[12] Bowman, L., Characterization of tissue growth into pellets

and partial sections of porous ceramics implanted in bone, M.S. Thesis, Clemson University (1971).

[13] Hulbert, S. F., Klawitter, J. J., Sauer, B. W., and Matthews, J. R., Characterization of tissue ingrowth into porous bioceramics, Technical Report No. 2, Office of Naval Research Contract No. N00014–71-A-0339-0001 (May 1972).

[14] Talbert, C. D., A basic investigation into the potential of ceramic materials as permanently implantable skeletal prostheses, M. S. Thesis, Clemson University (1969). [15] Welsh, R. P., Pilliar, R. M., and Macnab, Ian., Surgical implants, Journal of Bone and Joint Surgery 53-A, No. 5, pp. 963-977 (July 1971).

[16] Greenlee, T. K., Jr., Beckham, C. A., Crebo, A. R., and Malmorg, J. C., Glass ceramic bone implants, H. Biomed. Mater. Res. 6, pp. 235-244 (1972).

[17] Hamner, James E., III, Reed, O. M., and Greulich, R. C., Ceramic root implantation in baboons, J. Biomed. Mater. Res. Symposium, No. 3, pp. 1-13 (1972).

[18] Bartels, John C., Dental ceramics, Journal of Prosthetic Dentistry 11, No 3, pp. 537-551 (1961).

[19] Monroe, E. A., Votava, Ward, Bass, D. B., and McMullen, James, New calcium phosphate ceramic material for bone and tooth implants, Journal of Dental Research 50, No. 4 (1971).

[20] Benson, J., Elemental carbon as a biomaterial, J. Biomed. Mater. Res. Symposium 5, No. 2, Part 1, pp. 41-47 (1971). [21] Hench, L. L., and Paschall, H. A., Direct chemical bond of bioactive glass-ceramic materials to bone and muscle, Symposium on Materials and Design Considerations for the Attachment of Prostheses to the Musculo-Skeletal System, Clemson University (1972).

[22] Klawitter, J. J., and Hulbert, S. F., Application of porous ceramics for attachment of load bearing internal orthopedic appliances, J. Biomed. Mater. Res. 5, pp. 161-231 (1971). [23] Peterson, C. D., Miles, J. S., Solomons, C., Predecki, P. K., and Stephen, J. S., Union between bone and implants of open pore ceramic and stainless steel, a histological study, Journal of Bone and Joint Surgery 51-A, pp. 805 (1969). [24] Topazian, R. G., Hamner, W. B., Talbert, C. D., and Hulbert, S. F., The use of ceramics in augmentation and replacement of portions of the mandible, J. Biomed. Mater. Res. Symposium, No. 2, pp. 311-332 (1972).

[25] Leonard, R. B., Sauer, B. W., Hulbert, S. F., and Per-Lee, J. H., Use of porous ceramics to obliterate mastoid cavities, Symposium on Materials and Design Considerations for the Attachment of Prostheses to the MusculoSkeletal System, Clemson University (1972).

[26] Stanitski, C. L., and Mooney, Vert., Osseous attachment to vitreous carbons, Symposium on Materials and Design Considerations for the Attachment of Prostheses to the Musculo-Skeletal System, Clemson University (1972). [27] Laing, P. G., Ferguson, A. B., and Hodge, E. S., Tissue reaction in rabbit muscle exposed to metallic implants, J. Biomed. Mater. Res. 1, pp. 135–149 (1967).

[28] Ferguson, A. B., Laing, P. G., and Hodge, E. S., The ionization of metal implants in living tissues, Journal of Bone and Joint Surgery 42-A, No. 1, pp. 77–90 (January 1960). [29] Wood, N. K., Kaminski, E. J., and Oglesby, R. J., The significance of implant shape in experimental testing of biologi cal materials: disc vs. rod, J. Biomed. Mater. Res. 4, pp. 1-12 (1970).

[30] Stinson, N. E., The tissue reaction induced in rats and guineapigs by polymethylmethacrylate (acrylic) and stainless

steel (18/8/Mo), British Journal of Experimental Pathology 45, No. 1 (1964).

