Thoracic radiographic findings in cats with feline infectious peritonitis

Abstract

Objectives

The objective of this study was to describe thoracic radiographic findings and associated histopathological changes (where available) in cats with feline infectious peritonitis (FIP).

Methods

This was a retrospective descriptive study. Cats were included if they had a definitive diagnosis of FIP (based on histopathology and immunohistochemistry) or a presumptive diagnosis of FIP (based on case review by two veterinary internists), and contemporaneous orthogonal thoracic radiographs. Radiographs were reviewed retrospectively by a veterinary radiologist and veterinary radiology resident and assessed for the following: presence of pleural space disease; unstructured interstitial, bronchial, alveolar and/or nodular pulmonary patterns; lymphadenopathy; and cardiovascular abnormalities. Archived histopathologic specimens were reviewed by a veterinary pathologist.

Results

In total, 35 cats were included: 18 with definitive FIP and 17 with presumptive FIP. Radiographs were abnormal in 32/35 cats and normal in 3/35 cats. Pleural effusion was present in 13/35 cats and was either bilateral (11/13) or unilateral (2/13) in distribution. The lungs were radiographically abnormal in 25/35 cats, with the most common abnormality being an unstructured interstitial pattern (21/25), with bronchial (11/25) and alveolar (10/25) patterns less common. Pulmonary nodules were identified in 3/25 cats. Mixed pulmonary patterns were frequent (18/25). Sternal lymphadenopathy was present in 16/35 cats. An enlarged cardiac silhouette was noted in 6/35 cats, attributable to myocarditis (3/6), pericardial effusion (1/6), a high output state (1/6) or unrelated cardiomyopathy (1/6). Common histopathologic lesions included pulmonary edema (16/17), fibrinosuppurative pleuritis (13/17) and histiocytic vasculitis causing pneumonia (10/17); myocarditis (5/14); and lymphadenitis (2/2). Histologic lung changes were more common in patients with moderate to severe radiographic pulmonary changes.

Conclusions and relevance

Thoracic radiographic findings in cats with FIP may include variably distributed pleural effusion; interstitial, bronchial, and/or alveolar patterns; pulmonary nodules; lymphadenopathy; and cardiomegaly. FIP should be considered in cats with these radiographic changes and supportive clinical findings.

Introduction

Feline infectious peritonitis (FIP) is caused by a feline coronavirus (FCoV) and is a leading cause of death in juvenile cats. The most common form of FCoV is the ubiquitous enteric biotype (FECV).^{1 –5} FECV infects intestinal enterocytes primarily and is often asymptomatic or causes mild gastrointestinal signs.^4,5 In a small percentage of cats, FECV mutates in to a virulent biotype and leads to FIP.^{1 –5} Mutations that change the tropism of the virus from enterocytes to macrophages help facilitate the development of FIP, as viral replication in macrophages and monocytes allows the virus to spread systemically and, in combination with host immune responses, results in a severe, multisystemic, pyogranulomatous vasculitis.^{6 –11}

Two forms of FIP are described: the effusive or ‘wet’ form, causing pleural, peritoneal and/or pericardial effusion, and the non-effusive ‘dry’ form, which often results in non-specific illness with compromise of multiple organ systems.^{8,12 –19} A definitive diagnosis of FIP routinely requires histopathological examination of affected tissues and detection of mutated FCoV (FIPV) antigen by immunohistochemistry (IHC).^3,20,21 Making a definitive diagnosis without histopathology and IHC can be difficult, particularly in cats without effusion. Because of this, a presumptive diagnosis of FIP is frequently based on patient signalment, clinical history, bloodwork findings, imaging and exclusion of other disease processes that may present similarly.^3,22 Laboratory changes supportive of FIP include non-regenerative anemia, microcytosis, neutrophilia, lymphopenia, hyperglobulinemia, hypoalbuminemia, a low albumin:globulin ratio and hyperbilirubinemia.^22,23 Effusions in cats with FIP typically have high protein concentrations and low or moderate cell counts. ²⁴ RT-PCR for FCoV RNA can be performed on effusions but cannot differentiate between FECV and FIPV; thus, a positive result is supportive of a diagnosis of FIP but does not provide a definitive diagnosis. ²¹

Findings on imaging studies can also support a diagnosis of FIP. Central nervous system abnormalities, such as meningeal or ependymal contrast enhancement and ventriculomegaly, have been reported in cats with neurological involvement, and transabdominal ultrasonography may reveal effusion, lymphadenopathy and abnormalities in the gastrointestinal tract, liver, spleen or kidneys.^14,17,18,25 Pleural effusion is routinely noted in cats with effusive FIP,^12,13 and a few case reports have described thoracic radiographic findings, including pleuropneumonia and pyogranulomatous pneumonia in two individual cats and pericardial effusion in an additional case.^{26 –28} The purpose of this study was to describe thoracic radiographic findings in cats with a definitive or presumptive diagnosis of FIP.

