Post-Mastectomy Surgical Pocket Irrigation With Triple Antibiotic Solution vs Chlorhexidine Gluconate: A Randomized Controlled Trial Assessing Surgical Site Infections in Immediate Tissue Expander Breast Reconstruction

Abstract

Background

Post-mastectomy pocket irrigation solution choice is debated and primarily surgeon dependent. We compare triple antibiotic solution (TAS) with 0.05% chlorhexidine gluconate (CHG).

Objectives

The purpose of this study was to determine surgical site infection (SSI) rates after utilizing TAS vs CHG for breast pocket irrigation in immediate tissue expander (TE) breast reconstruction.

Methods

A prospective, blinded, randomized controlled trial was performed in patients (18-81 years old) who underwent bilateral mastectomy with TE reconstruction. In each patient, 1 mastectomy pocket was randomized to TAS and the other to CHG. Both the TE and the pocket were irrigated in the respective solution. The primary outcome was the incidence of SSI. Secondary outcomes were rates of mastectomy flap necrosis, hematoma, and seroma.

Results

A total of 88 patients who underwent bilateral immediate breast reconstruction were enrolled. Demographic and operative characteristics were equivalent because each patient served as their own control. Between the TAS and CHG groups, the incidence of SSI did not differ (5 [4.5%] vs 7 [8.0%], P = 0.35), including minor infections (2 [2.3%] vs 1 [1.1%], P = 0.56), major infections (2 [2.3%] vs 6 [6.8%], P = 0.15), and those resulting in explantation (2 [2.3%] vs 5 [4.5%], P = 0.25). Necrosis, hematoma, or seroma formation also did not differ. No patients who developed SSI received radiation.

Conclusions

This study does not demonstrate a statistically significant difference in SSI between TAS and CHG irrigation, though TAS approached statistical significance for lower rates of infectious complications.

Level of Evidence: 2

Immediate breast reconstruction with tissue expanders (TEs) has become a common and reliable form of reconstruction for patients with breast cancer because it decreases OR time, decreases donor site morbidity, and allows for a quick recovery.¹ Postoperative wound infection after the placement of TEs, however, can lead to devastating consequences for both patient and surgeon.^2,3 Surgical site infections (SSI) require additional interventions, frequently resulting in extended intravenous antibiotic therapy and wound debridement and occasionally in prosthetic explantation. Ensuring low postoperative infection rates during breast reconstruction with TEs is critical in providing optimal outcomes for our patients. Thus, there is a search for modifiable risk factors to reduce bacterial load, thereby reducing postoperative infections.

One method of infection prevention involves irrigating the post-mastectomy pocket with an antibiotic-containing solution. Current literature demonstrates that intraoperative irrigation of surgical wounds with an antibiotic-containing solution before insertion of breast TEs decreases postoperative infection rates.⁴ This approach is currently the standard of care within the plastic surgery clinical community. One of the most common solutions is a triple antibiotic solution (TAS) composed of cefazolin, gentamicin, and bacitracin.⁵

Chlorhexidine gluconate (CHG) 0.05% was approved by the FDA in 2012 for utilization in irrigating and cleaning various wounds. More recently, CHG has been reintroduced in the setting of surgical wound irrigation prior to closure with biomedical implants.⁶ Here, we investigate the incidence of SSI after irrigation of the post-mastectomy surgical pocket with TAS compared with CHG. This is the first study, to our knowledge, to address TAS vs CHG irrigation of the surgical pocket prior to the insertion of a biomedical implant in a blinded and randomized fashion. The purpose of this study was to determine SSI rates after utilizing TAS vs CHG for breast pocket irrigation in immediate TE breast reconstruction.

METHODS

A single-institution, prospective, blinded, randomized controlled trial was performed in patients who underwent bilateral mastectomy with immediate TE reconstruction from August 2015 to January 2020 with IRB approval from Vanderbilt University Medical Center, Nashville, TN.

