JOURNAL OF M A T E R I A L S S C I E N C E 22 (1987) 3 4 2 1 - 3 4 4 5

Review
Tissue-biomaterial interactions

D. F. W I L L I A M S
Institute of Medical and Dental Bioengineering, University of Liverpool,
Liverpool L69 3BX, UK

A wide variety of materials is being increasingly used in medical practice for the treatment of
patients in which the materials come into direct and often sustained contact with the tissues
of the body. The commonly-known examples of hip replacements, contact and intraocular
lenses and pace-makers serve to emphasize the importance of this subject. Because the body
is so well equipped to reject any intruding object, whether that is a bacterium or a splinter of
wood, the materials which are to stand any chance of success within this hostile yet sensitive
milieu must be chosen very carefully. The subject of biomaterials represents an almost unique
blend of physical and biological sciences, and it is becoming increasingly important that these
aspects are drawn together to help in the development of quality biomaterials that are able to
perform optimally in this environment. The key to this subject lies in the interactions that take
place between biomaterials and these tissues and this review is aimed at providing, for the
materials scientist, an understanding of the mechanisms of these interactions.

1. Introduction and a selection of materials ranging from ultrahigh

Biomaterials are those materials which are used in molecular weight polyethylene to austenitic stainless
medical, dental, veterinary or pharmaceutical appli- steel and alumina. At the other end of the spectrum is
cations and which come into intimate and sustained the total artificial heart [3], so far used in little more
contact with tissues of the body, generally (although than a handful of patients, and largely constructed
not exclusively) being implanted within these tissues. of polyurethane elastomer. In between these extremes
Put a little more succinctly, and as defined by a recent we find the replacement of the lens in the eye in
conference devoted to defining such terms [1], a bio- the treatment of cataracts [4], kidney machines [5],
material is a non-viable material used in a medical maxillofacial prostheses [6], neurological stimulators
device, intended to interact with biological systems. [7], drug depots [8], heart pacemakers [9] and many
They may be distinguished from other materials in others. The reader may refer to recent reviews for
that they possess a combination of properties, includ- further information [10-15]. The materials utilized in
ing chemical, mechanical, physical and biological these devices cover as wide a spectrum as found in any
properties that render them suitable for safe, effective industry, with examples of metals, ceramics, glasses,
and reliable use within a physiological environment, plastics, fibres, elastomers, composites and tissues
an environment that is both extremely hostile and yet themselves [16].
sensitive to and unforgiving of irritating foreign Although there are many factors which play a part
bodies. in determining the fate and effectiveness of a material
It is not uncommon, in journals such as this, which used in this way, including the mechanical and physi-
relate to materials science, to find papers dealing with cal properties, it is the interactions between the
the subject of biomaterials. Nor is it unusual for an material and the tissues that dominate the discussions.
industry that is involved with the development of new In all those applications referred to above, it is
materials, and academia that is concerned with the relatively easy to find materials that are able to satisfy
basic properties of these materials, to turn their atten- the functional requirements (e.g. strength, fatigue
tion to biomaterials for commercial or intellectual strength, rigidity, optical transparency, electrical
reasons. It is both relevant and timely, therefore, conductivity). It is very difficult, however, to find
to review some of the important aspects that are materials which are able to continue performing these
concerned with the transformation of an ordinary functions for a long period of time (more than twenty
material into a biomaterial. years in some cases) without deterioration of the
In order to familiarize the reader with the types of material itself or undesirable effects being induced in
situation that are involved here, a few examples may the tissues of the body.
be cited. Total joint replacements [2], such as the hip This review is concerned with the nature of these
and knee, are well known and represent some of the tissue-biomaterial interactions and is specifically
most successful surgical innovations of this century. aimed at the reader who has a materials science back-
Hundreds of thousands of patients receive new joints ground but who is interested in the fate of materials
worldwide each year; there is a multiplicity of designs when interfaced with tissues. These interactions are,
0022-2461/87 $03.00 + .12 © 1987 Chapman and Hall Ltd. 3421
 Figure 1 Aspects of biocompatibility: (l)
_ _ q

initial events at the surface, (2) local host

response, (3) corrosion or degradation of
material, (4) systemichost response.

of course, complex, and, it should be admitted, largely that are involved. This paper is, in fact, attempting to
unpredictable. There are many components and outline a hypothesis for the mechanism of tissue-
sequelae of the interactions, but for the purposes of biomaterial interactions.
introduction we may identify four features. These, as
shown in Fig. 1, are 2. Initial e v e n t s at t h e
tissue-biomaterial interface
(1) the initial events that take place at the inter-
For applications in the body, the duration of which
face, largely concerned with the physicochemical
may be measured in tens of years, the events which
phenomena that take place in times measured in
occur within a few seconds of contact with tissues
seconds or minutes following contact between bio-
might appear irrelevant. Indeed, for many such appli-
material and tissues;
cations there is little evidence that these initial events
(2) the effect that the presence of a foreign body has
are important. In some other situations, however, it is
on the tissue surrounding the implant, which may be
clear that the nature of this initial interaction plays a
seen at any time ranging from minutes to years;
significant role and it is necessary to review briefly the
(3) the changes seen in the material as a result of its
state of our knowledge.
presence in the tissues, usually described under the
The biomaterial-tissue interface that is established
headings of corrosion or degradation; and
on implantation is, almost inevitably, a biomaterial-
(4) the sequelae of the interfacial reaction that are
blood interface and the initial events are dominated by
seen systemically (that is, throughout the body) or at
the adsorption of proteins from the blood on to the
some specific but remote site.
surface. Baler and Dutton [17] demonstrated the
These four phenomena collectively constitute the ubiquitous and inevitable nature ot this process a
subject of biocompatibility. They may be treated number of years ago and Gendreau et al. [18] and
independently and, indeed, since the mechanism and Vroman et al. [19] determined that proteins were
principles may be so different, it is often convenient to already present on polymer surfaces within seconds
do so. On the other hand, they are all interrelated and of the exposure. At least three different driving forces
there is logic in the argument that they should be are at play here [20]. First, thermodynamically either
considered together. It is known, for example, that the enthalpy or entropy changes may be sufficiently great
corrosion of metals is influenced by pH and oxygen to provide a negative free energy change for adsorp-
potential. Since the presence of an implant in tissues tion of proteins on to polymers under physiological
can induce changes that on a very localized and micro- conditions [21, 22]. Secondly, the ambivalent polar/
scopic scale are concerned with such variables, the rate non-polar characteristics of proteins favour a concen-
of, or indeed mechanism of corrosion may vary. As tration of proteins at interfaces; and thirdly, proteins
the rate of corrosion varies, so the tissue response, are usually only sparingly soluble and adsorption
which as we shall see is partly mediated by the release increases as the solubility decreases.
of corrosion products, varies. Both processes are, The type of binding of proteins to foreign surfaces
therefore, dependent on each other. depends on the nature of the surfaces with hydro-
It is, however, impossible in a review of this nature philicity being the most frequently discussed par-
to cover all four of these aspects in any great depth. ameter. It is widely assumed that no specific covalent
Instead, the review will concentrate on the second of attachment occurs between proteins and polymers so
those listed above, the localized effects on the tissues, that all binding is secondary in nature, for example
since this is most relevant to the objectives of the hydrogen bonding. The amount that adsorbs similarly
review. This willbe preceded by brief sections on the varies; Brash and Uniyal [23] report variations
initial events and corrosion and degradation phenom- from 0.02 to 0.57/~gcm -2 for albumin and 0.034 to
ena in order to provide the basis for understanding the 1.09#gcm 2 for fibrinogen adsorbed under steady
way in which these phenomena may control the tissue state conditions on to polymers ranging from hydro-
response. The subject of systemic biocompatibility will philic to hydrophobic in character. In vivo, the situ-
not be addressed in this review. ation will probably be different for there may be an
It should be emphasized that the review will not initial maximum amount of protein adsorbed and a
attempt to catalogue the responses seen with specific subsequent decline as desorption takes place. There
materials but rather will describe the type of phenom- will almost certainly be conformational changes as the
ena that may be observed and review the mechanisms proteins organize themselves on the polymer surface.

3422
In a very extensive study of the mechanisms of protein products into the surrounding tissue, where even
adsorption on to hydrophilic polymer surfaces (con- minute amounts can have far-reaching consequences.
ducted in relation to soft contact lenses) Costillo and In passivated metals it is not necessary for breakdown
co-workers [24-27] have discussed the kinetics of of passivity to occur for metal ions will be slowly
adsorption of a range of proteins on to hydroxyethyl- released through oxide layers and cause an accumu-
methacrylate-methacrylic acid copolymers. They have lation of metal in the tissue. The effect of this corro-
shown structural changes in the adsorbed proteins, sion on the tissue response will be discussed in the next
especially the a-helix content of the protein reducing section, but it is sufficient here to note that even with
and the fl-sheet conformation increasing with time. commercially pure titanium, demonstrable amounts
When exposed to y-globulins, these materials rapidly of the metal can be found in the surrounding tissue
adsorb the protein, charge and hydrophobic interac- after relatively short periods of time [52].
tions as well as hydrogen bonding resulting in confor-
mational changes. It was suggested, and indeed seems 3.2. Polymers
very likely, that any protein or glycoprotein will In theory, polymers should have an advantage over
experience a conformational change in order to metals since the isotonic saline and protein solution
maximize these interactions. that comprises the extracellular fluid is not normally
The vast majority of work on protein adsorption associated with the degradation of synthetic high
has involved polymer surfaces and the reader is molecular weight polymers. Although all polymers are
referred to further review articles on this subject susceptible to degradation, the majority of degrada-
[28-31]. Very little has been performed in relation to tion processes involve the absorption of some kind of
metals, but the author has shown here that whilst energy that is able to cause disruption of primary
many pure metals adsorb proteins to the same extent covalent bonds to form free radicals, which then cause
as those polymers described above, other metals, such the propagation of molecular degradation by second-
as gold, silver and copper, adsorb far greater amounts, ary reactions. The conditions under which these
suggesting far stronger binding forces [32]. The signifi- processes take place include elevated temperatures,
cance of this is not yet clear. especially in the presence of oxygen, electromagnetic
radiation, mechanical stress at elevated temperatures
3. The corrosion and degradation of and ultrasonic vibration. Clearly the physiological
biomaterials environment within the human body does not offer
A considerable amount has been written on the subject any of these conditions to an implanted polymer, and
of the corrosion and degradation of biomaterials and thus high polymers such as certain polyolefins, acrylics
it is not intended to review all the information here. and halogenated hydrocarbon polymers should be
The reader is referred to a number of publications on very stable. As reviewed in this journal a few years
this subject for this detail [33-42]. ago, such is the case in practice [37].
On the other hand, some polymers contain linkages
3.1. Metals that are susceptible to hydrolysis. If such a material is
In the context of metals it is now appreciated that the hydrophilic and capable of absorbing water, then
physiological environment is extremely hostile, con- degradation is possible within the physiological
sisting of an aqueous solution of various anions, environment. This is the rationale for the choice of
cations and biological macromolecules. Corrosion many intentionally degradable polymer systems (for
phenomena in clinical practice are well known [43-47] drug delivery implants, for example) where aliphatic
and only the noblest of metals, such as gold and certain polyesters, poly(orthoesters) and poly(amino acids)
platinum group metals, or the most passive, such as may be selected [53-55]. Other polymers have been
titanium or chromium, stand any chance of keeping used for more permanent applications but suffered,
the corrosion rates within apparently acceptable albeit more slowly, from this type of hydrolytic
levels. Whereas it has for long been assumed, however, degradation, including some polyamides [56] and
that this corrosion is attributable to the chloride ion poly(ester urethanes) [57].
(which is unsurprising in view of the known aggressive- One interesting and relevant observation in this
ness of saline solutions), it has recently become respect concerns the effect that tissue enzymes may
apparent that the biological macromolecules, and have on this hydrolytic process. Enzymes are catalysts
specifically the proteins in extracellular fluids, are able for specific biochemical reactions, many of which are
to influence this corrosion considerably. This aspect hydrolytic in nature. There is now some evidence that
has been reviewed by the author [48] and in several certain enzymes are able to accelerate or induce hydro-
other papers published in the last year or so [49, 50]. lytic degradation of polymers such as poly(ether
In most engineering applications of metals and urethanes) and aromatic polyesters [58-6~] which,
alloys, the importance of corrosion lies in the effect it although containing hydrolysable ester bonds, are not
has on structural integrity; corro~ding structures fail hydrophilic and not associated with hydrolytic degra-
because they cannot support stresses, or they cause dation at low temperatures. It is also possible that
physical malfunctions. This is not the case with certain cells of the tissue response could be involved
biomaterials, for only exceptionally is a mechanical with degradative processes, either by their attachment
failure clearly associated with corrosion, excepting to polymer surfaces and the release of destructive
those cases where corrosion fatigue is important [51]. enzymes on to the surface [62] or by the ingestion of
Instead, the significance is on the release of corrosion fragments of the polymer.
3423
3.3. Ceramics and glasses It is most instructive, therefore, when considering
Ceramics, glasses and glass-ceramics present a the mechanisms by which implanted biomaterials influ-
spectrum of materials with wide-ranging character- ence the localized tissues, to consider the sequence of
istics as far as stability in physiological environments events that is seen in normal wound healing and then
is concerned. As discussed by Hench [36] and Hench to consider how this sequence is modified by the
and Etheridge [63] these materials may be divided into presence of an implant. Naturally, the precise features
those which are essentially inert, those which are sol- of the wound healing process depend on the nature of
uble, and those which display limited or controlled the tissues in question. For present purposes, we
surface reactivity. Into the former category come a may consider the situation of wound healing in soft
number of oxide ceramics, typically alumina [64] and connective tissue (for example muscle) and the response
certain hydroxylapatites, especially dense calcium of this tissue to implants. In real clinical use, implants
hydroxylapatite [65], the naturally occurring mineral may find themselves in situations other than those
phase of bones and teeth that can now be made syn- where they are encased in soft connective tissues and,
thetically. A close relative of this hydroxylapatite is indeed, the most important clinical areas in which
tricalcium phosphate but rather than being essentially biomaterials are prominent involve hard tissues (ortho-
inert, this is totally degradable, or soluble in the body paedic surgery and dentistry) and blood (cardio-
[66]. Naturally the essentially inert ceramics are vascular surgery). Some of the specific questions
used when permanency is required and the minimal or concerning the response of bone and blood to implants
undetectable degradation rate seen with oxide ceramics will, therefore, be covered later. In addition, we have
such as alumina is a very desirable property. The to consider separately the influence of other variables,
totally degradable ceramics have potential use for such as infection, on the response, and the develop-
short-term applications, especially where the implant ment of some very specific responses such as the
is being used as a matrix for new tissue regeneration. formation of tumours around implants.
The rate of degradation will depend on local con-
ditions and may involve either cellular processes or 4.1. Normal wound healing
direct solution, or both. Obviously this rate of The immediate response to injury, whatever its cause,
degradation and the nature of the degradation is inflammation. This involves vascular, neurological,
products have to be such that any deleterious effect of humoral and cellular responses and is essentially the
the degradation on the tissue response is minimized. same whether it is induced by mechanical trauma or
The concept of ceramics or glasses with controlled microbiological infection, or by electrical, chemical
surface activity [67] arises from the desire to achieve a or radiological energy [68]. This arises because the
degree of bonding to the tissues, by mechanisms which response is mediated by the same substances within
are described later. The surface reactivity is achieved the tissue in each case. The inflammatory process is
by the preferential leaching of certain components aimed at eliminating, or at least containing, the causa-
from the glass, which is comprised of SiO2, P205, CaO tive agent so that the tissue can be subsequently
and Na20 and other species, so that dissolution leaves repaired. The process of repair involves the replace-
a stable silica-rich surface. The reaction that does take ment of lost or destroyed cells by vital cells and of
place, involving the release of calcium and phosphate, damaged tissue by new tissue, and is, therefore, the
results in the formation of new bone at the implant second component of wound healing. Although the
surface. functions of the inflammatory and repair processes are
quite distinct and chronologically the containment of
4. The localized response of tissues to the injurious agent has to precede the repair, the repair
biomaterials process itself can begin during the inflammatory phase
It has to be said at this stage that there is no simple, and the two are very much closely interwoven. This
well-understood and universally applicable mech- becomes an important factor when the modification
anism by which tissues respond to the presence of an of this tissue response by implants is considered,
implanted device. The response, instead, is a dynamic since the implant represents an injurious agent that
phenomenon, a sequence of events each of which may cannot normally be eliminated and often cannot be
be triggered by its predecessor but which may also be contained.
influenced by the environmental features created by the The nature and timing of this transition from
implant. These events centre around the activity of inflammation to repair, and the extent of the clinical
different cell types, this activity being mediated by a manifestation of the inflammation, will depend on the
variety of biochemical substances which in turn are nature and severity of the injury, even though qualita-
influenced by the physical or chemical presence of the tively the processes that arc involved are non-specific
implant. We can, in fact, consider an implant as a to the causative agent. Thus, if the injury is minimal
source of irritation, or a stimulus to the tissue. This and the cause transient, the process of inflammation
stimulus provided by the implant is not a great deal does not have to deal with the containment of a harm-
different to that provided by other insults to the tissue ful agent and the repair process can proceed rapidly.
(for example trauma or infection) in terms of the With more extensive injuries, the inflammatory pro-
response that it incites, and most features of the tissue cess will be intense and may not be confined to the
response to an implant will bear a close similarity to the locality of the injury, while the repair process may
classical features of wound repair after trauma or of result in varying degrees of scarring and possibly the
the cellular and humoral response to invading bacteria. loss of specialized functions.

