Degradation of poly(D,L)lactide implants with or without addition of calciumphosphates in vivo

Biomaterials 22 (2001) 2371}2381

Degradation of poly(D,L)lactide implants with or without addition of calciumphosphates in vivo Wolfgang Heidemann *, Stephanie Jeschkeit , Kurt Ru$eux , JuK rgen Hartmut Fischer , Mathias Wagner , G. KruK ger , Erich Wintermantel , Klaus Louis Gerlach Department of Maxillofacial Surgery, Otto-von-Guericke-University, Leipziger Str. 44, D-39120 Magdeburg, Germany Institute of Experimental Medicine, University of Cologne, Robert-Kochstr. 10, D}50931 Cologne, Germany Degradable Solutions AG, Wagistr. 23, CH-8952 Schlieren, Switzerland Institute of Pathology, University of Cologne, Joseph-Stelzmann-Str. 9, D-50931 Cologne, Germany Department of Materials, Biocompatible Materials Science and Engineering, ETH Zurich, Wagistr. 23, CH-8952 Schlieren, Switzerland

Abstract The study was aimed at examining the in vivo degradation of pure poly(D,L)lactide (PDLLA) or PDLLA with an admixture of calciumphosphates. One rod (20;3;2 mm) and one cube (3;2;2 mm) of pure PDLLA, PDLLA with tricalciumphosphate (PDLLA#TCP) or PDLLA with calciumhydrogenphosphate (PDLLA#CHP), respectively, were implanted into the dorsal muscles of 50 male Wistar Albino rats. After de"nite intervals (from 2nd to 72nd week), pH measurements were performed in the environment of the implants. Afterwards, the cubes with their surrounding tissues were excised for histological examinations, measurements of the outer dimensions and mechanical analyses of the explanted rods were performed. No drop of more than 0.1 pH units was detectable in the tissue surrounding any type of implants. No advantageous e!ect of the calciumphosphates could be proved. A mild foreign body reaction could be observed around PDLLA implants. After 72 weeks, pure PDLLA had been totally resorbed from the extracellular space, the degradation of calciumphosphate-enriched PDLLA was still in progress. A large amount of in#ammations occurred in the tissues surrounding PDLLA with an admixture of slowly degrading TCP or CHP, leading to two abscesses and four "stulas at PDLLA#TCP, and two abscesses and three "stulas at PDLLA#CHP implantation site. Bending strength of pure PDLLA was constant up to the 4th week post-implantation and reduced to 60% of the initial value up to the 12th week. No traces of crystallinity could be observed during the degradation of PDLLA. As a conclusion of the study, complete resorption from the extracellular space and tissue tolerance of pure PDLLA is proved. An admixture of small calciumphosphate particles is not suitable to improve the biocompatibility of PDLLA but leads to a decrease in the mechanical characteristics. 2001 Elsevier Science Ltd. All rights reserved.

Keywords: Poly(D,L)lactide; Mechanical analysis; Calciumphosphates

1. Introduction Use of metallic osteosynthesis material is a common practice in traumatology and orthopaedic surgery, leading to good clinical results by stable "xation of the bony fragments. But the de"ciency of load transmission during the process of bone healing, due to stress protection by the rigid metallic osteosynthesis plates [1,2] and some possible disadvantages of long-lasting metallic "xation as * Correspondence address: UniversitaK tsklinik fuK r Mund-, Kiefer- und Gesichtschirurgie, Leipziger Str. 44, D-39120 Magdeburg, Germany.

Tel.:#49-391-6715170; fax:#49-391-6715172.

E-mail address: wolfgang.heidemann@medizin.uni-magdeburg.de (W. Heidemann).

in#ammative reactions of the surrounding tissues or allergic reactions, caused by corrosion products [3}5], furthermore, the disturbance of CT and, especially, of MRT examinations and "nally the transcranial migration of osteosynthesis screws, seen occasionally after craniomaxillofacial surgery of infants [6}8], require the surgical removal of the metallic implants after fracture healing.

To prevent an atrophy of bone just as the second operative procedure for the removal of the metallic implants, multiple biodegradable osteosynthesis materials were developed, especially, as modi"cations of di!erent poly- -hydroxy acids [9}21], which are less sti!er than the metallic implants and should be completely resorbed after a de"nite period. Especially, poly(L)lactide (PLLA)

0142-9612/01/$ - see front matter 2001 Elsevier Science Ltd. All rights reserved.

PII: S 0 1 4 2 - 9 6 1 2 ( 0 0 ) 0 0 4 2 4 - 5

2372

W. Heidemann et al. / Biomaterials 22 (2001) 2371}2381

was of interest because of its proven biocompatibility and high initial mechanical strength and was tested in di!erent clinical studies [10,20,22}25]. But the long-lasting degradation time and late foreign body reactions * due to crystalline remnants or a decrease of the pH value during the degradation * did not allow a general introduction for clinical use [26}29].

A solution to these problems might be seen in the use of amorphous poly(D,L)lactide (PDLLA) which degrades in vitro without generation of any crystalline remnants [30]. PDLLA is divided during its degradation into a fast degrading center and a slowly degrading outer layer, which stays intact and retains degradation products until the swelling of the implants or mechanical disorders cause it to break, after about 32 weeks in vitro [30].

A faster degradation in the interior is caused by the concentration of the acidic degradation products, leading to further stimulation of degradation. The disruption of the outer layer leads to a rapid release of acidic residues with a subsequent lowering of pH in vitro [30}34].

