Large Porous Particles for Pulmonary Drug Delivery

See allHide authors and affiliations

Science  20 Jun 1997:
Vol. 276, Issue 5320, pp. 1868-1872
DOI: 10.1126/science.276.5320.1868


A new type of inhalation aerosol, characterized by particles of small mass density and large size, permitted the highly efficient delivery of inhaled therapeutics into the systemic circulation. Particles with mass densities less than 0.4 gram per cubic centimeter and mean diameters exceeding 5 micrometers were inspired deep into the lungs and escaped the lungs' natural clearance mechanisms until the inhaled particles delivered their therapeutic payload. Inhalation of large porous insulin particles resulted in elevated systemic levels of insulin and suppressed systemic glucose levels for 96 hours, whereas small nonporous insulin particles had this effect for only 4 hours. High systemic bioavailability of testosterone was also achieved by inhalation delivery of porous particles with a mean diameter (20 micrometers) approximately 10 times that of conventional inhaled therapeutic particles.

Inhaled aerosols are effective therapeutic carriers for the treatment of respiratory inflammation (1), cystic fibrosis (2), and other lung disorders (3); they also offer potential for noninvasive systemic delivery of peptides and proteins (4). Local and systemic inhalation therapies can often benefit from a controlled release of the therapeutic agent (5), as is achievable with the use of biodegradable polymeric materials (6). Slow release from an inhaled therapeutic particle can prolong the residence of an administered drug in the airways or acini and can diminish the rate of a drug's appearance in the bloodstream (7). Also, patient compliance increases when dosage frequency is reduced (7).

The human lungs, however, have efficient means of removing deposited particles over periods ranging from minutes to hours. In the upper airways, ciliated epithelia contribute to the “mucociliary escalator” (8), by which particles are swept from the airways toward the mouth. In the deep lungs, an army of alveolar macrophages is capable of phagocytosing particles soon after their deposition (9). An effective slow-release inhalation therapy therefore requires a means of avoiding or suspending the lungs' natural clearance mechanisms until encapsulated drugs have been effectively delivered.

Until now, therapeutic dry powder aerosols have been made with particle mass densities (ρ) of ∼1 ± 0.5 g/cm3 and mean geometric diameters (d) of <5 μm to avoid excessive deposition in the dry powder inhaler (DPI) and oropharyngeal cavity (5, 10). Here, we show that very light particles (ρ < ∼0.4 g/cm3) with d > 5 μm can be deposited in the lungs. As a consequence of their large size and low mass density, porous particles can aerosolize from a DPI more efficiently than smaller nonporous particles, resulting in higher respirable fractions of inhaled therapeutics. Also by virtue of their size, large particles can avoid phagocytic clearance from the lungs until the particles have delivered their therapeutic dose; this attribute can be particularly useful for controlled-release inhalation therapies.

To assess the merits of large porous particles for pulmonary drug delivery, we encapsulated model therapeutics inside porous particles (Fig. 1A) composed of 50:50 poly(lactic acid-co-glycolic acid) (PLGA). Double- and single-emulsion solvent evaporation techniques (11) were used to prepare porous and nonporous PLGA particles, respectively. Porous and nonporous particles of similar aerodynamic diameter (12), loaded with ∼15 weight % model therapeutic (testosterone), were aerosolized into a cascade impactor system (13) from a Spinhaler DPI for 30 s at an airflow rate of 28.3 liter/min. The cascade impactor provides an in vitro system for estimating the respirable fraction of a dry powder; it consists of a closed chamber within which flat plates are arranged perpendicular to the airflow, such that particles deposit stagewise in a manner reflective of their aerodynamic diameters. After deposition on the stages of the impactor, particles were collected (13) and total particle mass was assessed stagewise; the respirable fraction was determined as the percent of total particle mass exiting the DPI, recovered from the terminal, “respirable” stages of the impactor. Nonporous particles [d = 3.5 μm, ρ = 0.8 g/cm3 (14)] exhibited a respirable fraction of 20.5 ± 3.5%, whereas 50 ± 10% of porous particles (d = 8.5 μm, ρ = 0.1 g/cm3) were respirable, even though the aerodynamic diameters (12) of the two particle types are nearly identical. The large porous particles' high efficiency can be attributed to their smaller surface-to-volume ratio. Large particles aggregate less than small particles, all other considerations being equal (15, 16); thus, while both have identical aerodynamic diameters, the large particles tend to exit the DPI more generally as single particles. The smaller particles aggregate more, leading to their deposition by gravity and inertia before reaching the “respirable” stages of the impactor.

