Accepted Manuscript Pharmacokinetic data reduce uncertainty in the acceptable daily intake for benzoic acid and its salts K. Zu, D.M. Pizzurro, T.A. Lewandowski, J.E. Goodman PII:
S0273-2300(17)30208-8
DOI:
10.1016/j.yrtph.2017.07.012
Reference:
YRTPH 3874
To appear in:
Regulatory Toxicology and Pharmacology
Received Date: 14 April 2017 Revised Date:
13 July 2017
Accepted Date: 14 July 2017
Please cite this article as: Zu, K., Pizzurro, D.M., Lewandowski, T.A., Goodman, J.E., Pharmacokinetic data reduce uncertainty in the acceptable daily intake for benzoic acid and its salts, Regulatory Toxicology and Pharmacology (2017), doi: 10.1016/j.yrtph.2017.07.012. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
2
Pharmacokinetic Data Reduce Uncertainty in the Acceptable Daily Intake for Benzoic Acid and Its Salts
3 4
Zu, K1; Pizzurro, DM1; Lewandowski, TA2; Goodman, JE1*
1
6
1
Gradient. 20 University Road, Cambridge, MA, 02138, USA
7
2
Gradient. 600 Stewart Street, Suite 1900, Seattle, WA, 98101, USA
9
SC
8
Corresponding Author:
M AN U
10
Julie E. Goodman, Ph.D., DABT, ATS, FACE
12
Gradient
13
20 University Road
14
Cambridge, MA 02138
15
Email:
[email protected]
AC C
EP
TE D
11
16
RI PT
5
1
ACCEPTED MANUSCRIPT
5 6
RI PT
SC
4
Introduction ............................................................................................................ 4 Methods .................................................................................................................. 7 Interspecies Comparison of Pharmacokinetics of Benzoic Acid and Its Salts ......... 8 3.1 Qualitative Interspecies Comparison of Pharmacokinetics ........................ 9 3.2 CSAF Derivation for Interspecies Differences in Pharmacokinetics ......... 11 3.2.1 CSAF Derivation .......................................................................... 12 3.2.2 Alternative Approaches .............................................................. 15 3.2.3 Discussion ................................................................................... 18 Intraspecies Comparison of Pharmacokinetics of Benzoic Acid and Its Salts in Humans ............................................................................................................. 20 Evaluation of Human Clinical Data of Sodium Benzoate ...................................... 23 Proposed ADI for Benzoic Acid and Its Salts ......................................................... 26
M AN U
1 2 3
References .................................................................................................................................... 28
TE D
32 33
Page
EP
18 19 20 21 22 23 24 25 26 27 28 29 30 31
Table of Contents
AC C
17
2
ACCEPTED MANUSCRIPT
Abstract
35
The current acceptable daily intake (ADI) for benzoic acid and its salts as food additives is 0-5 mg/kg bw-
36
day. This accounts for a total uncertainty factor (UF) of 100, which includes a default factor of 10 for
37
interspecies differences. Based on pharmacokinetic data in rodents and humans, we derived a chemical-
38
specific adjustment factor (CSAF) of 2 for the pharmacokinetic component of the interspecies UF.
39
Additional analyses indicate that this CSAF is conservative and interspecies differences between rats and
40
humans are likely closer to unity. Human clinical studies indicate that the pharmacokinetics of benzoic
41
acid and its salts are similar in children and adults, and that there is a lack of adverse events in humans at
42
doses comparable to the no observed adverse effect level (NOAEL) in rodents; this suggests that the
43
pharmacokinetic UF for intraspecies variability, as well as the pharmacodynamic components of the UFs,
44
may also be reduced, although we did not calculate to what degree. In conclusion, the total UF can be
45
reduced to 50 (2 for interspecies differences in pharmacokinetics, 2.5 for interspecies differences in
46
pharmacodynamics, and 10 for intraspecies variability), which would increase the ADI to 0-10 mg/kg bw-
47
day.
48
50
Keywords:
EP
49
TE D
M AN U
SC
RI PT
34
Benzoic Acid; Sodium Benzoate; Potassium Benzoate; Pharmacokinetics; Chemical-specific Adjustment
52
Factor; Data-derived Exploration Factor; Uncertainty Factor; Acceptable Daily Intake
AC C
51
3
ACCEPTED MANUSCRIPT
1
Introduction
54
Benzoic acid and its salts, particularly sodium and potassium benzoates (Figure 1), are commonly used as
55
preservatives in acidic foods, beverages, pharmaceuticals, and cosmetic products because they inhibit
56
mold, yeast, and certain bacterial growth (Brul and Coote, 1999; Nair, 2001; JECFA, 2015, 2016).
57
Sodium benzoate, benzoic acid, and precursor compounds are evaluated for safety as a group because,
58
upon ingestion and rapid absorption, they all result in circulating benzoic acid.
SC
RI PT
53
59
Benzoic acid and its salts have a long history of safe use as food additives. A number of scientific and
61
regulatory bodies worldwide have conducted comprehensive assessments of available acute and chronic
62
toxicity, reproductive/developmental toxicity, carcinogenicity, and genotoxicity studies, and concluded
63
that these substances have low toxicity and are generally safe for human consumption (Adams et al.,
64
2005; IOMC, 2000; JECFA, 1997; OECD, 2001; Nair, 2001; SCCP, 2005; EFSA, 2016). For example,
65
benzoic acid and its salts are "generally recognized as safe" (GRAS) for current use in food by the United
66
States Food and Drug Administration (US FDA, 1973a,b, 2016), and Food Standards Australia New
67
Zealand (FSANZ) evaluated the safety of benzoates in foods and concluded that "there were no issues
68
with benzoates" (FSANZ, 2016).
TE D
EP
69
M AN U
60
In addition to its use as a preservative, sodium benzoate was used historically to assess liver function (by
71
measuring hippuric acid formation in urine) with oral doses on the order of 5,000 mg (Borgstrom, 1949).
72
Sodium benzoate is currently used clinically to treat urea cycle disorders because it can increase waste
73
nitrogen excretion and improve clinical outcomes (IOMC, 2000). Currently, there are three US FDA-
74
approved drugs that contain sodium benzoate as one of the main active ingredients (Ammonul®, a generic
75
version of Ammonul®, and Ucephan®) (US FDA, 2017). Therapeutic doses of sodium benzoate can be as
76
high as 250-500 milligrams per kilograms of body weight per day (mg/kg bw-day), equivalent to 211.9-
77
423.7 mg/kg bw-day benzoic acid (IOMC, 2000). Such treatment typically results in improved clinical
AC C
70
4
ACCEPTED MANUSCRIPT
outcomes with no serious side effects reported (IOMC, 2000). Notably, these human pharmaceutical uses
79
of sodium benzoate that result in systemic exposures up to nearly two orders of magnitude higher than
80
exposure levels from dietary intake (IOMC, 2000) have not been previously considered in any safety
81
evaluation of benzoic acid or its salts for use as a food preservative.
RI PT
78
82
The World Health Organization (WHO) Joint Expert Committee on Food Additives (JECFA) first
84
presented an ADI of 0-5 mg/kg bw1 for benzoic acid and sodium benzoate in 1962 based on a four-
85
generation rat study and a total UF of 100 (JECFA, 1962). This study was conducted in 1960, and no
86
adverse effects were observed at up to the highest dose tested (i.e., the no observed adverse effect level
87
[NOAEL] of 500 mg/kg bw-day) (Kieckebusch and Lang, 1960). The basis of the UF was not specified
88
in the 1962 report (JECFA, 1962). In 1997, JECFA maintained this ADI for benzoic acid and confirmed
89
its use for benzoic acid and its salts (calcium, potassium, and sodium), as well as for its precursors
90
(benzaldehyde, benzyl acetate, benzyl alcohol, and benzyl benzoate); the rationale for applying a UF of
91
100 was not discussed in this assessment (JECFA, 1997). In 2000, JECFA reaffirmed the group ADI of
92
0-5 mg/kg bw-day and specified that the total UF of 100 included 10 for database uncertainty and 10 for
93
interspecies differences (IOMC, 2000). It is unclear why JECFA did not consider a UF for intraspecies
94
variability at this time, as a default UF of 100 (based on a factor of 10 for interspecies differences and
95
another factor of 10 for intraspecies variability) has historically been used when JECFA derived an ADI
96
from a NOAEL based on animal toxicity data (IPCS, 2005, 2009).
AC C
EP
TE D
M AN U
SC
83
97 98
Recently, JECFA conducted a comprehensive dietary assessment on benzoate exposure in the general
99
population and determined that high-end dietary exposures to benzoates in children (1-17 years of age),
100
may exceed the upper bound of the current ADI (0-5 mg/kg bw-day) (JECFA, 2015, 2016).
101
Consequently, the 48th Codex Committee on Food Additives lowered the level of benzoates in beverages
102
by 50-75% in 2016 (Joint FAO/WHO Food Standards Programme, 2016). However, given the long 1
Traditionally, JECFA has developed ADIs as ranges, from 0 to an upper limit.
5
ACCEPTED MANUSCRIPT
103
history of safe use of benzoates as food additives and therapeutic agents, the current ADI may not
104
appropriately reflect the general safety of dietary exposures.
105 Current risk assessment guidelines, such as those from the WHO International Programme on Chemical
107
Safety (IPCS) and the US Environmental Protection Agency (US EPA), encourage the use of human and
108
animal data to develop adjustment factors (AFs) (e.g., chemical-specific adjustment factors [CSAFs] and
109
data-derived exploration factors [DDEFs]) to replace the default UF values, where possible (IPCS, 2005,
110
2009; US EPA, 2014a). Such factors allow for the incorporation of quantitative pharmacokinetic and
111
pharmacodynamic chemical-specific data on interspecies differences or human variability into the risk
112
assessment process. In doing so, these chemical-specific values enhance the assessment by more fully
113
utilizing the entirety of the dataset available for a particular substance and reducing the uncertainty
114
surrounding the ultimate safety criteria value. Development of data-derived AFs in place of default UFs
115
for health-based toxicity criteria derivation is becoming more of a standard practice, and such values are
116
increasingly incorporated into human health risk assessments by a variety of agencies. JECFA, for
117
example, has set a precedent for using such AFs in previous assessments and has subdivided the default
118
UFs in their evaluations of dioxins (JECFA, 2002) and methylmercury (JECFA, 2004).
119
assessments from US EPA incorporate pharmacokinetic data and modeling to adjust default uncertainty
120
values to establish chronic oral and inhalation reference concentrations for environmental exposures (e.g.,
121
for lead [US EPA, 2013], perfluorooctanoic acid [PFOA], and perfluorooctane sulfonate [PFOS] [US
122
EPA, 2014b,c]).
SC
M AN U
TE D
EP
Recent
AC C
123
RI PT
106
124
WHO/IPCS guidance divides UFs into specific pharmacokinetic and pharmacodynamic components. As
125
illustrated in Figure 2, based on empirical analyses, the interspecies UF default value of 10 can be
6
ACCEPTED MANUSCRIPT
126
subdivided into a factor of 4 for pharmacokinetics and 2.5 for pharmacodynamics, while the intraspecies
127
UF default value of 10 can be subdivided into two equal subfactors of 3.16 (IPCS, 2005).2
128 The database on the safety, toxicity, and pharmacokinetics of benzoic acid and its salts has expanded
130
considerably since the ADI was developed in the early 1960s. We evaluated the available animal and
131
human pharmacokinetic data, and human clinical data to determine whether the current default UFs
132
should be updated, and, if so, to develop CSAFs for deriving the ADI for benzoic acid and its salts.
