Cat scratch disease with fever and lymphadenopathy without definite cat exposure diagnosed via tissue metagenomic next-generation sequencing technology: a case report

Abstract

Background

Cat scratch disease (CSD), a disease caused by Bartonella henselae (B. henselae), is often associated with exposure to cats. However, in rare cases, a history of cat exposure is absent, and thus achieving a definitive diagnosis may be challenging. Metagenomic next-generation sequencing (mNGS) is a relatively new technology that can analyze pathogens without introducing inherent biases. In this report, we describe a case of CSD diagnosed via mNGS, which supports the use of mNGS in identifying rare types of pathogenic infection.

Case Description

A 54-year-old male exhibited fever and headache of unknown cause, with a history of working at a chicken farm and no clear cat exposure. Routine antibiotic treatment for infection was ineffective. Lymph node biopsy demonstrated granulomas accompanied by chronic suppurative inflammation and abscess formation. A biopsy sample from the parotid lymph node was subjected to mNGS, which revealed the presence of B. henselae and led to a subsequent diagnosis of CSD. After treatment with azithromycin, the patient recovered and was discharged. At the 3-month telephone follow-up, the patient was asymptomatic.

Conclusions

CSD in patients with prolonged fever plus culture-negative granulomatous lymphadenopathy or similar compatible findings, even when the exposure history is atypical，and mNGS, as an effective examination method, can aid in arriving at a definitive diagnosis.

Highlight box.

Key findings

• We report a case of Bartonella henselae (B. henselae) associated cat scratch disease (CSD) in a 54-year-old former poultry farm worker without cat exposure. The patient presented with fever and headache, and lymph node biopsy revealed granulomas with chronic suppurative inflammation and abscess formation. Metagenomic next-generation sequencing (mNGS) of parotid lymph node tissue successfully detected B. henselae, facilitating definitive diagnosis. The patient achieved clinical remission after standardized azithromycin therapy and remained asymptomatic at 3-month follow-up.

What is known and what is new?

• It is widely recognized that CSD typically occurs in patients with cat exposure.

• This case expands the susceptible population of B. henselae to occupational poultry workers lacking typical cat contact. For patients with prolonged fever and culture-negative granulomatous lymphadenopathy, CSD should still be suspected despite atypical exposure history.

What is the implication, and what should change now?

• mNGS serves as a critical tool for identifying pathogens in patients with fever of unknown origin accompanied by granulomatous lymphadenopathy. Timely molecular testing and standardized azithromycin anti-infective treatment for patients with atypical CSD yield favorable clinical outcomes, providing practical references for the diagnosis and management of occult lymph node lesions in such patients.

Introduction

Cat scratch disease (CSD) was first identified in the early 20th century (1-3), However, the etiology of CSD remained unclear until the 1990s, when it was finally proven that Bartonella henselae (B. henselae) (formerly known as ‘Rochalimaea henselae’) identified as was the causative pathogen (4). Bartonella an intracellular facultative Gram-negative bacterium with the following taxonomic classification: class, subgroup α2 of Alphaproteobacteria; order, Rhizobiales; family, Bartonellaceae; and phylum, Proteobacteria (5). More than 45 species of Bartonella have been reported thus far, among which 12 are associated with human infections. The most common types include B. henselae, Bartonella quintana, and Bartonella bacilliformis (6). B. henselae infection in humans causes CSD, a zoonotic disease. It is usually transmitted directly or indirectly to humans via arthropods such as fleas, sandflies, ticks, red ants, spiders, mites, and mosquitoes. The most frequently reported routes of transmission are contact with cats and cat scratches or bites (5,6). Patients with CSD can exhibit a wide variety of symptoms. The typical symptom is localized lymph node enlargement, while atypical symptoms can involve multiple systems and include fever, rash, thrombocytopenic purpura, encephalitis, meningitis, osteolytic lesions, arthritis, endocarditis, eye disorders (e.g., uveitis, optic neuritis and retinitis, vasculitis, retinal vascular occlusion, retinal detachment, and Parinaud oculoglandular syndrome), acute fibrous tissue pneumonia, and granulomatous disease of the liver and spleen (1,2,7-16). The incidence of CSD is higher in the autumn and winter due to seasonal fluctuations in zoonotic transmission between felines or changes in animal behavior and breeding times (17). Domestic cats are the primary reservoir for the pathogen, and these animals may have long-term, asymptomatic bacteremia. Cat fleas are involved in the transmission of B. henselae among cats, and humans, particularly those with compromised immune systems (17,18). Diagnosis can usually be made based on medical history and auxiliary examinations. However, given the diverse manifestations of CSD, it may be difficult for physicians to arrive at a definitive diagnosis in patients without cat exposure or those with atypical symptoms. Metagenomic next-generation sequencing (mNGS) can examine all the information present in the DNA or RNA of a sample and is thus capable of analyzing its entire microbiome, along with the host genome or transcriptome. This technology can quickly and accurately identify multiple pathogens in the detected sample, which is sometimes useful in identifying infections of rare pathogens (19).

