April 2003

A neutropenic 2-year-old male, who was undergoing chemotherapy for leukemia, was admitted to the hospital for evaluation of a febrile illness. Blood cultures were obtained and antibiotic therapy was started immediately with vancomycin and ceftriaxone (Rocephin, Roche). An infectious disease consult was obtained when further examination revealed that he had a sore on the right, lateral aspect of his tongue where he had recently bitten it (figure 1).

Except for the sore on his tongue and a fever of 102.8° F, his examination was unremarkable. His absolute neutrophil count (ANC) was less than 100.

What would you do now?

1. Continue current treatment
2. Add clindamycin
3. Add gentamicin
4. Change ceftriaxone for ceftazidime
5. Both C & D

Answer

Of the choices given, I think the best answer is E. Add gentamicin and replace the ceftriaxone with ceftazidime (an anti-Pseudomonas cephalosporin). Other combinations may be equally acceptable, but this is a potentially lethal situation that many experts would recommend using two anti-pseudomonas drugs with an anti-staphylococcal drug, pending cultures.

The anti-staphylococcal drug in this case was vancomycin, which has become the drug of choice for our oncologists because of the rising incidence of methicillin-resistant Staphylococcus aureus (MRSA) in our facility and community. Where your practice is located, something like nafcillin or methicillin might be more appropriate. There should always be a good reason to use vancomycin empirically. Initial therapy for bacterial meningitis is one reason. Empiric therapy in sick, febrile, neutropenic cancer patients in an environment where MRSA has been common may be another.

While it remains debatable, initial therapy in febrile, neutropenic cancer patients may be anti-staphylococcal penicillin or no specific “anti-staph” therapy at all. Some experts feel that most of the S. aureus threat can be covered by one of the other broad-spectrum antibiotics being used (as long as it is not MRSA).

Morbidity and mortality due to staphylococcus in these patients has historically been low, but has been rising in recent years, probably related to the almost universal use of central lines. When dealing with MRSA, treating known infections with vancomycin is obviously appropriate, but not for simple colonization states. It does not work, and will only add to the likelihood of further vancomycin resistance.

Another option for those who live in areas where community-acquired MRSA is common would be to consider clindamycin for empiric anti-staphylococcal therapy. Most of these strains are susceptible to clindamycin, and, of course, this choice would avoid some of the vancomycin use.

However, if a clindamycin-sensitive MRSA isolate is recovered, and is resistant to erythromycin, it could predict possible inducible resistance to clindamycin during therapy. To be sure, it is recommended that a “D” test be performed (The Pediatric Infectious Disease Journal. 2002; 21:530-534). If it is negative, then clindamycin can be used with confidence. Generally, an infectious disease specialist should be consulted in these situations.

Enterobacteriaceae and Pseudomonas aeruginosa pose a significant threat to these patients, especially when there is some mucous membrane disease, breaking the normally intact barrier. The majority of infectious disease experts recommend initiating therapy effective against pseudomonas with at least one, but usually two anti-pseudomonas drugs. A variety of acceptable combinations exist. Note, however, that some studies suggest that monotherapy with an extended-spectrum cephalosporin (ceftazidime) or a carbapenem (cefepime, Maxipime, Dura) are equally effective as the more traditional double or triple therapy (The Pediatric Infectious Disease Journal. 2001; 20:362-369 and 2002; 21:203-209). However, this is a very complex issue with numerous variables to factor into the decision-making process. These include, but are not limited to the ANC, duration of neutropenia, recent chemotherapy, presence of mucositis, presence of a central line, evidence of thrush, history of transplantation, etc. Even on appropriate, two-drug therapy, these patients are still at risk for pseudomonas sepsis, and should be observed very closely. If any signs or symptoms suggesting early sepsis are seen, the patient should be treated promptly and aggressively with ICU support.

A patient similar to the one presented started the same, with a tongue bite. But in spite of early, aggressive therapy with two anti-pseudomonas antibiotics and early ICU support, the patient succumbed in three days to the pseudomonas sepsis that appeared to start in the tongue with an accidental bite (figure 2 on hospital day two and figure 3 on day three). I cannot emphasize strongly enough the potential lethality of this organism in these immunocompromised patients. You should always take anecdotal data with a block of salt, but most of the bad outcomes I have personally seen in this business have been due to P. aeruginosa. It is one really bad bug that is very unforgiving to these vulnerable patients.

A block of salt, by the way, is something cattlemen (or cattlewomen) put out in the pasture for their cattle to lick for supplementary salt intake. I do not know if they really need it, but they all use it. Most ranchers call it a saltlick. It is basically about one cubic foot of compressed salt and weighs about 50 pounds. The white ones are pure salt. The yellow salt blocks have additional minerals added. The salt block is much bigger than the proverbial grain of salt, therefore, carries with it more doubt or skepticism. When I was a young boy, helping my father put these out for his small herd of cattle, I would lick on it before the cattle got to it. It really is salty. I also used to take a bite of the feed that he put out in the winter. It had molasses in it, and actually tasted pretty good. This may explain some of the odd dietary habits I have now. Nowadays, the average cattle rancher gives a wide variety of nutritional supplements. It’s probably more nutrition than the average child receives.

Contrary to what you might hear, however, none of the ranchers I know give their cattle antibiotics unless the cow has an infection. They do give a variety of immunizations, including a combination shot of bovine rhinotracheitis, viral diarrhea, Parainfluenza 3, and respiratory syncytial virus (RSV). Yes, that’s right, RSV. I am sure this is why large animal hospitals do not fill up with infant calves every winter with RSV bronchiolitis.

To read more about infections in cancer patients, I recommend the excellent review in Long, Pickering, and Prober’s Principles and Practice of Pediatric Infectious Diseases, 2nd Edition, 2003, chapter 103, by Andrew Y. Koh and Philip A. Pizzo. I just got my copy and have already found it to be an excellent resource for pediatric infectious diseases topics. I highly recommend it to anyone who treats children. It’s a bargain at $195.