Biodegradable
Content
Biodegradable polymers for use in s u r g e r y poly(glycolic)/poly(lactic acid) homo and copolymers: 2. In vitro degradation
A. M. Reed* a n d D. K. G i l d i n g *
Bioengineering Department, University of Liverpool, Liverpool UK (Received 25 January 1980; revised 2 July 1980)
The degradation mechanism of a series of polyglycolic/polylactic acid, (PGA/PLA), homo and copolymers synthesized as in Part 11, has been studied in vitro. An in vitro test model similar to that described in a previous study 9, was used. The effects of time, temperature and pH on the rate and mechanism of degradation were elucidated. The degree of degradation was monitored molecularly by gel permeation chromatography (g.p.c.), tensile strength determination and mass loss measurements. The mechanism of degradation is shown to be by hydrolysis. The copolymers of PGA and PLA are shown to have a wide range of degradation rates governed by the hydrophilic/hydrophobic balance and crystallinity of the respective copolymer. The effect of the glass transition temperature (Tg) of PGA on its sensitivity to degradation is also demonstrated.
INTRODUCTION Part I of this series 1 described the synthesis and structural characterization of a range of homo and copolymers of polyglycolic (PGA) and polylactic (PLA) acid. These polymers are known to undergo degradation in the physiological environment 2 - 7 This paper describes a study to elucidate the mechanism and kinetics of degradation of the various members of this family of materials 8. This type of study can best be done in vitro. Although PGA has been commercially available for several years as Dexon R surgical sutures 2 there is little basic information in the literature. The need for the development of meaningful in vitro models for testing candidate biomaterials has been discussed elsewhere 9. The data presented in Part I t (see Figure 1) show that PGA (Dexon R) and P-I-LA (as an annealed film) have crystallinities of 50 and 37~o respectively by X-ray diffraction (XRD) and differential scanning calorimetry (d.s.c.) techniques. Crystallinity decreases with the increase of either comonomer, and copolymers in the range of compositions between 25 and 70~o GA are amorphous. This structural variation would be expected to produce an interesting range of properties, since the two homopolymers represent hydrophilic (PGA) and hydrophobic (PLA) crystalline matrices each containing ester bonds susceptible to hydrolysis. The amorphous range of compositions represents a series of copolymers of varying hydrophilicity. If dl-lactic acid is used instead of I-lactic acid, then the amorphous range is extended from 0-70 m~o GA, since P-dI-LA is amorphous. The in vitro test model employed was similar to that described for the PEO/PET series of copolymers 9. In this case the polymers under investigation were incubated for predetermined periods (0-26 weeks) in buffer solutions of * Present address: MilliporeCorp., Ashby Road, Bedford, MA 01730 0032-3861/81/040494~05502.00 © 1981 IPC Business Press
494 POLYMER, 1981, Vol 22, April
varying pH (pH 5-9), at several temperatures (25°-50°C). On removal from the incubation medium the polymers were taken and the following parameters determined as a function of degradation time and composition: (1) Loss of tensile strength (in the case of PGA, Dexon R sutures). (2) Mass loss. (3) Loss in molecular weight ( M W ) and changes in molecular weight distribution ( M W D ) , Change in mechanical properties of PGA (Dexon R) sutures are the key parameters from a practical point of view. Mass loss is significant from the point of view of absorption and removal of the polymers from the body. However the most sensitive parameters for studying degradation effects are M W and M W D . A significant percentage of polymer chains can be cleaved before there are changes in mechanical properties. Mass loss was expected to be slower than the loss in M W and tensile properties, as it is necessary to reduce both M W and mechanical strength to a considerable degree before any material will fragment and/or dissolve. The effect of pH on the above three parameters was studied over the range pH 5 - 9 , for selected materials. Similarly the effect of temperature between 25 ° 50°C was also investigated. The effect of enzymes on PGA sutures has been studied simultaneously and independently by Williams et al. 1° and these experiments were not repeated in the present study. A series of in vivo experiments were undertaken in rats, the results from which will be compared with the in vitro data in Part 3 lt'8 along with the histological and toxicity findings. EXPERIMENTAL On the basis of data presented in Part 11 three compositions were chosen for thorough study:
5° I o 40 P _~ 3c
•
2c
O
u
//
2O
Biodegradable polymers. A. M. Reed and D. K. Gilding
the initial weight (m0), where: % mass loss - t o o - m e
m 0
x 100
(1)
L JO 60 Z Mot % gtycolid¢ in copotyrner
4
c'
100
Figure 1 % Crystallinities for G A / L A copolymers as a function of composition (determined by X-ray and d.s.c, measurements)
(1) PGA; in the form of Dexon R 0, O'S', 3 0 and 3-0 'S' sutures. (2) P-1-LA; in the form of 0.5 mm thick films cast from CHCI 3 9. (3) 50/50 m% G A / L A copolymer: in the form of 0.5 mm thick film, cast from CHC13 9. In shorter studies, to evaluate specific effects, P-dI-LA and the 70/'30 m% GA/LA compositions were studied.