A similar drop in the pH would probably cause in#ammatory tissue reactions in vivo [35], and would limit the clinical use of PDLLA; the lowering of pH might be prevented by an admixture of bu!ering agents to PDLLA. The aim of this in vivo study was to test the reaction of soft tissue on PDLLA implants. Furthermore, it was evaluated, whether any pH decrease could be detected in the tissue surrounding pure PDLLA or whether an addition of TCP or CHP as bu!ering agents to PDLLA might be able to reduce the acidity of degrading PDLLA. Finally, the mechanical characteristics of the implants were examined by measuring the molecular weight (M ) and the mechanical properties as well as by

fracture and surface analysis using scanning electron microscopy.

2. Material and methods 2.1. Production of PDLLA implants with or without an admixture of TCP or CHP TCP (Fluka AG, Switzerland, purity 99%) and CHP (Fluka AG, Switzerland, purity 97%) particles were immersed in ethanol solution and ground by a mill to average particle size of 2.2 m (Ru$eux, 1997). The size of the particles was measured using a laser}optical spectrometer (Sympatec, type Sucell Cl, ETH Zurich). The ethanol was removed in a drying process at room temperature for 24 h and at 1003C for 3 h.

Racemic PDLLA granules (RESOMER威 R 207; Boehringer-Ingelheim, Germany; M "220,000 g/mol) con taining 50% D-lactide monomers and 50% L-lactide monomers were injection-molded in a machine-type KloK ckner FX 75.9% (molar) of TCP and due to its higher

solubility, only 4.6% (molar) of CHP was added to the polymer during the injection moulding process. Sterilization of the produced rods (20;3;2 mm) and cubes (3;2;2 mm) was performed by -irradiation (25 kGy).

2.2. Implantation procedure Under general anaesthesia, with 40 mg/kg body weight, sodium-pentobarbital (Nembutal威 0.06 g/ml) and ether/O the back skin of 50 rats (body weight 150}180 g)

was shaved and disinfected with Polyvidon-Iod (Betaisodona威-solution, Mundipharma GmbH, Limburg (Lahn), Germany). After incision of the skin, one rod and one cube of PDLLA, PDLLA#TCP or PDLLA# CHP, respectively, were implanted under sterile conditions into the dorsal muscles of each rat. The implants were perforated at one end and "xed in the tissue with non-resorbable 5-0 Polypropylen sutures (Prolene威 blue). Closure of the skin was performed using degradable atraumatic 3-0 Polyglactin 910-sutures (Vicryl威).

At the end of the operation, Flunixin-Meglumin (Finadyne威 Essex, Tierarznei, MuK nchen) 1.1 mg/kg body weight, was administered as an analgesic drug. Postoperatively, the rats got standard diet (Altromin威, Lage, Germany) and water ad libitum.

2.3. Determination of pH value At regular time intervals (4, 6, 8, 12, 16, 20, 24, 28, 32, 40, 48, 56, 64, and 72 weeks after implantation) three randomly sampled rats were narcotized with ether/O .

The pH values of the normal muscles, at least 2 cm away from the implants, and the pH values of the tissue surrounding the implants were measured using a puncture micro pH electrode (Ingold Co., Weinheim, Germany) and a pH meter (Wissenschaftlich Technische WerkstaK tten (WTW) Co., Weinheim, Germany). To examine the in#uence of respiratory activity on the pH value of the tissue during the anaesthesia, an intracardial withdrawal of blood was performed after having "nished the measurements of the tissue pH. The pH value of the blood was measured using an Acid}Base Laboratory, ABL 300 (Radiometer Copenhagen, Denmark).

Finally, after sacri"cing the animals with an ether overdose, the rods were taken out for mechanical tests and the cubes with their surrounding tissue were excised en bloc for histological examinations.

2.4. Histologic examinations The samples were stored for at least 30 days in 4% bu!ered formalin. To detect even minute amounts of of intra- and extracellular PDLLA or PDLLA catabolites by microscopy using polarized light and #uorescence microscopy ("lter set: Green H546-FT580-LP590; Axiophot威 photo microscope, Carl Zeiss, Oberkochen,

W. Heidemann et al. / Biomaterials 22 (2001) 2371}2381

Germany), samples were embedded in Paraplast威 tissueembedding medium (Sherwood Medical Co., St. Louis, MO, USA; melting point: 563C) up to the 40th postimplantation week, where visible macroscopic residues of PDLLA cubes could be found. The sections were cut with a microtome (Hn 40, Reichert-Jung GmbH, Nussloch, Germany) to obtain nine representative slides (i.e., three serial slides of 5}7 m thickness in a set; three sets; set-to-set distance: '50 m). The slides were subsequently depara$nized by two consecutive treatments with xylene followed by a rehydration procedure (sequential immersion in graded alcohol) and submitted to conventional histological staining (hematoxylin and eosin, H & E) for light microscopy. The periodic acid Schi!'s (PAS) cytochemical reaction for carbohydrates, and Elastica}van Gieson's staining (EvG) that reveals collagen were applied to alternating sections * in each of the three sets: slide 1, H & E; slide 2, EvG; slide 3, PAS. From the 48th week up to the 72nd week, samples were dehydrated using 70, 85, 96 or 100% alcohol and were embedded for about 24 h at 43C in a 1 : 1 mixture of butylmetacrylate and methylmetacrylate. Sections were cut with a microtome (Hn 40, Reichert-Jung GmbH, Nussloch, Germany) to obtain serial slides of 3 m thickness. Depolymerization of the sections was performed using -methoxymethylacetate, the slides were submitted to toluidine blue, 5 min for histological staining.

The number of leukocytes and macrophages was determined per high-power "eld magni"cation (HPF; 400;); *6 HPFs/slide (n ) were analyzed semi-quant itatively and then averaged. Cells were classi"ed as either absent, scarce (i.e., '0 but )5 cells/slide), occasional (i.e., 1 leukocyte or macrophage to 1/3 of all cells per HPF are leukocytes or macrophages in *4 HPFs), moderate (i.e., '1/3}2/3 of all cells per HPF in *4 HPFs), or abundant (i.e., '2/3}3/3 of all cells per HPF in *4 HPFs).

