auscultation, were noted. The signalment, duration of clinical signs, length of time in the shelter, intake type (stray or owner surrender), and, if known, the vaccination status and any antibiotic therapy were recorded. One oropharyngeal swab sample and one conjunctival cotton swab sample were collected and pooled (17) for each cat immediately after euthanasia. These samples were frozen (within 2hofcollection) at 20 C until submission for qpcr testing. Necropsy. A full necropsy was performed within 1 h of euthanasia with a collection protocol that included sections of palpebral conjunctiva, third eyelid, tongue, pharyngeal mucosa, tonsils, epiglottis, proximal and distal trachea, retropharyngeal and mandibular lymph nodes, multiple sections of lung, conjunctiva, nasal planum/nares, nasal valve, and rostral and caudal nasal/maxillary turbinates immediately placed in 10% buffered formalin. Tissue around the nares/nasal valve was collected by making a transverse cut through the cartilaginous portion of the nose approximately 3 mm from its rostral end. Sections of rostral and caudal nasal cavity were collected after sectioning of the head at midline. Nasal and maxillary turbinates were then divided into equal rostral and caudal halves and placed whole in separate cassettes for histopathologic evaluation. Lung sections were taken from 3 separate lobes and pressure inflated by hand with 10% buffered formalin. Formalin-fixed tissues were routinely processed and embedded in paraffin within 4 days of necropsy, sectioned at 5 m, and routinely stained with hematoxylin-eosin (H&E) for histologic evaluation. Histologic grading system. Tissues from the rostral and caudal nasal cavities were separately graded 1 to 5 on the basis of characterization of the inflammatory changes (Fig. 1). Grade 1 lesions were characterized by infiltration of the subepithelial stroma with small to moderate numbers of predominantly mononuclear cells and fewer admixed neutrophils or eosinophils with no infiltration or disruption of the mucosal epithelium. In grade 2 lesions, there was an increased neutrophilic inflammatory component, either within the subepithelial stroma and infiltrating the mucosal epithelium or mixed with mucus and debris in the nasal cavity lumen (mucopurulent exudate), with minimal infiltration or disruption of the mucosal epithelium. Grade 3 was characterized by epithelial damage affecting 10% of the tissue and consisting of dysplasia or early squamous metaplasia with loss of cilia, epithelial, or goblet cell hyperplasia or intranuclear inclusion bodies. The accompanying inflammatory infiltrate varied in intensity, frequently heavily infiltrating and disrupting the epithelium, and was generally neutrophilic or mixed. If present, epithelial necrosis was superficial or limited to few foci affecting 10% of the tissue. Sections of nasal turbinate were considered to have a grade 4 if approximately 10 to 50% of the mucosal epithelium was necrotic, lost (ulceration), or replaced by a mixture of inflammatory cells, fibrin, karyorrhectic cellular debris, and sloughed epithelial cells. In grade 5 lesions, 50 to 100% of the mucosal epithelium was similarly affected. Conjunctivitis was subjectively graded 0, 1, or 2 on the basis of severity. Lesions in bulbar and palpebral conjunctiva were graded 0, 1, or 2 on the basis of severity. Conjunctivae with grade 0 had no or low numbers of lymphocytes and plasma cells scattered throughout the subepithelial stroma that were considered to be within normal limits. Grade 1 lesions also had neutrophils that transmigrated the conjunctival epithelium with or without lymphofollicular hyperplasia. In grade 2 lesions, the moderate to severe neutrophilic and lymphoplasmacytic infiltrate was accompanied by more significant conjunctival epithelial changes, including erosions or