Polyglycolic acid (PGA) Dexon R suture
Approximately 50 cm lengths of PGA (Dexon R) suture of each suture gauge (0 or 3-0) and each filament diameter (6 and 2 denier. 'S' denotes the finer filament) were used as test pieces. Each test piece was weighed and placed in a McCartney screw capped phial containing 15 ml of buffer. The pH was monitored as a function of time and remained constant indicating that the buffer capacity was not exceeded. Degradation was carried out in a thermostatically controlled water bath (Grant SB10, Cambridge). The initial study was carried out over a period of 26 weeks in 0.2 M citrate-phosphate buffer of pH 7 at 37 _+ I'~C. Citrate-phosphate buffer was used as it allowed the pH to be varied over a wide range (pH 2.6-7.0) without the potential complication of changing to different buffer systems.
The fragments of samples remaining from the above mechanical and mass loss determinations were reused for molecular characterization by g.p.c. Molecular characterization by g.p.c, was carried out using chloroform at 25°C as the eluting solvent for the polymers containing up to 70 m % GA. Above this level of GA, PGA type crystallinity (see Figure 1) is present and hexafluoroisopropanol was required as the g.p.c, solvent 8. Two 1 m columns containing 105 and 103 A Styragel Rwere used for this work with a flow rate of I ml rain- 1. Although comparative M W s and M W D s have been used in this study, methods of calibrating the g.p.c, for use with these combinations of solvents and copolymers are discussed at length elsewhere 12. The effect of pH on the degradation kinetics of some of the materials was determined by incubation in buffers of pH 5, 7 and 9 at 37_+ 1 C as a function of time. Tensile and mass loss profiles were determined as previously described. In the case of PGA (Dexon ~) sutures we examined the dependence of the kinetics of degradation on temperature. The rate of degradation was followed by the deterioration of tensile properties and mass loss. Temperatures above and below the glass transition temperature (Tg), i.e. 3 6 C for Dexon R sutures, (measured by d.s.c.), were used. The experiments were carried out at pH 7 over a 12 week period.
RESULTS A N D D I S C U S S I O N
Polyglycolic acid (PGA) Dexon R suture
The effect of degradation in 0.2 M citrate-phosphate butter o f p H 7 at 3T'C on the change in tensile properties is shown as a function of time in Figure 2. The data show no effect of suture gauge (i.e. 0 vs 3 0) or filament denier in the 2-6 denier range. In each case ~ 80% of the initial tensile strength is maintained for two weeks. From the practical viewpoint this is important as a wound has little strength above that of coagulated fibrin up to l week, but is rapidly reinforced by collagen synthesis between 1 and 3 weeks. Most of the suture's strength is lost over the 2-4 week period. Figure 3 shows the change in stress/strain curve during this time. Elongation decreases simultaneously with tensile strength indicating that the material is becoming more brittle as degradation proceeds. Mass loss, as a function of time, is shown for two of the Dexon R configurations (0 and O'S') in Figure 4. Again there is no effect of filament denier size. Mass loss begins in the fourth week, and is complete by 10 12 weeks providing the test medium is not agitated. Agitation will cause total mass loss within 8 weeks. The observed asyncronization of tensile and mass loss profiles indicates that the degradation process may be complex, and the molecular loss profiles for 'O' Dexon R shown in Figure 5 provide the answer. The M W loss profiles are essentially identical for both suture gauges and both filament sizes studied. The M I47) for 0, 1, 2 and 3 week degradation periods show a steady decrease in the M Wof the peak of the distribution, with the high M Wtail
Glycolic acid/lactic acid, (GA/IA), copolymers and poly-llactic acid, (P-I-LA), polymers
1 cm diameter discs of P-1-LA, 50/50 m°//o G A / L A and 70/30 m% GA/LA copolymers were used for degradation studies. These discs were cut from films cast from CHCI 3 onto glass plates. As above, each test piece was weighed and placed in a McCartney screw capped phial containing 15 ml buffer. The pH was monitored as a function of time and remained constant indicating that the buffer capacity was not exceeded. Degradation was carried out as described for PGA (DexonRl sutures over similar time periods at pH 7 and 37_+ 1C. On removal from the incubation buffer all samples were washed with distilled water to remove residual buffer and dried over P205 under vacuum. In the case of the PGA sutures, tensile measurements were made on the samples. Between 5 and 10 replicates were used for each measurement. The tensile measurements were made using a 2 cm gauge length and a 100% /rain strain rate on a Howden Tensiometer. Mass loss was determined for each sample by comparing the dry weight (me) remaining at a specific time with
POLYMER, 1981, Vol 22, April
495
Biodegradable polymers." A. M. Reed and D. K. Gilding
100~
a
I
Poly-I-lactic acid (P-I-LA)
P-1-LA represents a more complex material than the 50/50 PGA/PLA copolymer as this is a 37% crystalline hydrophobic matrix. The mass and molecular loss profiles are shown in Figure 7. The results show that over a 16 week period, only 10-15% of the mass of the 0.5 mm thick disc is lost, whereas Mw is reduced to ~ 50% of its initial value. Although tensile properties were not measured, gross observations indicated relatively small changes in mechanical properties over a 26 week period. Over this period of time there is no change in crystallinity (as
801 6O
4o !
