Pharmacokinetics, Distribution, Metabolism, and Excretion of Omadacycline following a Single Intravenous or Oral Dose of 14C-Omadacycline in Rats

ABSTRACT The absorption, distribution, metabolism, and excretion (ADME) of omadacycline, a first-in-class aminomethylcycline antibiotic with a broad spectrum of activity against Gram-positive, Gram-negative, anaerobic, and atypical bacteria, were evaluated in rats. Tissue distribution was investigated by quantitative whole-body autoradiography in male Long-Evans Hooded (LEH) rats. Following an intravenous (i.v.) dose of 5 mg/kg of body weight, radioactivity widely and rapidly distributed into most tissues. The highest tissue-to-blood concentration ratios (t/b) were observed in bone mineral, thyroid gland, and Harderian gland at 24 h post-i.v. dose. There was no evidence of stable accumulation in uveal tract tissue, suggesting the absence of a stable binding interaction with melanin. Following a 90 mg/kg oral dose in LEH rats, the highest t/b were observed in bone mineral, Harderian gland, liver, spleen, and salivary gland. The plasma protein binding levels were 26% in the rat and 15% to 21% in other species. Omadacycline plasma clearance was 1.2 liters/h/kg, and its half-life was 4.6 h; the steady-state volume of distribution (Vss) was 6.89 liters/kg. Major circulating components in plasma were intact omadacycline and its epimer. Consistent with observations in human, approximately 80% of the dose was excreted into the feces as unchanged omadacycline after i.v. administration. Fecal excretion was primarily the result of biliary excretion (∼40%) and direct gastrointestinal secretion (∼30%). However, urinary excretion (∼30%) was equally prominent after i.v. dosing.

Following a single oral 90 mg/kg dose of 14 C-omadacycline administered to LEH rats, peak tissue concentrations were generally observed at 1 to 7 h postdose in most tissues, with measurable radioactivity. The radioactivity concentration in measurable tissues was higher than in blood (C max ϭ 125 ng · eq/ml), except in brain, spinal cord, and eyes (Table 2 and Table 3). At 24 h postdose, the radioactivity in blood and about two-thirds of tissues was below the lower limit of quantification (LLOQ), whereas a substantial amount of radioactivity was measured in bone mineral. The highest tissueto-blood concentration ratios (t/b Ͼ 5), calculated as values corresponding to the area under the concentration-time curve (AUC), were observed in bone mineral, Harderian gland, liver, spleen, and salivary gland (Table 3). Radioactivity exposure with a t/b of 1 to 5 was observed in bone marrow, kidney (cortex, pelvis, and medulla), thymus, heart, adrenal cortex, lung, thyroid gland, and pancreas (Table 3).
Drug-related radioactivity after a 90 mg/kg oral dose of 14 C-omadacycline showed a low distribution to the central nervous system (brain and spinal cord), since the concentrations were low (Ͻ80 ng · eq/g) and ϽLLOQ for most time points, although the tissue-to-blood concentration ratios were 0.62 (based on C max ) and ϳ0.19 (based on AUC). The tissue-toblood concentration ratios of 1.37 (based on C max ) and 0.37 (based on AUC) suggested moderate distribution of drug-related radioactivity to the testis (Table 3).
In vitro protein binding. Omadacycline was weakly bound to plasma proteins of all tested species, with no major species differences. In the omadacycline concentration range of 10 to 10,000 ng/ml, no obvious concentration dependency of plasma protein binding was found. The mean unbound protein fractions in plasma were 84.7% Ϯ 5.3% in mouse, 73.9% Ϯ 12.1% in rat, 78.8% Ϯ 7.3% in monkey, and 78.7% Ϯ 9.7% in human. Pharmacokinetics, metabolism, and excretion of omadacycline. Following a 90 mg/kg oral dose of 14 C-omadacycline, the peak radioactivity concentration in plasma (C max , 172 ng · eq/ml) was attained at between 0.25 and 2 h (Fig. 1). The peak plasma concentration of unchanged omadacycline (C max , 47.5 ng/ml) was attained at 0.5 h after oral dosing (Fig. 2), further suggesting rapid absorption ( Table 4). Absorption of omadacycline was estimated to be 2.9% based on the oral-to-i.v. ratio of the values corresponding to the dose-normalized area under the concentration/time curve at last observation (AUC last ) of radioactivity in plasma. The oral bioavailability was very low (0.23%) compared to the absorption (Table 4), suggesting significant first-pass elimination. The oral bioavailability was much lower than ϳ35% in humans (5).
