Effect of Body Weight on Cefazolin and Vancomycin Trabecular Bone Concentrations in Patients Undergoing Total Joint Arthroplasty

Patients and Methods

This was a prospective Institutional Review Board (IRB)-approved study. Thirty-four patients who underwent TJA were included in the study. Per standard of care prophylaxis against SSI at the University Medical Center, all patients received cefazolin (1 g if patient weighs <70 kg, 2 g if patient weighs >70 kg) within 60 min prior to skin incision and vancomycin (15 mg/kg) within 120 min of incision [5]. The 15 mg/kg dosing was used as an estimate to determine which of the four standard doses of vancomycin patients would receive: 0.75 g, 1 g, 1.25 g, or 1.5 g. For example, a patient weighing 80 kg at 15 mg/kg would receive 1.25 g of vancomycin. If allergic to penicillin or vancomycin, patients would receive levofloxacin and linezolid, respectively [5]. Vancomycin was used (only one pre-operative dose) as part of the standard prophylaxis protocol at our institution, as recommend by the antibiotic stewardship committee as a result of a high prevalence of MRSA infection in the institution. During the patients' operative procedure, discarded bone from the femoral head during THA and tibial and femoral cuts during TKA were collected. Using a surgical rongeur, trabecular bone was extravagated. Upon collection, trabecular bone was rinsed with 0.9% normal saline solution then dried with gauze and frozen at −80°C. Samples were spun at 2,000 rpm for 10 min and stored in a freezer at −80°C. Trabecular bone was then ground and mixed into a uniform consistency.

Vancomycin bone concentrations were determined using a validated high-performance liquid chromatography (HPLC) method at the Center for Anti-Infective Research and Development (CAIRD), Hartford Hospital [13]. Vancomycin bone samples were run with a cross-matrix validation on the serum standard curve. The assay was linear over a range of 1 to 80 mcg/mL (R² = 0.998). Intra-day (n = 10) coefficients of variation for the low (2 mcg/mL) and high (60 mcg/mL) serum quality control samples were 2.6% and 2.5%, respectively. Intra-day (n = 6) coefficients of variation for the low and high bone quality control samples were 5.3% and 2.4%, respectively. Inter-day (n = 5) coefficients of variation for the low- and high-quality control samples were 2.5% and 4.3%, respectively, for the serum samples and 4.6% and 6.2%, respectively, for the bone samples.

Cefazolin concentrations were determined using a validated HPLC method, as previously reported at CAIRD, Hartford Hospital [14]. Cefazolin bone samples were run with a cross-matrix validation on the serum standard curve. The serum assay was linear over a range of 0.5 to 50 mcg/mL (R² = 0.999). Intraday (n = 10) coefficients of variation for the low (1 mcg/mL) and high (40 mcg/mL) serum quality control samples were 3.6% and 3.4%, respectively. Intra-day (n = 6) coefficients of variation for the low and high bone quality control samples were 5.5% and 1.9%, respectively. Inter-day (n = 5) quality control samples were 4.3% and 8.3%, respectively, for the serum samples and 3.6% and 2.3% for the bone.

Furthermore, patient information including demographic data and intra-operative data were also collected from patient charts and operative data. Patients were separated into different groups based on their BMI (<24.9, 25–29.9, 30–34.9, and >35) as well as the pre-operative dosage of cefazolin and vancomycin that they received. Of the 34 patients in the study, three did not receive cefazolin because of their penicillin hypersensitivity (levofloxacin was administered instead) so cefazolin concentration was determined for the remaining 31 [5]. The concentration of levofloxacin in bone was not determined. Similarly, two of the 34 patients did not receive vancomycin because of hypersensitivity (linezolid was administered instead). The concentration of linezolid in bone was not determined.

Based on recent surveillance data, the susceptibility target value for MSSA to cefazolin is 2 mcg/mL; the target value for MRSA and S. epidermidis to vancomycin is also 2 mcg/mL [12].

Results

Thirty-four consecutive patients who underwent TJA were included in the study. Twelve patients underwent THA and 22 patients underwent TKA. The average age was 66.8 (range, 38–86) with 19 females and 15 males. The average patient weight was 82.49 kg (range, 50.35–124.69 kg), with an average BMI of 29.5 (range, 19.8–41.8).

Of the 31 patients who received cefazolin, 28 patients achieved bone concentrations above the target MSSA MIC value of 2 mcg/mL. There was no statistical difference between concentrations of cefazolin in bone, average dose of cefazolin per kilogram, or average concentration of cefazolin per kilogram between the four BMI separated groups (Table 1A). Regarding pre-operative dose of cefazolin, four patients received 1 g of cefazolin (compared with 2 g for the other 27 patients). Three of these four patients (75%) achieved bone cefazolin concentrations above the MIC levels for MSSA compared with 25 of 27 of the patients who received 2 g pre-operatively (92.6%). There was a significant decrease in the average dose of cefazolin per kilogram (p = 0.003) in patients who received 1 g of the drug compared with those who received 2 g. No difference was noted in average concentration of cefazolin per kilogram between the different dose groups (Table 1B).

