Pediatric Puzzler: Fever and abdominal pain (cat-scratch disease)
GEORGE K. SIBERRY, MD, MPH, SECTION EDITOR
Enter stage left: You walk into the emergency room to see your patient, whose presenting symptoms are fever and abdominal pain. But the 3 1/2-year-old boy you find there awaiting you is sitting up, playing, laughing, and acting cheerful. His mother, however, has been cast in a different role: She is, clearly, distraught.
She tells you that her son has been having persistent, spiking fevers, as high as 104° F, every day for the past month. The fevers are accompanied by stomach pains, so severe that she has become fearful, she tells you, that he will die of them. The pain is in his lower abdomen, usually comes on around midnight, lasts two to four hoursand then resolves spontaneously. During the day, his appetite is completely normal. He attends preschool and has displayed a normal activity level. She reports that he has had no change in stooling pattern, no rash, and no vomiting or diarrhea.
The child has been to your ED twice already this month with this complaint, his mother informs you. The first time, he was given a diagnosis of gastroenteritis and a prescription for trimethoprim-sulfamethoxazole (TMP/SMX). At the second visit, stool studies were positive for Clostridium difficile toxin and for Entamoeba coli, and metronidazole was prescribed. Neither medicine had the slightest effect on the fever or abdominal pain, she tells you. (No surprise: TMP/SMX doesn't treat most types of gastroenteritis, C difficile toxin alone does not signal pathogenic infection, and Entamoeba coli isn't a pathogen!) You've heard enough about this drama; it's time for the next act.
On physical examination, the boy's abdomen is completely benign. The remainder of the exam is likewise within normal limits. He is afebrile, and vital signs are normal. You examine the record of his previous visits to the ED. But as you're reading the chart, the nurse summons you: "Come quicklyyou have to see the patient." You rush back to the examining room. He is doubled over, crying, apparently in pain, and trying not to move. You palpate the abdomen; between his bouts of crying and holding the abdomen tightly, there is only softness to your touch. Bowel sounds remain normal. But his temperature is now 101° F. The change in his appearance is striking. You're convincedyou need to admit this child for further workup.
You take a more detailed history from the boy's mother. The family lives on an Indian reservation, and she and her husband work at home, smoking fish over alderwood. They have four other healthy children at home, and dogs who live outside. They have never traveled out of the area. Approximately three months ago, she explains, a dental mirror broke in the house, and the boyyour patientate the pieces! You ruminate: Might this have caused microperforations that led to microabscesses? That could explain the intermittent fevers and severe abdominal pain, but the boy did not have pain thereafter until the current episodethree months after the incident.
By the time you finish taking the detailed history, your patient has fallen asleep. Carefully, you awaken him. Now he denies any abdominal pain whatsoever!
You begin to probe the causes of fever of unknown origin. The main subcategories of culprit disease are rheumatologic, oncologic, and infectious. As for a rheumatologic cause, the erythrocyte sedimentation rate is elevated at 95 mm/hr, but the patient has no joint symptoms, no weight loss, and no other suspect signs. Might it be cancer? The complete blood count is completely normal, and a computed tomographic (CT) scan of the abdomen and pelvis that you ordered does not reveal any adenopathy (neither did your physical exam). You decide to pursue an abdominal cause. You order a KUB radiograph, which is read as showing some stool present but no free air masses. A barium enema is normal. What to do now?
You decide to review the CT scan with a radiologist, who keenly detects several small, hypodense areas in the liver and spleen. You order ultrasonography andsure enoughthere are small lesions consistent with cysts (see figure). Now you can narrow your focus: What causes cysts in the liver and spleen as well as fevers? Echinococcus can cause liver abscesses, but the lesions are usually larger than what this child exhibits. Bacteria can also cause liver abscesses, but you have multiple negative blood cultures to argue against bacterial illness. What about cat-scratch disease? Although lymphadenopathy is a typical finding, your search of the literature reveals that the rare patient can manifest the disease with nothing more than hepatosplenic pathology.
You call the hospital's gastro team, who agree to help. They perform endoscopy and colonoscopyboth are normal. They attempt to aspirate the small hepatic cysts under ultrasound guidance, but fail to retrieve very much fluid. The material they do get is yellowish, and it is sent for culture and susceptibility testing.