In this report, we describe an adult patient with fever, headache and lymphadenopathy and no prior cat exposure. The diagnosis of CSD was confirmed via mNGS analysis of lymph node pathological tissue. This case indicates that clinicians should suspect CSD in patients with prolonged fever combined with culture-negative granulomatous lymphadenopathy or compatible lesions, even when exposure history is atypical. mNGS detection of lesional tissue improves the clinical diagnosis of CSD. We present this article in accordance with the CARE reporting checklist (available at https://acr.amegroups.com/article/view/10.21037/acr-2026-0079/rc).

Case presentation

All procedures performed in this study were in accordance with the ethical standards of the institutional and/or national research committee’s and with the Declaration of Helsinki and its subsequent amendments. Written informed consent was obtained from the patient for the publication of this case report and accompanying images. A copy of the written consent is available for review by the editorial office of this journal.

A 54-year-old male patient visited the clinic due to intermittent headaches for 20 days, and fever for 4 days. Over the previous 10 years, the patient had a history of hypertension and intermittent headaches, with a maximum systolic blood pressure of 180 mmHg, and was taking combination of amlodipine and benazepril to manage his hypertension. During this period, the patient experienced intermittent headaches without obvious triggers, which were not taken seriously and no examinations or treatments were conducted. He had been working at a chicken farm for 6 months and has no other significant medical history.

Twenty days before presentation, the patient reported that they had worsening headache without any obvious cause. This was experienced as a persistent tightening pain at the temples and top of the head, accompanied with blurry vision. The episodes occurred irregularly. He was admitted to a local hospital for diagnosis and treatment. On the second day after admission, four days prior to his visit to our hospital, he developed fever, chills and rigors with a peak temperature of 42 ℃. His body temperature could return to normal after antipyretic and anti-infective therapy. At the local hospital, tests for influenza A and B virus antigens were negative, and a head computed tomography (CT) routine scan showed no significant abnormalities. The patient was treated with a combination of levofloxacin and piperacillin-tazobactam for infection (specific dosages unknown) but continued to have fever. He was transferred to our hospital after spending 4 days at the outside hospital without improvement, for further diagnosis and treatment.

Upon admission, the patient had stable vitals, normal level of consciousness and showed no signs of distress. There was an approximately 2 cm long wound on his forehead, painless enlarged left pre-auricular and submental lymph nodes. He stated that the wound was caused by accidental collision with a harvester during farm work.

Tests were conducted and results show, laboratory results revealed that the neutrophil percentage (79.8%), erythrocyte sedimentation rate (36 mm/h), procalcitonin level (0.90 ng/mL), and ferritin level (534.01 ng/mL) were all elevated. Albumin level (29.3 g/L) was decreased, gamma-glutamyltransferase level (112 U/L) was increased, Epstein-Barr virus DNA (blood) was positive (759 copies/mL), and the D-dimer level (3,821 ng/mL) was increased. Sputum culture and Gram staining detected Gram-positive cocci and Gram-negative bacilli. Peripheral blood specimens were incubated in aerobic and anaerobic blood culture bottles for a routine 5-day period, and no bacterial growth was observed in either bottle.

Head magnetic resonance imaging, including plain scanning, enhanced examination, and diffusion-weighted imaging, revealed a high signal in the white matter (Fazekas grade 1) Doppler ultrasound of the deep veins in the lower limbs revealed no significant abnormalities, while echocardiography demonstrated slight aortic valve insufficiency. Doppler ultrasound of the bilateral superficial temporal arteries showed mild thickening of the middle and distal segments of the bilateral superficial temporal artery walls, with potentially inflammatory enlargement of the left submandibular lymph node.

