European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 409426 www.elsevier.

com/locate/ejpb

Review article

In situ-forming hydrogelsreview of temperature-sensitive systems

Abstract In the past few years, an increasing number of in situ-forming systems have been reported in the literature for various biomedical applications, including drug delivery, cell encapsulation, and tissue repair. There are several possible mechanisms that lead to in situ gel formation: solvent exchange, UV-irradiation, ionic cross-linkage, pH change, and temperature modulation. The thermosensitive approach can be advantageous for particular applications as it does not require organic solvents, co-polymerization agents, or an externally applied trigger for gelation. In the last 2 decades, several thermosensitive formulations have been proposed. This manuscript focuses on aqueous polymeric solutions that form implants in situ in response to temperature change, generally from ambient to body temperature. It mainly reviews the characterization and use of polysaccharides, N-isopropylacrylamide copolymers, poly(ethylene oxide-b-propylene oxide-bethylene oxide) (poloxamer) and its copolymers, poly(ethylene oxide)/(D ,L -lactic acid-co-glycolic acid) copolymers, and thermosensitive liposome-based systems. q 2004 Elsevier B.V. All rights reserved.
Keywords: Thermosensitivity; Implants; In situ; Injectables; Drug delivery

1. Introduction In the past few years, an increasing number of in situforming systems have been reported in the literature for various biomedical applications, including drug delivery, cell encapsulation, and tissue repair. These systems are injectable uids that can be introduced into the body in a minimally invasive manner prior to solidifying or gelling within the desired tissue, organ, or body cavity. Injectable gel-forming matrices offer several advantages over systems shaped into their nal form before implantation. For example, injectable materials do not require a surgical procedure for placement (and withdrawal if not biodegradable), and various therapeutic agents can be incorporated by simple mixing. When they are used to ll a cavity or a defect, their owing nature enables a good t. In situ implant formation can occur as a result of either a physical or chemical change of the system.
* Corresponding author. Canada Research Chair in Drug Delivery, Faculty of Pharmacy, University of Montreal, P.O. Box 6128 Succ CentreVille, Montreal (Que) H3C 3J7, Canada. Tel.: 1-514-343-6455; fax: 1514-343-7738. E-mail address: [email protected] (J.-C. Leroux). 0939-6411/$ - see front matter q 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.ejpb.2004.03.019

There are several possible mechanisms leading to in situ implant formation. The solvent exchange approach consists of dissolving a water-insoluble polymer in a water-miscible, biocompatible solvent. Upon contact with body uids, the solvent diffuses out of the polymer while water permeates the liquid polymer matrix. Due to its insolubility in water, the polymer precipitates, resulting in the formation of a solid polymeric implant [1 5]. However, incomplete implant formation can be observed in vivo resulting in a high initial release and local or systemic toxicity. Also, the organic solvent used to solubilize the polymer can physically denaturate labile compounds such as proteins. Photopolymerization has also been proposed to prepare in situ implants. This approach has been taken to produce depot formulations [6], biological adhesives for soft tissues [7 10], and orthopaedic biomaterials [11]. However, photopolymerization requires the presence of a photoinitiator at the gelation site, which can be toxic. Furthermore, the penetration capacity of the radiation source limits the number of application sites, and the reaction can evoke enough heat to damage surrounding tissues. In situ-forming systems that do not require organic solvents or copolymerization agents have gained increasing attention. These are liquid aqueous solutions before administration, but gel under physiological conditions.

410

E. Ruel-Gariepy, J.-C. Leroux / European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 409426

Gelation can occur in situ by ionic cross-linking [12,13] or after a change in pH [14,15] or temperature. The latter approach exploits temperature-induced phase transition. Some polymers undergo abrupt changes in solubility in response to increases in environmental temperature (lower critical solution temperature, LCST). This phase separation is generally viewed as a phenomenon governed by the balance of hydrophilic and hydrophobic moieties on the polymer chain and the free energy of mixing [16 18]. The free energy of association varies with enthalpy, entropy and temperature DG DH 2 TDS: As the positive enthalpy term DH is smaller than the entropy term DS; an increase in temperature results in a larger TDS; making DG negative and favoring polymer chain association. The temperature dependence of certain molecular interactions, such as hydrogen bonds and hydrophobic effects, contribute to phase separation. At the LCST, hydrogen bonding between the polymer and water becomes unfavorable, compared to polymer polymer and water water interactions, and an abrupt transition occurs as the solvated macromolecule quickly dehydrates and changes to a more hydrophobic structure [18,19]. Alternatively, some amphiphilic polymers, that self-assemble in solution, show micelle packing and gel formation because of polymer polymer interactions when temperature is increased [20]. The ideal system would be a solution that is a freeowing, injectable liquid at ambient temperature. It should then gel at body temperature with minimal syneresis. Moreover, loading with drugs or cells should be achieved by simple mixing. When administered parenterally, these systems should exhibit a pH close to neutrality and should be bioresorbable. This paper focuses on polymeric solutions that can form implants in situ in response to temperature change, from ambient to body temperature. It mainly reviews the characterization and use of polysaccharides, N-isopropylacrylamide (NIPAM) copolymers, poly(ethylene oxide-b-propylene oxide-b-ethylene oxide) (PEO-PPOPEO) and its copolymers, poly(ethylene oxide)/(D ,L -lactic acid-co-glycolic acid) (PEO/PLGA) copolymers, and thermosensitive liposome-based systems (Fig. 1).

2. Polysaccharides 2.1. Cellulose derivatives Thermoreversible gels can be prepared with naturally occurring polymers. Most natural polymer aqueous solutions form a gel phase when their temperature is lowered. Classic examples of natural polymers exhibiting a sol gel transition include gelatin and carrageenan. At elevated temperatures, these polymers adopt a random coil conformation in solution. Upon cooling, a continuous network is formed by partial helix formation [21,22]. Some cellulose derivatives are an exception to this gelation mechanism. At low concentrations (1 10 wt%), their aqueous solutions are

liquid at low temperature, but gel upon heating. Methylcellulose (Fig. 1A) [23,24] and hydroxypropyl methylcellulose (HPMC) (Fig. 1B) [24] are typical examples of such polymers. Methylcellulose solutions transform into opaque gels between 40 and 50 8C [23,24], whereas HPMC shows phase transition between 75 and 90 8C [24]. These phase transition temperatures can be lowered by chemical or physical modications. For example, NaCl decreases the transition temperature of methylcellulose solutions to 32 34 8C [23,24]. Similarly, by reducing the hydroxypropyl molar substitution of HPMC, its transition temperature can be lowered to , 40 8C [24]. Gelation of methylcellulose or HPMC solutions is primarily caused by the hydrophobic interaction between molecules containing methoxy substitution. At low temperatures, the macromolecules are hydrated, and there is little polymer polymer interaction other than simple entanglement. As the temperature is raised, the polymers gradually lose their water of hydration, which is reected by a decline in relative viscosity. Eventually, when sufcient but not complete dehydration of the polymer occurs, polymer polymer associations take place, and the system approaches an innite network structure, as reected experimentally by a sharp rise in relative viscosity [23]. This sol gel transformation has been exploited to design in situ gelling systems. Tate et al. [25] evaluated methylcellulose-based constructs as potential tissue engineering scaffolds for the repair of brain defects. These systems exhibited low viscosity at 23 8C and formed soft gels intracerebrally at 37 8C. The gels were biocompatible both in the presence of cultured cells and in the injured rat brain. Aqueous solutions of ethyl(hydroxyethyl)cellulose (EHEC) (Fig. 1C) also exhibit thermosensitive behavior. However, their viscosity decreases with temperature, which is not appropriate for the preparation of in situ-forming implants. At the end of the 1980s, Carlsson et al. [26,27] reported that the addition of an ionic surfactant, like sodium dodecyl sulphate or cetyl triammonium bromide, to semidilute (1 4 wt%) EHEC solutions completely changed their thermal behavior. These systems underwent sol gel phase transition upon heating from room temperature to 30 40 8C, resulting in the formation of stiff and clear gels. The rheological properties of such gels were further investigated by Nystrom et al. [28]. The surfactant was found to interact with EHEC by a strongly cooperative process implying the formation of micelle-like surfactant clusters on the polymer [27,29,30]. Binding increased with rising temperature, and gelation was attributed to the ability of micelle-like clusters on the polymer chain to couple with segments on other chains [27,29]. Scherlund et al. [31] evaluated the EHEC/surfactant system for the local delivery of anesthetic agents to the periodontal pocket. They incorporated small amounts of lidocaine and prilocaine into the solution without affecting gelation behavior. The tested formulations showed

 E. Ruel-Gariepy, J.-C. Leroux / European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 409426

411

Fig. 1. Chemical formulas of hydrogel-forming polymers.

412

E. Ruel-Gariepy, J.-C. Leroux / European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 409426

Fig. 1 (continued )

sustained drug release over a minimum of 60 min, making them interesting for short-term pain control. From a toxicological point of view, the need for inclusion of an ionic surfactant in such a formulation may, however, impair its clinical development.

