Prophylaxis of Acute Osteomyelitis with Absorbable Ofloxacin-Impregnated Beads

Abstract

We investigated the use of an ofloxacin-impregnated bioabsorbable composite for the prevention of acute Staphylococcus aureus osteomyelitis. New Zealand White rabbits were anesthetized, the femur was exposed, and a cortical hole was drilled. Animals were randomly given drug-free composites or ofloxacin-impregnated composites; the composites were placed at the site of injury, and the incision was closed. One hour later, all animals were intravenously inoculated with 5 × 10⁴ CFU of S. aureus and observed for 28 days. Bone culture data revealed that S. aureus was isolated from 3 of 12 rabbits in the ofloxacin composite group and 9 of 11 animals in the control group (P = 0.02). Radiographic evaluation revealed that the drug-free group had a significantly (P = 0.01) greater degree of radiographic evidence of infection than the group given ofloxacin composites. Although a limited number of histologic samples were available, these data also paralleled the radiographic and culture data. This study demonstrates the effectiveness of the implantable ofloxacin bioabsorbable composites to prevent the development of acute osteomyelitis.

Osteomyelitis is a disease which can be defined as bone and bone marrow inflammation. At least two predisposing factors are required for the development of osteomyelitis. The first is trauma, and the second is the introduction of bacteria, which usually occurs via the systemic circulation or directly from open-fracture wounds (2–5). Although infection is an infrequent accompaniment of most orthopedic surgical procedures, the consequences of bone infection are often devastating, requiring both prolonged antibiotic therapy and the possibility of extensive surgical intervention. Therefore, the prevention of infection is one of the primary objectives in the orthopedic management of the patients at increased risk, such as those undergoing treatment of open fractures or total joint arthroplasty procedures (10, 12). Even though debridement for an open fracture is considered completely acceptable, the concern of postoperative infection and its associated morbidity is justification for the use of local prophylactic antimicrobial therapy. In addition, although the use of preventive antibiotics in patients undergoing prosthetic joint arthroplasty is still debated, it is now accepted practice.

Previously, we have reported the in vitro elution characteristics of bioabsorbable lactide-glycolide polymer composites impregnated with ofloxacin at physiologic temperature (11). During these studies, high ofloxacin concentrations were initially observed after the onset of the study and were followed by a more gradual and sustained release of the antimicrobial over a 60-day period. As a result of these data it was concluded that the antibiotic-containing beads may be suitable for in vivo drug delivery. The high initial concentrations may be appropriate for infection prophylaxis, while the prolonged ofloxacin concentration profile of the composites may have therapeutic benefits in the treatment of active infection. The purpose of this study was to investigate the effectiveness of an ofloxacin-impregnated bioabsorbable composite for the prevention of acute osteomyelitis after systemic bacterial inoculation.

MATERIALS AND METHODS

Twenty-four conventional female New Zealand White rabbits (Pines Acres Rabbitry, West Brattleboro, Vt.) were obtained and cared for according to the guidelines provided by the U.S. Department of Health and Human Services (9). Prior to implementation, the protocol was approved by our institutional animal care and use committee. Animals were allowed food and water ad libitum throughout the study; however, animals were made to fast for the 12 h prior to surgery.

To simulate systemic bacterial exposure which may occur in the postoperative period, a localized injury was created in the femur, with subsequent bacterial inoculation. As described by Morrissy and Haynes (8) these procedures often result in acute osteomyelitis and may be a suitable model for the determination of therapies directed at the prevention of hematogenous infection. Morrissy and Haynes used a method which incorporated a minor epiphyseal plate fracture; our modification of their technique consisted of a cortical hole in the femur. The femur was chosen as the anatomical site because it permitted the easy insertion of antibiotic composites into the intramedullary cavity.

Surgical procedure.