SC
RI PT
129
133
2
135
We conducted literature searches for the available pharmacokinetic studies of benzoic acid and its salts in
136
humans, experimental animals, and in vitro, as well as other human clinical studies, using PubMed and
137
Scopus to identify clinical and laboratory studies published through October 2016. We also considered
138
studies of precursor compounds including benzyl alcohol, benzyl acetate, and benzaldehyde, which are
139
exclusively metabolized to benzoic acid in both humans and rodents. We also searched the Drugs@FDA
140
online database, which includes detailed chemistry, pharmacology, and toxicology profiles for drug
141
products approved by US FDA to identify any potentially unpublished studies of benzoates. In addition,
142
we searched Tox Planet, an online database that compiles all safety assessment documents for chemicals
143
from national and international regulatory and research agencies, to identify existing comprehensive
144
assessments for benzoates. We cross-referenced the bibliographies of review articles and assessment
145
documents to identify additional relevant studies.
TE D
EP
AC C
146
Methods
M AN U
134
2 Based on an empirical analysis of the variability between various animal species and humans for a variety of pharmacokinetic parameters, the default UF for interspecies differences of 10 is subdivided into a factor of 4 (100.6) for pharmacokinetic and 2.5 (100.4) for pharmacodynamic differences (Renwick, 1993; IPCS, 1994). However, the database for intraspecies differences is not considered sufficient to justify uneven subdivision of the default UF of 10. This UF is therefore divided evenly into factors of 100.5 (3.16), which is supported by a later analysis of various kinetic parameters for 60 chemicals in humans (Renwick and Lazarus, 1998).
7
ACCEPTED MANUSCRIPT
147
We evaluated and integrated the available pharmacokinetic literature on benzoic acid and its salts in
148
humans, animals, and in vitro to qualitatively compare interspecies (i.e., human and animal) and
149
intraspecies (i.e., age-related) pharmacokinetics for these substances.
150
evaluation, we selected key studies for the CSAF derivation analysis if they contained sufficient
151
quantitative information to allow for quantitative analysis of key pharmacokinetic parameters (e.g.,
152
plasma benzoate area under the curve [AUC], maximum biotransformation rate [Vmax], or clearance rate)
153
for humans and rats.
SC
RI PT
Based on this review and
154
In addition, we evaluated the human clinical studies of sodium benzoate with a specific focus on the
156
therapeutic doses and any associated observed/reported adverse effects. We determined whether the data
157
were sufficiently robust for identifying a NOAEL or deriving CSAFs for the pharmacodynamic
158
components of the UFs.
M AN U
155
159
3
Interspecies Comparison of Pharmacokinetics of Benzoic Acid and Its Salts
161
In its guidelines for risk assessment, WHO/IPCS specifies that the evaluation of qualitative and
162
quantitative differences in absorption, distribution, metabolism, and excretion (ADME) between
163
experimental animals and humans greatly informs hazard identification and risk characterization (IPCS,
164
2009). The uncertainty of extrapolating animal toxicity data to humans is reduced when the same ADME
165
processes occur in both species. In addition, identifying key ADME determinants and corresponding
166
differences in both species provides a mechanistic basis for estimating a CSAF. Thus, below are our
167
qualitative and quantitative evaluations of the robust pharmacokinetic data available in humans and
168
animals for benzoic acid, its salts, and its precursor compounds.
AC C
EP
TE D
160
169
8
ACCEPTED MANUSCRIPT
170
3.1
Qualitative Interspecies Comparison of Pharmacokinetics
171
Absorption.
172
completely absorbed from the gastrointestinal tract (Tremblay and Qureshi, 1993; US FDA, 1972). In
173
humans, peak plasma benzoic acid levels are reached 1-2 hours after oral ingestion of sodium benzoate
174
(Kubota et al., 1988; Kubota and Ishizaki, 1991) or after a 90-minute "bolus" intravenous (i.v.) infusion
175
of sodium benzoate (administered as Ammonul® (10% sodium benzoate/10% sodium phenylacetate) up to
176
5.5 g/m2 (US FDA, 2004a; MacArthur et al., 2004). In rats, peak benzoic acid plasma levels were
177
reached within 3 hours of administration after oral gavage administration of benzyl acetate (500 mg/kg
178
bw) (Yuan et al., 1995). In both humans and rats, administered doses of benzoic acid are recovered
179
almost entirely in urine as the metabolite hippuric acid, suggesting that essentially none of the
180
administered parent compound passes through the body unabsorbed (Kubota et al., 1988; Kubota and
181
Ishizaki, 1991; Bridges et al., 1970; Thabrew et al., 1980).
M AN U
SC
RI PT
The available information suggests that, after ingestion, benzoic acid is quickly and
TE D
182
Distribution. Following systemic exposure of benzoic acid, its salts, or precursor compounds, there is no
184
evidence of tissue accumulation or sequestration in humans (Bridges et al., 1970; Kubota et al., 1988;
185
Kubota and Ishizaki, 1991) or rats (reviewed by US FDA, 1972; FASEB, 1973; IOMC, 2000; OECD,
186
2001; JECFA, 1996; SCCP, 2005; EFSA, 2016). The observed volumes of distribution for benzoates in
187
human adults were 0.14-0.18 L/kg bw,3 which is consistent with a distribution primarily in the plasma
188
compartment in the absence of any sequestration in tissues (US FDA, 2004a). Similarly, the volume of
189
distribution measurements in rats administered sodium benzoate via i.v. injection (28-288 mg/kg bw)
190
ranged from 0.46 to 0.61 L/kg bw for different doses (with no apparent relationship with increasing
191
doses), which also supports a lack of tissue sequestration at all doses tested (Gregus et al., 1992).
AC C
EP
183
192
3
Converted from units of L/m2 based on a 60 kg adult and a body surface area of 1.62 m2 (Nair and Jacob, 2016).
9
ACCEPTED MANUSCRIPT
Metabolism. In humans and rats, benzoic acid and its salts are primarily metabolized in the liver by
194
conjugation with glycine, resulting in the formation of hippuric acid (Tremblay and Qureshi, 1993; US
195
FDA, 1972). The conversion of benzoate to hippurate also takes place in the kidneys, although to a lesser
196
extent than it does in the liver (Tremblay and Qureshi, 1993). Benzoic acid is rapidly converted to the
197
intermediate, benzoyl-CoA, which is subsequently conjugated with glycine in a reaction catalyzed by the
198
enzyme glycine N-acyltransferase (Tremblay and Qureshi, 1993; US FDA, 1972). Studies have shown
199
that the formation of hippuric acid is a saturable process following Michaelis-Menten kinetics and is
200
limited by glycine availability at high benzoate/benzoic acid doses (Tremblay and Qureshi, 1993; US
201
FDA, 1972; Adams et al., 2005; Kubota et al., 1988; Kubota and Ishizaki, 1991; MacArthur et al., 2004;
202
reviewed in IOMC, 2000 and JECFA, 1996). This is supported by studies showing that percent benzoic
203
acid clearance decreased as the dose increased (ranging from 40 to 160 mg/kg bw), while percent hippuric
204
acid clearance was stable with increasing dose. In these studies, peak plasma benzoic acid concentrations
205
increased with increasing dose while peak hippuric acid concentrations did not change. Multiple kinetic
206
studies conducted in rats demonstrate that there is a similar saturable process for benzoate metabolism at
207
high exposures (i.e., > 120 mg/kg bw) (Schwab et al., 2001; Gregus et al., 1992; Simkin and White, 1957;
208
Bray et al., 1951). Together, these data indicate that the saturable metabolism of benzoic acid to hippuric
209
acid due to saturation of the glycine conjugation pathway at high doses is the key pharmacokinetic driver
210
for potential benzoic acid toxicity.
EP
TE D
M AN U
SC
RI PT
193
AC C
211 212
After saturation of the glycine conjugation pathway and depletion of glycine after high doses of sodium
213
benzoate (e.g., > 500 mg/kg bw), the benzoyl glucuronide metabolite appears; however, this only occurs
214
at high doses that are orders of magnitude greater than typical levels in the diet (US FDA, 1972; IOMC,
215
2000; Kubota and Ishizaki, 1991; as reviewed in Tremblay and Qureshi, 1993 and Adams et al., 2005).
216
Thus, saturation of the glycine pathway does not appear to be significant in the context of typical human
217
exposures.
218
10
ACCEPTED MANUSCRIPT
Excretion. The amount of benzoate excreted as hippuric acid is similar in rats and humans: 75-100% of
220
benzoic acid is excreted as hippuric acid within 6-24 hours in both species (IOMC, 2000; Bridges et al.,
221
1970; Thabrew et al., 1980). Benzoic acid sometimes appears in its free form in urine (Bridges et al.,
222
1970).
RI PT
219
223
Overall, the general pharmacokinetic pathways of benzoic acid and its salts are very similar in humans
225
and rats (summarized in Table 1).
SC
224
226
3.2
228
Evaluation of interspecies pharmacokinetic differences requires careful consideration and selection of key
229
kinetic dose metrics. As recommended by WHO/IPCS guidance (IPCS, 2005, 2010), these considerations
230
include the toxicologically active substance (e.g., parent chemical or metabolite), internal dose levels,
231
exposure route, duration, intensity (e.g., peak or average dose), and biological matrix (e.g., blood, target
232
tissue) that are consistent with the mode of action of the substance of interest. In this case, the active
233
substance is benzoic acid, and the systemic exposures resulting from chronic, oral exposure to benzoic
234
acid and its salts are most relevant for their use as food additives. While the exact mode of action of
235
benzoic acid toxicity is not certain, the suspected mode of action is related to the saturation of the glycine
236
conjugation pathway at very high doses in both animals and humans.
TE D
EP
AC C
237
CSAF Derivation for Interspecies Differences in Pharmacokinetics
M AN U
227
238
It is important to note that while toxicity is not observed at levels below saturation, evidence of saturation,
239
alone, is not evidence of toxicity; studies in rats and humans demonstrate some level of saturation (as
240
indicated by AUC values that increase in a greater than dose-proportional manner) at doses greater or
241
equal to 120 and 160 mg/kg bw, respectively. However, no evidence of toxicity is observed in either
242
species at or significantly above these doses. Our review of the literature indicates lowest observed
243
adverse effect levels (LOAELs) of 800 mg/kg bw in humans and greater than or equal to 750 mg/kg bw in
11
ACCEPTED MANUSCRIPT
rats (discussed further in Section 5) (IOMC, 2000).
245
metabolism to hippuric acid can be viewed as the primary detoxification pathway, the pharmacokinetic
246
determinants that best characterize the potential toxicity of benzoic acid and its salts are the overall
247
systemic exposure to benzoic acid over time and the metabolic capacity for glycine conjugation.
248
3.2.1
249
For these reasons, and because benzoic acid
RI PT
244
CSAF Derivation
To derive a CSAF for interspecies differences in pharmacokinetics (i.e., PKinterCSAF), we considered
251
studies that reported sufficient quantitative data on several dose metrics, such as plasma AUC, clearance,
252
and maximum biotransformation rate (Vmax). We selected AUC as the key dose metric because it
253
provides a specific comparison of chronic and systemic benzoic acid exposure based on in vivo human
254
and animal data, is reflective of metabolic saturation with increasing dose, and is also a conservative
255
metric due to the greater human variability in AUC compared to other metrics (e.g., Cmax4) (IPCS, 2005).
M AN U
SC
250
TE D
256
We primarily relied on pharmacokinetic data obtained from studies of i.v. exposures to sodium benzoate
258
or benzoic acid in humans and rats to derive the PKinterCSAF because extensive pharmacokinetic data for
259
oral exposure are not available in rats. For the human study (US FDA, 2004a), we first converted i.v.
260
doses of sodium benzoate (in units of mg/m2) to oral doses (in units of mg/kg bw) by dividing by a
261
conversion factor of 37, which is based on the default body surface area for a 60 kg adult5 (US FDA,
262
2005). For consistency across studies, we converted the administered sodium benzoate dose in animals
263
and humans to an internal benzoic acid equivalent dose for each dose group in each study by multiplying
264
the administered dose (in units of mg/kg bw) by the ratio of the molecular weights of benzoic acid
265
(122.123 g/mol) to sodium benzoate (144.11 g/mol).