You're baffled by the fact that the boy's abdominal pain has disappeared since the bowel clean-out. Looking back at his old films, you can almost convince yourself that constipation played a role in his abdominal pain and that the bowel clean-out before the scoping was the most effective treatment you've offered to him so far.
But what about the month of spiking fevers? Because cat-scratch disease is in the differential diagnosis, you revisit the medical history with the boy's mother. She recalls that, one or two months before this admission, the boy was scratched by a kitten at home. You decide to treat the most likely causecat-scratch diseaseso you begin with gentamicin. The fever drops quickly. You switch to oral ciprofloxacin (having weighed the benefit against the known risk of cartilage damage). He remains afebrile, with no return of the abdominal pain. By the time he is ready to be discharged, among the many lab tests that have finally returned are the cat-scratch disease titers. They are whoppingly positive! When you see the boy during rounds on his last day in the hospital, his mother tells you tearfully: "Thank you, Doctor, for saving my son's life." Rave reviews for your quick-thinking performance!
Cat-scratch disease (CSD) is caused by the slow-growing gram-negative bacterium Bartonella henselae. The typical clinical appearance of CSD is a tender regional lymphadenopathy, with fever and mild systemic symptoms in as many as one third of patients1not the case at all with your little patient! His atypical presentationabdominal pain with fever but without lymphadenopathyhas been clearly described previously.2 An atypical presentation occurs in approximately 10% of patients, in whom hepatosplenic abscess is found less than 1% of the time (that is, not even as often as one of every 1,000 CSD patients).3
Symptoms of CSD develop one to seven weeks after contactusually by way of a bite or scratchwith an immature cat.4 The course is usually self-limited and lasts two to four months.5 Diagnostic testing involves assay of titers of B henselae antibody or culture of the bacterium from tissue or blood (the more difficult route). In cases of hepatosplenic cat-scratch disease, the additional step of an abdominal sonogram allows proper diagnosis.2 The patient whose case is discussed here underwent aspiration of one of the hepatic cysts; in addition, biopsy of the liver might have shown granulomatous hepatitis consistent with CSD.
Most cases of CSDparticularly those in which adenopathy and low-grade fever are the only manifestationscan be managed with symptomatic measures. For severely or acutely ill patients, however, any one of several antibiotics can be given: TMP/SMX (even though this was given to your patient early in the course without success), rifampin, erythromycin, clarithromycin, azithromycin, doxycycline, ciprofloxacin, and gentamicin.1,6,7 Continue treatment until either lymph nodes decrease to approximately 10 mm in diameter, the patient has been afebrile for longer than one week, or he has not exhibited flu-like symptoms for five to 10 days.3 For a patient such as this one, in whom liver and spleen were involved, antibiotic therapy is generally recommended because there is evidence that treatment may shorten the course of illness. In that instance, the antibiotic is continued in the hospital until the fever has subsided and then for two additional weeks as an outpatient treatment. Abdominal ultrasonography is then repeated in one or two months to look for resolution of abscesses.2
Poet (not physician) T. S. Eliot summed it up: "So this is this, and that is that: And there's how you AD-DRESS A CAT."*
*These lines, and the subheadings within the text of this article, arise from Eliot's Old Possum's Book of Practical Cats, a collection of light verse and the literary source of the stage musical Cats.
1. Schutze GE: Diagnosis and treatment of Bartonella henselae infections. Pediatr Infect Dis J 2000;19:1185
2. Dunn HW, Berkowitz FE, Miller JJ, et al: Hepatosplenic cat-scratch disease and abdominal pain. Pediatr Infect Dis J 1997;16:269
3. Smith DL: Cat-scratch disease and related clinical syndromes. Am Fam Physician 1997;55:1783
4. Dangman BC, Albanese BA, Kacica MA, et al: Cat scratch disease in two children presenting with fever of unknown origin: Imaging features and association with a new causative agent, Rochalimaea henselae. Pediatrics 1995;5:767
5. Midani S, Ayoub EM, Anderson B: Cat-scratch disease. Adv Pediatr 1996;43:397
6. 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
7. Windsor JJ: Cat-scratch disease: epidemiology, aetiology, and treatment. Br J Biomed Sci 2001;58:101
George Siberry, ed. Yinka Davies, Etal Etal. Pediatric Puzzler: Fever and abdominal pain.