Introduction

Bartonellosis is a vector-borne zoonotic disease caused by species of the genus Bartonella.¹ Bartonella species infections are well documented in mammals, including cats, dogs and humans, and are associated with various clinical manifestations and disease severity.¹

The cat is the main reservoir for Bartonella henselae, the aetiological agent of cat scratch disease in humans.^1–3 B henselae has been isolated with relative frequency in cavitary effusions in human and canine patients, being sometimes difficult to determine if it is an aetiological agent or an opportunistic pathogen.^1–4 This case report describes a cat with pleural and pericardial effusions likely caused by B henselae infection.

Case description

A 7-year-old female spayed domestic shorthair cat presented with a 2-day history of dyspnoea, anorexia and lethargy. No previous diseases were reported and its retrovirus status was negative. The other cat in the household was admitted to our hospital 2 weeks earlier with fever. Investigations revealed a positive blood PCR for Bartonella species. Both cats had indoor/outdoor lifestyles.

A physical examination revealed pale mucous membranes, a grade II heart murmur, mild restrictive respiratory pattern (40 breaths/min), 170 beats/min, 40 rpm, fever (39.5°C), lethargy and a marked flea infestation. Blood tests showed mild normocytic normochromic non-regenerative anaemia (packed cell volume 22.4%) and a complete serum biochemistry revealed no abnormalities.

Thoracic radiographs revealed moderate pleural effusion (Figure 1), and echocardiography identified significant pericardial effusion with cardiac tamponade (Figure 2). Abdominal ultrasound showed mild ascites, but it was not sufficient enough to be sampled.

The cat was sedated with butorphanol (Torbugesic; Zoetis) at 0.3 mg/kg IV and dexmedetomidine (Dexdomitor; Ecuphar) at 1 µg/kg IV, and pericardiocentesis and thoracocentesis were performed.

Figure 1

Right lateral thoracic radiograph showing pleural effusion

Figure 2

Echocardiography showing pericardial effusion

A small sample was obtained during the initial attempt at pericardiocentesis, sufficient only for fluid analysis and cytology. Subsequently, the pericardial effusion immediately resolved, presumably owing to the drainage of pericardial fluid into the pleural space. Thoracocentesis was then performed, yielding 50 ml of fluid. The nucleated cell count and protein concentrations in the fluid from each location were consistent with a protein-rich transudate. Cytological examination revealed the presence of macrophages and non-degenerate neutrophils in similar numbers, with no evidence of microorganisms or neoplastic cells, consistent with macrophagic-neutrophilic inflammation in both sampled areas. Post-procedure echocardiography and ECG showed no structural, functional or cardiac rhythm abnormalities.

The mixture of pericardial/pleural fluid was submitted for aerobic/anaerobic bacterial and fungal cultures as well as RT-PCR for feline coronavirus (FCoV). Peripheral blood, and the mixture of pleural/pericardial effusions, were analysed in the Veterinary Clinical Hematology Service Laboratory of the Universitat Autònoma de Barcelona (UAB) by conventional and real-time PCR for Bartonella species.⁵ A positive result was obtained in peripheral blood (conventional and real-time PCR) and pleural/pericardial effusions (real-time PCR). The PCR products were sequenced, analysed and compared using the GenBank database, and a 100% identity match with B henselae was obtained in both samples. No growth was obtained in the cultures and FCoV real-time PCR was negative.

The patient was hospitalised with marbofloxacin (5 mg/kg SC q24h, Marbocyl 1%; Vetoquinol) and amoxicillin–clavulanate (62.5 mg/cat IV q12h, amoxicillin–clavulanate 1000 mg/200 mg; Normon) along with flea treatments (first, fipronil spray [Frontline spray; Boehringer Ingelheim] and, later, imidacloprid and moxidectin spot-on [Advocate; Bayer]). Over the following days, the cat showed improvement and the fever resolved. After 3 days, follow-up radiography and echocardiography revealed no effusions, leading to the cat being discharged with marbofloxacin (5 mg/kg PO q24h, Marbocyl; Vetoquinol) and amoxicillin–clavulanate (62.5 mg/cat PO q12h, Clavaseptin; Vetoquinol). Amoxicillin-clavulanate was discontinued after the culture results came back negative, but marbofloxacin was continued.