AC C
EP
257
266 4
Cmax is the maximum concentration of a drug achieved in a specified body compartment (usually plasma) after the drug's administration and before the administration of a second dose. 5 The WHO standard body weight of an adult human is 60 kg (IPCS, 2009).
12
ACCEPTED MANUSCRIPT
As discussed above, absorption after oral exposure is rapid and essentially 100% both in humans and in
268
rats. In addition, available data in healthy adult humans indicate that mean AUCs after oral exposure are
269
similar, if not lower, to those obtained after i.v. bolus infusions of comparable doses of benzoic acid (e.g.,
270
385.4 µg × hour/mL at 80 mg/kg bw and 564.6 µg × hour/mL at 86 mg/kg bw,6 respectively) (Kubota and
271
Ishizaki, 1991; US FDA, 2004a). Therefore, we determined that the pharmacokinetics for benzoic acid
272
and its salts are similar after exposure via the i.v. and oral routes and that the use of pharmacokinetic data
273
obtained from i.v. exposures in humans and animals is appropriate for PKinterCSAF derivation.
SC
RI PT
267
274
We identified two independent pharmacokinetic studies in healthy human volunteers conducted by US
276
FDA with the drug Ammonul® (10% sodium benzoate/10% sodium phenylacetate) (US FDA, 2004a;
277
MacArthur et al., 2004), from which an AUC for plasma benzoate could be derived.
278
administration, phenylacetate conjugates with glutamine in the liver and kidneys to form
279
phenylacetylglutamine, via acetylation, and is then excreted by the kidneys. The metabolic process of
280
phenylacetate is distinct from that of benzoate and the pharmacokinetics of benzoate and phenylacetate in
281
combination are similar to those for the chemicals when administered individually (e.g., Kubota and
282
Ishizaki, 1991; reviewed in MacArthur et al., 2004). Therefore, the pharmacokinetic data of these two
283
substances can be evaluated independently.
Upon
EP
TE D
M AN U
275
284
In one study, 20 healthy adult volunteers (9 males, 11 females) were administered Ammonul® via i.v.
286
infusion at doses of 5.5 g/m2 or 3.75 g/m2 (approximately 150 or 100 mg/kg bw sodium benzoate,
287
respectively) 7 and pharmacokinetic measurements were obtained after an initial 90-minute loading
288
infusion (US FDA, 2004a). Sodium benzoate demonstrated saturable elimination kinetics with decreased
289
clearance with increased dose. The authors also noted a greater than dose-proportional increase in both
290
AUC and Cmax for benzoic acid. The AUC for the primary metabolite, hippurate, increased proportionally
AC C
285
6
3.75 g/m2 × 1,000 × (1/37) = 101 mg/kg bw sodium benzoate × (122.123/144.11) = 86 mg/kg bw benzoic acid equivalent dose. A dose of 5.5 g/m2 caused severe emesis in the first three patients; the dose was then reduced to 3.75 g/m2 for the remaining 17 subjects.
7
13
ACCEPTED MANUSCRIPT
with dose, though the increase in Cmax was less than dose-proportional. The authors noted no marked sex
292
differences. In the second study, similar results were obtained for parent compounds and metabolites
293
with escalating dose (1, 2, and 4 g/m2) after a 90-minute administration using similar methods (US FDA,
294
2004a). Administered i.v. doses, benzoic acid equivalent doses, and dose-normalized AUCs for these
295
studies are presented in Table 2.
RI PT
291
296
The dose-normalized AUCs increased with increasing doses, indicating that the clearance of circulating
298
benzoic acid is nonlinear (equal dose-normalized AUCs would be expected for substances with linear
299
elimination kinetics). Standard errors (i.e., standard deviations divided by the square root of the subject
300
numbers) for all except the highest dose group are less than or equal to 10-11% of mean AUC values, and
301
thus, mean AUCs from these studies are considered adequate to provide an accurate measure of the
302
central tendency (IPCS, 2005). The largest inter-individual variability was observed in the highest dose
303
group, as indicated by the standard deviations in the AUC values and standard error of approximately
304
17% of the mean AUC. This is likely a result of the relatively small sample size (n = 3) and general
305
human variability in AUC measurements.
M AN U
TE D
306
SC
297
Gregus et al. (1992) conducted the most relevant animal study in adult male Wistar rats with sufficient
308
quantitative pharmacokinetic data for calculating the AUCs for plasma benzoate. As with the human
309
data, we converted administered i.v. doses from Gregus et al. (1992) from mmol/kg to oral doses (in units
310
of mg/kg bw) based on a molecular weight of sodium benzoate of 144.11 g/mol, and converted sodium
311
benzoate to benzoic acid equivalent doses based on the molecular masses of sodium benzoate and benzoic
312
acid (122.123 g/mol). Administered i.v. doses, benzoic acid equivalent doses, and dose-normalized
313
AUCs for each dose group in this study are presented in Table 3. Similar to humans, the clearance of
314
circulating benzoic acid in rats is nonlinear, as indicated by increasing dose-normalized AUCs with
315
increasing doses, and standard errors of mean AUC values indicate that sample size is acceptable as they
316
are all less than 5% of the mean.
AC C
EP
307
14
ACCEPTED MANUSCRIPT
317 We then compared dose-normalized AUCs at the most comparable doses of benzoic acid from the US
319
FDA human studies (US FDA, 2004a) and the rat study (Gregus et al., 1992) to calculate the
320
PKinterCSAF. Candidate values derived this way range from 2.1 at the lowest dose (i.e., 23 mg/kg bw and
321
24 mg/kg bw, respectively) to 6.7 at the highest dose (i.e., 126 mg/kg bw and 122 mg/kg bw,
322
respectively). Owing to the large inter-individual variability in the AUC results from the highest dose
323
group in humans from US FDA (2004a), we do not consider the high-dose comparison to be reliable for
324
CSAF derivation. Based on a comparison of AUCs at dose levels that are most comparable to the ADI
325
and levels experienced in the general population (i.e., the lowest dose groups, here), we ultimately
326
selected a CSAF of 2 for interspecies differences in pharmacokinetics.
M AN U
SC
RI PT
318
327 328
3.2.2
Alternative Approaches
Additional analyses further support a reduction of the pharmacokinetic component of the interspecies UF
330
and, together with the qualitative assessment presented in Section 3.1, indicate that the CSAF derived in
331
Section 3.2.1 is conservative.
332
TE D
329
First, to further inform potential species-related differences in glycine conjugation—a key step in the
334
metabolic pathway of benzoic acid—we evaluated in vitro metabolic studies of the maximum rate of
335
hippuric acid formation (i.e., Vmax) from benzoic acid in adult human and rat liver tissues. We compared
336
interspecies Vmax values because the saturation of this biotransformation pathway is presumably a vital
337
governing process in the clearance and potential toxicity of benzoic acid. Thus, understanding any
338
species-related differences in this pathway is essential to adequate evaluation of potential interspecies
339
pharmacokinetic differences for benzoic acid and its salts.
AC C
EP
333
340
15
ACCEPTED MANUSCRIPT
In human liver specimens, an average Vmax value of 204 nmol/min/mg liver was estimated in one study
342
(Pacifici et al., 1991), and the rate of hippuric acid formation at a high concentration (200 µM) was
343
reported to be 254 nmol/min/mg liver in another (Temellini et al., 1993). Because Temellini et al. (1993)
344
only used a single dose to determine this value, it is unclear whether the rate of hippuric acid formation
345
they calculated reflects a state where the enzyme is fully saturated (i.e., Vmax); however, plots of hippuric
346
acid versus benzoic acid concentration from Pacifici et al. (1991) suggest that the 200 µM dose by
347
Temellini et al. (1993) was in the saturated range. In rats, two studies examined glycine conjugation in
348
adult rat liver samples using single-pass perfused adult rat liver preparations and reported Vmax values of
349
101 and 78 nmol/min/g liver, respectively (Chiba et al., 1994; Schwab et al., 2001).
M AN U
SC
RI PT
341
350 351
Comparing the Vmax values from rat liver studies to those observed in the human liver studies yields a
352
mean ratio of 0.4 (Table 4). This indicates that humans have an approximately 2.5-fold greater metabolic
353
capacity to conjugate benzoic acid than rats at levels spanning metabolic saturation.
TE D
354 355
We next explored chemical-specific interspecies differences through the use of default allometric scaling
356
methods for a comparison to the CSAF.
357
pharmacokinetic UF default value is meant to address two different concepts: the known and expected
358
differences across species that are due to differences in basal rates of metabolism (i.e., rodents have a
359
faster metabolism and elimination of compounds relative to humans), and unusual and unanticipated
360
differences when humans are more sensitive than rodents even beyond the expected differences in
361
metabolism (e.g., due to differences in metabolic pathways or enzyme efficiencies). We compared AUCs
362
between humans and rats after allometric scaling of the rat-administered doses to explore whether there
363
were differences in benzoic acid metabolism across species beyond what would be expected based on
364
differences in basal metabolism. For this analysis, we converted the benzoic acid equivalent doses to
365
human equivalent doses (HEDs) via body weight scaling to the ¾ power (US EPA, 2011; Rhomberg and
366
Lewandowski, 2006) to place the rodent and human doses on a "basal metabolism equivalent" basis,
AC C
EP
When applied to derive a safety value, the interspecies
16
ACCEPTED MANUSCRIPT
367
which allows for a more direct and accessible evaluation of chemical-specific differences in metabolism
368
across species (Table 5).8
369 If metabolism between species were identical (i.e., same pathways, same saturation potential, same
371
enzyme affinities) and fully explained by the default allometric scaling approach, then the ratios of the
372
benzoic acid equivalent dose and HED presented in Table 6 should be equal or close to 1. Here, the ratios
373
range between 0.3 and 0.4, which indicates that there are metabolic differences between humans and rats
374
beyond the expected difference due to different body size or metabolic rate and that the default scaling is
375
too conservative.
M AN U
SC
RI PT
370
376
From these data, we can derive a DDEF for benzoic acid, consistent with US EPA (2014a) methodology,
378
by combining the allometric scaling value reflective of inherent slower metabolism in humans (i.e., by
379
dividing by 0.269 or multiplying by its inverse, 3.8) with the chemical-specific metabolic factor derived in
380
Table 6 that suggests more efficient benzoic acid metabolism in humans than rats (i.e., 0.3) to yield an
381
overall DDEF for benzoic acid and its salts of 1.1. This essentially means that humans' apparent greater
382
ability to metabolize benzoic acid almost completely offsets the rat's higher basal metabolism, and
383
therefore there is no significant difference between the overall pharmacokinetics of benzoic acid and its
384
salts in animals and humans.
EP
AC C
385
TE D
377
386
Lastly, WHO/IPCS guidance (2005) states that one can compare AUC values in animals at the dose used
387
for risk assessment (e.g., the NOAEL for the point of departure [POD]) to the AUC in humans at doses
388
expected after typical human exposures. While we do not have data in rats at the POD of 500 mg/kg bw,
389
metabolic saturation is observed at the higher dose groups in Gregus et al. (1992) of 122.1 and 244.2 8 As per US EPA (2011) guidance: HED = Dose (mg) × (BWa/BWh)(3/4), where Dose is the administered dose in benzoic acid equivalents (in units of mg); BWh is 60 kg, the standard adult human body weight (IPCS, 2009); and BWa is 0.3 kg, the average rat body weight based on data reported in Gregus et al. (1992). While this scaling approach was developed based on oral routes of exposure, its use is acceptable here because sodium benzoate is still extensively metabolized in both species and AUCs for sodium benzoate administration are comparable after oral and i.v. administration. 9 Scaling factor = (BWa/BWh)¼ = (0.3 kg/70 kg)¼ = 0.26 (US EPA, 2011).