Following a 5 mg/kg i.v. dose of 14 C-omadacycline, the mean total radioactivity concentration (4,110 ng · eq/ml) at 5 min postdose in blood rapidly declined to ϳ18% at 4 h and was ϳ3% at 24 h. At 4 h, plasma concentrations of 14 C-omadacycline-related radioactivity declined rapidly to ϳ12% of the initial concentration (3,170 ng · eq/ml) (Fig. 3). Systemic plasma clearance (CL; 1,200 ml/h/kg) appeared to be moderate compared to the level in the hepatic blood flow in the rat (3.3 liters/h/kg), assuming equal distributions of 14 C-omadacycline in the blood and plasma. The plasma volume of distribution at the steady state of the unchanged compound (V ss ; 6.89 liters/kg) was larger than the body water volume (0.6 liters/kg), suggesting that omadacycline was extensively distributed to tissues. a Tissue-to-blood concentration ratios (t/b) were calculated using either AUC last or C max . NA: not applicable.
The extraction recovery of 14 C-omadacycline and its metabolites in intact rat plasma was approximately 70% (i.v. dose only). The predominant radioactive components in rat plasma after a single i.v. dose were unchanged (omadacycline/C-4 epimer, 89.9% AUC). The level of M30 metabolite represented 2.6% of the AUC, and impurity-2 represented 6.3% AUC. Due to low levels of detected radioactivity in intact rat, the oral dose in plasma was not analyzed.
After the i.v. dose, the majority of radioactivity (73.4% to 85.8% mean, 80.4%) was excreted in the feces (Table 5). Following i.v. administration, approximately 30% of the dose was recovered in bile, ϳ35% in feces, and ϳ40% in urine (Table 6). Equal levels of excretion from bile and direct gastrointestinal secretion yielded about 70% to 80% fecal excretion of omadacycline in rats. Urine excretion accounted for about 28% to 38% of the dose in intact or bile duct-cannulated (BDC) rats (Table 5 and Table 6). Omadacycline/C-4 epimer together were unchanged and were the major components in bile, urine, and feces, accounting for 27.3%, 39.7%, and 30.2% of the i.v. dose and for 95.9% of the i.v. dose. Thus, the metabolism of omadacycline/C-4 epimer was limited, and omadacycline/C-4 epimer together were primarily eliminated unchanged in rats. Several minor metabolites were derived from N-demethylation and mono-oxygenation. The results suggested that omadacycline was mainly eliminated via excretion and not via metabolism in rats. After the oral dose, most (ϳ120%) radioactivity was recovered in feces, primarily due to unabsorbed material. Only trace radioactivity (0.29% of the dose) was detected in the urine sample due to poor absorption. The mass balance recovered in excreta within 168 h postdose was complete for both dose groups.
The mean radioactivity in urine accounted for ϳ30% of the i.v. dose and 0.29% of the oral dose. The major radioactivity peaks were those of omadacycline and the C-4 epimer, representing 20.6% and 4.3% of the i.v. dose. Three minor metabolites (M25, M30, and M37) each accounted for Ͻ1% the i.v. dose (Fig. 4). Because only 0.29% of the oral dose was recovered from intact rat urine, these data were not analyzed. Unchanged omadacycline and its C-4 epimer were the major components of intact rat feces at 56.7% and 84.4% of the i.v. and oral dose, respectively. Metabolites M25 and M37 were detected at 15.7% and 1.7% of the i.v. dose, respectively. The M37 metabolite also was detected in the feces at 7.0% of the oral dose.

