Evaluation of amoxicillin plasma and tissue levels in pediatric patients undergoing tonsillectomy

Tonsillectomy is the most common surgery performed in the pediatric and young adult populations. Although recent guidelines based on meta-analysis suggest that perioperative chemoprophylaxis plays a role in reducing bacteraemia-related post-tonsillectomy complications, there is no evidence or agreement upon which specific antibiotic, dosage or administration route should be preferred. Since few previous studies have assessed the effectiveness of prophylaxis by direct measurement of antibiotic levels both in plasma and tissue, we designed an experimental study to quantitatively evaluate amoxicillin concentrations in children ready for tonsillectomy and compare these plasma and tissue levels with the Minimal Inhibitory Concentrations (MIC) of the bacteria more commonly involved in the upper airway infections.

Thirty-three pediatric patients under 14 years of age (median 5.0, IQR 4-7, range 3-11; M:F 18:15) with recurrent tonsillitis were treated with 3 doses (established on patient's weight) of amoxicillin-clavulanic acid given orally the day before plus a further dose 2h before tonsillectomy. Amoxicillin concentrations on both homogenated tonsillar cores and plasma were measured by HPLC-UV. Bacterial epidemiology and susceptibility were derived respectively from survey data collected by Microbiology Unit and MIC according to the National Committee for Clinical Laboratory Standards (NCCLS).

Median plasma and tissue amoxicillin concentrations were respectively 4.7 microg/ml (IQR 2.1-8.0; min-max 0.4-14.3) and 1.1 microg/g (IQR 0.4-2.1; min-max 0.4-12.9), considerably below the selected target MIC of pathogens involved in the upper respiratory tract infections (S. aureus, H. influenzae, M. catarrhalis). 20 Children showed undetectable amoxicillin levels in one or both tonsils. Interestingly, 7 out of these patients (35%) had plasma concentrations higher than the target MIC (8 microg/ml). No patient displayed plasma concentrations under the limit of sensitivity of the method. Poor core-plasma and left-right core correlation was observed among patients, suggesting that fibrosis developed after recurrent tonsillitis may hamper antibiotic penetration.

Based upon direct measurement of antibiotic levels in plasma and tissue, this study suggests that a revision of the oral prophylaxis in children is required in order to reduce microbial charge in the operative field and accordingly improve the recovery after tonsillectomy.