ulcerations, dysplasia, and squamous metaplasia. IHC. Immunohistochemistry (IHC) for FHV-1 was performed on serial sections of rostral nasal turbinates in cats qpcr positive for FHV-1. IHC for detection of FHV-1 was also performed on one section each of affected nasal planum/nares (case 22) and tonsil (case 21) from two cats with inflammation in these tissues. Immunohistochemistry for FCV was performed on sections of rostral nasal turbinates from a subset of 5 cats (cases 5, 8, 10, 11, and 17) that were FCV PCR positive, and IHC for both FHV-1 and FCV was performed on sections in which the nasopharynx was inflamed (cases 10 and 11) and where glossal ulcers were present (cases 1 and 10). The primary antibodies were mouse monoclonal antibodies 7-7 (FHV) and S1-9 (FCV) (Custom Monoclonal Antibodies International, West Sacramento, CA). Deparaffinized sections were rehydrated, and endogenous peroxidase activity was blocked in 0.3% H 2 O 2 in methanol for 30 min. Antigen retrieval was achieved by pretreatment with proteinase K (Dako, Carpinteria, CA) for 7 min at room temperature. After a blocking step with a serum-free Dako protein blocker, the primary antibodies were applied at a dilution of 1:400 in Dako diluent for 120 min (FHV-1) or 1:100 in Dako diluent for 60 min (FCV) at room temperature. Negative controls were prepared by omitting the primary antibody. The secondary antimouse antibody was developed with Dako EnVision system horseradish peroxidase-labeled polymer. The sections were counterstained with hematoxylin (Gill s formula 3; Fischer, Pittsburg, WA) and sealed with Immunomount (ThermoShandon, Pittsburgh, PA), and a coverslip was placed. Real-time quantitative PCR. (i) Validation. Real-time PCR tests, including a previously published feline herpesvirus type 1 real-time PCR (18), were developed according to an industry standard for probe-based real-time PCR. The validation criteria included amplification efficiency (required, 95%), dynamic range (at least 6 orders of magnitude), analytical sensitivity (at least 10 molecules per PCR; within-run and in-between-run reproducibility CP values, 3% and with linearized values of 20%), r-square value of the curve (0.993 or better), a signal-to-noise ratio of the fluorophore release in positive PCRs (10-fold), and confirmation of analytical specificity by resequencing positive clinical material using external sequencing primers. All real-time PCRs passed these criteria to ensure uniform performance characteristics necessary to run PCR tests in parallel for individual samples. (ii) Assays. Real-time PCR assays from IDEXX Laboratories (test code 2512; RealPCR FURD panel) targeting 5 infectious agents that potentially contribute to feline upper respiratory disease (FURD) were used. These infectious agents are Bordetella bronchiseptica, Chlamydophila felis, FHV-1, FCV, and Mycoplasma felis. All assays were designed and validated according to industry standards (User Bulletin 3; Applied Biosystems). Target genes for each application were as follows: for B. bronchiseptica, hemagglutinin fusion protein gene (FhaB; GenBank accession no. AF140678); for C. felis, outer membrane protein A (OmpA; GenBank accession no. AP006861); for feline herpesvirus type 1, glycoprotein B (18); for feline calicivirus, ORF 1 (GenBank accession no. AF109465); and for M. felis, single-stranded rrna internal transcribed region 1 (ITS-1; GenBank accession no. AF443608). RESULTS Shelter cat population. During the study period, 60 cats were sampled within 24 h of arrival at the shelter (at intake) and 79 cats with clinical signs of URI were sampled. There were no differences in pathogen prevalence in cats sampled at intake and cats sampled with clinical URI disease except for FHV-1. Shelter cat population results are compared to those for the pathology study group in Fig. 2.