[31] Cohen, Jonathan, Assay of foreign-body reaction, The Journa of Bone and Joint Surgery 41-A, No. 1, pp. 152–1 (January 1959).

[32] Brown, D. E., Tissue reaction to plastic and metal implants. Arch Otolaryng. 88, pp. 283-287 (September 1962 [33] Fitzpatrick, Barry, A comparative study of some implant materials, Australian Dental Journal 13, pp. 422-434 (December 1968).

[34] Roydhouse, R. H., Implant testing of polymerizing materials, J. Biomed. Mater. Res. 2, pp. 265-277 (1968).

[35] Johnsson-Hegyeli, R. I. E., and Hegyeli, A. F., Interaction of blood and tissue cells with foreign surfaces, J. Biomed Res. 3, pp. 115-144 (1969).

[36] Johnsson, R. I., and Hegyeli, A. F., Tissue culture techniques for screening of prosthetic materials, Annals of the New York Academy of Sciences 146, pp. 66-76(1968).

[37] Homsy, C. A., McDonald, K. E., Akers, W. W., Short. C. and Freeman, B. S., Surgical suture-canine tissue inter action for six common suture types, J. Biomed. Mater Res. 2, pp. 215-230 (1968).

[38] Autian, John, Toxicologic aspects of implants, J. Biomed Mater. Res. 1, pp. 433-449 (1967).

[39] Rosenbluth, S. A., Weddington, G. R., Guess, W. L., and Autian, J., Tissued culture method for screening texerts of plastic materials to be used in medical practice. Journa of Pharmaceutical Sciences 54, No. 1, pp. 156–15× (January 1965).

[40] Homsy, C. A., Ansevin, K. D., O'Brannon, W., Thompson S. C., Hodge, R., and Estrella, M. E., Rapid in vitro screea ing of polymers for biocompatibility, J. Macromol. Sci. Chem. A4, No. 3, pp. 615-634 (May 1970).

[41] Cohen, Jonathan, Tissue reactions to metals - the influence of surface finish, The Journal of Bone and Joint Surgery 43-A, No. 5, pp. 687-699 (July 1961).

[42] Evarts, C. M., Steffee, A. D., and McCormack. L. J., Ir vestigation of canine tissue reaction to TEE-fluorocarts # resin, to high-density polyethylene and to vitalium Journal of Surgical Research 10, No. 2, pp. 91-44 (February 1970).

[43] Ludwigson, D. C., Today's prosthetic metals, Journal Metals 16, pp. 226-231 (1964).

[44] Hulbert, S. F., Morrison, S. J., and Klawitter, J. J., Problems associated with determining compatibility of biomaterias Symposium-Workshop on Research Techniques in B materials Evaluation, Clemson University (March 19711 [45] Richbourg, Henry Leroy, Evaluation of a metal-ceram composite hip prosthesis, M.S. Thesis, Clemson Universa (1973).

[46] Laing, P. G., Ferguson, A. B., and Hodge, E. S., Spectr chemical determination of trace metals in normal strated muscle in the rabbit, The Journal of Bone and Joint Surgery 41-A, No. 4, pp. 737-744 (June 1959).

[47] Ferguson, A. B., Akahoshi, Y., Laing, P. G., and Hodge, E. S. Trace metal ion concentration in the liver, kidney, spleet and lung of normal rabbits, The Journal of Bone and Joint Surgery 44-A, No. 2, pp. 317–322 (March 1962[48] Hulbert, S. F., Klawitter, J. J., and Richardson, W. C Jr. Soft tissue response to a dense ceramic and varying surfac characteristics, 3rd Annual Meeting of the Biomed, a Engineering Society, Baltimore, Maryland (April 7, 197 [49] Homsy, Charles A., Stanley, Rufus F., Anderson, M. Sidnes and King, Joe W., Reduction of tissue and bone adhes, to cobalt, alloy fixation appliances, J. Biomed. Mater. Res 6, pp. 451-464 (1972).

(Issued May 1975).

