Carbadox and Olaquindox as Feed Additives (updated 13 Apr.)

This document pulls together information from several sources and provides a background into the issues revolving around the use of Carbadox and Olaquindox for feed additive use in the light of the recent statment made by the Canadina Health Authority that the use of Carbadox as a growth promoter in swine should be banned. In Canada its use requires a 30 day withdrawal period and in the US 45 days. Both substances are banned in the EU as growth promoters.
calendar icon 13 April 2001
clock icon 34 minute read
This document has been created has as a reference point for information on the subject in the light of the Canadian Health Authorities wish to ban the use of Carbadox as a growth promoting substance in swine. The use of these two substances for growth promotion in the EU is already banned and is a major consideration in a disagreement between the EU, US and Canada over trade restrictions.

In addition to information from the Canadian Pork Council, there are two main articles of EU origin. The first published in 1997 is titled Toxicological and related effects of antibacterial feed additives (AFA's) and provides a good background understanding of the issues. The second published in July 1998 puts the opinion of the European Union Scientific Committee for Animal Nutrition on Possible risks for the consumer, the animal and the users (operators) from the use of Carbadox and Olaquindox as Feed Additives. We also have an extract on the US position over the use of these two substances, presented first.

Please be aware these are historical documents, and may not reflect the current situation. The information is presented to provide a background understanding of the issues only and is not aimed to confirm any particular position or opinion. If you feel there are inaccuracies that need addressing please let us know.

A lot of references are made to the "JECFA Report" (Joint FAO/WHO Expert Committee on Food Additives Report Series 799), unfortunately we have not managed to access an online version of this report for inclusion.

Canadian Pork Council

Pork Council Emphasizes Food Safety of Canadian Pork

Ottawa, April 12, 2001 -- “Canadian pork is a safe, wholesome product, which is sold around the world”, said Edouard Asnong, President of the Canadian Pork Council (CPC), the national association of hog farmers.

Mr. Asnong was responding to some recent media reports concerning a Health Canada review of carbadox, a veterinary medication which is effective at treating intestinal ailments in young pigs.

The federal health department’s concerns with the drug are primarily with any risks that might exist for those handling the product (such as on farms and in feed mills) as well as potential environmental impacts.

On Tuesday, April 10, on learning of Health Canada’s intentions to propose removing approval of the product, the Canadian Pork Council advised hog producers to stop using the drug, at least until official word comes from Health Canada on further action, which it is hoped will be very soon. “The Canadian hog industry’s Canadian Quality Assurance (CQA ® ) on-farm food safety program is designed to assure prudent use in swine production of all veterinary drugs”, added Mr. Asnong.

“We are very confident that our regulatory systems and our industry practices promote safe, wholesome pork.”

CPC Fact sheet on Carbadox (pdf)

US Position

Below is an extract from the WHO Pharmaceuticals Newsletter Nos. 5 & 6, May & June 1998 giving the US position on the use of Carbadox.

Carbadox - tolerance established

United States of America. The Food and Drug Administration has approved the use of carbadox (MecadoxR: Pfizer) for the control of swine dysentery, bacterial swine enteritis, increased rate of weight gain and improved feed efficiency. Additionally, a revised finite tolerance for residues of carbadox and its metabolites in edible swine tissues has been established.

A tolerance for no residues of carbadox and its metabolites and the method to determine these residues had been previously established. Because better and more accurate regulatory procedures are found in general use, the analytical procedure is no longer codified. The method of analysis has now been removed and a tolerance of 30 parts per billion is established for residues of quinoxaline-2-carboxylic acid (marker residue) in liver (target tissue) of swine.
Reference: Federal Register 63(53): 13337 (1998).

Toxicological and related effects of antibacterial feed additives (AFA's)

Below is an extract from European Union Document: SOU 1997:132 report: Toxicological and related effects of antibacterial feed additives (AFA's) which provides a basic understanding of the issues relating to AFA's. The extract below relates specifically to carbadox and Olaquindox. To view the whole document (pdf) which refers to many other products also, click the link above.


As animals are fed AFA for long periods of time, knowledge on bioaccumulation, chronic toxicity and problems associated with residues is imperative. AFA are defined chemical substances and might as such have toxic or allergenic properties.

Toxic effects, if any, could be seen either in target animal species (or non- target animals as a result of accidental intake) or in humans.

Animals and humans may be exposed via residues in animal products and humans also when handling products containing the substances. If the product has unwanted properties, such as organ toxicity, mutagenicity or allergenicity, both ways of exposure could be harmful.

Some AFA are poorly absorbed from the gut. Residues are therefore normally not an issue for these substances. Other AFA such as ardacin and avilamycin are absorbed to some extent, and tylosin, spiramycin, olaquindox and carbadox are well absorbed after oral administration.

