First draft prepared by Dr J. Gry¹, Professor A.G. Renwick² and Professor I.G. Sipes^3
1Institute of Food Safety and Nutrition, Danish Veterinary and Food Administration, Ministry of Food, Agriculture and Fisheries, Soborg, Denmark
²Clinical Pharmacology Group, University of Southampton, Southampton, United Kingdom
³ Department of Pharmacology and Toxicology, College of Pharmacy, University of Arizona, Tucson, Arizona, USA

The Committee evaluated a group of 38 flavouring agents that included alpha-methylbenzyl alcohol (No. 799), acetophenone (No. 806), and 36 structurally related aromatic secondary alcohols, ketones, and related esters (see Table 1) by the Procedure for the Safety Evaluation of Flavouring Agents (see Figure 1, p. 132). All members of this group are considered to be aromatic secondary alcohols, ketones, or related esters. The aromatic ring may contain additional alkyl substituents or a methoxy group, and the aliphatic side chain may be unsaturated or contain additional oxygenated functional groups.

The Committee previously evaluated three members of the group. alpha-Methylbenzyl alcohol (No. 799) was evaluated at the forty-first meeting (Annex 1, reference 107), when an ADI of 0–0.1 mg/kg bw was established. At its twenty-fourth meeting, the Committee reviewed data on alpha-isobutylphenethyl alcohol (No. 827) (Annex 1, reference 53), and at its twenty-third and twenty-fifth meetings it reviewed data on methyl beta-naphthyl ketone (No. 811) (Annex 1, references 50 and 56). No ADI was allocated for either of these two flavouring agents.

Of the 38 aromatic substituted secondary alcohols, ketones, and related esters considered, 16 have been reported to occur naturally in foods (Maarse et al., 2000). For instance, alpha-methylbenzyl alcohol (No. 799) has been detected in cheese, fruit, and tea. The corresponding ketone acetophenone (No. 806) is a natural component of berries, seafood, beef, and nuts.

The total annual production of the 38 aromatic secondary alcohols, ketones, and related esters in this group is approximately 4200 kg in Europe (International Organization of the Flavor Industry, 1995) and 17 000 kg in the USA (Lucas et al., 1999). About 58% of the total annual production in Europe is accounted for by two agents in the group (alpha-methylbenzyl acetate, No. 801, and acetanisole, No. 810). The estimated daily intakes per person of alpha-methylbenzyl acetate and acetanisole in Europe are 200 µg and 150 µg, respectively. In the USA, approximately 80% of the total production arises from the use of four agents (alpha-methylbenzyl acetate (No. 801), acetophenone (No. 806), 4-(para-methoxyphenyl)-2-butanone (No. 818), and ethyl benzoylacetate (No. 834)), with estimated daily intakes of 650, 170, 840, and 140 µg, respectively. The estimated daily intake of each flavouring agent in the group is reported in Table 2.

Generally, the flavouring agents in this group are rapidly absorbed from the gut. The aromatic secondary alcohols (and aromatic ketones after reduction to the corresponding secondary alcohols) are then either conjugated with glucuronic acid and excreted primarily in the urine or are further oxidized and excreted mainly as glycine conjugates. As aromatic esters are generally hydrolysed in vivo by the catalytic activity of carboxylesterases, which predominate in hepatocytes, it is anticipated that the 10 esters in the group of 38 flavouring agents will be hydrolysed to their parent aromatic or aliphatic alcohols and carboxylic acids. The eight aromatic secondary alcohols formed are excreted as their glucuronides or are further metabolized and excreted in the urine. The corresponding eight simple aliphatic carboxylic acids are all metabolized completely by well-known pathways. Two esters (Nos 834 and 835) are hydrolysed to ethanol and aromatic keto-carboxylic acids (3-oxo-3-phenyl-propanoic acid and 3-oxo-5-phenyl-pentanoic acid, respectively), which are anticipated to be further metabolized and excreted in the urine, like other aromatic ketones, described above. Simple aromatic ring substitution with methyl, isopropyl, or methoxy groups (Nos 805, 807–810, 813, 817, 818, 826, and 829) is predicted to have little influence on the principal metabolic pathways. It is more difficult to predict the metabolic fate of two of the flavouring agents (Nos 811 and 812) on the basis of the available data, and especially to what extent they are distributed in the tissues and eliminated. One of these substances (No. 812) might accumulate in human adipose tissue (Müller et al., 1996; Rimkus & Wolf, 1996).

Step 1. Of the 38 flavouring agents in this group, 34 are simple saturated or unsaturated methoxy- or alkyl-substituted benzene derivatives containing a secondary alcohol, corresponding ketone, and/or related ester functional group. In the Procedure, 28 of the 38 aromatic flavouring agents were classified in structural class I (Nos 799–810, 813–826, and 834–835) (Cramer et al., 1978). Six were assigned to structural class II, one (No. 833) because it is a vicinal diketone and the other five (Nos 812 and 827–830) because they contain a fused non-aromatic carbocyclic ring (No. 812) or aliphatic substituent chains with more than five carbon atoms (Nos 827–830). Four (Nos 811, 831, 832, and 836) of the 38 flavouring agents were assigned to structural class III because they contain more than one aromatic ring and cannot be hydrolysed to mononuclear residues.

Step 2. At current levels of estimated intake, 36 of the 38 flavouring agents in this group are predicted to be metabolized to innocuous products and would not be expected to saturate the available metabolic pathways. Evaluation of these substances therefore proceeds via the left-hand side of the decision tree in the Procedure. The remaining two flavouring agents in the group (Nos 811 and 812) cannot be predicted to be metabolized to innocuous products, and therefore their evaluation proceeds via the right-hand side of the decision tree.

Step A3. The daily per capita intakes in Europe and the USA of all 36 of the flavouring agents that are metabolized to innocuous products are below the human intake threshold for each class (1800 µg for class I, 540 µg for class II, and 90 µg for class III), indicating that they pose no safety concern when used at current levels as flavouring agents.

Step B3. The daily per capita intakes of the two agents (Nos 811 and 812) in Europe and the USA are both below the human intake threshold for their class (540 µg for class II and 90 µg for class III).

