Introduction

Top Abstract Introduction Methods Results Discussion References

rhGH, a 191 amino acid protein, is used to treat short stature caused by growth hormone deficiency, and promising clinical results have been obtained in the treatment of Turner's syndrome and growth failure due to chronic renal insufficiency (Daughaday, 1995, Daughaday and Harvey, 1995). In addition, because of its anabolic effects, there is clinical evidence that hGH may be useful in treating trauma, clinical malnutrition, osteoporosis and to facilitate wound healing.

hGH is stored in the anterior pituitary and secreted in a pulsatile fashion mainly during sleep (Finkelstein et al., 1972, Veldhuis and Johnson, 1986). Because it is a protein, it is not absorbed orally to any significant extent (Lee, 1995, Pearlman and Bewley, 1993) and thus must be administered by injection. rhGH is currently administered by daily or thrice weekly injections over a period of several years. However, recent clinical studies have shown that continuous infusion of rhGH via a pump results in growth velocity and IGF-I levels comparable to those achieved with daily injections (Jorgensen et al., 1990, Laursen et al., 1994, Laursen et al., 1995, Tauber et al., 1993). This result demonstrates that continuous, as well as pulsatile, administration of rhGH is efficacious, although the full range of potential differences between these two regimens of growth hormone administration has not yet been fully investigated.

One method to produce injectable, sustained-release formulations of proteins, including rhGH, is to encapsulate the drug into injectable microspheres of biodegradable PLGA from which the drug is released slowly by diffusion and as the polymer degrades (Cleland, in press, Schwendeman et al., 1996). Microspheres made from PLGA are biocompatible and biodegrade into lactic and glycolic acid and thus do not have to be removed. These polymers are also used to make sutures, bandages and bone plates (Austin et al., 1995, Pihlajamaki et al., 1992, Winde et al., 1993). Depending on different polymer properties, such as polymer chain length, lactide:glycolide ratio, and the presence of polymer end-group modification, the degradation rate and hence the rate of drug release can be controlled.

The processes commonly used to produce PLGA microspheres, such as that used to make the currently marketed sustained-release formulations of LHRH (Ogawa et al., 1988a, 1988b), were developed to encapsulate relatively small and stable molecules. They use elevated temperatures, surfactants or aqueous/organic solvent interfaces, conditions that denature and inactivate many proteins. To accomodate the stability needs of proteins, we have developed a process that is carried out at cryogenic temperatures, uses no water and hence avoids oil-aqueous interfaces and requires no surfactants (Johnson et al., 1996). This process, specifically designed to encapsulate relatively labile macromolecules, results in PLGA microspheres that release protein with physical, chemical and biological properties essentially identical to those before encapsulation.

Our study assesses the pharmacokinetics, the biologic effect and the potential immunogenicity of rhGH released from three different PLGA microsphere formulations. We find that all exhibit sustained-release and induce sustained biological effect of rhGH and one of the formulations, chosen for more extensive investigation, demonstrates no greater immunogenicity than when the protein in solution is administered by frequent injections. In addition, there are no adverse effects either systemically or at the site of injection. Moreover, when an equivalent total dose is administered in a continuous fashion rather than by frequent injections, rhGH elicits a greater effect as measured by IGF-I levels. These results suggest that, as has been demonstrated with nonprotein drugs (Langer, 1990), sustained-release formulations of therapeutic proteins have the potential to improve the convenience of use and the safety and efficacy of this increasingly important class of drugs.

	Methods

Top Abstract Introduction Methods Results Discussion References

Preparation of microspheres. Microspheres were fabricated as described (Johnson et al., 1996). Briefly, to form the Zn-hGH complex, six molar equivalents of zinc acetate were added to rhGH (Nutropin, Genentech, Inc., South San Francisco, CA) in 4 mM sodium bicarbonate pH 7.2. After complex formation, the suspension was sprayed through a sonic nozzle into liquid nitrogen, placed at -80°C and lyophilized. After lyophilization the protein powder was suspended in a solution of polymer in methylene chloride. The polymer used for formulation I (table 1) was from Birmingham Polymers (Birmingham, AL) (lot 115-56-1, internal viscosity = 0.23 dl/g), and polymers RG502H (internal viscosity = 0.17 dl/g) and RG503H (internal viscosity = 0.4 dl/g) (Boehringer Ingelheim, Petersburg, VA) were used for formulations II and III, respectively. Zinc carbonate was added (6% w/v for formulation I and 1% w/v for formulations II and III) and the suspension was then sprayed through a sonic nozzle into liquid nitrogen overlaying frozen ethanol and placed at -80°C for 24 hr. at which time an equal volume of -80°C ethanol was added. After 48 hr the microspheres were recovered using a 0.65-µ filter and dried under vacuum. The microspheres have a mean diameter of approximately 50 µ and are suspended in an aqueous vehicle (3% w/v carboxymethyl cellulose, low viscosity; 1% v/v polysorbate 20 and 0.9% w/v NaCl) before injection.

