Recommendation for a non-animal alternative to rat caries testing

Transcription

1 Position Article Recommendation for a non-animal alternative to rat caries testing JOHN D.B. FEATHERSTONE, MSC, PHD, GEORGE K. STOOKEY, PHD, MICHAEL A. KAMINSKI, PHD & ROBERT V. FALLER, BS ABSTRACT: Purpose: As a requirement of the Food & Drug Administration s final monograph on Anticaries drug products for over-the-counter human use, the toothpaste industry has been conducting animal caries tests on every fluoride-containing toothpaste introduced into the U.S. market since The practice of testing in animals, although required by law, is in stark conflict with the corporate policy of many U.S. and global toothpaste manufacturers, in which, if possible, alternatives to animal testing are utilized. A provision does exist within the regulation which allows the use of an alternative method to demonstrate efficacy. However, to take advantage of this provision, a petition must be submitted to the FDA and in this petition data demonstrating the alternative provides results of equivalent accuracy must be included. After many years of research, model development and model comparisons, we have identified one particular laboratory model that demonstrated excellent correlation with the currently accepted animal caries models. This model, known as the Featherstone ph cycling model, is discussed in this paper. Methods: The Featherstone ph cycling model has been shown to produce results of equivalent accuracy to the animal caries model by: (1) demonstrating a clinically relevant fluoride dose response similar to that shown in the animal caries model (including 1100 ppm F, 250 ppm F and placebo); (2) demonstrating similar results to the animal caries model for clinically proven dentifrice formulations relative to positive and negative controls; (3) demonstrating discriminating ability in strong agreement with the animal caries model for differentiating between a dentifrice formulation with attenuated fluoride activity and a USP standard; and (4) providing a clinically relevant representation of the caries process, as demonstrated by orthodontic banding studies. In addition, the model sufficiently addresses both salivary and abrasive/anticalculus agent interference concerns. Results: For more than 50 years, fluoride has been the first defense in the fight against dental caries. The clinical effectiveness of fluoride is well accepted and documented extensively in the literature. The mechanism through which fluoride provides its benefit is very straightforward and well understood. The proposed laboratory model effectively simulates the effect of the caries process and has been shown to demonstrate equivalent accuracy to animal caries. This indicates that there are strong scientific grounds for the use of this model as an alternative to the animal caries test. Based on the strength of the data and the correlations noted between the two models, we recommend that the scientific community and the toothpaste industry broadly accept the Featherstone laboratory ph cycling model as an appropriate alternative to animal testing, particularly for ionic fluoride based dentifrices. (Am J Dent 2011;24: ). CLINICAL SIGNIFICANCE: The proposed Featherstone ph cycling laboratory model effectively simulated the effect of the caries process and has been shown to demonstrate equivalent accuracy to animal caries. This indicated that there are strong scientific grounds for the use of this model as an alternative to the animal caries test. : Dr. John D. B. Featherstone, School of Dentistry, University of California San Francisco, San Francisco, CA 94143, USA. E- : john.featherstone@ucsf.edu Introduction In 1995, the U.S. Food and Drug Administration (FDA) issued the Anticaries Drug Products for Over-the-Counter (OTC) Human Use final rule, establishing the requirement that all OTC anticaries dentifrice drug product formulations be tested in the animal caries reduction test. 1 Although the FDA stated that the other tests included in the monograph (fluoride availability, in vitro enamel solubility reduction and in vitro fluoride uptake) were good predictors of potential effectiveness, the agency believed at that time that the in vivo animal test provided additional assurance of anticaries efficacy. FDA s agreement that the overall performance testing requirements set forth in the monograph provide a reasonable expectation of effectiveness has allowed industry to develop and market new product formulations that are more consumer appealing. This is important because dental caries still remains a problem of global concern, and development of new formulations designed to enhance consumer use is fundamental for the successful delivery of long-term anticaries benefits. Maintaining the standard of a reasonable expectation of effectiveness is a key component of the monograph system, and it should be a primary consideration in the development of any model designed to supplement or replace those models already accepted. In addition to the laboratory and animal testing requirements of the final anticaries monograph, some companies have incorporated other models, such as ph cycling models, in the development of new dentifrice technologies. 