17
ACCEPTED MANUSCRIPT
mg/kg bw benzoic acid, and thus reflects the nonlinear pharmacokinetics expected at higher doses closer
391
to the POD. A comparison of AUC values at these top doses in rats to the AUC in humans at the level
392
closest to the upper end of expected human exposures (i.e., the 22.9 mg/kg bw benzoic acid group) yields
393
human to rat ratios10 of 0.2-0.5. In other words, humans are capable of clearing the upper end of expected
394
human exposures of benzoic acid and its salts approximately 2 to 5 times faster than the rat clears the high
395
levels closer to the NOAEL used for basis of the ADI. This indicates that additional adjustment of the
396
POD for interspecies pharmacokinetic extrapolation is unnecessary for the purposes of the ADI and thus
397
should be reduced to 1.
SC
RI PT
390
3.2.3
399
Discussion
M AN U
398
We derived a PKinterCSAF of 2 in accordance with WHO/IPCS methodology from comparisons of AUC
401
values based on in vivo human data from US FDA (2004a) and rat data from Gregus et al. (1992). The
402
human data were obtained from a high-quality clinical study with a sufficient number of subjects in the
403
relevant dose groups. The key animal study had a sufficient number of rats, used appropriate controls,
404
and provided adequate documentation of the experimental procedures and results. While the dose ranges
405
used in these studies are lower than that of the POD for the ADI, these dose ranges are greater than
406
typical exposures in the general population and the current and proposed ADIs. Further, metabolic
407
saturation was observed in both species at the doses utilized in the available pharmacokinetic studies, and
408
these levels (including that of the POD) are not associated with toxicity in either species.
409 410
Several lines of evidence indicate that a PKinterCSAF of 2 is conservative, and that a value of 1 might be
411
more appropriate:
AC C
EP
TE D
400
412
10
At 122.1 mg/kg bw: Human AUCNorm/Rat AUCNorm = (Human AUC/Dose)/(Rat AUC/Dose) = (20.3/22.9)/(1,100/244.2) = 0.2. At 244.1 mg/kg bw: Human AUCNorm/Rat AUCNorm = (Human AUC/Dose)/(Rat AUC/Dose) = (20.3/22.9)/(239.5/122.1) = 0.5.
18
ACCEPTED MANUSCRIPT
413
414 415
Vmax comparisons from in vitro liver tissues indicate that humans more efficiently metabolize benzoic acid than rats;
AUC comparisons indicate that humans are capable of clearing the upper end of expected human exposures of benzoic acid faster than the rat clears the higher, saturated levels closer to the
417
NOAEL that is the basis of the ADI; and
418
RI PT
416
AUC ratios between humans and rats after allometric scaling of the rat-administered doses indicate that humans' apparent greater ability to metabolize benzoic acid almost completely
420
offsets the rat's higher basal metabolism, and that additional pharmacokinetic adjustment is
421
unnecessary.
M AN U
SC
419
422
We also note that Hoffman and Hanneman (2017) recently reported that physiologically based
424
pharmacokinetic (PBPK) modelling indicated human-to-rat margin-of-exposure (MOE) ratios of 0.3 to
425
0.4 based on predicted plasma concentrations of benzoic acid after simulated repeated daily oral
426
exposures. The authors ultimately concluded that benzoic acid is cleared more quickly in humans than
427
rats at a steady-state or AUC-based exposure level, and that the PKinterCSAF could be reduced to 1 for
428
ADI derivation.
EP
429
TE D
423
In conclusion, a PKinterCSAF of 2 is consistent with WHO/IPCS methodology, and several lines of
431
evidence indicate that it is conservative.
432
PKinterCSAF of 2 should be applied to the default interspecies UF for ADI derivation.
433
AC C
430
Given its scientific support and conservative nature, a
19
ACCEPTED MANUSCRIPT
434 435
4
Intraspecies Comparison of Pharmacokinetics of Benzoic Acid and Its Salts in Humans
Because of its history of therapeutic uses to treat urea cycle disorders (IOMC, 2000), clinical studies of
437
therapeutic drugs containing sodium benzoate involving infants, children, and adults are available and
438
provide robust pharmacokinetic information that can be used to evaluate potential sensitivities across life
439
stages. Comparable pharmacokinetic data for other potentially sensitive populations (e.g., the elderly or
440
those with specific polymorphisms that may affect metabolism) are not available.
SC
RI PT
436
M AN U
441
The available pharmacokinetic studies in humans indicate that ADME patterns of benzoic acid and its
443
salts are generally similar in children and adults. However, quantitative pharmacokinetic data in neonates
444
and infants are limited to those with severe genetic defects in urea synthesis (Green et al., 1983; Brusilow
445
et al., 1984; Simell et al., 1986). They therefore are not appropriate for a robust quantitative evaluation of
446
such differences for children in the general population without these conditions. For these reasons, and
447
because data on other potentially sensitive populations are not available, we did not derive an CSAF for
448
the intraspecies pharmacokinetic component of the UF. The available data support the use of a default
449
intraspecies pharmacokinetic UF value of 3.16 until more robust data are available. Below, we provide a
450
qualitative discussion of potential age-related pharmacokinetic differences for benzoic acid and its salts
451
based on available pharmacokinetic information in humans.
EP
AC C
452
TE D
442
453
Absorption. US FDA (2004b) noted that peak plasma benzoate levels from seven children (aged 3-26
454
months) following administration of 250-500 mg/kg bw of sodium benzoate (equivalent to 212-424
455
mg/kg bw of benzoic acid) for 1-2 hours were qualitatively similar to those in adults. In a clinical study
456
of children with inborn errors of urea synthesis, Brusilow et al. (1984) documented the time course of
457
plasma benzoate concentrations in two infants (aged 5 months and 12 months) after receiving a single
20
ACCEPTED MANUSCRIPT
dose of sodium benzoate (250 mg/kg bw) and sodium phenylacetate (250 mg/kg bw) via i.v.
459
administration. The plasma benzoate concentrations in these two infants rose quickly and peaked at 2.5
460
mmol/L and 3.2 mmol/L at 2.7 and 1 hours, respectively, following i.v. infusion. Similarly, Simell et al.
461
(1986) reported peak plasma benzoic acid concentrations of 5.1-6.2 mmol/L at approximately 2 hours
462
following i.v. administration of 2 mmol/kg bw (approximately 288 mg/kg bw) sodium benzoate plus L-
463
alanine (6.6 mmol/kg) to five Finnish children aged 2.8-12.6 years (mean of 8.8) with a urea cycle
464
disorder (lynsinuric protein intolerance).11
SC
RI PT
458
465
In a study of four neonates with hyperammonemia administered 3.5 mmol/kg bw-day (460 mg/kg bw-
467
day) sodium benzoate via i.v. injection, Green et al. (1983) found large inter-patient variability, observing
468
an eight-fold range in serum benzoate concentrations (mean ± SD: 7.38 ± 6.39 mmol/L; range: 2.14-16.0
469
mmol/L) after sodium benzoate administration for at least 24 hours. These levels are higher than those
470
observed in infants and children and may be a result of a small sample size and/or decreased metabolism
471
or excretion in these health-compromised neonates.
TE D
M AN U
466
472
Overall, the available information in infants and children demonstrates that peak plasma benzoic acid
474
levels are achieved between 1 and 3 hours after i.v. administration, which is similar to that of adults (1-2
475
hours; see Section 3.1). Data on neonates are limited and may not currently support assuming the same
476
degree of absorption as seen in adults.
AC C
EP
473
477 478
Distribution. There are limited data on the distribution of benzoic acid and its salts in children. The
479
mean volume of distribution reported for neonates was 0.14 L/kg bw with an SD of 0.07 L/kg bw (Green
480
et al., 1983), which is almost identical to those reported for adults (US FDA, 2004a; MacArthur, 2004)
481
and is consistent with a lack of tissue sequestration. No data are available for infants and older children.
482 11
Data obtained from figures provided in Brusilow et al. (1984) and Simell et al. (1986) with the use of GetData Software.
21
ACCEPTED MANUSCRIPT
Metabolism. Evidence regarding whether a particular age group may be more sensitive than others with
484
respect to metabolic capability remains equivocal. In neonates, some studies reported reduced glycine
485
conjugation capacity after i.v. exposure to benzoate (reviewed in Dorne et al., 2005), while others
486
reported capacities after oral exposure that are similar to those of adults (Gow et al., 2001). In two infants
487
aged 5 months and 12 months, sodium benzoate appeared to be completely metabolized within 20 and 10
488
hours of i.v. infusion, respectively (Brusilow et al., 1984), potentially supporting an age-dependent effect
489
on benzoate metabolism in infants less than 1 year of age. The wide variability of benzoate metabolism
490
in infants and neonates has been suggested to indicate the immaturity of the acylation system in the liver
491
and kidneys (Batshaw et al., 2001), as studies have shown that mitochondrial glycine N-acyltransferase
492
activity in the liver varied from 5% to 40% of peak activity between birth and 7 months of age, and peak
493
activity was observed by 18 months of age (Mawal et al., 1997, as reviewed in Batshaw et al., 2001).
M AN U
SC
RI PT
483
494
In older children (i.e., 1 year of age or older), some studies reported higher rates of metabolic clearance
496
compared to healthy adults for a number of metabolic pathways, including for glycine conjugation
497
(e.g., Dorne et al., 2005), but other studies reported no significant differences among the mean rate of
498
glycine conjugation in children, healthy adults, or the elderly12 (e.g., Dorne et al., 2004; reviewed in
499
Badenhorst et al., 2013). The latter finding is supported by evidence indicating that adult-level kidney
500
filtration rates (and thus typical clearance rates) are achieved by approximately 7 months of age in
501
humans (Besunder et al., 1988).
AC C
EP
TE D
495
502 503
Further, Dorne et al. (2005) analyzed metabolic pathway-related UFs for a number of compounds.
504
They determined that the UFs for the pharmacokinetic component of intraspecies variability for glycine
505
conjugation pathway were consistently less than the default value of 3.16 for children, younger adults,
12
Dorne et al. (2004) stratified the age groups as follows: neonates (< 1 month old), infants (> 1 month to < 1 year old), children (> 1 to < 16 years old), healthy adults (16 to < 70 years old), and elderly (healthy adults > 70 years old).
22
ACCEPTED MANUSCRIPT
506
and older adults. Specifically, they reported that the 99th percentile UFs for adults and children were 1.6
507
and 1.8, respectively.13
508 Excretion. As discussed in Section 3.1, i.v. and orally administered sodium benzoate or benzoic acid is
510
almost completely excreted as hippuric acid in adult volunteers. The same is observed in studies of
511
children, ranging from neonates (i.e., 3 days old) to 12.6 years of age (Simell et al., 1986; Green et al.,
512
1983).
SC
RI PT
509
513
Taken together, the evidence indicates that the overall ADME for adults and infants, and children are
515
similar. Metabolic capacity for benzoate may be limited in neonates because of their immature acylation
516
systems, but human variability in benzoate pharmacokinetics appears to be negligible after the age of 1
517
year. The clinical studies in infants and neonates are limited in that the study populations were of poor
518
health status and likely had impaired renal function and clearance as well as limited benzoate conjugation
519
capacity (reviewed by Batshaw et al., 2001; Brusilow et al., 1984; Simell et al., 1986). It is worth noting
520
that the doses given in these studies are quite large compared to potential dietary intake. Healthy
521
neonates and infants in the general population have little, if any, potential dietary exposure to benzoic
522
acid or its salts, so their lower metabolizing rates should not substantially contribute to uncertainties in
523
deriving the ADI for benzoic acid and its salts. Also, the JECFA dietary assessment for benzoate did not
524
include children younger than 1 year old (JECFA, 2015).
