A realistic ex vivo model, the isolated perfused rat lung (IPRL), was used to investigate tobramycin’s pulmonary disposition at typical therapeutic concentrations. Different nominal doses were administered in aqueous solution to the airways alongside nonbinding absorption markers, fluorescein and mannitol. The mean fraction of each administered dose reaching the perfusate (Fp) was determined as a function of time following administration. Dynamic dialysis was also used to quantify the kinetics of tobramycin binding and/or tissue retention in the IPRL immediately after drug administration. Whereas the absorption markers fluorescein and mannitol both showed monoexponential dose-independent increases in Fp with time, tobramycin’s pulmonary absorption into the perfusate was biexponential and dose-dependent due to tissue binding or retention. Best estimates for the first-order rate constants of tobramycin absorption appeared dose-independent (0.065–0.070 min−1), with values close to the mean for fluorescein (0.076 min−1). The rate constant for dissociation from IPRL tissue was also relatively constant (0.018–0.022 min−1), whereas that for association decreased from 0.16 to 0.07 min−1 with increasing airway dose from 0.002 to 2 mg. Dynamic dialysis data from sliced IPRL tissue following identical airway administration were consistent with those from the intact IPRL, confirming tobramycin’s “slow on, slow off” binding and sequestration by the rat lung. Overall, tobramycin absorption was fast following airway administration. However, dose- and concentration-dependent slow-onset tissue binding extended the duration of tobramycin’s presence in the rat lung. These findings may explain, in part, the apparent success of inhaled tobramycin therapy when treating pulmonary infections.
Administration of tobramycin directly to the airways via a nebulizer has brought great improvements to the health and well-being of cystic fibrosis patients and others suffering from chronic pulmonary infections with pathogens such as Pseudomonas aeruginosa (Chuchalin et al., 2009).The efficacy and safety brought about by topical use of this drug is well recognized in clinical circles, although it remains unclear whether the antibiotic’s physicochemical properties lead to slow absorption, tissue binding, and/or lung retention to sustain the drug’s anti-infective properties for periods between dosing (Patton et al., 2004; Li and Byron, 2012). Notably, tobramycin is believed to cross epithelial barriers poorly, and it is therefore marketed primarily as an injectable (Jaresko and Alexander, 1995; Phillips and Shannon, 1997). Our recent meta-analysis of pharmacokinetic data in humans following intravenous and inhalation administration found that it was not possible to come to a statistically sound conclusion about tobramycin's bioavailability and possible binding to, or sequestration in, lung tissue following inhalation administration due largely to the variance associated with the available data (Li and Byron, 2012), even though tobramycin and other aminoglycoside antibiotics are known to manifest some of their toxicity through tissue binding or sequestration and the creation of “deep compartments” (Mingeot-Leclercq and Tulkens, 1999; Nagai and Takano, 2004; Stepanyan et al., 2011). It is noteworthy that, and probably because of the physiological existence of these drugs as polycations, transfer into cells is slow and intracellular sequestration can be persistent. For these reasons, we have used the isolated perfused rat lung (IPRL) to systematically investigate the disposition of tobramycin in the lung following its administration directly to the airways. The IPRL has been used similarly before (Byron et al., 1986; Niven and Byron, 1988; Niven et al., 1990; Sakagami et al., 2002). In the most simple cases, when solutes such as fluorescein are administered in solution to the airways of this preparation, they behave in accordance with Scheme 1. Absorption of fluorescein then occurs in a dose-independent, first-order fashion with no evidence of binding or metabolism. However, the nominal dose, or dose loaded into a dosing cartridge (Fig. 1), usually results in a smaller administered dose (D) (Scheme 1), only part of which, the absorbable amount, A0, (A0 = D − U), where U is the untransferable (unabsorbable) amount of each dose, can be transferred to the perfusate (P). This is because the bronchial circulation in the IPRL is severed, and only solute that deposits proximal to the actively perfused pulmonary circulation can be absorbed into the perfusate (Byron et al., 1986; Niven and Byron, 1988). In such a case, solute in the perfusate will increase monoexponentially toward an asymptotic value equal to the initial condition, A0. The fraction of the administered dose (D) transferred to the perfusate (Fp) for a solute described by Scheme 1 is given as a function of time (t):(1)where Fa is the absorbable fraction (= A0/D) and ka is the apparent first-order rate constant for absorption.
