ANTiMIcROBAL AGENTS AND CHEMOTHERAPY, Aug. 1978, p. 172-177 0066-4804/78/0014-0172$02.

00/0 Copyright e) 1978 American Society for Microbiology

Vol. 14, No. 2 Printed in U.S.A.

Pharmacokinetics of Cefaclor and Cephalexin: Dosage Nomograms for Impaired Renal Function
DANIEL A. SPYKER,' BRUCE L. THOMAS,2 MERLE A. SANDE,3 AND W. KLINE BOLTON2*

Divisions of Clinical Pharmacology,' Infectious Disease,3 and Nephrology,2 Department of Internal Medicine, University of Virginia School of Medicine, Charlottesville, Virginia 22901

Received for publication 1 March 1978

The pharmacokinetics of cefaclor and cephalexin were characterized in patients with creatinine clearances ranging from 0 to 147 ml/min. Each of 24 fasted subjects received a single 500-mg oral dose of cefaclor, and 13 of these subjects later received 500 mg of cephalexin. Serum and urine levels of the antibiotics were measured by bioassay. The serum half-lives were highly correlated with corrected creatinine clearance (cefaclor r = 0.92, cephalexin r = 0.94). Linear regression estimates of the half-life of cefaclor were 2.3 h in the anephric patient and 40 min in the patient with a corrected creatinine clearance of 100 ml/min. For cephalexin, corresponding half-lives were 15.4 h and 58 min. We present a dosage nomogram for calculating the appropriate adjustments to the loading dose based on patient weight and maintenence dose based on corrected creatinine clearance.
Cefaclor [3-chloro-7-D-(2-phenylglycinamido)3-cephem-4-carboxylic acid] is a new, orally effective cephalosporin antibiotic, similar in structure and spectrum of activity to cephalexin. It differs from cephalexin only in the substitution of a chlorine for the methyl group in the 3' position, but exhibits improved antibacterial activity against Streptococcus pneumoniae, Escherichia coli, Klebsiella pneumoniae, Haemophilus influenzae, and most enteric pathogens (9); Cefaclor appears to be well absorbed orally in humans; after a 250-mg oral dosage, cefaclor -achieves serum levels approximately one-half those of cephalexin (7). The objective of the present study was to characterize the pharmacokinetics. of orally administered cefaclor, particularly as altered by iunpaired renal function. Ir addition, we sought to develop a simple, useful dosage strategy for administering cefaclor and cephalexin to patients with various degrees of renal impairnent.
calcium, and phosphorus) were performed before the study. The endogenous Ccr was determined from a 24h urine collection before the study. Informed consent was obtained from all subjects. Administration of drugs. Each of 24 fasting subjects received a single 500-mg oral dose of cefaclor. Thirteen of the subjects also received a single 500-mg dose of cephalexin at least 2 weeks later. Blood samples were taken immediately before and at 0.5, 1, 2, 3, 4, 6, and 8 h after dosing. Antibiotic assays. Blood was collected in evacuated glass tubes without anticoagulant, and serum was obtained by centrifugation. Samples were stored in duplicate and frozen at -50C until assayed. The cephalosporin concentration of the sera was determined by an agar well diffusion technique, using Bacillus subtilis as previously described (7). The sensitivity of this assay is 0.3 Atg/ml, and complete standard curves were prepared for each assay by using human

serum. Pharmacokinetic analysis. Data for each subject were entered via a computer terminal, stored directly on disk files, and verified. Thereafter, data were manipulated only by pharmacokinetic and statistical computer programs. MATERIALS AND METHODS The drugs were given orally, and 95% absorption Subjects. The subjects were adult male volunteers was assumed for volume of distribution (Vd) analysis. ranging in age from 22 to 76 years and in weight from The data were described in terms of a linear one62 to 169 kg. Their creatinine clearances (Ccr) ranged compartment model with first-order absorption lag. from "normal" (greater than 100 ml/min) to less than An iterative least-squares method was used to find the 5 ml/min in those patients in our dialysis program four parameters- Vd, absorption rate constant (K.), (Tables 1 and 2). A complete medical history excluded elimination rate constant (K.), and absorption lag hematological or hepatic disease, or any history of (TTa)-which best described the observed serum levels sensitivity to penicillins. A hematocrit, leukocyte for each subject (12). As an initial step in the analysis, count, differential, urinalysis, and blood chemistry individual plots were generated by computer and exscreen (including blood urea nitrogen, creatinine, bili- amined by D.S. and K.B. rubin, alkaline phosphatase, glutamic oxalacetic transStatistical analysis. The Vd was examined for aminase, sodium, potassium, chloride, bicarbonate, each drug as a function of measured weight, body 172