525
AC C
EP
TE D
M AN U
514
526
5
527
As discussed above, studies in humans and laboratory animals demonstrate that benzoate metabolism to
528
hippurate by glycine conjugation in the liver and excretion in urine diverts nitrogen from urea production,
529
and thus provides an alternative pathway for the elimination of waste nitrogen (Tremblay and Qureshi, 13
Evaluation of Human Clinical Data of Sodium Benzoate
This analysis was based on data from various substrates that undergo glycine conjugation and is not specific to benzoate.
23
ACCEPTED MANUSCRIPT
1993). Because of this, sodium benzoate has been used as a therapeutic agent to treat hyperammonemia
531
(Tremblay and Qureshi, 1993). More recently, sodium benzoate has also been tested as a treatment for
532
certain neuropsychological conditions (Lane et al., 2013) due to its capability in inhibiting D-amino acid
533
oxidase, a flavoenzyme of peroxisomes in the central nervous system, via benzoyl CoA, the common
534
intermediate to hippuric acid for benzoate.
RI PT
530
535
We evaluated available clinical studies of sodium benzoate, including case reports (Trijbels et al., 1974;
537
Brusilow et al., 1979; Kodama et al., 1981; Brubakk et al., 1982; Mizutani et al., 1983; Takeda et al.,
538
1983; Call et al., 1984; Qureshi et al., 1984; Van de Bor et al., 1984; Letarte et al., 1985; McCormick et
539
al., 1985; Sharp and Lang, 1987; Walter et al., 1992), case series (Brusilow et al., 1980, 1984; Batshaw
540
and Brusilow, 1980; Batshaw et al., 1981, 1982; Watson et al., 1985; Sakuma et al., 1992; Tuchman et
541
al., 1992; Feoli-Fonseca et al., 1996; Maestri et al., 1996; Enns et al., 2007), and clinical trials
542
(Mendenhall et al., 1986; Uribe et al., 1990; Sushma et al., 1992; Lane et al., 2013; Lin et al., 2014), to
543
determine treatment-related adverse effects in these studies associated with doses above the current ADI.
544
The clinical studies involved a number of individuals across life stages, including infants, small children,
545
adolescents, adults, and older adults. Therapeutic doses of sodium benzoate varied considerably, ranging
546
from 4.2 to 500 mg/kg bw-day (equivalent to 3.6 to 423.7 mg/kg bw-day benzoic acid), and were well
547
tolerated in the study subjects with no apparent adverse effects. Observed toxicity, such as vomiting and
548
irritation, occurred only after accidental exposure at doses around 800 mg/kg bw-day (equivalent to 677.9
549
mg/kg bw-day benzoic acid) and ceased once the exposure was discontinued. These studies are described
550
in detail in Table S1.
M AN U
TE D
EP
AC C
551
SC
536
552
Most of the clinical studies were observational in nature (vs. experimental), involved only a small number
553
of subjects, and were not designed to evaluate long-term adverse effects in a systematic manner.
554
Therefore, these studies are not informative regarding a POD for deriving an ADI for benzoates and are
555
not suitable for the development of pharmacodynamic CSAFs. Setting this aside, these studies support a
24
ACCEPTED MANUSCRIPT
NOAEL of 500 mg/kg bw-day (423.7 mg/kg bw-day benzoic acid equivalent) and a lowest observed
557
adverse effect level (LOAEL) of 800 mg/kg bw-day (677.9 mg/kg bw-day benzoic acid equivalent) in
558
humans, which are comparable to the rodent NOAEL of 500 mg/kg bw-day of benzoic acid. This
559
provides supporting evidence – at least qualitatively – that the UF for interspecies differences in
560
pharmacodynamics should be reduced or possibly eliminated.
RI PT
556
561
In addition, because a therapeutic use of sodium benzoate is to treat genetic disorders of urea synthesis,
563
the recipients of treatment are generally children – the population of interest in the most recent JECFA
564
assessment on benzoic acid and its salts. The therapeutic doses are generally higher and the treatment
565
durations are often longer in children than in adults. In terms of safety, the clinical reports do not show
566
that children are more sensitive to the potential toxicity of sodium benzoate. In fact, as discussed in
567
Section 4, there is evidence suggesting that there are no significant differences in benzoate metabolism,
568
the key driver of potential benzoate toxicity, between adults and children over 1 year of age.
M AN U
SC
562
TE D
569 570
Similarly, we investigated studies of individuals with impaired liver function as a possible sensitive
571
population to the potential toxic effects of sodium benzoate.
572
(Mendenhall et al., 1986; Uribe et al., 1990; Sushma et al., 1992) conducted in patients with liver
573
dysfunction showed that there were no treatment-related adverse effects from doses of 5,000-10,000
574
mg/day of sodium benzoate (equivalent to approximately 70-140 mg/kg bw-day of benzoic acid),
575
indicating that sodium benzoate at doses 14-28 times the current ADI is well tolerated in this potentially
576
sensitive population.
EP
AC C
577
These clinical intervention studies
578
Observed NOAELs in children with urea cycle defects and in individuals with impaired liver functions
579
are similar and are at least one order of magnitude higher than the current ADI of 5 mg/kg bw-day,
580
indicating that benzoate exposures at levels considerably higher than the current ADI do not pose any
581
health risk to potentially sensitive populations. Collectively, the clinical evidence indicates that the UF
25
ACCEPTED MANUSCRIPT
582
for human variability in pharmacodynamics should be reduced or possibly eliminated, but as discussed
583
above, the data are not robust enough to derive a specific CSAF.
584
6
586
In 2000, JECFA identified a group ADI of 0-5 mg/kg bw-day for benzoic acid, its salts, and precursor
587
compounds based on a four-generation rat study conducted in 1960 in which no adverse effects were
588
observed at up to the highest dose tested (i.e., the NOAEL of 500 mg/kg bw-day), and a total UF of 100:
589
10 for interspecies differences and 10 for database uncertainty (IOMC, 2000). Many studies on benzoic
590
acid and its salts have been conducted since the ADI was first derived, including toxicity,
591
pharmacokinetic, and clinical studies. For example, the animal toxicity database for oral exposures to
592
benzoic acid, its salts, and related precursor compounds is extensive, and includes acute and chronic
593
toxicity, reproductive/developmental toxicity, carcinogenicity, and genotoxicity studies. These studies
594
consistently report a lack of toxicity at doses up to 1,000 mg/kg bw-day (SCCP, 2005; EFSA, 2016).
595
Human clinical data on sodium benzoate as a therapeutic agent provide additional evidence supporting the
596
lack of toxicity at doses up to 500 mg/kg bw-day. This indicates that the UF for database uncertainty is
597
not warranted and should be changed to 1.
SC
M AN U
TE D
EP
598
Proposed ADI for Benzoic Acid and Its Salts
RI PT
585
The pharmacokinetics of sodium benzoate and benzoic acid are nearly identical in humans and rats.
600
Quantitative comparisons of AUCs support a CSAF of 2 for interspecies differences in pharmacokinetics
601
(in place of the default value of 4), and is supported by a comparison of the maximum rate of metabolic
602
transformation in rat and human liver tissues in vitro. In addition, while information from studies of
603
healthy adults and children with urea cycle defects support similar pharmacokinetic profiles across adults
604
and children, additional studies are needed to assess the generalizability of these findings before a CSAF
605
for the pharmacokinetic component of the intraspecies UF can be derived.
AC C
599
606 26
ACCEPTED MANUSCRIPT
Clinical studies of sodium benzoate in children and adults also support reducing the UFs for interspecies
608
and intraspecies variability in pharmacodynamics, but the data are not sufficient to derive specific CSAFs.
609
Thus, the pharmacodynamic components of the UFs for interspecies and human variability should remain
610
at their default values of 2.5 and 3.16, respectively, until more robust data are available.
RI PT
607
611
Taken as a whole, the current safety database of benzoic acid and its salts supports a total UF of 50;
613
individual UF components are illustrated in Figure 3. Applying this revised total UF of 50 to the NOAEL
614
of 500 mg/kg bw-day established by JECFA results in an ADI of 0-10 mg/kg bw-day for benzoic acid and
615
its salts.
M AN U
SC
612
616 617
In this analysis, we demonstrated that the incorporation of chemical-specific information enhances ADI
618
derivation by reducing uncertainty and developing health-protective values based on the best available
619
science.
TE D
620
Acknowledgments
622
We thank Dr. Lorenz R. Rhomberg, Dr. Isaac Mohar, and Dr. Chester Rodriguez for their technical
623
insights, Ms. Jessica Goldstein and Ms. Carla Walker for editorial support, and Mr. Aaron Bazzle for
624
graphics support.
AC C
625
EP
621
626
Funding Source
627
The authors are employed by Gradient, a private environmental consulting firm. The work reported in
628
this paper was conducted during the normal course of employment. This work was funded by the
629
American Beverage Association. The authors have the sole responsibility for the writing, content, and
630
conclusions of this article.
27
ACCEPTED MANUSCRIPT
RI PT
References
Adams, TB; Cohen, SM; Doull, J; Feron, VJ; Goodman, JI; Marnett, LJ; Munro, IC; Portoghese, PS; Smith, RL; Waddell, WJ; Wagner, BM. 2005. "The FEMA GRAS assessment of benzyl derivatives used
SC
as flavor ingredients." Food Chem. Toxicol. 43(8):1207-1240. doi: 10.1016/j.fct.2004.11.014.
Badenhorst, CP; van der Sluis, R; Erasmus, E; van Dijk, AA. 2013. "Glycine conjugation: Importance in
M AN U
metabolism, the role of glycine N-acyltransferase, and factors that influence interindividual variation." Expert Opin. Drug Metab. Toxicol. 9(9):1139-1153. doi: 10.1517/17425255.2013.796929. Batshaw, ML; Brusilow, S; Waber, L; Blom, W; Brubakk, AM; Burton, BK; Cann, HM; Kerr, D; Mamunes, P; Matalon, R; Myerberg, D; Schafer, IA. 1982. "Treatment of inborn errors of urea synthesis:
TE D
Activation of alternative pathways of waste nitrogen synthesis and excretion." N. Engl. J. Med. 306(23):1387-1392. doi: 10.1056/NEJM198206103062303. Batshaw, ML; Brusilow, SW. 1980. "Treatment of hyperammonemic coma caused by inborn errors of
EP
urea synthesis." J. Pediatr. 97(6):893-900. doi: 10.1016/S0022-3476(80)80416-1. Batshaw, ML; MacArthur, RB; Tuchman, M. 2001. "Alternative pathway therapy for urea cycle
AC C
631
disorders: Twenty years later." J. Pediatr. 138(1 (Suppl.)):S46-S55. doi: 10.1067/mpd.2001.111836. Batshaw, ML; Painter, MJ; Sproul, GT; Schafer, IA; Thomas, GH; Brusilow, S. 1981. "Therapy of urea cycle enzymopathies: Three case studies." Johns Hopkins Med. J. 148(1):34-40.
28
ACCEPTED MANUSCRIPT
Besunder, JB; Reed, MD; Blumer, JL. 1988. "Principles of drug biodisposition in the neonate. A critical evaluation of the pharmacokinetic-pharmacodynamic interface (Part I)." Clin. Pharmacokinet. 14(4):189-
RI PT
216. doi: 10.2165/00003088-198814040-00001. Borgstrom, B. 1949. "Detoxification of benzoic acid by glycuronic acid under normal conditions and in liver disease." Acta Med. Scand. 133(1):7-18. doi: 10.1111/j.0954-6820.1949.tb09332.x.
SC
Bray, HG; Thorpe, WV; White, K. 1951. "Kinetic studies of the metabolism of foreign organic compounds. 1. The formation of benzoic acid from benzamide toluene, benzyl alcohol and benzaldehyde
M AN U
and its conjugation with glycine and glucuronic acid in the rabbit." Biochem. J. 48(1):88-96. Bridges, JW; French, MR; Smith, RL; Williams, RT. 1970. "The fate of benzoic acid in various species." Biochem. J. 118(1):47-51. doi: 10.1042/bj1180047.