Connective tissue was classi"ed as either absent, scarce (i.e., minute amounts of connective tissue in (4 HPFs), occasional (i.e., )1/3 of a HPF is connective tissue in *4 HPFs), moderate (i.e., '1/3}2/3 of a HPF is connective tissue in *4 HPFs), or abundant (i.e., '2/3}3/3 of a HPF is connective tissue in *4 HPFs).

2.5. Material analysis To assesss macroscopic changes, the implant's length, volume and wet weight were determined. The molecular weight (M ) was measured using gel-permeation

chromatography (GPC, Knauer, Germany) in tetrahydrofurane comparing the results with polystyrene standards. The bending strength was measured in a 3-point bending test at room temperature using a universal testing machine (Zwick威 1456). Testing speed was 1 mm/min, supporting length was 11 mm. Glass transition temperature of the samples (5}12 mg) was determined by a

2373

di!erential scanning calorimeter (DSC 910, Du Pont Instruments) with a heating rate of 103C/min up to 1503C. Scanning electron microscopic (SEM) examinations (7 kV, C2500 with RE-Detector, Hitachi) were performed after coating the samples with platinum (10 mA, 600 s).

3. Results 3.1. Macroscopic degradation process All rats recovered well from the implantation procedure; post-operative healing was achieved in all the rats without complications.

While the pure PDLLA rods were transparent before the implantation, 4 weeks after implantation the clearness of PDLLA implants changed into opacity. In the following weeks, only gradual macroscopic changes occurred up to the 16th week post-implantation. In the 20th week plump extrusions were observed at the surface of pure PDLLA rods; cracks arose in the small solid outer coat in the 24th week, leading to an increased fragility of the implants. Two of three pure PDLLA rods broke during the removal in the 24th week and gelatinous content leaked out. Thirty-two weeks post-implantation only small PDLLA fragments could be obtained, the surrounding capsule was not removed in the subsequent weeks to prevent damaging of the gelatinous content.

A rapid dissolution of the pure PDLLA rods took place in the following period up to the 56th week; thereafter, only a thin scar was macroscopically detectable at the implantation site of PDLLA rods.

In PDLLA#CHP or PDLLA#TCP implants no remarkable macroscopic alterations were found up to the 20th week post-implantation; in the 24th week an unevenness occurred at the former smooth surface of PDLLA#CHP or PDLLA#TCP implants which developed into extrusions at PDLLA#CHP-rods in the 28th week. Similar changes occurred at the surface of PDLLA#TCP rods in the 32nd week post-implantation. After the 48th week, the implants containing TCP were obtained together with the surrounding capsule because of the increasing fragility of PDLLA#TCP rods. Maximum size of PDLLA#CHP rods was observed in the 64th week post-implantation; at the end of the observation time in the 72nd week post-implantation, PDLLA#TCP rods had a gelatinous-soft consistency similar to what was found inside the pure PDLLA rods in the 24th week. The PDLLA#CHP rods were yet solid, but fragile; small fragments crumbled o! during the removal.

3.2. pH values A signi"cant decrease of pH values in the tissues surrounding the implants could not be measured during the

2374

W. Heidemann et al. / Biomaterials 22 (2001) 2371}2381 Table 1 Tissue reaction to pure PDLLA. Results of histological examination

Macrophages Giant cells Lymphocytes Connective tissue Blood vessels

Weeks 4}8

Weeks 28}32

*Week 56

## ## # # (#)

## # ## #/## ##

(#) ! ## # #

!"absent, (#)"scarce, #"occasional, ##"moderate, #/##"abundant.

Fig. 1. Di!erences of pH between the normal muscle tissue and the tissue, surrounding the implants. Each datapoint represents the mean value $SD for a group of 3 rats, respectively.

period of observation; in Fig. 1 the pH values are given as di!erences between the values near to the implants and the values in normal muscle tissue. The mean pH values at the surface of pure PDLLA implants was between 7.25 and 7.48, the mean pH values near PDLLA#CHP rods ranged between 7.29 and 7.45 and the mean pH values close to PDLLA#TCP rods was between 7.18 and 7.46.

Measurements of the blood pH showed reliable agreement with the pH values in the normal tissues; the di!erences of the pH values in normal muscles and in the blood ranged between !0.04 and 0.22 units with a mean of 0.09 units (SD"0.07). No remarkable in#uence of TCP or CHP addition on the pH value with di!erences of more than 0.2 pH units, could be detected.

3.3. Tissue reactions After 4 weeks the implants were totally encapsulated by a thin layer of "broblasts, building a "brous tissue capsule up to the 6th week. The cellular reactions to pure, TCP-enriched, and CHP-enriched PDLLA implants showed minute di!erences in quantity (Tables 1}3).

Leukocytes, (foamy) macrophages, polynuclear giant cells and some intra- and extracellular PDLLA particles and/or PDLLA catabolites composed a regular foreign body reaction. Multinuclear giant cells were located at the inner and outer side of a maturing "brous capsule which was formed of homogeneous layers of connective tissue surrounding the implantation site. Fibrous bridges were seen around isolated PDLLA particles and/or PDLLA catabolites as "broblasts and collagen encapsulated the implants. The amount of connective tissue formation and vascularization did not di!er signi"cantly among the PDLLA variants. The auto#uorescence of the polymer particles appeared to be unpredictable in that only few scraps emitted bright green light (Fig. 2) and in

Table 2 Tissue reaction to CHP-enriched PDLLA. Results of histological examination

Macrophages Giant cells Lymphocytes Connective tissue Blood vessels

Weeks 4}8

Weeks 28}32

*Week 56

## ## # # (#)

## #/## # #/## (#)

(#) (#) (#) #/## #

(#)"scarce, #"occasional, ##"moderate, #/##"abundant.