Animals were anesthetized with a 1.25 ml of ketamine-acepromazine per kg of body weight in a 10:1 (Ketaset [100 mg/ml]-PromAce [10 mg/ml]; Aveco, Fort Dodge, Iowa), the left femur was exposed under sterile conditions, and a cortical hole was drilled (3-mm diameter) 2 cm from the distal end. After the creation of the bone defect, animals were randomly given either drug-free composites (control) or ofloxacin-impregnated composites. Ofloxacin-impregnated bioabsorbable composites containing 10 mg ofloxacin (Ortho Pharmaceuticals, Raritan, N.J.) plus 30 mg of dl-lactide–glycolide polymer (85:15; Medisorb Technologies International LP, Cincinnati, Ohio) were prepared as previously described (11). A total of three composites of either the drug-free or ofloxacin-impregnated beads were implanted into each animal (two in the intramedullary cavity and one outside the bone). The polymer used in the drug-free composites and that containing active drug were handled in an identical manner. At the conclusion of the procedure, the fascia, subcutaneous tissue, and skin were closed by conventional surgical techniques. One hour later, animals were inoculated (as described below) and were observed for 28 days.

Bacterial inoculation.

The isolate used in this study was a methicillin-susceptible strain of Staphylococcus aureus (ATCC 29213). The MIC of ofloxacin for this isolate is 0.125 to 1 μg/ml (16). A fresh culture was prepared by inoculating cation-adjusted Muller-Hinton broth and incubating the solution overnight at 37°C. The overnight culture was then diluted in 0.9% saline, and 1 ml of this S. aureus suspension containing 5 × 10⁴ CFU was injected intravenously via the marginal ear vein. The CFU counts of all bacterial suspensions were confirmed with quantitative techniques.

Roentgenogram.

Anteroposterior and lateral radiographs of each left femur were obtained 1 day (day 27) before the rabbits were sacrificed. All films were read separately by two observers without knowledge of the mode of treatment. Five parameters were assessed on each film: periosteal reaction (involucrum formation), loss of cortical thickness, density changes, development of a radiolucent line at the bone-beads interface, and presence of sequestra. Each parameter was assigned a numerical score (from 0 [normal] to 6 [extensive]) as previously described (15).

Collection and processing of bone and soft tissue.

Animals were sacrificed on day 28, and the infected femora were surgically extracted by aseptic procedures. After each bone was cleaned of all soft-tissue debris a 2-cm segment (1 cm proximal and distal to initial injury site) was removed. A portion of the sample was sent for histologic evaluation in 10% buffered formalin fixative; the portion was decalcified in 5% nitric acid, embedded in paraffin, and cut into 4-μ-thick sections, which were stained with hematoxylin and eosin. All samples were examined by one pathologist without knowledge of the mode of treatment. Histologic analysis evaluated the adequacy of the sample, depending on the sample size and amount of bone marrow space and the presence or absence of acute and chronic osteomyelitis. Acute osteomyelitis was diagnosed when acute inflammation, characterized by neutrophils, abscess formation, bone necrosis or a combination of these features, was identified. Chronic osteomyelitis was diagnosed by observing the presence of chronic inflammation with lymphocytes and plasma cells usually in a fibrotic background. Fracture callus was identified by the presence of granulation tissue, richly vascularized loose connective tissue, and/or cartilage and new bone formation.

The remaining bone was ground to a powder with a freezer mill, and approximately 500 mg was added to Trypticase soy broth and incubated at 37°C for 24 h. After the incubation period, broth was directly plated onto blood agar plates, incubated, and assessed for growth. For the purposes of this study, a positive culture was indicative of osteomyelitis.

Residual bone powder was used for ofloxacin concentration determinations. Soft tissue in the proximity of the initial defect was also collected, ground, and used for the determination of drug concentrations. All bone and soft-tissue samples were stored at −80°C until the time of analysis.

Ofloxacin concentrations in bone and soft tissue.