VOL. 14,1978

CEFACLOR PHARMACOKINETICS

173

TABLE 1. Description of subjects receiving cefaclor

Subject
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
b
Age

(yr)

(cm)

Ht

wt

Estimated

(kg)
70.0 68.2 67.1 66.0 68.6 86.4 72.7 61.8 86.4 88.8 63.6 70.0 81.0 85.5 83.1

wea

SeCh
(mg/dl)

(kg)
80.2 72.2 70.9 67.4 74.9 79.8 82.2 72.1 84.6 76.8 81.4 75.3 76.8 80.1 71.9 92.4 86.9 83.8 74.9 83.8 84.0 70.0 71.5 58.4

CorCcr
0 0 0 0 8 11 20 22 25 30 34 35 38 40 52 55 59 89 102 113 122 141 142 147

Vdd
(liters)

K. (per h)
0.203 0.330 0.249 0.249 0.330 0.333 0.300 0.427 0.437 0.655 0.583 0.829 0.811 0.476 1.063 0.793 0.610 0.619 1.015 1.407 1.217 1.467 1.357 1.216

24 23 22 76 48 62 39 57 48 38 29 22 22 48 34 51 63 37 48 28 30 23 33 33

188 178 177 170 170 178 182 167 183 175 186 183 185 177 170 193 187 185 170 190 189 175 170 152

88.6 81.8 169.1 78.1 90.0

94.0

a Based on height, age, and sex SeCr, Concentration of serum creatinine. c Corrected creatinine clearance in milliliters d Asumig 95% absorption of oral dose.

77.3 72.7 84.1 (3).

13.5 14.5 14.2 7.2 10.6 7.9 5.8 3.6 4.0 2.0 2.8 2.3 2.1 2.2 1.4 1.6 1.2 1.4 1.0 1.0 1.3 0.9 0.9 1.1

16.1 29.7 19.5 15.8 23.0 36.5 16.1 21.3 38.7 44.0 22.0 26.7 26.8 25.9 33.0 41.1 16.3 54.6 24.0 29.1 24.7 15.1 13.2 20.2

per

minute per 70 kg.

TABLE 2. Descruption of subjects receiving cephalexin

Sbet Subject
1 2 6 7 14

Age Agr)e
24 23 62 39 48 29 22 22 23 33 28 30 33

Ht (cm)

Wt (kg

Etmed

imted (kg)

(rng/dl) (g/)

SeCrb

CorCcf
0 0 6 13 21 34 35 39 93 106 112 126 148

VdKd(erh K. (per h) (ltvers)

23.7 26.8 33.6 25.8 27.6 25.8 35.3 30.7 31.2 27.3 30.6 30.4 17.8

188 67.5 80.7 11.1 178 71.2 70.9 10.7 178 81.9 81.5 10.0 182 72.0 77.6 7.6 177 83.3 82.7 3.7 11 186 65.3 81.4 3.1 12 183 69.8 75.2 2.7 13 185 80.0 75.2 2.1 22 175 74.3 70.9 1.1 24 170 83.3 80.9 1.1 20 190 84.6 85.1 1.0 21 189 94.0 81.1 1.3 23 170 68.4 71.6 0.8 "Based on height, age, and sex (3). b SeCr, Concentration of serum creatinine. e Corrected creatinine clearance in milliliters per minute per 70 kg. d Assuming 95% absorption of oral dose.

0.037 0.034 0.087 0.069 0.075 0.206 0.320 0.391 0.886 0.864 0.947

0.738 0.828

surface area (4), and lean body weight (LBW) by using least-squares linear regression. We considered LBW to be the expected weight for obese subjects as suggested by Hull and Sarubbi (6). Expected weight was determined based on height, age, and sex from the Geigy table of average weights (3). LBW was thus defined as the lesser of meaured or expected weight.

Linear regression analysis was performed, and scatter plots were eamined for Vd versus height, measured weight, exected weight, and LBW. Since drug half-life and, heiice, K. depend on the ratio of the clearance of the drug and the volume of distribution, we examined the predictive value Ccr corrected for weight, body surface area, and LBW.

174

SPYKER ET AL.

ANTIMICROB. AGENTS CHEMOTHER.

