Disseminated cat scratch disease with vertebral osteomyelitis and epidural abscess.

Abstract: A 5-year-old boy with cat scratch disease presented with fever of unknown origin and osteomyelitis of the thoracic spine and epidural abscess. He did not have localizing signs or symptoms. Computed tomography of the abdomen, which was initially negative, showed hepatosplenic disease. Cat scratch disease has variable systemic presentations and should be included in the differential diagnosis of fever of unknown origin if an epidemiologic risk factor is present.

Key Words: cat scratch disease, epidural abscess, osteomyelitis

**********

Bartonella henselae is a small, fastidious, slow-growing, Gram-negative bacillus that has been identified as the causative agent of cat scratch disease (CSD). Children are more likely to be infected by this organism than adults. Most cases of CSD occur in the fall and winter and may be related to the feline reproductive cycle and flea activity. Kittens younger than 12 months old are 15 times more likely to transmit the disease than adult cats. Kittens are more likely to have B. henselae bacteremia and are more likely to scratch than older cats. B. henselae is transmitted to humans by a kitten scratch, bite, lick, or other intimate contact. The cat flea transmits B. henselae from one kitten to another and may possibly have a role in direct transmission of the organism to humans. (1-5)

Children with CSD typically present with regional lymphadenopathy at or distal to the site of a cat scratch associated with variable systemic manifestations. (1) The disease is usually self-limiting and the course is indolent. However, atypical presentations of CSD occur and include Parinaud oculoglandular syndrome, hepatosplenic disease associated with microabscesses, osteomyelitis, neurologic involvement such as encephalitis and stellate macular retinitis, and various skin manifestations. (1-3) Although cases of CSD associated with osteomyelitis have been described, it is still an uncommon presentation and merits awareness. Previously reported cases of osteomyelitis were associated with local signs and symptoms of the affected area. We present the case of a healthy child with CSD who presented with fever of unknown origin and was found to have thoracic spinal osteomyelitis with epidural abscess as well as hepatosplenic disease.

Case Report

A 5-year-old, previously healthy boy was admitted to Children's Hospital of Michigan with a 3-week history of fever, headache, and intermittent abdominal pain. He had received a course of oral amoxicillin for presumed streptococcal pharyngitis without improvement. Subsequently, he was noted to have torticollis and neck stiffness. He was admitted to an outlying hospital, where a computed tomography (CT) scan of the abdomen was nondiagnostic and a lumbar puncture revealed cerebrospinal fluid findings of 11 white cells/[mm.sup.3] (all monocytes), 13 mg/dL of protein, and 60 mg/dL of glucose. He was diagnosed with viral meningitis. After discharge, his fever spiked to 104[degrees]F, and the abdominal pain continued. The pain was localized to the epigastric area, was intermittent, and at times was severe enough to wake him from sleep. During the 5 days before admission, he developed abdominal distension. His medical history was unremarkable and his immunizations were up to date. Physical examination revealed a healthy-appearing child. The temperature was 104[degrees]F, pulse rate was 85/min, respiration rate was 18/min, and blood pressure was 106/71 mm Hg. His general physical examination was normal except for a palpable liver edge. The white blood cell count was 10,400/[mm.sup.3] (bands, 5%; neutrophils, 64%; lymphocytes, 22%; monocytes, 6%; eosinophils, 2%), hemoglobin level was 10.2 g/dL, and platelet count was 668,000/m[m.sup.3]. Blood, urine, and spinal fluid cultures were negative. A chest radiograph was normal.

A body gallium scan showed an area of increased activity in the upper thoracic spine. Magnetic resonance imaging revealed a mass on the right of the T4-T7 vertebrae, suggestive of epidural abscess with associated vertebral bone changes, suggestive of osteomyelitis (Figure). The patient was started on intravenous ceftriaxone and vancomycin treatment.

