effect of annealing temperature on the degradation of reinforcing fibers for absorbable implants

Student Research Award in the Undergraduate, Master Candidate, or Health Science Degree Candidate Category,

21st Annual Meeting of the Society for Biomaterials, San Francisco, CA, March 18-22, 1995

Effect of annealing temperature on the degradation of reinforcing fibers for absorbable implants

Jack Choueka,'<* Jose Luis Charvet,' Harold Alexander,' Young H. Oh,' Gary Joseph,' Norman C. Blumenthal,' and William C. Lacourse' 'Department of Bioengineering, Hospital for Ioint Diseases Orthopaedic Institute, New York, New York; 21nstitu te for Glass Science and Engineering, NYS College of Ceramics, Alfred, New York

Calcium phosphate fibers designed for reinforcement of bioabsorbable fracture fixation devices were evaluated for their properties upon annealing. The composition of these fibers were 54% PO,, 27% Ca, 12% ZnO, 2.5% NaPO,, and 4.5% Fe,O,, and they were either not annealed, annealed at 250°C, or annealed at 420°C. Chemical degradation, mass loss, and morphology upon degradation were studied. Chemical degradation was performed in Tris-buffered HCl, while mass loss and morphologic studies were performed in both physiologic and nonphysiologic solutions. The results showed that degradation rates for fibers were inversely proportional to the annealing temperature. Mass loss analysis of fibers immersed in the two physiologic solutions (calf serum and simulated body fluid) revealed little change in fiber diameter up to 60 days. Morphologic examination revealed little change in fibers immersed in the two physiologic solutions until 60 days, after which thin shells

were found to be peeling off the outer coating of the fiber. Samples in tris-buffered HC1 revealed a dramatic difference in mode of degradation among the three fibers. Fibers not annealed and those annealed at lower temperatures underwent a delaminating type of degradation that appeared to destroy the overall integrity of the fiber, whereas fibers annealed at 420°C underwent crater-like deterioration in which the overall alignment of the fiber remained intact. It is therefore concluded that annealing fibers at higher temperatures slows degradation. The fibers annealed at the highest temperatures also undergo a mode of degradation that allows them to maintain their structural integrity. Al- though annealing of fibers close to glass transition temperature may produce an initially weaker fiber, chemical and physical degradation occur much slower, making these fibers most suitable for reinforcement of biodegradable implants. 0 1995 John Wiley & Sons, Inc.

INTRODUCTION

Rigid internal fixation of fractures is conventionally accomplished with metallic plates and screws. Dur- ing the early stages of fracture healing, rigid internal fixation maintains alignment and promotes primary osseous union by Unfortunately, as healing progresses, rigid fixation from bone plating

*To whom correspondence should be addressed at the Department for Orthopaedics, Hospital for Joint Diseases Orthopaedic Institute, Bernard Aronson Plaza, 301 East 17th Street, New York, NY 10003.

can cause bone in the regions of the plate to undergo stress protection atrophy. This can result in significant loss of bone mass and o s t e o p o r ~ s i s . ~ ~ In addi- tion, there is a basic mechanical incompatibility between the metal implants and bone. The elastic mod- ulus of cortical bone ranges from 17 to 24 GPa, depending upon the age and location of the speci- men,* while the commonly used alloys have moduli ranging from 110 GPa (titanium alloys) to 210 GPa (316L steel). This large difference in stiffness can result in relative motion between the implant and bone upon loading as well as high stress concentrations at bone-implant junctions.

Journal of Biomedical Materials Research, Vol. 29, 1309-1315 (1995) 0 1995 John Wiley & Sons, Inc. CCC 0021-9304195111 1309-07

CHOUEKA ET AL. 1310

Another potential problem is that the currently used alloys corrode to some degree. Ions so released have been reported to cause adverse local tissue reactions as well as allogenic responses’ and are impli- cated in adverse effects on bone mineralization as well as such adverse systemic responses as local tu- mor formation.’-12 As a consequence, it is usually recommended that a second operation be performed to remove hardware. Alternatively, the use of inter- mediately stiff plating devices has also been pro- posed4,13-15; such plates trade a certain loss in initial stability in return for some reduction in stress protection atrophy.

Research involving biodegradable polymers for fracture fixation has received great attention in recent years. Such implants potentially provide certain ad- vantages over metal implants. Absorbable implants obviate the need for a second surgical procedure to remove hardware, and they allow for the gradual transference of loading to the healing bone, thereby eliminating the problem of stress shielding.

