In October 2002, a review article was published in the Journal of Occupational and Environmental Medicine, Volume 44, pages 968 to 975.
The authors of this article are Peter M. J. Bos, M.Sc., Marloes Busschers, M.Sc. and Josje H. E. Arts, Ph.D.
The title of the article is "Evaluation of the Sensory Irritation Test (Alarie Test) for the Assessment of Respiratory Tract Irritation".
The conclusion of Bos, Busschers and Harts is: the Alarie test is worthless to evaluate irritation. This is supported by their colleagues, C. de Heer and V. Feron. These individuals are from TNO Nutrition and Food Research, Department of Toxicological Risk Assessment, Zeist, The Netherlands.
The following is the summary of their review, as printed by the Journal of Occupational and Environmental Medicine:
"Within the framework of risk assessment of existing substances in the EC the irritating properties on the respiratory tract should be considered. Since no standardized test is available it was studied whether the Alarie test could be used for this purpose, as proposed by the Technical Guidance Document for new and existing substances. The available literature on respiratory tract irritation, seen as a local inflammatory response and/or tissue damage, after single and repeated (few-day) exposure was evaluated and compared with data on sensory irritation. No relation was found between the sensory irritation potential (as measured by the Alarie test) and local tissue damage (histopathological changes) in the respiratory tract after single or repeated exposure. It was concluded that the Alarie test is inappropriate to evaluate respiratory tract irritation. In addition, the available data do not support a quantitative potency ranking for man based on the RD50 obtained with experimental animals. (J. Occup. Environ. Med. 2002; 44: 968-976)."
If you read the title of this article again, it is an "evaluation" of the Alarie test and this "evaluation" is published in 2002. One expects Bos, Busschers and Arts to provide up-to-date information on the Alarie test. None is presented.
I will now correct this dreadful 2002 "review" by Bos, Busschers and Arts. No picayune corrections here. It simply can be shown that the claims made are demonstrably false.
Bos, Busschers and Arts wrote "Within the framework of risk assessment of existing substances in the EC the irritating properties on the respiratory tract should be considered. Since no standardized test is available, it was studied whether the Alarie test could be used for this purpose, as proposed by the Technical Guidance Document for new and existing substances".
Did the TGD propose anything? Nothing could be further from the truth. A click on the following link will bring you to the TGD document quoted by Bos, Busschers and Arts. The relevant part starts on page 77 of the displayed PDF file at this site.
On page 80 of this document, you will find the following statement:
"There are non-Annex V test methods for respiratory irritation (e.g. the Alarie test, which may provide useful information on sensory irritation of the upper respiratory tract). It should be noted that the use of data from the Alarie test to derive limit values has been criticized (e.g. inter-laboratory differences and inconsistent intra and inter-species differences were noted) and care should be taken in the interpretation of the results of this test (Bos, 1992)".
So, we have the TGD quoting Bos, warning that inter-laboratory differences and inconsistent intra and inter-species differences were found.
Here are the results showing the intra-species differences when sulfur dioxide was used with a variety of mice. The RD50 values obviously show intra-species differences:
Source: Abbreviated from Alarie et al. 1980
If there is a factor of 10 between the most and least sensitive strains of mice as given in the table above, what do Bos et al. expect? That the sensitivity of the XYZ rat should match any of the mice strains noted above? Do these authors think that a rat is just a big mouse and that no inter-species or intra-species differences should exist? Do they think that all strains of mice should have the same sensitivity? How absurd. Does it surprise anybody that if Chemical A is tested in Laboratory B with mouse strain X (the most sensitive) vs. the same chemical being tested in Laboratory C with mouse strain Z (the least sensitive) the results will be different? Do I need to go further? The Alarie test uses male Swiss-Webster mice. The quantitative extrapolation to humans is based on the results obtained with these mice. Not rats, dogs or rabbitts. An investigator can substitute a different strain of mice. As long as their sensitivity is close to male Swiss-Webster or, they can use a different strain of mice for a specific purpose. Rats are not appropriate for evaluation of sensory irritation, at least the strains tested so far. Their sensitivity is too low. Rats can be used with the Alarie test as presented below for asthma models, provided that the correct strain is used. This is beyond the comprehension of Bos, Busschers and Arts. They have not yet grasped Toxicology 101.
We now come to the second point of the review by Bos, Busschers and Arts.
From their summary quoted above we have:
"The available literature on respiratory tract irritation, seen as a local inflammatory response and/or tissue damage, after single and repeated (few day) exposure was evaluated and compared with data on sensory irritation. No relation was found between the sensory irritation potential (as measured by the Alarie test) and local tissue damage (histopathological changes) in the respiratory tract after single or repeated exposure".