Thyroid ultrasound revealed no echo areas of the bilateral lobes (Thyroid Imaging Reporting and Data System category 2). Parotid ultrasound revealed a mixed echogenic mass in the left parotid gland which prompted further examination. Superficial lymph node ultrasound suggested that the lymph nodes in left neck region IV had an abnormal structure (Figure 1); the bilateral axillary lymph nodes are visible, with partial cortical thickening. No significant abnormalities were found for the following: four tests for infectious diseases; detection of respiratory tumor markers; cryptococcal capsular antigen test; triplex nucleic acid detection for Pneumocystis jirovecii, Aspergillus, and Cryptococcus; routine fecal occult blood tests; 1,3-beta-D-glucan test; detection of high-sensitivity cardiac troponin I and brain natriuretic peptide precursor; eight respiratory pathogen antibody tests; seven respiratory viral nucleic acid tests; complete set of autoimmune antibody panel; three anticardiolipin tests; human leukocyte antigen B27 test; anti-neutrophil cytoplasmic antibody profile; anti-glomerular basement membrane antibody test; and immunoglobulin anti-O rheumatoid.

Figure 1.

Ultrasound of the superficial lymph and parotid gland nodes. (A) Multiple low-echo light masses were visible in the region IV of the left neck, with unclear lymphatic borders, increased cortical echo, and small patchy low echoes inside; with one low-echo light mass measuring approximately 1.1 cm × 0.4 cm in size was also present; (B) the echo distribution of the left parotid gland was uneven, there were multiple mixed echo masses visible, the contours were clear, and the internal echoes were uneven, with one exhibiting blood flow and measuring approximately 2.1 cm × 1.3 cm in size.

Given the medical history and examination results, the provisional status was “fever with lymphadenectasis pending investigation”. The patient was started on anti-infective treatment (meropenem and moxifloxacin, which was later adjusted to meropenem and ganciclovir) along with gastric protection and antihypertensives. The patient continued to experience recurrent fever despite the treatment. After a focused discussion, it was considered necessary to rule out special infections, hematologic tumors, tuberculosis, etc. A comprehensive positron emission tomography (PET)-CT examination was performed, and the results were as follows: (I) there was a slightly enlarged spleen and a slightly low-density nodule within the spleen with increased metabolic abnormality; this suggested the possibility of an infectious lesion and ruled out lymphoma. (II) There were multiple nodules in the left parotid gland and below the parotid gland with increased metabolic abnormality, suggesting the possibility of pleomorphic adenoma. A recommendation was thus made to perform surgical biopsy to exclude malignant lesions (Figure 2).

Figure 2.

Positron emission tomography-computed tomography images. (Left) Multiple nodules were observed within the left parotid gland and below the parotid gland, with the largest measuring approximately 19 mm × 9 mm in size; there was abnormal accumulation of the imaging agent, with an SUVmax of 4.5–12 (arrow:lesion in the left parotid gland and subparotid region). (Right) The spleen was enlarged, and a slightly low-density nodule measuring approximately 8 mm × 7 mm in size was observed within the spleen, with a concentrated distribution of the imaging agent and an SUVmax of 4.6. In the splenic hilum, a nodule measuring approximately 22 mm × 17 mm in size was observed, with no abnormal accumulation of the imaging agent. SUVmax, maximum standardized uptake value (arrow: splenic lesion).

The patient was updated on findings, informed consent obtained and surgical biopsy of the lymph gland. Pathology report, showed granulomas accompanied by chronic purulent inflammation and abscess formation, suggesting an infectious lesion. Epstein-Barr encoding region in situ hybridization was negative. Special staining, including periodic acid-Schiff, acid-fast, and silver staining, revealed no abnormalities. Mycobacterium tuberculosis nucleic acid testing was also negative (Figure 3). As there was no evidence of tuberculosis, infection with a rare origin was considered. In order to identify the underlying pathogen, mNGS of the lymph node tissue was conducted. The results indicated infection with B. henselae. within the Bartonella genus, show that the Bartonella genus has a specific DNA sequence of 481, and B. henselae has 308 specific DNA sequences (Table 1). Upon further inquiry into the patient’s medical history, the patient works at a chicken farm. Although he has no recollection of cat exposure, he may not be sure if cats visit the farm.

Figure 3.

Pathology of parotid gland tissue and lymph nodes (left side parotid gland area tumor resection specimen). Microscopic examination revealed parotid gland tissue and lymph nodes with granulomas accompanied by chronic suppurative inflammation and abscess formation, suggesting an infectious lesion. Epstein-Barr encoding region in situ hybridization was negative. Special staining, including periodic acid-Schiff, acid-fast, and silver staining, revealed no abnormalities. Mycobacterium tuberculosis nucleic acid testing was also negative. After discussion in the pathology department, this was considered to be a case of infectious lesion. (Magnification: left, 100×; right, 40×).