He underwent surgical resection of a vascular-appearing epidural mass. Incision of the mass revealed yellow pus, a sample of which was sent for cultures. CT of the chest and abdomen on the fifth postoperative day revealed multiple low attenuation areas throughout the liver and spleen compatible with microabscesses. Permeative changes in the posterior aspect of the body and right pedicle of T5 vertebra suggestive of osteomyelitis were also noted. On microscopic examination, the mass showed organizing inflammation with focal necrosis and abscess formation consisting of neutrophils, lymphocytes, and macrophages. Bony fragments showed aggregates of neutrophils suggestive of necrosis and osteomyelitis. Gram, fungal, acid-fast, and Warthin-Starry stains were nonrevealing. Aerobic and anaerobic bacterial cultures as well as mycobacterial and fungal cultures were negative. Serum antibody titers for Epstein-Barr virus, cytomegalovirus, Toxoplasma gondii, Histoplasma capsulatum, and Blastomycosis dermatitidis were negative. B henselae serum titers showed negative immunoglobulin (Ig)M but positive IgG titer at 1:1,024. Intravenous ceftriaxone and vancomycin were stopped, and the patient was discharged on oral clarithromycin treatment. Further questioning revealed that a stray cat had delivered 3 kittens in the family home 4 months before onset of the patient's symptoms. After 2 weeks, treatment was changed to oral trimethoprim/sulfamethoxazole, and he completed a total of 10 weeks of oral antibiotic therapy. Follow-up magnetic resonance imaging of the thoracic spine 4 weeks after surgery showed mild residual soft tissue enhancement with no evidence of mass or swelling. He remained asymptomatic, and the hepatomegaly resolved. B henselae titer 8 weeks after the initial serology showed a decrease of IgG titer to 1:256.

Discussion

Patients with CSD generally present with a history of a scratch, bite, or close contact with a cat or kitten. Typical disease in immunocompetent individuals includes the development of a papule 3 to 10 days after the inoculation of the organism followed by regional, self-limited lymphadenopathy. (5) Atypical manifestations occur in 5 to 25% of cases. (1,2) The most common atypical manifestation of CSD is Parinaud oculoglandular syndrome consisting of unilateral conjunctivitis and ipsilateral preauricular lymphadenopathy. Other atypical clinical manifestations in immunocompetent patients include disseminated CSD with hepatosplenic abscesses, endocarditis, Leber stellate neuroretinitis, and CSD encephalopathy. (1-3) In immunocompromised individuals, the tendency for hematogenous dissemination is greater, and patients may have development of bacillary angiomatosis in different organs, including peliosis hepatis. B henselae stimulates endothelial proliferation, producing vasoformative lesions that can enlarge, forming large tumors in the skin, bone, bone marrow, respiratory or gastrointestinal tract, lymph nodes, or the central nervous system. Large blood-filled cysts may also form in the liver or spleen in severely immunocompromised patients. (6)

Our patient had disseminated CSD with atypical manifestations, namely osteomyelitis of the spine and epidural abscess, in addition to hepatosplenic microabscesses. The diagnosis was confirmed by elevated B henselae IgG titer.

The absence of adenopathy at the time of presentation, the elevated B henselae IgG-specific titer, and the absence of IgG-specific titer indicate that the infection had progressed over a period of time.

Bone involvement in CSD is uncommon. In 23 previously published cases of osteomyelitis associated with CSD, (7-11) the vertebrae were most commonly involved. Osteomyelitis involving the skull, sternum, ilium, rib, femur, tibia, and metatarsals has also been reported. Single focus or multifocal osteomyelitis can occur. Concomitant lymphadenopathy has been a common finding; however, osteomyelitis may develop after several weeks of lymphadenopathy. A variety of antibiotics were used for treatment including those not known to be effective against B henselae. In some cases, complete healing was reported after surgery alone. Prognosis was excellent in all cases.

A severe case of osteomyelitis was reported by Modi et al. (7) in a 4-year-old girl who had six sites of bony involvement involving the axial skeleton and ribs associated with chest wall abscess. She was treated with azithromycin and rifampin for 1 month after drainage of the abscess.

