Single-dose pharmacokinetics of imipenem-cilastatin in neonates

ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Apr. 1985, p. 511-514 Vol. 27, No. 4 0066-4804/85/040511-04$02.00/0 Copyright © 1985, American Society for Microbiology Single-Dose Pharmacokinetics of Imipenem-Cilastatin in Neonates WILLIAM C. GRUBER,'* MARCIA A. RENCH,' JOSEPH A. GARCIA-PRATS,' MORVEN S. EDWARDS,' CAROL J. AND BAKER"3 Department of Microbiology and Immunology3 and Sections of Infectious Diseases' and Neonatology,' Department of Pediatrics, Baylor College of Medicine, Houston, Texas 77030 Received 5 November 1984/Accepted 15 January 1985 The single-dose pharmacokinetics of imipenem (N-formimidoyl thienamycin), a I8-lactam antibiotic, used in combination with cilastatin, a renal dehydropeptidase I inhibitor, were evaluated in 10 neonates 1 to 8 days of age. The imipenem-cilastatin combination was given intravenously over a 15-min period at a dose of 15 or 25 mg/kg. Drug concentrations in serum, urine, and cerebrospinal fluid (when available) were determined by high-pressure liquid chromotography, and plasma disposition of the drugs was described by a two-compartment open model. The mean peak plasma levels of imipenem 30 min postinfusion were 55.4 and 27.2 ,ug1ml, and the mean t1i20 values were 2.1 and 1.8 h at doses of 25 and 15 mg/kg, respectively. The calculated volume of distribution was 0.41 liters/kg. In two patients from whom cerebrospinal fluid was obtained 1.5 h postinfusion, imipenem levels were 5.6 and 1.1 ,ug/ml at doses of 25 and 15 mg/kg, respectively, representing 10 and 4% of the 1-h serum levels. No side effects attributable to a single dose of imipenem-cilastatin were noted. imately 0.8 ml of whole blood was obtained through indwelling umbilical venous or arterial catheters at 0, 0.$, 1, 2, 4, 8, and 12 h after completion of the infusion. Urine was collected during the first 6 h after infusion, and the external collection bag was emptied at each void. No attempt was made to empty the bladder before drug administration or at the end of 6 h. CSF samples from those infants requiring lumbar puncture as a routine part of their management were saved for study. Antibiotic assay. Serum was removed after the centrifugation of clotted whole blood. CSF and serum samples were immediately stabilized with equal volumes of 50% (vol/vol) ethylene glycol-0.5 M MES (morpholineethanesulfonate) buffer (pH 6.0), quick-frozen, and stored at -70°C until assayed. Urine samples were diluted with equal volumes of 0.5 M MOPS (morpholinepropanesulfonate) and frozen at -70°C. The concentrations of imipenem and cilastatin in these specimens were determined by high-pressure liquid chromatography as described by Myers and Blumer (8). The reproducibility of this assay for imipenem and cilastatin and the temperature stability of these drugs in serum (8) and urine (J. L. Blumer, personal communications) has been established. Side effects. During the infusion period and for the next 12 h postadministration, each patient was closely monitored for clinical evidence of an adverse reaction. In addition, all patients had standard laboratory studies done before and within 5 days after imipenem-cilastatin administration. These studies included hemoglobin, hematocrit, leukocyte count and differential, platelet count, urinalysis, serum glucose, blood urea nitrogen, serum creatinine, direct and indirect bilirubin, alkaline phosphatase, serum oxalacetic transaminase, lactate dehydrogenase, and total serum protein. Pharmacokinetic analysis. For each patient, the concentrations of imipenem and cilastatin were evaluated with a two-compartment open model (2). Semilogarithmic plots of plasma concentrations of imipenem and cilastatin versus time indicated that disposition was biexponential with a rapid distribution phase which was complete in 30 min. Therefore, the data were analyzed by the equation C = Re-tf + Se-P, where R, S, a, and P are hybrid constants The currently recommended