Step B4. The NOEL in a 90-day study in rats given methyl beta-naphthyl ketone (No. 811) orally was greater than 33 mg/kg bw per day. This provides a safety margin of greater than 100 000 and 10 000 in relation to the estimated intakes in Europe and the USA, respectively. Therefore, methyl beta-naphthyl ketone does not pose a safety concern at current levels of intake. No data were available on the toxicity of the remaining agent (No. 812) or for relevant structurally related substances. Accordingly, the evaluation of this substance proceeded to step B5.

Step B5. As the estimated level of intake in Europe of 4-acetyl-6-tert-butyl-1,1-dimethylindan (No. 812) of 6 µg/person per day exceeds the threshold level of 1.5 µg/person per day, further data are required for a safety evaluation. The Committee concluded that this flavouring agent cannot be classified as of ‘no safety concern at current level of intake’.

Table 1 summarizes the evaluations of alpha-methylbenzyl alcohol and acetophenone and 36 structurally related flavouring agents.

In the unlikely event that all foods containing all the flavouring agents in structural classes I and II were to be consumed simultaneously on a daily basis, the estimated combined intake would exceed the human intake threshold for class II (540 µg). However, the agents are expected to be metabolized efficiently and would not saturate the available metabolic pathways. On the basis of an evaluation of all the data, there would be no safety concerns associated with combined intake.

The Committee concluded that 37 of this group of 38 aromatic secondary alcohols, ketones, and related esters would not pose a safety concern when they were used at currently estimated levels of intake as flavouring agents.

The Committee noted that the data on toxicity that were available were consistent with the results of the safety evaluation. Data on toxicity were required for two agents (Nos 811 and 812) in application of the Procedure. Relevant data were available for one of these substances (No. 811), which gave a large safety margin in relation to the estimated intake.

The Committee required additional data to evaluate the safety of one agent, 4-acetyl-6-tert-butyl-1,1-dimethylindan (No. 812), which could not be predicted to be metabolized to innocuous products, for which satisfactory data on toxicity were not available, and of which the estimated daily intake, 6 µg/person in Europe, exceeded the threshold of 1.5 µg/person per day.

This monograph summarizes the key data relevant to the safety evaluation of 38 aromatic substituted secondary alcohols, ketones, and related esters used as flavouring agents (Table 1). The structural feature common to all members of the group is a secondary alcohol, ketone, or ester functional group in the aliphatic side-chain bonded to a benzene ring.

Data on the natural occurrence and total annual production of these agents (see Table 2) are reported in sections 1.1 and 1.2, respectively.

Quantitative data on natural occurrence were available for three substances in the group (VCF, 2000). The consumption of alpha-methylbenzyl alcohol (No. 799) from foods is approximately equivalent to that from its intake as a flavouring agent. Consumption of 4-methylacetophenone (No. 807) and benzophenone (No. 831) as flavouring agents is greater than that from natural foods (Stofberg & Kirschman, 1985; Stofberg & Grundschober, 1987) (Table 2).

2.3.1 Biochemical data

Acetophenone (No. 806), alpha-methylbenzyl alcohol (No. 799), and structurally related aromatic ketones and alcohols have been shown to be absorbed rapidly from the gut, metabolized efficiently by the liver, and excreted primarily in the urine and to a very small extent in the faeces.

Early studies (Thierfelder & Daiber, 1923; Quick, 1928; Smith et al., 1954a,b) showed that acetophenone (No. 806) and alpha-methylbenzyl alcohol (No. 799) are absorbed, metabolized, and excreted as polar metabolites within 24 h. Approximately half of a dose of 480 mg/kg bw of acetophenone or 460 mg/kg bw of alpha-methylbenzyl alcohol administered to rabbits by gavage was present in the urine after 24 h (Smith et al., 1954b). Likewise, approximately half of a dose of 500 mg/kg bw of acetophe-none added to the food of dogs was accounted for within the first 24 h and in subsequent 12-h urine samples (Quick, 1928).

In a recent study (Sauer et al., 1997a), groups of three male Fischer 344/N rats were given a single oral dose of 200 mg/kg bw of [¹⁴C-ring](E)-4-phenyl-3-buten-2-one (trans-methyl styryl ketone) (No. 820) by gavage. Urine, faeces, and blood were collected periodically up to 48 h after dosing, and tissues were collected at 48 h. More than 70% of the radiolabel was excreted in urine within 6 h and > 97% within 48 h. After 48 h, only 4.8% of the radiolabel was found in the faeces, while < 0.2% was retained in the tissues. No parent ketone was detected in blood at any time during the experiment. Intravenous administration of 20 mg/kg bw of the labelled ketone resulted in a strikingly similar pattern of absorption and excretion. The blood concentrations were below the limit of detection after 60 min. Essentially 100% of the radiolabel was accounted for in the urine and faeces 48 h after dosing. The short disposition half-time (18 min), the relatively small volume of distribution (0.89 L/kg bw), and the high systemic clearance (70 ml/kg.min, approximately equivalent to hepatic clearance) suggested that the ketone undergoes essentially complete first-pass hepatic clearance (Sauer et al., 1997a).

In a parallel study, groups of three female B6C3F₁ mice were given a single oral dose of 200 mg/kg bw of [¹⁴C-ring](E)-4-phenyl-3-buten-2-one (No. 820) by gavage by the protocol described above. More than 84% of the radiolabel was excreted in the urine within 6 h and > 94% within 48 h. After 48 h, 1.2% of the radiolabel was found in the faeces and 0.3% in exhaled air. In contrast to rats, the mice had parent ketone in the blood, albeit accounting for only 2.6% of the total dose. After intravenous administration of a dose of 20 mg/kg bw, the concentration of ketone in blood was below the limit of detection by 30 min. The disposition half-time (8.9 min), the volume of distribution (3.3 L/kg bw), and the high rate of systemic clearance (540 ml/kg.min) indicated that the ketone was cleared more rapidly from the blood of mice than of rats. The larger apparent volume of distribution in mice suggests that one or more tissues accumulate the parent ketone to a greater extent in mice than in rats. The appearance of the parent ketone in the blood of mice could be due to the higher rate of intestinal absorption than in rats (Sauer et al., 1997b).

Both studies suggest that orally administered (E)-4-phenyl-3-buten-2-one undergoes essentially complete first-pass hepatic clearance in both mice and rats. The absence of the ketone in blood after oral dosing and the observation that the systemic clearance of ketone is approximately equivalent to hepatic clearance support this conclusion (Sauer et al., 1997a,b).