Patient Selection

Study inclusion criteria included female patients aged 18 to 81 years who were undergoing bilateral mastectomy and were candidates for immediate breast reconstruction with TEs. Exclusion criteria were female patients undergoing unilateral mastectomy and reconstruction, bilateral reconstruction employing other techniques, allergy to 1 or more components of the antibiotic solution, or allergy to CHG. Patients with a history of previous breast surgery, radiation, or chemotherapy were not excluded from the study. Patients who satisfied the inclusion and exclusion criteria were selected for this study and provided informed consent prior to inclusion. Each patient served as their own control. In each patient, 1 mastectomy pocket was randomized preoperatively to TAS and the other to CHG irrigation. Randomization of laterality of irrigation solution was performed through a computer-generated method in a 1:1 ratio.

Database

A prospective database of patient demographics and reconstructive details obtained via electronic medical record review was created. Demographics are listed in Table 1. Race, BMI, smoking status, alcohol utilization, tamoxifen history, and prior breast surgery were obtained during an initial evaluation with the reconstructive surgeon. Smoking status was classified as yes vs no for current tobacco utilization. Prior breast surgery was defined as prior reconstruction, augmentation, reduction mammoplasty, mastopexy, or breast biopsy. Electronic medical records were thoroughly reviewed to identify oncologic treatment history (chemotherapy, neoadjuvant vs adjuvant radiation, hormonal therapy) along with medical comorbidities (diabetes, hypertension, deep vein thrombosis, bleeding disorders). Operative details included date of surgery, concomitant lymph node procedure (sentinel lymph node biopsy or complete axillary lymph node dissection), breast reconstruction details (acellular dermal matrix [ADM] utilization, pre- vs sub-pectoral TE placement, and number of drains placed), and postoperative complications (SSI, flap necrosis, hematoma, and seroma). SSI were defined per CDC criteria.⁷ Patients were followed-up from the initial breast reconstruction with TE operation to the first postoperative visit after TEs were exchanged for implants. Study data were collected and managed employing Research Electronic Data Capture tools hosted at Vanderbilt University Medical Center (Nashville, TN, USA). Research Electronic Data Capture is a secure, web-based software platform designed to support research studies, providing (1) an intuitive interface for validated data capture, (2) audit trails for tracking data manipulation and export procedures, (3) automated export procedures for seamless data downloads to common statistical packages, and (4) procedures for data integration and interoperability with external sources.^8,9

Operative Technique

The surgical oncologists and plastic surgeons performing the operations varied throughout the study. Cefazolin preoperative intravenous antibiotic was given to each patient at least 30 minutes prior to the procedure. Intraoperatively, skin preparation was performed utilizing 2% chlorhexidine/70% isopropyl alcohol applicators (ChloraPrep, Becton, Dickinson and Company, Franklin Lakes, NJ, USA) and allowed to dry. Immediate breast reconstruction was performed after mastectomies were completed. At the completion of the mastectomies, the surgical field was prepared again with povidone-iodine and re-draped with sterile towels. The skin prep was allowed to dry before reconstruction.

CHG consisted of commercially prepared 0.05% CHG solution (IrriSept, 0.05% CHG in sterile water, Irrimax Corporation, Lawrenceville, GA, USA). TAS contained 1 g of cefazolin, 50,000 U of bacitracin, and 80 mg of gentamicin in 500 mL of normal saline. The solutions were opened into new, unlabeled containers on the surgical back table prior to entrance of the plastic surgeon. The surgeon was blinded to the solutions because they are both clear and undifferentiable. Both the TE and mastectomy pocket were bathed in the respective solutions, and a 1-minute dwell time was employed based on data presented on the IrriSept website reporting similar efficacy with 1- or 5-minute exposure times for most organisms.¹⁰ TE size and type varied per patient dimensions, patient preference, and surgeon discretion. Similarly, ADM utilization, pre- vs sub-pectoral TE placement, intraoperative filling of the TE, and drain placement varied; however, all patients had at least 1 drain placed in each pocket. The incisions were then closed in multiple layers.