DISCUSSION
Results from this study showed that absorption of omadacycline in the rat was rapid and bioavailability was low. The poor absorption and bioavailability may be the result of bile salt interaction with omadacycline in the small intestine; omadacycline achieved high concentrations in bile and small intestine after i.v. administration and high concentrations in small intestine after oral administration. In humans, bioavailability was estimated to be approx-  imately 35% (5). In contrast to humans and monkeys, rats lack gallbladder organs; therefore, bile flow in rats is continuous. In the presence of a relatively higher concentration of bile salt in rat small intestine, omadacycline may interact with bile salt and be entrapped in bile salt micelles (15). Consequently, the absorption may be greatly reduced in rats. Omadacycline displayed moderate clearance and a high V ss level in rats which were consistent with the pharmacokinetics (PK) parameters in healthy volunteers after i.v. administration (5,500 ml/kg) (16). After a 300-mg oral dose of omadacycline was administered to healthy subjects, the maximum concentration of drug in plasma (C max ) was 0.5 g/ml, the time to maximum concentration of drug in plasma (T max ) was 3.0 h, the half-life was 16.8 h, and the area under the concentration-time curve from 0 h to infinity (AUC 0 -∞ ) was 10.3 g · h/ml, which are substantially greater than the values seen after oral administration of omadacycline in the present study (4). Omadacycline was widely distributed into tissues and bile in rats, with tissue-toblood ratios exceeding 1 in most tissues. The highest tissue-to-blood ratios were observed in bone, Harderian gland, liver, spleen, and salivary gland. Of note, tissue levels exceeded blood levels in lung, kidney, skin, bone, and epididymis and/or testis after both oral and i.v. single-dose administration. The tissue-to-blood ratio of Ͼ2 in lung is of particular relevance given the clinical development of omadacycline for the treatment of CABP. Further, because these assessments were based on blood radioactivity concentrations, which were approximately 1.3-fold higher than those seen in plasma at 5 min after i.v. administration (Fig. 4), and because omadacycline concentrations are measured in plasma in humans, it would be expected that the human tissue-to-plasma concentration ratio for omadacycline would be higher. The lack of stable binding to uveal tract tissue suggests no stable interaction with melanin tissue.
Excretion of omadacycline, consisting of equal levels of biliary excretion and direct gastrointestinal secretion, was predominantly seen in the feces, although approximately 30% of systemic omadacycline was eliminated in the urine. A primary involvement of P-glycoprotein in omadacycline transport across Caco-2 monolayers was observed, which suggests that biliary excretion may occur via P-glycoprotein in vivo. Parent omadacycline and its C-4 epimer, which represents a tetracycline impurity (17), were the primary components in plasma following both oral and i.v. administration.  These results indicate that elimination of omadacycline occurs via excretion rather than metabolism in rats. Two inactive metabolites, M25 and M37, were recovered, but these represented 15% or less of the administered dose of omadacycline. No measurable levels of metabolites of omadacycline have been identified in humans to date, an absence which contributes to a low risk of drug-drug interactions. In contrast to omadacycline, tigecycline undergoes extensive metabolism to eight metabolites in humans such that parent tigecycline represents only 27% of total excretion over a 48-hour period (18). The M5 and M6 metabolites of tigecycline demonstrated antimicrobial activity but at a lower level than the parent compound. The impact of extensive metabolism on antimicrobial activity, pharmacokinetics, and safety and tolerability may be difficult to predict.
In summary, in rats, omadacycline is characterized by high biliary, fecal, and renal excretion, low absorption, low lipophilicity (clog P ϭ Ͻ1), and high aqueous solubility. Omadacycline distributes to tissues associated with common infectious diseases such as pneumonia and infections of the skin and urinary tract. In addition, plasma protein binding was Ͻ30%, which may have been beneficial because the free, unbound fraction of an antibiotic typically is most closely correlated with antimicrobial activity. The limited metabolism in rats is encouraging in regard to human metabolism and drug interactions. These results suggest the potential use of omadacycline in the treatment of a variety of human infections caused by susceptible bacterial pathogens.

MATERIALS AND METHODS
Chemicals. 14 C-Omadacycline was synthesized by the Isotope Laboratory of Novartis Pharmaceuticals Corporation (East Hanover, NJ, USA). The specific activities were 20 Ci/mg (i.v.) and 1.65 Ci/mg (oral). The purity of the compound was Ͼ97%. The internal standard for the analysis of unchanged omadacycline was [ 13 C 6 ], provided by Paratek Pharmaceuticals (Boston, MA, USA). The chemical structure of radiolabeled omadacycline and position of the radiolabel are shown in Fig. 5.
Tissue and body fluid distribution via QWBA. Distribution of radioactivity of omadacycline into tissues, organs, and body fluids was investigated by quantitative whole-body autoradiography (QWBA). Pigmented Long-Evans Hooded (LEH) rats were administered a single oral dose (n ϭ 8) of 14 Comadacycline at 90 mg/kg (1.92 Ci/mg) via gavage and a single i.v. dose (n ϭ 2) of 14 C-omadacycline at 5 mg/kg (20 Ci/mg) with a 30-min infusion via a jugular catheter. Animals were sacrificed 0.5, 1, 3, 7, 24, 48, and 168 h following a single oral dose and 0.083 and 24 h following a single i.v. dose.
