Symptom-Based Diagnosis in Pediatrics (CHOP Morning Report) 1st Ed.

CASE 19-5

Eleven-Year-Old Boy




An 11-year-old boy was well until the day of admission, when he stood up and fell, striking his head on a desk at school. Approximately 1 hour later, he developed head twitching, eye blinking, and tonic-clonic movements of his right arm that lasted approximately 20 minutes. During this episode, he had bowel and bladder incontinence. The school nurse found him confused, combative, and unable to follow simple commands. He was taken to the emergency department for evaluation. His family reported that the patient had been intermittently febrile during the past 2 weeks but denied any history of headache, vomiting, rash, visual problems, alteration in gait, travel, ill contacts, and use of alcohol or illicit drugs.


The boy did not have any underlying medical conditions. His growth and development had been normal. His older brother had died from complications related to acquired immunodeficiency syndrome several years earlier, after becoming infected with human immunodeficiency virus (HIV) through intravenous drug use. There was no family history of seizures, metabolic disorders, or sickle cell disease. His maternal grandmother and father both died of complications related to hypertension. He lived at home with his mother and stepfather. They had two healthy kittens at home but no reptiles or birds. He had received all required immunizations, including the meningococcal conjugate vaccine. He did not require any medications.


T 38.6°C; HR 110 bpm; RR 18/min; BP 136/80 mmHg; SpO2, 96% in room air

Weight 50th percentile; Height 75th percentile

On examination, he was agitated and combative. There was a small contusion over his left eyebrow but no other signs of trauma. There were no scleral hemorrhages. There was no hemotympanum. Pupils were symmetrically reactive (6 mm-4 mm). His neck was supple. There was a 2-cm right cervical lymph node without surrounding erythema or drainage (Figure 19-6). There were no murmurs or rubs on cardiac examination. Lungs were clear to auscultation. No other lymphadenopathy was appreciated. His abdomen was soft without hepatosplenomegaly. He was incontinent of urine and stool. There were no petechiae. Neurologic assessment was difficult to obtain. The child cried and shouted but his speech was unintelligible. He did not follow simple commands. He thrashed his arms and legs spontaneously and purposefully.


FIGURE 19-6. Photo of the patient’s neck.


Complete blood count revealed 5300 WBCs/mm3 (51% segmented neutrophils, 32% lymphocytes, 13% monocytes, and 5% eosinophils); hemoglobin, 11.6 g/dL; and 333 000 platelets/mm3. Serum ALT and AST were 152 U/L and 114 U/L, respectively. Serum electrolytes, blood urea nitrogen, calcium, magnesium, phosphorus, creatinine, and albumin were normal. Serum glucose was 106 mg/dL. Serum and urine toxicology screens were negative. CT of the head did not reveal any hemorrhage, mass lesions, or ventriculomegaly. Examination of the CSF revealed 24 WBCs/mm3; 1 RBC/mm3; glucose, 77 mg/dL; and protein, 19 mg/dL. There were no organisms on Gram staining of the CSF.


The patient was admitted to the intensive care unit and received vancomycin, cefotaxime, and acyclovir as empiric therapy for encephalitis while the results of additional CSF testing were pending. Two hours after admission, he had several additional right-sided motor seizures with occasional generalization. He received phenytoin and multiple doses of lorazepam, with resolution of each seizure. He ultimately required endotracheal intubation for respiratory failure. The EEG showed a right temporal focus that spread mostly through the right hemisphere, with occasional spread to the left hemisphere. Continuous EEG monitoring demonstrated a total of 75 clinical and subclinical seizures over the next 24 hours, despite the addition of Tegretol (carbamazepine), valproate, and phenobarbital. CSF PCR for Borrelia burgdorferi (Lyme disease), HSV (repeated twice), and enterovirus were negative, as were cultures for bacteria and stains for fungi and acid-fast bacilli. Viral culture of the CSF was also ultimately negative. Serum acute and convalescent IgM and IgG titers were undetectable for Lyme disease, Rocky Mountain spotted fever, and human granulocytic ehrlichiosis. Metabolic studies including serum and urine organic acids, serum amino acids, and serum and CSF lactate and pyruvate were within normal limits. Findings on physical examination combined with results of the abdominal CT suggested a diagnosis (Figure 19-7).


FIGURE 19-7. CT of the patient’s abdomen demonstrates lesions in the A. Spleen. B. Liver.