Toxicological aspects

Carbadox and olaquindox are known to cause adrenal damage at growth promoting levels. Lowered aldosterone production is found in porcine adrenal glands exposed to carbadox IN VITRO. Van Der Molen (1988) demonstrated a dose-response as well as a time-response relationship between in-feed carbadox and adrenal damage in pigs. After 10 weeks of carbadox at 25 ppm or more, damage to adrenal glomerular cells could be observed histologically at post-mortem examination. Adrenal damage leads to profound hormonal disturbances. This means that the use of carbadox as a feed additive probably results in, at least to some degree, Addisons disease (a syndrome caused by impaired function of the adrenal glands) in pigs.

The clinical signs observed are dose-related, ranging from mild to severe. At 50 ppm, mild effects of increased faecal dryness may be observed. Other signs include symptoms indicating salt imbalance (urine drinking), decreased abdominal volume and lowered haematocrit values. Further, changes in hair quality, with hair becoming longer and withered, irritable behaviour and a decrease in feed intake and weight gain may be observed.

Olaquindox has the same toxic effects, although less pronounced, at dosages of 100 ppm or more. Also in other respects, there is a difference in activity between the substances. Carbadox is permitted in feed at a concentration of 50 ppm, whereas olaquindox is used at 100 ppm. Accidental overdosing of olaquindox and carbadox has been reported to cause death and severe adrenal damage in piglets.

As the main early sign of intoxication with quinoxalines, i.e. dry faeces, may be mistaken for recovery from enteric disease, mild intoxications are expected to be overlooked by farmers and farm workers.

Most AFA are not expected to cause toxic reactions in target species at the levels used.

Quinoxalines and carbadox in particular, cause adrenal damage in pigs at doses used for growth promotion. This poses a serious risk for the well-being of the animals.

Toxicological aspects: Summary comments

For humans, the use of AFA is an occupational hazard. The risks involved are allergy and, for quinoxalines, genotoxicity. The population at risk includes farmers, farm workers, feed mill workers and other persons handling products containing the substances.

In target animals, the use of quinoxalines results in adrenal damage (Addisons disease). This is deleterious for animal well being.

MRLs and withdrawal times for Carbadox withinthe EU.
Substance Animal Maximum Withdrawal MRL Tissue Comments
Carbadox piglets 4 months, 28 days 52
MRL based on
sensitivity of
analytical method


During storage, carbadox is rapidly decomposed to desoxycarbadox in kidney and liver samples, but is stable in eggs and muscle. Carbadox has shown dose-related increases of benign and malignant liver tumours in long-term feeding studies in rats, and gave positive results in 14 out of 15 mammalian and non-mammalian genotoxicity studies.

Quinoxaline-2-carboxylic acid (QCA) is another major residual metabolite of carbadox. Studies on QCA elimination from liver after feeding of 50 ppm carbadox yielded residues above 30 µg/kg in liver and kidney for 4 to 5 weeks. The authors' opinion was that a withdrawal time of 8 weeks should be recommended. In an experiment by Rutalj (1996), QCA was still present at around 10 µg/kg at 62 days after cessation of feeding of carbadox at 50 ppm. Detection of QCA is dependent on the extraction method used, and it has been suggested that this is due to other intermediate metabolites in the pathway of carbadox to QCA (Baars ET AL , 1991).

In the 36th report of JECFA (FAO/WHO, 1990) it was concluded that an ADI could not be established, due to the carcinogenic and genotoxic nature of carbadox and some of its metabolites. MRLs set in 1990 (FAO/WHO, 1990) for QCA as marker substance were apparently based on the detection limit of the analytical method.


Olaquindox is extensively metabolised in the animal. The metabolites found vary between tissues and between animal species. One of the metabolites, 3-methylquinoxaline-2-carboxylic acid (MQCA), is known to be responsible for the IN VITRO mutagenicity of other quinoxaline derivatives and has therefore been chosen as a marker compound (FAO/WHO, 1995). The drug is rapidly absorbed from the gut and mainly excreted via urine. No bound residues appear to be present in tissue (FAO/WHO, 1995). In pigs given 60 ppm in-feed olaquindox up to 16 weeks of age, and with a withdrawal period of 28 days, olaquindox residues were below 0.005 ppm in muscle and below 0.01 ppm in kidney (FAO/WHO, 1995). An ADI could not be allocated by JECFA 1995, because of the genotoxic potential of the parent compound and the absence of specific toxicity studies on the metabolites. No MRL has been set.

Some comments on the quinoxalines

Carbadox has, and olaquindox is suspected of having carcinogenic and genotoxic properties. Carcinogenic and genotoxic effects in consumers could be possible at very low intake levels, especially if the substance in question is ingested regularly over a number of years. Farm and feedmill workers are a special risk group, frequently exposed to AFA when handling animal feed. If appropriate protection cannot be ensured, the handling of animal feed containing quinoxalines and other AFA with potentially toxic effects must be regarded as an occupational hazard. A conservative approach is often recommended for genotoxic substances in order to prevent underestimation of the risks.