Following sacrifice, animal carcasses were frozen and then sectioned in specimens of 40-m thickness. Sections analyzed included adrenal gland (cortex and medulla), artery, bile, blood, bone (marrow and mineral), brain, colon wall, epididymis, esophagus, eye, fat (brown and white), Harderian gland, heart, kidney (cortex, medulla, and pelvis), lachrymal gland, liver, lung, lymph node, muscle, pancreas, pituitary gland, salivary gland, seminal vesicles, skin, small intestine wall, spinal cord, spleen, stomach (nonglandular and glandular linings), testis, thymus, thyroid gland, and uveal tract. A block of 14 C-radiolabeled standards, prepared in blood and assayed by liquid scintillation counting, was sectioned in the same manner and on the same days as each rat was sectioned.
To assess the actual concentrations in calibration standards and quality control samples, replicate 50-l aliquots of spiked blood were analyzed for total radioactivity. The values (expressed in numbers of disintegrations per minute [DPM] per milliliter) obtained from the radio assay of the spiked blood samples and digital analysis of the spots on the section blocks (molecular dynamic count [MDC] per square millimeter ϫ 10 3 ) were used to generate the calibration curve used to calculate the actual 14   For analysis of metabolites, pooled (n ϭ 3 for i.v. and oral dose groups) plasma and selected pooled urine, bile, and feces samples were analyzed for unchanged omadacycline and metabolites. 14 C-Omadacycline and its metabolites in the plasma and excreta were analyzed by high-pressure liquid chromatography (HPLC) with offline radioactivity detection. Omadacycline and metabolite concentrations in rat plasma (expressed as nanograms per equivalent per milliliter) were calculated by multiplying the percent peak of a metabolite in the radiochromatogram (expressed as a fraction) by the radioactivity in plasma expressed as nanograms per equivalent per milliliter. Levels of metabolites and parent omadacycline in urine, bile, and feces were quantified as percentages of the dose. No correction for the extraction recovery was included.
Omadacycline and its internal standard were isolated from sodium heparin-pooled rat plasma samples (50 l) using a 96-well solid-phase extraction procedure. For the initial extraction steps, samples were thawed, subjected to vortex mixing for 1 min, and centrifuged for 10 min at 3,000 rpm. A 500-l volume of 2% Na 2 EDTA(aq) was added to each well in a 2.4-ml 96-deep-well collection plate. A 25-l volume of internal standard (ISTD) working solution (5 g/ml of omadacycline-d6 -water) was added to all wells except control blanks. A 25-l volume of Milli-Q water (prepared in-house) was added to control blank wells. The plate was centrifuged at 3,000 rpm for up to 1 min to ensure that all liquid was in the bottom of the wells. Volumes (50 l) of standard control (C), quality control (QC), or pooled rat plasma samples were then added to the appropriate wells. The plate was sealed with a cap mat and subjected to vortex mixing to ensure thorough mixing.
An Oasis hydrophilic-lipophilic-balanced (HLB) 96-well extraction plate (Waters, Milford, MA, USA) (10 mg) was conditioned on a Tomtec Quadra system (Hamden, CT, USA) with 400 l of methanol, followed by conditioning with 400 l of Milli-Q water. Samples were then transferred to the extraction plate and drawn through using low vacuum. The wells were washed with 400 l of Milli-Q water, followed by another wash with 400 l of methanol/water (5:95 [vol/vol]). The plate was placed on high vacuum for approximately 2 min to remove remaining water. The samples were then eluted with 400 l of methanol into a 1-ml 96-deep-well block.
The eluent was evaporated to dryness under nitrogen at 45°C and reconstituted with 100 l of Milli-Q water. The reconstituted samples were sealed and subjected to vortex mixing for 1 min and then centrifuged for 5 min at 3,000 rpm. Samples were then transferred to a Corning Costar half-height block (Tewksbury, MA, USA) using a personal pipettor for analysis on the mass spectrometer.
All pharmacokinetic parameters were calculated with the computer program WinNonlin (S3; Certara, Princeton, NJ). The highest average plasma omadacycline and radioactivity concentrations (C max ) and corresponding times (T max ) were recorded. For the i.v. dose, the first sampling time was at 0.083 h. The concentration profiles of the radioactivity of omadacycline in blood and plasma were analyzed using WinNonlin and the pharmacokinetic parameters, including the area under the concentration curve from h 0 to infinity (AUC 0 -∞ ), and the terminal half-lives were estimated by a noncompartmental analysis. Clearance (CL) and the steady-state volume of distribution (V ss ) of omadacycline were calculated using data from the i.v. dose. The fraction or percentage of the dose absorbed was calculated based on blood or plasma radioactivity data, assuming a proportional relationship between AUC and dose. For the major metabolites in plasma, C max and T max were recorded as observed. The AUC from h 0 to h 48 (AUC 0 -48 ) was calculated using the linear trapezoidal rule.