Predicting clinical biological responses to dental materials

d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 23–40

Available online at www.sciencedirect.com

journal homepage: www.intl.elsevierhealth.com/journals/dema

Predicting clinical biological responses to dental materials John C. Wataha ∗ Department of Restorative Dentistry, University of Washington, D770B, Box 357456, 1959 NE Pacific Street, Seattle, WA 98195-7456, United States

a r t i c l e

i n f o

a b s t r a c t

Article history:

Objectives. Methods used to measure and predict clinical biological responses to dental mate-

Received 29 June 2011

rials remain controversial, confusing, and to some extent, unsuccessful. The current paper

Received in revised form

reviews significant issues surrounding how we assess the biological safety of materials, with

25 August 2011

a historical summary and critical look at the biocompatibility literature. The review frames

Accepted 26 August 2011

these issues from a U.S. perspective to some degree, but emphasizes their global nature and universal importance. Methods. The PubMed database and information from the U.S. Food and Drug Administra-

Keywords:

tion, International Standards Organization, and American National Standards Institute were

Biocompatibility

searched for prominent literature addressing the definition of biocompatibility, types of

In vitro tests

biological tests employed, regulatory and standardization issues, and how biological tests

Dental materials

are used together to establish the biological safety of materials. The search encompassed

Biomaterials

articles published in English from approximately 1965–2011. The review does not compre-

Biological safety

hensively review the literature, but highlights significant issues that confront the field.

Biological regulation

Results. Years ago, tests for biological safety sought to establish material inertness as the

Controlled clinical trials

measure of safety, a criterion that is now deemed naive; the definition of biocompatibility

Biological response

has broadened along with the roles for materials in patient oral health care. Controversies persist about how in vitro or animal tests should be used to evaluate the biological safety of materials for clinical use. Controlled clinical trials remain the single best measure of the clinical response to materials, but even these tests have significant limitations and are less useful to identify mechanisms that shape material performance. Practice-based research networks and practitioner databases are emerging as important supplements to controlled clinical trials, but their final utility remains to be determined. Significance. Today we ask materials to play increasingly sophisticated structural and therapeutic roles in patient treatment. To accommodate these roles, strategies to assess, predict, and monitor material safety need to evolve. This evolution will be driven not only by researchers and manufacturers, but also by patients and practitioners, who want to use novel materials in new ways to treat oral disease. © 2011 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.

1.

Introduction

We have spent much effort and money to design tests that predict the clinical biocompatibility of dental materials. Yet even

∗

today, the fundamental problem with these tests is that they do not work as well and cannot be as broadly applied as we would like. For several decades, the U.S. Food and Drug Administration, other international governing bodies, researchers in the biocompatibility field, and standards organizations have

Tel.: +1 206 543 5948; fax: +1 206 543 2881. E-mail addresses: [email protected], [email protected] 0109-5641/$ – see front matter © 2011 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.dental.2011.08.595

24

d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 23–40

Fig. 1 – The classic paradigm for biocompatibility assessment of new materials. A new material is first tested with in vitro tests (bottom of pyramid). Only materials that ‘pass’ the in vitro tests move forward for further testing. Presumably some materials are ‘screened out’ during this first phase of testing, represented by the narrower area at the animal stage. Some materials also will be screened out by animal tests, and relatively few will reach the state of clinical or usage tests. Materials that perform favorably in clinical tests will be introduced for clinical use to the public. This paradigm was designed to efficiently evaluate a large number of new materials while maintaining economic and ethical feasibility (see text), and limiting the potential pain or suffering of animals and humans.

subscribed to a stepwise paradigm of in vitro, animal, and usage (clinical) tests to predict the clinical biological performance of new materials (Fig. 1) [1,2]. Originally proposed in 1970 by Autian [3], this paradigm was adopted to ensure the safe introduction of new biomaterials for patient care while accommodating the need for an efficient, ethically sound, and financially viable assessment process. The tests employed within the paradigm each have strengths and weaknesses (see Section 3), but a key tenet of the stepwise strategy has been that each level of test will eliminate unsuitable materials from further testing, hence the pyramidal shape in Fig. 1. In this manner, high-risk materials should be ‘screened out’ early in the process using faster, less expensive tests, thereby saving time, money, and potential pain and suffering of animals and humans. In spite of the potential utility and advantages of the stepwise paradigm of Fig. 1, evidence suggests that it does not work well at all. Early studies demonstrated unequivocally that in vitro, animal, and usage tests did not function as intended within the stepwise paradigm using dental materials relevant to that time [4,5]. Silicate cements, which tested favorably in in vitro tests, were consistently irritating to the pulp in usage tests. Conversely, zinc-oxide eugenol cement, which was uniformly cytotoxic in vitro, performed favorably in usage tests and via clinical experience. These data spurred others to modify animal or especially in vitro tests to be more clinically relevant and therefore predictive. For over 30 years, the quest for better in vitro tests has embraced the philosophy espoused

by Langeland in 1978: if we knew enough and developed better in vitro tests, the stepwise paradigm could work [6]. We have made progress. We have modified both in vitro and animal test conditions to better mimic clinical use. Early efforts explored different types of in vitro or animal tests [7,8]. The concept of the barrier test was developed [5] and refined to include dentin barriers [9–12]. Strategies for aging materials have been applied to in vitro tests [13]. Subcellular mechanisms [14,15], inflammatory responses [16], mutagenic responses [17], oxidative stress [18] and lower, more clinically relevant concentrations of released chemicals have been investigated [19]. More relevant cell lines for testing dental materials have been developed [20,21]. And there are many other areas of progress far too numerous to reiterate here. Yet in spite of our progress, the fundamental problem remains: we have little evidence that in vitro or animal tests can reliably predict which materials will be successful in clinical use. The stepwise paradigm introduced over 30 years ago seems substantively no more viable today than it was then. So how are we to assess and predict the biological response to a new material in a reliable, economically viable, and ethical way? What tools do we have? What new tools do we need? All would agree that we cannot regress to the period when our first use of a material was to try it in a patient. The goals of the current review are to first present concepts that are central to biocompatibility assessment, then to critically discuss the types of tests we currently use to measure biocompatibility. Finally, an alternative paradigm will be presented on how we might progress in our 30+ year quest to adequately assess the biological risks and benefits of a new dental material for clinical use.

Principles and basic ideas about 2. biological responses to materials 2.1.

A definition for biocompatibility

The most-cited definition for biocompatibility was proposed in 1987 by Williams [22]: “ability of a material to perform with an appropriate host response in a specific situation.” Although this definition seems vague and unhelpful at first glance, it represented a quantum leap forward at the time of its introduction. Prior to this definition, the prevailing view was that successful materials played largely inert roles the body. A long list of ‘non-properties’ had evolved for ‘successful’ biomaterials: non-toxic, non-immunogenic, non-thrombogenic, non-carcinogenic, and so forth [22]. The above definition required that materials not only provide some function, but also recognized that the interface created by introduction of the material will elicit a biological response. Thus, the idea that the material could be truly inert was essentially rejected with the adoption of this definition. Given today’s level of understanding of our bodies as sophisticated, complex biological environments, the idea that one could place a foreign material without some sort of response seems naive. Recognition of an active interface between biomaterials and biological systems led to several important basic ideas

d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 23–40

about biocompatibility. These ideas persist today [2,22] and comprise the core ‘dogma’ of biocompatibility assessment. The first idea is that the interactions at the material–tissue interface occur for both; the material elicits a response from the body and the body elicits a response from the material. All materials will be changed at some level by their introduction into a biological environment—either via corrosion, chemical modification, deposition of substance, degradation, or other mechanism. This exchange of responses leads to a second idea: that the material–tissue interface is dynamic. As the material and biological tissue are modified by each other, the changes themselves may trigger other changes. Thus, the interface is not static, but is changing over its lifetime. Furthermore, because we are always changing—we age, develop systemic or local diseases, adopt new activities, eat differently, etc.—any equilibrium established at a material–tissue interface is subject to perturbation. A third idea is that reactions at the material–tissue interface are a function of the tissue where the interface is created. For example, implantation of a material next to dentin creates a different interface than one created in bone or soft tissue. Consequently the effect of the tissue on the material and vice versa will be fundamentally different in each environment. Therefore, favorable material-biological responses in one environment do not assure the same in other environments. Put simply, a material that functions well to restore tooth structure lost through caries may not be a good endosseous implant. And a material that serves well as an endosseous implant may not serve well as a partial denture framework or casting. A fourth idea about biological–tissue interfaces recognizes the nearly obvious, but often forgotten fact that the materials we use do not belong there. Biomaterials are foreign bodies, and biological responses to these materials are characterized by foreign body responses [22–24]. Most evidence supports the concept that materials promote nonspecific protein absorption, which then triggers formation of foreign body giant cells from monocytes and macrophages that mediate a collagenous fibrous a-vascular barrier between the material and body [23,24]. Avoiding or limiting the foreign body response has been a major goal of material development in the past decade [23]. Finally, the most recent idea about biocompatibility is that it is possible to customize interactions at the material–tissue interface [23,25]. As we ask materials to play more sophisticated, longer-term roles in tissues, we seek to customize and optimize the material–tissue interface to assure the best long-term clinical outcomes. We may modify the surface of a material to limit nonspecific protein absorption, add peptide sequences to encourage native protein or cell interactions, or provide a three-dimensional structure to encourage matrix formation. We may place a material that degrades by design over time, but not before it directs tissue responses via embedded cells, proteins or drugs. To accommodate the bioactive dimension of materials described above, Williams recently (2008) updated his original definition of biocompatibility [22]: “ability of a biomaterial to perform its desired function with respect to a medical therapy, without eliciting any undesirable local or systemic effects in the recipient or beneficiary of that therapy, but generating the most appropriate beneficial cellular

25

or tissue response to that specific situation, and optimising the clinically relevant performance of that therapy.”

2.2.