These authors are chasing a possible correlation between tissue inflammation or corrosion and sensory irritation potency. What would be the basis for such a correlation? No basis is presented. Instead of doing all this work to chase an impossible correlation, they could have read the Alarie articles. No such claim has ever been made and no such correlation can ever exist as already stated (Alarie, 1981).
However, there is a major point to consider here. This is the fact that, as stated by Bos, Busschers and Arts, the TGD is lacking in a standard test to evaluate inflammatory reactions on the respiratory tract. Bos, Busschers and Arts are extremely concerned about this.
Back in the 1970s my laboratory was involved in evaluating the toxicity of smoke from burning materials. It became quite clear that smoke from any burning material produced sensory irritation. It also became quite clear that while smoke from burning wood or burning polyvinylchloride produced sensory irritation, smoke from polyvinylchloride was corrosive to the eye and upper respiratory tract while smoke from wood was not. With Helen Lucia, M.D., a pathologist, a systematic approach was developed. Four sections of the nose were made and examined. A rating system for corrosive damage was arrived at (Lucia et al. 1978).
Since then, I have collected over 150 articles on histopathological examination of the nose in mice or rats. Bos, Busschers and Arts quoted a total of three! If they are so concerned about providing the TGD with good methods to evaluate inflammatory reactions, they should spend their time on reviewing the superb literature on nasal histopathological examinations and come up with a recommendation.
Now, for the last issue. As stated in the above summary of the Bos, Busschers and Arts article, we have the following:
"It was concluded that the Alarie test is inappropriate to evaluate respiratory tract irritation. In addition, the available data do not support a quantitative potency ranking for man based on the RD50 obtained with experimental animals".
These are unequivocal statements. Let us see if they can stand up to scrutiny.
The first attempt to develop a quantitative relationship between the potency (RD50) of sensory irritants measured in male Swiss-Webster mice for prevention of irritation in exposed industrial workers was published in 1979 (Kane et al., 1979).
In this article, we did not propose a specific relationship between RD50 and an occupational exposure level (OEL) such as a Threshold Limit Value (TLV). Rather, with our limited data (RD50 for 11 industrial chemicals), we proposed that a TLV should be between 0.01 and 0.1*RD50 to prevent irritation in industrial workers.
The second phase of relating RD50 to TLVs came in 1981 when we had RD50 values obtained in male Swiss-Webster mice for 21 chemicals. Here we proposed that 0.03*RD50 would be appropriate to set a TLV to prevent sensory irritation (Alarie 1981). The third phase came in 1986, now with 46 chemicals (Alarie and Luo, 1986). The final phase came in 1993 (Schaper 1993), now with 89 chemicals, updated in 2000 (Alarie et al. 2000). These articles included not only the data on male Swiss-Webster mice, but other strains of mice of comparable sensitivity.
Bos, Busschers and Arts wrote "A strong correlation (r2= 0.78) was reported between 0.03*RD50 and 1991 TLVs based on a database of 89 substances". Yet, the summary of their article stated "In addition, the available data do not support a quantitative potency ranking for man based on the RD50 obtained in experimental animals". If there is such a strong correlation between 0.03*RD50 and the TLV for 89 chemicals, how can it be that "In addition, the available data do not support..."?
Enough said about these "scientists" at the TNO.
Time to move on.
Between 1992 and 2002 the Alarie test has undergone a variety of refinements, to make it a much more valuable test to evaluate irritation of airborne chemicals on the respiratory tract. The Alarie test can now detect irritation of the conducting airways of the lung (bronchi and bronchioles) as well as irritation at the alveolar level (pulmonary irritation); in addition to sensory irritation. The Alarie test has also been fully computerized and limits of detection established to prevent any false positive. These innovations have been reviewed (Alarie et al. 2000; Alarie, 2000). As further elaborated below, the Alarie test is now used in animal models of asthma, bronchial hyperreactivity and very recently, for sensory irritation hyperreactivity. A large number of articles were published between 1992 and 2002 related to the Alarie test. These are included in the List of References given below.
The mechanism of sensory irritation, in both mice and humans received great attention between 1992 and 2002. Published articles on this topic are also included in the List of References given below.
Two articles were published (Vijayaraghavan et al. 1993; 1994) establishing the fundamental principles to detect sensory irritation, bronchial irritation and pulmonary irritation with the use of a computerized breathing pattern analysis system; extending the capabilities of the Alarie test.
Two articles were published (Boylstein et al. 1995; 1996) using the computerized Alarie test, demonstrating how mixtures of airborne chemicals can induce a variety of effects including sensory irritation, bronchoconstriction and pulmonary irritation. A group of only four male Swiss-Webster mice was needed to demonstrate irritation at three different levels of the respiratory tract. Many investigators have use this computerized version of the Alarie test to evaluate a wide variety of mixtures, as listed in the references below.
