J. BIOMED. MATER. RES.
VOL. 10, PP. 391-397 (1976)
Tissue Response to Implanted Polymers: The Significance of Sample Shape BARBARA F. MATLAGA, LEWIS P. YASENCHAK, and THOMAS N. SALTHOUSE, Histology and Histochemistry Section, Ethicon Research Foundation, Somerville, New Jersey 08876
Summary Studies were designed to demonstrate the need for standardization of shape of samples used as implants to evaluate histotoxicity of polymer materials. Six medical-grade polymers (polypropylene, polyethylene, polyurethane, silicone rubber, poly(viny1 chloride), and Teflon) were extruded as rods with circular-, triangular-, and pentagonal-shaped crosss ections, and were implanted in rat gluteal muscles for 14 days. Evaluation of the tissue response was assessed by quantitating cellular lysosomal acid phosphatase enzyme activity by using microspectrophotometry. All triangular-shaped implants showed the highest enzyme activity and cellular response; pentagon shapes showed less, and circular rods showed the lowest activity. The results demonstrate the need for standard sample shape for valid comparative studies of tissue response to implanted polymers.
INTRODUCTION The need for safety evaluations of plastic materials has become evident during the past years as the number of new medical, dental, and pharmaceutical devices has increased. Many methods have been devised to test these materials before they come in contact with the body. Generally, they include an in vivo muscle implantation test with subsequent histological evaluation of the tissue Often, protocols describing these techniques lack adequate descriptions of the size and shape of samples to be implanted. Kaminski and Oglesby4 and Wood et al.5 compared rod and disk shapes of stainless steel implants in rabbit muscles. However, little work has been done to compare the effects of different shapes of polymeric materials on the tissue response. 39 I @ 1976 by John Wiley & Sons, Inc.
392
MATLAGA, YASENCHAK, AND SALTHOUSE
MATERIALS AND METHODS Poly(viny1 chloride) was received as flexible poly(viny1 chloride), Unichem Lot # ?815A, from Colorite Plastics, Inc., Ridgefield, New Jersey. Polyethylene (Bakelite DYNH-1, Blend ET 8952) was a product of the Union Carbide Corp., Bound Brook, New Jersey. Polypropylene (suture resin, lot # 2A 27, virgin) and Polyurethane (BIOMER segmented polyether polyurethane, lot # G170.1, virgin) were both received from Ethicon, Inc., Somerville, New Jersey. Teflon (Grade 100, lot 10572 FEP Fluorocarbon Resin) was supplied by Du Pont Plastics Dept., Wilmington, Delaware. Silicone rubber was a product of Ja-Bar Silicone Corp., Andover, New Jersey. The above medical-grade polymers were extruded by using a C. W. Brabender Modular Extrusion Instrument which allowed for optimum processing conditions such as screw compression, type of feed, zone temperature, die temperature, and screw speed rpm, for each type of polymer. All samples were extruded through stainless steel dies to yield rods with the following cross-sectional dimensions; a 1 mm diameter circle, an equilateral triangle 1 mm on each side, and a pentagon 0.6 mm on each side. All samples had identical cross-sectional perimeters (Fig. 1). In this way, upon implantation, the cellular environment about the implants would be exposed to
Fig. 1. Photomicrograph of cross sections of polymer shapes: triangle, circle, and pentagon. (X 28).
SAMPLE SHAPES OF IMPLANTED POLYMERS
393
the same amount of polymer surface, the only difference being the shape of the surface. Fifteen-millimeter rods of each polymer configuration were cut, washed, dried, and ethylene oxide-sterilized in a Cryothermunit (AMSCO) with a 46 1. capacity chamber. Exposure conditions: 6 hr; 900-1100 mg/l. ethylene oxide (12-88 ethylene oxide-Freon 12 mixture); 130 f 5°F; 50 f 5y0 relative humidity. Degassing time was 8 hr a t 130°F and 30 in. Hg vacuum. Polymers known to carry over high amounts of ethylene oxide residues, the polyurethene and poly (vinyl chloride), were analyzed by using the head-space gaschromatography method.6 No residues were detected in these samples. The sterile 15 mm lengths were implanted into the gluteal muscles of female Long Evans rats with a 16 gauge needle and stylet. For sham controls, only the needle and stylet were inserted and immediately withdrawn. Rats received halothane anesthesia. Skin incisions were closed with Michel clips. Three different configurations of each polymer type were implanted according to the scheme in Figure 2. This one-way analysis of variance design allowed us to make valid comparisons of the three different shapes of the same polymer type; 4 implants per shape, 6 rats per polymer type. To measure the cellular response to the implants, a method for quantitating cellular enzyme activity was used, rather than relying solely on the routine subjective histological evaluation of hematoxylin-eosin-stained paraffin sections. This method, which is based on the lysosomal acid phosphatase enzyme activity of the macroIMPLANTATION SCHEME Sample C o n f i g u r a t i o n
O O A L
Rat#l R
Rat#2 L
Rat#2 R
Rat#3 L
Rat#3 R
Rat#4 L
Rat#4 R
Rat#5 L
Rat#5 R
Rat#6 L
Rat#6 R
Rat#l Polymer Type
Fig. 2. A one-way ANOVA design used to give 4 muscle implants per shape, 6 rats per polymer type. (L) left gluteal muscle; (R) right gluteal muscle.
