Infection: studies reviewed

This report is based on the submission of some thirty publications for review and addresses several issues:-

• What is the evidence for bacterial colonisation/infection of mammary prostheses?
• What is the nature of the colonising/infecting bacterium?
• What are the likely consequences (if any) of such colonisation/infection?
• How can colonisation/infection be reduced or prevented?

1. General points
At the outset it should be clearly realised that the implantation of foreign material whether it be a mammary implant, a hip replacement, an intravenous catheter or an prosthetic heart valve constitutes a risk of introduction of endogenous bacteria from the patient's own skin. This means that of the resident bacteria, Staphylococcus epidermidis (S. epidermidis) (or other coagulase-negative staphylococci) or more rarely Staphylococcus aureus (S. aureus) are likely to be involved. Therefore the isolation of such bacteria species from the prosthetic material is not unexpected. In addition, it is also possible that the nature of the response to prosthetic implant per se may be inflammatory. One cannot therefore construe from the simultaneous isolation of staphylococci and the presence of inflammatory cells that there is a real infection taking place.

Care has to be taken in the procurement of appropriate specimens for bacteriological analysis and the interpretation of any laboratory findings. The finding of only a small number of colonies of staphylococci may or may not be meaningful and certainly cannot be taken as infection. Contamination or transient colonisation may only be inferred from such a finding.

S. epidermidis is primarily a skin commensal and only rarely is it deemed pathogenic whereas S. aureus is carried in the anterior nares and skin but has considerable pathogenic potential. The `seeds of infection' will arise only at the time of surgery and therefore the onset of a late infection some months or years after mammary implantation is unlikely to have arisen de novo but to be due to the reactivation of a small number of surviving bacteria.

It is clear from the papers submitted for examination that the risk of colonisation/infection is variable, with some investigators achieving isolation rates of staphylococci approaching 50% whereas others had much lower rates of isolation. Most of the papers are concerned with elective cosmetic augmentation procedures. It is not clear how many studies have been performed using patients in whom mammary reconstruction has taken place. Prior chemotherapy in the case of patients having augmentation post mastectomy as well as the different surgical techniques used, etc. may influence the outcome with respect to the incidence of colonisation/infection with staphylococci. In most of the published studies S. aureus or S. epidermidis have been isolated with a single report (but involving 17 cases) implicating Mycobacterium fortuitum. The relevant papers are summarised in Table 1.

Table 1: Incidence of 'infection' in patients undergoing mammary prostheses implantation

Date	Author	Incidence of infection	Main pathogen
1979	Coutiss et al	21/29	S. aureus
1983	Ransjo et al	18/25	S. epidermidis
1983	Clegg et al	17 patients	Mycobacterium fortuitum
1989	Freeman and Jackson	1 patient	TSST producing S. aureus
1992	Hakelius and Olsen	none/25	-
1992	Virden et al	23/40	S. epidermidis
1993	Brand	<1%	staphylococci
1994	Poblete	5 cases	TSST producing S. aureus
1995	Dobke et al	81/150	S. epidermidis
1996	Ahn et al	65/139	Proionibacterium acnes

It is unclear from many of these papers whether merely colonisation or true infection had taken place. Insufficient details of the baseline on which the microbiological diagnosis is made are provided.

If one were seeking to discriminate clearly between colonisation and infection then a quantitative microbiological analysis would need to be performed taking into account the nature of the host response. It is clear that the incidence of infection is much lower than the incidence of fibrosis and capsular contraction suggesting that no cause and effect relationship exists between the two.

Is there anything to be learned from other clinical consequences of infection involving other prosthetic devices? In joint replacement loosening associated with infection of staphylococci is unwelcome and requires revision arthroplasty after clearance of infection. Infected intravenous catheters may need to be replaced and/or the patient put on vancomycin therapy to cure infection. Similarly, in infected prosthetic heart valves combination antibiotic therapy is needed whilst the risk of infected fragments of vegetation being released should not be underestimated. Valve replacement may be required in some cases.

The situation for mammary prostheses is less clear. Capsular contracture on the evidence presented may be exacerbated by concomitant bacterial infection but is not caused by it. Therefore bacterial eradication per se is unlikely to restore the mammary volume or displacement.

There have been a few experiments reported which have tried to illustrate how the mammary implant may be colonised in viva. However, the model systems chosen are (using mice, guinea pigs and rabbits) quite different from the human situation and the results obtained add a very little to our understanding of the role of microbial colonisation of implants and the development of capsular contraction. Somehow the model system needs to reflect more accurately the surgical procedures used in the human situation.