20
1oo 80 6c 4c
"F_
6
~
20
x:
C
N~oo
©
w
1'5
' ~ S
~ 80
~
6c
4C 20 0 100 80 60 40 2C
o i 2 3 4
xl.O
%
v ' weeks
0"5
s
7
Time (weeks)
Figure 2 In-vitro tensile data, Dexon R sutures, (PGA) (Degraded at pH 7 and 37 o C). (a), '0' Dexon R ;(b), "0' Dexon R %';(e),'30' DexonR; (d), '30' Dexon R %'
'6' weeks
o
16
2b
°/o elongation
ab
Figure 3 Stress/strain curves for degraded Dexon R, (PGA), sutures ('0' Dexon R) (Degraded at 37°C at pH7)
of the distribution being reduced. The low MWtail of the MWD increases up to 3 weeks, and begins to decrease from that point, indicating that mass loss is taking place. The material remains highly crystalline over the initial 6 week period indicating, as expected, that degradation is taking place in the amorphous regions and the remaining mass is contained in small crystallites. At 56 days (Figure 5) the majority of the Dexon R is present as oligomers.
100 -
:
:
:
80 6C
4C i~
o
L.
2oi
50/50 tool % glycolic acid~lactic acid (GA/LA) copolymer
This copolymer is amorphous and more hydrophobic than the PGA. The mass and MWloss profiles are shown in Figure 6 and are found to be virtually superimposable, with the major loss in each occurring over the 2 4 week period. This arises due to the amorphous nature of the matrix, and confirms, as was suggested in Part 1 x, that this composition should be an excellent sustained drug release matrix, as it should disappear from the body within 3 months after the drug has been delivered. As the material is inherently hydrophilic 1, i.e. the equilibrium water level is ~25%, it will probably be useful for the more hydrophobic drugs.
~ lOOI
£ 8O
- _._....
b
~
60'
40
20
O
Figure 4
1
3
4 5 Time (weeks)
6
In vitro mass loss data, Dexon R sutures (Degraded at pH 7 and 37°C). (a) '0' DexonR; (b), '0' Dexon R %'
496
POLYMER, 1981, Vol 22, April
Biodegradable polymers. • A. M. Reed and D. K Gilding
/--\. // 56 days \'\ 14 days 7 days. X
Low MW
High MW
gradations similar to those described previously for PGA at pH 7and 37~C were set up at 2 5 , 3 0 , 3 5 , 4 0 , 4 5 ' and 50C. Three suture sizes, '0', ' 4 0)' and '5 0', 6 denier/filament Dexon R were used. The tensile and mass loss profiles over 12 weeks are shown for '0' Dexon R as a function of temperature in Figures 8a and h respectively. As expected, temperature
Figure 5 In vitro molecular weight distributions for PGA ('0' Dexon R) sutures, (expressed as a function of degradation time.) Degradation at pH 7 and 37°C. Degradation time expressed in days
c~ c
8O
"
~
(]
1001~.~.
E 60 ~
c~ c
40
E
(3
E3
P
20
E
O
2
4
6
8
10
12
14
Time(weeks)
Figure 7 In vitro mass and molecular weight loss for P - 1 - - L A polymer (Degraded at pH 7 and 37°C), ( - - I - ) % mass remaining; (-- - - A - - ) % MW remaining in vitro
0
2
4
6 Ti me (weeks)
8
10
12
loo
_c
c~
a!
Figure 6 In vitro mass and molecular weight loss for 50/50 GA/LA copolymer (Degraded at pH 7and 37°C). ( - Q - ) % mass remaining, (-- --A-- --) % MW remaining in vitro
.~ 80 O
25°C
60
measured by d.s.c.) indicating that the mechanical properties are controlled by the crystalline phase.