In the case of other nonmetabolized drugs that are absorbed by apparent first-order kinetics, binding or sequestration to lung tissue can be modeled in accord with Scheme 2. This scheme differs from Scheme 1 because of the addition of a “bound drug” compartment (B) and the association and dissociation rate constants (k12 and k21). In Scheme 2, which is analogous to a two- compartment pharmacokinetic model, the drug in A and B should behave in a similar fashion to drug in plasma and drug in tissue, respectively, whereas drug in P should behave in the same way as cumulative elimination. Thus, Fp for Scheme 2 is given by a rearranged form of eq. 216 from Gibaldi and Perrier (1975), and the data for drug in the perfusate should conform to(2)whereand
In the current study, the dynamic dialysis techniques, first described by Meyer and Guttman and used subsequently by other groups to study solute-ligand binding (Meyer and Guttman, 1968, 1970; Bottari et al., 1975; Pedersen et al., 1977; Hiji et al., 1978; Sparrow et al., 1982; Hashimoto et al., 1984), were first used to support the use of the IPRL binding and sequestration model described in Scheme 2. When dynamic dialysis under sink conditions is used to extract solutes from sliced IPRL tissue following initial airway dosing to that tissue, a model analogous to Scheme 2, Scheme 3 applies (Meyer and Guttman, 1968). In contrast to Scheme 2, Scheme 3 lacks an untransferable component, i.e., U = 0, because all of D that is administered to the airways becomes available for dialysis, so the amount released into the receiver solution, R, becomes equal to D as t tends to infinity. Furthermore, the binding or sequestration rate constants, k12' and k21', differ from those in Scheme 2 because sliced tissue and drug release from the dialysis sac differ from the intact IPRL. The first-order rate constant for solute release (ke) from a dialysis sac into R depends largely on the diffusive properties of the solute and the conditions of dialysis. However, the fraction of administered dose remaining in the sac (At /A0), at time t, should be a reflection of the solute’s tissue binding, and should theoretically conform to eq. 218 from Gibaldi and Perrier (1975), where(3)in whichand
Materials and Methods
Tobramycin, mannitol, and fluorescein were administered in different aqueous solution formulations, by forced intratracheal instillation, to the airways of the isolated perfused rat lung preparation (Fig. 1) using a previously described method (Byron and Niven, 1988). Using this technique, studies were performed to explore the kinetics and mechanisms responsible for tobramycin’s pulmonary disposition. IPRL absorption studies were performed to elucidate the effects of concentration on airway-to-perfusate transfer (absorption). Prior to the analysis of data from those studies, IPRL binding studies were performed using dynamic dialysis (Meyer and Guttman, 1968). Those binding experiments used identical surgical preparation, perfusion, and airway dosing as those for IPRL absorption studies, after which the extent and rate of tobramycin and/or mannitol dialysis from lung tissue were determined.
Specific pathogen-free, 300–400-g, male Sprague-Dawley rats (Hilltop Laboratory Animals Inc., Scottsdale, PA) were used throughout. The animals were housed with access to water and food at 30–70% relative humidity and 18–26°C, and a 12-hour light/dark cycle for ≥2 days before sacrifice. All procedures were approved by the Institutional Animal Care and Use Committee at Virginia Commonwealth University.
Chemicals and Materials.
