Abstract
To the Editor:
In our study, AGNB infections represented 43% of recorded infections. It should be noted that five of seven infections (71%) occurred beyond the timing (i.e., >30 d) for surgical site infection (SSI), and three of seven exceeded the timing that defines early prosthetic infection in other surgical procedures (i.e., >2 mo). To further investigate these findings, following the renovation of the study unit at our hospital, we started another still-ongoing study. In an interim analysis of 450 patients, we observed 35 infections, with only two infections (5%) due to AGNB. We believe these infections are associated with either a point source of environmental contamination or endogeneous translocation.
With regard to AGNB infections and selective digestive decolonization, overgrowth and translocation of bacteria across the digestive tract or across the skin during chemotherapy or radiotherapy may be a source of implant contamination. Peri-operative selective digestive decolonization combined with standard intravenous antibiotics reduces the rate of post-operative infectious complications and anastomotic leakage in elective gastrointestinal surgery [6]. Selective digestive decolonization combined with systemic antibiotic prophylaxis also may be useful in preventing blood stream infections and mortality in ICU patients [6–8]. No data are available on selective digestive decolonization in breast implant surgery, and future sufficiently-powered randomized trials would be welcome in settings where AGNB infections are prevalent.
Selective decontamination, together with standard parenteral antibiotic prophylaxis is effective in preventing SSIs involving Staphylococcus aureus. Studies provide strong evidence for staphylococcal decolonization of cardiothoracic surgery patients and weaker but still relevant evidence for decolonization of orthopedic surgery patients [3–5]. Currently, there is no evidence to justify screening or universal mupirocin/chlorhexidine decontamination to prevent S. aureus infection of breast implants in cancer patients. Randomized trials are needed to support this suggestion.
The major findings of our study were not only the relatively high incidence of infections due to AGNB, as pointed out correctly by Silvestri et al., but also the occurrence of infections well beyond the timing that defines SSIs [2]. Indeed, six prosthetic infections were observed within 30 d of surgery and could be included in the SSI definitions, but 10 occurred beyond 30 d. In the case of breast implant infections occurring more than 1 mo after surgery—thus inconsistent with SSI definition—the use of systemic antibiotic prophylaxis for SSI is pointless, and selective digestive decolonization combined with systemic antibiotic prophylaxis is unproved and not likely to be effective. Selective digestive decolonization has proved to be effective in the prevention of SSI and blood stream infection in critically ill patients and in patients who undergo some elective surgical procedures, but not in the specific prevention of early prosthetic infection (i.e., 1–2 mo after surgery). We believe there is still considerable confusion about definitions in this area, which should be addressed thoroughly in future work. Improved attention needs to be paid by the clinical and scientific community in categorizing breast implant infections not only as SSIs (<30 d) but also as early (<2 mo) and late (>2 mo) prosthetic infection. For example, studies reviewing factors associated with SSI infection in breast implants do not address late breast implant infection more than 2 mo after surgery nor early prosthetic infection occurring between 1–2 mo after surgery [9].
In patients who undergo other implant operations, such as orthopedic or cardiothoracic surgery, extensive efforts are undertaken to prevent late bacterial translocation during procedures that may cause bacteremia, including endoscopy and oropharyngeal and odontostomatologic procedures. This may not be the case among breast cancer patients who undergo breast reconstruction. We suggest that proper antibiotic prophylaxis for odontostomatologic, orthopedic, urologic, gastrointestinal, or gynecologic procedures might decrease further the high rate of late infections observed after breast implant surgery.
In conclusion, improved antibiotic prophylaxis and possibly oropharyngeal decontamination or selective digestive decolonization (depending on local epidemiology) against both gram-positive and gram-negative bacteria might help to prevent a fraction (as much as 40% to 60%) of breast implant infections. In addition, active surveillance in the follow-up period (i.e., 1 y after surgery and possibly beyond) and advising antibiotic prophylaxis for procedures favoring bacterial translocation similar to what is suggested for other prosthetic devices might help address and curb the large fraction of breast implant infections that occur beyond the SSI-defined period.