JAC Advance Access originally published online on October 24, 2007
Journal of Antimicrobial Chemotherapy 2007 60(6):1298-1301; doi:10.1093/jac/dkm408
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Antibiotics for the eradication of Propionibacterium acnes biofilms in surgical infection
1 Biomaterials-Related Infection Group, School of Medical and Surgical Sciences, University of Nottingham, Nottingham, UK 2 Centre for Spinal Studies and Surgery, Queen's Medical Centre, University Hospitals NHS Trust, Nottingham, UK
* Correspondence address. Biomaterials-Related Infection Group, School of Medical and Surgical Sciences, Division of Orthopaedic and Accident Surgery, C Floor West Block, University Hospitals QMC, Nottingham NG7 2UH, UK. Tel: +44-115-8231115; Fax: +44-115-8231118; E-mail: roger.bayston{at}nottingham.ac.uk
Received 20 June 2007; returned 27 July 2007; revised 1 October 2007; accepted 2 October 2007
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
Objectives: Propionibacterium acnes is increasingly recognized as a cause of delayed infection after spinal instrumentation or shunting for hydrocephalus. Biofilm development by this organism has recently been demonstrated. We therefore investigated the effect of two different courses of three antibiotics (penicillin, rifampicin and linezolid) on mature P. acnes biofilms in vitro. Outcomes were eradication or regrowth after withdrawal of antibiotics, simulating successful treatment and relapse.
Methods: P. acnes biofilms were grown on titanium discs for 6 days until mature, then exposed to the antibiotics for either 7 or 14 days before sonication and culture. Further, discs were similarly exposed, but after each course, they were reincubated for a further 9 days to check for regrowth.
Results: Penicillin, linezolid and linezolid plus rifampicin eradicated P. acnes biofilms after 14 days, but only penicillin had this effect after 7 days. ‘Relapse’ was prevented only by 14 day courses of penicillin or linezolid plus rifampicin, but not by linezolid alone.
Conclusions: For P. acnes spinal instrumentation infections, either penicillin or linezolid plus rifampicin might be equally effective. For shunt infections, as penicillin does not give therapeutic cerebrospinal fluid concentrations, rifampicin plus linezolid might be the treatment of choice. Linezolid alone appears not to be as effective as penicillin against P. acnes biofilms.
Keywords: orthopaedic implants , spinal instrumentation , hydrocephalus shunts , P. acnes
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
Propionibacterium acnes, a major component of the skin commensal flora, is implicated in an increasing number of cases of biomaterials-related infection (BRI) involving arthroplasty, cerebrospinal fluid (CSF) shunts, spinal instrumentation and others.1–5 This is partly due to the greater use of implantable devices and also to wider recognition of its involvement in infections associated with them. Such infections are notable for their delayed presentation after surgery, and Richards and Emara6 reported P. acnes from 52% of their delayed spinal infection cases with an average time from operation to presentation of 25 months (range 11–79). CSF shunt infections due to P. acnes are also usually delayed, typically presenting several years after shunt insertion.5,7,8 In those occurring in ventriculoatrial shunts, the delay can give rise to immune complex nephritis.8–10 There is now clear evidence that biofilm development by P. acnes is a key factor in both orthopaedic and neurosurgical infections.7,11 In order to eradicate the infections, device removal is recommended. P. acnes is usually susceptible to a wide range of common antibiotics, although some, particularly in acne patients, have been found to be resistant to clindamycin, erythromycin and tetracyclines.12 Despite their susceptibility, they are sometimes remarkably difficult to eradicate, and long courses are often recommended. In addition, where the implant is used as a structural stress-bearing component as in fracture repair or spinal instrumentation, it may be highly desirable to retain the implant until fusion is complete and to administer suppressive therapy until such time that the implant can be safely removed. In view of the biofilm formation and difficulty in eradicating the infections, we decided to investigate the effect of three antibiotics on mature P. acnes biofilms in vitro and to determine which gave the most promising results that might guide clinical practice. Benzylpenicillin is currently the drug of choice for P. acnes BRI, but we considered that linezolid, to which P. acnes strains have been shown to be susceptible,13 might be a more effective alternative, especially for neurosurgical infections in view of its effective CSF penetration. Rifampicin has been reported to have good anti-biofilm activity,14 but should not be used alone; therefore, it was tested in combination with linezolid to determine whether faster eradication could be achieved by this combination than with either penicillin or linezolid alone. Also, as a common feature of biofilm infections is that they relapse after the course of treatment has finished, we therefore determined which of the drugs/combinations gave the lowest ‘relapse’ rate.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
A clinical isolate of P. acnes (NB692) from a hip arthroplasty infection was chosen as the test strain. This strain had previously been shown to be representative of the isolates in our collection. It was resurrected from frozen by culturing anaerobically on sheep blood agar (Oxoid, Basingstoke, UK) and was characterized by conventional methods including Gram stain and API Rapid ID 32A (bioMérieux, Marcy-l'Étoile, France). Antibiotic susceptibility testing by agar incorporation on Brucella blood agar (Oxoid)15 showed full susceptibility to all agents tested including penicillin (MIC 0.03 mg/L), rifampicin (MIC 0.0075 mg/L) and linezolid (MIC 0.94 mg/L).