Ofloxacin bone and soft-tissue concentrations were determined by reverse-phase high-pressure liquid chromatography (HPLC) methods using fluorescent detection (λ_ex = 282 nm), a Nucleosil C₁₈ column (10 μm, Alltech Associates, Deerfield, Ill.), and a modification of the procedure previously described (7). The procedure was modified to use 0.07 M sodium phosphate as the back-extraction solution. Briefly, 125 mg of unknown bone or soft tissue or blank bone powder was added to 125 μl of phosphate buffer, vortexed, and sonicated for 5 min. Once the mixture was homogenized, 50 μl of the appropriate aqueous ofloxacin standard was added to the blank bone samples and 50 μl of HPLC water was added to all unknown samples. Then 50 μl of the internal standard (pipemidic acid; 40 μg/ml) was added to all samples and the samples were vortexed. Chloroform (3.5 ml) was added, and the mixtures were shaken for 10 min and centrifuged at 3,000 × g for 10 min. The organic phase was removed, and 200 μl of 0.07 M sodium phosphate solution was added; the mixtures were shaken for 20 min and centrifuged at 3,000 × g for 15 min. The aqueous layer was removed, and 20 μl was injected into the HPLC system. The standard curve for the bone and soft-tissue assays ranged from 0.1 to 10.0 μg/g of tissue. The interday coefficients of variation (CV) of both assays for the low (0.2 μg/ml) and the high (8.0 μg/ml) quality control samples were less than 5%, while the intraday CV were less than or equal to 4%.

Statistical analysis.

Radiography scores from different treatment groups were compared by the Mann-Whitney test. The percentage of sterile femora in the study was assessed by the chi-square test. A P value of ≤0.05 was considered significant.

RESULTS

Twenty-three animals completed the four week protocol, 11 in the drug-free composite group (control) and 12 in the ofloxacin composite treatment group. One rabbit in the drug-free composite group was sacrificed after the animal was found to have a femoral fracture which limited mobility and the ability to feed. Soft-tissue swelling was only occasionally observed one week after surgery, and all the wounds of both groups healed well. During the 28-day observation period all animals continued to maintain their weight and showed no apparent signs of systemic infection.

Bone culture.

Gross inspection of the bone samples in the control group revealed evidence of bone destruction and irregularities which were not apparent in the treatment group. Bone culture data are reported in Table 1. Ofloxacin composite-treated animals had a significantly greater rate of sterilization than those in the drug-free composite group.

TABLE 1.

Bone culture, radiography, and ofloxacin concentration data 28 days after intravenous inoculation with S. aureus

Treatment	Bone culture dataa (no. positive/total no. tested [%])	Radiography scoreb (mean ± SD)	Ofloxacin concn (μg/g) (median [range])
Drug-free composite (n = 11)	9/11 (82)	6.1 ± 3.3
Ofloxacin composite (n = 12)	3/12 (25)	2.7 ± 2.2	32.9 (1.7–160.0)c
			3.5 (0.2–9.3)d

^aP = 0.02. 

^bP = 0.01. 

^cValue for bone. 

^dValue for soft tissue. 

Radiography studies.

Radiographic examination revealed some evidence of osteomyelitis. In the control group the bone-bead interface was characterized by the appearance of a lucent zone. Involucrum formation characterized by sclerosis of the cancellous bone adjacent to the beads was also observed in this group. In the ofloxacin composite group the radiographic changes were slight with the exception of those for one rabbit. A comparison of the group data by using the radiographic evaluation system is displayed in Table 1. The drug-free group had a significantly greater degree of radiographic evidence of infection than the ofloxacin treatment group.

Ofloxacin concentrations in bone and soft tissue.

Concentration data for the animals receiving the ofloxacin-containing composites are shown in Table 1.

Histology studies.

Histologic samples were evaluated for 11 of 23 cases, 5 from the control group and 6 from the ofloxacin composite-treated animals. In the control group all samples were adequate for evaluation; four of five showed acute and chronic osteomyelitis, and one of five showed only chronic osteomyelitis. In the ofloxacin composite group one sample was inadequate for evaluation, three were adequate but suboptimal due to small size, and two were fully evaluable. In this group, two of five samples showed acute and chronic osteomyelitis and three of five showed only granulation tissue, presumably representing fracture callus. No evidence of acute and chronic inflammation was noted in the latter three samples; however, the sample size may be too small for the bone lesion to be apparent. These findings suggest that the placement of the ofloxacin composites reduces the appearance of osteomyelitis in this model. Although the limited number of histologic samples in each group precludes the performance of reliable statistical analysis, a parallel between the histologic findings, the radiographic findings, and the culture data is noted.

