Canine bone response to tyrosine-derived polycarbonates and poly(L-lactic acid)

Jack Choueka,’ Jose L. Charvet,’ Kenneth J. Koval,’ Harold Alexander,‘ Kenneth S. James; Kimberly A. Hooper; and Joachim Kohn*f ‘Department of Bioengineering, Orthopaedic Institute, Hospital for Joint Diseases, New York, New York; ZDepartment of Chemistry, Rutgers, State University of New Jersey, Piscataway, New Jersey 08855

Tyrosine-derived polycarbonates are a new class of degrad- able polymers developed for orthopedic applications. In this study the long-term (48 week) in vivo degradation kinet- ics and host bone response to poly(DTE carbonate) and poly(DTH carbonate) were investigated using a canine bone chamber model. Poly(L-lactic acid) (PLA) served as a control material. Two chambers of each test material were retrieved at 6-, 12-, 24-, and 48-week time points. Tyrosine-derived polycarbonates were found to exhibit degradation kinetics comparable to PLA. Each test material lost approximately 50% of its initial molecular weight (M,) over the 48-week test period. Poly(DTE carbonate) and poly(DTH carbonate)

test chambers were characterized by sustained bone in- growth throughout the 48 weeks. In contrast, bone ingrowth into the PLA chambers peaked at 24 weeks and dropped by half at the 48-week time point. A fibrous tissue layer was found surrounding the PLA implants at all time points. This fibrous tissue layer was notably absent at the interface be- tween bone and the tyrosine-derived polycarbonates. Histo- logic sections revealed intimate contact between bone and tyrosine-derived polycarbonates. From a degradation- biocompatibility perspective, the tyrosine-derived polycar- bonates appear to be comparable, if not superior, to PLA in this canine bone chamber model. 0 1996 John Wiley & Sons, Inc.

INTRODUCTION

Degradable, polymeric bone fixation devices are an intriguing alternative to metallic implants, since they can alleviate bone stress shielding associated with metal implants and eliminate the need for surgical removal of the metallic devices after completion of the healing process. Additional advantages include the possibility of using degradable polymeric implants for the local delivery of osteogenic factors and/or their use as cellular scaffolds to aid in the healing process.

Degradable polymers being investigated for ortho- pedic applications include poly(L-lactic acid) (PLA),1-3 poly(glyco1ic acid) (PGA),4,5 PLA-PGA ~opolymers,4.~,~ polydio~anone,8.~ polycaprolactone,10 poly(ortho es- ter)s,”#” and poly(ethy1ene oxide)-poly(buty1ene ter- ephtalate) copolymer^.'^ In particular, PLA has gar- nered much attention, since it is biocompatible in the short term and able to provide adequate strength and stiffness for use in small bone fixation.’j3j5 However, PLA implants have been associated with late inflam- matory and bone resorption responses? Concerns have

*To whom correspondence should be addressed at De- partment of Chemistry, Rutgers University, P.O. Box 939, Piscataway, NJ 08855.

been raised regarding the acidic nature of the degrada- tion products and the particulate debris formed dur- ing erosi~n.’~J~

Tyrosine-derived polycarbonates are part of a newly synthesized class of degradable Like PLA, these materials are based on natural metabolites (the amino acid tyrosine), have favorable mechanical properties, are readily processible, and degrade on the order of months to years. Unlike PLA, tyrosine-derived polycarbonates are completely amorphous. Hence, degradation to crystalline particulate debris is not of practical concern for these materials. Similarly, the degradation products of tyrosine-derived polycarbon- ates are nonacidic. In vitro cytotoxicitylE and short- term in vivo evaluations in ratsI9 and rabbits8 have shown tyrosine-derived polycarbonates to be generally bio- compatible.