2-5 Historically, many companies have taken a tiered approach to testing. The first tier involves testing a new formulation in routine chemical availability and stability tests. If positive results are obtained, the formulation is then placed into a second level. This second level would include the simple fluoride availability tests required in the anticaries monograph, such as laboratory enamel solubility reduction testing (ESR) 6 or fluoride (F) uptake testing. 7 With products showing success after these tests, a more involved test would be run (ph cycling) to provide confidence that the formulation is active under more robust testing conditions. This approach would then culminate in a confirmatory animal caries reduction test to satisfy monograph requirements prior to marketing. Our experience with this approach has provided the key finding that all monograph level fluoride dentifrice formulations that pass all three levels of testing will also pass

2 290 Featherstone et al American Journal of Dentistry, Vol. 24, No. 5, October, 2011 the animal caries reduction test. 8 As a result, animal caries testing has now become a confirmatory effort to meet the Caries Monograph marketing requirement for routine monograph fluoride dentifrice formulas, rather than a vital test to confirm anticaries effectiveness. Animal caries reduction models have a long history of successful use in the dentifrice industry. Multi-center testing has confirmed that different laboratories with varying animal models produce similar observed results. 9,10 As such, they have served as a standard to compare and validate new laboratory methods designed to mimic clinical performance of fluoride containing products. New information strongly suggests the continued requirement for conducting routine animal tests is unnecessary. After many years of research, model development, and model comparisons, we have identified one particularly robust laboratory ph cycling model that demonstrated excellent correlation with the currently accepted animal caries models. This model, known as the Featherstone ph cycling model, 4 is discussed in this paper. The authors present an argument against the continued use of animal models for confirmation of product effectiveness, as is currently required, based not only on their many years of experience with the models involved, but also a deep understanding of the caries process and ways to effectively mimic this process in the laboratory. We have confirmed a highly significant correlation between the performances of fluoride containing products in the FDA required animal models 11 and the Featherstone 4 laboratory ph cycling model. Based on the strength of the data and the correlations noted between the two models described in this paper, we recommend broad acceptance of the Featherstone laboratory ph cycling model as a viable alternative to animal caries reduction testing, particularly for ionic fluoride-based dentifrices. Materials and Methods THE CARIES PROCESS AND THE USE OF PH CYCLING MODELS It is generally accepted that the two most important factors of fluoride action in preventing or reversing caries are: (1) the inhibition of demineralization, and (2) the enhancement of remineralization Wide recognition of these key mechanisms has led researchers to develop experiments in which both deand remineralization processes are taken into consideration. Laboratory test methods play a major role in fluoridated dentifrice profile testing. These tests are designed to measure the availability of fluoride in solution, 18 the effect of fluoride on enamel solubility reduction, 6 the uptake of fluoride into demineralized enamel 7 and the ability of fluoride to physically and chemically alter tooth mineral content Laboratory methods that once measured only simple mechanistic activity (e.g. analysis of fluoride availability, fluoride uptake and enamel solubility reduction) have evolved into protocols capable of determining the ability of a dentifrice formulation to result in the loss or gain of mineral (demineralization and remineralization). These more recent state-of-the-art laboratory experiments are often referred to as ph cycling methods. 