Penicillin G (Crystapen, Britannia Pharmaceuticals, Redhill, UK), linezolid (Zyvox, Pharmacia, Milton Keynes, UK) and rifampicin (Rifadin, Aventis Pharma, West Malling, UK) were used to make stock solutions that were then stored at –20°C. Final concentrations used were 12, 20 and 8 mg/L, respectively.
Titanium discs (6 mm diameter) were cut from a 0.25 mm sheet (Goodfellow, Cambridge, UK) and sterilized by autoclaving. They were then placed in thawed fresh frozen human plasma (National Blood Service, Sheffield, UK) on a rocker at 37°C for 30 min for a plasma conditioning film to form.
Bacterial adherence and biofilm development
Each disc was then rinsed in sterile PBS and placed in a 2 mL microtube containing 1 mL of bacterial suspension made from an overnight culture of NB692 in anaerobe basal broth (ABB, Oxoid), diluted in ABB to give 
107 cfu/mL. The microtubes were then incubated at 37°C for 1 h. Sixty tubes for each drug and 60 for control were prepared. Each disc was then aseptically rinsed in sterile 2% ABB to remove surplus bacteria and incubated for 6 days in 1.5 mL of full-strength ABB on a roller drum (LEEC, Cardiff, UK) at 37°C. P. acnes had previously been shown by experiment to develop mature biofilms under these conditions.7
After 6 days of incubation on the roller, each disc was aseptically transferred to a microtube containing 1.5 mL of fresh ABB, and the antibiotics were added to give the final concentrations required. The control discs were transferred to tubes containing ABB without antibiotic. All tubes were incubated at 37°C on the roller drum. Each day, the ABB was aspirated and replaced with fresh ABB/antibiotic solution. One set of 4 x 30 discs was ‘treated’ for 7 days and a second set for 14 days. After each ‘treatment’ period, the discs were transferred to 1 mL of 2% ABB and sonicated to detach any remaining biofilm. A 100 µL aliquot of each sonicate was spread onto sheep blood agar and incubated at 37°C anaerobically for 6 days. Any growth was recorded quantitatively by colony counting. The detection limit of the assay was determined to be 10 cfu/mL by experiment. The MIC for any bacteria surviving the exposure to antibiotics was determined anaerobically on Brucella blood agar (Oxoid) by agar incorporation.15
Two further sets of 30 discs for each antibiotic/combination and control, as in the ‘treatment’ phase, were prepared and exposed to antibiotics as above, for either 7 or 14 days. After this, they were transferred to 1.5 mL of ABB in microtubes and incubated in the absence of antibiotic on the roller drum for a further 9 days to check for regrowth (‘relapse’). They were then sonicated and cultured as described earlier.