Animals. Juvenile (prepubescent) male rhesus monkeys (Macaca mulatta) were housed at Corning Hazleton, Inc. (Madison, WI). At the initiation of treatment, they were between 11 and 27 mo old and weighed between 1.9 and 3.8 kg. Monkeys were maintained on a 14-hr light/10-hr dark cycle. Details of dose administration are given in the legend to table 2. To determine the local effects of the microspheres, the s.c. tissue surrounding the injection sites was isolated, fixed in 10% neutral buffered formalin, embedded in glycomethacrylate, sectioned at 2 to 3 µ and stained with H&E or with an immunohistochemical stain for hGH.

Quantitation of serum protein concentrations. Serum concentrations of rhGH in the monkeys were determined using an immunoradiometric assay (RADIM Group, Rome, Italy). Endogenous monkey GH was detected by this assay. Total IGF-I was measured using a radioimmunoassay (Liberman et al., 1992) and IGFBP-3 concentrations were determined by an immunoradiometric assay (Diagnostic Systems Laboratories, Inc., Webster, TX). All concentrations reported in the figures are means ± S.E.M.

Detection of anti-hGH antibodies. The presence of anti-hGH antibodies in the serum was determined using a radioimmunoprecipitation assay. Nonimmune serum was used as a negative control. Serum samples were incubated with ¹²⁵I-rhGH (ICN Pharmaceuticals, Inc., Costa Mesa, CA) and the antibody-bound ¹²⁵I-rhGH precipitated with 16% PEG 8000. Serum samples at a dilution of 1:10 having a value less than twice the negative control value were scored as negative.

Pharmacokinetic analysis. Noncompartmental pharmacokinetic analysis was performed on the rhGH concentration vs. time profile, using RSTRIP (MicroMath Scientific Software, Salt Lake City, UT), a program designed for exponential stripping and parameter estimation. All predose concentrations and all postdose concentrations lower than the lowest standard (1.5 ng/ml) were considered to be zero. The following pharmacokinetic parameters were determined: area under the curve (by trapezoidal integral) from time zero to day 2 (AUC_0-2); AUC from time zero to the last day of sampling (AUC_0-); maximum blood concentration (C_max), time to the maximum concentration (T_max), cumulative release and absolute bioavailability (F).

	Results

Top Abstract Introduction Methods Results Discussion References

Preparation of rhGH containing microspheres. The PLGA microspheres were made by precipitating rhGH from solution by the addition of zinc acetate and encapsulating the Zn-protein powder using a nonaqueous, cryogenic process (see "Methods"). Zinc, which reversibly complexes with hGH and causes the dimerization of the protein (Cunningham et al., 1991), is present at high concentrations in the secretory granules of the anterior pituitary (Thorlacius-Ussing, 1987). A zinc complex is believed to be the form in which the protein is naturally stored. The three microsphere formulations used in this study, which were selected from a panel of 11 formulations based on in vitro and in vivo (rat) release rate and the protein integrity (Johnson et al., in press), differed in the content of zinc carbonate and polymer composition (table 1). The lactide:glycolide ratio of all three polymers was 50:50. The rhGH extracted from all three microsphere formulations was shown to be identical to that before encapsulation by size exclusion chromatography, reverse phase HPLC, and anion exchange chromatography (Johnson et al., in press). This indicates that the encapsulation process caused no detectable change in the protein.

