regulatory toxicology and pharmacology · 2017. 1. 6. · (us national library of medicine),...

Regulatory Toxicology and Pharmacology 59 (2011) 310–323

Contents lists available at ScienceDirect

Regulatory Toxicology and Pharmacology

journal homepage: www.elsevier .com/locate /yr tph

Integrated risk assessment of a hydroxyapatite–protein-compositefor use in oral care products: A weight-of-evidence case study

Julia Scheel ⇑, Martina HermannHenkel AG & Co. KGaA, Corporate Product Safety, Department of Human Safety Assessment, D-40191 Düsseldorf, Germany

a r t i c l e i n f o a b s t r a c t

Article history:Received 25 August 2010Available online 26 November 2010

Keywords:Risk assessmentWeight-of-evidenceParticlesHydroxyapatiteBiocomposite

0273-2300 � 2010 Elsevier Inc.doi:10.1016/j.yrtph.2010.11.003

Abbreviations: ANOVA, analysis of variance; ECEcotoxicology and Toxicology of Chemicals; ECVAMValidation of Alternative Methods; EPR, electron paGood Clinical Practice; gd, glycerol/water dispersionhydroxyapatite-fine/dull needles; HA-NN, hydroxyaphydroxyapatite-nano, plate-like; HA-NR, hydroxyaphen‘s egg test-chorioallantoic membrane; GIT, gastrocorneal epithelium; HPC, hydroxyapatite proteincoupled plasma; ITT, intend-to-treat; LPS, lipopolysacylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; SCConsumer Products (former name of SCCS); SCCConsumer Safety; SGF, simulated gastric fluid; TEM, tcopy; WoE, Weight of Evidence; WST-1, water-solusodium-30-[1-(phenyl amino-carbonyl)-3,4-tenitro)benzene sulfonic acid hydrate.⇑ Corresponding author. Fax: +49 211 798 12413.

E-mail address: [email protected] (J. Scheel

Open access under CC B

Risk assessment of cosmetic ingredients represents a regulatory standard requirement in Europe andother regions. An integrated approach was designed to assess the safety of HPC, a particulate compositeof hydroxyapatite and protein (gelatin) for use in oral care products, employing a weight-of-evidenceassessment and considering specific physico-chemical properties and exposure conditions. An initialevaluation of the constituents suggested that their chemical nature does not represent a particular healthhazard per se. Hydroxyapatite is the main component of teeth and bones in mammals; gelatin is used infood and assumed to be safe once a BSE/TSE risk has been excluded. In vitro screening tests were chosento further evaluate the biocompatibility: Hen’s egg test-chorioallantoic membrane (HET-CAM) to assessirritating effects towards mucous membranes; MTT cytotoxicity test with 3T3 fibroblasts; human cornealepithelial models to investigate inflammatory mediators and cytotoxicity; macrophage assays to measurecytotoxicity, inflammatory mediators and oxidative stress. Together with results from clinical studies,exposure estimates and analyses of kinetic properties, the presented information provides sound evi-dence to support the safe use of HPC. This is an example of a risk assessment for cosmetic use of smallparticles without the need for additional animal studies.

� 2010 Elsevier Inc. Open access under CC BY-NC-ND license.

1. Introduction

Sensitive teeth still present an unsolved problem in practicaldentistry. It is commonly accepted that exposure of dentine tothe conditions in the oral cavity is the main reason for this phe-nomenon. It has been reported that approximately 36% of adultsin the United States and Western Europe may suffer from sensitiveteeth (Addy, 2002) and it is likely that the incidence will increasebecause of the demographic development. The hydroxyapatite–protein-composite HPC represents an active ingredient in

ETOC, European Centre for, European Center on the

ramagnetic resonance; GCP,; HA, hydroxyapatite; HA-FN,

atite-nano, needles; HA-NP,atite-nano, rods; HET-CAM,intestinal tract; HCE, humancomposite; ICP, inductivelycharide; MTT, 3-(4,5-dimeth-CP, Scientific Committee onS, Scientific Committee onransmission electron micros-ble tetrazolium salt 1; XTT,trazolium]-bis(4-methoxy-6-

).

Y-NC-ND license.

toothpaste especially suited to protect sensitive teeth. HPC hasbeen demonstrated to adsorb to the surface of exposed dentine,inducing the crystallization of calcium and phosphate from the sal-iva to grow a thin protective layer which covers the dentin tubulithat conduct unpleasant stimuli to the nerve, thereby preventingpain induction caused by external stimuli like hot/cold or sweet/sour food and drinks (Henkel, 2005).

The natural mineralization repair mechanism of the saliva thusis used in a self-organizing process which can be referred to as neo-mineralization. In general, engineered hydroxyapatite materialsand their composites are intended for a variety of biomedical appli-cations, often serving as bone substitute in intraosseous implanta-tion or as implant coating materials, including small particle sizesdown to the nano range (Arts et al., 2006; Huber et al., 2007,2006c; Palmer et al., 2008; Webster and Ahn, 2007).

According to European Cosmetics legislation (EC, 1976), a cos-metic product is only allowed to be marketed if it has been provento be safe for the consumer. Toothpaste, like any other cosmeticproduct, therefore requires a thorough risk assessment which isessentially based on the safety of its ingredients. Information onthe properties of the formulation (e.g. clinical data) can be includedin the assessment. Legal restrictions to testing of ingredients andformulations are provided in Directive 2003/15/EC (EC, 2003),where a phasing out of animal testing is laid down (animal testingban).

http://dx.doi.org/10.1016/j.yrtph.2010.11.003mailto:[email protected]://dx.doi.org/10.1016/j.yrtph.2010.11.003http://www.sciencedirect.com/science/journal/02732300http://www.elsevier.com/locate/yrtphhttp://creativecommons.org/licenses/by-nc-nd/4.0/http://creativecommons.org/licenses/by-nc-nd/4.0/

J. Scheel, M. Hermann / Regulatory Toxicology and Pharmacology 59 (2011) 310–323 311

The discussion about safety concerns associated with small par-ticles is ongoing for many decades and is to a large extent relatedto potential risks following inhalative exposure, e.g. (Schmid et al.,2009). Especially ultrafine or nanoparticles are in the focus of thedebate, e.g. (ECETOC, 2005; Holsapple et al., 2005), usually meantto have one or more size dimensions between 0.1 and 100 nm.

Due to the diversity of particulate materials and their applica-tions, no fixed scheme is applicable to assess the potential risk ofsmall particles. Consequently, a specific, taylor-made riskassessment and testing strategy needed to be designed for HPC,consisting of several building blocks. This strategy comprises aweight-of-evidence (WoE) analysis which is often used to assessthe potential hazard, including multiple elements like physico-chemical characterization, available information on constituents,in vitro screening tests and clinical studies.

The assessment of biocompatibility is a key request in this re-gard. When selecting suitable test systems for tooth paste ingredi-ents, specific attention needs to be paid to the local compatibilitytowards mucous membranes as the primary site of contact.

Beyond that, it needs to be investigated whether a particulatematerial can penetrate into the body leading to systemic exposure.In case part of the material should become systemically available,the investigation of potential interactions with cells is of particularinterest, mainly the responses of macrophages as the first line ofdefense of the body. Cytotoxicity, inflammation and oxidativestress through reactive oxygen species (ROS) formation are consid-ered to be relevant to the potential development of several chronicdiseases (Borm et al., 2006; Brown et al., 2001; Donaldson et al.,2005; Hirvonen et al., 1996; Lewinski et al., 2008; Unfried et al.,2007) and consequently need to be included in the analysis. To fi-nally assess the relevance of effects, information on the biodegrad-ability of the material under consideration is also of importance.

This study aims at describing the whole process of the riskassessment of HPC, comprising the initial description of scientificobjectives, the identification of information requirements and datagaps, the selection of suitable non-animal methods to generatemissing data and the subsequent integration of all relevant infor-mation into a comprehensive assessment.

2. Materials and methods

2.1. Integrated risk assessment and weight-of-evidence (WoE) analysis

‘‘Integrated risk assessment’’ is not a standardized term and isused in a number of different contexts (Bridges and Bridges,2004). In our study it describes a qualitative, expert judgment-based integration of data from different sources to assess the safetyof consumers for the intended and foreseeable use of a cosmeticingredient.

In particular for the hazard assessment part, the risk assessmentis making use of a WoE analysis, a term which has also been usedin many different ways in the public literature (Weed, 2005). Werefer to WoE as a systematic, expert-judgment based approach toassess available information relevant for human safety and to iden-tify potential data gaps, triggering a tiered testing approach. Test-ing results in turn were included in the overall WoE analysis, andexposure estimates were performed. The process was continueduntil a responsible safety decision could be made based on the cur-rent knowledge (with the general option for re-evaluation as soonas new relevant knowledge becomes available).

2.2. Strategy of literature search and review

In addition to retrieval of supplier information on the constitu-ents, a literature search on the biocompatibility and toxicological

profile of gelatin and hydroxyapatite or composites thereof wasperformed in PubMed and other databases or networks (TOXNET(US National Library of Medicine), Ariel™/ChemEXPERT, otherinternet sources, in-house databases). Hits were first screened forrelevance to the issue by reading the abstracts. Additional refer-ences were identified through cross-references and additionalsearches on specific topics. About 80 references were selected insummary for detailed review with a particular focus on paperswith well-documented materials and procedures which are appli-cable to this assessment.