Drugs vs. devices

The U.S. Food and Drug Administration (FDA) recognizes several broad types of substances that are used to promote human health; the two most common of these are drugs and devices. Most dental biomaterials are classified as devices, including filling materials, diagnostic aids, cements, bonding agents, and implants. The FDA defines a device as [26]: “an instrument, apparatus, implement, machine, contrivance, implant, in vitro reagent, or other similar or related article, including a component part, or accessory which is: (1) recognized in the official National Formulary, or the United States Pharmacopoeia, or any supplement to them, (2) intended for use in the diagnosis of disease or other conditions, or in the cure, mitigation, treatment, or prevention of disease, in man or other animals, or intended to affect the structure or any function of the body of man or other animals, and (3) which does not achieve any of it’s primary intended purposes through chemical action within or on the body of man or other animals and which is not dependent upon being metabolized for the achievement of any of its primary intended purposes.” The European Union (EU) directive 93/42 (MDD) defines a medical device as a material that [1]: “achieves its principal way of action not by pharmacological, immunological, or metabolic means, but which may be assisted in its function by such means.” Put simply, drugs are distinct from devices because drugs use chemical actions for their primary effect whereas devices do not. Drugs and devices are regulated by different branches of regulatory agencies, and approval of drugs requires that they show both safety and efficacy for their intended chemical action. Devices, on the other hand, must show only safety in their intended use to cure, treat, or prevent human disease. According to FDA regulations, if the device emits radiation of any type (including blue light), it must also comply with section 531 of the Federal Food Drug & Cosmetic Act [27]. The FDA classifies devices as Class I, II, or III types, with Class III reserved for materials that pose the greatest risks to human health [28]; classification is based on both the intended use and indications for use [29]. The complexity of the approval process and regulations post-approval depend on classification, Class III materials being scrutinized the most both pre- and post-approval. Class III devices require a rigorous and expensive pre-market approval (PMA) process to gain FDA approval. The Class I or II designations make it possible for companies to use the so-called 510(k) process to introduce ‘new’ materials relatively rapidly onto the market by claiming they are substantially equivalent to a material already on the market (see Section 3.5 [29]). If Class I and II devices are ‘exempt’, then even the 510(k) requirement may be waived [30]. Most dental devices are Class I or II, including casting alloys, amalgam (with special consideration), implants, composites, cements, and bonding agents (Table 1). The

26

d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 23–40

Table 1 – FDA classification of common dental devices (sorted approximately by Class).a Material

Section

Class

Facebow Dental articulator Dental cements

872.3220 872.3150 872.3275

Intraoral dental wax Alloy, base metal Alloy, noble metal Amalgam Calcium hydroxide liner Endosseous implant

872.6809 872.3710 872.3060 872.3070 872.3250 872.3640

Impression materials Pit and fissure sealants Porcelain powder for clinical use Pre-formed plastic dental teeth Resin bonding agent Root canal filling resin

872.3660 872.3265 872.6660 872.3590 872.3200 872.3820

Tooth shade resin material

872.3690

I I I II I II II II II II III II II II II II II III II

a b

Comment

- Zinc oxide eugenol, temporary - Non-zinc oxide eugenol, permanent - Special controls in guidance document 08/23/2004b - Special controls in guidance document 08/23/2004 - Special controls in guidance document 07/28/2008 - Root forms, special controls in guidance document 02/12/2004 - Blade forms - Special controls in guidance document 04/22/2003

- Formulations without chloroform - Formulations with chloroform

FDA website: http://www.fda.gov/MedicalDevices/default.htm; this list is not complete; see website for a complete list. Guidance documents are issued periodically and provide information on conditions that might limit or modify the overall classification of the device; available at FDA website.

classification of devices holds importance for manufacturers because it directly determines the costs and time required to get a material to market. Post-market surveillance also depends on the classification. Thus, the ability to introduce new materials in dental practice is significantly facilitated because of their classification as devices, and their classification as Type I or Type II devices. The view by the FDA of dental biomaterials as devices may change in the coming years. Even today, several dental biomaterials clearly have intended chemical effects. For example, any anti-cariogenic effect of glass ionomers mediated by fluoride release would be considered the effect of a drug by the FDA. The ability of coated implants to trigger osteo-inductive responses, or the ability of MTA or calcium hydroxide to stimulate tertiary dentinogenesis are other examples. At present, these material effects are outside the drug regulations of the FDA because they are not portrayed by manufacturers as the ‘primary use’ of the materials. But as our use of materials evolves to focus on their therapeutic effects, we can expect a more complex regulatory environment that will change the availability of materials in dentistry [31].

2.3.

Ethics in biological testing of materials

Historically, clinical testing of dental biomaterials has not always been guided by robust ethical motives. Until the 1970s to 1980s, new techniques and dental materials were sometimes tested in humans with limited regard for biological safety or consequences. A few examples illustrate this point. According to Clarke, a book published by Wooffendale in 1783 describes the treatment of the exposed tooth pulp with oil of cinnamon, cloves, and turpentine ‘or any chemical oil’. An alternative was the delivery of crude opium on ‘lint’ for 8–10 days to cauterize the pulp [32]. Lead was a common filling material because of its malleability, low melting point (328 ◦ C), and ease of filing intraorally [33]. Prosthetics

commonly incorporated human teeth, animal teeth, ivory, and various metallic springs, without consideration for their decay or corrosion in the mouth, let alone their infectious risk [34]. In 1803, Fox described ‘fusible metal’, a combination of bismuth, lead, and tin, that was heated to 100 ◦ C, then poured directly, molten, into the cavity preparation [35]. Fusible metal was a precursor of modern amalgam; its major drawback was reported to be the difficulty of pouring the molten alloy directly into maxillary cavity preparations! Violation of patient rights and interests also has occurred in medicine. The inappropriate use of humans for medical experimentation by some over the years has been well documented [36,37]. These relatively rare but extremely unfortunate events triggered many changes that persist today. The Nuremburg Code (1947), the Declaration of Geneva (1948) and the Helsinki Declaration (originally in 1964, most recent revision in 2008) were seminal documents in formulating ethical guidelines to govern the use of humans in research. From these and other documents stemmed the U.S. National Research Act in 1974, which formed a commission to identify basic ethical principles for research involving human subjects [36]. The product of this commission was the now famous Belmont Report of 1979. The Belmont Report outlined all the principles and procedures for the ethical treatment of humans in research that we use today, including informed consent, the idea of assessing risks and benefits fairly and objectively, the directive to ‘do no harm’, and the protection of special vulnerable populations. The ethical treatment of humans in research today is regulated in the U.S. by the Food and Drug Administration within the Department of Health and Human Services. In 1981, the FDA issued a code of regulations regarding the public welfare, protection of human subjects, development of foods and drugs, and institutional review boards to review research procedures. In 1991, these concepts were adopted by nearly all U.S. agencies and departments that conduct or fund human

d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 23–40

research, and the regulations are best known today as the Common Rule [36]. The Common Rule includes several basic tenets: it (1) requires compliance by research institutions, (2) requires obtaining and documenting informed consent, (3) regulates institutional review board activities, and (4) provides additional protections for vulnerable populations such as pregnant women, prisoners, and children. All biocompatibility research today that involves humans falls under the jurisdiction of the Common Rule in the U.S. The use of animals in research also has a sometimes unfortunate history that has led to many regulations to assure animal welfare and humane treatment, but detailed discussion of these regulations is beyond the scope of the current review. These regulations are dictated by government agencies that fund research, and like regulations for human research, are administered by Universities in the U.S or government or state agencies in other countries. A principal concern with animals is that, unlike humans, animals cannot express discomfort, pain, or other distress. In this sense, assuring the welfare of animals is more difficult than for humans. For biocompatibility assessment of new materials, the above history and current regulations pose a quandary. As Section 3 will discuss, the final, best measure of a material’s biological response is currently obtained via clinical use. All biological tests in vitro and in animals attempt to predict the clinical ‘gold standard’ result, but as we shall see, these other tests are, almost by nature, inadequate. Yet, ethics, expense, and practicality dictate that we cannot test a new material in a clinical environment without considerable pre-testing with these ‘inadequate’ tests. This is a fundamental and currently unresolved problem for biocompatibility assessment.

2.4.

Knowledge gaps and patient exploitation

Patients and practitioners often are at a disadvantage created by a gap in knowledge between what is known about material’s corrosion products and the biological response to those products. It is accepted that no material can be assured to be 100% safe biologically [38,39], and most research in the field suggests that the preponderance of adverse effects of materials are mediated by substances released from the material (‘corrosion’) [40,41]. Yet, our ability to measure the levels of substances released from a material, particularly alloys, often exceeds our knowledge about the biological effects of those released substances. Examples in dentistry are mercury, nickel, lead, bis-phenol A, and trace radioactive elements. This knowledge gap provides opportunities for unscrupulous individuals to make speculative and sensationalistic claims about the effects of low concentrations of corrosion products. These individuals typically promote the idea that low levels of released substances cause or exacerbate diseases that have no known cause. They promote either expensive testing or analyses to determine the ‘real’ risks to a patient, removal of the ‘offending’ materials, or treatment with ‘safer’ alternative materials at great cost. A common strategy is to exploit patients who are especially vulnerable—with chronic physical or emotional diseases untreatable by medicine—with promises of hope, treatment, and even cures. These patients are compelled to pay formidable sums for these services in hopes they can recover from diseases for which traditional

27

medicine offers little. In the absence of well controlled studies, our profession finds itself in a defensive position, unable to prove that the claimed treatments are ineffective because data do not exist. Aside from the unethical nature of these types of activities, the fear of materials or material constituents generated from these activities often requires significant and expensive responses from academia, industry, or government agencies. Such responses misdirect focus and funds from more fruitful research efforts to understand the biological responses to the materials.

3.

Measuring biological responses

3.1.