3424
 In those cases where the injurious agent is persistent
and cannot be eliminated, or where the inflammatory
response cannot effectively deal with a harmful agent
( a ~ (~ s°some
because of interference to normal mechanisms, the
inflammatory-reparative process becomes consider-
ably more complex, with prolonged or chronic inflam-
~"x,~ /Fo~re~igin particle
mation and excessive fibroblastic repair. It is in this
context that we have to consider implanted devices.
Nucleus
The situation becomes especially complex if the implant,
instead of acting as an immovable but otherwise non- (c)
irritant foreign body, becomes a persistent source of ~ Lysosome and
particlefuse
irritation due to the release of corrosion or degrada-
tion products or leachables, in which case the inflam-
matory response may never subside and the repair
process may never proceed to completion[
For convenience, the sequence of events may be
divided into acute inflammation, chronic inflam-
mation and repair, although there is obviously a
considerable overlap between them.

4.2. Acute inflammation

This, by definition, is the immediate response to
injury. Associated with the acute inflammation are the
mechanisms by which the two principal forces in the
Figure 2 Process of phagocytosis: (a, b) the cell ingests a foreign
body's defence capability, the leucocytes and anti- particle; (c) the particle and lysosome fuse; (d) the phagosome that
bodies, are brought into prominence and action. Since is thereby created digests the particle if possible, leaving (e) a
both of these are blood-borne, changes in the vascu- residual body.
lature are the main features of acute inflammation.
present in only small numbers. However, they are
4.2. 1. Vascular changes attracted to the area in increasing numbers and have
Immediately after localized injury, the microvascu- a longer lifespan so they soon outnumber the neutro-
lature undergoes considerable change, the vessels phils in the tissue.
becoming dilated and filling with excess blood [69]. Once in the extravascular tissue, these cells have to
This is rapidly followed by a stagnation of flow since the migrate to the precise locations where they carry out
blood becomes more viscous as water is lost into the their functions. This migration is directional and is
surrounding tissue through the now abnormally per- mediated by a chemical process known as chemotaxis
meable capillaries, yielding a situation of stasis and [74].
increased pressure. In this way, far more blood than One of the most important functions of the cells is
normal is available to the injured tissue. The blood that of phagocytosis, a defence against invading
rapidly starts to clot, while leucocytes (the white blood microbes and particles; both neutrophils and macro-
cells) and plasma proteins pass through the capillary phages are phagocytic. The process of phagocytosis
wall into the surrounding tissue. These changes in the [75] (Fig. 2) is initiated by the attachment of the
permeability that allow such processes to occur are particle to the surface of the cell, and this is facilitated
mediated by histamine, a substance that is rapidly if the particle becomes coated (in the tissues) with one
released into the tissue upon injury. of the substances (such as the immunoglobulin IgG)
for which the cell surface has receptors. Since bacteria
4.2.2. White cell activity may be coated with IgG antibodies from the serum,
During this dilation of the blood vessels, the red cells this receptor-mediated attachment, or recognition,
tend to occupy central positions within the flowing may result in a far more effective clearance rate than
blood, while the leucocytes (white cells) accumulate at with, for example, the biologically unrecognizable
the periphery [70]. There are many different types of degradation products of a biomaterial.
white cell, the most prominent being the neutrophils, After attachment, the cell engulfs the object. This
monocytes and lymphocytes [71]. The immediate con- may be easy in the case of bacterium, which becomes
tact between them and the endothelial surfaces allows enclosed within a membrane that was hitherto part of
them to stick to the latter. Emigration of the white the cell membrane, producing a phagosome. The cyto-
cells through the vessel walls then takes place, with plasmic granules of the cell then fuse with the
neutrophils and monocytes being more active in this phagosome and shed their collection of powerful
respect than the slower lymphocytes. The neutrophils enzymes to kill and cause the disintegration of the
[72] tend to dominate the extravascular spaces in the entrapped bacterium.
early stages since they are also the most numerous The release of enzymes can have a number of effects
white cells in the blood. The monocytes [73], which since they may find their way into the surrounding
are known as macrophages or histiocytes when extra- tissues and cause destruction there. This may be
vascular, are also reasonably mobile but are initially particularly important in the genesis of prolonged

3425
responses to inert foreign bodies such as implanted cells. The term is frequently used to describe the
materials. In such cases, the offending particles, unlike response to biomaterials and indeed, several specific
bacteria, may be of varying dimensions and not amen- responses have been given distinctive terms in this
able to complete and effective engulfment. This is context. For example, the release of particulate poly-
especially significant if phagocytes attempt to engulf tetrafluoroethylene from the first generation of total
particles but, while failing to do so completely, get hip prostheses gave rise to a very severe chronic
close enough to cause degranulation, with much of the response, similar to that induced by particulate PTFE
released enzyme finding its way into the surrounding in other situations, now referred to as a Teflon
tissue, thereby mediating and promoting inflammatory granuloma [85].
responses.
4.4. Reparative processes
4.2.3. Chemical mediators of inflammation The extent and nature of tissue regeneration is depen-
The development of an inflammation following an dent upon the ability of the cells within that tissue
injury is dependent upon chemical mediators [71, 76]. to replicate. Labile cells have a very large regener-
These are substances derived from either the plasma ative capacity, being able to multiply at any time
or directly from the tissue, which are able to activate throughout life in order to replace those lost during
one or more of the processes described above and normal physiological processes, an example being the
which are then deactivated, by enzymes or antagonists, cells of the epithelial surface that give rise to the
once their function has been performed. These continual replacement of skin. Stable cells may not
mediators may be divided into those which control actively replicate in the same way because there is not
vascular permeability and those which have chemo- normally any need for them to do so, but they do have
tactic capability. Histamine [77, 78], serotonin, com- the potential for replication should the need arise.
plement [79] and prostaglandins [80] are particularly Thus mesenchymal cells may differentiate (or trans-
effective mediators .of dilation and increased per- form) into chondroblasts (cartilage-producing cells)
meability of blood vessels. or osteoblasts (bone-producing) should either of these
tissues be involved in order to bring about their repair
4.3. Chronic inflammation [86]. Fibroblasts are, perhaps, the best examples of
The vascular and exudative changes that constitute mesenchymal cells th~,t are extensively involved in
the acute inflammation will subside and lead directly repair processes [87].
to the repair stage if the injurious agent is mild and Permanent cells have no replicating ability at all,
rapidly eliminated. If, however, the agent is not tran- their destruction representing permanent loss. Repair
sient but persistent, or very severe, the acute inflam- here merely involved unspecialized connective tissue
mation will be followed by a chronic inflammation, formation and scarring. Neurons in nerve tissue and
which will often occur simultaneously with the repair skeletal muscle cells are good examples, the repair of
process. The chronic inflammation is a proliferative these tissues being effected by fibroblasts which merely
rather than exudative response and the tissue is lay down collagen instead of the original type of tissue
characterized by fibroblasts associated with repair and [88].
an accumulation of leucocytes that attempt to carry
on the work of the cells of the acute response. These 4.5. Incisional w o u n d healing
cells of the chronic response largely consist of the Having considered the general mechanisms by which
macrophages, plasma cells [81] and lymphocytes [82]. tissue responds to injury, we shall not consider the
Macrophages are able to transform to several specific situation in which a surgical incision is made
derivatives, the most important of which in this situ- into the soft connective tissue, this being the normal
ation is the multinucleated foreign body giant cell. precursor to the implantation of a biomaterial. If the
They are derived, in fact, from macrophages fusing wound margins can be accurately joined without any
together in an attempt to increase their effectiveness tissue loss, healing is said to take place by primary
against larger and more resistant foreign bodies [83]. union, or by first intention. If there is some degree of
The lymphocytes and plasma cells may be seen in tissue loss, secondary union takes place. With primary
large numbers in chronic responses, especially when union (Fig. 3) the incision will initially fill with blood,
the immune system is involved. They may therefore be which will clot. The acute inflammatory response is
seen in the reaction to implants which, although rapidly initiated with the release of a neutrophil
outwardly non-immunogenic, may still be capable of exudate. The fibrin meshwork of the blood clot pro-
eliciting an immune response, which is discussed vides a basis for both re-epithelialization of the surface
briefly later. and sub-epithelial fibrous tissue growth to occur.
Capillary buds start to migrate through the clot and
4.3. 1. Granulomas fibroblasts at the edge start to become active within a
When the stimulus to inflammation is particularly day or so. Cells at the epithelial margin extend their
resistant to degradation by phagocytic cells, a special- processes across the wound surface and a thin epithelial
ized form of inflammatory tissue may develop, which surface, possibly only one cell thick, is established.
is known as granulation tissue or a granuloma [84]. The acute inflammation will start to subside after
This is defined as a focus of cells in a chronic inflam- about three days, macrophages taking over from
matory response that predominantly consists of macro- neutrophils to clear the area of dead cells and the
phages and their derivatives, the epithelioid and giant remaining fibrin. Blood vessel sprouts grow into

3426
 (a) (b) Figure 3 (a~t) Process of wound
.:.....-:-,..:-..:..'::-:..,, ~: ~..':--?"..::'.-":O'. -: :.'.-: healing, in the case of a simple
Fibrin~i'.'O'.'~T/ Capillary~!~:. "';~" incisional wound.

ii!i
::®':: ' : i : .!-"
• -::J:.
"" " ".:-~ Fibroblas~.O. "

bl°°dlsj/X.\.O.:;'. "" ~;~[--)LE ryth rocytes . It"~.:'.~.1~l

vessels ~[.:e:'.';llli ~ Iltl~: 0" "."~i~c o IIa g e n

~ ~a g ~ " O ' £ ' ~ " " "..