2.3. Test samples

Synthesis and characterization of HPC was described previously(Albrecht et al., 2009). In brief, HPC is synthesized by a co-precip-itation of calcium chloride dihydrate, collagen (gelatin with anaverage molecular weight of 140,000 Da with a molecular weightdistribution between approx. 6000 and >400,000 Da) and ammo-nium phosphate. The calcium/phosphorus ratio was analyzed withinductively coupled plasma (ICP) spectroscopy and equals to1.95 ± 0.28. The protein content is 36.2%. Particle morphologyand size were analyzed by transmission electron microscopy(TEM), showing irregular shaped particles in the micrometer rangewith average sizes of 1200 � 2100 nm; pictures shown in (Albrechtet al., 2009). The specific surface is 67 m2/g as analyzed by theBrunauer–Emmet–Teller (BET) method. For formulation into cos-metic products, HPC (9.5%) is provided in a preserved dispersionof glycerol (15%) and water (73%). In order to specifically investi-gate the properties of HPC, the material was used in cellular testsystems without preservative (due to the known cytotoxic effectsof preservatives) and also was isolated from the sediment. The finalconcentration of HPC in a marketed toothpaste is approx. 0.1%.Trade names for HPC are Nanit or Nanit�active (glyceroldispersion).

Depending on the specific test system, a number of controls andreference substances were applied along with HPC. An overview oftest samples is provided in Table 1. Pure hydroxyapatite withoutgelatin was produced in an analogous precipitation process asHPC but without the presence of gelatin (HA-NP). This materialwas used either as sediment (macrophage tests) or as 5.3% glyceroldispersion (HCE and 3T3 experiments). Two other HA materials,HA-NR (hydroxyapatite – nano, rods) and HA-FN (hydroxyapatite– fine/dull needles), were prepared with different sizes and mor-phologies as previously described (Albrecht et al., 2009) and in-cluded in macrophage experiments. Ostim� (Heraeus Kulzer,Hanau, Germany; abbreviated as HA-NN (hydroxyapatite-nano/needle-shaped) in the following), a bone grafting material ap-proved for clinical use, was used as a benchmark control in mostin vitro tests. The calcium/phosphate ratio of this material is 1.67,the surface area/mass is 106 m2 g�1 (Huber et al., 2007).

2.4. HET-CAM (Hen’s egg test-chorioallantoic membrane) to assessmucosal membrane irritation

The HET-CAM was carried out as previously described (Steilinget al., 1999) using the reaction time method for transparent andthe endpoint assessment for non-transparent test items. In brief,fertilized eggs were incubated for 9 days prior to use. Six eggs wereused for each test item: HPC (gd) and two dilutions thereof with 1%and 5% HPC, tooth gel with 1% HPC, standard reference toothpaste(Sensodyne Fresh Mint). Of the dispersions 300 ll were applied tothe CAM. The irritation potential is evaluated by occurrence of spe-cific effects to the membranes and/or vessels (hemorrhage (H), ly-sis (L), coagulation (C)) which are interpreted in comparison to 5%sodium magnesium lauryl-myristyl-6-ethoxysulphate (TexaponASV, Cognis, Germany). This internal reference compound is

Tabl

e1

Ove

rvie

wof

test

sam

ples

.

Mat

eria

lA

bbre

viat

ion

Part

icle

dim

ensi

ons

(nm

)

Sou

rce

Ref

eren

cefo

rde

tail

edsp

ecifi

cati

ons

HET

-C

AM

3T3,

MTT

HC

E-m

odel

s,LD

H

HC

Em

odel

s,IL

-alp

ha

Mac

roph

ages

RA

W26

4.7

Mac

roph

ages

NR

8383

Mac

roph

ages

prim

ary

rat

Hyd

roxy

apat

ite–

prot

ein

-com

posi

te(i

rreg

ula

rly

shap

ed)

(gd

wit

h9.

5%H

PC)

HPC

1200�

2100

SusT

ech

Dar

mst

adt

Alb

rech

tet

al.(

2009

)x

(gd)

ax

(gd)

x(g

d)x

(gd)

xx

x

Nan

o-h

ydro

xyap

atit

e,pl

ate-

like

(gd

wit

h5.

3%H

A-N

P)H

A-N

P3�

20�

45Su

sTec

hD

arm

stad

tA

lbre

cht

etal

.(20

09)

–x

(gd)

x(g

d)x

(gd)

xx

x

Nan

o-h

ydro

xyap

atit

e,ro

d-li

keH

A-N

R5�

90Su

sTec

hD

arm

stad

tA

lbre

cht

etal

.(20

09)

––

––

xx

x

Nan

o-h

ydro

xyap

atit

e,n

eedl

e-sh

aped

HA

-NN

3�

20�

100

Her

aeu

s(O

stim

�)

Hu

ber

etal

.(20

07)

–x

xx

xx

x

Fin

eh

ydro

xyap

atit

e,bl

un

t-en

ded

nee

dles

HA

-FN

95�

740

SusT

ech

Dar

mst

adt

Alb

rech

tet

al.(

2009

)–

––

––

xx

aPr

eser

ved.

312 J. Scheel, M. Hermann / Regulatory Toxicology and Pharmacology 59 (2011) 310–323

included in each study and known to be a moderately irritatingcompound to the rabbit eye in vivo. In the reaction time methodany reactions occurring within 5 min are observed and an irritationquotient (Q) is calculated; in the endpoint assessment, the sub-stance is rinsed after 30 s before observations are made and thesum of scores (S) for the six eggs is determined. The most pro-nounced effects (highest S-value) are then used to translate resultsinto four categories which are defined as follows: Q 6 0.8 or S 0–5:slightly irritating, Q > 0.8 to >1.2 or S 6–12: moderately irritating,Q P 1.2 to


release). One percent SDS was used as a positive control, untreatedHCE models as a negative control.

The LDH assay was performed with a commercially availableLDH-Cytotoxicity Detection Kit (Roche, Mannheim, Germany)according to the instructions of the manufacturer. IL-1a releasewas detected by using the Human IL-1a/IL-1F1 Quantikine ELISAKit (R&D Systems, Minneapolis, USA) according to the instructionsof the manufacturer. For all test samples and controls P5 replicaswere used (for details see Section 3).

Histological analysis was performed for HPC, HA-NP and SDStreated models. Models were fixed in 10% formalin and embeddedin paraffin according to standard procedures. Paraffin slices werestained with hematoxylin/eosin.

Statistical analysis of experiments described in Sections 2.5.1and 2.5.2 was performed with Microsoft�Office Excel 2003(1985–2003, Microsoft Corporation) software for basic evaluationand graphical editing; Kruskal–Wallis-ANOVA, a non-parametricmethod for testing equality of population medians among groups,and the Mann–Whitney U test, a non-parametric significance test,were performed with STATISTICA 8.0 software (StatSoft, Inc. 1984–2008, Tulsa, USA).

2.5.3. Macrophage assaysAssays with murine RAW264.7 macrophages, rat NR8383 and

primary macrophages were performed as described previously(Albrecht et al., 2009; Scheel et al., 2009). In brief, RAW264.7 cellswere treated with different concentrations of HPC, HA-NP, HA-NR,HA-NN and DQ12 quartz and LPS as positive controls to determinecell viability (XTT-test), TNF-a and NO release with commerciallyavailable test kits. For details see (Scheel et al., 2009).

NR8383 macrophages were treated with different concentra-tions of HPC, HA-NP, HA-NR, HA-NN, HA-NF and DQ12 quartz andLPS as positive controls to determine cell viability (WST-1 assay)and TNF-a release with commercially available test kits; oxidativestress was determined by EPR. The WST-1 test and EPR were addi-tionally performed with primary macrophages. For details see(Albrecht et al., 2009).

2.5.4. Clinical studiesTwo randomized, double blind, parallel-group, positive-stan-

dard controlled studies have been commissioned and performedaccording to Good Clinical Practice (GCP).

In Study 1, two toothpaste formulations with 0.1% net content ofHPC were compared against a toothpaste for sensitive teeth as abenchmark. Besides the assessment of sensitivity as major efficacyparameter, the criteria relevant for the evaluation of safety and tol-erability were as follows: adverse events collection, bucco-dentalexamination, gingival index, change of the teeth color and theamount of plaque, cosmetic acceptance questionnaire. The analysisof the included 90 subject’s characteristics and safety data was per-formed using individual data listings and descriptive statistics(planned: 90, screened/enrolled: 121, in safety analysis: 90; groupsof 30 subjects tested 1 toothpaste each, completed trial: 89 (1withdrawal in one of the test toothpaste groups)).

In Study 2 planned with 90 subjects (enrolled: 111, completedtrial: 105) the combination of a standard toothpaste with a HPCtooth-gel formulation (1.5% HPC; treatment group size 38 personsin the beginning/37 in the end of the study) and a HPC gel-strip (4%HPC; treatment group size 37/37) were tested against a toothpastefor sensitive teeth as a benchmark (treatment group size 36/31). Inanalogy to Study 1, safety and tolerability were assessed in additionto the sensitivity assessment using results of bucco-dental exami-nation and overall cosmetic questionnaire, and by incidence andseverity of adverse events. Safety (oral and dental examination)and tolerability evaluations (adverse events) were listed bytreatment group and subject number and described by treatment

group using frequency tables on the ‘‘Intent to Treat’’ (ITT)population.

2.6. Exposure and kinetics

2.6.1. General exposure estimate and overviewA general analysis of consumer exposure was calculated accord-

ing to average values proposed by the EU Commission’s scientificadvisory panel for issues concerning safety of consumer productsSCCS (formerly termed SCCP) (SCCP, 2006). Further ADME consid-erations were based on practical (Section 2.6.2) and theoreticaltoxicokinetic considerations, literature data and expert consulta-tion, as well as penetration behavior of HPC in mucosa-like tissuemodels as analyzed by Cryo-TEM (Section 2.6.3).

2.6.2. Dissolution behavior of HPC in simulated gastric fluidGeneral information on the stability/solubility of HA and gelatin

was initially obtained from general knowledge and literature. Thedissolution of HPC was further investigated in simulated gastricfluid (SGF), an aqueous dilution of HCl. For preparation of SGF,1.5 ml HCl (Roth, Rotipuran, p.a., >32%) were added to 1.5 l H2O(demineralized) to result in a pH value of approx. two.