Tobramycin (≥98% free base) and bovine serum albumin [BSA; ≥98% (agarose gel electrophoresis) lyophilized powder; Sigma-Aldrich, St. Louis, MO] were stored according to USP 2007 (United States Pharmacopeia Convention, 2007). Krebs-Henseleit solution (KHS) containing 4% (w/v) BSA (K4) was freshly prepared and used as perfusate in the isolated perfused rat lung preparation. [3H]Tobramycin and [3H]mannitol solutions (0.8 Ci/mmol, 1 mCi/ml and 10–30 Ci/mmol, 1 mCi/ml, respectively) were purchased from Moravek Biochemicals (Brea, CA) and PerkinElmer Life and Analytical Sciences (Waltham, MA), and sodium fluorescein was obtained from Acros Organics (Geel, Belgium). Ecoscint XR scintillation cocktail (National Diagnostics, Atlanta, GA) was used for liquid scintillation counting. Other chemicals were purchased from Thermo Fisher Scientific (Pittsburgh, PA). Dialysis tubing (Snakeskin, 10K molecular weight cut-off; Thermo Fisher Scientific) was washed with water and immersed in aqueous buffer solutions for ≥30 minutes prior to use.
Aqueous radiolabeled tobramycin solutions were prepared in different concentrations (100, 20, 2, 0.2, 0.04, and 0.02 mg/ml) containing a nominal radiolabel content of 18 µCi/ml in 0.45% w/v NaCl and adjusted to pH 7.4 with H2SO4. Radiolabeled mannitol was prepared with and without tobramycin in different concentrations at 12 µCi/ml, whereas sodium fluorescein (unlabeled) solutions were prepared in pH 7.4 phosphate buffer and used at a concentration of 0.2 mg/ml. Although the majority of the experiments involving tobramycin were performed with dosing solutions containing 0.45% w/v NaCl (1/2 normal saline), osmolality was varied in some investigations by using different concentrations of sodium chloride.
IPRL Binding Studies.
A dynamic dialysis method (Meyer and Guttman, 1968) was modified to investigate solute binding or sequestration to lung tissue. Tobramycin binding was investigated across doses (0.002, 0.004, 0.2, and 2 mg) by comparing its interaction with lung tissue to that of mannitol (a nonbinding solute). In binding studies, the IPRL preparation was used and dosed in exactly the same way as described later for IPRL absorption studies. After dosing, the isolated lung preparation was taken out of the artificial glass thorax (AGT), sliced into 60–90 rectangular pieces (maximum linear dimension: 4 mm), and transferred to a dialysis sac (dry length: 15 cm, washed and immersed in KHS solution prior to use) with 10 ml of K4. To determine the intrinsic diffusive properties of tobramycin and mannitol in this system, solute dialysis was also studied in the absence of lung tissue and BSA, from sacs with identical total volumes of KHS. Sacs were closed, with minimal headspace, by ligating with cotton thread, after which they were immersed at time zero in the perfusate reservoir containing 200 ml of magnetically stirred receiver solution (KHS or K4) maintained at 37°C and pH 7.4. The solute mass enclosed in the sac at time zero (A0) was determined by subtracting the solute lost during lung slicing (by assay of rinse solutions used in the procedure) from the administered dose, D (determined as described in the IPRL absorption studies). One-milliliter samples were removed for assay from the receiver solution at 5, 10, and 20 minutes, after which a 100-ml sample was withdrawn at each time point of 30, 60, 90, 120, and 180 minutes and hourly thereafter. An equal volume of solute-free receiver solution (37°C) was added to replace each sampling aliquot and restore the receiver volume to 200 ml. The amount of solute remaining in the sac as a function of time (At) was calculated assuming mass balance.
IPRL Absorption Studies.
