In Vivo Evaluation of a High-Strength, High-Ductility Stainless Steel for Use in Surgical Implants BARRY C. SYRETT and EDWARD E. DAVIS, SRI International, Menlo Park, California 94025
Summary A high-strength, high-ductility, austenitic stainless steel has been evaluated for use in surgical implants by performing ~n uiuo tests in rats, rabbits, dogs, and rhesus monkeys. This stainless steel, a TRIP (TRansformation Induced Plasticity) steel containing about 4% Mo, was compared with two alloys in current clinical use: Type 316L stainlsss steel and cast Vitallium. Compared with the other two alloys, cast Vitallium generally had higher resistance to corrosion and superior biocompatibility in all animals. The tests in rats and dogs indicated that the corrosion resistances of the TRIP steel and the Type 316L stainless steel were similar and that the tissue reactions caused by these alloys were also similar. However, in rhesus monkeys, the TRIP steel was shown to be susceptible to stress-corrosion cracking and much more susceptible to crevice corrosion than Type 316L stainless steel. Limited tests in rabbits supported the observation that the TRIP steel is susceptible to stress-corrosion cracking. These inconsistencies in the in vioo tests underline the need for a reevaluation of the popular test techniques and of the animals commonly chosen for assessing the suitability of candidate implant materials. The “worst case” results from the rhesus monkey tests were entirely consistent with previous results obtained from in uitro studies. However, further work must be performed before the behavior of metals in humans, rhesus monkeys, or any other animal, can be predicted with confidence from an in uitro test program.
INTRODUCTION
A family of high-strength, high-ductility stainless steels, called TRIP (TRansformation Induced Plasticity) steels, has been proposed for use in surgical implants. These stainless steels have been evaluated in our laboratory using both i n vitro and in vivo test procedures. This paper presents the results of the in viuo studies, and other redescribe the progress made in concurrent in uitro studies. The scope of the in vitro test program was relatively wide, enJournal of Biomedical Materials Research, Vol. 13,543-556 (1979) 0021-9304/79/0013-0543$01.00 0 1979 John Wiley & Sons, Inc.
544
SYRETT AND DAVIS
compassing all important forms of localized corrosion (pitting, crevice corrosion, stress corrosion, corrosion fatigue, and fretting corrosion). The resistances of several TRIP steels to these various forms of corrosion were compared with the corrosion resistances of Type 316L stainless steel and cast Vitallium* (a Co-Cr-Mo alloy), two materials in current clinical use. Some, but not all, of the TRIP steels were shown to have a high resistance to these forms of corrosion, comparing very favorably with Type 316L stainless steel. The resistance of cast Vitallium to localized corrosion was generally somewhat higher than either Type 316L stainless steel or the TRIP steels. The present study was designed to evaluate the biocompatibility of three of the alloys (one of the TRIP steels, Type 316L stainless steel, and cast Vitallium) that were studied in the in uitro test program mentioned above. Because of time constraints, it was necessary to initiate this in uiuo study before any in uitro test data were available. By chance, the TRIP steel used in the animal experiments was subsequently shown3 to have an in uitro resistance to localized corrosion that was inferior to Type 316L stainless steel. Thus, the results of the present investigation were expected to reflect this inferior corrosion resistance and confirm the predictive capability of the in uitro experiments. The TRIP steel chosen for the in uiuo program contained nominally 4% Mo and was used in the warm-rolled (not cold-worked) condition. This steel is termed 4 Mo TRIP No. 1-0%CW so as to be consistent with the nomenclature used elsewhere.1-6 Similarly, the Type 316L stainless steel, which was tested in the as-received (cold-worked) condition, is termed 316L-CW, and the Vitallium, which was also tested in the as-received condition, is termed Vitallium-as-cast.
EXPERIMENTAL The in uiuo studies consisted of implanting powdered metals and metal coupons into muscle of rats, metal coupons into muscle and bone of rabbits, metal coupons into muscle and bone of dogs, and metal bone plates on the long bones of rhesus monkeys. Details of the alloy compositions, preparation of the metal specimens, and surgical procedures follow. *Vitallium is a trademark of Howmedica, Inc.
STAINLESS STEEL FOR SURGICAL IMPLANTS
545
Materials The chemical compositions of the 316L-CW, Vitallium-as-cast, and 4 Mo TRIP No. 1-0%CW used in this study are shown in Table I. Fabrication histories, heat treatments, and mechanical properties are given in ref, 3.
