VOL. 10, PP. 471-481 (1976)
J. BIOMED. MATER. RES.
In vivo Degradation of Silicone Rubber Poppets in Prosthetic Heart Valves EDWARD F. CUDDIHY, JOVAN MOACANIN, and E. JOHN ROSCHKE, Propulsion and Materials Research Section, Jet Propulsion Laboratory, Pasadena, California 91103, and EARL C. HARRISON, Department of Medicine, Cardiology Section, University of Southern California Medical Center, Los Angeles, California 90035
Summary Dynamic shear modulus G' was measured throughout the volume of three nonvariant silicone rubber poppets which were recovered from aortic prosthetic heart valves that had been implanted for 4 days, 52 days, and 8 years. Similar measurements were obtained for two unused silicone rubber poppets. Although the recovered poppets exhibited no obvious physical evidence of damage, the silicone rubber had undergone in vivo degradation throughout the poppet volume as indicated by decreases in modulus. The measurements also indicate that the poppet surface degrades at a rate faster than the core. Further, comparison with data reported in the literature suggests that the surfaces of variant poppets degrade a t a rate faster than the surfaces of nonvariant poppets.
INTRODUCTION Gross physical changes such as swelling, discoloration, and surface and internal cracking have been observed in silicone rubber poppets recovered from patients with early-model caged-ball heart valve prostheses.' When these physical and chemical changes are sufficient to cause valvular dysfunction, the poppet is said to be variant.2 Although many authors classify any observable poppet change as variant, Hylen et a1.2 classify changes not leading to valvular dysfunction as incipient ball variance. The distinction is important because the percentage of cases leading t o mechanical valvular dysfunction is quite small. The overall incidence of variance in implanted poppets was reported as approximately 0.04%, but was as 47 1 @ 1976 by John Wiley & Sons, Inc.
472
CUDDIHY ET AL.
high as 291, prior to 1966.3 Most manufacturers of silicone rubber poppets changed the curing cycle around the years 1965-1966. Fatalities due to poppet degeneration have occurred from excessive swelling of poppets, thus restricting or preventing their freedom of m ~ t i o n ,or ~ ?from ~ poppet migration from the cage due to fragmentation or shrinkage caused by a b r a ~ i o n . ~The . ~ cause of variance is believed to be related to infiltration and absorption by the poppet of body fluids, particularly lipid^.^^^^^ The incidence of variance is higher for aortic valve prostheses as compared to mitral valve prostheses, a result sometimes ascribed to differences in valve hemodynamics.IO It is noteworthy that Carmen and Mutha" reported that the presence of high levels of oxygen and oxidation products in the blood markedly enhance the degradation of silicone rubber poppets. Recently, the chemical characteristics of lipids required for interaction with, and absorption by, silicone rubber was recognized.a Raible et a1.I2 measured mechanical properties at the surfaces of variant poppets, and found reductions in tensile strength as high as 47y0, resulting from a combination of a permanent reduction of 23y0,and an additional reduction from the softening action on silicone rubber from absorbed materials. The permanent reduction resulted from in vivo degradation of the silicone rubber. Because lipids, and perhaps other body chemicals, have the chemical characteristics which can interact with, and become absorbed by, the poppets, the occurrence of in vivo degradation may be general to all implanted poppets as suggested by Carmen and Mutha.'l Further transformation into variance may be related to certain manufacturing batches of poppets, or to some adverse alterations in blood chemistry or hemodynamics in particular patients, but the exact causes of variance are still unknown. A study was carried out to determine if degradation in recovered, nonvariant silicone rubber poppets could be detected and measured. Dynamic shear modulus G' was measured throughout the volume of three nonvariant aortic poppets which had been implanted 4 days, 52 days, and 8 years, and for two unused poppets serving as controls. The purpose of this study was to measure the level of degradation that had occurred in vivo, the depth of penetration into the poppet, and, from the time-dependence of the modulus changes found for the three recovered poppets, to provide clues on the rates and extent of
I N VIVO DEGRADATION OF SILICONE RUBBER POPPETS 473
infiltration into the poppets of the body fluids involved in the degradation reactions.
