The objective of this study was to evaluate the impact of bariatric surgery on the pharmacokinetics of amoxicillin.

Methods: This study was performed as a randomized, open-label, single-dose clinical trial, with two periods of treatment, in which obese subjects (n=8) received amoxicillin 500 mg capsule orally before and 2 months after the Roux-en-Y gastric bypass surgery. The amoxicillin plasma concentration was determined by liquid chromatography coupled to mass spectrometry (LC-MS/MS).

Study population

Eight obese volunteers of both sexes, aged between 18 and 60 years old, were selected from the obesity program of Hospital General Dr. César Cals with indication of bariatric surgery based on the criteria established by the Federal Medical Council (2015) and who were operated using the RYGB surgery technique.

All subjects signed the informed consent form. The study was approved by the institutional review board (IRB) of the Federal University of Ceará, accredited by the National Commission for Research Ethics - National Health Council, Ministry of Health and the Scientific Committee of the Hospital General Dr. César Cals. The study was performed according to the ethical principles of the Declaration of Helsinki as revised in 2013 and Good Clinical Practice Guidelines.

Study design and pharmacokinetic sampling

This study was performed as a monocentric, open-label, randomized, single-dose study with two treatment periods - before the surgery (Time 0 - T0) and 2 months after the bariatric surgery (Time 2 - T2). In each period, the obese volunteers received a single oral dose of amoxicillin 500 mg capsule (Amoxil®, GlaxoSmithKline Brazil Ltda.) with 200 mL water after an overnight fast (approximately 8 h). The subjects remained fasted for up to 3 h after taking the medication. Water and food restriction was held up to 3 h after the intake of the formulation. Standard meals were offered to all subjects. The volunteers who needed to use any medication (antihypertensive, antidiabetic and others) were instructed to use it 3 h after the intake of amoxicillin. Blood samples (6 mL) for determination of plasma concentration of amoxicillin were collected via a venous catheter into heparinized tubes at pre-dose (0 h), 0.33, 0.66, 1, 1.5, 2, 2.5, 3, 4, 6 and 8 h post-dose. The plasma samples for chromatographic analysis were obtained by centrifugation (2000 g x 12 min at 4 ºC) of the blood samples. Plasma samples were stored at -20 °C until the analysis.

Anthropometric parameters before and after surgery

At each visit (T0 and T2), the weight and height of all volunteers were measured, followed by the calculation of their body mass index (BMI) for anthropometric comparison before and after bariatric surgery. The body weight was measured using a digital scale, and the volunteers were wearing light clothes, barefoot, with empty pockets and without accessories. The height was measured using a stadiometer with 2 meters extension. The subjects were barefoot, with legs and feet parallel, weight distributed on both feet, arms relaxed at the side of the body and palms facing the body and head positioned on the Frankfurt plane.

Bariatric surgery procedure

All volunteers included in this trial were operated using the RYGB surgery technique, performed by the same surgical team, following a standardization for the alimentary loop size (1.20 m length) and the biliopancreatic loop (1 m length). The gastrojejunoanastomosis was calibrated by a Fouchet 32F probe, resulting in a diameter of 12 mm.

Analytical method

Amoxicillin plasma concentration was quantified by liquid chromatography coupled to mass spectrometry (LC-MS/MS) and a liquid-liquid extraction.

To sample preparation, 100 µL of each plasma sample, 50 µL of the internal standard (IS) cephalexin (2.5 µg/mL) and 500 µL of acetonitrile (100 %) were added into a glass tube. The samples were vortexed (30 sec) and centrifuged (2000 g x 4 min at 4 ºC). The organic phase was transferred to another glass tube, which was added of 500 µL of deionized water and vortexed (10 sec). The samples were added of 10 mL of dichloromethane (100 %), vortexed (10 sec) and centrifuged (2000 g x 4 min at 4 ºC). The organic phase was transferred to the auto-injector PCR plates.

The high pressure liquid chromatography (HPLC) system (Agilent, Germany) consisted of a LC-10AD pump, autoinjector model CTC HST PAL/ 110695 and a G1316A/DE03018295 oven. The extracted solutions (10 µL) was injected into a 150 x 4.6 mm (5 μm particles) Inertsil ODS-C18 column maintained at 65 ◦C and the mobile phase was methanol/water (30/70, v/v) + 0.1 % formic acid at a flow rate of 900 μL/min with a split ratio of 1:3. The mass spectrometer used was a Quattro Micro (Micromass, Manchester, UK) with electrospray in positive mode for MRM (Multiple Reaction Monitoring) to monitor the transitions 366.00 >348.90 and 348.00 >158.00 for amoxicillin and cephalexin, respectively.