4,16,19-21 The term ph cycling is used to describe laboratory experiments in which either sound teeth or teeth containing artificially induced caries lesions are exposed to multiple denti- frice treatments, periods of demineralization and periods of remineralization. This is in contrast to the earlier methods that utilized only single exposures to product treatment. Demineralization periods, extended periods of acid exposure in the ph cycling protocols, are designed to mimic the destructive effects of acid challenge to the tooth surface and subsurface that occurs clinically, while remineralization periods mimic the protective and repairing effects of saliva. These methods are quite sensitive to fluoride dose (i.e., concentration), showing a distinct separation between the clinically proven, positive control and a fluoride-free, negative control. 4,12,19-21 Clearly, the most meaningful ph cycling experiment is one in which specimens receiving a fluoride-free control dentifrice treatment progress toward cavitation. This ensures the test dentifrice will prevent cavities by: (1) enhancing the naturally occurring remineralization process, and (2) simultaneously providing protection against the inevitable acid attack. Models such as this are often referred to as lesion progression models because specimens are included in the study that are treated with a fluoride-free control or low-dose fluoride treatment and progress toward cavitation, which has been shown to occur clinically. The Featherstone ph cycling model falls into this particular category. 4,12,19,20,22 THE NEED FOR A DEMONSTRATION OF EQUIVALENT ACCURACY In the anticaries monograph, the FDA allows alternative methods to be considered with no specific limits other than providing equivalent accuracy to existing models. 1 It is our collective position that a demonstration of equivalent accuracy includes the following four criteria: 1. Fluoride dose response. First, the alternate model should show a response to fluoride dose with clear separations in fluoride levels previously shown to provide different levels of anticaries efficacy in the animal caries models (i.e., 1100 ppm F > 250 ppm F > 0 ppm F). 2. Correlation with clinically-proven formulations. Second, the alternate model should be capable of showing that clinically proven formulations perform similarly, relative to controls, as do the animal caries models. 3. Ability to discriminate products with attenuated effectiveness in animal caries testing. Third, the alternate model should also be able to show results similar to the animal caries results for formulations that show attenuated fluoride activity. 4. Meaningful representation of caries processes. Lastly, in addition to simply showing the same results as the animal caries model (equivalent accuracy), it is prudent to demonstrate that the alternate model also provides a meaningful representation of the caries process, in correlation with known clinical results. The data that demonstrate fulfillment of the above four requirements by the Featherstone ph cycling model are outlined in the following sections. FLUORIDE DOSE RESPONSE Results and Discussion The Featherstone ph cycling model has been described in detail in several publications and exhibited a clear response to fluoride dose. 23 Featherstone and Zero 22 further demonstrated

3 American Journal of Dentistry, Vol. 24, No. 5, October, 2011 Table 1. Featherstone ph cycling results for fluoride dose response. Mean ΔZ ± SE 2800 ppm F 491 ± 80 a* 1100 ppm F 1039 ± 192 b 250 ppm F 2334 ± 179 c 0 ppm F (placebo) 3381 ± 331 d * Mean Delta Z (µm x volume % min ± SE) values with different letter designations are significantly different P< 0.05) by ANOVA/Newman Keuls. Table 3. Featherstone ph cycling results for clinically proven formulas. Mean ΔZ ± SE 5.0% ppi/1100 ppm F as NaF 572 ± 122 a* 4.0% ppi/1100 ppm F as NaF 270 ± 212 a 3.3% ppi/1100 ppm F as NaF 465 ± 167 a 1100 ppmf as NaF 785 ± 512 a Placebo (0 ppm F) dentifrice 3213 ± 337 b no treatment 2806 ± 558 b * Mean ΔZ (µm x vol % min ± SE) values with different letter designations are significantly different (P< 0.05) by ANOVA/Newman Keuls. Data reproduced with the permission of the author, and by permission of Oxford University Press. 