Breast Implant Infections after Surgical Reconstruction in Patients with Breast Cancer: Assessment of Risk Factors and Pathogens over Extended Post-Operative Observation

Abstract

Background:

Infection is a severe potential complication of breast implant positioning in women with cancer. There still is some degree of uncertainty regarding optimal antibiotic prophylaxis regimens, infecting pathogens, and risk factors associated with infection during long-term followup of these patients.

Methods:

We performed a systematic clinical review to assess infecting microorganisms and risk factors among patients undergoing reconstructive procedures for breast cancer between January 2005 and February 2007. A randomly selected group of infection-free patients treated over the same time span was considered as a control.

Results:

Among 240 women undergoing implant procedures performed and followed up as outpatients, 16 patients with prosthetic infections were observed (infection rate 6.7%). Infection was recorded within six months from surgery in 94% of the cases, with an overall mean time to infection of 95 days. The time interval between surgery and infection did not support a diagnosis of hospital-acquired infection in most cases. Gram-negative microorganisms were identified in seven cases. A higher proportion of patients with implant infection underwent radiotherapy or chemotherapy after surgery for advanced tumors compared with the control patients without infection.

Conclusions:

Extended post-operative surveillance is indicated, at least for the first six months after breast implant placement, particularly for women who need radiotherapy or chemotherapy after implant surgery. Gram-negative bacilli may be involved more often in late infections than otherwise expected. This finding may influence initial empiric antibiotic treatment.

Breast implants in women undergoing surgery for cancer represent a valuable aesthetic and psychological support after mutilating procedures. In these patients, reconstruction may produce several clinical problems, including seroma, capsular contracture, skin necrosis, hematoma, and infection.

Surgical site infection (SSI) is a major complication that may lead to infection of the implant and necessitate additional surgery, prosthesis removal, and longer or repeated hospitalization. The rate of infection after breast reconstruction ranges from 1.5% to 12% and is significantly higher when dermal matrix is used [1 –4]. Some reports based on retrospective analyses of extended time-frames (as long as 20 years) found low infection rates (1.8%–3%), probably as a result of retrospective study bias [5].

Several factors have been linked to a higher risk of infection, including chemotherapy, immediate or bilateral reconstruction [2], radiotherapy, and axillary node dissection [3]. Tissue scarring and skin atrophy following chemotherapy or radiotherapy affect postoperative tissue ischemia and delay wound healing, thus explaining their association with SSI in cancer patients with breast implants [6].

With regard to the infecting microorganisms, there is theoretical and evidence-based agreement on the role played by staphylococci and other cutaneous gram-positive microorganisms [7]. However, there are reports of infections by microorganisms that are not part of the usual skin flora. In fact, small series describe the occurrence of implant infections caused by non-tuberculous mycobacteria, including Mycobacterium fortuitum [8] or M. jacuzzii [9], and gram-negative bacilli such as Serratia marcescens [10].

Recent reports recommend administering antibiotic prophylaxis for breast surgery in cancer patients [11] even for a theoretically clean procedure. However, few specific data or recommendations are available for patients having breast surgery and reconstruction [12].

Following the occurrence of a series of breast implant infections at the National Cancer Research Institute, Genoa, Italy, we decided to review the clinical experience over a two-year period in order to assess the type of infecting microorganisms, to identify possible risk factors associated with infection, and to review the adequacy of institutional antibiotic prophylaxis and of empiric antibiotic treatment protocols.

Patients and Methods

We considered all breast cancer patients (n=240) who underwent mastectomy and immediate or delayed reconstruction at the Breast Cancer Surgery and Plastic Surgery Units of the National Cancer Research Institute in Genoa, Italy, between January 1, 2005, and February 1, 2007. Outpatient observation was extended for six months thereafter. All patients underwent surgery by the same surgical team, and in all cases, the same type of expanders and prostheses were used (McGhan Medical Corporation, Santa Barbara, CA). The expanders consisted of silicone rubber envelopes filled with saline after insertion, whereas the implants were silicone gel within a silicone rubber envelope.

