Calcium sulphate mixed with antibiotics does not decrease efficacy against Cutibacterium acnes (formerly Propionibacterium acnes), in vitro study

Abstract

Background

This study explored the in vitro efficacy of antibiotics mixed with calcium sulfate (ACS) against Cutibacterium acnes (C. acnes).

Methods

C. acnes isolates from orthopaedic infection sites were tested for antimicrobial susceptibility with ACS. Minimal inhibitory concentrations (MIC) were determined with a gradient diffusion method (Etest® strips).

Results

When tested with Etest®, all 22 isolates were susceptible to penicillin, ceftriaxone, vancomycin, and two were resistant to clindamycin (MICs of 4 and 8 mg/L). Penicillin and rifampin had the largest inhibition zone diameters.

Conclusions

Antibiotics retained activity against C. acnes when mixed with calcium sulfate.

1. Introduction

Cutibacterium acnes, formerly named Propionibacterium acnes, has gained prominence in the past decade among the pathogens found in deep post-operative infections of the shoulder.1, 2, 3, 4, 5 This facultative anaerobic Gram-positive rod is a commensal skin bacteria, found in the deep layers of the skin.⁶ The shoulder has been recognized as a region particularly susceptible to infection by C. acnes, when compared to the knee or the hip.7, 8, 9 Although a standardized method to conduct antimicrobial susceptibility testing (AST) for C. acnes is lacking, the bacteria is usually susceptible to penicillin G, third-generation cephalosporins, vancomycin and rifampin.10, 11, 12 It also shows at least some resistance to clindamycin.¹³

In the treatment of deep post-operative infections, local antibiotics diffusion may be used as an adjuvant to intravenous antibiotics.¹⁴ Various methods can be employed, such as spreading antibiotic powder locally, antibiotic mixed with cement, or antibiotic mixed with synthetic graft.^15,16 Stimulan® (Biocomposites, Keele Science Park, Staffordshire, United Kingdom) is a synthetic biocompatible bone graft material of calcium sulfate,^15,17 which can be moulded into small beads (3 mm, 4.8 mm or 6 mm in diameter). These beads are then used to fill a bone defect (for example, chronic non-healing fractures, osteomyelitis) or a dead space. During surgery, the orthopaedic surgeon mixes the Stimulan® powder and solution with the antibiotic and fills the mould under sterile conditions (Fig. 1). However, this method is not validated against the most frequent germ in shoulder post operative infection: C. acnes. Dunne et al. reported that, for gentamicin, the advantage of combining calcium sulfate to an antibiotic is its slower release, providing a longer antibiotic presence.¹⁸ The drawback is the possibility that antibiotic efficacy could be affected by the calcium sulfate. Indeed, Bistolfi et al. reported poor release of vancomycin when mixed with bone cement.¹⁹ Stimulan® is currently approved to be mixed with the following antibiotics: vancomycin (powder), tobramycin (powder and liquid) and gentamicin (liquid).

Fig. 1.

a) example of patient with chronic shoulder arthroplasty infection and humeral stem subsidence. b) postoperative temporary antibiotic cement spacer and stimulant beads with vancomycin.

The main goal of the present study was to explore the in vitro efficacy of a variety of antibiotics against Cutibacterium acnes when mixed with calcium sulfate in beads and compare it to inhibition zones obtained by the Kirby-Bauer (KB) method and minimal inhibitory concentration (MIC) assessed by Etest®. Since much higher concentrations of antibiotics are usually present in calcium sulfate beads than in disks, it was hypothesized that, for susceptible C. acnes isolates, the inhibition zone of calcium sulfate beads would be larger than the one of standardized disks used for KB methods. The secondary goal was to study the behavior of calcium sulfate when mixed with each antibiotic and identify any potential solidification or texture problem.

2. Methods

2.1. Identification of clinically significant isolates

All consecutive C. acnes isolates identified from orthopaedic surgical specimens between 2013 and 2016 were sent to the microbiology laboratory of Hôpital du Sacré-Coeur de Montréal and kept frozen at −80 °C. C. acnes was identified either by rapid identification (Gram-positive coryneform rods, positive 15% catalase and positive indol spot test) or API Rapid ID 32A test (bioMérieux, Marcy-L’Étoile, France). For inconclusive results, matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF) (VITEK MS, bioMérieux, Marcy-L’Étoile, France) was used to confirm identification (IVD 2.0 database). Isolates were considered significant and included in the present study if identified 1) in more than one culture of deep tissue sample from a single surgery or 2) in one sample if the patient presented with a clinical infection and no other causative pathogen was identified. Medical records were reviewed to confirm the clinical significance of C. acnes isolates. Isolates identified in only one sample, in the absence of clinical infection, were considered as colonizing the skin and were excluded from this study. Only one isolate per patient per episode was included for further testing.

