Pharmacokinetics of gentamicin after iv infusion or iv bolus

Journal of Antimicrobial Chemotherapy (1987) 19, 225-231 Pharmacokinetics of gentamicin after iv infusion or iv bolus F. Meunier", P. Van der Auwera", H. Schmitt*, V. de Maertelaer', and J. Klastersky* "Service de Medecine, Institute Jules Bordet, Centre des Tumeurs de I'Universite Libre de Bruxelles, rue Heger Bordet 1, 1000 Brussels; b Brussels; c Institut de Recherches Interdisdplinaires en Biologie humaine et nucleaire; Hopital Erasme, Universite Libre de Bruxelles, Brussels, Belgium A cross-over randomized comparative study was undertaken in ten healthy subjects to study the pharmacokinetics of 80 mg of gentamicin given as a bolus iv injection or as an infusion over 15 min using a commercially available 100 ml solution. Mean peak values at the end of the two modes of injection were respectively: 17-8 and 7-8mg/1 (P< 0002, Wilcoxon ranked pairs test). Mean peak values extrapolated at r = 0 for each volunteer were respectively 4-2. and 3-8mg/1 (not statistically different). Mean serum values 1 h after administration were respectively 4-2 and 3-7 mg/1 (not statistically different). Mean trough values after 8 h were: 0-45 and 0-44 mg/1. Mean elimination half-lives were 139 and 142 min (P > 01). Introduction Gentamicin is an aminoglycoside antibiotic used frequently in the treatment of infections caused by aerobic Gram-negative bacteria. Methods of administration of aminoglycosides vary widely and the data concerning the feasibility, practicability and potential toxicity of various procedures remain controversial. There is a trend now towards diluting the dose in 100-200 ml of saline or dextrose solution and then infusing the drug into the patient over a 30 min-2 h period (Dahlgren, Anderson & Hewitt, 1975; Hull & Sarubbi, 1976; Mendelson et al., 1976; Schentag et al., 1977; Zaske, Cipolle & Strate, 1980; Laskin et al, 1983). On the other hand, a slow (Bailey & Lynn, 1974; Michel et al., 1974; Stratford, Dixson & Cobcroft, 1974; Riff & Moreschi, 1977) or rapid (Korner, 1973; Dobbs & Mawer, 1976) iv bolus injection has also been sometimes advocated. Such a wide range of infusion times will be reflected in varying peak plasma levels (Errick et al., 1982) and may result in a decrease in efficacy or in the production of toxic effects. A recent survey of the administration of 80 mg gentamicin in 100 ml dextrose solution, has shown that infusion times ranged from 5 to 110 min (mean 40 min (Armitstead & Nahata, 1983). Another study showed that bolus injections are frequently given much more rapidly than recommended; while a 3-5 min injection time was advised, more than half of the injections were done in less than 2 min (lies & Newman, 1975). A similar uncertainty exists about the precise timing of peak serum gentamicin measurement and this value is often used to adjust dosage to individual patient requirements. It is generally advised to measure peak concentrations in the serum 1 h after im injection (Jackson & Riff, 1971; Kaye, Levinson & Labovitz, 1974; Noone 225 0305-7453/87/020225 + 07 $02.00/0 1987 The British Society for Antimicrobial Chemotherapy

226 F. Meunier et al. et al, 1974; Schentag et al, 1977). For iv administration there is no clear guideline and 'peaks' have been measured from 0 to 60 min after infusion (Noone et al, 1974; Dahlgren et al, 1975; Hull & Sarubbi, 1976; Mendelson et al, 1976; Schentag et al, 1977; Zaske et al, 1980; Armitstead & Nahata, 1983; Laskin et al, 1983; Moore, Smith & Lietman, 1984a, b) and from 1 to 15 min after bolus injection (Korner, 1973; Bailey & Lynn, 1974; Michel et al, 1974; Stratford et al, 1974; Dobbs & Mawer, 1976; Riff & Moreschi, 1977). Peak levels are thus difficult to interpret