394
MATLAGA, YASENCHAK, AND SALTHOUSE
phage population around the implant, has been reported in great detail by Salthouse et al.79s It will, therefore, only be described briefly here. After 14 days, all animals containing the same polymer type were killed, muscles excised, and an area through the midportion of the polymer rod was quick-frozen with the implant in cross section. Remaining muscle tissue was fixed in 10% buffered formalin and processed through paraffin or methacrylate embedding techniques. Frozen sections of the implants were cut at 10 p in a cryostat and treated to demonstrate acid phosphatase activity, giving an insoluble red-colored product at the site of the enzyme. This dye product was then quantitated by microspectrophotometry using a Zeiss MPM unit. Each implant cross section was measured and an optical density reading was recorded. The data was then analyzed using a one-way analysis of variance to discover significant differences among the three sample shapes.
RESULTS The one-way analysis of variance computations showed statistical differences among sample shapes at the 5% level. The triangularshaped rods of all polymer types showed the highest activity of lysosomal acid phosphatase in cells surrounding the implant. The
Fig. 3. Photomicrographs of cryostat sections treated to demonstrate acid phosphatase activity around the three configurations of PVC implants at 14 days. (X 45).
SAMPLE SHAPES OF IMPLANTED POLYMERS
395
.13 0
.12
+
POLYETHYLENE
.11
d .1P 0
.09
.oa
A
9
+
A A 0
+ a
0
POLYPROPYLENE P V C
0
TEFLON POLYURETHANE
A
SILICONE
RUBBER
0
+
07
4
w
A
0
Fig. 4. Acid phosphatase activity at 14 days measured by microspectrophotometry and recorded as optical density (O.D.) at 530 nm. Each point is an average of four readings obtained from 4 implants. All triangular-shaped rods showed the highest enzyme activity; pentagon shapes showed less, and circular rods showed the lowest enzyme activity.
pentagon-shaped rods showed less activity and the circular rods revealed the lowest activity (Fig. 3). All 14 day sham controls showed no enzyme activity or cellular reaction whatsoever. Results are recorded in Figure 4. The foreign body reaction at 14 days (Fig. 5) contains mostly macrophage cells with an outer ring of fibroblasts and is typical of what we normally see around nontoxic implants in rat gluteal muscle. The greatest number of cells was seen around the triangular-shaped rods.
DISCUSSION Our results show that implant shape can have a significant effect on the in vivo tissue reaction in rat gluteal muscle. This was demonstrated through the accurate measurement of the tissue response around rod implants of three different cross-sectional configurations, a circle, triangle, and pentagon, using microspectrophotometric quantitation of acid phosphatase activity as a parameter. The triangularshaped implants of all polymer types showed the greatest reaction, both in numbers of cells and enzyme activity. Although this effect may be the result of a variety of factors, the most probable cause is
396
MATLAGA, YASENCHAK, AND SALTHOUSE
Fig. 5 . Photomicrographs of H&E-stained sections. Triangular cross section on the left, pentagon in the center, and circular on the right. All implants are PVC a t 14 days. (X 450).
the movement of the implant in the muscle, the triangular shape having the most acute angles and initiating the greatest tissue response. The consequence of implant movement must be considered in any implantation procedure employing muscle tissue. I t was thought that the surface-to-volume ratio of the configurations implanted may have also had an effect on the resultant tissue reaction. However, the triangular implant had the smallest ratio and, in all cases, gave the largest tissue response. Although rat muscle is not a common site for implantation, it was used in this study because it was large enough to accommodate our samples and was found to give less variation in response than rabbit muscle. “Satellite” necrosis, which has been reported to occur in tissue surrounding implants in rabbit^,^ is rarely seen in rat muscle tissue, even with severely toxic implants. Fourteen-day sham sites in rat gluteal muscle revealed no tissue response or enzyme activity. However, Kaminski et al.9reported evidence of surgical trauma, consisting of a minimal inflammatory response, in rabbit thigh muscle 6 weeks after surgery. For the reasons cited above, the authors believe that rat gluteal muscle offers some advantages over rabbit tissue for the type of implantation studies described in this report.