Brubakk, AM; Teijema, LL; Blom, W; Berger, R. 1982. "Successful treatment of severe OTC
TE D
deficiency." J. Pediatr. 100(6):929-931. doi: 10.1016/S0022-3476(82)80518-0. Brul, S; Coote, P. 1999. "Preservative agents in foods. Mode of action and microbial resistance mechanisms." Int. J. Food Microbiol. 50 (1-2) :1-17.
EP
Brusilow, S; Tinker, J; Batshaw, ML. 1980. "Amino acid acylation: A mechanism of nitrogen excretion
AC C
in inborn errors of urea synthesis." Science 207(4431):659-661. doi: 10.1126/science.6243418. Brusilow, SW; Danney, M; Waber, LJ; Batshaw, M; Burton, B; Levitsky, L; Roth, K; McKeethren, C; Ward, J. 1984. "Treatment of episodic hyperammonemia in children with inborn errors of urea synthesis." N. Engl. J. Med. 310(25):1630-1634. doi: 10.1056/NEJM198406213102503. Brusilow, SW; Valle, DL; Batshaw, M. 1979. "New pathways of nitrogen excretion in inborn errors of urea synthesis." Lancet 314(8140):452-454. doi: 10.1016/S0140-6736(79)91503-4.
29
ACCEPTED MANUSCRIPT
Call, G; Seay, AR; Sherry, R; Qureshi, IA. 1984. "Clinical features of carbamyl phosphate synthetase-I deficiency in an adult." Ann. Neurol. 16(1):90-93. doi: 10.1002/ana.410160118.
RI PT
Chiba, M; Poon, K; Hollands, J; Pang, KS. 1994. "Glycine conjugation activity of benzoic acid and its acinar localization in the perfused rat liver." J. Pharmacol. Exp. Ther. 268(1):409-416.
Dorne, JL; Walton, K; Renwick, AG. 2004. "Human variability for metabolic pathways with limited data
42(3):397-421. doi: 10.1016/j.fct.2003.10.003.
SC
(CYP2A6, CYP2C9, CYP2E1, ADH, esterases, glycine and sulphate conjugation)." Food Chem. Toxicol.
M AN U
Dorne, JL; Walton, K; Renwick, AG. 2005. "Human variability in xenobiotic metabolism and pathwayrelated uncertainty factors for chemical risk assessment: A review." Food Chem. Toxicol. 43(2):203-216. doi: 10.1016/j.fct.2004.05.011.
Enns, GM; Berry, SA; Berry, GT; Rhead, WJ; Brusilow, SW; Hamosh, A. 2007. "Survival after
TE D
treatment with phenylacetate and benzoate for urea-cycle disorders." N. Engl. J. Med. 356(22):22822292. doi: 10.1056/NEJMoa066596.
European Food Safety Authority (EFSA). 2016. "Scientific Opinion on the re-evaluation of benzoic acid
EP
(E 210), sodium benzoate (E 211), potassium benzoate (E 212) and calcium benzoate (E 213) as food additives." Panel on Food Additives and Nutrient Sources (ANS). EFSA J. 14(3):4433. doi: 10.2903/
AC C
j.efsa.2016.4433. 110p.
Federation of American Societies for Experimental Biology (FASEB). 1973. "Evaluation of the Health Aspects of Benzoic Acid and Sodium Benzoate as Food Ingredients." Life Sciences Research Office. Report to US Food and Drug Administration (US FDA). NTIS PB-223837; SCOGS-7. 24p., January.
30
ACCEPTED MANUSCRIPT
Feoli-Fonseca, JC; Lambert, M; Mitchell, G; Melancon, SB; Dallaire, L; Millington, DS; Qureshi, IA. 1996. "Chronic sodium benzoate therapy in children with inborn errors of urea synthesis: Effect on carnitine metabolism and ammonia nitrogen removal." Biochem. Mol. Med. 57(1):31-36. doi: 10.1006/
RI PT
bmme.1996.0006. Food Standards Australia New Zealand (FSANZ). 2016. "Sulphites." June. Accessed at
SC
http://www.foodstandards.gov.au/consumer/additives/sulphite/Pages/default.aspx.
Gow, PJ; Ghabrial, H; Smallwood, RA; Morgan, DJ; Ching, MS. 2001. "Neonatal hepatic drug
M AN U
elimination." Pharmacol. Toxicol. 88(1):3-15. doi: 10.1034/j.1600-0773.2001.088001003.x. Green, TP; Marchessault, RP; Freese, DK. 1983. "Disposition of sodium benzoate in newborn infants with hyperammonemia." J. Pediatr. 102(5):785-790. doi: 10.1016/S0022-3476(83)80259-5. Gregus, Z; Fekete, T; Varga, F; Klaassen, CD. 1992. "Availability of glycine and coenzyme A limits
TE D
glycine conjugation in vivo." Drug Metab. Dispos. 20(2):234-240.
Hoffman, TE; Hanneman, WH. 2017. "Physiologically-based pharmacokinetic analysis of benzoic acid in rats, guinea pigs and humans: Implications for dietary exposures and interspecies uncertainty." Comput.
EP
Toxicol. doi: 10.1016/j.comtox.2017.06.002.
AC C
International Programme on Chemical Safety (IPCS). 1994. "Environmental Health Criteria No. 170: Assessing human health risks of chemicals: Derivation of guidance values for health-based exposure limits." WHO EHC 170. 73p.
31
ACCEPTED MANUSCRIPT
International Programme on Chemical Safety (IPCS). 2005. "Chemical-Specific Adjustment Factors for Interspecies Differences and Human Variability: Guidance Document for Use of Data in Dose/Concentration-Response Assessment." World Health Organization (WHO); International Labour
RI PT
Organisation; United Nations Environmental Programme. IPCS Harmonization Project Document No. 2. 100p. Accessed at http://www.who.int/ipcs/methods/harmonization/areas/uncertainty/en.
International Programme on Chemical Safety (IPCS). 2009. "Environmental Health Criteria 240:
SC
Principles and Methods for the Risk Assessment of Chemicals in Food." United Nations Environmental Programme; International Labour Organisation; World Health Organization (WHO) Environmental
M AN U
Health Criteria 240. 731p. Accessed at http://www.who.int/foodsafety/publications/chemical-food/en. International Programme on Chemical Safety (IPCS). 2010. "Characterization and Application of Physiological Based Pharmacokinetic Models in Risk Assessment." World Health Organization (WHO) IPCS
Harmonization
Project
Document
No.
9.
97p.
Accessed
at
TE D
http://www.inchem.org/documents/harmproj/harmproj/harmproj9.pdf.
Inter-Organization Programme for the Sound Management of Chemicals (IOMC). 2000. "Benzoic Acid and Sodium Benzoate." World Health Organization (WHO) Concise International Chemical Assessment
AC C
pdf?ua=1.
EP
Document (CICAD) 26. 52p. Accessed at http://www.who.int/ipcs/publications/cicad/cicad26_rev_1.
Joint FAO/WHO Expert Committee on Food Additives (JECFA). 1962. "Evaluation of the Toxicity of a Number of Antimicrobials and Antioxidants: Sixth Report of the Joint FAO/WHO Expert Committee on Food Additives." World Health Organization (WHO), Geneva, Switzerland, WHO Technical Report Series No. 228. 104p. Accessed at http://apps.who.int/iris/bitstream/10665/40518/1/WHO_TRS_228.pdf.
32
ACCEPTED MANUSCRIPT
Joint FAO/WHO Expert Committee on Food Additives (JECFA). 1996. "Toxicological Evaluation of Certain Food and Additives: Report of the Forty-Sixth Meeting of the Joint FAO/WHO Expert Committee on Food Additives (JECFA)." World Health Organization (WHO), Geneva, Switzerland,
RI PT
WHO Food Additives Series 37. Accessed at http://www.inchem.org/documents/jecfa/jecmono/ v37je01.htm.
Joint FAO/WHO Expert Committee on Food Additives (JECFA). 1997. "Evaluation of Certain Food
SC
Additives and Contaminants. Forty-Sixth Report of the Joint FAO/WHO Expert Committee on Food Additives." World Health Organization (WHO), Geneva, Switzerland, WHO Technical Report Series
M AN U
868. 80p. Accessed at http://apps.who.int/iris/bitstream/10665/41962/1/WHO_TRS_868.pdf. Joint FAO/WHO Expert Committee on Food Additives (JECFA). 2002. "Polychlorinated dibenzodioxins, polychlorinated dibenzofurans and coplanar polychlorinated biphenyls." WHO Technical Report Series 909. In Evaluation of Certain Food Additives and Contaminants: Fifty-seventh report of the Joint
TE D
FAO/WHO Expert Committee on Food Additives (JECFA). World Health Organization (WHO), Geneva, Switzerland. p121-149. Accessed at http://apps.who.int/iris/bitstream/10665/42578/1/WHO_TRS_ 909.pdf.
EP
Joint FAO/WHO Expert Committee on Food Additives (JECFA). 2004. "Methylmercury." WHO Technical Report Series 922. In Evaluation of Certain Food Additives and Contaminants: Sixty-
AC C
first Report of the Joint FAO/WHO Expert Committee on Food Additives (JECFA). World Health Organization (WHO), Geneva, Switzerland. p132-143. Accessed at http://apps.who.int/iris/bitstream/ 10665/42849/1/WHO_TRS_922.pdf.
33
ACCEPTED MANUSCRIPT
Joint FAO/WHO Expert Committee on Food Additives (JECFA). 2015. "Safety Evaluation of Certain Food Additives and Contaminants: Prepared by the Eightieth Meeting of the Joint FAO/WHO Expert Committee on Food Additives (JECFA)." World Health Organization (WHO), Geneva, Switzerland.
RI PT
WHO Food Additives Series 71. 142p. Joint FAO/WHO Expert Committee on Food Additives (JECFA). 2016. "Evaluation of Certain Food Additives and Contaminants: Eightieth Report of the Joint FAO/WHO Expert Committee on Food
SC
Additives." World Health Organization (WHO), Geneva, Switzerland, WHO Technical Report Series
M AN U
995. 132p.
Joint FAO/WHO Food Standards Programme. 2016. "Report of the Forty-Eighth Session of The Codex Committee on Food Additives, Xi'an, China, 14-18 March, 2016." Codex Alimentarius Commission, REP16/FA,
101p.,
March.
Accessed
at
http://www.fao.org/fao-who-codexalimentarius/sh-
proxy/en/?lnk=1&url=https%253A%252F%252Fworkspace.fao.org%252Fsites%252Fcodex%252FMeet
TE D
ings%252FCX-711-48%252FReport%252FREP16_FAe.pdf.
Kieckebusch, W; Lang, K. 1960. "The tolerability of benzoic acid in chronic feeding experiments."
EP
Arzneimittelforschung 10:1001-1003.
Kodama, H; Nose, O; Tajiri, J; Maki, I; Sano, T; Harada, T; Kai, H; Yabuuchi, H. 1981. "Treatment with
AC C
sodium benzoate in a patient with ornithine transcarbamylase deficiency." Pediatr. Int. 23(3):399. doi: 10.1111/j.1442-200X.1981.tb01658.x. Kubota, K; Horai, Y; Kushida, K; Ishizaki, T. 1988. "Determination of benzoic acid and hippuric acid in human plasma and urine by high-performance liquid chromatography." J. Chromatogr. 425(1):67-75. doi: 10.1016/0378-4347(88)80007-0.