19 Mean Delta Z in ph Cycling (Seven studies) Delta Z by Animal Caries Score Mean Animal Caries Score (Four studies) Figure. Featherstone ph cycling as a function of animal caries study results. that the model was capable of exhibiting a fluoride dose response not only at high fluoride concentrations, similar to those experienced during tooth brushing, but also at lower fluoride concentrations, similar to the level of fluoride in saliva between tooth brushings. Tables 1 and 2 show a direct comparison of results obtained using the Featherstone ph cycling model and an FDA recognized animal caries model. In both experiments, the fluoride dose was varied, and in each case clear separation in response values was noted over the range of ppm F. In addition to demonstrating significant differences over a wide range of fluoride doses, the response in the Featherstone ph cycling model (mean lesion severity ± standard deviation) plotted as a function of the response in the animal caries model (mean total caries score ± standard deviation) demonstrated a positive correlation between the two models. The results of seven individual Featherstone ph cycling studies and three animal caries model studies were used to create a historical demonstration of this positive correlation (Figure). The mean values are plotted with error bars representing standard deviation between studies for both animal caries results (x-axis) and ph cycling results (y-axis). Only those studies that included placebo, 250 ppm F, and 1100 ppm F test groups are included Alternatives to animal caries testing 291 Table 2. Animal caries results for fluoride dose response. % Reduction vs. placebo 2800 ppm F 62 a* 1450 ppm F 49 b 1100 ppm F 47 b 500 ppm F 24 c 250 ppm F 14 c 0 ppm F (placebo) -- d * Groups with different letter designations are significantly different (P< 0.05) by ANOVA/Newman Keuls. Table 4. Animal caries results for clinically proven formulas. % Reduction vs. placebo 5.0% pyrophosphate/1100 ppm F as NaF 62 a* 4.0% pyrophosphate/1100 ppm F as NaF 61 a 3.3% pyrophosphate/1100 ppm F as NaF 53 a 1100 ppm F as NaF 58 a Placebo (0 ppm F) dentifrice -- b *Mean caries scores with different letter designations are significantly different (P< 0.05) by ANOVA/Newman Keuls. Data reproduced with the permission of the author, and by permission of Oxford University Press. 19 Placebo 250 ppm F 1100 ppm F in this figure. Data points above 1100 ppm F (i.e ppm) were not included in the correlation because they are above the levels allowed in the US FDA monograph. The Featherstone ph cycling model exhibited a response to fluoride dose similar to the dose response seen in the animal caries model. Both models are capable of clearly separating 1100 ppm F from 250 ppm F and 0 ppm F placebo. When the historical response to a fluoride dose for each of these models is compared, the correlation is quite strong. Based on the variances and response to dose observed in Fig. 1, it is reasonable to conclude that the Featherstone ph cycling model demonstrated equivalent accuracy to the animal caries model. PERFORMANCE OF CLINICALLY PROVEN FORMULATIONS The previous section showed that results using the Featherstone ph cycling model correlate with findings from animal testing models. In addition, studies show that laboratory results from the Featherstone ph cycling model correlate with findings from human clinical trials. For example, the anticaries effectiveness of an anticalculus dentifrice established in a human clinical trial 24 was confirmed using the Featherstone ph cycling model and an FDA recognized animal caries model study. 19 Specifically, the Featherstone ph cycling model was used to confirm the results of human clinical trials for two proven Crest a formulas [Advanced Formula Crest (0.243% NaF) and Crest Tartar Control (0.243% NaF/3.3% pyrophosphate)]. Consistent with the human study, both formulas were found to be significantly better at inhibiting caries initiation and progression than the placebo dentrifrice using the laboratory Featherstone ph cycling model (Table 3). These findings were also confirmed using the animal caries model (Table 4).

4 292 Featherstone et al Table 5. Featherstone ph cycling results for clinically proven formulas. Enamel Roots ΔZ (SD) ΔZ (SD) 1100 ppm F, stabilized SnF 2 /silica 520 ± 160 a* 934 ± 319 a* 1100 ppm F, NaF/silica 861 ± 593 a 918 ± 350 a 1000 ppm F, unstabilized SnF 2 /silica 1165 ± 459 b 1379 ± 350 b 250 ppm F 1574 ± 788 b 1461 ± 650 b Placebo (0 ppm F) 4517 ± 1871 c 2150 ± 492 c * Mean ΔZ (µm x vol % min ± SE) values with different letter designations are significantly different (P< 0.05) by ANOVA/Newman Keuls. Data reproduced with the permission of the author, and The Journal of Clinical Dentistry. 