Pharmacokinetics of rhGH release. To evaluate the pharmacokinetics and pharmacodynamics of rhGH release, juvenile rhesus monkeys were used because of their low level of endogenous growth hormone. This experiment included three groups of animals receiving the microsphere formulations and three additional groups receiving rhGH solution by different means; all animals in each of these six groups received a total of 24 mg of rhGH (table 2). The animals receiving the microspheres (groups 1-3) received a single s.c. injection of 160 mg of microsphere formulations I to III (24 mg rhGH), respectively; group 4 received the entire amount of rhGH as a single s.c. bolus of protein in solution and group 5 received rhGH in solution administered daily for 28 days (0.86 mg/day). To mimic the expected release rate from the microspheres,¹ group 6 received 15% of the 24 mg (3.6 mg) as a s.c. bolus (to mimic the initial release) and the remaining 20.8 mg were delivered continuously via a 28-day osmotic pump surgically implanted s.c. A seventh group received only the microsphere vehicle, which resulted in no change in the growth hormone, IGF-I or IGFBP-3 levels (data not shown).

View larger version (13K): [in this window] [in a new window]   Fig. 1.   rhGH serum levels in rhesus monkeys for 2 days after s.c. administration of encapsulated rhGH and rhGH solution. Treatment groups (see table 2) were: microsphere formulation I (; group 1), formulation II (; group 2) or formulation III (; group 3); 24 mg of rhGH in solution (; group 4); 3.6 mg of rhGH solution followed by implantation of an osmotic pump containing rhGH (; group 6).

View larger version (20K): [in this window] [in a new window]   Fig. 2.   rhGH serum levels in rhesus monkeys for 2 mo after s.c. administration of encapsulated rhGH and rhGH solution. Treatment groups, excluding group 4, and symbols are as in figure 1.

Biological effect of hGH. The IGF-I and IGFBP-3 serum levels induced by the three microsphere formulations, the osmotic pump and daily injections of rhGH solution are shown in figures 3 and 4, respectively. Whereas the single injection of the entire amount of rhGH (group 4) resulted in no change in the levels of either of these proteins (not shown), each of the microsphere formulations and the osmotic pump induced sustained elevated levels of both proteins. Consistent with the shorter sustained level of rhGH from formulation I, the IGF-I and IGFBP-3 levels were elevated for a shorter duration relative to the other formulations. In addition, as with the rhGH levels in the animals receiving formulation III, there was an initial elevation followed by a dip in the levels of these two proteins. Notably, continuous administration of rhGH by either the osmotic pump or the microsphere formulations resulted in significantly higher serum concentrations of both of these proteins than when an equivalent overall dose of protein as solution was administered by daily injections. The continuous concentration of rhGH that appeared necessary to sustain an elevation in IGF-I was approximately 5 ng/ml.

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Serum levels of total IGF-I. A, Groups 5 and 6; B, Groups 1 to 3. Symbols are as in the legend to figure 1 with the addition that group 5 (animals receiving daily injections of rhGH solution) are . The IGF-I concentrations were significantly different between groups 1 and 5 on days 4 to 8 (P  .01), between groups 2 and 5 on days 4 to 20 (P  .04), and between groups 5 and 6 on days 2 to 14 (P  .05).

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Serum levels of IGFBP-3. A, Groups 5 and 6; B, Groups 1 to 3. Treatment groups and symbols as are in figure 3. The IGFBP-3 concentrations were significantly different between groups 1 and 5 on days 4 and 6 (P  .05), between groups 2 and 5 on day 6 (P = .002) and between groups 5 and 6 on days 2-20 (P  .05).

Effect of sequential administration of microencapsulated rhGH. Because formulation II gave the most consistent and longest lasting levels of rhGH, IGF-I and IGFBP-3, the effect of three monthly doses of this formulation was evaluated. Juvenile monkeys were given the same dose per body weight as in the previous experiment, and the levels of rhGH (Figure 5A) and IGF-I (Figure 5B) were measured. (Because changes in IGFBP-3 levels paralleled those of IGF-I in the first experiment, IGFBP-3 levels were not measured in this experiment). Levels of rhGH after each of the three doses were similar to each other and were maintained between 10-20 ng/mL throughout most of the three month period and above the pre-dose level throughout the entire period. As predicted from the results of the previous experiment, thrice weekly doses of an equivalent amount of rhGH in solution resulted in little, if any, elevation of IGF-I levels.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Serum levels of rhGH and IGF-I in monkeys receiving three monthly injections of rhGH microsphere formulation II. Microspheres (50 mg/kg) were administered as in the legend to figure 1 on days 0, 28 and 56. A, hGH levels. B, IGF-I levels () of monkeys receiving monthly microsphere administrations. Animals receiving thrice weekly injections of rhGH solution are also shown () in which the total amount of protein administered as solution and as microspheres was equivalent (7.5 mg/kg/mo; 22.5 mg/kg total).

