May 2002

A 4-month-old boy presented to the clinic with fever and a small, red bump on his abdomen. The bump first appeared a few days earlier and fever was first noted yesterday. The bump began as an erythematous papule that progressed to a vesicle or pustule that spontaneously drained, revealing some yellowish fluid by history.

He was otherwise in his usual state of health, which is severely compromised by a thoracic spina bifida with the associated neurologic deficits. He also has a ventriculoperitoneal (VP) shunt due to hydrocephalus, which has been revised multiple times. His most recent revision was about three weeks ago. His immunizations are up to date, and he has no other complaints.

His examination reveals him to be a small-for-age 4-month-old male with the above problems. It is noted that he has a sore on his abdominal wall as shown in figures 1 and 2. His admitting lab tests included a complete blood count that revealed a white blood cell (WBC) count of 26,000 with a “left shift.” After further exam of the bump on the abdomen, it was apparent that this was actually the VP shunt eroding through the abdominal wall (figure 3). His abdominal exam was otherwise unremarkable with normal bowel sounds. Further lab tests revealed that the cerebrospinal fluid (CSF) taken from the end of the shunt had no red blood cells but 185 WBCs with a predominance of segmented neutrophils, a protein and glucose of 65 and 43 mg/dl, respectively. The CSF Gram’s stain showed numerous gram-positive cocci in clusters and rare coccobacilli.

What’s Your Diagnosis?

1. Nafcillin + gentamicin
2. Ampicillin + gentamicin + clindamycin
3. Vancomycin + an aminoglycoside
4. Vancomycin alone

Answer

You may disagree, but I feel like the best answer is #3, vancomycin plus an aminoglycoside. This is the choice recommended by most experts for VP shunt infections until the etiology is determined by culture. Many recommend using a third-generation cephalosporin in place of the aminoglycoside for better penetration through the blood-brain barrier and less toxicity potential. I think this is just as acceptable as the choice offered above.

The fact that this infection was associated with the shunt eroding through the abdominal wall should not alter these recommendations provided the patient does not have any evidence of peritonitis. In that event, it might be reasonable to add clindamycin to treat anaerobes that may be contributing to the peritonitis. This appeared not to be the case in this patient. His CSF culture grew Staphylococcus epidermidis and a few colonies of two types of Corynebacterium species, neither diphtheria nor jeikeium (group JK), which were likely contaminants. Coagulase-negative staph is by far the most common cause of VP shunt infections, and the rate of methicillin (and nafcillin) resistance is so high, vancomycin is the most widely recommended empiric choice. One needs to remember that the recommended dose for central nervous system infections is higher than other infections treated with vancomycin; usually starting at about 60 mg/kg/day ÷ q 6 hours. Gram-negative rods can occasionally be the cause, therefore, most experts recommend adding an aminoglycoside or third-generation cephalosporin pending culture results.

In this case the entire device was removed and the patient underwent serial ventricular taps until the CSF returned to normal (about three weeks), at which time a new VP shunt was placed. Most patients don’t need to be treated that long before replacing the shunt. This is generally left up to the neurosurgeon, and some are much more conservative than others. However, most would recommend replacing the infected shunt with a ventriculostomy tube first and place a new shunt seven to 10 days later if the infection is clearing. Many neurosurgeons externalize the shunt at the abdomen instead (figure 4) pending clearing the infection. The problem with VP shunt infection management is a paucity of good scientific data from which to base recommendations. These cases are usually best managed with the help of an infectious disease specialist or a pediatrician familiar with shunt infections, along with the neurosurgeon. The patient presented above remains well, with the shunt still functioning after two years of follow up.

An excellent resource for a review of this problem is Ram Yogev, Central Nervous System Shunt-Related Infections, Chapter 53, Pediatric Infectious Diseases Principles and Practice, second edition, 2002, Jenson & Baltimore, W.B. Saunders Co. Dr. Yogev is not only a recognized expert in the area of shunt infections, he is also a member of the editorial board of this magazine.

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