Back to Annual Meeting Program

Purpose: Tissue expander (TE) breast reconstruction-related infections that necessitate implant removal represent a devastating complication to both patient and surgeon. Once a superficial soft tissue infection is diagnosed appropriate empiric antimicrobial therapy is initiated to potentially prevent the TE from becoming secondarily infected and salvage the reconstruction. To identify whether appropriate empiric antimicrobial therapy was initiated, we compared antimicrobial susceptibility panels of the causative microorganisms obtained from intraoperative culture data of the explanted infected TE’s to the empiric antimicrobial treatment regimen.
Methods: All patients who underwent post-mastectomy TE reconstruction and developed a postoperative infection requiring TE removal at the MD Anderson Cancer Center between 2004-2010 were included in this study. Microrganisms isolated intraoperatively were stratified, to skin flora, non-skin flora or mixed flora group. The antimicrobial susceptibility panel of these organisms was compared to the provided empiric antimicrobials.
Results: Overall 97 patients underwent TE explantation, 75 had positive intraoperative cultures, 14 had sterile intraoperative cultures and 8 patients did not undergo intraoperative fluid culture. Fifty-one patients (68%) were infected with skin flora, 17 (23%) non-skin flora and 7 (9%) had mixed flora. Patient characteristics and risk factors including smoking status, diabetes, obesity, pre-op chemotherapy and/or radiation therapy, and time to infection were not significantly different among the three bacterial groups. Fifteen patients underwent bioprosthetic mesh reconstructions and were found to have a statistically higher incidence of infection with non skin flora as compared to patients without mesh (46.7% vs. 16.7%, p=0.04).
Empiric therapy consisted of Vancomycin/Zosyn in 29.3% of patients, Vancomycin in 16% and Unasyn in 14.7% of patients. Once final culture results were available antibiotics were modified in 26 patients (34.7%), which did not differ among the three bacterial groups (p=0.17). The incidence of appropriate antibiotic therapy targeted towards the isolated microbes ranged from preoperative prophylaxis 16% of cases, postoperative prophylaxis 4% and empiric therapy 53.3% of cases. When aggregate bacterial sensitivity was evaluated by antibiotic type sensitivity in greater than 70% of cases was seen with Vancomycin, Zosyn, Bactrim, Ciprofloxacin and Rifampin and antimicrobial resistance in greater than 65% of cases was seen with Unasyn and Cephalosporins. (Table 1)
Conclusion: Once a periprosthetic TE infection is diagnosed targeted antimicrobial therapy beyond standard skin flora is warranted including resistant Gram-positive organisms and Gram-negative non-skin flora especially in patients undergoing reconstruction with bioprosthetic mesh. Despite our best efforts, appropriate empiric therapy was achieved in only 53% of patients in this series. Given the overall sensitivities of bacterial pathogens identified targeted empiric antimicrobial therapy towards both skin flora and non-skin flora with dual drug therapy provide the best option for appropriate treatment. Broad-spectrum antimicrobial combinations are available based on oral or intravenous routes and potential patient drug allergies: Vancomycin/Zosyn, Vancomycin/Ciprofloxacin, Bactrim/Rifampin and Ciprofloxacin/Rifampin to broaden empiric therapy and increase the potential to salvage an infected tissue expander breast reconstruction.

Table 1 Overall Aggregate Antimicrobial Sensitivities
Antibiotic Sensitivity	Overall	Skin Flora	Non Skin Flora	Mixed	P Value
Unasyn
R	33 (64.7%)	26 (68.4%)	1 (14.3%)	6 (100%)
S	18 (35.3%)	12 (31.6%)	6 (85.7%)	0 (0%)	<0.01
Cefazolin
R	22 (78.6%)	16 (72.7%)	0 (0%)	6 (100%)
S	6 (21.4%)	6 (27.3%)	0 (0%)	0 (0%)	0.29
Zosyn
S	19 (100%)	1 (100%)	12 (100%)	6 (100%)	---
Clindamycin
R	19 (40.4%)	18 (43.9%)	0 (0%)	1 (16.7%)
S	28 (59.6%)	23 (56.1%)	0 (0%)	5 (83.3%)	0.38
Ciprofloxacin
R	13 (24.1%)	13 (36.1%)	0 (0%)	0 (0%)
S	41 (75.9%)	23 (63.9%)	12 (100%)	6 (100%)	0.01
Vancomycin
S	64 (100%)	46 (100%)	11 (100%)	7 (100%)	---
Rifampin
S	49 (100%)	42 (100%)	0 (0%)	7 (100%)	---
Bactrim
R	16 (28.1%)	12 (27.9%)	1 (14.3%)	3 (42.9%)
S	41 (71.9%)	31 (72.1%)	6 (85.7%)	4 (57.1%)	0.48

1