• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Dose-Response Relationships and Threshold Levels in Skin and Respiratory Allergy

Posted on: Friday, 21 April 2006, 06:00 CDT

By Arts, Josje H E; Mommers, Carolien; de Heer, Cees

A literature study was performed to evaluate dose-response relationships and no-effect levels for sensitization and elicitation in skin- and respiratory allergy. With respect to the skin, dose- response relationships and no-effect levels were found for both intradermal and topical induction, as well as for intradermal and topical elicitation of allergenic responses in epidemiological, clinical, and animal studies. Skin damage or irritation may result in a significant reduction of the no-effect level for a specific compound. With respect to the respiratory tract, dose-response relationships and no-effect levels for induction were found in several human as well as animal studies. Although dose-response relationships for elicitation were found in some epidemiological studies, concentration-response relationships were present only in a limited number of animal studies. Reported results suggest that especially relatively high peak concentrations can induce sensitization, and that prevention of such concentrations will prevent workers from developing respiratory allergy. Moreover, induction of skin sensitization may result in subsequent heightened respiratory responsiveness following inhalation exposure. The threshold concentration for the elicitation of allergic airway reactions in sensitized subjects is generally lower than the threshold to induce sensitization. Therefore, it is important to consider the low threshold levels for elicitation for recommendation of health-based occupational exposure limits, and to avoid high peak concentrations. Notwithstanding the observation of dose-response relationships and no-effect levels, due to a number of uncertainties, no definite conclusions can be drawn about absolute threshold values for allergens with respect to sensitization of and elicitation reactions in the skin and respiratory tract. Most predictive tests are generally meant to detect the potential of a chemical to induce skin and/or respiratory allergy at relatively high doses. Consequently, these tests do not provide information of dose-response relationships at lower doses such as found in, for example, occupational situations. In addition, the observed dose- response relationships and threshold values have been obtained by a wide variety of test methods using different techniques, such as intradermal exposure versus topical or inhalation exposure at the workplace, or using different endpoints, which all appear important for the outcome of the test. Therefore, especially with regard to respiratory allergy, standardized and validated dose-response test methods are urgently required in order to be able to recommend safe exposure levels for allergens at the workplace.

Keywords Dose-Response, Elicitation, Respiratory Allergy, Sensitization, Skin Allergy

I. INTRODUCTION

Allergy is a diverse family of diseases caused by untoward immune reactions that may ultimately lead to tissue inflammation and organ dysfunction. Allergy comprises a two-phase process: (1) a generally symptom-free sensitization phase and (2) a symptomatic effector phase. After initial encounters with a particular allergen, a primary immune response is mounted, resulting in a state of heightened responsiveness to this allergen (induction or sensitisation). Upon subsequent exposures (challenge), a vigorous immune response is provoked that can result in clinically manifest adverse health effects (challenge or elicitation).

Two important types of allergic diseases at the workplace are skin and respiratory allergy. These are caused by exposure to exogenous substances, that is, certain low-molecular-weight (LMW) chemicals (5000 Da) or high-molecular-weight (HMW) compounds (usually proteins, >5000 Da; Chan-Yeung and Malo, 1995). Occupational exposure of humans to LMW allergens via the skin may be considerable, such as found in auto body shop workers exposed to isocyanates despite protective clothing (Liu et al., 2000). Occupational skin allergy induced by allergenic proteins is less frequently reported since such chemicals do not easily penetrate the skin. Occupational asthma resulting from respiratory sensitization can be life-threatening (Fabbri et al., 1988; Ehrlich, 1994), and the majority of patients do not completely recover even when removed from exposure for several years (Saetta etal., 1992, 1995; Chan- Yeung and Malo, 1995; O'Neill, 1995). The number of known respiratory allergens is increasing and includes both HMW compounds and LMW chemicals. LMW compounds are reactive chemicals and act as an allergen when bound to an appropriate carrier protein; that is, a complete hapten-carrier antigen is formed. The antigenic determinant may be the hapten, or it may be a new antigenic determinant formed by the chemical reaction between the hapten and the carrier protein. This requirement for haptenation is an important distinction between protein and LMW-induced allergy (Cullen et al., 1990; Chan-Yeung and Malo, 1995).

Given the serious health problems of skin and respiratory allergy together with the continuous introduction of new compounds into workplaces, early identification of chemical sensitizers is very important. With respect to skin allergy, validated predictive tests exist. For respiratory allergy, however, validated tests are lacking. Although a number of animal test protocols have been published to detect respiratory allergy (see for reviews Briatico- Vangosa et al., 1994; Pauluhn et al., 1999; Pauluhn and Mohr, 2005), none of these are widely applied or fully accepted. Moreover, most predictive tests carried out are meant to detect potential rather than potency.

Determination of potency is important since the development of sensitivity to allergens and the severity of symptoms may be directly related to exposure levels. Exposure-response data are needed to assess the effective dose(s) and "no-observed-effect" or threshold levels for sensitization and elicitation upon challenge to specific skin and respiratory allergens. No-effect levels are to be used as a basis for risk assessment and management to prevent allergic sensitization and specific hyperreactivity in exposed workers. Established thresholds may thus form a basis for the recommendation of health-based occupational exposure limits.

The objective of this survey was to review the available literature on dose-response relationships and no-effect levels for both skin and airway sensitization and elicitation that could be used to obtain safe exposure concentrations for allergens at the workplace. Epidemiological, clinical, and animal studies were reviewed.

II. SKIN ALLERGY

Skin allergy generally is associated with a cell-mediated immune reaction, formerly described as type IV hypersensitivity reaction according to the classification of Cell and Coombs (Roitt et al., 1998), and is then called allergic contact dermatitis. In this type of reaction, allergens come into contact with Langerhans cells in the skin, which carry the allergen through the lymphatics to the regional or local lymph nodes, where the allergen is presented to naive CD4+ T lymphocytes. After several days, clones of these T cells become sensitized to the allergen and circulate as memory cells in the bloodstream, and some reside in the skin. Upon reexposure to the allergen, the specifically sensitized memory T lymphocytes infiltrate the site of contacted skin, where they release cytokines and other factors that attract the numerous inflammatory cells that are responsible for the epidermal cell injury. Clinically, the contacted skin becomes erythematous and swollen with vesiculation within 6 hours to several days after reexposure. Hence the term "delayed reaction" for the cell-mediated response. Other allergic skin reactions require in the sensitization phase the production of allergen-vasoactive mediators from mast cell granules. Clinically, this results in urticaria (raised pruritic paules and plaques) and angioedema, which develop acutely and may fade within hours. Persistent plaques, however, can last for more than 24 hours. Because of the involvement of immunoglobulin E (IgE) antibody and the acute response, urticaria and angioedema are called antibody- or IgE-dependent, anaphylactic or immediate-type reactions, formerly called type I hypersensitivity reaction according to the classification of Gell and Coombs. Urticaria occurs in atopic as well as in nonatopic individuals.