Neutrophils Macroph

(,) (a)

~--~'2"----- -- agen
FIbrobl

subepithelial space at a rapid rate such that by the fifth the irritant, but different components of the response
day there is a richly vascularized fibroblastic connec- may take on greater or lesser significance with vari-
tive tissue within the wound. Fibroblasts lay down ations in the conditions.
more and more collagen, which will be immature and
weak initially. By the end of the second week, the 4. 6. 1. M i n i m a l fibrosis
inflammation will have subsided, as will the intense In the case of a monolithic solid consisting of a single
vascularity. Maturation of the collagen will continue material that is neither toxic to the host, in the normal
for many weeks as it slowly increases in strength. sense of that word, nor degraded by the tissues, the
During secondary union, a more extensive repair inflammatory response and repair processes may take
process is necessary because of the greater amount of place virtually unaffected. The site of implantation
dead tissue requiring removal and the greater amount fills with blood and a fibrin network forms a basis for
of tissue that needs replacement. The edges of a large subsequent fibrous tissue growth to occur. There is an
wound are filled with granulation tissue and it is this initial interaction between the blood and the implant,
which forms a base for epithelialization to take place as indicated in Section 2, with plasma proteins being
with the migration of the epithelial cells over its adsorbed on the implant surface, the features of which
surface. New blood vessels grow in from the edges and depend to a certain extent on the chemistry and
scar tissue starts to form as fibroblast proliferation is surface morphology.
also initiated. An acute inflammatory response is initiated and it is
unlikely that the exudate will be different to that
4.6. The response of soft connective tissue produced by the incision in the absence of the implant.
to implantation Macrophages take over from the neutrophils fairly
The above discussion has shown that there are many rapidly, as before. Capillary buds migrate through the
cells and chemical mediators involved in the process of clot, but their passage across the wound is disrupted
inflammation and repair that constitute the response by the implant and the blood vessel network will
to trauma and irritation in tissues. If a surgical inevitably be different. Fibroblasts will also become
implant is placed within an incisional wound, then the active, laying down collagen as the fibrin clot is
response may be considered as a modification to one resorbed, although this will not be able to traverse the
of these processes. There are many points at which the incision in the region of the implant, resulting in an
normal sequence of events may be modified by vari- altered morphology. It is probable that the presence of
ations in the nature of the irritant. It should be the implant will prolong the inflammation and repair
emphasized that the same mechanisms exist whatever processes and the cellular infiltrate will persist for a
3427
 (a) (b)

\ ~.:'~." • . . • • • • • 11 b u d s~-'...E~ll "" •..'.. • 1#

Fibrin~--'~. '' 0"'" "." "0 t / i / Y ~ : lii'll~'" "~/'!/
• . ~ . • • . • • • . .. ~l ~ • • .

Y i\@ --- rUll ].l "--]:e//CoIlagen

'~i/a~ed ~ l i ~ ' I : ~ E r y t hrocytes ~'.~1 1"0~

tl ~ . ' / / # Macrophages X~'~

NeuIr°p hill AI l~?f/ ~V//Ei~>'°b'asts
(<) (d)

Capil : " -'~. blasts

'-_-'--.-. --~=_ ~_~-~ Co.agen

"II I'- ;i i Ill

Ii:-VI I,i!JX 7,
" I1~:1:~'-?-.~ Macr° Phages kk ~- - --~/7/

Figure 4 (a~t) Woundhealingin the presenceof a foreignbody, such as an implant.

longer time than in the normal incisional case. Within response is somewhat more severe and slightly pro-
four to eight weeks, however, the tissue response longed (Fig. 6a); the fibrous capsule that forms is
should have stabilized, leaving a zone of tissue that is slightly thicker and takes longer to stabilize.
rather similar to the normal scar tissue, with perhaps Secondly, the reaction may be more extensive again
differing patterns of vasculature and collagen fibres such that the acute response is severe and progresses
running parellel to the implant surface. This sequence to a significant chronic response. The repair process
is depicted in Fig. 4. may be initiated at an early stage but requires more
This response is, however, rarely seen. It is described time to deal with the extensively damaged tissue; it
as the classical fibrous encapsulation of implants, but results in a fibrous capsule that is different in size and
this term is too simplistic. Absolute inertness of a characteristics (Fig. 6b) with variations in cell popu-
biomaterial, as already described, is rare and other lation, blood capillary density, tissue destruction and
material and device characteristics can be super- so on. It may be that the capsule takes several months
imposed on inertness criteria to modify this response. to settle down.
Of the currently used biomaterials, pure titanium [89], Thirdly, the acute response may be minimal
high purity alumina [90] and some special grades of such that the repair process is quickly effected but
polymers such as polytetrafluoroethylene [91], ultra- then long-term interactions (corrosion, degradation,
high molecular weight high density polyethylene [92] abrasion) take place such that the implant becomes a
and silicone rubber [93] may elicit this minimal fibrous persistent stimulus to the tissue and chronic inflam-
encapsulation under some conditions, with examples mation ensues. Cells typical of both acute and chronic
being shown in Fig. 5. inflammation may be seen, often in association with
reaction products derived from the implant which they
4.6.2. Deviations from the minimal response are trying to eliminate (for example by phagocytosis).
There are many ways in which the minimal fibrous Fibroblasts will be active, attempting to make good
response may be modified by the presence of an the damage, but they may be fighting a losing battle if
implant. the implant provides a continuum of irritants. The
First, the initial reaction between the implant and long-term result is a granuloma (Fig. 6c), especially
the tissue may be more pronounced than that seen with numerous foreign body giant cells, which may
with a "totally inert" material such that the acute visibly be associated with oedema (swelling) and in the

3428
Figure 5 Stages in the developmentof a fibrous capsule: (a) after 4 weeks, with implant of titanium (normal muscle to the left, implant to
the right, relativelyacellular capsule developing between implant and muscle); (b) more cellular response at 9 weeks to nickel-titanium
implant; (c) thin capsuledevelopedwith titanium at 10 weeks; (d) thin capsule developedwith thermoplasticmaterial, polyetheretherketone,
at 10 weeks, x 200.

clinical sense will result in pain. If the reaction is broad headings. First, and probably, foremost, is the
sufficiently severe, cells, and indeed the tissue, will die influence of implant chemistry. Secondly, there is the
(become necrotic), a process which further aggravates role of physical factors, including size and shape of the
the situation. implanted device, its surface morphology and texture
This discussion implies that there is a progression of and its mechanical relationship with the tissues.
tissue responses, from the minimal to the necrotic, Thirdly, there is the influence of biological variations
which are seen in the context of implanted biomaterials such as implant size, host species, age and sex, the
and indeed, we shall see below some examples of how state of health of the host and pharmacological status.
different materials in different forms initiate these
various responses. It should not be assumed at this 4.6.3. The effect of surface chemistry
stage, however, that the ideal biomaterial in every There are two ways in which the implant surface
clinical situation is that which produces the minimal chemistry can influence the tissue response. First, by
response described above and that all the others are the processes we have already discussed in Section 2,
getting progressively less desirable as far as clinical the nature of the surface chemistry, by virtue of the
devices are concerned. It may be said that the pro- properties such as surface energy, dielectric constant
longed chronic response with granuloma, or worse, is or equilibrium potential that are dependent on this
universally undesirable. On the other hand, minimal chemistry, will influence the initial events at the
fibrosis, which effectively means that the implant is implan~tissue surface; that is, the characteristics of
being ignored by the tissue, is not necessarily a good protein adsorption will depend on the chemical nature
thing because it does not lend itself to full incorpor- of the surface. If there were no subsequent chemical
ation and acceptance of the biomateriat into the tissue. interaction between biomaterial and tissue, the extent
We shall see later how a modified tissue response may to which the wound healing process would be modified
be beneficial in terms of the full integration of the by the implant would be entirely dependent (for con-
implant into the body. stant physical and biological conditions) on the nature
Since there are so many mediators of these tissue of this adsorbed layer. However, as we have seen in
responses, it is not surprising that numerous factors Section 3, the prospect of zero chemical interaction
may be involved in their modification in the presence within the human body is remote and thus the second
of an implant. We may consider these under three way in which the chemistry influences the tissue
3429
 Figure 6 More severe responses at implants (com-
pared to Fig. 4d): (a) more extensive capsule;
(b) more extensive and cellular capsule, with per-
sistent macrophagic response alongside fibroblastic
response; (c) extensive chromic inflammation
(including foreign body giant cells and epithelioid
cells) with granulation tissue.
, \\ L~. h Lw,-===-,-~ .~ I ~_I
I ~illlJ!lll l~ ifl~"'/l /
\ //I,,,,,,,,J l,:,',
'"'"
iiiii I

Capillari~...

I /
Macrophages ri11,_,!l,I V / /
I-~]ii,L___~rh"tijI I g"
I ',I'I !l •~g~="=-:--~J
t iL"=-=-=~---.W~ v i J _ ~ . l
~/ t
/
~a.~- ---.'--~'--~-~-~';5!'~'/( Fi b r ob la st s
" -"- =_--2-2 ~Z=-~" l~j

1') - _&o
~~-~" -- llagen
¢ei is~..~L~_ "-
Epithelioid .., 7/o-~=,.~O__~ i
~~Macrophages
_ _ Z O ~ ~"
O f---- --lOl[~i I ~ [Fibroblasts
" "" el l°l ~',](~I1".'I"1./
Foreign L
cells / 1; k \" ~ ~J~h ~:;/;mSim°:tt°Fn
;n::ua/iacent

response is through the nature (amount, chemical A far greater amount of information is available con-
nature, physical form, solubility, toxicity, etc.) of cerning the role of the interaction between materials
the reaction products. Naturally, there are many and tissues on this response, although the complexity
mechanisms by which these reaction products can and multiplicity of events still makes a rational and
exert this influence. universal understanding of the phenomenon rather
In spite of its obvious importance, very little is difficult.
known about the extent of the effect of adsorbed Let us first of all consider the morphology of
proteins on the soft-tissue response. In a later section the tissue response and use metals as an example.
the vast amount of information on the effect on inter- McNamara and Williams [98-100] have reported
actions with blood will be reviewed but, apart from a studies in which the local tissue response induced by
few speculative papers [94, 95] the mediation of con- discs of some ultra-pure metals (5 mm diameter, 2 mm
nective tissue reactions by this layer remains unclear. thick) implanted intramuscularly in rats, was examined.
It is certainly known that in many in vivo conditions, Pure cobalt, nickel, lead, aluminium, copper and
the very important phenomenon of cell adhesion to titanium were studied. A morphologically distinct
foreign surfaces is controlled by the nature of the pattern of response developed for each metal, although
intervening protein layer [96, 97] and this must be all within a general framework. In the intramuscular
assumed to be important in vivo. On the other hand, site a reaction zone, analogous to the fibrous capsule
the adsorbed protein layer in vivo will not remain described in the minimal response situation, develops
unchanged for long and it is difficult to see how a layer and separates the implant from the normal surround-
that is probably constantly undergoing metallic turn- ing muscle. This capsule, which obviously matures
over maintains a long-term influence on the tissue with time, varies in thickness, in its organization and
response. its relationship to the adjacent tissue. In general, as

3430
 Figure 7 Generalized fibrous capsule
forming in response to intramuscularly
implanted metal. Some or all of these
features may be present depending on the
metal.

:o/;il;I,'.;I II
- $I
I ,l - I

Area of cellular of
L-~osely wov il~ang~n inf~rngt i iot e o~i t y necrosis

shown in Fig. 7, the zone may contain an area of copper was the very high degree of vascularity, which
necrosis adjacent to the implant, surrounded by a may correlate with the speed of systemic distribution
region of chronic cellular infiltration. Often this of the metal ions. A sterile exudate was normally
is adjacent to a band of densely packed oriented present adjacent to the implant. Haemosiderin-laden
collagen, which itself will be surrounded by a more macrophages were always in evidence, as well as cells
loosely packed collagen zone; which may contain containing black pigmented material (Fig. 8b). Large
blood vessels, vacuoles, fatty tissue, muscle fragments areas of yellow-brown pigmented tissue were also
and discrete cellular populations. In some cases the seen, a feature unique to copper. With time, progres-
capsule has a well-defined boundary but in other cases sive vacuolization and resorption of the muscle fibres
it extends irregularly and diffusely into the surround- occurs within the affected parts.
ing muscle. The reaction to the cobalt was most interesting.
With aluminium, for example, there was a distinct There was no clear capsular edge but rather a gradual
necrotic region adjacent to the implant, but the change in appearance of the tissue. The most notice-
capsule was very compact and only a short distance able feature was the presence of significant areas of
away the muscle was healthy (Fig. 8a). There were a lymphoid tissue (Fig. 8c), consisting principally of
few layers of collagen but these were not extensive. plasma cells, which indicate some form of immuno-
There was little neutrophil infiltration after the acute logical activity. It is likely that proteins, not in them-
stage, but many macrophages were present, actively selves antigenic, are rendered "foreign" by contact
secreting hydrolytic enzymes. The capsule was poorly with the metal. Cobalt may corrode fairly freely in the
vascularized. tissues, such that corrosion products intermingle with
The response to lead, in spite of its known toxicity the cells in the reaction zone (Fig. 8d).
in other situations, was much less severe. Some fibro- There are probably many factors involved in the
blastic proliferation was seen and a dense collagenous development of this range of responses. The rate of the
layer developed relatively quickly. However, cellular interfacial reaction inevitably is important, and the
infiltration was minimal and no necrosis was observed. extent and longevity of the chronic responses to
This comparison between aluminium and lead is copper and nickel are in part associated with these
interesting and important. Aluminium is normally kinetics. Reaction products can take many forms. In
thought to be relatively non-toxic [101] but such a the case of passivated metals, discrete and discernible
conclusion has been derived from classical toxicologi- corrosion products may not be observed, but rather
cal studies where the test substance is administered the process involves the diffusion of metal ions
to animals by the oral route. Since aluminium is so through an oxide film, these ions immediately being
poorly absorbed within the gastrointestinal tract, very bound to some organic or inorganic species in the
little gets into the bloodstream and hence it appears surrounding milieu. Some such products could pre-
non-toxic. If, on the other hand, the metal is presented cipitate locally, whilst others are easily transported
directly to the tissues (and by implication, the blood) away in vascular or lymphatic systems. With titanium,
it is able to exert strong biological activity, which is for example, although the corrosion rate is almost
seen here resulting in tissue and cell death. In recent immeasurably small, titanium complexes are pre-
years the complacency over the safety of aluminium cipitated locally and give a distinct discoloration to
has also been brought into question with several the tissue (Fig. 8e) [103]. This question of the binding
problems of toxicity, leading to a variety of symptoms of metal ions to tissue components has been addressed
and occasional mortality in patients undergoing in a number of recent papers [104-6].
kidney dialysis, due to the traces of aluminium in the Bearing in mind that some of the metals involved
water that is used and which gains direct access to the are essential trace elements in animal tissues and are
bloodstream [102]. naturally present at levels often in the p.p.b, or p.p.m.
The most noticeable feature of the response to range [107], whilst others have no physiological func-
3431
 Figure 8 Some examples of the response
to metals: (a) necrosis seen adjacent to
aluminium x 100, (b) pigmented tissue
around copper implant x I00; (c) plasma
cells and lymphocytes in response to cobalt
x 200; (d) cells, corrosion products and
proteins on metal surface SEM x 3000; (e)
discoloration of tissue adjacent to titanium
x 150.