2456 BURNS ET AL. J. CLIN. MICROBIOL. FIG. 2. Presence of pathogens by qpcr. Pathogens of the shelter cat population were monitored by PCR testing over the study period. The study group of cats was asymptomatic at intake and sampled within 24 h of arrival at the shelter (n 60). The study group URI cats were shelter cats with clinical URI disease (n 79). FHV-1 was the most commonly detected pathogen. The subset of URI cats that were submitted to the necropsy service (Pathology/URI) harbor a representative distribution of detectable pathogens. Pathology study group. Over a 9-month period, 22 cats euthanized with clinical signs of URI were necropsied. History, signalment, and clinical data of the cats are summarized in Table 1. There were equal numbers of females and males, and 4 of the males were castrated. Seven cats were juveniles ( 6 months old), with the youngest being approximately 10 weeks old, and 15 were adults ( 1 yr), with the maximum known age being 8 years. Fifteen of 22 cats were vaccinated at intake with a subcutaneous injection of Merial PUREVAX feline 3 containing modified live FHV-1, FCV, and feline panleukopenia virus. Cats were in the shelter an average of 7.7 5.8 days (range, 0 to 29 days) prior to the onset of clinical signs and were typically euthanized 1 to 6 days (average, 2.3 1.8 days) after onset. Seven of the cats were treated with doxycycline, Case Age In shelter to onset of clinical signs No. of days: With signs before euthanasia TABLE 1. History, signalment, and clinical data of the cats Vaccination at intake a Antibiotic treatment a and one was treated with amoxicillin-clavulanic acid (Clavamox). Clinical signs. In order of frequency, clinical signs included sneezing (15 cats), serous to mucopurulent nasal discharge (10 cats), serous to mucopurulent ocular discharge (8 cats), head shaking (5 cats), harsh lungs sounds (2 cats), and hypersalivation (1 cat). Pathological findings. Gross lesions included focal ulcerations of the tongue or mucocutaneous junctions (5 cats), enlargement of the retropharyngeal lymph nodes (4 cats), mild crusting at the medial canthi of the eyes and around the nares (2 cats), and loss or red to red-brown discoloration of the nasal turbinates (2 cats). Because histologic changes were consistently (22/22) and often exclusively (9/22) seen in the nasal cavity, a grading scheme for this tissue was designed, and these lesions were graded on the basis of the type and extent of inflammation and degree of epithelial changes (Fig. 1). In all 22 cases, rhinitis was present bilaterally and occurred concurrently in both the rostral (rostral maxillary turbinates) and caudal (caudal maxillary, ethmoid turbinates) portions of the nasal cavities, with no difference in the grade of inflammation between these regions. The two cats negative for FHV-1 by qpcr had lowgrade (grade 1 or 2) rhinitis. Among the 20 cats qpcr positive for FHV-1, 16 had severe rhinitis with ulceration (grade 4 or 5), 2 cats had moderate (grade 3) rhinitis, and 2 had low-grade (grade 1 or 2) rhinitis. The presence of two or more pathogens did not correlate with a higher grade of rhinitis since cats that were qpcr positive for FHV-1 alone had an average maximum grade of 4.45 ( 1.2), which is higher than the average maximum grade in cats that were qpcr positive for one or more pathogens in addition to FHV-1 (3.89 1.1). Intraepi- PCR positive results C T value for FHV-1 Rhinitis grade Rostral Cauda Conjunctivitis grade 1 3 mo 8 5 Y Y FHV-1 27.97 4 3 0 2 1 yr 5 4 Y Y FHV-1 26.76 5 5 0 3 4 mo 11 1 Y N FHV-1 29.04 1 1 0 4 1 yr 29 1 N N M. felis 40 1 2 1 5 1 yr 12 2 N N FHV-1, FCV 33.13 3 3 0 6 4 mo 0 6 N N FHV-1, C. felis, M. felis 20.61 4 4 2 7 1 yr 9 4 N N FHV-1, M. felis 25.19 4 4 0 8 6 mo 11 1 Y N FHV-1, FCV 27.69 4 4 1 9 1 yr 7 1 N N FHV-1 27.52 5 3 0 10 1 yr 7 1 N N FHV-1, FCV, M. felis 25.58 5 3 1 11 1 yr 0 7 Y Y FCV 40 2 2 2 12 1 yr 6 1 Y Y FHV-1 21.22 5 4 1 13 6 mo - Y Y FHV-1 23.38 5 5 2 14 6 mo 5 2 Y Y FHV-1 20.57 5 4 1 15 1 yr 5 2 Y Y FHV-1 21.03 5 4 0 16 8 yr 11 1 Y N FHV-1, M. felis 37.51 3 3 0 17 4 mo 5 2 Y Y FHV-1, FCV 35.27 2 2 0 18 1 yr 6 1 N N FHV-1, M. felis 27.10 5 5 2 19 1 yr 7 2 Y N FHV-1 28.01 5 5 1 20 4 yr 7 1 Y N FHV-1 22.83 3 3 1 21 1 yr 5 2 Y N FHV-1, M. felis 23.62 4 4 2 22 1 yr 5 2 Y N FHV-1 28.54 3 3 1 a Y, yes; N, no.