In this study, we investigated these materials over a longer in vivo degradation time course. A canine bone chamber model was used to investigate the inter- action of bone with polycarbonates derived from desaminotyrosyl-tyrosine ethyl ester [poly(DTE car- bonate)] and desaminotyrosyl-tyrosine hexyl ester [poly(DTH carbonate)]. Poly(L-lactic acid) was in- cluded in the study for comparison and served as a control material. Here we present 48 week in vivo deg-

Journal of Biomedical Materials Research, Vol. 31, 35-41 (1996) 0 1996 John Wiley & Sons, Inc. CCC 0021-9304/96 /010035-07

36 CHOUEKA ET AL.

radation kinetics for poly(DTE carbonate), poly(DTH carbonate), and PLA. In addition, the host bone re- sponse to these materials and the nature of the bone- implant interface were explored.

MATERIALS AND METHODS

Materials and implant preparation

Poly(DTE carbonate) ( M , = 121,000, M , = 62,000) and poly(DTH carbonate) ( M , = 220,000, M , = 114,000) were synthesized according to published proce- d ~ r e s . ' ~ J ~ Medical-grade poly(L-lactic acid) ( M , =

191,000, M , = 113,000) was obtained from Boehringer Ingelheim in pellet form and used as obtained.

Compression molded films (41 X 38 X 0.4 mm, 800 mg) were prepared using a Carver Laboratory Press (Fred S. Carver Inc., Menomonee Falls, WI) and a stainless-steel mold. Poly(DTE carbonate) and poly(DTH carbonate) were molded at 130°C and 110"C, respectively, corresponding to 50-60°C above their re- spective glass transition temperatures. Poly(L-lactic acid) was molded at its melting temperature (160°C). The films were pressed under two metric tons of pres- sure applied for 5 min for poly(DTH carbonate) and 10 min for poIy(DTE carbonate) and poly(L-lactic acid).

Each of the fabricated films was cut into approxi- mately 20 coupons (8 X 7 X 0.4 mm). The coupons were sonicated in sterile water, dried under vacuum for 1 week, and inserted into either a poly(DTE carbon- ate), poly(DTH carbonate), or poly(1actic acid) desig- nated bone chamber. The bone chamber (8 X 25 X 10 mm) has been described previ~usly.~ Briefly, coupon pairs were inserted into grooves in the medical grade UHMW polyethylene implant housing to form 10 bone ingrowth channels measuring 1 X 5 X 10 mm.

The assembled chambers were individually bagged, sterilized using ethylene oxide (Anprolene; Anderson Products, Chapel Hill, NC), and allowed to degas for 2 weeks.

poly(DTH carbonate); dog 2: poly(L-lactic acid), poly(DTE carbonate); and dog 3: poly(L-lactic acid), poly(DTH carbonate).

Specimen retrieval

At the designated time intervals the dogs were sacri- ficed and the femurs harvested. The bone chambers were isolated with a diamond saw leaving at least 3 mm of surrounding host bone to assure that the bone-biomaterial interface would not be disturbed.

Bone ingrowth

High-resolution X-rays of the intact implants with the surrounding bone were taken perpendicular to the tissue-coupon interface using a Faxitron imaging sys- tem operated at 30.0 kV, 2.5 mA, for 3.0 min (Faxitron series 43805N X-ray system; Hewlett-Packard, Rock- ville, MD). These X-rays were captured onto a Macin- tosh IIfx computer for image analysis (TCL Image) with a videocamera linked to a microscope. Quantitative analysis of the Faxitrons was performed by measuring the percent area of each channel occupied by calci- fied bone.

Polymer degradation

Polymer coupons removed from the retrieved cham- bers were examined for changes in appearance and residual molecular weight was measured by gel per- meation chromatography (GPC). A piece of the proxi- mal coupon from each chamber was dissolved in THF and injected into a GPC system consisting of a Perkin- Elmer Model 410 pump, a Waters Model 410 Refractive Index Detector, and Perkin-Elmer Model 2600 compu- terized data station. Two PL-gel GPC columns (lo5 and lo3 A pore size, 30 cm in length) were operated in series at a flow rate of 1 ml/min in THF. Molecular weights were calculated relative to polystyrene stan- dards (Polymer Laboratories, Inc.).

In vivo testing scheme Bone-implant interface

We used 12 male hound dogs, 3-5 years old, in this study. The dogs were divided equally into four groups: 6-, 12-, 24-, and 48-week experimental time points.

