TTC NON-CANCER ORAL DATABASES Dr Sue Barlow Consultant in toxicology & risk assessment suebarlow@mistral.co.uk EUROTOX CEC on TTC 13 September 2015
Overview of presentation Rationale for TTC values for non-cancer endpoints Development of databases for TTC values for non-cancer endpoints The first database Munro Are the Munro TTC values robust? Other databases Summary and conclusions
RATIONALE FOR TTC VALUES FOR NON-CANCER ENDPOINTS Based on the fundamental principles that: For endpoints other than cancer and genotoxicity, chemical toxicity is related to dose and duration of exposure For any individual chemical, there is a threshold below which non-cancer toxicity does not occur
RATIONALE FOR TTC VALUES FOR NON-CANCER ENDPOINTS For chemicals on which we have toxicity data, that share broadly similar functional groups, can human exposure threshold values be identified below which toxicity for non-cancer effects is unlikely? For chemicals of unknown toxicity, can we use these human exposure thresholds as values below which there would be a low probability of adverse effects on health?
DEVELOPMENT OF DATABASES FOR TTC VALUES FOR NON-CANCER EFFECTS Collate oral toxicity data on repeat-dose study types (subchronic, chronic, reproduction, development) for a large number of chemicals Establish criteria for inclusion/exclusion of studies in the database to ensure included data are reliable Were the study methods sound? Were the results adequately reported? Do the data support the authors NO(A)EL? Conduct quality control checks for each study against the criteria
DEVELOPMENT OF DATABASES FOR TTC VALUES FOR NON-CANCER EFFECTS Classify each chemical into one of the three Cramer structural classes For each Cramer structural class, plot potency distribution curve as the lowest NO(A)EL for each chemical Determine the 5 th percentile value of each cumulative plot Divide the 5 th percentile value by an uncertainty factor (100) to derive a TTC value for each Cramer class TTC values can be expressed as µg/person/day (60kg adult) or µg/kg bw/day (preferable)
MUNRO DATABASE Oral toxicity data from sub-chronic, chronic, reproductive, and developmental toxicity studies 613 substances (2941 NOELs) covering industrial chemicals, agrochemicals, food chemicals, consumer chemicals Focused on rodents and rabbits (species with few animals per dose group excluded) For NOELs from subchronic studies, adjustment factor of 3 applied to approximate chronic NOELs Munro I, Ford RA, Kennepohl E, Sprenger JG. Correlation of structural class with no-observed effect levels: a proposal for establishing a threshold of concern. Food Chem. Toxicol. 34, 829-867, 1996
MUNRO DATABASE Derivation of TTC values Number of chemicals in each Cramer structural class: class I 137 class II 28 class III 448 Plotted distributions of toxic potencies, expressed as the lowest NOEL for each chemical, for each Cramer class
Munro et al. 1996 Class I II III 5%ile NOEL (mg/kg/day) 3.0 0.91 0.15 Human threshold (µg per day) * 1800 540 90 * NOEL/100 X 60kg bwt
MUNRO TTC VALUES Cramer class µg/person **/day TTC value µg/kg bw/day I 1800 30 II 540 9 III 90 1.5 OPs & carbamates * 18 0.3 * Value set by Munro et al. 1999 ** For 60 kg adult
STRENGTHS & WEAKNESSES OF THE Weaknesses MUNRO DATABASE Most studies published before 1990 NOELs proposed by study authors or EPA IRIS accepted without further review Includes some substances with structural alert for genotoxicity Later studies may have used more up-to-date protocols and may show lower NO(A)ELs For many chemicals, lowest NOEL is from subchronic study (as in other TTC databases) Few substances in Cramer Class II (as in all TTC databases) Wide overlap in NOEL distribution curves between Cramer classes
STRENGTHS & WEAKNESSES OF THE MUNRO DATABASE Strengths Studies selected from established reports/databases containing well-validated chemical structures and toxicity data (i.e. NTP, JECFA, IRIS, DART) Included most of the published chronic toxicity studies available at that time; acute and short-term studies excluded Oral studies also cover metabolite toxicity Good separation of 5 th percentile NOEL values between Cramer classes Conservative TTC values Electronic version available from EFSA: http://www.efsa.europa.eu/en/supporting/pub/159e
ROBUSTNESS OF MUNRO TTC VALUES FOR PARTCULAR ENDPOINTS OR TYPES OF CHEMICAL Do the Munro TTC values cover all chemical domains? Do they adequately cover chemicals with particular types of use? Cosmetics-related Food contact materials Pesticides Do later databases and analyses produce TTC values similar to Munro? Do the Munro TTC values adequately cover reproductive and developmental endpoints?