Materials and methods

Study design

A retrospective descriptive study design was utilized. The sample period was 2007–2022. Retrospective evaluation of patient medical records was permitted through informed consent signed by each client upon admission. Approval by the institutional Animal Care and Use Committee was not needed.

Case selection

The electronic medical record system at Texas A&M Veterinary Medical Teaching Hospital was used to identify cats with a definitive or presumptive diagnosis of FIP. This was performed by searching the master problem list and free text in all documents, including client communications, for the terms ‘feline infectious peritonitis’ and ‘FIP’. A search of the institution’s veterinary pathology database utilizing the same terms identified cats diagnosed with FIP via IHC.

Medical records were initially reviewed by a diagnostic imaging resident (KR). The extracted data for all patients included the following: breed; date of birth; sex; body weight; age; presenting clinical signs; duration of clinical signs; presence of fever; respiratory rate and effort; presence of a heart murmur or gallop rhythm; echocardiographic findings, if performed; fluid analysis and cytology of pleural effusion, if performed; and histopathology and IHC results, if performed. Cats with histopathologic findings consistent with FIP and positive IHC for FIPV were classified as definitive cases.

For the remaining cats without confirmatory histopathology and IHC, the following additional data were recorded: presence of chorioretinitis; presence of complete blood count abnormalities, including anemia, microcytosis, neutrophilia, lymphopenia or thrombocytopenia; presence of biochemical abnormalities, including hyperglobulinemia (globulin >4.5 g/dl), albumin:globulin ratio <0.4, hyperbilirubinemia or elevated liver enzymes; presence of abdominal ultrasonographic findings suggestive of infiltrative disease (eg, lymphadenopathy; liver, spleen or kidney enlargement; diffuse changes in echogenicity in the liver or spleen; presence of nodules in the liver or spleen; presence of a renal subcapsular hypoechoic rim; or presence of gastrointestinal mass lesions); cerebrospinal fluid or aqueous humor fluid analysis and cytology results, if performed; and results of RT-PCR for FCoV, if performed. Case information was reviewed independently by two ACVIM board-certified small animal internists (GA and AC), one of whom is additionally ABVP board certified in feline practice (AC). The extracted data were considered holistically, and the complete electronic medical record was subsequently reviewed if questions arose about the exclusion of other appropriate differential diagnoses. Cases were categorized on the preponderance of the available evidence as presumptive FIP or not. Cases considered to meet the criteria for a presumptive diagnosis of FIP by both internists were included in the final data analysis. Orthogonal (two- or three-view) thoracic radiographic studies performed within 7 days of presentation for clinical signs attributable to FIP were also required for inclusion in the study population.

Radiographic evaluation

Digital thoracic radiographs (Toshiba DC-1050F2; Canon Medical Systems or Del Medical CM-80; Del Medical Inc) were evaluated utilizing DICOM viewers (eFilm Workstation; Merge Healthcare or eUnity; Mach7 Technologies Canada Inc) by two observers, including one ACVR board-certified veterinary radiologist (CG) and one diagnostic imaging resident (KR). Each evaluator reviewed all images independently and recorded thoracic radiographic findings using a predesigned data collection form. When interpretation discrepancies were present, a consensus was reached between both observers.