DISCUSSION

Infection remains the most serious complication following orthopedic surgery, and the delivery of antibiotics to the operative area is one method to reduce the infection frequency. Buchholz and Engelbrecht first reported on the clinical utility of the addition of antibiotic to bone cement in 1970 (1). Antibiotic concentrations attainable in the tissue surrounding the implanted drug-cement composite were a good indication of efficacy for the prevention of infection due to surgical site contamination.

Other than the use of antibiotic-containing cement, there is no other delivery system which achieves high local concentrations that exceed, for many weeks, the concentrations required to kill commonly infecting bacteria at the site of infection (13). Although individually prepared beads have been used clinically for more than 20 years, these products are not commercially available in the United States. In an attempt to improve this drug delivery system for bone, several investigators have reported on the development and use of biodegradable vehicles (6, 14, 17, 18). These newer delivery systems not only provide the advantages of high local drug concentrations but also can be absorbed and therefore do not require a second surgery for removal. In addition, the lactide-glycolide polymer used in the present study causes minimal tissue reaction at the site of implantation (17).

Previously, we have reported the in vitro elution characteristics of bioabsorbable ofloxacin-polymer composites (11). From our preliminary studies it appeared that the ofloxacin-polymer composites may be suitable for in vivo drug delivery, since their release pattern of ofloxacin was very similar to that of antibiotic-impregnated cement beads.

In the present study, we used a modification of a previously described technique for the development of acute osteomyelitis (8). In an attempt to minimize the systemic sequelae of the initial bacterial inoculation and subsequent infection over the 1-month study period, we used an initial bacterial load which was substantially lower than that originally used by Morrissy and Haynes. Despite the reduction in CFU, the majority of control animals in our study had culture-positive bone samples and local evidence of bone abnormalities at the site of initial injury. However, it should be noted that the original experiments of Morrissy and Haynes were conducted over only a 7-day period; thus, a smaller inoculum may be sufficient to cause bone infection with minimal systemic progression if the incubation period is prolonged as in the present study. Although the inoculum employed by the originators of the model produced a higher rate of infection (94 to 100%) than that in our study (82%), our infection rate in the untreated control animals was sufficient to determine an ofloxacin-impregnated composite effect with the study sample size.

In this study, the implantation of ofloxacin-impregnated bioabsorbable composites provided a high local concentration of antibiotic in both the bone and soft tissue relative to the susceptibilities of common infecting pathogens. Ofloxacin concentrations in bone were above the MIC for our S. aureus isolate at the conclusion of the 28-day treatment period for all animals. In general, bone ofloxacin concentrations exceeded those of soft tissue, probably because bone ofloxacin concentrations were the result of placement of the intramedullary and extramedullary composites, whereas soft tissue ofloxacin concentrations were the sole result of extramedullary placement. Despite lower values, drug concentrations in the soft tissue exceeded the MIC for the organism in 11 of 12 animals studied. Although drug delivery was apparently effective, considerable variation in ofloxacin concentration was observed and is likely due to irregularities in the preparation of the composites between batches or the movement of the composites in and around the bone during the study period. This disparity in bone concentrations may in part be due to one of the three positive cultures, since the bone ofloxacin concentration in this animal was the lowest observed (1.7 μg/g). However, the bone ofloxacin concentrations for the other two prophylaxis failures were 136 and 39 μg/g, values that are well above the MIC for the pathogen.

During the first hours of infection, a large number of bacteria are killed by the local defense mechanisms. However, the subsequent course of an infection is determined by the number of organisms that survive this period. Antibiotics may augment the local defenses by enhancing this early killing of bacteria. The observed prophylactic effects of the implantable ofloxacin-impregnated composites in our study are likely due to the early and sustained released of ofloxacin, which results in high local concentrations which reduce the number of viable bacteria to a level at which local defense mechanisms function adequately to eliminate bacteria (19).

The pharmacokinetic data provided by this study further suggest that the utilized polymer is a suitable vehicle for the delivery of high local concentrations of ofloxacin. The radiographic, histologic, and culture data in this study demonstrate the effectiveness of implantable ofloxacin-bioabsorbable composites to prevent the development of acute osteomyelitis after systemic bacterial exposure.