2.2. Antimicrobial susceptibility testing (AST) and inhibition zone reading

AST was performed for each isolate with a panel of six antibiotics (penicillin G, ceftriaxone, vancomycin, tobramycin, rifampicin and clindamycin) by using a gradient diffusion method with Etest® strips. For each isolate, a 0.5 McFarland inoculum was inoculated on a CDC anaerobe blood agar and incubated in a GasPak EZ anaerobe container system (BD Corp., Franklin Lakes, NJ) at 37 °C for 48 h. MICs were interpreted according to breakpoints published by CLSI (M100-S29), whenever possible,¹¹ and the EUCAST standard was used for the vancomycin breakpoint.²⁰ The isolates were then tested using the Kirby-Bauer method for all antibiotics (except rifampicin - disks not available) under the same conditions. The diameter of the inhibition zone was reported but not interpreted, since breakpoints have not been defined with this method. The same process was later used to record the inhibition zone diameter values for the antibiotic-calcium sulfate beads, which had a similar contact surface as the Kirby-Bauer disks, i.e. 6 mm.

2.3. Calcium sulfate beads preparation

In accordance with the manufacturer's monograph, 1.6 million IU of penicillin and 1000 mg of the other antibiotics, in powder form, were mixed with a 10 cc Stimulan® pack Rapid Cure under sterile conditions. The mixture was spread out on a mould to create 6 mm diameter beads (Fig. 2). Therefore, the base of each bead had a flat surface with a 6 mm diameter. Once solidified, the beads were transferred into six sterile containers, one for each of the six antibiotics. with antibiotic calcium sulfate beads and antibiotic-impregnated disks. Each bead contained 2.5 mg of antibiotic, whereas the penicillin beads contained 4167 IU per bead.

Fig. 2.

Diameter of a calcium sulfate bead (6 mm).

To assess the workability of the mixtures, the following characteristics were recorded by the research team: color of beads and setting time.

2.4. MIC and inhibition zone distribution

MIC₅₀ and MIC₉₀ were computed for each antibiotic. The 50th and 90th percentile of inhibition zones for both KB disks and calcium sulfate beads were also calculated.

3. Results

3.1. Antimicrobial susceptibility testing

Between 2013 and 2016, 41 isolates of C. acnes were retrieved from orthopaedic surgical samples in our institution. Of those, 22 were deemed clinically significant, including 15 specimens from shoulder surgeries (Table 1). Clinically significant isolates were further tested for AST with Etest® strips. According to CLSI and EUCAST breakpoints, all isolates were considered susceptible to penicillin G (MIC₉₀ = 0.064 mg/L), ceftriaxone (MIC₉₀ = 0.125 mg/L) and vancomycin (MIC₉₀ = 0.5 mg/L). Two isolates were considered resistant to clindamycin with MIC of 4 and 8 mg/L. Although no breakpoint is available, overall MICs to rifampin were very low (MIC₉₀ = 0.008 mg/L) and MICs to tobramycin were very high (MIC₉₀ = >256 mg/L) (Table 2).

Table 1.

Demographic data.

Clinically significant isolates (n = 22)
Age (range)	64 (27–98)
Gender
Male (%)	15 (68)
Female (%)	7 (32)
Surgery site
Shoulder (%)	15 (68)
Spine (%)	4 (18)
Other (%)	3 (14)

Table 2.

Comparison of inhibition zone diameter values of antibiotic bead vs antibiotic disk of clinically significant isolates (n = 22).

MIC (mg/L)				Inhibition zone diameter (mm)
	Etest®			Antibiotic bead				Antibiotic disk
Antibiotic	MIC50	MIC90	MIC range	Concentration antibio/bead	Median	90th percentile	Bead range	Concentration antibio/disk	Median	90th percentile	Disk range
Penicillin	0.032	0.064	0.016–0.25	4176 UI	60	68	56–68	10 UI	50	50	44–52
Clindamycina	0.032	0.064	0.016–8	2.5 mg	54	60	38–60	2 ug	40	44	6–46
Rifampin	0.008	0.008	0.008–0.064	2.5 mg	50	56	42–56	–	–	–	–
Vancomycin	0.25	0.5	0.25–0.5	2.5 mg	43	46	40–48	30 ug	38	42	30–42
Ceftriaxone	0.064	0.125	0.016–0.125	2.5 mg	40	42	32–52	30 ug	46	48	40–54
Tobramycin	>256	>256	128- >256	2.5 mg	26	32	18–34	10 ug	6	6	6–6

^a2 isolates found to be resistant.