The following data were recorded for each case: presence of unstructured interstitial, bronchial or alveolar pulmonary patterns and their subjective severity (mild, moderate, severe); distribution of pulmonary patterns (focal, multifocal, diffuse); number, size and distribution of pulmonary nodules/masses, if present; and presence, subjective severity and distribution of pleural fluid or gas. If pleural gas was present, history of thoracocentesis was recorded. The presence of sternal, cranial mediastinal or tracheobronchial lymph node enlargement and its subjective severity was also recorded. Sternal lymphadenomegaly was defined as a broad-based, ovoid soft tissue opacity dorsal to the second and third sternebrae in the lateral view. Cranial mediastinal lymphadenomegaly was defined as an ovoid, increased soft tissue opacity ventral to the trachea and cranial to the cardiac silhouette in the lateral view with or without concurrent soft tissue widening of the cranial mediastinum in the ventrodorsal view. Tracheobronchial lymphadenomegaly was defined as increased soft tissue opacity caudodorsal to and resulting in ventral displacement of the carina in the lateral view, with concurrent increased opacity between the principal bronchi, with or without lateral displacement of the principal bronchi, in the ventrodorsal view. The presence of cardiomegaly and its severity and sidedness (left, right or generalized) was determined subjectively. If pleural effusion precluded assessment of mediastinal structures, including the intrathoracic lymph nodes or cardiac silhouette, this was recorded. The presence of enlargement or tortuosity of pulmonary lobar arteries and veins was also recorded and was determined subjectively, with the cranial lobar pulmonary vasculature assessed in lateral views and caudal lobar vasculature assessed in ventrodorsal views.

Histopathology

In cats undergoing necropsy or surgical biopsy, tissues were processed for routine formalin-fixed paraffin-embedded histopathology. Institutional protocol was followed for sampling of organs on necropsy. For cases undergoing necropsy after 2011 (n = 13), visible gross lesions in the lungs and additional pieces of cranial and caudal lung (ideally all lobes) were sampled. In cases before 2011 (n = 4), visible lung lesions were sampled; if no lesions were visible, a single sample of lung from one lobe would be collected routinely. Tissues mentioned as being affected in radiology reports were typically included in the histopathologic assessment. All cases were reviewed blinded by a single veterinary anatomic pathologist (LB). Lung, heart and lymph node samples were evaluated where available, and pulmonary histopathologic lesions were subjectively scored. IHC targeting the nucleocapsid of FIPV types 1 and 2 was performed in all cases. Briefly, 4 μm formalin-fixed, paraffin-embedded sections were incubated with mouse pan monoclonal primary antibody (clone FIPV3-70; Invitrogen), developed with 3,3′diaminobenzidine and counterstained with hematoxylin.

Results

Cases, presenting signs, and clinical findings related to the thoracic cavity

The initial search identified 54 cats. Eight cats were excluded owing to lack of consensus among internists for a diagnosis of presumptive FIP, seven were excluded as a result of insufficient evidence for a diagnosis of presumptive FIP agreed upon by both internists, and four cats were excluded because of negative IHC results. A total of 35 cats, 18 with definitive FIP and 17 with presumptive FIP, met the inclusion criteria. The median age was 2.7 years (range 4.7 months to 14.6 years). The median weight was 3.2 kg (range 1.6–6.6 kg). The study population consisted of 22 castrated males, nine spayed females, three intact males and one intact female. Breeds included 17 domestic shorthairs, three domestic mediumhairs, three domestic longhairs, two Siamese, two Bengals, one Pixie-Bob, one Birman, one British Shorthair, one Cornish Rex, one Maine Coon, one Persian, one Ragdoll and one Scottish Fold. The most common presenting complaints described by the owners were decreased appetite (27/35, 77.1%), lethargy (21/35, 60.0%), weight loss (10/35, 28.6%), fever (8/35, 22.9%) and increased respiratory effort (7/35, 20.0%). The median duration of clinical signs before presentation was 14 days (range 1–56). Tachypnea was noted on physical examination in 15/35 (42.9%) cats and 6/35 (17.1%) cats were dyspneic. Heart murmurs were present in 5/35 (14.3%) cats and ranged from grades 1–2/6 to 4/6. Pleural fluid analysis with cytology was performed in seven cats. The fluid was classified as an exudate in five cats, with total protein in the range of 6.7–9.3 g/dl and total nucleated cell counts of 2600–50,000 cells/µl. Two cats were reported to have a modified transudate, with total protein of 2.6 and 5.0 g/dl and total nucleated cell counts of 201 and 1471 cells/µl, respectively.