20 Table 7. Featherstone ph cycling results for product with attenuated effectiveness. ΔZ (SEM) 2800 ppm F 1593 ± 112 a* 1100 ppm F 1870 ± 143 a Marketed remineralizing toothpaste 2733 ± 164 b 250 ppm F 2776 ± 292 b Placebo (0 ppm F) 3833 ± 418 c * Mean ΔZ (µm x vol % min ± SEM) values with different letter designations are significantly different (P< 0.05) by ANOVA/Newman Keuls. Data reproduced with the permission of the author and SAGE Publications, Inc. 27 Two experimental test products were also included in the laboratory and animal testing. As shown in Tables 3 and 4, the results for the experimental tartar control variants with 4.0% and 5.0% pyrophosphate also showed similar performance in the Featherstone ph cycling model and the animal caries model. The Featherstone ph cycling model has also been shown to reflect clinical differences measured for different fluoride species. In human clinical studies, Zacherl 25 and Beiswanger et al 26 directly compared the original SnF 2 /calcium pyrophosphate Crest formulation (unstabilized SnF 2 ) with the currently marketed NaF formula. The NaF formulation was shown to provide ~17% benefit over the unstabilized SnF 2 /calcium pyrophosphate formulation in anticaries efficacy, as measured by DMFS (Decayed, Missing and Filled Surfaces) scores. The Featherstone ph cycling model was able to demonstrate a similar outcome, separating the earlier unstabilized SnF 2 formulation from the improved NaF/silica formulations. 20 Although the ph cycling experiment used a different SnF 2 /abrasive combination than the Zacherl 25 and Beiswanger et al 26 studies, both studies demonstrated that an unstabilized SnF 2 formulation was inferior to a NaF/silica formulation. In addition to showing a correlation with human clinical results, Faller et al 20 again demonstrated the ability of the Featherstone ph cycling model to mimic results of the animal caries model. That is, the unstabilized SnF 2 formulations were shown to exhibit lower anticaries efficacy in both the Featherstone ph cycling model and the rat caries model (Tables 5,6). The above studies demonstrated that the Featherstone ph cycling model provided results of anticaries effectiveness which were consistent with human clinical studies and the animal caries model. ABILITY TO DISCRIMINATE PRODUCTS WITH ATTENUATED FLUORIDE ACTIVITY An alternative model to animal caries reduction testing should also be capable of discriminating between formulations that meet and those that do not meet the Anticaries Monograph American Journal of Dentistry, Vol. 24, No. 5, October, 2011 Table 6. Animal caries results for proven formulations (two independent studies). Study 1 Study 2 % Reduction % Reduction vs. placebo vs. placebo 1100 ppm F, stabilized SnF 2 /silica 48 a* 39 a* 1100 ppm F, NaF/silica 42 a 37 a 1000 ppm F, unstabilized SnF 2 /calpyro 25 b 12 b Placebo (0 ppm F) -- c -- c * s with different letter designations are significantly different (P< 0.05), ANOVA/Newman Keuls. These data were reproduced with the permission of the author and and The Journal of Clinical Dentistry. 20 Table 8. Animal caries results for product with attenuated effectiveness. % Reduction vs. placebo 1450 ppm F 53 a* 1100 ppm F 51 a Marketed remineralizing toothpaste 29 b* 250 ppm F 10 c Placebo (0 ppm F) -- c * Mean caries scores with different letter designations are significantly different (P< 0.05), ANOVA/Newman-Keuls. Data reproduced with the permission of the author, and The Journal of Clinical Dentistry. 28 requirements based on the animal caries models. This was demonstrated for the Featherstone ph cycling model in a study of a marketed remineralizing dentifrice. 27 In this case, the remineralizing dentifrice was designed to simultaneously provide a monograph level of sodium fluoride in addition to soluble calcium and phosphate ions in order to maximize remineralization. Researchers evaluated the dentifrice formulation in several experiments to determine if the formulation would, in fact, provide a benefit above and beyond that of the appropriate USP reference standard dentifrice. The remineralizing dentifrice was also compared to clinically tested fluoride dentifrices and a placebo control. Results from the Featherstone ph cycling model demonstrated that the remineralizing dentifrice performed significantly lower than 1100 ppm F (USP reference standard) and was comparable to that of a 250 ppm F product (Table 7). The compromised level of efficacy measured for the remineralizing toothpaste in the Featherstone ph cycling model was confirmed 28 when this same product was included in the animal caries model (Table 8). Overall, results in Tables 7 and 8 illustrate that both models clearly separate the lower fluoride dose controls from the USP reference standard (1100 ppm F). In addition, the remineralizing toothpaste was shown to perform similarly to the 250 ppm F control formulation and significantly lower than the USP reference standard (1100 ppm F NaF/silica) in both the laboratory Featherstone ph cycling model and the animal-based testing model. It should be noted that these results for the remineralizing dentifrice were in conflict with several publications in the literature. Zero et al 29 published work that indicated the remineralization response of the remineralizing toothpaste was equal to that of the USP reference standard. This conclusion was based on results obtained using a short-term intra-oral appliance (IOA) model. This model uses surface microhardness techniques to evaluate surface softened human and bovine enamel substrates after being subjected to treatments. Findings from the IOA