EXPERIMENTAL Mea.surements were obtained by using a Rheovibron DDV-II-B viscoelastometer test machine (Toyo Instruments, Inc., Japan), a device that can measure the dynamic shear modulus G’ of specimens as small as 3.2 mm square by 1.5 mm thick. All five poppets .had a visible mold line on the surface; this mold line was defined as the equator of the spherical poppets. A slab 1.5 mm thick was first sliced out of each of the poppets at the equatorial plane. Because the poppets varied in diameter between 1.5-1.9 cm, it was possible to cut 3.2 mm square specimens from the slabs to allow modulus measurements at three locations in the equatorial plane : core, intermediate, and surface (Fig. 1). In addition, specimens of the same dimensions were cut out of each of the poppets at the polar surfaces. A special cutting device was developed so that the small specimens could be cut accurately and uniformly.
RESULTS AND DISCUSSION Physical Appearance The results of this investigation are summarized in Table I. Four of the poppets were manufactured by Cutter Laboratories (Smeloff Cutter valves), whereas the poppet implanted for 8 years was manufactured by Edwards Laboratories (Starr-Edwards valves). The recovered poppets were spherical, and there was no indication of swelling. Their equatorial and polar diameters were within the tolerance measured for the unused poppets. The surfaces of all three recovered poppets appeared visually as smooth as the unused poppets. The poppets implanted for both 4 days and 52 days retained the same color and translucent quality of the unused poppets, but the poppet implanted for 8 years was noticeably opaque and creamy in color. No internal cracks or fissures were observed in the recovered poppets when they were sliced during sample preparation. In general, these recovered poppets appeared nonvariant, not having any of the usual physical characteristics of variant poppets. Koorajian3 reported that the incidence of va;riant poppets in implanted
Supplier
Smeloff-Cutter Smeloff-Cutter Smeloff-Cutter Smeloff-Cutter Starr-Edwards
Poppet implant history
Unused Unused Implanted for 4 days Implanted for 52 days Implanted for 8 years
TABLE I
1.907 1.621 1.720 1.519 1.532
Polar 1.902 1.618 1.725 1.519 1.526
Equator
Diameter, em
1.905 1.620 1.723 1.519 1.529
Average
1.50 1.35 1.39 1.24 1.47
Core
1.56 1.39 1.45 1.19 1.34
1.59 1.43 1.43 1.22 1.22
1.060 1.059 1.029 0.984 0.830
Surface/core W/C) Intermediate Surface modulus ratio
Dynamic shear modulus G' x lo-' dyne/cm2
Properties of the Silicone Rubber Poppets
I N VIVO DEGRADATION OF SILICONE RUBBER POPPETS 475
. ,CENTER SLICE
SPECIMEN
LOWER SPHERICAL CAP
Fig. 1. Pattern for cutting specimens from spherical silicone rubber poppets.
prosthetic heart valves was slightly less than 2% prior to 1966, and that following a valve redesign, the incidence of variant poppets to date has been reduced to 0.04%.
Equatorial Shear Modulus For the unused poppets, the magnitude of the modulus was found to decrease with distance from the surface; in both cases, the modulus
476
CUDDIHY ET AL.