At our institution, it is protocol for mastectomy patients to undergo overnight observation. Intravenous cefazolin was continued postoperatively for 24 hours or until discharge on postoperative day 1. All patients were prescribed a 14-day course of postoperative oral sulfamethoxazole/trimethoprim DS 800/160 mg twice per day or 100 mg doxycycline twice per day for those with sulfa allergies.

Postoperative Evaluation

Patients were followed-up for subsequent postoperative visits occurring weekly to bi-monthly. Each drain was removed when its output was ≤30 mL/d. Expander fills were performed at subsequent clinic visits, beginning approximately 3 to 4 weeks postoperatively. Patients who underwent adjuvant chemotherapy continued with expansion throughout the duration of chemotherapy, whereas those who required radiation postoperatively were expanded rapidly followed by radiation.

At all postoperative visits with the reconstructive surgeon, patients were thoroughly examined for presence of SSI (cellulitis, purulent drainage), necrosis, seroma, or hematoma. SSI was defined as any instance where antibiotics were restarted or where erythema, cellulitis, warmth, purulent drainage, or fever occurred. SSI was further classified in 3 categories: minor when requiring oral antibiotics, major when requiring inpatient intravenous antibiotics or operative washout, and finally, explantation. The determination of outpatient vs inpatient antibiotic therapy was primarily clinical—that is, for patients with an initial presentation and minimal to mild symptoms (eg, focal area of cellulitis limited to a small portion of the total incision), these were typically treated with outpatient per oral antibiotics. Persistent or extensive cellulitis as well as systemic symptoms such as fever were treated with inpatient intravenous antibiotics.

Other complications requiring intervention (hematoma or seroma drainage, wound debridement) for each visit were recorded. Review of chart for clinical evaluations was concluded on second-stage reconstruction, need for explantation, or death.

Statistical Analysis

Statistical analysis was performed with IBM SPSS Statistics 23.0 software (IBM Corporation, Armonk, NY, USA). An a priori power analysis based on a prior retrospective study¹¹ comparing TAS and CHG was performed. This analysis indicated that a sample size of 359 breast pockets for each irrigation solution, and thus 359 total patients who acted as their own control, would be required to reject the null hypothesis and show a statistically significant difference in SSI between the 2 solutions, if there was a difference at all. Chi-square tests were employed to detect significantly different outcomes between the 2 groups. A probability of type I error of 5% (P < 0.05) was utilized to determine statistical significance.

RESULTS

A total of 88 patients who underwent bilateral mastectomy were enrolled. All 88 patients received TAS irrigation on one side and CHG irrigation on the contralateral side. Thus, demographic characteristics and operative details were equivalent in the TAS and CHG cohorts because each patient served as their own control. Demographics and operative details are outlined in Tables 1 and 2.

The mean age of the patients enrolled was 47 years, and 81 (92%) patients were Caucasian and 6 (6.8%) were African American. Four (4.5%) patients reported current tobacco utilization, 50 (56.8%) reported alcohol utilization, 5 (5.7%) were diabetic, 15 (17%) had hypertension, 2 (2.3%) had a history of deep vein thrombosis, 2 (2.3%) had history of a bleeding disorder, and 26 (29.5%) had a history of hormonal therapy with Tamoxifen. A total 76 (86.4%) patients underwent bilateral mastectomy for cancer; 71 (80.7%) had unilateral cancer, and 5 (5.7%) had bilateral breast cancer. The remaining 12 (13.7%) patients underwent bilateral mastectomy for risk reduction. A total of 12 (13.6%) patients received radiation—2 (2.3%) in the neoadjuvant period and 10 (11.4%) in the adjuvant period—and 35 (39.8%) received chemotherapy in their treatment regimen.