Using SPSS V14 software, paired sample t-test was used to determine statistical significance.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
The results are shown in Table 1 and Figure 1. Culture of the samples following 7 days of ‘treatment’ with penicillin G showed that none of the samples grew P. acnes, whereas 15 and 20 out of 30 samples grew P. acnes following 7 days of ‘treatment’ with the combination and linezolid alone, respectively, and the average counts were 970 and 2010 cfu/mL, respectively. None of the samples showed growth following 14 days of ‘treatment’ for each of the three regimens. However, in the ‘relapse’ series, 14 days of exposure to penicillin alone and to the combination of rifampicin and linezolid was necessary to prevent regrowth, whereas linezolid alone did not prevent regrowth even after 14 days of exposure.
|
|
No change in MIC of any of the three antibiotics was seen after exposure as biofilms.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
In a laboratory study of 24 isolates of P. acnes from non-implant-associated intracranial infections, Mory et al.13 showed that all were susceptible to penicillin, rifampicin and linezolid on conventional testing, but did not test them as biofilms. Our results show that penicillin is at least as effective as linezolid plus rifampicin and much more effective than linezolid alone against P. acnes biofilms. However, there are some limitations to such a study. The large standard deviations betray the considerable between-sample variation in controls and tests. This is inherent in studies of biofilms attached to biomaterials, mainly due to the ease with which parts of the biofilm detach during manipulation and rinsing. Despite this, the results are clear. In addition, the exposure to antibiotics was constant rather than pulsed, as would be found during clinical treatment. The concentrations used were based on the serum Cmax.16,17 The results can therefore serve as a guide to choice of antibiotic for P. acnes biofilm infections, and they also suggest that a course of at least 2 weeks might be necessary to avoid relapse, although in practice, this might need to be lengthened further. However, from our results, it appears that penicillin alone is superior to linezolid and is equal to a combination of linezolid and rifampicin, leading to eradication without regrowth. The diagnosis of P. acnes BRI is usually delayed, often by at least several months, by which time the infection has become chronic. From what is known of BRI in other settings, it is highly likely that mature biofilms are well established by then. The published antimicrobial regimens for P. acnes BRI vary considerably. Hahn et al.1 writing of infection in spinal instrumentation recommended ‘non-specific wide-ranging parenteral therapy’ for a few days, followed by culture-directed oral therapy for up to 9 weeks. They used ciprofloxacin alone or combined with rifampicin or teicoplanin, or amoxicillin. Richards et al.6 used intravenous cephalosporin (usually cefazolin) for a mean of 6.5 days (2–14) followed by oral cephalosporin for a mean of 20 days (9–42). Clark and Shufflebarger18 recommended much shorter courses, but all recommendations of choice of antibiotic and length of course were linked to the need to remove all the spinal instrumentation. However, management of spinal instrumentation infections differs depending on whether they present early or late.19 In late infections, presenting more than
18 months after surgery, where bony fusion has occurred, the instrumentation can be safely removed in most cases (although some loss of correction might occur). In early infections where fusion is incomplete, infection is managed by surgical debridement, with or without irrigation, and antibiotics. In such cases, the antibiotic therapy is often termed ‘suppressive’ as the hope is that the patient will remain asymptomatic for long enough for fusion to progress so that, if necessary, instrumentation can be removed. However, our results suggest that, where P. acnes is the cause, a shorter course might actually eradicate the infection. Either penicillin or linezolid plus rifampicin could be used, and the former is obviously preferable on grounds of cost and potential toxicity. In the case of CSF shunts, penicillin does not penetrate into the CSF in therapeutic quantities in low-grade infections, and the combination of linezolid plus rifampicin, both of which give therapeutic CSF concentrations in conditions of low inflammation, would be preferable. It is also interesting to note that the risk of thrombocytopenia from linezolid can be reduced by the addition of rifampicin.20 To the best of our knowledge, no attempt has yet been made to treat P. acnes shunt infections without shunt removal, although Thompson and Albright5 suggested that, in such cases, antibiotic courses might be considerably curtailed. If this were attempted, then our results suggest that a combination of rifampicin and linezolid (but not linezolid alone) could be used, with the option of oral administration. |
Funding |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
The Wade Charitable Trust and the Association for Spina Bifida and Hydrocephalus provided financial support. Funding for W. A. was in the form of a grant from industry, but not in support of research in this area.
|
Transparency declarations |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
None to declare.
