048 - Amikacin for the Treatment of Gram-negative Infections in Neonates: Patients Characteristics, Clinical Response and Pharmacokinetics

Publication Number: 48.3962

Soledad Urzua, Universidad Catolica de Chile, Santiago, Region Metropolitana, Chile; Alberto Toso, Pontificia Universidad Catolica de Chile, Santiago, Region Metropolitana, Chile; Liliana Marcela. Leguizamón, Hospital Clinico Universidad Católica de Chile, Santiago, Region Metropolitana, Chile; Andrea Maccioni, Pontificia Universidad Católica de Chile and Complejo Asistencial Dr. Sótero del Río, Santiago, Region Metropolitana, Chile; Rocio Inojosa, Complejo Asistencial Dr Sotero del Rio, Santiago, Region Metropolitana, Chile; Valeria De Toro, Pontificia Universidad Catolica de Chile, Santiago de Chile, Region Metropolitana, Chile; Nicole Le Corre, Pontificia Universidad Católica de Chile, Santiago, Region Metropolitana, Chile; Nicolas Severino, Pontificia Universidad Catolica de Chile, Santiago, Region Metropolitana, Chile

Background: Amikacin as monotherapy in the treatment of Gram-negative (GN) infections is controversial due to its complexity and adverse effects, especially in sepsis. In neonatology, the use of amikacin in the treatment of GN infections remains an effective option.

Objective: To describe the characteristics, clinical response, and pharmacokinetics of neonates with Gram-negative infections treated with amikacin as monotherapy.

Design/Methods: Prospective, descriptive, and analytical observational study conducted in 2 tertiary care hospitals’ neonatal units between 2019-2021. Patients older than 3 days with suspected GN infection were included if they started empirical treatment with amikacin. In patients with confirmed sepsis by GN, MIC for the bacterium was determined and followed up with amikacin blood levels. General epidemiologic parameters, clinical response, and pharmacokinetic parameters were recorded. Non-favorable clinical response was defined as a patient who persists with positive cultures or dies during treatment. Patients with central nervous system infections were excluded. Parametric statistical tests were conducted, including t-tests, chi-squared tests, linear regression analysis, and Pearson correlation. Significance was determined with p< 0.05. In addition, ROC curves for MIC and Cmax/MIC according to clinical response were performed.

Results: A total of 104 patients with suspected GN infection were included. GN infection was confirmed in 40 patients (38%). Table 1 describes the baseline characteristics of the study population. Table 2 shows patients with favorable and non-favorable clinical responses. Patients with a non-favorable clinical response were more associated with a positive peritoneal culture. Indeed, the only patient who died during treatment was a patient with necrotizing enterocolitis. Regarding pharmacokinetic parameters, patients with a favorable response had lower MICs and, as expected, more days on amikacin (Table 2). In addition, compared to dry weight, body real weight is a better variable in predicting amikacin peak level, with a significant positive association (r: 0.2 p = 0.01; coef. B: 0.9, p = 0.02). Moreover, Figure 1 shows ROC curves for MIC and Cmax/MIC in amikacin-treated patients with Gram-negative infection (cutoff MIC > 4; cutoff Cmax/MIC > 5.2). No alterations in renal function were observed in the study population.

Conclusion(s): Amikacin treatment of GN infections in neonates appears safe and effective when supported by MIC and pharmacokinetic parameters. A MIC < or = 4 and Cmax/MIC > 5.2 seem as reliable predictors of favorable treatment response.

TABLE 1. Baseline neonates characteristics1 (n=104).
GN: gram-negative microorganism; NEC: necrotizing enterocolitis. 1 Values are mean ± SD or % (n). 2 Differences between groups were analyzed using T-test for independent samples and Chi-square tests were used for categorical variables, statistical significance p<0.05. 3 Overall mortality, not necessary related to infection.

TABLE 2. Gram-negative infection episodes treated with amikacin with favorable and non-favorable response (n=49).
Non-favorable response was defined as a patient who persists with positive cultures or dies during treatment.
Cmax: antibiotic maximum concentration MIC: bacterial minimum inhibitory concentration.

1 Values are mean ± SD or % (n). 2 Differences between groups were analyzed using T-test for independent samples and Chi-square tests were used for categorical variables, statistical significance p<0.05.

FIGURE 1. ROC curves for MIC and Cmax/MIC in patients with gram-negative infection treated with amikacin (n=49).
(A) ROC curve for predicting non-favorable response to amikacin (AUC 0.9375), with a cutoff MIC > 4 (Sensitivity 100%, Specificity 76.1%). (B) ROC curve for prediciting favorable response to amikacin (AUC 0.9128), with a cutoff Cmax/MIC > 5.2 (Sensitivity 95.4%, Specificity 75%).