Table 3 Tissue reaction to TCP-enriched PDLLA. Results of histological examination

Macrophages Giant cells Lymphocytes Connective tissue Blood vessels

Weeks 4}8

Weeks 28}32

*Week 56

## ## ## # #

## #/## # ## #

# # # #/## #

#"occasional, ##"moderate, #/##"abundant.

that the occurrence of this luminescence did not correlate with any other examined feature (e.g.: intra- or extracellular position of the particle, week of removal, enriched or pure PDLLA, its behavior in H & E, PAS, and/or EvG staining). Extracellular signaling (in those cases that showed auto#uorescence) was preponderant in regions formed by the cones prior to wax embedding.

Signs of extra- and intracellular PDLLA particles and/or PDLLA catabolites such as caves or vacuoles remained visible in pure PDLLA samples up to the 48th week post-implantation (Fig. 3). In the 72nd week, no residues of the former cube are demonstrable; solely foamy macrophages and giant cells, containing minute amounts of polymeric debris whose resorption was under progress (Fig. 4), were surrounded by "brous tissue near

W. Heidemann et al. / Biomaterials 22 (2001) 2371}2381

Fig. 2. Pure PDLLA, 16 weeks post-implantation, H & E##uorescence, 200;. Various light-emitting extra- and intracellular (P) PDLLA scraps are visible in the tissue surrounding the PDLLA cube.

Fig. 3. Pure PDLLA, 48 weeks post-implantation, toluidine blue, 400;. Multiple vacuoles (P) can be seen, containing PDLLA residues or catabolites which have been removed by the embedding procedure.

to the non-resorbable Prolene威 suture. Neither extracellular remnants of polylactides nor caves or vacuoles could be discovered.

In samples, containing TCP-enriched PDLLA, residues of the cube are macroscopically detectable in the 72nd week. The polymer is surrounded by a broad wall of phagocytizing macrophages and giant cells, intracellular

2375

Fig. 4. Pure PDLLA, 72 weeks post-implantation, toluidine blue, 400;. Foamy macrophages and lymphocytes, containing residues of polymer catabolites (P), compose a mild foreign body reaction in the region of the former pure PDLLA cube. A part of the mono"le suture which was used to "x the cube in the tissue can be seen in the left upper corner.

Fig. 5. TCP-enriched PDLLA, 72 weeks post-implantation, toluidine blue, 400;. Residues of the cube are detectable in the upper part, surrounded by a thick layer of phagocytizing cells, containing granular material (P).

vacuoles contain granular polymeric material (Fig. 5).

CHP-enriched PDLLA samples show only a few giant cells at the inner side of a "brous capsule (Fig. 6) in the 72nd week post-implantation.

2376

W. Heidemann et al. / Biomaterials 22 (2001) 2371}2381

Fig. 6. CHP-enriched PDLLA, 72 weeks post-implantation, toluidine blue, 400;. The cube has been lost due to the embedding procedure; only a few phagocytizing cells (P) are visible inside of the capsule.

3.4. Inyammations Altogether, two abscesses (12 weeks, 48 weeks postimplantation) and 4 "stulas (20 weeks, 28 weeks, 32 weeks, 40 weeks post-implantation) at PDLLA#TCP implantation site and 2 abscesses (20 weeks, 36 weeks post-implantation) and 3 "stulas (28 weeks, 32 weeks, 40 weeks post-implantation) at PDLLA#CHP implantation site were observed. Only one pure PDLLA implant was found to be lost in the 40th week by a "stula. But in this special case the origin of the in#ammation process could not be localized, as the PDLLA#TCP and the PDLLA#CHP implants of the same animal had also been lost by "stulas.

3.5. Changes of dimensions After implantation, a pronounced length reduction of the rods which was combined with an increase in the

implant's cross section was observed. In pure PDLLA rods, a contraction in length of 25% was measured in the 2nd week post-implantation, the length of PDLLA# TCP rods or PDLLA#CHP rods decreased around 8.3 or 16.6%, respectively. A further reduction in length, of up to a total of 12.5% was observed in PDLLA#TCP rods in the following 2 weeks. The volume of pure PDLLA rods increased from 120 mm before the implantation to 158.2 mm (SD"7.5) in the 16th week, also an increase of weight from 137 mg up to the maximum (149.8 mg; SD"0.5) in the 16th week post-implantation was measured. Thereafter, a rapid decrease of volume and weight was observed.

The PDLLA#TCP rods continuously increased in volume up to the 28th week (175 mm ; SD"8.3) and in weight up to the 32nd week post-implantation (199 mg; SD"2.3); then a reduction of volume and weight took place in PDLLA#TCP rods. The volume of PDLLA#CHP implants increased by the 64th week post-implantation up to 501 mm and decreased to 337 mm in the 72nd week. Maximum weight of PDLLA#CHP was 176.4 mg (SD"4.2) in the 32nd week, initial weight was 144 mg.

No distinct alteration of the length of PDLLA#TCP and PDLLA#CHP implants occurred between the fourth and 32nd week, then a length decrease of PDLLA#TCP rods was observed while PDLLA# CHP rods expanded up to the 64th week.