Possible risks for the consumer, the animal and the users (operators) from the use of Carbadox and Olaquindox as Feed Additives

This article has been heavily edited for conciseness and may not reflect the full picture. To get the complete perspective and view the whole 25 page PDF document (including all minority opinions and the references) CLICK HERE (PDF Format).

1. Introduction

Carbadox and olaquindox have been approved feed additives in the EU since 1974 and 1976 respectively.

Since then SCAN’s evaluations of carbadox have been published as Opinions in 1978 and 1982. Its toxicologists further considered the safety of carbadox in 1988 in preparation for receiving a United Kingdom delegation on the matter of worker safety.

For carbadox and olaquindox the pig is the only approved target species. For both, use is permitted up to 50 mg/kg of complete feedingstuff for pigs up to 4 months of age and a 4 week minimum withdrawal period is required. (NOTE: Both products are now banned for use as growth promoters)

2. New elements (since previous assessments)

The “new elements” comprise summaries of the earlier evaluations and highlight the adverse properties of the drugs. In addition, there is new open literature information on the metabolism and kinetics of carbadox and olaquindox, on their phototoxicity, on their target animal safety and on microbial resistance. Finally, companies, sometimes not the original sponsors, have performed new studies on worker exposure to both additives.

3. Assessment of the new Elements


3.1.1 Carbadox

While no tumours were found in a 2-year monkey study at doses up to 25 mg/kg b.wt./day, the hepatotoxicity and tumorigenicity and carcinogenicity of carbadox in 4 long term studies in rats established the liver as the primary target at 25 mg/kg b.wt./day (malignant transformation in 3 of 13 rats). Both a 2 year rat and a 7 year dog relay toxicity studies, in which liver (rats, 10% of diet) and pork (dogs, 200g/day) from pigs fed at 4x the user rate of carbadox (without withdrawal period) was fed at 10x the expected human consumption rate, were free of histological abnormality.

Comment: Carbadox, probably via its metabolite desoxycarbadox, is a rodent carcinogen. The negative methyl carbazate studies support the view that any hydrazine formed from it is free of carcinogenicity at intakes which greatly exceed those of consumers exposed to carbadox residues.

The predictive utility of rodent liver tumours for man is often questioned by toxicologists in particular as for Carbadox when associated with hepatotoxicity and regenerative hyperplasia.

3.1.2 Olaquindox

In 4 studies in rats and 3 in mice no evidence of carcinogenicity was found. A small increase in adenomas of the adrenal gland was reported in one rat study, but this was not sufficient to increase the total of animals with tumours. One mouse study showed small increases in adenomas (adrenal, lung and ovary) in top dose males.

Comment: Olaquindox is not a carcinogen in rats or mice, but has increased benign tumours in one species, the mouse, in only one sex and only at the highest dose tested 54 mg/kg b.wt./day.


3.2.1 Carbadox

Positive findings occurred in 10 microbial and 3 of 4 mammalian tests, both in vitro and in vivo. By contrast, desoxycarbadox was negative in 14/18 microbial assays (including 6 host-mediated) and 2 of 4 mammalian tests in vitro. QCA was negative in an Ames test and in human lymphocytes in vitro. Methyl carbazate was negative in microbial systems and in human lymphocytes in vitro. Hydrazine was positive in 2 of 2 bacterial tests and in mouse lymphoma cells.

Comment: Carbadox is genotoxic.

3.2.2 Olaquindox

In 14 microbial tests, olaquindox gave positive findings. In mammalian tests it was positive in 1 of 1 in vitro and 12 of 18 in vivo tests addressing various endpoints, including one weak positive in Chinese hamster spermatogonia. It does not, however, bind to DNA and several mammalian positive results were obtained at near toxic dose levels or with olaquindox whose purity differed from that of the original sponsor’s product.

Comment: Olaquindox is genotoxic and possibly a germ cell mutagen. It is noted that no evidence to support the latter possibility was found in the mammalian, multi-generation reproductive toxicity test.

The predictive utility of genotoxicity tests for rodent carcinogenicity is far from absolute and little is known of their predictive value even for genotoxicity in man. While it is unlikely that further genotoxic substances would be developed for any class of product, that property, while having impact on the labelling, is not of itself regarded as sufficient to require the removal of existing chemical products from their markets.


This potentially worrying but unspecified statement arose at the Scientific Conference on the Use of Growth Promoters in Meat production. Responding to it allows a comment on the metabolism and fate of the drugs.