The pharmacokinetics of benzophenone (No. 831) in plasma was studied in male and female Fischer 344 rats and B6C3F₁ mice given a single intravenous dose of 2.5 mg/kg bw for rats and 15 mg/kg bw for mice, or a single dose by oral gavage of 2.5, 5, or 10 mg/kg bw for rats and 15, 30, or 60 mg/kg bw for mice. Plasma benzophenone concentrations were also followed in mice and rats maintained on diets containing 312 or 1200 mg/kg benzophenone for 7–8 days. The oral bioavailability of benzophenone was found to be high (100%) in rats and low (50%) in mice. After administration by oral gavage, the mice cleared benzophenone more rapidly than rats, suggesting more extensive first-pass metabolism in mice (Dix et al., 1997a). When benzophenone was administered at the same concentration in the feed, mice received significantly more benzophenone than rats, but the plasma concentrations of benzophenone were significantly higher in rats than in mice (Dix et al., 1997b).

Generally, the aromatic secondary alcohols (and aromatic ketones after reduction to the corresponding secondary alcohols) are conjugated with glucuronic acid and excreted in the urine, or the alcohols and ketones are further oxidized and conjugated primarily with glycine to form hippuric and phenaceturic acids and excreted. The aromatic esters in the group are hydrolysed either to aromatic secondary alcohols and simple aliphatic carboxylic acids or to ethanol and keto-carboxylic acids. Several aromatic ketones and corresponding alcohols have been shown to be interconvertible in vivo. The available studies on the metabolism, including hydrolysis, of the flavouring agents in the group are summarized below.

In general, aromatic esters are expected to be hydrolysed in vivo by the catalytic activity of the carboxylesterases found throughout mammalian tissues, but predominantly in hepatocytes (Heymann, 1980).

No data were available on the hydrolysis of the 10 aromatic esters in the group (Nos 800–804, 814, 816, 823, 834, and 835). Hydrolysis of the five alpha-methylbenzyl esters (Nos 800–804) yields the alpha-methylbenzyl alcohol and simple aliphatic carboxylic acids. This conclusion is supported by data on the hydrolysis in vitro of structurally related benzyl esters (benzyl acetate, benzyl 2-methylbutanoate, benzyl cinnamate, and benzyl phenylacetate), which indicate that significant ester hydrolysis is expected before absorption (Leegwater & van Straten, 1974). After absorption, rapid hydrolysis is expected in the blood and liver in vivo. Benzyl acetate was readily hydrolysed in pig liver homogenate (Heymann, 1980). The plasma half-times for the hydrolysis of a series of four alkyl benzoates (including methyl benzoate, ethyl benzoate and propyl benzoate) and two aralkyl benzoates (benzyl benzoate and phenethyl benzoate) in 80% human blood plasma in vitro were 15–210 min (Nielsen & Bundgaard, 1987).

alpha-Methylbenzyl alcohol (No. 799) and acetophenone (No. 806): As alpha-Methylbenzyl alcohol (No. 799) and acetophenone (No. 806) are interconvertible, similar excretion products may be formed. Reduction of acetophenone to alpha-methylbenzyl alcohol and oxidation of alpha-methylbenzyl alcohol to acetophenone has been reported in rat hepatocytes (Hopkins et al., 1972; Maylin et al., 1973). The reduction and oxidation steps have been shown to be stereoselective in vitro and in vivo (Culp & McMahon, 1968; Callaghan et al., 1973; Sullivan et al., 1976). The alcohol is conjugated mainly with glucuronic acid and excreted, while the ketone can also undergo omega-oxidation and subsequent oxidative decarboxylation to yield benzoic acid, which is excreted mainly in the urine as hippuric acid. Little or no oxidation of the aromatic ring has been reported.

The reduction and oxidation pathways have been observed in various animal species. Acetophenone administered to rabbits via a variety of routes or to dogs in the diet was primarily reduced to alpha-methylbenzyl alcohol. In other studies, alpha-methylbenzyl alcohol was oxidized to acetophenone in rats and to a metabolite of acetophenone in rabbits (El Masry et al., 1956; Kiese & Lenk, 1974). Incubation of acetophenone with carbonyl reductase from rabbit kidney resulted predominantly in the formation of S-(–)-alpha-methylbenzyl alcohol (Culp & McMahon, 1968). Conversely, incubation of (+)- or +(–)-alpha-methylbenzyl alcohol in rat liver cytosolic preparations containing NADP⁺ resulted in stereospecific oxidation of the (–)-isomer only, to yield acetophenone (Callaghan et al., 1973).

In Chinchilla rabbits, about 28% of a single dose of 244 mg/kg bw of alpha-methylbenzyl alcohol administered via a stomach tube was excreted in the urine as hippuric acid within 24 h. Rabbits given acetophenone at a single dose of 240 mg/kg bw excreted 19% as hippuric acid (El Masry et al., 1956). Also in rabbits, about 50% of a single oral dose of 450 mg/kg bw alpha-methylbenzyl alcohol was excreted as the glucuronic acid conjugate in urine within 24 h. Other urinary metabolites included hippuric acid (30%) and mandelic acid (1–2%). Under similar conditions, acetophenone underwent essentially the same metabolic fate. An oral dose of 450 mg/kg bw acetophenone was excreted in 24-h urine as the glucuronic acid conjugate of alpha-methylbenzyl alcohol (47%) and as hippuric acid (17%) (Smith et al., 1954a).

The mode of administration and the species have little effect on the metabolic fate of the alcohol or ketone. The major urinary metabolites were still the glucuronic acid conjugate of alpha-methylbenzyl alcohol (35%) and hippuric acid (24%) when rabbits were given a single subcutaneous dose of acetophenone at 500–1400 mg/kg bw. Small amounts were excreted as mandelic acid or unchanged (Thierfelder & Daiber, 1923). When dogs were given a single oral dose of 500 mg/kg bw, 35% was recovered in the urine as the glucuronic acid conjugate of alpha-methylbenzyl alcohol, while 20% was excreted as hippuric acid. Much of the remainder was excreted unchanged (Quick, 1928).