3.6. Material analysis 3.6.1. Bending strength The bending strength of pure PDLLA rods was 101.6 N/mm (SD"1.3) up to the 4th week post-implantation and was equivalent to the initial values before the implantation (Fig. 7). After 6 weeks 93.1 N/mm (SD"2.7) were measured corresponding to 90% of the

Fig. 7. Decrease of the bending strength during the period of observation (n"3). The PDLLA implants retained at least 60% of their initial value for 12 weeks.

W. Heidemann et al. / Biomaterials 22 (2001) 2371}2381

initial values, after 8 weeks 68 N/mm (SD"10.1) and after 12 weeks 61.2 N/mm (SD"13.6) were measured, corresponding to 68 and 61% of the initial values, respectively. An enhanced decrease of the bending strength of pure PDLLA rods took place in the following weeks; only 9.2 N/mm (SD"3.0) were measured in the 16th week.

In PDLLA#CHP rods 82.7 N/mm (SD"3.6), in PDLLA#TCP 82.8 N/mm (SD"1.5) were measured 4 weeks post-implantation. A further regular decrease of the bending strength was observed in PDLLA#CHP or PDLLA#TCP up to the 28th week post-implantation, where the bending strength was found to be 28.1 N/mm (SD"2.3) in PDLLA#CHP rods and 12.7 N/mm (SD"3.2) in PDLLA#TCP rods. In the following weeks, values of bending strength were too small to be measurable (Fig. 7).

3.7. Modulus of elasticity The modulus of elasticity of pure PDLLA rods was constant for 4 weeks post-implantation; 2250 N/mm were measured before the implantation, 2270 N/mm (SD"304) were measured after 4 weeks. Then, a reduction of the modulus of elasticity to 1643 N/mm (SD"285) took place in the 6th week with constant values up to the 12th week; a further reduction to 217 N/mm (SD"19) occurred up to the 20th week post-implantation.

In PDLLA rods, with addition of calciumphosphates higher initial values were observed which were constant up to the 6th week post-implantation (PDLLA#TCP: 2305 N/mm $20; PDLLA#CHP: 2255 N/mm $249 after 6 weeks). Two weeks later strong reductions of the modulus of elasticity occurred in PDLLA#TCP (1670 N/mm $390) or PDLLA#CHP (1442 N/mm $ 139). Constant values were measured in the 12th week,

2377

then a continuous reduction of the modulus of elasticity took place.

3.8. Molecular weight In pure PDLLA rods, the molecular weight was reduced from 141,000 to 95,000 g/mol within 4 weeks postimplantation, a further slight reduction to 78,000 g/mol took place in the following 8 weeks; 32 weeks postimplantation, 4300 g/mol were measured (Fig. 8).

In PDLLA rods with an admixture of calciumphosphates a slower degradation was observed, 129,000 g/mol were measured in PDLLA#TCP rods and 131,000 g/mol in PDLLA#CHP rods 4 weeks postimplantation. After the 16th week di!erences in the degradation process were detectable between PDLLA#TCP and 56,000 g/mol in PDLLA#CHP; the reduction of the molecular weight was faster in PDLLA#TCP. Thirty-two weeks post-implantation, the molecular weight was 33,000 g/mol in PDLLA#TCP or PDLLA#CHP, respectively; after 56 weeks 1100 g/mol were found in PDLLA#CHP, in PDLLA#TCP 1600 g/mol were measured 72 weeks post-implantation.

Glass transition temperature of pure PDLLA was 54.63C after 4 weeks and was reduced to 41.43C, 20 weeks post-implantation. In PDLLA rods with TCP or CHP no distinct alterations of glass transition temperature were observed up to the 20th week; in PDLLA#TCP rods it measured 54.93C in the 4th week, 53.43C in the 20th week post-implantation. In PDLLA#CHP rods, it was 54.33C after 4 weeks, and measured 53.63C after 20 weeks.

3.9. Morphological characterization by SEM In PDLLA implants with an admixture of calciumphosphates, about 45 pores/100 m were visible on the

Fig. 8. Decrease of the molecular weight between the 4th and 72nd week post-implantation (n"3). Pure PDLLA implants degraded faster as implants containing TCP or CHP.

2378

W. Heidemann et al. / Biomaterials 22 (2001) 2371}2381

Fig. 9. SEM picture of the cross section of a PDLLA#CHP sample 12 weeks post-implantation in survey (1000; magni"cation) and detail (5000; magni"cation). In contrary to pure PDLLA implants, pore formation can be detected in the surface layer.

Fig. 10. SEM picture of the cross section of a PDLLA#TCP sample 12 weeks post-implantation in survey (1000; magni"cation) and detail (5000; magni"cation). Pores are evenly distributed over the cross section.

surface of PDLLA#TCP, 55 pores/100 m were perceptible on the surface of PDLLA#CHP rods 4 weeks post-implantation with continuous increase of the number of pores in the subsequent weeks. In the 6th week post-implantation, pores were detectable for the "rst time in the outer layer of pure PDLLA rods.

In PDLLA#CHP or PDLLA#TCP implants, no remarkable macroscopic alterations of the cross section were found up to the 20th week post-implantation; the SEM picture of the cross section of a PDLLA#CHP rod which was performed after measuring the bending strength in the 12th week post-implantation showed, multiple pores underneath the surface of the implant, homogeneously distributed small particles and the lacking of an undegraded solid outer layer (Fig. 9). Multiple pores were also visible in the electron microscopic picture of the surface and the center of a PDLLA#TCP implant in the 12th week; in addition many homogeneously distributed small TCP particles ((1 m) and also a few larger particles (&7 m) are detectable (Fig. 10).

In general, a homogenous distribution of CHP or TCP particles was found; the average size of CHP particles seemed to be less than 2 m, TCP particles were measured between 0.3 and 10 m. In special cases, agglomerations of CHP and TCP particles were observed; Fig. 11 shows an accumulation of long, squared CHP particles in the center

of a sample 16 weeks post-implantation. In Fig. 12, a conglomerate of rhombic TCP particles is visible.