at the surface was about 6% higher than the value a t the core. This trend presumably results from the thermal history during cure because the surface achieves cure temperature faster than does the core. This skin effect is a common occurrence in the curing of elastomers. For the recovered poppets, there was a noticeable and systematic time-dependent trend in the reduction of the surface modulus relative to the core. The modulus a t the surface of the poppet implanted for 4 days was about 3% above the value a t the core. I n contrast, the modulus a t the surface of the poppet implanted for 52 days was 2% below the value at the core. For the poppet implanted for 8 years, the modulus a t the surface was 17y0below the value a t the core, a complete reversal in trend compared with the results obtained for unused poppets. These data indicate the occurrence and trend of in vivo degradation of the silicone rubber poppets. It should be pointed out that phonocardiographic techniques employed t o monitor prosthetic heart valve sounds in patients could be influenced by these modulus reductions and reversals, even though the poppets are nonvariant and still functional. The modulus data for the four Smeloff-Cutter poppets yielded a n unexpected and surprising relationship when plotted in an appropriate manner. Figure 2 is a plot of modulus a t the core, intermediate, and surface locations versus poppet diameter. The modulus dependency for each location yields a straight-line relationship with poppet diameter, and the three lines associated with each location arc essentially parallel. Note that the modulus values for the intermediate and core location of the poppet implanted for 4 days fall almost perfectly on the tie lines generated by the two unused poppets, but its surface modulus is below the surface tie line. The modulus values a t all locations of the Smeloff-Cutter poppet implanted for 52 days lay below the linear extrapolation of the tie lines (3 data point, a t the lower left-hand corner). These results, plus those for the poppet implanted for 8 years, indicate clearly that the modulus of silicone poppets decreases during human implantation, and that in time, the initial modulus distribution which increased from core t o surface is reversed in trend, as is noted for the modulus distribution of the Starr-Edwards poppet. The poppet implanted for 52 days would appear to be in an intermediate stage of modulus reversal.
I N VZVO DEGRADATION OF SILICONE RUBBER POPPETS 477
0 SURFACE
A INTERMEDIATE
0
CORE
r!l
6 h
1.1
t
-
UNFLAGGED SYMBOLS FLAGGED SYMBOLS
UNUSED POPPETS IMPLANTED POPPETS
i) FLAG DOWN
- 4dayr HUMAN IMPLANTED
i i ) FLAG UP
- 52 days HUMAN IMPLANTED
I
I
I
1S O
I
1.60
1.70
1.80
POPPET DIAMETER,
I
1.90
2.
cm
Fig. 2 . Dynamic shear modulus G’ vs. diameter for the four Smeioff-Cutter poppets.
The data in Figure 2 reveal that the surface of silicone rubber poppets undergo modulus reductions within 4 days of implantation, and that modulus reductions propagate throughout the entire volume of the poppet within 52 days. This implies that the infiltration rate of the body fluids must exceed the kinetic rate of the chemical reaction in order to reach the core without being completely consumed along the diffusion path from surface to core.
Time Dependence The initial surface modulus values of the 4 and 52 day implanted poppets may be estimated by utilizing the data in Figure 2. Thus,
CUDDIHY ET AL.
478 30
I
I
m
3 3
i
I
/
P
/
20
/ / / 1
3 2
/ / / 0
I
0 0
z 0 -
P’ 0-0 THIS STUDY
d!
M REFERENCE 13
---A
0
1
I
I
10
100
ESTIMATE FOR 8 YEARS
1
1
.ooo
lo$
IMPLANT TIME, D A Y S
Fig. 3. I n vivo reduction of the surface modulus of silicone poppets.
it was calculated that the surface modulus of the 4 day implanted poppet decreased by about 3.7% and the 52 day implanted poppet decreased by about 11%. These values are plotted in Figure 3 as open circles. Raible et a1.I2 reported a permanent 23% reduction in surface properties of variant poppets, but did not state the specific implant time. They reported only that the implant time for the various poppets studied ranged from 3.5-38 months. This data point of 23y0 is also plotted in Figure 3 for that reported time span. An examination of Figure 3 suggests the possibility that in vivo degradation may proceed faster for variant poppets than for nonvariant poppets. That is, poppets that will exhibit variance become degraded in vivo at a faster rate. Whether this result is unique for a particular batch of poppets or instead is related to differences in the patients is a matter of conjecture. The conclusion is tentative because of the paucity of data. By extrapolating the two data points plotted in Figure 3, it was estimated that the 8 year implanted poppet underwent about a 22% reduction in surface modulus (Fig. 3). Alternatively, another
I N VZVO DEGRADATION OF SILICONE RUBBER POPPETS 479
time-dependent relationship can be generated from the modulus data for the three recovered poppets. Assuming that the initial ratio of surface to core modulus (X/C)o for the recovered poppets was 1.06, as was observed for the unused poppets, then a linear log-log plot of { l - (S/C)/(X/C),} versus time may be constructed as shown in Figure 4, where the (S/C) values are the modulus ratios given in Table I for each of the recovered poppets. This treatment of data facilitates comparison between silicone rubber samples that may have been slightly different in their original state due to varying processing procedures during manufacture. A satisfactory explanation for the linear log-log relation in Figure 4 has not been determined. The plot, however, is useful for estimates of the degradation magnitude of nonvariant poppets at intermediate implantation times.