3.2. Inhibition zones for discs and calcium sulfate beads

The inhibition zone diameter values for antibiotic beads were very large with median values ≥ 40 mm for all antibiotics, except tobramycin. The 3 calcium sulfate antibiotic beads with the largest inhibition zone diameter values against C. acnes were: penicillin (90th perc = 68 mm), clindamycin (90th perc = 60 mm), and rifampin (90th perc = 56 mm).

Compared to the antibiotic disks, inhibition zone diameters obtained with the calcium sulfate antibiotics beads were slightly larger for all antibiotics, except ceftriaxone. This was also observed in the clindamycin-resistant strains. Calcium sulfate tobramycin beads showed an increase in the inhibition zone diameters despite all isolates being resistant with tobramycin disk (range from 18 to 34 mm with beads as compared to 6 mm with disks). Inhibition zone diameters obtained with the calcium sulfate ceftriaxone beads were smaller than the disks, despite retaining large inhibition zone diameters (90th perc = 42 mm with beads as compared to 90th perc = 46 mm with disks).

3.3. Practical characteristics

The different characteristics of each antibiotic mixed with calcium sulfate were noted during preparation and were found satisfactory. All antibiotic beads were white except for the rifampin, which was brown. Setting time varied from one antibiotic to another: rifampin, penicillin and tobramycin took 5–7 min, 6–8 min, and 10 min respectively to set, while the other antibiotics took between 3 and 4 min (Table 3).

Table 3.

Characteristics of antibiotics mixed with calcium sulfate.

Antibiotics	Concentration of antibiotic per bead of 6 ml	Color of beads	Approximate setting time (min)
Penicillin	4176 UI	White	6 to 8
Ceftriaxone	2.5 mg	White	3 to 4
Vancomycin	2.5 mg	White	3 to 4
Clindamycin	2.5 mg	White	3 to 4
Rifampin	2.5 mg	Brown	5 to 7
Tobramycin	2.5 mg	White	10

4. Discussion

This study aimed to explore the efficacy of different antibiotics mixed with calcium sulfate against C. acnes and to compare the inhibition zone diameters of calcium sulfate beads mixed with antibiotic and standard antibiotic-impregnated disks used for the Kirby-Bauer method. The inhibition zone diameters of calcium sulfate beads containing penicillin, ceftriaxone, clindamycin and vancomycin was comparable or better than for the equivalent antibiotic disks. Of the 6 antibiotics tested with the beads, penicillin, clindamycin and rifampin had the largest inhibition zone diameters. However, given the large difference between the antimicrobial concentration of beads and disks, it is very likely that there is some effect of the calcium sulfate on antimicrobial diffusion and/or effectiveness. Our study did show, however, that antibiotics mixed with calcium sulfate in beads retain some activity against C. acnes. Aiken et al. studied the antibiotic diffusion of 6 mm-diameter calcium sulfate beads with combined tobramycin, vancomycin, rifampin and calcium sulfate. They found that the calcium sulfate matrix delivered effective local antibiotic concentrations for a period of at least 42 days at 37 °C.²¹ On the other hand, Roberts et al. reported a time-diffusion of 20 days for vancomycin-loaded beads.²² However, the amount of antibiotic contained in the beads as well as the diameter of the beads were different, making the comparison between the diffusion results difficult. Furthermore, antibiotics combined with calcium sulfate beads solidified in a way that would make them useable in a real surgical setting. Calcium sulfate is already used in orthopedics to fill bone deficits, and while some studies have examined the efficacy of the calcium sulfate antibiotic mixture against different bacteria, none have tested for C. acnes.^15,16

Among the antibiotics tested, some have been used against C. acnes with suboptimal results according to the literature, especially for clindamycin.^3,13,23 Piggott et al. reported a 6% resistance to clindamycin, compared to 9% for Khassebef et al.^1,13 This antibiotic has indeed shown variability in susceptibility against C. acnes, especially in strains previously exposed to clindamycin.^3,23 It is therefore not surprising that we found two resistant isolates in our sample. Interestingly, when using clindamycin calcium sulfate beads at the recommended concentration (i.e. 2.5 mg per bead), the 2 clindamycin-resistant strains showed large inhibition zone diameters (38 and 60 mm). We can hypothesize that the concentration achieved locally through diffusion was high enough to partly overcome the resistance observed with the standard concentration used in the disk. However, there were no control beads to evaluate what the effect of antibiotic-free calcium sulfate beads may be on C. acnes growth inhibition, and this should be further studied.