Use of the IPRL preparation (Fig. 1) to study solute absorption was carefully controlled as described previously (Byron and Niven, 1988; Sun et al., 1999). In brief, a rat lung was surgically removed and housed in an AGT maintained at 37°C. KHS with 4% (w/v) BSA was used as perfusate (K4; 200 ml) and recirculated through the pulmonary circulation via the pulmonary artery at a constant flow rate of 15 ml/min. A metal dosing cartridge containing 0.1 ml of dosing solution was inserted into the trachea via a tracheal cannula. A metered dose inhaler (25 μl of drug-free chlorofluorocarbon propellants per actuation) was connected to the dosing cartridge and actuated once. The dosing solution was propelled into the lung as a coarse spray, and the lung was inflated simultaneously to ~6 ml. The dosing cartridge was removed and the lung allowed to deflate (Byron and Niven, 1988). The administered dose (to the airways of the IPRL) was determined from the initial weight of the primed dosing cartridge, solution density, and concentration after subtracting the mass remaining in the dosing cartridge after administration (determined by assay). The perfusate samples were taken from the well mixed reservoir at time zero (blank sample, immediately prior to dosing) and subsequently at 1, 3, 5, 10, 15, 20, 30, 40, 50, 60, 75, 90, 105, and 120 minutes following dosing. Solute concentrations were determined as described under “Assay Methods.” Sufficient IPRL studies were performed to yield more than four fully viable preparations for each dosing solution, as evidenced by the absence of any signs of edema onset over 120 minutes (preparations that were “nonviable” at times ≤120 minutes were discarded). We have shown previously, and observed consistently with this preparation, that “signs of edema onset” occur when the (blood-free) lungs change in outside color and texture from smooth white to a gray and/or patchy appearance; this appearance change is an early indicator of the preparation’s declining viability, shortly after which values for the wet lung/dry lung weight ratio, epithelial permeability to solutes, and other effects change markedly (Byron et al., 1986; Niven et al., 1990). At the end of each IPRL absorption study, the lung tissues were collected, homogenized, and assayed for solute remaining in both the airways and the lung tissue.
[3H]Tobramycin and [3H]mannitol concentrations were determined by scintillation counting relative to standards prepared freshly for each experiment (Liquid Scintillation Analyzer, Tri-Carb 2800-TR; PerkinElmer Life and Analytical Sciences) after first validating the radioactivity assay for tobramycin with high-performance liquid chromatography–mass spectrometry to ensure that chemical degradation and/or metabolism did not occur in lung tissue or perfusate over the duration of a typical experiment. In brief, perfusate samples containing radiolabeled solutes were used neat or diluted in K4. One-milliliter aliquots were added to 5 ml of scintillation cocktail (Ecoscint XR; National Diagnostics) in 7-ml polypropylene scintillation vials and well mixed. Total radioactivity was expressed as disintegrations per minute. Disintegrations per minute from accurately prepared [3H]tobramycin or [3H]mannitol standard solutions in K4 were measured independently for each assay series to calculate solute concentrations in unknown test solutions. Fluorescein concentrations in the perfusate were assayed by spectrofluorophotometer (RF-5301 PC; Shimadzu Corporation, Kyoto, Japan; λex and λem = 490 and 520 nm, respectively), as described earlier (Byron et al., 1986; Byron and Niven, 1988; Sakagami et al., 2002).
Kinetic analysis was performed according to the theory described earlier and Schemes 1–3. The data from the IPRL binding studies using dynamic dialysis were expressed in accord with Scheme 3, as the fraction of the solute dose (At/A0) remaining in the sac as a function of time. The ke value for tobramycin was first determined by studying its dialysis kinetics in the absence of IPRL components (tissue and protein). In this case, linear regression analysis of the first-order data for ln[At/A0] versus t was performed to determine its value (Meyer and Guttman, 1968). The binding rate constants, in the presence of lung tissue, k12' and k21' were then determined by fitting data for [At/A0] versus t to eq.3 by least mean square nonlinear regression analysis. The value for ke was fixed, whereas k12' and k21' were allowed to float. Goodness-of-fit was assessed using the calculated r2 and model selection criterion by Scientist 3.0 (MicroMath Scientific Software, Salt Lake City, UT). Data from IPRL absorption studies were grouped in accord with the nominal dose initially added to the dosing cartridge. Each nominal dose resulted in a mean value for D (Scheme 1), due to solution retention in the cartridge. Fluorescein absorption into the perfusate was expressed as the mean fraction of each administered dose reaching the perfusate, Fp, versus time. Best estimates for fluorescein’s mean absorbable fraction, Fa (= A0/D; Scheme 1), and its apparent first-order rate constant for absorption, ka, were obtained by fitting the unweighted data for Fp versus time to eq.1 using Scientist 3.0. Because fluorescein that penetrates the airways proximal to the circulating perfusate is known to be completely absorbed, this provided a mean value for the solute’s eventually absorbable fraction, Fa, that could be used across solutes. Tobramycin data were analyzed according to Scheme 2, where best estimates for the rate constants describing binding and absorption (k12, k21, ka) at different nominal doses were determined by curve-fitting tobramycin Fp (fraction of administered dose in the perfusate) versus time data to eq. 2.