Polar Modulus After the 1.5 mm thick equatorial slabs were sliced from each of the poppets, a n effort was made to slice test specimens from the polar locations of the hemispheres remaining in each case (Fig. 1). This proved difficult t o accomplish and, as a consequence, the specimens that were prepared had noticeable irregularities. The quality of data produced by the Rheovibron is sensitive t o geometrical irregularities in the case of very small specimens.
0.011 1
I
10
I
100
I 1000
10,oc
TIME, days
Fig. 4. Plot of the function [I - (S/C)/S/C),] vs. time for the three nonvariant implanted poppets.
480
CUDDIHY ET AL.
Despite this uncertainty, there was a general trend in the data indicating that the modulus at the polar surfaces was consistently less than the modulus at the equatorial surfaces. Thus, the silicone rubber appears to be mechanically anisotropic along the poppet surface. The cause of the anisotropy is not known. However, it is possible that surface anisotropy might contribute to nonuniform swelling of the poppet from absorption of body fluids such as lipids. Nonuniform swelling has been reported as a characteristic of variant poppets.
CONCLUSIONS Silicone rubber poppets undergo in vivo degradation throughout their entire volume. In vivo degradation occurs in both variant and nonvariant poppets, but there is an indication that degradation proceeds faster for variant poppets. Mechanical properties a t the surfaces of silicone poppets degrade faster than at the poppet cores. The infiltration of body fluids into the silicone rubber poppets proceeds at a rate faster than the rate of in vivo degradation. This paper represents one phase of research performed by the Jet Propulsion Laboratory, California Institute of Technology sponsored by the National Aeronautics and Space Administration, Contract NAS7-100.
References 1. W. C. Roberts and A. G. Morrow, Amer. J . Cardiol., 22, 614 (1968). 2. J. C. Hylen, J. C. Kloster, A. Starr, and H. E. Griswold, Ann. Znt. Med., 72, 1 (1970). 3. S. Koorajian, in New Industries and Applications for Advanced Materials Technology, SAMPE Symposium Publication, 19, 566 (1974). 4. E. G. Laforet, New Engl. J . Med., 276, 1025 (1967). 5. T . J. Donovan, H. B. C. Low, C . A. Bucknam, and P. V. Stroughton, Vasc. Surg., 5, 48 (1971). 6. A. Krosnick, J. Amer. Med. Assoc., 191, 105 (1965). 7. K. Hameed, S. Ashfaq, and D. 0. W. Waugh, Arch. Path., 86, 521 (1968). 8. J. Moacanin, D. D. Lawson, H. P. Chin, E. C. Harrison, and D. H. Blankenhorn, Biomat. Med. Dev. Artif. Organs, 1, 183 (1973). 9. H. P. Chin, E. C. Harrison, D. H. Blankenhorn, and J. Moacanin, Circulation, Suppl. I to 43 and 44, 1-51 (1971). 10. W. C. Roberts, G. E. Levinson, and A. G. Morrow, Arch. Znt. Med., 126, 517 (1970).
I N VIVO DEGRADATION OF SILICONE RUBBER POPPETS 481 11. R. Carmen and S. C. Mutha, J . Biomed. Mater. Res., 6,327 (1972). 12. D. A. Raible, D. P. Keller, W. R. Pierie, S. Koorajian, and E. G. Partridge, Rubber Chem. Technol., 39, 1276 (1966).