initial antimicrobial therapy for suspected bacterial sepsis in neonates includes a penicillin and an aminoglycoside. The latter class of antibiotics has the potential problems of nephrotoxicity (6), ototoxicity (7), poor meningeal penetration (5), and the in vivo development of resistant strains (5). Imipenem-cilastatin represents the combination of a P-lactam carbapenem and a renal dipeptidase inhibitor with no intrinsic antimicrobial activity. Because of its potent in vitro activity against both gram-positive and gram-negative organisms (4), imipenem (N-formimidoyl thienamycin) may be a useful single drug alternative to the penicillin-aminoglycoside combination for the treatment of neonatal sepsis and meningitis. Although imipenem has been extensively studied in adults, it has not been evaluated previously in neonates. Accordingly, the pharmacokinetics of imipenem-cilastatin given as a single intravenous dose of 15 or 25 mg/kg were assessed in 10 neonates. Drug concentrations were determined in serum, urine, and cerebrospinal fluid (CSF), when available, and patients were monitored by both clinical and laboratory parameters for possible side effects or toxicity attributable to these drugs. MATERIALS AND METHODS Subjects. Ten neonates receiving standard antimicrobial therapy for suspected or proven bacterial infection were identified, and their parents gave written, informed consent for their infants to be evaluated. Patients were excluded from the study if they were >28 days of age, weighed <1.2 kg at birth, were severely ill, or exhibited abnormal renal or liver function on the day of enrollment. Imipenem and cilastatin were supplied in standard vials by Merck Sharp & Dohme, West Point, Pa. The cilastatin concentrate was added to the intravenous bottle containing the imipenem powder, and then sterile isotonic saline was added to form a final concentration of 5 mg/ml for each drug. A single dose of 15 or 25 mg/kg (five patients each) was administered intravenously to each subject over 10 to 15 min in a line separate from that from which blood was collected for antibiotic assay. Approx* Corresponding author. 511 GRUBER ET AL. 512 ANTIMICROB. AGENTS CHEMOTHER. 100. z :- la 0 z 10. z 0 'U z uAJ 1.0 9 3 10 7 0 1 2 8 4 12 0 1 2 4 8 12 TIME (hr) FIG. 1. Serum concentrations of imipenem in 10 neonates (patient numbers at right of diagram) given a single dose of 25 mg/kg (A) or 15 mg/kg (B). 0 time indicates the end of a 15-min intravenous infusion. and t is time, using the method of residuals and linear least-squares regression of the component segments (2, 3). Final estimates of R and S were corrected for the infusion time according to the equations A = RTotI(l - e-') and B = STPI(1 - eC), where T equals the infusion time (2, 3). From A, a, B, and P the following parameters of imipenem and cilastatin disposition were calculated (2): (i) elimination half life (t1/2p) = 0.693/1; (ii) total plasma clearance (CLs) = dose/[(A/a) + (B/P)]; and (iii) volume of distribution in elimination phase (V) = CLs/p, volume of distribution at steady state (Vss) = dose{[(A/a2) + (B/p2)I[(A/a) + (B/13)]2}, and renal clearance (CLR) = [U(t2 - tj)]/(ft,2Cdt), where U02 - t1) is the amount of drug excreted in the time interval t2 - t1 and f',Cdt is the area under the curve during the same time interval. Cumulative results were calculated as means standard deviations. RESULTS Ten neonates ranging from 1 to 8 days of age and weighing from 1.34 to 4.82 kg on the day of study were evaluated. On day 3 of life, 24 h after imipenem-cilastatin administration, patient 4 showed deteriorating renal function due to an obstructive uropathy (posterior urethral valves). He was excluded from comparative analysis with the remaining patients. The characteristics and final diagnoses of individual patients are summarized in Table 1. TABLE 1. Imipenem pharmacokinetics in neonates Wto(kg)Vol of (h) distribution Final diagnosisa Wt(kg) (mg/kg) (days) Patient Patient no.no. (dagys) Age 