The most common skin allergens at the workplace are a number of low-molecular-weight chemicals, such as Kathon CG, aldehydes, nickel salts, certain isocyanates, DNCB, DNFB, oxazolone, PPD, and rubber chemicals. Far less is reported on effects of skin allergenic proteins, probably because these HMW compounds do not easily penetrate the intact skin.

With respect to skin effects, a compound is usually classified as a skin sensitizer (EC risk phrase R43) upon a positive skin sensitization test such as the guinea pig maximization test (GPMT; Magnusson and Kligman, 1969) or the Buehler test (Buehler and Griffith, 1975), incorporated in the Organization for Economic Cooperation and Development (OECD) guidelines (OECD 406; OECD, 1992). However, according to European Commission (EC) guidelines, scientific justification should be given when the Buehler test is used. The sensitivity of the GPMT can be increased by means of intradermal injection with Freund's complete a\djuvant. Therefore, different labeling criteria are used: A positive result in the GPMT is obtained when 30% or more of the animals react in a test involving adjuvant, or when 15% or more of the animals react without using adjuvant. Two other tests are also used: the mouse ear swelling test (MEST; Gad et al., 1986) and the local lymph node assay (LLNA; Dearman et al., 1992a, 1992b). The LLNA investigates sensitization potential by measuring cell proliferation in the lymph nodes draining the area of application that is the ear. Initially, these tests were used as a first stage in the assessment of skin sensitization potential. If a positive result was seen in either assay, a test substance was classified as a sensitizer and it was not necessary to conduct a further guinea pig test. However, if a negative result was seen, a guinea pig test still had to be conducted. Recently, the LLNA (OECD 429; OECD, 2000) has been indicated as the preferred skin sensitization test.

A. Dose-Response Relationships and Thresholds for Skin Sensitization

1. Low-Molecular-Weight Chemicals

a. Epidemiological or Clinical Studies (Table 1 ). Multiple patches for induction of sensitization, followed by a 2-week rest period and a subsequent challenge with a patch at a new skin site, are generally used to determine skin allergy in humans (Stotts, 1980). This method has shown dose-response relationships for several skin sensitizers like benzocaine, mafenamide, bronopol, formaldehyde, glutaraldehyde, and PPD (Marzulli and Maibach, 1974). A linear relationship was found between the degree of sensitivity and the sensitizing dose (in log) using DNCB. The 50% sensitizing dose (ED50) calculated from the dose-response curve was 116 g (Friedmann et al., 1983). Patch tests with Kathon CG showed that the induction threshold for sensitization was between 10 and 20 ppm (0.001-0.002%), similar to the value found in animal studies (Cardin et al, 1986). With respect to the dose per unit area, it was shown that high nickel concentrations applied to limited skin areas (i.e., a local high concentration, comparable to skin occlusion) were needed to sensitize a substantial number of individuals (Menn and Calvin, 1993). A nickel concentration of 0.5 g/cm^sup 2^ has been suggested as a no-effect level for sensitization, based on a wide range of studies (Menn, 1994).

The vehicle used may play a significant role. Upon testing the sensitization potential of pure hydroxycitronellal, none out of 39 subjects was sensitized at 5%, 1 out of 38 subjects was sensitized at 7.5%, while 6 out of 40 subjects were positive at 10% (Steltenkamp et al, 1980). Similar results were reported by Ford et al. (1988): No reactions were induced at 5%, while 10% sensitized 2 out of 25 subjects, and a 12% concentration sensitized 8 out of 11 subjects. In contrast, hydroxycitronellal in petrolatum at 5% or 12% did not induce sensitization (Opdyke, 1974). For cinnamic aldehyde, results also depended on the vehicle used. Using ethanol as vehicle, cinnamic aldehyde did not induce sensitization at concentrations of 0.1 or 0.5%. At 1%, 5 out of 41 subjects were positive, and at 1.25%, 5 out of 10 subjects were sensitized. Using petrolatum as a vehicle, in contrast, subjects treated with 20 (i.e., 4 sets of 5) consecutive 48-h patch test exposures responded already to 0.5% cinnamic aldehyde (Danneman et al., 1983).

b. Animal Studies (Table 1).

Skin sensitization measured by delayed-type skin reactions after dermal challenge. For several compounds, doseresponse relationships were obtained. In a GPMT in which both induction and challenge application concentrations were varied in order to determine relative sensitization potencies of substances, groups of animals were sensitized by intradermal and topical induction of various concentrations of 1 of 15 chemicals. Three weeks after the initial injection, animals were challenged with one of various concentrations by an open patch test. The sensitization rate was found to be related to the sensitization dose for all 15 chemicals tested. The minimum induction concentrations that induced a positive response were, for instance, 20 ppm (0.002%) DPTU, 500 ppm (0.05%) MBT, and 10 ppm (0.001%) DNCB. No-effect levels were 2 ppm (0.0002%) DPTU, <500 ppm (0.05%) MBT, and 1 ppm (0.0001%) DNCB (Nakamura et al., 1994; see also Table 1).

