Mezlocillin pharmacokinetics in patients with normal and impaired renal functions.

The pharmacokinetics of intravenous bolus doses of 1.0 g of mezlocillin were studied in 13 persons with normal and reduced renal functions. In renal failure a moderate increase was observed for the terminal serum half-life(t1/2 beta). This changed from a mean of 1.1 h at a glomerular filtration rate of 100 ml/min to 1.6 h at 10 ml/min. The difference was not statistically significant. The excretion of unchanged drug in urine during 24 h was reduced from a mean of 59.4% (range, 52 to 77) in subjects with glomerular filtration rate above 50 ml/min to 10% (range, 7.9 to 12.1) in two patients with glomerular filtration rate of 10 to 20 ml/min. The volume of distribution during the beta-phase, Vd,b, was 14% of the body weight. Much of the antibiotic was metabolized, and this proportion increased upon reduction in renal function.

The present communication reports on pharmacokinetic changes with reduction in renal function.
MATERIALS AND METHODS Antibiotic. Mezlocillin was received from Bayer Chemie AG, Leverkusen, West Germany. One dose of 1.0 g was given as a 10% aqueous solution by intravenous bolus (2-min) injection.
Patients. Studied were 13 informed volunteer patients referred to the Medical Department B for assessment of renal function (Table 1). Other antibacterial drugs were not given concomitantly. The patients were supine during the study.
Samples. Blood samples were withdrawn from a cubital vein (not same arm as used for injection) at 0, 0.5, 1, 2, 3.5, 6, 8, 12, and 24 h after medication. The zero-time sample was taken immediately after the injection was completed. Urine was collected for the periods 0 to 3, 3 to 4, 4 to 8, 8 to 12, and 12 to 24 h postprandially. A permanent urethral catheter and bladder lavage ensured complete recovery of urine for each time interval. Urine volumes were measured, and 10-ml samples were taken from each portion. Serum was separated within 1.5 h. All samples were stored at -20°C to avoid inactivation before assay.
Assay. The concentrations were determined by an agar diffusion technique, with well reservoirs filled with 0.5-ml samples and Escherichia coli as the test strain (3). Square plates (16 by 16 cm) with 16 holes and random distibution (Latin square) of standards and samples in triplicate were used. Incubation temperature was 37°C. The assay could measure down to 0.4 ug/ml. The experimental error was 3 to 7% of the observed levels.
Calculations. For pharmacokinetic evaluations, curve fitting was done with the AUTOAN 2/NONLIN program of Wagner, Sedman, and Metzler (9). For patients 5, 7, and 12, the best-curve fit was achieved by the first-order, one-compartment open model, whereas the majority of the curve fitting was best with the first-order, two-compartment open model. The values obtained from the two-compartment program were the rate constants for transfer from the central to the peripheral compartment (k,2), for transfer in the opposite direction (k2l), elimination from the central compartment (kE), and sum of intercepts for each model compartment (cj). The intercepts (A, B), hybrid rate constants (a, ,B), serum half-life during fl-phase (tlhA/), total area under the serum curve till infinity (F), and total body clearance (CIB) were calculated by the method of Wagner (9) from the values derived from AUTOAN 2/NONLIN. For this purpose, an additional program developed by the Norwegian Computer Center, Oslo, was used. The renal clearance of mezlocillin was estimated as described previously (2). Variance is indicated by standard deviation.

RESULTS
Serum concentrations. Initial concentrations ranged from 62.5 to 170 ,ug/ml, dropping to subdetectable levels after 8 to 24 h.
Urine concentrations. Urinary recovery of unchanged drug was lower with reduced than with normal renal function. Reduced urinary recovery of unchanged mezlocillin was only noted in patients with glomerular filtration rates 651  Table 2 shows the hybrid intercepts and rate constants of curves fitted to the two-compartment model. The rates of transport between compartments and distribution volumes were all within the same range regardless of renal function. The rate constant for passage from the central to the peripheral compartment, k12, was 0.996 ± 0.501 (R, 0.303 to 2.000) h-', and the constant for passage in the opposite direction, k2b, was 0.322 ± 0.504 (R, 0.050 -0.678) h-'.
The constant of the overall elimination from the body (kE), terminal serum half-life (t'lf2?), total area under the serum curves (F), overall body clearance (CIB), and renal clearance (Clmez) are shown in Table 3. It is apparent that the elimination rate is diminished only slightly in reduced renal function. The tf3,8 values for the patients with the most reduced GFR were statistically not significantly different from those with normal renal functions. The relationship between the tl/2fl and the glomerular filtration rate measured by inulin clearance is illustrated in Fig. 1    the urinary recovery in all but two subjects. These discrepancies are related to overestimates of amounts excreted. Proportionately, the CIB values are higher in relation to the plain renal clearances, Clmez, in patients with low GFR than in those with normal renal functions.

DISCUSSION'
The serum levels reached in the present study correspond to concentrations noted in previous studies (1,5,7,8). The relative volume of distribution, 14% of the body weight, was comparable to 15.2 and 14.4% reported previously.
The serum half-life of mezlocillin is only moderately increased upon reduction in renal function. We found an increase in t½f38 from a regres-sion mean of 1.1 h at GFR = 100 ml/min to 1.6 h with a GFR of 10 ml/min. Few data are available for comparison. Kosmidis et al., after administering 5.0 g intravenously, found a t/2 of 1.0 h for patients with a creatinine clearance (Cl,,) above 80 ml/min and 3.0 h for the interval 10 to 30 ml/min (6). The pharmacokinetic estimations in our study of 1.0 g and that of Kosmidis et al. of 5.0 g were not carried out by identical approaches since a computer-assisted curve fitting was not reported for the 5.0-g study. The fact that lower serum half-life values were found with renal impairment in our patients than in those studied by Kosmidis et al. may, however, also be related to the fact that we gave a 4.0-g-lower dose. This is because the rate of elimination has exhibited a slight dose dependence in subjects with normal renal function, although this problem has not been studied directly in renal function impairment (1).
Since mezlocillin and azlocilhin have a similar chemical structure and comparable pharmacokinetics in normals, it is interesting to see that the two compounds also appear similar in renal impairment. Thus, after administering 2.0 g of azlocillin, Fielder and Becker (4) observed a mean tlA of 0.9 h for Ccr = 100 ml/min and 2.3 h for 10 ml/min.
The reason for the modest impact of renal function on elimination is probably that a major portion of the drug is eliminated or inactivated by extrarenal mechanisms. In normal volunteers, only 45 to 70% of the mezlocillin dose is excreted unchanged in the urine (1).
On the basis of our results with intravenous boluses of 1.0 g of mezlocillin intravenously, modification of the dosage would appear to be necessary only with a renal insufficiency below GFR = 10 ml/min. Then serum concentrations should be determined until further experience has been accumulated. This conclusion presupposes both that the toxicity of mezlocillin parallels that of other penicillins and that mezlocillin metabolites (not determined here) do not accumulate more than the parent compound.