Cat Scratch Disease (Bartonella Infection)

Are You Confident of the Diagnosis?

What you should be alert for in the history

A history of a cat exposure (scratch, bite, other contact) can be a clue to the diagnosis, as 80-90% of children with cat scratch disease (CSD) report contact with a cat. Kittens are most commonly implicated, as they more often harbor the etiologic agent.

Characteristic findings on physical examination

A primary inoculation papule or pustule may be found on the skin in about two thirds of children who have CSD (Figure 1). The primary inoculation site may be the ocular conjunctivae and this may produce aconjunctival granuloma (Figure 2).

Figure 1.

Papule and adenitis.

Figure 2.

Conjunctival granuloma.

The presence of regional lymphadenitis, usually of the lymph nodes draining the initial site of inoculation, is seen in virtually all children with apparent infection (Figure 3). The lymphadenopathy is usually tender, may be erythematous, and the affected nodes can range in size from a few millimeters to several centimeters. Axillary and cervical lymph nodes are most commonly involved, but submandibular, femoral, and inguinal nodes can also be involved.

Figure 3.

Preauricular adenitis.

Preauricular adenopathy may be found when the inoculation site is the ipsilateral ocular conjunctiva (Parinaud oculoglandular syndrome) (Figure 3). The degree of tenderness and erythema of the lymph nodes is variable and can range from absent or little pain and erythema to exquisite tenderness and suppuration. Spontaneous suppuration occurs in about 25% of patients. Enlarged nodes may persist for weeks or months.

Constitutional symptoms such as fever, headache, myalgias, and anorexia occur in about one-third of patients. They are usually short-lived, lasting only several days.

Expected results of diagnostic studies

The diagnosis is usually made clinically when the presentation is classic and there is a history of cat or kitten exposure. The diagnosis can be confirmed using commercially available serologic tests (indirect immunofluorescent antibody assays) that test for the presence of serum IgG antibodies to the etiologic agent, Bartonella henselae.

When used in the appropriate clinical setting, these tests are highly specific for Bartonella infection, and a single positive IgG titer in the appropriate clinical context can confirm the diagnosis. These tests may not be sensitive early in the course of the infection but may become positive when repeated in 2-3 weeks.

Polymerase chain reaction assays are available for detection of B henselae on tissue specimens, such as lymph node material. These assays in general have poor sensitivity for the detection of infection. Moreover, tissue specimens are usually not available for most cases of uncomplicated infection.

The organism may be visualized in tissue specimens when a Warthin-Starry silver stain is used; however the finding of bacilli on such stains is not specific for CSD. Pathologic examination of tissue specimens in which CSD is present classically shows lymphocytic proliferation with epithelioid granuloma formation (early stages) and may show necrotizing granuloma formation and stellate microabscesses (later stages of infection).

Diagnosis confirmation

The differential diagnosis includes other causes of lymphadenopathy in children, including pyogenic infections caused by Staphylococcus aureus and Streptocccus pyogenes (group A), tularemia, and nontuberculous mycobacteria.

Pyogenic infections caused by S aureus and S pyogenes commonly have a more abupt onset of illness than does cat scratch adenitis. The lymph nodes are more exquisitely tender and erythematous, and there is frequently an associated or preceding upper respiratory tract infection or cellulitis of the adjacent skin. Definitive diagnosis can be made if the lymph node suppurates and a specimen is sent for staining and cultures.

The ulceroglandular form of Francisella tularensis can produce a clinical illness that is difficult to distinguish from CSD. An inoculation site (from tick bite) and regional adenitis can be found. This infection is most often transmitted from a tick vector but can be acquired from direct contact with infected animals such as rabbits. Thus a careful exposure history is very helpful.

Nontuberculous mycobacteria is a common cause of regional lymphadenopathy in immunologically normal young children. This infection is most commonly found in toddlers aged 1-4 years, who commonly present with a subacute or chronic, firm, nonerythematous swelling of 1 or more cervical lymphnodes. Systemic manifestations such as fever are uncommon.