In vitro tests

The biological response to dental materials can be measured a number of ways, but the most fundamental of these ways is via in vitro tests. By current definition, in vitro tests occur outside an organism, in a vessel of some sort, using cultures of cells or cell constituents. In vitro tests are distinct from ex vivo tests because the latter use an intact tissue or organ that is maintained for a short time (usually < 24 h) in a culture vessel. In their most sophisticated form, in vitro tests use multiple cells, barriers, or special culture conditions to attempt to replicate conditions in vivo (see next section). The term ‘tissue-culture’ is often but inappropriately used to describe the cell-culture techniques employed for in vitro biocompatibility tests. Cell-culture techniques have revolutionized our knowledge of cellular biology—revealing many dimensions of cellular and subcellular function over the past 40 years. In vitro tests are the youngest of the testing strategies to determine the biological response to materials, developed nearly 30 years after tissue and cell-culture techniques emerged in 1926 [41]. The first tests were technically ex vivo, using chick embryonic tissues, but in 1968, Kawahara reported a method to use true cell-cultures to test the ‘cytotoxicity’ of materials, including dental materials [42]. Several years later, Leirskar and Helgeland reported the use of L-929 cells to assess the biocompatibility of standard sized disks of amalgam, early resins, silicate cement, and a gold-based alloy [43]. In 1973, the first quantitative measure of biological response in vitro was reported using the 51 Cr assay. Based on the ability of viable cells with intact membranes to sequester and retain 51 Cr, this test quantified the release of 51 Cr when the toxic effects of a material compromised membrane integrity [44]. Similarly, the development of the agar overlay test was a milestone in the development of the idea of interposing a barrier between the material and cells to simulate in vivo conditions [5,45]. Since 1973, the development and use of in vitro tests has been prodigious, and today’s tests are sophisticated and can assess nearly any aspect of cell function or metabolism, including gene expression, signaling activation, cell cycle and division, inflammatory activation, protein expression, oxidative stress, and many others (see discussion below). The primary strengths of in vitro tests are (a) the ability to control the environment of the cells and their interface with materials, and (b) the ability to measure cell response in detail and with precision. Because of these strengths, in vitro tests

28

d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 23–40

are well suited to identifying mechanisms of cellular response, even to the level of phosphorylation of a single amino acid in a protein or mutation of a single base pair of a gene. In vitro tests also are faster, less expensive, more reproducible, and more scalable than other types of tests. However, in vitro tests may suffer from a lack of relevance to the clinical use of materials, and this weakness is not trivial. Because in vitro tests have been used as screening vehicles for the biological safety of materials, this weakness creates immense problems for biocompatibility testing paradigms ([2]; see Section 3.4). Yet, adaptation of in vitro tests has provided some successes in correlating with the clinical performance of materials. Zinc oxide-eugenol-based cements in intracoronal applications serve as an example. Early in vitro tests of ZOE cements reported severe toxicity [5,45], but this result was incongruous with a sustained track record of successful clinical use in dentistry in cavity preparations. Certainly, no clinical experience suggested the kind of acute, severe toxicity that the in vitro tests predicted, and in vivo pulpal studies verified a relatively low toxicity [5]. Although novel and creative at their introduction, the use of Millipore® filter or agar ‘barriers’ between the ZOE and cells to simulate the clinical use were equivocal [5]; Millipore® filter barriers predicted a lower toxicity whereas agar filters predicted a high clinical toxicity. The juxtaposition of a dentin barrier in in vitro models via intact teeth [46], dentin slices [11,47] or dentin chips [48] reduced the apparent toxicity of ZOE. Results such as these created both skepticism about the usefulness of in vitro tests to screen new materials for biological safety [5] and a belief that in vitro tests might be more useful if they could be constructed with more appropriate clinical relevance [6]. Ironically, the major strength of in vitro tests—to control nearly every aspect the cell–material interface—also is a disadvantage. The number of variables one must define in even a simple in vitro test is formidable (Table 2). With so many variables to manage (e.g., Fig. 2), one can easily expect variation in results among different research groups. Indeed, the level of congruence of reported conclusions about the biological response to many materials is remarkable given the number of variables that must be defined. The diversity of methods in Table 2 also demonstrate the need for standardization of testing methods if these tests are to be screening tests; one cannot be surprised that standards (discussed below) have had problems addressing these many variables. The problem is even more challenging because appropriate variables change depending on the type of material and its clinical use, and we generally employ experimental strategies that assess the effects of only 1–2 variables at a time [49]. In spite of the shortcomings above, the ability to control and measure the cell–material interface in the in vitro environment has given us a remarkable insight into how materials affect biological systems. In vitro work done on nickel-based alloys and compounds serves as one example. Nickel-based alloys have been controversial nearly since their introduction to dentistry because of the known allergenicity [50,51] and carcinogencity [52,53] of nickel compounds, and the potential of these alloys to corrode in the oral cavity [54–58]. Although the clinical implications are not always clear, in vitro tests have revealed how nickel ions or compounds alter cellular oxidative balance [53,59,60], induce carcinogenesis [53,61], activate

Fig. 2 – Various methods of structuring the material–cell interface during in vitro tests. All these methods have been used extensively during in vitro biological assessment of materials. The design chosen has major implications for the test’s outcomes and possible clinical importance.

cellular hypoxia [53,62,63], activate cellular redox responses [31,19], trigger inflammatory pathways [64–68], alter cytokine secretion [16,69] and induce allergic reactions [70–72], among others. In some cases, the level of knowledge is at the atomic level. In vitro tests have helped determine how nickel enters cells [60,61,73], where it is distributed [61,74,75], and how it binds to cellular proteins [76]. The work done on nickel compounds illustrates the strengths and potential of in vitro tests to uncover mechanisms that influence the interface between materials and tissues. Although they are not, strictly speaking, in vitro tests, and although they have not been adopted by current standards organizations or regulatory agencies, computerized modeling techniques have been proposed to predict biological responses [77–79]. These models are based on known biological reactions to molecules with similar geometries or reactive groups and sophisticated computer modeling and statistical analyses of possible interfaces between molecules and cellular constituents. This area continues to evolve and may one day play a role in biocompatibility assessment.

3.2.

Tests in animals

Despite their expense, controversy, and ponderous bureaucratic challenges, animal tests are critical for assessing the biological responses to a new material before it is used in humans. Many aspects of clinical biological responses cannot currently be modeled by in vitro tests, including blood interactions, wound healing, infection, hypersensitivity responses, carcinogenesis, and chronic inflammation, among others [24]. Animal tests provide information about these types of effects without putting humans at risk. Animal tests may be structured to mimic human clinical use to some degree, are generally less expensive than human clinical trials, can be

d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 23–40

29

Table 2 – Variables common in in vitro tests for biological response. Variables Material vs. components

Possibilities-1st level Material Material components

Contact between material and cells

Direct Indirect Barrier

Culture conditions

Ambience Cell environment

Contact duration Cell type(s)

Multiple cells Morphology Cell division Cell metabolism Cell Function Cell membrane integrity

Cell environment Material effect

*Mixing conditions, setting time, surface condition, aging, sterilization procedures, geometry, size scale (nano vs. micro vs. gross). *Chemical form, oxidation state, complexes, concentration, use of solubilizing agents (e.g., DMSO), single vs. multiple components tested together. *Cells onto material, cells around material, material onto cells, surface area of material to volume of medium, method of securing material in relation to cells. *Extracting medium, dilution scheme and diluent. *Type of barrier, thickness of barrier, interface between barrier and material, volume of diluent below barrier. *CO2 content, temperature. *Cell density (cells per mm2 substrate), pre-incubation on substrate (e.g., for attachment), culture vessel (flask, plate, well), volume of medium in culture. *Time relative to division rate of cells.

Cell-line Primary cells

Cell effect

Possibilities-2nd level

Degradation, fouling Corrosion Absorption/adsorption

*Source tissue, species, propagation scheme (e.g., viral or 3T3 method). medium and supplements, substrate (including 3-D). *Source tissue, species, disaggregation method, isolation method, passage number, medium and supplements (particularly with or without serum), substrate (including 3-D). *Layers, 3-D culture, ratios of cell types. *Number, attachment (detachment), pathologic changes, magnification (light vs. electron microscopy). *Number, DNA content (including flow cytometry), DNA condition (strand breaks, mutations, apoptotic changes). *Carbohydrate content, synthetic activity (DNA, RNA, protein), mitochondrial activity, ATP content, glutathione content, ROS content, oxidative markers. *Secretion (e.g., cytokines), gene expression, intracellular protein content or expression, protein compartmentalization, protein phosphorylation, signaling. *Release of natural component (e.g., LDH), release of preloaded molecule (e.g., neutral red, 51 Cr), uptake of molecule normally excluded (trypan blue), membrane inversion (phosphytidal serine), membrane markers. *Stimulation (LPS, TNF␣), antioxidant addition, chelating agents *Imaging (e.g., pitting). Change in contact angle, surface composition. *Release of substances, electrochemical changes (metals), changes in chemical structure, changes in crystal structure. *Change in mass (increase or decrease).

completed more quickly in many cases, and can be controlled to a greater degree. Animals may be exposed to materials or their degradation products with routes of administration or doses that would be unethical to consider in humans. Animal tests may be used to determine responses such as healing post-trauma, bone formation, or dentinogenesis that are difficult or impossible to ethically test in humans. Finally, animals may be tested at various stages of life (for example, embryos or ‘children’) in a manner that is not possible in humans. In spite of the potential for animal tests to yield important information about the biological response to dental materials, several disadvantages limit their usefulness. Above all, the congruity of animal responses to human responses cannot be assumed, and may be, at worst, misleading. Species differences in biological responses to materials can be dramatic [24]. Second, the ability structure an animal test to accurately mimic the human–material interface may be limited to the point that relevance is compromised. For example, the ability to mimic human occlusal forces in the evaluation of endosseous implants is limited in most species, yet occlusal forces are critical to assessing the osseointegrative response

between bone and implant. Third, interpretation of responses is complex in animal tests because many overlapping complex events are occurring simultaneously. For example, inflammatory effects from the trauma of placement, inflammation from bacterial contamination, toxicity from corrosion products, and the biological responses to oral forces are all occurring simultaneously in an animal test. The end result of the overall tissue response may be clear, but the contribution of these mechanisms may be difficult to dissect. Finally, the bureaucracy for approval of animal tests and associated costs are formidable and may even exceed those for human tests, particularly in some species such as primates. As with in vitro tests, many variables must be considered and controlled in animal tests if they are to yield useful information (Table 3). The number of variables are perhaps fewer than for in vitro tests, but still, there are many considerations including the form of the material, how to expose the material to the animal, the nature of the animal itself (species, age, gender, etc.), the contact duration of the material with the animal, and formidably, how to assess the response [49]. Although animal tests are not currently formally divided as

30

d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 23–40

Table 3 – Variables common in animal tests for biological response. Variables Material form

Possibilities-1st level Material intact Material not intact Material leachables

Possibilities-2nd level *Mixing/manufacturing conditions, setting time, surface treatment/condition, aging, sterilization procedures. *Grinding or processing method, particle size, particle shape. *Extracting medium, dilution scheme, diluent.