Initial research in absorbable implants centered on the use of poly(glyco1ic acid) (PGA) and poly(1actic acid) (PLA) implants. Unreinforced PGA was found to be too brittle and absorbed too rapidly to be ade- quate for osteosynthesis.16 Further studies of PGA involved self-reinforced PGA (SRPGA) rods. Vainion- paa et al.I7 found that after 5 weeks in distilled water at 37”C, only 5% of the flexural strength of PGA was retained. Laiho et a1.’* were able to produce SRPGA rods that retained 55 and 14% of their flexural strengths after 5 and 9 weeks, respectively. However, Vasenius et al.” determined that SRPGA rods placed in subcutaneous tissue of rabbits lost nearly 50% of their strength in 2 weeks. A review of 516 patients treated with SRPGA rodsZ0 revealed complication rates of 1.2% for failure of fixation necessitating re- operation, 1.7% for bacterial infection of the operative wound, and 7.9% for late noninfectious inflammatory tissue response that warranted operative drainage. The unique complication of these implants thus ap- pears to be late inflammatory reaction.

Unreinforced PLA implants have also been studied. PLA is highly resistant to hydrolysis and therefore degrades much more slowly than PGA. Vert et a1.21 reported that poly(L-lactic acid) (PLLA) bone plates showed no loss of strength after 5 months in uivo, whereas composites of 50% PLLA and 50% poly(DL-lactic acid) (PDLLA) lost 69% of their strength after only 1 month. Leenslag et al.22,23 man- ufactured bone plates and screws and found that at 3 weeks, in vitvo specimens retained 71% of their strength while in vivo specimens retained only 33% of their strength. This discrepancy was thought to be due to stress cracking of the degrading implant.

Several clinical studies have revealed problems with the currently available polymers used for bone

fixation device^.'^ Late inflammatory reaction as a result of release of acidic degradation products represents the most serious complication of absorbable fixation devices derived from PGA and PLA. The results of recent research synthesizing a new class of biore- sorbable polymers based on the amino acid tyrosine offer the possibility of optimizing internal fixation of bone f r a ~ t u r e . ~ ~ , ‘ ~ Tyrosine-based polymers are more hydrophobic than PGA or PLA and do not release acidic degradation products or particulate debris upon degradation.

Compared to stainless steel, however, unreinforced biodegradable polymers are as much as 36% as strong in tension and 54% as strong in bending, but only 3% as stiff in either test mode.27 Low implant stiffness allows too much bone motion for satisfactory healing. Consequently, reinforcement is essential for the development of a generally acceptable absorbable fracture fixation implant material. Calcium phosphate based glass fibers have been proposed for this purpose. The solubility and degradation properties of these materials are crucial when assessing their use- fulness and biological application.

Much research on glass fibers involves surface treatments and coupling agents to enhance bonding with the host polymer.28 Another avenue of fiber modification involves annealing. The rapid cooling and pulling of fibers that occur during production cause the material to be highly stressed.” Reheating of fibers to temperatures below glass transition temperatures can decrease the chemical reactivity and increase the strength of glass fibers by reforming bonds that may have been broken during production and allowing bonds that were frozen in nonequilibrium states to rotate into more stable configurations.

This study focused on the characterization of fibers intended for the development of a bioabsorbable composite with tyrosine-based polymers. Fiber ex- perimentation, performed in vitro, focused on modification of the fibers by annealing. Several experiments were designed to evaluate the degradation properties of these fibers in physiologic and nonphysiologic solutions. These included chemical degradation analysis, analysis of mass loss, and morphologic examination of degraded fibers.

MATERIALS AND METHODS

All fibers were of the composition 54% POa, 27% Ca, 12% ZnO, 4.5% Fez03, and 2.5% NaPO,. Fibers were drawn by standard technique using bulk glass placed in a platinum crucible heated to 1000°C until a meniscus of glass appeared at the crucible opening. A glass rod was used to capture the melted glass and draw it onto a spinning drum, which proceeded to

ANNEALING TEMPERATURE OF REINFORCING FIBERS 1311

wind a single continuous fiber at room temperature in an open-air environment. Drum speed, crucible temperature, crucible opening diameter, and room conditions were kept constant to maintain consistent fiber properties. The fibers collected were separated into three groups. The first group was reheated to 250°C, the second group was reheated to 420"C, and a third group received no further treatment. Fibers were not pretensioned prior to annealing. A11 fibers were kept in a vacuum desiccator until the time of testing to avoid degradation.