4. Discussion Light-microscopic examinations showed that the degradation of pure PDLLA from the extracellular space is totally "nished 72 weeks post-implantation, only small intracellular remnants of a foreign body material could be found at higher magni"cation. In TCP-enriched PDLLA, a rapid phase of degradation took place in the 72nd week post-implantation, multiple phagocytizing cells were visible in the environment of the implant.

Because of the strong cellular reaction with in"ltration of cells into the cube, the polymer is not lost by the embedding procedure, in contrast to former degradation stages.

In CHP-enriched PDLLA samples the cellular resorption of the polymer has just started at the end of the observation time; only a few phagocytizing cells were detectable. The alterations in the consistency of PDLLA#TCP and PDLLA#CHP implants point to the fact that the degradation might be completed after two or three years.

While degradation of the PDLLA implants was uncomplicated, a large number of in#ammations was found

W. Heidemann et al. / Biomaterials 22 (2001) 2371}2381

Fig. 11. SEM picture of a PDLLA#CHP sample 16 weeks postimplantation in 6000; magni"cation. Pore formation around an agglomerate of CHP particles probably due to water uptake.

in the tissue around the implants containing TCP (13.4%) or CHP (11%). Only one abscess in the PDLLA site was found in the 40th week at a rat, where a con#uent in#ammation presumably had started at the PDLLA# TCP and PDLLA#CHP sites, leading to "stulas. An importation of contagious germ by the implants can be excluded, because the implants were sterilized by -irradiation (25 kGy). A contamination of the implants during the operation seems to be improbable because the implantation was performed under sterile conditions and comparable in#ammations were not seen around pure PDLLA implants. A factor which could be responsible for the in#ammations are tiny slowly resorbable TCP or CHP particles, which might have separated from the surface of the implants and concentrated in the surrounding tissue and "nally leading to in#ammatory reactions.

The results of the present study are contrary to the results of Lin et al. [36,37] who reported on uncomplicated healing of PDLLA/TCP composite bone screws in rabbits, whereas, they used larger TCP particles (50}100 m) and they did not implant into soft tissues.

No signi"cant e!ect of TCP or CHP on the pH value in the tissue around the implants was found in our study nor could a drop in the pH values around pure PDLLA implants be detected. Obviously, acid ions, produced during the degradation of the implants, could be bu!ered

2379

Fig. 12. SEM picture of a PDLLA#TCP sample 16 weeks postimplantation in 6000; magni"cation.

by the surrounding tissue #uid in vivo and did not reach clinical importance. These in vivo "ndings are in clear contrast to the earlier in vitro results, reported by Ruf"eux [30], who measured a drop in pH from 7.4 to 2.0 in the unbu!ered surrounding media, containing samples of pure PDLLA 32 weeks post-incubation. Samples of PDLLA with an addition of calciumphosphates did not show signs of pH drop in vitro after 32 weeks of incubation, certainly resulting from the retarded degradation of this material.

Taylor et al. [31] reported on a drop of pH in incubation solutions, containing PGA or low molecular weight PLA samples 16 weeks post-incubation. No signi"cant drop in the pH was detected in the incubation solutions containing samples of high molecular weight PLA. Drawback of this study is the short observation period of 16 weeks which might be adequate for the assessment of rapidly degrading PGA or low molecular weight PLA but is not su$cient for the examination of slowly degrading high molecular PLA. In agreement to our results, Vasenius et al. [38] could not detect a signi"cant drop of blood pH in a 60 week follow up after implantation of intramedullary rods made by self-reinforced poly-D,L/L-lactic acid (40/60) or pure poly-L-lactic acid in both femurs of 10 rabbits.

Di!erences of the pH values between the tissue surrounding the implants and the normal muscle tissue were

2380

W. Heidemann et al. / Biomaterials 22 (2001) 2371}2381

maximal 0.2 units in our study. This is in accordance with the results of Martin et al. [34], who measured pH values in bone chambers containing poly(D,L-lactide-co-glycolide) which were implanted in the right tibia of female rabbits. Statistically signi"cant pH di!erences were found between the unloaded and PDLLG bearing bone chambers, but they were also only of the order of 0.2 pH units.

Osteosynthesis plates for use in craniomaxillofacial surgery should have initial bending strength of more than 100 N/mm or about 60 N/mm 4 weeks post-implantation to achieve undisturbed healing of facial fragments [39] * a value, which could be met by all the tested rods easily even up to 8}12 weeks. Measurements of the bending strength during the process of degradation showed constant values up to the 4th week post-implantation in pure PDLLA-rods of about 102 N/mm , 90% of the initial bending strength was measured in the 6th week, 60% of the initial values were obtained up to the 12th week. The admixture of calciumphosphates led to a decrease of the bending strength from the beginning; the higher particle content in PDLLA#TCP rods causes a stronger decrease of the bending strength in comparison to PDLLA#CHP implants.

The modulus of elasticity of pure PDLLA rods was constant for 4 weeks post-implantation, then a reduction took place in the 6th week with constant values up to the 12th week. A higher sti!ness was observed in PDLLA rods with the addition of calciumphosphates which is quite normal for particle "lled polymers. The initial modulus of elasticity was constant up to the 6th week post-implantation.

Molecular weight of pure PDLLA was reduced from 141,000 to 4300 g/mol after 32 weeks. Samples containing TCP or CHP showed an almost linear decrease in M as

compared to pure PDLLA samples which indicates that autocatalysis in the interior is hindered. Molecular weight of samples containing CHP decreased slower than the samples containing TCP, probably due to the higher solubility of CHP and therefore having a better neutralizing e!ect.

