Newborns and gentamicin--how much and how often?

Aminoglycosides are widely used in the treatment of suspected or proven severe bacterial infections. Combined with β-lactam antibiotics, gentamicin provides synergic activity against the most commonly encountered pathogens in the first month of life, including the frequently isolated coagulase-negative staphylococci. The clinical use of aminoglycosides is limited, unfortunately, by their potential ototoxicity and nephrotoxicity.1 Nevertheless, their concentration-dependent bactericidal action and post-antibiotic effect, antibacterial synergy with β-lactams and low cost, and the better understanding of risk factors associated with the use of these agents, are favourable points for their continued use. The risk of aminoglycoside-induced nephrotoxicity and ototoxicity, although controversial in newborns, must still be considered a potential hazard related to drug accumulation.2–4 Until substantial evidence proves otherwise, maintaining serum gentamicin concentrations within the accepted therapeutic range in newborns continues to be prudent. In order to obtain optimal therapeutic effect and to prevent potential toxicity associated with high post-dose and pre-dose gentamicin levels, gentamicin dosing has been determined taking into consideration the newborn's weight, gestational age, post-natal age, post-conceptional age (gestational age + post-natal age), renal function [primarily glomerular filtration rate (GFR)], plasma creatinine concentration and last, but not least, the pharmacokinetics of the drug. The mean values (± s.d.) of the following pharmacokinetic parameters have been calculated for gentamicin: elimination rate constant, half-life, volume of distribution and clearance. The correlation between estimated gestational age or birth weight and these parameters has also been determined by regression analysis. Various target gentamicin concentrations based on standard practice guidelines to maximize efficacy and to prevent toxicity are:5 (i) pre-dose <2 mg/L; (ii) pre-dose <1 mg/L; (iii) pre-dose <0.5 mg/L; (iv) post-dose 4–8 mg/L and pre-dose <2 mg/L; (v) post-dose 5–10 mg/L and pre-dose <2 mg/L. A pre-dose level of <0.5 mg/L and a post-dose level of 4 mg/L may be considered to be sub-therapeutic.6 Faura et al.7 divided the newborns they studied into four groups: gestational age ≥37 weeks, gestational age <37 weeks, age ≥7 days and age <7 days. They were given a standard dose of gentamicin (5.2 ± 1.0 mg/kg/day). It was found that these groups not only had different gentamicin serum concentrations, including low post-dose levels of <4 mg/L or potentially toxic pre-dose levels of >2 mg/L, but also different pharmacokinetic parameters. There was marked variation in the pharmacokinetic parameters, such as gentamicin half-life and distribution volume. The newborn's weight, gestational age and serum creatinine were factors of importance in the variability of serum levels. Relatively safe dosage regimes, therefore, should be calculated from individual pharmacokinetic properties or by division of newborns into sub-groups. In the study conducted by Lopez-Samblas et al.,8 newborns received gentamicin 3.0 mg/kg every 24 h or 2.5 mg/kg every 18 h, depending on their weight and post-conceptional age. Although there was no significant difference in post-dose concentrations between the two groups, pre-dose concentrations of <2 mg/L differed significantly, thereby establishing that a protocol based on post-conceptional age was reliable in achieving therapeutic gentamicin serum concentrations in newborns. The authors believed that the accuracy of such a protocol is based on the fact that the proposed dosing guidelines reflect the pharmacokinetic variability resulting from normal developmental changes in renal function. Unlike other protocols, it provides a simple and accurate approach to empirical dosing, using only two different schedules that do not involve complicated calculations or inconvenient dosing intervals. On the other hand, Izquierdo et al.,9 after measuring multiple linear regressions, concluded that gestational age was the variable most predictive of gentamicin clearance in newborns. They also found that body weight had a good predictive value in estimating the apparent volume of distribution in normal term newborns, though the predictive value diminished in premature babies. Shopnik et al.10 studied 20 full-term newborns, who were randomly assigned to a once-daily or twice-daily dosage regimen with gentamicin (4 mg/kg/day). The potentially toxic serum concentration of 12 mg/L was never reached in the once-daily regimen, and post-dose and pre-dose concentrations showed no significant difference.11 Murphy & Carter12 established that estimated gestational age and birth weight were highly correlated with each other, as in previous studies. As gentamicin dosing recommendations based only on post-natal age or gestational age in premature newborns have been demonstrated to produce pre-dose concentrations >2 mg/L,6,13 the authors extrapolated a gentamicin dosing schedule for premature infants based on the work of Kasik et al.7,14 They determined the ability of a protocol based on dosing according to post-conceptional age to produce therapeutic steady-state serum gentamicin concentrations post-dose (4–10 mg/L) and pre-dose <2 mg/L,15 and compared the protocol with published dosing recommendations.16–20 The highest accuracy (89%) was achieved by this study's guidelines based on post-conceptional age. Although numerous studies have been reported concerning gentamicin pharmacokinetics in newborns,14,16,21–25 few guidelines have been updated to reflect the current literature, which shows that the development of renal function, including the elimination half-life of gentamicin, is better correlated with post-conceptional age.14,26,27 On the basis of pharmacokinetic parameters, weight, gestational or post-conceptional age, or normal or impaired renal function, several workers recommended a loading dose of between 3.5 and 5 mg/kg for all newborns followed by a once-daily dose of between 2.6 and 4 mg/kg/day.28,29 Others determined the subsequent dosing by individual pharmacokinetic analysis.30 Ohler et al.,31 on the other hand, devised their dosing schedule of 5 mg/kg every 24, 36 or 48 h based on post-conceptional age and normal or impaired renal function. They all concluded that a once-daily regimen was superior to the old ones of administering the same amount of