Thirty-four (38.6%) patients had a lymph node procedure on the side of CHG irrigation and 32 (36.3%) patients had a lymph node procedure on the side of TAS irrigation. Fifteen patients did not have a lymph node procedure, and 7 patients had bilateral lymph node procedures. Two of 34 patients (5.8%) who had a lymph node procedure on the side with CHG irrigation had an infection or explant compared with 1 of 32 (3.1%) patients who had a lymph node procedure on the side with TAS irrigation (P = 0.59). None of the patients who underwent axillary lymph node dissection had an infectious complication.

A total of 5 attending surgeons performed the operations, including the irrigations as described. The senior author (K.K.H.) performed 65 (73.9%) of the operations, with the remaining 4 surgeons performing 11 (12.5%), 6 (6.8%), 5 (5.7%), and 1 (1.1%) each. ADM was utilized in 78 (88.6%) patients bilaterally. Twelve (13.6%) patients had 1 drain placed in each breast pocket, and the remaining 76 (86.4%) had 2 drains in each pocket. Of the 76 patients who underwent mastectomy and reconstruction for breast cancer, 5 had bilateral breast cancer for a total of 81 breasts with cancer pathology; 41 (50.6%) of these were randomized to CHG irrigation, and the remaining 40 (49.4%) were randomized to TAS.

The mean follow-up time at either second-stage reconstruction, explant, or death was found to be 207.8 days (range, 28-606 days). Eighty patients underwent second-stage reconstruction with a mean follow-up time to reconstruction of 213.1 days, 6 patients had either unilateral or bilateral explantation with a mean follow-up time to explant of 135.2 days, and 2 patients deceased with a mean follow-up time to death of 446.5 days. There was no statistically significant difference in the incidence of SSI, including major infections, minor infections, and those resulting in explantation, between the TAS and CHG groups. One patient underwent bilateral TE explantation secondary to bilateral non-infectious persistent rash not attributed to either solution and was therefore not counted as an SSI. In the TAS group, there were 4 (4.5%) SSI vs 7 (8.0%) in the CHG group (P = 0.35). Minor SSI were identified in 2 (2.3%) patients in the TAS group vs 1 (1.1%) in the CHG group (P = 0.56), major SSI in 2 (2.3%) patients in the TAS group vs 6 (6.8%) in the CHG group (P = 0.15), and explantation was required in 2 (2.3 %) patients in the TAS group vs 5 (4.5%) in the CHG group (P = 0.25). There was also no difference in the development of necrosis (20 [22.7%] in TAS vs 16 [18.2%] in CHG, P = 0.45), hematoma (1 [1.1%] in TAS vs 2 [2.3%] in CHG, P = 0.56), or seroma (2 [2.3%] in TAS vs 3 [3.4%] in CHG, P = 0.65) (Table 3). Table 4 depicts bacteria cultured from breast pockets with major SSI. No patients who suffered any of the above infectious complications had a history of radiation. No patients had any allergic reaction to either solution.

Discussion

Implant-based reconstruction continues to be the most common mode of breast reconstruction, representing 69,921 operations and approximately 69% of all breast reconstructive procedures in the Unites States in 2018.¹² TE-related infection rates have been reported to range from 2.5% to 24%.¹³ Infections in implant-based reconstruction are due to varying reasons, such as exposure to bacteria from the mastectomy, inclusion of a foreign body, disruption of soft-tissue vascularity, lymph node dissection, as well as any neoadjuvant or adjuvant radiation treatments.^14,15 Infections can lead to substantial consequences such as the need for antibiotics, hospital readmission, and explantation. Most importantly, infections can have important psychological consequences and delay cancer treatment for patients. Therefore, it is paramount to employ strategies to help minimize the risk of infection and prevent these devastating consequences.