Radiographic findings

Radiographic studies included three-view (28/35; 80.0%) and two-view (7/35; 20.0%) studies. Three-view studies consisted of right lateral, left lateral and ventrodorsal views in all 28 patients, and two-view studies consisted of right lateral and ventrodorsal views in all seven patients. A total of 32/35 (91.4%) cats had abnormal radiographs and 3/35 (8.6%) were radiographically normal. Pleural effusion was present in 13/35 (37.1%) cats and was bilaterally symmetric (9/13), bilateral but asymmetric (2/13) or unilateral (2/13) (Figure 1) in distribution. Unilateral pleural effusion was distributed in the right hemithorax in one cat and the left hemithorax in the other. Pleural effusion was characterized as mild in 7/13 cats, moderate in 4/13 cats and severe in 2/13 cats. A small-volume pneumothorax (presumably iatrogenic) was present in one cat with a history of recent thoracocentesis. A total of 7/13 cats with pleural effusion had mixed unstructured interstitial and alveolar patterns; the remaining concurrent lung changes included an unstructured interstitial pattern alone (2/13), interstitial and bronchial pattern (1/13), interstitial, bronchial and alveolar pattern with nodules (1/13) bronchial and alveolar pattern (1/13), and the remaining cat had no other radiographic abnormalities.

Figure 1

Ventrodorsal radiograph (presumed case). Radiographic findings for this case included unilateral right-sided pleural effusion and an interstitial and alveolar pattern in the right lung lobes. Given the large volume of effusion present, atelectasis of the right lung lobes is likely

The pulmonary parenchyma was abnormal in 25/35 (71.4%) cats and normal in 10/35 (28.6%) cats. The most common pulmonary pattern present was an unstructured interstitial pattern (21/25) (Figures 2–6) followed by a bronchial pattern (11/25) (Figures 2, 3, 5), an alveolar pattern (10/25) (Figure 5) and pulmonary nodules (3/25) (Figures 4–6). Pulmonary nodules were well defined in 2/3 cats and ill defined in 1/3 cats. All three cats with pulmonary nodules had multiple small nodules distributed throughout multiple lung lobes, with the largest nodule measuring 10 mm in one cat (Figure 6). Mixed pulmonary patterns were observed in 18/25 cats. The breakdown of pulmonary patterns was as follows: unstructured interstitial and alveolar pattern (8/25); unstructured interstitial and bronchial pattern (6/25) (Figures 2 and 3); unstructured interstitial pattern alone (4/25); bronchial pattern alone (3/25); unstructured interstitial pattern with nodules (2/25); a combination of all three pulmonary patterns with nodules (1/25) (Figure 5); and bronchial and alveolar patterns (1/25). Unstructured interstitial patterns were mild in 12/21 cats, moderate in 6/21 cats and severe in 3/21 cats. Bronchial patterns were either mild (9/11) or moderate (2/11). Alveolar patterns were graded as mild (7/10), moderate (1/10) or severe (2/10). The most common distribution of pulmonary patterns was diffuse (17/25), with interstitial patterns being diffuse in 16/21 cats and bronchial patterns being diffuse in 10/11 cats. The remaining interstitial patterns were characterized as multifocal (4/21) or focal (1/21), and the final bronchial pattern was characterized as multifocal (1/11). Alveolar patterns were either multifocal (8/10) or focal (2/10) in distribution. Lateralized pulmonary patterns occurred in 2/25 cases, both of which had unilateral pleural effusion.

Figure 2

Left lateral thoracic radiograph (confirmed case, cat 14). A moderate diffuse unstructured interstitial and mild bronchial pattern is present throughout the lungs

Figure 3

Left lateral thoracic radiograph (confirmed case, cat 12, same case as Figure 7a). Radiographic findings for this case included a moderate diffuse unstructured interstitial pattern and bronchial pattern, mild cardiomegaly, mild enlargement of pulmonary lobar arteries and mild pleural effusion

Figure 4

Right lateral thoracic radiograph (confirmed case, cat 3). Predominant radiographic findings for this case included a severe diffuse interstitial pulmonary pattern with an ill-defined miliary to nodular component and moderate cardiomegaly

Figure 5

Right lateral thoracic radiograph (confirmed case, cat 15). Findings for this case included a severe diffuse interstitial pattern with numerous small nodules throughout all lung lobes, mild diffuse bronchial pattern, mild ventral alveolar pattern and mild pleural effusion. Pleural effusion silhouettes with the cranial and ventral margins of the cardiac silhouette, resulting in border effacement in those regions

Figure 6

Right lateral thoracic radiograph (confirmed case, cat 7). There are multifocal well-defined pulmonary nodules, a mild diffuse interstitial pattern and moderate cardiomegaly in this patient

Sternal lymphadenomegaly was present in 16/35 (45.7%) cats. The region of the sternal lymph nodes could not be assessed in six cats secondary to the presence of pleural effusion. Cranial mediastinal lymphadenomegaly was present in 1/35 (2.9%) cats and was seen only in conjunction with sternal lymphadenomegaly. No cats had radiographic tracheobronchial lymphadenomegaly.