Figure 1

Confocal microscopy images of (A) porous PLGA and (B) porous PLAL-Lys particles. Fluorescein isothiocyanate–dextran was encapsulated in the PLGA particle to render the pore spaces of the particle visible in the fluorescent confocal image. The PLAL-Lys particles were fluorescently labeled through the reaction of rhodamine isothiocyanate with lysine amine groups on the surface of the particles. The PLGA and PLAL-Lys particles are highly porous, as evidenced by the appearance of fluorescence throughout the particle structure.

To assess the influence of particle composition, we aerosolized a second type of porous particle (Fig. 1B), composed of poly(lactic acid-co-lysine-graft-lysine) (PLAL-Lys) (11). The PLAL-Lys particles exhibit some hygroscopicity, possibly a result of their lysine content, whereas the PLGA particles do not. Porous PLAL-Lys aerosols (d = 8.2 μm, ρ < 0.1 g/cm3) exhibited an in vitro respirable fraction (57 ± 1.9%) similar to that of the porous PLGA particles (50 ± 10%), which suggests that absolute particle mass density, rather than particle chemistry or hygroscopicity, is the prime determinant of the relatively high respirable fractions observed for the large porous particles. The values of 50 ± 10% and 57 ± 1.9% for porous particles exceed comparable respirable fractions obtained in recent aerosolization studies (15) performed with mannitol (4 ± 0.3%) and recombinant human granulocyte colony-stimulating factor (blended with mannitol) (34 ± 2%) powders using a Spinhaler DPI at a similar airflow rate (30 liter/min).

To determine whether the relatively efficient in vitro aerosolization of large porous particles translates into improved respirable fractions in vivo, we aerosolized porous and nonporous particles into the airways of rats (17). During forced ventilation, rats were exposed to porous or nonporous particles; bronchoalveolar lavage was used to collect particles deposited in the trachea as well as in the airways and acini (18). The nonporous particles deposited primarily in the trachea (∼79% of all particle mass that entered the trachea), whereas only 46% of the porous particle mass deposited in the trachea. Particles remaining in the rat lungs after bronchoalveolar lavage were obtained by careful dissection of the individual lobes of the lungs in subsequent experiments (19) (Fig. 2). The absolute number of porous particles remaining in the lungs was approximately an order of magnitude greater than the corresponding number of nonporous particles.

Figure 2

Total particle recovery in rat lungs after bronchoalveolar lavage. Lobe numbers correspond to (1) left lung, (2) anterior, (3) median, (4) posterior, and (5) postcaval. For porous PLAL-Lys particles, d = 6.9 ± 4.2 μm and ρ = 0.1 g/cm3. For nonporous PLA particles, d = 6.7 ± 3.2 μm and ρ = 0.94 g/cm3. Means and SEs are based onn = 4.

The role of low mass density in rendering large particles respirable can be understood in terms of the particles' mean aerodynamic diameter (12). Relatively large particles with high porosity have the same aerodynamic diameter as smaller, nonporous particles; these larger particles can enter the lungs because particle mass dictates the location of aerosol deposition in the lungs. The increased aerosolization efficiency of large, light particles lowers the probability of deposition losses before particle entry into the airways, thereby increasing the systemic bioavailability of an inhaled drug.

To test whether large particle size can increase systemic bioavailability, we encapsulated insulin into porous and nonporous polymeric particles. We designed the mass densities and mean diameters of the two particles such that they each had an aerodynamic diameter (∼2 μm) suitable for deep lung deposition (12); the mean diameters of the porous and nonporous particles were >5 μm and <5 μm, respectively (Fig. 3, A to C). Identical masses of the porous or nonporous particles were administered to rats as an inhalation aerosol or injected subcutaneously (controls) (20). Serum insulin concentrations were monitored as a function of time after inhalation or injection. For both porous (Fig.3A) and nonporous (Fig. 3B) particles, blood levels of insulin reached high values within the first hour after inhalation. Only with large porous particles did blood levels of insulin remain elevated (P < 0.05) beyond 4 hours, with a relatively constant insulin release continuing to at least 96 hours (0.04 <P < 0.2). These results were confirmed by serum glucose values (Fig. 3C), which show falling glucose levels for the first 10 hours after inhalation of the porous insulin particles, followed by relatively constant low glucose levels for the remainder of the 96-hour period [for small nonporous insulin particles, initially suppressed glucose values rose after 24 hours (21)]. Similar biphasic release profiles of macromolecules from PLGA polymers have been reported (22). For the large porous particles, insulin bioavailability relative to subcutaneous injection (23) was 87.5%, whereas the small nonporous particles yielded a relative bioavailability of 12% after inhalation. By comparison, bioavailability (relative to subcutaneous injection) of insulin administered to rats as an inhalation liquid aerosol has been reported as 37.3% using a similar endotracheal method (24). The absolute bioavailability of insulin inhaled into rat lungs in the form of a lactose-insulin powder through a DPI connected to an endotracheal tube has been reported as 6.5% (25). The longest sustained insulin release previously reported (6 hours) was achieved using liposomes intratracheally instilled into rat lungs (26).

