Free serum concentrations of antibiotics determined by ultrafiltration: extensive evaluation of experimental variables

Abstract

Aim: To assess the impact of experimental conditions on free serum concentrations as determined by ultrafiltration and HPLC-DAD analysis in a wide range of antibiotics.Materials & methods: Relative centrifugation force (RCF), temperature, pH and buffer were varied and the results compared with the standard protocol (phosphate buffer pH 7.4, 37°C, 1000 × g).Results: Generally, at 10,000 × g the unbound fraction (f_u) decreased with increasing molecular weight, and was lower at 22°C. In unbuffered serum, the f_u of flucloxacillin or valproic acid was increased, that of basic or amphoteric drugs considerably decreased. Comparable results were obtained using phosphate or HEPES buffer except for drugs which form metal chelate complexes.Conclusion: Maintaining a physiological pH is more important than strictly maintaining body temperature.

Keywords: HPLC-UV; anti-infectives; azole antimycotics; beta-lactams; plasma protein binding; therapeutic drug monitoring; unbound fraction; valproic acid.

Figure 1.

Autosampler stability (>20 h, 6°C) of processed samples following ultrafiltration of buffered or unbuffered serum. Dotted line: 90% stability threshold. AMP: Ampicillin; CFC: Cefiderocol; ETP: Ertapenem; MEM: Meropenem; PIP: Piperacillin.

Figure 2.

Non-specific binding (%) of the analytes to the ultrafiltration device, calculated as (A) (1 - C_UF/C_RS) × 100%, (B) (1 - C_UF/C_ret) × 100% and (C) (1 - C_{UF_HCl}/C_ret) × 100%. C_UF: Concentration in the ultrafiltrate; C_RS: Concentration in the spiked Ringer's solution; C_ret: Concentration in the retentate (upper container); C_{UF_HCl}: Concentration in the acidified ultrafiltrate; MIN: Minocycline; TGC: Tigecycline; TZD: Tedizolid.

Figure 3.

Bland-Altmann plot for comparison of f_u as determined at 22°C compared with 37°C. The dotted lines indicate relative deviation of the f_u by ±15%, which was taken as limit for acceptable imprecision or bias [20]. CFZ: Cefazolin; CLI: Clindamycin; ETP: Ertapenem; FXN: Flucloxacillin; SMX: Sulfamethoxazole; TZD: Tedizolid.

Figure 4.

Ratio of the unbound fraction as determined at 10000 × g compared with 1000 × g plotted against the molecular weight of the drugs. ETP: Ertapenem; MIN: Minocycline; TGC: Tigecycline; VAN: Vancomycin.

Figure 5.

Bland-Altmann plots for comparison of f_u in “old” serum compared with “fresh” serum, (A) without buffering and (B) with phosphate buffering prior to ultrafiltration. The dotted lines indicate relative deviation of the f_u by ±15%, which was taken as limit for acceptable imprecision or bias [20]. CLI: Clindamycin; MIN: Minocycline; MXF: Moxifloxacin; TGC: Tigecycline; VPA: Valproic acid.

Figure 6.

Bland-Altmann plot for comparison of f_u as determined in (A) phosphate buffered serum or (B) HEPES buffered serum compared with unbuffered serum. The dotted lines indicate relative deviation of the f_u by ±15%, which was taken as limit for acceptable imprecision or bias [20]. CIP: Ciprofloxacin; CFZ: Cefazolin; CLI: Clindamycin; CRO: Ceftriaxone; ETP: Ertapenem; FXN: Flucloxacillin; LVX: Levofloxacin; MIN: Minocycline; MXF: Moxifloxacin; TGC: Tigecycline; VPA: Valproic acid.

Figure 7.

Difference between the f_u (%) determined in HEPES buffered and phosphate buffered serum. Horizontal line and whiskers represent mean ± standard deviation. CIP: Ciprofloxacin; CFC: Cefiderocol; ETP: Ertapenem; LVX: Levofloxacin; MXF: Moxifloxacin; TGC: Tigecycline.