Table 1. mNGS results.

Category	Type	Genus		Species
		Genus	Number of DNA sequences	Species	Number of DNA sequences
Bacteria	G−	Bartonella	481	Bartonella henselae	308
Fungus	−	−	−	−	−
Virus	−	−	−	−	−
Parasite	−	−	−	−	−
Mycobacterium tuberculosis complex	−	−	−	−	−
Mycoplasma/chlamydia	−	−	−	−	−
Bacterial resistance genes	−	−	−	−	−
Virulence genes	−	−	−	−	−
Background bacteria	G+	Cutibacterium	84	Cutibacterium acnes	70
	G+	Staphylococcus	35	Staphylococcus warneri	12
	G+	Staphylococcus	35	Staphylococcus hominis	12
	G+	Micrococcus	29	Micrococcus luteus	28
	G−	Moraxella	28	Moraxella osloensis	28
	Virus	Lymphocryptovirus	3	Human gammaherpesvirus 4	2

mNGS, metagenomic next-generation sequencing.

Overall, the patient presented with fever and localized lymph node enlargement. Lymph node biopsy demonstrated granulomas accompanied by chronic suppurative inflammation and abscess formation. mNGS identified DNA sequence fragments specific to B. henselae, and other infectious etiologies were excluded. These findings supported a diagnosis of CSD, and azithromycin was administered accordingly (an initial dose of 500 mg is administered on the first day, with a maintenance dose of 250 mg taken once per day afterwards). Two days after azithromycin treatment, the patient’s fever subsided, and he was discharged 3 days later. He was instructed to take azithromycin for another 2 weeks to prevent disease recurrence. At the 3-month telephone follow-up, the patient was asymptomatic.

Discussion

B. henselae is an intracellular facultative Gram-negative (5). It is typically transmitted to humans directly or indirectly through arthropods such as fleas, with the most frequently reported cases involving contact with cats and cat scratches or bites (6).

CSD is usually caused by an infection from B. henselae and is most commonly spread through percutaneous contact, scratches, bites, or contact with fleas from infected cats (17). About half of the patients hospitalized with CSD are 18 years old or younger, with males accounting for around 60% of these cases. The incidence varies by season, with approximately 60% of patients becoming ill between September and January (20).

Among patients diagnosed with CSD, about 26.3–84% recall a history of cat exposure, scratches, or bites. A history of exposure supports the diagnosis but is not a necessary condition. As mentioned above, there are also reports indicating that fleas, sandflies, ticks, red ants, spiders, mites, and mosquitos can serve as transmission vectors for this bacterium, which indirectly suggests that the disease may involve indirect exposures that are not clearly recognized (6,21,22).

The core pathogenic mechanism of B. henselae involves its invasion of red blood cells and endothelial cells. It primarily targets CD34⁺ stem cells, releasing the virulence factor lipid A to exert toxicity, spreading through blood vessels and lymphatic vessels to multiple organ systems, causing disease. Additionally, it can form protective vesicles within CD34⁺ stem cells to evade the immune system, further aggravating disease manifestations (6,23).

After skin exposure, patients usually begin to develop symptoms within 1–10 days and undergo 3 general stages: vesicular, erythematous, and papular (6). When lymph node involvement occurs, lymphadenopathy often lasts 1–4 months and can even exceed 1 year (24), progressing through several stages: an early stage of lymphocyte proliferation and neutrophil aggregation, a middle stage of irregular granulomas and stellate abscesses indicating central necrosis, and a late stage characterized by inflammation and necrosis and subsequent spindle-shaped fibroblast proliferation. Pathologically, CSD manifests as granulomatous lymphadenitis, which is typically accompanied by central microabscesses. The granulomatous tissue surrounds a necrotic core, encircled by a fence-like arrangement of histiocytes, and bordered by lymphocytes and fibroblasts on the periphery (24).

B. henselae can affect multiple organs, which accounts for the diverse clinical manifestations in patients with CSD. In one study, among patients diagnosed with CSD after being hospitalized for unexplained fever, about 60% had ≥1 affected organ, with liver and spleen involvement (35%), lymphadenopathy (20%), ocular lesions (18%), and multifocal osteomyelitis (6%) being particularly common. Of these patients, 21% had initial imaging diagnoses suggesting malignant tumors, particularly lymphoma, and due to the difficulty in making a definite diagnosis, 32% underwent invasive procedures to assist in the diagnosis (25). Although CSD symptoms vary, contact with or scratches from cats are important diagnostic clues, and thus, cases without cat exposure are relatively rare.