Powder Preparation Each powder sample was prepared by abrading the appropriate bulk material with a carefully cleaned and degreased file that had not been used previously. The metal powders produced by filing were collected, degreased, and dried in warm air. To remove traces of the file material, the powders were washed with 20-50 vol % nitric acid. When all reaction had ceased, the powders were flushed thoroughly with deionized water, rinsed with alcohol, and dried in warm air. Each powder was sieved to obtain the +325/-200 fraction (a particle size of 45-75 pm). The resulting powder particles were of irregular shape and their surfaces contained many sharp protrusions.
Coupon and Bone Plate Preparation Cylindrical coupons of four sizes were used in this program. The largest, having nominal dimensions of 6.4 mm diam X 25 mm long, were implanted in the perivertebral muscles of dogs. Coupons nominally 3.2 mm diam X 12.5 mm long were implanted into the femurs of dogs and into the perivertebral muscles of rabbits. Coupons nominally 1.6 mm diam X 12.5 mm long were implanted into the femurs of rabbits. The smallest cylindrical coupons, nominally 1.6 mm TABLE I Chemical Compositions of the Test Alloys Allov
Fe
Ni
Bal 13.8 316L-CW Vitallium-as-cast 0.18 0.29 4 MoTRIP No. Bal 7.27 1-0% cw ND = Not determined. Bal = Balance.
Cr
Composition (wt. %) Mo C Mn Si
P
17.4 2.14 0.026 1.62 0.57 0.020 27.4 5.10 0.20 0.85 0.80 ND 12.5 3.88 0.29 0.011 0.05 0.008
S
Co
0.002 ND ND Bal 0.002 ND
SYRETT AND DAVIS
546
diam x 6.3 mm long, were implanted into the perivertebral muscles of rats. Each 316L-CW and 4 Mo TRIP No. 1-0% CW coupon was machined a little oversize, then electropolished to size in a bath of 60 vol % H3P04 40 vol % H2S04 a t 6OOC. The electropolishing solution was contained in a platinum beaker that also served as the cathode of the electropolishing cell. Satisfactory results were obtained when the voltage and current density were maintained in the ranges of 7-9 V and 1-5 A/cm2, respectively. Vitallium-as-cast could not be electropolished satisfactorily, probably because it is a multiphase material. Thus, all machined coupons of Vitallium-as-cast were mechanically polished with successively finer grades of abrasive paper and were given a final polishing with a soft cloth impregnated with 3-pm jeweler’s rouge. Sherman-type bone plates were attached with mating screws onto each humerus and each femur of four rhesus monkeys. The Vitallium-as-cast bone plates, provided by Howmedica, Inc., each had a length of 2.97 cm and a total surface area of 4.19 cm2. The 316L-CW and 4 Mo TRIP No. 1-0%CW bone plates, fabricated in our laboratory, each had a length of 2.98 cm and a total surface area of 4.79 em2. Each bone plate was equipped with two mating screws, about 0.95 cm long, fabricated from the same material as the bone plate. The Vitallium-as-cast bone plates were used in the condition (unpolished) they were received from the manufacturer. Bone plates fabricated from the other materials were mechanically polished and then electropolished before use, as described above. All metal coupons, bone plates, and screws were steam-sterilized for 20 min a t 127°C before use.
+
Surgical Procedures Twenty-five male and twenty-five female Fischer-344 rats were used for each of the three test alloys to test for possible carcinogenicity, and an equal number of rats was used as a control group. The right thigh muscle of each of the 200 rats was injected either with 5 mg of a powdered metal suspended in 0.2 ml of pure, synthetic trioctanoin or, in the case of the control group, with 0.2 ml of the vehicle alone. The rats were weighed daily during the first week, again after three and after five weeks, and monthly thereafter. All animals were palpated twice a month to check for the presence of nodules. At the
STAINLESS STEEL FOR SURGICAL IMPLANTS
547