Minor urinary metabolites in rabbits given large doses (total dose, 5.4 g) of acetophenone by intraperitoneal injection included omega-hydroxyacetophenone, meta-hydroxyacetophenone, and para-hydroxyacetophenone. These metabolites accounted for approximately 1% of the dose (Kiese & Lenk, 1974). Studies of ethyl benzene, chiral and achiral forms of alpha-methylbenzyl alcohol (No. 799), acetophenone (No. 806), and omega-hydroxyacetophenone suggested that chiral mandelic acid is formed from alpha-methylbenzyl alcohol via acetophenone, benzoic acid is also formed directly from acetophenone, and omega-hydroxyacetophenone is an intermediary metabolite in the formation of chiral mandelic acid and benzoic acid from acetophenone or alpha-methylbenzyl alcohol (Sullivan et al., 1976).

When rats were given a single intraperitoneal dose of racemic, labelled [³H-C₁]alpha-methylbenzyl alcohol, the urinary mandelic acid was chiral ((–)form) but did not contain the ³H label, suggesting that the alcohol was oxidized (dehydrogenated) to acetophenone before formation of mandelic acid. Acetophenone thus appears to be the precursor of optically active mandelic acid, given that either stereoisomer or the racemic form of alpha-methylbenzyl alcohol forms only the (–) form of mandelic acid. Formation of benzoic acid from acetophenone was confirmed when groups of eight male rats exhaled 30% of a single dose of 100 mg/kg bw of [methyl-¹⁴C]acetophenone as ¹⁴CO₂ within 30 h. The intermediate role of omega-hydroxyaceto-phenone in the formation of benzoic acid and mandelic acid is indicated by the observation that incubation of acetophenone with rat hepatocyte microsomes yields mainly omega-hydroxyacetophenone (Sullivan et al., 1976).

These observations indicate that, in animals, alpha-methylbenzyl alcohol and acetophenone are interconvertible. alpha-Methylbenzyl alcohol may be excreted in the urine predominantly as the glucuronic acid conjugate. Acetophenone undergoes omega-oxidation to yield alpha-hydroxyacetophenone. Subsequent stereoselective reduction of the ketone function and oxidation of the terminal alcohol yields mandelic acid, while simple oxidation of the terminal alcohol yields the corresponding ketoacid, which may undergo oxidative decarboxylation to yield benzoic acid, which is excreted as hippuric acid (see Figure 1).

An increase in chain length does not significantly alter the metabolic fate of these agents. The ketone and alcohol are interconvertible. In major metabolic pathways, the ketone is stereoselectively reduced to the corresponding alcohol, which is subsequently excreted as the glucuronic acid conjugate. If the alkyl chain is even-numbered, the ketone may undergo oxidation and cleavage to yield a phenylacetic acid derivative. If the alkyl chain is odd-numbered, oxidative cleavage yields mainly a benzoic acid derivative. The acids are excreted almost exclusively as glycine conjugates (i.e., phenaceturic acid and hippuric acid).

Metabolic interconversion of 1-phenyl-1-propanol (No. 822) and propiophenone (No. 824) has been observed in vitro. Metabolic reduction of propiophenone produced 19% and 24% 1-phenyl-1-propanol in NADPH- and NADH-fortified male rat liver preparations, respectively. Male rabbit liver homogenate incubated with propiophenone and fortified with NADPH- and NADH-generating systems produced 75% and 61% 1-phenyl-1-propanol, respectively. Other minor metabolic pathways detected included hydroxylation of the alpha-methylene group to produce 2-hydroxy-propiophenone and oxidation of 1-phenyl-1-propanol or propiophenone to yield acetophenone (4–8%). A larger amount of propiophenone (17–18%) was produced in both rat liver preparations (Coutts et al., 1981). In a follow-up experiment, 93–97% of the 1-phenyl-1-propanol produced from propiophenone in rat and rabbit preparations occurred as the S (–)-isomer. The remainder occurred in the R (+) form. NADPH- and NADH-generating systems were equally efficient in the two species (Prelusky et al., 1982).

The presence of alpha,beta-unsaturation in trans-4-phenyl-3-buten-2-one does not significantly alter the metabolic fate of this ketone when compared with that of other aromatic ketones. The glycine conjugate of phenylacetic acid, phenaceturic acid (65%), was the major urinary metabolite collected 48 h after male Fischer 344 rats were given a single dose of 200 mg kg bw trans-4-phenyl-3-buten-2-one (No. 820) by oral gavage. Minor urinary metabolites included hippuric acid (9.9%) and glutathione conjugates of the parent ketone (5.6%) and alcohol (2.2%). Presumably, hippuric acid (benzoylglycine) is formed from hydration of the double bond, subsequent retro-aldol reaction to form benzaldehyde, and then oxidation to benzoic acid. The parent ketone was not detected in blood after dosing. The principal blood metabolite after intravenous administration of the same dose was the corresponding alcohol, 4-phenyl-3-butene-2-ol, which represented 4.4% of the total dose. Almost all the administered dose was recovered within 48 h (Sauer et al., 1997a).

In a similar experiment in female B6C3F₁ mice, the principal urinary metabolites included the glycine conjugates of phenylacetic acid (35%) and benzoic acid (19%), the glutathione conjugate of the ketone (6.7%), and unchanged ketone (8.6%). The principal blood metabolites after intravenous administration of the same dose were the corresponding alcohol and the hydrated ketone 4-hydroxy-4-phenyl-2-butanone, which represented 5.4% and 2.3% of the total dose, respectively. Only about 1.2% of the administered dose was present in the faeces. Approximately 96% was recovered within 48 h (Sauer et al., 1997b).

The majority (46–61%) of a single dose of benzophenone of 364 mg/kg bw administered to rabbits by stomach tube was excreted as the glucuronide conjugate of the corresponding alcohol, benzhydrol within 48 h (Robinson, 1958). Incubation of a solution of 8 mmol/L benzophenone with rabbit liver homogenate and NADPH resulted in the formation of 20% benzhydrol within 1 h (Leibman, 1971).

2.3.2 Toxicological studies

The available LD₅₀ values for mice treated orally ranged from 500 to 3100 mg/kg bw (Rohrbach & Robineau, 1958; Caprino et al., 1976; Moreno, 1982; Schafer & Bowles, 1985; Proctor & Gamble Co., 1992) (Table 3).