IPRL Binding Studies.
Tobramycin dialysis from the sac is shown as mean At /A0 (fraction of administered dose remaining in the sac) versus time in Fig. 2A. Biphasic (biexponential) profiles resulted from experiments in which drug release from sliced lung tissue occurred after airway dosing to the IPRL at five different dose levels; smaller doses showed greater values for tissue retention with time. In the absence of lung tissue, control experiments showed monoexponential (apparent first-order) release of tobramycin from the dialysis sac under sink conditions (Meyer and Guttman, 1968). Following the data analysis, the mean dialysis rate constant, ke (Scheme 3), for tobramycin was found to be dose-independent (0.0107 ± 0.0021 min−1; Fig. 2A). Mannitol, a nonbinding solute, produced monoexponential and dose-independent At /A0 profiles in the presence and absence of IPRL tissue (Fig. 2B). Tobramycin dialysis data from IPRL tissue (Fig. 2A) was fitted to eq. 3 to produce the best estimates of the rate constants in Scheme 3. These are shown in Table 1 for each dose; dashed curves in Fig. 2A were produced by simulation using the rate constant values shown in the table. Tobramycin’s dissociation rate constant (k21') was effectively dose-independent, whereas k12' appeared saturable and decreased with ascending dose.
IPRL Absorption Studies.
Fp versus time data for fluorescein (Fig. 3) was fitted to eq. 1 to obtain best estimates of the absorbable fraction (Fa) and the first-order rate constant for absorption (ka) for this nonbinding solute; the values were 0.75 and 0.076 min−1, respectively (Table 2). Although mannitol absorption was too slow to reach an asymptote within the IPRL’s viable lifetime (Sun et al., 1999), its ka values shown in Table 2 were derived by fitting mannitol data to eq. 1 assuming that the absorbable fraction, Fa, was the same as that of fluorescein (Fa = 0.75). The approach was supported by the absence of a statistical difference between mannitol’s Fp (fraction of administered dose in perfusate) versus time profiles at different nominal doses (0.02 and 2 mg; t test, P < 0.05; data not shown) and the agreement of ka values (Table 2) with reports in the literature for mannitol in the in situ rat lung (Brown and Schanker, 1983). In contrast to fluorescein and mannitol, tobramycin’s absorption was clearly dose-dependent, the rate and extent of its pulmonary absorption increasing with the magnitude of the administered dose (Fig. 3). At 120 minutes following tobramycin administration, the values for mean Fp (±S.D.; nominal dose) were 0.449 (±0.024; 0.002 mg), 0.505 (±0.021; 0.02 mg), 0.587 (±0.028; 0.2 mg), and 0.612 (±0.050; 2 mg). Tobramycin’s Fp versus time data showed good agreement with Scheme 2 and eq. 2, as indicated by the continuous curves shown in Fig. 3. Best estimates of the rate constants used to generate these curves are reported in Table 2 alongside the values for the coefficient of determination, r2, and the model selection criterion. The rate constant for absorption appeared to be dose-independent (range from 0.065 to 0.070 min−1) and close to that of fluorescein (0.076 min−1). The rate constant for dissociation from intact IPRL tissue (k21) was also relatively constant (0.018–0.022 min−1), whereas that for association (k12) decreased from 0.164 to 0.072 min−1 with increasing airway dose from 0.002 to 2 mg. As a result, the ratio of k12/ k21 decreased from 8.9 to 3.4 when the nominal dose was increased from 0.002 to 2 mg.