2.2. Xyloglucan Xyloglucan (Fig. 1D), a polysaccharide derived from tamarind seed, forms thermoresponsive gels in water, under certain conditions. Xyloglucan is composed

 E. Ruel-Gariepy, J.-C. Leroux / European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 409426

413

of a (1-4)-b-D -glucan backbone chain (GLU) which presents (1-6)-a-D -xylose branches (XYL) partially substituted by (1-2)-b-D -galactoxylose (GAL). Tamarind seed xyloglucan is composed of three units of xyloglucan oligomers with heptasaccharide, octosaccharide and nanosaccharide, which differ in the number of galactose side chains. When xyloglucan is partially degraded by b-galactosidase, the resultant product exhibits thermally reversible gelation in dilute aqueous solutions. Such behavior does not occur with native xyloglucan. Gelation is only possible when the galactose removal ratio exceeds , 35% [32]. The transition temperature is inversely related to polymer concentration [33] and the galactose removal ratio [32]. For example, the sol gel transition of xyloglucan was shown to decrease from 40 to 5 8C when the galactose removal ratio increased from 35 to 58%. Xyloglucan gels have been evaluated for the rectal delivery of indomethacin in rabbits [33]. They provided a broader absorption peak and longer residence time than commercial suppositories. Moreover, morphological studies of rectal mucosa after a single administration showed no evidence of tissue damage. Intraperitoneal administration of mitomycin C in a 1.5-wt% xyloglucan gel to rats resulted in a broad concentration time prole, as opposed to a narrow peak and rapid disappearance from the peritoneal uid and plasma when the drug was given as a solution [34]. In two other studies, the gels were investigated as vehicles for the oral delivery of indomethacin [35] and theophylline [36]. The bioavailability of indomethacin from xyloglucan gels was increased approximately three-fold compared to the control suspension. Likewise, theophylline bioavailability was 1.7 2.5 times higher than that of the commercially available oral, sustained-release liquid dosage form. Xyloglucan formulations were also assessed for ocular delivery of pilocarpine, using poloxamer 407 as a positive thermosensitive control [37]. The 1.5 wt% xyloglucan formulation enhanced the miotic response to a degree similar to that of a 25 wt% poloxamer 407 gel. More recently, xyloglucan gels were evaluated for the percutaneous administration of nonsteroidal anti-inammatory drugs [38]. After topical application, the xyloglucan formulations performed better than poloxamer 407 gels in improving the bioavailabilities of ibuprofen and ketoprofen. As for cellulose derivatives, xyloglucan solutions gel at low concentrations (1 2 wt%), and this may be advantageous from a toxicological viewpoint as the amount of administered polymer is low. In addition, xyloglucan is approved for use as a food additive. However, its relatively low transition temperature (22 27 8C) makes handling at room temperature problematic. 2.3. Chitosan and glycerophosphate Chenite et al. [39,40] developed a novel approach to produce thermally sensitive neutral solutions based on

chitosan/polyol salt combinations. Chitosan (Fig. 1E) is the deacetylated derivative of chitin, a natural component of shrimp and crab shells. It is a biocompatible, pH-dependent cationic polymer, which is soluble in water up to pH 6.2. Basication of chitosan aqueous solutions above this pH leads to the formation of an hydrated gel-like precipitate. Phase separation ensues from the neutralization of chitosan amine groups and the consequent elimination of repulsive interchain electrostatic forces, which subsequently allow for extensive hydrogen bonding and hydrophobic interactions between chains. pH-gelling, cationic chitosan solutions have been transformed into thermally sensitive, pHdependent, gel-forming systems by the addition of polyol salts (e.g. b-glycerophosphate, GP). These formulations possess a neutral pH, remain liquid at or below room temperature, and form monolithic gels at body temperature. The stability of the sol at room temperature and the gelation time increase as the chitosan degree of deacetylation decreases [41]. Like other polysaccharide systems, the gels are obtained at low polymer concentrations (, 2 wt%). Solubility at low temperatures is probably due to hydration of the chitosan promoted by GP. Upon heating, the chitosan chains lose their water of hydration, bonding between chains can occur and gelation proceeds. Three types of interactions may be involved in the gelation process: (1) electrostatic attraction between the ammonium groups of chitosan and the phosphate group of GP; (2) hydrogen bonding between polymer chains as a consequence of reduced electrostatic repulsion after neutralization of the chitosan solution with GP; and (3) chitosan chitosan hydrophobic interactions. Readers are referred to the original papers for more details on the gelation mechanism of chitosan/GP solutions [40,41]. In parallel, our group evaluated the chitosan/GP system for pharmaceutical applications [41 43]. Stability studies showed that solutions made with 84% deacetylated chitosan could be stored for 3 months at 4 8C without apparent change in viscosity. In vitro release experiments revealed that the system could deliver macromolecules (Mw 12,000 148,000) over a period of several hours to a few days [41]. However, due to high gel porosity, low molecular weight hydrophilic compounds (, 1000 g/mol) were completely released within hours. To achieve sustained delivery that would be independent of the drugs molecular weight, hydrophilic compounds were rst loaded into liposomes, which were then mixed with the thermosensitive solution [42]. This approach substantially slowed the release rate (Fig. 2), and the release prole of the incorporated compound could be controlled by adjusting liposome characteristics, such as size and composition. Recently, we proposed to use the thermosensitive hydrogel for the local administration of paclitaxel, a hydrophobic anticancer drug, at tumor resection sites to prevent local tumor recurrence [43]. In vitro release proles demonstrated controlled drug delivery for over 1 month. In mice, intratumoral

414

E. Ruel-Gariepy, J.-C. Leroux / European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 409426

Fig. 2. In vitro cumulative release of carboxyuorescein, free (triangles) or encapsulated in 100- (circles) or 280-nm (squares) egg phosphatidylcholine/cholesterol liposomes, from the chitosan/GP gel in PBS at 37 8C. After 14 days, Triton X-100 was added to the release medium to disrupt the liposomes still entrapped in the gel and release the remaining carboxyuorexein. Each point represents the mean value ^ SEM n 5: The solid lines represent the simulated release kinetics. Reprinted with permission from Ref. [42].

injection of the paclitaxel-loaded hydrogel was as efcacious as four intravenous injections of the commercially available Taxolw formulation in inhibiting the growth of EMT-6 cancer cells, and proved to be less toxic (Fig. 3). The chitosan/GP system was also evaluated as a tool for cartilage repair. Hoemann et al. [44] reported that primary calf chondrocytes could proliferate in solidied chitosan/GP solutions both in vitro and in vivo. Mechanical testing of 3-week aged in vitro implants demonstrated the initiation of functional matrix deposition. After injection into bone defects in rabbits, the chitosan/GP solution adhered to both bone and cartilage. Later, a hybrid implant, made of the chitosan/GP solution and whole blood, was developed to improve cartilage healing [45]. This new system, named

CarGelw, was investigated in rabbits, sheep, and horses. The in vitro mixture of chitosan/GP with whole blood resulted in the accelerated formation of a more adhesive and voluminous clot scaffolding. In healing articular defects, specic repair responses were observed, including enhanced chemotaxis of bone marrow stromal cells, and temporarily increased vascularity of the repair tissue. No local or systemic toxic effects were seen in CarGelw-treated animals. Recently, human studies were undertaken with this new product. Preliminary biocompatibility experiments showed that the local injection of chitosan/GP solutions in the rat hind paw led to an acute inammatory response, which was inversely related to the chitosans degree of deacetylation [46]. This response was substantially more important than

Fig. 3. In vivo antitumor effect in Balb/c mice bearing an EMT-6 tumor. Saline 0.2 ml/day intravenous 4 days (closed circles, n 8), Taxolw 10 mg/kg/day intravenous 4 days (open circles, n 8), chitosan/GP solution 10 ml intratumoral (closed squares, n 8) and paclitaxel/chitosan/GP solution (64 mg/ml paclitaxel) 10 ml (40 mg/kg paclitaxel) intratumoral (open squares, n 8). Reprinted with permission from Ref. [43].

 E. Ruel-Gariepy, J.-C. Leroux / European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 409426

415

that observed after the transdermal injection of similar solutions in the rat dorsal region [47], suggesting that the extent of the inammatory response depended on the injection site. Hind paw injections may have elicited a more pronounced reaction because of differences in blood ow, tight geometry of the injection space, etc. Jarry et al. [47,48] evaluated the effects of sterilization on chitosan/GP solutions. Initially, steam sterilization experiments were conducted on chitosan in the dry state. However, it was noted that, upon dissolution in water and the addition of GP, viscosity of the solutions increased dramatically compared to non-sterilized samples, compromising injectability of the system. Accordingly, it was decided to sterilize the polymer in solution. Steam sterilization of chitosan solutions induced molecular weight reduction, which led to a decrease in dynamic viscosity, gelling rate, and mechanical strength of the chitosan/GP hydrogels. Thus, this sterilization process may be considered when high mechanical performance is not critical or essential for implant function, such as in the case of controlled drug delivery. In conclusion, chitosan/GP hydrogels constitute promising implants for pharmaceutical and bioengineering applications. Because of their high porosity, they are more adapted to the delivery of macromolecules and poorly water-soluble drugs.