RepDose DATABASE Developed at the Fraunhofer Institute of Toxicology & Experimental Medicine (ITEM), Germany, under CEFIC LRI Publicly accessible at http://fraunhofer-repdose.de/ Bitsch A, Jacobi S, Melber C, Wahnschaffe U, Simetska N, Mangelsdorf I. REPDOSE: a database on repeated dose toxicity studies of commercial chemicals - a multifunctional tool. Regul. Toxicol. Pharmacol. 46, 202-210, 2006
RepDose DATABASE Oral and inhalational toxicity data on subacute, subchronic and chronic studies Currently includes c.1200 existing industrial chemicals, 3450 studies Includes defined commercial organic chemicals with limited number of functional groups Excludes complex molecules, pharmaceuticals, inorganics, metal compounds, mixtures Includes oral and inhalational studies on rats, mice, dogs Guideline studies or studies prepared for regulatory purposes preferred, supplemented by other studies
RepDose DATABASE Applied following adjustment factors to obtain NOELs approximating to chronic dosing: 3 for LOEL:NOEL 2 for subchronic:chronic (REACH-recommended ) 6 for subacute:chronic (REACH-recommended ) Applied adjustment factors to obtain TTC values: Allometric assessment factors of 4/7 for rat/mouse for interspecies differences Further factor of 25 applied to account for remaining interand intra-species differences i.e. overall adjustment factor is 100 for rats (same as Munro UF), 175 for mice
RepDose DATABASE Oral TTC values have been derived from 561 chemicals in the RepDose database (40% overlap of chemicals with Munro database) Number of chemicals in each Cramer structural class: class I 109 class II 12 class III 400 Plotted distributions of toxic potencies, expressed as LOELs, for each Cramer class Tluczkiewicz I, Buist HE, Martin MT, Mangelsdorf I, Escher SE. Improvement of the Cramer classification for oral exposure using the database TTC RepDose A strategy description. Regul. Toxicol. Pharmacol. 61, 340-350, 2011
Munro & RepDose Databases Cumulative distributions of LOELs (mmol/kg bw/d) Comparison RepDose & Munro Munro RepDose oral Repdose inhalational RepDose oral studies only Cramer class I Cramer class II Cramer class III Plots taken from Escher et al. The TTC concept for industrial chemicals. Are inhalation thresholds needed? EUROTOX Poster, 2008
RepDose TTC VALUES Molar TTC values converted back to µg/person/day for comparison with Munro by multiplying by median MW of all the chemicals in the database (220 g/mol) Cramer class * TTC threshold value (µg/person/day) RepDose (from NOELs in molar units) RepDose (from NOELs in mass units) Munro (from NOELs in mass units) I 1930 1500 1800 III 74 93 90 * TTC values only for Cramer classes I and III (insufficient data in Class II)
Weaknesses STRENGTHS & WEAKNESSES OF THE REPDOSE DATABASE Database searchable by individual substance or by duration of studies Difficult to find list of all the chemicals currently in the database For many chemicals, lowest NOEL is from subchronic study (as in other TTC databases) No reproductive or developmental studies included Few substances in Cramer Class II (as in all TTC databases)
STRENGTHS & WEAKNESSES OF THE REPDOSE DATABASE Strengths Chemicals selected by availability of toxicity data from peerreviewed, national and international documents German MAK documents, Reports of German Advisory Board for Existing Chemicals (BUA), EU RARs, WHO EHCs, HPV chemicals Study quality (A,B,C,D) is evaluated and recorded Excludes organophosphates Excludes substances with structural alert for genotoxicity Good separation of 5 th percentile NOEL values between Cramer classes Includes analyses in which doses converted to molar amounts to eliminate data variability in potency due to MW differences
COSMOS TTC DATABASE Focus on cosmetics-related chemicals Oral toxicity data on subchronic, chronic, reproduction, developmental, neurotoxicity and immunotoxicity studies Rat, mouse, monkey, dog, rabbit Includes 558 chemicals for which N(L)OELs available Applied following adjustment factors to obtain NOELs approximating to chronic dosing (except developmental and reproductive organ effects): 3 for LOEL:NOEL 3 for subchronic:chronic 6 for subacute:chronic Overlap of 179 substances between COSMOS and Munro