1 2 3 3 8 2 1 6 4b 5 Mean + 6 7 8 9 10 t1/20 (m/Dkg) 4.82 2.67 2.16 4.03 1.34 24.9 25.1 25.0 25.0 25.0 GBS meningitis NPI NPI NPI NPI SD 2 2 1 3 2 1.38 2.29 2.54 3.08 2.81 15.0 15.0 15.0 15.0 14.9 NPI NPI NPI GBS sepsis NPI Mean ± SD a GBS, Group B streptococcus; NPI, no pathogen isolated. b ~~~~~~~~~~~~~~(liters/kg) CLs CLR V,, (liters/kg) (ml/kg-min) (ml/kg-min) 1.57 1.52 2.00 2.63 2.13 0.263 0.337 0.435 0.358 0.356 0.251 0.322 0.418 0.345 0.345 1.93 2.56 2.57 1.57 1.93 0.49 1.10 0.70 0.05 0.36 1.81 ± 0.31 0.348 ± 0.071 0.334 ± 0.069 2.23 ± 0.35 0.67 ± 0.33 2.58 1.78 2.02 1.95 1.90 0.395 0.481 0.376 0.441 0.403 0.352 0.470 0.354 0.421 0.327 1.76 3.10 2.15 2.61 2.45 0.09 1.06 0.67 1.09 0.74 2.05 ± 0.31 0.419 ± 0.042 0.385 ± 0.059 2.41 ± 0.50 0.73 ± 0.41 Bilateral hydronephrosis secondary to posterior urethral valves, therefore excluded from mean determination because of renal dysfunction. IMIPENEM-CILASTATIN IN NEONATES VOL. 27, 1985 513 TABLE 3. Levels of imipenem-cilastatin in serum Mean ± SD levels (~Lg/ml) in serum at doses of: Imipenem Cilastatin Time (h) 11 E 25 mg/kg z 0 P- 0 0.5 1 2 4 8 12 4 4 z uJ z 0 z 3 Ca 4 8 7 01 15 mg/kg ± 19.4 + 8.6 ± 6.4 + 4.4 ± 3.0 ± 1.2 ± 0.5 66.0 27.2 22.0 15.3 8.3 2.2 0.6 ± ± ± ± ± ± ± 25 mg/kg 28.3 3.4 2.8 2.7 1.8 1.0 101.0 69.1 62.5 54.8 41.6 24.9 14.4 ± ± ± ± ± ± ± 13.8 11.3 12.3 12.3 10.7 9.1 7.0 15 mg/kg ± 29.4 ± 3.3 ± 3.9 ± 4.1 ± 4.1 ± 3.3 ± 3.6 80.0 36.9 34.9 29.4 24.9 15.8 9.5 0.4 a Excludes patient 4 because of progressive renal dysfunction. 6 2 97.3 55.4 45.5 30.6 14.4 3.1 0.9 than for imipenem (Fig. 2). The mean renal clearance accounted for 75% (25 mg/kg) to 60% (15 mg/kg) of the systemic clearance. The individual pharmacokinetic values for cilastatin are summarized in Table 2. The mean standard deviation levels of imipenem and cilastatin in serum at each collection interval are summarized in Table 3. CSF was obtained from two patients at 1.5 h after drug administration. The imipenem levels in CSF were 5.58 ,ug/ml (25-mg/kg dose) and 1.10 ,ug/ml (15-mg/kg dose) and were approximately 10 and 4% of the corresponding 1-h serum levels. The cilastatin levels in CSF in these two patients were 1.83 and 0.753 ,ug/ml, respectively. No clinical or laboratory evidence of toxicity attributable to imipenem or cilastatin was noted in these 10 neonates. Patient 3 had neutropenia (absolute neutrophil count, 992/mm3) 2 days after the administration of imipenem-cilastatin, but this resolved within 5 days (absolute neutrophil count, 3,600/mm3), at which time the infant was doing well. However, 3 days later this infant suddenly developed hypotension and consumptive coagulopathy. Despite vigorous supportive care, he expired. A postmortem liver tissue culture grew herpes simplex virus type 2. At his initial newborn examination, patient 4 had a suprapubic mass, which was decompressed by bladder aspiration, and renal function was normal. On day 3 of life, 1 day after the administration of imipenem-cilastatin, a renal ultrasound test revealed bilateral hydronephrosis, and an intravenous pyelogram showed poor visualization of the right kidney. The level of creatinine in serum rose to 3.1 mg/dl by day 6 of life, and cutaneous ureterostomies were performed on day 9. The 9 10 ± 12 TIME (hr) FIG. 2. Serum concentrations of cilastatin in 10 neonates (patient numbers at right of diagram) given a single dose of 25 mg/kg (A) or 15 mg/kg (B). 