One month later, the cat remained stable, with resolved anaemia and no effusions detected. However, a repeat blood PCR for Bartonella species, while the cat was still on marbofloxacin treatment, was positive. Marbofloxacin was discontinued and doxycycline was started (5 mg/kg PO q12h, Doxycare; Ecuphar). Six weeks later, while the cat was still receiving doxycycline, a follow-up blood PCR test for Bartonella species was negative; therefore, doxycycline treatment was stopped.

Three years later, the patient remained healthy and the recheck echocardiography indicated no pericardial or pleural effusions.

Discussion

Epidemiological studies indicate a high prevalence of Bartonella species infection in cats.^1,5 However, the few clinical cases described in the literature indicate that accurately diagnosing Bartonella species infections remains challenging.^1–4

Clinical signs, lesions and laboratory abnormalities reported in association with Bartonella species infection in cats are fever, lymphadenomegaly, anaemia, eosinophilia, hyperglobulinaemia, thrombocytopaenia, diaphragmatic myositis, endocarditis, pyogranulomatous myocarditis, endocardial fibrosis complex, mild neurological signs, uveitis, conjunctivitis, keratitis, corneal ulcers, arthritis and osteomyelitis.⁶ However, to the best of the authors’ knowledge, pleural and/or pericardial effusion associated with B henselae infection has not been reported in cats without concurrent heart failure.^1,6,8

One possible explanation for the pericardial effusion in our patient is infectious pericarditis caused by B henselae. Feline bacterial pericarditis is rarely documented, with a retrospective study of pericardial effusion in 146 cats finding no cases of septic pericarditis.⁹ Another study of 83 cats with pericardial effusion identified hypertrophic cardiomyopathy, neoplasia and systemic infection as the main causes, and there have been few case reports of cats with bacterial pericarditis caused by bacteria other than Bartonella species.^8,10–13 This raises the question of whether Bartonella species could produce a similar clinical presentation.

The most common causes of pleural effusion in cats include congestive heart failure, feline infectious peritonitis (FIP), neoplasia, pyothorax or chylothorax.¹⁴ Our patient presented with cardiac tamponade, which could explain the pleural effusion. Another possibility is pleuritis resulting from haematogenous spread or vasculitis due to B henselae-induced endothelial damage.^15–22

In human medicine, Bartonella species infections have been associated with pericardial and pleural effusions.^15–19 In dogs, B henselae and Bartonella vinsonii subspecies berkhoffii have been identified in pleural and peritoneal effusions, though their pathogenic role remains unclear.^20–22 These observations suggest a comparable scenario might occur in cats. On the other hand, a PCR study for vector-borne pathogens in dogs with pericardial effusion did not detect Bartonella species.²³

The effusions in our patient were classified as protein-rich transudates. An exudate typically indicates an inflammatory response to bacterial infection; however, this may not always apply to Bartonella species infections. The lipopolysaccharide of Bartonella quintana has demonstrated anti-inflammatory properties, while studies suggest that B vinsonii subspecies berkhoffii can lead to immunosuppression.^21,24,25

Research on the pathophysiology of Bartonella species infection reveals their ability to invade endothelial cells, potentially increasing vascular permeability and leading to effusions without a significant inflammatory response. In addition, isolated cases in dogs have shown protein-rich transudate effusions where Bartonella species were detected.^21,24–26

Fluid cytology analysis of the pericardial and pleural effusions revealed a protein-rich transudate with mixed macrophagic-neutrophilic inflammation. Common infectious agents associated with this type of inflammation in cats include FCoV, mycobacteria and systemic fungal infections, with occasional involvement from Actinomyces, Nocardia, Rhodococcus, Streptomyces, Francisella, Bartonella and Leishmania species, as well as rare immunomediated processes such as idiopathic sterile pyogranuloma.²⁷

Although the FCoV real-time PCR (IDEXX Laboratory) results were negative, this test is not the gold standard, so a false negative is possible. However, given the patient’s improvement and resolution of clinical signs, FIP is unlikely. In addition, culture and cytology tests did not reveal any fungal pathogens or other bacterial infections.