1. Introduction

A successful orthopaedic implant must possess sufficient mechanical strength to resist the stresses it encounters in service and it should be compatible with its host environment. A yield strength of 100,000 psi (.07 GPa), tensile strength of 150,000 psi (0.1 GPa) and an elongation of 10 percent probably represent minimal characteristics. High yield and tensile strengths are needed to prevent bending and breaking due to tensile overload and fatigue, while adequate elongation is needed to avoid brittleness. In addition, alloys should be easy to fabricate into complex shapes and consequently, those which can be strengthened by heat treatment after machining hold special interest. To be compatible, an alloy should corrode very slowly when implanted and its corrosion products should not affect adjacent tissue (inflammatory response) and produce no adverse reactions in the rest of the organism (systemic effects). It is especially important that the alloy resists crevice corrosion since many implants are multicomponent devices containing shielded areas.

None of the current implant alloys meet the above requirements. Specifically, Type 316 stainless steel lacks sufficient strength and corrosion resis

tance, especially crevice corrosion. Cast Vitallium ' although much more inert and consequently more compatible with adjacent tissue, tends to exhibit brittle behavior. Also, the fact that it can be fabricated only by casting is a limitation. Haynes 25 ("Ductile Vitallium") has the strength of cast Vitallium, together with sufficient ductility to prevent brittleness and to permit fabrication by conventional methods. Its corrosion resistance, although not as good as Vitallium, is superior to stainless steel. Ideally, its strength should be somewhat higher. The actual biological compatibility of this alloy is not known since the manufacturer has not distinguished it from cast Vitallium in his products. Titanium is very corrosion-resistant and may be the least toxic although this has not been confirmed by quantitative clinical tests. Unfortunately, the strength of pure titanium is essentially identical to stainless steel.

Summarizing, all of the presently employed implant alloys are inadequate. Alternate materials are needed since only small improvements can be

1 Certain commercial materials and instruments may be identified in this publication in order to adequately specify the experimental procedure. In no case does such identification imply recommendation or endorsement by the National Bureau of Standards, nor does it imply that the equipment or instruments identified are necessarily the best available for the purpose.

accomplished with current implant alloys. Cold working is utilized by most manufacturers to increase the yield strength of Type 316 stainless steel, but this is done at the expense of decreasing ductility and it cannot be conveniently applied to all devices. Coating stainless steel implants to improve their corrosion resistance is limited by two factors. First, orthopaedic implant surfaces are subjected to high tensile and abrasive forces. Coating failures would most likely occur at the most critical areas (e.g., at bone plate-screw interfaces). More importantly, even if an ideal coating were applied to stainless steel, it would not alter its basically inferior mechanical properties. Similarly, there is no way to significantly increase the ductility of Vitallium or the strength of titanium. The only approach capable of producing a significant improvement is a modification of alloy composition (i.e., substitution).

Historically, the development of implant alloys has utilized empirical in vivo implantation tests. The American Society for Testing and Materials (ASTM) is now standardizing such a test for the purposes of evaluating new implant materials [1].2 Cylindrical specimens are inserted in the muscles and bones of rabbits, rats and/or dogs and histological examinations made at various intervals up to 104 weeks. However, this test is insufficiently sensitive to detect significant differences between any of the present implant alloys, as shown by the studies of Laing et al. [2]. Current implant materials, in the absence of crevices, corrode so slowly that histologically observable tissue reaction does not occur. Thus, crevice-free samples of present orthopaedic alloys are sufficiently inert to render conventional implantation methods obsolete. The solution to this is the implantation of large numbers of multicomponent appliance samples for extended periods (e.g., 5 years or more). This of course would be prohibitively expensive and time-consuming except for more promising candidate materials. Sensitive, accurate screening techniques for selecting potential alloys are therefore required for future implant alloy development. In the future, as implants remain in place for longer periods, the possi bilities of local inflammatory complications and toxic systemic effects become increasingly important.