LD₅₀ values in rats treated orally have been reported for 17 of the 38 agents in this group. The values ranged from 400 to > 10 000 mg/kg bw, but most values were > 1000 mg/kg bw, indicating that aromatic substituted secondary alcohols, ketones, and related esters have little toxicity (Smyth & Carpenter, 1944, 1948; Brown et al., 1955; Rohrbach & Robineau, 1958; Linet et al., 1962; Jenner et al., 1964; Trubek Labs, 1964; Smyth et al., 1969; Fogleman & Margolin, 1970; Calandra, 1971; Posner, 1971; Levenstein & Wolven, 1972; Denine & Palanker, 1973; Levenstein, 1973; Moreno, 1973; Russell, 1973; Carpenter et al., 1974; Levenstein, 1976; Avon Products, Inc., 1977; Moreno, 1977; Reagan & Becci, 1984; Burdock & Ford, 1990).

The results of short-term studies were available for 12 of the 38 aromatic substituted secondary alcohols, ketones, and related esters in this group (Brown et al., 1955; Trubek Labs, 1956, 1958; Oser et al., 1965; Hagan et al., 1967; Posternak et al., 1969; Gaunt et al., 1974; National Toxicology Program, 1980; Ford et al., 1983; National Toxicology Program, 1990; Burdock et al., 1991; Freeman et al., 1994). These studies cover a range of structures, including the parent alcohol alpha-methylbenzyl alcohol (No. 799) and its corresponding ketone, acetophenone (No. 806) and acetate ester (No. 801), one naphthyl ketone (methyl beta-naphthyl ketone (No. 811)), three aromatic secondary alcohols or related esters (1-phenyl-1-propanol (No. 822), the butyric acid ester of 1-phenyl-2-propanol (No. 814), and alpha-isobutylphenethyl alcohol (No. 827)), two para-methoxyphenyl-substituted ketones (4-(para-methoxyphenyl)-2-butanone (No. 818) and 1-(para-methoxyphenyl)-1-penten-3-one (No. 826)), two ketones containing two aromatic rings (benzophenone (No. 831) and benzoin (No. 836)), and one aromatic diketone (1-phenyl-1,2-propandione (No. 833)). The results of these studies are summarized in Table 4 and described below.

alpha-Methylbenzyl alcohol (No. 799): Groups of four or five male and five female B6C3F₁ mice were given alpha-methylbenzyl alcohol at a dose of 0, 125, 250, 500, 1000, or 2000 mg/kg bw per day by gavage in corn oil on 5 days per week for 16 days. All four males and all five females receiving the highest dose and three of four males and four of five females receiving 1000 mg/kg bw per day died before the end of the study. Histopathological evaluation showed no treatment-related effects (National Toxicology Program, 1990).

In a subsequent study, groups of 10 male and 10 female B6C3F₁ mice were given alpha-methylbenzyl alcohol at a dose of 0, 47, 94, 190, 380, or 750 mg/kg bw per day by gavage in corn oil on 5 days per week for 13 weeks. No compound-related deaths occurred, and the final body weights were comparable to those of controls. Complete histological examinations performed on all male and female mice given the highest dose revealed no treatment-related effects (National Toxicology Program, 1990).

Benzophenone (No. 831): Benzophenone was administered in the diet to groups of male and female B6C3F₁ mice (numbers not specified) at a concentration of 0.12, 0.5, 1, or 2% for 13 weeks, corresponding to average daily intakes of 0, 190, 750, 1500, and 3000 mg/kg bw. Significant decreases in terminal body weight were reported in male mice receiving 1500 mg/kg bw per day and female mice at the three higher doses. Increased serum concentrations of bile salts, increased absolute and relative liver weights, and an increased incidence of centrilobular hepatocellular hypertrophy indicated hepatic toxicity (doses not specified). Thymic necrosis and atrophy and splenic lymphoid depletion were common in treated mice. The authors concluded that these effects were probably the result of stress and inanition rather than the lymphotoxicity of the test agent and suggested that a time-dependent interaction of palatability and cumulative toxic effects could explain the results (Freeman et al., 1994).

Benzoin (No. 836): Groups of five male and five female B6C3F₁ mice were given diets containing benzoin at concentrations providing a dose of 0, 220, 460, 1000, 2200, or 4600 mg/kg bw per day for 14 days. At necropsy, enlarged lymph nodes and spleens were reported in males and enlarged lymph nodes in females at the highest dose (National Toxicology Program, 1980).

In a subsequent study, groups of 10 male and 10 female B6C3F₁ mice were given diets containing benzoin at a concentration of 0, 620, 1200, 2500, 5000, or 10 000 mg/kg for 90 days, corresponding to average daily intakes of 0, 93, 190, 380, 750, and 1500 mg/kg bw, respectively. All surviving animals were killed and necropsied. Microscopic examination of major organs and tissues and gross lesions revealed no evidence of any compound-related effect (National Toxicology Program, 1980).

alpha-Methylbenzyl alcohol (No. 799): Groups of five male and five female Fischer 344 rats were given alpha-methylbenzyl alcohol at a dose of 0, 120, 250, 500, 1000, or 2000 mg/kg bw per day by gavage in corn oil on 5 days per week for 16 days. Two males and four at the highest dose died before the end of the study. Significantly reduced final body weights (males, 21%; females, 15%) and haemorrhagic gastrointestinal tracts (males, 2/5; females, 1/5) were also observed in this group. Histopathological evaluation of two males and two females at 1000 mg/kg bw per day revealed no lesions related to administration of the test agent (National Toxicology Program, 1990).

In a subsequent study, groups of 10 male and 10 female Fischer 344/N rats were given alpha-methylbenzyl alcohol at a dose of 0, 93, 190, 380, 750, or 1500 mg/kg bw per day by gavage in corn oil on 5 days per week for 13 weeks. One male and three females receiving the highest dose died before the end of the study. Weekly measurement of body weights revealed a significant reduction in males (12%) and females (7%) at the highest dose. At necropsy, the relative weights of the livers of males at the three higher doses and of all treated females were greater than those of controls. Histopathological examination of all rats at 750 and 1500 mg/kg bw per day and of males at 375 mg/kg bw per day revealed minimally to mildly elevated concentrations of haemosiderin in spleen macrophages of 9/10 male and 6/10 female rats receiving 1500 mg/kg bw per day and all 10 male rats receiving 750 mg/kg bw per day (National Toxicology Program, 1990).

alpha-Methylbenzyl acetate (No. 801): Four groups of 15 male and 15 female rats were given alpha-methylbenzyl acetate at a dose of 0, 15, 50, or 150 mg/kg bw by gavage daily for 13 weeks. Daily observations showed no changes in the appearance or behaviour of treated or control rats. Weekly measurements of body weight and of food and water consumption showed that males at the highest dose had a statistically significant increase in food and water intake. Analysis of the urine of this group showed an increased number of cells at week 6, but not at week 13. At necropsy, the absolute and relative weights of the kidney and liver of males at the highest dose were higher than those of controls, although no abnormalities were observed histopathologically. The weights of the kidney and liver of animals at 50 mg/kg bw per day were slightly but not statistically significantly increased (Gaunt et al., 1974).