2. Specific points
One research group in particular (Dobke, San Diego, California) has provided the best evidence of `infection' being associated with capsular contracture (1.2). Their findings albeit from a fairly small number of patients do show that staphylococci and in particular S. epidermidis may cause capsular contracture. When one examines their methodology one discovers that positive cultures were obtained by enrichment culture involving 72 hours incubation in culture broth with shaking before inoculation onto solid culture media to isolate bacterial colonies. Thus, it would be possible for only a very small number of viable bacteria to have been originally present within the explanted material and so this may not truly represent infection but rather colonisation or biofilm formation as discussed earlier. However, their use of scanning electron microscopy would lend credence to the idea that adherent bacteria were present on the silicone mammary implant surface. However, this does not imply viability or ability to cause capsular contraction.

The paper of Ahn et al. (3) is also worthy of mention. Although a high association of `infection' was seen in women who had undergone mammary implantation the association with capsular contracture failed to reach significance.

In this connection a questionnaire-based survey of 73 US plastic surgeons with large mammary implantation practices revealed only 60 early and late implant infections among 54,661 implantations (4). Different types of implant had similar infection-rates. S. aureus was the most common bacterium isolated.

The other main piece of evidence suggesting that subclinical infection may determine incidence of capsular contracture comes from the studies of Burkhardt (Tuscon, Arizona) (5.6). His team used intraluminal antibiotic (cephalothin) at operation with/without povidone iodine irrigation. Both procedures reduced the incidence of Class III and IV capsule contracture possibly (but not proven) to be due to reduced numbers of bacteria present peri- and post-operatively.

Of the animal experiments two are worthy of comment, the first by Shah et al. (7) used miniature silicone implants implanted into the flanks of rabbits followed by variable number of S. epidermidis. Their results appear to show that an inoculum as low as 10 cfu were able to induce capsulation around the implants. Much higher numbers (107 cfu) caused extrusion of the implants. The second study by Kossovsky et al. (8) used implants immersed in approximately 105 cfu S. aureus. Larger more palpable capsules were evidence 14 days later with the contaminated implants although the control implants also induced a tissue response. However, similar findings have been described for a number of other implant such as catheters, collodion bags and even cotton dust. So the presence of a foreign body is sufficient per se to induce a measurable host tissue reaction suggesting that in the human situation capsular contracture and fibrosis may occur even in the absence of any bacteria.

3. Summary
Bearing in mind the large number of mammary implantations performed world-wide even a low incidence of infection would mean in reality a relatively large number of potential therapeutic problems to prevent systemic effects which might even include toxic shock syndrome. Nor is it clear whether infection should not be taken to mean contamination or low level biofilm formation in several of the publications.

Key references cited
1. Virden, C P Dobke, M K Stein, P Parsons, C L and Frank. D H. Subclinical infection of the silicone breast implant surface as a possible cause of capsular contracture, Aesth Plast Surg, 1992; 16; 172; 179.

2. Dobke, M K Svahn, J K Vastine, V L Landon, B N Stein and Paraon C L. Characterisation of microbial present at the surface of silicone mammary implants. Ann Plast Surg, 1995; 34; 563-571.

3. Ahn. C Y Ko C Y Wagar, E A Wong RS and Shaw, W W. Microbial evaluation: 139 implants removed from symptomatic patients. Plast Reconstr Surg. 1996; 98; 1225-12-29.

4. Brand, K G. Infection of mammary prostheses a survey and the question of prevention. Ann Plast Surg. 1993; 30; 289;295.

5. Burkhardt, B R. Fibrous capsular contracture around breast implants; the role of subclinical infection. Infections in Surgery, June 1985; 469-474.

6. Burkhardt, B R and Demas, C P. The effect of siltex texturing and povidone-iodine irrigation on capsular contracture around saline inflatable breast implants. Plast Reconstr Surg 1992: 93 123-128.

7. Shah. Z Lehman J A and Tan J. Does infection play a role in breast capsular contracture. Plast Reconstr Surg 1981: 68: 34-38

8. Kossovsky, N Heggers, J P Parson, R W and Robson, M C. Acceleration of capsule formation around silicone implants by infection in a guinea pig. Plast. Reconstr Surg. 1984;73:91-96.