Assembled bone chambers were implanted in surgi- cally created longitudinal cortical defects in the lateral metaphysis of the right and left distal femurs of each dog as per S~ganuma.~ The ingrowth channels faced the intact surfaces of the anterior and posterior cortices. The three dogs in each group were implanted with bone chambers according to the following scheme (left femur, right femur): dog 1: poly(DTE carbonate),

The retrieved implants were fixed en bloc in 10% phosphate-buffered formalin, dehydrated, and methacrylate-embedded for undecalcified light mi- croscopy analysis of the bone-implant interface. The embedded specimens were cut perpendicularly through the center of the bone chamber openings. A 5-pm section was taken from this central region of each bone chamber, stained with Masson trichrome, von Kossa, hematoxylin, and eosin staining, and examined by light microscopy.

TYROSINE-DERIVED POLYCARBONATES

- 3 a2

$

+

9 20

37

and 48 (c) weeks. The PLA channel area occupied by bone peaked at 24 weeks with a marked decrease in channel bone area observed at 48 weeks (original mag. X5).

24 weeks. At this time point, some channels were com-

- -

RESULTS

In vivo polymer degradation

The coupons made from each of the polymers re- mained physically intact for the duration of the ex- periment. The tyrosine-derived polycarbonates were cloudy at 6 weeks and became increasingly opaque at the later time points. In contrast, the PLA coupons remained transparent out to 48 weeks.

Figure 1 depicts the reduction in the (weight aver- age) molecular weight observed for all three test poly- mers. At 48 weeks, poly(DTE carbonate), poly(DTH carbonate), and poly(L-lactic acid) had respectively de- graded to 40,46, and 47% of their initial (weight aver- age) molecular weight.

Bone ingrowth

Faxitron images from PLA containing bone cham- bers are shown in Figure 2. Corresponding Faxitron images from poly(DTE carbonate) bone chambers are shown in Figure 3. Qualitatively identical images were obtained from bone chambers containing poly(DTH carbonate) (images not shown). A quantitative mea- sure of bone ingrowth was obtained by calculating the percentage of the channel area occupied by mineral- ized bone (Fig. 4).

Bone ingrowth into the implant channels was readily discernible from the Faxitron images of all three test polymers as early as 6 weeks postimplantation [Figs. 2(a) and 3(a)]. Bone ingrowth into the chambers contin- ued to progress for each of the polymers throughout

38 CHOUEKA ET AL.

Figure 3. Faxitron X-ray images of poly(DTE carbonate) chambers at 6 (a), 24 (b), and 48 (c) weeks. Bone ingrowth into poly(DTE carbonate) chambers was sustained through 48 weeks with chambers at 24 and 48 weeks being fully ingrown with bone (original mag. X5).

derived polycarbonates [Figs. 2(c) and 3(c)]. The tissue response to poly(DTE carbonate) and poly(DTH car- bonate) was, on average, characterized by continued ingrowth into the polymer test chambers. Of the two 48-week poly(DTE carbonate) and poly(DTH carbon- ate) implants, one chamber of each material was found to be fully ingrown with bone (>95% bone area mea- sured). In contrast, for the PLA implants, the channel area occupied by bone decreased to less than half of the 24-week ingrowth average.

0 poly(DTI1 carbonate) poly(DTE carbonate)

80 ' " i poly(1:lactic acid)

h

5 60 3 r2 Q 2 40 0 m

20

0 6 12 24 48

Time Point (weeks)

Figure 4. Average percent area occupied by bone in the PLA, poly(DTE carbonate), and poly(DTH carbonate) test chambers. At 6 and 12 weeks, ingrowth into the chambers was comparable for each material. At 24 weeks, ingrowth into the PLA chambers peaked and decreased dramatically at 48 weeks. In contrast, for the tyrosine-derived polycar- bonates, there is a trend toward sustained bone ingrowth at the later time points. Average and standard deviation re- ported for two test chambers per time point, 10 channels per test chamber.