VOL. 49, 2011 HISTOLOGIC AND MOLECULAR CORRELATION OF FELINE URI 2457 thelial, intranuclear inclusion bodies were seen in 80% (16/20) of FHV-1-positive cases. Other changes in the nasal cavities included bony remodeling of the osseous cores of the nasal turbinate and fibroplasia. Bony remodeling occurred only in FHV-1-positive cats. Both productive and resorptive bone changes were present in all seven of the cats that were juvenile ( 6 months old). In adult cats, bone changes were present only in cats with high-grade rhinitis and included segmental loss of turbinate bone with osteoclastic activity (Fig. 3D, inset) and no detectable boney production. In both young and adult cats with bone changes, the cambium layer around the central bone was variably to markedly thick (Fig. 3D). Nine of 10 cats with lesions in the stratified nonkeratinized and keratinized squamous epithelium of the nares, nasal valve, or mucocutaneous junctions were qpcr positive for FHV-1, and the presence of herpesviral antigen in this region was confirmed by intranuclear inclusions or by IHC (case 22; Fig. 3C, with inset). One cat (case 4) had a chronic mucosal ulcer- ation of the lip that had marked eosinophilic inflammation in the underlying submucosa. This cat was negative for all tested pathogens except M. felis. Histologic lesions were seen in the tongues of two cats. Case 1 had a small focus of full-thickness epithelial necrosis, and case 10 had a grossly visible chronic ulceration on the lingual frenulum. In both cases, the surrounding epithelium was FHV-1 positive by IHC and FCV negative by IHC. All of the cats positive for M. felis and/or FCV in which nasopharyngeal tissue was examined (4 and 3 cases, respectively) had neutrophilic and lymphoplasmacytic cellular infiltrates of the pharyngeal mucosa; however, all but one of these cats (case 11) were additionally qpcr positive for FHV-1. Inclusions and FHV-1 antigen were detected in this region in only one case (case 21), and no nasopharyngeal inflammation was seen in examined sections from cats that were positive for FHV-1 alone. Nonetheless, we cannot establish causation by FCV or M. felis in these cases; IHC for FCV in cases 10 and 11 was negative in this and other tissues. Lesions in the tonsils, consisting of multifocal epithelial cellu- FIG. 3. (A) Tonsil, H&E staining: segmental, necrotizing tonsillitis. (B) Tonsil, immunohistochemistry, anti-fhv-1 antibody. Virus is abundant within the cytoplasm of the tonsillar epithelium. (C) Nasal valve, H&E staining: segmental region of acute epithelial necrosis predominantly affecting the basilar epithelium. Epithelial cells are isolated from neighboring cells and many contain intranuclear inclusions. (Inset) Inclusions are present in the basilar epithelium. (D) Nasal turbinate, H&E staining: there is intense submucosal inflammation, and a wide cambial layer surrounds residual fragments of bone. (Inset) Osteoclastic resorption.