The calcium content of HPC was analyzed by ICP (inductivelycoupled plasma) spectrometry. 1.5 g HPC (sediment) was pre-incubated with 100 ml H2O (demineralized) and stirred for25 min at 400 rpm at room temperature, then transferred to centri-fuge tubes and centrifuged for 15 min at 1000g. The supernatantwas removed and analyzed for its calcium content by IPC. The sed-iment was added to 500 ml SGF and stirred for 2 h. The calciumcontent of the solution was again analyzed by ICP. The totalamount of dissolved HPC was calculated based on the calcium con-tent in HPC as determined by IPC and by referencing the stoichiom-etry of HA. Measurements were carried out in duplicate.

2.6.3. HCE penetration screening (Cryo-TEM)HCE models (treated with HPC, HA-NP and HA-NN as described

in Section 2.5.2) were investigated in Cryo-TEM analysis. The watercontained in the HCE models was exchanged stepwise against eth-anol (Merck, Darmstadt, Germany) by incubating the tissue modelsfor 30 min each with an aqueous dilution of 30% ethanol v/v, 50%ethanol, 70% ethanol, 90% ethanol and 100% ethanol (3 � 30 min).Then ethanol is exchanged against propylene oxide (Merck,Darmstadt, Germany) by incubating with 100% propylene oxide2� for 15 min, then propylene oxide against Epon 812 (Serva, Hei-delberg, Germany) for final embedding of the tissue. Ultrafineslices were prepared from the samples, contrasted with uranyl ace-tate (Serva, Heidelberg, Germany) and subjected to TEM analysis(Philips CM12). Samples were prepared and analyzed in triplicateby inspecting particles in the different compartments of the tissuemodels without statistical evaluation.

3. Results

3.1. Assessment and testing strategy for hazard

The design of the assessment and testing strategy took the fol-lowing elements into account:

– What is the available hazard information for HPC including dataon starting materials; do these hazard characteristics suggest asafety concern?

– Is there need to further evaluate the potential hazard of HPC?Which additional tests are suitable to address the relevantquestions? Which are appropriate benchmark and controlsamples?


– Is HPC comparable to the hazard profile to other forms of HA(including nanoscale material)?

– Which overall conclusion can be drawn regarding the safety ofHPC under normal conditions of use?

The main building blocks of the risk assessment are displayed inFig. 1. Following initial assessment of available data, in vitro studieswere chosen to further assess the biocompatibility of HPC. WoE ex-pert judgment was used in each step to take a decision if availableinformation is sufficient or not and to choose suitable further testswhere necessary. In the test design, exposure considerations wereincluded to target typical concentrations of HPC in oral care prod-ucts (e.g. approx. 1% HPC (gd) or 0.1% HPC in toothpaste). Results ofthese tests were compared to observations from clinical studies.The specific rationale for each element of the assessment is de-scribed in more detail in the section below.

3.1.1. Basic toxicological profile of the constituents. Review of supplierinformation and publicly available data

Rationale: The logical first step in any assessment of hazard orrisk is to review available information. The outcome of the expertevaluation will be an initial assessment and a judgment whetherinformation is still missing or not.

3.1.1.1. Hydroxyapatite (HA) and composites thereof. Hydroxyapatiteis a naturally occurring phosphate mineral, which is the main com-ponent of teeth and bones in mammals. It is a well-establishedmedicinal product which is considered suitable and safe materialintended for intraosseous implantation and which is widely usedin practical dentistry (Heraeus Kulzer GmbH, 2005; Huber et al.,2006a,b,c,d; Kumta et al., 2005; Thorwarth et al., 2004).

A large number of studies on hydroxyapatite in biomedicalapplications report good biocompatibility and biodegradability ofthe material, which is used in various specifications (e.g. dense,coating, granules, powders/dispersions, various sizes and shapes)and under different conditions, e.g. implants, bone filling (for pri-mary hits in a recent PubMed search see Fig. 2). In general, thematerial causes no inflammatory reaction at the site of administra-tion and does neither induce acute nor chronic toxic effects.(Bloebaum et al., 1998; Cui et al., 1996; Guo et al., 1999; Huanget al., 2004; Liu et al., 1997; Rumpel et al., 2006; Trofimov et al.,1996; Wenisch et al., 2003). Specifically, data on HA-collagen com-posites has been reviewed and supports the biocompatibility bothin humans and animals (Wahl and Czernuszka, 2006).

Commercially available hydroxyapatite is not classified as haz-ardous; the acute toxicity is extremely low: LD50oral,rat >25,350 mg/kg (Merck, 2000). A nano-sized HA dispersion, Ostim�, has under-gone comprehensive preclinical evaluation including biocompati-bility and genotoxicity testing (Rudin et al., 1994) and receivedregulatory approval for biomedical applications in Europe (CE cer-tificate) and the US (FDA approval). Several studies for clinicalapplications have been published (Arts et al., 2006; Heraeus KulzerGmbH, 2005; Huber et al., 2006a,b,c,d, 2007; Kilian et al., 2002;Kumta et al., 2005; Laschke et al., 2007; Rauschmann et al.,2005; Thorwarth et al., 2004). Several nanoscalar HA materialswere investigated and used for dentistry/implantation, local drugdelivery and cosmetics, e.g. as root filling material (Busso, 2009).

Besides Ostim�, a number of other HA materials have beentested for mutagenicity in various in vitro and in vivo assays whilepart of the assays were performed directly with the HA or HA com-posite materials, others with eluates thereof. Reports include theprotocol of Ames (Guo et al., 1999; Jantova et al., 2008; Liu et al.,1997; Suzina et al., 2004), micronucleus test (MNT) (Guo et al.,1999; Liu et al., 1997; Ye et al., 2004), Comet assay (Jantovaet al., 2008), HPRT (Jantova et al., 2008) and chromosome aberra-tion (Kannan et al., 2004a,b). Though studies were usually not

performed according to internationally approved protocols (likeOECD technical guidelines), there is strong evidence that genotoxicand mutagenic effects of HA can be considered unlikely. Somein vivo studies explicitly investigated possible carcinogenicity withno positive findings (Landi et al., 2008; Takamura et al., 1994). Noreports on tumor formation due to HA exposure could be found inthe literature which is also true for other adverse subchronic orchronic effects.

Another observation is that HA has a well-known bindingcapacity for DNA and proteins, a fact that is used extensively inDNA and protein chromatography as well as in pharmaceutical re-search for delivery agents of proteins and DNA (Kumta et al., 2005).Regarding the relevance for the assessment of genotoxic effects,however, there is no evidence that adverse reactions occur uponcontact with cells through direct interaction with nuclear DNA.Also, in contrast to intercalating agents like ethidium bromide,DNA binding to HA is non-covalent, pH-dependent and allows re-elution (e.g. with increased concentrations of phosphate buffer).

Upon uptake in macrophage cells, it has been demonstratedthat nano HA particles were dissolved progressively in lysosomes.Enhanced cytotoxicity towards macrophages was observed withextremely small particles (approx. 50 � 20 nm) compared to largerparticles on the basis of mass/volume. It was speculated thatincreasing intracellular calcium level following HA dissolutioncould be the main reason for cytotoxicity and apoptosis observedin such studies (Motskin et al., 2009).

No allergic skin reactions have been reported in clinical studiesincluding those with specific investigation of potential allergeniceffects (Kannan et al., 2004a,b; Li et al., 2008; Pankratov et al.,1994; Rajab et al., 2004; Suzina et al., 2004).

Physical particle properties like surface area, size and morphol-ogy were reported to potentially influence the biocompatibility ofdifferent hydroxyapatite materials on specific cellular systems(Evans, 1991; Grandjean-Laquerriere et al., 2005; Laquerriereet al., 2003). Needle-shaped particles, for instance, were reportedto induce larger production of cytokine IL-18 compared to othermorphologies (Grandjean-Laquerriere et al., 2004).

Summary of literature review and implications for testing needs:HA is a naturally occurring phosphate mineral, which is the maincomponent of teeth and bones in mammals and is used in a varietyof biomedical applications. Literature shows that also small dimen-sions do not necessarily produce a risk to human health. However,some studies have associated cytotoxic effects, enhanced inflam-matory mediators and other variations of biomarkers with HAparticle size in specific cell cultures. There is not yet enough infor-mation available to deduce robust rules from hydroxyapatite spec-ifications to its toxicologically relevant properties. Since there is nostandard, officially accepted method or approach to assess the haz-ard of small particles, including the choice of the cell or tissue sys-tem and application of the test samples, the relevance of the testconditions to the physiological situation needs to be critically as-sessed. An individual assessment of each material is needed whichmight comprise at least some in vitro testing, in particular to assessinflammatory effects and cytotoxicity towards cells. Ostim� as anofficially approved medical product seems to be a valuablebenchmark to be included in the testing of particulate HAmaterials.

3.1.1.2. Gelatin. Gelatin is a naturally occurring polymer/protein.The used gelatin is of no BSE (Bovine Spongiform Encephalopathy)source as ensured by a BSE-Certificate from the manufacturer. Also,being produced from animal sources, it was assured that the mate-rial does not present a TSE (Transmissible Spongiform Encephalop-athy)/BSE risk applying state-of-the-art criteria. No hazards areexpected based on the nature of gelatin and its long-standing use

RISK ASSESSMENT

Toxicological information on the constituents (gelatin and hydroxyapatite): Literature data and supplier’s information for the bulk material (basic assessment); data search to also include information on nanoscalar hydroxyapatite to complement the assessment

In vitro Screening Tests to further assess/support the assumptions derived from the initial evaluation: test for irritation, inflammation, cytotoxicity, oxidative stress

Clinical studies to approve biocompatibility in humans

Weigth-of-evidence HAZARD ASSESSMENT of HPC

EXPOSURE ASSESSMENT of HPC

Generic exposure estimate according to SCCS guidance

Consideration of physico-chemical properties (e.g. solubility in acid medium)

Consideration of potential mucosal penetration Literature/theoretical considerations + in vitro screening experiment

Physicochemical Characterization

Fig. 1. Building blocks of the integrated safety assessment strategy for HPC.


in food products. An extensive literature search for potential ad-verse effects of gelatin therefore was not considered necessary.