0 time indicates the end of a 15-min intravenous infusion. Mean peak standard deviation levels of imipenem in min after infusion for doses of 25 and 15 mg/kg 8.6 and 27.2 3.4 ,ug/ml, respectively. Semilogarithmic plots of plasma concentration versus time showed a biexponential decay (Fig. 1). The mean standard deviation t4/20 values for imipenem at doses of 25 and 15 mg/kg were 2.05 0.31 and 1.81 0.31 h, respectively. Mean renal clearance constituted approximately 30% of systemic clearance but showed considerable interpatient variability. Additional pharmacokinetic parameters are listed in Table 1. The mean standard deviation t1/20 values for cilastatin doses of 25 and 15 mg/kg were longer than those for imipenem; the values were 5.09 1.32 and 6.10 1.53 h, respectively. The elimination phase showed greater variability for cilastatin serum 30 were 55.4 ± ± ± ± ± ± ± ± ± TABLE 2. Cilastatin pharmacokinetics in neonates Patient no. Dose (mg/kg) t,,2G (h) Vol of distribution 1 2 3 4 5 24.9 25.1 25.0 25.0 25.0 5.87 3.13 5.87 24.24 5.50 0.376 0.378 0.310 0.329 0.283 Meana + SD 6 7 8 9 10 Mean ± SD 5.09 15.0 15.0 15.0 15.0 14.9 + 1.32 8.83 6.70 6.70 CLR V'CLs (r mlk (ml/kg-min) (liter/kg)(ltr/g(m/gmn(mkgin 0.337 + 0.048 0.370 0.372 0.306 0.328 0.279 0.332 + 0.047 0.74 1.38 0.61 0.16 0.60 0.83 + 0.37 0.69 1.09 0.53 0.08 0.17 0.62 ± 0.38 4.42 4.64 0.370 0.405 0.359 0.312 0.392 0.357 0.401 0.351 0.306 0.356 0.53 0.70 0.62 0.81 0.98 0.15 0.44 0.42 0.58 0.62 6.10 ± 1.53 0.368 ± 0.036 0.354 ± 0.034 0.73 ± 0.17 0.44 ± 0.18 a Excludes patient 4 because of progressive renal dysfunction. 514 GRUBER ET AL. prolonged tl2p and reduced clearance of imipenem demonstrated in this patient were undoubtedly a reflection of poor renal function, and the parameters for cilastatin were even more extensively altered. DISCUSSION There is currently no single antimicrobial agent with satisfactory in vitro activity against the pathogens causing neonatal sepsis and meningitis (group B streptococcus, enterococci, coliform bacteria, Haemophilus influenzae, and Listeria monocytogenes). Consequently, combinations of antibiotics, almost always including an aminoglycoside, are recommended for initial therapy with the attendant risks of nephrotoxicity (6), ototoxicity (7), and poor penetration of the blood-brain barrier (5). Imipenem, the N-formimidoyl derivative of thienamycin, is a natural P-lactam antibiotic produced by Streptomyces cattleya (1) which has potent activity against both gram-positive and gram-negative bacteria (4). Furthermore, it appears to penetrate uninflamed meninges at levels above the 90% MICs for most susceptible bacteria. Therefore, imipenem may be a useful alternative agent for the treatment of neonatal bacterial infection, especially for meningitis. Cilastatin is a potent inhibitor of a renal dipeptidase, dehydropeptidase I. When administered concomitantly with imipenem, cilastatin increases the urinary recovery of imipenem and eliminates the nephrotoxicity induced by high doses of imipenem in the rabbit model and that caused by subacute multiple doses in the rhesus monkey model (11). Furthermore, in adults, the coadministration of imipenem and cilastatin increased the urinary recovery of imipenem from 43% or less to 72% (9). Before evaluation of the efficacy of imipenem-cilastatin in neonatal bacterial infections, its safety and pharmacokinetics must be defined. Therefore, this combination was given as a single dose to 10 neonates receiving standard antibiotic therapy for possible or proven bacterial sepsis. No side effects or toxicity attributable to imipenem were observed. The alteration of renal function in patient 4 was easily explained by the structural renal abnormality and was perhaps enhanced by the concurrent administration of an aminoglycoside. Neutropenia in patient 3 was transient and may have been related to the subsequently proved herpes simplex infection. The distribution phase of imipenem and cilastatin appears complete by 30 min as demonstrated by the initial rapid fall in drug levels in all the patients (Fig. 1). However, the characterization of the a phase is incomplete, and additional testing of samples during the first 30 min postinfusion is required to more clearly define this rapid phase. The imipenem half-life of approximately 2 h was longer than that reported for adults (1 h) (10). Freij et al. have reported similarly prolonged half-lives in neonates given doses of 10, 15, and 20 mg/kg (B. J. Freij, G. H. McCracken Jr., K. D. Olsen, and N. Threlkeld, Program Abstr. 