2458 BURNS ET AL. J. CLIN. MICROBIOL. lar swelling, necrosis, or loss or moderate infiltration with neutrophils, were present in 7/19 cases. In two cases, the presence of FHV-1 antigen was confirmed by IHC (Fig. 3A and B). Both proximal (level of the thyroid glands) and distal (level of carina) sections of trachea were examined from each cat. No significant lesions were seen in the proximal trachea. Two cats (qpcr positive for FHV-1) had distal tracheal lesions. The tracheal epithelium was multifocally disrupted with loss of cilia, individual cell necrosis, and infiltration with small numbers of neutrophils and lymphocytes. The retropharyngeal lymph nodes exhibited mild to moderate reactive hyperplasia in 15 cats. All sections of lung were unremarkable except for mild interstitial lymphoid hyperplasia around vessels and bronchi. Sections of palpebral conjunctiva from the lower and third eyelids of both eyes were examined in each cat. Lesions were most consistently present on the bulbar surface of the third eyelid, as opposed to its palpebral surface or on the lower palpebral conjunctiva itself, although lesions were usually present in all areas. Changes specific for FHV-1 (inclusions) were seen in only a single case (case 6). Eleven cats were graded 1 or 2, and all but one was qpcr positive for a pathogen commonly associated with feline conjunctivitis, i.e., FHV-1, M. felis, and C. felis. The average conjunctival grade was higher in cats with a positive qpcr result for M. felis or for both M. felis and C. felis at 1.14 ( 0.9) (median, 1) than both the overall average grade (0.73 0.83) (median, 0.5) and the average of cats with a negative result for these pathogens (0.53 0.74) (median, 0). Three cats reported to have ocular discharge had grade 0 conjunctival lesions, and all had received antimicrobial treatment. Immunohistochemistry. One cat (case 5) that was qpcr positive for FHV-1 was not immunoreactive for FHV-1. In IHC-positive cats, immunostaining was primarily intracytoplasmic in epithelial cells within acutely necrotic areas and within epithelial cells at the periphery of ulcerated regions. Areas of positive immunostaining varied from a few small, surface clusters of epithelial cells to large confluent regions of intact and necrotic epithelium. Some areas with complete loss of epithelium were negative, although strong, granular positive immunostaining was often present in neutrophils and admixed with other inflammatory cellular debris in necrotic areas and the nasal cavity lumen. IHC for FCV was performed on sections of rostral nasal cavity and nasopharynx from the 5 cats that were positive for FCV by qpcr. All of these were negative by IHC. Quantitative real-time PCR. From the pooled oropharyngeal and conjunctival swabs, feline herpesvirus type 1 was the most common pathogen detected (91%, 20/22), followed by M. felis (32%, 7/22), feline calicivirus (23%, 5/22), and C. felis (5%, 1/22). None of the cats were positive by PCR for B. bronchiseptica. Nine cats (41%) were positive for two or more pathogens. None of the bacterial pathogens (M. felis, C. felis, or B. bronchiseptica) were detected by any method in cats treated with antibiotics. Because significant and specific histopathologic findings in this group of cats were exclusively referable to FHV-1 infection, we correlated the raw data (cycle threshold [C T ]) values obtained from the qpcr for FHV-1 with the histologic grade of the nasal turbinates. For the 20 cases that tested positive by qpcr for FHV-1, the nasal grades for rostral and caudal were tallied for each case and compared to the FHV-1 C T value. There was a significant inverse association (r 0.56; P 0.009, computed using the Pearson correlation in the Excel program) between the FHV-1 C T value and histological grade of inflammation. These results indicate that as C T values increased (i.e., less virus was present), the histologic grade was correspondingly lower. The test coefficient was used to compare FHV-1 qpcr results to histopathological FHV-specific lesions ( 0.62 [95% confidence interval, 0.16, 1]; P 0.026), indicating a significant positive correlation (agreement) between the presence of FHV-specific lesions and a positive qpcr result for FHV-1. All cats with grade 4 or 5 rhinitis had a CP value for FHV-1 of 28, while all those with a grade of 2 had a CP value of 29. DISCUSSION Multiple pathogens, singly or in combination, and a multitude of host and environmental factors can cause and sustain upper respiratory infection in a feline population. Primers designed to identify FHV-1, FCV, M. felis, C. felis, and B. bronchiseptica are commonly used in real-time PCR panels designed to identify potential causative agents in feline respiratory disease cases. A