1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
|
---
author: Christoph Helma
date: October 12, 2022
title: Variability of chronic rodent bioassays
---
# Content
Rodent Carcinogenicity
: E Gottmann, S Kramer, B Pfahringer and C Helma\
*Data quality in predictive toxicology: reproducibility of rodent
carcinogenicity experiments*\
Environ Health Perspect 109:509--514 (2001)\
<https://doi.org/10.1289/ehp.01109509>
Lowest observed adverse effect level (LOAEL)
: C Helma, D Vorgrimmler, D Gebele, M Gütlein, B Engeli, J Zarn,
B Schilter and E Lo Piparo\
*Modeling Chronic Toxicity: A Comparison of Experimental Variability
With (Q)SAR/Read-Across Predictions*\
Front Pharmacol 9 (2018)\
<https://doi.org/10.3389/fphar.2018.00413>
# Carcinogenicity Data
- Carcinogenic Potency Database(CPDB, Gold 1997)
- 1,289 unique compounds
- 2 Subsets
- National Toxicology Program (NTP)
- General literature
- 121 common compounds in both subsets
# Carcinogenicity Classification
- **57%** concordant classifications (69/121 compounds, 39
carcinogens, 30 non-carcinogens)
Rats
: 62% concordant classifications
Mice
: 49% concordant classifications
Multi species carcinogens
: 58% concordant classifications
Multi organ carcinogens:
: 52% concordant classifications
- poor reproducibility of sex, species and organ specific effects
# Carcinogenicity TD50's
# Carcinogenicity caveats
- low sample size
- no standardized protocols for literature data
Gold et al. (1987)
: - 38 compounds from the literature
- 93% reproducibility for rats
- 76% for mice
- 34 studies were published by the same authors (!)
# LOAEL Data
Chronic (\>180 days) lowest observed effect levels (LOAEL) for rats
(Rattus norvegicus) after oral (gavage, diet, drinking water)
administration
Nestlé Database
: 567 LOAEL values for 445 unique chemical structures from the
literature (Mazzatorta et al., 2008)
Swiss Food Safety and Veterinary Office (FSVO) Database
: 493 rat LOAEL values for 381 unique chemical structures from
pesticide evaluations (Zarn et al., 2011, 2013)
- European Food Safety Authority (EFSA) (EFSA, 2014)
- Joint FAO/WHO Meeting on Pesticide Residues (JMPR) (WHO, 2011)
- US EPA (US EPA, 2011)
Combined dataset
: - compounds that occur in both databases
- 375 LOAEL values for 155 unique chemical structures
# LOAEL Variability
**Both** datasets contain substances with multiple measurements
All datasets have almost the same experimental variability (standard
deviations: 0.56 mg/kg_bw/day (Nestlé), 0.57 mg/kg_bw/day (FSVO), 0.56
mg/kg_bw/day (combined))
# LOAEL Correlation
As both databases contain duplicates medians were used for the
correlation plot and statistics
# LOAEL Experiments vs Predictions
# Conclusions
- Carcinogenicity classifications seem to be poorly reproducible (57%
concordant classifications for repeated experiments)
- Experimental LOAEL values have a variablity of approximately 1.5 log
units (orders of magnitude)
- Variability in chronic *in vivo* bioassays might be caused by
- biological complexity
- long term experimental conditions
- evaluation complexity
- statistical limitations (low number of animals/treatment)
- Good *in-silico* models have the same accuracy as biological
experiments (*in-vivo* and *in-vitro*) for **compounds in their
applicability domain**
\
\
<https://in-silico.ch/presentations/epa-nam-2022/>
|