Received August 24, 1975 Revised November 10, 1975
J. BIOMED. MATER. RES.
In vivo Degradation of Silicone Rubber Poppets in Prosthetic Heart Valves EDWARD F. CUDDIHY, JOVAN MOACANIN, and E. JOHN ROSCHKE, Propulsion and Materials Research Section, Jet Propulsion Laboratory, Pasadena, California 91103, and EARL C. HARRISON, Department of Medicine, Cardiology Section, University of Southern California Medical Center, Los Angeles, California 90035
Summary Dynamic shear modulus G' was measured throughout the volume of three nonvariant silicone rubber poppets which were recovered from aortic prosthetic heart valves that had been implanted for 4 days, 52 days, and 8 years. Similar measurements were obtained for two unused silicone rubber poppets. Although the recovered poppets exhibited no obvious physical evidence of damage, the silicone rubber had undergone in vivo degradation throughout the poppet volume as indicated by decreases in modulus. The measurements also indicate that the poppet surface degrades at a rate faster than the core. Further, comparison with data reported in the literature suggests that the surfaces of variant poppets degrade a t a rate faster than the surfaces of nonvariant poppets.
INTRODUCTION Gross physical changes such as swelling, discoloration, and surface and internal cracking have been observed in silicone rubber poppets recovered from patients with early-model caged-ball heart valve prostheses.' When these physical and chemical changes are sufficient to cause valvular dysfunction, the poppet is said to be variant.2 Although many authors classify any observable poppet change as variant, Hylen et a1.2 classify changes not leading to valvular dysfunction as incipient ball variance. The distinction is important because the percentage of cases leading t o mechanical valvular dysfunction is quite small. The overall incidence of variance in implanted poppets was reported as approximately 0.04%, but was as 47 1 @ 1976 by John Wiley & Sons, Inc.
472
CUDDIHY ET AL.
high as 291, prior to 1966.3 Most manufacturers of silicone rubber poppets changed the curing cycle around the years 1965-1966. Fatalities due to poppet degeneration have occurred from excessive swelling of poppets, thus restricting or preventing their freedom of m ~ t i o n ,or ~ ?from ~ poppet migration from the cage due to fragmentation or shrinkage caused by a b r a ~ i o n . ~The . ~ cause of variance is believed to be related to infiltration and absorption by the poppet of body fluids, particularly lipid^.^^^^^ The incidence of variance is higher for aortic valve prostheses as compared to mitral valve prostheses, a result sometimes ascribed to differences in valve hemodynamics.IO It is noteworthy that Carmen and Mutha" reported that the presence of high levels of oxygen and oxidation products in the blood markedly enhance the degradation of silicone rubber poppets. Recently, the chemical characteristics of lipids required for interaction with, and absorption by, silicone rubber was recognized.a Raible et a1.I2 measured mechanical properties at the surfaces of variant poppets, and found reductions in tensile strength as high as 47y0, resulting from a combination of a permanent reduction of 23y0,and an additional reduction from the softening action on silicone rubber from absorbed materials. The permanent reduction resulted from in vivo degradation of the silicone rubber. Because lipids, and perhaps other body chemicals, have the chemical characteristics which can interact with, and become absorbed by, the poppets, the occurrence of in vivo degradation may be general to all implanted poppets as suggested by Carmen and Mutha.'l Further transformation into variance may be related to certain manufacturing batches of poppets, or to some adverse alterations in blood chemistry or hemodynamics in particular patients, but the exact causes of variance are still unknown. A study was carried out to determine if degradation in recovered, nonvariant silicone rubber poppets could be detected and measured. Dynamic shear modulus G' was measured throughout the volume of three nonvariant aortic poppets which had been implanted 4 days, 52 days, and 8 years, and for two unused poppets serving as controls. The purpose of this study was to measure the level of degradation that had occurred in vivo, the depth of penetration into the poppet, and, from the time-dependence of the modulus changes found for the three recovered poppets, to provide clues on the rates and extent of
I N VIVO DEGRADATION OF SILICONE RUBBER POPPETS 473
infiltration into the poppets of the body fluids involved in the degradation reactions.