Bone-implant interface

Histologic analysis focused on the interface between the bone and the test materials at the 24- and 48-week time points (Fig. 5). A representative and consistent feature of all observed bone-PLA interfaces was the presence of an intervening fibrous tissue layer [stained red in Fig. 5(b)] which prevented direct bone apposi- tion to the PLA surface. Such a fibrous tissue lining layer was consistently absent in the histologic sections of the poly(DTE carbonate) chambers [Fig. 5(d)]. This represents a fundamental difference in the host bone response to PLA and poly(DTE carbonate) coupons. For poly(DTH carbonate), direct bone-implant apposi- tion was likewise seen, albeit, to a lesser extent than that observed for the poly(DTE carbonate) implants.

Microscopic analysis of the bone present at the cham- ber channel openings revealed evidence of bone re- sorption at the entrances to the PLA-bounded channels at 48 weeks [Fig. 6(a)l. This corresponds with the bone ingrowth X-ray data, which indicated a significant de- crease in the degree of bone ingrowth by 48 weeks. Bone resorption was not observed for the tyrosine- derived polycarbonates. For both poly(DTE carbonate) and poly(DTH carbonate), bone was seen completely filling the entrances of the channels [Fig. 6(b)].

DISCUSSION

Tyrosine-derived polycarbonates represent a new class of degradable polymers derived from the natu-

TYROSINE-DERIVED POLYCARBONATES 39

Figure 6. Channel openings of PLA (a) and poly(DTE car- bonate) (b) bone chambers at 48 weeks. Scalloped bone (blue) present at the entrance of the PLA channels is indicative of bone resorption. In contrast, bone was observed filling the entrances of the poly(DTE carbonate) chambers with no evi- dence of late bone resorption (original mag. X15; stain: Mas- son trichrome, von Kossa, hematoxylin, and eosin).

rally occurring amino acid L-tyrosine. With respective strengths and stiffnesses upwards of 60 MPa and 2.1 GPa,s,'s the mechanical properties of tyrosine- derived polycarbonates are comparable to other de- gradable polymers currently being investigated for or- thopedic applications.'"

(d)

Figure 5. Histologic sections of the test chambers with the most bone ingrowth for PLA (A and B; 24 weeks) and poly(DTE carbonate) (c and d; 48 weeks). For the staining used, bone stains blue, fibrous tissue red or brown, and the implant is colorless. Arrows point to the fibrous tissue

observed at the PLA-bone interface (b) and the poly(DTE carbonate)-bone interface, which is free of an intervening fibrous tissue layer (d) (original mags. X5 and X510; stain: Masson trichrome, von Kossa, hematoxylin, and eosin).

40 CHOUEKA ET AL.

The canine bone chamber model provides a unique tool to investigate the host bone response to different materials. It has been successfully used to characterize the response of femur endosteal bone to metals, hy- droxyapatite coatings, and poly(L-lactic a ~ i d ) . ~ , ~ ~ , ~ ~ By creating 10 channels bounded by the materials under investigation, the amount of subsequent bone in- growth into these channels and the biomaterial- implant interface can be examined by micro X-ray and histologic analysis. Additionally, polymer samples can be easily retrieved from the implant site for material characterization.

For this model and implantation site, poly(DTE car- bonate) and poly(DTH carbonate) were found to de- grade in vivo at a rate comparable to PLA. Consistent with previous in vitro studies,‘* full resorption of tyrosine-derived polycarbonates was not observed over the time period of this experiment. Like PLA im- plants, resorption of tyrosine-derived polycarbonates may take as long as 2-3 years. Detailed studies of strength and stiffness retention of the tyrosine-derived polycarbonates are currently in progress.

The Faxitron X-rays yielded an effective average of bone ingrowth over the depth of the implant. The qual- itative trends observed in the Faxitron X-ray images support the quantitative assessment of the bone area present in the test chamber channels. Poly(DTE carbon- ate) and poly(DTH carbonate) chambers were charac- terized by sustained bone ingrowth throughout the 48-week duration of the study. Growth into the PLA chambers, however, peaked at 24 weeks and dropped dramatically by the 48-week time point.

This bone ingrowth sequence for PLA mimics the events seen in an earlier study that used the canine chamber model specifically to address the bone re- sponse to PLA. Suganuma et al.3 observed initial in- growth into PLA chambers 6 weeks postimplantation followed by significant bone resorption at 12- and 24- week time points. Most likely, the bone resorption phase occurred earlier in the Suganuma study because the molecular weight of the starting material was half that of the PLA used in the current study.

