March 2002

A 10-year-old girl presented to a clinic in Sierra Vista, Ariz., for evaluation of a fever of unknown origin. Some malaise, occasional loose nonbloody stool, slight cough and mild sore throat accompanied this fever. The onset was five days prior to presentation. The initial visit failed to reveal anything unusual, but her fever persisted. Her initial lab tests included a throat culture that was positive for nongroup A streptococcus, a complete blood count that had a white count of 6,700 with 70% neutrophils, 0 bands, 18% lymphocytes, 2% atypical lymphocytes and 334,000 platelets. A urinalysis was negative as were cultures of blood and urine. Her chest radiograph was clear, and a variety of serologic tests and chemistries were done over the course of the next two weeks that were all negative, including Epstein-Barr virus, mycoplasma, parvovirus, coccidioides immitis, antinuclear antibody, rheumatoid factor and liver enzymes. However, her erythrocyte sedimentation rate was 80. She also had a negative purified protein derivative skin test. She was initially treated with clindamycin about nine days into the illness because of the sore throat.

Her past medical history and family history were unremarkable. Social history was positive for numerous sick contacts at school, most of whom were diagnosed with Mycoplasma pneumoniae. Based on this, she was empirically treated with azithromycin (Zithromax, Pfizer) at about two weeks into the illness, without apparent benefit.

There’s no travel out of Arizona or insect (tick) exposure, and no pets at home. However, her aunt had a kitten that she thinks scratched her about three months earlier.

Seventeen days after the onset she was admitted to the hospital because of ongoing fever and vague abdominal complaints. Her examination revealed a fever of 104.5° F and some mild abdominal pain. Otherwise, she appeared normal. Specifically, there were no scratch marks, lymphadenopathy, rashes or weight loss. An abdominal ultrasonography was normal. An abdominal CT scan is shown in figures 1–5. Notice close-ups of her liver and spleen in figures 4 and 5.

What's Your Diagnosis?

1. Cat scratch disease
2. Disseminated coccidioidomycosis
3. Disseminated histoplasmosis
4. Amebiasis

Answer

The answer is #1, cat scratch disease (CSD). When the abdominal CT scan results were known, Bartonella henselae titers were done revealing an immunoglobulin G (IgG) of 1:2848 and an IgM of 1:16. This case serves to illustrate one of the varied ways cat scratch disease can present. I don’t think that too many people would have ordered Bartonella titers early on with this child’s history and physical findings. It makes you wonder how many similar cases are missed that don’t go beyond the normal abdominal ultrasound without doing a CT scan.

Hepatosplenic cat scratch disease is being reported more frequently as CSD is being considered more frequently in the evaluation of fevers of unknown origin in children and diagnostic capability improves.

By now, most everyone is familiar with the chronic adenitis (figures 6 and 7) and oculoglandular syndrome (figures 8 and 9) associated with CSD, and has been discussed in this column on a couple of occasions before. However, CSD can be associated with a variety of other findings. CSD should be at least considered in all children undergoing evaluation for fever of unknown etiology, encephalitis and aseptic meningitis, granulomatous hepatitis, microabscesses in the liver and/or spleen with or without hepatosplenomegaly, osteolytic bone lesions, internal lymphadenitis (mesenteric, retropharyngeal, etc.), subacute iritis, uveitis, idiopathic thrombocytopenic purpura, erythema nodosum, thyroiditis and pneumonia. All these entities have been associated with or implied to be caused by CSD. Improved serologic techniques are being used to confirm the cause of these entities with CSD. A negative cat exposure history should never be used to rule out CSD.

Hepatosplenic CSD was recently reviewed by Estrada, et al in The Pediatric Infectious Disease Journal (1996;15:720-721). I would recommend this paper to read more about this entity. For some treatment data, I would also recommend the only randomized, double blind, placebo-controlled treatment study to date, by Bass, et al, in the same journal (1998;17:447-452).

Disseminated histoplasmosis and coccidioidomycosis can be a cause of extrapulmonary disease, but this is primarily bone and joint, cutaneous and central nervous system involvement, not hepatosplenic. Dissemination is a fairly rare condition in a normal child, but can be seen in normal infants. Histoplasmosis can be ruled out by a lack of travel since it is primarily seen in the southeastern and central United States. Coccidioidomycosis, however, is endemic in Arizona. But with a negative IgM and IgG serology, that diagnosis was unlikely.

Amebiasis can certainly cause liver abscesses, but is not likely to result in splenic abscesses. It can be seen in the United States, but is more likely to be imported from underdeveloped countries and/or from lower socioeconomic groups. This child’s history and splenic involvement mitigated against that diagnosis.

Acknowledgement: Thanks go to Dr. Lisa Cunningham and Dr. Brent Barber of Sierra Vista, Ariz., for contributing this case.

For Your Information:

• James H. Brien, DO, Pediatric Infectious Disease, Scott and White's Children's Health Center and Texas A&M University, College of Medicine, Temple, Texas. E-mail: jhbriend@aol.com