Resident rounds

May 2007

A 10-year-old boy was referred for further evaluation by his primary care physician due to failure to respond to outpatient management of cervical adenitis.

He initially presented six days prior with a two-day history of unilateral cervical adenitis without fever for which he was empirically treated with amoxicillin-clavulanic acid. Since that time, he had been afebrile and developed no new symptoms, such as cough, conjunctivitis, pharyngitis, nausea, vomiting, diarrhea or rash. No weight loss was noted.

The patient complained of tenderness overlying the adenopathy but was in no significant distress. The family reported a gradual increase in erythema since their last visit but that the size of the neck mass was essentially unchanged. He was on no other medications. The patient had no prior history of significant illness and had received all routine immunizations. His last tuberculin skin test was reportedly several years prior and was negative. His family owned a cat. There were no sick contacts or history of foreign travel.

Physical examination results showed that the patient’s temperature was 99.3° F. His pulse was 82, and his blood pressure was 92 mm Hg/54 mm Hg. The patient was alert and in no apparent discomfort. His head, eyes, ears, neck and throat were normal except for an enlarged left anterior cervical lymph node measuring approximately 4 cm in diameter. There was mild tenderness and mild overlying edema and erythema. Inferior and anterior to the lymph nodes was a partially healed abrasion/laceration with a central scaling papule measuring approximately 5 mm in diameter. There was no apparent tenderness, erythema or edema at this location (Figure 1). A single submental lymph node measuring approximately 1 cm without any overlying edema or erythema was also noted.

Other physical findings included heart: no murmur. Lungs: clear to auscultation. Abdomen: no masses, organomegaly or tenderness. Skin: no other rashes or adenopathy noted, except as detailed above. A number of superficial abrasions and lacerations in varying stages of healing were noted.

Figure 2. An ultrasound of the neck had been performed, which showed adenopathy without suppuration measuring approximately 3.7 cm in diameter. A complete blood count revealed: white blood cell count, 6.3x103/UL; neutrophils, 51%; lymphocytes, 32.3%; monocytes, 11.3%; eosinophils, 4.8%; hemoglobin, 14.3 g/dL; hematocrit, 41.2%; platelet, 504x103/UL. Based on the history and physical findings shown in figure 2, a presumptive diagnosis was made, and a new therapeutic intervention was initiated. A lab test confirmed the diagnosis.

Source: Patrick Hickey, MD, FAAP, DTMH, and Michael Rajnik, MD

Answer

The problem of cervical lymphadenitis is common in pediatrics. For most patients with an acute presentation of unilateral cervical adenitis, initial empiric diagnosis and treatment can be targeted at Staphylococcus aureus, Streptococcus pyogenes or – for patients with significant co-morbid dental disease – anaerobic oral flora. Less common organisms, such as Francisella tularensis or Yersinia pestis, could be considered based on epidemiological clues in the history.

For patients who neither respond nor significantly worsen despite appropriate empiric therapy, as well as those who present subacutely, the differential diagnosis should be expanded to include Bartonella henselae, non-tuberculous mycobacterium (Mycobacterium avium complex and M. scrofulaceum in particular), tuberculosis, toxoplasmosis (typically with generalized adenopathy), actinomycosis and nocardiosis. Nonlymphatic masses caused by congenital anomalies, malignancies and masses of salivary or thyroid gland origin should also be considered.

This patient presented with classic historical and physical exam findings of infection with Bartonella henselae leading to cat scratch disease (CSD), which was confirmed serologically. Detailed questioning revealed that the cat was, in fact, a 6-month-old kitten, which the family had adopted as a flea infested stray several months prior. The patient was known to play roughly with the kitten, resulting in multiple scratches including the papule on his neck, which was at the site of an old scratch. The B. henselae titers were immunoglobulin M 1:32 and IgG 1:512. Pending the serology, the patient was treated empirically with azithromycin and experienced improvement in symptoms over the subsequent week.

Epidemiology

CSD is caused by B. henselae, a fastidious gram-negative bacillus. Humans become infected following inoculation from the scratch or bite of a cat that was infected with Bartonella by fleas. Epidemiological studies have shown that cats typically acquire the infection as kittens and maintain an asymptomatic bactericidal. By adulthood, cats are generally no-longer bacteremic and have serologic evidence of past infection.

Feral cats, which implicitly have a more intense exposure to fleas, are at greatest risk of contracting this infection. By extension, people that adopt, feed or handle feral cats put themselves at increased risk relative to those that have contact with domesticated cats that have proper veterinary care and flea control measures in place.

Clinical manifestations

Classic CSD manifests as localized adenitis, which may be accompanied by fever in approximately one-third of cases. A papule, representing the presumed site of inoculation, is often noted at the site of an old scratch. Axillary adenopathy is most common reflecting inoculation on the hand and forearm. Cervical, epitrochlear and inguinal nodes are also typical sites of infections. Adenopathy resolves spontaneously within two to three months in most cases, but longer times to resolution are also described.

This organism can present in a myriad of rare or unusual ways. An excellent review of this topic, with an emphasis on unusual manifestations organized by organ system, was presented by Massei et al. Parinaud’s oculoglandular syndrome results from inoculation of the eyelid conjunctiva (presumably through direct contact with fleas and their excreta) leading to preauricular lymphadenopathy and ipsilateral granulomatous conjunctivitis.

