Linezolid Penetration into Bone and Joint Tissues Infected with Methicillin-Resistant Staphylococci

Abstract

Penetration of linezolid into bone and joint tissues was studied by high-performance liquid chromatography in 13 patients suffering from implant-associated infections with methicillin-resistant staphylococci. Mean concentrations of linezolid in infected tissues were greater than 10 mg/liter in a sampling time range of 35 to 124 min after administration of the preoperative dose, except in bone specimens, where they reached 3.9 ± 2.0 mg/liter.

Infection of bone and joint tissues remains one of the most severe problems in orthopedic and trauma orthopedic surgery. Multiresistant gram-positive cocci, including methicillin-resistant Staphylococcus aureus (MRSA) and methicillin-resistant Staphylococcus epidermidis (MRSE), are frequently identified pathogens (4). Linezolid is the first available oxazolidinone antibiotic with potent activity against gram-positive cocci (1). Linezolid was clinically and microbiologically proven as effective as standard vancomycin therapy for patients with MRSA infections (8). Oral and intravenous (i.v.) administrations of linezolid are completely bioequivalent (3, 10). Adequate penetration of an antibiotic into bone and joint infection sites is a prerequisite for antibiotic therapy. Recently, penetration of linezolid into noninfected bone, muscle, and hematoma fluid was determined during routine total hip and knee replacements (5, 7). However, the penetration of linezolid into infected osteoarticular tissues has not been studied. We therefore determined the linezolid concentrations in the bone and joint tissues of patients with confirmed infections due to methicillin-resistant staphylococci.

This study was approved by the local medical research ethics committee, and all patients gave written informed consent. The study extended from April 2002 to November 2002. Patients were eligible for enrollment if they met the following criteria: (i) bone and/or joint infection; (ii) infection caused by MRSA, MRSE, or vancomycin-resistant enterococci; and (iii) indication for surgical debridement of infected or necrotic bone and soft tissues.

The exclusion criteria were (i) concomitant use of a drug known to show adverse effects in combination with linezolid and (ii) untoward effects of linezolid. Enrolled were 13 patients (7 male, 6 female) with an average age of 66 ± 11 (range, 47 to 81) years and an infected total hip (n = 7) or knee (n = 2) endoprosthesis or exacerbation of chronic osteomyelitis in the presence of orthopedic implants (n = 4). Infection was diagnosed 6 to 30 months before study enrolment. Table 1 shows the demographic and clinical data of the patients. A dose of 600 mg of linezolid (Zyvoxid; Pharmacia GmbH, Erlangen, Germany) was administered i.v. for 30 min during induction of anesthesia. Three patients were pretreated with linezolid for 24 h (two i.v. doses of 600 mg). In the postoperative period, 600 mg of linezolid was administered every 12 h i.v. or orally for different periods, as clinically indicated.

TABLE 1.

Patients' demographic and clinical data

Patient no.	Age (yr)/gender	Diagnosisa	Pathogen(s)	C-reactive protein concn (mg/dl)	Soft-tissue destruction	pOPb	Concomitant disease(s) or factorc
1d	48/male	K	MRSE, Enterococcus faecalis, Staphylococcus haemolyticus	1.8	Fistula	8	HBP, smoking
2	69/male	H	MRSA	5.0	Fistula	4	HBP, COLD, CHD, smoking
3	64/male	H	MRSA	0.6	Defect	5	ALC, PAVD, CHD, smoking
4	58/female	K	MRSE, Enterococcus faecalis, Staphylococcus haemolyticus	0.7	Defect	4	HBP, CHD, smoking
5e	82/female	H	MRSE, Pseudomonas aeruginosa	9.9	Defect	4	DM, HBP, CHD, PAVD
6	76/female	H	MRSA, MSSAf	0.6	Fistula	4	DM, HBP, CHD
7	69/female	H	MRSA	1.5	Fistula	3	DM, HBP, CHD
8	78/female	H	MRSE	3.5	Fistula	3	HBP, CHD
9d	69/male	O	MRSE, Enterococcus faecalis	0.7	Fistula	2	HBP, CHD
10	52/male	O	MRSA	27.5	Fistula	8	Smoking
11	55/male	O	MRSA	0.6	Fistula	4	COLD, HBP, CHD, PAVD
12d	51/male	O	MRSA	20.2	Fistula	6	Smoking
13	69/female	H	MRSE	1.5	Fistula	3	COLD, HBP, CHD, PAVD

^aK, knee endoprosthesis; H, hip endoprosthesis; O, exacerbated osteomyelitis in the presence of orthopedic implant.

^bpOP, previous operations.

^cHBP, high blood pressure; COLD, chronic obstructive lung disease; CHD, coronary heart disease; ALC, alcoholism; PAVD, peripheral arterial vessel disease; DM, diabetes mellitus.

^dTreated with linezolid for 24 h before surgery.

^eAdditionally treated with imipenem-cilastatin.

^fMSSA, methicillin-sensitive Staphylococcus aureus.