With respect to the type of the dose-response curve, a modified GPMT using five dose groups (multiple dose design) resulted in nonlinear, sigmoid, sensitization dose-response curves for formaldehyde, cinnamic aldehyde, MBT and Kathon CG. The intracutaneous sensitization EC50s were found to be 0.96% for formaldehyde, 0.04% for cinnamic aldehyde, 0.07% for MBT, and <3 ppm (0.0003%) for Kathon CG. For MBT a noeffect level of 0.03% was found, and for the other three compounds a no-effect level could not be established (Andersen et al., 1995). A sigmoid dose-response relationship for sensitization with saturation at higher doses has also been described for other LMW chemicals like DNCB (Buehler and Griffith, 1975), sultones (Ritz et al., 1975), p-nitrobenzyl compounds (Roberts et al., 1983), and chlorocresol (Andersen and Hamann, 1984). Also using "Buehler's occluded epicutaneous patch test," a sigmoid sensitization dose-response curve was obtained for Kathon biocide. There was evidence of a very steep slope in this sigmoid response curve, resulting in a substantial increase in sensitization with relatively small changes in concentration. A no- response concentration was observed for several combinations of induction and challenge concentrations. The EC^sub 50^ for induction was 88 ppm (0.0088%) at a challenge concentration of 2000 ppm (0.2%; Chan et al., 1983). In a cumulative contact enhancement test (CCET), a test method using multiple topical applications, a dose-related sensitization potential with a bell-shaped curve and threshold levels could be detected for MDBGN (Wahlkvist et al., 1999a). Dose- related sensitization potential (monotone or nonmonotone curves [i.e., increases at lower concentrations, decreases at higher concentrations]) was also reported for potassium dichromate and hydroxycitronellal using the CCET or the FCAT, but not the GPMT. However, for all three tests, EC^sub 50^ values and threshold concentrations could be calculated (Wahlkvist et al., 1999b).

Different types of sensitization dose-response curves have been observed in different species using the same compounds. Guinea pigs topically sensitized with various doses of HMDI and PiCl showed a normal dose-response curve upon challenge; that is, reactions were more severe in animals receiving the higher sensitization doses. A no-observed-effect level was found in animals receiving 0.01 mg HMDI or PiCl. In mice, in contrast, the sensitization dose-response to both HMDI and PiCl showed a bell-shape curve; that is, the response increased with increasing dose, showed a plateau at higher doses, and decreased at the highest exposure concentrations tested (Stadler and Karol, 1985).

Different results have been obtained using intradermal versus topical exposure. In the GPMT, intradermal concentrations ranging from 0.01% to 3% and topical (second induction) concentrations ranging from 0.5% to 20% formaldehyde were tested. A nonlinear, sigmoid, sensitization dose-response relationship was obtained following challenge with 1% formaldehyde, with saturation at high induction doses. The incidence of contact sensitivity depended on the intradermal application (first induction) dose but not on the topical (second induction) dose. The formaldehyde concentration at which contact sensitivity was seen in 50% of the animals (EC^sub 50^) was 0.061% and 0.024%, using two guinea pig strains, respectively (Andersen et al., 1985). In guinea pigs, using an open epicutaneous application method, animals were sensitized 5 days a week for 4 weeks with 0.3-3% nickel sulfate using adjuvant. They were challenged 14 and 21 days after the last treatment using 2% nickel sulfate in water and 1% in petrolatum, respectively. A nonlinear dose-response curve in incidence was obtained indicating that there was an optimal induction dose (Nielsen et al., 1992). From the same group, another study was reported also using nickel sulfate. Six intradermal (0.01-3.0%) and six topical (0.25-10.0%) concentrations were used for induction, while 1% and 2% concentrations were used for patch test challenges. Intradermal induction was performed on day 0, topical induction on day 7, and subsequent challenges were performed on days 21 and 35. After the first challenge with 1% nickel on day 21, a linear relationship was obtained between the intradermal induction dose and the number of animals developing contact dermatitis, resulting in a significant maximum contact dermatitis rate of 42% (at an intradermal induction with 3%). The intradermal nickel dose was decisive for the development of contact allergy, while variation in the topical induction concentration had no effect. Another patch challenge with 2% nickel on day 51 revealed a nonlinear (bell-shaped) dose- response relationship following a topical booster with 10% nickel on day 35. A maximum response frequency of 45% was found after initial intradermal induction with 3%, and initial topical induction with 2% nickel. Thus, the dose-response relationship was linear at the initial challenge and nonlinear after the rechallenge following a topical booster. This test, however, was not found to be an efficient animal model for experimental nickel contact allergy, as it was not possible to achieve strong, lasting, and sensitive nickel sensitization in the majority of the test animals (Rohold et al., 1991).

TABLE 1

Summary of methods to determine dermal sensitization dose- response relationships

TABLE 1

Summary of methods to determine dermal sensitization dose- response relationships

Also, the number of applications may play a significa\nt role. In the aforementioned study of Wahlkvist et al. (1999a), using MDBGN, no statistically significant sensitization was induced in the GPMT using one intradermal application followed by a topical application. In contrast, in the CCET, using multiple topical applications, a dose-related sensitization potential could be detected for MDBGN.

Changing the dose by changing the volume is not necessarily the same as changing the concentration, which may lead to differences in surface concentrations. In a MEST, used to investigate the relationship between sensitizing dose and skin reactions to DNFB, groups of mice were sensitized on the shaved back skin with various volumes of a fixed concentration of DNFB. Upon challenge, ear swelling reactions were greater at lower DNFB sensitization doses. Reactions were even observed at the lowest dose tested (Brown and Shivji, 1991). For DNFB, DCC, and oxazolone, ear swelling dose- response relationships with a cubic trend (optimum for DNFB and DCC, plateau for oxazolone) were obtained after cutaneous sensitization with different concentrations followed by a challenge with the same compound (Flint et al., 2003).

Skin sensitization measured directly by proliferation in skin- draining lymph nodes. In the LLNA, dose-response relationships and sensitization thresholds have been obtained for Kathon CG and two other isothiazolinone-based chemicals (Botham et al., 1991b) and a number of other chemicals such as DNCB, citral, DNBS, chloromethylisothiazolone, diphencyclopropenone, ethylene glycol dimethacrylate, formaldehyde, glutaraldehyde, TDI, MDI, HDI, TMA, PA, isopropyl myristate, MDBGN, PPD, PDC, AMT, nickel sulfate, eugenol, isoeugenol, geraniol, hydroxycitronellal, isopropyl myristate, linalool, cinnamic aldehyde, alpha-hexylcinnamaldehyde, BMHCA, TCSA, squaric acid dibutyl ester, tetramethylthiuram disulfide, and a piperidinyl chlorotriazine (PCT) derivative (see Table 5). A dose-response relationship was also found for oxazolone, but no threshold was reported as more than threefold increases above control were obtained even at the lowest tested concentration of 0.0025% (Loveless et al., 1996; Ulrich et al., 1998). Thresholds were neither found for MA and HHPA (Plitnick et al., 2003). In the study of Ryan et al. (2002), different vehicles were tested, among others a vehicle for use with water-soluble compounds.