The diagnosis of CSD is mainly based on the following four key criteria proposed in a 1985 study: history of cat contact, lesions at the site of contact, regional lymphadenopathy, and positive skin test results for Bartonella. A study in 2000 further suggested that CSD can be considered if at least three of the following factors are present: contact with a cat or fleas, exclusion of other infectious causes, positive serology, or biopsy showing granulomatous inflammation consistent with CSD (26).

Patients with CSD typically exhibit fever, skin lesions, and localized lymphadenopathy. When lymphadenitis occurs, it needs to be differentiated from infections such as tuberculosis or atypical mycobacterial infections, viral lymphadenitis, Staphylococcus aureus, and pyogenic streptococcal infections.

The adult male patient described in this report had no history of cat exposure, developed symptoms in the autumn, presented with fever and lymphadenopathy, yet no positive serological results were obtained. PET-CT revealed splenic lesions suggestive of infectious lesions, while lymphoma could not be excluded temporarily. Lymph node biopsy was further performed, which demonstrated granulomas accompanied by chronic suppurative inflammation and abscess formation, indicating an infectious etiology. That ruled out malignant tumors and tuberculosis. Up to this point, sequential anti-infective regimens combining meropenem with moxifloxacin, and then meropenem with ganciclovir, had failed to improve the patient’s condition. Rare pathogenic infection was highly suspected clinically; therefore, mNGS was performed on the biopsy tissue, which detected B. henselae. Collectively, based on the patient’s occupational exposure history at a chicken farm, forehead trauma history, and all clinical manifestations and examination results, CSD was definitively diagnosed.

There is no gold standard method for the diagnosis of CSD, and the rates of misdiagnosis and missed diagnosis are quite high (27).

CSD is often diagnosed clinically and supported by laboratory tests. Various testing methods can aid in diagnosis, including serology, pathogen culture, ultrasound, histopathology, molecular diagnostic techniques, and recombinant protein serology. Of these methods, serology is preferred method due to its accessibility. After human exposure, the seroconversion rate is nearly 28%. As with other immune responses, antibody titers increase fourfold within 2–3 weeks, which may indicate CSD, but most patients with CSD do not produce antibodies, resulting in low sensitivity (6,9,11,26,28). Immunofluorescent antibody testing, enzyme immunoassays, and Western blotting can be used to detect B. henselae antibodies to, but their specificity and sensitivity vary (17,29,30). Immunoassays and immunofluorescence tests are limited in their ability to detect chronic infections with low bacterial loads (26). Studies suggest that immunological tests for B. henselae have cross-reactivity with various pathogens such as Rickettsia, Orientia tsutsugamushi, Mycoplasma pneumoniae, and Brucella, among others, with some even exceeding 50% (10,26). If a patient is infused with immunoglobulin before sampling, false-positive results are likely (10). Western blotting can be used for a definitive diagnosis of CSD but, requires a certain amount of tissue and is cumbersome to perform (17,29,30). Recombinant serology for proteins such as GrrEL, 17-kDa, Pap31, SucB, P26, BadA, OMP89, and OMP43, among others, may be able to assist in diagnosis, but their application remains in the research stage and is not widespread in clinical practice (26).

Cultivation of B. henselae for clinical diagnosis is difficult due to its fastidious growth requirements and culture cycle (17,29,30). Specifically, isolation of bacteria from the blood, saliva, or wound samples has an extremely high rate of failure; the culture environment is restrictive, requiring media under conditions of 30–37 ℃, 5% carbon dioxide enrichment, high humidity, and the presence of hemin and other supplements; and the culture cycle is long, with colonies only forming after to 45 days (5,26).

A few studies have asserted that ultrasound of the superficial lymph nodes can assist in diagnosis, revealing oval hypoechoic lesions, and can dynamically monitor their changes to aid in diagnosis and evaluate treatment effectiveness (31). The widespread and reliable gene marker citrate synthase has also been used in B. henselae diagnosis (28).

Histopathology requires procedures such as biopsies, which are mostly invasive and generally not preferred as the first choice. Clinically, histopathology presents in multiple stages, and studies have shown that CSD may coexist with malignant diseases or infections caused by branch bacteria (24), and thus lymphoid tissue biopsy and molecular testing are essential.

mNGS has been applied in the diagnosis of CSD. As a fast and impartial diagnostic method, mNGS can facilitate investigations into pathogens without introducing biases. It allows for early diagnosis and the initiation of targeted antibiotic treatment by simultaneously detecting a variety of pathogens in a timely manner, particularly benefiting the identification of rare, atypical, and complex infectious diseases (19,27,32). Clinicians should suspect CSD in patients with prolonged fever combined with culture-negative granulomatous lymphadenopathy or compatible lesions, even when exposure history is atypical.