2. Research Approach

New, improved implant alloys together with rapid, sensitive methods for predicting their in vivo performances are urgently needed. Actually, a survey of present commercial alloys shows that many materials with properties superior to these now used for implants already exist. This is especially true of the superalloys - high strength, corrosion resistant alloys developed for jet turbine applications. For example, Elgiloy and MP-35N closely

* Figures in brackets indicate the literature references at the end of this paper.

resemble wrought Vitallium on the basis of composi tion and corrosion resistance. Both possess markedly superior strength characteristics. Elgiloy is used as lead wires in cardiac pacemakers because of its high strength, resistance to fatigue fracture and excellent tissue compatibility. There is little doubt that these alloys could function as improved orthopaedic implants.

Although high yield and tensile strengths together with good ductility are desirable for screws and bone plates, there may be optimum limits for these mechanical properties. There is some evidence to indicate that less ductility is required as tensile and yield strength are increased. Also, it is likely that beyond a certain tensile strength level it is impossible to torsionally fracture bone screws during conventional orthopaedic procedures. If so, then higher mechanical properties could be "traded off" for better corrosion resistance, lower toxicity or other features.

Since there are many potentially useful implant alloys, selection can be made on the basis of minimum long term inflammatory and systemic effects. To accomplish this, three major coordinated research studies are used in this program. Corrosion measurements are being used to determine the rate of metal solution into adjacent tissue. These data are then utilized to determine inflammatory and systemic effects. Also, optimum mechanical properties for bone screws are being determined by laboratory and in vivo experiments. The interactions between these different studies is schematically illustrated in figure 1. The corrosion rates of most current and potential implant alloys are so low that loss of mechanical strength is not an important consideration. Consequently, the corrosion rates of most implanted metals are meaningless, per se. Corrosion rate only has biological significance, and this can be described in terms of inflammatory and systemic effects. As noted in the diagram, there is feedback between these two general biological effects and the alloy corrosion rate. These effects are influenced by both the specific metallic elements and their rate of discharge (corrosion rate) under in vivo conditions.

Although mechanical properties can be used to qualitatively select potentially useful alloys for orthopaedic applications, the optimum properties can only be defined by empirical mechanical tests with actual appliances (e.g., bone screw and plates). Figure 1 illustrates this aspect by showing that mechanical properties can be used to select suitable alloys but the selection process can be improved by feedback dialog between engineers and orthopaedic surgeons. It follows that the selection of new implant alloys requires an interdisciplinary research team, competent in the areas of metallurgy, corrosion, orthopaedic surgery, inflammatory response and systemic effects.

3. Corrosion Studies

3.1. Background

There are two general features important in evaluating the corrosion characteristics of material. These are the rate and type of corrosive attack. For implant applications, the corrosion rate determines the amount of foreign metal compounds released into surrounding tissues.

To avoid significant local reaction, an implant must corrode at a very slow rate. In general, rates two or three orders of magnitude lower than those commonly employed in conventional industrial practice are necessary [3]. The measurement of such very low rates introduces experimental difficulties since conventional weight loss methods which are usually employed are insufficiently sensitive. Fortunately, electrochemical methods [3-8], particularly linear polarization techniques and charging curve analyses [3], are ideally suited for the measurements of ultra low rates. Briefly, these methods involve the measurement of the potentialcurrent relationships of a corroding specimen near its corrosion potential. Over a potential range of 10 mv or less, there is a linear relationship between voltage and applied current. The slope of this linear region can be used to calculate the instantaneous corrosion rate of a metal or alloy. The accuracy of these techniques have been verified under in vitro and in vivo conditions. Corrosion rates, especially those of alloys protected by passive films, usually vary with time. Thus, the ability to measure corrosion as a function of time is very valuable in comparing implant performance.

The evaluation of resistance to localized corrosion such as crevice attack can also be accomplished by electrochemical techniques. The initiation of crevice attack is readily detected by monitoring the corrosion rate of a sample via linear polarization techniques or charging curve analyses [9]. Also, it has been shown that characteristic changes in the corrosion potential also accompany initiation of localized corrosion [9]. Localized attack, because it is concentrated at isolated points, can be detected by microscopic (light and electron) tech

niques. Such methods are also very useful for examining specimens after in vivo exposure.