3. N-isopropylacrylamide copolymers Poly(N-isopropylacrylamide) (PNIPAM) (Fig. 1F) is a non-biodegradable polymer with a LCST , 32 8C in water [17], and cross-linked gels of this material collapse around this temperature [49,50]. The PNIPAM LCST can be controlled by copolymerization with other monomers. The addition of hydrophilic monomers typically increases the LCST whereas the incorporation of more hydrophobic units has the opposite effect [51]. 3.1. Poly(N-isopropylacrylamide-co-acrylic acid) In the 1990s, Han and Bae [52] reported that an aqueous solution of high molecular weight P(NIPAM-co-AA) (Fig. 1G), synthesized in benzene (with 2 mol% AA in feed), showed reversible gelation at , 32 8C, above a critical concentration. Aqueous solutions of this copolymer did not just precipitate or result in a shrunken mass when the temperature was elevated above its LCST, like aqueous solutions of PNIPAM. A typical gelation process with a 5-wt% solution in phosphate-buffered saline (pH 7.4) is shown in Fig. 4A and B. The clear polymer solution (phase A) became cloudy as temperature increased to 27 8C (cloud temperature). At this temperature, the solution was still freely mobile (phase B). With further heating up to the gelation temperature (35 8C), the polymer solution became immobile (phase C). The initially formed gel was

translucent, and turned opaque with increasing temperature. At 43 8C (gel-shrinking temperature), the gel started to shrink and expel water (syneresis). Upon cooling, it reverted to the sol state with no hysteresis between sol gel and gel sol temperatures (Fig. 4B). An interesting feature was that the gel did not dissolve or change its hydration state upon the addition of water (no swelling). The authors speculated that as temperature increases, polymer chains with lower AA content precipitate rst, while those with higher AA content remain as partially collapsed globules or in a fully expanded coil state. The completely collapsed globules, entangled with the expanded coils, are thought to produce hydrophobic aggregates that form physical junctions. P(NIPAM-co-AA) solutions were investigated as cell matrices in rellable bioarticial pancreas [53 58]. The idea originated from the observation that islets of Langerhans tend to malfunction at some point after transplantation and have to be replaced. Also, aggregation of the islets can lead to necrosis. The thermosensitive solutions were proposed as an extracellular matrix for islets placed in an immunoprotecting pouch installed in diabetic patients (Fig. 5). An islet/polymer solution can be injected into the pouch where the polymer forms a gel, immobilizing the islets. To replace the islets, the polymer matrix is cooled below the critical temperature, causing the gel to redissolve. The solution is then withdrawn, and a fresh islet/polymer suspension is injected. It was shown that the copolymer efciently immobilized rat islets without impairing insulin secretion [53 55]. Compared to free islets, entrapment in the gel resulted in prolonged insulin secretion [55]. Moreover, P(NIPAM-co-AA)-entrapped rat islets responded faster and with greater magnitudes to changes in glucose concentration than islets in an alginate gel. The addition of glucagon-like peptide-1, a potent islet stimulant, was also investigated [56]. Its presence in the preparation signicantly enhanced insulin secretion without affecting islet viability. Similarly, a hemoglobin PEO conjugate was used, as it is believed that the low functionality and viability of immunoprotected islets are closely associated with low oxygen tension [57]. Insulin secretion functions, as well as islet viability, were signicantly enhanced by the addition of hemoglobin. The same approach was later taken to develop a biohybrid articial liver [59]. Hepatocytes cultured as spheroids in the thermoreversible matrix exhibited greater viability and enhanced liver-specic functions vs. control cells. These studies demonstrated the potential of the system as a temporary cell-supporting matrix, but the non-biodegradability of the polymer may limit its biomedical applications. 3.2. Poly(N-isopropylacrylamide)/poly(ethylene oxide) A new family of polymers that self-assemble to form gels in a thermoreversible fashion has been proposed recently by Lin and Cheng [60]. It consists of block linear and star

416

E. Ruel-Gariepy, J.-C. Leroux / European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 409426

Fig. 4. Phase transitions (A) and viscosity (ow time through a CannonFenske viscometer) of an aqueous solution of P(NIPAM-co-AA) as a function of temperature. Adapted with permission from Ref. [52].

copolymers with a central hydrophilic PEO segment and temperature-responsive PNIPAM terminal segments. Linear and star copolymers of PEO and PNIPAM form liquid aqueous solutions at room temperature that transform to relatively strong elastic gels upon heating. Multiple-arm copolymers yield gels via physical cross-links between aggregates of PNIPAM segments, whereas diblock copolymers gel by a micellar aggregation mechanism (see Section 4). The rheological properties of the hydrogels were found to depend on molecular architecture, with the star structure having four PNIPAM terminal segments showing superior qualities. These copolymer solutions presented low to moderate injection viscosities, high gel strengths, low degrees of syneresis, and rapid gelation kinetics. So far, the systems have not been investigated for biomedical purposes.

are ABA-type triblock copolymers composed of PEO (A) and PPO units (B). The poloxamer series covers a range of liquids, pastes, and solids, with molecular weights and ethylene oxide-propylene oxide weight ratios varying from 1100 to 14,000 and 1:9 to 8:2, respectively. Concentrated aqueous solutions of poloxamer form thermoreversible gels.

4. Poloxamer systems 4.1. Poloxamer (Pluronicw) The poloxamers (Fig. 1H) consist of more than 30 different non-ionic surface-active agents. These polymers
Fig. 5. Schematic diagram illustrating the concept for a rellable cell delivery system with a thermally reversible extracellular matrix. Reprinted with permission from Ref. [54].

 E. Ruel-Gariepy, J.-C. Leroux / European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 409426

417

The gelation mechanism of poloxamer solutions has been investigated extensively, but is still being debated. Ultrasonic velocity, light-scattering and small-angle neutron scattering measurements of aqueous poloxamer solutions have clearly indicated a micellar mode of association [61 66]. Micelle formation occurs at the critical micellization temperature as a result of PPO block dehydration [63,67]. With increasing temperature, micellization becomes more important, and at a denite point, micelles come into contact and no longer move. In addition, the formation of highly ordered structures, such as cubic crystalline phase, has been proposed as the driving force for gel formation [20,64 66], but this hypothesis has been questioned recently. Thus, packing of micelles and micelle entanglements may be possible mechanisms of poloxamer solution gelation with increased of temperature [68]. Poloxamer 407 (Pluronicw F127) was found to gel at a concentration of 20 wt% at 25 8C, which is less than that of the other members of the poloxamer series. At room temperature (, 25 8C), the solution behaves as a mobile viscous liquid, which is transformed into a semi-solid transparent gel at body temperature (37 8C). Preliminary toxicity data indicate that this copolymer is well tolerated [69]. Taken together, these results have prompted the use of poloxamer 407 in the design of medical, pharmaceutical, and cosmetic systems. Early studies evaluated poloxamer 407 thermosensitive solutions for the treatment of burns [69], topical administration of anticancer agents [70], and sustained delivery of drugs after extravascular parenteral injection [71]. Over the past 15 years, this copolymer has been investigated extensively for various applications, and it is not the intent of this paper to review them all. Some recent applications are reported in Table 1 and discussed below. After parenteral injection, poloxamer gels can prolong drug release compared to solutions, but the delivery period rarely exceeds a few days [72 76]. This characteristic makes poloxamer gels interesting for short-term therapies like pain management [74], infection treatment [73,76], and fertility control [75]. Besides injectables, other administration routes have been evaluated, such as rectal [77,78], vaginal [79,80], transdermal [81,82] and ophthalmic [83,84]. Poloxamer formulations generally increased drug residence time at application sites, resulting in improved bioavailability and efcacy (Fig. 6). Poloxamer 407 is usually regarded as non-toxic. After intramuscular injection in rabbits, poloxamers 238 and 407 displayed musculoirritancy/toxicity comparable to that of traditional intramuscular vehicles, such as saline and peanut oil [85]. However, other studies have reported systemic side effects. Wout et al. [86] demonstrated that poloxamer 407 injected intraperitoneally into rats (1.5 g/kg) resulted in sustained hypercholesterolemia and hypertriglyceridemia (. 96 h). Palmer et al. [87] obtained similar results in mice. It was suggested that the predominant mechanism for this effect was inhibition of heparin-releasable lipoprotein lipase [88]. Blonder et al. [89] examined whether lower but

clinically useful doses of poloxamer 407 gels induced hyperlipidemia in rabbits. The highest dose (137.5 mg/kg) signicantly increased serum triglycerides and cholesterol in both male and female rabbits, with peak values observed 2 days after injection. The lower doses (5.5 27.5 mg/kg) did not alter serum lipids. Hence, the amount of administered polymer should be kept to a minimum, especially when repeated dosing is required. Potential drawbacks of poloxamer gels include their weak mechanical strength, rapid erosion (i.e. dissolution from the surface), and the non-biodegradability of PEOPPO-PEO, which prevents the use of high molecular weight polymers that cannot be eliminated by renal excretion. To circumvent the biodegradability issue, new polymers were synthesized by linking together a few (usually 3) poloxamer 407 monomers via degradable carbonate linkage [90]. As the carbonate linkages were hydrolyzed under physiological conditions, the hydrogel degraded into soluble poloxamer 407 units and carbonate. Variation of the polymer concentration enabled modication of the gels dissolution time (25 80 days). Another interesting feature is that oligomerization decreased the critical gel concentration by two-fold. To address their poor mechanical properties, Cohn et al. [91] proposed two different strategies: (i) to polymerize poloxamers using hexamethylene diisocyanate as a linker; and (ii) to use phosgene to covalently bind PEO and PPO chains. These new materials achieved viscosities that were at least 15 times higher than poloxamer 407 at 37 8C. Moreover, the critical gel concentration was lowered to 5 wt%. Also, a 30% poly(poloxamer 407)4 gel delivered drug over a 40-day-period vs. 7 days for poloxamer 407. 4.2. Poloxamer/poly(acrylic acid) High molecular weight poly(acrylic acid) (PAA) is a bioadhesive polymer known to adhere to the hydrated mucosal cells coating the eye, nose, lungs, gastrointestinal tract, and vagina. It is often incorporated into drug delivery systems that come in contact with mucosal surfaces in order to prolong the residence time. As described earlier, random copolymers of NIPAM and AA can gel with an increase in temperature (Section 3.1). However, thermosensitivity is lost at physiological pH when the AA content is increased so as to obtain bioadhesive properties. To circumvent this problem, graft copolymers were developed [92]. At pH 7.4, the graft copolymers exhibited phase transition at a temperature similar to that of PNIPAM. Nevertheless, NIPAM copolymers are not generally recognized as safe (GRAS). This fact prompted investigators to synthesize similar copolymer structures using GRAS polymers. Since poloxamer and PAA components have approved regulatory status, it is expected that their graft-copolymers would be safe. Synthesis of the graft copolymer, PAA-g-poloxamer (Fig. 1J), was reported by Hoffman and co-workers [93 95]. The proposed synthetic route involved a three-step procedure with several intermediate steps of purication,