Chemical degradation of fibers was performed in tris-buffered HCl solution maintained at neutral pH. We placed 100 mg of each fiber in a mesh capsule and immersed it in 200 cc of tris-buffered HCl solution. A closed, continuously circulating environment was produced using a Masterflex pump with solution circulating in Tygon tubing. An isothermal environment was maintained at 39°C to mimic a physiologic environment. Three pump heads were employed so that all three fibers could be tested at once, thus control- ling for environmental variables. A calcium electrode (Microelectrics, Londonderry, NH) was used to measure the calcium ion concentration of the solution at any given time. An operational amplifier was inte- grated between the electrode and the voltmeter to amplify the output signal. This was calibrated to known concentrations of calcium in solution. Calibra- tion solutions were maintained in the same heated bath at the same temperature and pH. The calcium present in the solution was assumed to be a product of the degraded fibers. Control tests were performed to confirm that no other source of calcium was present. Calibration curves were generated before each reading to quantify the calcium concentration of the test solution. Readings were obtained every 10 min for the 1st h, hourly for the next 6 h, and every 6 h for the remainder of the experiment. Each fiber type was tested in five experimental runs to obtain data for statistical analysis. The kinetics of calcium degradation for each fiber was thus determined. Lin- ear regression analysis was used to obtain a best-fit line, with a Pearson correlation coefficient of 0.9 con- sidered to be acceptable. The slope of the degradation curve was thus determined to be the degradation rate. Unpaired t tests were used for statistical analysis of the slopes.

To determine the effects of fiber degradation in the physiologic environment, mass loss analysis was performed in both calf serum and simulated body fluid. Tests were also performed in tris-buffered HC1 for a nonphysiologic comparison. Test capsules were made by placing 10 mg of fiber in a porous peg jacket. Seven capsules of each fiber type were prepared for each solution. Each group was then immersed in one of the three test solutions. One capsule of each fiber type was removed from each solution at seven differ-

ent time intervals for a total of nine tests per time period (three fiber types in three solutions). pH of the test solution was determined at each time period to ensure that degradation was not the result of bacterial overgrowth. Sodium azide was added to the test solutions at the outset of the experiment to prevent bacterial overgrowth. Fibers were analyzed on days 1, 2, 4, 7, 11, 14, 36, 60, and 90 by examination under light microscopy using a computerized enhancement soft- ware package (TCL image) to measure fiber diameter. Ten independent fibers were examined from each capsule. Mass loss analysis was determined by comparison of degraded to undegraded fiber diameters.

The fibers remaining in each capsule after mass loss analysis were examined under scanning electron microscopy for morphologic changes as a result of degradation. Fibers were visualized both longitudinally and in cross section for accurate determination of the degradation mode.

RESULTS

The calcium degradation rate of the fibers, mea- sured as a function of calcium ion release into the solution, was linear (? > 0.98) for the first 75 h of testing, after which little change in calcium concentration appeared in the fluid. In the linear region, rates of calcium uptake by the solution (in millimoles per minute) for the three fiber groups were 4.9 x

for fibers not annealed, 3.6 X l o p 5 for fibers annealed at 250"C, and 2.4 x for fibers annealed at 420°C (Fig. 1). Statistical significance existed among all these groups. Rates of fiber degradation were inversely proportional to annealing temperature.

Mass loss analysis of fibers immersed in the two physiologic solutions (calf serum and simulated body fluid) revealed an initial increase in diameter in both, which likely represents swelling or water retention (Figs. 2 and 3 ) . Following day 1, there was little change in fiber diameter up to 60 days. Diameters for the 90-day samples were unobtainable because of morphologic changes in the fibers (described below). The pH of the solutions remained essentially constant throughout the testing period. Accurate analysis of the diameter of fibers immersed in tris-buffered HC1 was not possible because of the mode and rate of degradation; even within the 1st day of testing, there was marked destruction of fiber structure that pre- cluded accurate measurements. The results for these samples are better analyzed morphologically.

Morphologic examination revealed virtually no change in fiber character in the two physiologic solutions up to the first 60 days of testing (Fig. 4). At 60-90 days, fibers in both solutions revealed cracking and pealing of an outer shell (Fig. 5). The thickness of

1312

Rate of' Calcium Uptake in Tris-Buffered HCI

CHOUEKA ET AL.