COSMOS TTC DATABASE Derivation of TTC values Work in progress TTC values due to be available end 2015
STRENGTHS & WEAKNESSES OF THE COSMOS TTC DATABASE Strengths All data obtained from publicly available reports or databases (except CERES FCS), guideline studies preferred Munro database; US FDA PAFA (Priority-based Assessment of Food Additives) database; FDA CERES (Chemical Evaluation and Risk Estimation System) FCS database; USEPA Toxicity Reference Database (ToxRefDB); ILSI Research Foundation Developmental Toxicity Database (ILSI DevTox); Opinions of the European Commission s Scientific Committee on Consumer Safety (SCCS); NTP database; REACH Registered Substance Database of the European Chemicals Agency
STRENGTHS & WEAKNESSES OF THE COSMOS TTC DATABASE Strengths Data meticulously curated before entry into database Rigorous inclusion/exclusion criteria applied Detailed QC against original publications for final decision on 20% of the NOELs in the database on substances below the 10 th percentile for potency (NOELs) on substances with conflicting NOELs assigned by different authors/bodies Quality comments and evaluations recorded
FOOD CONTACT MATERIALS DATABASE Database of 232 substances used to manufacture FCMs, for which an ADI or TDI established by EC SCF or EFSA Number of chemicals in each Cramer structural class: class I - 112, class II - 7, class III - 113 8 substances present in both FCM and Munro databases Comparison of distribution curves of NOELs for Munro + FCMs with Munro alone showed: Cramer class I curves very close; Cramer class III curves superimposed 96% of the 845 substances had ADIs/TDIs lower than the relevant TTC value, confirming TTC approach is conservative Pinalli R, Croera C, Theobald A, Feigenbaum A. Threshold of toxicological concern approach for the risk assessment of substances used for the manufacture of plastic food contact materials. Trends Food Sci. Technol. 22, 523-534, 2011
PESTICIDES DATABASE 279 active pesticide substances, all in Cramer class III Database built from EU peer review evaluations published by EFSA Inorganic substances and substances without an ADI excluded 5 th percentile NOEL value: 0.2 mg/kg bw/day, compared with Munro 0.15mg kg bw/day Difference due to exclusion of substances with neurotoxicity alert (AChE inhibition) from the pesticides database Feigenbaum A, Pinalli R, Giannetto M, Barlow S. Reliability of the TTC approach: Learning from inclusion of pesticide active substances in the supporting database. Food Chem. Tox 75, 25-38, 2015
COMPARISON OF CLASS III NOEL DISTRIBUTION CURVES FOR PESTICIDES, FCM, MUNRO
REPRODUCTIVE & DEVELOPMENTAL TOXICITY EFSA (2012) opinion analysed EU database on industrial chemicals classified for reproductive and developmental toxicity Confirmed findings of Kroes et al. 2004, Bernauer et al. 2009, van Ravenzwaay et al. 2010, on reproductive toxicants that Munro TTC values adequately protective
KALKHOF DATABASE Data obtained from German regulatory database on new industrial chemicals data submissions required under EU chemicals law 1982-2002 All chemicals >90% purity 28-day OECD GL studies on 776 chemicals 90-day OECD GL studies on 85 chemicals Applied REACH recommended adjustment factors to obtain NOELs approximating to chronic dosing: 6 for subacute (28-d):chronic 2 for subchronic (90-d):chronic
KALKHOF DATABASE Confirmed that Munro TTC values are conservative 5 th percentile NO(A)EL value (mg/kg bw/d) Cramer class Kalkhof 28-d Kalkhof 90-d Munro I 2.5 24.5 3.0 II 2.5 0.91 III 1.6 1.0 0.15 Kalkhof H, Herzler M, Stahlmann R, Gundert-Remy U. Threshold of toxicological concern values for non-genotoxic effects in industrial chemicals: re-evaluation of the Cramer classification. Arch. Toxicol. 86, 17-25, 2012
SUMMARY & CONCLUSIONS Human exposure threshold values originally derived by Munro et al. have been confirmed as suitably conservative by subsequent analyses using different databases They remain the most widely used oral TTC values If exposure uncertain, or worst case exposure cannot be estimated, do not use TTC approach It is a probability-based tool A substance with an exposure below the relevant non-cancer TTC value may still pose a potential risk, with a probability estimated to lie between zero and 5%