3.1.2. In vitro screening testsAs a result of the initial assessment, it was concluded that the

biocompatibility of a new HA material should be specifically testedwith a focus on potential inflammation, oxidative stress and cyto-toxic effects since these effects are of major concern in relation tosmall particle sizes. The interpretation of effects should be done incomparison with appropriate benchmark controls and might in-clude a comparison with HA in other physical forms.

When selecting the in vitro test battery for HPC and defining theindividual test designs, the foreseeable conditions of use were ta-ken into account. Test concentrations were chosen to cover thetypical concentration in products. In addition, technical limitationswere considered, like in the case of cell culture experiments, theapplied concentrations should not produce overload effects in or-der not to create misleading results and artifacts. The followingtest systems were selected which represent valuable routine

protocols to provide relative estimates of the properties of testsamples: Section 3.1.2.1 estimation of the irritating potential to-wards mucous membranes in the HET-CAM, Section 3.1.2.2 cellviability (MTT) assay in 3T3 fibroblasts, Section 3.1.2.3 LDH andIL-1a measurement in HCE models, Section 3.1.2.4 investigationof inflammatory and cytotoxic effects as well as induction of oxida-tive stress in three different macrophage test systems. The ratio-nale for the choice of each test system is described below inmore detail.

3.1.2.1. HET-CAM. Rationale: Assessment of potential irritatingeffects towards mucous membranes as an important aspect of bio-compatibility. The CAM represents a sensitive, highly vascularizedmembrane suitable to detect potential irritating effects towardseyes and mucous membranes. A good concordance of HET-CAM re-sults with in vivo data of the Draize test could be found (Steilinget al., 1999). Henkel has a 20 years in-house experience with thistest for the assessment of a variety of raw materials andformulations.

(“hydroxyapatite” OR “hydroxylapatite”OR “hydroxy apatite” OR “hydroxyl apatite”)

16000 hits

AND

AND

3174189"biodegradation"

1101517256("biodegradability" OR "biodegradable")

00002"free" AND "radical" AND "formation"

1062271"inflammation"

011137("reactive" AND "oxygen" AND "species" OR ROS)

022016"oxidative" AND "stress"

000030("carcinogenicity" OR "carcinogenic")

000024("mutagenicity" OR "mutagenic")

000019("genotoxicity" OR "genotoxic")

7610114693823("biocompatibility" OR "biocompatible")

12194199("cytotoxicity" OR "cytotoxic")

52164341("toxicity" OR "toxic")

NTNMNPNC

AND

140NT = “nanotechnology”

23NM = ("nanomaterial" OR "nanomaterials")

275NP = ("nanoparticle" OR "nanoparticles")

130NC = ("nanocomposite" OR "nanocomposites")

Fig. 2. PubMed search for toxicological information on hydroxyapatite. Example of a PubMed search in December 2009 for toxicologically relevant information onhydroxyapatite, displaying the initial search terms and hits.


Results of the HET-CAM assay with different concentrations ofHPC (gd) are displayed in Table 2. None of the formulations is clas-sified as irritating or severely irritating in this test system. HPC (gd)was found to be ‘‘slightly irritating’’ (lowest category) up to thehighest concentration and less irritating than a tooth gel formula-tion and a standard toothpaste formulation.

3.1.2.2. Viability of 3T3 fibroblasts (MTT test). Rationale: Assessmentof biocompatibility, particularly cell viability. 3T3 cells represent asensitive test system in combination with the MTT assay. Even be-fore fatal destruction of cells, this assay is able to detect adversecellular events, namely metabolic dysfunction of enzymes leadingto decreased performance as detected by a colorimetric readout.This test might be complemented by another assay as many differ-ent conditions can increase or decrease metabolic activity. TheMTT assay have been frequently used for the assessment of the

biocompatibility of dental materials, e.g. (Schweikl and Schmalz,1996; Torrado et al., 2005).

Cell vitality was measured following treatment of 3T3 cells withdifferent concentrations of test samples. The application of HPC(gd) and the other test samples tested up to a concentration of5% did not markedly affect cell viability. Results are displayed inFig. 3.

3.1.2.3. Mucosa-like tissue models (LDH and IL-1a release). Rationale:Assessment of biocompatibility towards mucous membranes, par-ticularly cell viability and inflammation. LDH release indicatessevere and fatal cell damage (disruption of outer cell membranes).IL-1a is a common marker for inflammation and considered to be asignificant factor for the development of a number of chronic ef-fects. The three-dimensional HCE models are suited to investigatemucous membrane irritations and resemble more closely the phys-iological conditions than 3T3 cell culture. Amongst a number of

Table 2Results of the HET-CAM.

Test samples Reaction time method (Q) End point assessment (S) Evaluation

H L C

HPC (gd)a, dilution 1 (1% HPC) 0.02 (±0) – – – slightly irritatingHPC (gd)a, dilution 2 (5% HPC) – 0 0 0 slightly irritatingHPC (gd), undiluted (9.5% HPC) – 0 0 0 slightly irritatingToothgel with 1% HPC – 6 3 0 moderately irritatingStandard reference toothpaste (Sensodyne Fresh Mint) – 6 0 0 moderately irritatingTexapon ASV 5% (reference) 1.00 (±0.04) 12 10 0 moderately irritating

a Preserved.


commercially available mucosa models, the SkinEthic HCE modelwas chosen due to a broad experience with this model in our in-house laboratories and well-established and reliable protocols.

HCE models were treated topically for 3 min with HPC (gd), HA-NP, and HA-NN in concentrations of 0.5% and 1.0%. Following 18 hof incubation, no significant inflammatory or cytotoxic effects ofHPC or the other test samples were observed (Figs. 4 and 5). Inthese experiments treatment with SDS as positive control lead toan increased mean LDH value and a substantial amount of releasedIL-1a in the culture medium. Although the mean value is clearlyelevated, this enhancement was not shown to be statisticallysignificant.

The adverse effects of SDS could however be clearly demon-strated in the histological evaluation, where SDS was shown to dis-rupt the upper layers of the HCE model. The test samples, incontrast, provided a vital picture composed of intact cells in differ-ent layers as also seen in the untreated control (Fig. 6).

3.1.2.4. Macrophage assays. Rationale: Assessment of biocompati-bility and information on biodegradability in case that part of thematerial would be taken up systemically. Macrophage response

0

20

40

60

80

100

120

140

5% 2.5% 1% 0.5

%0.1

%

vita

lity

[%]

Fig. 3. MTT assay with 3T3 cells. Cell vitality was measured following treatment of 3T3 c(n = 8 for HA-NP (gd) 1%, HPC (gd) 1%, HA-NN 1%. Statistical significance of effects comparsign-test for paired samples with alpha-correction. Following Friedman-ANOVA, significasignificance: p < 0.0045 (alpha/11)).

is considered essential since these cells represent the first line ofdefense and can mediate further processes. Macrophage assaysare well established to investigate the cytotoxic and inflammatorypotential of particles (Aam and Fonnum, 2007; Albrecht et al.,2007; Balduzzi et al., 2004; Shamsuria et al., 2004). Activation ofmacrophages is generally important to ensure that particles canbe taken up and degraded. The reaction however should not beuncontrolled and not be persistent in order not to lead to adverseeffects.

Results of the macrophage assays were published previously. Ashort summary of the results will be given here; for details we liketo refer to the original articles.

In a first study (Scheel et al., 2009), RAW 264.7 macrophageswere incubated with HA-NR, HA-NP, HA-NN, and HPC in concentra-tions from 50 to 5000 lg/ml. Cells were analyzed for viability (XTT-test), cytokine production (TNF-a) and induction of nitric oxide(NO) after 18 and 42 h with DQ12 quartz and lipopolysaccharide(LPS) as positive controls. Viability was not considerably impairedup to concentrations of 500 lg/ml by the test samples at both timepoints and was overall about one order of magnitude higher thanwith comparable concentrations of quartz. TNF-a release was

SDS

0.01

%

untre

ated

contr

ol

HPC(gd)

HA-NP(gd)

HA-NN

ells with different concentrations of test samples. Data are presented as mean ± SDed to the untreated control was tested by Friedmans ANOVA and subsequently by ant differences between groups could not be confirmed in the sign test (criterium for

0,865 0,84 0,856 0,859 0,8440,771

1,132

0,862

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

HPC(

gd) 0

,5%

HPC(

gd) 1

%

HA-N

P(gd

) 0,5

%

HA-N

P(gd

) 1%

HA-N

N 0,5

%

HA-N

N 1%

SDS1

%

untre

ated

contr

ol

LD

H r

elea

se [

OD

]

Fig. 4. LDH release from HCE models. LDH release from HCE models was detected following 3 min exposure with test samples and subsequent 18 h incubation. Data arepresented as mean ± SD (n = 8 for HA-NP (gd) 1%, HPC (gd) 1%, HA-NN 1%, untreated control; n = 5 for all other samples). Statistical significance of effects in the differenttreatment groups was tested by Kruskal–Wallis-ANOVA. No significant differences between groups could be identified.

