Repository logo

Investigation into optimal amikacin dosing in children.

Thumbnail Image



Journal Title

Journal ISSN

Volume Title



Aminoglycoside antibacterial agents, such as amikacin, continue to play an important role in the treatment of Gram-negative infections. However, although extremely effective, they are not without potential adverse events, the most important of which being nephro- and ototoxicity. Research into factors thought to influence both the efficacy and toxicity, has challenged the rationale upon which these agents have classically been dosed. Various studies in adult patients have found that a new approach to dosing (use of single daily administration) has equal or greater efficacy or safety compared to the standard multiple daily dosing of these agents. Similar studies comparing regimens in children are few, and as yet no comparative investigation has been performed using amikacin in children (as a separate and distinct group). Additionally, in evaluating the impact of altering dose regimens, it is imperative that the documented age-related aminoglycoside pharmacokinetic alterations, be taken into account. Amikacin pharmacokinetic parameters (determined using traditional methods) have been previously published for various (usually small) groups of children. However, population parameters are not currently available for South African children . This study therefore aimed to investigate optimal amikacin dosing in children by studying: a) the comparative efficacy and toxicity of two dosing regimens, and b) the population pharmacokinetic parameters derived using one of the alternative approaches capable of utilising routine, sparse serum drug concentration time data. This investigation was conducted in the paediatric surgical and burns wards of King Edward VIII Hospital , Durban. Study patients (0.6-12 years) received amikacin either once daily (15mg/kg) or twice daily (7.5 mg/kg) by slow intravenous bolus. Concomitant medication was given as prescribed. Regimen efficacy (favourable, unfavourable or indeterminate outcome) was assessed by patient temperatures, clinical improvement and white cell counts. Clinical nephrotoxicity was evaluated by changes in serum creatinine, and renal tubular damage (investigated in a small subgroup of patients) was indicated by detection of urinary low molecular weight proteins. Ototoxicity (cochleotoxicity) was assessed by pure tone audiometry. Pertinent demographic and treatment details (amikacin concentration time data) were recorded for the population pharmacokinetic analysis. The Nonlinear Mixed Effects Model (NONMEM) programme was used to derive appropriate models describing clearance (CL) and volume of distribution (V), as well as mean values of these pharmacokinetic parameters for this population. Fifty four patients were entered into the regimen assessment. Patients in the single daily regimen (n=27) had significantly greater (p<0.05) mean (SO) peak (±0.5 hour post-dose) serum amikacin levels (37.7 (6.9) mg/L) as well as cumulative dose (91.5 (26.5) mg/kg) and duration of therapy (5.7 (1 .5) days) when compared with those of the twice daily group (19.5 (3.7) mg/L, 70.1 (26.1) mg/kg and 4.6 (1 .6) days respectively). No statistically significant differences were found between the groups in terms of outcome (18/24 and 22/25 patients in the once and twice daily dosing groups had favourable outcomes; there were no unfavourable outcomes). Pure tone audiometry (evaluated post-therapy , in 20 patients from each dosing regimen) revealed no statistically significant differences between the number of patients in the two groups with possible drug-related ototoxicity. None of the patients assessed (including an additional 14 patients with burn injury) developed clinical nephrotoxicity. Urinalysis was performed in 17 amikacin treated patients (9 and 8 from the once and twice daily dosing regimens respectively) and 9 control subjects. Low molecular weight proteinuria was absent in all of the latter patients except one, in whom pre-existing renal disease was suspected. Tubular dysfunction ascribed to possible drug effect was detected in similar numbers of patients in the two treatment groups (3 and 2 patients in the once and twice daily dosing groups respectively). In the pharmacokinetic assessment (156 serum levels from 82 patients) using a one compartment model, the final models which best described the data were as follows : CL (Uhr) = 0.271 x age(yrs) + 2.46 x body surface areatrrr'), V (L) = 7.34 x body surface areatrn") Other fixed effects tested, which did not render the data more probable, included serum creatinine measurements at the start of treatment, gender, presence of burn injury and drug regimen. Interpatient variation was 15% and 18% for CL and V respectively, with intrapatient variation or residual error of 10%. The weight adjusted population parameter estimates (95% Confidence Interval) for this group were CL =0.180 (0.175 ,0.185) Uhr/kg and V =0.293 (0.286, 0.300) Ukg, which are within the range of values published previously for other children of similar ages. The findings of this investigation , consistent with those of other similar studies, indicate that daily amikacin administration (in combination with a B-lactam), to children with normal renal function, has similar efficacy to, and no greater toxicity than multiple daily dosing. However, the role, if any, of the significantly greater cumulative dose and duration of therapy in the daily dosing group is unknown. As uncertainty remains regarding the precise duration of certain post-exposure events (and hence, the ideal duration of the interdose interval), and with the rapid drug clearance in this group of patients , future in vitro and in vivo investigations may shed even further light on the optimal dosing approach in these patients.


Thesis (M.Med.)-University of Natal, Durban, 1996.


Aminoglycosides--Therapeutic use., Antiviral agents., Infection in children--Chemotherapy., Theses--Therapeutics and medicines management.