The method validation was performed according to the United States Food and Drug Administration (FDA) bioanalytical method validation guidance [22] and the Brazilian National Sanitary Surveillance Agency (ANVISA) [23]. To evaluate the specificity of the method, eight different blank plasma lots (4 normal, 2 haemolysed and 2 lipemic) were tested for its interferences using the proposed extraction procedure and the chromatographic or spectroscopic conditions and compared with those obtained in the samples processed from the lower limit of quantification (LLOQ). The calibration curves were prepared by assaying standard plasma samples at eight concentrations of amoxicillin (20-5000 ng/mL) in triplicate and the linearity of each calibration curve was determined by plotting the peak area ratio (y) of amoxicillin/IS vs. nominal concentration of analyte. The calibration curve was constructed by weighted (1/x) least squares linear regression. The accuracy and precision of assay were evaluated by intra- and inter-assay studies. Seven aliquots of LLOQ and each quality control (QC) plasma samples (60, 850, 2500, 4000 and 12000 ng/mL) were run in three validation batches on three different days. Inter and intra-day precisions were determined as coefficient of variation, CV (%) = (SD/M) x 100 and the accuracy as the percentage relative error, RE (%) = [(E - T)/T] x 100, where M is the mean, SD is the standard deviation of M, E is the experimentally determined concentration and T is the theoretical concentration. The matrix effect experiments were performed using the ratio between amoxicillin (60 and 4000 ng/mL) and IS injected directly into the mobile phase and standard solutions added to blank plasma extracts (4 normal, 2 haemolysed and 2 lipemic). Each sample was obtained by a matrix factor normalized by IS (MFN) according to the following formula: MFN = (response of the analyte in matrix/internal standard response matrix)/(response of the analyte in solution/response of the internal standard solution). Stability QC plasma samples (60 and 4000 ng/mL) were subjected to short-term (6.0 h) room temperature, three freezes-thaw (-20 to 25 °C) cycles and 46.5 h autosampler (25 ºC) stability tests in triplicate. The stability results were compared with the nominal values.

Pharmacokinetic analysis

The pharmacokinetic parameters were obtained based on the plasma concentration vs. time curves analyzed by a non-compartmental model using the Phoenix WinNonlin® software, version 5.0 (Pharsight Corp, MoutainView, CA, USA). Values for the maximum plasma concentration (Cmax) and time to Cmax (tmax) were obtained directly from the observed individual plasma concentration vs. time curves. The area under the plasma concentration vs. time curve from time zero to the time of the last quantifiable concentration was calculated by the linear trapezoidal rule. Elimination rate constant (ke) was obtained by log-linear regression analysis, and the elimination half-life (t1/2) was calculated with the equation [t1/2=ln(2)/ke].

Statistical analysis

The quantitative variables (continuous and discrete) were initially analyzed by the Kolmogorov-Smirnov test to verify the normality of the distribution. As this requirement was verified in most cases, mean and standard deviation (SD) were calculated for descriptive statistics. Parametric tests were used for analytical statistics.

To compare the pharmacokinetic parameters of the obese volunteers obtained before (T0) and after (T2) bariatric surgery, the t-test for paired data was used. The mean difference (or mean of differences) and their respective 95% confidence interval (95% CI) were also determined. Furthermore, the ratio between the parameter values verified before and after bariatric surgery was calculated, being expressed as geometric mean and its 95% CI. The degree and sign of the linear correlation between two variables were evaluated by the Pearson correlation coefficient, accompanied by its 95% CI.

In all analyses, two-tailed tests were used, establishing the level of significance at 0.05. The GraphPad Prism software, version 5.00 (GraphPad Software, San Diego, California, USA) was used for both statistical analyses and graphics.

Inclusion Criteria:

• Volunteers of both sexes;
• Aged between 18 and 60 years old;
• Indication of bariatric surgery based on the criteria established by the Federal Medical Council (2015);
• Volunteers operated using the RYGB surgery technique.

Exclusion Criteria:

• Hypersensitivity to amoxicillin or to chemically related compounds;
• History of serious adverse reactions;
• Hospitalization for any reason during the eight weeks before the beginning of the study;
• Blood donation or other blood loss of more than 450 mL within three months prior to individual enrolment of the subject;
• Surgery of digestive tract prior to bariatric surgery;
• History or presence of liver, kidney, gastrointestinal or serious heart disease or other conditions that could interfere with the absorption, distribution, excretion or metabolism of the drug;
• Continuous use of the studied medication;
• Pregnancy and/or breastfeeding in the case of female volunteers.