34
ACCEPTED MANUSCRIPT
Kubota, K; Ishizaki, T. 1991. "Dose-dependent pharmacokinetics of benzoic acid following oral administration of sodium benzoate to humans." Eur. J. Clin. Pharmacol. 41(4):363-368. doi: 10.1007/
RI PT
BF00314969. Lane, HY; Lin, CH; Green, MF; Hellemann, G; Huang, CC; Chen, PW; Tun, R; Chang, YC; Tsai, GE. 2013. "Add-on treatment of benzoate for schizophrenia: a randomized, double-blind, placebo-controlled trial of D-amino acid oxidase inhibitor." JAMA Psychiatry 70(12):1267-1275. doi: 10.1001/
SC
jamapsychiatry.2013.2159.
ornithine
transcarbamylase
deficiency."
3476(85)80359-0.
M AN U
Letarte, J; Ouellet, R; Qureshi, IA; Godard, M. 1985. "Chronic benzoate therapy in a boy with partial J.
Pediatr.
106(5):794-797.
doi:
10.1016/S0022-
Lin, CH; Chen, PK; Chang, YC; Chuo, LJ; Chen, YS; Tsai, GE; Lane, HY. 2014. "Benzoate, a D-amino acid oxidase inhibitor, for the treatment of early-phase Alzheimer disease: A randomized, double-blind,
TE D
placebo-controlled trial." Biol. Psychiatry 75(9):678-685. doi: 10.1016/j.biopsych.2013.08.010. MacArthur, RB; Altincatal, A; Tuchman, M. 2004. "Pharmacokinetics of sodium phenylacetate and
EP
sodium benzoate following intravenous administration as both a bolus and continuous infusion to healthy adult volunteers." Mol. Genet. Metab. 81(Suppl. 1):S67-S73. doi: 10.1016/j.ymgme.2003.12.011.
AC C
Maestri, NE; Brusilow, SW; Clissold, DB; Bassett, SS. 1996. "Long-term treatment of girls with ornithine transcarbamylase deficiency." N. Engl. J. Med. 335(12):855-859. McCormick, K; Viscardi, RM; Robinson, B; Heininger, J. 1985. "Partial pyruvate decarboxylase deficiency with profound lactic acidosis and hyperammonemia: Responses to dichloroacetate and benzoate." Am. J. Med. Genet. 22(2):291-299. doi: 10.1002/ajmg.1320220211.
35
ACCEPTED MANUSCRIPT
Mendenhall, CL; Rouster, S; Marshall, L; Weesner, R. 1986. "A new therapy for portal systemic encephalopathy." Am. J. Gastroenterol. 81(7):540-543.
RI PT
Mizutani, N; Maehara, M; Hayakawa, C; Kato, T; Watanabe, K; Suzuki, S. 1983. "Hyperargininemia: Clinical course and treatment with sodium benzoate and phenylacetic acid." Brain Dev. 5(6):555-563. doi: 10.1016/S0387-7604(83)80059-X.
Basic Clin. Pharm. 7(2):27-31. doi: 10.4103/0976-0105.177703.
SC
Nair, AB; Jacob, S. 2016. "A simple practice guide for dose conversion between animals and human." J.
M AN U
Nair, B. 2001. "Final report on the safety assessment of benzyl alcohol, benzoic acid, and sodium benzoate." Int. J. Toxicol. 20(Suppl. 3):23-50. doi: 10.1080/10915810152630729. Organisation for Economic Co-operation and Development (OECD). 2001. "SIDS Initial Assessment Report for SIAM 13 on Benzoates: Benzoic Acid (CAS No. 65-85-0), Sodium Benzoate (CAS No. 532-
UNEP Publications. 320p.
TE D
32-1), Potassium Benzoate (CAS No. 582-25-2), and Benzyl Alcohol (CAS No. 100-51-6)."
Pacifici, GM; Mogavero, S; Giuliani, L; Rane, A. 1991. "Conjugation of benzoic acid with glycine in the
EP
human fetal and adult liver and kidney." Dev. Pharmacol. Ther. 17(1-2):52-62.
AC C
Qureshi, IA; Letarte, J; Ouellet, R; Batshaw, ML; Brusilow, S. 1984. "Treatment of hyperargininemia with sodium benzoate and arginine-restricted diet." J. Pediatr. 104(3):473-476. doi: 10.1016/S00223476(84)81120-8.
Renwick, AG. 1993. "Data-derived safety factors for the evaluation of food additives and environmental contaminants." Food Addit. Contam. 10(3):275-305.
36
ACCEPTED MANUSCRIPT
Renwick, AG; Lazarus, NR. 1998. "Human variability and noncancer risk assessment - An analysis of the default uncertainty factor." Regul. Toxicol. Pharmacol. 27(1):3-20.
RI PT
Rhomberg, LR; Lewandowski, TA. 2006. "Methods for identifying a default cross-species scaling factor." Hum. Ecol. Risk Assess. 12:1094-1127.
Sakuma, T; Sugiyama, N; Wada, Y. 1992. "The urinary acylcarnitine profile in three cases of transient
SC
hyperammonemia of the newborn." Acta Paediatr. 81(5):436-438. doi: 10.1111/j.1651-2227.1992. tb12264.x.
M AN U
Schwab, AJ; Tao, L; Yoshimura, T; Simard, A; Barker, F; Pang, KS. 2001. "Hepatic uptake and metabolism of benzoate: A multiple indicator dilution, perfused rat liver study." Am. J. Physiol. Gastrointest. Liver Physiol. 280(6):G1124-G1136.
Scientific Committee on Consumer Products (SCCP). 2005. "Opinion on Benzoic Acid and Sodium
TE D
Benzoate." SCCP/0891/05. European Commission, Health & Consumer Protection Directorate-General. 30p., June 21. Accessed at http://ec.europa.eu/health/ph_risk/committees/04_sccp/docs/sccp_o_015.pdf. Sharp, RA; Lang, CC. 1987. "Hyperammonaemic encephalopathy in chronic myelomonocytic leukaemia
EP
(Letter)." Lancet 329(8536):805. doi: 10.1016/S0140-6736(87)92827-3.
AC C
Shen, DD. 2013. "Toxicokinetics." In Casarett and Doull's Toxicology: The Basic Science of Poisons (Eighth Edition). (Ed: Klaassen, CD), McGraw-Hill Education, New York, NY. p367-390. Simell, O; Sipila, I; Rajantie, J; Valle, DL; Brusilow, SW. 1986. "Waste nitrogen excretion via amino acid acylation: Benzoate and phenylacetate in lysinuric protein intolerance." Pediatr. Res. 20(11):11171121. doi: 10.1203/00006450-198611000-00011.
37
ACCEPTED MANUSCRIPT
Simkin, JL; White, K. 1957. "The formation of hippuric acid; The influence of benzoate administration on tissue glycine levels." Biochem. J. 65(3):574-582.
RI PT
Sushma, S; Dasarathy, S; Tandon, RK; Jain, S; Gupta, S; Bhist, MS. 1992. "Sodium benzoate in the treatment of acute hepatic encephalopathy: A double-blind randomized trial." Hepatology 16(1):138-144. doi: 10.1002/hep.1840160123.
SC
Takeda, E; Kuroda, Y; Toshima, K; Watanabe, T; Naito, E; Miyao, M. 1983. "Effect of long-term administration of sodium benzoate to a patient with partial ornithine carbamoyl transferase deficiency."
M AN U
Clin. Pediatr. (Phila) 22(3):206-208.
Temellini, A; Mogavero, S; Giulianotti, PC; Pietrabissa, A; Mosca, F; Pacifici, GM. 1993. "Conjugation of benzoic acid with glycine in human liver and kidney: A study on the interindividual variability." Xenobiotica 23(12):1427-1433. doi: 10.3109/00498259309059451.
TE D
Thabrew, MI; Bababunmi, EA; French, MR. 1980. "The metabolic fate of [14C] benzoic acid in proteinenergy deficient rats." Toxicol. Lett. 5(6):363-367. doi: 10.1016/0378-4274(80)90016-8. Tremblay, GC; Qureshi, IA. 1993. "The biochemistry and toxicology of benzoic acid metabolism and its
AC C
7258(93)90022-6.
EP
relationship to the elimination of waste nitrogen." Pharmacol. Ther. 60(1):63-90. doi: 10.1016/0163-
Trijbels, JM; Monnens, LA; van der Zee, SP; Vrenken, JA; Sengers, RC; Schretlen, ED. 1974. "A patient with nonketotic hyperglycinemia: Biochemical findings and therapeutic approaches." Pediatr. Res. 8(5):598-605. doi: 10.1203/00006450-197405000-00009.
38
ACCEPTED MANUSCRIPT
Tuchman, M; Mauer, SM; Holzknecht, RA; Summar, M; Vnencak-Jones, CL. 1992. "Prospective versus clinical diagnosis and therapy of acute neonatal hyperammonaemia in two sisters with carbamyl
RI PT
phosphate synthetase deficiency." J. Inherit. Metab. Dis. 15(2):269-277. doi: 10.1007/BF01799641. Uribe, M; Bosques, F; Marin, E; Cervera, E; Gil S; Luis Poo, J; Garcia Compean, D; Santoyo, R; Huerta, E; Garcia-Ramos, G; Cadena, M; Merikansky, A; Robledo, J; Cortes, M; Guevara, A; Lopez, N; Aguilar, A; Alejandra Ochoa, E. 1990. "Sodium benzoate in portal-systemic-encephalopathy-induced
SC
blood ammonia normalization and clinical improvement. Interim report of a double-blind multicenter
M AN U
trial." Rev. Invest. Clin. 42(Suppl.):149-154.
US EPA. 2011. "Recommended Use of Body Weight(3/4) as the Default Method in Derivation of the Oral Reference Dose (Final)." Risk Assessment Forum. EPA/100/R11/0001. 50p. US EPA. 2013. "Integrated Science Assessment for Lead (Final)." National Center for
TE D
Environmental Assessment (NCEA). EPA/600/R-10/075F. 1885p., June. US EPA. 2014a. "Guidance for Applying Quantitative Data to Develop Data-Derived Extrapolation Factors for Interspecies and Intraspecies Extrapolation." Risk Assessment Forum, Office of the Science
EP
Advisor, EPA/100/R-14/002F, 109p., September.
US EPA. 2014b. "Health Effects Document for Perfluorooctanoic Acid (PFOA) (Draft)." Office of
AC C
Water, Health and Ecological Criteria Division. EPA 822-R-14-001. 268p., February. US EPA. 2014c. "Health Effects Document for Perfluorooctane Sulfonate (PFOS) (Draft)." Office of Water, Health and Ecological Criteria Division. EPA 822-R-14-002. 208p., February.
39
ACCEPTED MANUSCRIPT
US Food and Drug Administration (US FDA). 1972. "GRAS (Generally Recognized as Safe) Food Ingredients - Benzoic Acid and Sodium Benzoate." Prepared by Informatics, Inc. , NTIS PB-221208,
RI PT
Rockville, MD, 135p., September. US Food and Drug Administration (US FDA). 1973a. "Select Committee on GRAS Substances (SCOGS) Opinion: Benzoic Acid (CAS No. 65-85-0)." Accessed at http://www.fda.gov/Food/IngredientsPackaging
SC
Labeling/GRAS/SCOGS/ucm260036.htm.
US Food and Drug Administration (US FDA). 1973b. "Select Committee on GRAS Substances (SCOGS)
M AN U
Opinion: Sodium Benzoate (CAS No. 532-32-1)." Accessed at http://www.fda.gov/Food/Ingredients PackagingLabeling/GRAS/SCOGS/ucm260074.htm.
US Food and Drug Administration (US FDA). 2004a. "Clinical Pharmacology and Biopharmaceutics Review(s) for Ammonul." Center for Drug Evaluation and Research. 16p.
TE D
US Food and Drug Administration (US FDA). 2004b. "Pharmacology/Toxicology Review and Evaluation for Ammonul." Center for Drug Evaluation and Research. 16p.. December 8. Accessed at http://www.accessdata.fda.gov/drugsatfda_docs/nda/2005/020645s000_PharmR.pdf.