The treatment of CSD is tailored according to the severity of clinical symptoms. Typical or mild cases of CSD are usually self-limiting and spontaneously resolve within 2–4 months (17,32). However, severe forms of CSD require therapeutic intervention. B. henselae is a facultative Gram-negative bacterium. However, studies have shown that penicillin has been clinically proven to be ineffective against this bacterium, even though B. henselae appears sensitive to various antibiotics in vitro, such as commonly used penicillins, cephalosporins, aminoglycosides, chloramphenicol, tetracyclines, macrolides, rifampin, quinolones, and trimethoprim-sulfamethoxazole. This apparent discrepancy has been attributed differences in the drug efficacy between in vitro and in vivo conditions (6). Clinically available antibiotics include erythromycin, doxycycline, azithromycin, ciprofloxacin, gentamicin, rifampin, and minocycline, among others (5,13,21,24,33,34). For patients with clinical manifestations of widespread lymphadenopathy, administration of erythromycin, doxycycline, or azithromycin is more appropriate. Azithromycin does not shorten the overall duration of symptoms, but it can accelerate the regression of enlarged lymph nodes and is often the first choice in clinic (6,24). The treatment of complex CSD mainly includes azithromycin, doxycycline, and rifampin, either as monotherapy or in combination (17,32). Some patients with eye involvement may even require combined steroid hormone therapy (13), but the majority of patients can expect a good prognosis, with 95% recovering completely and 5 % experiencing ocular sequelae (25).

A review of CSD cases manifesting fever, lymphadenopathy and parotid involvement published in the past 5 years shows that most patients achieve a favorable prognosis after timely standardized treatment once diagnosed. Nevertheless, atypical CSD, especially in patients without cat exposure history, is often diagnosed late (35,36). In this case, histopathology combined with tissue mNGS successfully confirmed the etiology (Table 2).

Table 2. Comparison of case characteristics of parotid-involved cat-scratch disease.

Publication year	2026	2024	2025
Age	54-year-old adult	Pediatric patient (specific age unreported)	9-year-old boy
Exposure history	Prior exposure to poultry farm environments	Pet exposure history; denied history of cat scratches	Close contact with a kitten; healed scratch scar on the right forearm
Key findings	Fever, headache, parotid lymphadenopathy; conventional antibiotics ineffective	Isolated parotid gland mass mimicking parotid neoplasm	Fever and right postauricular pain and swelling; fever alleviated with antibiotics, accompanied by local abscess and erythema
Diagnostic method	Tissue mNGS detected B. henselae	Postoperative pathological confirmation: histopathology revealed epithelioid granulomas with lymphocytic infiltration (preoperatively suspected parotid tumor)	Abscess puncture and culture, bartonella serology: IgM titer 1:80, IgG titer 1:512
Treatment	Standard oral azithromycin monotherapy	Subtotal parotidectomy	Abscess incision and drainage combined with oral azithromycin
Outcome	Fever resolved on day 2 of treatment; asymptomatic at 3-month telephone follow-up	Good prognosis, recovery time unknown	Swelling resolved fully in 3 weeks

B. henselae, Bartonella henselae; IgG, immunoglobulin G; IgM, immunoglobulin M; mNGS, metagenomic next-generation sequencing.

Conclusions

The details of this report indicate that for patients with fever of unknown origin, CSD can still be considered even if there is no history of cat contact or exposure to cat scratches or bites. The clinical diagnosis of CSD is difficult due to the numerous influencing factors and the low sensitivity of pathogen culture, serology, and immunological testing. Moreover, CSD may coexist with malignancies or infections caused by mycobacteria (24), and thus lymphoid tissue biopsy and molecular testing are essential. mNGS, as an effective examination method, is also critical to arriving at a definitive diagnosis. It is hoped that this report can provide assistance in the clinical diagnosis and treatment of CSD.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All procedures performed in this study were in accordance with the ethical standards of the institutional and/or national research committee(s) and with the Declaration of Helsinki and its subsequent amendments. Written informed consent was obtained from the patient for the publication of this case report and accompanying images. A copy of the written consent is available for review by the editorial office of this journal.