U3
40
EJ]'ect qt' pH on degradation kinetics In vitro degradation was carried out in citratephosphate buffers at pH 5 and pH 7 and boric acid borax buffer at pH 9 (ref 8). In each case there was no significant change in degradation kinetics with change in pH 8. This was surprising as polyesters are known to be base sensitive. We considered that the reason accelerated degradation was not observed at pH 9 may have been due to the combination of hydrophobicity and crystallinity. The 70/30 GA/LA copolymer has the most hydrophilic composition of the amorphous structures, and it was expected to show greater sensitivity to pH than any other in the family. A limited study showed the 70/30 GA/LA copolymer beginning to fragment at 12 days, and undergo 1007,,~ mass loss within 4 weeks at pH 9 (ref 8). This confirms that any sensitivity of these copolymers to pH is dominated by the combined effects of crystallinity and hydrophobicity, and is relatively unimportant, which is converse to the P E O / P E T system described previously 9. Effect of temperature on degradation kinetics It was shown by d.s.c, in Part I that the glass transition temperature (7~) for PGA is 36°C. As temperature coefficients of diffusion and hydrolysis rate constants would be expected to show a discontinuity at ~, an experiment was designed to investigate this phenomenon. Six de-
I//\
2 4
\
6 8 Time (weeks) 10
mperatures
0
12
Figure 8a Tensile loss profiles for PGA sutures ('0' Dexon R) as a function of time
100
8O
c
6o
\\\
2 4 6 8 10 Time (weeks) 12
O
Figure 8b Mass loss profiles for PGA sutures ('0" Dexon R) as a function of time
POLYMER, 1981, Vol 22, April
497
Biodegradable polymers. A. M. Reed and D. K. Gilding
10(
12
8c
lO
6c
F-
o
CD
8
E
-6 4 0 .-E
.~
o
E
6
"6
~
2o
-6
tL_
4
40
F--
2
0 25 30 35 40 45 Temperature of incubation (°C) at pH 7 50
20
E 0 25
--
Figure 9
Slopes of the linear portions of the tensile loss profiles plotted as a function of temperature PGA suture: '0' Dexon R
30 35 45 Temperature of incubation (oC)at pH 7
L
I
4'0
I
50
-n j 0
has a dramatic effect on the rate of loss of tensile strength and mass. The slopes of the linear portions of the tensile loss profiles are plotted as a function of temperature in Figure 9. The rates are expressed as ~o initial strength loss per week and show the classical degradation behaviour of a constant temperature coefficient above and below To with a clear discontinuity at TO. The temperature coefficient above Tg is ~ 7 times greater than that below. This indicates that the glassy state protects PGA from hydrolysis as all short term chain motions are frozen. Above TO,water diffusion, and therefore hydrolysis, is more facile. The mass loss profiles in Figure 8b shows that not only is the rate of mass loss dependent on temperature, but also the threshold time before its onset. Figure 10 shows threshold time prior to mass loss, and rate of mass loss, as a function of temperature. Clear discontinuities are seen at 36.5°C for each of the functions indicating the position of TO. The differences in slope represent differences in the diffusion constants of the degradation products in the glassy and rubbery states. CONCLUSIONS This study has shown that the glycolic acid/lactic acid family of homo and copolymers has a wide range of degradation rates that are governed by both hydrophobic/hydrophilic balance and crystallinity. The mechanical properties and property loss profiles during degradation of both homopolymers are governed by the crystalline phase. The degradation mechanism is one of simple hydrolysis, which is insensitive to pH for most compositions.
Figure 10
Threshold time to mass loss ( - o - ) and rate of mass loss (-- --e-- --) plotted as a function of temperature PGA suture:
'0' Dexon R
The effect of the To of PGA on its sensitivity to degradation has been demonstrated. These simple in vitro modelling experiments will be compared with in vivo behaviour in Part 3 (ref I I). ACKNOWLEDGEMENT We are grateful to the Science Research Council and ICI Corporate Laboratory for the CASE award to A.M.R. during the course of this work. REFERENCES
1 2 3 4 5 6 7 8 9 10 11 12 13 Gilding, D. K. and Reed, A. M. Polymer 1979, 20, 1459 Frazza, E. and Schmitt, E. E. J. Biomed. Mater. Res. 1971, 1, 43 Dardik, H., Dardik, I. and Laufman, H. Am. J. Surg. 1971, 121, 656 Miln, D. C., O'Connor, J. and Dalling, R. Scott. Med. J. 1972, 17, 108 Jackanicz, T. M., Nash, H. A., Wise, D. L. and Gregory, J. B. Contraception 1973, 8, 227 Kulkarni, R. K., Pani, K. C., Neuman, C. and Leonard, F. J. Biomed. Mater. Res. 1971, 5, 169 Miller, R. A., Brady, J. M. and Cutright, D. E. J. Biomed. Mater. Res. 1977, 11,711 Reed, A. M. Ph.D. Thesis, University of Liverpool, 1978 Reed, A. M. and Gilding, D. K. Polymer 1981 (to be published) Williams, D. F. and Mort, E. J. Bioeng. 1977, 1,231 Wilson, J. G., Gilding, D. K. and Reed, A. M. 1980 (to be published) Gil.ding, D. K., Reed, A. M. and AskilI, I. N. Polymer 1981 (to be published) Wasserman, D. US Pat 3839297, 1975
498
P O L Y M E R , 1 9 8 1 , Vol 22, A p r i l
A. M. Reed* a n d D. K. G i l d i n g *
Bioengineering Department, University of Liverpool, Liverpool UK (Received 25 January 1980; revised 2 July 1980)
The degradation mechanism of a series of polyglycolic/polylactic acid, (PGA/PLA), homo and copolymers synthesized as in Part 11, has been studied in vitro. An in vitro test model similar to that described in a previous study 9, was used. The effects of time, temperature and pH on the rate and mechanism of degradation were elucidated. The degree of degradation was monitored molecularly by gel permeation chromatography (g.p.c.), tensile strength determination and mass loss measurements. The mechanism of degradation is shown to be by hydrolysis. The copolymers of PGA and PLA are shown to have a wide range of degradation rates governed by the hydrophilic/hydrophobic balance and crystallinity of the respective copolymer. The effect of the glass transition temperature (Tg) of PGA on its sensitivity to degradation is also demonstrated.
INTRODUCTION Part I of this series 1 described the synthesis and structural characterization of a range of homo and copolymers of polyglycolic (PGA) and polylactic (PLA) acid. These polymers are known to undergo degradation in the physiological environment 2 - 7 This paper describes a study to elucidate the mechanism and kinetics of degradation of the various members of this family of materials 8. This type of study can best be done in vitro. Although PGA has been commercially available for several years as Dexon R surgical sutures 2 there is little basic information in the literature. The need for the development of meaningful in vitro models for testing candidate biomaterials has been discussed elsewhere 9. The data presented in Part I t (see Figure 1) show that PGA (Dexon R) and P-I-LA (as an annealed film) have crystallinities of 50 and 37~o respectively by X-ray diffraction (XRD) and differential scanning calorimetry (d.s.c.) techniques. Crystallinity decreases with the increase of either comonomer, and copolymers in the range of compositions between 25 and 70~o GA are amorphous. This structural variation would be expected to produce an interesting range of properties, since the two homopolymers represent hydrophilic (PGA) and hydrophobic (PLA) crystalline matrices each containing ester bonds susceptible to hydrolysis. The amorphous range of compositions represents a series of copolymers of varying hydrophilicity. If dl-lactic acid is used instead of I-lactic acid, then the amorphous range is extended from 0-70 m~o GA, since P-dI-LA is amorphous. The in vitro test model employed was similar to that described for the PEO/PET series of copolymers 9. In this case the polymers under investigation were incubated for predetermined periods (0-26 weeks) in buffer solutions of * Present address: MilliporeCorp., Ashby Road, Bedford, MA 01730 0032-3861/81/040494~05502.00 © 1981 IPC Business Press
494 POLYMER, 1981, Vol 22, April
varying pH (pH 5-9), at several temperatures (25°-50°C). On removal from the incubation medium the polymers were taken and the following parameters determined as a function of degradation time and composition: (1) Loss of tensile strength (in the case of PGA, Dexon R sutures). (2) Mass loss. (3) Loss in molecular weight ( M W ) and changes in molecular weight distribution ( M W D ) , Change in mechanical properties of PGA (Dexon R) sutures are the key parameters from a practical point of view. Mass loss is significant from the point of view of absorption and removal of the polymers from the body. However the most sensitive parameters for studying degradation effects are M W and M W D . A significant percentage of polymer chains can be cleaved before there are changes in mechanical properties. Mass loss was expected to be slower than the loss in M W and tensile properties, as it is necessary to reduce both M W and mechanical strength to a considerable degree before any material will fragment and/or dissolve. The effect of pH on the above three parameters was studied over the range pH 5 - 9 , for selected materials. Similarly the effect of temperature between 25 ° 50°C was also investigated. The effect of enzymes on PGA sutures has been studied simultaneously and independently by Williams et al. 1° and these experiments were not repeated in the present study. A series of in vivo experiments were undertaken in rats, the results from which will be compared with the in vitro data in Part 3 lt'8 along with the histological and toxicity findings. EXPERIMENTAL On the basis of data presented in Part 11 three compositions were chosen for thorough study:
5° I o 40 P _~ 3c
•
2c
O
u
//
2O
Biodegradable polymers. A. M. Reed and D. K. Gilding
the initial weight (m0), where: % mass loss - t o o - m e
m 0
x 100
(1)
L JO 60 Z Mot % gtycolid¢ in copotyrner
4
c'
100
Figure 1 % Crystallinities for G A / L A copolymers as a function of composition (determined by X-ray and d.s.c, measurements)
(1) PGA; in the form of Dexon R 0, O'S', 3 0 and 3-0 'S' sutures. (2) P-1-LA; in the form of 0.5 mm thick films cast from CHCI 3 9. (3) 50/50 m% G A / L A copolymer: in the form of 0.5 mm thick film, cast from CHC13 9. In shorter studies, to evaluate specific effects, P-dI-LA and the 70/'30 m% GA/LA compositions were studied.
Polyglycolic acid (PGA) Dexon R suture
Approximately 50 cm lengths of PGA (Dexon R) suture of each suture gauge (0 or 3-0) and each filament diameter (6 and 2 denier. 'S' denotes the finer filament) were used as test pieces. Each test piece was weighed and placed in a McCartney screw capped phial containing 15 ml of buffer. The pH was monitored as a function of time and remained constant indicating that the buffer capacity was not exceeded. Degradation was carried out in a thermostatically controlled water bath (Grant SB10, Cambridge). The initial study was carried out over a period of 26 weeks in 0.2 M citrate-phosphate buffer of pH 7 at 37 _+ I'~C. Citrate-phosphate buffer was used as it allowed the pH to be varied over a wide range (pH 2.6-7.0) without the potential complication of changing to different buffer systems.
The fragments of samples remaining from the above mechanical and mass loss determinations were reused for molecular characterization by g.p.c. Molecular characterization by g.p.c, was carried out using chloroform at 25°C as the eluting solvent for the polymers containing up to 70 m % GA. Above this level of GA, PGA type crystallinity (see Figure 1) is present and hexafluoroisopropanol was required as the g.p.c, solvent 8. Two 1 m columns containing 105 and 103 A Styragel Rwere used for this work with a flow rate of I ml rain- 1. Although comparative M W s and M W D s have been used in this study, methods of calibrating the g.p.c, for use with these combinations of solvents and copolymers are discussed at length elsewhere 12. The effect of pH on the degradation kinetics of some of the materials was determined by incubation in buffers of pH 5, 7 and 9 at 37_+ 1 C as a function of time. Tensile and mass loss profiles were determined as previously described. In the case of PGA (Dexon ~) sutures we examined the dependence of the kinetics of degradation on temperature. The rate of degradation was followed by the deterioration of tensile properties and mass loss. Temperatures above and below the glass transition temperature (Tg), i.e. 3 6 C for Dexon R sutures, (measured by d.s.c.), were used. The experiments were carried out at pH 7 over a 12 week period.
RESULTS A N D D I S C U S S I O N
Polyglycolic acid (PGA) Dexon R suture
The effect of degradation in 0.2 M citrate-phosphate butter o f p H 7 at 3T'C on the change in tensile properties is shown as a function of time in Figure 2. The data show no effect of suture gauge (i.e. 0 vs 3 0) or filament denier in the 2-6 denier range. In each case ~ 80% of the initial tensile strength is maintained for two weeks. From the practical viewpoint this is important as a wound has little strength above that of coagulated fibrin up to l week, but is rapidly reinforced by collagen synthesis between 1 and 3 weeks. Most of the suture's strength is lost over the 2-4 week period. Figure 3 shows the change in stress/strain curve during this time. Elongation decreases simultaneously with tensile strength indicating that the material is becoming more brittle as degradation proceeds. Mass loss, as a function of time, is shown for two of the Dexon R configurations (0 and O'S') in Figure 4. Again there is no effect of filament denier size. Mass loss begins in the fourth week, and is complete by 10 12 weeks providing the test medium is not agitated. Agitation will cause total mass loss within 8 weeks. The observed asyncronization of tensile and mass loss profiles indicates that the degradation process may be complex, and the molecular loss profiles for 'O' Dexon R shown in Figure 5 provide the answer. The M W loss profiles are essentially identical for both suture gauges and both filament sizes studied. The M I47) for 0, 1, 2 and 3 week degradation periods show a steady decrease in the M Wof the peak of the distribution, with the high M Wtail
Glycolic acid/lactic acid, (GA/IA), copolymers and poly-llactic acid, (P-I-LA), polymers
1 cm diameter discs of P-1-LA, 50/50 m°//o G A / L A and 70/30 m% GA/LA copolymers were used for degradation studies. These discs were cut from films cast from CHCI 3 onto glass plates. As above, each test piece was weighed and placed in a McCartney screw capped phial containing 15 ml buffer. The pH was monitored as a function of time and remained constant indicating that the buffer capacity was not exceeded. Degradation was carried out as described for PGA (DexonRl sutures over similar time periods at pH 7 and 37_+ 1C. On removal from the incubation buffer all samples were washed with distilled water to remove residual buffer and dried over P205 under vacuum. In the case of the PGA sutures, tensile measurements were made on the samples. Between 5 and 10 replicates were used for each measurement. The tensile measurements were made using a 2 cm gauge length and a 100% /rain strain rate on a Howden Tensiometer. Mass loss was determined for each sample by comparing the dry weight (me) remaining at a specific time with
POLYMER, 1981, Vol 22, April
495
Biodegradable polymers." A. M. Reed and D. K. Gilding
100~
a
I
Poly-I-lactic acid (P-I-LA)
P-1-LA represents a more complex material than the 50/50 PGA/PLA copolymer as this is a 37% crystalline hydrophobic matrix. The mass and molecular loss profiles are shown in Figure 7. The results show that over a 16 week period, only 10-15% of the mass of the 0.5 mm thick disc is lost, whereas Mw is reduced to ~ 50% of its initial value. Although tensile properties were not measured, gross observations indicated relatively small changes in mechanical properties over a 26 week period. Over this period of time there is no change in crystallinity (as
801 6O
4o !
20
1oo 80 6c 4c
"F_
6
~
20
x:
C
N~oo
©
w
1'5
' ~ S
~ 80
~
6c
4C 20 0 100 80 60 40 2C
o i 2 3 4
xl.O
%
v ' weeks
0"5
s
7
Time (weeks)
Figure 2 In-vitro tensile data, Dexon R sutures, (PGA) (Degraded at pH 7 and 37 o C). (a), '0' Dexon R ;(b), "0' Dexon R %';(e),'30' DexonR; (d), '30' Dexon R %'
'6' weeks
o
16
2b
°/o elongation
ab
Figure 3 Stress/strain curves for degraded Dexon R, (PGA), sutures ('0' Dexon R) (Degraded at 37°C at pH7)
of the distribution being reduced. The low MWtail of the MWD increases up to 3 weeks, and begins to decrease from that point, indicating that mass loss is taking place. The material remains highly crystalline over the initial 6 week period indicating, as expected, that degradation is taking place in the amorphous regions and the remaining mass is contained in small crystallites. At 56 days (Figure 5) the majority of the Dexon R is present as oligomers.
100 -
:
:
:
80 6C
4C i~
o
L.
2oi
50/50 tool % glycolic acid~lactic acid (GA/LA) copolymer
This copolymer is amorphous and more hydrophobic than the PGA. The mass and MWloss profiles are shown in Figure 6 and are found to be virtually superimposable, with the major loss in each occurring over the 2 4 week period. This arises due to the amorphous nature of the matrix, and confirms, as was suggested in Part 1 x, that this composition should be an excellent sustained drug release matrix, as it should disappear from the body within 3 months after the drug has been delivered. As the material is inherently hydrophilic 1, i.e. the equilibrium water level is ~25%, it will probably be useful for the more hydrophobic drugs.
~ lOOI
£ 8O
- _._....
b
~
60'
40
20
O
Figure 4
1
3
4 5 Time (weeks)
6
In vitro mass loss data, Dexon R sutures (Degraded at pH 7 and 37°C). (a) '0' DexonR; (b), '0' Dexon R %'
496
POLYMER, 1981, Vol 22, April
Biodegradable polymers. • A. M. Reed and D. K Gilding
/--\. // 56 days \'\ 14 days 7 days. X
Low MW
High MW
gradations similar to those described previously for PGA at pH 7and 37~C were set up at 2 5 , 3 0 , 3 5 , 4 0 , 4 5 ' and 50C. Three suture sizes, '0', ' 4 0)' and '5 0', 6 denier/filament Dexon R were used. The tensile and mass loss profiles over 12 weeks are shown for '0' Dexon R as a function of temperature in Figures 8a and h respectively. As expected, temperature
Figure 5 In vitro molecular weight distributions for PGA ('0' Dexon R) sutures, (expressed as a function of degradation time.) Degradation at pH 7 and 37°C. Degradation time expressed in days
c~ c
8O
"
~
(]
1001~.~.
E 60 ~
c~ c
40
E
(3
E3
P
20
E
O
2
4
6
8
10
12
14
Time(weeks)
Figure 7 In vitro mass and molecular weight loss for P - 1 - - L A polymer (Degraded at pH 7 and 37°C), ( - - I - ) % mass remaining; (-- - - A - - ) % MW remaining in vitro
0
2
4
6 Ti me (weeks)
8
10
12
loo
_c
c~
a!
Figure 6 In vitro mass and molecular weight loss for 50/50 GA/LA copolymer (Degraded at pH 7and 37°C). ( - Q - ) % mass remaining, (-- --A-- --) % MW remaining in vitro
.~ 80 O
25°C
60
measured by d.s.c.) indicating that the mechanical properties are controlled by the crystalline phase.
U3
40
EJ]'ect qt' pH on degradation kinetics In vitro degradation was carried out in citratephosphate buffers at pH 5 and pH 7 and boric acid borax buffer at pH 9 (ref 8). In each case there was no significant change in degradation kinetics with change in pH 8. This was surprising as polyesters are known to be base sensitive. We considered that the reason accelerated degradation was not observed at pH 9 may have been due to the combination of hydrophobicity and crystallinity. The 70/30 GA/LA copolymer has the most hydrophilic composition of the amorphous structures, and it was expected to show greater sensitivity to pH than any other in the family. A limited study showed the 70/30 GA/LA copolymer beginning to fragment at 12 days, and undergo 1007,,~ mass loss within 4 weeks at pH 9 (ref 8). This confirms that any sensitivity of these copolymers to pH is dominated by the combined effects of crystallinity and hydrophobicity, and is relatively unimportant, which is converse to the P E O / P E T system described previously 9. Effect of temperature on degradation kinetics It was shown by d.s.c, in Part I that the glass transition temperature (7~) for PGA is 36°C. As temperature coefficients of diffusion and hydrolysis rate constants would be expected to show a discontinuity at ~, an experiment was designed to investigate this phenomenon. Six de-
I//\
2 4
\
6 8 Time (weeks) 10
mperatures
0
12
Figure 8a Tensile loss profiles for PGA sutures ('0' Dexon R) as a function of time
100
8O
c
6o
\\\
2 4 6 8 10 Time (weeks) 12
O
Figure 8b Mass loss profiles for PGA sutures ('0" Dexon R) as a function of time
POLYMER, 1981, Vol 22, April
497
Biodegradable polymers. A. M. Reed and D. K. Gilding
10(
12
8c
lO
6c
F-
o
CD
8
E
-6 4 0 .-E
.~
o
E
6
"6
~
2o
-6
tL_
4
40
F--
2
0 25 30 35 40 45 Temperature of incubation (°C) at pH 7 50
20
E 0 25
--
Figure 9
Slopes of the linear portions of the tensile loss profiles plotted as a function of temperature PGA suture: '0' Dexon R
30 35 45 Temperature of incubation (oC)at pH 7
L
I
4'0
I
50
-n j 0
has a dramatic effect on the rate of loss of tensile strength and mass. The slopes of the linear portions of the tensile loss profiles are plotted as a function of temperature in Figure 9. The rates are expressed as ~o initial strength loss per week and show the classical degradation behaviour of a constant temperature coefficient above and below To with a clear discontinuity at TO. The temperature coefficient above Tg is ~ 7 times greater than that below. This indicates that the glassy state protects PGA from hydrolysis as all short term chain motions are frozen. Above TO,water diffusion, and therefore hydrolysis, is more facile. The mass loss profiles in Figure 8b shows that not only is the rate of mass loss dependent on temperature, but also the threshold time before its onset. Figure 10 shows threshold time prior to mass loss, and rate of mass loss, as a function of temperature. Clear discontinuities are seen at 36.5°C for each of the functions indicating the position of TO. The differences in slope represent differences in the diffusion constants of the degradation products in the glassy and rubbery states. CONCLUSIONS This study has shown that the glycolic acid/lactic acid family of homo and copolymers has a wide range of degradation rates that are governed by both hydrophobic/hydrophilic balance and crystallinity. The mechanical properties and property loss profiles during degradation of both homopolymers are governed by the crystalline phase. The degradation mechanism is one of simple hydrolysis, which is insensitive to pH for most compositions.
Figure 10
Threshold time to mass loss ( - o - ) and rate of mass loss (-- --e-- --) plotted as a function of temperature PGA suture:
'0' Dexon R
The effect of the To of PGA on its sensitivity to degradation has been demonstrated. These simple in vitro modelling experiments will be compared with in vivo behaviour in Part 3 (ref I I). ACKNOWLEDGEMENT We are grateful to the Science Research Council and ICI Corporate Laboratory for the CASE award to A.M.R. during the course of this work. REFERENCES
1 2 3 4 5 6 7 8 9 10 11 12 13 Gilding, D. K. and Reed, A. M. Polymer 1979, 20, 1459 Frazza, E. and Schmitt, E. E. J. Biomed. Mater. Res. 1971, 1, 43 Dardik, H., Dardik, I. and Laufman, H. Am. J. Surg. 1971, 121, 656 Miln, D. C., O'Connor, J. and Dalling, R. Scott. Med. J. 1972, 17, 108 Jackanicz, T. M., Nash, H. A., Wise, D. L. and Gregory, J. B. Contraception 1973, 8, 227 Kulkarni, R. K., Pani, K. C., Neuman, C. and Leonard, F. J. Biomed. Mater. Res. 1971, 5, 169 Miller, R. A., Brady, J. M. and Cutright, D. E. J. Biomed. Mater. Res. 1977, 11,711 Reed, A. M. Ph.D. Thesis, University of Liverpool, 1978 Reed, A. M. and Gilding, D. K. Polymer 1981 (to be published) Williams, D. F. and Mort, E. J. Bioeng. 1977, 1,231 Wilson, J. G., Gilding, D. K. and Reed, A. M. 1980 (to be published) Gil.ding, D. K., Reed, A. M. and AskilI, I. N. Polymer 1981 (to be published) Wasserman, D. US Pat 3839297, 1975
498
P O L Y M E R , 1 9 8 1 , Vol 22, A p r i l