Similarly, all tobramycin-resistant strains showed increased inhibition zone diameters when exposed to tobramycin beads (from 6 mm to ranging from 18 to 34 mm), despite tobramycin being generally considered useless against C. acnes.²⁴ This assumption might need to be revisited when higher concentrations are achievable, as is the case in therapies administered directly to the surgical site.

Rifampin also appears to be a good choice when combined with calcium sulfate. Due to the absence of a rifampin impregnated disk, no comparison was performed. Nevertheless, all our isolates were known to harbor very low MICs with the E-test method.²⁵ CLSI has also recently reported 80 isolates with MIC values ≤ 0.03 mg/L using the agar dilution method,¹¹ which is in accordance with the very low MIC we observed in our study. According to Olsson J et al., 25 isolates, from various anatomical regions, showed a MIC_50/90 of ≤0.016/≤0.016 for rifampin.²⁵ Moreover, Furustrand T et al. report a MIC₅₀ equal to 0.007 for rifampin with C. acnes strain ATCC 11827²⁶. A potential advantage of using rifampin against C. acnes is that, unlike other antibiotics, it has an anti-biofilm activity.²⁶ However, due to the potential emergence of acquired drug resistance, particularly to Staphylococcus aureus, rifampin should never be used in monotherapy.²⁷ Thus, when mixed with calcium sulfate, rifampin should probably be used in combination with another effective antibiotic.

Lastly, contrary to the other antibiotics, strains of C. acnes showed smaller inhibition zone diameters with ceftriaxone beads than with disks. One hypothesis is that the diffusion of ceftriaxone is impeded when mixed with calcium sulfate. Mousset et al. calculated the diffusion time of cylindrical calcium sulfate implants of 6 mm by 3 mm, to which a third-generation cephalosporin was mixed. Cephalosporins were released in 2–3 days, compared to quinolone and glycopeptides that were still detectable after two weeks.²⁸ Nevertheless, ceftriaxone beads seem to retain some efficacy with inhibition zones ranging from 32 to 52 mm.

This exploratory in vitro study has several limitations. First, our main comparator is based on the Kirby-Bauer method which has yet to be standardized for C. acnes. However, having the same contact surface (i.e. 6 mm diameter beads and disks) could have helped create comparable experimental conditions. Second, the contact area of the antibiotic bead with the blood agar was limited to its flat surface, whereas in clinical practice most of the bead is in contact with the tissues on more than one surface. Therefore, it seems reasonable to assume that the in vivo efficacy would probably be better, since its diffusion surface would be greater. Third, it would have been interesting to vary the amount of antibiotic in the mixture to further optimize its effect. Lastly, since the rifampin disk was not available, it was not possible to test rifampin with the Kirby-Bauer disk. It would have been very interesting to compare the inhibition zone diameter obtained by the rifampin bead with the impregnated disk, since rifampin has a small MIC value against C. acnes.²⁵

This study suggests that antibiotic-impregnated beads retain some activity against Cutibacterium acnes when tested in vitro against Kirby-Bauer disks. Of the 6 antibiotics tested, penicillin and rifampin had the largest inhibition zone diameters. Isolates resistant to tobramycin and clindamycin also showed some inhibition zones. Further testing should focus on different concentrations to achieve the optimal dose and dilution tests to determine how long the local antibiotic concentration remains therapeutic. Further research on the potential synergy of the antibiotics most effective against C. acnes strains when mixed with calcium sulfate will also need to be explored further.

Ethical review statement

IRB approval was waved for this study by the Comité d’Éthique à la Recherche CIUSSS Nord-de-l’îles de Montréal.

Declaration of competing interest

Research material provided by Smith & Nephew, no involvement in any aspect of the research. Dominique M Rouleau is a consultant for Bioventus and Wright Medical. Jean-Michel Leduc has received speaker honorarium from Biomerieux. Benoit Benoit is a consultant for Bioventus. Stéphane Leduc is a consultant for Stryker. Departmental funding to the institution for educational and research purposes for one or more of the authors from: Arthrex, Conmed, Depuy, Linvatec, Smith & Nephew, Stryker, Synthes, Tornier, Wright, Zimmer Biomet.