Of the 32 patients who received vancomycin, 27 achieved bone concentrations above the target MIC values of 2 mcg/mL for both MRSA and S. epidermidis. There was a significant difference in average dose of vancomycin per kilogram of body weight between the different BMI groups, with an ANOVA F-value of 4.094 (p = 0.02). Similarly, there was a statistically significant difference between average concentration of vancomycin per kilogram of body weight with an ANOVA F-value of 4.885 (p = 0.007) (Table 2A). Regarding the different doses of vancomycin received pre-operatively, there was no difference in average concentration of vancomycin in bone per kilogram between the different dosage groups (Table 2B).

A comparison was also made between the bone concentrations of vancomycin and cefazolin between patients undergoing THA and TKA. Regarding cefazolin, there was a significant increase in average bone concentration (p = 0.005) and average concentration per kilogram (p = 0.006) in patients undergoing THA compared with TKA (Table 3A). Regarding vancomycin, there was no significant difference in average dose of vancomycin per kilogram (p = 0.36) or average concentration of vancomycin per kilogram (p = 0.10) in patients undergoing THA compared with TKA (Table 3B).

The pre-operative dose of drug was also compared between patients who met MIC values and those who did not. For both cefazolin (Table 4A) and vancomycin (Table 4B), there was no difference in pre-operative dose or dose per kilogram between patients who were above MIC values and those who were below MIC values. Because seven of the eight patients who did not achieve adequate bone concentrations underwent TKA, an analysis of the length of time between tourniquet inflation and time of bone cuts was performed for patients undergoing TKA. For cefazolin, the average time was 48 min (standard deviation [SD] = 11.5 min) for those who did not exceed MIC targets and 38 min (SD = 18.7 min) for those who did achieve adequate concentrations; this was not significant (p = 0.32). For vancomycin, the average time was 37 min (SD = 12.3 min) for those who did not achieve MIC target concentrations and 34 min (SD = 17.7 min) for those who did; this was not statistically significant (p = 0.74)

Discussion

The antibiotic concentration in bone during TJA and its effect on the incidence of SSI has not been explored fully. Whereas the concentration of cephalosporins in cortical and trabecular bone during THA has been documented to approach levels nearing the free serum levels, to our knowledge, no study has taken into account patient weight to determine its effect on antibiotic bone concentration [8]. Furthermore, whereas the concentrations of vancomycin in cortical and trabecular bones of patients undergoing TJA have been reported to be above the MIC, variability of such concentrations in the obese and morbidly obese populations have not been explored [8]. The few studies that have investigated the relation between obesity and the concentration of prophylactic antibiotics within tissues have come from analysis of gastric bypass patients [15]. When Forse et al. [8] noted an increase in the rate of SSIs in their severely obese patients after gastric bypass surgery, the researchers set out to compare antibiotic concentrations in serum and adipose tissues of their patients to those of “normal-weight” patients. As expected, they found that blood and tissue concentrations of antibiotics (cefazolin) were lower in all severely obese patients [8]. Furthermore, the concentrations of antibiotics in adipose tissue of these patients were noted to be below the MIC for most pathogens associated with SSIs. Minimum inhibitory concentration is defined as the lowest concentration of an antimicrobial that will inhibit the visible growth of an organism after plated incubation in the laboratory. Minimum inhibitory concentrations are used by diagnostic laboratories to assess the potency of a given antibiotic against a certain micro-organism. Only when the investigators doubled the dose of prophylactic antibiotics did they achieve adequate tissue concentrations. Moreover, after a change in policy and routine use of greater doses of prophylactic antibiotics, the rates of SSIs decreased from 16.5% to 5.6% over four months [15]. Edmiston et al. [16] investigated the antimicrobial activity in adipose tissues of obese patients (BMI >50). Despite the use of greater dose of antibiotics (cefazolin) as suggested by Forse et al. [15], the investigators failed to achieve therapeutic tissue concentrations in nearly 90% of patients with the high BMI patients compared with 50% of those with BMI 40–49 [16].