Antibiotic prophylaxis consisted of amoxicillin/clavulanic acid 1.2 g intravenously (IV) before skin incision followed by two doses of 1.2 g IV at 8-h intervals. Clindamycin was used for prophylaxis in patients allergic to beta-lactam antibiotics.

After discharge, all patients were followed up as outpatients once weekly for one month and every two months for one year thereafter. Outpatient visits were more frequent if there had been post-operative complications. Cases and control patients were followed similarly.

In the present study, we considered all infections requiring medical attention, including all patients needing hospitalization for antibiotic therapy or surgery. Those treated as outpatients were not considered. Infection was defined as a complication occurring after breast implant surgery that was characterized by three or more of the following findings: Pain, local swelling, erythema, pus, fever, seroma, wound dehiscence, or perforation of the skin. According to the U.S. Centers for Disease Control and Prevention (CDC)/National Healthcare Safety Network definition, all our patients had infections.

In all cases, the following data were collected: Age, timing of reconstruction (immediate or delayed), cancer stage (TNM), chemotherapy/radiotherapy before and after implant reconstruction, antibiotic prophylaxis, fever, time between surgery and infection, infecting microorganism, and antibiotic sensitivity pattern of the isolated microorganism(s). Cultures of incision site samples, drainage fluids, peripheral blood, periprosthetic seroma, and the prosthesis itself (if removed) were performed. All causative microorganisms were identified using routine microbiologic methods. Disk susceptibility testing was performed and interpreted according to the guidelines published by the Clinical and Laboratory Standards Institute [13]. For breast implants that could not be kept in place, replacement was offered at least three months after successful treatment.

To identify factors associated with implant infection, we randomly selected from the database of patients seen over the same time span 39 age- and procedure-matched control patients (mean age 52.3 years; range 31–76 years) using 1:2 (case:control) randomization criteria.

Statistical analysis was performed using SPSS 16.0 (SPSS Inc., Chicago, IL).The Mann–Whitney U test for unpaired data was used to test the differences in time to infection and the Fisher exact test was used to examine the association between the study covariables and the occurrence of infection.

Results

Sixteen implant-associated infections requiring medical attention were recorded among 240 surgical procedures, a rate of 6.7% (Table 1). All patients (mean age 46.3 years; range 32–70 years) had undergone immediate unilateral reconstruction. The median time from surgery to implant infection was 67 days (range 2–389 days). Only one infection occurred beyond the first six months. Patients were treated with various antibiotic regimens: Ciprofloxacin (n=3), ceftazidime (n=2), piperacillin-tazobactam (n=1), meropenem (n=2), vancomycin (n=2), and teicoplanin (n=2).

Table 1.

Epidemiologic and Microbiologic Characteristics of Infections after Breast Implant Placement in Cancer Patients

Patient	Age at infection (years)	Organism	Days after surgery
1	64	Streptococcus agalactiae	95
2	36	None	2
3	55	MSSA	174
4	37	Pseudomonas aeruginosa	167
5	38	MRSE	3
6	70	Enterobacter cloacae	38
7	41	P. aeruginosa	35
8	45	Serratia marcescens	7
9	57	None	5
10	46	MSSE; Acinetobacter baumannii	130
11	43	None	3
12	56	MRSE	389
13	47	MSSA	148
14	33	P. aeruginosa; MSSA	155
15	49	None	178
16	33	S. marcescens	5

MRSE=Methicillin-resistant Staphylococcus epidermidis; MSSA=methicillin-susceptible Staphylococcus aureus; MSSE: methicillin-susceptible Staphylococcus epidermidis.

Surgical removal of the implants was required in 11 patients (69%). Among these patients, secondary reconstruction was successful in three cases.