Acetophenone (No. 806): Groups of five male and five female weanling Osborne-Mendel rats received a diet containing acetophenone at a concentration of 1000, 2500, or 10 000 mg/kg for 17 weeks, corresponding to average daily intakes of 0, 100, 250, and 1000 mg/kg bw, respectively. Histopathological examination and measurements of the weights of the liver, kidneys, spleen, heart, and testis showed no dose-related effects (Hagan et al., 1967).

alpha-Methylphenethyl butyrate (No. 814): Groups of 10–16 male and 10–16 female rats were fed diets containing alpha-methylphenethyl butyrate for 90 days at concentrations reported to result in daily intakes of 3.1 and 3.5 mg/kg bw per day, respectively. A significant change in leukocyte count at week 7 was considered by the authors to be of no toxicological significance; it was no longer present at the end of the study. Measurements of growth, clinical chemistry, organ weights, and tissues revealed no adverse effects (Posternak et al., 1969).

4-(para-Methoxyphenyl)-2-butanone (No. 818): Groups of six male albino rats were given diets containing 4-(para-methoxyphenyl)-2-butanone at a concentration of 0, 0.5, 1, or 2% for 2 weeks, corresponding to average daily intakes of 0, 500, 1000, and 2000 mg/kg bw, respectively. The body-weight gains of groups at the two higher doses were significantly lower than those of controls. The authors noted that the test compound had probably reduced the palatability of the diet. Gross examination at necropsy showed no abnormal results (Trubek Laboratories, 1956).

In a follow-up study, groups of 10 male and 10 female rats (strain unspecified) were given diets containing 4-(para-methoxyphenyl)-2-butanone at concentrations resulting in an average daily intake of 56 or 110 mg/kg bw per day for 90 days. No statistically significant differences were found between control and test animals with regard to body weight, general appearance or behaviour, haematological or urinary parameters or histopathological appearance (Trubeck Laboratories, 1958).

1-Phenyl-1-propanol (No. 822): Five male and five female rats were given diets containing 1-phenyl-1-propanol for 4 months at concentrations providing an average daily intake of 480 mg/kg bw for females and 420 mg/kg bw for males. Measurements of organ and body weights and histological studies revealed no adverse effects. No other details were given (Brown et al., 1955).

alpha-Isobutylphenethyl alcohol (No. 827): Groups of 15 male and 15 female rats were given alpha-isobutylphenethyl alcohol in the diet for 90 days at concentrations providing an average daily intake of 0, 10, 40, or 160 mg/kg bw. The food intake of treated groups was generally lower than that of controls. A decrease in serum glucose concentration was observed in males at 40 mg/kg bw per day, but this effect was considered of questionable toxicological significance. The following statistically significant changes were found in animals at the highest dose: a reduction in weight gain (possibly due to reduced palatability) in both sexes, mild proteinuria in females, increased relative liver weight in males, increased relative caecal weights in both sexes, increased relative spleen weight in females, a reduction in serum glucose concentration in both sexes, and a reduced reticulocyte count in both sexes. Significantly increased relative spleen weights were also observed in males and females receiving 40 mg/kg bw per day. No histopathological changes were found in the liver or spleen of any treated animal (Ford et al., 1983).

1-(para-Methoxyphenyl)-1-penten-3-one (No. 826): No treatment-related gross or histological effects were observed after administration of 1-(para-methoxyphenyl)-1-penten-3-one in the diet to groups of 15 male and 15 female FDRL rats at concentrations calculated to provide intakes of 13 and 15 mg/kg bw per day for males and females, respectively, for 90 days (Oser et al., 1965).

1-Phenyl-1,2-propanedione (No. 833): Groups of eight male and eight female rats were fed diets containing 1-phenyl-1, 2-propanedione for 90 days at concentra-tions reported to provide doses of 18 mg/kg bw per day for males and 17 mg/kg bw per day for females. Measurements of growth, haematological and clinical chemical end-points, organ weights, and histological appearance revealed no adverse toxicological effects (Posternak et al., 1969).

Benzophenone (No. 831): Groups of 10–16 male and 10–16 female Sprague-Dawley rats were given diets containing benzophenone at concentrations providing a dose of 0, 20, 100, or 500 mg/kg bw per day for 28 days. Weekly measurements of body weight and food consumption revealed a significant decrease in the final mean body weights of both sexes at the highest dose. Haematological, clinical chemical, and urinary determinations conducted in all test and control rats on day 28 showed significantly decreased erythrocyte count, haemoglobin concentration, and erythrocyte volume fraction in males and females at the two higher doses. At necropsy, significantly increased absolute and relative weights of the liver were found in males and females at the two higher doses and significantly increased relative kidney weights in males at the two higher doses and in all treated females. Females at the lowest dose had significantly increased relative liver weights. Histopathological examinations of the liver and kidney of all test and control animals and 18 additional organs from five males and five females randomly selected from the control and high-dose groups revealed significant increases in the incidence of slight to moderate hepatocyte hypertrophy in males and females at the two higher doses. No other treatment-related abnormalities were observed (Burdock et al., 1991).

In a related study, 16 male and 16 female Sprague-Dawley rats were given diets containing benzophenone at a concentration providing a dose of 20 mg/kg bw per day for 90 days. The actual intake was reported to be 19 mg/kg bw per day for males and 22 mg/kg bw per day for females. Haematological, clinical chemical, and urinary analyses conducted on day 90 revealed no apparent differences between test and control animals. Gross necropsy of all test and control animals and histopathological examination of 20 organs (including liver and kidney) from 12 males and 12 females randomly selected from the treated and control groups revealed no findings attributable to treatment (Burdock et al., 1991).