Mass Loss of Fihers in Simulated Body Fluid

XSe-4

3 02-4

2 . 5 4

h - 2 Oe-4 E - 2 I.Se-4

I .oe-4

/---

- N O ~ Hear Treated ?SO C Hcai Treated 420°C Heat Treated

, , ~ ~ , , ~ , , ~ , ~ ~ ~ , ~ ~ , , ~ ~

.. 5 Oe-5 ".*... a or+a

o 25 50 7 5 loo 125 150 175 200 225

Timc (houi\)

Figure 1. Calcium uptake in the solution over time. In- creased rate of calcium uptake in solution corresponded to increased degradation rates. In the linear region annealing temperature was inversely proportional to degradation rate. Statistical significance, P < .05, existed between each group.

the shell appeared to be quite minimal in comparison to the overall diameter of the fiber. Fibers of all three groups shared this method of degradation. Samples in tris-buffered HC1 revealed a dramatic difference in mode of degradation among the three fibers. Fibers not annealed and those annealed at 250°C underwent a delaminating type of degradation that appeared to destroy the overall integrity of the fiber [Fig. 6(a) and (b)], whereas fibers annealed at 420°C underwent cra- terlike deterioration in which the overall alignment of the fiber remained intact (Fig. 7). This latter mode of

Mass Loss of Fihers in Calf Serum

-c- Non Heat Treated Fihers ...... 0 ....... .. .b.. .

25042 Heat Treated Fibers 420.C Heat Treated Fihers

- N 1 z

Time (days)

Figure 2. Fiber diameter over time with degradation in calf serum. Each point represents the average diameter of 10 fibers. There was no statistical difference, P < .05, in changes in diameter of each fiber type throughout the time interval tested.

............. ........ *..... .................. I

.... 0 .... 250C Heat Treated Fiber\

.. .&. .. 420C Heat Treated Fihers

'lime (days)

Figure 3. Fiber diameter over time with degradation in simulated body fluid. Each point represents the average diameter of 10 fibers. There was no statistical difference, P > .05, in changes in diameter of each fiber type throughout the time interval tested.

degradation appeared by day 1, and the fibers were nearly completely degraded by day 7 of testing.

DISCUSSION

This study was conducted to explore the use of postfabrication annealing and its applicability to fiber enhancement. Experiments were performed in physiologic solutions to simulate the biologic environment, and in nonphysiologic solutions to expedite the degradation process for the purpose of examining differences among fibers.

Characterization of fiber degradation is essential to developing fully absorbable composites. All fibers used in this experiment differed only in their post-

Figure 4. Non-heat-treated fiber degraded in calf serum for 2 weeks. Note that the surface of the fiber is virtually unaffected by immersion in a physiologic solution.


Figure 5. Heat-treated fibers in serum at 90 days. An outer shell is seen peeling from the outer layer of the fiber. This may represent a precipitate that formed on the surface, or the outer layer of the fiber itself. In either case, the overall structure of the fiber is maintained.

fabrication heat treatment; otherwise they were produced in an identical manner. Degradation rates were inversely proportional to annealing temperature (i.e., higher temperatures increased resistance to degradation). Degradation rates were calculated for the first 75 h of testing because after this time the curves be- gan to level off because of either supersaturation of the solution or complete degradation of the fibers.

The mode of degradation in tris-buffered HC1 corresponded well with the chemical degradation studies. Fibers that were not annealed underwent a more destructive form of degradation than those that were annealed. It is clear from the morphologic analysis that greater volumes of nonannealed fiber were degraded at a faster rate. This would correspond to increased rates of calcium dissolution.

Evaluation of mass loss in the two physiologic solutions did not show appreciable changes for either fiber group. A common ion effect, in which dissolution of calcium and phosphate into solutions that have significant concentrations of these elements is inhibited, is the probable cause. Fibers dissolved in calf serum and simulated body fluid viewed under SEM showed virtually no change in fiber morphology until the last test period. At 90 days, there was only minor cracking and peeling of a thin outer shell. It is unclear whether this shell was the outer coating of the fiber or a precipitate that formed on the fiber surface. In either case, the overall integrity of the fiber appeared to be intact at this time period.