Microbiological characteristics and epidemiology

A total of 14 infecting microorganisms were isolated from 12 of the 16 patients (81.2%) (Table 1). Microbiologic sampling was negative in three of six patients (50%) presenting with infection within one month, whereas success in pathogen identification was greater for infections presenting later than one month after surgery. Pathogens were grown from multiple sites, including the implant in one case, drainage fluids in seven patients, and multiple blood cultures in two cases. In five patients, the same pathogen was identified in cultures from more than one site.

Time to infection was investigated by stratifying patients according to infecting microorganism. Implant infections with gram-negative bacteria occurred earlier than infections with gram-positive bacteria (mean 54.2±76.7 [standard deviation] days and 140.7±137.9 days after surgery, respectively) although the difference was not statistically significant (Table 2).

Table 2.

Days between Breast Reconstruction and Infection

	N	Mean±SD	Median	p value ^a
Gram-positive	5	140.7±137.9	122
Gram-negative	5	54.2±76.7	23	0.9
Both	2	142.5±17.7	143
No growth	4	47.0±87.3	4
Total	16	95.9±106.1	67

Mann–Whitney statistic.

SD=standard deviation.

Gram-positive bacteria were susceptible to glycopeptides and linezolid in all cases and rifampicin in six of seven cases. Gram-negative bacilli had a similarly favorable pattern of susceptibility, as none showed resistance to more than one of the surveyed antibiotics, which were selected for potential activity against Pseudomonas aeruginosa (amicacin, meropenem, piperacillin-tazobactam, ceftazidime, cefepime, ciprofloxacin).

Factors associated with implant infection

There was no statistical association between infection and cancer stage, although 75% of the infections occurred among patients with stage II, III, and IV breast cancer, whereas these stages were less represented among the uninfected control group (Table 3).

Table 3.

Association between Covariables and Infection Occurrence

Covariable	No. (%) infected (N=16)	No. (%) not infected (N=39)	p value ^a
Cancer stage
0, I	3 (19)	12 (31)
II–IV	9 (56)	13 (33)
Unknown	4 (25)	14 (36)	0.16
Neoadjuvant chemotherapy
Yes	7 (44)	12 (31)
No	9 (56)	27 (69)	0.89
Neoadjuvant radiotherapy
Yes	0	7 (18)
No	16 (100)	32 (82)	0.076
Antibiotic prophylaxis
Yes	10 (62.5)	30 (77)
No	4 (25)	9 (23)
Unknown	2 (12.5)		0.78
Adjuvant chemotherapy
Yes	5 (31)	5 (13)
No	11 (89)	34 (87)	0.22
Tumor stage II–IV (vs. I)^b
Yes	7 (44)	5 (13)
No	9 (56)	34 (87)	0.018
Radiotherapy after surgery before infection
Yes	6 (37.5)	5 (13)
No	10 (62.5)	34 (87)	0.037

Fisher exact test.

“Advanced tumor stage”=needing adjuvant chemotherapy at any time after surgery, including after treatment of implant infection.

Radiotherapy or chemotherapy before the procedure was not associated with a risk of implant infection. A higher risk of implant infection was observed in those patients who underwent adjuvant radiotherapy after implant positioning, with infection occurring after surgery and after beginning radiotherapy. Adjuvant chemotherapy was performed between surgery and infection in five of seven patients and in a lower proportion of patients without infection (12.8%) (p=not significant). When any need for adjuvant chemotherapy was considered, thus including an additional two patients who received chemotherapy after clearance of infection because of advanced tumor stage, this increased proportion was significant compared with the control cohort (p=0.018).