Who is at Risk for Developing this Disease?

CSD is one of the most common causes of subacute orchronic lymphadenopathy in children. Classic CSD most commonly occurs in the pediatric and adolescent age groups. Eighty percent to 90% of individuals with CSD report a history of recent contact with healthy cats, often kittens. CSD is more common in autumn and winter and is seen more commonly in warm, humid climates.

What is the Cause of the Disease?

B henselae, a pleomorphic gram-negative bacillus, is the organism associated with CSD. This organism can be found as normal oral flora in cats, especially kittens, and it is not unusual for kittens to have bacteremia related to this organism. The transmission of this organism between cats occurs by the cat flea, but it is not clear whether this flea is responsible for the transmission from cats to humans. Stray cats harbor the organism much more commonly that do pet cats.


The incubation period from the time of the cat scratch to the development of the skin lesions or rash is about 7-12 days. Regional lymphadenopathy occurs 5-50 days (median 12 days) after the intial skin lesion.

Systemic Implications and Complications

Atypical presentations and complications of CSD are being recognized more commonly. Recent studies show that 25-30% of cases come to medical attention for reasons other than lymphadenopathy. Focal or diffuse inflammatory responses to CSD may involve neurologic, systemic, or skeletel systems.

Neurologic complications of CSD include encephalopathy, neuroretinitis, status epilepticus, and transverse myelitis. A history of cat or kitten exposure with regional lymphadenopathy may or may not be seen at the time of presentation.

A recent study found that CSD causes about 5% of cases of prolonged fever of unknown origin in children. This may or may not be accompanied by multifocal hepatosplenic microabscesses and osteolytic bone lesions.

Immunocompromised patients, especially those who have AIDS, may manifest bacillary angiomatosis or bacillary peliosis as a result of B henselae infection. Bacillary angiomatosis is a vascular proliferation of the skin and subcutaneous tissue and can manifest as nontender, firm, papular, or nodular lesions. Bacillary peliosis is characterized by vasoproliferation, mainly within the liver and spleen.

Treatment Options


Warm compresses

Antipyretics and analgesics


Needle aspiration for suppurative lymph nodes


Optimal Therapeutic Approach for this Disease

Because uncomplicated CSD is self-limited, the mainstay of therapy for such cases is symptomatic relief of the symptoms with warm compresses applied to the site of adenopathy and analgesic medications and antipyretic agents.

Surgical management of the lymphadenopathy is rarely required in the treatment of CSD. Needle aspiration of a tender lymph node may be required for relief of the pain associated with a suppurative node, and relief of severe pain is the only indication for surgical management of CSD. Surgical removal of the lymph node is not required under most circumstances.

Occasionally, when the diagnosis is in doubt, removal of the lymph node for pathologic study may be required to differentiate CSD from lymphoma or nontuberculous mycobacterial infection. However this differential diagnosis can usually be made clinically and supported with laboratory methods, obviating the need to obtain tissue material. Fine needle aspiration of the lymph node is not recommended, as this procedure may increase the risk of suppuration and development of a fistulous tract to the skin, with resultant chronic drainage.

The efficacy of systemic antibiotics for uncomplicated CSD has not been established. The only randomized, prospective, placebo-controlled trial in patients who had typical CSD revealed that azithromycin therapy decreased the size of the lymph node in the first 30 days of illness but no difference was found after 30 days.

Other data regarding the efficacy of antimicrobial therapy for typical and atypical CSD come from case studies and retrospective reviews. These data have shown that patients who have severe, acute, or disseminated disease may recover more quickly with antimicrobial therapy, which should be considered for such patients. Antimicrobial agents that have been usedto treat typical and atypical CSD and include erythromycin, ciprofloxacin, trimethoprim-sulfamethoxazole, rifampin, and intravenous gentamicin.