24th Intersci. Conf. Antimicrob. Agents Chemother., abstr. no. 581, 1984). The half-life of cilastatin was approximately 2.5 to 3.2 times that of imipenem. Because of the different half-lives of these two drugs, multiple doses of imipenem-cilastatin in the neonate might lead to the accumulation of the latter. However; no accumulation of cilastatin was noted by Freij et al. (24th ICAAC) in their study of a small number of neonates. ANTIMICROB. AGENTS CHEMOTHER. Although toxicity from cilastatin has not been demonstrated in adults (9), a multiple-dose study of the imipenem-cilastatin combination is obviously necessary to exclude this possibility in the newborn infant. The measurement of renal clearance for imipenem and cilastatin was potentially inaccurate owing to limitations of the collection method, and these data should be interpreted with caution. Data from patient 4 suggest that renal failure may decrease the clearance of imipenem and cilastatin, since the half-life for both agents was prolonged. However, the accurate assessment of imipenem-cilastatin pharmacokinetics in this setting awaits additional studies. Both the peak serum levels and the CSF levels of imipenem demonstrated in our newborn patients exceeded the 90% MICs for the pathogens implicated in neonatal sepsis and meningitis by severalfold (4). These findings and the lack of toxicity of this drug combination when given to neonates as a single intravenous dose suggest that multiple-dose and efficacy studies are indicated. LITERATURE CITED 1. Albers-Schonberg, G. B., H. Arrison, 0. D. Hensens, J. Hirschfield, K. Hoogsteen, E. A. Kaczka, R. E. Rhodes, J. S. Kahan, F. M. Kahan, R. W. Ratcliffe, E. Walton, L. J. Ruswinkel, R. B. Morin, and B. G. Christensen. 1978. The structure and absolute configuration of thienamycin. J. Am. Chem. Soc. 100:6491-6499. 2. Gibaldi, M., and D. Perrier. 1975. Multicompartment models, p. 45-96. In J. Swarbrikled (ed.), Pharmacokinetics. Marcel Dekker, Inc., New York. 3. Gibaldi, M., and D. Perrier. 1975. Appendix C. Method of residuals, p. 433 444. In J. Swarbrikled (ed.), Pharmacokinetics. Marcel Dekker, Inc., New York. 4. Kropp, H., J. G. Sundelof, J. S. Kahan, F. M. Kahan, and J. Birnbaum. 1980. MK0787 (N-formimidoyl thienamycin): evaluation of in vitro and in vivo activities. Antimicrob. Agents Chemother. 17:993-1000. 5. McCracken, G. H., and S. G. Mize. 1976. A controlled study of intrathecal antibiotic therapy in gram negative enteric meningitis of infancy. Report of the Neonatal Meningitis Cooperative Study Group. J. Pediatr. 89:66-72. 6. McCracken, G. H., and N. Threlkeld. 1976. Kanamycin dosage in newborn infants. Pediatrics 89:313-314. 7. Moore, R. D., C. R. Smith, and P. S. Lietman. 1984. Risk factors for development of auditory toxicity in patients receiving aminoglycosides. J. Infect. Dis. 149:23-30. 8. Myers, C. M., and J. L. Blumer. 1984. Determination of imipenem and cilastatin in serum by high-pressure liquid chromatography. Antimicrob. Agents Chemother. 26:78-81. 9. Norrby, S. R., K. Alestig, B. BjornegArd, L. A. Burman, F. Ferber, J. L. Huber, K. H. Jones, F. M. Kahan, J. S. Kahan, H. Kropp, M. A. P. Meisinger, and J. G. Sundelof. 1983. Urinary recovery of N-formimidoyl thienamycin (MK0787) as affected by coadministration of N-formimidoyl thienamycin dehydropeptidase inhibitors. Antimicrob. Agents Chemother. 23:300-307. 10. Norrby, S. R., K. Alestig, F. Ferber, J. L. Huber, K. H. Jones, F. M. Kahan, M. A. P. Meisinger, and J. D. Rogers. 1983. Pharmacokinetics and tolerance of N-formimidoyl thienamycin (MK0787) in humans. Antimicrob. Agents Chemother. 23: 293-299. 11. Norrby, S. R., K. Alestig, F. Ferber, F. M. Kahan, J. S. Kahan, H. Kropp, and M. A. P. Meisinger. 1982. Enhanced urinary recovery of N-formimidoyl thienamycin (MK 0787) on administering an inhibitor of the renal dipeptidase responsible for antibiotic metabolism, p. 743-745. In P. Periti and C. Grassi (ed.), Current chemotherapy and immunotherapy. American Society for Microbiology, Washington, D.C.