Material–tissue interface

Route of administration Tissue Capsule

*Oral, surface, implanted, intramuscular, intravenous, others. *Mucosa, muscle, connective tissue, bone, blood, organ, others. *Material encased in second material for implantation (e.g., polyethylene tube) or not.

Animal factors

Species, subspecies Age Gender, reproductive status Diet, nutrition Activity Special

*Mouse, rat, rabbit, pig, goat, cow, sheep, cat, dog, baboon, monkey, others. *Embryonic, newborn, growth period, adult. *Male, female, sterile. *Normal vs. special, special supplements (e.g., antioxidant, drug). *Static, dynamic (application of forces), relevant to clinical use or not. *Animal genetic status (e.g., gene knockout, known mutation, recombinant), disease status (e.g., diabetes, cardiovascular, osteoporotic).

Contact duration Assessment

*Time relative to use of material, time relative to the lifespan of the animal. Survival-growth Gross changes Histological changes

Cell function changes Release of material components

*Survival long- or short-term, weight over time, size. *Visible tissue changes, ulcers, tumors, infections, other. *Tissue architecture, cells numbers, cell types or subtypes, changes over time (e.g., wound healing), necrosis, neoplasm, cell organelle changes, cell surface markers. *Secretion, internal protein levels or distribution, gene expression, apoptosis, mitosis/meiosis. *Local vs. systemic; target organ levels.

such [49], most can be categorized into two types: (1) those to establish the safety of a material (safety-oriented); and (2), those for some specific attribute necessary for clinical success (function-oriented). For example, a dentin bonding agent may be tested for its acute systemic toxicology, chronic systemic toxicology, hypersensitivity potential, ability to cause ulceration of broken or unbroken mucosal membranes, genotoxicity, reproductive toxicology, and other safety-oriented issues [49]. In these types of tests, the form of the material, its doses, and its routes of administration are designed to push limits that may have little clinical relevance. However, an animal test also may be structured to assess the ability of a material to permit secondary dentin formation through a remaining dentin layer, cause odontoblastic damage or destruction, or stimulate tertiary dentin if placed directly on pulpal tissue. For many dental materials, the safety-oriented tests tend to be similar for all new materials (assuming similar use and contact profiles (see Section 3.5)), but the functionoriented tests will, by nature be customized to the specific use and perhaps even the composition of the material. Animal tests pose several unique challenges. Appropriate controls are paramount to maximize the interpretability of the results, but can be difficult to design. For example, if a material is to be implanted, several controls might be appropriate: a non-surgical site to assess normal tissue architecture, a sham surgery site to account for the effects of the surgery itself, a site with a negative control such as Teflon® to account for the physical presence and shape of the material, and a positive control to account for the degree of species specific adverse reaction. To complicate matters further, consideration must be given to the need for control sites in each animal to account for differences in activity, diet, and interaction

with other animals. Second, often it is difficult to quantitatively assess the material–tissue interface without disturbing it and risking inappropriate interpretation—a sort of ‘Heisenberg Uncertainty’ effect. For example, evaluating an intact bone–implant interface histologically was nearly impossible before special techniques were developed in the late 1980s that permitted sectioning of the implant and bone simultaneously in a manner that preserved the interface to the light or electron microscopic level [80]. Or, analyzing tissue adjacent to an alloy crown for release elements generally involves digestion of tissue that destroys information about the tissue status at the site [81]. Third, the ethical and political issues surrounding animal tests are particularly volatile and difficult, and as the nature of these tests become more complex, these issues become more complex. Finally, statistical analysis of responses in animals can be problematic; by nature, the analysis of a section or specific area of tissue is subject to selection bias. This type of problem is particularly acute with imaging techniques. Problems with the definition of independent variables also can occur. For example, if we use multiple teeth in a given animal to study the effects of a restorative material, the teeth are (at least theoretically) not statistically independent samples. Yet use of a one-tooth-per-animal strategy would require an extraordinary number of animals that would be rebuffed by agencies trying to assure the welfare of animals. Cautious interpretation of these experiments is therefore often warranted. Advances in animal testing routinely used for the study of drugs and disease have been slow to be adopted in assessing biocompatibility. Animal models for diseases such as osteoporosis, cardiovascular disease, and diabetes [82] are available and are improving. Animals genetically engineered

d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 23–40

to silence the expression of a specific gene (knockout) or animals with added genes are increasingly supplementing the germ-free and immuno-compromised models that have been available for years. New technologies to analyze and image animal responses, sometimes in three dimensions, are increasingly used, such as micro-CT, infrared scanning, in vivo hybridization, and laser ablation-inductively coupled mass spectrometry. It seems likely that these models and techniques will be applied to analyze dental material–tissue interfaces in the future.

3.3.

Clinical tests (in humans)

Human clinical tests are generally acknowledged as the ‘gold standard’ for assessing any dimension of dental material performance, including biological response [83]. Clinical tests may be of several designs, which are only superficially introduced here. Each design has strengths and weaknesses in terms of its difficulty to conduct, its cost, and the utility of the information it generates [84]. The simplest and least expensive clinical test is the retrospective test, where patient records are reviewed ‘after the fact’ to assess material performance. Although retrospective tests are efficient and less expensive because they do not require direct patient examination, they suffer from the quality of information that may be collected. These tests also rely heavily on the completeness and accuracy of the patient record. For example, if the incidence of pulpitis post-placement of composite were the outcome to be measured, one could look in the patient record for past history of pulp capping, pulp tests, post-restoration endodontic treatment, and patient symptoms, but all of this information must be in the record and must be accurate. Retrospective tests abdicate control of the data quality to past clinicians. And although constraints and safeguards may be applied in collecting or using such data, there are risks of selection bias, inappropriate conclusions, and misinterpretation. In many respects, retrospective tests are uncontrolled, but they provide a relatively efficient and low-cost means to assess clinical material biocompatibility. Cross-sectional tests examine a patient cohort at one point in time. Cross-sectional tests benefit from the ability to define patient inclusion or exclusion criteria and to collect specific data in a standardized way, but they cannot control how a dental material was used or the variables that may have been important but were unrecorded at the time of treatment. In the case of evaluating the tendency of a composite to cause pulpitis, a clinician would have the ability to do pulp tests, take a radiograph, and otherwise examine the patient after treatment, but may not know if the material was placed with appropriate bonding technique, if all decay was removed, or even the age of the restoration. The examiner may not even be sure of what material was used! Cross-sectional tests predict failure rates, but cannot explain the cause of failure [84]. The prospective or longitudinal test is the most powerful clinical tool for determining a material’s clinical biological performance. Also called controlled clinical trials (CCTs) or randomized controlled trials, these types of tests use sophisticated procedures to assure blinding, randomization, and a placebo group to try to account for variables that often confound interpretation of results in human studies [85]. In CCTs,

31

a stringently screened group of patients are treated with one of several strategies (interventions), then followed over time for defined outcomes. In the best designs (which may not be achievable in all cases), there is total blinding: the patient does not know the treatment received, the treating dentist does not know, and the assessor of the outcome (separate from the treating clinician) does not know the treatment. CCTs generate the most reliable and interpretable information, yet they too suffer from several significant problems. CCTs are expensive, time consuming, and require a major commitment of resources. More importantly, the strict definitions used to select cases, treat disease, and assess outcomes may not be clinically relevant [83]. The skill of the treating operator also may not represent the ability of the average clinician, and the disease stage treated may not be relevant to clinical practice [83,84]. Because of the limitations of CCTs, other designs have been explored for clinical testing, including ‘simple’ clinical trials and practice-based research networks (PBRNs) [83,86–88]. Simple clinical trials offer a clinical view of material performance, but without stringency for controls, blindedness, or randomized designs [84]. Rigorous conclusions are impossible, but these tests are faster and less expensive and offer some insight into clinical safety. In a PBRN, the goal is to have a broad cross-section of practicing dentists, diagnose and treat disease and collect data on selection and outcomes. The hope with these networks is to gain clinical information on material performance that is more predictive than traditional CCTs. PBRN strategies are only beginning to be applied to question of dental material biocompatibility [84]. PBRNs are likely most useful if practitioners in the network are calibrated for assessing outcomes and trained to adhere to similar protocols; unfortunately, these goals are hard to realize [88]. Of course, the ultimate clinical trial is the generalized use of a material in clinical practice. Although this strategy is not ethical for dental materials without previous human use, there are countries that have created national databases for practitioners to report adverse events post-market introduction, with the goal of using a very large sample size to dissect problems caused by a material from the ‘noise’ of other variables inherent in clinical use (e.g., disease state, operator skill level and technique used, patient compliance, etc.) [83,84]. It should be clear from the above discussion that there is no ideal strategy for conducting clinical tests; some balance is required among costs, time, and quality of information generated [84], and most materials will be tested by a variety of strategies over their clinical lifetime. When clinically assessing the biological properties of dental materials, there are additional challenges to those mentioned above. Biological outcomes are inherently more difficult to quantify than some other clinical outcomes. Ryge’s 1971 USPHS guidelines for clinical evaluation focused on five clinical criteria: color match, cavo-surface marginal discoloration, anatomic form, marginal adaptation, and caries [89]; many other criteria have since been added. The USPHS criteria, as subjective as they might appear, are generally measurable by direct observation of a patient by a trained examiner. Biological outcomes are inherently more challenging in several ways. First, unlike the USPHS criteria listed above, there is often nothing to directly observe clinically when estimating