5 American Journal of Dentistry, Vol. 24, No. 5, October, 2011 Alternatives to animal caries testing 293 model are in conflict with the responses observed in both the FDA recognized animal caries model 28 and the Featherstone ph cycling model. 27 It has been hypothesized that the surface microhardness measurement in the IOA model was biased by calcium precipitates on the surface of the specimen which were found to be clinically irrelevant in the monograph animal caries and the Featherstone ph cycling models; both of which measure a product s effect on the more clinically relevant subsurface lesion itself. This example clearly illustrates the danger of relying on the results from laboratory models that have not been validated across a broad range of products and correlated with accepted animal testing models and/or human clinical studies. Results presented here clearly demonstrated that the Featherstone ph cycling model was capable of discriminating products with attenuated fluoride activity which have been confirmed using the FDA-approved animal caries reduction model. These results further support the fact that the Featherstone ph cycling model meets the monograph requirement for demonstrating equivalent accuracy to the animal caries reduction test. MEANINGFUL REPRESENTATION OF CARIES PROCESSES In human orthodontic banding/bracket studies, O Reilly & Featherstone 30 studied the amount of demineralization that occurs around orthodontic appliances and the ability of commercially available fluoride products to inhibit or reverse this demineralization process. They confirmed that demineralization was occurring around these orthodontic appliances and that the extent of demineralization could, in fact, be quantified; also, the demineralization could be completely inhibited and/or reversed with regular use of commercially available fluoride products. Preliminary results from the clinical orthodontic banding studies, as well as several other experiments, lead Featherstone et al 4 to develop a laboratory ph cycling experiment capable of producing lesion conditions quantitatively similar to those found in the clinical banding studies. This particular ph cycling model was developed to mimic the results found clinically during 1 month of demineralization around orthodontic appliances. 30 Specifically, the change in mineral values (ΔZ) seen in the orthodontic banding studies was reproduced in the ph cycling model. That is, the 14-day treatment phase in the ph cycling model resulted in ΔZ values similar to those seen in the 1 month clinical orthodontic banding studies. The clinical study that led to the creation of this ph cycling model was repeated. In the more recent study, 31 premolars had brackets bonded with either a fluoride free composite resin or a fluoride-releasing glass-ionomer. Both groups were given 1100 ppm F (NaF) dentifrice to use twice daily ad lib. The use of the resin and ionomer (bonding agents) was considered to be comparable to the use of metal brackets in the earlier O Reilly & Featherstone 30 study. Gorton & Featherstone 31 again showed the use of a fluoride dentifrice alone for 1 month resulted in significantly more demineralization around the fluoride-free composite used as the control bonding agent (ΔZ = 805 ± 98 vol% x µm) versus the fluoride-releasing glass-ionomer bonding agent (ΔZ = 160 ± 101) and was comparable to previous results. The Featherstone ph cycling model was developed to pro- vide a meaningful representation of the caries process. These results show that the model mimics demineralization occurring around orthodontic brackets in human mouths, which has been shown to occur clinically in orthodontic patients. 32 QUESTIONS PERTAINING TO ABRASIVES AND ANTICALCULUS AGENTS Although the FDA stated that the other tests included in the monograph (fluoride availability and laboratory enamel solubility reduction and laboratory fluoride uptake) were good predictors of potential effectiveness, the agency believed at that time that the animal test was needed to provide additional assurance of anticaries efficacy. In the preamble to the Anticaries Final Monograph, the FDA expressed concern that the chemical and laboratory tests available at that time might not be capable of detecting the reduced availability of fluoride ions in situ that could arise due to introduction of new abrasives and anticalculus agents in dentifrice formulations. It was stated: These chemical tests may not always reflect the true anticaries effectiveness of fluoride dentifrices with or without additives in situ when diluted in the mouth by saliva or exposed to the subtle reactions between dentifrice ingredients and salivary components. Although these in vitro tests may show positive results that are predictive of anticaries activity, during actual use in the mouth, the product may not provide the same amount of anticaries effectiveness. 