Analysis of microsphere injection sites. To determine the local effect of the microspheres, the sites of injection of the animals receiving each of the three formulations were examined histologically. The injection site from the animals receiving formulations I and II were surgically recovered at day 61 after injection whereas the site from animals receiving formulation III, because of its delayed release pattern, were recovered at day 84. Injection sites of all four animals receiving formulation I were approximately 12 mm in diameter and up to 4-mm thick (not shown). These contained a small amount of polymer and rhGH immunoreactive material and were surrounded by a slight foreign body inflammation. In contrast, the reaction at injection sites of animals receiving formulation II was less pronounced; the injection site could be identified in only one of the four animals and those in the other three animals had completely resolved. This site was smaller (approximately 4 × 0.1 mm) and contained less polymer than the sites of animals receiving formulation I. A few small cystic spaces, less than 100 µm in diameter, were surrounded by a mild foreign body inflammatory reaction and a small amount of rhGH immunoreactive material was present. Injection sites of two of the four animals receiving formulation III could not be identified. The other two showed traces of immunoreactive rhGH and minimal inflammation. The size of the immunoreactive area was approximately 5 × 0.2 mm.

View larger version (147K): [in this window] [in a new window]   Fig. 6.   Macroscopic (A-D) and microscopic (E-H) appearance of s.c. tissue at the injection sites from monkeys receiving microsphere formulation II. Monkeys received three injections (on days 0, 28 and 56) of microspheres (see legend to fig. 5) and all of the injection sites were collected on day 92. Photographs of injection sites from one of the animals, which are representative of the group, are shown. A and E show the site at which vehicle was injected at day 0 (hence the site was analyzed 92 days after injection); B and F show the site at which microspheres were injected at day 0 (hence the site was analyzed 92 days after injection); C and G show the site at which microspheres were injected on day 28 (hence the site was analyzed 64 days after injection); and D and H show the site at which microspheres were injected on day 56 (hence the site was analyzed 36 days after injection). The macroscopic and microscopic appearance of the injection site 92 days after injection is normal (cf. A and B, and E and F). The injection site 64 days after injection is macroscopically normal (C) whereas the microscopic analysis (G) reveals a condensed focus of macrophages and lymphocytes (arrow) within the s.c. connective tissue. The macroscopic appearance of the injection site 36 days after injection (D) shows a raised area approximately 5 × 7 × 1 mm (outlined by small arrowheads) and an adjacent area of reddish discoloration (large arrow). Microscopically (H), this injection site consists chiefly of a discrete mass of macrophages and multinucleated giant cells. Cytoplasmic vacuolar structures contain ingested polymer. The bar in A indicates 0.5 cm (same scale for B-D) and that in E indicates 200 microns (same scale for F-H). The microscopic sections were stained with H&E. Subcutaneous connective tissue is indicated by  and adipose tissue by *.

Immunogenicity of microsphere formulations. The immunogenicity of the rhGH solution and rhGH released from the microspheres was evaluated by measurement of anti-hGH antibodies. There is potential for anti-hGH antibody formation in rhesus monkeys because the monkey and human growth hormone sequences differ by four amino acids (Li et al., 1986). Seroconversion to rhGH was determined by incubating sera with radiolabeled rhGH followed by precipitation with polyethylene glycol. The results, as well as the endpoint titers of the positive sera, are shown in table 3.

	Discussion

Top Abstract Introduction Methods Results Discussion References

Our data demonstrate that each of the rhGH microsphere formulations provided sustained release and a sustained biological effect. This occurred with all three formulations, each of which elevated IGF-I levels for 3 wk or more. In contrast, even daily injections of the same total amount of rhGH in solution gave a lower increase in the level of IGF-I. However, there were differences in the profile of rhGH release from the three different microsphere formulations and these differences are consistent with the differences in the polymer comprising the three formulations. For example, although both formulations I and II show a spike in rhGH concentration followed by 3 or more wk of sustained serum levels, formulation III, after the initial release, showed a dip in the rhGH serum levels and a concomitant dip in the IGF-I and IGFBP-3 levels. This was followed by a period of an increased release rate and elevated IGF-I and IGFBP-3 levels. Release of encapsulated molecules from PLGA microspheres occurs by two principal mechanisms: 1) release of the drug by diffusion through pores formed in the polymer matrix after hydration and 2) release of the drug as the polymer hydrolyzes and the microsphere degrades (Cleland, in press). The initial spike in rhGH serum levels (0-48 hr after administration), observed with all three formulations, is principally due to protein release by diffusion whereas the remaining release is due to both diffusion and polymer degradation. In general, the higher the molecular weight of the polymer the longer it takes to degrade and release the protein. The delay in the release of rhGH observed after treatment with formulation III can be explained by the higher molecular weight of the polymer used in this formulation.