Encephalitis occurs in 0.5 per 100 000 individuals in the United States. In late childhood, a variety of agents cause encephalitis, although viruses are most commonly implicated. HSV is a serious but potentially treatable cause of viral encephalitis. Viral culture is frequently negative in cases of HSV encephalitis; therefore, HSV PCR is considered the gold standard diagnostic test. Enteroviruses are the most common cause of CNS infection in children, but predominant encephalitis without meningitis is an unusual manifestation. Encephalitis occurs in approximately 1 of every 1000 cases of Epstein-Barr virus-associated mononucleosis; children frequently have a history of antecedent fatigue and pharyngitis. Other viral causes of encephalitis in the United States include human herpesvirus 6, arboviruses (Eastern and Western equine encephalitis), hepatitis viruses A and B, HIV, and rabies virus. Postinfectious encephalitis often occurs after respiratory infections, especially those caused by influenza virus or M. pneumoniae. The incidence of encephalitis related to measles, mumps, and varicella has declined due to widespread immunization. Encephalitis in the summer months should raise suspicion for Rocky Mountain spotted fever, Lyme neuroborreliosis, and enterovirus infection. Exposure to kittens raises the possibility of B. henselae encephalitis. Few pyogenic bacteria cause encephalitis without overt meningitis. Syphilis, leptospirosis, brucellosis, tuberculosis, and listeriosis are rare causes of encephalitis but should be considered when an appropriate exposure has been documented.


The abdominal CT revealed multiple hypodense hepatic and splenic lesions (Figure 19-7). These findings, in conjunction with cervical lymphadenopathy (Figure 19-6) and history of contact with a kitten, suggested infection with B. henselae, the causative agent of cat-scratch disease (CSD). Bartonella henselae antibody titers were 1:2,048. These findings supported the diagnosis of hepatosplenic CSD with encephalitis. The patient’s seizures continued for more than 1 week despite aggressive anticonvulsant therapy. He received rifampin and macrolide antibiotics for his CSD. Unfortunately, his hospital course was complicated by disseminated fungal infection, including Candida albicans retinitis and meningitis, and chronic respiratory failure requiring a tracheostomy. He was discharged after 8 months of hospitalization with impaired vision, developmental delay, and recurrent seizures. His poor neurologic outcome was most likely related to prolonged seizures and disseminated candidiasis.


CSD, a well-recognized and self-limited cause of regional lymphadenitis in children, is often associated with systemic symptoms such as fever and malaise. Approximately 24 000 cases are reported in the United States each year; the highest age-specific incidence occurs in children younger than 10 years of age. CSD predominantly occurs in the fall and winter, possibly related to temporal changes in animal behavior and reproduction. History of contact with a cat can be established in more than 90% of cases of CSD, and clustering of cases has occurred within families upon acquisition of new pet cats. The infection may be transmitted through licks, scratches, or bites. Cats which transmit CSD are not ill and have no distinctive features, although most are kittens (younger than 1 year of age). In a study of 1200 patients with CSD, 64% reported contact with only a kitten, and 25% reported contact with both kittens and adult cats. Up to 50% of domestic cats demonstrate antibodies to B. henselae.

Approximately 10% of children with CSD develop atypical features, including Parinaud’s oculoglandular syndrome, neuroretinitis, erythema nodosum, pulmonary nodules, osteomyelitis, and encephalitis or encephalopathy. Encephalitis or encephalopathy complicates approximately 2% of cases of CSD.


The clinical manifestations of CSD depend on the site of inoculation and infection. Fever and localized, tender lymphadenopathy is the most common presentation of CSD. The most frequently involvedsites include the axillary and epitrochlear nodes (46%), the neck and jaw (28%), and the groin (18%).

Children with hepatosplenic CSD often complain of fever, malaise, and abdominal pain. Physical examination findings include hepatomegaly (30%), hepatosplenomegaly (16%), and lymphadenopathy (25%)—usually a single enlarged lymph node. A scratch or papule identifies the probable site of inoculation in 60%-90% of cases. Recently, B. henselae has been identified as an important cause of fever of unknown origin (FUO), accounting for up to 5% of patients with FUO.

Children with CSD-associated encephalitis often present with complaints of fever, malaise, and lymphadenopathy. Two to three weeks later, they develop headache, alterations in mental status, and seizures. Patients may also present with weakness, nuchal rigidity, and alterations in tone and reflexes. Less common manifestations of this rare event include hallucinations, hemiplegia, aphasia, cerebellar ataxia, and sixth cranial nerve palsy. In Florida, a cluster of five children with acute CSD encephalopathy occurred within a 6-week period in 1994. All presented with status epilepticus and required endotracheal intubation but subsequently recovered without sequelae. In one series of 76 patients with neurologic symptoms attributable to CSD, encephalopathy occurred 1-6 weeks after onset of lymphadenopathy; seizures occurred in 46% of these patients. More than 90% of children with CSD encephalitis or encephalopathy experience complete and spontaneous resolution of symptoms, although recovery has often taken up to 8 weeks and can take as long as 1 year.


Bartonella henselae IgG antibodies. Serologic testing is the standard method of diagnosis and should be considered for patients who present with status epilepticus, adenopathy, and history of feline contact. A single elevated indirect immunofluorescence assay titer or enzyme immunoassay value for IgG antibodies is usually sufficient to confirm CSD, because initiation of a humoral immune response usually precedes or is concurrent with symptom onset. IgG levels rise during the first 2 months after onset of illness, followed by a gradual decline. The sensitivity is reported to be greater than 98%; however, IgG titers remain elevated for up to 1 year following acute infection, often complicating the identification of acute infection. An IgG titer greater than 1:64 is consistent with recent infection. An increase in antibody titers between acute and convalescent (2 weeks later) may be confirmatory. Paired sera are particularly useful in cat owners who may have a background rate of B. henselae antibody seropositivity greater than the general population. IgM levels are less useful in the diagnosis of CSD, since IgM antibodies have a short duration of detection and have frequently cleared by the time of presentation.