end of the 20- to 24-month test period, the tissue surrounding the powdered metal was examined for tumors. Clean surgical procedures were used to implant coupons of 4 Mo TRIP No. 14%CW and of the control materials into the perivertebral muscles of nineteen rats. Normally, each rat was implanted with one 316L-CW7one Vitallium-as-cast, and two 4 Mo TRIP No. 1-0%CW coupons but, in one instance, two 316L-CW and two 4 Mo TRIP No. 1-0% CW coupons were implanted and, in another instance, one 316L-CW, two Vitallium-as-cast, and one 4 Mo TRIP No. 1-0%coupons were implanted. Two weeks after implantation of the coupons, four rats were sacrificed so that the coupons and surrounding tissue could be examined. Similar procedures were followed four and eight weeks after implanting the coupons. The remaining seven rats were sacrificed 35 to 38 weeks after implantation. In each instance, the tissue was fixed in 10% formalin for 48 hr before the coupons were retrieved; the tissue was given a histopathological examination and the coupons were examined optically and microscopically. These tissue and metal specimen handling procedures were also employed after sacrificing all other animals used in this study. Our original protocol, modeled on ASTM protocol F361-72, called for a series of implants in muscle and bone of rabbits. However, because our rabbits had a low surgical survival rate, associated with their poor tolerance of general anesthetics, this aspect of the program was severely curtailed. Using aseptic surgical procedures, two Vitallium-as-cast, two 316L-CW, and four 4 Mo TRIP No. 1-0%CW coupons were successfully implanted into the perivertebral muscles of each of two rabbits; in addition, one Vitallium-as-cast, one 316L-CW, and four 4 Mo TRIP No. 1-0%CW coupons were implanted into the femurs of one of these rabbits. The rabbit that had coupons only in muscle was sacrificed after 4 weeks and the rabbit with coupons in muscle and bone was sacrificed after 104 weeks. Aseptic surgical procedures were adopted for implanting metal coupons into the perivertebral muscles and femurs of six mongrel dogs (both male and female). Normally, two Vitallium-as-cast, two 316L-CW, and eight 4 Mo TRIP No. 1-0% CW coupons were implanted into the muscles of each dog and a similar (but physically smaller) set of twelve coupons was implanted into holes drilled into the femurs of each dog. However, in two instances, one fewer 316L-CW coupon and one fewer 4 Mo TRIP No. 1-0% CW coupon were implanted into the dog muscles, and, in another case, one fewer
548
SYRETT AND DAVIS
316L-CW coupon was implanted into the dog bones. Four of the dogs were sacrificed 40 to 44 weeks after implantation of the coupons and two were sacrificed after 104 weeks. Aseptic surgical procedures were used to implant one 316L-CW, one Vitallium-as-cast, and two 4 Mo TRIP No. 1-0%CW boneplatehrew assemblies into each of four adult male Macaca mulatta. One bone plate was attached with its two mating screws to each humerus and to each femur of these rhesus monkeys. The animals were sacrificed 100 to 104 weeks after implantation of the metal specimens.
RESULTS Rats The rats injected with the metal powders gained weight during the test period at the same rate as the control animals. No nodules were detected during the biweekly examinations, nor were any tumors visible after the animals were sacrificed. All metal coupons retrieved from rat muscle were clean and bright and showed no signs of corrosion either visually or under 9X magnification. There was no clearly visible tissue reaction at any implant site, but histopathological examination indicated the occurrence of some reaction at a few 4 Mo TRIP No.14%CW and control material sites. In samples taken 2 weeks after implantation of the coupons, one 316L coupon and one 4 Mo TRIP No. 1-0% CW coupon had connective tissue capsules that were thicker than the capsules surrounding any of the other implants; both of these thicker capsules were observed in the same animal. In the 4-week samples, one Vitallium-as-cast capsule was thicker and more cellular than any other capsule, and a t one 4 Mo TRIP No.1-0%CW implant site, a foreignbody granuloma was observed immediately exterior to the capsule. In samples taken 8 weeks after implantation of the coupons, one Vitallium-as-cast implant site had a thicker capsule and three small granulomas were observed adjacent to it. Three (one of each material) of the 28 samples retrieved after 35 to 38 weeks had moved from the sites of implantation: two were found in the fatty tissue between the skin and muscle and one was found in the abdominal cavity. Of the 25 samples that remained in the muscle for 35 to 38 weeks, one Vitallium-as-cast sample and one 4 Mo TRIP No.1-0%CW sample
STAINLESS STEEL FOR SURGICAL IMPLANTS
549
were surrounded by relatively acellular capsules that were about twice the thickness of any other capsule. Thus, some tissue reaction was observed at three of the thirty-six 4 MqTRIP No. 1-0%CW implant sites examined, three of the eighteen Vitallium-as-cast sites, and one of the nineteen 316L-CW sites. It appears that 4 Mo TRIP No. 1-0%CW performed as well as, if not better than, the control materials in these tests.
