Staphylococcal issues: VRSA; mupirocin and surgical site infections

August 2002

We knew it would happen, and it finally happened. The lead report in the July 5th issue of the Morbidity and Mortality Weekly Report related the story of the first recovery of vancomycin-resistant Staphylococcus aureus (VRSA) from a patient in the United States. It was isolated from a catheter site infection in a 40-year old patient in Michigan (the same state in which the first U.S. isolate of vancomycin-intermediate S. aureus [VISA] was recovered) who had diabetes, peripheral vascular disease and chronic renal failure, maintained on outpatient dialysis. Since April 2001, the patient had multiple foot ulcerations and had received multiple courses of antibiotics, including vancomycin. The organism was first recovered when a temporary dialysis catheter was removed but was then subsequently recovered from the chronic foot ulcer as well. Nasal cultures were negative for VRSA. The patient was reported to be clinically stable and responding to local wound care and treatment with sulfa/trimethoprim. The patient was managed in the dialysis center with contact precautions and a dedicated dialysis machine. Thus far, there has been no evidence of spread to health care personnel, to other patients or to the patient’s family.

The culprit in this strain of S. aureus is the presence of the vanA gene, presumably from enterococci. Transfer of this vancomycin resistance determinant from enterococci to S. aureus has been demonstrated in vitro, but this is the first time such transfer has apparently occurred “in the wild.”

What is likely to happen next? No one knows, of course, but I would predict that we will slowly see an increase in recovery of these strains in various places around the country, and indeed, around the world – just as we are seeing with VISA. Gradually, over the course of several years, these strains may appear more commonly and gradually become endemic in our hospitals, and we will likely see evidence of nosocomial spread. If we are fortunate and strive to eradicate these organisms whenever we encounter them, we may be able to slow this process down, but I don’t believe we can really stop it. It may, however, be 10 years or more before we see VRSA as a widespread endemic problem. Fortunately, this particular strain was susceptible to chloramphenicol, linezolid (Zyvox, Pharmacia), minocycline, quinupristin/dalfopristin (Synercid, Aventis), tetracycline and sulfa/trimethoprim. Not every isolate, however, can be expected to allow that many antimicrobial options.

On a different staphylococcal topic: Within the last two months two studies have appeared dealing with intranasal mupirocin ointment (Bactroban, GlaxoSmithKline) and prevention of postoperative staphylococcal infection. The first of these appeared in The New England Journal of Medicine (June 13) and was carried out by Trish Perl and her colleagues when she was at the University of Iowa. Their study population consisted of cardiothoracic, general, oncologic, gynecologic and neurologic surgery patients. The second report was published in the Clinical Infectious Diseases electronic edition (Aug. 15) and was carried out by investigators in Rotterdam, The Netherlands. This study was in orthopedic surgery patients, so there was no overlap in these two reports.

I will not detail the methodology used in these two randomized, double-blinded and placebo-controlled studies, but they appeared, as described, to be appropriate, with careful surveillance and follow-up. In the Iowa study, an infection was defined as occurring within the 30 days following the operation; in the orthopedic surgery study, however, there was telephone follow-up even after 30 days, though the authors did not state how many infections occurred after the 30th postoperative day.

Results were similar in both studies. Both confirmed the efficacy of intranasal mupirocin in eradicating nasal carriage of staphylococci when applied twice daily for three to five days. Both studies confirmed the efficacy of intranasal mupirocin in preventing endogenous staphylococcal infection in the postoperative period, defined in both studies as staphylococcal infection of the same pulsed-field gel electrophoretic pattern as the strain found in the patient’s nares. In neither study, however, was there a significant reduction in the postoperative staphylococcal surgical site infection rate. In both studies, the rate of postoperative staphylococcal surgical site infection was quite low, less than that expected a priori by the investigators.

What next?

What can one conclude from these studies? First, universal use of intranasal mupirocin preoperatively in all nonemergent surgery is not at all justified on the basis of these and previous studies. Second, recognizing that the study by Perl and her associates did indeed find a significant reduction in postoperative nosocomial staphylococcal infections at any site in patients who were nasal carriers in the mupirocin-treatment group, there is room to consider that some patients may benefit by this approach. Additional studies may be needed to identify with more clarity just which patients might benefit the most. One would think most easily of high-risk patients undergoing implant procedures of various kinds or those with a history of staphylococcal infections in the past. At the very least, a preoperative screening nasal swab culture would seem to be appropriate.

The emergence of resistance was not a problem in either study, probably since the treatment courses were kept quite short, five days or less. Nonetheless, resistance emergence is known to occur, and it would be a grave mistake to use mupirocin so widely as to generate a nosocomial population of mupirocin-resistant staphylococci.

The Dutch study could really make no recommendations at all about mupirocin prophylaxis, since there were no significant differences found in their study. They did make the interesting observation that the intensive surgical site infection surveillance and attention to the details of the protocol may itself have had a salutary effect on the surgical site infection rate. This is a phenomenon well known to clinical and epidemiological investigators and is perhaps an analogue of the Hawthorne effect, well known to industrial psychologists. In effect, the very process of measuring the effect of some intervention on a clinical phenomenon, such as the surgical site infection rate, may itself bring about some change in the surgical site infection rate. Some readers may recall the first U.S. prelicensure studies of pneumococcal vaccine, carried out in geriatric residential communities; the pneumococcal pneumonia rate dropped to very low levels, such that it was impossible to demonstrate a beneficial effect of the vaccine!

Finally, recall that not all prevention efforts need involve an antimicrobial agent. An investigational staphylococcal vaccine last year was reported to show considerable promise in preventing staphylococcal infection when evaluated in dialysis patients, a high-risk group. One hopes that it is or will soon be studied in high-risk surgical patients, such as those undergoing implant surgery or cardiac valve insertion.