Discussion

Infection represents a major concern associated with breast reconstruction, as it may necessitate additional hospitalization, surgery, and antibiotic treatment. The rate of breast implant infection in the present work (6.7%) was similar to that observed in previous reports [1 –4]. This confirms its occurrence at a higher incidence than after breast implant surgery in patients without cancer [1]. It is noteworthy that 94% of the infections (15/16) occurred within six months after implant positioning. The mean time to implant infection was 96 days, with variability depending on the causative pathogen. Other investigators report much shorter intervals (e.g., 18–25 days) [14]. This difference is to be ascribed to the short post-operative observation period (two months) in previous studies. Indeed, if had we used such a short observation period, we would have been prevented from identifying the more than one-half (56%) of the infections that occurred more than two months after surgery. Other studies do not mention the timing of prosthetic infection [1,3,5,8,11,12,15]. Thus, the first six months after surgery represent the period of highest risk of infection, during which patients may benefit from closer clinical monitoring. We therefore suggest that post-implantation clinical surveillance protocols for reconstruction in breast cancer patients extend beyond standard SSI monitoring.

Although followup of these patients for 12 months may represent good clinical practice standards followed by several groups, it usually is not found in guidelines [16], nor is it observed routinely. This protocol is not recommended, for example, by the American Society of Plastic Surgeons (www.plasticsurgery.org/reconstructive-procedures/breast-reconstruction.html) or the CDC.

Because we found that one of our 16 patients had infection >12 months after surgery, it is possible that a proportion of patients will have infection >1 year after infection. Although this case definitely is not enough to suggest active patient followup beyond 12 months, this possibility should be considered by physicians/surgeons, discussed with patients, and addressed by future studies.

In the present work, both gram-positive and gram-negative microorganisms were associated with implant infection without any clear pattern in the rate or clustering and a wide range in time to infection for both types of organisms. Several single-case reports describe the isolation of unusual pathogens from infected breast implants in cancer patients [8,10]. A predominance of infections caused by skin-derived gram-positive microorganisms would be expected for breast implant surgery [7], as is the case for cardiothoracic or orthopaedic surgery. However, our finding of frequent gram-negative infections is not unprecedented [10,14].

The source of the gram-negative pathogens isolated in our series could not be determined. Routine epidemiologic analysis of surgical surface contamination did not provide evidence of matching contamination. The presence of three P. aeruginosa, two Serratia spp., one Enterobacter cloacae, and one Acinetobacter spp. may be interpreted as contamination of the surgical site during and after the procedure with pathogens that represent environmental contamination or breaks in prevention procedures. This hypothesis would be confirmed if infection had occurred exclusively early after surgery, but in only three of these cases did the infection occur within the first 40 days after surgery, whereas in the other cases, the lag was more than four months. Although conclusions are limited by the size of the infected cohort, we observed that radiotherapy after surgery is associated with a risk of implant infection. This is in line with other recent reports [3]. Altogether, these observations indicate that radiotherapy may determine minimal residual cutaneous damage with possible barrier breaks and bacterial seeding. Adjuvant chemotherapy cannot be associated with or responsible for an infection occurring after implant positioning but before the start of adjuvant chemotherapy. Indeed, we observed that, even if the difference was not statistically significant, 45% of the patients with infection underwent adjuvant chemotherapy as opposed to only 13% of those without subsequent infection. Because the need for adjuvant chemotherapy commonly is associated with a more advanced tumor, an estimate of any need for adjuvant chemotherapy—either before infection or after clearance of infection—may be made at the time of surgery. For this reason, we conclude that the higher risk of implant infection may be related to more advanced breast tumor stage rather than to bacterial translocation following gastroenteric mucosal damage during chemotherapy [17].

From a clinical point of view, the observation of prosthetic implant infections caused by gram-negative bacteria is relevant for the choice of empiric antibiotic treatment. Optimal antibiotic regimens for the treatment of established infection after breast reconstruction have been outlined recently considering only gram-positive pathogens [12]. When confirmed by local epidemiology, febrile episodes after breast implant positioning and suspected implant infection could be managed with agents active against gram-negative bacteria. Future work on this topic might need to account for the possible occurrence of mixed gram-negative/-positive infections.

In conclusion, women with more advanced or aggressive breast cancers undergoing post-reconstructive chemotherapy or radiotherapy need more intensive and longer followup to prevent or quickly identify late local infection. Continuous local microbiological surveillance in these cases may help to identify the need to target gram-negative pathogens as well as skin flora.