418

E. Ruel-Gariepy, J.-C. Leroux / European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 409426

Table 1 Poloxamer temperature-sensitive solutions for drug delivery applications: some examples since 1997 Poloxamer 407 Concentration (wt%) 36 Drug Recombinant human growth hormone Objective of the study Controlled release of human growth hormone following intramuscular or subcutaneous administration. Subcutaneous delivery of peptides and proteins having short half-lives. Prolonged residence time of vancomycin in a body site with a high infection risk. Controlled release of ibuprofen for epidural analgesia. Intratumoral administration of paclitaxel. Intramuscular sustained release of deslorelin and GnRH to induce the release of luteinizing hormone and formation of luteal tissue in cattle. Sustained release gel formulation of ceftiofur for treating foot infections in cattle. Increased bioavailability using an in situ gelling and mucoadhesive liquid suppository. Increased bioavailability using an in situ gelling and mucoadhesive liquid suppository. Prolonged antifungal effects using an in situ gelling and mucoadhesive vaginal gel. Enhanced ocular bioavailability of timolol maleate. Development of a thermosetting gel with a suitable phase transition temperature for ocular delivery. Enhanced efcacy of piroxicam following percutaneous absorption. Poloxamer gels as release vehicles for percutaneous administration of fentanyl. Ref. [72]

407 407

20 or 30 25

Insulin Vancomycin

[132] [73]

407 407 407

25 20 25

Ibuprofen Paclitaxel Deslorelin or GnRH

[74] [133] [75]

407

25, 30 and 35

Ceftiofur

[76]

407 and/or, 188 and additives

15, 15 and 20

Acetaminophen

[77]

407 and 188 and additives

15

Propranolol

[78]

407, 188

15, 15 and 20

Clotrimazole

[79,80]

407 and thickening agents 407, 188

15, 20 or 25 21

Timolol maleate None

[83] [84]

407 46

Piroxicam Fentanyl

[81] [82]

making the process potentially problematic on an industrial scale. Subsequently, the synthesis of another graft copolymer, i.e. poloxamer-g-PAA (Fig. 1I), was described by Bromberg et al. [96 98]. This single-step synthetic procedure was reproducible and scaleable. Work on this copolymer (Smart Hydrogele) has been reviewed recently by Bromberg and Ron [99]. While the PNIPAM-g-PAA graft copolymers precipitated out of solutions when temperature was raised above the LCST, dilute solutions (0.5 3 wt%) of PAA-g-poloxamer and poloxamer-g-PAA formed clear gels when the temperature increased from 4 to 37 8C at pH 7.4 [94,97,98] (Fig. 7). Moreover, the critical gel concentration was considerably lower than that of the parent poloxamer. Bromberg thoroughly studied the rheological properties of poloxamer-g-PAA hydrogels and showed that temperature-induced sol gel transition and gel strength were concentration-dependent [100 102].

An increase in polymer concentration resulted in lower transition temperature and higher gel strength. Besides polymer concentration, pH and salts inuenced gelation properties [102]. With rising pH, the onset of gelation was shifted to lower temperatures, and gel strength increased. The data obtained by Bromberg suggested that gelation was driven by the formation of micelles that act as thermoreversible cross-links. The onset of gelation in poloxamer-g-PAA solutions occurred approximately at the critical micellization temperature of the parent poloxamer [97]. Polymer aggregation indeed coincided with the onset of the relative viscosity increase [101]. A negative change in heat capacity upon gelation suggested a decrease in the exposure of PPO segments to water [102]. The bioadhesive nature of this system [103] makes it interesting for a wide variety of applications. Hydrogels of

 E. Ruel-Gariepy, J.-C. Leroux / European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 409426

419

Fig. 6. Timolol concentration in aqueous humor after instillation of timolol aqueous solution (circles), timolol-poloxamer solution (squares) and timololpoloxamer-methylcellulose solution (triangles). Reprinted with permission from Ref. [83].

poloxamer-g-PAA have been proposed for vaginal [104] and topical [105] drug administration. The pharmacokinetic prole of estradiol after vaginal delivery of poloxamer-gPAA formulation was equivalent to that of Estracew vaginal

cream, which contains ve times more drug. Potts et al. [106] examined this hydrogel in the treatment of gastrooesophageal reux by measuring its adherence to the oesophageal mucosa. This study revealed that 15% of

Fig. 7. Phase diagram for aqueous solutions of poloxamer-g-PAA at pH 7.0. Tgel and cmt stand for the gel point and critical micellization temperature determined by the WinterChambon method and DSC, respectively. Reprinted with permission from Ref. [102].

420

E. Ruel-Gariepy, J.-C. Leroux / European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 409426

the administered dose displayed prolonged retention in the oesophagus. In another investigation, it was shown that Smart Hydrogele could sustain the release of luteinizing hormone-releasing hormone (LHRH) and human insulin [107]. At a polymer concentration of 5%, the release kinetics of LHRH followed a zero-order rate. Finally, preliminary animal toxicological studies showed the nontoxic nature of the copolymer administered topically, orally or to the eye [99].

5. Poly(ethylene oxide)/poly(D ,L -lactic acid-co-glycolic acid) Jeong and co-workers described different thermosensitive, biodegradable hydrogels based on poly(lactic acid). Block copolymer solutions of PEO and poly(L -lactic acid) were shown to be in the sol state at 45 8C, and in the gel state at body temperature [108]. However, the need to heat the solution limits the nature of the drugs that can be incorporated in this delivery system, and makes the injection procedure not practical. Later, PEO-b-(D ,L -lactic acid-co-glycolic acid)-b-PEO (PEO-PLGA-PEO) (Fig. 1K) triblock copolymer solutions were found to form, at room temperature, a free-owing sol that became a transparent gel at 37 8C [109,110]. PEO-PLGA-PEO (550-2810-550) exhibited a critical gel concentration of approximately 16 wt%. PEO-PLGA-PEO polymers are amphiphilic and yield micelles in aqueous environments with the hydrophilic PEO occupying the outer shell region, and the hydrophobic PLGA constituting the inner core. Micelle formation was indeed conrmed by 13C-NMR and dye solubilization studies [109]. It was shown that the gels formed via a micellar aggregation mechanism. Abrupt changes in micellar diameter and aggregation number were demonstrated above 30 8C, and the second virial coefcient decreased sharply at this temperature, indicating a sudden increase in polymer polymer attraction [109]. Other studies established that micelle micelle clusters were formed through hydrophobic interactions between core phases [111]. The latter mechanism differentiates these copolymers from poloxamers where gel formation is only associated with simple micellar entanglements. The gelling properties are inuenced by PLGA or PEO proportions. The more hydrophobic the polymer, the lower the temperature and polymer concentration required to achieve the desired liquid crystalline structure [110]. In vitro studies revealed that the gel could sustain the release of ketoprofen (hydrophilic drug) and spironolactone (hydrophobic drug) for 2 weeks and 2 months, respectively [112]. The subcutaneously injected solution formed a hydrogel in situ [113]. The latter maintained a threedimensional shape at the implant site, indicating that gelation was fast. The implant was strong enough to be easily taken out with forceps and persisted for more than

1 month with little or no tissue irritation at the injection site. As opposed to poloxamers, PEO-PLGA-PEO gels do not erode rapidly. Furthermore, due to their PLGA segments, they can be hydrolyzed in vivo. Since the transition temperature of the triblock copolymers is quite sensitive to PEO length, it is difcult to modify polymer structure and still maintain gelation temperature around 37 8C. To overcome the molecular weight limitation, Jeong et al. [114 117] developed graft copolymers based on PEO and PLGA. As with the PEO-PLGA-PEO system, PEO-g-PLGA (Fig. 1L) and PLGA-g-PEO (Fig. 1M) micellized in water [114 116]. The critical gel concentrations of both polymers were approximately 16 [114] and 22 wt% [116], respectively. The applicability of this new material as an in situ-forming device was assessed after subcutaneous injection into rats [116]. Again, the implant was round-shaped, indicating fast gelation kinetics. In vivo, the PLGA-g-PEO depot persisted for about 3 months, whereas PEO-g-PLGA disappeared within a week. By mixing two polymers of PEO-g-PLGA and PLGA-g-PEO, the duration of depot could be adjusted between 1 week and 3 months. The grafted copolymer systems were evaluated for sustained insulin release and as a cell delivery matrix in cartilage repair [117]. Insulin depots prepared from a 50/50 mixture of PEO-g-PLGA and PLGA-g-PEO showed a hypoglycemic effect over 5 days vs. 16 days for PLGA-gPEO. Injection of a chondrocyte cell suspension in PLGA-gPEO into a cartilage defect effectively promoted cartilage repair in rabbits. Low molecular weight PLGA-PEO-PLGA copolymers also exhibit reverse thermal gelation properties at concentrations between 10 and 30 wt% [118]. This system, known as ReGelw, was evaluated as a pharmaceutical vehicle by Zentner et al. [119]. In vivo biodegradation studies in rats revealed that the formulation was completely resorbed from the injection site within 4 6 weeks. In vitro release experiments showed sustained paclitaxel release over 50 days. This was in sharp contrast to the complete release of paclitaxel from poloxamer 407 hydrogel within 1 day. Direct intratumoral injection of paclitaxel-loaded ReGelw resulted in slow drug clearance from the injection site with minimal distribution into any organ. The efcacy of ReGelw/paclitaxel (OncoGelw) against human breast tumor xenografts (MDA231) was also superior to that achieved with a maximum tolerated dose of the commercial paclitaxel product (i.e. Taxolw). Moreover, the ReGelw/ paclitaxel formulation was less toxic as treated animals showed no drug-related adverse effects, whereas in the Taxolw group, there was weight loss and two acute toxic deaths within 2 days of dosing. OncoGelw is MacroMeds lead product; it entered Phase II studies in the second quarter of 2003 [120]. The ReGelw system has also been tested for the parenteral delivery of insulin [121]. After one injection in rats, elevated plasma insulin levels were maintained for at least 15 days (Fig. 8).

 E. Ruel-Gariepy, J.-C. Leroux / European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 409426

421

Fig. 8. Plasma insulin levels of SpragueDawley rat after subcutaneous injection of the insulin/ReGelw formulation. Reprinted with permission from Ref. [121].

6. Thermosensitive liposomes as a physical barrier between reactive species In a series of studies, Messersmith and co-workers [122 126] exploited a rise in temperature to initiate a reaction cascade leading to the formation of a biomaterial

in situ. Their approach was based on the use of temperature-sensitive liposomes to compartmentalize reactive species (Fig. 9A). The physical barrier imposed by the vesicle membrane prevented the chemical reaction from proceeding. Destabilization of the lipid membrane upon an increase in temperature triggered

Fig. 9. (A) Schematic illustration of thermally triggered alginate cross-linking. Adapted with permission from Ref. [122]. (B) Thermally triggered gelation of TGase/brinogen/thermally sensitive, calcium-loaded liposome precursor uid. Reprinted with permission from Ref. [123].

422

E. Ruel-Gariepy, J.-C. Leroux / European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 409426

the reaction as reactive species escaped from the liposomes. The same strategy can be adopted to design thermoresponsive gels. Alginates are linear, water-soluble polysaccharides composed of 1,4-linked b-D -mannuronic and a-L -glucuronic acid units. One of the main biofunctional characteristics of alginates is their ability to form transparent or translucent gels in the presence of multivalent ions. Thus, thermoresponsive calcium-loaded liposomes were prepared and mixed with an alginate solution [123]. Ca2 sequestration within the liposomes prevented the ionic cross-linking of the polymer. The Ca-liposome/alginate mixture remained uid when stored at 20 8C. Gelation occurred upon heating the system above the liposome phase transition temperature due to the leakage of entrapped Ca2. At room temperature, however, after only 1 day of storage, a gradual increase in viscosity was observed, which eventually led to full gelation within 4 5 days. It was found that alginate triggered slow release of calcium from the liposomes at 20 8C. Although the lack of stability imposes constraints on the storage time of the mixture, it does not necessarily preclude the use of such a system, since the two components can be mixed prior to injection. Many biological processes are regulated by Ca2 concentration, indicating that Ca2 release can be exploited to produce different types of biomaterials. Protein-based hydrogels were prepared by taking advantage of Ca2-dependent enzymes to catalyze enzymemediated crosslinking of macromolecular substrates. One such family of Ca2-dependent proteins are the transglutaminases. These enzymes catalyze the formation of crosslinks between glutamine- and lysine-containing proteins. Calcium-loaded, thermally sensitive liposomes were used to activate hrFXIII, a calcium-dependent enzyme that catalyzes the crosslinking of brin during blood coagulation [123,124]. When calcium-loaded liposomes, brinogen, and hrFXIII were combined and heated to 37 8C, gelation occurred within 10 min (Fig. 9B). Sandborn et al. [125] reported a system comprising thermosensitive, calcium-loaded liposomes, hrFXIII, thrombin and an enzymatic substrate based on a four-arm PEO terminated with a 20mer peptide sequence derived from the g-chain of brin. Upon heating the precursor solution to 37 8C, a solid, enzymatically crosslinked hydrogel formed within 9 min. Gelation was not reversible as the material remained in a gelled state after cooling to room temperature. Another study investigated the potential of the system to trigger the self-assembly of FEK16 [126], a 16-amino acid peptide that is highly soluble in pure water, but forms selfassembled aggregates in the presence of NaCl, KCl, and CaCl2. Hydrogel formation was achieved when temperature-sensitive, calcium-loaded liposomes were suspended in aqueous solutions of FEK16 and subsequently warmed to 37 8C. The main advantage of using peptide substrates as hydrogel crosslinks lies in their biodegradability.

However, a possible immune response to the implant is a potential concern that should be addressed in the future.

7. Miscellaneous 7.1. Poly(organophosphazene) derivatives Recently, poly(organophosphazene) (PPZ) derivatives were shown to exhibit sol gel phase transitions as a function of temperature. PPZ bearing methoxy-PEO and amino acid esters as substituents were synthesized by Song et al. (Fig. 1N) [127,128]. The polymers were hydrolytically degradable and displayed a LSCT in the 25.2 98.5 8C range. The same group demonstrated that oligomeric cyclophosphazenes with proper orientation of substituents were also thermosensitive (Fig. 1P) [129]. PPZ bearing a-amino-vmethoxy-PEO and hydrophobic L -isoleucine ethyl ester as side groups exhibited reversible sol gel transition between 29 and 61 8C depending on the structure (Fig. 1O) [130]. Gelation was attributed to hydrophobic interactions between the side-chain fragments ( CH(CH3) CH2CH3) of L -isoleucine ethyl ester. 7.2. Poly(1,2-propylene phosphate) Lately, a new injectable temperature-sensitive hydrogel system was proposed by Wang et al. [131]. The system is based on a synthetic polyanion, poly(1,2-propylene phosphate) (Fig. 1Q), that can be crosslinked by calcium ions. Solutions of poly(1,2-propylene phosphate) in water did not exhibit temperature-dependent phase transition. In the presence of CaCl2, the polymer solutions remained liquid at room temperature, but gelled rapidly when heated. The sol gel transition temperature was a function of polymer and Ca2 concentrations. The in vitro release study of lysozyme followed zero-order kinetics after a lag time of 1 h. Release was accompanied with decross-linking of the gel, which was most likely due to Ca2 diffusion into the buffer. 8. Conclusion Over the last decade, an impressive number of novel, thermosensitive, in situ gel-forming solutions have been described in the literature. Each system has its own advantages and drawbacks. The choice of a particular hydrogel depends on its intrinsic properties and envisaged therapeutic use. For instance, the formation of a transparent gel is particularly important when ophthalmic applications are considered. Non-biodegradable gels could prove useful for administration routes other than parenteral. Poloxamer hydrogels perhaps represent the most extensively studied systems. However, despite the clinical acceptance of poloxamers as solubilizer and thickening agents, these polymers have not met initial expectations as pharmaceutical

 E. Ruel-Gariepy, J.-C. Leroux / European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 409426

423

and biomedical implants, mainly due to their non-biodegradability and inability to provide sustained drug delivery over more than just a few days. Recent developments in the design of poloxamer derivatives showed that some of these problems could be worked out. Polysaccharides usually demonstrate good biocompatibility and/or biodegradability, and their solutions are thermosensitive at low polymeric concentrations. However, these systems may not be adapted for the sustained release of hydrophilic, low molecular weight drugs because their large, water-lled pores permit rapid diffusion. On the other hand, they provide adequate scaffolds for cell growth and tissue repair. PEO/PLGA hydrogels are particularly attractive systems for pharmaceutical applications. They are biodegradable and generally have a good safety prole. Their composition can be tailored to provide drug delivery over weeks or months after parenteral extravascular administration.

[11]

[12]

[13]

[14] [15]

[16]

[17]

Acknowledgements Financial support from the National Sciences and Engineering Research Council of Canada, BioSyntech Inc. and the Canada Research Chair Program is acknowledged.

[18] [19]

[20]

References
[21] [1] R.L. Dunn, J.P. English, D.R. Cowsar, D.P. Vanderbilt, Biodegradable in-situ forming implants and methods of producing the same, US Patent 4,938,763, 1990. [2] M.L. Shively, B.A. Coonts, W.D. Renner, J.L. Southard, A.T. Bennett, Physico-chemical characterization of a polymeric injectable implant delivery system, J. Control. Release 33 (1995) 237243. [3] R.E. Eliaz, J. Kost, Characterization of a polymeric PLGA-injectable implant delivery system for the controlled release of proteins, J. Biomed. Mater. Res. 50 (2000) 388396. [4] B.O. Haglund, R. Joshi, K.J. Himmelstein, An in situ gelling system for parenteral delivery, J. Control. Release 41 (1996) 229235. [5] R.E. Eliaz, F.C. Szoka, Robust and prolonged gene expression from injectable polymeric implants, Gene Ther. 9 (2002) 1230 1237. [6] Y. An, J.A. Hubbell, Intraarterial protein delivery via intimallyadherent bilayer hydrogels, J. Control. Release 64 (2000) 205215. [7] K. Ono, Y. Saito, H. Yura, K. Ishikawa, A. Kurita, T. Akaike, M. Ishihara, Photocrosslinkable chitosan as a biological adhesive, J. Biomed. Mater. Res. 49 (2000) 289295. [8] M. Ishihara, K. Nakanishi, K. Ono, M. Sato, M. Kikuchi, Y. Saito, H. Yura, T. Matsui, H. Hattori, M. Uenoyama, A. Kurita, Photocrosslinkable chitosan as a dressing for wound occlusion and accelerator in healing process, Biomaterials 23 (2002) 833840. [9] M. Ishihara, K. Obara, T. Ishizuka, M. Fujita, M. Sato, K. Masuoka, Y. Saito, H. Yura, T. Matsui, H. Hattori, M. Kikuchi, A. Kurita, Controlled release of broblast growth factors and heparin from photocrosslinked chitosan hydrogels and subsequent effect on in vivo vascularization, J. Biomed. Mater. Res., 64A (2002) 551559. [10] K. Obara, M. Ishihara, T. Ishizuka, M. Fujita, Y. Ozeki, T. Maehara, Y. Saito, H. Yura, T. Matsui, H. Hattori, M. Kikuchi, A. Kurita, [22]