3.3.1 Carbadox

The compound is extensively metabolised to yield essentially the same urinary metabolite profile in rats and monkeys as in pigs (13 of 15 metabolites), validating the predictive value of the rat long term toxicity studies for other species. Urine plus faecal label accounts for 88.2% of an oral dose of ring-labelled carbadox at 50mg/kg feed within 72 hours. No N- oxides were present in urine and no carbadox was present in faeces. The rapid metabolism of carbadox is reflected in the presence only of QCA in pig liver 24 hours after a single dose of labelled carbadox. Free methyl carbazate-derived hydrazine, was detected in pig urine only after dosing at 7 mg/kg b.wt. (i.e. about 3 x user level).

After 14 days, 0.1mg/kg carbadox equivalent radioactivity remained in pig liver. With side-chain labelling, 0.1-0.34 mg/kg carbadox equivalent radioactivity remained in liver at 5 days. In neither case was this very low-level residue fully characterised but some labelled carbon was shown to have been incorporated into physiological substances.

Neither carbadox nor desoxycarbadox formed part of the residue, however.

Comment: Because residues were detectable at 14 days, they were in that sense “appreciable”. However, the concentration present (even had it been confirmed by the subsequent study) was very low and in that sense could not be considered “appreciable”.

Neither feed additives (Council Directive 70/524/EEC) nor animal medicines have to achieve zero residues in order to be approved. Of importance here is that the kinetics and the metabolism of carbadox, as well as the toxicity of its individual metabolites have been well and extensively studied. The early studies established that beyond 24 hours after dosing, none of the carcinogenic entities remain identifiable in pig tissues.

Subsequent open literature studies cited above have not been quantitatively consistent but none have yielded results which suggest that the 28 day withdrawal period was unsafe for the consumer. Unusually, several metabolites have been subject to separate safety studies. From these, the safety of the major, long-lived metabolite QCA has been established. This information, plus the unusually long mandatory withdrawal period should provide for the consumer freedom from concern over carcinogenic residues.

3.3.2 Olaquindox

SCAN previously reported that rapid and extensive absorption followed by excretion into the urine characterises the fate of oral olaquindox in rat and pig. At user levels, muscle, liver and kidney residues were below the limit of detection of olaquindox (0.05mg/kg) by day 2. Around 70% of an oral dose in the pig is excreted unchanged in the urine accompanied by 16% as mono-N-oxides with 3 probable carboxylic acid derivatives accounting for the balance. Later studies confirmed these findings and completed the characterisation of the metabolites.


Both additives have caused field episodes of severe toxicity including deaths when included in pig diets at accidentally high (i.e. unauthorised) concentration. The narrow safety margin of both additives emphasises the need strictly to observe the approved conditions of use.

Effects mediated via adrenal cortex depression or damage were seen both in the field episodes and in laboratory studies. Similar effects were later shown to be present at approved inclusion levels but were weak and inconsistent. In production units such levels rather than causing impairment of production indices, allow their enhancement.

3.4.1 Carbadox

Slight histological change in adrenal zona glomerulosa cells was seen in 6-week Dutch studies at inclusion levels as low as 25 and 50 mg/kg of piglet diet. At these levels the lesions were completely reversible. The authors considered that their findings indicated adrenally-mediated impairment of fluid and electrolyte balance.

3.4.2 Olaquindox

As for carbadox except that effects were not consistently dose dependent. Comment The effects of both products in pigs are much the same at the same exposure levels. The weak effects seen at user levels would not be judged as adverse from the standpoints of condition and performance while the corresponding clinical biochemistry findings were difficult to interpret, including the corticosteroid measurements. The facts that adrenal and other endocrine effects were seen in rats, but always at higher doses and a longer duration of exposure than those effective in the pig and were absent from a 2 year study in monkeys at up to 20 mg/kg/day (the highest dose tested), suggest the possibility of a special sensitivity of the porcine adrenal to both additives. While it is plain that both additives are toxic to pigs at dietary inclusion levels greater than those authorised, the significance of the observations and findings at authorised levels remains unclear.


3.5.1 Carbadox

JECFA selected QCA as the marker residue, established MRLs and considered the residues after 28 days of withdrawal to be acceptable for consumers, but did not allocate an ADI. When MRLs have been established by other means, the absence of an ADI, although unconventional, is without practical regulatory effect.

3.5.2 Olaquindox

In the case of olaquindox, JECFA has renewed an earlier decision that the residues were temporarily acceptable, has again declined to set an ADI on the grounds that metabolic data were insufficient and has required information on its nominated marker residue in liver and kidney. Finding itself unable to set an ADI is also a JECFA device for encouraging firms to provide missing data. However, an MRL for muscle of 4mg/kg olaquindox-derived residues measured as methyl quinoxaline carboxylic acid (MQCA) was assigned.

Comment: The “new element” ignores the fact that Directive 70/524/EEC does not call for the identification of MRLs for feed additives. Further, it implies that without an ADI, residues are unacceptable, which is contrary to the SCAN Opinions and to JECFA decisions on carbadox and olaquindox. While it is conventional first to establish an ADI and from that to derive MRLs, that is not the only route by which an MRL can be established, nor is it always the sole determinant of the concentrations adopted as MRLs. The MRL of 30 mg/kg liver for carbadox (measured as QCA) established by the FDA on 19 March, 1998 (Sundlof, 1998) was selected because it ensures, using a linear extrapolation method on the results of carcinogenicity studies, that meat residues are free of carcinogenic risk.