The production of fibers places inherent stresses on the glass. Rapid cooling in combination with the pulling process causes some bonds to be frozen in lengths greater than equilibrium, some bond angles to be in nonequilibrium, and some bonds actually to be broken. This results in a highly reactive glass. Reheating

Figure 6. Non-heat-treated fibers in tris-buffered HCl at 2 days. (a) The destructive delaminating mode of degradation is apparent. (b) In cross section, the delamination oc- curring is more evident.

the glass to temperatures below the glass transition temperature can obviate some of the effects of stress- ing and cooling by reforming bonds and allowing some bonds to rotate into more stable configurations. The chemical reactivity is therefore decreased. An-

Figure 7. Fibers annealed to 420°C in tris-buffered HC1 at 2 days. A crater-like mode of degradation is seen. The structure of the fiber is maintained.

1314 CHOUEKA ET AL.

nealing of fibers close to the glass transition temperature will stabilize more bonds; however, the strength may be decreased as the properties of the fiber approach those of bulk glass. Therefore, a near completely annealed fiber, although initially weaker, will undergo chemical and physical degradation at a much slower rate, allowing it to retain its original strength for longer time periods, a feature appropri- ate for a degradable implant.

The next generation of biodegradable implants will rely on fiber reinforcement for effectiveness as fracture fixation devices. The fibers should withstand degradation in the early stages of implantation, undergo degradation in a manner that least effects their integrity to withstand forces, and be fully degradable. The results of this study indicate that annealing fibers enhances these properties. Mechanical testing of degraded fibers as well as composite studies should be performed to further evaluate these fibers.

SUMMARY

Annealing fibers at higher temperatures slows degradation by 50% in tris-buffered HCl. Fibers annealed at the highest temperatures also undergo a mode of degradation that allows them to maintain their structural integrity for longer time periods. These properties make high-temperature-annealed fibers most suitable for reinforcement of biodegradable implants. On the basis of this study, further evaluation is warranted of a composite composed of poly(DTE carbon- ate) and calcium phosphate based fibers annealed at high temperatures.

References

1.

2.

3.

4.

5.

6.

P. Hutzchenreuiter, S. M. Perren, S. Steinmann, V. Geret, and M. Klebl, ”Some effects of rigidity of internal fixation on the healing pattern of osteotomies,”

M. D. Muller, M. Allgower, and H. Willenegger, eds., Manual of Internal Fixation, Springer-Verlag, Berlin, 1970. H. K. Uhtoff and F. L. Dubuc, “Bone structure changes in the dog under rigid internal fixation,” Clin. Orthop. Rel. Res., 81, 165-170 (1971). W. H. Akeson, S. L.-Y. Woo, R. D. Coutts, J. V. Mathews, M. Gonsalves, and D. Ameris, “Quantita- tive histological evaluation of early fracture healing of cortical bones immobilized by stainless steel and composite plates,” Calcif. Tiss. Res., 19, 27-37 (1975). W. H. Akeson, S. L.-Y. Woo, L. Rutherford, R. D. Coutts M. Gonsalves, and D. Amiel, “The effects of rigidity of internal fixation plates on long bone remod- eling,” Acta Orthop. Scand., 47, 241-249 (1976). A. J. Tonino, C. L. Davidson, P. J. Hopper, andL. A.

Injury, 1, 77-81 (1969).