In an LLNA by van Och et al. (2000), 10 chemical allergens were tested, each chemical at 4-6 concentrations. For each compound, concentration-related increases were obtained and the dose-response data were analyzed by nonlinear regression analysis. Four different concentration-response curves were obtained: exponentially shaped curves (two types) were seen for six of the chemicals (benzocaine, diethylamine, MBT, TMTD, ZDMC, and oxazolone), three of the chemicals showed a sigmoidally shaped curve (DNCB, TDI and TMA), whereas one curve showed a logarithmic dose-response profile (phthalic anhydride). Oxazolone showed the strongest sensitizing potency, followed in this order by DNCB, TDI, TMA, PA, TMTD, ZDMC, MBT, benzocaine, and diethylamine.

In the LLNA, irritating compounds were not found to induce cell proliferation (Dearman et al., 1992a, 1992b), although some studies suggest this distinction between irritants and allergens was not always evident. For instance, the nonsensitizing skin irritant sodium lauryl sulfate induced a positive response in this assay (Loveless et al., 1996), whereas the nonsensitizing skin irritant methyl salicylate induced a nearly positive response (increases of up to 2.9-fold above control at 20%; Kimber et al., 1996). As a result, compounds with strong irritating properties could falsely be classified as allergens, or the allergenicity of chemicals with both allergenic and irritating properties could be overestimated (Robbins et al., 1991; Basketter and Scholes, 1992; Edwards et al., 1994; Montelius et al., 1994). Arts et al. (1997) reported that formaldehyde (which is both an irritant and a sensitizer) caused a dose-dependent and a slightly more than threefold activation of the lymph nodes, but irritant properties might have substantially contributed to the LLN cell proliferation observed. Basketter et al. (1996) discussed the possible mechanisms involved in low level activity in the LLNA by some but not all irritants, and noted that a weak response in the LLNA to a known strong irritant should prompt further consideration.

Skin sensitization measured directly by cytokine profiles. Increases in cytokine levels, induced by sensitizers, can be measured using different assays such as enzyme-linked immunosorbent assay (ELISA) (Ryan et al., 1998; Dearman and Kimber, 1999; Dearman et al., 1999; Vandebriel et al., 2000; Ulrich et al., 2001), reverse- transcription polymerase chain reaction (RT-PCR) (Ryan et al., 1998; Ulrich et al., 1998,) or a multiprobe ribonuclease protection assay (RPA; Plitnick et al., 2002). Changes in cytokine profiles are generally measured in mice upon flank application of the chemicals on days O and 5, followed by ear application on days 10, 11, and 12. At various times following ear exposure mice are sacrificed, draining lymph nodes removed, and one of the different assays applied. In one of these studies (Plitnick et al., 2002), dose- response relationships in several cytokine levels (interleukin [IL]- 4, IL-5, IL-10, IL-13) were obtained using TMA. However, this was only observed when the TMA concentration was varied during the ear exposures. Variation in concentration during flank application did not result in dose-response relationships; variation in concentration during both exposures resulted in an initial increase at lower doses followed by a subsequent decrease at higher doses. Dearman et al. (1999) found a dose-related increase in IL-10 using different concentrations of glutaraldehyde. Concentrations were similar at each flank or ear application. Interestingly, the second series of applications on days 10-12 were considered by the investigators (Plitnick et al., 2003) to be the elicitation/ challenge phase of the exposure, which would imply that the cytokine profiles test should be considered a full test, including sensitization and elicitation. However, as cytokine levels are no effect parameters, that is, they do not reflect clinical symptoms whatsoever, this test was not considered to be an endpoint test and was therefore not incorporated in the elicitation section.

Skin sensitization measured by total serum IgE or specific antibody levels (as function of immediate reaction). Groups of mice received a topical application of various concentrations of the chemical under investigation. Changes in serum IgE after topical reexposure to the chemicals 7 days later enabled the examination of dose-response relationships. Exposure to TDI, MDI, HDI, IPDI, and TMA caused a significant dose-related increase in serum IgE- concentration measured 14 days after the initiation of exposure. The no-effect level was lower than 1 % w/v for the diisocyanates and lower than 10% w/v for TMA. In contrast, exposure of mice to DNCB, oxazolone, or glutaraldehyde produced only a relatively small elevation in serum IgE levels at a high concentration only, whereas formaldehyde did not induce an IgE antibody response (Hilton et al., 1995; Potter and Wederbrand, 1995). In another experiment with TDI, various concentrations of TDI were administered 15 times over a 3- week period, or 30 times over a 6-week period. The apparent TDI threshold for IgE antibody production increased with an increase in the number of TDI applications. In contrast, multiple applications with high TDI concentrations increased total IgE antibody production when TDI was administered in 15 or 30 doses rather than in 2 doses (Potter and Wederbrand, 1995).

Blaikie et al. (1995) investigated a single intradermal injection model in the guinea pig to establish the relationship between the sensitizing dose and the antibody response. Four strong immediate- type (type I) allergens were screened: TMA, PA, MDI, and TDI. The skin sensitizer DNCB was used as a negative control. Guinea pigs were sensitized on day 1 and different groups were sensitized with a range of concentrations. Sensitization was assessed on day 19 by serological analysis measuring the presence of antigen-specific antibodies in the serum of treated animals. All test compounds but DNCB induced high serum levels of specific antibodies. There was a sensitization dose-related increase in specific antibody levels for TMA, PA, and MDI (for TDI, only one sensitization dose was used). For MDI, a no-effect level for the presence of antibodies was found at a sensitization dose of 0.01%. No-effect levels were not found for TMA and PA, as the lowest tested doses of 0.003% and 0.03%, respectively, were still effective. For TDI, a threshold was not found either at the only sensitization concentration of 0.1%.