The MRL of 4mg/kg olaquindox residues in muscle, measured as 3-methyl quinoxaline-2- carboxylic acid, was established because its quantitative relationship to olaquindox and its other metabolites was known and it is compatible with the amounts known to be present in muscle when, as expressed by JECFA, “olaquindox is used according to good practice in the use of veterinary drugs.”


These toxicities are not specifically required to be investigated in pre-marketing studies on feed additives. The information available has arisen following the post-marketing exposure of some workers in pig production. SCAN is aware of the problem and, has recognised that although severe when it occurs, that it is of low prevalence and preventable.

3.6.1 Carbadox

The new elements contained no reports of such reactions to carbadox in people.

3.6.2 Olaquindox

The 1985 and 1986 reports of 2 cases in Italy involved unprotected workers who hand mixed pig feed. The feed mill regulations now in place were designed to exclude the possibility of such avoidable exposures to feed additives. Further these cases predate the current anti-dusting specifications for olaquindox. The 1996 German report covers cases accumulated over 6 years (1987-1993) and lists 15 cases in pig farmers, 2 of which had contact with finished feed and 13 had contact with a 1g/kg concentrate. Of these, 12 were habitual and 2 occasional home mixers and only 1 wore gloves prior to sensitisation.

Eleven patients used product which did not carry the required label warning against the generation of dust. Many of the patients had dermatitis before presenting with olaquindox photoallergy. Regrettably, as dusting is formulation dependent, no information is given as to the origin(s) of the additives. It is noteworthy, however, that the originator’s patent on olaquindox expired in 1985, followed by a decline in market share with obvious implication for product identity within the German case report.

Comment: The evidence of clinical cases implicates olaquindox, but not carbadox in the claims of those whose comments lead to this re-evaluation. Importantly, nearly all of the patients had failed to use protective clothing and they had practised hand-mixing.

While instances of photoallergy and phototoxicity have occurred, it seems inappropriate to cite this in evidence against olaquindox when so many of the affected workers had not observed safe practice in the handling of a known sensitiser.

The in vitro study which showed a plausible chemical route to possible sensitisation for both additives is of unknown value as a predictor for skin disease in humans.


This mainly concerns workers exposed to carbadox during its production, or to the manufacture of carbadox as an additive or to the additive in use before or during feed assembly. The concern is proper because, as for the farm worker, exposure is to parent carbadox and not, as for the consumer, to trace amounts of a known non-carcinogenic metabolite. Those who change air filter mats in feed plants are cited as at special genotoxic or carcinogenic risk following possible exposure by the dermal and/or inhalatory routes.

3.7.1 Carbadox

The possibility for absorption by workers if exposed to carbadox in factory or farm must exist. Animal tests show that <0.1% of 100mg held to the skin for 24 hours is absorbed. Human skin is a more efficient barrier than rat and rabbit skin and the particle size of the originally approved formulation is such as to make percutaneous absorption highly unlikely. The animal tests show that carbadox is well absorbed across the mucous membrane of the gut. Hence, if not absorbed by the lung, carbadox, if inhaled, is likely to be absorbed following mucociliary clearance and ingestion. Studies on dust exposed workers and non-supplemented control pigs in production units revealed the absence of QCA from urine samples, however.

3.7.2 Olaquindox

Supplementary concentrate (25 %) with higher concentrations of feed additives mixed at farms. The risk of phototoxicity and photoallergy for workers, therefore, does exist and cases are reported.".12 While no specific claim was made for olaquindox in this category, the expectations must be as for carbadox. However, olaquindox applied to the skin of 2 volunteers for 48 hours was not detectable in urine samples.

Comment: Existing, enforceable good manufacturing and working practices should exclude these risks and are a matter for surveillance and enforcement at the level of incorporation of additive into pre-mixtures and of pre-mixtures into feeds as well as during the process of feeding. A recent study implied failures in these areas. See also Section 3.6, 3.8 and 3.9


This relates to the risks outlined in 3.7.

3.8.1 Carbadox

Directive 70/524/EEC specifies already a maximum permitted rate of formation of dust from the anti-dusting formulations whose performance was assessed in the approval process of both additives. A specific exposure avoidance warning is required on the packings of both products (ref I). See also the comment in section 3.9.

3.8.2 Olaquindox

As above.

Comment: The call for short-term measures to be taken at Community level seems not to have taken into account the existing Commission provisions for avoiding worker exposure. See also 3.9.