and misleading unless accurate and consistent administration is combined with careful timing of blood sampling. However, the monitoring of serum levels seems to be mandatory since a favourable outcome of Gram-negative septicaemia or pneumonia may be related to adequate levels of aminoglycosides, particularly for granulocytopenic patients (Moore et al, 1984a, b). The recent availability of an 80 mg dose of gentamicin premixed in 100 ml of normal saline might provide a convenient and safe gentamicin solution ready for iv administration. We decided to standardize the infusion time of this solution to 15 min, a practical length of time for nursing staff and one that is generally used in our unit. The pharmacokinetics of gentamicin by this method was compared to iv bolus in ten volunteers in a randomized cross-over study. Materials and methods Ten healthy volunteers, from 22 to 34 years of age participated in the study. There were five men and five women. Their weights ranged from 55 to 85 kg (mean 64-2 kg). Their heights ranged from 1-68 to l-86m (mean 1-75 m). Written informed consent was obtained from each volunteer. Each received a standard dose of 80 mg gentamicin. The dosage ranged from 0-94 to l-45mg/kg with a mean of 1-27 + 019 (s.d.) mg/kg. Subjects were randomized to receive on day 1 80 mg of gentamicin either prediluted in 100 ml normal saline (Travenol, Belgium) infused iv over 15 min or in 2 ml (Essex, Belgium) injected as an iv bolus in less than 1 min. Subjects were crossed over on day 3 to receive the alternate mode of administration (bolus or infusion). Venous blood was drawn from the arm contralateral to the infusion site. Samples were obtained before infusion, 7-5 min after the beginning of the infusion, and at 2, 5, 10, 30 min, 1, 2, 4, 6, 8 h after completion of the infusion or after iv bolus injection. The samples were analysed in duplicate by fluorescence polarization immunoassay (Abbott, Belgium). Pharmacokinetic analysis was carried out for each volunteer, from the data describing the logarithm of the gentamicin serum levels versus time. A one compartment model was assumed for the linear part of the corresponding curve, i.e. between 2 and 6 h (before 2 h, the drug was not completely distributed; from 8 h, accumulation of gentamicin in a second tissue compartment slows down the elimination rate). Least squares regression analysis was performed on that part of each curve (Sawchuk & Zaske, 1976; Sawchuk et al, 1977; Zaske et al, 1980), and the halflife was determined from the calculated slope. Individual calculated peak (at / = 0), trough values (at t = 8 h) and half-lives of gentamicin were then averaged to calculate the mean peak values and half-lives as shown in Figure 2. The peak values and the

Pharmacokinetics of gentamicin modes of administration 227 0 I 2 8 Time ( h ) Figure 1. Pharmacokinetics in ten volunteers of 80 mg gentamicin by bolus injection ( ) or 15 min infusion half-lives were compared between bolus injection and infusion with the use of the distribution-free signed rank test of Wilcoxon (Hollander & Wolfe, 1973). Results The pharmacokinetics of 80 mg gentamicin injected by iv bolus in 1 min or by a 15 min infusion are shown in Figures 1 and 2. For the ten volunteers receiving bolus injections, the mean peak value was 17-8 (s.d.: 4.6) mg/1 at the end of injection. In contrast those volunteers receiving the drug by infusion had peak values below 9-2mg/1 with a mean of 7-8 (s.d.: 0-7) mg/1 at the end of infusion (P < 0002). The mean measured trough value at / = 8 h was respectively 0-45 (s.d.: 014) mg/1 and 0-44 (s.d.: 013) mg/1 for iv bolus injection and iv infusion. In the /?-phase, equilibrium is reached between the blood and tissue compartments.