32

d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 23–40

biological response. If evaluating pulpitis as discussed earlier, the examiner must rely on feedback from the patient using rudimentary tests (percussion, cold, electric pulp tests, palpation, hot, etc.) to estimate the outcome. In many cases, it is the patient who ultimately does the rating, not the examiner, and this difference makes reliable data collection difficult. Second, even if signs are observable, attribution to the material is not straightforward; differential causes must be considered [2,83]. Is observed gingival redness and swelling from allergy, plaque retention, material toxicity, food trauma, prophylaxis trauma, other cause, or combinations of these factors? Finally, biological responses may occur at sites distant from the restoration—in the pulp, periapical tissues, periodontium, or even in distant organs—a classic example being nickel allergy. So an oral examiner may have no chance to observe or measure the biological effect of a material. On a positive note, the most recent FDI system for clinical evaluation of materials includes biological properties as a cornerstone [90]. Tooth vitality and hypersensitivity, changes in the periodontal or adjacent mucosal tissues, and general health are part of the evaluation. Although clearly a step forward, the inherent problem is the objective quantification of these criteria in the assessment of clinical biocompatibility. In my opinion, the inherent difficulties in measuring clinical biological responses make all but the most basic clinical outcomes difficult to measure via CCTs; the reality is that many thousands of dollars and years of effort using the most rigorous methods may not provide the definitive assurance of clinical biocompatibility that the public demands and researchers seek. Yet, we have no better tool for gathering unbiased data. Bayne [84] has eloquently summarized the concept of failure analysis in the context of clinical assessment of dental materials (Fig. 3). Clinical efficacy is an estimate of the probability of failure using CCTs whereas clinical effectiveness is the rate of failure in clinical practice, commonly twice the rate of efficacy. Put simply, the degree of success of a dental material across a broad cross-section of practitioners will never equal the success rate of the same material in a CCT. This concept can be extended to biocompatibility—the frequency of biological problems in clinical practice with a material likely exceeds that of controlled studies because the day-to-day use of these materials amplifies material weaknesses in ways that affect biological performance [83]: microleakage, corrosion, hydrolysis, etc. (Fig. 3). The practitioner does not have the luxury of inclusion or exclusion criteria: restorations are placed in teeth with compromised pulpal status or in environments with active periodontal disease; and individuals have diets or habits that may alter material–tissue interfaces and therefore biological response. The use of PBRNs, and national or international practitioner databases, coupled with the power of sophisticated statistical analyses, may help define the true biological risks of dental materials that are not able to be accurately measured via CCTs.

3.4. Combinations of biocompatibility tests; order of testing As discussed in the introduction (Section 1), the classic paradigm of sequential in vitro—animal-clinical tests to assess

Fig. 3 – Bayne [84] has observed that the clinical efficacy of a material, assessed by controlled clinical trials is nearly always greater clinically than realized in private clinical practice (effectiveness) because of the stringent conditions under which CCTs are conducted (top figure). A similar principle can be applied to biological assessments; materials used in CCTs will generate fewer biological problems than in private practice because of factors such as case selection, material manipulation, and operator skill level (bottom figure). In this sense, CCTs are not sufficient to accurately predict the biocompatibility of dental materials in clinical practice.

the biocompatibility of a new material is not always successful (Fig. 1). Although attractive in principle, this paradigm suffers from the inability of in vitro and animal tests to sufficiently mimic the clinical environment [5–8]. As a result, materials that may ultimately fail clinically are not eliminated in the early stages of testing, and materials that have known clinical utility do not perform well in in vitro or animal tests [2]. In spite of its shortcomings, parts and some principles of the paradigm in Fig. 1 are still in use because no better alternative has emerged. A perennial question facing the field is what paradigm to use in place the one of Fig. 1? Alternative testing paradigms have been proposed and are used to some extent in some parts of the world (Fig. 4). Some experts have advocated primarily clinical tests, citing the minimal impact laboratory (in vitro and animal) tests have had on the practice of dentistry over the past 30 years [87]. If one looks at the evidence, it is difficult to refute this argument. Yet, it also is not possible to eliminate in vitro and animal tests because we cannot use clinical tests to evaluate novel

d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 23–40

Fig. 4 – Alternate paradigms to combine biological tests to assess the biocompatibility of dental materials [1,2]. These proposals have acknowledged the shortcomings of the classic paradigm in Fig. 1 and advocate combination of in vitro, animal, and clinical tests with more vague boundaries. Each of these models proposes a continued role for in vitro and animal tests as a material progresses to market.

materials—legally, ethically, or scientifically. To do so would be a regression to the human experimentation that we have long since judged to be unethical. Several authors have proposed other paradigms (Fig. 4) [1,91]. These alternatives generally advocate less reliance on the ‘screen out’ philosophy of Fig. 1 and acknowledge a continued role for in vitro and animal tests to initially assess a material, as well as to resolve problems that come up during clinical evaluation. In some sense, these newer paradigms are already in use; researchers have commonly used in vitro or animal tests to evaluate and compare materials on the market, sometimes with surprising results and significant controversy [92]. Others have used this strategy to show ‘substantial equivalence’ of materials (see next section). Still other groups use in vitro tests to try to estimate clinical risk. It is safe to say that no alternative testing paradigm has fully solved the problem of accurate prediction of biocompatibility.

3.5. Standards for biocompatibility testing and FDA 510(k) clause Standards for tests to assess the biological properties of materials have evolved significantly over the years since

33

ANSI/ADA Specification #41 was introduced in 1979 [93]. Today, biological testing standards are maintained by several organizations, including the American National Standards Institute (ANSI, via the American Dental Association, ADA), the American Society for Testing and Materials (ASTM-International), the Committee on European Normalization (CEN), and most importantly, the International Organization for Standardization (ISO) (Table 4). Other important groups such as the Nordic Institute of Dental Materials (NIOM) and the European Union (EN) do not issue standards for biocompatibility directly, but play critical roles in organizations (ISO) that do, and still others such as the ADA coordinate their activities closely with the ISO. The U.S. Food and Drug Administration (FDA) and equivalent governmental organizations in other countries generally use standards from these organizations as a framework for their regulatory decisions. Although the complete listing of standards governing biocompatibility tests is beyond the scope of this manuscript, a partial listing of major standards and standards organizations is listed in Table 4. Organizations update these standards on an ongoing basis. Although much coordination of effort occurs through the ISO, there are organizations, such as the Global Harmonization Task Force (GSTF) that seek to achieve more uniformity among national regulatory agencies for medical devices and better patient safety and access to safe materials for medical devices. The GSTF, formed in 1992, is a partnership with five founding members: the European Union, United States, Canada, Australia, and Japan. When a new material is introduced as a device into dental practice, the manufacturer must decide what testing for safety is necessary, based on the material–tissue interface and the duration of contact. ISO standards provide guidance for these decisions and strategies [49]. Devices applied to the external skin require less testing than those that are on mucosa or breached mucosa. Devices that communicate externally or those that are completely implanted require still more rigorous testing. Furthermore, the ISO divides durations of contact into three groups: limited (≤24 h), prolonged (>24 h to 30 days), and permanent (>30 days); devices with permanent interfaces must be tested most rigorously. The governmental agency will receive the data from the manufacturer and determine if the panel of tests selected and the data from those tests support approval of the device. Thus, it is the governmental agency that approves use of dental devices, but standards organizations provide the testing ‘framework’ for data collection. The manufacturer must build a case to support use of a material in a biological context. The use of standards for biocompatibility testing such as those in Table 4 is not without controversy. On the positive side, standards provide uniformity of testing methods, assurance of appropriate controls during testing, reduced likelihood of testing bias, and better interpretation of data by all parties. Furthermore, standards organizations provide a forum for multiple parties to have input into how materials are tested, thereby assuring that practitioners, academicians, industry, the lay public, and governmental agencies all participate in assuring the biological safety of materials. Some negative aspects of standards are that they tend to be slow to change (often lagging behind the most current technologies for research), may sacrifice the most technically relevant

34

d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 23–40

Table 4 – Summary of major standards governing use of dental materials. Organization ANSI/ADA ASTM-Int

CEN ISO

Standard

Year

Title

41

2005

Recommended standard practices for biological evaluation of dental materials

F1027-86

2007

F1538-03 F748-06

2009 2010

F1609-08 F1876-98

2008 2003

F2026-10

2010

F1441-30 F2211-04 F523-99

2009 2004 2006

EN 1640-1642 10993 10993-1 10993-2 10993-3 10993-4 10993-5 10993-6 10993-7 10992-8 10993-9

2009

Standard practice for assessment of tissue and cell compatibility of orofacial prosthetic materials and devices Standard specification for glass and glass ceramic biomaterials for implantation Standard practice for selecting generic biological test methods for materials and devices Standard specification for calcium phosphate coatings for implantable materials Standard specification for polyetherketoneetherketonekone (PEKEKK) resins for surgical implant applications Standard specification for polyetheretherketone (PEEK) polymers for surgical applications Standard specification for soft-tissue expander devices Standard classification for tissue engineered medical products (TEMPs) Standard guide for pre-clinical in vivo evaluation in critical size segmental bone defects Medical devices for dentistry (4 parts)

10993-10 10993-11 10993-12 10993-13

2010 2006 2007 1998

10993-14 10993-15

2001 2000

10993-16 10993-17 10993-18 10993-19

1997 2002 2005 2006

10993-20 7405 14155

2006 2008 2011

14971

2007

2009 2006 2003 2002/2006 2009 2007 2008 2001 1999

Part 1: Evaluation and testing Part 2: Animal welfare requirements Part 3: Tests for genotoxicity, carcinogenicity and reproductive toxicity Part 4: Selection of tests for interactions with blood Part 5: Tests for in vitro cytotoxicity Part 6: Tests for local effects after implantation Part 7: Ethylene oxide sterilization residuals Part 8: Selection of reference materials Part 9: Framework for identification and quantification of potential degradation products Part 10: Tests for irritation and delayed-type hypersensitivity Part 11: Tests for systemic toxicity Part 12: Sample preparation and reference materials Part 13: Identification and quantification of degradation products from polymeric medical devices Part 14: Identification and quantification of degradation products from ceramics Part 15: Identification and quantification of degradation products from metals and alloys Part 16: Toxicokinetic study design for degradation products and leachables Part 17: Establishment of allowable limits for leachable substances Part 18: Chemical characterization of materials Part 19: Physico-chemical, morphological and topographical characterization of materials Part 20: Principles and methods for immunotoxicology testing of medical devices Evaluation of biocompatibility of medical devices used in dentistry Clinical investigation of medial devices for human subjects—good clinical practice. Application of risk management to medical devices

ANSI/ADA, American National Standards Institute/American Dental Association (http://www.ansi.org or http://www.ada.org); ASTM-Int, American Society for Testing and Materials International (http://www.astm.org); CEN, Comité Européen de Normalisation (http://www.cenorm.be); ISO, International Organization for Standardization (http://www.iso.org). Other Organizations: EN, European Union (http://europa.eu); FDA, U.S. Food and Drug Administration (http://www.fda.gov/medicaldevices); GHTF, Global Harmonization Task Force (http://www.ghtf.org); NIOM, Nordic Institute for Dental Materials (http://www.niom.no).

information in the name of simplicity and speed, may be compromised by competing interests or disagreements within the standards committees, and may not be relevant to known clinical performance. The discussion in Section 1 is a testament to this last point. The shortcomings of standards and standards organizations are frustrating, but one must only envision a world without them to appreciate their contribution to the public safety and the coordination of efforts among competing interests around the globe. At present, standards appear most useful for providing tests to compare materials [83].