Rabbits None of the coupons retrieved after 4 weeks in rabbit muscle was corroded or tarnished, but one of the four 4 Mo TRIP No. 1-0%CW implant sites showed prominent muscle fiber regeneration. The other seven implant sites were completely normal and indistinguishable from one another. Similarly, uniform and indistinguishable capsules were observed at all but one of the 104-weekimplant sites in muscle. At this site, one of the 4 Mo TRIP No. 1-0%CW sites had a focus of pigmented macrophages between the capsule wall and the surrounding muscle; a metallographic examination of this 4 Mo TRIP No. 1-0%CW sample revealed intergranular cracks (see Fig. 1)emanating from the metal surface. It seems likely that residual stresses introduced during sample preparation had been high enough to induce stress-corrosion cracking in this sample. This was the only metal sample implanted in rabbit muscle or bone that showed any sign of corrosion. The six samples examined after 104 weeks in rabbit bone were all essentially within normal limits: all three materials had a similar effect in that they caused slight thickening of the femurs in the implanted region. Thus, compared with the control materials, 4 Mo TRIP No. 1-0% CW appears to have somewhat inferior corrosion resistance and biocompatibility when tested in rabbit muscle. This contrasts the equal (possibly superior) performance of 4 Mo TRIP No. 1-0%CW in rats. However, it must be emphasized that tests were performed in only two rabbits so there are insufficient data for a sound statistical analysis. Dogs None of the metallic coupons retrieved after 40 to 44 weeks of implantation in dog muscle were obviously corroded, and no tissue re-
550
SYRETT AND DAVIS
Fig. 1. Intergranular stress-corrosion cracks in a 4 Mo TRIP No. 14%CW coupon implanted for 104 weeks in rabbit muscle.
action was observed at any implant site. Histopathologicalevaluation of these tissue samples confirmed that all implant materials had behaved similarly. However, on close inspection at low magnification, some of these coupons were found to be tarnished or slightly tarnished in very small areas either on the ends or on the sides of the coupons. Of the thirty 4 Mo TRIP No. 1-0%CW coupons implanted, nine were tarnished and eleven were slightly tarnished in very small areas; of the six 316L-CW coupons implanted, four were tarnished and one was slightly tarnished. One of the eight Vitallium-as-cast coupons was tarnished and one was slightly tarnished. Similarly, none of the coupons retrieved after 104 weeks of implantation in dog muscle were obviously corroded and both visual and histopathologicalexamination indicated no occurrence of tissue reaction and no pigmentation at any implant site. However, close inspection again revealed some tarnishing in a few small areas. Of the sixteen 4 Mo TRIP No. 1-0%CW coupons, five were tarnished and one was slightly tarnished; of the four 316L-CW coupons, two were tarnished and the other two were slightly tarnished. None of the four Vitallium coupons were tar-
STAINLESS STEEL FOR SURGICAL IMPLANTS
551
nished. If any importance is to be attached to the tarnishing, it must be concluded that 4 Mo TRIP No. 1-0%CW performed as well as, if not better than, 316L-CW but that Vitallium-as-cast was somewhat superior to both the other alloys. All the coupons implanted in dog bone for 40 to 44 weeks appeared clean and bright but they were all held tightly by the bone growth that had begun to encapsulate the coupons. A t two 4 Mo TRIP No. l-Wo CW implant sites in one dog, a histopathologicalexamination revealed a large osteoid mass adjacent to the implant. As with the muscle coupons, a careful examination revealed that small areas on some of the coupons were tarnished or slightly tarnished. Of the thirty-two 4 Mo TRIP No. 1 4 %CW coupons implanted for 40 to 44 weeks, three were tarnished and one was slightly tarnished; of the eight 316L-CW coupons, one was tarnished. None of the eight Vitallium-as-cast coupons were tarnished. The coupons implanted in dog bone for 104 weeks also showed no obvious signs of corrosion, but new bone growth had caused a complete and very tight encapsulation of all coupons in these two animals. No other tissue reaction was observed visually or during histopathological examination. Examination at low magnification showed that one of the sixteen 4 Mo TRIP No. 1-0% CW had tarnished and that three were slightly tarnished. None of the four Vitallium-as-cast coupons were tarnished, but one of the three 316L-CW coupons was slightly tarnished. Again, the 4 Mo TRIP No. 1-00’ CW appears to have performed as well as 316L-CW but not quite as well as Vitallium-as-cast. Weight-loss measurements on a representative number of muscle and bone coupons confirmed that corrosion had been minimal (10.1 mg weight loss).