The CEAS Report in which this issue was raised was written in 1991. Its point was that while the worker safety aspect of the Directive 70/524/EEC approval for carbadox and olaquindox had been based on the assessment of the original manufacturers formulations, the patents on both additives had long since lapsed. Olaquindox and carbadox of various origins are now on the market in additional copy products, none of which has been subject to assessment by the Commission. This situation is legal provided that the copies comply with the specifications in Directive 70/524EEC. There is in the “new elements” recent data which suggests that some samples studied did not comply with the Annex-specified anti-dusting criterion.

Comment: The Commission was advised by SCAN in its 1982 Opinion of health risks from un-evaluated formulations which might be poor copies of the originals. This was rectified by Council Directive 96/51/EC which beginning 1 April 1998 will place both carbadox and olaquindox in an Annex in order to link the authorisation to the person responsible for putting them into circulation. This will require proof that copy products at least meet the specifications of composition and performance of the authorised originals. Those which do not may not be sold after 1 Oct, 1999. Worker safety has also been regulated in by Directive 95/69/EC which addresses exposure in the establishments, farms included, in which additives, including carbadox and olaquindox, are required to be incorporated into animal feeds by specifying equipment, protection and enforcing compliance through inspection and monitoring. For both products for which data was supplied to SCAN for this review compliance with the Stauber-Heubach dusting test was supported by data.


When carbadox was first approved it was on the basis of its growth promotional properties. It later became apparent that the compound was also effective in preventing swine dysentery. It is now the case that the substance is used in pig production in some regions in order to achieve both benefits. The proposition that good husbandry could remove the need for carbadox has to be supported by evidence. The RIKILT-DLO document supporting the Dutch ban on carbadox states that “the applications of carbadox have hardly any alternative.” It goes on to state that “in NL it remains fully effective and free of resistance problems in S. hyodysenteriae.” However, the document quoted only one reference in support of its claim that the efficacy of carbadox as a growth promoter is nil when husbandry is optimal. That reference described a toxicity study, not one designed to measure the efficacy of carbadox on production indices under commercial circumstances.

Conventional pig production can be maintained without carbadox only if other substances of comparable effectiveness remain available. For the growth promotional aspect other authorised antimicrobial feed additives have the same effect.


Although it is understandable that many would respond in the negative to this question, the only scientific contribution proper to such a general question would be to provide on a case by case basis the quantitative risk assessment necessary for the appropriate Risk Analysis. The SCAN evaluations of these hazards in the cases of carbadox and olaquindox presented in Sections 3.1, 3.2, 3.3, 3.4, 3.6, 3.7, 3.8, and 3.9 and are discussed in Sections 4 and 6. Also appropriate to SCAN would be to provide to the risk-benefit consideration component of risk analysis, information which specifies the efficacy-determined part of the overall benefit.

Comment: The SCAN response to this large question is given in the following sections

4. Human Safety/Risk

The issue of human safety can be approached by deciding whether SCAN’s previous safety assessments of carbadox and olaquindox should be re-considered in the light of the foregoing 11 points. It depends also on the evidence which establishes the expected level of exposure under the circumstances of concern. The more obvious of these circumstances are dietary exposure of consumers and workplace exposure for those involved in product manufacture and use.

In its previous opinions SCAN has not established ADIs for carbadox or olaquindox, apparently because the parent molecules were absent from animals at slaughter after 28 days of withdrawal. It did, however, express its view that the ingestion of the low-level, derived residues remaining after 28 days of withdrawal would not be harmful to the consumer. Furthermore, based on measurements of real exposure levels and the requirement for anti-dusting properties, it was considered that the risk of adverse effects for exposed workers would be negligible.


4.1.1 Carbadox

While carbadox and its metabolites desoxycarbadox and hydrazine are proven rodent carcinogens, the consumer is not exposed to these entities because of their relatively short persistence in pig tissues following withdrawal of the supplemented feed. This evidence is then reinforced by the added safety measure of requiring the exceptionally long withdrawal period of 28 days to be applied in pig production. The most persistent metabolite, quinoxaline carboxylic acid, the form in which the consumer would be exposed to residues of carbadox, is a non-carcinogen, was free of toxicity in a 2 year study in rats from which a NOEL of 100 mg/kg b.wt. day, the highest dose tested could be derived. It is the JECFA and FDA nominated marker residue and its concentration in pig liver, the tissue of its longest persistence, is of the order of 20 mg/kg at 28 days of withdrawal.

Eating 100g liver provides an intake of 2mg and 300g muscle would yield <1mg QCA. These amounts fall within the JECFA MRLs of 30mg/kg of liver and 5mg/kg for muscle for carbadox-derived residues. Compared with the rat NOAEL of 100 mg/kg for QCA, this provides a large safety margin. To this could be added the re-assurance that the 2 and 7 year relay toxicity studies in rats and dogs respectively fed daily with zero-withdrawal pig tissues to achieve 10 times the expected daily human exposure to carbadox residues were free of evidence of carcinogenicity and other adverse effects.