The approval of dental devices for use by regulatory agencies via standardized testing is significantly influenced by the 510(k) or ‘grandfather’ clause of the U.S. FDA. For Class I or Class II devices (see previous discussion, Section 2.2), the FDA allows a manufacturer to avoid the expense and time of a premarket approval process (PMA) if they can show ‘substantial equivalence’ to a device already approved for the same indication. A full discussion of substantial equivalence is beyond the scope of this text, but the process centers around the following questions [28]:

d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 23–40

(a) does the new device have the same intended use as the equivalent device?, (b) does the new device have technological characteristics that raise new types of safety or effectiveness concerns?, and (c) does performance data demonstrate equivalence? It is the third criteria above that involves the standards in Table 4. When applying for a substantial equivalence exemption, manufacturers must provide data to the FDA that will demonstrate equivalence to an existing material; it is up to the manufacturer to select tests and provide data to create a compelling case for granting a 510(k) exemption from the PMA process. It should come as no surprise that manufacturers will seek 510(k) approval for a new dental material whenever possible; this strategy saves substantial sums of money and time for a product to reach the market. And many ‘new’ materials are introduced into dental practice on the basis of the 510(k) exemption. For example, a new dental alloy or composite might be introduced on the basis that its composition differs by only a few percentage points from composites already used in dentistry. A manufacturer would try to show that this compositional change does not alter the safety of the device via data on constituent release, perhaps some in vitro toxicity tests, and a limited demonstration of product utility in practice. This strategy sounds reasonable on paper. Yet, the 510(k) clause causes concern among some, who worry that this clause discourages manufacturers from introducing entirely new materials (which is much more expensive) and that 510(k) approval is an abused mechanism that is a step-wise pathway toward material in-equivalence and higher risks of biological failure [28,94,95].

4. Considerations for the future of biocompatibility testing 4.1. An alternate strategy to assess biological responses to materials Given the complexities, failures, and limitations of current testing strategies to predict biological responses to new materials, alternative strategies are desirable. This section presents one possible strategy, which relies on an initial assessment of the hazard and risk associated with a new material, followed by controlled tests in humans, followed by approval to use in the general marketplace with surveillance. By way of definition, a hazard is defined as the potential for the material to cause harm in a biological context. For example, if a dental alloy corrodes in the oral environment and releases metal ions, the ions are a hazard because nearly all known adverse effects are mediated by released ions [2]. An alloy that corrodes therefore poses a hazard of causing adverse effects. Risk, on the other hand, is defined as the probability that the corrosion of the alloy will have adverse clinical effects. An alloy that corrodes may or may not trigger adverse clinical effects—it depends on the ions released, concentrations, exposure routes, and many other factors. In this example, clinical tests will define the risks of any corrosion of the alloy. The

35

remainder of this section will introduce this alternative strategy for materials testing, using the concepts of hazard and risk (Fig. 5). In the strategy presented in Fig. 5, in vitro tests would play a different role than their role in Fig. 1 of screening in or out a new material prior to animal tests. The in vitro tests in Fig. 5 would help characterize release of any constituents from the material and their hazards and risks. Constituent release might be measured directly (e.g., by various spectrographic methods) or indirectly (e.g., by mass loss). The rationale for this focus is that most adverse effects from materials are mediated by released constituents [2]. The potential importance of the biological effects of released constituents could be measured by cell responses in short-term or long-term cell-culture tests, with or without barriers. All the variables in Table 2 could be used to customize these tests to reflect some aspects of the clinical use of the material, but in the strategy of Fig. 5, the goal of the in vitro test would not be not to predict clinical performance. Rather the goal would be to determine how much mass of the material is labile, the dynamics of any mass release, and the potential for released constituents to alter cellular metabolism or function. Put more succinctly, these tests would estimate the hazards of the material. Used appropriately, in vitro tests could also play a significant role in risk assessment in Fig. 5. Although in vitro tests cannot fully replicate clinical conditions, consideration of the relevant aspects of the tests to clinical conditions can provide valuable insight to clinical risks. A good example occurs with zinc oxide eugenol (ZOE) cements, referenced earlier as markedly cytotoxic to cells in direct contact format. Using in vitro data presented earlier (Section 3.1), one could extrapolate that the risks of using ZOE for direct pulp caps are too high, but where a dentin barrier intervenes, the risk would be low enough to progress to subsequent evaluations in Fig. 5. To some extent this role for in vitro tests is already employed by many experts in the field, and is codified in some countries (e.g., ISO 14971, Table 4). The ability to use in vitro tests for relevant risk assessment in Fig. 5 would depend on the efforts made to model clinical use. In the strategy shown in Fig. 5, animal tests would serve two functions. First, they would supplement the hazard assessment of the in vitro tests. Animal tests such as acute or chronic toxicity from ingestion, carcinogenicity, genotoxicity, or reproductive toxicity (safety-oriented tests) would estimate hazards of the material (see Section 3.2 for more description of animal tests). These tests expose animals to materials at concentrations, by routes of administration, and in contexts that are not relevant to clinical use, but do define hazards of the material. In a parallel track, efforts should be made to update some of these tests (e.g., for mutagenicity) that are in current standards but no longer accepted as sound by many. Second, animal tests could be used to define biological risks via tests for sensitization, pulpal inflammation, bone formation, or other cellular responses (function-oriented tests); animal tests for risk assessment would be customized, as much as possible, to the clinical use of the material using variables in Table 3. As with in vitro tests, the goal of animal tests in this paradigm would not be to screen materials in or out from clinical use, but to assess the hazards and risks of the materials.

36

d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 23–40

Fig. 5 – An alternate paradigm for biological assessment. In vitro and animal tests would be used collectively to assess the hazards and risks of human tests of a new material. Ideally, these tests would be integrated with development of the material (two-way arrow). Once complete, the data would be used by a governmental agency to decide if human testing was justified. If approved for human testing, clinical tests would be used to decide if an introduction to the market is justified. If justified, the material would receive ongoing surveillance for safety or efficacy by a national or international database system. Modified materials would be introduced into this process depending on the risk that modifications pose in the opinion of the governing agency.

After a palette of in vitro and animal tests were completed, a decision would be made if this material should be tested initially in humans (1st decision point, Fig. 5). The palette of in vitro and animal tests would comprise a hazard-risk profile that could be evaluated collectively to make this decision.

In this manner, the limitations and difficulties of having any one in vitro or animal test screen out materials from further testing would be reduced. The decision to proceed could be made by governmental agencies that manage approval of materials now. If the data did not support risking provisional

d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 23–40

human testing, then further in vitro or animal tests could be required to help reformulate the material and strengthen confidence that the hazards and risks are acceptable for human testing. Once the material received provisional approval for human testing, a limited CCT (Section 3.3) could be designed in humans to assess its safety. The CCT could be designed to maximize efficiency and minimize costs. In the paradigm of Fig. 5, it would be this type of test that would serve as the first indicator of human risk during use. In this manner, materials would be assessed for safety primarily in controlled, human tests. If the material performed adequately, then human testing could be extended using a PBRN or other type of clinical test (Section 3.3). Secondary clinical tests would focus on longer-term evaluations and more relevant, more variable clinical techniques used by a practicing clinicians. These tests also would evaluate material performance in patients who might not meet rigid inclusion and exclusion criteria required in CCTs but who constitute a real part of clinical dental practice. After clinical evaluation, a new material would be evaluated by a governing agency for provisional acceptance onto the marketplace (2nd decision point, Fig. 5). The agency would judge the performance of the material using human data, using as needed any in vitro or animal data. But the primary decision would be made using human data from clinical trials. If the material did not appear acceptable, further human, animal, or in vitro tests could be used to prescribed to resolve concerns. After approval for the marketplace, all materials would ideally be evaluated continuously by clinicians using national or international surveillance databases of adverse incident reports; this type of surveillance is already being used effectively in some parts of the world (e.g., the European Union). Problems that were transparent during initial clinical testing might be identified during clinical practice in this manner. Such a system might require clinicians to report adverse incidents to a government agency as a prerequisite for licensure to practice and as an ethical obligation to the profession. Developing problems could be investigated with any of the testing methods previously discussed; government agencies could either withdraw the material or issue warnings during such testing (3rd decision point, Fig. 5). The philosophy of post-marketing surveillance would be that the ‘final’ biocompatibility of a material is never fully determined, requiring ongoing vigilance of the profession. Long-term data for biological safety and efficacy would be monitored during ongoing clinical use and reporting; this type of monitoring is rare for dental materials today, but is emerging in some parts of the world.

4.2.