The presence of a fibrous layer surrounding the PLA coupons is characteristic of a mild foreign-body re- sponse. That such a fibrous layer is not as distinctly formed for poly(DTE carbonate) and poly(DTH car- bonate) chambers is an important characteristic of these polymers. Intimate contact between the bone and implant, even at 48 weeks postimplantation, is a strong indicator of the biocompatibility of the tyrosine- derived polycarbonates.

One difference between the degradation products of PLA and tyrosine-derived polycarbonates is that poly(DTE carbonate) and poly(DTH carbonate) release nonacidic degradation products.Is It has been sug- gested that the acidity of lactic acid released during the degradation of PLA can be deleterious to cells.’*J5

This fundamental difference in the nature of the degra- dation products may be partly responsible for the marked differences observed in bone ingrowth into the material test chambers and the bone-implant interface.

An earlier study employing a rabbit model com- pared the local bone response to fabricated poly(DTH carbonate) pins and clinically used polydioxanone (PDS) Orthosorb pins.8 In this 26-week study, pins were implanted transcortically in the distal femur and proxi- mal tibia of New Zealand White rabbits. The PDS im- plants were surrounded by a fibrous capsule at all time points. In contrast, the bone tissue response to poly(DTH carbonate) was characterized by a lack of fibrous capsule formation and very close contact with bone tissue. At 26 weeks, bone growth into cracks and crevices along the periphery of the implant was clearly visible. The favorable bone response to the tyrosine- derived polycarbonates in the present canine study confirms the earlier observations in the rabbit model.

Although tyrosine-derived polycarbonates do not have the peptide structure of conventional poly(amino acid)s, they are, nevertheless, amino acid-derived polymers. Thus, in the context of implantable materi- als, the immunologic properties of tyrosine-derived polycarbonates become an important concern. The 48- week in vivo response to poly(DTE carbonate) and po- ly(DTH carbonate) described is not indicative of a cell- mediated immune reaction to these materials.

Although limited by its small sample size, this study suggests that tyrosine-derived polycarbonates exhibit an in vivo bone response that is fundamentally different from the response elicited by degradable polyesters (such as PLA or polydioxanone) for which long-term, direct bone apposition is not observed. In addition, prior to this study, full bone ingrowth into the bone chambers was only seen with osteoconductive, hy- droxyapatite materials.21

Based on the results of this study, tyrosine-derived polycarbonates appear to be promising as a new class of degradable materials for orthopedic applications and are currently being evaluated in larger, longer- term animal studies.

The authors thank Dr. Sylvie I. Ertel for valuable assistance during the initiation of the project. This study was supported by the following grants: NIH GM39455, NIH GM00550, OREF 93-027, and NSF BES-9310070.

References

1. R. W. Bucholz, S. Henry, and M. B. Henley, “Fixation with bioabsorbable screws for the treatment of fractures of the ankle,” 1. Boneloint Surg. Am., 76,319-324 (1994). H. Pistner, R. Gutwald, R. Ordung, J. Reuther, and J. Muhling, ”Poly(L-lactide): A long-term degradation study in vivo. I. Biological results,” Biomuteriuls, 14,671- 677 (1993).

2.

TYROSINE-DERIVED POLYCARBONATES 41

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

J. Suganuma and H. Alexander, ”Biological response of intramedullary bone to poly-L-lactic acid,” J . Appl . Biomater., 4, 13-27 (1993). 0. Bostman, E. Hirvensalo, S. Vainionpaa, A. Makela, K. Vihtonen, P. Tormala, and P. Rokkanen, “Ankle frac- tures treated using biodegradable internal fixation,”

0. M. Bostman, ”Absorbable implants for the fixation of fractures,” J . Bone Joint Surg., 73, 148-153 (1991). H. Winet and J. 0. Hollinger, ”Incorporation of polylactide-polyglycolide in a cortical defect: Neoos- teogenesis in a bone chamber,” J . Biomed. Muter . Res.,