32,055 31,10829,463 30,001 30,171

35,964

45,909

29,35

1,340

10

20

30

40

50

60

70

HPC(

gd) 0

,5%

HPC(

gd) 1

%

HA-N

P(gd

) 0,5%

HA-N

P(gd

) 1%

HA-N

N 0,5

%

HA-N

N 1%

SDS

1%

untre

ated

contr

ol

cultu

re m

edium

IL-1

α [

pg]

Fig. 5. IL-1a release from HCE models. IL-a release from HCE models was detected following 3 min exposure with test samples and subsequent 18 h incubation. Data arepresented as mean ± SD (n = 8 for HA-NP (gd) 1%, HA-NN 1%, untreated control; n = 7 for HPC (gd) 1%; n = 5 for all other samples). Statistical significance of effects in thedifferent treatment groups was tested by Kruskal–Wallis-ANOVA. Differences in significance were initially indicated but not confirmed in a subsequent Mann–Withney U test(criterium for significance: p 6 0.0083 (alpha/6)).


induced in all samples after 18 h, with HA-NR showing the mostpronounced induction at 100 lg/ml, still clearly below the LPS sig-nal. No or little induction was observed after 42 h. NO productionwas low at both time points.

In a second study (Albrecht et al., 2009), rat NR8383 macro-phages and rat primary alveolar macrophages were incubated withHA-NR, HA-NP, HA-NN, HA-FN and HPC in concentrations from 50 to

5000 lg/ml with LPS and DQ12 quartz as positive controls. In WST-1 and LDH assays with NR8383 cells, no cytotoxicity was observedfor HPC and other HA samples up to 3000 lg/ml, while HA-FNshowed a significant effect at the highest dose in the LDH assay.No cytotoxicity was observed in primary cells with all samplesup to 300 lg/ml. ROS generation was significantly enhanced withHA-NN and HPC in NR8383 cells in an EPR analysis. No effect was

Fig. 6. Histology of HCE models. Histology of HCE models treated with (a) 1% HPC(gd), (b) 1% HA-NP (gd), (c) 1% SDS, (d) untreated. Paraffin slices were stained withhematoxylin/eosin. The microscopic magnification was 20�.


detected in primary cells, which are considered more relevant tophysiological conditions. All HA samples elicited TNF-a releasefrom NR8383 cells with significantly lower potency than the posi-tive controls.

The combined results of both studies support the conclusionthat the tested HA materials exhibit a good biocompatibility andare safe to use. Since the benchmark Ostim� (HA-NN) as officiallyapproved material is readily biodegradable and since HPC showssimilar effects, it can be assumed that HPC is biodegradable as well.

3.1.3. Clinical studiesRationale: (a) Investigation of product efficacy of various formu-

lations with HPC (including toothpaste with 0.1% HPC) in the treat-ment of dentine hypersensitivity and (b) confirmation the resultsof the initial risk assessment with regard to safety and tolerability(the clinical studies were not designed to assess the hazard). Withregard to the scope of this publication only findings from (b) are re-ported in the following.

Study 1: Only one adverse event (lumbago), which was not re-lated to the treatment, was reported during the study. No clinicallyrelevant abnormalities were observed in bucco-dental examina-tion during the study. The variations of gingival index were iso-lated and were, therefore, not considered as clinically relevant.Thus, both test toothpastes may be considered as having no inflam-matory effect on gingiva. The change in teeth color was reported bya small number of subjects equally distributed among the threetreatment groups whereas the amount of plaque remained un-changed during the study.

The cosmetic acceptability questionnaire did not elicit negativeresults with regard to side effects. Thus the overall acceptabilitywas considered to be very good for both toothpastes containingHPC.

Study 2: No clinically relevant abnormalities were observed inbucco-dental examination during the study. No adverse eventswere reported during the study. The use of the tooth-gel formula-tion and the gel-strip product were safe and well tolerated.

In summary, no clinically relevant abnormalities were observedin bucco-dental examination during both studies. The use of theproducts was without unwanted side effects and well tolerated.

3.1.4. Summary of hazard evaluationBased on an initial assessment of supplier information and liter-

ature data no major concern was identified with regard to the safeuse of the constituents of HPC, hydroxyapatite and gelatin. How-ever, there is evidence from the literature that in some test systemssmall sizes of HA particles and specific morphologies could resultin different behavior like an increased inflammatory response (cf.Section 3.1.1.1). Findings from a series of in vitro studies withHPC and other HA materials supported the biocompatibility ofHPC with regard to cell viability, irritating effects, inflammationand oxidative stress. Additional information was obtained fromclinical studies with HPC-containing formulations, such as tooth-paste. No clinically relevant abnormalities were observed in buc-co-dental examination during the studies. The use of theproducts was without unwanted side effects and well tolerated.The hazard to human health for the intended use in oral care for-mulations is therefore considered to be low.

3.2. Exposure analysis/toxicokinetics

The exposure analysis took the following elements intoaccount:

– What is the amount of HPC that is applied on a basis of dailyuse?

– How much of HPC is likely to be swallowed?– If swallowed, what is the fate of the material in the gastrointes-

tinal tract/in the body?– If not swallowed, is it possible or likely that the material pene-

trates directly through the oral mucosa?

The concentration in currently marketed toothpastes of approx.0.1% (corresponding to approx. 1% HPC (gd)) is used as a represen-tative concentration for exposure calculations. Exposure throughinhalation is not considered since it is not expected under normaluse conditions. The building blocks of the assessment are outlinedin Fig. 1.

3.2.1. General exposure estimate according to regulatory proposalRationale: Generic exposure estimate. The major part of the

product taken up by the consumer is expected to occur via the oralroute through potential swallowing of a small portion of the prod-uct during application. To estimate this amount, the consumerexposure was calculated according to average values provided bythe EU Commisson’s advisory panel (SCCP, 2006) which are similarto the results of a study carried out by European cosmetic manu-facturers (Hall et al., 2007).

Toothpaste with 0.1% HPC net content (adult): 1.4 g; two times/day = 2.8 g/day.

Retention factor: 0.17.Resulting amount: 0.48 g toothpaste/person/day.Maximum amount of HPC after swallowing: 0.48 g � 0.001 =

0.5 mg/person/day.As a result, approximately 0.25 mg HPC per application may

reach the stomach.


3.2.2. Consideration of mucosal penetration potential (direct contactwith oral mucosa)3.2.2.1. Theoretical calculation. Rationale: This ‘‘worst-case’’ calcula-tion was performed to estimate the maximum amount of HPC thatcomes into contact with the mucosa in the oral cavity. In case of100% penetration this would represent the maximum amount thatcan be taken up systemically.

As outlined above (Section 3.2.1), the daily average amount ofHPC to come into contact with oral cavity can be calculated as fol-lows: 2.8 g � 0.001 = 3 mg. This represents the maximum amountthat could be taken up in the (theoretical) case of complete pene-tration through the oral mucosa.

The potential distribution of HPC in the oral cavity can be calcu-lated as well. The material is not expected to be significantly dis-solved in the oral cavity, where the pH typically ranges betweenapprox. 6 and 7.5. The total surface area (epithelium) of the oralcavity in adult humans has been reported to be about 200 cm2

(Aframian et al., 2006). Provided that the complete epitheliumwould come into contact with 3 mg HPC and would be evenlyspread, an amount per mucosal area of 0.015 mg HPC/cm2/daycan be calculated. (For comparison, the amount per area in theHCE mucosa model experiments is within a range of 0.5–1 mg/cm2.) It has to be considered that this represents a maximumamount, since saliva (approx. 1 ml) will dilute the product, and alarge part of the product is expected to be disgorged after applica-tion, or partly swallowed. This calculation is largely theoretical andis not considered sufficient for the exposure estimate.

3.2.2.2. Considerations regarding penetration based on literature andexpert consultation. Rationale: Estimation of the likelihood of directsystemic uptake through the mucosa of the oral cavity via litera-ture search and expert consultation.

Much of the literature dealing with the permeability of the oralmucosa has been concerned with drug absorption; for review see(Werner, 2003). Being more permeable, the non-keratinized areasof the mucosa are of special interest for drug delivery. Buccal mu-cosa contains intercellular lipids which are responsible for itsphysical barrier properties, resulting in poor permeability for lar-ger drug molecules such as peptides and proteins, requiringabsorption enhancers at molecular weights above 500–1000 Da.The gelatin component in HPC has an average molecular weightof 140,000 Da with a molecular weight distribution between ap-prox. 6000 and >400,000 Da. It can therefore be assumed thatHPC agglomerates are not likely to penetrate oral mucosa.

Another important factor for enforced absorption/penetration(of drugs) is a specific bioadhesivity towards the mucosa to over-come washing with saliva. In our experiments, the dispersion couldbe readily rinsed from the mucosa models and did not apparentlyadhere to the cells, as assessed by optical examination.

Though evidence is given that under normal conditions HPC isnot absorbed through healthy oral mucosa, the question could beraised whether systemic exposure with small amounts of HPCmay occur, especially in case of mucosa damage, e.g. fresh woundsafter dental extraction and tissue damage following scaling, sub-gingival irrigation or other therapeutical procedures, e.g.(Schroeder, 2000). In such cases it cannot be ruled out that tracesof the material enter the blood circulation though normal physio-logical defense mechanisms are likely to counteract potential up-take through bleeding and coagulation. It can also be consideredthat some oral diseases like periodontitis, gingivitis, neutropeniaand ulcerative mucositis can affect the permeability barrier ofthe oral mucosa. For example, there are case reports that patho-genic bacteria can reach the circulation of patients, e.g. (Ulstrupand Hartzen, 2006).