No traces of crystallinity were detected in PDLLA implants, with or without addition of calciumphosphates, during the degradation period.

Due to residual tensions inside injection-moulded PDLLA a relaxation occurred after the implantation in the tissue, connected to the increase of temperature to 373C, leading to length reduction in combination with an increase in cross-sectional area; this phenomenon was also described by Tschakalo! et al. [40] after implantation of prebent PDLLA (R 207) plates in rabbits. The extent of length reduction depended on the compound of the implants; the length of pure PDLLA rods decreased around 25%, the length of PDLLA#CHP and PDLLA#TCP rods decreased around 16% within 4 weeks. The higher the content of particles, the more was

the relaxation hindered. Later in vitro tests showed that the length reduction of injection-moulded PDLLA can be prevented by performing a controlled relaxation after injection moulding.

Multiple pores were observed in scanning electron microscopy at the surface of PDLLA implants with the addition of calciumphosphates in the 4th week postimplantation on and also at cross sections of pure PDLLA rods, 6 weeks post-implantation. Number and size of pores are indicators for water uptake of the implants, leading to the observed volume increase. It is assumed that water accumulates around the CaP particles by an increased osmolarity due to a partial dissolution of CaP, thus forming pores. Close underneath the surface, the pores may burst open and expel CaP particles.

The distribution of TCP and CHP particles was homogeneous in general, but in a few cases agglomerations of CHP and TCP particles were observed. During the degradation of the polymeric parts of PDLLA#CaP, TCP and CHP particles might be released from the surface of the implants and deposited in the surrounding tissue.

The results of the study show that an admixture of calciumphosphates to PDLLA for implantation in soft tissues is disadvantageous due to the in#ammations in the surrounding tissue of many implants, the size change and the longer degradation period. In a subsequent study, the biocompatibility of PDLLA#CaP should be examined after implantation in bone.

For the moment, the clinical application of PDLLA with an admixture of calciumphosphates cannot be recommended because of the poor mechanical characteristics and the frequent in#ammative tissue reactions, whereas, pure PDLLA implants seem to have favorable mechanical properties to be used for bone fragment "xation, at least in non-loaded areas of the skull.

Acknowledgements The authors are grateful to Prof. Dr. A. Roessner, Institute of Pathology, Otto-von-Guericke University Magdeburg, Germany, for the preparation of the butyl/methylmetacrylate-embedded slides.

References [1] Diehl K, Mittelmeier H. Biomechanische Untersuchungen zur ErklaK rung der Spongiosierung bei der Plattenosteosynthese.

Z Orthop 1974;112:235.

[2] Uhtho! HK. Aktuelle biomechanische Fragen der Osteosyntheseforschung. Z Orthop 1975;113:760.

[3] Perez Macias JM, Sanchez Palacios A. Artifact lesions and metal sensitivity after insertion of osteosynthesis material. Contact Dermatitis 1986;14(4):260.

W. Heidemann et al. / Biomaterials 22 (2001) 2371}2381 [4] Balato N, Lembo G, Patruno C, Ayala F. Generalized dermatitis due to an osteosynthesis screw. Contact Dermatitis 1991; 24(4):310.

[5] Merle C, Vigan M, Devred D, Girardin P, Adessi B, Laurent R. Generalized eczema from vitallium osteosynthesis material. Contact Dermatitis 1992;27(4):257}8.

[6] HoK nig JF, Merten HA, Luhr HG. Passive and active intracranial translocation of osteosynthesis plates in adolescent minipigs.

J Craniofac Surg 1995;6(4):292}8.

[7] Papay FA, Hardy S, Morales Jr L, Walker M, Enlow D. `Falsea migration of rigid "xation appliances in pediatric craniofacial surgery. J Craniofac Surg 1995;6(4):309}13.

[8] Goldberg DS, Bartlett S, Yu JC, Hunter JV, Whitaker LA. Critical review of micro"xation in pediatric craniofacial surgery.

J Craniofac Surg 1995;6(4):301}7.

[9] Bos RRM, Rozema FR, Boering G, Nijenhuis A, Pennings AJ, Verwey AB. Bio-absorbable plates and screws for internal "xation of mandibular fractures. A study in six dogs. Int J Oral Maxillofac Surg 1989;18:365}9.

[10] Gerlach KL. Treatment of zygomatic fractures with biodegradable poly(L-lactide). In: Heimke G, Soltesz U, Lee AJC, editors.

Clinical implant materials. Amsterdam: Elsevier Science, 1990. p. 573}8.

[11] Rozema FR, Bos RRM, Pennings AJ, Jansen HBW. Poly(Llactide) implants in repair of defects of the orbital #oor: an animal study. J Oral Maxillofac Surg 1990;48:1305}9.

[12] Pistner H, Gutwald R, Ordung R, Reuther J. Poly(L-lactide): a long-term degradation study in vivo. Part I: biological results.

Biomaterials 1993;14(9):671}7.

[13] Pistner H, Stallforth H, Gutwald R, MuK hling J, Reuther J, Michel C. Poly(L-lactide): a long-term degradation study in vivo. Part II: physico-mechanical behaviour of implants. Biomaterials 1994;15(6):439}50.

[14] Pistner H, Bendix DR, MuK hling J, Reuther JF. Poly(L-lactide): a long-term degradation study in vivo. Part III: analytical characterization. Biomaterials 1993;14(4):291}8.

[15] Salyer KE, Bardach J, Squier CA, Kelly KM. A comparative study of the e!ects of biodegradable and titanium plating systems on cranial growth and structure: experimental study in beagles.

Plast Reconstr Surg 1994;93:705}13.