Perspective, conclusions, final comments

The strategy in Fig. 5 would try to accommodate our long experience that using in vitro or animal tests to accurately predict the clinical biologic responses to dental materials has not been productive. It seems unlikely we can ever identify non-clinical, non-human tests (singly or in combination) to accurately predict the complex interface between a material and human tissues; human testing is likely the only test fully

37

capable of the accurate clinical assessment of biocompatibility. Yet, we are ethically obligated to make every effort to assure that those human tests are as safe as possible, particularly for novel materials or in novel applications. Using in vitro and animal tests to construct a hazard-risk profile prior to human testing eliminates the difficulties of the linear model of Fig. 1, gives us means to try to assure patient safety during human tests, and uses each type of test for its strengths. The difference between the strategy of Figs. 1 and 5 are that in Fig. 5, the tests are used in a collective fashion—no one test carries ‘veto’ power. Fig. 5 also accommodates the emerging trend that future materials will use components released by design to achieve therapeutic endpoints. Release of these components may be misleading in cell-culture tests. For example, the release of fluoride ions from glass ionomers, which are toxic in culture [96,97], may be therapeutic in vivo. In Fig. 5, these ideas could be considered in deciding the appropriateness of human tests at the first decision point. At the second decision point of Fig. 5, the clinical tests could be structured to assess safety, then efficacy, as it done now for drugs. The decision about introducing the materials onto the marketplace would integrate both safety and efficacy data for these types of materials. If a material were not new, but were modified from a formulation already in clinical use, the paradigm of Fig. 5 also could be employed. The manufacturer would present data to support a decision at either the first or second decision points, depending on the modification. The governing agency would use these data to assess what level of risk that the modification poses and dictate where in Fig. 5 clinical testing should begin. Ideally, a modified material would be required to pass PBRNtype testing at a minimum. In this way, no material, new or modified, would reach the market without the scrutiny of controlled human testing of some sort. This approach would be quite different from the FDA’s 510(k) clause in effect at present, where it is possible to introduce a modified material deemed to be ‘substantially equivalent’ into clinical use via in vitro and animal tests alone (Section 3.5). A final trend to consider is the interface between testing for biocompatibility and other properties of the material. At present, biological response is assessed mostly ‘after the fact’, after a material is already fairly far along in its development. Although this strategy is understandable, it casts biological testing into the role of ‘spoiler’ for a material, and pressures manufacturers, marketing departments, and academicians into uncomfortable decisions with huge financial consequences. It would seem to be more desirable to integrate biological testing temporally with the development of the material to the extent possible (two-way arrow at the entry point for materials in Fig. 5). Such a trend seems reasonable, particularly as dental materials inherit therapeutic roles and testing of biological responses evolves from a focus on safety to one on both safety and efficacy. In dentistry, we are blessed with many materials that are used to significantly improve patient health. The current system for predicting and assessing biocompatibility, even with its foibles, moments of un-gracefulness, and difficulties, has guided us in providing materials for safe clinical use. And this system is certainly better than nothing. Yet, we are faced with a revolution of new biomaterials, many of which may

38

d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 23–40

be used to treat oral disease. Materials will become as much therapeutic as structural, and the needs for biological testing will extend beyond safety. As these exciting changes transpire, the demands for accurate and efficient biological assessment of materials will grow, and will likely require that we restructure the way we currently measure and predict biocompatibility.

Acknowledgements The author thanks his mentors, Drs. Robert G. Craig and Tom Hanks, for their training and guidance and introduction to the field of biological testing and biocompatibility so many years ago. He also thanks the Academy of Dental Materials for their invitation to contribute this manuscript, and the many, many individuals in the biocompatibility field, too numerous to mention here—but all important, who have contributed so much to its understanding. Finally, the perspective and suggestions of Dr. Steve Bayne, as well as the Editor and reviewers of Dental Materials are greatly appreciated.

references

[1] Schmalz G. Concepts in biocompatibility testing of dental restorative materials. Clin Oral Investig 1997;1:154–62. [2] Wataha JC. Principles of biocompatibility for dental practitioners. J Prosthet Dent 2001;86:203–9. [3] Autian J. The use of rabbit implants and tissue culture tests for the evaluation of dental materials. Int Dent J 1970;20:481–90. [4] Mjör IA, Hensten-Pettersen A, Skogedal O. Biologic evaluation of filling materials. A comparison of results using cell culture techniques, implantation tests and pulp studies. Int Dent J 1977;27:124–9. [5] Wennberg A, Mjör IA, Hensten-Pettersen A. Biological evaluation of dental restorative materials: a comparison of different test methods. J Biomed Mater Res 1983;17:23–36. [6] Langeland K. Correlation of screening tests to usage tests. J Endod 1978;4:300–3. [7] Spångberg LS. Correlation of in vivo and in vitro screening tests. J Endod 1978;4:296–9. [8] Tronstad L, Wennberg A, Hasselgren G. Screening tests for dental materials. J Endod 1978;4:304–7. [9] Hume WR. A new technique for screening chemical toxicity to the pulp from dental restorative materials and procedures. J Dent Res 1985;64:1322–5. [10] Hanks CT, Craig RG, Syed SA, Adams ER. Dentin ‘filtration’ reduces cytotoxic effects of setting composites in vitro. J Dent Res 1987;66(287). Abstr. No. 1444. [11] Schmalz G, Schuster U, Nuetzel K, Schweikl H. An in vitro pulp chamber with three-dimensional cell-cultures. J Endod 1999;25:24–9. [12] Bouillaguet S, Virgillito M, Wataha J, Ciucchi B, Holz J. The influence of dentine permeability on cytotoxicity of four dentine bonding systems, in vitro. J Oral Rehabil 1998;25:45–51. [13] Wataha JC, Lockwood PE, Nelson SK, Rakich D. In vitro cytotoxicity of dental casting alloys over 8 months. J Oral Rehabil 1999;26:379–87. [14] Demirci M, Killer K-A, Bosl C, Galler K, Schmalz G, Schweikl H. The induction of oxidative stress, cytotoxicity, and genotoxicity by dental adhesives. Dent Mater 2008;24:362–71.

[15] Schweikl H, Spagnuolo G, Schmalz G. Genetic and cellular toxicology of dental resin monomers. J Dent Res 2006;85:870–7. [16] Lewis JB, Messer RLW, Pitts L, Hsu SD, Hansen JM, Wataha JC. Ni(II) ions dysregulate cytokine secretion from human monocytes. J Biomed Mater Res B Appl Biomater 2009;88:358–65. [17] Schweikl H, Schmalz G, Weinmann W. The induction of gene mutations and micronuclei by oxiranes and siloranes in mammalian cells in vitro. J Dent Res 2004;83:17–21. [18] Englemann J, Volk J, Leyhausen G, Geurtsen W. ROS formation and glutathione levels in human oral fibroblasts exposed to TEGDMA and camphorquinone. J Biomed Mater Res B Appl Biomater 2005;75:272–6. [19] Wataha JC, Lockwood PE, Schedle A, Noda M, Bouillaguet S. Ag, Cu, Hg, and Ni ions alter the metabolism of human monocytes during extended low-dose exposures. J Oral Rehabil 2002;29:133–9. [20] Costa CA, Vaerten MA, Edwards CA, Hanks CT. Cytotoxic effects of current dental adhesive systems on immortalized odontoblast cell line MDPC-23. Dent Mater 1999;15:434–41. [21] Boland EJ, MacDougall M, Carnes DL, Dickens SH. In vitro cytotoxicity of a remineralizing resin-based calcium phosphate cement. Dent Mater 2006;22:338–45. [22] Williams DF. On the mechanisms of biocompatibility. Biomaterials 2008;29:2941–53. [23] Ratner BD, Bryant SJ. Biomaterials: where we have been and where we are going. Annu Rev Biomed Eng 2004;6:41–75. [24] Anderson JM. Biological responses to materials. Annu Rev Mater Sci 2001;31:81–110. [25] Ratner BD. Replacing and renewing: synthetic materials, biomimetics, and tissue engineering in implant dentistry. J Dent Educ 2001;65:1340–7. [26] U.S. Food and Drug Administration Website: http://www.fda.gov/MedicalDevices/DeviceRegulationand Guidance/Overview/ClassifyYourDevice/ucm051521.htm. [27] U.S. Food and Drug Administration Website: http://www.fda.gov/MedicalDevices/DeviceRegulationand Guidance/Overview/ClassifyYourDevice/ucm051504.htm. [28] Monsein LH. Primer on medical device regulations: Part II. Regulation of medical devices the by U.S. Food and Drug Administration. Radiology 1997;205:10–8. [29] U.S. Food and Drug Administration Website: http://www.fda.gov/MedicalDevices/DeviceRegulationand Guidance/Overview/ClassifyYourDevice/ucm051530.htm. [30] U.S. Food and Drug Administration Website: http://www.fda.gov/MedicalDevices/DeviceRegulationand Guidance/Overview/ClassifyYourDevice/ucm051549.htm. [31] St. John KR. Biocompatibility of dental materials. Dent Clin North Am 2007;51:747–60. [32] Clarke JH. Dentistry in western civilization 3500 B.C. to 1905 A.D. Oregon Health Sciences University; 1981. p. 63. [33] Clarke JH. Dentistry in western civilization 3500 B.C. to 1905 A.D. Oregon Health Sciences University; 1981. p. 52. [34] Clarke JH. Dentistry in western civilization 3500 B.C. to 1905 A.D. Oregon Health Sciences University; 1981. p. 66. [35] Clarke JH. Dentistry in western civilization 3500 B.C. to 1905 A.D. Oregon Health Sciences University; 1981. p. 76. [36] Byerly WG. Working with the institutional review board. Am J Health Syst Pharm 2009;66:176–84. [37] McCallum JM, Arekere DM, Green L, Katz RV, Rivers BM. Awareness and knowledge of the U.S. Public Health Service syphilis study at Tuskegee: implications for biomedical research. J Health Care Poor Underserved 2006;17: 716–33. [38] Mackert JR, Berglund A. Mercury exposure from dental amalgam fillings: absorbed dose and the potential for adverse health effects. Crit Rev Oral Biol Med 1997;8:410–36.