tion and cannot normally be detected therein, a multi- a m o u n t s o f metals such as v a n a d i u m can radically
plicity o f mechanisms exist whereby these metals alter the kinetics o f enzyme activity associated with
derived from implants can influence biochemical and cells o f the inflammatory response. These metals
physiological reactions. The metals may, for example, m a y also influence the chemotactic mechanisms that
compete with the n o r m a l cations present for the bind- are involved in the attraction o f cells to the area;
ing sites on proteins. If these proteins are enzymes aluminium is found to be strongly positively chemo-
then these catalysts m a y readily be inactivated, a tactic whilst cobalt is negatively chemotactic (repels
classical mechanism for toxicity [108]. Williams and cells), possibly explaining the n a r r o w but densely
Crowley [109] have recently shown that very small packed reaction zone to aluminium c o m p a r e d

3432
 Figure 8 Continued.

with the wide invasive capsule around cobalt [98]. If fibrous capsule, examples being PTFE [91], polyethy-
discrete particles are derived from the corrosion lene [92], silicone rubber [93], polymethylmethacrylate
process, as indeed is seen in clinical practice in the case [118] and polyetherurethane [119]. Few systematic
of stainless steel [110, 111], then the tissue response studies of the influence of these surface variables on
may be modified by the physical effects as well as the response have been reported, however.
chemical effects, as explained in a later section. Some years ago Gilding et al. [120] presented a
When working with pure metals, it is possible to series of abstracts on work of this type, describing the
interpret the biological response in terms of single tissue response to a series of polymers with controlled
cation species only. When working with alloy systems, hydrophilicity and surface charge, but the full results
as indeed is the practice in the surgical application appear not to have been published. Some other studies
of these materials, the interpretation is much more have involved far fewer polymers or variables, or have
difficult. References of the soft-tissue response to used quantifiable in vitro techniques rather than
many different metallic systems have been published the more descriptive in vivo methods. For example,
including stainless steel [112], titanium and its alloys Lentz et al. [121] have studied macrophage adhesion
[52, 113], cobalt-chromium alloys [114], nickel-based to hydroxyethylmethacrylate-ethylmethacrylate co-
alloys [115], noble and platinum group metals [116] polymers and hydroxystyrene-styrene copolymers.
and dental amalgam [117]. The data showed that there was a time delay between
In the case of polymers, a somewhat different story the contact and adhesion of cells to surfaces that
emerges for several reasons. On the one hand it is varied with hydrophilicity within the former group, but
far more practical to prepare polymers of minimal not in the latter. The results were explained in terms of
reactivity, as we have seen. At the same time, however, exclusion volumes relating to the swelling of these
the chemical and physicochemical features of the sur- hydrophilic polymers.
faces can be varied more extensively with ranges of The general scheme of the tissue response to
hydrophilicity/hydrophobicity, crystallinity, surface implanted materials as outlined above has been
charge, reactive groups and so on. As indicated above, discussed in detail with respect to some polymers
there are several polymers which can be prepared with by Anderson and co-workers [122, 123], who have
sufficient inertness to allow the formation of the thin designed an experimental model for the quantitative
3433
study of certain parameters of this interaction. A we can do no more than bear them in mind at this
stainless steel cage containing the polymer of interest time.
is implanted subcutaneously in rats and exudate that
forms within the cage as part of the inflammatory 4.7. The response of hard tissues
response is aspirated and analysed for protein content, to implantation
enzyme activity and white cell content. The surfaces of Bone responds to the presence of implants in several
the polymers are also examined for cell attachment. In different ways, the nature of which depend on the
an extensive study with the polyetherurethane Blamer ® type of bone, the previous history of the bone, the
the total white cell count in the exudate decreased with geometrical and morphological relationships between
time (4, 7 and 21 days), as would be expected from the the bone and implant, the method of attachment
resolution of the acute inflammatory response, while of one to the other, the mechanical stress system
the differential cell count showed a decrease in the acting on the implant-bone system and the material
proportion of neutrophils and an increase in macro- chemistry.
phages and lymphocytes, again as expected. Those Let us consider the two most important aspects
cells attached to the polymer surface were largely of the bone-implant interaction, the effect of the
macrophages, with a slow increase in the number of material on the response of bone to damage or to loss
foreign body giant cells. of tissue, and the influence of an implant on the stress
system within the bone.
4.6.4. The effect of surface topography
Not all implants, either experimental or clinical, have 4.7. 1. Biomaterials and bone growth or
smooth surfaces and attention must be given to the remade~ling
influence of surface topography on the development Bone will normally come into contact with a bio-
of the tissue response. Leaving aside the separate material in one of three ways. First there is the situ-
question of tissue ingrowth into porous materials, it is ation where a bone fractures and an implant is used,
now known that minor variations in surface texture as a plate, nail or other device, to hold fragments
can influence this response [124-126]. Perhaps the together while bone healing takes place. In this situ-
most interesting observations are those of Gibbons ation, the healing process is controlled by the mech-
and co-workers [127, 128], who have studied the tissue anics of fixation and the material has little direct
response to polymers given different textures by ion- effect on the bone since it is not normally placed
beam milling. Using P T F E and a texture containing within the healing zone. Secondly there may be some
conical projections of height 12#m and base width defect within a bone which requires repair. This could
4/~m, they have been able to show increased cell be a bacterially-derived resorptive process, an ageing
adhesion and increased lysosomal enzyme activity in process, the result of trauma or surgical intervention,
the cells of the response. The macrophages in the and here the implant is used to facilitate the healing of
response, compared to that for smooth surfaces, had bone within this defect. In this case the nature of the
differing structural characteristics, especially increased material is vitally important. Thirdly the implant may
cytoplasmic-to-nuclear ratios and vacuolization. The be inserted into a bone for the purpose of recon-
rate at which the capsule developed also changed: at struction in adjacent tissues, the bone itself acting as
8 weeks, capsules associated with textured surfaces a secure fixation point. Examples here include the
were some 30% smaller than those around smooth- intramedullary fixation of joint prostheses, bone
surfaced implants, possibly because of reduced fibre- screws, and the location of dental implants in mandible
blast proliferation. Further experiments showed that or maxilla. Here again the material is an important
protein adsorption also differed between smooth and factor in determining the response of the bone to the
textured surfaces. implant.
In addition to the surface topography, the actual We may consider these latter two together and
shape of the implant may influence the response. Little Fig. 9 indicates the general scheme. In Fig. 9a we see
is known of this phenomenon, although Matlaga et al. the progression of healing in a bony defect without the
[129] have demonstrated clear differences in the aid of any biomaterial, where bridging the defect
fibrous capsule that develops in response to various can occur spontaneously and completely. Initially an
cross-sectional shapes of polymeric implants. exudate will form within the defect (for example a
blood clot), which will slowly reorganize. In the
4.6.5. The effect of physiological variables process of osteogenesis, new bone may grow, either
The general scheme described above for the develop- directly from the existing bony walls (osteoconduction)
ment of the tissue response assumes a uniform and or from isolated areas within the reorganizing tissue
consistent host site. In addition, most of the infor- should the appropriate bone cells (osteoblasts) and
mation on the tissue-biomaterial interaction has been growth factors be available (osteoinduction). Ulti-
derived from studies on young healthy animals. For us mately new solid bone may form, although there will
to be able to extrapolate from these animal studies to inevitably be intervening periods where there are both
the human clinical situation, we should ideally take areas of bone and of unmineralized soft connective
into account variations of the tissue response that tissue present. There is an upper limit to the size of
arise with differences of species, sex, age, state of defect which can be bridged in this way, above which
health, pharmacological status and other factors. the defect will merely fill with unmineralized tissue.
Unfortunately, very little is known of these effects and If now a solid object is placed in the bony cavity

3434
(Q) Defect (b) Oefect
/

Osteoblast proliferation _ BIoodj clot Osteoblasts Blood clot

:/I-" .~. Implant-- ,!~
- ' lq)" " I
:o : 0
v v v ~__

New bone f o r m e d by o s t e o i n d u c t i o n .Osteoblasts Complete bone r e g e n e r a t i o n

I- I
_ _ _

N e w bone formed by o s t e o c o n d u c t l o n

Soft tissue i n t e r f a c e New bone

.(. ]--
__~.

Complete bone regeneration

(c) Defect Figure 9 Scheme of healing in bone: (a) sequence in absence of any

/ implant, leading to complete bone regeneration; (b) response in

presence of monolithic implant, leading either to complete bone
regeneration or a soft tissue interface; (c) response in the presence
of particulate implant, leading either to complete bone regeneration
or a zone of soft fibrous tissue around the particles.

Blood clot Particles of biomaterial

defect will completely fill with bone such that there
will be intimate bone-implant contact. This is a
matter of considerable current controversy, for it
_ _

appears that most materials will not allow such con-

tact but instead a soft fibrous tissue interface will exist
Osteoblasts
between bone and implant. This, of course, has con-
siderable functional consequences if the objective is
I for the implant to remain secure in the bone. The
i 0 ~ - - 0 r
evidence for and against direct bone contact will be
I l - - reviewed below.
If, instead of placing a monolithic solid within the
Complete bone r e g e n e r a t i o n bone, some particulate matter is used (Fig. 9c), then a
combination of the two processes seen in Figs 9a
Zone of fibrous tissue and b may take place. Several permutations are, in
fact, possible; the particles may actively encourage
osteogenesis such that the defect fills faster, or larger
defects will fill with bone rather than unmineralized
\
tissue. Alternatively the particles may have a neutral
New " ~ grow th
oone effect such that bone growth occurs around them,
while it is also possible for the particles to prevent the
total conversion to new bone and instead promote the
(Fig. 9b) the healing will take place in the space formation of fibrous tissue envelopes around each of
between the implant and the bone. If this gap is small them. Again, the chemical nature of the material has
there will be a tendency for new bone to form, but this been assumed critical in determining which event will
process can be modified by mechanical and chemical occur.
influences. Again there will be a time when this space
is filled, partly with new bone and partly with fibrous 4. 7.2. The biomaterial-bone interface
unmineralized tissue. Under some circumstances the As noted above, the question of whether any bio-
3435
material can allow the formation of new bone up to rounded by proteoglycan layers of up to 2000 and
and contiguous with its surface is highly controversial. 300nm, respectively. Since a layer of proteoglycans
Several terms have been used in this context to up to 20 nm thick may be seen between individual
describe the performance of materials and it is import- collagen filaments or at cell-collagen interfaces, it has
ant that the ideas behind these are understood, even if been argued that titanium, and no other biomaterial,
their definitions have not been widely accepted. First is behaving as a natural tissue. It is not clear why
we have to recognize that there are materials which are titanium should behave uniquely in this way. It is
indifferent as far as bone is concerned; they may be claimed that titanium is binding by irreversible elec-
essentially inert in the body but have no positive or trostatic forces to the bone and that the physical
negative effects on bone growth. These are sometimes characteristics of the titanium oxide surface are
referred to as bioinert, although the mention of that important but such hypotheses have yet to be proved.
term here does not imply endorsement of its use. There is no suggestion that new bone growth occurs
Secondly, there are materials which appear to actually spontaneously on the surface of these materials in the
encourage the formation of new bone at their surface; absence of growth from adjacent, pre-existing bone. A
that is, osteogenesis takes place right at the implant- few materials have, however, been described as bio-
tissue interface without necessarily any bridging active, that is, having this property of inducing osteo-
to pre-existing bone. Such materials are sometimes genesis in their vicinity. Although many such claims
described as bioactive. Thirdly, there are materials have been made, we shall refer to three such situ-
which allow sustained growth from the surrounding ations, involving calcium hydroxylapatite ceramics,
bone up to their surfaces with direct bone-material certain glasses or glass-ceramics of controlled surface
contact. Such a process has been described as osseo- activity, and certain porous surfaces.
integration. Calcium hydroxylapatite is an interesting bio-
In spite of many attempts to do so, it has not yet material that owes its use to the fact that it is the
proved possible to define the conditions under which synthetic analogue to the mineral phase of bone itself.
this bone-material contact will occur. Much early It has been prepared, under a number of trade names
comment was related to the use of orthopaedic devices and in various physical forms, as a bone reconstruc-
such as bone screws and intramedullary stems [130], tive material and, along with its near neighbours in
but the complex interrelationship between mechanical the calcium phosphate group of ceramic structures, it
stresses, surgical technique and implant material made has been the subject of a number of recent reviews
sensible interpretation of the data very difficult. In [134-136]. Of special interest here is the observation
some cases bone could be seen forming right up to that when certain samples of calcium hydroxylapatite
(say) screws inserted into cortical bone, whilst in other are placed within a bony environment, they may
cases, the bone would actually resorb away. More become well incorporated into the bone, without
recently attention has been directed towards the role intervening inflammatory or fibrous tissue elements.
of the material and especially on the claim that one Whilst this is not a universally recognized phenom-
metal in particular, titanium, is highly conducive to enon, and the precise conditions under which this can
this direct bone-implant contact. Most of this work occur are not understood, there is no doubt that new
has been reported by Branemark, Albrektsson and bone can intimately merge with the ceramic, with
co-workers [131-133] and is reviewed briefly. individual crystals of biological apatites epitaxially
Titanium implants placed within bone have been deposited on the surface of the synthetic material
shown to become surrounded by bone tissue without [1371.
interposed fibrous structure and without chronic Glasses or glass-ceramics of controlled surface
inflammatory reactions. Although the structure of the activity appear to provide a different mechanism.
bone distant from the implants was normal, the These glasses, first developed by Hench et al. [138],
arrangement of bone lamellae changed near the surface are typically based on the S i O 2 - C a O - N a 2 0 phase
such that they became oriented parallel to the implant diagram and contain a small quantity of P~O5. The
surface. At a distance of 0.5/~m from the implant significance is that, within a narrow compositional
surface, as reviewed by transmission electron micro- range, these glasses have a reactive surface that
scopy, the collagen bundle arrangement gave way to initially leaches out calcium and phosphate ions.
randomly arranged collagen filaments, which could be These are able to promote the formation of new
observed down to 20 nm from the surface. There was bone at their surface, since once again this calcium
a partly calcified amorphous ground substance, con- phosphate is chemically equivalent to the mineral
sisting of proteoglycans and glycosaminoglycans phase of bone. This surface reaction is self-limiting
within the 20 nm layer at the surface. Although in because of the protective effect of the silica-rich
places the degree of calcification was less than normal, surface layer that is left after the preliminary leaching,
bone was observed up to the limit of 3 nm detection. but the transition zone is of such a character that it
Cellular processes (from osteoblasts) approached the is able to incorporate the molecules of the organic
surface but were separated from them by a 20 to 30 nm matrix of new bone as it is forming, hence providing
thick proteoglycan layer. for a chemical bond between the glass and the new
Other metals were evaluated using the same tech- bone [67, 139, 140].
nique and were found to have proteoglycan layers of It has been known for some time that if a porous-
50 to 500 nm. Other materials, such as polymethyl- surfaced material is placed within tissues, tissues
methacrylate or certain glass-ceramics, were sur- can grow into the pores and provide for some