1 At the time, the available laboratory performance tests included in the final monograph represented early and relatively simplistic methodologies based on single treatments to test availability and activity of fluoride in dentifrice products. Since 1995, more sophisticated laboratory models have been developed that incorporate multiple treatments and that realistically and accurately reflect the effect of fluoride on dental enamel exposed to the acid challenge and salivary buffering cycles that lead to the formation of caries in the oral cavity. By comparing the results from improved laboratory models to those from the required animal testing, researchers have been able to evaluate the predictability of their models. Specifically, the Featherstone ph cycling model involves exposure to multiple dentifrice treatments, periods of simulated saliva interaction using calcium phosphate solutions, and periods of demineralization. Because the activity of fluoride against dental caries is primarily governed by the physical chemical reactions of fluoride, calcium and phosphate within the tooth, an appropriately designed ph cycling model is expected to be capable of detecting any chemical or physical surface interferences that may occur between fluoride and abrasives and/or anticalculus agents. Results presented here clearly demonstrated that the Featherstone ph cycling model was capable of providing results consistent with those obtained using the animal caries reduction model and human clinical studies across a broad range of dentifrice formulations. In conclusion, the Featherstone ph cycling model has been shown to produce equivalent accuracy to the animal caries model by: (1) demonstrating a clinically relevant fluoride dose response similar to the dose response shown in the animal caries model (including 1100 ppm F, 250 ppm F and placebo); (2) demonstrating similar results to the animal caries model for clinically proven dentifrice formulations relative to positive and

6 294 Featherstone et al American Journal of Dentistry, Vol. 24, No. 5, October, 2011 negative controls; (3) demonstrating discriminating ability similar to the animal caries model for differentiating between a dentifrice formulation with attenuated fluoride activity and a USP standard; and (4) providing a clinically relevant representation of the caries process, as demonstrated by orthodontic banding/bracketing studies. In addition, the model sufficiently addresses both salivary and abrasive/anticalculus agent interference concerns. Based on the wealth of available data, comparing results for a wide range of both effective and noneffective products tested in both the FDA accepted animal caries reduction models and the Featherstone ph cycling model, we recommend broad acceptance of the Featherstone model as a viable alternative to current FDA animal testing requirements. a. Procter & Gamble, Cincinnati, OH, USA. Acknowledgements: To Dr. Anita Guy for assistance with the manuscript and its development. Disclosure statement: Dr. Featherstone is the developer of the Featherstone ph cycling model. Dr. Stookey declared no conflict of interest. Dr. Kaminski is and Mr. Faller was employed by the Procter & Gamble Company. No conflicts of interest were stated. Dr. Featherstone is Dean and Professor in the Department of Preventive and Restorative Dental Sciences in the School of Dentistry at the University of California San Francisco, San Francisco, California, USA. Dr. Stookey is President and CEO of Therametric Technologies, Inc. in Indianapolis, Indiana, USA. Dr. Kaminski is a Senior Scientist at The Procter & Gamble Company in the Health Care Research Center in Mason, Ohio, USA. Mr. Faller is a retired Principal Scientist at The Procter & Gamble Company in the Health Care Research Center in Mason, Ohio, USA. References 1. Food and Drug Administration. Final monograph on anticaries drug products for over-the-counter human use. Federal Register 1995;60: ten Cate JM, Duijsters PPE. Demineralization and remineralization of artificial enamel lesions. Caries Res 1982;16: ten Cate JM, Shariati M, Featherstone JDB. Enhancement of salivary remineralization by 'dipping' solutions. Caries Res 1985;19: Featherstone JDB, O'Reilly MM, Shariati M, Brugler S. Enhancement of remineralization in vitro and in vivo. In: Leach SA. Factors relating to demineralisation and remineralisation of the teeth. Oxford: IRL Press Ltd, 1986; White DJ. The application of in vitro models to research on demineralization and remineralization of the teeth. Adv Dent Res 1995;9: Food and Drug Administration. Biological testing procedures for fluoride dentifrices. Food and Drug Administration Docket No. 80N-0042: Test Methods Food and Drug Administration. Biological testing procedures for fluoride dentifrices. Food and Drug Administration Docket No. 80N-0042: Test Methods Pfarrer AM, Kaminski MA. Revisiting the need for biological testing of fluoride-containing dentifrices. Caries Res 2006;40:(Abstr 107). 