For cefaclor, the average Vd was 26.4 liters (average LBW, 74.0 kg; range, 58.4 to 88.6 kg). The correlation coefficient between Vd and LBW was 0.572 (P = 0.0037). The correlation coeffiRESULTS cient between measured weight and Vd was The Vd and Ke for each patient are presented 0.718. This result for cefaclor was heavily influin Tables 1 and 2. Figure 1 presents a graphic enced by a single subject (subject number 18). If Vd is restricted to 0 for zero weight, then summary of the average serum levels stratified according to renal function. For each drug, the the best estimate is 35% of LBW, i.e., 0.35 itop panels show mean serum levels for patients ters/kg. Simnilarly, for the 13 subjects receiving with seriously compromised renal function (cor- cephalexin, the Vd was 28.2 liters, with a correrected creatinine clearance [CorCcr] less than sponding LBW of 74.7 kg, or 38% of LBW. These 25 ml/min per 70 kg). The bottom curves rep- figures are based on an assumed 95% absorption resent patients with normal renal function fraction. The Ke was found to be dependent on Ccr. (CorCcr greater than 60 ml/min per 70 kg). The absorption rate constants for cefaclor and ceph- Since drug half-life depends on the ratio of drug alexin were found to be 3.75 and 2.83/h, respec- clearance to Vd (11), we define CorCcr as Ccr x tively. Corresponding peak serum levels of 23.1 70/LBW. and 27.6 jig/ml in all patients (corrected for Figure 2 shows the least-squares linear regresweight of 70 kg) occurred at 0.88 and 1.35 h, sion of Ke on CorCcr. The best fit is Ke = 0.302 + 0.00754 x CorCcr. The corresponding correrespectively.
TABLE 3. Average pharmacokinetic parametersa
Excretion half-time

That is, linear regression and scatter plots were done for Ke versus Ccr corrected for weight, surface area, and LBW.

Absorption
delay
(h; P 0.041)
Cefaclor n = 24 Cephalexin
0.36 0.174

half-life (h; P= 0.54)

0.185 0.172

Absorption

serum

Peak level

(jug/ml per
P
70 kg;,
=

(hXP

Time to peak

Vd

(lit/kg

0.017)

P = 0.56)

Ce O
(min)
39

Standard

Ccr
0 (h) 2.3

0.077)
0.88 0.33
0.357 0.114

estimate
(h)
0.162

error of

23.1 7.7 27.6 6.4

0.49 0.191

0.245 0.309

1.35 0.83

0.377 0.056

58

15.4

0.123

n= 13 a Values indicate mean standard deviation.

I 00

I
I

U6

0or
20
10

c.r cc,%o

4
do"

Tim fm

(he)

how

do.

-)

FIG. 1. Observed serum levels of cefaclor and cephalexin stratified according to renal function. Top graph is for all subjects with CorCr <25, middle graph for subjects with Ccr of 25 to 60, and bottom for subjects with Ccr of >60 ml/min per 70 kg LBW. Bars represent mean standard error, and curve is the best least-squares fit to the means.

VOL. 14, 1978

CEFACLOR PHARMACOKINETICS

175

LINEAR REGRESSION
2
3

10
A-

The maintenance dose was chosen to replace the amount of drug lost during the dosage interval and thus depends on the loading dose and drug half-life. A line connecting the loading dose (bar 3) and CorCcr (bar 5) gives the best estimate of the correct maintenance dose (bar 4). Figure 4 represents an equivalent dosage nomogram for cephalexin. It is different only in the
CEFACLOR DOSAGE NOMOGRAM eW CeWe CCr BoY Weih
Fro Lee

i
r.
0

LaW
'am
0

SERUM
LEVEL

LOAD MANENANCa Deo Dom

4

Co

C C,

he

/ag/d

mi/mi

CC, Correted for LBW 40FIG. 2. Least-squares linear regression and scat2 ter plot of Ke versus CorCcr for 24 patients receiving s0 ~ ~ ~~~~~~~I cefaclor. o~ Center line is regression line, ~ outer lines are 95% confidence limits on the estimate. Regression equation is K. = 0.302 + 0.00754 x CorCcr, with m0 corresponding correlation coefficient of 0.917 (P .0 20
so