3.2. Procedures

Pretreatments of implants, such as sterilization procedures, cause marked changes in the corrosion resistance of a given material under both in vitro and in vivo environments [8, 10]. Also, the surface finish of an implant (e.g., polished versus grit blasted) has a major influence on corrosion characteristics [11]. Thus, in screening possible implant alloys it is important that they be tested with a representative surface finish after exposure to typical clinical sterilization techniques. Where possible, highly polished surfaces which have been exposed to high temperature steam or dry heat sterilization for long periods have been employed since these conditions yield surfaces with the greatest corrosion resistance [10].

The corrosion rates of various implant alloys are being measured as a function of time using linear and transient linear polarization methods together with charging curve analyses [3]. Also, when applicable, passive current-time measurements [10] are used to monitor corrosion rates. Passive current measurements are simpler than linear polarization techniques and they have the added advantage of permitting continuous monitoring of corrosion rate during the test period. Corrosion rates are monitored for periods ranging between 100 hours and several months depending on the behavior of a particular system. The corrosion rate of most metals decreases exponentially with time [3, 10] in a regular fashion. Therefore, the relative resistances of various materials can be compared on the basis of short time exposures. However, the most promising materials will be subjected to long term tests in both in vitro and in vivo environments.

Resistance to crevice corrosion can be evaluated with the above electrochemical measurements. Since the corrosion potential must be monitored during linear polarization measurements, potential changes which accompany the initiation of crevice corrosion can be detected. Thus, the experimental procedures used to determine corrosion-time characteristics are also capable of detecting localized attack.

Isotonic saline solutions at average body temperature can be used to qualitatively measure the corrosion rate of implant alloys [8]. Although the absolute value of the corrosion rates measured under these conditions may not correspond to in vivo environments, they are sufficiently accurate to determine if a potential alloy has a sufficiently low corrosion rate to be considered for further testing. The most promising materials will be subjected to in vivo tests utilizing specially modified electrochemical apparatus and procedures [4, 8]. Rats, rabbits, dogs, and monkeys will be employed. Of special interest are comparisons of different in vivo

environments. Previous studies [8] suggest that they are very similar, but additional data are needed.

3.3. Results

Since crevice corrosion is the most common form of corrosion damage observed on surgical implants, we have devoted special attention to this aspect of the program. A new technique for quantitatively evaluating the crevice corrosion susceptibility of alloys has been developed [11] and is presently being refined further. The method employs a paper covered metal sample immersed in a suitable electrolyte. The presence of the paper simulates the restrictive convection conditions within a crevice without the resistance drop ("IR") errors normally associated with shielded areas. Crevice corrosion susceptibility is directly related to the change in dissolution currents observed in the presence of a paper substrate.

Figure 2 illustrates the construction of the paper shielded electrode used in these measurements, and figure 3 shows the effect of paper shielding on Type 304 stainless steel-a material which is very susceptible to crevice attack. Photographic paper with its gelatin coating provides a very effective convective diffusion barrier and its influence on alloy dissolution rate is very pronounced. Titanium, which is almost completely resistant to crevice attack, shows the opposite effect - paper shielding decreases dissolution current (fig. 4). This decrease in current is probably due to the increase in pH caused by the accumulation of sulfate ions. Additional metals and alloys are now being compared using this procedure in several different electrolytes.

Passive currents versus time are being determined in isotonic saline at 37 °C using newly developed mini potentiostats [12]. These devices, based on solid state operational amplifiers, are ideally suited for multiple, long time measurements because of

THREADED MOUNTING ROD

FIGURE 4. Polarization curves of titanium in sulfuric acid solution.

No sensitivity to crevice attack in the active region is indicated.

their simplicity and low cost. These potentiostats together with specially designed polarization cells and a shielded wiring system are being used to collect passive current-time data. MP35N alloy possesses passive corrosion characteristics similar to wrought Vitallium (HS-2 alloy). This together with its outstanding mechanical properties indicates potential usefulness as an implant alloy. However, the recent systemic kinetic analyses of nickel distribution (see below) suggest that although nickel base alloys such as MP35N may appear promising, their use should be approached with caution because of possible toxic and carcinogenic effects.