Dose-dependent increases in prevalence and concentration of specific IgE and IgG were found in BN rats treated dermally with different doses of dry powder TMA under occlusion on days 0,7, 14 and 21. Specific antibodies were detectable 2 weeks after the first application and peaked between 3 and 4 weeks; no threshold was found (Zhang et al., 2002). Vanoirbeek et al. (2003) found a significant increase in total serum IgE following treatment with 20% of a PCT derivative but not at 10%. Klink and Meade (2003) found a dose- dependent elevation in total serum IgE upon dermal exposure to 5, 15, or 25% AMT, 5 days a week for 68 days. Significantly elevated concentrations were obtained at 25% at 26 days after initiation, and at all concentrations at 40 days after initiation.

Skin sensitization measured by pulmonary reactions after inhalation or intratracheal challenge. In the aforementioned study by Blaikie et al. (1995; see preceding subsection), the relationship between the various intradermal sen\sitization doses and pulmonary response (endpoint) was also investigated upon inhalation challenge (on day 22) at a fixed concentration of each chemical. No clear dose- response relationship between sensitizing dose and pulmonary reactions was noticed for TMA, PA, and MDI-that is, there was no substantial increase in the incidence or severity of these pulmonary responses with the increase in the sensitizing dose. For MDI, a no- effect level for the occurrence of pulmonary reactions was found at a sensitization concentration of 0.01%, which was in agreement with the detected no-effect level for the presence of specific antibodies. For TMA and PA again no thresholds were found; for TDI, only one sensitization dose was used.

Arakawa et al. (1995) used intradermal induction of sensitization to two concentrations of TMA to assess airway responses (endpoint). Guinea pigs intradermally sensitized with 300 g TMA and intratracheally challenged with 50 g of TMA (5 mg/ml) responded with both airflow obstruction and airway plasma exudation, while the low sensitization dose (3 g) failed to induce airway responses to TMA. Antigen-specific or total antibody levels to TMA were not measured. Using the same sensitization and challenge method, guinea pigs responded with both immediate effects on lung resistance and Evans blue dye extravasation, which was dependent on the sensititization dose of HHPA (0.05, 0.5, or 5%). A no-effect level was not found (Zhang et al., 1998).

Pauluhn (2003) showed that BN rats treated topically with 5 or 25% TMA showed a dose-related increase in the number of animals showing changes in breathing patterns and histopathological pulmonary changes upon a challenge with approximately 25 mg/m^sup 3^ TMA, whereas a dose of 1% was ineffective. In the aforementioned study of Klink and Meade (2003), it was shown that dermal exposure of BALB/c mice to concentrations of 5, 10, 15, 20, or 25% AMT, 7 days a week for 35 days, resulted in concentration-related pulmonary histopathological effects (alveolitis) following intratracheal challenge with AMT; a threshold was not found.

2. High-Molecular-Weight Compounds

a. Epidemiological or Clinical Studies. No studies have been found.

b. Animal Studies (Table 1 ). Arakawa et al. (1995) investigated the effect of two intradermal sensitization concentrations of Dermatophagoidesfarinae (Df: mite) allergens on airway responses (endpoint). In animals sensitized with 300 g of Df mite allergen, intratracheal challenge produced a significant airway plasma exudation in the airways of all animals, while no significant exudation was produced after sensitization with the lower dose of Df (30 g) and similar challenge. Levels of specific antibodies to Df were not measured.

Different types of vehicle may influence the results. Although guinea pigs intradermally sensitized with various doses of ovalbumin (OA) in either saline or corn oil developed dosedependent levels of antigen-specific IgG1 antibodies, an intratracheal challenge with OA of animals sensitized with OA in corn oil induced an inversely dose- dependent airflow obstruction and airway plasma exudation. In contrast, animals sensitized with OA in saline had dose-dependent airway responses to OA (Arakawa et al., 1995).

B. Dose-Response Relationships and Thresholds for Skin Elicitation

1. Low-Molecular-Weight Chemicals

a. Epidemiological or Clinical Studies (Table 2). Concentration- dependent allergic contact dermatitis was reported for Kathon biocide (Pasche and Hunziker, 1989), DNCB (Friedmann et al., 1983), hydroxycitronellal (Ford et al., 1988), cinnamic aldehyde (Johansen et al., 1996), formaldehyde (Flyvholm et al., 1997), MBT and MBTS (Emmet et al., 1994), PPD (McFadden et al., 1998), potassium dichromate (Eun and Marks, 1989), chloroatranol (Johansen et al., 2003b), hydroxyisohexyl 3-cyclohexene carboxaldehyde (Lyral; Johansen et al., 2003a), and nickel (Allenby and Goodwin, 1983; Eun and Marks, 1989; Menn and Calvin, 1993; Uter et al., 1995).

With respect to the shape of the dose-response curves, information was limited, most probably due to the small numbers of concentrations tested. For DNCB, linear and almost linear log-dose- response curves were observed in the proportion of DNCB-sensitized subjects showing clinically detectable reactivity to an occluded patch test. Also, a proportionate increase in the degree of reactivity was detected (Friedmann et al.. 1983).

For Kathon CG, the threshold for occluded challenge was in the range of 25 ppm (0.0025%; Weaver et al., 1985), 10-20 ppm (0.001- 0.002%; Cardin et al., 1986), or 7 ppm (0.0007%; Pasche and Hunziker, 1989). For cinnamic aldehyde, the minimum effect level was 0.02% (occluded patch testing) and 0.1 % (repeated open application; Johansen et al., 1996). The highest concentration of MET at which no reaction was observed was 0.0032% (1.45 g/cm^sup 2^); the lowest concentration producing a positive reaction was 0.01% (4.5 g/cm^sup 2^). At concentrations of 0.316% or 1% MBT, all patients reacted. For MBTS, questionable results were obtained at a concentration of 0.0316%, a positive reaction at 0.1 %, and at 1 %, the highest level tested, at least 5 out of 12 patients reacted (Emmet et al., 1994). In studies in nickel-sensitive patients, no-effect levels varied from 100 ppm (0.01%; Menn and Calvin, 1993) to 5 ppm (0.0005%; Uter et al., 1995), although it was stated that in the latter study, only a few weak reactions were observed at 10 ppm (0.001 %) and 50 ppm (0.005%), and moderate reactions at 100 ppm (0.01%). In another dose- response study, the minimum elicitation level was approximately 112 ppm nickel (0.05% nickel sulfate) for the majority of nickel- sensitive individuals (14 out of 25; Allenby and Goodwin, 1983). Overall, the minimum threshold concentration for nonoccluded challenge in nickel-sensitized subjects has been established at 100 ppm (0.01%; Emmet et al., 1988; Uter et al., 1995; Allenby and Goodwin, 1983; Menn, 1994). There was a marked interindividual variability with regard to provocation threshold concentrations among the sensitized patients (Emmet et al., 1994; McFadden et al., 1998). It was even noted that the variability in the provocation threshold was 100-fold among 12 tested sensitive individuals (Emmet et al., 1994).

