In the field of Plastic Surgery, the use of breast implants (BI) is a fundamental pillar, both in reconstructive and aesthetic surgery. According to the International Society of Aesthetic Plastic Surgery (ISAPS), every year nearly two million women undergo breast augmentation with implants throughout the world; in Spain the number reaches 50,000 women operated on annually, compared to over 400,000 women a year in the United States (ISAPS, 2019).

In breast surgery with implants, infection of the surgical site as a consequence of bacterial colonization of the BI is one of the most relevant complications. While in cosmetic breast augmentation its incidence is around 0.1–1.5%, it rises to 5.8–28% in breast reconstruction, depending on the study (ISAPS, 2019). Consequences of these infections are a delay in starting adjuvant chemotherapy or radiotherapy treatment and additional surgeries with a significant increase in the risk of reconstructive failure and/or implant loss (Ranganathan et al., 2018; Ngaage et al., 2020).

In the context of BI colonization by pathogenic microorganisms, the presence of bacterial biofilm has been correlated with the development of capsular contracture, which is the most frequent complication in breast implant surgery (Drinane et al., 2017). Such bacterial contamination of the implant shell has also been proposed as one of the factors involved in the pathogenesis of breast implant-associated anaplastic large cell lymphoma (BI-ALCL) (Jones et al., 2018).

In relation to the type of breast implants (shell architecture), multiple studies have shown that textured BI have higher rates of bacterial growth in vitro, compared to those with a lower degree of surface texturing (Jones et al., 2018; James et al., 2019). However, no consistent association has been demonstrated between smooth or textured implants and capsular contracture, but has so with the severity (Poeppl et al., 2007). In general, textured implants are preferable because they have a lower rate of pocket over-dissection, rotation and displacement, and they adhere more to the surrounding tissues. Smooth implants have greater long-term complications such as seroma and capsular contracture.

As for BI colonizing microorganisms, the most frequently isolated in implant culture is Staphylococci (S. epidermidis and S. aureus) and anaerobes, which are associated with treatment-resistant infection and implant loss (Del Pozo et al., 2009; Deva et al., 2013; Rieger et al., 2013; del Pozo and Auba, 2015; Ngaage et al., 2020).

Several therapeutic measures have been proposed to reduce the bacterial load of implants during surgery, such as washing the breast pocket with antiseptic solutions, being the most common a triple antibiotic dilution (Cefazolin 1 g, Bacitracin 50,000 U, Gentamicin 80 mg) followed by povidone iodine (PI) (Chopra et al., 2017; Drinane et al., 2017; Banerjee and Featherstone, 2019; Carvajal et al., 2019). Another alternative is to irrigate or immerse the implant before its placement; but to date there is no agreement on the optimal product to use (Ta et al., 2008; Dang et al., 2020). Despite this lack of consensus, a recent study by Ngaage et al. (2020) has shown that PI is the most effective irrigating formula in reducing contamination of smooth breast implants by methicillin-resistant S. aureus and S. epidermidis, shown in a 30-min in vitro test. However, in surgery, it is not feasible to wait 30 min for the solution to act, as the implants cannot be submerged during surgery for that time. Despite usually implants are pre-selected, the final choice is made at the end of the surgery, when the dissection of the prosthetic pocket is completed; so waiting an additional 30 min would mean such a prolongation of the surgery time, leading to increased costs and morbidity.

So it is necessary to determine whether shorter times could be equally effective, as it has been demonstrated in other fields (Ta et al., 2008).

In this regard, given that in Spain the majority of implants used are textured, we determined using an in vitro model of the efficacy of 10% PI applied for 3 different time lengths, against bacterial biofilm formation of S. aureus and S. epidermidis strains on breast implants with different degrees of texture.