[FIGURE OMITTED]

The diagnosis of CSD is based on clinical features, positive serology, and tissue findings of granulomas and multinucleated giant cells. (5,12,13) Isolation of B henselae by culture is very difficult. Warthin-Starry stain has been used to identify Bartonella bacilli in tissue samples, including vascular masses associated with angiomatosis. Histology and staining cannot reliably distinguish between species of Bartonella. The most reliable and widely used test to diagnose CSD is the detection of IgM and IgG antibodies to Bartonella, using immunofluorescent assay. (14) Diagnosis can readily be made by using commercially available immunofluorescent assay kits (Focus Technologies, Cypress, CA) that are highly sensitive (up to 93%) and specific (98%) for IgG testing. A titer of IgG greater than 512, or 4-fold rise of IgG titer over 2 to 4 weeks, has been proposed as diagnostic of acute CSD. (14) Recently, polymerase chain reaction (PCR) has been shown to be superior to histology or culture for B henselae detection in tissues. PCR to several gene-specific DNA sequences to Bartonella is extremely accurate for diagnosis. PCR can also be used to differentiate different genotypes within the species of B henselae. (15) However, PCR testing is not readily available. The pathologic changes in the epidural inflammatory mass in our patient were consistent with a nonspecific inflammatory response consisting of acute and chronic inflammatory cells. These findings, together with the absence of localized lymphadenopathy and the presence of B henselae IgG and absence of IgM in the serum, all point to delayed manifestations of CSD-associated osteomyelitis.

Necrotizing granulomas of the liver and spleen associated with CSD have been increasingly implicated in the pathogenesis of fever of undetermined orgin in children. (16-18) Associated abdominal pain, usually severe and episodic, is reported in 60% of cases of hepatosplenic CSD. (18) Fever has been reported in all patients diagnosed with hepatosplenic CSD. Liver involvement is more frequently diagnosed than splenic involvement, and hepatomegaly is seen in 31% of cases. Our patient had an initial CT scan of the abdomen that was normal. However, he had subsequent development of hepatomegaly, and a follow-up CT of the abdomen showed evidence of hepatosplenic hypodense lesions. These findings suggest that hepatosplenic involvement may be an evolving process during the course of systemic CSD, and radiographic findings of hepatosplenic CSD may lag behind clinical manifestations. (5,19) In hepatosplenic CSD, the natural course is for the immune system to clear the microabscesses over a period of weeks or months. (20)

Bass et al (21) suggested that azithromycin may hasten the recovery of cervical lymphadenopathy associated with CSD. No prospective and controlled studies are available regarding the role of antibiotics in the treatment of immunocompetent children with CSD-associated osteomyelitis or hepatosplenic disease. However, some observational studies have suggested that antibiotic therapy may shorten the course of illness in hepatosplenic CSD. (17,18) Patients with severe CSD have been treated with different antibiotics including azithromycin, doxycycline, rifampin, or trimethoprim/sulfamethoxazole with rifampin. Rifampin may shorten the duration of fever in hepatosplenic CSD. Other antibiotics that have also been shown to be effective are ciprofloxacin and gentamicin. Only gentamicin has been shown to be bactericidal in vitro. (22)

Our findings indicate that osteomyelitis associated with CSD can be an indolent process that may not be associated with specific signs and symptoms. CSD osteomyelitis should be included in the differential diagnosis of vertebral osteomyelitis in children. The destructive bony changes and the presence of inflammatory mass in our patient and in other reports indicate that the inflammatory process can progress at low grade over time.

Conclusion

Although the most common presentation of CSD is regional adenopathy, it may present as fever of unknown origin. CSD can be a systemic disease rather than a localized infectious process. Disseminated infection in the form of hepatosplenic disease may occur with other organ-system involvement such as spinal osteomyelitis and epidural abscess, as in our case. Hepatosplenic disease may be missed early in the disease process. CSD should always be suspected in patients with fever of unknown origin if an epidemiologic risk factor such as exposure to cats or kittens is present.

References

1. Carithers HA. Cat-scratch disease: an overview based on a study of 1,200 patients. Am J Dis Child 1985;139:1124-1133.

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

3. Massei F, Messina F, Talini I, et al. Widening of the clinical spectrum of Bartonella henselae infection as recognized through serodiagnostics. Eur J Pediatr 2000;159:416-419.