Although reports suggest that antimicrobial therapy is effective, the role of antibiotics in the treatment of these infections is not clear. Antimicrobial therapy, however, has been shown to be beneficial and is recommended for immunocompromised patients with bacillary angiomatosis and bacillary peliosis. Azithromycin (12mg/kg/day; max 600mg/day), erythromycin (40mg/kg/day; max 2g/day), and doxycycline (2-4mg/kg/day; max 200mg/day) have been shown to be effective for these entities. Doxycycline is generally not used for children younger than 8 years of age because of the risk of permanent dental staining.

Patient Management

Typical CSD is a benign, self-limited illness. Spontaneous resolution of the symptoms occurs in 2-4 months. Supportive care and reassurance about the benign nature of the illness are important.

Unusual Clinical Scenarios to Consider in Patient Management

Removal of the cat from the home is usually not necessary after a case of CSD. Careful play with the cat can be recommended to prevent bites or scratches and declawing cats may also be helpful. Elimination of fleas may help decrease transmission of the organism between cats.

What is the Evidence?

English , R. ” Cat-scratch disease”. Pediatr Rev . vol. 27. 2006. pp. 123-8. (An in-depth review of the clinical features of CSD, including a nice discussion of the etiology and epidemiology of the disease. Also highlights the features of atypical disease in immunocompetent and immunocompromised hosts.)

Reynolds , MG, Holman , RC, Curns , AT, O’Reilly , M, McQuiston , JH, Steiner , CA. ” Epidemiology of cat-scratch disease hospitalizations among children in the United States”. Pediatr Infect Dis J . vol. 24. 2005. pp. 700-4. (This analysis of a national inpatient database examined the CSD-associated hospitalization rate among children for a specific year. The authors found that despite the increasing incidence of cat ownership from previous years, the overall hospitalization rate for CSD did not increase. They also found that early diagnosis by serologic testing obviated the need for unnecessary intervention.)

Bass , JW, Cary Freitas , BC, Freitas , AD. “Prospective randomized double blind placebo-controlled evaluation of azithromycin for treatment of cat-scratch disease”. Pediatr Infect Dis J . vol. 17. 1998. pp. 447-52. (This often-referenced study is the only controlled study that examined the benefit of antimicrobial therapy for uncomplicated cat scratch adenitis. The results showed a modest benefit of therapy, manifested by a total decrease in lymph node volume within the first month of treatment.)

Jacobs , RF, Schutze , GE. ” Bartonella henselae as a cause of prolonged fever and fever of unknown origin in children”. Clin Infect Dis . vol. 26. 1998. pp. 80-4. (This report established CSD as a relatively common cause of prolonged fever and fever of unknown origin in children. This was a prospective study of 146 children with prolonged fever and fever of unknown origin conducted over a 6-year period in Arkansas. Bartonella infection accounted for 10% of cases of fever of unknown origin in this series of patients.)

Zangwill , KM, Hamilton , DH, Perkins , BA. “Cat scratch disease in Connecticut. Epidemiology, risk factors and evaluation of a new diagnostic test”. N Engl J Med . vol. 329. 1993. pp. 8-13. (This case control study examined the epidemiologic features of CSD and defined its risk factors. Owning a kitten was identified as the major risk factor, and the indirect fluorescent antibody test for the detection of antibodies to the etiologic agent was established.)

Florini , TA, Zaoutis , TE, Zaoutis , LB. ” Beyond cat scratch disease: widening spectrum of Bartonella henselae infection”. Pediatrics. vol. 121. 2008. pp. e1413-25. (A very thorough review of the history, microbiologic features, clinical manifestations, pathogenesis, diagnostic techniques, and management of B. henselae infection.)

Margileth , AM. “Recent advances in diagnosis and treatmentof cat scratch disease”. Curr Infect Dis Rep . vol. 2. 2000. pp. 141-6. (This is a review of the contemporary methods used for the diagnosis of CSD.)