Given the fact that hitherto the exact cause for capsular contraction has not yet been determined sufficiently, the biofilm hypothesis is one of the major corner stones to explain fibroblast proliferation. Chronic inflammation is somehow considered to play a role in this entity and hence numerous authors have tried to circumvent biofilm occurrence utilizing various disinfectants or antibiotic solutions, applied directly to implants, and into the implant pockets (Dang et al., 2020; Jewell et al., 2021; Nai et al., 2021).

In the spring of 2000, the US Food and Drug Administration (FDA) issued a ban on the use of Betadine (Purdue Frederick, Stamford, Connecticut) with BI due to concerns about an adverse effect on shell integrity that could lead to implant deflation or rupture (Wiener, 2013). In 2017 the United States FDA reviewed and approved a request by a BI manufacturer for a change in directions for use that removed warnings regarding the use of Betadine (PI) 10% solution. It has been proposed that Betadine would not have a significantly negative impact on shell integrity but might help influence to reduce capsular contraction (Wiener, 2007).

Nevertheless, meanwhile it has been shown in an in vitro study that in principle Betadine is not harmful to silicone implant surfaces when administered in a clinically usual time frame (Schmitz et al., 2019).

Despite several preventive measures exist to reduce BI infection, being the irrigation of the implant with local antiseptics one of the most widely used (Chopra et al., 2017; Drinane et al., 2017; Banerjee and Featherstone, 2019; Carvajal et al., 2019), there is no consensus on which antiseptic, at what dose, and for how long it should be used to obtain optimal prevention. In a recent study by Ngaage et al. (2020) it was shown that 10% PI had the best results compared to other antiseptics when applied for 30 min on smooth implants colonized by S. aureus and S. epidermidis. However, since that time of exposure was too long to be applied in the real clinical practice and since implants with different degrees of texture are used in Spain, we proceeded to optimize the efficacy of PI at 3 shorter exposure times, as well as studying different textures using both conventional culture and CLSM.

In general, we observed that PI at the 3 exposure times was able to significantly reduce (mostly inhibit) cfu counts after culture of the sonicate from disks implants, regardless of the microorganism tested, compared to the control group treated with SS. All textures, except for smooth implants had sterile cultures (100% reduction) from the disks sonicate (mean percentage reduction of smooth implants ranged between 51.9 and 82.9%). However, these differences could not be demonstrated when analyzing the percentage of living cells. Although mean reduction ranged from 0.1 to 14.4%, it was not enough to reach clinical impact (>25%).

Regarding bacterial density, Mentor® implants treated with PI showed an augmentation in number of bacteria/μm² (except for 10 μm—smooth). However, Polytech® implants (except for macrotextured 50–100 μm) showed a significant mean reduction of bacterial density, with 25–50 μm Polytxt® the one that showed better reduction rates at each PI exposure time (PI1′: 73.3%, PI3′: 74.4%, and PI5′: 62.1%).

A possible explanation for this lack of correlation between reduction of cfu counts and cell viability/bacterial density could be that the cells lost their ability to grow in microbial culture and, therefore were not recoverable, the so-called “viable but not culturable cells.” The impact of this cell state is still under debate, although several authors favor the theory that these cells are in the pre-death phase, so they are still viable but not recoverable. Despite having all the nutrients necessary for their growth, they are not able to grow and sooner or later they will die (Li et al., 2014; Zhao et al., 2017).