SAMPLE SHAPES OF IMPLANTED POLYMERS
397
We believe the results of our implantation studies point out the need for standardization of sample shape in order to obtain valid comparative results in implantation studies. Our experiences with the quantitation of lysosomal acid phosphatase activity as a measure of tissue response around implants indicate the value of this approach to obtain objective data for the in vivo evaluation of polymeric materials. The authors wish to thank Mr. David Braze1 for preparation of the photographs and Mrs. Ruth Boyle for the typing of this manuscript.
References 1. W. L. Guess and J. Autian, Amer. J . Hosp. Pharm., 21,261 (1964). 2. J. Autian, J . Biomed. Muter. Res., 1, 433 (1967). 3. J. E. Turner, W. H. Lawrence, and J. Autian, J . Biomed. Muter. Res., 7, 29 (1973). 4. E. Kaminski and R. J. Oglesby, Res. Dent. Med. Muter., Proc. Sump., 1968, 113 (pub. 1969). 5. N. K. Wood, E. J. Kaminski, and R. J. Oglesby, J . Biomed. Muter. Res., 4, 1 (1970). 6. S. Romano, J. A. Renner, and P. M. Leitner, Anal. Chem., 45, 2347 (1973). 7. T. N. Salthouse and B. F. Matlaga, Biomater. Med. Devices Artij. Organs, 3, 47 (1975). 8. T. N. Salthouse, B. F. Matlaga, and R. K. O’Leary, 5”oxicol. Appl. Pharmacol., 25, 201 (1973). 9. E. J. Kaminski, R. J. Oglesby, N. K. Wood, and J. Sandkirk, J . Biorned. Muter. Res., 2, 81 (1968).
Received June 27, 1975 Revised August 27, 1975
VOL. 10, PP. 391-397 (1976)
Tissue Response to Implanted Polymers: The Significance of Sample Shape BARBARA F. MATLAGA, LEWIS P. YASENCHAK, and THOMAS N. SALTHOUSE, Histology and Histochemistry Section, Ethicon Research Foundation, Somerville, New Jersey 08876
Summary Studies were designed to demonstrate the need for standardization of shape of samples used as implants to evaluate histotoxicity of polymer materials. Six medical-grade polymers (polypropylene, polyethylene, polyurethane, silicone rubber, poly(viny1 chloride), and Teflon) were extruded as rods with circular-, triangular-, and pentagonal-shaped crosss ections, and were implanted in rat gluteal muscles for 14 days. Evaluation of the tissue response was assessed by quantitating cellular lysosomal acid phosphatase enzyme activity by using microspectrophotometry. All triangular-shaped implants showed the highest enzyme activity and cellular response; pentagon shapes showed less, and circular rods showed the lowest activity. The results demonstrate the need for standard sample shape for valid comparative studies of tissue response to implanted polymers.
INTRODUCTION The need for safety evaluations of plastic materials has become evident during the past years as the number of new medical, dental, and pharmaceutical devices has increased. Many methods have been devised to test these materials before they come in contact with the body. Generally, they include an in vivo muscle implantation test with subsequent histological evaluation of the tissue Often, protocols describing these techniques lack adequate descriptions of the size and shape of samples to be implanted. Kaminski and Oglesby4 and Wood et al.5 compared rod and disk shapes of stainless steel implants in rabbit muscles. However, little work has been done to compare the effects of different shapes of polymeric materials on the tissue response. 39 I @ 1976 by John Wiley & Sons, Inc.