Our results in the rhesus monkeys showed that rhGH has similar immunogenicity when delivered from microsphere formulation II or when delivered by multiple injections as a protein solution. This was confirmed in the rhGH transgenic mice, which did not develop antibodies after administration of either microspheres or protein solution. In addition, the immunogenicity was not increased when three monthly injections were given. This is significant because the probability of an immune response is enhanced when an antigen is formulated as a depot formulation and given multiple times over a several week period, e.g., as in vaccination (Cleland, 1995, Janeway and Travers, 1994). Our results show that, although it acts as a depot, this microsphere formulation does not act as an adjuvant even after repeated administration. Thus, although PLGA formulations of vaccines have been shown to display enhanced immunogenicity (Cleland et al., in press, Eldridge et al., 1991, Hazrati et al., 1993, O'Hagen et al., 1993), our results show that this is not the case for homologous proteins.

Our results, which are similar to previous studies on the local response to PLGA microspheres (Visscher et al., 1985, 1987), showed that there were inflammatory cells and a foreign body reaction at the site of microsphere injection that resolved between 2 and 3 mo. There were no microscopically or macroscopically visible lasting effects, including any lasting fibrosis or scarring, even after repeated microsphere administration. Given that the dose of rhGH was relatively high (24 mg), it is likely that any amount of nonnative protein would have elicited a more severe response. This is particularly true after the third dose at which time the prior two doses would have primed such an immune response.

The daily approved dose of rhGH for growth hormone deficient children is 0.026 to 0.043 mg/kg that, over a 1-mo period, equals between 0.8 and 1.3 mg/kg. The amount of protein given in the form of microspheres in our experiments was higher, 7.5 mg/kg (50 mg/kg of microspheres). Although it is not yet known what dose of microencapsulated rhGH will be necessary to induce a therapeutic effect in humans, our data in monkeys (figs. 2 and 3) indicate that elevated IGF-I levels are induced by hGH serum levels of 5 ng/ml or more. Although there is high inter-individual variability, the threshold level for the biological effect of hGH in humans appears to range from 1 to 5 ng/ml (Daughaday, 1995). Given that the clearance of this protein is slower in humans than in monkeys (2.1 vs. 3.9 ml/hr/kg) (Moore and Wroblewski, 1992), it is reasonable to expect that a lower dose per kilogram in humans will elicit similar levels of hGH, IGF-I and IGFBP-3. This prediction is supported by the fact that about 1 mg/day of hGH is released from the pituitary of a growing child (Daughaday, 1995, Pearlman and Bewley, 1993) and only 2-fold less (approximately 0.4 mg/day) is released, after the initial release, from the microspheres in our study.

Our data show that delivery of rhGH from the microspheres resulted in differences between frequent injections of protein solution. First, a lower maximum serum concentration is obtained with an equivalent dose of protein. This would be an advantage for proteins other than rhGH with dose-limiting toxicity, such as cytokines. Another difference is that our data showed a greater effect (i.e., IGF-I levels) when the protein was delivered in a continuous fashion from the microspheres than when the same dose was delivered by frequent injections. This suggests that a lower dose of rhGH, when delivered in a sustained fashion, may be sufficient to achieve the same therapeutic effect as when administered by serial injections of protein in solution. This has been seen, for example, with an LHRH agonist to treat prostate cancer in which the dose is 1 mg/kg when given daily in solution (30 mg/kg/month) and 7.5 mg/kg (four-fold lower) when given monthly as the PLGA formulation (Arky, 1996). The eventual clinical utility of an injectable sustained-release formulation of rhGH is yet to be determined and such an investigation will require a careful evaluation of the dose and the frequency of administration to insure that the proper level of IGF-I is achieved and that undesireably high levels (e.g., those with the potential to cause acromegaly) are not maintained. Although sustained-release formulations will not be appropriate for all therapeutic proteins, our results with rhGH gives encouragement that formulations of proteins can be developed that induce a sustained biological effect.