Uncommon manifestations of CSD include osteomyelitis; hepatosplenic granulomatosis, which may present as fever of unknown origin; encephalopathy; aseptic meningitis; retinitis and culture negative endocarditis. This organism is also the causal agent of bacillary angiomatosis and bacillary peliosis hepatitis (typically in patients with advanced HIV/AIDS).

Diagnosis

Perhaps the most important aspect of diagnosing B. henselae is to consider it as a possibility. Parents may be unaware of cat contacts outside the home, and children may be reluctant to admit to forbidden contacts with feral cats. Clinicians may limit their consideration of Bartonella infection to classic CSD, misdiagnosing its more unusual manifestations. Beyond the identifiable contact history or presence of scratches, papules or adenitis on exam, laboratory evidence of disease can be found using culture, histology, serology and molecular diagnostics.

Bartonella can be cultured from tissue and blood samples; however, yields are low, and growth times required an average of approximately 30 days. Biopsy samples that show granulomas containing bacilli on Warthin-Starry stain reaction suggest the diagnosis but are not pathognomonic. PCR has shown excellent specificity; however, depending on the pretest probability derived from suggestive diagnostic criteria (such as cat contact or a positive serology), sensitivity can be 80% or less on single samples, leading some to recommend three tests on three different tissue sample sites. The most clinically useful and readily available test method is serology. Indirect fluorescent antibody (IFA) titers of greater than 1:64 are considered elevated, but background IgG seropositivity among cat owners leads some to suggest considering a single IgG of greater than 1:512 or a fourfold rise between acute and convalescent samples as reflective of recent disease while others support using an IgG of greater than 1:256.

Enzyme immunoassays are also available commercially. Cross reactivity with other Bartonella species may also occur. So, review of available diagnostics and their interpretation with your laboratory is recommended. In patients thought to have CSD, serologic evidence should be obtained prior to pursuing tissue biopsy or excision of lymph nodes, in which these interventions are not necessary diagnostically or therapeutically.

Management

Treatment of CSD adenitis remains controversial, and optimum therapy has not been well studied. Expectant management without antibiotics as well as efficacy using azithromycin, erythromycin, ciprofloxacin, trimethoprim-sulfamethoxazole and rifampin, has been described. Only one prospective, randomized, double blind, placebo-controlled study has been performed. That study, by Bass et al., examined patients with typical lymphadenitis CSD and showed 50% of azithromycin treated patients achieved an 80% reduction in lymph node size during the 30 days following treatment compared with approximately 7% of placebo control patients. This study used a dosage of 10 mg/kg/day for day one and 5 mg/kg/day for day two to five if patients weighed less than 45.5 kg and 500 mg on day one followed by 250 mg on day two to five for those weighing greater than 45.5 kg. The difference in overall duration of any lymphadenopathy was not statistically significant.

Incision and drainage or excisional biopsy is generally not recommended as part of therapy. Needle aspiration of large suppurating nodes at risk for spontaneous rupture may benefit patients with significant discomfort in functionally sensitive areas, such as the axilla. Analgesia should be provided as needed.

In patients with severe systemic symptoms, particularly people with hepatosplenic disease, endocarditis, severe adenitis and immunocompromised people, treatment is recommended. An oral agent or parenteral gentamicin may be considered in these cases; however, the optimal duration of therapy is not known. Patients with endocarditis should receive an effective therapy that includes gentamicin for at least two weeks. Therapy for patients with bacillary angiomatosis or bacillary peliosis is recommended with azithromycin or doxycycline for several months duration to prevent relapse in the immunocompromised patient.

Prevention

Pet cats should be prophylactically treated for fleas and feral animals avoided. Removing implicated cats from the home is unnecessary as the infectious stage is transient, asymptomatic infections may have already occurred in other family members and simple hygiene of washing scratches will likely reduce the risk of transmission.

For more information:

• Patrick Hickey, MD, FAAP, DTMH, Major, U.S. Army, Assistant Professor of Pediatrics, F. Edward Hebert School of Medicine, Uniformed Services University of the Health Sciences Bethesda, MD.
• Michael Rajnik, MD, FAAP, Major, U.S. Air Force, Assistant Professor of Pediatrics, Director, Pediatric Infectious Disease Fellowship F. Edward Hebert School of Medicine Uniformed Services University of the Health Sciences Bethesda, MD.
• Information related to risk reduction for immunocompromised patients is available at www.cdc.gov/healthypets/extra_risk.htm.
• American Academy of Pediatrics. Red Book: 2006 Report of the Committee on Infectious Diseases. 27th ed. Elk Grove Village, IL: American Academy of Pediatrics, 2006: 246-249.
• Bass JW, Freitas BC, Freitas AD, et al. Prospective randomized double blind placebo-controlled evaluation of azithromycin for treatment of cat-scratch disease. Pediatr Infect Dis J. 1998;17:447-452.
• Demers DM, Bass JW, Vincent JM, et al. Cat-scratch disease in Hawaii: etiology and seroepidemiology. J Pediatr. 1995;127:23-26.
• Hansmann Y, DeMartino S, Piemont Y, et al. Diagnosis of cat scratch disease with detection of Bartonella henselae by PCR: a study of patients with lymph node enlargement. J Clin Microbiol. 2005;43:3800-3806.
• Massei F, Gori L, Macchia P, Maggiore G. The expanded spectrum of bartonellosis in children. Infect Dis Clin North Am. 2005;19:691-711.