Tolerability and outcomes of kinetically guided therapy with gentamicin in critically ill neonates during the first week of life: an open-label, prospective study

Abstract

Background: Aminoglycosides are bactericidal antibiotics used worldwide for the treatment of serious infections in critically ill patients, including neonates. Critically ill neonates constitute a unique challenge in dosing owing to the pathologic alterations that accompany severe illness and the rapidly changing conditions of these patients.

Objectives: The main objective of this study was to analyze the kinetically guided dosage adjustment of gentamicin in neonates critically ill during the first week of life based on plasma concentrations after the first dose and to identify the impact of covariates (eg, fluid intake, body fluid retention) with respect to gestational age (GA). Tolerability of therapy was also assessed.

Methods: This 10-day, open-label, prospective study included neonates critically ill during the first week of life admitted to the neonatal intensive care unit of a children's hospital between January 2006 and July 2009. Hearing and renal assessments were conducted over a 24-month follow-up period. The patients were treated with gentamicin for suspected sepsis, proven sepsis, or pneumonia as an early sign of sepsis. The first and second doses of gentamicin 4 mg/kg were adjusted according to birth weight and GA: group 1 (GA < 34 weeks), 48-hour interdose intervals; group 2 (GA 34-38 weeks), 36 hours; and group 3 (GA > 38 weeks), 24 or 48 hours. Individual pharmacokinetic parameters were estimated after the first dose (given in 30-minute intravenous infusions) using 4 concentrations. Individual pharmacokinetic parameters were estimated by fitting the parameters of a 2-compartment model into 4 concentrations. The last 2 blood samples were taken 30 minutes before the fourth infusion (C(trough,3)) and 1 hour after its start (C(max,4)). Dosing was individualized to reach target ranges for the C(trough,3) (0.5-2.0 mg/L) and C(max,4) (6-10 mg/L) values. If needed, initial dosing was changed after the second dose by adjusting (reducing or increasing) the third and subsequent doses, or by adjusting (prolonging or shortening) the interdose intervals. C(trough,3) and C(max,4) were assessed to determine differences between predicted and assayed values. Fluid retention was registered as the difference between fluid intake and urine output at different intervals related to the first dose per kilogram of birth weight, and from the start of the first infusion (0 hour) to the day of the fourth infusion. The C(max)/minimum inhibitory concentration (MIC) ratio was determined for assessment of optimal response. Tolerability was evaluated during the 24-month follow-up period using renal sonography to screen for nephrocalcinosis and transient evoked otoacoustic emission recordings to evaluate hearing abnormalities.

Results: A total of 84 neonates (all white; 53 males, 31 females; birth weight range, 0.8-4.56 kg; GA range, 24-42 weeks) were enrolled in 3 groups: group 1, GA < 34 weeks, n = 27; group 2, GA 34-38 weeks, n = 22; and group 3, GA > 38 weeks, n = 35. The C(max) value detected 1 hour after the start of the first infusion (C(max,1)) reached the target range of 6-10 mg/L in 66 of the 84 neonates (79%). After the initial dose, C(max,1) was variable (%CV, 29%); the failure rate to reach 6 mg/L was 13%. V(d) decreased with GA (r = -0.30, P < 0.01) and achieved mean (SD) rates of 0.51 (0.10), 0.48 (0.13), and 0.40 (0.15) L/kg in groups 1, 2, and 3, respectively. Neither C(max) nor V(d) was correlated with fluid intake relative to the first infusion. Mean gentamicin clearance measured after dose 1 (0.47 [0.23], 0.66 [0.26], and 0.76 [0.32] mL/min/kg) increased with GA (r = 0.45, P < 0.001). The interdose interval was prolonged after the second and subsequent infusions in 8 of 84 neonates (10%) or by decreasing the third dose and subsequent doses in 51 neonates (61%). The target C(max,4) and C(trough,3) values occurred in 63% (22 of 35) and 83% (29 of 35) of full-term patients (GA >38 weeks), respectively. In preterm neonates, the target range for C(max,4) was reached in 11 of 27 patients (41%) in group 1 and 11 of 22 patients (50%) in group 2; for C(trough,3), the target range was reached in 25 patients (93%) in group 1 and in 16 (73%) in group 2. C(trough,3) >2 mg/L was detected in 1 full-term neonate, and gentamicin was withdrawn. Suspected fluid retention within the time period of 0 hour to the day of the fourth infusion was well correlated with actual body weight (r = 0.58, P < 0.001), but it was negatively correlated with C(max,4) (r = -0.25, P = 0.02). Thirteen of the 84 neonates (15%) had confirmed sepsis. C(max)/MIC was >12 except for 2 resistant staphylococcal infections (C(max)/MIC = 0.4); amikacin and vancomycin were substituted for gentamicin in these cases. Clinical signs and laboratory data indicative of suspected sepsis disappeared in 5 to 10 days in 68 of 71 neonates. In 1 neonate, gentamicin was withdrawn after dose 4 because of a high C(trough,3) value. In the 3 remaining neonates, C-reactive protein was decreased >10 days without changing therapy. Two neonates died, 1 of severe hypoxic-ischemic encephalopathy as a consequence of perinatal asphyxia and another of stage IV intraventricular hemorrhage. Transient renal dysfunction attributable to gentamicin was detected in 1 case. No signs of late toxicity (nephrocalcinosis) were found during the second year of follow-up. Two neonates were diagnosed with unilateral hearing loss, a secondary phenomenon of hypoxic-ischemic encephalopathy thought to be related to the severe perinatal asphyxia.

Conclusions: The initial dose of gentamicin 4 mg/kg for these critically ill premature and mature neonates with sepsis during the first week of life was high enough to reach bactericidal C(max,1) within 6-10 mg/L. C(max,1) <6 mg/L occurred in 13% of neonates. The interdose interval modified according to the recommendation resulted in C(trough) values within the target range of 0.5-2.0 mg/L in all but 2 neonates. The kinetically guided maintenance dosing of gentamicin based on plasma concentrations after the first dose should be optimized, taking into account actual body weight. (EudraCT number: 2005-002723-13).