9. Larson RH, Amsblaugh SM, Navia JM, Rosen S, Schuster GS, Shaw JH. Collaborative evaluation of a rat caries model in six laboratories. J Dent Res 1977;56: Stookey GK, Warrick JM, Miller LL, Greene AL. Animal caries models for evaluating fluoride dentifrices. Adv Dent Res 1995;9: Food and Drug Administration. Biological testing procedures for fluoride dentifrices. Food and Drug Administration Docket No. 80N-0042: Test Method Featherstone JD. Remineralization of artificial carious lesions in vivo and in vitro. In: Leach SA, Edgar WM. Demineralization and remineralization of the teeth. Oxford: IRL Press, Ltd, 1983; Arends J, Gelhard T. Enamel remineralization in vivo. In: Fearnhead RW, Suga S. Tooth enamel IV. Amsterdam: Elsevier Science Publishers, 1984; White DJ, Featherstone JDB. A longitudinal microhardness analysis of fluoride dentifrice effects on lesion progression in vitro. Caries Res 1987;21: White DJ. Reactivity of fluoride dentifrices with artificial caries. I. Effects on early lesions: F uptake, surface hardening and remineralization. Caries Res 1987;21: White DJ. Reactivity of fluoride dentifrices with artificial caries. II. Effects on subsurface lesions: F uptake, F distribution, surface hardening and remineralization. Caries Res 1988;22: ten Cate JM. In vitro studies on the effects of fluoride on deremineralization. J Dent Res 1990;69: Food and Drug Administration. Biological testing procedures for fluoride dentifrices. Food and Drug Administration Docket No. 80N-0042: Methods Featherstone JDB, White DJ, Bowman WD, Shariati M, Stookey GK, Lo Re M, Nancollas GH. Effects of anticalculus dentifrices on caries processes: In vitro and in vivo studies. In: ten Cate JM. Recent advances in the study of dental calculus. Oxford: IRL Press, Ltd., 1989: Faller RV, Best JM, Featherstone JDB. Barrett-Vespone N. Anticaries efficacy of an improved stannous fluoride toothpaste. J Clin Dent 1995;6 (Sp Is): Faller RV, Pfarrer AM, Eversole SL, Cox ER, Landrigan WF, Wang Q. The comparative anticaries efficacy of Crest toothpaste relative to some marketed Chinese toothpastes. Results of in vitro ph cycling testing. Int Dent J 1997;47: Featherstone JDB, Zero DT. Laboratory and human studies to elucidate the mechanism of action of fluoride-containing dentifrices. In: Embery G, Rølla G. Clinical and biological aspects of dentifrices. Oxford: University Press, 1992; Featherstone JDB. Glena R, Shariati M, Shields CP. Dependence of in vitro demineralization of apatite and remineralization of dental enamel on fluoride concentration. J Dent Res 1990;69: Lu KH, Yen DJC, Zacherl WA, Ruhlman CD, Sturzenberger OP, Lehnhoff RW. The effect on dental caries of a fluoride dentifrice containing an anticalculus agent. J Dent Child 1985;52: Zacherl WA. A three-year clinical caries evaluation of the effect of sodium fluoride-silica abrasive dentifrice, Pharmacol Ther Dent 1981;6: Beiswanger BB, Gish CW, Mallatt ME. A three-year study on the effects of a sodium fluoride-silica abrasive dentifrice on dental caries. Pharmacol Ther Dent 1981;6: Eversole SL, Faller RV, Bitten ER Featherstone JDB. Conventional and remineralizing toothpastes compared in two ph cycling models. J Dent Res 1998;77 (Sp Is):(Abstr 1695). 28. Landrigan WF, Eversole SL, Best JM, Faller RV. Animal caries efficacy of conventional and remineralizing toothpastes. J Dent Res 1998; 77 (Sp Is B): Zero DT, Fu J, Hayes A, Scott-Anne K, Glena R, Proskin H, Triol C, Winston A. Anticaries efficacy of calcium phosphate dentifrice under different in situ model conditions. J Dent Res 1999;78 (Sp Is):(Abstr 529). 30. O Reilly MM, Featherstone JDB. Demineralization and reminerali-zation around orthodontic appliances: An in vivo study. Am J Orthod Dentofac Orthop 1987;92: Gorton J, Featherstone JDB. In vivo inhibition of demineralization around orthodontic brackets. Am J Orthod Orthofac Orthop 2003;123: Basdra EK, Huber H, Komposch G. Fluoride release from orthodontic bonding agents alters the enamel surface and inhibits enamel demineralization in vitro. Am J Orthod Dentofac Orthop 1996;109: