**1. Introduction**

254 Recent Advances in Arthroplasty

Springfield DS. Massive autogenous bone grafts. Orthop Clin North Am. Apr; 18(2):249-56;

Sugihara S, van Ginkel AD, Jiya TU, van Royen BJ, van Diest PJ, Wuisman PI.

Tägil M. The morselized and impacted bone graft. Animal experiments on proteins, impaction and load [Thesis]. Lund (Sweden): Lund University Hospital; 2000.

Thien TM, Welten ML, Verdonschot N, Buma P, Yong P, Schreurs BW. Acetabular revision

Ullmark G, Obrant KJ. Histology of impacted bone-graft incorporation. J Arthroplasty. Feb;

Update: Allograft-associated bacterial infections - United States, 2002. Nov., 4th, 2002. http://www.medscape.com/viewarticle/430131 [accessed Jun, 9th, 2011]. Urist MR, Hernandez A. Excitation transfer in bone. Deleterious effects of cobalt 60 radiation-sterilization of bank bone. Arch Surg. Oct; 109(4):486-93; 1974. Vajaradul Y. Bone banking in Thailand. A 10-year experience (1984-1994).Clin Orthop. Feb;

Viceconti M, Toni A, Brizio L, Rubbini L, Borrelli A. The effect of autoclaving on the

Zhang Q, Cornu O, Delloye C. Ethylene oxide does not extinguish the osteoinductive

mechanical properties of bank bovine bone [abstract]. Chir Organi Mov. Jan-Mar;

capacity of demineralized bone. A reappraisal in rats. Acta Orthop Scand. Apr;

cases at 5 to 9 years' follow-up. J Arthroplasty. Aug; 16(5):666-70; 2001. Turek SL. Orthopaedics: Principles and applications. 4th ed. Editora Manole Ltda, Rio de

Taylor D. Inactivation of the BSE agent. C R Acad Sci III. Jan; 325(1):75-6; 2002.

Janeiro, RJ, Brazil. p.756; 1991 (in Portuguese).

Histopathology of retrieved allografts of the femoral head. J Bone Joint Surg Br.

with impacted freeze-dried cancellous bone chips and a cemented cup: a report of 7

1987.

Mar; 81(2):336-41; 1999.

17(2):150-7; 2002.

(323):173-80; 1996.

81(1):63-8; 1996.

68(2):104-8; 1997.

A majority of individuals reporting temporomandibular joint (TMJ) disorders have joint damage, inflammation or arthritis (Manfredini, Chiappe & Bosco, 2006; Plesh, Sinisi, Crawford & Gansky, 2005). The joint will show loss of extra-cellular matrix components in the articular cartilage and subchondral bone resulting in destruction of cartilage and bone, such processes lead to inflammation and exacerbation of joint tissue catabolism (Tanaka, Detamore & Mercuri, 2008). Matrix metalloproteinases 1 and 9 (MMP-1 and MMP-9) are two major enzymes that contribute to tissue catabolism and have been observed in patients with TMJ disorders (Kanyama, Kuboki, Kojima, Fujisawa, Hattori, Takigawa & Yamashita, 2000; Srinivas, Sorsa, Tjaderhane, Niemi, Raustia, Pernu, Teronen & Salo, 2001; Yoshida, Takatsuka, Hatada, Nakamura, Tanaka, Ueki, Nakagawa, Okada, Yamamoto & Fukuda, 2006). Reversal of these disease processes and treatment of the joint to reduce pain are effective in the early stages of the disease, but treatment often fails to alleviate the severe, chronic pain caused by advanced joint degeneration (Gerwin, Hops & Lucke, 2006; Tanaka, Detamore & Mercuri, 2008).

Treatment of TMJ osteoarthritis with intra-articular injections of non-steroidal antiinflammatory drugs (NSAIDs) and opiates into the superior joint space have shown efficacy (Bryant, Harrison, Hopper & Harris, 1999; Swift, Roszkowski, Alton & Hargreaves, 1998; Zuniga, Ibanez & Kozacko, 2007). Intra-articular administration versus systemic administration of NSAID or opiates would be advantageous for treatment of TMJ inflammation and pain because local administration avoids the ectopic effects seen with NSAIDs like rofecoxib (i.e., Vioxx) (Lin, Weisdorf, Solovey & Hebbel, 2000) or opiates. For example, nonselective NSAIDS can cause intestinal bleeding, whereas some selective cyclooxygenase-2 inhibitors have significant cardiovascular and renal safety risks (Davies & Jamali, 2004; Mukherjee, Nissen & Topol, 2001). Moreover, opioids frequently cause constipation, sedation, nausea, vomiting, and respiratory depression (Mercadante, 1999).

Intra-articular injection remains controversial in light of decades of mixed reports demonstrating intra-articular injections either accelerate or trigger destruction of tissues within the TMJ and the surrounding area (Bjornland, Rorvik, Haanaes & Teige, 1994; Sugisaki, Ikai & Tanabe, 1995; Westesson, Eriksson & Liedberg, 1986). Recent reports

Cross-Linked Gelatin Microcapsules for Drug Delivery in a Arthritic TMJ 257

2.5µl/ml of 1mM Alexa 488 dye (Invitrogen, Carlsbad CA) dissolved in toluene. Or a saturated solution of ibuprofen in canola oil was used containing ~15% (w/v) ibuprofen. A solution of Type A gelatin with a Bloom strength of 300 was added after sonication. The collagen gelatin was allowed to coat the oil droplets such that 80% of the microcapsule mass was canola oil and 20% was a gelatin shell by weight. Amino acids in the gelatin coated oil droplets were then cross-linked by addition of 15% 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide) [EDC] (dry wt. of gelatin) to induce stable microcapsules. EDC is a water soluble carbodiimide usually obtained as the hydrochloride that can generally be used as a carboxyl activating agent for the coupling of primary amines to yield amide bonds between a free amine group (e.g., protein- or peptide-bound lysine) and the gamma-carboxamid group of protein- or peptide-bound glutamine. The cross-linking reaction was performed at room temperature for 12 hr. After crosslinking the microcapsules were washed 4 times with Tris Buffer pH 7.0, dried, sterilized with ethylene oxide and then hydrated with 0.9% saline. The morphine capsules were produced in a similar manner with some modifications. Morphine alkaloid crystals were ground to a small size (less than 50 microns) and suspended in canola oil. This suspension was emulsified in a solution of the gelatin at 60°C. A complexing agent (hexametaphosphate) was added to the aqueous phase and the pH was adjusted to approximately 4.75 with acetic acid to allow coacervation to occur. The

Known weights of beads were dissolved in acetonitrile and analyzed via high performance liquid chromatography (HPLC) as follows. A six point calibration curve was prepared by serial dilution and HPLC analysis. The stock solution of 1mg/mL was prepared in 1:1 deionized water (pH 2.5) and acetonitrile. Concentrations of 1, 5, 10, 25, 50 and 100 µg/mL were made in methanol. Samples of 20 µL were injected onto a C-18 column (Supelco Discovery C-18, 4.6 x 15 cm). The isocratic separation was carried out at 30°C using a mobile phase of 1:1 deionized water (pH 2.5) and acetonitrile. The flow rate was 2.0 ml/min with a run time of 10 minutes. Absorbance at 214 nm was taken using a Waters 486 UV detector. The absorbance of the sample was compared to a six-point standard curve from which

In 24 well tissue culture plates approximately 80,000 microcapsules were placed in each well with either 0.9% saline, 15 µg/ml CFA (paraffin oil/15g *Mycobacterium tuberculosis*) mixed in water, or a 50:50 mixture of 0.9% saline and 15 µg/ml CFA. A homogenate of the water and oil was created by passing the solution through a 20 gauge needle 5 times. The plates were placed in a 5% CO2 chamber at 37◦C. The microcapsules were counted in triplicate in a hemacytometer chamber 0, 5 and 10 days following addition to the different solutions.

The Baylor College of Dentistry Institutional Animal Care and Use Committee approved the experimental protocol. Male (250 grams) Sprague-Dawley rats from Harlan Industries, Houston, TX were kept on a 12:12 light/dark cycle with lights on at 08:00 hours. The animals were housed individually in our computerized feeding modules and given food and water ad libitum. They were acclimated to the surroundings for two days before

microcapsules were crosslinked, lyophilized until dry and sterilized.

concentration and percent loading in capsules was determined.

**2.3 In vitro assay for microcapsule degradation** 

**2.4 TMJ Injections** 

**2.2 Detection of Ibuprofen and morphine by HPLC** 

demonstrate that the risks of intra-articular TMJ injection can be minimized by utilizing magnetic resonance imaging (MRI) for visualizing the needle during the injection procedure (Fritz, Thomas, Tzaribachev, Horger, Claussen, Lewin & Pereira, 2009; Hayakawa, Kober, Otonari-Yamamoto, Otonari, Wakoh & Sano, 2007). Current treatment protocols that utilize intra-articular drug delivery can require frequent injections, increasing the risk of infection and damage to the joint and, in the event MRI or some other form of imaging is utilized, dramatically increase the expense of the procedure. Alternatively, if drug delivery could be sustained after a single injection these risks from a multi-injection protocol could be minimized.

A variety of methods have been tested for sustained drug delivery in a joint space, including encapsulating or incorporating drugs into nano- or microparticles consisting of organic polymers (Deasy, 1994; Mountziaris, Kramer & Mikos, 2009). In this report we test a crosslinked fluid filled gelatin capsule for intra-articular injection. This capsule has a gelatin shell with a core of oil containing the drug ibuprofen or morphine. Other studies used uncrosslinked gelatin microcapsules, in contrast to our crosslinked capsules. The capsules were given orally but only released drug for 24-48 hours (Jizomoto, Kanaoka, Sugita & Hirano, 1993). Most intra-articular studies testing sustained release of NSAIDs utilized poly(lactide-co-glycolide) (PLGA) (Bozdag, Calis, Kas, Ercan, Peksoy & Hincal, 2001; Fernandez-Carballido, Herrero-Vanrell, Molina-Martinez & Pastoriza, 2004a; Fernandez-Carballido, Herrero-Vanrell, Molina-Martinez & Pastoriza, 2004b; Liggins, Cruz, Min, Liang, Hunter & Burt, 2004; Puebla, Pastoriza, Barcia & Fernandez-Carballido, 2005; Tuncay, Calis, Kas, Ercan, Peksoy & Hincal, 2000). PLGA has a rigid structure and may be irritating to the joint when injected intra-articularly, particularly when smaller than 20 microns (Liggins, Cruz, Min, Liang, Hunter & Burt, 2004). Importantly, in vivo studies injecting solid gelatin spheres have shown that they do not induce an inflammatory response in vivo suggesting gelatin in these studies should not induce an immune response (Brown, Leong, Huang, Dalal, Green, Haimes, Jimenez & Bathon, 1998).

Testing the rate of gelatin microcapsule degradation, as a result of MMP degradation, and the level of inflammation induced by the microcapsules in the TMJ are necessary. Also an important question to address was does sustained release improve pain reduction when administering NSAIDs or opiates to treat a TMJ disorder? To address these questions we used a rat model for TMJ arthritis.

In this model we inject the TMJ with the adjuvant complete Freund's adjuvant (CFA) and the rat's meal duration will lengthen in male and female rats as a result of this injection (Kerins, Carlson, Hinton, Grogan, Marr, Kramer, Spears & Bellinger, 2005; Kerins, Carlson, McIntosh & Bellinger, 2003; Kerins, Carlson, McIntosh & Bellinger, 2004; Kramer & Bellinger, 2009; Thut, Hermanstyne, Flake & Gold, 2007). Microcapsules containing vehicle or the NSAID ibuprofen or morphine were injected into the TMJ of rats that had no joint arthritis or had adjuvant (i.e., CFA) induce arthritis. Degradation of the microcapsules was monitored and the level of inflammatory cytokine IL-1β in the joint tissues was quantitated. In addition, the nociceptive response to the microcapsules loaded with vehicle or NSAID or morphine were measured in a non-arthritic and arthritic TMJ after intra-articular injection.

### **2. Materials and methods**

### **2.1 Microcapsule production**

To produce the loaded microcapsules canola oil was sonicated in water to induce formation of droplets 20-50 micrometers in size. Before sonication the canola oil was mixed with

demonstrate that the risks of intra-articular TMJ injection can be minimized by utilizing magnetic resonance imaging (MRI) for visualizing the needle during the injection procedure (Fritz, Thomas, Tzaribachev, Horger, Claussen, Lewin & Pereira, 2009; Hayakawa, Kober, Otonari-Yamamoto, Otonari, Wakoh & Sano, 2007). Current treatment protocols that utilize intra-articular drug delivery can require frequent injections, increasing the risk of infection and damage to the joint and, in the event MRI or some other form of imaging is utilized, dramatically increase the expense of the procedure. Alternatively, if drug delivery could be sustained after a single injection these risks from a

A variety of methods have been tested for sustained drug delivery in a joint space, including encapsulating or incorporating drugs into nano- or microparticles consisting of organic polymers (Deasy, 1994; Mountziaris, Kramer & Mikos, 2009). In this report we test a crosslinked fluid filled gelatin capsule for intra-articular injection. This capsule has a gelatin shell with a core of oil containing the drug ibuprofen or morphine. Other studies used uncrosslinked gelatin microcapsules, in contrast to our crosslinked capsules. The capsules were given orally but only released drug for 24-48 hours (Jizomoto, Kanaoka, Sugita & Hirano, 1993). Most intra-articular studies testing sustained release of NSAIDs utilized poly(lactide-co-glycolide) (PLGA) (Bozdag, Calis, Kas, Ercan, Peksoy & Hincal, 2001; Fernandez-Carballido, Herrero-Vanrell, Molina-Martinez & Pastoriza, 2004a; Fernandez-Carballido, Herrero-Vanrell, Molina-Martinez & Pastoriza, 2004b; Liggins, Cruz, Min, Liang, Hunter & Burt, 2004; Puebla, Pastoriza, Barcia & Fernandez-Carballido, 2005; Tuncay, Calis, Kas, Ercan, Peksoy & Hincal, 2000). PLGA has a rigid structure and may be irritating to the joint when injected intra-articularly, particularly when smaller than 20 microns (Liggins, Cruz, Min, Liang, Hunter & Burt, 2004). Importantly, in vivo studies injecting solid gelatin spheres have shown that they do not induce an inflammatory response in vivo suggesting gelatin in these studies should not induce an immune response (Brown, Leong, Huang,

Testing the rate of gelatin microcapsule degradation, as a result of MMP degradation, and the level of inflammation induced by the microcapsules in the TMJ are necessary. Also an important question to address was does sustained release improve pain reduction when administering NSAIDs or opiates to treat a TMJ disorder? To address these questions we

In this model we inject the TMJ with the adjuvant complete Freund's adjuvant (CFA) and the rat's meal duration will lengthen in male and female rats as a result of this injection (Kerins, Carlson, Hinton, Grogan, Marr, Kramer, Spears & Bellinger, 2005; Kerins, Carlson, McIntosh & Bellinger, 2003; Kerins, Carlson, McIntosh & Bellinger, 2004; Kramer & Bellinger, 2009; Thut, Hermanstyne, Flake & Gold, 2007). Microcapsules containing vehicle or the NSAID ibuprofen or morphine were injected into the TMJ of rats that had no joint arthritis or had adjuvant (i.e., CFA) induce arthritis. Degradation of the microcapsules was monitored and the level of inflammatory cytokine IL-1β in the joint tissues was quantitated. In addition, the nociceptive response to the microcapsules loaded with vehicle or NSAID or morphine were measured in a non-arthritic and arthritic TMJ after intra-articular injection.

To produce the loaded microcapsules canola oil was sonicated in water to induce formation of droplets 20-50 micrometers in size. Before sonication the canola oil was mixed with

multi-injection protocol could be minimized.

Dalal, Green, Haimes, Jimenez & Bathon, 1998).

used a rat model for TMJ arthritis.

**2. Materials and methods 2.1 Microcapsule production**  2.5µl/ml of 1mM Alexa 488 dye (Invitrogen, Carlsbad CA) dissolved in toluene. Or a saturated solution of ibuprofen in canola oil was used containing ~15% (w/v) ibuprofen. A solution of Type A gelatin with a Bloom strength of 300 was added after sonication. The collagen gelatin was allowed to coat the oil droplets such that 80% of the microcapsule mass was canola oil and 20% was a gelatin shell by weight. Amino acids in the gelatin coated oil droplets were then cross-linked by addition of 15% 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide) [EDC] (dry wt. of gelatin) to induce stable microcapsules. EDC is a water soluble carbodiimide usually obtained as the hydrochloride that can generally be used as a carboxyl activating agent for the coupling of primary amines to yield amide bonds between a free amine group (e.g., protein- or peptide-bound lysine) and the gamma-carboxamid group of protein- or peptide-bound glutamine. The cross-linking reaction was performed at room temperature for 12 hr. After crosslinking the microcapsules were washed 4 times with Tris Buffer pH 7.0, dried, sterilized with ethylene oxide and then hydrated with 0.9% saline. The morphine capsules were produced in a similar manner with some modifications. Morphine alkaloid crystals were ground to a small size (less than 50 microns) and suspended in canola oil. This suspension was emulsified in a solution of the gelatin at 60°C. A complexing agent (hexametaphosphate) was added to the aqueous phase and the pH was adjusted to approximately 4.75 with acetic acid to allow coacervation to occur. The microcapsules were crosslinked, lyophilized until dry and sterilized.