[23]

[24] [25]

[26]

[27]

[28]

[29]

[30]

[31]

Photocrosslinkable chitosan hydrogel containing broblast growth factor-2 stimulates wound healing in healing-impaired db/db mice, Biomaterials 24 (2003) 34373444. A.K. Burkoth, K.S. Anseth, A review of photocrosslinked polyanhydrides: in situ forming degradable networks, Biomaterials 21 (2000) 2395 2404. A. Rozier, C. Mazuel, J. Grove, B. Plazonnet, Gelrite: A novel, ionactivated, in-situ gelling polymer for ophthalmic vehicles. Effect on bioavailability of timolol, Int. J. Pharm. 57 (1989) 163 168. S. Cohen, E. Lobel, A. Trevgoda, Y. Peled, A novel in situ-forming ophthalmic drug delivery system from alginates undergoing gelation in the eye, J. Control. Release 44 (1997) 201 208. S. Kumar, B.O. Haglund, K.J. Himmelstein, In situ-forming gels for ophthalmic drug delivery, J. Ocul. Pharmacol. 10 (1994) 47 56. B. Srividya, R.M. Cardoza, P.D. Amin, Sustained ophthalmic delivery of ooxacin from a pH triggered in situ gelling system, J. Control. Release 73 (2001) 205211. L.D. Taylor, L.D. Cerankowski, Preparation of lms exhibiting a balanced temperature dependence to permeation by aqueous solutionsa study of lower consolute behavior, J. Polym. Sci. Polym. Chem. Ed. 13 (1975) 25512570. M. Heskins, J.E. Guillet, Solution properties of poly(N-isopropylacrylamide), J. Macromol. Sci. Chem. A2 (1968) 14411455. H.G. Schild, Poly(N-isopropylacrylamide): experiment, theory and application, Prog. Polym. Sci. 17 (1992) 163 249. A.S. Hoffman, Intelligent polymers, in: K. Park (Ed.), Controlled Drug DeliveryChallenges and Strategies, American Chemical Society, Washington, 1997, pp. 485 498. K. Mortensen, J.S. Pederson, Structural study on the micelle formation of poly(ethylene oxide)-poly(propylene oxide)-poly(ethylene oxide) triblock copolymer in aqueous solution, Macromolecules 26 (1993) 805812. W.F. Harrington, P.H. Von Hippel, The structure of collagen and gelatin, Adv. Protein Chem. 16 (1961) 1138. N.S. Anderson, J.W. Campbell, M.M. Harding, D.A. Rees, J.W.B. Samuel, X-ray diffraction studies of polysaccharide sulphates: double helix models for k- and i-carrageenans, J. Mol. Biol. 45 (1969) 8599. E. Heymann, Studies on sol gel transformations, I. The inverse sol gel transformation of methylcellulose in water, Trans. Faraday Soc. 31 (1935) 846 864. N. Sarkar, Thermal gelation properties of methyl and hydroxypropyl methylcellulose, J. Appl. Polym. Sci. 24 (1979) 10731087. M.C. Tate, D.A. Shear, S.W. Hoffman, D.G. Stein, M.C. LaPlaca, Biocompatibility of methylcellulose-based constructs designed for intracerebral gelation following experimental traumatic brain injury, Biomaterials 22 (2001) 11131123. A. Carlsson, G. Karlstrom, B. Lindman, O. Stenberg, Interaction between ethyl(hydroxyethyl)cellulose and sodium dodecyl sulphate in aqueous solution, Colloid Polym. Sci. 266 (1988) 1031 1036. A. Carlsson, G. Karlstrom, B. Lindman, Thermal gelation of nonionic cellulose ethers and ionic surfactants in water, Colloids Surf. 47 (1990) 147 165. B. Nystrom, H. Walderhaug, F.K. Hansen, Rheological behavior during thermoreversible gelation of aqueous mixtures of ethyl(hydroxyethyl)cellulose and surfactants, Langmuir 11 (1995) 750757. A. Carlsson, B. Lindman, T. Watanabe, K. Shirahama, Polymersurfactant interactions. Binding of N-tetradecylpyridinium bromide to ethyl(hydroxyethyl)cellulose), Langmuir 5 (1989) 12501252. B. Lindman, A. Carlsson, G. Karlstrom, M. Malmsten, Nonionic polymers and surfactantssome anomalies in temperature dependence and in interactions with ionic surfactants, Adv. Colloid Interface Sci. 32 (1990) 183 203. M. Scherlund, A. Brodin, M. Malmsten, Nonionic cellulose ethers as potential drug delivery systems for periodontal anesthesia, J. Colloid Interface Sci. 229 (2000) 365 374.

424

E. Ruel-Gariepy, J.-C. Leroux / European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 409426 [51] H. Feil, Y.H. Bae, J. Feijen, S.W. Kim, Effect of comonomer hydrophilicity and ionization on the lower critical solution temperature of N-isopropylacrylamide copolymers, Macromolecules 26 (1993) 24962500. [52] C.K. Han, Y.H. Bae, Inverse thermally-reversible gelation of aqueous N-isopropylacrylamide copolymer solutions, Polymer 39 (1998) 28092814. [53] B. Vernon, A. Gutowska, S.W. Kim, Y.H. Bae, Thermally reversible polymer gels for biohybrid articial pancreas, Macromol. Symp. 109 (1996) 155 167. [54] Y.H. Bae, B. Vernon, C.K. Han, S.W. Kim, Extracellular matrix for a rechargeable cell delivery system, J. Control. Release 53 (1998) 249 258. [55] B. Vernon, S.W. Kim, Y.H. Bae, Insulin release from islets of Langerhans entrapped in a poly(N-isopropylacrylamide-co-acrylic acid) polymer gel, J. Biomater. Sci. Polym. Ed. 10 (1999) 183 198. [56] H. Gappa, M. Baudys, J.J. Koh, S.W. Kim, Y.H. Bae, The effect of zinc-crystallized glucagon-like peptide-1 on insulin secretion of macroencapsulated pancreatic islets, Tissue Eng. 7 (2001) 35 44. [57] S.Y. Chae, S.W. Kim, Y.H. Bae, Effects of cross-linked hemoglobin on functionality and viability of microencapsulated pancreatic islets, Tissue Eng. 8 (2002) 379394. [58] S.Y. Chae, S.W. Kim, Y.H. Bae, Bioactive polymers for biohybrid articial pancreas, J. Drug Target. 9 (2001) 473 484. [59] K.H. Park, Y.H. Bae, Phenotype of hepatocyte spheroids in Arg-GlyAsp (RGD) containing a thermo-reversible extracellular matrix, Biosci. Biotechnol. Biochem. 66 (2002) 14731478. [60] H.H. Lin, Y.L. Cheng, In-situ thermoreversible gelation of block and star copolymers of poly(ethylene glycol) and poly(N-isopropylacrylamide) of varying architectures, Macromolecules 34 (2001) 37103715. [61] J. Rassing, D. Attwood, Ultrasonic velocity and light-scattering studies on the polyoxyethylene-polyoxypropylene copolymer Pluronic F127 in aqueous solution, Int. J. Pharm. 13 (1983) 4755. [62] Z. Zhou, B. Chu, Anomalous association behavior of an ethylene oxide/propylene oxide ABA block copolymer in water, Macromolecules 20 (1987) 30893091. [63] Z. Zhou, B. Chu, Light-scattering study on the association behavior of triblock polymers of ethylene oxide and propylene oxide in aqueous solution, J. Colloid Interface Sci. 126 (1988) 171 180. [64] G.Wanka,H. Hoffman,W. Ulbricht, Theaggregationbehaviorof poly(oxyethylene)-poly-(oxypropylene)-poly-(oxyethylene)-block-copolymers in aqueous solution, Colloid Polym. Sci. 268 (1990) 101 117. [65] K. Mortensen, Block copolymer in aqueous solution: Micelle formation and hard-sphere crystallization, Progr. Colloid Polym. Sci. 93 (1993) 7275. [66] K. Schillen, O. Glatter, W. Brown, Characterization of a PEO-PPOPEO block copolymer system, Progr. Colloid Polym. Sci. 93 (1993) 66 71. [67] M. Bohorquez, C. Koch, T. Trygstad, N. Pandit, A study of the temperature-dependent micellization of pluronic F127, J. Colloid Interface Sci. 216 (1999) 3440. [68] A. Cabana, A. Ait-Kadi, J. Juhasz, Study of the gelation process of polyethylene oxide-polypropylene oxide-polyethylene oxide copolymer (poloxamer 407) aqueous solutions, J. Colloid Interface Sci. 190 (1997) 307312. [69] I.R. Schmolka, Articial skin 1. Preparation and properties of Pluronic F-127 gels for treatment of burns, J. Biomed. Mater. Res. 6 (1972) 571 582. [70] S. Miyazaki, S. Takeuchi, C. Yokouchi, M. Takada, Pluronic F-127 gels as a vehicle for topical administration of anticancer agents, Chem. Pharm. Bull. 32 (1984) 4205 4208. [71] T.P. Johnston, S.C. Miller, Inulin disposition following intramuscular administration of an inulin/poloxamer gel matrix, J. Parent. Sci. Technol. 43 (1989) 279 286.