|
Acknowledgements |
|---|
We are grateful to the Wade Charitable Trust and the Association for Spina Bifida and Hydrocephalus for financial support.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
1 Hahn F, Zbinden R, Min K. Late implant infections caused by Propionibacterium acnes in scoliosis surgery. Eur Spine J (2005) 14:783–8.[CrossRef][Web of Science][Medline]
2 Tunney MM, Patrick S, Gorman SP, et al. Improved detection of infection in hip replacements: a currently underestimated problem. J Bone Joint Surg B (1998) 80:568–72.[CrossRef][Medline]
3 Page Le L, Podglajen I, Chemla E, et al. Molecular diagnosis of a vascular prosthesis infection, due to Propionibacterium acnes, by amplification and sequencing of 16s rDNA. Clin Microbiol Infect (2003) 9:1125–7.[CrossRef][Web of Science][Medline]
4 Tribus CB, Garvey KE. Full thickness thoracic laminar erosion after posterior spinal fusion in association with late presenting infection. Spine (2003) 28:E194–7.[CrossRef][Web of Science][Medline]
5 Thompson TP, Albright AL. Propionibacterium acnes infections of cerebrospinal fluid shunts. Childs Nerv Syst (1998) 14:378–80.[CrossRef][Web of Science][Medline]
6 Richards BS, Emara KM. Delayed infections after posterior TSRH spinal instrumentation for idiopathic scoliosis: revisited. Spine (2001) 26:1990–6.[CrossRef][Web of Science][Medline]
7 Bayston R, Ashraf W, Barker-Davies R, et al. Biofilm formation by Propionibacterium acnes on biomaterials in vitro and in vivo: impact on diagnosis and treatment. J Biomed Mater Res A (2007) 81A:705–9.[Medline]
8 Rekate HL, Ruch T, Nulsen FE. Diphtheroid infections of cerebrospinal fluid shunts. The changing pattern of infection in Cleveland. J Neurosurg (1980) 52:553–6.[Web of Science][Medline]
9 Setz U, Frank U, Andina K, et al. Shunt nephritis associated with Propionibacterium acnes. Infection (1994) 22:99–101.[CrossRef][Web of Science][Medline]
10 Balogun RA, Palmisano J, Kaplan AA, et al. Shunt nephritis from Propionibacterium acnes in a solitary kidney. Am J Kidney Dis (2001) 38:E18.
11 Rammage G, Tunney MM, Patrick S, et al. Formation of Propionibacterium acnes biofilms on orthopaedic biomaterials and their susceptibility to antimicrobials. Biomaterials (2003) 24:3221–7.[CrossRef][Web of Science][Medline]
12 Oprica C, Nord CE. European surveillance study on the antibiotic susceptibility of Propionibacterium acnes. Clin Microbiol Infect (2005) 11:204–13.[CrossRef][Web of Science][Medline]
13
Mory F, Fougnot S, Rabaud C, et al. In vitro activities of cefotaxime, vancomycin, quinupristin/dalfopristin, linezolid and other antibiotics alone and in combination against Propionibacterium acnes isolates from central nervous system infections. J Antimicrob Chemother (2005) 55:265–8.
14
Svensson E, Hanberger H, Nilsson M, et al. Factors affecting development of rifampicin resistance in biofilm-producing Staphylococcus epidermidis. J Antimicrob Chemother (1997) 39:817–20.
15 National Committee for Clinical Laboratory Standards. Methods for Antimicrobial Susceptibility Testing of Anaerobic Bacteria—Sixth Edition: Approved Standard M11-A6 (2004) Villanova, PA, USA: NCCLS.
16
Kutscha-Lissberg F, Hebler U, Muhr G, et al. Linezolid penetration into bone and joint tissues infected with methicillin-resistant staphylococci. Antimicrob Agents Chemother (2003) 47:3964–6.
17 Kucers A, Crowe SM, Grayson ML, et al. The Use of Antibiotics: A Clinical Review of Antibacterial, Antifungal and Antiviral Drugs (1997) London: Hodder Arnold.
18 Clark CE, Shufflebarger HL. Late-developing infection in instrumented idiopathic scoliosis. Spine (1999) 24:1909–12.[CrossRef][Web of Science][Medline]
19 Kowalski TJ, Berberi EF, Huddleston PM, et al. The management and outcome of spinal implant infections: contemporary retrospective cohort study. Clin Infect Dis (2007) 44:913–20.[CrossRef][Web of Science][Medline]
20
Soriano A, Ortega M, Garcia S, et al. Effects of pyridoxine, rifampin and renal function on haematological adverse events induced by linezolid: a comparative study. Antimicrob Agents Chemother (2007) 51:2559–63.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||