3436
 Figure 10 Tissue ingrowth into porous
polyethylene, x lO0.

attachment. Generally, and as indicated previously, considerable clinical importance and is likely to have
porous materials placed within soft connective tissue an influence on the response to any material within
will act as hosts for ingrowth of soft tissue, although bony tissue.
there have been occasions when they will initiate the
spontaneous formation of bone under unusual circum- 4.8. The r e s p o n s e of b l o o d to b i o m a t e r i a l s
stances [141]. Placed adjacent to existing bone, porous Although the history of using biomaterials and
surfaces may promote the ingrowth of new bone. This devices within the cardiovascular system does not
process is largely independent of chemistry, but is, stretch as far back as that involved with some soft and
instead, controlled by the morphology of the pores. If hard tissue applications, it is probably by now the
the porosity is interconnecting, and the minimum pore most extensive in terms of the allocation of resources
size is in the region of 150/~m, then bone can grow into to tackle the very significant problems inherent within
the surface area and promote attachment [142-145] this system. The first serious attempts to replace blood
(Fig. 10). There is no inference that the materials are vessels can be traced to the mid 1940s, but most
acting in any osteogenic manner here, but rather the significant developments did not take place until
open porosity allows for a conduction of new bone open-heart surgical techniques evolved during the
growth emanating from the surrounding bone. 1950s. Since that time numerous materials and designs
have been introduced into areas such as heart valve
4. 7.3. B i o m e c h a n i c a l compatibility replacement [151], circulatory assist devices [152], the
Although all tissues are subjected to mechanical total artificial heart [153], oxygenators [154], liver
stresses and each behaves in a different but always support systems [155], renal dialysis [156], blood vessel
complex way, the reaction of bone is perhaps the most replacement [157] and so on. With these devices the
significant. The mechanical properties of bone itself difficulties, and the constraints on progress, are largely
have been reviewed on many occasions (e.g. [146]), but related to the interactions between the blood and the
it is of most relevance to note here that the structure devices. Blood compatibility is, therefore, of the
and indeed viability of bone are determined by the utmost importance in determining the performance of
stress system to which it is subjected. The magnitude devices within the cardiovascular system and has to be
and nature of this stress system determines the degree considered somewhat separately from other aspects of
of mineralization and the extent of any porosity, and biocompatibility.
alterations in stress can, by a feedback mechanism It is possible to discuss this subject at a number
which influences cellular activity, produce changes in of levels, including specific reactions associated with
the bone structure, either of a growth or resorptive different materials and different devices. In this review
nature [147]. we shall only discuss, and briefly at that, the basic
The significance of this is that implanted devices principles of the interactions between blood and syn-
which are integrated into bone (e.g. by bone screws or thetic materials. This will be done under the headings
other fixation) will usually have quite different elastic of the essential characteristics of blood, the mech-
properties and will, therefore, totally alter the stress anisms of blood clotting, the effects of materials on
distribution [148]. In particular, since metals and clotting proteins, material-platelet interactions and
ceramics are of substantially higher elastic modulus, mechanical damage to blood. In addition to the refer-
their attachment to bone must inevitably result, under ences quoted in the sections that follow, the reader is
constant strain conditions, in reductions in stress referred to some general texts on blood compatibility
levels within the adjacent bone. This results in a [158-162].
remodelling which is most likely to be resorptive in
nature, giving a porous, less richly mineralized tissue 4.8. 1. Essential characteristics of blood
with grossly inferior mechanical properties [149, 150]. Blood is a suspension of cells in plasma, an aqueous
This problem of disuse atrophy or osteoporosis is of solution containing a variety of organic and inorganic

3437
T A B L E I The composition of h u m a n plasma (major com- ing state. The most noticeable effects of the activation
ponents only)
are adhesion of the cells to sites of blood vessel wall
Surface Concentration Molecular weight injury, their aggregation and the fusion of granules
(gdl -~ ) with the plasma membrane to facilitate the release of
Water 90 to 92 granular contents. This is particularly significant in
Proteins
the blood clotting process.
Serum albumin 3.3 to 4.0 69000 As indicated in Table I, the plasma is a very complex
Fibrinogen 0.34 to 0.43 340 000 mixture of proteins, anions and cations. The plasma
~-globulins 0.3l to 0.32 44000 to 200000 proteins [164] include those which provide nutrients to
e2-globulins 0.48 to 0.52 150000 to 300000 the cells, principally albumin and lipoproteins, those
/3-globulins 0.78 to 0.81 90000 to 1300000
which are involved in the transport of hormones and
y-globulins 0.66 to 0.74 160000 to 320000
other chemicals, such as transferrin (carrying iron),
Cations
ceruloplasmin (copper), vitamin-binding proteins and
Na + 0,31 to 0.34
K+ 0.016 to 0,021
steroid-binding proteins, and those which are involved
Ca 2+ 0.009 to 0.011 in defence, especially the immunoglobins, comple-
Mg 2+ 0.002 to 0.003 ment, and, of special relevance to this discussion, the
Anions proteins of the clotting process.
Chloride 0.36 to 0.39
Bicarbonate 0.20 to 0.24 4.8.2. Mechanisms of blood clotting
Phosphate 0.003 to 0.004 In normal healthy humans, blood flows through the
vascular system and does not clot. Blood does, how-
ever, have the ability to form clots when a blood vessel
molecules, the major components being shown • is injured, so that bleeding can be arrested. The mech-
Table I. Thare are three groups of blood cell, as in ~. anism by which this haemostasis takes place is very
cated in Table II. complex and there are many opportunities for inter-
The erythrocytes or red cells are clearly the most ference. Thus, the process may be initiated in patients
numerous. They have a lifespan of 105 to 120 days, when physiological parameters change for reasons
becoming more fragile as they age. The main effect of other than the arrest of bleeding, or when the vascular
blood-material interactions in relation to the red cell system is interrupted by some foreign surface.
is that of accelerated ageing, or premature mechanical Haemostasis is achieved by the formation of a mass
destruction. This is discussed in Section 4.8.5. of platelets and fibrin that is deposited in such a
There are far fewer white blood cells than red cells way that it is impervious to the flow of blood. There
per unit volume. Unlike red cells or platelets, these are two crucial events in this process, involving first
are not confined to blood and, as we have seen in the cellular components of blood and secondly the
Section 4.2, are released into the tissue where they plasma proteins. Injury to a vessel that involves dis-
carry out certain specific functions. These ceils are not ruption of the endothelium, initiates a sequence of
of great significance in blood compatibility. events which allow platelets, normally non-adherent to
Platelets, on the other hand, are of enormous signi- endothelium, to adhere to the damaged surface. The
ficance. When resting, they are small discoid cells first stage is the platelet-collagen interaction, in which
of diameter 2 to 3 #m, with a very complex cell contact with collagen causes the platelet membrane to
membrane [163] which has numerous receptors for undergo a series of changes. This stimulated platelet
interaction with key proteins contained in the plasma. releases a number of substances, principally adenosine
The interior of the platelet consists of a number of diphosphate (ADP), which causes the platelets to
granules that contain a variety of proteins which con- aggregate and, ultimately, form a platelet mass [165].
trol the ability of the platelets to aggregate and The other crucial event is the activation of the
interact with other structures. sequence of events known as the coagulation cascade
Under certain conditions, platelets are activated, [166] in the plasma proteins, that leads to the forma-
resulting in significant functional, biochemical and tion of a thrombus. The blood coagulation proteins
structural alterations to the cell, compared to its rest- are a series of enzymes that function sequentially,

TABLE II The cells of circulating blood

Cell Concentration Normal shape Volume

(Number per m m 3) (%)
in blood

Erythrocytes 4 to 6 x 106 Biconcave disc, 8 # m x 1 to 3 # m 45

Leukocytes
Neutrophils 1.5 to 7.5 x 103 Spherical, 7 to 22 # m diameter 1
Eosinophils 0 to 4 x 102
Basophils 0 to 2 x 102
Lymphocytes 1 to 4.5 x 103
Monocytes 0 to 8 x 102

Platelets 250 to 500 × 103 R o u n d e d or oval, 2 to 4 # m

3438
 Vascular injury tion is not clear, but a considerable amount of evi-
"--"''-----~. ael=ase of dence has been obtained. We may first of all refer
¼ Exposure of Collagen ~ tissue factor
Platelet XlI--~-Zila,= I briefly to the proposed mechanisms by which surfaces
adhesion XI--~/~XIa may influence the factors of the intrinsic pathway, and
then briefly describe the known interactions between
L
Platelet
aggregation"F--ADP
IX ~"~Z
Ca+xa
VIII---~gIII*---~VIIIa
Caz+ VII
surfaces and plasma proteins in general.
The available evidence suggests that it is negativelY
Phos+ I I charged surfaces which possess the ability to initiate
t
Platelet X
coz%
"-~Xa~-z--X coagulation, and that these surfaces perform three
release ~Phos vital functions. First, they introduce a structural
V~ H Ca2+
reaction change in Factor XII such that the surface-bound
a II
Factor XII becomes susceptible to proteolytic activa-
Platelet
plug /
formation -
I I
Fib~rin
formation tion. Secondly, the surface promotes an interaction
between Factor XII and the inactive molecule pre-
Consolidation of kallikrein which results in the reciprocal activation of
platelet plug
each. Thirdly, the surface promotes the activation of
Haemostasis Factor XI by surface-bound Factor XIIa.
Figure 11 The coagulationcascadein relation to blood clotting. Although a great deal is now known about the
biochemical changes which are taking place here, little
is known of the way in which different biomaterials
with the final event being the polymerization of the are able to influence these events. On the other hand,
fibrinogen monomer, brought about by the action of much experimental data has been accumulated con-
thrombin, to form a cross-linked biological macro- cerning the ability of foreign surfaces to adsorb and
molecule, fibrin. The fibrin strands begin to reinforce interact with plasma proteins in general [170-173].
the primary platelet plug, consolidating it into an Most available information concerns the adsorption
impermeable mass. Several substances released during of single isolated proteins on to polymer surfaces in the
degranulation are involved in this sequence. The fibrin absence of the competition that would arise in plasma.
is important in the platelet plug since it stabilizes the It is generally believed that the Langmuir model of
platelets irreversibly; without the fibrin, the platelets adsorption applies and that surfaces which adsorb
would soon disintegrate. albumin are generally thromboresistant, while those
The coagulation cascade may be initiated by one of which adsorb fibrinogen, gamma globulin and fibro-
two mechanisms [167]. The first of these involves the nectin tend to be more thrombogenic [174, 175].
presence of the so-called tissue factor, a glycoprotein
associated with phospholipid, and the activation of the 4.8.4. Material-platelet interactions
extrinsic pathway. The second involves the exposure In spite of the obvious importance of material plat~lm
of a non-endothelial surface, such as collagen, and the interactions in the development of thrombogenesis,
activation of the intrinsic pathway. In either case, and the vast amount of information about the adhesion
there is a localized conversion of inactive molecules to of platelets to foreign surfaces, the surface charac-
proteolytic enzymes, in a sequential pattern that culmi- teristics which are responsible for the attraction of
nates in the conversion of prothrombin to thrombin, platelets have not really been identified. Several
which catalyses the polymerization of fibrinogen to theories have been proposed but the literature shows
fibrin. As shown in Fig. 11, the two pathways come that it is quite possible to argue that opposing charac-
together with the conversion of Factor X to Xa and teristics are equally important. Thus, it has been shown
thereafter they follow a common pathway to the on the one hand that platelet adhesion decreases
formation of fibrin. with decreasing interfacial free energy [176], while on
Initiation of the extrinsic pathway is a relatively the other hand that it decreases with increasing inter-
straightforward matter [168]. Tissue factor is an facial free energy [177], the role of adsorbed proteins
enzymatically inert glycoprotein that is present on the apparently being crucial here.
surface of many cells, but not in plasma proteins. It is This subject has been reviewed at great length
released when tissue is damaged, interacts with Factor recently by Anderson and Kottke-Marchant [178]
VII, and together they catalyse the conversion of from which it is concluded that no one theory is able
Factor X. The intrinsic pathway is more complex and to explain the interaction of platelets with foreign
more relevant to the use of biomaterials. The enzyme surfaces. One point of interest, however, is that
central to this intrinsic, surface-initiated coagulation since natural blood Vessels do not themselves cause
pathway is Factor XII, the Hageman factor. This is platelet ,adhesion until' damaged, there must be sub-
activated by the presence of one or more co-factors and stances within the vessel walls that repel the platelets.
some non-endothelial surfaces, including collagen and Several such substances have been isolated, includ-
the synthetic surfaces of biomaterials [169]. ing prostacyclin, one of the prostaglandins. One
of the many attempts to prepare non-thrombogenic
4.8.3. The effect of materials on the clotting materials has involved the attachment of prostacyclin
proteins and other plasma proteins or prostacyclin-like substances to the surface and this
The precise nature of the mechanism by which foreign has indeed shown some success in respect of anti-
surfaces (e.g. those of biomaterials) initiate coagula- platelet activity [179].