The cardiac silhouette was unable to be assessed in one cat owing to the degree of pleural effusion present. Radiographic cardiomegaly was present in 6/35 (17.1%) cats, with all cases characterized as generalized. Of these six cats, two had subsequent echocardiograms performed, which diagnosed pericardial effusion with tamponade in one and increased left ventricular internal dimensions and left atrial enlargement attributed to a high output state, given the presence of anemia, in the other. Of the six cats with cardiomegaly, four underwent necropsy and were diagnosed with myocarditis (n = 3), one of which was the cat diagnosed with pericardial effusion and tamponade on echocardiography, and an unrelated cardiomyopathy (n = 1). One cat with cardiomegaly did not have an echocardiogram or necropsy. Echocardiography was performed in an additional 3/35 cats as a result of the presence of a heart murmur and/or inability to assess the cardiac silhouette on radiographs due to silhouetting with pleural effusion. The remaining echocardiographic diagnoses included diffuse thickening of the atrial and ventricular walls with pericardial effusion (1/3), mild left ventricular hypertrophy and left atrial enlargement with pericardial effusion (1/3), and normal (1/3). Radiographic abnormalities of the pulmonary lobar vasculature were present in 5/35 (14.3%) cats, most frequently affecting the arteries (enlargement 4/5, tortuosity 1/5) with concurrent pulmonary venous enlargement in one case. Of these cats, three had necropsies performed, which revealed moderate to severe vasculitis in 3/3 cats and myocarditis in 2/3 cats.

Histopathologic results

Histopathology was available in 18/35 cases. Of the 35 cats, 17 were humanely euthanized and submitted for necropsy. For a single cat, only a mandibular lymph node biopsy was available for histologic examination as the client declined necropsy at the time of euthanasia. The lung was sampled for histopathology in 17 cats (cats 1–17), the heart in 14 cats (cats 1–14), tracheobronchial lymph node in one cat (cat 5) and mandibular lymph node in one cat (cat 18). Strong cytoplasmic immunolabeling was observed with IHC in all cats, confirming FIPV infection. Gross and histologic pathology varied considerably across cases. All cats had mild to moderate evidence of post-euthanasia alveolar proteinosis (barbiturate artifact). The most commonly observed pulmonary histologic lesions were edema (16/17), fibrinosuppurative pleuritis (13/17) and histiocytic vasculitis (10/17) causing pneumonia in the adjacent parenchyma (Figure 7). Nodules corresponded to moderate to severe granulomatous inflammation or necrotic foci histopathologically. One cat with histiocytic pneumonia due to FIP also had bronchial smooth muscle hypertrophy suggestive of concurrent asthma. Additional findings included myocarditis (5/14) and lymphadenitis (2/2). Two cats had evidence of cardiomyocyte degeneration without myocarditis, likely unrelated to FIP infection. Alveolar and perivascular macrophages were frequently positive for FIPV. Among the 4/17 necropsied cats without pulmonary radiographic abnormalities, mild to moderate perivascular edema, mild vasculitis or pleuritis were observed. Table 1 shows a summary of radiographic findings and histopathologic changes for each patient.

Figure 7

Selected pulmonary histopathology: (a) lung from cat 12 (same cat as in Figure 3), H&E, ×100 magnification, showing severe septal and alveolar edema and vasculitis; (b) cat 5, H&E, × 100 magnification, showing severe fibrinosuppurative pleuritis; (c) cat 16, H&E, × 200 magnification, demonstrating mild histiocytic pneumonia, type II pneumocyte hyperplasia and edema; (d) FIPV immunohistochemistry of cat 13, demonstrating strong cytoplasmic labeling of alveolar macrophages and necrotic cellular debris. FIPV = mutated feline coronavirus (the cause of feline infectious peritonitis); H&E = hematoxylin and eosin