limitation in the value of identifying the presence of nucleic acid by qpcr is that these pathogens can be, and often are, present in the absence of disease; therefore, detection alone does not imply a disease state and in diseased cats does not determine causation. In this study we have correlated quantitative detection of pathogens with pathological findings in cases of feline upper respiratory disease. This study was designed as part of a large study designed to evaluate whether URI is associated with selected environmental risk factors, and the prevalence and spectrum of detectable pathogens in the background population of URI and unaffected (intake) cats are presented in Fig. 2. The cats that were necropsied for this study were from a single shelter during the 9-month period of time and were a representative subset of cats that had clinically defined URI disease. Among the cats chosen for necropsy, 20/22 had entered the shelter without clinically detectable URI, and either the source of pathogen or the trigger for recrudescence of latent infection (FHV-1) was assumed to be the shelter environment. Regardless, given the turnover and euthanasia practices of this single open-intake shelter, we were primarily examining cats recently introduced to the shelter (mean 8 days from intake; Table 1). Euthanasia in all examined cases occurred within a week of the onset of clinical signs. Clinical presentation commonly included sneezing (15/22 cats) but otherwise varied substantially (lung sounds, nasal discharge, ocular discharge). Regardless of clinical findings, in a complete set of upper and lower respiratory tract tissues that were collected, the gross and histologic changes were limited to the upper respiratory tract, and acute inflammation was common only in the nasal cavity (Table 1). We had anticipated variability in the pathogens causative for URI during this time frame; however, in this study, by PCR analysis, FHV-1 was both the most commonly identified agent (20/22 cats) and, moreover, was present in all cases of coinfection (9/22 cats). The next most common agents were M. felis (7/22) and FCV (5/22), and 2 cats, one each, were singly infected with M. felis or FCV. In these two cases, lesions were nonspecific, with no lesion specifically

VOL. 49, 2011 HISTOLOGIC AND MOLECULAR CORRELATION OF FELINE URI 2459 attributable to these agents, no visible bacteria, and no detectable FCV by immunohistochemistry. No cats treated with antibiotics had Mycoplasma or Chlamydophila detectable by PCR, and it is possible that these bacteria were present and contributory to the lesions prior to antibiotic treatment. During this time period, at this shelter, acute rhinitis was attributable to infection with FHV-1, as suggested by the prevalence of respiratory pathogens detected at intake in cats without clinical signs compared to that in cats with clinical signs (Fig. 2) and confirmed by histopathology (pathology study group). Cats infected with FHV-1 alone had an equal or higher histologic grade of rhinitis than cats infected with FHV-1 and a copathogen or pathogens. There was no difference in the presence or severity of FHV-1 lesions in cats that were or were not vaccinated at intake in this study. Therefore, no influence of the subcutaneously administered modified live vaccine, either in prevention or in induction of disease, was detected. The vaccination status of the cats prior to entry into the shelter was unknown. Because the PCR primers used did not differentiate between vaccine and wild-type strains, no conclusions on the source of the FHV-1 infection in these cases could be made. Lesions attributable to FHV-1 infection were rhinitis, tonsillitis, and glossitis. The respiratory epithelium overlying the maxillary turbinates was consistently targeted, with a subset of cats having inflammation or ulceration that extended rostrally to the nasal vestibule or caudally to the ethmoid turbinates. FHV-1 infection was identifiable histologically by the presence of inclusions (Fig. 2C) and/or by immunohistochemistry (e.g., Fig. 3B) in 19/20 cases. By immunohistochemistry, virus was present in the nasal cavity in epithelial cells bordering regions of necrosis and within the sloughed necrotic debris. Virus was also present in segmental stretches of tonsillar epithelium. Because alterations in nasal turbinate architecture are cited as a potential predisposition to chronic inflammation, we were interested in whether we could identify boney changes in infected cats. Turbinate bone remodeling was both proliferative and resorptive in the 7 ( 6-month-old) young cats both within regions and remote from regions of inflammation. In adult cats, high-grade (grade 4/5) inflammation was associated with segmental resorption