60

aso

4@

0.CKX)01).
o -

900

lation coefficient is 0.917 (P = 0.000001). This indicates that the half-life of cefaclor is 0.693/0.302 (ca. 2.3) h in the anephric subject and 0.693/1.056 h (ca. 0.66 h or 40 min) in a subject with normal renal function (100 ml/min per 70 kg LBW). For the 13 patients receiving cephalexin, the regression of Ke on CorCcr was Ke = 0.045 + 0.00667 x CorCcr, with a corresponding correlation coefficient of 0.944 (P = 0.00002). The expected half-life for cephalexin would thus be 58 min and 15.4 h for CorCcr of 100 and 0 ml/min per 70 kg, respectively. Total urinary recovery of active cefaclor was 62% in six subjects with CorCcr above 100 ml/min per 70 kg. Dosage nomogram. The adjustment of dose for patient size and renal function can be accomplished on an alignment chart (Fig. 3) by drawing two straight lines. The dose for each patient should be specified in terms of a desired peak serum drug level. Based on in vitro susceptibility studies, most strains of susceptible pathogens will be inhibited by concentrations of more than 4 ,g of cefaclor per ml (9). A single dose of 250 mg in a 70-kg patient would be expected to give a peak serum level of 10 ,ug/ml, 500 mg would give 20 ,ug/ml, etc. The appropriate loading dose depends only on desired peak serum level and LBW. It is calculated by locating the LBW of the patient (bar 1) and the desired peak serum level (bar 2) and extending a straight line to the loading dose (bar 3).

FIG. 3. Dosage nomogram for cefaclor. First alignment chart (left three bars) gives loading dose from patient LBW and desired serum level. Second alignment chart (right three bars) gives maintenance dose

from loading dose and CorCcr.

CEPHALEXIN DOSAGE NOMOGRAM
From

Leon Body Weigt

Ceoecod Cc,

LBW
kg

SERUM LEVEL

LOAD Dow
orn
4

MAINTENANCE De
gq6b
.-4

Cow

c,

lb

me/mI

mi/min
no loo=
*0

-a3
340
-

40
-.2

ISO
en-

30

2-:
20
- I

0me

so

ISO5
I-

10-

Go

90-

-25

so

F00
sme
O
-

FIG. 4. Dosage nomogram for cephalexin. Loadand maintenance doses, from patient LBW, target dose, and CorCcr as described in legend to fig. 3.
ing

176

SPYKER ET AL.

ANTIMICROB. AGENTS CHEMOTHER.