During surgical debridement of necrotic and infected tissues, samples were collected in a time range of 35 to 124 min after linezolid infusion began. Simultaneously, 2 ml of peripheral venous EDTA-anticoagulated blood was drawn separately. Transport of tissue and blood samples to the laboratory followed immediately. Plasma was obtained by blood centrifugation (5 min, 2,000 × g). All samples were frozen at −80°C prior to assay. Thawed tissue specimens (0.2 to 0.7 g) were rapidly rinsed with phosphate-buffered saline until a clear supernatant was obtained. Bone samples free of connective tissue were crushed with an analytic mill (IKA A11 basic; IKA-Werke GmbH, Staufen, Germany). Other tissue samples were sliced into small pieces (up to 5 mm). The hacked-up tissue samples were transferred into a porcelain mortar, covered with fluid nitrogen, and further ground up with a pestle. Subsequently, the ground material was transferred to a 20-ml glass tube, 5 ml of acetonitrile-methanol (50/50, vol/vol) was added, and the mixture was stirred for 2 h at room temperature with crosshead magnetic stirring bars. Linezolid was extracted from plasma by addition of 1.9 ml of acetonitrile-methanol (50/50, vol/vol) to a 100-μl sample volume. This mixture was vigorously vortexed and allowed to settle for 2 h at room temperature. Finally, the mixture was centrifuged at 16,000 × g (Biofuge pico; Kendro Laboratory Products, Hanau, Germany) and aliquots of 20 μl of the supernatants were subjected to reversed-phase high-performance liquid chromatography (HPLC). Isocratic HPLC was performed with a Waters 2690 separation module (Waters, Eschborn, Germany) consisting of a multiple-solvent delivery system and an automatic sample injection module (Waters Alliance System). Samples were separated by a Waters Symmetry C₁₈ reversed-phase column (4.6 by 150 mm, 5-μm particles) that was protected by a Waters Sentry C₁₈ guard column (3.9 by 20 mm, 5-μm particles). The absorbance of the column effluent was monitored with a Waters 2487 dual-wavelength absorbance detector adjusted to 254 nm. The peak areas were calculated with a chromatography manager program (Millennium 2.15.01; Waters). The linezolid detection solvent system was 30% methanol supplemented with 1% ortho-phosphoric acid and heptanesulfonic acid (2 g/liter), which was adjusted to pH 5.0 with sodium hydroxide (10 M) as previously described by Tobin et al. (9). The flow rates were maintained at 1 ml/min. The solvent was automatically degassed and constantly stirred during HPLC analysis. Identification and quantification of linezolid were performed by external standardization. The linearity of peak areas to drug concentration was r = 0.9999 across a concentration range of 0.02 to 200 mg of linezolid per liter. The lower limit of quantification (9) was calculated to be 0.01 mg/liter. Linezolid recovery from spiked tissue samples was in the range of 95 to 110%, similar to that reported by others (2, 5). Assay reproducibility was as follows: intraday, <5%; interday, <11%. Data are expressed as means ± standard deviations and data ranges (minimal and maximal values).

The linezolid concentrations in infected tissues (from either knee or hip joints) rapidly reached mean values greater than 10 mg/liter after administration of the preoperative 600-mg i.v. dose (sampling time range of 35 to 124 min after infusion start), except for bone specimens, in which they reached a mean of 3.9 ± 2.0 mg/liter (Table 2) in spite of sufficient parallel concentrations in plasma (>11 mg/liter). Mean penetration of linezolid into noninfected bone tissue has also been determined: 6.3 or 8.5 mg/liter (5, 7). Since our analytical linezolid detection method is more sensitive than previously described methods (2, 9) and we obtained linezolid tissue recoveries similar to those previously reported (2, 5, 9), we suggest that differences in linezolid concentrations in bone may be due to differences in sample preparation or to an inflammation-related decrease in the blood supply to the infected bone. In addition, two patients were admitted for revision surgery 9 and 15 days after the initial surgery. These patients continuously received standard linezolid therapy between the two operations. The linezolid concentrations in their bone samples were 4.6 and 2.5 mg/liter, respectively (not listed in Table 2). Furthermore, two pure corticalis bone specimens free of any adherent tissue and extracted without the washing step revealed linezolid concentrations of 0.8 and 1.4 mg/liter (not listed in Table 2).

TABLE 2.

Resection time and linezolid concentration in tissue samples

Tissue	Timea (min)		nb	Linezolid concn (mg/liter)
	Mean ± SD	Range		Mean ± SD	Range
Fascia	48 ± 26	35-100	10	13.1 ± 5.4	5.6-23.7
Joint capsule	49 ± 10	35-62	7	12.6 ± 2.9	7.6-16.3
Bone (cancellous)	54 ± 18	35-89	11	3.9 ± 2.0	1.2-8.3
Bone marrow	71 ± 35	35-124	5	10.5 ± 2.2	8.0-14.1
Granulation tissue (periprosthetic)	55 ± 10	43-70	7	13.6 ± 7.7	7.9-30.1
Granulation tissue	61 ± 37	35-87	2	10.3 ± 1.4	9.3-11.3
Subcutis	49 ± 19	35-62	3	13.2 ± 1.1	12.0-14.1
Tendon	84 ± 50	49-119	2	14.5 ± 3.5	12.0-16.9
Plasma	54 ± 19	35-89	13	17.1 ± 5.1	11.1-26.6

^aTime after start of linezolid infusion.

^bn = number of tissue specimens from different patients.

In summary, our data indicate that linezolid rapidly reaches infected tissue compartments of joints and the tissues surrounding bone in concentrations greater than twice the MIC for 90% of the strains tested (1 to 2 mg/liter for MRSE, 1 to 4 mg/liter for MSSA, MRSA, and Enterococcus faecalis) (6). However, intra-bone tissue concentrations of linezolid below the MIC for 90% of the strains tested also occur; thus, an aggressive surgical approach to bone infection should be taken if possible.