EXPERIMENTAL Mea.surements were obtained by using a Rheovibron DDV-II-B viscoelastometer test machine (Toyo Instruments, Inc., Japan), a device that can measure the dynamic shear modulus G’ of specimens as small as 3.2 mm square by 1.5 mm thick. All five poppets .had a visible mold line on the surface; this mold line was defined as the equator of the spherical poppets. A slab 1.5 mm thick was first sliced out of each of the poppets at the equatorial plane. Because the poppets varied in diameter between 1.5-1.9 cm, it was possible to cut 3.2 mm square specimens from the slabs to allow modulus measurements at three locations in the equatorial plane : core, intermediate, and surface (Fig. 1). In addition, specimens of the same dimensions were cut out of each of the poppets at the polar surfaces. A special cutting device was developed so that the small specimens could be cut accurately and uniformly.
RESULTS AND DISCUSSION Physical Appearance The results of this investigation are summarized in Table I. Four of the poppets were manufactured by Cutter Laboratories (Smeloff Cutter valves), whereas the poppet implanted for 8 years was manufactured by Edwards Laboratories (Starr-Edwards valves). The recovered poppets were spherical, and there was no indication of swelling. Their equatorial and polar diameters were within the tolerance measured for the unused poppets. The surfaces of all three recovered poppets appeared visually as smooth as the unused poppets. The poppets implanted for both 4 days and 52 days retained the same color and translucent quality of the unused poppets, but the poppet implanted for 8 years was noticeably opaque and creamy in color. No internal cracks or fissures were observed in the recovered poppets when they were sliced during sample preparation. In general, these recovered poppets appeared nonvariant, not having any of the usual physical characteristics of variant poppets. Koorajian3 reported that the incidence of va;riant poppets in implanted
Supplier
Smeloff-Cutter Smeloff-Cutter Smeloff-Cutter Smeloff-Cutter Starr-Edwards
Poppet implant history
Unused Unused Implanted for 4 days Implanted for 52 days Implanted for 8 years
TABLE I
1.907 1.621 1.720 1.519 1.532
Polar 1.902 1.618 1.725 1.519 1.526
Equator
Diameter, em
1.905 1.620 1.723 1.519 1.529
Average
1.50 1.35 1.39 1.24 1.47
Core
1.56 1.39 1.45 1.19 1.34
1.59 1.43 1.43 1.22 1.22
1.060 1.059 1.029 0.984 0.830
Surface/core W/C) Intermediate Surface modulus ratio
Dynamic shear modulus G' x lo-' dyne/cm2
Properties of the Silicone Rubber Poppets
I N VIVO DEGRADATION OF SILICONE RUBBER POPPETS 475
. ,CENTER SLICE
SPECIMEN
LOWER SPHERICAL CAP
Fig. 1. Pattern for cutting specimens from spherical silicone rubber poppets.
prosthetic heart valves was slightly less than 2% prior to 1966, and that following a valve redesign, the incidence of variant poppets to date has been reduced to 0.04%.
Equatorial Shear Modulus For the unused poppets, the magnitude of the modulus was found to decrease with distance from the surface; in both cases, the modulus
476
CUDDIHY ET AL.