3439
4.8.5. Mechanical damage to blood materials, the development of allergies to biomaterials
As noted above, red cells, which are responsible for (the immunological response), and the influence of
carrying the haemoglobin in blood, have a finite life infection.
span. Their eventual removal from the system involves
a disintegration of the cell membrane; new cells are 5.1. Tumour induction by biomaterials
continually produced to offset this controlled removal. Tumours are the result of excessive and uncontrolled
The red cell membrane is, in fact, a highly deformable proliferations of cells, which may be benign, in which
material, since it needs to change its discoid shape as case they are confined to their site of origin, or
it passes through narrow vessels, especially in the malignant, which grow locally by infiltration and
spleen, and it is quite susceptible to damage induced expansion and have the ability to spread via body
by shear stresses [180]. fluids. Tumours occur at a wide variety of anatomical
The interposition of hard or rigid materials in the sites and show a wide variation in type. Although for
vascular system usually perturbs the haemodynamic many tumours the cause is not known, for others, the
regime and high shear stresses are easily generated. It agents responsible may be identified. Such agents
is not well appreciated that the interaction of flowing include radiation, viruses, hormones and exogenous
blood with a variety of devices leads to premature chemical substances.
ageing of the cells and, therefore the development of It has been known for some time that implantable
a haemolytic anaemia [181-183]. The nature of the biomaterials may cause tumours under certain circum-
material is not particularly important, but rather the stances [192]. It is logical to assume that the chemical
design and location of the device. nature of the biomaterial may be important in this
context (i.e. the material may act as a chemical
4.8.6. Blood-compatible materials carcinogen) but it is also clear that other factors, such
It is clear that the future of biomaterials within the as physical factors, may also be important.
cardiovascular system depends on the understanding The experimental data on tumour induction is
of their interactions with the blood, especially those derived mainly from laboratory animals. As reported
leading to the resistance to clot formation. For by Pedley et al. [192] very many materials, including
many years attempts have been made to define the examples of polymers, metals, ceramics and minerals
ideal thromboresistant surface in terms of physico- have been shown to be carcinogenic under some con-
chemical parameters. Some theories have, for example, ditions in experimental animals, especially rodents. It
suggested that negatively charged surfaces are ideal is known that a latent period for induction of the
[184], others have. defined surface energy parameters tumours is generally applicable in these animals (that
[185] and yet others have suggested that water- is there is a time lapse between implantation and the
containing surfaces of hydrogels are most appropriate occurrence of the tumour, measured in months in the
[186]. For many years, however, the choice of the best rat, for example) and that considerable variations are
blood-compatible surfaces has been dominated by the seen with different species of animal.
perceived need for inertness so that carbon, titanium, The experiments also indicate that both physical
PTFE, silicone rubber and similar materials have been and chemical variables influence the extent of tumour
extensively used. induction. With polymers which are relatively inert
More recently, serious attempts have been made to chemically, it is likely that chemical carcinogenesis is
tackle this problem from a more logical biological playing a significant part and the most important
view. As well as the anti-platelet substances such as variable appears to be the size of the implant. Large
described above, there have been attempts to coat monolithic objects tend to be the most effective agents,
polymer surfaces with phospholipids that mimic the whereas small particles are generally tess active
cell membrane [187], or with heparin or heparin-like tumorigenic agents. With metals the same situation
substances that are themselves known to be anti- does not necessarily prevail, since small particles, with
thrombogenic [188-190]. More natural materials in their much enhanced surface area, are able to act as a
their own right are also being used in some appli- source of metal ions, some of which (e.g. nickel) are
cations, such as the glutaraldehyde-treated pericardial known to be chemical carcinogens, so that chemical
tissue derived from cows or pigs used in some heart effects dominate particle size effects.
valves [191]. Two questions dominate the discussion of "solid-
state carcinogenesis". First, is it possible to pre-
5. Variations in biocompatibility dict which materials will be carcinogenic when
phenomena implanted? Secondly, are materials which appear to
Having described in some detail the various mech- be carcinogenic in rodents also carcinogenic in
anisms by which biomaterials may interact with the humans? Clear answers can be given to neither
tissues of the body, and giving the impression that question. It is likely that tumours can be induced
some unified theory of biocompatibility, however under some conditions with any material on implan-
complex, may be possible, it is necessary at this tation, although rates of induction and species speci-
stage to introduce a few other phenomena which may ficity may vary. However, although some recent
be superimposed on the whole process and which journals have borne some evidence of tumours associ-
make interpretation of the overall response very diffi- ated with implants in humans [193], the chances of this
cult. Three particular phenomena come into this being a significant problem are quite low.
category: the initiation of tumours by implanted bio- In studying the biocompatibility of potential