As a result of financial constraints, mycobacterial-specific culture and/or PCR, and Leishmania species PCR or serology, were not conducted. However, considering the patient’s clinical presentation and disease progression with the given treatment, Leishmania species infection is unlikely.²⁸ In addition, considering the acute onset of clinical signs and outcome, a mycobacterial infection was also deemed unlikely, as mycobacterial infections usually only respond after prolonged periods with triple antibiotic therapy.

Bartonella species are difficult to detect directly in blood samples with traditional culture methods because of their low-level, intermittent bacteraemia and the requirement for prolonged incubation periods. Therefore, despite being the ‘gold standard’ method for diagnosing, a negative culture does not necessarily rule out their presence.^6,29,30

The most reliable methods for confirming Bartonella species infection involve specialised culture techniques such as lysis centrifugation, cell culture isolation and insect-based growth enrichment.¹ Bartonella species can be cultured from reservoir hosts and occasionally from accidental hosts using agar plates with 5% defibrinated rabbit or sheep blood.¹ A combined method using pre-enrichment liquid culture (Bartonella/alpha-Proteobacteria growth medium [BAPGM]), PCR and subculture on agar plates has been used to improve the detection and isolation of Bartonella species in cats, dogs and humans.²⁹ In this case, traditional culture media likely resulted in no growth from the pleural/pericardial effusion.

We obtained positive PCR results for Bartonella species in both peripheral blood (conventional and real-time PCR) and pleural/pericardial effusion (real-time PCR). While commercial veterinary laboratories offer conventional and quantitative PCR (qPCR) for detecting Bartonella species, the accuracy may vary depending on the primers and protocols used.⁶

Cats are natural reservoirs for B henselae, complicating serological diagnosis. Serology is primarily used for exclusion rather than confirmation; therefore, it was not performed for this patient.¹ A multimodal testing approach is advised for diagnosing feline bartonellosis, as one negative result is not conclusive.^29,30

Recognising an active infection in cats can be challenging because they can be subclinically infected. Treatment is advised only for those cats suspected of exhibiting clinical signs due to Bartonella species or for those living with children or immunocompromised individuals owing to zoonotic risks.¹ In our case, the patient was not immunosuppressed, no other causes for the clinical signs were identified and flea infestation was noted.

There is no standardised treatment protocol for Bartonella species and the ideal therapy is still unknown.¹ Prolonged treatment is generally recommended, with recent trends suggesting a combination of fluoroquinolone and doxycycline, although in vitro studies have reported conflicting results on the need for multiple antibiotics.^{1,6,7,31–35}

In our case, initial treatment with amoxicillin–clavulanate and marbofloxacin for 2 weeks, followed by 4 weeks of marbofloxacin alone, did not resolve the infection. However, after 6 weeks of doxycycline at 5 mg/kg PO q12h, a negative blood PCR was achieved. Although other doxycycline-responsive pathogens could not be completely ruled out, this was considered unlikely based on the type of inflammation present and clinical signs.

Amoxicillin–clavulanate and marbofloxacin were initially used for broad-spectrum coverage before confirming B henselae. Marbofloxacin was continued owing to the patient’s positive response and the good response to fluoroquinolones seen in some studies.^7,32–34

Conclusions

This case emphasises the importance of considering Bartonella species as a differential diagnosis in cats with protein-rich transudate effusions, neutrophilic-macrophagic inflammation and fever. The clinical signs were likely linked to B henselae infection, as no other causes were identified and the patient responded positively to treatment. Further research and case studies are needed to better understand the effects of this pathogen on healthy, non-immunocompromised cats and to characterise its morbidity. In addition, the importance of using combined diagnostic tools to detect or rule out the pathogen as the cause of clinical conditions is reinforced.

© The Author(s) 2025

This article is distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 License ( https://creativecommons.org/licenses/by-nc/4.0/) which permits non-commercial use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages ( https://us.sagepub.com/en-us/nam/open-access-at-sage).