Under most chronic disease conditions and with open woundshowever, the flux is usually to the outside, thus counteracting

uptake. In chronic inflammation states of the periodontium, the sul-cus-flux-rate (SFFR) is elevated – plasma release at the crevicularepithelial cells (Sauerwein, 2000) – and considered to be an indica-tor for inflammation. Various investigations (Del Fabbro et al.,1998; Fjaertoft et al., 1992) have demonstrated that, on the onehand, the number of inflammatory cells is increased in the connec-tive tissue, and on the other hand, proteins can enter the gingiva-sulcus through osmotic pressure with the emerging pressure gradi-ent. Therefore, in the presence of chronic alterations of the perio-dontium, it is very unlikely that HPC will reach the blood stream.

In conclusion, the risk of systemic exposure through lesions ofthe periodontium or wounds is therefore considered very low.(Dr. med. dent. L. Weber, Alderney Dental Practice, Poole, UnitedKingdom, personal communication.)

3.2.2.3. Screening for possible translocation of HPC into cells or celllayers using Cryo-TEM. Rationale: Exposure estimate through po-tential direct penetration through the oral mucosa. This in vitroscreening experiment was performed to substantiate consider-ations described under Section 3.2.2.2 as well as an observationfrom the experiments with HCE models described in Section 3.1.2,where optical inspection indicated that the test samples includingHPC can be easily rinsed off the models and are not sticky towardsthe epithelium. As a consequence it was assumed that the penetra-tion was not promoted and might even be absent or negligible.

A possible translocation of HPC into cells or cell layers in HCEmodels was analyzed by Cryo-TEM following treatment with testsamples and rinsing as described in Section 2.5.2. Since the SkinE-thic HCE model used for this investigation is non-keratinized, it canbe assumed that it resembles the non-keratinized (= more perme-able) regions of the oral cavity.

Result: No indication for penetration of HPC into tissue and cellscould be detected in the TEM pictures, and only rarely with HA-NN(pictures not shown).

3.2.3. Physico-chemical properties (dissolution kinetics)Rationale: Generate information about the fate of HPC/HA in the

gastrointestinal tract. After swallowing the material entersthe stomach. Knowing that HA readily dissolves in acid medium,the dissolution behavior of HPC in simulated gastric fluid wasinvestigated.

It is commonly known that HA is nearly insoluble in aqueousmedium at alkaline or neutral pH but increasingly dissolves ifthe pH goes down to the acid range. In particular, starting frompH values of approx. 5 HA becomes increasingly soluble (Larsenand Pearce, 1997), which is also utilized by macrophages for bio-degradation of the material in the lysosomes (Bloebaum et al.,1998; Motskin et al., 2009).

While the dissolution behavior may vary to a certain extentdepending on the presence of ions in the saliva (phosphate, fluo-ride, etc.), a ‘‘critical pH’’ for dissolution of hydroxyapatite in dentalenamel was determined to be at pH 5.1 (Ericsson, 1949); the mostoften used reference value is 5.5 (Bardow et al., 2004; Larsen andPearce, 1997).

The average pH values in the acid medium of human stomachare 1.7 (fasted), 1.9 (median posterior portion), 2.7 (median ante-rior portion) or 5 (fed) (ECHA, 2007). The gastric emptying half-time of solid meal was reported to be around 2 h (Yamada,1999). It is therefore assumed that HA largely dissolves in thestomach, even though the dissolution rate at fed conditions is ex-pected to be lower than in the non-fed state. The gelatin compo-nent, like other proteins, is expected to be cleaved by proteolyticenzymes in the stomach.

A test with stimulated gastric fluid (SGF) was conducted to ana-lyze the specific dissolution behavior of HPC. No precipitationcould be detected following incubation with SGF. From the initial


43/45 mg calcium in the two HPC sediment samples, 1.8/1.9 mgwere identified in the aqueous medium and 39.5/42 mg in SGF.The average recovery rate was 96.9%. It can therefore be expectedthat the material readily dissolves in the stomach.

3.2.4. Summary exposure assessmentSystemic exposure through either oral uptake or direct contact

with the oral mucosa is considered to be very low. Exposurethrough inhalation is not expected under normal use conditions.Small amounts that are taken up through swallowing will be disin-tegrated in the acid environment of the stomach and not be persis-tent in a particulate form. A direct penetration through the mucosaof the oral cavity is very unlikely as demonstrated by literaturedata and supported by the results of an in vitro screening experi-ment. In case trace amounts of particles would be taken up system-ically, they can be degraded by macrophages in the acidenvironment of lysosomes.

3.3. Risk assessment

Risk assessment is generally accomplished by a holistic view onboth hazard and exposure assessment and by coming up with aconclusive statement on the estimated risk. As summarized above(Section 3.1.4) the hazard of HPC can be considered as low due tothe various pieces of evidence generated by basic assessment ofavailable data, in vitro experiments and clinical studies. No concernwas raised in the basic assessment when considering the chemicalcomposition of the material and the toxicological profile of its par-ent compounds which both are naturally occurring and representthe main component of teeth and bones in mammals. It could besubstantiated by the in vitro experiments and information fromclinical studies that the specific physico-chemical properties ofHPC are not leading to additional, unwanted effects at typical useconcentrations. Nevertheless, a thorough exposure assessmentwas performed, leading to the conclusion that systemic exposurecan be considered negligible. Small amounts that might be swal-lowed are likely to dissolve in the acid medium of the stomach. Di-rect mucosal penetration seems to be unlikely, and in case traces ofHPC would become systemically available it can be assumed thatmacrophages as the first line of defense are well capable to processand degrade the particles without leading to adverse effects. It isconcluded that the use of HPC in oral care formulations can be con-sidered to be safe for the consumer.

4. Discussion

A taylor-made risk assessment was designed for a hydroxyapa-tite–protein-composite intended for use in oral care formulationswith a typical concentration of approx. 0.1% in toothpaste. Hazardwas analyzed by making use of a weight-of-evidence approach,taking into account the physico-chemical properties, toxicologicalinformation on the constituents, in vitro screening tests and clinicalstudies with HPC without the need to perform new animal testing.A major reason why such an approach could be successfully ap-plied in this case is the fact that the constituents of HPC are well-known substances with a long tradition of safe use and whichnaturally occur in bones and teeth of mammals. In addition, con-cerns often raised with potentially enhanced cytotoxic and inflam-matory properties of small or nanosized particles could bedispelled by the results of relevant in vitro screening experimentswhich were additionally confirmed in clinical studies.

Pure hydroxyapatite materials in the typical size range of ananomaterial (i.e. one or more dimensions below 100 nm) were in-cluded in the majority of the in vitro tests. These materials behavedlargely similar in our test systems compared to particle sizes in the

micrometer range. Interestingly, a meta-analysis of 49 studiesdealing with effects of particulate matter on cytokine productionin vitro revealed that results across cell types are relatively consis-tent especially in the typically chosen concentration range be-tween 50 and 100 lg/ml (Mitschik et al., 2008).

Broad experience is available regarding in vitro methods usedfor the assessment, though individual methods may not be offi-cially approved. For example, with regard to classification purposesunder chemicals law for eye and mucous membrane irritation, onlyfew in vitro tests, including the HET-CAM, are accepted by author-ities for screening but of severe irritants only (EU, 2009). In vitromethods to detect these kind of effects nevertheless have a long-standing history in the safety assessment of cosmetics, as reviewedin (Eskes et al., 2005; McNamee et al., 2009). In addition, severalstudies were published using the HET-CAM test to assess nanopar-ticles and suspensions thereof in medical applications (Araujoet al., 2009; Gupta et al., 2009; Pardeike and Muller, 2009; Wolfet al., 2009).

Benchmark-based testing designs (in the current study we wereable to make use of an officially approved medical product) andinclusion of appropriate controls are necessary tools to make rea-sonable use of the results, in particular in cases where no absolutethresholds are established to judge on adversity of effects.

Obviously, known or potential limitations of individual testmethods need to be considered. Electron microscopy, for example,commonly provides detail snapshots on the analyzed items. Suit-able TEM approaches for quantifying particles in tissues are beingexplored and so far have proven to be very complex (Mayhewet al., 2009). Also, methods to measure cell viability can be influ-enced by various parameters. Therefore combination of differentcell viability methods and cellular systems which may differ insensitivity enhances the overall quality of information. Cell viabil-ity was not severely impaired in any of the different test systemsused in this study but slight differences could be observed in somecases. Regarding the comparability of MTT and LDH, others re-ported that MTT and LDH results did not correlate well whenassessing fine and nanoparticle effects in lung cell cultures (Sayeset al., 2007).

The use of in vitro systems will benefit from further standardi-zation of protocols, data interpretation and validation, especiallywith regard to a quantitative extrapolation to physiological effectsin vivo. This is important to promote their broad use and accep-tance in regulatory risk assessments. It was demonstrated thatalready today they can serve as useful elements in a weight-of-evi-dence approach.

Research on the effects of particulates is ongoing within the sci-entific community comprising the investigation of particulate up-take into living organisms and fate in the human body likeaccumulation in specific organs. A thorough exposure analysis asincluded in the current assessment therefore is indispensable evenif a low hazard is expected. Beyond the standard exposure esti-mate, additional considerations on biodegradability were includeddue to the particulate nature of the material. Numerous publica-tions confirm the biodegradability of the hydroxyapatite compo-nent, while gelatin can be assumed to be cleaved in the stomach.Additional information might however be needed in such caseswhere particles are persistent and therefore may have the poten-tial to induce prolonged inflammation, oxidative stress or dislocateto different compartments in the body.

With the integrated assessment as it was described here forHPC, one of the major scientific and ethical objectives of the EUand beyond, i.e. the use of alternative methods to the currentanimal tests for risk assessment is pursued (cf. DirectiveEC86/609). For each new material it needs to be investigatedif a similar approach is appropriate based on the actual state-of-the-art.


5. Conclusions

Based upon the presented experimental results and consideringthe data on the constituents, it can be concluded that HPC does notpose a concern to human health for the intended use in oral careformulations. This strategy provides an example of a risk assess-ment for a cosmetic use of small particles without the need to per-form additional animal studies. In vitro screening tools are alreadycapable to serve as useful elements of weight-of-evidence assess-ments and should be further developed and used for suchpurposes.