Binding of tobramycin and other aminoglycosides to various tissues has been used to explain the formation of deep compartments in pharmacokinetic studies (Schentag et al., 1978; Vozeh et al., 1979; Winslade et al., 1987), some epithelial cells of which (e.g., kidney proximal tubule and inner ear hair cells) are associated with nephro- and ototoxicity (Just and Habermann, 1977; Hiel et al., 1993; Todd and Hottendorf, 1995; Nagai and Takano, 2004). It is noteworthy that, and probably because of the physiological existence of these drugs as polycations, transfer into cells is slow, and intracellular sequestration can be persistent. In the hope of elucidating the drug’s binding to and/or intracellular sequestration by lung tissue, dynamic dialysis was performed both in the absence and presence of tissue and from the sliced IPRL, after dosing the airways using the same technique as that used for the IPRL absorption studies. The open circles in Fig. 2A show the apparent first-order, dose-independent dialysis of tobramycin in the absence of lung tissue in which there was no evidence of drug binding. The comparator data shown in Fig. 2B, for the nonbinding solute mannitol, was monoexponential and dose-independent (r2 > 0.98) in both the presence and absence of tissue, indicating only that the presence of sliced lung tissue slightly reduced the value of ke for diffusive release from the sac (Scheme 3; ke values for mannitol were 0.025 and 0.019 min−1 in the absence and presence of sliced IPRL tissue, respectively) by hindering diffusion. Most notably, the tobramycin data for At /A0 (fraction of administered dose remaining in the sac) versus time at different doses in the presence of lung tissue were described well by eq. 3 and Scheme 3 (r2 > 0.998), showing clear evidence of dose-dependent lung tissue binding or sequestration, as summarized by the decreasing values for the forward binding rate constant, k12', and the ratio k12'/k21' as doses were increased (Scheme 3; Table 1). These values for k12' and k21' implied that tissue binding occurred relatively slowly after airway administration, and that all three rate constants in Scheme 3 had a similar order of magnitude. Efforts to determine whether significant binding/sequestration was due to tissue or the constituents of the IPRL airway lining fluid implied that lung tissues themselves were largely responsible for antibiotic sequestration. Bronchoalveolar lavage samples, diluted with KHS containing low tobramycin concentrations (consistent with doses of approximately 0.002 mg), showed no difference in drug dialysis kinetics from the IPRL tissue-free control data for At/A0 in Fig. 2A.
This “slow-on” and “slow-off” binding behavior seen during dynamic dialysis supported the kinetic analysis of Fp versus time data in the IPRL absorption studies. When the ex vivo IPRL was used, the values of the aminoglycoside’s binding constants in Scheme 2 (Table 2) were quantitatively different from those from the dialysis experiments (association and dissociation occurred faster under conditions involving absorption into the perfusate), although the overall trend in the rate constant values was the same as that seen during dialysis. Neither fluorescein nor mannitol showed dose-dependent binding behavior, and both showed monoexponential absorption properties (ka and Fa values) consistent with reports in the literature (Brown and Schanker, 1983; Byron and Niven, 1988). Tobramycin, however, showed decreasing rates of fractional absorption (transfer from airways to the perfusate) as a function of decreasing dose and increasing binding/sequestration. Notably, dosing solution osmolality appeared to have no effect on Fp versus time data. For example, Fp at 120 minutes for a tobramycin dose of 0.2 mg in 0.9% (w/v) NaCl showed no statistical difference from a dose of 0.2 mg in 0.45% (w/v) NaCl (t test, P < 0.05). However, the data for Fp versus time at different doses were described well by eq. 2 and Scheme 2 (r2 > 0.998) with clear evidence of dose-dependent and possibly saturable lung tissue binding summarized by the decreasing values for the forward binding rate constant, k12, and the ratio k12/k21 as doses were increased (Table 2). The significant difference between the retention of this solute in lung tissue was related to tobramycin’s slow binding or sequestration following its administration. Values for k21 appeared unrelated to dose, but values of k12 decreased as the dose was increased, but without evidence of capacity limitation. Although the exposure of different tissue sites was clearly possible for absorption and dialysis experiments (IPRL was sliced immediately following dosing), the trend in the data for the lung tissue binding or sequestration constant, k12 (increases with decreasing dose) and k21 (effectively dose-independent), was consistent with the rate constant data for the sliced IPRL.