d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 23–40

[39] Stanley HR. Biological evaluation of dental materials. Int Dent J 1992;42:37–46. [40] Wataha JC. Biocompatibility of dental casting alloys: a review. J Prosthet Dent 2000;83:223–34. [41] Hanks CT, Wataha JC, Sun ZL. In vitro models of biocompatibility: a review. Dent Mater 1996;12:186–93. [42] Kawahara H, Yamagami A, Nakamura M. Biological testing of dental materials by means of tissue culture. Int Dent J 1968;18:443–67. [43] Leirskar J, Helgeland K. A methodologic study of the effect of dental materials on growth and adhesion of animal cells in vitro. Scand J Dent Res 1972;80:120–33. [44] Spångberg L. Kinetic and quantitative evaluation of material cytotoxicity in vitro. Oral Surg Oral Med Oral Pathol 1973;35:389–401. [45] Hensten-Pettersen A, Helgeland K. Evaluation of biologic effects of dental materials using four different cell culture techniques. Scand J Dent Res 1977;85:291–6. [46] Hume WR. An analysis of the release and the diffusion through dentin of eugenol from zinc oxide-eugenol mixtures. J Dent Res 1984;63:881–4. [47] Hanks CT, Craig RG, Makinen PK, Pashley DH. Characterization of the in vitro pulp chamber using cytotoxicity of phenol. J Oral Pathol 1989;18:97–107. [48] Meryon SD, Jakeman KJ. The effect in vitro of zinc released from dental restorative materials. Int Endod J 1985;18: 191–8. [49] International Standards Organization. Biological evaluation of medical devices. Part 1: guidance on selection of tests. ISO 10993-1, 1st ed. 1992-04-15. See the ISO website at: http://www.iso.org/iso/iso catalogue.htm. [50] Hildebrand HF, Veron C, Martin P. Nickel, chromium, cobalt dental alloys and allergic reactions: an overview. Biomaterials 1989;10:545–8. [51] Rapp U, Stiesch M, Reh H, Kapp A, Werfel T. Investigation of contact allergy to dental metals in 206 patients. Contact Dermat 2009;60:339–43. [52] Costa M. Perspectives on the mechanism of nickel carcinogenesis gained from models of in vitro carcinogenesis. Environ Health Perspect 1989;81:73–6. [53] Costa M, Davidson TL, Chen H, Ke Q, Zhang P, Yan Y, et al. Nickel carcinogenesis: epigenetics and hypoxia signaling. Mutat Res 2005;592:79–88. [54] Wylie CM, Schelton RM, Fleming GJP, Davenport AJ. Corrosion of nickel-based dental casting alloys. Dent Mater 2007;23:714–23. [55] Wataha JC, O’Dell NL, Singh BB, Ghazi M, Whitford GM, Lockwood PE. J Biomed Mater Res B Appl Biomater 2001;58:537–44. [56] Schmalz G, Garhammer P. Biological interaction of dental cast alloys with oral tissues. Dent Mater 2002;18:396–406. [57] Wataha JC, Lockwood PE, Noda M, Nelson SK, Mettenburg DJ. Effect of toothbrushing on the toxicity of casting alloys. J Prosthet Dent 2002;87:94–8. [58] Wataha JC, Lockwood PE, Khajotia SS, Turner R. Effect of pH on element release from dental casting alloys. J Prosthet Dent 1998;80:691–8. [59] Hansen JM, Zhang H, Jones DP. Differential oxidation of thioredoxin-1, thioredoxin-2, and glutathione by metal ions. Free Radic Biol Med 2006;40:138–45. [60] Salnikow K, Donald SP, Bruick RK, Zhitkovich A, Phang JM, Kasprazk KS. Depletion of intracelluar ascorbate by the carcinogenic metals nickel and cobalt results in the induction of hypoxic stress. J Biol Chem 2004;279:40337–44. [61] Kasprzak KS, Sunderman FW, Salnikow K. Nickel carcinogenesis. Mutat Res 2003;533:67–97. [62] Li Q, Chen H, Huang X, Costa M. Effects of 12 metal ions on iron regulatory protein 1 (IRP-1) and hypoxia-inducible

[63] [64]

[65]

[66]

[67]

[68]

[69]

[70]

[71]

[72]

[73]

[74] [75]

[76]

[77]

[78]

[79] [80]

39

factor-alpha (HIF-1␣) and HIF-regulation genes. Toxicol Appl Pharmacol 2006;213:245–55. Maxwell P, Salnikow K. HIF: an oxygen and metal responsive transcription factor. Cancer Biol Ther 2004;3:29–35. Jeong H-J, Chung H-S, Lee B-R, Kim S-J, Yoo S-J, Hong S-H, et al. Expression of pro-inflammatory cytokines via HIF-1␣ and NF-␬B activation on desferrioxamine-stimulated HMC-1 cells. Biochem Biophys Res Commun 2003;306:805–11. Lewis JB, Randol TM, Lockwood PE, Wataha JC. Effect of subtoxic concentrations of metal ions on NF␬B activation in THP-1 human monocytes. J Biomed Mater Res A 2003;64:217–24. Lewis JB, Wataha JC, McCLoud V, Lockwood PE, Messer RLW, Tseng W-Y. Au(III), Pd(II), Ni(II) and Hg(II) alter NF␬B signaling in THP1 monocytic cells. J Biomed Mater Res A 2005;74:474–781. Goebeler M, Meinardus-Hager G, Roth J, Goerdt S, Sorg C. Nickel chloride and cobalt chloride, two common contact sensitizers, directly induce expression of intercellular adhesion molecule-1 (ICAM-1), vascular cell adhesion molecule-1 (VCAM-1), and endothelial leukocyte adhesion molecule-1 (ELAM-1) by endothelial cells. J Invest Dermatol 1993;100:759–65. Viemann D, Schmidt M, Tenbrock K, Schmid S, Müller V, Klimmek K, et al. The contact allergen nickel triggers a unique inflammatory and proangiogenic gene expression pattern via activation of NF-␬B and hypoxia-inducible factor 1␣. J Immunol 2007;178:3198–207. Wataha JC, Lewis JB, Volkmann KR, Lockwood PE, Messer RLW, Bouillaguet S. Sublethal concentrations of Au(III), Pd(II), and Ni(II) differentially alter inflammatory cytokine secretion from activated monocytes. J BIomed Mater Res B Appl Biomater 2004;69:11–7. Schmidt M, Raghavan B, Müller V, Vogl T, Fejer G, Tchaptchet S, et al. Crucial role for human Toll-like receptor 4 in the development of contact allergy to nickel. Nat Immunol 2010;11:814–20. Silvennoinen-Kassinen S, Ikäeimo I, Tuiilikainen A. TAP1 and TAP2 genes in nickel allergy. Int Arch Allergy Immunol 1997;114:94–6. Goebeler M, Gutwald J, Roth J, Meinardus-Hager G, Sorg C. Expression of intercellular adhesion molecules-1 in murine allergic contact dermatitis. Int Arch Allergy Appl Immunol 1990;93:294–9. Chen H, Davidson T, Singleton S, Garrick MD, Costa M. Nickel decreases cellular iron level and converts cytosolic aconitase to iron-regulatory protein 1 in A549 cells. Toxicol Appl Pharmacol 2005;206:275–87. Wataha JC, Hanks CT, Craig RG. Uptake of metal cations by fibroblasts in vitro. J Biomed Mater Res 1993;27:227–32. Edwards DL, Wataha JC, Hanks CT. Uptake and reversibility of uptake of nickel by human macrophages. J Oral Rehabil 1998;25:2–7. Bridgewater JD, Lim J, Vachet RW. Transition metal-peptide binding studied by metal-catalyzed oxidation reactions and mass spectrometry. Anal Chem 2006;78:2432–8. Pande V, Sharma RK, Inoue J, Otsuka M, Ramos MJ. A molecular modeling study of inhibitors of nuclear factor kappa-B(p60)—DNA binding. J Comput Aided Mol Des 2003;17:825–36. Hendry LB, Rueggeberg RA, Davis III RE. Endocrine activity of bisphenol-A evaluated by computer modeling. J Dent Res 1997;297(76):236. Abstr. No. 1781. Holder AJ, Kilway KV. Rational design of dental materials using computational chemistry. Dent Mater 2005;21:47–55. Palmquist A, Emauelsson L, Branemark R, Thomsen P. Biomechanical, histological and ultrastructural analyses of laser micro- and nano-structured titanium implant after 6

40

[81]

[82]

[83] [84]

[85] [86]

[87] [88]

[89]

d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 23–40

months in rabbit. J Biomed Mater Res B Appl Biomater 2011;97:289–98. Garhammer P, Schmalz G, Hiller KA, Reitinger T. Metal content of biopsies adjacent to dental cast alloys. Clin Oral Investig 2003;7:92–7. Mozaffari MS, Abdelsayed R, Liu JY, Wimborne H, El-Remessy A, El-Maraby A. Effects of chromium picolinate on glycemic control and kidney of the obese Zuker rat. Nutr Metab 2009;6:51, doi:10.1186/1743-7076-6-51. Mjör IA. Minimum requirement for new dental materials. J Oral Rehabil 2007;34:907–12. Bayne SC. Dental restoration for oral rehabilitation-testing of laboratory properties versus clinical performance for clinical decision making. J Oral Rehabil 2007;34:921–32. Cochrane Collaboration Website: http://www.cochrane.org. Mjör I. Controlled clinical trials and practice-based research in dentistry. J Dent Res 2008;87:605 [see also J Dent Res 2008;87:800, for comment and author response]. Mjör IA. Practice-based dental research. J Oral Rehabil 2007;34:913–20. DeRouen TA, Cunha-Cruz J, Hilton TJ, Ferracane J, Berg J, Zhou L, et al. What’s in a dental practice-based research network? J Am Dent Assoc 2010;141:889–99. Cvar JF, Ryge G. Reprint of criteria for the clinical evaluation of dental restorative materials. Clin Oral Investig 2005:215–32.

[90] Hickel R, Peschke A, Tyas M, Mjör I, Bayne S, Peters M, et al. FDI World Dental Federation: clinical criteria for the evaluation of direct and indirect restorations – update and clinical examples. Clin Oral Investig 2010;14:349–66. [91] Yamamoto A, Kohyama Y, Hanawa T. Mutagenicity evaluation of forty-one metal salts by the UMU test. J Biomed Mater Res 2002;59:176–83. [92] Messer RLW, Lockwood PE, Wataha JC, Lewis JB, Norris S, Bouillaguet S. In vitro cytotoxicity of traditional versus contemporary dental ceramics. J Prosthet Dent 2003;90:452–8. [93] American Dental Association. ADA specification #41 for recommended standard practices for biological evaluation of dental materials. J Am Dent Assoc 1979;99:697–8. [94] Alexander H. The editor’s corner: substantially equivalent to what? J Appl Biomater 1994;5:375. [95] Baier RE. Biomaterials applicability: establishing suitable ‘materials equivalency’ protocols. J Appl Biomater 1994;5:377–8. [96] Geursten W. Substances released from dental resin composites and glass ionomer cements. Eur J Oral Sci 1998;106:687–95. [97] Stanislawski L, Daniau X, Lautié A, Goldberg M. Factors responsible for pulp cell cytotoxicity induced by resin-modified glass ionomer cements. J Biomed Mater Res 1999;48:277–88.