Subclinical bacterial contamination has been implicated in clinical problems such as breast implant–associated anaplastic large cell lymphoma and capsular contracture, and eradication of these causative bacteria at the time of surgery may limit later presenting issues.^16,17 One method of decreasing bacterial burden is breast pocket irrigation.^18-20 The choice of solution for irrigating pockets has been widely debated and is dependent on surgeon preference or tradition. A national survey performed by Gowda et al demonstrated that the majority of surgeons performing implant-based breast reconstruction will soak the implant in TAS and irrigate the mastectomy pocket with povidone-iodine (Betadine, Purdue Products LP, Stamford, CT, USA) solution prior to implantation.²¹ Notably, in the year 2000, the FDA issued a ban on Betadine for concern of delamination of implant shell integrity and potential for implant rupture. Although multiple studies have suggested this to be an unlikely event, it has been recommended to immediately washout the Betadine solution with saline.^22-26 However, this technique would seemingly decrease the antimicrobial effectiveness of Betadine. TAS has been utilized in multiple studies because it is well tolerated and relatively inexpensive. In addition, its minimum inhibitory concentration is exceedingly low for the bacteria most commonly responsible for breast implant infections, such as S. aureus and S. epidermidis.²⁷ Adams et al found low expander infection rates employing TAS for expander immersion and pocket irrigation.⁴

Cefazolin and gentamicin can be utilized in combination to create a synergistic effect, where cefazolin destroys the cell wall during bacterial replication, allowing gentamicin intracellular access to inhibit bacterial protein synthesis at the level of RNA translation. Bacitracin can be employed to directly inhibit the cell wall, but Gram-negative bacteria have been shown to be universally resistant to bacitracin. Although antibiotic resistance remains an area of concern, it is also prudent to note that the pharmacokinetics and pharmacodynamics of the irrigation process have not been widely studied. In addition, antimicrobials require both sufficient contact time to penetrate the microbial cell wall along with a persistent drug concentration. Both of these factors may be limited during post-mastectomy pocket irrigation and are unlikely to be satisfied during quick evacuation of the fluid from the pocket.^5,28,29

CHG has been widely employed as a topical antiseptic in various clinical settings for the past 50 years, including handwashing and oral products. CHG is a chlorinated-cationic biguanide that has broad-spectrum activity against a variety of pathogens, including multi–drug-resistant Gram-positive and Gram-negative organisms, certain fungi, and lipid-enveloped viruses such as HIV. CHG is able to disrupt the cell wall very rapidly, with maximal effect occurring within 20 seconds.^6,30 Furthermore, 0.05% CHG has been effective in reducing the risk of methicillin-resistant S. aureus biofilm-mediated mesh infections utilizing in vivo animal models.^28,29 A retrospective study comparing the complication profiles of TAS vs 0.05% CHG solution in implant-based breast reconstruction noted significantly lower incidence of total complications, including infection, in the CHG cohort.¹¹ To our knowledge, our study is the first randomized controlled trial to compare the 2 solutions.

In January 2020, the FDA requested the withdrawal of bacitracin, a component of TAS, from the market. Bacitracin was only approved for utilization in infants with pneumonia or empyema, but due to concerns of nephrotoxicity, anaphylaxis, and the need for repeat injections, it has been withdrawn.³¹ Unfortunately, the withdrawal of bacitracin resulted in the need for our study to be terminated because it is a component of TAS and therefore became not readily available. Although the study necessitated premature termination, there are a few observations that are important to be pointed out. Overall SSI, major SSI, and most importantly explantations were less frequent in the TAS group compared with CHG. Although these findings were not statistically significant, it was particularly notable the only 2 patients that required explantations on the TAS side also required a contralateral (CHG) explantation secondary to infection. This poses the question whether a factor beyond just the solution influenced the infectious complication for these 2 patients.

The authors believe the results of this study warrant further investigation into bacitracin-containing TAS for TE breast reconstruction. The reasons for discontinuation of bacitracin include anaphylaxis and nephrotoxicity. Anaphylaxis requires prior exposure to bacitracin, and nephrotoxicity is reported with intramuscular injections and not wound irrigation. Therefore, these potential events can be readily mitigated. If TAS can achieve lower rates of major SSI and explantation, its utilization should be reconsidered with careful consideration of these benefits compared against the risks for adverse events.