No toxicological bioassay has a limit of detection, except for the Alarie test. Limit of detection for sensory irritation already existed, but limit of detection for sensory irritation with the computerized Alarie test were defined, as well as limits of detection for bronchial irritation and pulmonary irritation (Alarie, 1998). This eliminates the possibility of false positives.
Asthma is obviously a major problem in humans. Since the computerized Alarie test can detect airflow limitation along the conducting airways of the lung (due to either bronchoconstriction, inflammatory reaction or mucus accumulation) it has been used to evaluated immediate and delayed allergic reactions in mice, as well as bronchial hypperreactivity (Neuhaus-Steinmetz et al. 2000; Glaab et al. 2001)
The Brown Norway rat is attractive for an asthma model because of the type of antibodies produced when exposed to allergens and a fairly well developed bronchial musculature. The Alarie test was adapted to use this strain of rat for research on asthma (Glaab et al. 2002) and the measurements made reviewed (Hantos and Brusasco, 2002).
A recent article (Path et al. 2002) also used the Alarie test to evaluate a mouse model of asthma. These researchers used transgenic mice overexpressing nerve growth factor (NGF) and showed an augmentation of allergic early-phase bronchial reaction in these mice. They also showed an augmentation of the sensory irritation reaction to capsaicin (a potent sensory irritant) in these mice. This is the first demonstration of sensory irritation hyperreactivity. It is a very important finding. Numerous reports of humans having a higher than expected sensitivity to airborne sensory irritants have appeared in the literature over the past 15 years. Many hypotheses have been advanced. Why do some humans feel much more intense sensory irritation than others, particularly at low concentrations of airborne chemical mixtures such as with indoor air problems? These findings in mice may open the door for experimental and fundamental explorations in this area.
Sensory irritation is such a common reaction in humans and so many chemicals are sensory irritants that establishing the mechanism(s) by which it occurs is important. Alarie (1973) reviewed the early literature on this aspect and this was amplified in another review by Nielsen (1991). Major advances were made between 1990 and 1998, as presented by Alarie et al. (2000).
In short, sensory irritation by airborne chemicals occurs via two different mechanisms. The first one is applicable to non-reactive chemicals (mainly commonly used solvents) and their potency as sensory irritants can be fully described by a set of physicochemical descriptors. This is an extremely useful finding because when a mixture of these chemicals is present (a common finding in industry or indoor air) the potency the mixture can be estimated and there is no need for animal testing of such mixtures (Alarie et al. 2000). This mechanism was already proposed in 1982 (Nielsen and Alarie, 1982). It was then firmly established with the help of Michael H. Abraham, Ph.D. and the data base prepared by M. Schaper, Ph.D.
The second mechanism is applicable to reactive chemicals. Here, the findings are of more limited practical use, in the sense that the potency of closely related chemicals can be estimated to be within a factor of about three or so, with reservation that the reactivity of a related chemical must be evaluated with great care (Alarie et al. 2000).
During the last ten years, many new developments have taken place with the Alarie test. Researchers have used it as a basic tool in different areas, beyond what Alarie imagined it could be used for when first published (Alarie, 1966).
In the entire world of toxicology, there is only one validated bioassay to predict toxicity to humans, other than lethality. A chemical found to be a positive sensory irritant in male Swiss-Webster mice (or a strain of similar sensitivity) will be positive in human at a similar exposure concentration. A chemical found to be a non-sensory irritant will be negative in humans (Alarie, 1973; Alarie et al. 2000).
In the entire world of toxicology, there is only one calibrated bioassay to predict safety to humans. When the potency (RD50) of a sensory irritant is obtained, an Occupational Exposure Level (OEL) can never be higher than 0.03*RD50. Some modifications can be made to this general rule of 0.03*RD50, when appropriate. For example, not all chemicals will produce a plateau response of sensory irritation during exposure, or the slope of the concentration-response curve may be steeper or more shallow for a particular chemical. Nevertheless, 0.03*RD50 is the established starting point.
Also, while 0.03*RD50 has been the established starting point to prevent irritation in industrial workers, a more conservative 0.001*RD50 has been used to prevent indoor air problems (Alarie et al. 2000). Nielsen (see list of references below) has discussed in great details how the starting points of 0.03*RD50 or 0.001*RD50 should be modified, depending on the specific findings with the Alarie test, as well as what is available from the toxicological literature for each tested chemical.
The references cited above can be found in the list of references provided below. This list of references also includes references on the Alarie test published between 1992 and 2002, something that Bos, Busschers and Arts omitted in their “evaluation” of the Alarie test.
If you have questions or comments on the Alarie test, please submit them via e-mail at firstname.lastname@example.org.