SCAN endorses the JECFA and the FDA safety evaluations of carbadox for the consumer and endorses also the MRL concentrations assigned.

4.1.2 Olaquindox

Olaquindox is proportionally much less extensively metabolised than carbadox and is therefore excreted unchanged to a very much greater extent (about 70% of a single dose in urine in 24 h). Because they have not been subject to individual evaluation, less is known of the safety of the metabolites of olaquindox. However, the main metabolites in the pig have been shown to be non-mutagenic. Further, the long term predictive toxicology studies will have evaluated simultaneously the toxicity of olaquindox and all of its metabolites. There is no parent olaquindox in the carcass after 28 days of withdrawal and bound residues are absent or negligible. The multigeneration reproductive toxicity assay yielded a NOAEL of 2.5mg/kg b.wt./day.

Liver at 28 days of withdrawal contains olaquindox equivalent to <5mg/kg radioactivity, kidney <10 mg/kg and muscle 4mg/kg.

These concentrations would yield a residue intake of 0.5mg from liver, 0.5mg from kidney and 1.2mg from muscle, giving a total intake of 2.2mg/day.

Because of the large difference between this market basket-based exposure of consumers and the smallest study derived NOEL (a factor of 70.000) and based on its previous evaluations of the toxicity of olaquindox, SCAN endorses the JECFA safety evaluations of olaquindox for the consumer and endorses also concentration of 4µg/kg muscle assigned as the MRL.

Comment: The calculations reveal the expected low level dietary intakes of residues to be free of unacceptable risk for the consumer.

The possibility that the 28 day withdrawal period necessary to achieve low residue levels may not always be observed is sometimes raised as an additional safety concern, despite the fact that this is a time-point in pig production in Europe at which a change in ration composition is customarily made. Because of this possibility, it is comforting to refer to a report which claims that parent carbadox is no longer present in carbadox-containing meat after 20 minutes of cooking (Hassett et al, 1990). It is also noted that the metabolism studies have confirmed the absence of carbadox and desoxycarbadox from edible tissues by 3 days of withdrawal and this has been essentially confirmed in several open literature studies with the exception of one study which reported the detectability of desoxycarbadox up to day 14 of withdrawal.


Worker safety requires the compliance of the product with the 0.1mg active substance Stauber-Heubach test limit as a means of excluding unacceptable levels of airborne exposure at the workplace. Exposure estimation presumes an 8h shift and a respired volume of 10 cu.m. throughout which an ADI, if available, could be distributed.

Although airborne particles up to 50 microns can deposit in the nasal tract, it is only those.16 smaller than 10 microns which enter and deposit in the lungs. Particles larger than 50 microns are not expected to enter the respiratory tract.

4.2.1 Carbadox

Following discussions with SCAN in the early 80’s, the product of the original manufacturer of carbadox was tested extensively for worker exposure at several locations in various feed mills as well as on farm. These studies were included in the dossiers submitted to SCAN for its 1982 review. The worst case exposure was claimed to have provided a 1000-fold safety margin. SCAN, in turn, expressed itself satisfied that the potential daily exposure of workers of up to 0.05mg airborne carbadox (presumably incorrectly expressed in the 1982 Opinion as 0.05 mg/kg) per 24h did not threaten its 1978 opinion that the risk to health of workers was negligible.

In coming to its 1982 conclusion that exposure of workers to airborne carbadox would pose a negligible risk to health SCAN was provided with linear extrapolation risk assessments based both on the “one-hit” and the “multi-hit” models for predicting the “virtually safe dose” (VSD), i.e. the daily dose which carried a life-time increased risk of cancer which was no more than one in a million. The VSDs were 1.2-4.7 x 10 -4 mg/kg day for the “one-hit” model and 2400 x10 -4 mg/kg b.wt./day for the “multi-hit” model.

These doses were greater than the measured range of feedmill worker exposures (< 0.08- 0.25 x 10 -4 mg/kg b.wt./day). The additive formulation in 1982 was shown to comply with the Stauber-Heubach dust release test limit of 0.1µg. This claim was repeated in the data submitted for present review.

4.2.2 Olaquindox

The maximum permitted intake of 2.5mg/kg/day based on the NOEL of 2.5mg/kg/b.wt transforms into a workplace air concentration of 17.5 mg/cu.m., presuming continuous exposure for the entire 8h shift, respiration of all particles and their complete subsequent absorption. In a feed mill study summary, the sponsors showed that actual release was less than 17.5 mg/cu.m.

The additive on which data was supplied for this review complies with the mandatory Stauber-Heubach limit of active substance in dust.

5. Target Animal Safety

Concern for target animal safety has arisen because of reports in the open literature which recorded episodes of overdose toxicity and because of studies which reported on the occurrence of adrenal cortex damage as a possible mechanism of that toxicity. While an overdose effect is irrelevant to the proper use of feed additives, it is apparent that a weak effect on the adrenals is possible at user levels.