4. Anderson BE, Neuman MA. Bartonella spp as emerging human pathogens. Clin Microbiol Rev 1997;10:203-219.

5. Zangwill KM, Hamilton DH, Perkins BA, et al. Cat scratch disease in Connecticut: epidemiology, risk factors, and evaluation of a new diagnostic test. N Engl J Med 1993;329:8-13.

6. Spach DH, Koehler JE. Bartonella-associated infections. Infect Dis Clin North Am 1998;12:137-155.

7. Modi SP, Eppes SC, Klein JD. Cat-scratch disease presenting as multifocal osteomyelitis with thoracic abscess. Pediatr Infect Dis J 2001;20:1006-1007.

8. Robson JM, Harte GJ, Osborne DR, et al. Cat-scratch disease with paravertebral mass and osteomyelitis. Clin Infect Dis 1999;28:274-278.

9. Ulzebos CV, Koetse HA, Kimpen JL, et al. Vertebral osteomyelitis associated with cat-scratch disease. Clin Infect Dis 1999;28:1310-1312.

10. Keret D, Giladi M, Kletter Y, et al. Cat-scratch disease osteomyelitis from a dog scratch. J Bone Joint Surg Br 1998;80:766-767.

11. Fernandez M, Carrol CL, Baker CJ. Discitis and vertebral osteomyelitis in children: an 18-year review. Pediatrics 2000;105:1299-1304.

12. Nadal D. Zbinden R. Serology to Bartonella (Rochalimaea) henselae may replace traditional diagnostic criteria for cat-scratch disease. Eur J Pediatr 1995;154:906-908.

13. Muszynski MJ, Eppes S, Riley HD Jr. Granulomatous osteolytic lesion of the skull associated with cat-scratch disease. Pediatr Infect Dis J 1987;6:199-201.

14. Sander A, Berner R, Ruess M. Serodiagnosis of cat scratch disease: response to Bartonella henselae in children and a review of diagnostic methods. Eur J Clin Microbiol Infect Dis 2001;20:392-401.

15. Zeaiter Z, Fournier PE, Raoult D. Genomic variation of Bartonella henselae strains detected in lymph nodes of patients with cat scratch disease. J Clin Microbiol 2002;40:1023-1030.

16. Jacobs RF, Schutze GE. Bartonella Henselae as a cause of prolonged fever and fever of unknown origin in children. Clin Infect Dis 1998;26:80-84.

17. Dunn MW, Berkowitz FE, Miller JJ, et al. Hepatosplenic cat-scratch disease and abdominal pain. Pediatr Infect Dis J 1997;16:269-272.

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

19. Estrada B, Silio M, Begue RE, et al. Unsuspected hepatosplenic involvement in patients hospitalized with cat-scratch disease. Pediatr Infect Dis J 1996;15:720-721.

20. Lenoir AA, Storch GA, DeSchryver-Kecskemeti K, et al. Granulomatous hepatitis associated with cat scratch disease. Lancet 1988;1:1132-1136.

21. 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.

22. Musso D, Drancourt M, Raoult D. Lack of bactericidal effect of antibiotics except aminoglycosides on Bartonella (Rochalimaea) henselae. J Antimicrob Chemother 1995;36:101-108.

It is our choices ... that show what we truly are, far more than our
abilities.
--J. K. Rowling

Nahed Abdel-Haq, MD, Walid Abuhammour, MD, Hossam Al-Tatari, MD, and Basim Asmar, MD

From Children's Hospital of Michigan and Wayne State University School of Medicine, Detroit, MI, and Hurley Medical Center, Michigan State University, Flint, MI.

Reprint requests to Dr. Nahed Abdel-Haq, Children's Hospital of Michigan, 3901 Beaubien Blvd., Detroit, MI 48201. Email: [e-mail]

Accepted January 14, 2005.

RELATED ARTICLE: Key Points

* Cat scratch disease may be a systemic disease rather than localized lymphadenitis.

* In addition to hepatosplenic involvement, disseminated cat scratch disease may include focal bone and nervous system disease.

* Cat scratch disease should be suspected in children with fever of unknown origin and epidemiologic risk factors such as cat exposure.