Rhesus Monkeys A visual inspection of the bone plates and surrounding tissue in rhesus monkeys showed that one of the four Vitallium-as-cast plates was partially encapsulated by new bone growth and one of the four 316L-CW was completely encapsulated; both of these plates were in the same animal. Seven of the eight 4 Mo TRIP No. 1-00?CW plates were partially or, in three instances, almost completely encapsulated. Visual observation suggested that the extent of encapsulation was related to the extent of corrosion. Each of the eight bone-plate assemblies fabricated from 316L-CW or Vitallium-as-cast, whether or not they had been encapsulated,
552
SYRETT AND DAVIS
experienced less than 0.7 mg weight loss (equivalent to
Summary A high-strength, high-ductility, austenitic stainless steel has been evaluated for use in surgical implants by performing ~n uiuo tests in rats, rabbits, dogs, and rhesus monkeys. This stainless steel, a TRIP (TRansformation Induced Plasticity) steel containing about 4% Mo, was compared with two alloys in current clinical use: Type 316L stainlsss steel and cast Vitallium. Compared with the other two alloys, cast Vitallium generally had higher resistance to corrosion and superior biocompatibility in all animals. The tests in rats and dogs indicated that the corrosion resistances of the TRIP steel and the Type 316L stainless steel were similar and that the tissue reactions caused by these alloys were also similar. However, in rhesus monkeys, the TRIP steel was shown to be susceptible to stress-corrosion cracking and much more susceptible to crevice corrosion than Type 316L stainless steel. Limited tests in rabbits supported the observation that the TRIP steel is susceptible to stress-corrosion cracking. These inconsistencies in the in vioo tests underline the need for a reevaluation of the popular test techniques and of the animals commonly chosen for assessing the suitability of candidate implant materials. The “worst case” results from the rhesus monkey tests were entirely consistent with previous results obtained from in uitro studies. However, further work must be performed before the behavior of metals in humans, rhesus monkeys, or any other animal, can be predicted with confidence from an in uitro test program.
INTRODUCTION
A family of high-strength, high-ductility stainless steels, called TRIP (TRansformation Induced Plasticity) steels, has been proposed for use in surgical implants. These stainless steels have been evaluated in our laboratory using both i n vitro and in vivo test procedures. This paper presents the results of the in viuo studies, and other redescribe the progress made in concurrent in uitro studies. The scope of the in vitro test program was relatively wide, enJournal of Biomedical Materials Research, Vol. 13,543-556 (1979) 0021-9304/79/0013-0543$01.00 0 1979 John Wiley & Sons, Inc.
544
SYRETT AND DAVIS
compassing all important forms of localized corrosion (pitting, crevice corrosion, stress corrosion, corrosion fatigue, and fretting corrosion). The resistances of several TRIP steels to these various forms of corrosion were compared with the corrosion resistances of Type 316L stainless steel and cast Vitallium* (a Co-Cr-Mo alloy), two materials in current clinical use. Some, but not all, of the TRIP steels were shown to have a high resistance to these forms of corrosion, comparing very favorably with Type 316L stainless steel. The resistance of cast Vitallium to localized corrosion was generally somewhat higher than either Type 316L stainless steel or the TRIP steels. The present study was designed to evaluate the biocompatibility of three of the alloys (one of the TRIP steels, Type 316L stainless steel, and cast Vitallium) that were studied in the in uitro test program mentioned above. Because of time constraints, it was necessary to initiate this in uiuo study before any in uitro test data were available. By chance, the TRIP steel used in the animal experiments was subsequently shown3 to have an in uitro resistance to localized corrosion that was inferior to Type 316L stainless steel. Thus, the results of the present investigation were expected to reflect this inferior corrosion resistance and confirm the predictive capability of the in uitro experiments. The TRIP steel chosen for the in uiuo program contained nominally 4% Mo and was used in the warm-rolled (not cold-worked) condition. This steel is termed 4 Mo TRIP No. 1-0%CW so as to be consistent with the nomenclature used elsewhere.1-6 Similarly, the Type 316L stainless steel, which was tested in the as-received (cold-worked) condition, is termed 316L-CW, and the Vitallium, which was also tested in the as-received condition, is termed Vitallium-as-cast.
EXPERIMENTAL The in uiuo studies consisted of implanting powdered metals and metal coupons into muscle of rats, metal coupons into muscle and bone of rabbits, metal coupons into muscle and bone of dogs, and metal bone plates on the long bones of rhesus monkeys. Details of the alloy compositions, preparation of the metal specimens, and surgical procedures follow. *Vitallium is a trademark of Howmedica, Inc.
STAINLESS STEEL FOR SURGICAL IMPLANTS
545
Materials The chemical compositions of the 316L-CW, Vitallium-as-cast, and 4 Mo TRIP No. 1-0%CW used in this study are shown in Table I. Fabrication histories, heat treatments, and mechanical properties are given in ref, 3.