In our prospective pilot study, we were able to quantify the concentration of both cefazolin and vancomycin in trabecular bone during TJA and found no variability with respect to patient BMI for cefazolin, but did find variability with respect to vancomycin with lower BMI groups achieving greater trabecular bone concentrations. The average concentration of cefazolin noted in the 31 patients who received the drug as prophylaxis was 7.94 mcg/mL, which was substantially greater than the MIC of MSSA. Of the 31 patients receiving cefazolin, only three did not meet the 2 mcg/mL threshold. It is unclear why these patients were not able to achieve this value, because there was no difference in BMI or dose of cefazolin per kilogram compared with those patients who met the MIC value. Whereas laboratory error cannot be ruled out, tourniquet inflation, as well as the difficulty in the collection of trabecular bone from the thin discarded proximal tibial plateau and distal femur for patients undergoing TKA (all three patients underwent TKAs) are possible causes. Our study did show a difference between cefazolin concentrations in bone for patients undergoing TKA compared with those undergoing THA to support the latter finding. Whereas the tourniquet was inflated after antibiotics were administered in every patient, we performed an analysis of the time between bone cuts and tourniquet inflation to see if there was a correlation between tourniquet time and decreased trabecular concentration. Although a viable hypothesis, we did not observe a difference in the time separating bone cuts and tourniquet inflation between patients who met MIC target levels and those who did not. Whereas we used a target value of 2 mcg/mL, recent data show that 99.8% of MSSA in the United States are inhibited by a concentration of ≤1 mcg/mL [14]. It is also important to note that there were no post-operative infections in any of the patients who did not meet the cefazolin MIC threshold.

Regarding vancomycin, a significant difference was noted in average vancomycin concentration per kilogram of body weight with the highest values noted in patients with the lowest BMI (<25) and lowest values noted in patients with the highest BMI (>35). This difference may be attributed to the difference in doses received between the subgroups; the standard dosing mechanism of vancomycin (15 mg/kg to determine if patients are to receive 0.75 g, 1 g, 1.25g, or 1.5 g) led to patients with lower BMIs receiving in effect a greater dose per kilogram (p = 0.02). The average concentration of vancomycin in the 32 patients who received the drug was 3.42 mcg/mL, which is greater than the MIC for MRSA. Of these 32 patients, five were unable to achieve the 2 mcg/mL threshold. Similar to the patients who did achieve a MIC of cefazolin, four of the five patients who did not achieve the MIC of vancomycin had TKAs (the fifth patient had a THA and had the highest concentration of vancomycin of the five patients at 1.98 mcg/g). No complications or SSIs were noted in these patients either. It is also important to note that the three patients who did not achieve MIC for cefazolin and the five patients who did not achieve MIC for vancomycin were mutually exclusive; thus, no patient was below the MIC level for both antibiotics. This may have played a role in the absence of SSIs in this subset of patients.

Regarding the lower average trabecular bone concentrations of vancomycin (3.42 mcg/mL) compared with cefazolin (7.94 mcg/mL), this can be explained by the different pharmacokinetic properties of the two antibiotics. Vancomycin has a volume of distribution of 0.39 compared with 0.12 for cefazolin indicating the increased deposition of vancomycin in soft tissues [17]. As such, it is expected that a far smaller percentage of the injected vancomycin ends up in trabecular bone compared with what would be expected for cefazolin.

There were a few limitations to the study. The method to extract trabecular bone from discarded cuts of the tibia and femur did not yield a large amount of bone. After the rinse and processing, it is unclear how much the antibiotic concentration was affected before the specimens were stored at −80°C. Furthermore, the HPLC processing was performed at a facility different than from where the bone itself was collected. Whereas all samples were immediately stored at −80°C until shipment, were shipped in dry ice, and were maintained at −80°C upon arrival at the processing facility until they were assessed, the time lapse between these collection and analysis could have played a role in the concentration variability. Last, the concentration of trabecular bone that is quantified in this study represents an isolated concentration. For adequate protection, one would like the concentrations of antibiotic to be above that of the MIC for the duration of the surgery and immediate post-operative period. Given that the half-life of cefazolin is only 1.8 h (vancomycin is 5.6 h), further studies would be required to determine if bone concentrations of cefazolin are above the MIC level for the entire duration of a TJA procedure.

Our assessment of trabecular bone concentration of cefazolin and vancomycin during total joint replacement surgery showed no differences with respect to patient weight for cefazolin but did note variability with respect to vancomycin with lower BMI groups achieving greater trabecular bone concentrations. Using the current weight-based protocol of antibiotic prophylaxis, 84% of patients receiving vancomycin and 87% of patients receiving cefazolin achieved adequate bone antibiotic concentration above the MIC. Because there were no SSIs in our patient population, further studies may be required to determine if increasing the pre-operative dosage of antibiotics is mandated given the findings of this pilot study. Furthermore, it is imperative to determine whether bone concentrations of antibiotics is a more important indicator of effective prophylaxis against SSI compared with serum and soft tissue concentrations.