D. K. Gilding and A. M. Reed, ”Biodegradable poly- mers for use in surgery-poly(glycolic)/poly(lactic acid) homo- and copolymers: 1,” Polymer, 20, 1459-1464 (1979). S. I. Ertel, J. Kohn, M. C. Zimmerman, and J. R. Parsons, ”Evaluation of poly(DTH carbonate), a tyrosine-derived degradable polymer, for orthopaedic applications,” 1. Biomed. Mater. Rcs., 29, 1337-1348 (1995). J. A. Ray, N. Doddi, D. Regula, J. A, Williams, and A. Melveger, ”Polydioxanone (PDS), a novel mono- filament synthetic absorbable suture,” Surg. Gynecol. Obstet., 153, 497-507 (1981). C. G. Pitt, ”Poly-E-caprolactone and its copolymers,” in Biodegradable Polymers as Drug Delivery Systems, M. Chasin and R. Langer (eds.), Marcel Dekker, New York, 1990, pp. 71-120. J. Heller, R. V. Sparer, and G. M. Zentner, ”Poly(ortho esters),” in Biodegradable Polymers as Drug Delive y Sys- tems, M. Chasin and R. Langer (eds.), Marcel Dekker, New York, 1990, pp. 121-162. A. U. Daniels, K. P. Andriano, W. P. Smutz, M. K. 0. Chang, and J. Heller, “Evaluation of absorbable poly(ortho esters) for use in surgical implants,” J . Appl . Biomat., 5, 51-64, (1994). A. M. Radder, H. Leenders, and C. A. van Blitterswijk, “Bone-bonding behavior of poly(ethy1ene oxide)- polybutylene terephthalate copolymer coatings and

Clin. Ortkop., 238, 195-203 (1989).

27, 667-676 (1993).

14.

15.

16.

17.

18.

19.

20.

21.

22.

bulk implants: A comparative study,” Biornaterials, 16,

A. U. Daniels, M. S. Taylor, K. P. Andriano, and J. Hel- ler, “Toxicity of absorbable polymers proposed for frac- ture fixation devices,” Proc. Orthop. Res. SOC., 17, 88 (1 992). M. S. Taylor, A. U. Daniels, K. P. Andriano, and J. Hel- ler, ”Six bioabsorbable polymers: In vitro acute toxicity of accumulated degradation products,” J. Appl . Bio- rnater., 5, 151-157 (1994). J. Kohn, “Pseudo-polyfamino acids),“ Drug News Per- spect., 4, 289-294 (1991). S. Pulapura and J. Kohn, “Tyrosine derived polycarbo- nates: Backbone modified, ”pseudo”-poly(amino acids) designed for biomedical applications,” Biopolyrners, 32,

S. I. Ertel and J. Kohn, “Evaluation of a series of tyrosine- derived polycarbonates for biomaterial applications,” J. Biomed. Muter. Res., 28, 919-930 (1994). F. H. Silver, M. Marks, Y. P. Kato, C. Li, S. Pulapura, and J. Kohn, ”Tissue compatibility of tyrosine derived polycarbonates and polyiminocarbonates: An initial evaluation,” J. Long-Term Effects Med. Implants, 1, 329- 346 (1992). A. U. Daniels, M. K. 0. Chang, K. P. Andriano, and J. Heller, ”Mechanical properties of biodegradable poly- mers and composites proposed for internal fixation of bone,” J . Appl. Biomater., 1, 57-78 (1990). J. L. Ricci, J. M. Spivak, N. C. Blumenthal, and H. Alex- ander, “Modulation of bone ingrowth by surface chem- istry and roughness,” in The Bone-Biomaterial Interface, J. Davies (ed.), University of Toronto Press, Toronto,

J. M. Spivak, N. C. Blumenthal, J. L. Ricci, and H. Alex- ander, “A new canine model to evaluate the biological effects of implant materials and surface coatings on intramedullary bone ingrowth,” Biomaterials, 11,

507-513 (1995).

411-417 (1992).

1991, pp. 334-349.

79-82 (1990).

Received August 24, 1995 Accepted September 5, 1995