Table 1

Selected radiologic and post-mortem findings in 18 cats diagnosed with feline infectious peritonitis

Number	Radiology						Histopathology
	Pleural effusion	Interstitial	Bronchial	Alveolar	Nodules	Cardiomegaly	Edema	Vasculitis	Pleuritis	Myocarditis
1	++	++	–	+	–	–	++	++	+++	–
2	–	–	–	–	–	–	++	–	–	–
3	–	+++	–	–	+	+	++	+++	+	+
4	–	–	–	–	–	–	–	–	–	–
5	+++	+	–	+++	–	+	++	–	+++	+
6	+	+	–	–	–	–	+	–	++	+
7	–	+	–	–	+	+	+++	+++	+++	–
8	++	+	–	+	–	–	+++	–	+	–
9	+	–	+	+	–	–	+	+	–	–
10	++	++	–	++	–	–	+	–	–	+
11	–	+	–	–	–	–	++	+	+	–
12	+	++	++	–	–	+	+++	++	++	+
13	++	+	–	+	–	–	+	+++	+	–
14	–	++	+	–	–	–	++	+++	+	–
15	+	+++	+	+	+	–	+	++	+	NA
16	–	–	–	–	–	–	+	+	+	NA
17	+	–	–	–	–	–	+	–	+	NA
18*	–	–	+	–	–	–	NA	NA	NA	NA

*

Only the mandibular lymph node was available for histopathology in this case

+

 = present or mild; ++ = moderate; +++ = severe; – = not present; NA = not applicable

Discussion

The objective of this retrospective study was to describe the thoracic radiographic changes in cats diagnosed with FIP. Radiographs were abnormal in 32/35 (91.4%) cats and normal in 3/35 (8.6%) cats. As expected, pleural effusion was a common finding, present in 13/35 (37.1%) cats in this study. Of the 13 cats with pleural effusion, 12 had pulmonary parenchymal abnormalities, including unstructured interstitial, bronchial, alveolar and nodular pulmonary patterns. The presence of pleural effusion confounds interpretation of the lung changes in these cats, specifically interstitial and alveolar patterns, owing to a reduction of pulmonary volume and summation of pleural fluid with the lungs. Atelectasis of lung parenchyma secondary to pleural effusion was likely a contributing factor to the interstitial and alveolar pulmonary patterns observed in patients with pleural effusion, but this depends on the volume of effusion present. For cats with small volumes of effusion, the pulmonary changes observed are more likely to represent concurrent underlying primary pulmonary pathology, while in cases with large volumes of effusion, it is likely that atelectasis was a major factor in the pulmonary patterns observed. Unilateral pleural effusion was present in two cats; unilateral distribution of pleural fluid can occur secondary to viscid exudative and/or inflammatory effusions, as occurs with FIP. ²⁹

Lung changes were the most common radiographic abnormality and were identified in 25/35 (71.4%) cats. In cats with abnormal lungs, the most common radiographic finding was an unstructured interstitial pulmonary pattern (21/25, 84.0%), followed by bronchial (11/25, 44.0%) and alveolar (10/25, 40.0%) patterns, and, finally, pulmonary nodules (3/25, 12%). A combination of pulmonary patterns was frequently observed (18/25, 72.0%) and most pulmonary patterns had a diffuse distribution (17/25, 68.0%). Common histopathologic changes observed in the lungs in this study included pulmonary edema, fibrinosuppurative pleuritis and histiocytic vasculitis causing pneumonia, similar to those found in another recent study. ³⁰ Nodules corresponded to moderate to severe granulomatous inflammation or necrotic foci histopathologically. Moderate to severe pulmonary radiographic lesions correlated with histopathology in our study, while mild radiographic changes were often not appreciable on histologic evaluation. It is possible that the milder pulmonary patterns may have been more likely to be associated with atelectasis considering this, and this should be taken into account when interpreting the data for the cases with mild pulmonary changes without histopathology. Some cats with radiographically normal pulmonary parenchyma, however, had mild perivascular edema and/or vasculitis on histopathology, highlighting that a lack of radiographic pulmonary abnormalities does not rule out mild lung involvement in FIP. One cat in our study had histologic evidence of concurrent asthma. Bronchial pulmonary patterns were fairly common in our patient population (11/25, 44.0%). Bronchial patterns may occur secondary to thickening of the bronchial wall or infiltration of cells or fluid within the immediate peribronchial space. ³¹ Given the histopathologic findings observed in many of our patients, pathology within the lung interstitium surrounding the bronchi may be an explanation for this radiographic finding. However, given the lack of histopathology in all cases, it is possible that other comorbidities, such as feline asthma, could have also contributed to bronchial patterns in a portion of our study population. The radiographic pulmonary changes described in this study align with previous case reports of cats with FIP. In one case report, a right caudal lung lobe mass was described, ²⁶ and in another, diffuse lung consolidation was seen, which was most severe in the caudal lung lobes. ²⁷ Similar findings were seen in our study population, with the exception of a single large mass in the pulmonary parenchyma of one cat. Instead, three cats in our study had multiple pulmonary nodules that were smaller in size, a finding that has not been described with FIP previously.