Footnotes

Acknowledgments

The authors thank Doctors Laura Papa, Letizia Di Stefano Valeria Puggioni, Guido Di Vito, Michaela Adami, Carlo Vecchio, Alessandra Catturich, and Maura Bonvento for help with the patients, and Doctors Michele Mussap and Maria Pia Molinari for microbiologic analysis.

Author Disclosure Statement

All authors contributed to the work and approved the manuscript. There are no conflicts of interest for any author, actual or potential, and the present work is not being submitted elsewhere.

References

Pittet

, Montandon

, Pittet

. Infection in breast implants. Lancet Infect Dis, 2005; 5:94–106.

Lanier

, Wang

, Chen

et al. The effect of acellular dermal matrix use on complication rates in tissue expander/implant breast reconstruction. Ann Plast Surg, 2010; 64:674–678.

Armstrong

, Bolding

. Infection following breast reconstruction. Ann Plast Surg, 1991; 27:284–288.

Nahabedian

, Tsangaris

, Momen

, Manson

. Infectious complications following breast reconstruction with expanders and implants. Plast Reconstr Surg, 2003; 112:467–476.

Pinsolle

, Grinfeder

, Mathoulin-Pelissier

, Faucher

. Complications analysis of 266 immediate breast reconstruction. J Plast Reconstr Aesth Surg, 2006; 59:1017–1024.

Gabriel

, Woods

, O'Fallon

et al. Complications leading to surgery after breast implantation. N Engl J Med, 1997; 336:677–682.

Ahn

, Ko

, Wagar

et al. Microbial evaluation: 139 implants removed from symptomatic patients. Plast Reconstr Surg, 1996; 98:1225–1229.

Macadam

, Mehling

, Fanning

et al. Non-tuberculous mycobacterial breast implant infections. Plast Reconstr Surg, 2007; 1419:337–344.

Rahav

, Pillik

, Amitai

et al. An outbreak of Mycobacterium jacuzzii infection following insertion of breast implants. Clin Infect Dis, 2006; 43:823–830.

10.

Pegues

, Shireley

, Riddle

et al. Serratia marcescens surgical wound infection fol1owing breast reconstruction. Am J Med, 1991; 91:173S–178S.

11.

Penel

, Yazdanpanah

, Chauvet

et al. Prevention of surgical site infection after breast cancer surgery by targeted prophylaxis antibiotic in patients at high risk of surgical site infection. J Surg Oncol, 2007; 96:124–129.

12.

Bunn

, Cunningham

, Handscomb

. Prophylactic antibiotics to prevent surgical site infection after breast cancer surgery. Cochrane Systematic Rev J, 2006CD005360.pub2.

13.

Clinical and Laboratory Standards Institute (CLSI). Performance Standards for Antimicrobial Disk Susceptibility Tests. Approved Standard. CLSI document M02-A10 (ISBN 1-56238-688-3) Tenth. Wayne, PA: The Institute, 2009.

14.

Vilar-Compte

, Roldan-Marin

, Robles-Vidal

, Volkow

. Surgical site infection (SSI) rates among patients who underwent mastectomy after the introduction of 551 prevention policies. Infect Control Hosp Epidemiol, 2006; 27:829–834.

15.

Olsen

, Chu-Ongsakul

, Brandt

et al. Hospital-associated costs due to surgical site infection after breast surgery. Arch Surg, 2008; 143:53–60.

16.

Mangram

, Horan

, Pearson

et al. Centers for Disease Control and Prevention (CDC) Hospital Infection Control Practices Advisory Committee. Infect Control Hosp Epidemiol, 1999; 20:250–278.

17.

Tsuji

, Hiki

, Nomura

et al. Simultaneous onset of acute inflammatory response, sepsis-like symptoms and intestinal mucosal injury after cancer chemotherapy. Int J Cancer, 2003; 107:303–308.