392
MATLAGA, YASENCHAK, AND SALTHOUSE
MATERIALS AND METHODS Poly(viny1 chloride) was received as flexible poly(viny1 chloride), Unichem Lot # ?815A, from Colorite Plastics, Inc., Ridgefield, New Jersey. Polyethylene (Bakelite DYNH-1, Blend ET 8952) was a product of the Union Carbide Corp., Bound Brook, New Jersey. Polypropylene (suture resin, lot # 2A 27, virgin) and Polyurethane (BIOMER segmented polyether polyurethane, lot # G170.1, virgin) were both received from Ethicon, Inc., Somerville, New Jersey. Teflon (Grade 100, lot 10572 FEP Fluorocarbon Resin) was supplied by Du Pont Plastics Dept., Wilmington, Delaware. Silicone rubber was a product of Ja-Bar Silicone Corp., Andover, New Jersey. The above medical-grade polymers were extruded by using a C. W. Brabender Modular Extrusion Instrument which allowed for optimum processing conditions such as screw compression, type of feed, zone temperature, die temperature, and screw speed rpm, for each type of polymer. All samples were extruded through stainless steel dies to yield rods with the following cross-sectional dimensions; a 1 mm diameter circle, an equilateral triangle 1 mm on each side, and a pentagon 0.6 mm on each side. All samples had identical cross-sectional perimeters (Fig. 1). In this way, upon implantation, the cellular environment about the implants would be exposed to
Fig. 1. Photomicrograph of cross sections of polymer shapes: triangle, circle, and pentagon. (X 28).
SAMPLE SHAPES OF IMPLANTED POLYMERS
393
the same amount of polymer surface, the only difference being the shape of the surface. Fifteen-millimeter rods of each polymer configuration were cut, washed, dried, and ethylene oxide-sterilized in a Cryothermunit (AMSCO) with a 46 1. capacity chamber. Exposure conditions: 6 hr; 900-1100 mg/l. ethylene oxide (12-88 ethylene oxide-Freon 12 mixture); 130 f 5°F; 50 f 5y0 relative humidity. Degassing time was 8 hr a t 130°F and 30 in. Hg vacuum. Polymers known to carry over high amounts of ethylene oxide residues, the polyurethene and poly (vinyl chloride), were analyzed by using the head-space gaschromatography method.6 No residues were detected in these samples. The sterile 15 mm lengths were implanted into the gluteal muscles of female Long Evans rats with a 16 gauge needle and stylet. For sham controls, only the needle and stylet were inserted and immediately withdrawn. Rats received halothane anesthesia. Skin incisions were closed with Michel clips. Three different configurations of each polymer type were implanted according to the scheme in Figure 2. This one-way analysis of variance design allowed us to make valid comparisons of the three different shapes of the same polymer type; 4 implants per shape, 6 rats per polymer type. To measure the cellular response to the implants, a method for quantitating cellular enzyme activity was used, rather than relying solely on the routine subjective histological evaluation of hematoxylin-eosin-stained paraffin sections. This method, which is based on the lysosomal acid phosphatase enzyme activity of the macroIMPLANTATION SCHEME Sample C o n f i g u r a t i o n
O O A L
Rat#l R
Rat#2 L
Rat#2 R
Rat#3 L
Rat#3 R
Rat#4 L
Rat#4 R
Rat#5 L
Rat#5 R
Rat#6 L
Rat#6 R
Rat#l Polymer Type
Fig. 2. A one-way ANOVA design used to give 4 muscle implants per shape, 6 rats per polymer type. (L) left gluteal muscle; (R) right gluteal muscle.
394
MATLAGA, YASENCHAK, AND SALTHOUSE
phage population around the implant, has been reported in great detail by Salthouse et al.79s It will, therefore, only be described briefly here. After 14 days, all animals containing the same polymer type were killed, muscles excised, and an area through the midportion of the polymer rod was quick-frozen with the implant in cross section. Remaining muscle tissue was fixed in 10% buffered formalin and processed through paraffin or methacrylate embedding techniques. Frozen sections of the implants were cut at 10 p in a cryostat and treated to demonstrate acid phosphatase activity, giving an insoluble red-colored product at the site of the enzyme. This dye product was then quantitated by microspectrophotometry using a Zeiss MPM unit. Each implant cross section was measured and an optical density reading was recorded. The data was then analyzed using a one-way analysis of variance to discover significant differences among the three sample shapes.
RESULTS The one-way analysis of variance computations showed statistical differences among sample shapes at the 5% level. The triangularshaped rods of all polymer types showed the highest activity of lysosomal acid phosphatase in cells surrounding the implant. The
Fig. 3. Photomicrographs of cryostat sections treated to demonstrate acid phosphatase activity around the three configurations of PVC implants at 14 days. (X 45).