Figure 3

Systemic concentrations in rats after administration of porous and nonporous therapeutic particles. (A) Serum insulin concentration after inhalation and subcutaneous injection of 9 mg of large porous insulin particles. No insulin particles were administered to nontreated controls. Porous particles contained insulin (20.0 weight %) and 50:50 PLGA (80.0 weight %) (11). Their mean d andd aer values were 6.8 μm and 2.15 μm, respectively. Means and SEs are based on n = 3. (B) Serum insulin concentration after inhalation and subcutaneous injection of 9 mg of small nonporous insulin particles. No insulin particles were administered to nontreated controls. Nonporous particles contained insulin (10.0 weight %) and 50:50 PLGA (90.0 weight %) (11). Their mean d andd aer values were 4.4 μm and 2.15 μm, respectively. Means and SEs are based on n = 3. (C) Serum glucose concentration after inhalation of 9 mg of large porous insulin particles or 9 mg of small nonporous insulin particles. No insulin particles were administered to nontreated controls. Means and SEs are based on n = 3. (D) Serum testosterone concentration after administration of 6 mg of porous testosterone particles (d = 20.4 μm) as an inhalation powder and as a subcutaneous control. Particles contained testosterone (15 weight %), 50:50 PLGA (76.5 weight %), and PLAL-Lys (8.5 weight %). For the dry powder, ρ = 0.1 g/cm3. (E) Same as (D) but with smaller porous testosterone particles (d = 10.1 μm).

Given the short systemic half-life of insulin (11 min) (27) and the 12- to 24-hour time scale of particle clearance from the central and upper airways (5), the appearance of exogenous insulin in the bloodstream several days after inhalation appears to indicate that large porous particles achieve long, nonphagocytosed lifetimes in the deep lungs. To test this hypothesis, we lavaged (18) the lungs of rats immediately after inhalation of the porous and nonporous insulin particles as well as 48 hours after inhalation. For nonporous particles, 30 ± 3% of phagocytic cells contained particles immediately after inhalation, and 39 ± 5% contained particles 48 hours after inhalation. By contrast, only 8 ± 2% of phagocytic cells contained large porous particles immediately after inhalation, and 12.5 ± 3.5% contained particles 48 hours after inhalation. For small nonporous particles, 17.5 ± 1.5% of the phagocytic cell population contained three or more particles 48 hours after inhalation, compared with 4 ± 1% for large nonporous particles. Inflammatory response was also elevated with small nonporous particles; neutrophils represented 34 ± 12% of the phagocytic cell population 48 hours after inhalation of the small nonporous particles, compared with 8.5 ± 3.5% for large porous particles (alveolar macrophages represented 100% of cells measured immediately after inhalation). These results are consistent with those of in vitro experiments showing that phagocytosis of particles diminishes precipitously as particle diameter increases beyond 3 μm (28).

To further determine whether increased bioavailability correlates with increasing size of porous particles, we encapsulated a second model drug, testosterone, in porous particles of two different mean geometric diameters (10.1 and 20.4 μm). An identical mass of powder was administered to rats as an inhalation aerosol or as a subcutaneous injection (controls). Serum testosterone concentrations were monitored as a function of time after inhalation or injection (Fig. 3, D and E). Blood levels of testosterone remained well above background levels (P < 0.05) for 12 to 24 hours, even though the systemic half-life of testosterone is 10 to 20 min (27). Testosterone bioavailability relative to subcutaneous injection was 177% for the 20.4-μm-diameter particles (Fig. 3D) and 53% for the 10.1-μm-diameter particles (Fig. 3E). The increase in testosterone bioavailability with increasing size of porous particles is especially notable given that the mean diameter of the 20.4-μm particles is ∼10 times that of nonporous conventional therapeutic particles (5, 10). The relatively short time scale of testosterone release observed for both the inhalation and subcutaneous controls is near the in vitro time scale of release (several hours) reported elsewhere for 50:50 PLGA microparticles of similar size encapsulating a therapeutic substance (bupivacaine) of similar molecular weight and lipophilicity (29).

Porous particles comprising therapeutics and pharmaceutical excipients can easily be formed by spray-drying (30), rapid expansion of supercritical fluids (31), and other particle formation technologies. Hence, they can immediately address a variety of needs as therapeutic carriers for inhalation therapies. Their potential for high aerosolization efficiency, long-term drug release, and increased systemic bioavailability makes large porous particles especially attractive for systemic inhalation therapies.

  • * To whom correspondence should be addressed. E-mail: dxe11{at} (D.A.E.), rlanger{at} (R.L.).


Stay Connected to Science

Navigate This Article