228 F. Meunier et al. Figure 2. Pharmacokinetic analysis in the equilibrium phase of 80 mg gentamicin administered by bolus injection ( ) or IS min infusion ( ) (semi-logarithmic scale). Extrapolation from concentration-time data, at / = 0, gives a calculated peak serum level of gentamicin (Sawchuk & Zaske, 1976; Sawchuk et al., 1977; Zaske, et al., 1980) As illustrated in Figure 2, the mean peak values extrapolated at / = 0 for each volunteer, is for bolus injection: 4-2 (S.D.: 0-9) mg/1 and for infusion 3-8 (S.D.: 0-9) mg/1. These values are not statistically different. The serum value measured 1 h after infusion was completed (3-5 mg/1) is close to the extrapolated peak value and could be used for practical adjustment of dosage regimens to individual patient needs. The half-lives calculated from the /3-phases of iv bolus and infusion were not statistically different. Their mean values were respectively 139 + 21 min and 142+17 min. Discussion The relation between the peak or trough serum concentrations and oto- or nephrotoxicity is highly controversial. Older studies have suggested that ototoxicity occurred more frequently when peak concentrations were consistently above 10-12 mg/1 (Barza & Lauermann, 1978; Jackson & Arcieri, 1971) and nephroxicity when trough levels were above 2 mg/1 (Barza & Lauermann, 1978; Dahlgren et al., 1975). Clinically inapparent nephrotoxicity is a constant feature of most

Pharmacokinetics of gentamicin modes of administration 229 aminoglycosides given at usual dosage and occurs early in the course of therapy; it consists of lysosomal phospholipidosis at the level of the proximal tubule of the kidney, the magnitude of which depends on the aminoglycoside (De Broe et al., 1984). A retrospective study of nephrotoxicity in patients treated with an aminoglycoside for serious infection concluded that neither clinical parameters nor the determination of serum concentrations were able to predict or prevent aminoglycoside nephrotoxicity (Schentag, Cerra & Plaut, 1982). Similarly to nephrotoxicity, clinically inappafent ototoxicity was demonstrable using brainstem auditory evoked potentials in patients receiving tobramycin (Wilson & Ramsden, 1977) or gentamicin (Guerit et al., 1981). It was suggested that the alterations were more likely to occur when gentamicin serum levels were 8-11 mg/1, however most of the observed alterations were reversible. Two recent prospective studies concluded that neither the dose (mg/kg) nor the serum levels were significant parameters in predicting ototoxicity (Fee, 1980; Moore et al., 1984c); they also concluded that ototoxicity was independent of nephrotoxicity. The specific toxic effects of aminoglycosides on the inner ear are related to their penetration into the inner ear fluid and ability to cause hair cell damage. The rate of entry is slow, but the elimination rate is even slower. In one study, intervals between doses of less than four half-lives led to significant accumulation in the perilymph until equilibrium (Manuel, Tran Ba Huy & Meulernans, 1982). Stupp et al (1973) have suggested the existence of a toxic threshold in the inner ear. Beyond a certain dose the kinetics of aminoglycosides were modified and the concentration increased suddenly in the perilymph. Other authors found a linear correlation between the dose and perilymph concentrations (Manuel et al., 1982; Fox et al., 1980; Desjardins-Giasson & Beaubien, 1984) suggesting that ototoxicity could be better correlated with the duration of treatment than with the dose. This is in good agreement with two prospective clinical studies (Fee, 1980; Moore et al., 1984c). A peak concentration of 5 mg/1 at steady state is required for therapeutic effectiveness in Gram-negative bacteraemias (Jackson & Riff, 1971; Moore et al., 1984a; Noone et al., 1974). However, higher values are needed in case of Gramnegative pneumonia (Noone et al., 1974; Moore et al., 19846). In these studies, peak levels were measured either 1 h after iv infusion (Moore et al., 1984a, b); 1 h after im injection (Jackson & Riff, 1971), or 1 h after im and 15 min after iv injection (Noone et al., 1974). More recently, there has been a suggestion that a single high daily dose of aminoglycoside could be more effective by achieving high serum concentration. However, further evaluation of such a regimfch m large clinical trials is needed to establish the optimal mode of administration of aminoglycosides. As methods of gentamicin administration and timing of peak measurement vary widely in clinical practice, peak values are not always directly comparable and controversy remains as to the most effective and safe way of administration of gentamicin. A reliable method to individualize dosage regimens from peak levels at steady state requires pharmacokinetic calculations from three blood samples taken at various times (Zaske et al., 1980). This method may hot be practical in many hospital settings. In our study, infusion of gentamicin over 15 min gave a mean peak serum level of 7-8 mg/1. In another study comparing three different modalities of administration of sisomicin, we also demonstrated a higher peak value after a bolus injection than after im injection or an infusion (Meunier-Carpentier> Slaquet & Klastersky, 1976).