TABLE 2

Summary of methods to determine dermal elicitation dose-response relationships

No-effect levels were not obtained for hydroxycitronellal; when 28 sensitized subjects were exposed to 1%, 3 subjects were reacting (Ford et al., 1988). One out of 20 formaldehydesensitive patients reacted at the lowest concentration of formaldehyde tested, that is, 250 ppm (0.025%) in the occluded patch test (Flyvholm et al., 1997). Also, for potassium dichromate a no-effect level could not be found (Eun and Marks, 1989).

In the study of Flyvholm et al. (1997), both occluded and nonoccluded patch testing were used to establish threshold levels. Although a dose-response relationship was observed using occluded patch tests, no definite positive reactions were seen in the nonoccluded patch test, indicating a different skin penetrating ability. Menn (1994) also mentioned a large difference between occluded and nonoccluded exposure; that is, concentrations to induce nickel challenge reactions were about 15-150 times lower during occluded challenge than during nonoccluded exposure. Skin penetrating ability is even more increased in damaged skin. The elicitation threshold concentration of nickel sulphate appeared 100- 1000 times lower than the elicitation threshold concentration for normal skin; that is, less than 1 ppm (0.0001%) was sufficient to elicit reactions (Allenby and Basketter, 1993). Highly sensitized individuals might react to concentrations as low as 0.5 ppm (0.0075 g/cm^sup 2^) nickel when exposed on inflamed skin under occlusion (Basketter and Allenby, 1990). Reactions were also observed in a few nickelsensitive patients at the same concentration under occlusion after inducing a mild degree of skin irritation (Basketter and Allenby, 1990).

b. Animal Studies (Table 2).

Delayed-type skin reactions after dermal sensitization and challenge. Dose-response relationships for dermal elicitation reactions were investigated in sensitized guinea pigs using different tests. With DNCB, using the open epicutaneous test or the Buehler occlusive patch test, a gradation in response (percentage of animals showing a positive reaction) was clearly observed. Almost all animals responded at the highest DNCB challenge concentration and the observed response decreased with decreasing doses. No challenge reactions were observed at the lowest challenge concentration of 0.00001% DNCB (Buehler test; 100 animals) or 0.01% DNCB (open epicutaneous test; 20 animals). The maximization test was used to evaluate doseresponse relationships for PPD. At the highest challenge dose of 6%, 93% of the 250 guinea pigs showed a positive reaction; at the lowest dose of 0.01%, only 15% of 200 guinea pigs were positive. Similar dose-response curves were obtained for both DNCB and PPD. The higher doses produced a linear relationship, but at lower doses the curves flattened out (Bronaugh et al., 1994).

The previously mentioned study by Chan et al. (1983) using Kathon CG in the Buehler test also indicated a dose-response relationship for elicitation reactions. Results showed that the sensitizing potential of the biocide (measured as the incidence of erythema) was depending on both the induction and challenge concentration. The number of animals responding increased with increasing challenge concentrations when the induction concentration was kept equal. A sigmoid dose-response curve was obtained. A no-response concentration was observed for several combinations of induction and challenge concentrations. The EC^sub 50^ for challenge reactions at an induction concentration of 1000 ppm (0.1 %) was calculated to be 429 ppm (0.0429%).

Another \previously mentioned study in guinea pigs in which both induction and challenge application concentrations were varied was carried out by Nakamura et al. (1994). The dermal response, measured as the formation of erythema and/or edema, was found to be related to the sensitization dose (intradermal injection) as well as the challenge dose (topical application) for almost all of the 15 chemicals tested-that is, the higher the doses, the higher the number of responders.

In a third previously mentioned study, MDBGN was tested in both the GPMT and the CCET, a test method using multiple topical applications. Various challenge doses were used. Dose-related elicitation reactions (monotone curve) and threshold levels could be detected for MDBGN in the CCET; a few positive reactions were seen in the GPMT at the highest challenge concentration only (Wahlkvist et al., 1999a).

Mean ear thickness measured in sensitized mice increased dose- dependently using both DNCB and squaric acid dibutyl ester. Shallow linear increases with thresholds were found, and eventual plateaus at increasing doses. However, when also varying the sensitization concentrations, the minimum dose was not static but rather reflected a "sliding scale," that is, as the sensitization dose was increased, the concentration required to elicit a challenge response was decreased. Correspondingly, as the challenge dose was increased, the dose required for sensitization was lessened (Scott et al., 2002).

c. Immediate-type skin reactions after dermal sensitization and challenge. In a study using BN rats, animals were sensitized by an intradermal injection of TMA at a fixed concentration. Three weeks later, Evans blue dye was given, and dermal reactions (extravasation of dye in skin) were measured 30 min after intradermal challenge with various concentrations of TMA. A dose-dependent increase of Evans blue dye extravasation was observed. In addition, skin histology revealed a significant and dose-dependent increase in eosinophils after repeated TMA injections in similarly sensitized BN rats (Andius et al., 1996).

2. High-Molecular-Weight Compounds

a. Epidemiological or Clinical Studies (Table 2). In a review on workers involved in either manual or automated processing of seafood it was found that occupational dermal exposure as a result of unprotected handling of seafood and its by-products resulted in a prevalence of occupational protein contact dermatitis from 3 to 11%. Cross reactivity between various species within a major seafood group was also found. Limited evidence from dose-response relations indicated that the development of symptoms was related to duration or intensity of exposure. It was also found that disruption of the intact skin barrier seems to be an important risk factor for occupational protein contact dermatitis. Workers in the seafood processing industry are also exposed by inhalation, which may have contributed to the adverse health effects (see also at respiratory allergy; Jeebhay et al., 2001).

b. Animal Studies. No studies have been found.