Regarding the optimal exposure time of PI to reduce bacterial biofilm, we found no differences between the 3 times tested. Although the recent study by Ngaage et al. (2020) showed that 10% PI was effective when breast implants disks were left immersed for 30 min, this time is neither viable nor feasible to apply in clinical practice, since it is not possible to keep the patient in the operating room waiting 30 min for the solution to act. Therefore, according to our data, given that the variable that showed the greatest reduction rates regardless of time exposure tested was cfu count, we consider that there is no relationship between PI exposure time and its effect, which allows us to use it in clinical practice for < 30 min of exposure. In our opinion, optimization of the efficacy of PI is not so much due to the exposure time, but rather to its penetration capacity on the silicone material of the implants. In the recent study by Barnea et al. (2018), given that the surface of BI is hydrophobic, they pre-treat the implants with plasma to reverse their hydrophobic to hydrophilic condition, and their results are spectacular after subsequent immersion of the implants in the PI solution for only 5 s. Despite they demonstrated that treating BI with plasma to improve implant permeability to PI, our study is the first to assess that PI efficacy against Staphylococcal biofilms is not dependent of the time of exposure but of the implant surface. Therefore, this would support our hypothesis that the effectiveness does not depend so much on the exposure time but on the penetration of the solution, as it can be observed in Figure 4, where depending on the implants’ texture, the penetration ability of PI is different. We suggest that what Barnea et al. (2018) demonstrated is a promising sept for Plastic Surgery that needs to be further assess in clinical studies.

This is also related to the possible differences in the effectiveness of PI according to the texture of the implants. Previous studies have shown that colonization occur more frequently in smooth implants in comparison with macrotextured implants (Ersek, 1991; Wong et al., 2006). We have been able to confirm these results, as the implants with the least reduction of colonization degree were the smooth ones. However, contrary to what might be expected, the density of the biofilm in the Mentor® smooth texture (10 μm—smooth) was significantly lower than in the microtextured (28 μm Siltex™ and 36 μm CPG™ implants). Conversely, in the latter, although the reduction in biofilm density was lower, the reduction in cfu count per culture was greater. This could be explained by the fact that cells adhere more to the surface of smooth implants, which have less biofilm. With this adherence, cells are harder to detach, and treatment with PI would not optimally access and eliminate those cells. The opposite would occur in the microtextured ones, where the biofilm is thicker and the cells are less adhered and detach more easily, and PI can act more effectively (Li et al., 2014; Barnea et al., 2018; Ngaage et al., 2020). This was not observed in Polytech® implants.

Despite demonstrating good results with the application of PI during 1′ at BI, future studies are needed to validate the recent application of plasma to better penetrate silicone, as well as to test PI efficacy in other microorganisms, such as Cutibacterium. acnes, and with other types of implants. Moreover, since there is a complex bacterial ecosystem surrounding BI, further research is needed regarding the role of microbioma of the biofilms in the pathogenesis of BI infection and capsular contracture (Cook et al., 2021; Crowe and Simister, 2021).

One the main limitations of the study is that, in order to mimic the clinical scenario, we did not rinse silicone disks after the adsorption step and were immediately transferred to the bacterial solution, which may have influenced PI amount on surfaces. Moreover, regarding some concerns for tissue toxicity and cellular damage with PI, we used betadine solution (not soap) and the relative cytotoxicity of betadine is considered a beneficial effect around a breast implant, given that it would decrease the proliferation of fibroblasts (related to the periprosthetic capsule), which would imply a lower incidence of capsular contracture. This is one of the beneficial effects of betadine that is intended to be demonstrated in a further subsequent in vivo study in which we will assess not only the colonization of the implants upon explanting, but also the infection rates and the state of the resulting periprosthetic capsule after placing miniature implants in rodents after irrigation with betadine.

Our data need to be validated by investigating different microorganisms and with various other types of implants. It remains also true that the complex bacterial system of biofilms (quorum sensing) is by far not sufficiently understood and needs to be further investigated to better specify the optimal attack point against biofilms and capsular contraction.

We demonstrated that PI was able to inhibit bacterial growth applied on the surface of breast implants regardless of the exposure time. However, no significant reduction on living cells or bacterial density was observed. This lack of correlation may be caused by differences in texture that directly affect PI absorption.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

MG and ÁG-R were responsible for the organization and coordination of the trial. RP-C, JL, PM, and MG were the chief investigator and responsible for the data analysis. BF-I, MD-N, GI, AR, RH, RC, RP-C, and ÃI developed the trial design and data collection. All authors contributed to the writing of the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.