[32] M. Shirakawa, K. Yamatoya, K. Nishinari, Tailoring xyloglucan properties using an enzyme, Food Hydrocolloids 12 (1998) 2528. [33] S. Miyazaki, F. Suisha, N. Kawasaki, M. Shirakawa, K. Yamatoya, D. Attwood, Thermally reversible xyloglucan gels as vehicles for rectal drug delivery, J. Control. Release 56 (1998) 75 83. [34] F. Suisha, N. Kawasaki, S. Miyazaki, M. Shirakawa, K. Yamatoya, M. Sasaki, D. Attwood, Xyloglucan gels as sustained release vehicles for the intraperitoneal administration of mitomycin C, Int. J. Pharm. 172 (1998) 27 32. [35] N. Kawasaki, R. Ohkura, S. Miyazaki, Y. Uno, S. Sugimoto, D. Attwood, Thermally reversible xyloglucan gels as vehicles for oral drug delivery, Int. J. Pharm. 181 (1999) 227234. [36] S. Miyazaki, N. Kawasaki, K. Endo, D. Attwood, Oral sustained delivery of theophylline from thermally reversible xyloglucan gels in rabbits, J. Pharm. Pharmacol. 53 (2001) 1185 1191. [37] S. Miyazaki, S. Suzuki, N. Kawasaki, K. Endo, A. Takahashi, D. Attwood, In situ gelling xyloglucan formulations for sustained release ocular delivery of pilocarpine hydrochloride, Int. J. Pharm. 229 (2001) 29 36. [38] A. Takahashi, S. Suzuki, N. Kawasaki, W. Kubo, S. Miyazaki, R. Loebenberg, J. Bachynsky, D. Attwood, Percutaneous absorption of non-steroidal anti-inammatory drugs from in situ gelling xyloglucan formulations in rats, Int. J. Pharm. 246 (2002) 179186. [39] A. Chenite, C. Chaput, D. Wang, C. Combes, M.D. Buschmann, C.D. Hoemann, J.C. Leroux, B.L. Atkinson, F. Binette, A. Selmani, Novel injectable neutral solutions of chitosan form biodegradable gels in situ, Biomaterials 21 (2000) 21552161. [40] A. Chenite, M. Buschmann, D. Wang, C. Chaput, N. Kandani, Rheological characterization of thermogelling chitosan/glycerolphosphate solutions, Carbohydr. Polym. 46 (2001) 3947. [41] E. Ruel-Gariepy, A. Chenite, C. Chaput, S. Guirguis, J.C. Leroux, Characterization of thermosensitive chitosan gels for the sustained delivery of drugs, Int. J. Pharm. 203 (2000) 8998. [42] E. Ruel-Gariepy, G. Leclair, P. Hildgen, A. Gupta, J.C. Leroux, Thermosensitive chitosan-based hydrogel containing liposomes for the delivery of hydrophilic molecules, J. Control. Release 82 (2002) 373383. [43] E. Ruel-Gariepy, M. Shive, A. Bichara, M. Berrada, D. Le Garrec, A. Chenite, J.C. Leroux, A thermosensitive chitosan-based hydrogel for the local delivery of paclitaxel, Eur. J. Pharm. Biopharm. 57 (2004) 5363. [44] C.D. Hoemann, J. Sun, A. Legare, M.D. McKee, P. Ranger, M.D. Buschmann, A thermosensitive polysaccharide gel for cell delivery in cartilage repair, Trans. Orthop. Res. Soc. 26 (2001) 626. [45] C.D. Hoemann, M. Hurtig, J. Sun, D. McWade, E. Rossomacha, M.S. Shive, M.D. McKee, M.D. Buschmann, Full-thickness cartilage repair using an injectable chitosan scaffold, BST-CarGel, Third Annual Engineering Tissue Growth, 2003, Pittsburgh, PA. [46] G. Molinaro, J.C. Leroux, J. Damas, A. Adam, Biocompatibility of thermosensitive chitosan-based hydrogels: an in vivo experimental approach to injectable biomaterials, Biomaterials 23 (2002) 2717 2722. [47] C. Jarry, C. Chaput, A. Chenite, M.A. Renaud, M. Buschmann, J.C. Leroux, Effects of steam sterilization on thermogelling chitosanbased gels, J. Biomed. Mater. Res. (Appl. Biomater.) 58 (2001) 127135. [48] C. Jarry, J.C. Leroux, J. Haeck, C. Chaput, Irradiating or autoclaving chitosan/polyol solutions: Effect on thermogelling chitosan-bglycerophosphate systems, Chem. Pharm. Bull. 50 (2002) 1335 1340. [49] A.S. Hoffman, A. Afrassiabi, L.C. Dong, Thermally reversible hydrogels: II. Delivery and selective removal of substances from aqueous solutions, J. Control. Release 4 (1986) 213 222. [50] S. Hirotsu, Y. Hirokawa, T. Tanaka, Volume-phase transitions of ionized N-isopropylacrylamide gels, J. Chem. Phys. 87 (1987) 1392 1395.

 E. Ruel-Gariepy, J.-C. Leroux / European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 409426 [72] M. Katakam, W.R. Ravis, A.K. Banga, Controlled release of human growth hormone in rats following parenteral administration of poloxamer gels, J. Control. Release 49 (1997) 2126. [73] M.L. Veyries, G. Couarraze, S. Geiger, F. Agnely, L. Massias, B. Kunzli, F. Faurisson, B. Rouveix, Controlled release of vancomycin from poloxamer gels, Int. J. Pharm. 192 (1999) 183 193. [74] A. Paavola, I. Kilpelainen, J. Yliruusi, P. Rosenberg, Controlled release injectable liposomal gel of ibuprofen for epidural analgesia, Int. J. Pharm. 199 (2000) 8593. [75] J.G.W. Wenzel, K.S.S. Balaji, K. Koushik, C. Navarre, S.H. Duran, C.H. Rahe, U.B. Kompella, Pluronic F127 gel formulations of Deslorelin and GnRH reduce drug degradation and sustain drug release and effect in cattle, J. Control. Release 85 (2002) 5159. [76] L. Zhang, D.L. Parsons, C. Navarre, U.B. Kompella, Development and in-vitro evaluation of sustained release poloxamer 407 (P407) gel formulations of ceftiofur, J. Control. Release 85 (2002) 73 81. [77] H.G. Choi, J.H. Jung, J.M. Ryu, S.J. Yoon, Y.K. Oh, C.K. Kim, Development of in situ-gelling and mucoadhesive acetaminophen liquid suppository, Int. J. Pharm. 165 (1998) 3344. [78] J.M. Ryu, S.J. Chung, M.H. Lee, C.K. Kim, C.K. Shim, Increased bioavailability of propranolol in rats by retaining thermally gelling liquid suppositories in the rectum, J. Control. Release 59 (1999) 163172. [79] J.Y. Chang, Y.K. Oh, H.G. Choi, Y.B. Kim, C.K. Kim, Rheological evaluation of thermosensitive and mucoadhesive vaginals gels in physiological conditions, Int. J. Pharm. 241 (2002) 155 163. [80] J.Y. Chang, Y.K. Oh, H.S. Kong, E.J. Kim, D.D. Jang, K.T. Nam, C.K. Kim, Prolonged antifungal effects of clotrimazole-containing mucoadhesive thermosensitive gels on vaginitis, J. Control. Release 82 (2002) 39 50. [81] S.C. Shin, C.W. Cho, I.J. Oh, Enhanced efcacy by percutaneous absorption of piroxicam from the poloxamer gel in rats, Int. J. Pharm. 193 (2000) 213 218. [82] J. Liaw, Y.C. Lin, Evaluation of poly(ethylene oxide)-poly(propylene oxide)-poly(ethylene oxide) (PEO-PPO-PEO) gels as a release vehicle for percutaneous fentanyl, J. Control. Release 68 (2000) 273282. [83] A.H. El-Kamel, In vitro and in vivo evaluation of Pluronic F127based ocular delivery system for timolol maleate, Int. J. Pharm. 241 (2002) 4755. [84] G. Wei, H. Xu, P.T. Ding, S.M. Li, J.M. Zheng, Thermosetting gels with modulated gelation temperature for ophthalmic use: the rheological and gamma scintigraphic studies, J. Control. Release 83 (2002) 65 74. [85] T.P. Johnston, S.C. Miller, Toxicological evaluation of poloxamer vehicles for intramuscular use, J. Parent. Sci. Technol. 39 (1985) 8388. [86] Z.G.M. Wout, E.A. Pec, J.A. Maggiore, R.H. Williams, P. Palicharla, T.P. Johnston, Poloxamer 407-mediated changes in plasma cholesterol and triglycerides following intraperitoneal injection to rats, J. Parent. Sci. Technol. 46 (1992) 192 200. [87] W.K. Palmer, E.E. Emeson, T.P. Johnston, Poloxamer 407-induced atherogenesis in the C57BL/6 mouse, Atherosclerosis 136 (1998) 115123. [88] T.P. Johnston, W.K. Palmer, Mechanism of poloxamer 407-induced hypertriglyceridemia in the rat, Biochem. Pharmacol. 46 (1993) 10371042. [89] J.M. Blonder, L. Baird, J.C. Fulfs, G.J. Rosenthal, Dose-dependent hyperlipidemia in rabbits following administration of poloxamer 407 gel, Life Sci. 65 (1999) 261266. [90] X. Zhao, J. Yan, W. Battle, S. Allums, M. Bentley, Oligomers of poloxamer 407 as degradable thermal sensitive depot materials for sustained drug release, Proc. Int. Symp. Control. Rel. Bioact. Mater. 30 (2003) 167. [91] D. Cohn, A. Sosnik, A. Levy, Improved reverse thermo-responsive polymeric systems, Biomaterials 24 (2003) 37073714.