gentamicin in divided doses two or three times per day. Higher doses gave extended dosing intervals and produced higher post-dose levels, resulting in a lower incidence of neprotoxicity, and effectively achieving better bacteriological and clinical cure rates. Brion et al.32 evaluated the relationship between gentamicin pharmacokinetics and GFR in newborn infants. They determined the appropriate interval of administration in newborns with renal insufficiency. Infants with a post-conceptional age of ≥29 weeks and a plasma creatinine concentration of ≥88.4 μmol/L had significantly greater pre-dose and post-dose gentamicin levels than those with a plasma creatinine of <88.4 μmol/L. If gentamicin is indicated in patients with renal insufficiency where the interval of administration can be estimated from the prolonged gentamicin half-life and plasma creatinine, and then the interval can be adjusted according to pre-dose and post-dose gentamicin levels. Since gentamicin is excreted almost exclusively by the kidney, its pharmacokinetics are closely linked to GFR at all ages.33–36 Nevertheless, in one published study plasma creatinine concentration was found to contribute very little to the variance of the pre-dose gentamicin concentration in newborns.37 On a schedule based on gestational age, patients with abnormal renal function (post-conceptional age ≥29 weeks and plasma creatinine ≥88.4 μmol/L) are at greater risk of excessively high gentamicin levels. Patients with a post-conceptional age of ≤28 weeks, in whom a creatinine concentration was ≥88.4 μmol/L, did not share that risk if the interval of administration was 24 h.38 However, if the interval was 18 h, high pre-dose levels were more common.39 As the pharmacokinetic parameters of gentamicin show that elimination half-lives are longer in newborns, and especially in pre-term infants, compared with older children, a once-daily dosage regimen of gentamicin assures that post-dose serum concentrations are above the MICs for relevant bacteria. Pre-dose concentrations should also be below the potentially toxic concentration of 2 mg/L. Toxicity in newborns has been associated with increased serum concentrations and durations of therapy.13 Whilst serum creatinine, age, weight, height and sex are patient variables that best predict the pharmacokinetic parameter of gentamicin in older patients, they are not so well defined in this population of patients. Many infants will be exposed to pre-dose concentrations >2 mg/L if traditional dosing approaches are used (i.e. 2.5 mg/kg every 12 h). Since the newborns in this study were <7 days old at the start of therapy, serum creatinine was not one of the patient characteristics examined.12 Results of a study by Faura et al.7 indicated that the pharmacokinetics of gentamicin in infants did not change significantly over 1 week and doses do not require adjustment more frequently than weekly as long as newborns do not exhibit signs of inability to eliminate gentamicin (e.g. decreased urine output, increased serum creatinine or urea). Of the published protocols, that of Murphy & Carter12 (3.5 mg/kg every 24 h) produced the highest percentage of pre-dose levels of <2 mg/L (91%). The corresponding figure for Lopez-Samblas' protocol was 80%. However, the Murphy & Carter protocol led to 23% of the pre-dose levels being <0.5 mg/L, the highest such percentage for the published protocols, which might represent a subtherapeutic period. The Pons protocol resulted in the highest percentage of newborns with pre-dose levels of >2 mg/L (48%).40 Faura's and Lopez-Samblas' protocols performed similarly, providing pre-dose levels of between 1 and 2 mg/L in 54% and 47% of patients, respectively. Of the published protocols, that of Murphy & Carter produced the greatest percentage of newborns with post-dose levels of between 5 and 10 mg/L, and pre-dose levels of <2 mg/L. This was confirmed by others when the performance of 15 dosing protocols was evaluated.5 Le Brun et al.41 drew attention to the reported variations in the renal toxicity of aminoglycosides in experimental animals as well as in humans, the drug being more effective and less toxic when injected during the activity period. The once-daily administration of aminoglycosides, instead of the conventional two or three times daily dosing, is actually the most attractive alternative for reducing the renal toxicity of these agents in patients. The animal model indicated that extending the interval of injection was as effective as, and less toxic than, more frequent administration of aminoglycosides.42,43 Several meta-analyses of randomized clinical trials in humans have compared the clinical efficacy, the bacteriological cure rate and the incidence of nephrotoxicity of the same total dose of aminoglycosides administered once daily.44–50 These analyses indicated that the once-daily dosing with aminoglycosides in adults is at least as effective and safe as the multiple dose regimen. These studies, however, did not take into account the time of injection of these antibiotics throughout the day, as the greatest renal toxicity has been observed when aminoglycosides were administered in the middle of the rest period of the experimental animals and in humans, while lower toxicity was found when they were treated in the middle of the activity period.47 Based on evidence available in the current literature, it is quite clear that the renal toxicity of aminoglycosides can be reduced by giving the drug as a single daily injection when patients are active.51 However, this has little relevance so far as newborns are concerned. From all the above studies, in order to standardize and simplify gentamicin therapy, it might be recommended that for newborns up to 7 days old with normal renal function the dose of gentamicin should be 3.5 mg/kg iv once daily, though this protocol has been criticized for often producing pre-dose levels of <0.5 mg/L, which are considered to be sub-therapeutic. This advice is further strengthened by recent studies, although they advocate close monitoring.52,53 No patient developed renal impairment. In addition, with this regimen, with newborns that have normal renal function receiving <72 h of therapy, monitoring of gentamicin serum levels may not be necessary. However, for longer duration of therapy in patients with compromised renal function, drug concentration measurement is essential. Regardless of the dosing regimen used, most authors recommended monitoring gentamicin concentrations.