To the best of our knowledge, this is the first work to investigate tobramycin binding to and/or sequestration by lung tissue and explore its effects on the drug’s pulmonary disposition following airway administration. Dynamic dialysis, initially used for protein binding studies, was successfully used to study tissue binding and sequestration, and the results were consistent with those from the realistic ex vivo IPRL model. Based on the results, it was possible to calculate the apparent elimination half-life from the IPRL (equal to 0.693/β; eq. 2) as 2.3, 1.8, 1.3, and 1.2 hours for nominal doses of 0.002, 0.02, 0.2, and 2 mg, respectively. Notably, if binding did not occur, absorption into the perfusate (and thus, elimination from the IPRL) should occur with a half-life of ~10 minutes (0.693/ka; Table 2), similar to that for fluorescein anions (Niven and Byron, 1988; Table 2). The 1.8-hour half-life at a nominal dose of 0.02 mg is consistent with the in vivo lung elimination half-life of 2.1 hours reported by Valcke and Pauwels (1991) in rat alveolar lining fluid following aerosol administration of a similar dose.
In clinical treatment of humans, tobramycin inhalation solution (TOBI; Novartis, Basel, Switzerland), the commercial nebulizer formulation used for cystic fibrosis patients, is prescribed as an intermittent (28 days on/28 days off) treatment with 300 mg/5 ml nebulized and inhaled twice daily regardless of age (≥6 years) or body weight. Since the percentage of each nebulizer dose reaching the lung via the recommended PARI LC PLUS nebulizer has been reported to be ~15% (Lenney et al., 2011), a lung dose of ~45 mg could be expected in a 70-kg human. Based on weight scaling, this corresponds to a nominal dose of approximately 0.2 mg to the airways of the rat lungs used in the IPRL studies described here, where the k12/k21 ratio was 4.06 (Table 2). Although species differences may well exist, similar cell constituents and concordance with alveolar surface area, lung volume, capillary volume, and body weight have been reported across many species, including rats and humans (Crapo et al., 1983; Plopper, 1983; Cryan et al., 2007). Therefore, in conclusion, these studies in the rat lung appear to support the existence of tobramycin retention in lung tissue following airway administration at doses likely to produce airway concentrations seen in humans. As a result of this slow-on and slow-off tissue binding or sequestration, the antibiotic’s longevity in the lung is extended. It is possible that this may account, at least in part, for the apparent success of tobramycin inhalation therapy seen in clinical practice.
The authors thank Drs. Masahiro Sakagami, Michael Hindle, Douglas Sweet, and Jurgen Venitz for providing experimental help with the techniques described in this study or data interpretation during manuscript preparation.
Participated in research design: Li, Byron.
Conducted experiments: Li.
Contributed new reagents or analytic tools: Byron.
Performed data analysis: Li, Byron.
Wrote or contributed to the writing of the manuscript: Li, Byron.
- Received June 25, 2013.
- Accepted September 5, 2013.
The authors are faculty and students of Virginia Commonwealth University. No conflicts of interest exist.
M.L. received a stipend and tuition and fees during graduate studies from the School of Pharmacy, Virginia Commonwealth University. Supplies and equipment were provided by The Medical College of Virginia Foundation, Richmond, VA.
- the absorbable amount
- artificial glass thorax
- the fraction of administered dose remaining in the dialysis sac
- the amount of drug bound to, or sequestered in, tissue
- bovine serum albumin
- the administered dose
- the absorbable fraction
- mean fraction of each administered dose reaching the perfusate
- isolated perfused rat lung
- k12 (k12')
- the association rate constants for binding and/or retention in the intact IPRL (sliced IPRL, dynamic dialysis)
- k21 (k21')
- the dissociation rate constants for binding and/or retention in the intact IPRL (sliced IPRL, dynamic dialysis)
- Krebs-Henseleit solution containing 4% (w/v) BSA
- the apparent first-order rate constant for absorption
- the apparent first-order rate constant for dialysis
- Krebs-Henseleit solution
- the amount absorbed into the perfusate
- the amount released into the receiver solution
- the untransferable (unabsorbable) amount of each dose
- Copyright © 2013 by The American Society for Pharmacology and Experimental Therapeutics