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

Linclav, “Protection from stress in bone and its effects,” J . Bone Joint Surg., 58B, 107-113. H. K. Uhtoff and F. L. Dubuc, ”Bone structural changes in the dog under rigid internal fixation,” Clin. Orfhop. Re]. Res., 81, 165170 (1970). F. G. Evans and A. Vioncentelli, ”Relation of the com- pressive properties of human cortical bone to histological structure and calcification,” J . Biomechanics, 7, 1 (1974). J . Black, Biological Performance of Materials, Marcel De- kker, New York, 1981. K. Sinibaldik, H. Rosen, S.-K. Lin, and M. De Angelis, ”Tumors associated with metallic implants in ani- mals,” Clin. Orthop. Rel. Res., 118, 257-266 (1976). J . Lamovec, A. Zidar, and M. Cucek-Plenicar, “Syno- vial sarcoma associated with total hip replacement,” J . Bone Joint Surg., 70A, 155S1560 (1988). A. Martin, T. W. Bauer, M. T. Manley, and K. E. Marks, ”Osteosarcoma at the site of total hip replacement,” 7. Bone Joint Surg., 70A, 1561-1567 (1988). R. E. Coutts, W. H. Akeson, S. Woo, J . V. Mathews, M. Gonsalves, and D. Amiel, ”Comparison of stainless steel and composite plates in the healing of di- aphyseal osteotomies of the dog radius,” Ortho. Clin. North. A m . , 7 , 223-229. H. K. Uhthoff and D. Bardos, “The use of titanium-6 A1-4V plates in the treatment of fracture,” 24th An- nual Meeting of the Orthopedic Research Society, Dal- las, TX, Feb. 21-23, 1978. S. L.-Y. Woo, B. R. Simon, W. H. Akeson, M. A. Go- mez, and Y. Seguchi, ”A new approach to the design of internal fixation plates,” J. Biomed. Muter. Res., 17, 427439 (1983). P. Christel, F. Chabot, J . L. Leray, C. Morin, and M. Vert, “Biodegradable composites for internal fixation,” in Biomaterials 1980, G. D. Winter, D. F. Gib- bons, and H. Plenck, Jr. (eds), Wiley, 271-280 (1982). S. Vainionpaa, J. Kilpikari, J. Laiho, P. Helevirta, P. Rokanen, and P. Tormala, ”Strength and strength retention in vitro, of absorbable, self-reinforced polyglycolide (PGA) rods for fracture fixation,” Biomaterials 8, 46-48 (1987). J. Laiho, T. Mikkonen, and P. Tormala, “A comparison of in vitro degradation of biodegradable polyglycolide (PGA) sutures and rods,” Trans. Soc. Biomater., 11, 564 (1988). J. Vasenius, S. Vainionpaa, K. Vihtonen, M. Mero, J. Mikkola, P. Rokkanen, and P. Tormala, “Biodegrad- able self-reinforced polyglycolide (SR-PGA) composite rods coated with slowly biodegradable polymers for fracture fixation: Strength and strength retention in vitro and in vivo,” Clin. Muter., 4, 307-317 (1989). 0. Bostman, E. Hirvensalo, J . Makinen, and P. Rok- kanen, ”Foreign-body reactions to fracture fixation implants of biodegradable synthetic polymers,” I . Bone Joint Surg., 72B, 592-596 (1990). M. Vert, P. Christel, F. Chabot, and J. Leray, ”Biore- sorbable plastic materials for bone surgery,” in Mac- romolecular Biomaterials, G. W. Hastings and P. Duch- eyne (eds.) CRC Press, Boca Raton, FL, 1984. J. W. Leenslag, A. J. Pennings, R. R. M. Bos, F. R. Rozema, and G. Boering, ”Resorbable materials of poly(L-lactide). VI. Plates and screws for internal fracture fixation,” Biomaterials, 8, 7C-73 (1987). J. W. Leenslag, A. J. Pennings, R. R. M. Bos, F. R. Rozema, and G. Boering, “Resorbable materials of poly(L-lactide). VII. In vivo and in vitro degradation,” Biomaterials, 8, 311-314 (1987).


24.

25.

0. Bostman, ”Absorbable implants for the fixation of fractures,” J. Bone Joint Surg., 73, 14fk153 (1991). S. Pulapura and J. Kohn, ”Tyrosine derived polycar- bonates: Backbone modified, ’pseudo’ poly(amino acids) designed for biomedical applications,” Biopoly- mers, 32, 411 (1992).

26. J . Zhou, S. I. Ertel, H. M. Buettner, and J. Kohn, ”Evaluation of tyrosine-derived pseudopoly(amino acids): In vitro cell interactions,” 20th Annual Meeting of the Society for Biomaterials, Boston, 1994. A. U. Daniels, M. K. 0. Chang, andK. P. Andriano, ”Mechanical properties of biodegradable polymers

27.

and composites proposed for internal fixation of bone,” J. Appl. Biornater., 1, 57-78 (1990). Y. B. Peng, J. B. Koen, R. F. Brown, D. E. Day, and H. L. Collier, ”Bioabsorbable glass fiber polymer composite: Part 1, preparation and mechanical properties,” Trans. Soc. Biomater., 20, 38 (1994). A. K. Varshneya, Fundamentals of lnorganic Glasses, Ac- ademic Press, New York, 1993.

28.

29.

Received October 15, 1994 Accepted March 10, 1995

effect of annealing temperature on the degradation of reinforcing fibers for absorbable implants

Documents