Powder Preparation Each powder sample was prepared by abrading the appropriate bulk material with a carefully cleaned and degreased file that had not been used previously. The metal powders produced by filing were collected, degreased, and dried in warm air. To remove traces of the file material, the powders were washed with 20-50 vol % nitric acid. When all reaction had ceased, the powders were flushed thoroughly with deionized water, rinsed with alcohol, and dried in warm air. Each powder was sieved to obtain the +325/-200 fraction (a particle size of 45-75 pm). The resulting powder particles were of irregular shape and their surfaces contained many sharp protrusions.
Coupon and Bone Plate Preparation Cylindrical coupons of four sizes were used in this program. The largest, having nominal dimensions of 6.4 mm diam X 25 mm long, were implanted in the perivertebral muscles of dogs. Coupons nominally 3.2 mm diam X 12.5 mm long were implanted into the femurs of dogs and into the perivertebral muscles of rabbits. Coupons nominally 1.6 mm diam X 12.5 mm long were implanted into the femurs of rabbits. The smallest cylindrical coupons, nominally 1.6 mm TABLE I Chemical Compositions of the Test Alloys Allov
Fe
Ni
Bal 13.8 316L-CW Vitallium-as-cast 0.18 0.29 4 MoTRIP No. Bal 7.27 1-0% cw ND = Not determined. Bal = Balance.
Cr
Composition (wt. %) Mo C Mn Si
P
17.4 2.14 0.026 1.62 0.57 0.020 27.4 5.10 0.20 0.85 0.80 ND 12.5 3.88 0.29 0.011 0.05 0.008
S
Co
0.002 ND ND Bal 0.002 ND
SYRETT AND DAVIS
546
diam x 6.3 mm long, were implanted into the perivertebral muscles of rats. Each 316L-CW and 4 Mo TRIP No. 1-0% CW coupon was machined a little oversize, then electropolished to size in a bath of 60 vol % H3P04 40 vol % H2S04 a t 6OOC. The electropolishing solution was contained in a platinum beaker that also served as the cathode of the electropolishing cell. Satisfactory results were obtained when the voltage and current density were maintained in the ranges of 7-9 V and 1-5 A/cm2, respectively. Vitallium-as-cast could not be electropolished satisfactorily, probably because it is a multiphase material. Thus, all machined coupons of Vitallium-as-cast were mechanically polished with successively finer grades of abrasive paper and were given a final polishing with a soft cloth impregnated with 3-pm jeweler’s rouge. Sherman-type bone plates were attached with mating screws onto each humerus and each femur of four rhesus monkeys. The Vitallium-as-cast bone plates, provided by Howmedica, Inc., each had a length of 2.97 cm and a total surface area of 4.19 cm2. The 316L-CW and 4 Mo TRIP No. 1-0%CW bone plates, fabricated in our laboratory, each had a length of 2.98 cm and a total surface area of 4.79 em2. Each bone plate was equipped with two mating screws, about 0.95 cm long, fabricated from the same material as the bone plate. The Vitallium-as-cast bone plates were used in the condition (unpolished) they were received from the manufacturer. Bone plates fabricated from the other materials were mechanically polished and then electropolished before use, as described above. All metal coupons, bone plates, and screws were steam-sterilized for 20 min a t 127°C before use.
+
Surgical Procedures Twenty-five male and twenty-five female Fischer-344 rats were used for each of the three test alloys to test for possible carcinogenicity, and an equal number of rats was used as a control group. The right thigh muscle of each of the 200 rats was injected either with 5 mg of a powdered metal suspended in 0.2 ml of pure, synthetic trioctanoin or, in the case of the control group, with 0.2 ml of the vehicle alone. The rats were weighed daily during the first week, again after three and after five weeks, and monthly thereafter. All animals were palpated twice a month to check for the presence of nodules. At the
STAINLESS STEEL FOR SURGICAL IMPLANTS
547