Given the current unavailability of bacitracin for wound irrigation, the surgeon performing immediate TE breast reconstruction must make an informed decision on choice of breast pocket irrigation solution for their patients. Unfortunately, there is a paucity of randomized controlled trials comparing irrigation solutions. Based on this study and the current available literature, the authors recommend the utilization of 10% betadine along with double antibiotic solution consisting of cefazolin and gentamicin.^26,32 Another study by Adams et al notes that, in their in vitro experiments, none of the bacteria most commonly found in the environment of breast prostheses were cultured after utilization of this combination of irrigation solution.²⁶ Investigation of this hypothesis via randomized controlled trials will be necessary to make a more definitive conclusion about choice of irrigation in this patient population.

This study has some notable limitations. The sample size at the time of bacitracin withdrawal was inadequate to demonstrate a statistically significant difference in SSI between the TAS and CHG groups. Despite this study having a smaller sample size than our power analysis indicated would be required, the lower rate of major SSI with TAS still approaches statistical significance. The authors in no way propose that non-significant results can provide definitive conclusions. But when a surgeon is faced with a choice for irrigation solution and high-quality data are limited, this near-significant result—especially given that the sample size is much lower than the power analysis calculation—in combination with the fact that no explantations were required for the TAS side alone must be taken into consideration. Another potential limitation is that the definition of infection, especially minor SSI, is subjective and based on the clinician’s assessment; blinding of the surgeon along with randomization of laterality of the irrigation solutions helps to minimize this subjectivity from altering the results of the study. Additionally, this study aimed to minimize confounders by making contralateral post-mastectomy surgical pockets within the same patient the control variable. However, certain characteristics between the bilateral pockets may be different and thus might have effects not accounted for in this study. Finally, our group prescribes 14 days of oral antibiotics postoperatively, which is unlikely to be a universal practice. The literature has been evolving for practice of oral antibiotics in implant-based breast reconstruction over the past decade. There are no level 1 data to prescribe practice patterns for breast reconstruction. Our practice of 14 days of oral antibiotics is protocolized for our surgeons and is based on literature to suggest postoperative antibiotics from 5 to 14 days is beneficial. The evidence has been evolving for almost a decade since Clayton et al described the consequences of a single dose of perioperative antibiotics alone.³³ The optimal duration for postoperative antibiotics remains a source of study and discussion. The manuscript from Avashia et al also suggests postoperative antibiosis is related to a reduction in infectious complications.³⁴ A recent paper shows that postoperative antibiotics were beneficial, because the antibiotic cohort had significantly lower rates of TE loss (4.3% vs 17.0%, P = 0.003), unplanned operation (10.4% vs 24.5%, P = 0.007), and infection (7.0% vs 24.5, P < 0.001).³⁵ There is no defined standard of practice at this point. We follow this practice and have shown a significant reduction in our own infectious complication rates based on our local microbiome.

Conclusions

This study does not demonstrate a statistically significant difference in SSI between TAS and CHG irrigation, though TAS approached significance for lower rates of infectious complications. Continued enrollment would be necessary to determine if TAS leads to lower rates of major infection. Although no definitive conclusion can be made for CHG, the authors believe TAS irrigation for immediate breast reconstruction with TEs is a more optimal irrigation solution. Due to the unavailability of bacitracin, betadine in combination with double antibiotic solution should be investigated as a potential alternative.

Presented at: 63rd Annual Scientific Meeting of the Southeastern Society of Plastic and Reconstructive Surgeons (SESPRS) in Kiawah Island, SC in June 2020.

Disclosures

The authors declared no potential conflicts of interest with respect to the research, authorship, and publication of this article.

Funding

The authors received no financial support for the research, authorship, and publication of this article.

References