Bartonella henselae DNA polymerase chain reaction. PCR may be performed on tissue samples (lymph node aspirates, liver or spleen tissue, bone biopsies). It is available on a limited basis at some commercial laboratories as well as from the Centers for Disease Control and Prevention. PCR has specificity as high as 100% in the identification of B. henselae; however, sensitivity ranges from 43% to 76%, making it less useful in uncomplicated cases.

Lumbar puncture. CSF analysis may be normal, but some patients have a mild pleocytosis (fewer than 30 WBCs/mm3). CSF protein may be slightly elevated, but CSF glucose is normal. PCR assays to exclude other potential causes of encephalitis, including HSV, enteroviruses, and arboviruses, should be considered.

Abdominal CT. Abdominal CT is usually not warranted but may demonstrate granulomata in the liver and spleen which, as in this case, may help establish the diagnosis.

Central nervous system imaging. Head CT is normal but should be performed to exclude CNS lesions that can cause seizures. Head MRI may reveal focal or diffuse white matter changes that are not specific for CSD.

Other studies. Bartonella henselae can be cultured from blood, lymph nodes, and other tissues, but it grows slowly and requires a 6-week incubation period. Additionally, isolation of B. henselae from tissues can be difficult, particularly in patients with limited disease. A presumptive diagnosis can be made based on microscopic examination of tissues by demonstrating the bacilli with the use ofWarthin-Starry silver impregnation staining. Other histologic findings depend on duration of illness. Early findings in affected lymph nodes include lymphocytic infiltration and epithelioid granuloma formation. Later changes include neutrophil infiltration and necrotizing granulomas. The cat-scratch skin test is no longer used. Studies to exclude other causes of encephalopathy should be performed.


In the immunocompetent child with hepatosplenic CSD, the benefit of antimicrobial therapy is unclear, because the illness is self-limited in most cases. Patients who receive rifampin appear to have more rapid resolution of fever, but no controlled trial had been conducted, so the efficacy of any antibiotic regimen in the treatment of hepatosplenic CSD is not known. In a study by Bass et al., a 5-day course of azithromycin hastened resolution of lymphadenopathy, causing an 80% reduction in lymph node volume in 50% of azithromycin-treated patients compared with 7% of patients receiving placebo. However, there was no difference in time to complete resolution between the azithromycin- and placebo-treated groups. Although antimicrobial therapy is not recommended for mild-to-moderate infection, azithromycin can be considered in patients with extensive lymphadenopathy. Trimethoprimsulfamethoxazole, erythromycin, clarithromycin, doxycycline, ciprofloxacin, and gentamicin may also be effective in treating patients with severe illness. Beta-lactam agents appear ineffective despite favorable in vitro susceptibility. Immunocompromised patients with CSD show a much more significant response to antimicrobial therapy, and these patients should receive antibiotics for a minimum of 3 months. Erythromycin ethylsuccinate is the preferred agent in children, but doxycycline, isoniazid, azithromycin, gentamicin, and rifampin have shown benefit.

Experience with antimicrobial therapy in CSD encephalitis is limited to anecdotal reports. Most children appear to recover without CSD-specific therapy, and conservative, symptomatic management, including anticonvulsants for seizure activity, is recommended. The poor outcome for the patient in this case was probably related to disseminated fungal infection rather than CSD.


1. Arisoy ES, Correa AG, Wagner ML, Kaplan SL. Hepatosplenic cat-scratch disease in children: selected clinical features and treatment. Clin Infect Dis. 1999;28: 778-784.

2. Armengol CE, Hendley JO. Cat-scratch disease encephalopathy: a cause of status epilepticus in school-aged children. J Pediatr. 1999;134:635-638.

3. 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.

4. Bass JW, Vincent JM, Person DA. The expanding spectrum of Bartonella infections: II. Cat-scratch disease. Pediatr Infect Dis J. 1997;16:163-179.

5. Carithers HA, Margileth AM. Cat-scratch disease: acute encephalopathy and other neurologic manifestations. Am J Dis Child. 1991;145:98-101.

6. Florin TA, Zaoutis TE, Zaoutis LB. Beyond cat scratch disease: widening spectrum of Bartonella henselae infection. Pediatrics. 2008;121(5):e1413-e1425.

7. Klotz SA, Ianas V, Elliott SP. Cat-scratch disease. Am Fam Physicians. 2011;83(2):152-155.

8. Margileth AM. Antibiotic therapy for cat-scratch disease: clinical study of therapeutic outcome in 268 patients and review of the literature. Pediatr Infect Dis J. 1992;11:474-478.

9. Noah DL, Bresee JS, Gorensek MJ, et al. Cluster of five children with acute encephalopathy associated with cat-scratch disease in South Florida. Pediatr Infect Dis J. 1995;14:866-869.