3440
biomaterials, it is clearly important to consider the although it is not usual to continually re-expose the
carcinogenicity aspect. It remains, however, difficult body to repeated challenges from biomaterials, in
to interpret the data that are obtained. the same way as tissues are exposed to the micro-
organisms of contagious or infectious diseases for
5.2. Immunological r e s p o n s e example, the continued presence of a material in the
Even more diffficult to assess and understand is the tissues, leading to a persistent source of reaction
possibility of the host developing a hypersensitivity products, can provide a continuum in which this
response to a biomaterial. immunological memory can be developed.
The inflammatory response to the presence of a It should be borne in mind that the development of
foreign body (described earlier) has to be considered immunological reactions is idiosyncratic, as witnessed
as the first line of defence, but it lacks specificitv in by the great variability in the susceptibility of the
relation to the nature of the material and has no population to allergens, such as in hay fever and
built-in memory facility. In addition to this response, asthma, so that the testing for (and predictability of)
however, there is a second line of defence which the immunological properties of biomaterials is a very
possesses both specificity and memory; this is the difficult matter. Nevertheless, there is increasing
immunological response. Specificity arises because the evidence that biomaterials are involved in eliciting a
response is not activated by all materials; indeed, only response from the immune system and this aspect has
certain types of material can elicit an immune response to be taken very seriously in their evaluation. The
and the extent of this response will depend on the reader is referred to several recent reviews on this
nature of the material. The memory is involved because subject for further information [194-196].
on first exposure the tissues develop a recognition
pattern, and at this time may react only minimally, but 5.3. Biomaterial-bacterial interactions
this recognition can be fully brought into play when It will be recalled from Section 4 that the response
the tissues are exposed a second or subsequent times, of tissues to injury follows the same general path-
a very considerable response being called into action. way irrespective of the causative agent. One of the
Thus, the immunological defence mechanism (which most frequent causes of tissue damage is infection by
may differ in detail in different animal species) is micro-organisms (bacteria, fungi, viruses) and the
characterized by two important features. First, the ability of tissues to deal with such an invasion is well
reaction is directed against one specific foreign invader documented. It should be apparent, however, that
at a time. Following exposure to an invader (e.g. micro-organisms (and especially bacteria) and bio-
micro-organism or foreign particle) the host will react materials have a number of similarities, especially as
against that invader and will be left with a state of both may be considered as persistent sources of tissue
immunity to that body. That particular immune state irritation, so that the mechanisms and features of
will be ineffective against subsequent challenge with a the tissue response to them will have some common
different organism or material. Secondly, the immuno- features. This implies that there is some scope for
logical memory for the invader results in a far more interaction between bacteria-tissue and biomaterial-
rapid generation of immunological effectors on the tissue phenomena; this is indeed the case and such an
second exposure. interaction can very significantly influence the course
The type of immune response generated against any of events.
particular foreign body will depend to a large extent We may consider this subject from two different
on the nature of the body, which is referred to as the points of view. First there is the prospect that the
antigen. Generally it is proteinaceous substances that stability of materials in the physiological environment
are the most potent antigens, synthetic, man-made may be comprised by the presence of bacteria in an
substances themselves being unlikely antigens. The infection. This subject has been discussed at some
reaction itself may be either a humoral or a cell- length recently by the author [197]. It is relevant here
mediated response, or both. A humoral response is merely to point out that microbiological corrosion
one in which antibodies are developed by the host and degradation are phenomena directly relevant
which react with (i.e. bind to and inactivate) the invad- to biomaterials usage, where the rates and indeed
ing foreign body. A cell-mediated response is one mechanisms of degradative processes may be either
involving special lymphocytes, specifically endowed enhanced or reduced by the activity of bacteria,
with receptors or recognition facilities for (and the especially by the changes in the microenvironment,
capacity to kill) the particular invading substance. such as pH, oxygen and enzyme activity, introduced
The relevance of these immunological responses to by the bacteria.
biomaterials lies in a number of points. First, although Secondly, and as reviewed recently by Sugarman
antigens are usually complex organic molecules of and Young [198], biomaterials or prosthetic devices
relatively high molecular weight, some other substances are associated with increased susceptibility to infec-
may become antigenic when coupled to other molecules tion. At its simplest, it is known that the presence of
even if they themselves are non-antigenic. Thus, the a foreign material in a wound can reduce the number
vast majority of biomaterials are non-antigenic but of bacteria necessary to produce a clinical infection
they may activate the immune response because of the by orders of magnitude [199]. It also appears that
binding of products released from their interaction implanted devices can alter the pathogenicity or
with the environment of the body with some appro- virulence of micro-organisms such that bacteria which
priate carrier molecules in the tissue. Secondly, are normally non-pathogenic may become pathogenic
3441
in the presence of materials. Furthermore, and of 20. J. L. BRASH, in "Interactions of the Blood with Natural
and Artificial Surfaces", edited by E. W. Salzman (Dekker,
considerable clinical significance, implanted devices
New York, 1981) p. 37.
may act as host sites for infection years after the 21. N. N O R D E and J. L Y K L E M A , J. Coll. Interface Sei. 66
operative procedure to place the device and there is (1978) 295.
evidence that such late infections can occasionally 22. E. NYILAS, T. H. C H I U and G. A. H E R Z L I N G E R ,
arise from the spread of bacteria from other, possibly Trans. Amer. Soc. Artif. Int. Organs 20 (1974) 480.
harmless transient infections. It is difficult to assess the 23. J. L. BRASH and S. U N I Y A L , ibid. 22 (1976) 753.
24. E. J. C A S T I L L O , J. L. K O E N I G , J. M. A N D E R S O N
significance of this, and the assumption that infections and J. LO, Biomaterials $ (1984) 319.
arising from dental treatment may be involved has 25. E. J. C A S T I L L O , J. L. K O E N I G , J. M. A N D E R S O N
recently been challenged [200]. and J. LO, ibid. 6 (1985) 338.
26. E. J. C A S T I L L O , J. L. K O E N I G , J, M. A N D E R S O N
and N. J E N T O F T , ibid. 7 (1980) 9.
6. Concluding comments
27. E. J. CASTILLO, J. L. K O E N I G and J. M. A N D E R -
The biomaterials field, as measured by any parameter, SON, ibid. 7 (1986) 89.
is large and is growing fast. New materials are con- 28. T. A. H O R B E T T , C.M. CHENG, B.D. RATNER,
stantly, being introduced into various medical and A. S. H O F F M A N and S. R. H A N S O N , J. Biomed.
surgical applications. Their success depends upon the Mater. Res. 20 (1986) 739.
29. L. V R O M A N , A. L. A D A M S and G. C. F I S H E R , Adv,
manner with which they interact with the tissues of
Chem. 199 (1982) 265.
the body. This review has attempted to define the
30. S. U N I Y A L and J. L. BRASH, Thromb. Hemostas. 47
basic mechanisms by which this takes place and to (1982) 285.
indicate the variables which influence the course of the 31. M. DEYNE, A. B A S Z K I N , J. E. P R O U S T , E. PEREZ
reactions observed. and M. M. B O I S S O N N A D E , J. Biomed. Mater. Res. 20
(1986) 951.
32. D. F. W I L L I A M S , I. N. A S K I L L and R. SMITH, ibid.
References 19 (1985) 313.
1. D. F. W I L L I A M S , in Proceedings of Consensus Con- 33. D. F. W I L L I A M S , Ann. Rev. Mater. Sci. 6 (1976) 237.
ference of the European Society for Biomaterials, Chester, 34. Idem, in " F u n d a m e n t a l Aspects of Biocompatibility", Vol. I,
1987 (Elsevier, Amsterdam, in press). edited by D. F. Williams (CRC, Boca Raton, 1980) p. 11.
2. D. D O W S O N and V. W R I G H T , " A n Introduction to the 35. D. K. G I L D I N G , ibid. p. 43.
Biomechanics of Joints and Joint Replacement" (Mechanical 36. L. L. H E N C H , ibid. p. 67.
Engineering Publications, London, 1981). 37. D. F. W I L L I A M S , J. Mater. Sci. 17 (1982) 1233.
3. D. L. C O L E M A N , H.L.C. MEUZELAAR, T.R. 38. S. G. S T E I N M A N N , in "Evaluation of Biomaterials",
KESSLER, W.H. McCLENNEN, J. M. R I C H A R D S edited by G. D. Winter, J. L. Leray and K. de Groot (Wiley,
and D. E. G R E G O N I S , J. Biomed. Mater. Res. 20 (1986) London, 1980)p. 1.
417. 39. D. B R U N E , Biomaterials 7 (1986) 163.
4. M. F. R E F O J O , in Proceedings of the 5th European 40. K. J. B U N D Y , M.A. VOGELBAUM and v.H.
Congress of Biomaterials, Paris, September 1985, edited by DESAI, J. Biomed. Mater. Res. 20 (1986) 493.
P. Christel, A. Meunier and A. J. C. Lee (Elsevier, Amster- 41. C. C. CHU, C R C Crit. Rev. Biocompat. I (1985) 261.
dam, 1986)p. 159. 42. W. J. BAILEY and B. G A P U D , Ann. N.Y. Acad. Sci. 446
5. W. D R U K K E R , F. M. P A R S O N S and J. F. M A H E R , (1985) 42.
"Replacement of Renal Function by Dialysis" (Nijhoff, 43. D. F. W I L L I A M S , in "Biocompatibility of Orthopaedic
Boston, 1983). Implant Materials", Vol. I, edited by D. F. Williams (CRC,
6. D. F. W I L L I A M S , J. Dent. 14 (1986) 185. Boca Raton, 1983).
7. G. E. LOEB, J. M c H A R D Y , E. M. K E L L I H E R and 44. D. F. W I L L I A M S and G. M E A C H I M , J. Biomed. Mater.
S. B. B R U M M E R , in "Biocompatibility in Clinical Prac- Res. Symp. 5 (1974) 1.
tice", Vol. II, edited by D. F. Williams (CRC, Boca Raton, 45. P. SURY, Corros. Sci. 17 (1977) 155.
1982) p. 123. 46. W. R O S T O K E R , C.W. PRETZEL and J. O. GAL-
8. C. G. G E B E L E I N and C. E. C A R R A H E R , "Polymeric ANTE, 3". Biomed. Mater. Res. 8 (1974) 407.
Materials in Medication" (Plenum, New York, 1985). 47. T. P. H O A R and D. C. M E A R S , Proc. R. Soc. A294
9. P. SAMET and N. E L - S H E R I F , "Cardiac Pacing" (1966) 486.
(Grune and Stratton, New York, 1980). 48. D. F. W I L L I A M S , C R C Crit. Rev. Biocompat. 1 (1985) 1.
10. D. F. W I L L I A M S , "Biocompatibility in Clinical Practice", 49. D. F. W I L L I A M S and G. C. F. C L A R K , J. Mater. Sci.
Vols I and II (CRC, Boca Raton, 1981). 17 (1982) 1675.
11. J. W. BORETOS and M. EDEN, "Contemporary Bio- 50. S. A. B R O W N and K. M E R R I T T , J. Biomed. Mater.
materials" (Noyes, New Jersey, 1984). Res. 15 (1981) 479.
12. J. J. G R O T E , "Biomaterials in Otology" (Nijhoff, Boston, 51. P. D U C H E Y N E , P. de M E E S T E R , E. A E R N O U D I T ,
1984). M. M A R T E N S and J. C. M U L L E R , J. Biomed. Mater.
13. C. G. G E B E L E I N and F. F. K O B L I T Z , "Biomedical Res. Syrup. 6 (1975) 199.
and Dental Applications of Polymers" (Plenum, New York, 52. G. M E A C H I M and D. F. W I L L I A M S , ibid. 7 (1973) 555.
1981). 53. C. G. PITT, M. M. G R A T Z L , G. L. K U M M E L , F.
14. D. F. W I L L I A M S , Phys. Med. Biol. 25 (1980) 611. SVRIES and A. S C H I N D L E R , Biomaterials 2 (1981) 215.
15. D. F. W I L L I A M S and R. R O A F , "Implants in Surgery" 54. J. H E L L E R , Ann. N.Y. Acad. Sci. 446 (1985) 51.
(Saunders, London, 1973). 55. J. M. A N D E R S O N , ibid. 446 (1985) 67.
16. D. F. W I L L I A M S , "Biocompatibility of Clinical Implant 56. B. E. C O H A N , Amer. J. Ophthalmol. 88 (1979) 982.
Materials", Vols I and II (CRC, Boca Raton, 1981). 57. v. M I R K O V I T C H , T. A K U T S U and w . S. K O L F F ,
17. R. E. BALER and R. C. D U T T O N , J. Biomed. Mater. Trans. Amer. Soc. Artif. Int. Organs 8 (1962) 79.
Res. 3 (1969) 191. 58. R. SMITH and D. F. W I L L I A M S , J. Mater. Sci. Lett. 4
18. R. M. G E N D R E A U , R.I. LEININGER and R . J . (1985) 547.
J A K O B S E N , in Abstracts of First World Congress on Bio- 59. Idem, Polym. Testing 6 (1986) 243.
materials, Baden, Austria, 1980 (Wiley, New York, 1980). 60. R. SMITH, C. OLIVER and D. F. W I L L I A M S , J.
19. L. V R O M A N , A. L. A D A M S , A. L. K L I N G S , G . C . Biomed. Mater. Res. 21 (1987) in press.
F I S H E R , P. C. M U N O Z and R. P. SOLENSKY, Ann. 6l. R. E. M A R C H A N T , K.M. MULLER and J.M.
N.Y. Acad. Sci. 283 (1977) 65. A N D E R S O N , ibid. 18 (1984) 1169.