EP
US Food and Drug Administration (US FDA). 2005. "Guidance for Industry: Estimating the Maximum Safe Starting Dose in Initial Clinical Trials for Therapeutics in Adult Healthy Volunteers." Center for
AC C
Drug Evaluation and research (CDER). 30p., July. US Food and Drug Administration (US FDA). 2016. "Listing of specific substances affirmed as GRAS: Benzoic acid." 21 CFR 184.1021. Accessed at https://www.gpo.gov/fdsys/pkg/CFR-2016-title21vol3/pdf/CFR-2016-title21-vol3-sec184-1021.pdf.
40
ACCEPTED MANUSCRIPT
US Food and Drug Administration (US FDA). 2017. "Drugs@FDA: FDA Approved Drug Products." Accessed on January 19, 2017 at http://www.accessdata.fda.gov/scripts/cder/daf/
RI PT
Van de Bor, M; Mooy, P; van Zoeren, D; Berger, R; van Gelderen, HH; Teijema, HL. 1984. "Successful treatment of severe carbamyl phosphate synthetase I deficiency." Arch. Dis. Child. 59(12):1183-1185. Walter, JH; Allen, JT; Holton, JB. 1992. "Arginosuccinate synthetase deficiency: Good outcome despite
SC
severe neonatal hyperammonaemia." J. Inherit. Metab. Dis. 15(2):282-283. doi: 10.1007/BF01799645. Watson, AJ; Chambers, T; Karp, JE; Risch, VR; Walker, WG; Brusilow, SW. 1985. "Transient idiopathic
M AN U
hyperammonaemia in adults." Lancet 326(8467):1271-1274. doi: 10.1016/S0140-6736(85)91554-5. Yuan, JH; Goehl, TJ; Abdo, K; Clark, J; Espinosa, O; Bugge, C; Garcia, D. 1995. "Effects of gavage versus dosed feed administration on the toxicokinetics of benzyl acetate in rats and mice." Food Chem.
EP AC C
632
TE D
Toxicol. 33(2):151-158. doi: 10.1016/0278-6915(94)00123-6.
41
ACCEPTED MANUSCRIPT
Figure Titles
RI PT
Figure 1 Chemical Structures of Benzoates Figure 2 Subdivision of Inter and Intraspecies Uncertainty Factors. Adapted from IPCS (2005).
AC C
EP
TE D
M AN U
SC
Figure 3 Proposed Uncertainty Factors for Acceptable Daily Intake for Benzoic Acid and Its Salts
1
ACCEPTED MANUSCRIPT
Table 1 Pharmacokinetics of Benzoic Acid and Its Salts in Humans and Rats •
Distribution
•
Rapid distribution with no evidence of tissue sequestration (JECFA, 1996; US FDA, 2004a)
•
Rapid distribution with no evidence of tissue sequestration (JECFA, 1996; Gregus et al., 1992)
Metabolism
•
Rapidly and completely metabolized to hippuric acid (US FDA, 1972; IOMC, 2000; Tremblay and Qureshi, 1993)
•
Rapidly and completely metabolized to hippuric acid (IOMC, 2000; Bridges et al., 1970; Thabrew et al., 1980)
•
Peak plasma benzoic acid levels at 1-2 hours after oral administration (Kubota et al., 1988; Kubota and Ishizaki, 1991)
•
Peak plasma benzoic acid levels 3 hours after oral gavage administration (Adams et al., 2005; JECFA, 1996)a
•
At high doses (> 500 mg/kg), benzoyl glucuronide is a secondary metabolite (Kubota and Ishizaki, 1991; JECFA, 1996)
•
At high doses (> 500 mg/kg), benzoyl glucuronide is a secondary metabolite (Adams et al., 2005; JECFA, 1996)
•
Metabolism driven by conjugation with glycine; saturable process at high doses (i.e., ≥ 160 mg/kg) (Kubota et al., 1988; Kubota and Ishizaki, 1991; MacArthur et al., 2004)
•
Metabolism driven by conjugation with glycine; saturable process at high doses (i.e., > 120 mg/kg) (Schwab et al., 2001; Gregus et al., 1992; Simkin and White, 1957; JECFA, 1996)
•
Almost entirely excreted as hippuric acid within 6-24 hours (Kubota et al., 1988; Kubota and Ishizaki, 1991)
•
Almost entirely excreted as hippuric acid within 24 hours (Bridges et al., 1970; Thabrew et al., 1980)
Rat Essentially complete absorption after oral ingestion (e.g., US FDA, 1972; IOMC, 2000)
SC
•
b
M AN U
Excretion
Human Essentially complete absorption after oral ingestion (e.g., US FDA, 1972; IOMC, 2000)
RI PT
Endpoint Absorption
AC C
EP
TE D
Notes: (a) Data from a study of F344 rats administered 500 mg/kg benzyl acetate via gavage (Yuan et al., 1995). (b) Data from studies of oral, subcutaneous, or intraperitoneal (i.p.) administration of benzyl acetate to rats (reviewed in Adams et al., 2005 and JECFA, 1996).
1
ACCEPTED MANUSCRIPT
Table 2 Pharmacokinetic Parameters for Humans After Intravenous Administration of Sodium Benzoate Administered i.v. Dose (g/m2) 1 2 4 3.75 5.5
a
BAED (mg/kg bw) 22.9 45.8 91.6 85.9 126.0
AUC (μg × h/mL) Mean (SD) 20.3 (3.6) 114.9 (31.3) 562.8 (142.3) 564.6 (103.9) 1,599.1 (463.1)
AUC/BAED Mean (SD) 0.9 (0.2) 2.5 (0.7) 6.1 (1.6) 6.6 (1.2) 12.6 (3.6)
RI PT
Number of Subjects 6b b 6 b 6 c 17 c 3
AC C
EP
TE D
M AN U
SC
Notes: AUC = Area Under the Curve; BAED = Benzoic Acid Equivalent Dose; i.v. = Intravenous; SD = Standard Deviation; US FDA = United States Food and Drug Administration. (a) Calculated internal sodium benzoate doses from reported i.v. doses with a 2 conversion factor of 37 to convert mg/m to mg/kg bw for adult humans (US FDA, 2005). Converted sodium benzoate to benzoic acid equivalent doses based on the molecular masses of sodium benzoate (144.11 g/mol) and benzoic acid (122.123 g/mol). (b) US FDA Study 973600. (c) US FDA Study 951603. Source: US FDA (2004a).
1
ACCEPTED MANUSCRIPT
Table 3 Pharmacokinetic Parameters for Rats After Intravenous Administration of Sodium Benzoate
5-8 5-8 5-8 5-8
Administered i.v. Dose (mmol/kg) 0.2 0.5 1 2
a
BAED (mg/kg bw) 24.4 61.1 122.1 244.2
b
AUC (μg × h/mL) Mean (SD) 10.4 (0.8) 58.2 (6.3) 239.5 (22.5) 1,100.2 (89.2)
c
AUC/BAED Mean (SD) 0.4 (0.03) 1.0 (0.1) 2.0 (0.2) 4.5 (0.4)
RI PT
Number of Rats
AC C
EP
TE D
M AN U
SC
Notes: AUC = Area Under the Curve; BAED = Benzoic Acid Equivalent Dose; bw/BW = Body Weight; i.v. = Intravenous; SD = Standard Deviation. (a) Converted sodium benzoate to benzoic acid equivalent doses based on the molecular masses of sodium benzoate (144.11 g/mol) and benzoic acid (122.123 g/mol). (b) Calculated based on the formula: AUC = F × Dose/CL, where F = Fraction absorbed (1 for i.v. administration) and CL = Clearance (in units of mL/kg × min) (Shen, 2013). (c) Units are μg × h/mL per mg/kg dose. Source: Gregus et al. (1992).
1
ACCEPTED MANUSCRIPT
RI PT
Table 4 Comparison of the Relative Maximum In Vitro Metabolism in Adult Humans and Rat Livers Vmax Rats (SD) Vmax Humans (SD) Rat Vmax/Human Vmax (nmol/min/g liver) (nmol/min/g liver) 101 (NR)a 204 (47.8)b 0.5 c 78 (6) 204 (47.8)b 0.38 101 (NR)a 254 (90.5)d 0.4 c d 78 (6) 254 (90.5) 0.31
AC C
EP
TE D
M AN U
SC
Notes: NR = Not Reported; SD = Standard Deviation; Vmax = Maximum in vitro biotransformation rate of benzoic acid to hippuric acid. (a) Chiba et al. (1994); Tested concentrations: 0.004 to 600 µM benzoic acid. (b) Pacifici et al. (1991); Tested concentrations: 6 to 200 µM benzoic acid. (c) Schwab et al. (2001); Tested concentrations: 5 to 1,000 µM benzoic acid. (d) Temellini et al. (1993); Tested concentration: 200 µM benzoic acid.
1
ACCEPTED MANUSCRIPT
Table 5 Pharmacokinetic Parameters for Rats After Intravenous Administration of Sodium Benzoate a
b
BAED (mg/kg bw)
HED (mg/kg bw)
24.4 61.1 122.1 244.2
6.5 16.2 32.5 64.9
c
AUC (μg × h/mL) Mean (SD) 10.4 (0.8) 58.2 (6.3) 239.5 (22.5) 1,100.2 (89.2)
d
AUC/HED Mean (SD) 1.6 (0.1) 3.6 (0.4) 7.4 (0.7) 16.9 (1.4)
RI PT
Administered i.v. Dose (mmol/kg) 0.2 0.5 1 2
AC C
EP
TE D
M AN U
SC
Notes: AUC = Area Under the Curve; BAED = Benzoic Acid Equivalent Dose; bw/BW = Body Weight; HED = Human Equivalent Dose; i.v. = Intravenous; SD = Standard Deviation. (a) Converted from mmol/kg bw to mg/kg bw based on the molecular mass of benzoic acid (122.123 g/mol). 3/4 (b) Converted to HEDs based on allometric scaling of BW (US EPA, 2011; Rhomberg and Lewandowski, 2006). (c) Calculated based on the formula: AUC = F × Dose/CL, where F = Fraction absorbed (1 for i.v. administration) and CL = Clearance (in units of mL/kg × min) (Shen, 2013). (d) Units are μg × h/mL per mg/kg dose. Source: Gregus et al. (1992).
1
ACCEPTED MANUSCRIPT
Table 6 Comparison of Dose-normalized AUCs in Humans and Rats After Allometric Scaling a
b
HED in Rats (mg/kg bw) 16.2 32.5 64.9
Mean AUCNorm c in Humans 0.9 2.5 6.6
Mean AUCNorm c in Rats 3.6 7.4 16.9
Human AUCNorm/Rat AUCNorm 0.3 0.3 0.4
RI PT
BAED in Humans (mg/kg bw) 22.9 45.8 85.9
AC C
EP
TE D
M AN U
SC
Notes: AUCNorm = Dose-normalized Area Under the Curve; BAED = Benzoic Acid Equivalent Dose; bw = Body Weight; HED = Human Equivalent Dose. (a) Converted to benzoic acid equivalent doses. 3/4 (b) Converted to HEDs by scaling the rat benzoic acid equivalent doses by BW (US EPA, 2011; Rhomberg and Lewandowski, 2006). (c) Units are μg × h/mL per mg/kg dose. Sources: US FDA (2004a); Gregus et al. (1992).
1
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Benzoic Acid
Sodium Benzoate
Potassium Benzoate
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Benzoate pharmacokinetics are similar between humans and rodents.
A conservative CSAF of 2 is derived for interspecies variability in pharmacokinetics.
Use of chemical-specific data reduces the default UF of 100 for the ADI to 50.
An ADI of 10 mg/kg bw-day is proposed for benzoic acid and its salts.
Clinical data support the safety of benzoates at current dietary levels.
AC C
EP
TE D
M AN U
SC
RI PT
1