[16] Eppley BL, Sadove AM. A comparison of resorbable and metallic "xation in healing of calvarian bone grafts. Plast Reconstr Surg 1995;96:316}22.

[17] Eppley BL, Sadove MA, Havlik RJ. Resorbable plate "xation in pediatric craniofacial surgery. Plast Reconstr Surg 1997;100(1):1}7.

[18] Suuronen R. Comparison of absorbable self-reinforced poly-Llactide screws and metallic screws in the "xation of mandibular condyle osteotomies: an experimental study in sheep. J Oral Maxillofac Surg 1991;49:989}95.

[19] Suuronen R, Pohjonen T, Tech L, Vasenius J, VainionpaK aK S. Comparison of absorbable self-reinforced multilayer poly-L-lactide and metallic plates for the "xation of mandibular body osteotomies: an experimental study in sheep. J Oral Maxillofac Surg 1992;50:255}62.

[20] Suuronen R, Laine P, Pohjonen T, Lindquist C. Sagittal ramus osteotomies "xed with biodegradable screws: a preliminary report. J Oral Maxillofac Surg 1994;52:715}20.

[21] StaK helin AC, Weiler A, RuK fenacht H, Ho!mann R, Geissmann A, Feinstein R. Clinical degradation and biocompatibility of di!erent bioabsorbable interference screws: a report of six cases. Arthroscopy 1997;13(2):238}44.

[22] Bos RRM, Boering G, Rozema FR, Leenslag JW. Resorbable poly(L-lactide) plates and screws for the "xation of zygomatic fractures. J Oral Maxillofac Surg 1987;45:751}3.

2381

[23] Cordewener FW, Bos RRM, Rozema FR, Houtman WA. Poly(Llactide) implants for repair of human orbital #oor defects: clinical and magnetic resonance imaging evaluation of long-term results.

J Oral Maxillofac Surg 1996;54:9}13.

[24] Bessho K, Iizuka T, Murakami KI. A bioabsorbable poly-L-lactide miniplate and screw system for osteosynthesis in oral and maxillofacial surgery. J Oral Maxillofac Surg 1997; 55:941}5.

[25] Harada K, Enomoto S. Stability after surgical correction of mandibular prognathism using the sagittal split ramus osteotomy and "xation with poly-L-lactid acid (PLLA) screws. J Oral Maxillofac Surg 1997;55:464}8.

[26] BoK stman O, Hirvensalo E, MaK kinen J, Rokkanen P. Foreignbody reactions to fracture "xation implants of biodegradable synthetic polymers. J Bone Jt Surg [Br] 1990; 72-4:592}6.

[27] Bergsma EJ, Rozema FR, Bos RRM, de Bruijn WC. Foreign body reactions to resorbable poly(L-lactide) bone plates and screws used for the "xation of unstable zygomatic fractures. J Oral Maxillofac Surg 1993;51:666.

[28] Bergsma JE, de Bruijn WC, Rozema FR, Bos RRM, Boering G. Late degradation tissue response to poly(L-lactide) bone plates and screws. Biomaterials 1995;16:25}31.

[29] Verheyen CCPM, de Wijn JR, van Blitterswijk CA, Rozing PM, de Groot K. Examination of e!erent lymph nodes after 2 years of transcortical implantation of poly(L-lactide) containing plugs: a case report. J Biomed Mater Res 1993;27(8):1115}8.

[30] Ru$eux K. Degradables Osteosynthesesystem aus Polylactid fuK r die maxillofaciale Chirurgie: ein Beitrag zur Werksto!- und Prozessentwicklung. Diss., ETH Zurich, 1997.

[31] Taylor MS, Daniels AU, Andriano KP, Heller J. Six bioabsorbable polymers: in vitro acute toxicity of accumulated degradation products. J Appl Biomater 1994;5:151}7.

[32] von Recum HA, Cleek RL, Eskin SG, Mikos AG. Degradation of polydispersed poly(L-lactic acid) to modulate lactic acid release.

Biomaterials 1995;16(6):441}7.

[33] Mainil-Varlet P, Curtis R, Gogolewski S. E!ect of in vivo and in vitro degradation on molecular and mechanical properties of various low-molecular-weight polylactides. J Biomed Mater Res 1997;36:360}80.

[34] Martin C, Winet H, Bao JY. Acidity near eroding polylactide}polyglycolide in vitro and in vivo in rabbit tibial bone chambers. Biomaterials 1996;17:2373}80.

[35] Suganuma J, Alexander H. Biological response of intramedullary bone to poly-L-lactic acid. J Appl Biomater 1993;4:13}27.

[36] Lin FH, Chen TM, Lee CJ. The preliminary study and tentative animal study on the sintered PDLLA/TCP composites as bone fracture internal "xation. Biomed Sci Instrum 1997;34:76}81.

[37] Lin FH, Chen TM, Lin CP, Lee CJ. The merit of sintered PDLLA/TCP composites in management of bone fracture internal "xation. Artif Organs 1999;23(2):186}94.

[38] Vasenius J, Majola A, Miettinen EL, ToK rmaK laK P, Rokkanen P. Do intramedullary rods of self-reinforced poly-L-lactide or polyDL/L-lactide cause lactic acid acidosis in rabbits. Clin Mater 1992;10:213}8.

[39] Gerlach KL. In-vivo and clinical evaluations of poly(L-lactide) plates and screws for use in Maxillofacial traumatology. Clin Mater 1993;13:21}8.

[40] Tschakalo! A, Losken HW, von Oepen R, Michaeli W, Moritz O, Mooney MP, Losken A. Degradation kinetics of biodegradable DL-polylactid acid biodegradable implants depending on the site of implantation. Int J Oral Maxillofac Surg 1994;23:443}5.