to complete absence of bone, without boney production. All cats with turbinate destruction were FHV-1 positive; however, whether this was solely a consequence of inflammation or was in association with the virus is speculative. Additional studies would be necessary to determine if severity of disease or age of infection correlates with development of chronic sequelae. Quantitation of FHV-1 by qpcr was significantly associated with the presence and grade of rhinitis. The presence of other pathogens detected by qpcr did not correlate with consistent or specific histologic lesions. In the cases where FHV-1 and FCV were concurrently detected by qpcr, only FHV-1 was detected by immunohistochemistry. Quantitation of FHV-1 with correlation to disease using virus isolation, virus replication, and immune reaction has been published (10, 18). In one study (10), the amount of virus was used to delineate between the different clinical and histologic parameters. A C T value of 28 which differentiated between cats with and without FHV-1 lesions in the upper respiratory tract was established. This value also differentiated virus isolation-positive from virus isolation-negative cats. This observation suggested that in virus isolation-positive cats the virus was actively replicating and recoverable in cell culture, while in the virus isolation-negative cats, virus was not present or, if present, not able to replicate. The fact that qpcr was able to detect FHV-1 over 8 more weeks suggested that latent virus was shed in low numbers which did not trigger virus isolation to come up positive. Moreover, this study showed that only the FHV-1 RNA-positive cats (indicating replicating virus) showed significantly increased mrna transcriptional activity for immune cytokines, indicating that the immune mechanisms in the tissue responded to current infection and generation of non-self-antigens processed by immune cells. While these two studies were based on clinical, tissue culture, and immunological observations and correlating these findings to a particular amount of FHV-1 DNA or presence of FHV-1 RNA, they did not prove causation of the observations by a current FHV infection. The current study takes the mechanistic correlation of different observations one step further and links the presence of a particular amount of FHV-1 DNA with the existence of FHV-1 pathognomonic inclusions in the tissue and high-grade histologic changes, therefore establishing a causative link between the presence of FHV-1 and tissue damage causing clinical signs of URD. The use of quantitative real-time PCR results to delineate treatment strategies or prognosis is well known and used in human diagnostic medicine, in particular, for HIV-1 and hepatitis C virus. In particular, herpes simplex virus (HSV) infections that cause lower respiratory infections in humans are probably the closest example to FHV-1 infections that cause upper respiratory disease in cats. Gooskens et al. (7) used quantitative real-time PCR to establish HSV loads and were able to correlate the amount of virus with either mortality or survival within 28 days after hospital admission. In a different study using real-time PCR tests, including tests for HSV types 1 and 2, quantitative determination of viral load allowed early diagnosis of viral infections, monitoring of the kinetics of viral load during treatment, early recognition of antiviral resistance, and recognition of relapse before clinical signs occurred (19). In conclusion, our study has confirmed previously published observations about the link between FHV-1 DNA amount and clinical parameters, viral replication, isolation in cell culture, immune reaction, pathognomonic inclusions, and histological grades for tissue damage. qpcr therefore could be a key tool to differentiate between active and latent FHV-1 infection in cats and to determine if FHV-1 is the cause of URI in individual cats. ACKNOWLEDGMENT This project was supported by the Center for Companion Animal Health, School of Veterinary Medicine, University of California, Davis. REFERENCES 1. Bannasch, M. J., and J. E. Foley. 2005. Epidemiologic evaluation of multiple respiratory pathogens in cats in animal shelters. J. Feline Med. Surg. 7:109 119. 2. Burgesser, K. M., et al. 1999. Comparison of PCR, virus isolation, and indirect fluorescent antibody staining in the detection of naturally occurring feline herpesvirus infections. J. Vet. Diagn. Invest. 11:122 126. 3. Coyne, K., et al. 2006. Recombination of feline calicivirus within an endemically infected cat colony. J. Gen. Virol. 87(Pt 4):921 926. 4. Di Francesco, A., S. Piva, and R. Baldelliw. 2004. Prevalence of Chlamydophila felis by PCR among healthy pet cats in Italy. New Microbiol. 27:199 202.