Comment: Of importance is, that at approved use levels pigs have superior production indices while receiving the products. That implies that the animals suffer no net physiological let alone a toxicological impact because reduced rather than enhanced production indices would be the expected consequence were that so.

It is also pertinent to note, that the use of carbadox or olaquindox may reduce the prevalence of enteric disease and allows a greater proportional survival of pigs to slaughter weight. In that sense too, it is apparent that the overall influence of the products is beneficial because of decreased pig morbidity and mortality.

Given that the recommended dose should never be exceeded, SCAN recognises the additional risk posed by non-compliance with approved conditions of use of these additives because of doubts over the existence of a safety margin and recommends that the regulations relating to inclusion levels and their monitoring in animal feed are strictly observed.

6. Conclusion

Although the foundation studies on both additives may be considered dated and some would not meet current specifications of design, many were repeated (some several times) and so the compounds benefit from a considerable and in some respects unique body of safety studies. Furthermore, no modern study has differed in conclusion from the earlier ones to an extent which renders the original SCAN Opinions on safety provisions inaccurate.

The substances share also the property of genotoxicity, differing only in degree. Carbadox is a rodent carcinogen but offers metabolic data which disposes of carcinogenic risk to the consumer provided that its approved conditions of use are observed. Olaquindox is genotoxic, non-carcinogenic but is a rodent tumorigen. Both sponsored products are presented in well-researched formulations designed to protect the feed operative, but the market contains some products about whose safety performance nothing is known to SCAN. That is a problem which the recent generic products amendment to Directive 70/524/EEC removes. There is little doubt that the possible exposure of workers is a risk, especially from carbadox and it is in this area that the Commission should require increased vigilance at the national level to ensure compliance with the adequate and relatively recent safety measures now in place.

A valuable addition to safety which SCAN urges the Commission to make would be, in recognition of the change in feed mixing practices at the level of the farm, to ensure for the farmer the safety provided by anti-dusting formulations by removing from Directive 70/524/EEC the ambiguity of the anti-dusting provision’s applicability to finished feed (ref VII). Also acknowledging the changed circumstances of the feed industry, SCAN recommends that new, extensive studies of real exposures of workers to products of defined origins are now necessary.

When the Commission addresses the question of whether both additives be allowed to remain on the European market, SCAN recommends that within the risk management phase of the risk analysis, due attention be paid to the paucity of the evidence of disease in people additive-exposed as opposed to speculations based on theoretical risks.

The essence of this safety-assessment has been a re-examination of the adverse properties of both additives which had been evaluated previously by SCAN, JECFA and the FDA, and found to be tolerable at expected levels of exposure. For the consumer, there is good security because exposure to carbadox, desoxycarbadox or hydrazine will not occur if user guidelines are followed and, as is also the case for olaquindox, the concentration of total residues to which the consumer will be exposed is, after 28 days of withdrawal, very low and free of toxicological concern. In SCAN’s view, it is the potential for exposure at the workplace, and to carbadox in particular, which merited the greater attention. Again, presuming compliance with the anti-dusting specification and the recent restriction of the use of the additives to fully approved feed mills, SCAN was persuaded that security for the operatives was sufficient and that consequently, any risk of adverse health effects from exposure to additive pure substance was negligible.

While recognising that neither additive has an ideal safety profile in laboratory animal tests, that there is potential for worker exposure to parent substance and that there is evidence for a weak effect on adrenal cortex histology in some pigs at user levels in the diet, SCAN nonetheless concludes, based on the preceding analysis and discussion of those attributes, and in the absence of any evidence which supports the occurrence of adverse health effects in consumer, properly protected worker or target animal under approved conditions of use and despite decades of potential exposure to additive concentrations higher than should now be possible under current regulations, that it is able to maintain its earlier Opinions on the acceptability of the quinoxaline-N-dioxides, carbadox and olaquindox within their previously defined conditions of use 13 . SCAN recommends:
  • that the carbadox or olaquindox residue status of pigs slaughtered for consumption at less than 4 months of age be controlled as a matter of urgency
  • that workplace exposure both to carbadox and olaquindox be re-evaluated, that as a matter of urgency it be supplied with the results of the national monitoring programs of carbadox and olaquindox in animal feeds
  • that potential sponsors for these additives under Directive 96/51/EC be informed immediately of the need to commence relevant epidemiological studies of the health status of additive-exposed workers.

Further Information

For further information on Carbadox usage and residue concerns are noted below:

Antimicrobial use and related management practises among Ontario Swine Producers
Information from a mail study of Ontario Swine Producers to identify the types, frequency and motives for antimicrobial use. No date given for survey. Article has a horrid green background!
© 2000 - 2023 - Global Ag Media. All Rights Reserved | No part of this site may be reproduced without permission.