Thoracic lymph node enlargement was a common finding (16/35, 45.7%). All cases with lymphadenomegaly had sternal node enlargement and one case had concurrent cranial mediastinal node enlargement. As the sternal lymph nodes receive drainage from abdominal structures in addition to the thoracic wall, ³² it is possible that the high prevalence of sternal lymphadenomegaly could reflect concurrent pathology within the abdomen. The presence of intrathoracic lymphadenopathy may be underestimated in this study owing to the inclusion of cats with pleural effusion; silhouetting of pleural fluid with the mediastinal structures limits assessment of this region radiographically and was noted to be a limitation in six cats in this study. No cats had radiographically detectable tracheobronchial lymphadenopathy, although one cat did have lymphadenitis of the tracheobronchial lymph nodes on histopathology. Histopathologic lesions within the two lymph nodes examined in this study were both consistent with lymphadenitis, as would be expected with FIP.

Six cats had generalized radiographic cardiomegaly (6/35, 17.1%). Two cats with cardiomegaly had an echocardiogram performed. One of these cats had severe pericardial effusion with tamponade. The other cat was diagnosed with an increased left ventricular internal diameter and moderate left atrial enlargement attributed to a high output state, possibly due to anemia. ³³ Five patients were diagnosed with myocarditis on necropsy, three of which had radiographic cardiomegaly. Another patient with radiographic cardiomegaly was suspected to have an unrelated cardiomyopathy on necropsy; no evidence of myocarditis was seen on histopathology. The results of this study suggest that cardiomegaly may be observed on thoracic radiography of cats with FIP, for which recommended differentials would include myocarditis, pericardial effusion or an unrelated cardiomyopathy. Pulmonary arterial enlargement or tortuosity was seen in a total of 5/35 (14.3%) cats, with concurrent pulmonary venous enlargement in one. Three of these patients were examined by necropsy, all of which had moderate to severe vasculitis and 2/3 had myocarditis. It is unclear if the pulmonary vascular changes are secondary to severe pneumonia and vasculitis or myocarditis with secondary vascular congestion, although it was seen in one cat without concurrent cardiac pathology. Additional research would be required to further assess this.

There are a few limitations to our study. The first is lack of confirmation of FIP with IHC in all cases. It is therefore possible that not all of the presumptive cases had FIP. Radiographic grading parameters were based on subjectivity; the authors tried to mitigate this by having two observers reach a consensus, but this does provide an additional limitation through which the results of this study should be viewed. Another limitation is the retrospective nature of the study and the lack of standardization of diagnostics performed in this population of cats. Sampling bias likely also impacted the results of our study, as clinicians may not perform thoracic radiographs in cats with a straightforward diagnosis of FIP or those without respiratory signs. Cats ultimately included in the study population may therefore have had atypical disease or more severe thoracic involvement than a larger population of cats with FIP. Further prospective studies are needed to reliably define and describe thoracic radiographic findings in a large population of cats with FIP.

Conclusions

Cats with FIP may have a wide range of thoracic radiographic changes that include pleural effusion; unstructured interstitial, bronchial and/or alveolar patterns; multifocal, small, discrete or ill-defined pulmonary nodules; lymphadenopathy; and occasionally cardiomegaly. Cardiomegaly may be the result of myocarditis, pericardial effusion or unrelated cardiomyopathy. The radiographic changes identified in this study are non-specific and overlap with those seen in many other disease processes; however, our results suggest that FIP should be considered as a differential diagnosis in cats with these radiographic changes and supportive clinical findings.