III. RESPIRATORY-TRACT ALLERGY

Occupational exposure to various LMW or HMW compounds can cause sensitization of the respiratory tract, resulting in allergic pulmonary hypersensitivity reactions upon a subsequent encounter with the same compound. It is well established that respiratory allergy to proteins in humans is associated with, and mediated by, specific IgE antibodies. There is less certainty with respect to a similar requirement for IgE antibodies in the development of respiratory allergy to LMW chemicals, not least because specific IgE antibodies could not be demonstrated in a large number of symptomatic individuals sensitised to certain diisocyanates (Karol etal., 1978, 1979a, 1979b, 1980, 1994; Butcher et al., 1980, 1983; Karol and Alarie, 1980; White etal., 1980; Karol, 1980, 1981; Baur and Fruhmann, 1981; Danks etal., 1981, 1983; Baur etal., 1984, 1994; Butcher and Salvaggio, 1986; Carder et al., 1989; Kimber and Dearman, 1997; Beckett, 2000), as well as in a number of persons with acid anhydrides-induced asthma (Venables, 1989). Agents in the workplace that are able to induce allergen-specific airway reactions are HMW compounds such as flour allergens, grain allergens, feed dust, animal dander/urinary proteins, and a number of LMW chemicals such as diisocyanates, acid anhydrides, and reactive dyes (Kimber et al., 1996).

There are two risk phrases for respiratory effects-that is, respiratory sensitization (R42) and respiratory irritation (R37)- but for both types of effects, validated tests are lacking. Although a number of animal test protocols have been published to detect respiratory allergy (see for reviews Briatico-Vangosa et al., 1994; Pauluhn et al., 1999; Pauluhn and Mohr, 2005), none of these are widely applied or fully accepted, most probably because no large effort has yet been made for validation.

With respect to substances inducing occupational asthma, the current European Commission (EC) labeling criteria for dangerous substances (CEC, 1967; amended several times and adapted to technical progress for the 29th time recently), which are used to classify substances among others on their potential to induce respiratory allergy, include all chemicals that can induce asthma(- like) attacks. As stated: "The condition will have the clinical character of an allergic reaction. However, immunological mechanisms do not have to be demonstrated." In this guideline, it is further emphasized that "Substances that elicit symptoms of asthma by irritation only in people with bronchial hyperreactivity should not be assigned R42." In addition, the decision on classification needs to take into account "the size of the population exposed" and the "extent of exposure." Thus, a high-production-volume chemical found in large quantities in many workplaces throughout the world might not warrant the R42 phrase if only a few cases of asthma with its use have been reported over the years. In contrast, three cases of asthma among a workforce of 20 in contact with a specific chemical might well indicate the need for classification as a respiratory sensitizer (Evans, 1997).

A. Dose-Response Relationships and Thresholds for Respiratory- Tract Sensitization

1. Low-Molecular-Weight Chemicals

a. Epidemiological or Clinical Studies (Table 3). A few studies were found that reported dose-response relationships for respiratory- tract sensitization to LMW chemicals in workers. In a prospective study performed in TDI-exposed workers, it was shown that accidental exposure to high concentrations of TDI resulted in IgE antibody production. In contrast, exposure to low concentrations of TDI (at or below 0.02 ppm) for up to 3 years did not result in any cases of TDI sensitivity or in production of TDI-specific antibodies (Karol, 1981). In another prospective study of workers exposed to organic acid anhydrides (OAAs), 163 previously unexposed persons were exposed to epoxy resins containing HHPA, MHHPA, and MTHPA. The mean observation time was 32 (1-105 months). The levels of OAAs in air and of specific IgE and IgG in serum were recurrently monitored. The mean combined OAA exposure was 15.4 g/m^sup 3^ (range < 1-189 g/ m^sup 3^). An exposure-response relationship was demonstrated by an increasing risk of sensitization with increasing exposure. Specific IgE was demonstrated by 21 (13%) subjects with a mean induction time of 8.8 (1-35 months). The incidence of sensitization was 4.1 cases/ 1000 months at risk. The relative risk for atopies was 5.4 (1.9- 15.3; 95% confidence interval; Welinder et al., 2001). In a cross- sectional study, also with HHPA and MHHPA, there was an increasing risk of sensitization with increasing exposure. Specific IgE was found in 13% of the workers exposed to concentrations <10 g/m^sup 3^ HHPA, in 26% exposed to concentrations between 10 and 50 g/m^sup 3^ HHPA, and in 21% exposed to concentrations higher than 50 g/m^sup 3^ HHPA; for MHHPA, specific IgE levels were about similar, 15, 26, and 17%, respectively. Specific IgG levels were 2, 22, and 41%, and 4, 26, and 38% for these exposure categories of HHPA and MHHPA, respectively. Atopy did not significantly increase these risks (Nielsen et al., 2001). Drexler et al. (2000) also reported in a study of three plants using MTHPA that sensitization to OAAs increased with increasing exposure.

b. Animal Studies (Table 3).

Respiratory sensitisation measured by antibody levels. Guinea pigs were exposed to various TDI concentrations (0.12-10 ppm) for 3 hours a day on 5 consecutive days. TDI-specific antibodies were measured from day 22 onward. No antibodies were detected in animals exposed to 0.12 ppm TDI. Within the range of 0.12-0.96 ppm, a linear relationship was observed between log-concentration of TDI and the antibody response as well as the percentage of animals producing antibody to TDI. Half of the animals exposed to 0.36 ppm TDI or more showed TDI-specific antibodies in their sera. At exposure levels of 0.96 ppm or higher, all animals produced an antibody response (Karol, 1983). In a similar study (Karol etal, 1980), concentrations of 0.25 ppm also resulted in the production of antibody to TDI. No production of TDI-specific antibody was noted after exposure to a low TDI concentration of 0.02 ppm. Animals were exposed for 6 h/ day, 5 days/week, for a total of 70 days (total exposure: 8.67 ppm- h, compared to a total exposure of 3.75 ppm-h [0.25 ppm for 3 h/day for 5 days] that induced production of antibodies during short-term exposure). It was concluded that the exposure concentration (in combination with the duration of exposure) was important for establishment of antibody response, rather than total exposure (Karol, 1983). A single 15-min exposure of guinea pigs to various concentrations of polymeric MDI (5, 12, 32, 108, or 835 mg/m^sup 3^) resulted in a concentration-related increase in specific antibodies 3 weeks af\ter induction; a threshold was not reported (Pauluhn et al., 2000). Exposure of guinea pigs to various concentrations of two trimersofHDI(HDI-isocyanurate3.0,15.9,49.4, or 261 mg/m^sup 3^, or HDI-biuret 2.7, 9.5, 49.4, or 142 mg/m^sup 3^) 3 h/day for 5 days revealed a concentration-related increase in the number of animals showing increased antibody levels. The no-observed-effect levels (NOELs) were about 3 mg/m^sup 3^ with borderline effects at concentrations of 10-15 mg/m^sup 3^ (Pauluhn et al., 2002a). However, upon a challenge with a concentration of about 85 mg/m^sup 3^ with the free hapten, or upon challenge with a respective GPSA conjugate 1 week later, no specific functional or morphological pulmonary responses were seen in contrast to monomeric HDI (Pauluhn et al., 2002a).