end of the 20- to 24-month test period, the tissue surrounding the powdered metal was examined for tumors. Clean surgical procedures were used to implant coupons of 4 Mo TRIP No. 14%CW and of the control materials into the perivertebral muscles of nineteen rats. Normally, each rat was implanted with one 316L-CW7one Vitallium-as-cast, and two 4 Mo TRIP No. 1-0%CW coupons but, in one instance, two 316L-CW and two 4 Mo TRIP No. 1-0% CW coupons were implanted and, in another instance, one 316L-CW, two Vitallium-as-cast, and one 4 Mo TRIP No. 1-0%coupons were implanted. Two weeks after implantation of the coupons, four rats were sacrificed so that the coupons and surrounding tissue could be examined. Similar procedures were followed four and eight weeks after implanting the coupons. The remaining seven rats were sacrificed 35 to 38 weeks after implantation. In each instance, the tissue was fixed in 10% formalin for 48 hr before the coupons were retrieved; the tissue was given a histopathological examination and the coupons were examined optically and microscopically. These tissue and metal specimen handling procedures were also employed after sacrificing all other animals used in this study. Our original protocol, modeled on ASTM protocol F361-72, called for a series of implants in muscle and bone of rabbits. However, because our rabbits had a low surgical survival rate, associated with their poor tolerance of general anesthetics, this aspect of the program was severely curtailed. Using aseptic surgical procedures, two Vitallium-as-cast, two 316L-CW, and four 4 Mo TRIP No. 1-0%CW coupons were successfully implanted into the perivertebral muscles of each of two rabbits; in addition, one Vitallium-as-cast, one 316L-CW, and four 4 Mo TRIP No. 1-0%CW coupons were implanted into the femurs of one of these rabbits. The rabbit that had coupons only in muscle was sacrificed after 4 weeks and the rabbit with coupons in muscle and bone was sacrificed after 104 weeks. Aseptic surgical procedures were adopted for implanting metal coupons into the perivertebral muscles and femurs of six mongrel dogs (both male and female). Normally, two Vitallium-as-cast, two 316L-CW, and eight 4 Mo TRIP No. 1-0% CW coupons were implanted into the muscles of each dog and a similar (but physically smaller) set of twelve coupons was implanted into holes drilled into the femurs of each dog. However, in two instances, one fewer 316L-CW coupon and one fewer 4 Mo TRIP No. 1-0% CW coupon were implanted into the dog muscles, and, in another case, one fewer
548
SYRETT AND DAVIS
316L-CW coupon was implanted into the dog bones. Four of the dogs were sacrificed 40 to 44 weeks after implantation of the coupons and two were sacrificed after 104 weeks. Aseptic surgical procedures were used to implant one 316L-CW, one Vitallium-as-cast, and two 4 Mo TRIP No. 1-0%CW boneplatehrew assemblies into each of four adult male Macaca mulatta. One bone plate was attached with its two mating screws to each humerus and to each femur of these rhesus monkeys. The animals were sacrificed 100 to 104 weeks after implantation of the metal specimens.
RESULTS Rats The rats injected with the metal powders gained weight during the test period at the same rate as the control animals. No nodules were detected during the biweekly examinations, nor were any tumors visible after the animals were sacrificed. All metal coupons retrieved from rat muscle were clean and bright and showed no signs of corrosion either visually or under 9X magnification. There was no clearly visible tissue reaction at any implant site, but histopathological examination indicated the occurrence of some reaction at a few 4 Mo TRIP No.14%CW and control material sites. In samples taken 2 weeks after implantation of the coupons, one 316L coupon and one 4 Mo TRIP No. 1-0% CW coupon had connective tissue capsules that were thicker than the capsules surrounding any of the other implants; both of these thicker capsules were observed in the same animal. In the 4-week samples, one Vitallium-as-cast capsule was thicker and more cellular than any other capsule, and a t one 4 Mo TRIP No.1-0%CW implant site, a foreignbody granuloma was observed immediately exterior to the capsule. In samples taken 8 weeks after implantation of the coupons, one Vitallium-as-cast implant site had a thicker capsule and three small granulomas were observed adjacent to it. Three (one of each material) of the 28 samples retrieved after 35 to 38 weeks had moved from the sites of implantation: two were found in the fatty tissue between the skin and muscle and one was found in the abdominal cavity. Of the 25 samples that remained in the muscle for 35 to 38 weeks, one Vitallium-as-cast sample and one 4 Mo TRIP No.1-0%CW sample
STAINLESS STEEL FOR SURGICAL IMPLANTS
549
were surrounded by relatively acellular capsules that were about twice the thickness of any other capsule. Thus, some tissue reaction was observed at three of the thirty-six 4 MqTRIP No. 1-0%CW implant sites examined, three of the eighteen Vitallium-as-cast sites, and one of the nineteen 316L-CW sites. It appears that 4 Mo TRIP No. 1-0%CW performed as well as, if not better than, the control materials in these tests.