slightly larger Vd and anephric half-life of 15.4 h Dosage nomograms. The cefaclor nomo(versus 2.3 h for cefaclor). gram made very little adjustment for the maintenance dose, since the 6-h dosage interval was still more than three half-lives. The cephalexin DISCUSSION nomogram, in contrast, recommended considerThe present study shows that oral cefaclor able adjustment in maintenance dose to give the rapidly achieves therapeutic blood levels, as does same peak blood levels as Ccr decreased. Brocephalexin (Fig. 4) (10). Peak blood levels after gard et al. presented a table of recommended a 500-mg single oral dose ranged from 8.3 to 33.7 correction for cephalexin based on creatinine ,ug/ml, with a mean of 23.1 and standard devia- clearance and dosage interval (2). Our dosage tion of 7.7 ,ug/ml, compared with a mean of 27.6 nomograms provided a simple, rational method ,ug/ml for cephalexin. The absorption delay was for choosing the loading dose based on patient slightly less for cefaclor, and peak -serum levels weight and the maintenance dose based on occurred at 53 min, versus 81 min for cephalexin. CorCcr. The two drugs had similar kinetics in the Cefaclor exhibits a bacteriological spectrum patient with normal renal function (CorCcr of which compares favorably with that of cepha100 ml/min), with half-lives of 40 and 58 min for lexin (9). It appears to have the same low toxicity cefaclor and cephalexin, respectively. For ane- in acute and chronic studies. In this study, it was phric patients, however, the difference was much absorbed more rapidly and reached equivalent greater (2.3 versus 15.4 h, respectively). This serum levels. In addition, the large nonrenal rate means that, at a CorCcr of 100 ml/min, 29% of constant suggests that accumulation in the rencefaclor was probably being excreted by a non- ally compromised host would be much less for renal route, compared with only 6% for cepha- cefaclor than for cephalexin. Furthermore, cephlexin. This is in close agreement with the 62% alexin has been implicated in nephrotoxicity, urinary recovery in this study and our previous especially in combination with other drugs. The study in which we found 70% urinary recovery present findings suggest that cefaclor may be a for cefaclor and 96% recovery for cephalexin (7). safer antibiotic to use for selected bacterial inIn any event, the half-life of cefaclor of 2 to 3 h fections in the presence of preexisting renal imfor a CorCcr of 0 suggests that drug accumula- pairment when a cephalosporin is indicated. tion will be negligible even for a 6-h dosage ACKNOWLEDGMENTS regimen in anephric patients. The metabolism of cefaclor has been examWe thank Vickie Eglitis and Kip B. Courtney for their ined in three species of laboratory animals (13) expert technical assistance, and Cathy Stewart and Mary secretarial assistance. and in normal volunteers (7). Cefaclor is chem- Saumweber for their expert and The antibiotics cefaclor ically similar to other cephalosporins recently Eli Lilly & Co., Indianapolis, Ind.cephalexin were supplied by reviewed by Nightingale et al. (8). Cephalexin LITERATURE CITED exhibits biphasic blood levels (distribution phase) after rapid intravenous administration, 1. Bloch, R., J. J. Szwed, R. S. Sloan, and F. C. Luft 1977. Pharmacokinetics of cefaclor in normal subjects which is better described by a two-compartment and patients with chronic renal failure. Antimicrob. model such as that of Greene et al. (5). Those 12:730-731. Agents authors conclude, however, that use of the sim- 2. Brogard, Chemother.Pinget, M. Dorner, and J. LavilJ. M., M. pler one-compartment model will not produce laureix. 1975. Determination of cefalexin pharmacokinetics and dosage adjustments in relation to renal funcsignificant errors, since the dosage interval is tion. J. Clin. Pharmacol. 15:666-673. long compared with drug half-life. Documenta geigy. SciIntravenous cefaclor was not available for this 3. Diem, K., and C. Lentner. 1970. R. Geigy, Basel, Switzentific tables, 7th ed., p. 711. J. study, so early (distribution phase) kinetics were erland. not accessible. The absence of intravenous data 4. Dubois, D., and E. F. Dubois. 1916. A formula to estimate the approximate surface area if height and weight also means that only the ratio of Vd and fraction be known. Arch. Intern. Med. absorbed were measurable. Animal studies of 5. Greene, D. S., D. R. Flanagan,17:863-871. R. Quintiliani, and C. cefaclor and human studies of cephalexin show H. Nightingale. 1976. Pharmacokinetics of cephalexin: nearly complete (90 to 95%) oral absorption. In an evaluation of one- and two-compartment model pharmacokinetics. J. Clin. Pharmacol. 16:257-264. the interest of simplicity, we assumed 95% absorption and attributed all of the observed vari- 6. Hull, J. H., and F. A. Sarubbi. 1976. Gentamicin serum concentrations: pharmacokinetic predictions. Ann. Ination to Vd. This assumption does not affect tern. Med. 85:183-189. estimates of other kinetic parameters. 7. Korzeniowski, 0. M., W. M. Scheld, and M. A. Sande. 1977. Comparative pharmacology of cefaclor and cephBloch et al. found similar kinetics for cefaclor alexin. Antimicrob. Agents Chemother. 12:157-162. in chronic renal failure (1). Their results suggest 8. Nightingale, C. H., D. S. Greene, and R. Quintiliani. a half-life of 62 min for Ccr of 100 and 2.9 h for 1975. Pharmacokinetics and clinical use of cephalospoCcr of 0. rin antibiotics. J. Pharn. Sci. 64:1899-1927.

VOL. 14, 1978

9. Scheld, W. IL, 0. IL Korzeniowsk, and IL A. Sande.
1977. In vitro

CEFACLOR PHARMACOKINETICS
usceptibility studies
with cefaclor and

177

cephalexm. Antimicrob. Agents Chemother. 12: 290-292. 10. Speigt, T. KL, R. N. Brogden, and G. S. Avery. 1970. Cephalexm a review of its antibacterial, phannacological and therapeutic properties. Drus 3:9-78. 11. Spyker, D. A., and R. L Guerrant. 1977. Gentamicin dowge. Ann. Intem Med. 86:357.

12. Spyker, D. A., R. J. Ruglosk, R. L Vann, and W. IL O'Brie 1977. Pharmacokinetics of amosicillin dose dependence after intravenous, oral, and intramusular administrion. Antimicrob. Agents Chemother. 11:132-141. 13. Sullivan, H. R., S. L Due, D. L K. Kau, J. F. Quay, and W. IL Miller. 1976. Metabolism of ('4C]cefaclor, a cephalosporin antibiotic, in three species oflaboratory animals Antimicrob. Agents Chemother. 10:630-638.