In a further study, benzophenone was administered in the diet to groups of male and female Fischer 344 rats (numbers of animals not specified) at a concentration of 0.12, 0.5, 1, or 2% for 13 weeks, corresponding to average daily intakes of 120, 500, 1000, and 2000 mg/kg bw. Excessive weight loss led to moribundity of all males at the highest dose after 6 weeks, and only six females at this dose remained alive at day 78. The terminal body weights were significantly decreased in all dose groups, except for males at the lowest dose. An increased concentration of serum bile salts, increased absolute and relative liver weights, and an increased incidence of centrilobular hepatocellular hypertrophy provided evidence of hepatic toxicity (doses not specified). Cellular atrophy and hypercellularity of the bone marrow and papillary necrosis of the kidney were also reported (doses not specified). The authors suggested that the observed effects were the result of stress and inanition rather than the lymphotoxicity of the test agent and that a time-dependent interaction of palatability and cumulative toxic effects could explain the results (Freeman et al., 1994).

Benzoin (No. 836): Groups of five male and five female Fischer 344 rats were given diets containing benzoin at concentrations providing a dose of 0, 100, 320, 1000, 3200, or 10 000 mg/kg bw per day for 14 days. Three females at the highest dose died before the end of the study. A dose-related decrease in weight gain was observed in treated males (statistical significance not specified). None of the treated females gained significant weight, and those at the highest dose lost weight. Necropsy revealed a solid, silvery-white substance in the stomachs of rats at the two higher doses (National Toxicology Program, 1980).

In a subsequent study, groups of 10 male and 10 female Fischer 344 rats were given diets containing benzoin at a concentration of 0, 500, 1500, 5000, 15 000, or 50 000 mg/kg for 90 days, corresponding to average daily intakes of 0, 50, 150, 500, 1500, and 5000 mg/kg bw, respectively. The body-weight gain of males and females at the three higher doses were reduced by more than 10% with respect to that of controls (statistical significance not specified). At necropsy, greenish cortices were observed in the kidneys of four males and two females at 5000 mg/kg bw per day, one male at 500 mg/kg bw per day, and one female at 50 mg/kg bw per day. Discolouration was observed in the livers of one to four females at each dose. Interstitial nephritis was observed in all treated groups, and the severity was dose-related. Scattered vacuolated hepatocytes were observed in the livers of females at the two higher doses (National Toxicology Program, 1980).

In a second 90-day study to determine a dose that would not cause interstitial nephritis, groups of 10 male and 10 female Fischer 344 rats were given diets containing benzoin at a concentration of 0, 30, 60, 120, 250, or 500 mg/kg, corresponding to average daily intakes of 0, 3, 6, 12, 25, and 50 mg/kg bw, respectively. The experimental parameters were evaluated according to the same protocol use in the corresponding study in mice (see above). Interstitial nephritis was observed in the kidneys of males only at the two higher doses (National Toxicology Program, 1980).

Methyl beta-naphthyl ketone (No. 811): Fifteen male and 15 female rats were given diets containing methyl beta-naphthyl ketone for 90 days at concentrations reported to result in daily intakes of 33 mg/kg bw per day for males and 37 mg/kg bw per day for females. Measurements of growth, haematological and clinical chemical end-points, and gross and histological examination at necropsy revealed no significant adverse effects (Oser et al., 1965).

alpha-Methylbenzyl alcohol (No. 799): Groups of 50 male and 50 female B6C3F₁ mice were given alpha-methylbenzyl alcohol at a dose of 0, 380, or 750 mg/kg bw per day by gavage in corn oil on 5 days per week 103 weeks. All organs and tissues were evaluated for the presence of gross lesions, which were examined histopathologically. Tissues from all controls, animals at the high dose, and animals at the low dose that died during the first 21 months of the study were examined histologically. All organs targeted for neoplastic and non-neoplastic effects in treated mice were also examined histologically.

The survival rate of treated mice was comparable to that of controls. The mean body weights of females at the high dose were 8–16% lower than those of controls from week 72 to the end of the study. Significant lung congestion was observed in males (control, 0/50; low dose, 0/50, high dose, 7/50) and females (0/50, 0/50, 7/50) at the highest dose. Pulmonary haemorrhage and foreign matter were observed in 6/50 males at the high dose and 1/49 control males. These effects are commonly associated with the gavage technique. No other negative effects were observed. The authors concluded that there was no evidence of carcinogenic activity (National Toxicology Program, 1990).

Benzoin (No. 836): Groups of 50 male and 50 female B6C3F₁ mice were given diets containing benzoin at a concentration of 0, 2500, or 5000 mg/kg for 104 weeks, corresponding to average daily intakes of 0, 375, and 750 mg/kg bw, respectively. Gross lesions were examined microscopically, but some evaluations were limited by autolysis and cannibalization. The survival of treated mice was comparable to that of controls, and no compound-related clinical signs were observed. After week 44, the body weights of treated females were reduced by about 10% with respect to that of controls. The survival rates of treated groups did not differ significantly from that of controls. Neoplastic and non-neoplastic lesions were observed in treated mice at frequencies comparable to those seen in aged B6C3F₁ mice. The authors concluded that there was no evidence of carcinogenic activity (National Toxicology Program, 1980).

alpha-Methylbenzyl alcohol (No. 799): Groups of 50 male and 50 female Fischer 344/N rats were given alpha-methylbenzyl alcohol at a dose of 0, 380, or 750 mg/kg bw per day by gavage in corn oil on 5 days per week for 103 weeks. Evaluations and necropsies were performed according to the same protocol used in the corresponding 103-week study in mice (see above).

Death occurred before the end of the study in 49/50 males and 39/50 females at the higher dose, 42/50 males and 24/50 females at the lower dose, and 15/50 control males and 16/50 females. Eight males and 14 females at the higher dose, nine males and four females at the lower dose, and one male and one female control were killed accidentally. Most of the deaths occurred during the second year of the study.