Conflict of interest statement

The authors declare no conflict of interest. The authors areemployees of Henkel AG & Co. KGaA.

Funding source statement

The work was funded by Henkel AG & Co. KGaA and SusTechDarmstadt GmbH & Co. KG.

Acknowledgments

The authors would like to thank Elke Lehringer for assisting inthe literature search and for editorial assistance; Karin Michel foradvice on the literature search strategy; Burkhard Eschen andLothar Kintrup for providing TEM analyses; Joachim Kremer, AnjaFischer and Jürgen Kreutz for carrying out cell culture experimentsand HET-CAM; Birgit Eicker for statistical evaluation of cell cultureexperiments; Adolf-Peter Barth, Christel Adomat and PetraRenk-Schmehl for advice on clinical studies; Christiane Schüle,André Schirlitz, Badr Nfissi and former colleagues from SusTechDarmstadt (Thilo Poth, Holger Franke, Carola Braunbarth) for help-ful discussions and test sample preparation; Adolf-Peter Barth,Werner Schuh and Frederike Wiebel for manuscript review.

References

Aam, B.B., Fonnum, F., 2007. Carbon black particles increase reactive oxygen speciesformation in rat alveolar macrophages in vitro. Arch. Toxicol. 81, 441–446.

Addy, M., 2002. Dentine hypersensitivity: new perspectives on an old problem. Int.Dent. J. 52, 367–375.

Aframian, D.J. et al., 2006. The distribution of oral mucosal pH values in healthysaliva secretors. Oral Dis. 12, 420–423.

Albrecht, C. et al., 2007. Surface-dependent quartz uptake by macrophages:potential role in pulmonary inflammation and lung clearance. Inhal. Toxicol.19 (Suppl. 1), 39–48.

Albrecht, C. et al., 2009. Evaluation of cytotoxic effects and oxidative stress withhydroxyapatite dispersions of different physicochemical properties in ratNR8383 cells and primary macrophages. Toxicol. In Vitro 23, 520–530.

Araujo, J. et al., 2009. Effect of polymer viscosity on physicochemical properties andocular tolerance of FB-loaded PLGA nanospheres. Colloids Surf. B Biointerfaces72, 48–56.

Arts, C.J.J. et al., 2006. The use of a bioresorbable nano-crystalline hydroxyapatitepaste in acetabular bone impaction grafting. Biomaterials 27, 1110–1118.

Balduzzi, M. et al., 2004. In vitro effects on macrophages induced by noncytotoxicdoses of silica particles possibly relevant to ambient exposure. Environ. Res. 96,62–71.

Bardow, A. et al., 2004. Saliva. In: Miles, T.S. et al. (Eds.), Clinical Oral Physiology.Quintessence, Copenhagen. pp. 17–18, 30–33.

Bloebaum, R.D. et al., 1998. Dissolution of particulate hydroxyapatite in amacrophage organelle model. J. Biomed. Mater. Res. 40, 104–114.

Borm, P.J. et al., 2006. The potential risks of nanomaterials: a review carried out forECETOC. Part Fibre Toxicol. 3, 11.

Bridges, J.W., Bridges, O., 2004. Integrated risk assessment and endocrine disrupters.Toxicology 205, 11–15.

Brown, D.M. et al., 2001. Size-dependent proinflammatory effects of ultrafinepolystyrene particles: a role for surface area and oxidative stress in theenhanced activity of ultrafines. Toxicol. Appl. Pharmacol. 175, 191–199.

Busso, M., 2009. Calcium hydroxylapatite (Radiesse): safety, techniques and painreduction. J. Drugs Dermatol. 8, s21–s23.

Cui, F.Z. et al., 1996. Biodegradation of a nano-hydroxyapatite/collagen compositeby peritoneal monocyte macrophages. Cells Mater. 6, 31–44.

Del Fabbro, M. et al., 1998. Fluid dynamics of gingival tissues. J. Periodontal Res. 33,328–334.

Donaldson, K. et al., 2005. Combustion-derived nanoparticles: a review of theirtoxicology following inhalation exposure. Part Fibre Toxicol. 2, 10.

EC, 1976. Council Directive 76/768/EEC on the approximation of the laws of theMember States relating to cosmetic products. OJ L 262, 27.7.1976. Directive aslast amended by Commission Directive 2002/34/EC (OJ L 102, 18.4.2002, p. 19).169 ff.

EC, 2003. Directive 2003/15/EC of the European Parliament and of the Council of 27February 2003 amending Council Directive 76/768/EEC on the approximation ofthe laws of the Member States relating to cosmetic products. OJ L 66, 11.3.2003.26 ff.

ECETOC, Workshop on Testing Strategies to Establish the Safety of Nanomaterials.7–8 November 2005, Barcelona, Workshop Report No. 7, 2006.

ECHA, Guidance on information requirements and chemical safety assessment,2007. p. 182.

Ericsson, Y., 1949. Enamel apatite solubility. Acta Odont. Scand. 8, 60–71.Eskes, C. et al., 2005. Eye irritation. Altern. Lab. Anim. 33 (Suppl. 1), 47–81.EU, Guidance to Regulation (EC) No 1272/2008 on classification, labelling and

packaging (CLP) of substances and mixtures, ECHA Reference: ECHA-09-G-02-EN, 2009.

Evans, E.J., 1991. Toxicity of hydroxyapatite in vitro: the effect of particle size.Biomaterials 12, 574–576.

Fjaertoft, M. et al., 1992. Micropuncture measurements of interstitial fluid pressurein normal and inflamed gingiva in rats. J. Periodontal Res. 27, 534–538.

Grandjean-Laquerriere, A. et al., 2005. Importance of the surface area ratio oncytokines production by human monocytes in vitro induced by varioushydroxyapatite particles. Biomaterials 26, 2361–2369.

Grandjean-Laquerriere, A. et al., 2004. The effect of the physical characteristics ofhydroxyapatite particles on human monocytes IL-18 production in vitro.Biomaterials 25, 5921–5927.

Guo, X. et al., 1999. Biocompatibility of self-designed absorbable hydroxyapatite/poly (DL-lactide) composites. Sheng Wu Yi Xue Gong Cheng Xue Za Zhi 16, 135–139.

Gupta, H. et al., 2009. Sparfloxacin-loaded PLGA nanoparticles for sustained oculardrug delivery. Nanomedicine 6, 324–333.

Hall, B. et al., 2007. European consumer exposure to cosmetic products, aframework for conducting population exposure assessments. Food Chem.Toxicol. 45, 2097–2108.

Henkel, 2005. Small particles with great potential. Press release Henkel GmbH &Co. KGaA. http://www.henkel.com/cps/rde/xchg/SID-0AC8330A-4805D2D5/henkel_com/hs.xsl/12028_5860_COE_HTML.htm.

Heraeus Kulzer GmbH, 2005. Ostim, Product Information; Scientific Information.http://www.ostim-dental.de.

Hirvonen, M.R. et al., 1996. Heat shock proteins and macrophage resistance to thetoxic effects of nitric oxide. Biochem. J. 315 (Pt. 3), 845–849.

Holsapple, M.P. et al., 2005. Research strategies for safety evaluation ofnanomaterials, part II: toxicological and safety evaluation of nanomaterials,current challenges and data needs. Toxicol. Sci. 88, 12–17.

Huang, J. et al., 2004. In vitro assessment of the biological response to nano-sizedhydroxyapatite. J. Mater. Sci. Mater. Med. 15, 441–445.

Huber, F.X. et al., 2006a. First histological observations on the incorporation of anovel nanocrystalline hydroxyapatite paste OSTIM in human cancellous bone.BMC Musculoskelet. Disord. 7, 50.

Huber, F.X. et al., 2007. Evaluation of a novel nanocrystalline hydroxyapatite pasteand a solid hydroxyapatite ceramic for the treatment of critical size bonedefects (CSD) in rabbits. J. Mater. Sci. Mater. Med. 9, 33–38.

Huber, F.X. et al., 2006b. Open reduction and palmar plate-osteosynthesis incombination with a nanocrystalline hydroxyapatite spacer in the treatment ofcomminuted fractures of the distal radius. J. Hand Surg. [Br] 31, 298–303.

Huber, F.X. et al., 2006c. The use of nanocrystalline hydroxyapatite for thereconstruction of calcaneal fractures: preliminary results. J. Foot Ankle Surg.45, 322–328.

Huber, F.X. et al., 2006d. Void filling of tibia compression fracture zones using anovel resorbable nanocrystalline hydroxyapatite paste in combination with ahydroxyapatite ceramic core: first clinical results. Arch. Orthop. Trauma Surg.126, 533–540.

Jantova, S. et al., 2008. In vitro effects of fluor-hydroxyapatite, fluorapatite andhydroxyapatite on colony formation, DNA damage and mutagenicity. Mutat.Res. 652, 139–144.

Kannan, T.P. et al., 2004a. Chromosome aberration test for hydroxyapatite in sheep.Med. J. Malaysia 59 (Suppl. B), 168–169.

Kannan, T.P. et al., 2004b. In vivo chromosome aberration test for hydroxyapetite inmice. Med. J. Malaysia 59 (Suppl. B), 115–116.

Kilian, O. et al., 2002. Einfluss von Ostim kombiniert mit autologen thrombozytärenWachstumsfaktoren auf die Knochendefektheilung in vivo. Biomaterialien 3,70–73.

Kumta, P.N. et al., 2005. Nanostructured calcium phosphates for biomedicalapplications: novel synthesis and characterization. Acta Biomater. 1, 65–83.