Rabbits None of the coupons retrieved after 4 weeks in rabbit muscle was corroded or tarnished, but one of the four 4 Mo TRIP No. 1-0%CW implant sites showed prominent muscle fiber regeneration. The other seven implant sites were completely normal and indistinguishable from one another. Similarly, uniform and indistinguishable capsules were observed at all but one of the 104-weekimplant sites in muscle. At this site, one of the 4 Mo TRIP No. 1-0%CW sites had a focus of pigmented macrophages between the capsule wall and the surrounding muscle; a metallographic examination of this 4 Mo TRIP No. 1-0%CW sample revealed intergranular cracks (see Fig. 1)emanating from the metal surface. It seems likely that residual stresses introduced during sample preparation had been high enough to induce stress-corrosion cracking in this sample. This was the only metal sample implanted in rabbit muscle or bone that showed any sign of corrosion. The six samples examined after 104 weeks in rabbit bone were all essentially within normal limits: all three materials had a similar effect in that they caused slight thickening of the femurs in the implanted region. Thus, compared with the control materials, 4 Mo TRIP No. 1-0% CW appears to have somewhat inferior corrosion resistance and biocompatibility when tested in rabbit muscle. This contrasts the equal (possibly superior) performance of 4 Mo TRIP No. 1-0%CW in rats. However, it must be emphasized that tests were performed in only two rabbits so there are insufficient data for a sound statistical analysis. Dogs None of the metallic coupons retrieved after 40 to 44 weeks of implantation in dog muscle were obviously corroded, and no tissue re-
550
SYRETT AND DAVIS
Fig. 1. Intergranular stress-corrosion cracks in a 4 Mo TRIP No. 14%CW coupon implanted for 104 weeks in rabbit muscle.
action was observed at any implant site. Histopathologicalevaluation of these tissue samples confirmed that all implant materials had behaved similarly. However, on close inspection at low magnification, some of these coupons were found to be tarnished or slightly tarnished in very small areas either on the ends or on the sides of the coupons. Of the thirty 4 Mo TRIP No. 1-0%CW coupons implanted, nine were tarnished and eleven were slightly tarnished in very small areas; of the six 316L-CW coupons implanted, four were tarnished and one was slightly tarnished. One of the eight Vitallium-as-cast coupons was tarnished and one was slightly tarnished. Similarly, none of the coupons retrieved after 104 weeks of implantation in dog muscle were obviously corroded and both visual and histopathologicalexamination indicated no occurrence of tissue reaction and no pigmentation at any implant site. However, close inspection again revealed some tarnishing in a few small areas. Of the sixteen 4 Mo TRIP No. 1-0%CW coupons, five were tarnished and one was slightly tarnished; of the four 316L-CW coupons, two were tarnished and the other two were slightly tarnished. None of the four Vitallium coupons were tar-
STAINLESS STEEL FOR SURGICAL IMPLANTS
551
nished. If any importance is to be attached to the tarnishing, it must be concluded that 4 Mo TRIP No. 1-0%CW performed as well as, if not better than, 316L-CW but that Vitallium-as-cast was somewhat superior to both the other alloys. All the coupons implanted in dog bone for 40 to 44 weeks appeared clean and bright but they were all held tightly by the bone growth that had begun to encapsulate the coupons. A t two 4 Mo TRIP No. l-Wo CW implant sites in one dog, a histopathologicalexamination revealed a large osteoid mass adjacent to the implant. As with the muscle coupons, a careful examination revealed that small areas on some of the coupons were tarnished or slightly tarnished. Of the thirty-two 4 Mo TRIP No. 1 4 %CW coupons implanted for 40 to 44 weeks, three were tarnished and one was slightly tarnished; of the eight 316L-CW coupons, one was tarnished. None of the eight Vitallium-as-cast coupons were tarnished. The coupons implanted in dog bone for 104 weeks also showed no obvious signs of corrosion, but new bone growth had caused a complete and very tight encapsulation of all coupons in these two animals. No other tissue reaction was observed visually or during histopathological examination. Examination at low magnification showed that one of the sixteen 4 Mo TRIP No. 1-0% CW had tarnished and that three were slightly tarnished. None of the four Vitallium-as-cast coupons were tarnished, but one of the three 316L-CW coupons was slightly tarnished. Again, the 4 Mo TRIP No. 1-00’ CW appears to have performed as well as 316L-CW but not quite as well as Vitallium-as-cast. Weight-loss measurements on a representative number of muscle and bone coupons confirmed that corrosion had been minimal (10.1 mg weight loss).
Rhesus Monkeys A visual inspection of the bone plates and surrounding tissue in rhesus monkeys showed that one of the four Vitallium-as-cast plates was partially encapsulated by new bone growth and one of the four 316L-CW was completely encapsulated; both of these plates were in the same animal. Seven of the eight 4 Mo TRIP No. 1-00?CW plates were partially or, in three instances, almost completely encapsulated. Visual observation suggested that the extent of encapsulation was related to the extent of corrosion. Each of the eight bone-plate assemblies fabricated from 316L-CW or Vitallium-as-cast, whether or not they had been encapsulated,
552
SYRETT AND DAVIS
experienced less than 0.7 mg weight loss (equivalent to