The mean body weights at the end of the study were 12–32% lower than that of controls for surviving males at the higher dose, 10–20% lower for surviving males at the lower dose, and 12–19% lower for surviving females at the higher dose; the final body weights of surviving females at the lower dose were comparable to that of controls. Exacerbation of age-related renal nephropathy was reported in more than half of the treated and control females and nearly all treated and control males. However, this age-related renal disease was judged to be more severe in dosed male rats than in the controls. Renal tubule-cell adenomas occurred at significantly greater frequency in males at the higher dose (10%) than in control rats (0%). Significantly increased incidences of parathyroid hyperplasia, heart calcification, glandular stomach calcification, and fibrous osteodystrophy occurred in males at both doses. These effects were probably a secondary response to a mineral imbalance caused by impairment of renal function. Centrilobular necrosis of the liver was observed at significantly increased incidence in males at both doses when compared with controls. Inflammation of the forestomach was also observed at increased frequency in dosed males. Significantly increased incidences of lung congestion were found in females at both doses. Pulmonary haemorrhage and foreign matter were significantly more prevalent in males and females at the higher dose than in control rats. Inflammation of the nasal cavity and salivary gland were observed at significantly increased incidences in dosed males. These symptoms are commonly associated with accidents during gavage.

The authors concluded that, under the conditions of the study, there was some evidence that alpha-methylbenzyl alcohol had carcinogenic activity in male but not female Fischer 344/N rats, as shown by increased incidences of renal tubule-cell adenomas and adenomas or adenocarcinomas (combined). The authors remarked that the renal toxicity was characterized by severe nephropathy and related secondary lesions and that excessive deaths occurred during the last quarter of the study. The poor survival reduced the sensitivity of the study for detecting the presence of a carcinogenic response (National Toxicology Program, 1990).

alpha-Methylbenzyl alcohol was evaluated by the Committee at its forty-first meeting (Annex 1, reference 107), when it reviewed a series of studies, including that of the National Toxicology Program (1990) in mice and rats: "The Committee noted that alpha-methylbenzyl alcohol administered by gavage in corn oil was associated with a higher incidence of renal tubule-cell adenomas in male rats than in untreated controls, but not in female rats or in mice, at dose levels at or exceeding the maximum tolerated dose (MTD) and in the presence of factors that exacerbated a high incidence of age-related chronic progressive nephropathy. The intake of this compound from all sources is extremely low. On the basis of the evidence available, the Committee concluded that the higher incidence of benign neoplasms in the kidney of male rats is not relevant to humans. In view of the limited database, the Committee concluded that the available data could be used to set an ADI by application of a safety factor of 1000 to the minimal-effect level of 93 mg/kg of body weight per day with respect to liver weight increase in the absence of associated pathology in the 13-week study in rats. Accordingly, an ADI of 0–0.1 mg/kg of body weight per day was allocated for alpha-methylbenzyl alcohol."

Benzoin (No. 836): Groups of 50 male and 50 female Fischer 344 rats were given diets containing benzoin at a concentration of 0, 125, or 250 mg/kg for males and 0, 250, or 500 mg/kg for females, for 104 weeks, corresponding to average daily intakes of 0, 12, and 25 mg/kg bw for males and 0, 25, and 50 mg/kg bw for females. The experimental parameters were evaluated according to the same protocol used in the corresponding study in mice (see above).

Survival was significantly reduced in males at the lower dose (50%) with respect to controls (72%), whereas the survival of other treated groups was comparable to that of controls. No treatment-related clinical signs were observed. The mean body weights of treated rats were normal throughout the study. A dose-related increase in the incidence of lymphomas and leukaemias in male rats was not statistically significant. A dose-related increase in the incidence of hyperplasia of the adrenal medulla was observed in male rats (control, 8%; lower dose, 16%; higher dose, 38%). These foci were described by the authors as "very small collections of medullary cells with basophilic cytoplasm and nuclei smaller than those of normal pheochromocytes." A dose-dependent increase in the frequency of chronic nephritis was observed in animals of each sex (males: 33/49, 41/49, 45/50; females: 7/50, 19/49, 29/50). The effect was not significant in male rats, given the high incidence of chronic nephritis in the control group. The chronic inflammation observed in the kidney was qualitatively similar to that commonly observed in ageing rats. As the degenerative, proliferative, and inflammatory lesions occurred with comparable frequency in control and treated rats, the authors concluded that, under the conditions of the study, benzoin was not carcinogenic to male or female Fischer 344 rats (National Toxicology Program, 1980).

Fifteen aromatic substituted secondary alcohols, ketones, and related esters used as flavouring agents have been tested for genotoxicity in vitro (see Table 5). No reverse mutation was reported in the standard Ames assay with various strains of Salmonella typhimurium (TA97, TA98, TA100, TA102, TA1535, TA1537, TA1538, and TA2637) incubated with 33–6700 µg/plate of alpha-methylbenzyl alcohol (No. 799) (Zeiger et al., 1987; National Toxicology Program, 1990), 10–1000 µg/plate of acetophenone (No. 806) (Florin et al., 1980; Nohmi et al., 1985; Fujita & Sasaki, 1987), 3600 µg/plate of methyl beta-naphthyl ketone (No. 811), 3600 µg/plate of 4-(para-methoxyphenyl)-2-butanone (No. 818), 3600 µg/plate of 4-phenyl-3-buten-2-ol (No. 819), 3600 µg of alpha-propyl phenethylalcohol (No. 825), 3600 µg/plate of 1-(para-methoxyphenyl)-1-penten-3-one (No. 826) (Wild et al. 1983), 3–1000 µg/plate of benzophenone (No. 831) (Mortelmans et al., 1986), up to 150 µg/plate of 1-phenyl-1,2-propanedione (No. 833) (Dorado et al., 1992), 3600 µg/plate of ethyl benzoylacetate (No. 834) (Wild et al. 1983), 20–5000 µg/plate of benzoin (No. 836) (Baker & Bonin, 1985; Matsushima et al., 1985; Rexroat & Probst, 1985), and 5–500 µg/plate of 4-acetyl-6-tert-butyl-1,1-dimethylindan (No. 812) (Mersch-Sundermann et al., 1998a), with and without metabolic activation (see Table 5). Although 4-phenyl-3-buten-2-one did not induce reverse mutation in three strains of S. typhimurium (TA98, TA1535, and TA1537) (Prival et al., 1982), concentrations of 10–3000 µg/plate gave positive results in TA100 with metabolic activation but negative results without it.

    See Also:
       Toxicological Abbreviations