SAMPLE SHAPES OF IMPLANTED POLYMERS
395
.13 0
.12
+
POLYETHYLENE
.11
d .1P 0
.09
.oa
A
9
+
A A 0
+ a
0
POLYPROPYLENE P V C
0
TEFLON POLYURETHANE
A
SILICONE
RUBBER
0
+
07
4
w
A
0
Fig. 4. Acid phosphatase activity at 14 days measured by microspectrophotometry and recorded as optical density (O.D.) at 530 nm. Each point is an average of four readings obtained from 4 implants. All triangular-shaped rods showed the highest enzyme activity; pentagon shapes showed less, and circular rods showed the lowest enzyme activity.
pentagon-shaped rods showed less activity and the circular rods revealed the lowest activity (Fig. 3). All 14 day sham controls showed no enzyme activity or cellular reaction whatsoever. Results are recorded in Figure 4. The foreign body reaction at 14 days (Fig. 5) contains mostly macrophage cells with an outer ring of fibroblasts and is typical of what we normally see around nontoxic implants in rat gluteal muscle. The greatest number of cells was seen around the triangular-shaped rods.
DISCUSSION Our results show that implant shape can have a significant effect on the in vivo tissue reaction in rat gluteal muscle. This was demonstrated through the accurate measurement of the tissue response around rod implants of three different cross-sectional configurations, a circle, triangle, and pentagon, using microspectrophotometric quantitation of acid phosphatase activity as a parameter. The triangularshaped implants of all polymer types showed the greatest reaction, both in numbers of cells and enzyme activity. Although this effect may be the result of a variety of factors, the most probable cause is
396
MATLAGA, YASENCHAK, AND SALTHOUSE
Fig. 5 . Photomicrographs of H&E-stained sections. Triangular cross section on the left, pentagon in the center, and circular on the right. All implants are PVC a t 14 days. (X 450).
the movement of the implant in the muscle, the triangular shape having the most acute angles and initiating the greatest tissue response. The consequence of implant movement must be considered in any implantation procedure employing muscle tissue. I t was thought that the surface-to-volume ratio of the configurations implanted may have also had an effect on the resultant tissue reaction. However, the triangular implant had the smallest ratio and, in all cases, gave the largest tissue response. Although rat muscle is not a common site for implantation, it was used in this study because it was large enough to accommodate our samples and was found to give less variation in response than rabbit muscle. “Satellite” necrosis, which has been reported to occur in tissue surrounding implants in rabbit^,^ is rarely seen in rat muscle tissue, even with severely toxic implants. Fourteen-day sham sites in rat gluteal muscle revealed no tissue response or enzyme activity. However, Kaminski et al.9reported evidence of surgical trauma, consisting of a minimal inflammatory response, in rabbit thigh muscle 6 weeks after surgery. For the reasons cited above, the authors believe that rat gluteal muscle offers some advantages over rabbit tissue for the type of implantation studies described in this report.
SAMPLE SHAPES OF IMPLANTED POLYMERS
397
We believe the results of our implantation studies point out the need for standardization of sample shape in order to obtain valid comparative results in implantation studies. Our experiences with the quantitation of lysosomal acid phosphatase activity as a measure of tissue response around implants indicate the value of this approach to obtain objective data for the in vivo evaluation of polymeric materials. The authors wish to thank Mr. David Braze1 for preparation of the photographs and Mrs. Ruth Boyle for the typing of this manuscript.
References 1. W. L. Guess and J. Autian, Amer. J . Hosp. Pharm., 21,261 (1964). 2. J. Autian, J . Biomed. Muter. Res., 1, 433 (1967). 3. J. E. Turner, W. H. Lawrence, and J. Autian, J . Biomed. Muter. Res., 7, 29 (1973). 4. E. Kaminski and R. J. Oglesby, Res. Dent. Med. Muter., Proc. Sump., 1968, 113 (pub. 1969). 5. N. K. Wood, E. J. Kaminski, and R. J. Oglesby, J . Biomed. Muter. Res., 4, 1 (1970). 6. S. Romano, J. A. Renner, and P. M. Leitner, Anal. Chem., 45, 2347 (1973). 7. T. N. Salthouse and B. F. Matlaga, Biomater. Med. Devices Artij. Organs, 3, 47 (1975). 8. T. N. Salthouse, B. F. Matlaga, and R. K. O’Leary, 5”oxicol. Appl. Pharmacol., 25, 201 (1973). 9. E. J. Kaminski, R. J. Oglesby, N. K. Wood, and J. Sandkirk, J . Biorned. Muter. Res., 2, 81 (1968).
Received June 27, 1975 Revised August 27, 1975