230 F. Meunier et al. There was no significant difference for the half-life of gentamicin and its calculated peak levels at / = 0 between iv bolus or iv infusion, using the pharmacokinetic analysis developed by Zaske and colleagues (Zaske et al., 1980; Sawchuk & Zaske, 1977; Sawchuk, 1977). This shows that the method of administration does not alter the pharmacokinetics and the peak serum levels at equilibrium. The mean peak values at t = 0 were about 4 mg/1 and would increase after steady state is reached, after multiple doses of gentamicin (Schentag et al., 1977). It must also be noted that a short 15min infusion probably avoids the large variations in peak serum values observed when a longer infusion time (1 h) is used in patients with different gentamicin half-lives (Zaske et al., 1980). We think that infusion of gentamicin over 15 min provides a convenient way to reach effective serum levels of the drug. It avoids the large variations resulting from wide differences in the length of infusion. However, the 15 min infusion is inevitably more expensive than direct injection. Acknowledgements We thank Dr F. Gibson, Dr R. Van Essche and Ms Hastier for helpful discussions. References Armitstead, J. A. & Nahata, M. C. (1983). Effects of variables associated with intermittent gentamicin infusion on pharmacokinetic predictions. Clinical Pharmacokinetics 2, 153-6. Bailey, R. R. & Lynn, K. L. (1974). Serum levels of gentamicin after intravenous bolus injection. Lancet i, 730. Barza, M. & Lauermann, M. (1978). Why monitor serum levels of gentamicin. Clinical Pharmacokinetics 3, 202-15. Dahlgren, J. G., Anderson, E. T. & Hewitt, W. L. (1975). Gentamicin blood levels, guide to nephrotoxicity. Antimicrobial Agents and Chemotherapy 10, 58-62. De Broe, M. E., Paulus, G. J., Verpooten, G. A., Roels, F., Buyssens, N., Wedeen, R., et al. (1984). Early effects of gentamicin, tobramycin and amikacin on the human kidney. Kidney International 25, 643-52. Desjardins-Giasson, S. & Beaubien, A. R. (1984). Correlation of amikacin concentrations in perilymph and plasma of continuously infused guinea pigs. Antimicrobial Agents and Chemotherapy 26, 87-90. Dobbs, S. M. & Mawer, G. E. (1976). Intravenous injection of gentamicin and tobramycin without impairment of hearing. Journal of Infectious Diseases 134, SI 14-7. Errick, J. T., Torre, M. S., Coleman, J. B. & Hryciuk-Flaska, L. (1982). Pharmacokinetic considerations of variation in intermittent infusion-time duration. Clinical Pharmacokinetics 1, 61-2. Fee, W. E. (1980). Aminoglycoside ototoxicity in the human. The Laryngoscope XC, Supp. 24,1-19. Fox, K. E., Brummett, R. E., Brown, R. & Himes, D. (1980). A comparative study of the ototoxicity of gentamicin and gentamicin CIA. Archives of Otolaryngology 106, 44-9. Guerit, J. M., Mahieu, P., Houben-Giurgea, S. & Herbay, S. (1981). The influence of ototoxic drugs on brainstem auditory evoked potentials in man. Archives of Oto-Rhino-Laryngology 233, 189-99. Hollander, M. & Wolfe, D. A. (1973). Non Parametric Statistical Methods, pp. 26-33. Wiley & Sons, New York. Hull, J. H. & Sambbi, F. A. Jr. (1976). Gentamicin serum concentrations: Pharmacokinetic predictions. Annals of Internal Medicine 85, 183-9. lies, J. E. M. & Newman, M. S. (1975). Infusion therapy. Problems encountered by nurses. Nursing Times 15, 769. Jackson, G. G. & Arcieri, G. (1971). Ototoxicity of gentamicin in man, a survey and controlled analysis of clinical experience in the United States. Journal of Infectious Diseases 124, SI 30-7.