Landi, E. et al., 2008. Biomimetic Mg-substituted hydroxyapatite: from synthesis toin vivo behaviour. J. Mater. Sci. Mater. Med. 19, 239–247.

Laquerriere, P. et al., 2003. Importance of hydroxyapatite particles characteristics oncytokines production by human monocytes in vitro. Biomaterials 24, 2739–2747.

http://www.henkel.com/cps/rde/xchg/SID-0AC8330A-4805D2D5/henkel_com/hs.xsl/12028_5860_COE_HTML.htmhttp://www.henkel.com/cps/rde/xchg/SID-0AC8330A-4805D2D5/henkel_com/hs.xsl/12028_5860_COE_HTML.htmhttp://www.ostim-dental.de


Larsen, M.J., Pearce, E.I., 1997. A computer program for correlating dental plaque pHvalues, cH+, plaque titration, critical pH, resting pH and the solubility of enamelapatite. Arch. Oral Biol. 42, 475–480.

Laschke, M.W. et al., 2007. Injectable nanocrystalline hydroxyapatite paste for bonesubstitution: in vivo analysis of biocompatibility and vascularization. J. Biomed.Mater. Res. B Appl. Biomater. 82, 494–505.

Lewinski, N. et al., 2008. Cytotoxicity of nanoparticles. Small 4, 26–49.Li, H. et al., 2008. Preparation and biological safety evaluation of porous n-HA/PA66

composite. Sheng Wu Yi Xue Gong Cheng Xue Za Zhi 25, 1126–1129.Liu, C. et al., 1997. Evaluation of the biocompatibility of a nonceramic

hydroxyapatite. J. Endod. 23, 490–493.Mayhew, T.M. et al., 2009. A review of recent methods for efficiently quantifying

immunogold and other nanoparticles using TEM sections through cells, tissuesand organs. Ann. Anat. 191, 153–170.

McNamee, P. et al., 2009. A tiered approach to the use of alternatives to animaltesting for the safety assessment of cosmetics: eye irritation. Regul. Toxicol.Pharmacol. 54, 197–209.

Merck, Sicherheitsdatenblatt, Hydroxylapatit für die Biochromatographie (15 lm)(0–30 lm) Art. 105199, 2000.

Mitschik, S. et al., 2008. Effects of particulate matter on cytokine productionin vitro: a comparative analysis of published studies. Inhal. Toxicol. 20, 399–414.

Mosmann, T., 1983. Rapid colorimetric assay for cellular growth and survival:application to proliferation and cytotoxicity assays. J. Immunol. Methods 65,55–63.

Motskin, M. et al., 2009. Hydroxyapatite nano and microparticles: correlation ofparticle properties with cytotoxicity and biostability. Biomaterials 30, 3307–3317.

Palmer, L.C. et al., 2008. Biomimetic systems for hydroxyapatite mineralizationinspired by bone and enamel. Chem. Rev. 108, 4754–4783.

Pankratov, A.S. et al., 1994. The immunotropic and allergenic activity ofhydroxyapatite with an ultrahigh degree of dispersion. Stomatologiia (Mosk)73, 37–40.

Pardeike, J., Muller, R.H., 2009. Dermal and ocular safety of the new phospholipaseA2 inhibitors PX-18 and PX-13 formulated as drug nanosuspension. J. Biomed.Nanotechnol. 5, 437–444.

Rajab, N.F. et al., 2004. DNA damage evaluation of hydroxyapatite on fibroblast cellL929 using the single cell gel electrophoresis assay. Med. J. Malaysia 59 (Suppl.B), 170–171.

Rauschmann, M.A. et al., 2005. Nanocrystalline hydroxyapatite and calciumsulphate as biodegradable composite carrier material for local delivery ofantibiotics in bone infections. Biomaterials 26, 2677–2684.

Rudin, V.N. et al., 1994. Preparation of ‘‘Ostim Apatite’’ for stimulating growth inbone tissue. European Patent Application No. 0 664 133 A1.

Rumpel, E. et al., 2006. The biodegradation of hydroxyapatite bone graft substitutesin vivo. Folia Morphol. (Warsz) 65, 43–48.

Sauerwein, 2000. Zahnerhaltungskunde. Thieme, Stuttgart.Sayes, C.M. et al., 2007. Assessing toxicity of fine and nanoparticles: comparing

in vitro measurements to in vivo pulmonary toxicity profiles. Toxicol. Sci. 97,163–180.

SCCP, The SCCP’s (Scientific Committee on Consumer Products) Notes of Guidancefor the Testing of Cosmetic Ingredients and Their Safety Evaluation. 6thRevision, 2006.

Scheel, J. et al., 2009. Exposure of the murine RAW 264.7 macrophage cell line tohydroxyapatite dispersions of various composition and morphology:assessment of cytotoxicity, activation and stress response. Toxicol. In Vitro23, 531–538.

Schmid, O. et al., 2009. Dosimetry and toxicology of inhaled ultrafine particles.Biomarkers 14 (Suppl. 1), 67–73.

Schroeder, H.E., 2000. Orale Strukturbiologie. ISBN 3-13-540905-8.Schweikl, H., Schmalz, G., 1996. Toxicity parameters for cytotoxicity testing of

dental materials in two different mammalian cell lines. Eur. J. Oral Sci. 104,292–299.

Shamsuria, O. et al., 2004. In vitro cytotoxicity evaluation of biomaterials on humanosteoblast cells CRL-1543; hydroxyapatite, natural coral andpolyhydroxybutarate. Med. J. Malaysia 59 (Suppl. B), 174–175.

Steiling, W. et al., 1999. The HET-CAM, a useful in vitro assay for assessing the eyeirritation properties of cosmetic formulations and ingredients. Toxicol. In Vitro13, 375–384.

Suzina, A.H. et al., 2004. Mutagenicity of CORAGRAF and REKAGRAF in the Amestest. Med. J. Malaysia 59 (Suppl. B), 105–106.

Takamura, K. et al., 1994. Evaluation of carcinogenicity and chronic toxicityassociated with orthopedic implants in mice. J. Biomed. Mater. Res. 28, 583–589.

Thorwarth, W.M. et al., 2004. Untersuchung zur knöchernen Regeneration ossärerDefekte unter Anwendung eines nanopartikulären Hydroxylapatits (Ostim).Implantologie 12, 21–32.

Torrado, A. et al., 2005. Cytotoxicity of a new toothpaste based on an ion exchangeresin mixture. Am. J. Dent. 18, 267–269.

Trofimov, V.V. et al., 1996. A biological compatibility study of hydroxyapatite.Stomatologiia (Mosk) 75, 20–22.

Ulstrup, A.K., Hartzen, S.H., 2006. Leptotrichia buccalis: a rare cause of bacteraemiain non-neutropenic patients. Scand. J. Infect. Dis. 38, 712–716.

Unfried, K. et al., 2007. Cellular responses to nanoparticles: target structures andmechanisms. Nanotoxicology 1, 52–71.

Wahl, D.A., Czernuszka, J.T., 2006. Collagen–hydroxyapatite composites for hardtissue repair. Eur. Cell Mater. 11, 43–56.

Webster, T.J., Ahn, E.S., 2007. Nanostructured biomaterials for tissue engineeringbone. Adv. Biochem. Eng. Biotechnol. 103, 275–308.

Weed, D.L., 2005. Weight of evidence: a review of concept and methods. Risk Anal.25, 1545–1557.

Wenisch, S. et al., 2003. In vivo mechanisms of hydroxyapatite ceramic degradationby osteoclasts: fine structural microscopy. J. Biomed. Mater. Res. A 67, 713–718.

Werner, U., 2003. In Situ Gelling Nasal Inserts for Prolonged Drug Delivery.Fachbereich Biologie, Chemie, Pharmazie. Freie Universität Berlin, Berlin.

Wolf, N.B. et al., 2009. Influences of opioids and nanoparticles on in vitro woundhealing models. Eur. J. Pharm. Biopharm. 73, 34–42.

Yamada, T., 1999. Textbook of Gastroenterology. Lippincott Williams & Wilkins.Ye, L. et al., 2004. Genotoxicity of a new NanoHA-PA66 root filling material in vitro.

Hua Xi Kou Qiang Yi Xue Za Zhi 22, 93–95.

Integrated risk assessment of a hydroxyapatite–protein-composite for use in oral care products: A weight-of-evidence case studyIntroductionMaterials and methodsIntegrated risk assessment and weight-of-evidence (WoE) analysisStrategy of literature search and reviewTest samplesHET-CAM (Hen’s egg test-chorioallantoic membrane) to assess mucosal membrane irritationCytotoxicity and inflammationViability of 3T3 fibroblasts (MTT test)Mucosa-like human corneal epithelial (HCE) tissuMacrophage assaysClinical studies

Exposure and kineticsGeneral exposure estimate and overviewDissolution behavior of HPC in simulated gastric fluidHCE penetration screening (Cryo-TEM)

ResultsAssessment and testing strategy for hazardBasic toxicological profile of the constituents. Review of supplier information and publicly available dataHydroxyapatite (HA) and composites thereofGelatin

In vitro screening testsHET-CAMViability of 3T3 fibroblasts (MTT test)Mucosa-like tissue models (LDH and IL-1α releaseMacrophage assays

Clinical studiesSummary of hazard evaluation

Exposure analysis/toxicokineticsGeneral exposure estimate according to regulatory proposalConsideration of mucosal penetration potential (direct contact with oral mucosa)Theoretical calculationConsiderations regarding penetration based on literature and expert consultationScreening for possible translocation of HPC into cells or cell layers using Cryo-TEM

Physico-chemical properties (dissolution kinetics)Summary exposure assessment

Risk assessment

DiscussionConclusionsConflict of interest statementFunding source statementAcknowledgmentsReferences

regulatory toxicology and pharmacology · 2017. 1. 6. · (us national library of medicine),...

Documents