at the surface was about 6% higher than the value a t the core. This trend presumably results from the thermal history during cure because the surface achieves cure temperature faster than does the core. This skin effect is a common occurrence in the curing of elastomers. For the recovered poppets, there was a noticeable and systematic time-dependent trend in the reduction of the surface modulus relative to the core. The modulus a t the surface of the poppet implanted for 4 days was about 3% above the value a t the core. I n contrast, the modulus a t the surface of the poppet implanted for 52 days was 2% below the value at the core. For the poppet implanted for 8 years, the modulus a t the surface was 17y0below the value a t the core, a complete reversal in trend compared with the results obtained for unused poppets. These data indicate the occurrence and trend of in vivo degradation of the silicone rubber poppets. It should be pointed out that phonocardiographic techniques employed t o monitor prosthetic heart valve sounds in patients could be influenced by these modulus reductions and reversals, even though the poppets are nonvariant and still functional. The modulus data for the four Smeloff-Cutter poppets yielded a n unexpected and surprising relationship when plotted in an appropriate manner. Figure 2 is a plot of modulus a t the core, intermediate, and surface locations versus poppet diameter. The modulus dependency for each location yields a straight-line relationship with poppet diameter, and the three lines associated with each location arc essentially parallel. Note that the modulus values for the intermediate and core location of the poppet implanted for 4 days fall almost perfectly on the tie lines generated by the two unused poppets, but its surface modulus is below the surface tie line. The modulus values a t all locations of the Smeloff-Cutter poppet implanted for 52 days lay below the linear extrapolation of the tie lines (3 data point, a t the lower left-hand corner). These results, plus those for the poppet implanted for 8 years, indicate clearly that the modulus of silicone poppets decreases during human implantation, and that in time, the initial modulus distribution which increased from core t o surface is reversed in trend, as is noted for the modulus distribution of the Starr-Edwards poppet. The poppet implanted for 52 days would appear to be in an intermediate stage of modulus reversal.
I N VZVO DEGRADATION OF SILICONE RUBBER POPPETS 477
0 SURFACE
A INTERMEDIATE
0
CORE
r!l
6 h
1.1
t
-
UNFLAGGED SYMBOLS FLAGGED SYMBOLS
UNUSED POPPETS IMPLANTED POPPETS
i) FLAG DOWN
- 4dayr HUMAN IMPLANTED
i i ) FLAG UP
- 52 days HUMAN IMPLANTED
I
I
I
1S O
I
1.60
1.70
1.80
POPPET DIAMETER,
I
1.90
2.
cm
Fig. 2 . Dynamic shear modulus G’ vs. diameter for the four Smeioff-Cutter poppets.
The data in Figure 2 reveal that the surface of silicone rubber poppets undergo modulus reductions within 4 days of implantation, and that modulus reductions propagate throughout the entire volume of the poppet within 52 days. This implies that the infiltration rate of the body fluids must exceed the kinetic rate of the chemical reaction in order to reach the core without being completely consumed along the diffusion path from surface to core.
Time Dependence The initial surface modulus values of the 4 and 52 day implanted poppets may be estimated by utilizing the data in Figure 2. Thus,
CUDDIHY ET AL.
478 30
I
I
m
3 3
i
I
/
P
/
20
/ / / 1
3 2
/ / / 0
I
0 0
z 0 -
P’ 0-0 THIS STUDY
d!
M REFERENCE 13
---A
0
1
I
I
10
100
ESTIMATE FOR 8 YEARS
1
1
.ooo
lo$
IMPLANT TIME, D A Y S
Fig. 3. I n vivo reduction of the surface modulus of silicone poppets.
it was calculated that the surface modulus of the 4 day implanted poppet decreased by about 3.7% and the 52 day implanted poppet decreased by about 11%. These values are plotted in Figure 3 as open circles. Raible et a1.I2 reported a permanent 23% reduction in surface properties of variant poppets, but did not state the specific implant time. They reported only that the implant time for the various poppets studied ranged from 3.5-38 months. This data point of 23y0 is also plotted in Figure 3 for that reported time span. An examination of Figure 3 suggests the possibility that in vivo degradation may proceed faster for variant poppets than for nonvariant poppets. That is, poppets that will exhibit variance become degraded in vivo at a faster rate. Whether this result is unique for a particular batch of poppets or instead is related to differences in the patients is a matter of conjecture. The conclusion is tentative because of the paucity of data. By extrapolating the two data points plotted in Figure 3, it was estimated that the 8 year implanted poppet underwent about a 22% reduction in surface modulus (Fig. 3). Alternatively, another
I N VZVO DEGRADATION OF SILICONE RUBBER POPPETS 479
time-dependent relationship can be generated from the modulus data for the three recovered poppets. Assuming that the initial ratio of surface to core modulus (X/C)o for the recovered poppets was 1.06, as was observed for the unused poppets, then a linear log-log plot of { l - (S/C)/(X/C),} versus time may be constructed as shown in Figure 4, where the (S/C) values are the modulus ratios given in Table I for each of the recovered poppets. This treatment of data facilitates comparison between silicone rubber samples that may have been slightly different in their original state due to varying processing procedures during manufacture. A satisfactory explanation for the linear log-log relation in Figure 4 has not been determined. The plot, however, is useful for estimates of the degradation magnitude of nonvariant poppets at intermediate implantation times.
Polar Modulus After the 1.5 mm thick equatorial slabs were sliced from each of the poppets, a n effort was made to slice test specimens from the polar locations of the hemispheres remaining in each case (Fig. 1). This proved difficult t o accomplish and, as a consequence, the specimens that were prepared had noticeable irregularities. The quality of data produced by the Rheovibron is sensitive t o geometrical irregularities in the case of very small specimens.
0.011 1
I
10
I
100
I 1000
10,oc
TIME, days
Fig. 4. Plot of the function [I - (S/C)/S/C),] vs. time for the three nonvariant implanted poppets.
480
CUDDIHY ET AL.
Despite this uncertainty, there was a general trend in the data indicating that the modulus at the polar surfaces was consistently less than the modulus at the equatorial surfaces. Thus, the silicone rubber appears to be mechanically anisotropic along the poppet surface. The cause of the anisotropy is not known. However, it is possible that surface anisotropy might contribute to nonuniform swelling of the poppet from absorption of body fluids such as lipids. Nonuniform swelling has been reported as a characteristic of variant poppets.
CONCLUSIONS Silicone rubber poppets undergo in vivo degradation throughout their entire volume. In vivo degradation occurs in both variant and nonvariant poppets, but there is an indication that degradation proceeds faster for variant poppets. Mechanical properties a t the surfaces of silicone poppets degrade faster than at the poppet cores. The infiltration of body fluids into the silicone rubber poppets proceeds at a rate faster than the rate of in vivo degradation. This paper represents one phase of research performed by the Jet Propulsion Laboratory, California Institute of Technology sponsored by the National Aeronautics and Space Administration, Contract NAS7-100.
References 1. W. C. Roberts and A. G. Morrow, Amer. J . Cardiol., 22, 614 (1968). 2. J. C. Hylen, J. C. Kloster, A. Starr, and H. E. Griswold, Ann. Znt. Med., 72, 1 (1970). 3. S. Koorajian, in New Industries and Applications for Advanced Materials Technology, SAMPE Symposium Publication, 19, 566 (1974). 4. E. G. Laforet, New Engl. J . Med., 276, 1025 (1967). 5. T . J. Donovan, H. B. C. Low, C . A. Bucknam, and P. V. Stroughton, Vasc. Surg., 5, 48 (1971). 6. A. Krosnick, J. Amer. Med. Assoc., 191, 105 (1965). 7. K. Hameed, S. Ashfaq, and D. 0. W. Waugh, Arch. Path., 86, 521 (1968). 8. J. Moacanin, D. D. Lawson, H. P. Chin, E. C. Harrison, and D. H. Blankenhorn, Biomat. Med. Dev. Artif. Organs, 1, 183 (1973). 9. H. P. Chin, E. C. Harrison, D. H. Blankenhorn, and J. Moacanin, Circulation, Suppl. I to 43 and 44, 1-51 (1971). 10. W. C. Roberts, G. E. Levinson, and A. G. Morrow, Arch. Znt. Med., 126, 517 (1970).
I N VIVO DEGRADATION OF SILICONE RUBBER POPPETS 481 11. R. Carmen and S. C. Mutha, J . Biomed. Mater. Res., 6,327 (1972). 12. D. A. Raible, D. P. Keller, W. R. Pierie, S. Koorajian, and E. G. Partridge, Rubber Chem. Technol., 39, 1276 (1966).
Received August 24, 1975 Revised November 10, 1975