2460 BURNS ET AL. J. CLIN. MICROBIOL. 5. Di Martino, B., C. E. Di Francesco, I. Meridiani, and F. Marsilio. 2007. Etiological investigation of multiple respiratory infections in cats. New Microbiol. 30:455 461. 6. Egberink, H., et al. 2009. Bordetella bronchiseptica infection in cats. ABCD guidelines on prevention and management. J. Feline Med. Surg. 11:610 614. 7. Gooskens, J., et al. 2007. Quantitative detection of herpes simplex virus DNA in the lower respiratory tract. J. Med. Virol. 79:597 604. 8. Haesebrouck, F., L. A. Devriese, B. van Rijssen, and E. Cox. 1991. Incidence and significance of isolation of Mycoplasma felis from conjunctival swabs of cats. Vet. Microbiol. 26:95 101. 9. Helps, C. R., et al. 2005. Factors associated with upper respiratory tract disease caused by feline herpesvirus, feline calicivirus, Chlamydophila felis and Bordetella bronchiseptica in cats: experience from 218 European catteries. Vet. Rec. 156:669 673. 10. Johnson, L. R., and D. J. Maggs. 2005. Feline herpesvirus type-1 transcription is associated with increased nasal cytokine gene transcription in cats. Vet. Microbiol. 108:225 233. 11. Kang, B. T., and H. M. Park. 2008. Prevalence of feline herpesvirus 1, feline calicivirus and Chlamydophila felis in clinically normal cats at a Korean animal shelter. J. Vet. Sci. 9:207 209. 12. Maggs, D. J., and H. E. Clarke. 2005. Relative sensitivity of polymerase chain reaction assays used for detection of feline herpesvirus type 1 DNA in clinical samples and commercial vaccines. Am. J. Vet. Res. 66:1550 1555. 13. Pedersen, N. C., R. Sato, J. E. Foley, and A. M. Poland. 2004. Common virus infections in cats, before and after being placed in shelters, with emphasis on feline enteric coronavirus. J. Feline Med. Surg. 6:83 88. 14. Radford, A. D., et al. 2009. Feline calicivirus infection. ABCD guidelines on prevention and management J. Feline Med. Surg. 11:556 564. 15. Randolph, J. F., et al. 1993. Prevalence of mycoplasmal and ureaplasmal recovery from tracheobronchial lavages and of mycoplasmal recovery from pharyngeal swab specimens in cats with or without pulmonary disease. Am. J. Vet. Res. 54:897 900. 16. Thiry, E., et al. 2009. Feline herpesvirus infection. ABCD guidelines on prevention and management. J. Feline Med. Surg. 11:547 555. 17. Veir, J. K., R. Ruch-Gallie, M. E. Spindel, and M. R. Lappin. 2008. Prevalence of selected infectious organisms and comparison of two anatomic sampling sites in shelter cats with upper respiratory tract disease. J. Feline Med. Surg. 10:551 557. 18. Vögtlin, C., et al. 2002. Quantification of feline herpesvirus-1 DNA in ocular samples of clinically diseased cats by real-time TaqMan PCR. J. Clin. Microbiol. 40:519 523. 19. Watzinger, F., et al. 2004. Real-time quantitative PCR assays for detection and monitoring of pathogenic human viruses in immunosuppressed pediatric patients. J. Clin. Microbiol. 42:5189 5198. 20. Wills, J. M., T. J. Gruffydd-Jones, S. J. Richmond, R. M. Gaskell, and F. J. Bourne. 1987. Effect of vaccination on feline Chlamydia psittaci infection. Infect. Immun. 55:2653 2657. 21. Zicola, A., C. Saegerman, D. Quatpers, J. Viandier, and E. Thiry. 2009. Feline herpesvirus 1 and feline calicivirus infections in a heterogeneous cat population of a rescue shelter. J. Feline Med. Surg. 11:1023 1027. Downloaded from http://jcm.asm.org/ on October 10, 2018 by guest