3442
62. D. F. WILLIAMS, R. SMITH and C. OLIVER, in Pro- 97. F. GRINNELL and M. K. FIELD, J. Biol. Chem. 257
ceedings of the 5th European Conference on Biomaterials, (1982) 4888.
Paris, September 1985, edited by P. Christel, A. Meunier and 98. A. McNAMARA and D. F. WILLIAMS, Biomaterials 2
A. J. C. Lee (Elsevier, Amsterdam, 1986) p. 239. (1981) 33.
63. L. L. HENCH and E. C. ETHERIDGE, "Biomaterials: 99. ldem, ibid. 3 (1982) 160.
An Interfacial Approach" (Academic, New York, 1982). 100. ldem, J. Biomed. Mater. Res, 18 (1984) 185.
64. R. W. DAVIDGE, Biomaterials 5 (1984) 37. 101. C. R. BRADLEY, in "Systemic Aspects of Biocompati-
65. C. P. A. T. KLEIN, A. A. DRIESSEN, K. de GROOT bility", Vol. I, edited by D. F. Williams (CRC, Boca Raton,
and A. van den HOOFF, J. Biomed. Mater. Res. 17 (1983) 1981) p. 187.
769. 102. M. K. WARD, ibid. in "Systemic Aspects of Biocompati-
66. C. P. A. T. KLEIN, K. de GROOT, A. A. DRIESSEN bility, Vol. II, edited by D. F. Williams (CRC, Boca Raton,
and H. B. M~ van der LUBBE, Biomaterials 6 (1985) 189. 1981) p. 187.
67. L. L. HENCH and A. E. CLARK, in "Biocompatibility 103. G. MEACHIM and D. F. WILLIAMS, J. Biomed. Mater.
of Orthopaedic Implants", Vol. II, edited by D. F. Williams Res. 7 (1973) 555.
(CRC, Boca Raton, 1982) p. 129. 104. H. J. MUELLER, Biomaterials 4 (1983) 66.
68. I. H. LEPOW and P. A. WARD, "Inflammation: Mech- I05. A. KOEGEL and J. BLACK, J. Biomed. Mater. Res. 18
anisms and Control" (Academic, New York, 1972). (1984) 513.
69. G. RYAN and G. MAJNO, Amer. J. Pathol. 86 (1977) 106. K. MERRITT, S. A. BROWN and N. A. SHARKEY,
185. ibid. 18 (1984) 1005.
70. H. KELLER and G. O. TILL, in Proceedings of First 107. D. F. WILLIAMS, in "Fundamental Aspects of Biocom-
International Conference on Leucocyte Locomotion and patibility", Vol. II, edited by D. F. Williams (CRC, Boca
Chemotaxis, Gersau, 1982 (Birkhauser, Basel, 1983). Raton, 1981) p. 45.
71. T. RAE, in "Fundamental Aspects of Biocompatibility", 108. Y. RAE, in "Systemic Aspects of Biocompatibility", Vol. I,
Vol. II, edited by D. F. Williams (CRC, Boca Raton, 1981) edited by D. F. Williams (CRC, Boca Raton, 1981) p. 21.
p. 159. 109. D. F. WILLIAMS and H. CROWLEY, in "Biological
72. H. J. SENN and W. F. JUNGI, Semin. Haematol. 12 and Biomechanical Performance of Biomaterials", edited by
(1975) 7. P. Christel, A. Meunier and A. J. C. Lee (Elsevier, Amster-
73. S. J. LEIBOVICH and R. ROSS, Amer. J. Path. 78 dam, 1986) p. 391.
(1975) 71. 110. G. D. WINTER, J. Biomed. Mater. Res. Symp. 5 (1974)
74. S. H. ZIGMOND, J. Cell Biol. 77 (1978) 269. 11.
75. T. P. STOSSEL, New Engl. J. Med. 290 (1974) 717. 111. G. MEACHIM, in "Biocompatibility of Implant Materials",
76. M. DIROSA, J. P. GIROUD and D. A. WILLOUGH- edited by D. F. Williams (Sector, London, 1976) p. 120.
BY, J. Path. 104 (197I) 15. 112. G. MEACHIM and R. B. PEDLEY, in "Fundamental
77. L. S. GOODMAN and A. GILMAN, "The Pharmaco- Aspects of Biocompatibility", Vol. I, edited by D. F. Williams
logical Basis of Therapeutics", 5th Edn (MacMillen, New (CRC, Boca Raton, 1981) p. 107.
York, 1975) p. 590. 113. O. E. BEDER and G. LADE, Surgery39(1956) 470.
78. G. MAJNO and M. LEVENTHAL, Lancet 2 (1967) 99. 114. F. ESCALAS, J. GALANTE, W. ROSTOKER and P. S.
79. R. SNYDERMAN, J. L. PHILLIPS and S. E. MER- COOGAN, J. Biomed. Mater. Res. 9 (1975) 303.
GENHAGEN, J. Exp. Med. 134 (1971) 1131. 115. A. HENSTEN-PETTERSEN and N. JACOBSEN, in
80. R. B. ZURIER, Arch. Intern. Med. 133 (1974) 101. "Evaluation of Biomaterials", edited by G. D. Winter, J. L.
81. R. F. DIEGELMANN, J . K . COHNEN and A . M . Leray and K. de Groot (Wiley, London, 1980) p. 44.
KAPLAN, Plast. Reeonstr. Surg. 68 (1981) 107. 116. S. G. STEINEMANN, ibid. p. 1.
82. M. F. GREAVES, J. OWEN and M. RAFF, "T and B 117. I. A. MJOR, H. M. ERIKSEN, E. HAVGEN and O.
lymphocytes; their origins, properties and roles in the immune SKOGEPAL, Int. Dent. J. 27 (1977) 333.
response" (North-Holland, Amsterdam, 1973). 118. J. R. de WIJN and P. J. van MULLEM, in "Biocompati-
83. T. J. CHAMBERS, in "Fundamental Aspects of Biocom- bility of Clinical Implant Materials", Vol. II, edited by D. F.
patibility", Vol. I, edited by D. F. Williams (CRC, Boca Williams (CRC, Boca Raton, 1981) p. 99.
Raton, 1981)p. 145. 119. G. L. WILKES, in "Polymers in Medicine and Surgery",
84. K. S. WARREN, Ann. N.Y. Aead. Sei. 278 (1976) 7. edited by R. L. Kronenthal, Z. Oser and E. Martin (Plenum,
85. R. B. LEWY, Arch. Otorhinolaryn. 83 (1966) 355. New York, 1975) p. 45.
86. F. BENTLEY and R. B. DUTHIE, in "Scientific Founda- 120. D. K. GILDING, I. N. ASKILL, G. F. GREEN and
tion of Surgery", edited by C. Wells, J. Kyle and J. Dunphy G. H. PIGOTT, in Transactions of 5th Meeting of the
(Heinemann, London, 1974) pp. 143, 156. Society for Biomaterials, Clemson, South Carolina, 1979
87. R. ROSS, Phil. Trans. Soc. B271 (1975) 247. (Society for Biomaterials, San Antonio, Texas, 1979) p. 159.
88. B. WYKE, in "Scientific Foundations of Surgery", edited 121. A. J. LENTZ, T. A. HORBETT, L. HSU and B. D.
by C. Wells, J. Kyle and J. Dunphy (Heinemann, London, RATNER, J. Biomed. Mater. Res. 19 (1985) 1101.
1974) p. 242. 122. R. E. MARCHANT, A. HILTNER, C. HAMLIN, A.
89. D. F. WILLIAMS, J. Med. Eng. Tech. 9 (1977) 266. RABINOVITCH, R. SLOBODKIN and J. M. ANDER-
90. P. GRISS and G. HE1MKE, in "Biocompatibility in Clini- SON, ibid. 17 (1983) 301.
cal Practice", Vol. I, edited by D. F. Williams (CRC, Boca 123. R. E. MARCHANT, K . M . MILLER and J.M.
Raton, 1981)p. 155. ANDERSON, ibid. 18 (1984) 1169.
91. C. A. HOMSY, in "Biocompatibility in Clinical Practice", 124. R. H. RIGDON, ibid. 8 (1974) 97.
Vol. II, edited by D. F. Williams (CRC, Boca Raton, 1981) 125. R. L. CURRAN and J. A. M. AGER, Nature 121 (1962)
p. 59. 494.
92. V. GERET, B. A. RAHN, R. MATHYS, F. STRAU- 126. T. N. SALTHOUSE, in "Advances in Biomaterials",
MANN and S. M. PERREN, in "Evaluation of Bio- edited by G. D. Winter (Wiley, Chichester, 1983) p. 127.
materials", edited by G. D. Winter, J. L. Leray and K. de 127. S. R. TAYLOR and D. F. GIBBONS, J. Biomed. Mater.
Groot (Wiley, London, 1980) p. 351. Res. 17 (1983) 205.
93. R. C. SCHULTZ, in "Contemporary Biomaterials", edited 128. N. J. KORMAN, O. SUDILOVSKY and D. F. GIB-
by J. W. Boretos and M. Eden (Noyes, New Jersey, 1984) BONS, ibid. 18 (1984) 225.
p. 332. 129. B. F. MATLAGA, L. P. Y. ASENCHAK and T . N .
94. R. D. BAGNALL, Biomaterials 1 (1980) 97. SALTHOUSE, ibid. 10 (1976) 391.
95. J. M. CARTER, J. R. NATIELLA, R. E. BAIER and 130. J. CHARNLEY, Proc. R. Soc. Med. 50 (1957) 840.
R. R. NATIELLA, Artif. Organs 8 (1984) 102. 131. T. ALBREKTSSON, CRC Crit. Rev. Bioeompat. 1 (1985)
96. D. M. BRUNETTE, CRC Crit. Rev. Biocompat. 1 (1985) 53.
323. 132. P. I. BRANEMARK, R. ADELL, T. ALBREKTSSON,
3443
 U. LEKHOLM, S. LUNDQVIST and B. ROCKLER, 164. S. NATELSON and E. A. NATELSON, "Principles of
Biomaterials 4 (1983) 25. Applied Clinical Chemistry", Vol. 3 (Plenum, New York,
133. Y. ALBREKTSSON, J. Prosthet. Dent. 50 (1983) 255. 1980).
134. K. de GROOT, "Bioceramics of Calcium Phosphate" 165. E. W. SALZMAN, J. LINDON, D. BRIER and E. W.
(CRC, Boca Raton, 1983). MERRILL, Ann. N.Y. Acad. Sci. 283 (1977) 114.
135. D. F. WILLIAMS, in "Biocompatibility of Tissue 166. E. W. DAVIE and U. D. RATNOFF, in "The Proteins",
Analogs", Vol. II, edited by D. F. Williams (CRC, Boca Vol. III, edited by H. Neurath (Academic, New York, 1965).
Raton, 1985) p. 43. 167. L. POLLER (ed.), "Recent Advances in Blood Coagula-
136. J. N. KENT, Dent. Clin. N. Amer. 30 (1986) 231. tion" (Churchill Livingstone, Edinburgh, 1981).
137. M. JARCHO, Clin. Orthop. 157 (1981) 259. 168. T. H. SPAET (ed,), "Progress in Hemostasis and Throm-
138. L. L. HENCH, R. J. SPLINTER, W. C. ALLEN and bosis" (Grune and Stratton, New York, 1978).
Y. K. GREENLEE, J. Biomed. Mater. Res. Symp. 2 (1972) 169. J. H. G R I F F I N , Proc. Nat. Acad. Sci. USA 75 (1978)
117. 1998.
139. J. WILSON, G. H. PIGOTT, F. J. SCHOEN and L. L. 170. J. A N D R A D E , "Surface and Interfacial Aspects of Bio-
HENCH, J. Biomed. Mater. Res. 15 (1981) 805. medical Polymers", Vol. 2 (Plenum, New York, 1985).
140. U. GROSS and V. STRUNZ, ibid. 19 (1985)251. 171. J. L. BRASH and S. UNIYAL, J. Polym. Sei. 66 (1979)
141. G. D. WINTER and B. J. SIMPSON, Nature 223 (1969) 66.
88. 172. B. D. RATNER, T . A . HORBETT, D. SHUTTLE-
142. M. SPECTOR, in "Biocompatibility of Orthopaedic WORTH and H. R. THOMAS, J. Colloid. Interface Sci.
Implants", Vol. II, edited by D. F. Williams (CRC, Boca 83 (1981) 630.
Raton', 1982) pp. 55, 89. 173. J. L. BRASH, in "Biocompatible Polymers", edited by
143. M. SZYCHER, B. D. T. DALY and K. A. DASSE, in M. Szycher (Technomic, Lancaster, Pennsylvania, 1983) p. 35.
"Contemporary Biomaterials", edited by J. W. Boretos and 174. R. G. MASON, Prog. Hemost. Thromb. 1 (1972) 141.
M. Eden (NoTes, New Jersey, 1984) p. 493. 175. S. W. KIM, R. G. LEE, H. OSTER, D. COLEMAN,
144. P. DUCHEYNE, M. MARTENS, P. de MEESTER and J. D. A N D R A D E , D . L . LENTZ and D. OLSEN,
J. C. MULIER, ASTM Special Technical Publication 79b Trans. Amer. Soc. Artif Int. Organs 20 (1974) 449.
(ASTM, Philadelphia, 1983) p. 365. 176. D. L. COLEMAN, D . E . GREGONIS and J.D.
I45. J. D. BOBYN, R. M. PILLIAR, H. V. CAMERON and A N D R A D E , J. Biomed. Mater. Res. 16 (1982) 381.
G. C. WEATHERLY, Clin. Orthop. 150 (1980) 263. 177. T. OKANO, S. NISHIYAMA, I. SHINOHARA, T.
146. D. F. WILLIAMS, in "Biocompatibility of Orthopedic AKAIKE, Y. SAKURAI, K. KATAOKA and
Implants", Vol. I, edited by D. F. Williams (CRC, Boca T. TSURUTA, ibid. 15 (1981) 393.
Raton, 1982)p. 1. 178. J. M. ANDERSON and K. K O T T K E - M A R C H A N T ,
147. M. LANDRY and H. FLEISCH, J. Bone Jr. Surg. 46B CRC Crit. Rev. Bioeompat. 1 (1985) 111.
(1964) 764. 179. C. H. BAMFORD, I. P. MIDDLETON, Y. SATAKE
148. D. F. WILLIAMS, L. F. GORE and G. C. F. CLARK, and K. G. AL-LAMEE, in "Advances in Polymer Syn-
Biomaterials 4 (1983) 285. thesis", Vol. 31, editedby B. M. Culbertson and J. E. McGrath
149. H. K. U H T H O F F and F. L. DUBUC, Clin. Orthop. 81 (Plenum, New York, 1985).
(1971) 165. 180. D. F. WILLIAMS, in "Blood Compatibility", edited by
150. S. L. Y. WOO, W. H. AKESON, R. D. COUTTS, L. D. F. Williams (CRC, Boca Raton, in press).
RUTHERFORD, D. DOTY, G. F. JEMMOTT and 181. A. A. LEFEMINE, M. M U L L E R and G. C. PINDER,
D. A. AMIEL, J. Bone Jt. Surg. 58A (1976) 190. J. Thorac. Cardiovasc. Surg. 67 (1974) 857.
151. D. F. WILLIAMS, in "Blood Compatibility", edited by 182. G. C. FRANCOS, J. F. BURKE, A. BESARAB, J.
D. F. Williams (CRC, Boca Raton, in press). MARTINEZ, R. G. KIRKWOOD and L. A. H U M M E L ,
152. W. S. PIERCE, in Proceedings of U S A - U S S R Joint Sym- Trans. Amer. Soc. Artif. Intern. Organs 29 (1983) 140.
posium on Circulatory Assistance and the Artificial Heart, 183. P. R. FEBRES-ROMAN, W . C . BOURG, H.A.
Tbilisi, September 1979 (US Department of Health, Human CRONE, H. C. DAVIS and T. H. WILLIAMS, Amer. J.
Services, Washington, 1980; NIH Publication 80-2032). Cardiol. 46 (1980) 735.
153. T. AKUTSU, "Artificial Heart" (Springer-Verlag, Tokyo, 184. M. N. HELMUS, D. F. GIBBONS and R. D. JONES,
1986). J. Biomed. Mater. Res. 18 (1984) 165.
154. M. I. IONESCU, "Techniques in Extracorporeal Circu- 185. R. E. BAIER, in "Biologic and Synthetic Vascular Pros-
lation" (Butterworths, London, 1981). theses", edited by J. C. Slenley (Grune and Stratton, New
155. R. D. HUGHES, P. G. LANGLEY and R. WILLIAMS, York, 1982) p. 83.
in "Blood Compatibility", edited by D. F. Williams (CRC, 186. A. S. H O F F M A N , T. A. HORBETT and B. D. RAT-
Boca Raton, in press). NER, Ann. N.Y. Acad. Sci. 283 (1977) 372.
156. K. OTA, in "Contemporary Biomaterials", edited by 187. J. A. HAYWARD and D. CHAPMAN, Biomaterials 6
J. Bonetos and M. Eden (NoTes, New Jersey, 1984) p. 537. (1984) 135.
157. J. C. STANLEY, W . E . BURKEL, G . M . LIN- 188. W. E. H E N N I N K , J. FEIJEN, C . D . EBERT and
D E N A U E R , R. H. BARTLETT and J. E. TURCOTTE, S. W. KIM, Thromb. Res. 29 (1983) 1.
"Biological and Synthetic Vascular Prostheses" (Grune and 189. Y. MIURA, S. ADYUGI, Y. KUSADA and K. MIYA-
Stratton, New York, 1982). MOTO, J. Biomed. Mater. Res. 13 (1980) 619.
158. N. A. PEPPAS and S. L. COOPER, A C S Adv. Chem. 190. M. V. SEFTON, C. H. CHOLAKIS and G. LLANOS,
Ser. 194 (1982). in "Blood Compatibility", edited by D. F. Williams (CRC,
159. E. W. SALTZMAN, "Interaction of the Blood with Boca Raton, in press).
Natural and Artificial Surfaces" (Dekker, New York, 1981). 191. L. M. COHEN and V. GALLUCCI, "Cardiac Biopros-
160. P. DIDISHEIM, M . K . DEWANJEE, C . S . FRISK, theses" (Yorke, New York, 1982).
M. P. KAYE and D. N. FASS, in "Contemporary Bio- 192. R. B. PEDLEY, G. MEACHIM and D. F. WILLIAMS,
materials", edited by J. W. Boretos and M. Eden (NoTes, New in "Fundamental Aspects of Biocompatibility", Vol. II, edited
Jersey, 1984) p. 132. by D. F. Williams (CRC, Boca Raton, 1981) p. 175.
161. M. D. LELAH, L . K . LAMBRECHT and S.L. 193. J. BAGO-GRANELL, J. Bone Jt. Surg. 66 (1984) 540.
COOPER, J. Biomed. Mater. Res. 18 (1984) 475. 194. K. MERRITT, Biomaterials 5 (1984) 47.
162. V. I. SEVASTIANOV and E. A. TSEYTLINA, ibid. 18 195. M. W. ELVES, in "Fundamental Aspects of Biocompati-
(1984) 969. bility", Vol. II, edited by D. F. Williams (CRC, Boca Raton,
163. J. M. ANDERSON and K. KOTTKE-MARCHANT, in 1981) p. 159.
"Blood Compatibility", edited by D. F. Williams (CRC, Boca 196. K. MERRITT, in "Techniques of Biocompatibility Test-
Raton, in press). ing", Vol. II, edited by D. F. Williams (CRC, Boca Raton,

3444
 1986) p. 123. 200. J. J. J A C O B S E N , H. D. M I L L A R D , R. PLEZIA and
197. D. F. W I L L I A M S , CRC Crit. Rev. Biocomp. 1 (1985) 1. J. R. B L A N K E N S H I P , Oral Surg. 61 (1986) 413.
198. 8. S U G A R M A N and E. J. Y O U N G , "Infections Associ-
ated with Prosthetic Devices" (CRC, Boca Raton, 1984).
199. S. D. ELEK and P. D. C O N E N , Br. J. Exp. Pathol. 38 Received 1 August 1986
(1957) 573. and accepted 12 December 1986

3445