Pharmacokinetics of gentamicin modes of administration 231 Jackson, G. G. & Riff, L. J. (1971). Pseudomonas bacteremia. Pharmacologic and other bases for failure of treatment with gentamicin. Journal of Infectious Diseases 124, S185 9.1. Kaye, D., Levison, M. E. & Labovitz, E. D. (1974). The unpredictability of serum concentrations of gentamicin: Pharmacokinetics of gentamicin in patients with normal and abnormal renal function. Journal of Infectious Diseases 130, 150-4. Korner, B. (1973). Gentamicin therapy administered by intermittent intravenous injections. Ada Pathologica, Microbiologica, et Immunologica Scandinavia, Sect. B 81, Suppl. 241, 15-22. Laskin, O. L., Longstreth, J. A., Smith, C. R. & Lietman P. S. (1983). Netilmicin and gentamicin multidose kinetics in normal subjects. Clinical Pharmacology and Therapeutics 34, 644-50. Manuel, C, Tran Ba Huy, P. & Meulemans, A. (1982). Etude de la pharmacocinetique des aminosides dans la perilymphe et l'endolymphe chez l'animal d'experience. In Nephrotoxicite Ototoxicite Medicamenteuses (Fillastre, J. P., Ed.), pp. 407-33. INSERM, Vol. 102. Universite de Rouen, France. Mendelson, J., Portnoy, J., Dick, V. & Black, M. (1976). Safety of the bolus administration of gentamicin. Antimicrobial Agents and Chemotherapy 9, 633-8. Meunier-Carpentier, F., Staquet, M. & Klastersky, J. (1976). Comparative study of three routes of administration of sisomicin. Journal of Clinical Pharmacology 16, 625-30. Michel, J., Sacks, T., Stessman, J. & Licht, A. (1974). Serum gentamicin levels after intravenous bolus injection. Lancet ii, 525-6. Moore, R. D., Smith, C. R. & Lietman, P. S. (1984a). The association of aminoglycoside plasma levels with mortality in patients with gram negative bacteremia. Journal of Infectious Diseases 149, 443-8. Moore, R. D., Smith, C. A. & Lietman, P. S. (19846). Association of aminoglycoside plasma levels with therapeutic outcome in gram negative pneumonia. American Journal of Medicine 11, 657-62. Moore, R. D., Smith, C. R. & Lietman, P. S. (1984c). Risk factors for the development of auditory toxicity in patients receiving aminoglycosides. Journal of Infectious Diseases 149, 23-30. Noone, P., Parsons, T. M. C, Pattison, J. R., Slack, R. C. B., Garfield-Davies, D. & Hughes, K. (1974). Experience in monitoring gentamicin therapy during treatment of serious gram negative sepsis. British Medical Journal i, 477-81. Riff, L. J. & Moreschi, G. (1977). Netilmicin and gentamicin: comparative pharmacology in humans. Anitimicrobial Agents and Chemotherapy 11, 609-14. Sawchuk, R. J. & Zaske, D. E. (1976). Pharmacokinetics of dosing regimens which utilize multiple intravenous infusions: gentamicin in burn patients. Journal of Pharmacokinetics and Biopharmaceutics 4, 183-95. Sawchuk, R. J., Zaske, D. E., Cipolle, R. J., Wargin, W. A. & Strate, R. G. (1977). Kinetic model for gentamicin dosing with the use of individual patient parameters. Clinical Pharmacology and Therapeutics 21, 362-9. Schentag, J. J., Cerra, F. B. & Plaut, M. E. (1982). Clinical and pharmacokinetic characteristics of aminoglycoside nephrotoxicity in 201 critically ill patients. Antimicrobial Agents and Chemotherapy 21, 721-6. Schentag, J. J., Jusko, W. J., Vance, J. W., Cumbo, T. J., Abrutyn, E., DeLattre M. et al. (1977). Gentamicin disposition and tissue accumulation on multiple dosing. Journal of Pharmacokinetics and Biopharmaceutics 5, 559-77. Stratford, B. C, Dixson, S. & Cobcroft, A. J. (1974). Serum levels of gentamicin and tobramycin after slow iv bolus. Lancet i, 378-9. Stupp, H., Kupper, K., Lagler, F., Sous, H. & Quante, M. (1973). Inner ear concentrations and ototoxicity of different antibiotics in local and systemic application. Audiology 12, 350-63. Wilson, P. & Ramsden, R. T. (1977). Immediate effects of tobramycin dn human cochlea and correlation with serum tobramycin levels. British Medical Journal i, 259-61. Zaske, D. E., Cipolle, R. J. & Strate, R. G. (1980). Gentamicin dosage requirements: wide interpatient variations in 242 surgery patients with normal renal function. Surgery 87, 164-9. (Manuscript accepted 24 June 1986)