425

[92] G. Chen, A.S. Hoffman, Graft copolymers that exhibit temperatureinduced phase transition over a wide range of pH, Nature 373 (1995) 4952. [93] A.S. Hoffman, G. Chen, S. Kaang, D.T. Priebe, New bioadhesive polymer compositions for prolonged drug release in the eye, Proc. Int. Symp. Control. Rel. Bioact. Mater. 22 (1995) 159160. [94] A.S. Hoffman, G. Chen, X. Wu, Z. Ding, B. Kabra, K. Randeri, M. Schiller, E. Ron, N. Peppas, C. Brazel, Graft copolymers of PEO-PPO-PEO triblock polyethers on bioadhesive polymer backbones: Synthesis and properties, Polym. Prepr. 38 (1997) 524 525. [95] A.S. Hoffman, G. Chen, Block and graft copolymers and methods relating thereto, Int. Pat. Appl. W/O 95/24430, 1995. [96] L. Bromberg, E.C. Lupton, M.E. Schiller, M.J. Timm, G.W. McKinney, M. Orkisz, B. Hand, Responsive polymer networks and methods of their use, Int. Pat. Appl. WO 97/00275, 1997. [97] L. Bromberg, Novel family of thermogelling materials via CC bonding between poly(acrylic acid) and poly(ethylene oxide)-bpoly(propylene oxide)-b-poly(ethylene oxide), J. Phys. Chem. B 102 (1998) 1956 1963. [98] L. Bromberg, Polyether-modied poly(acrylic acid): synthesis and applications, Ind. Eng. Chem. Res. 37 (1998) 42674274. [99] L.E. Bromberg, E.S. Ron, Temperature-responsive gels and thermogelling polymer matrices for protein and peptide delivery, Adv. Drug Del. Rev. 31 (1998) 197 221. [100] L. Bromberg, Scaling of rheological properties of hydrogels from associating polymers, Macromolecules 31 (1998) 61486156. [101] L. Bromberg, Self-assembly in aqueous solutions of polyethermodied poly(acrylic acid), Langmuir 14 (1998) 58065812. [102] L. Bromberg, Properties of aqueous solutions and gels of poly(ethylene oxide)-b-poly(propylene oxide)-b-poly(ethylene oxide)-g-poly(acrylic acid), J. Phys. Chem. B 102 (1998) 1073610744. [103] L.E. Bromberg, M.J. Orkisz, E.S. Ron, Bioadhesive properties of polyethylene-b-polyoxypropylene-b-polyoxyethylene-g-poly(acrylic acid) polymers (Smart Hydrogele), Polym. Prepr. 38 (1997) 626 627. [104] E.S. Ron, L. Bromberg, S. Luczak, M. Kearney, D. Deaver, M. Schiller, Smart hydrogel: a novel mucosal delivery system, Proc. Int. Symp. Control. Rel. Bioact. Mater. 24 (1997) 407 408. [105] E.S. Ron, E.J. Roos, A.K. Staples, L.E. Bromberg, M.E. Schiller, Interpenetrating polymer networks for sustained dermal delivery, Proc. Int. Symp. Control. Rel. Bioact. Mater. 23 (1996) 128129. [106] A.M. Potts, S. Jackson, N. Washington, P. Gilchrist, E.S. Ron, M. Schiller, C.G. Wilson, In vivo determination of the oesophageal retention of smart hydrogel, Proc. Int. Symp. Control. Rel. Bioact. Mater. 24 (1997) 335 336. [107] L.E. Bromberg, T.H.E. Mendum, M.J. Orkisz, M.E. Schiller, E.C. Lupton, E.S. Ron, Polyoxyethylene-b-polyoxypropylene-b-polyoxyethylene-g-poly(acrylic acid) polymers (Smart Hydrogele) as a carrier in controlled delivery of proteins and peptides, Polym. Prepr. 38 (1997) 602 603. [108] B. Jeong, Y.H. Bae, D.S. Lee, S.W. Kim, Biodegradable block copolymers as injectable drug-delivery systems, Nature 388 (1997) 860 862. [109] B. Jeong, Y.H. Bae, S.W. Kim, Biodegradable thermosensitive micelles of PEG-PLGA-PEG triblock copolymers, Colloids Surf. B 16 (1999) 185 193. [110] B. Jeong, Y.H. Bae, S.W. Kim, Thermoreversible gelation of PEGPLGA-PEG triblock copolymer aqueous solutions, Macromolecules 32 (1999) 70647069. [111] K.W. Kwon, M.J. Park, Y.H. Bae, H.D. Kim, K. Char, Gelation behavior of PEO-PLGA-PEO triblock copolymers in water, Polymer 43 (2002) 33533358. [112] B. Jeong, Y.H. Bae, S.W. Kim, Drug release from biodegradable injectable thermosensitive hydrogel of PEG-PLGA-PEG triblock copolymers, J. Control. Release 63 (2000) 155 163.

426

E. Ruel-Gariepy, J.-C. Leroux / European Journal of Pharmaceutics and Biopharmaceutics 58 (2004) 409426 polysaccharide and protein hydrogel, Biomaterials 22 (2001) 453 462. E. Westhaus, P.B. Messersmith, Thermally triggered enzymatic crosslinking of proteins for in-situ gelation, Transaction of Society of Biomaterials, 25th Annual Meeting, 1999, p. 247. T.J. Sanborn, P.B. Messersmith, A.E. Barron, In situ crosslinking of a biomimetic peptide-PEG hydrogel via thermally triggered activation of factor XIII, Biomaterials 23 (2002) 27032710. J.H. Collier, B.H. Hu, J.W. Ruberti, J. Zhang, P. Shum, D.H. Thompson, P.B. Messersmith, Thermally and photochemically triggered self-assembly of peptide hydrogels, J. Am. Chem. Soc. 123 (2001) 94639464. S.C. Song, S.B. Lee, J.I. Jin, Y.S. Sohn, A new class of biodegradable thermosensitive polymers. I. Synthesis and characterization of poly(organophosphazenes) with methoxy-poly(ethylene glycol) and amino acid esters as side groups, Macromolecules 32 (1999) 2188 2193. S.B. Lee, S.C. Song, J.I. Jin, Y.S. Sohn, A new class of biodegradable thermosensitive polymers. 2. Hydrolytic properties and salt effect on the lower critical solution temperature of poly(organophosphazenes) with methoxypoly(ethylene glycol) and amino acid esters as side groups, Macromolecules 32 (1999) 78207827. S.B. Lee, S.C. Song, J.I. Jin, Y.S. Sohn, Thermosensitive cyclotriphosphazenes, J. Am. Chem. Soc. 122 (2000) 83158316. B.H. Lee, Y.M. Lee, Y.S. Sohn, S.C. Song, A thermosensitive poly(organophosphazene) gel, Macromolecules 35 (2002) 38763879. J. Wang, D.D.N. Sun, K.W. Leong, Injectable polyphosphate hydrogel, Proc. Int. Symp. Control. Rel. Bioact. Mater. 30 (2003) 140. J.M. Barichello, M. Morishita, K. Takayama, T. Nagai, Absorption of insulin from pluronic F-127 gels following subcutaneous administration in rats, Int. J. Pharm. 184 (1999) 189198. M.M. Amiji, P.K. Lai, D.B. Shenoy, M. Rao, Intratumoral administration of paclitaxel in an in situ gelling poloxamer 407 formulation, Pharm. Dev. Tech. 7 (2002) 195 202.

[113] B. Jeong, Y.H. Bae, S.W. Kim, In situ gelation of PEG-PLGA-PEG triblock copolymer aqueous solutions and degradation thereof, J. Biomed. Mater. Res. 50 (2000) 171177. [114] B. Jeong, M.R. Kibbey, J.C. Birnbaum, Y.Y. Won, A. Gutowska, Thermogelling biodegradable polymers with hydrophilic backbones: PEG-g-PLGA, Macromolecules 33 (2000) 83178322. [115] B. Jeong, L.Q. Wang, A. Gutowska, Biodegradable thermoreversible gelling PLGA-g-PEG copolymers, Chem. Commun. (2001) 1516 1517. [116] Y.M. Chung, K.L. Simmons, A. Gutowska, B. Jeong, Solgel transition temperature of PLGA-g-PEG aqueous solutions, Biomacromolecules 3 (2002) 511516. [117] B. Jeong, K.M. Lee, A. Gutowska, Y.H. An, Thermogelling biodegradable copolymer aqueous solutions for injectable protein delivery and tissue engineering, Biomacromolecules 3 (2002) 865868. [118] R.C. Rathi, G.M. Zentner, Biodegradable low molecular weight triblock poly(lactide-co-glycolide) polyethylene glycol copolymers having reverse thermal gelation properties, U.S. Patent 6,004,573, 1999. [119] G.M. Zentner, R. Rathi, C. Shih, J.C. McRea, M.H. Seo, H. Oh, B.G. Rhee, J. Mestecky, Z. Moldoveanu, M. Morgan, S. Weitman, Biodegradable block colpolymers for drug delivery of proteins and water-insoluble drugs, J. Control. Release 72 (2001) 203215. [120] K.D. Fowers, M. Baudys, R. Rathi, C. Shih, Thermally reversible gelling materials for safe and versatile depot delivery, Drug Del. Technol. 3 (2003) 5864. [121] Y.J. Kim, S. Choi, J.J. Koh, M. Lee, K.S. Ko, S.W. Kim, Controlled release of insulin from injectable biodegradable triblock copolymer, Pharm. Res. 18 (2001) 548550. [122] P.B. Messersmith, S. Starke, Thermally triggered calcium phosphate formation from calcium-loaded liposomes, Chem. Mater. 10 (1998) 117124. [123] E. Westhaus, P.B. Messersmith, Triggered release of calcium from lipid vesicles: a bioinspired strategy for rapid gelation of

[124]

[125]

[126]

[127]

[128]

[129] [130]

[131] [132]

[133]