TABLE 3

Summary of methods to determine respiratory sensitization dose- response relationships

In BN rats, sensitization-concentration-response relationships were observed using TMA. A daily 1-h exposure to 5 mg/m^sup 3^ for 1, 3, or 5 days resulted in an exposure-duration-related increased number of animals responding with specific IgG and IgE antibody levels. There was also some suggestion of a dose-response effect with an increased number of rats displaying induction of specific IgE antibody, namely, 1/5 rats after daily 1-h exposures, 2/5 rats after 3-h exposures, and 3/5 rats after 5-h exposures (Warbrick et al., 2002).

Respiratory sensitization measured by pulmonary reactions after inhalation challenge. Pulmonary reactions upon challenge with 1% TDI- guinea pig serum albumin aerosol, measured as an increase in respiratory rate, were not detected in animals sensitized to 0.12 ppm TDI but were present in guinea pigs exposed to TDI concentrations of 0.36 ppm and higher (Karol, 19830. Exposure of BN rats to either 25 or 120 mg/m^sup 3^ TMA 3 h/day for 5 days revealed a concentration-related increase in the respiratory response, and in lung-associated lymph node weights following challenge at a concentration of 23 mg/m^sup 3^ TMA (Pauluhn et al., 2002b).

Respiratory sensitization measured by skin reactions after dermal challenge. Guinea pigs and mice were sensitized by inhalation to various concentrations of HMDI for 2 h/day on 3 consecutive days and tested for dermal reactions by means of a topical challenge at a fixed dose of HMDI, 7 days following the initial HMDI inhalation exposure. In both species, a concentration-response relationship was observed between the inhalation sensitization concentration and both the severity of the dermal response and the number of animals responding. Guinea pigs developed skin reactions following exposure to concentrations ≥3 mg/m^sup 3^; exposure to 1.25 mg/m^sup 3^ did not result in respiratory sensitization. In mice, dermal reactions were noted upon exposure to ≥17 mg/m^sup 3^; no reactions occurred at sensitization concentrations ≤7 mg/ m^sup 3^. Specific antibody levels to HMDI were not measured (Stadler and Karol, 1984). Guinea pigs receiving a single intranasal application of TDI at various concentrations exhibited skin sensitivity in a dose-dependent manner upon a patch challenge with TDI. One of four responded to 0.6%, 2/4 responded to 1.2%, and 4/4 responded to 1.8% TDI. Specific antibody levels to TDI were not measured (Ebino et al., 2001).

2. High-Molecular-Weight Compounds

a. Epidemiological or Clinical Studies (Table 3). Dose-response relationships for respiratory-tract sensitization to HMW compounds were observed in several studies in workers. A cross-sectional epidemiological study was carried out to assess the relationship between a-amylase exposure and allergic sensitization. All workers were categorized into groups on the basis of their job histories and a-amylase exposure levels measured in personal dust samples. The prevalence of positive IgE tests to a-amylase tended to increase with intensity of exposure. After stratification for atopy, however, there was no clear exposure-response relationship. In contrast, prevalence ratios of a positive skin-prick test increased with increasing a-amylase exposure groups. Sensitization rate increased from 1.4% in the low exposed worker group (geometric mean [GM] of 0.7 ng/m^sup 3^), to 12.8% in the medium exposed worker group (GM 1.3 ng/m^sup 3^), and to 30.4% in the high exposed worker group (GM 18.1 ng/m^sup 3^). The investigators concluded that a strong and positive association was shown between a-amylase allergen exposure levels and a-amylase specific sensitization. Sensitization was already found at 0.25 ng a-amylase/m^sup 3^ air (Houba et al., 1996a). In another study in workers in bakeries and flour mills, 5% of the workers were sensitized to a-amylase. The prevalence increased with increasing exposure to a-amylase, that is, from 3.1% in the low exposed worker group (GM 0.7 ng/m^sup 3^), to 16.7% in the medium exposed worker group (GM 10.7 ng/m^sup 3^), and to 15.4% in the high exposed worker group (GM 46.7 ng/m^sup 3^; Nieuwenhuijsen et al., 1999).

Flour dust concentrations of 1-2.4 mg/m3 were found to be associated with a significantly elevated risk of sensitization to wheat antigens (Houba, 1996). Exposure-sensitization relationships for flour dust exposure and wheat aeroallergen exposure among workers in bakeries have been reported by Musk et al. ( 1989), Cullinan et al. ( 1994a), and Houba et al. ( 1998a, 1998b). It was also reported that sensitization risk will be negligible when exposure levels will be reduced to average exposure concentrations of 0.2 g/m^sup 3^ wheat allergen, or approximately 0.5 mg/m^sup 3^ inhalable dust during a work shift (Houba et al., 1998a).

In laboratory animal workers, the prevalence rate of sensitization to rat allergens was also clearly associated with exposure levels (Hollander et al., 1997a) but only in those workers with less than 4 years of working experience with laboratory animals (Hollander et al., 1997b). A positive correlation was found between the intensity and frequency of rat urinary allergen exposure and the frequency of positive skin-prick test results and specific sensitization (Cullinan etal., 1994b). Data from three cross- sectional studies in laboratory animal workers revealed similar results: The rat urinary allergen sensitization risk increased with increasing air allergen exposure. Atopic workers had a clearly elevated sensitization risk at low allergen exposure levels (Heederik et al., 1999a).

In workers in a sawmill, the prevalence of sensitization to saw dust and Trichoderma koningii was associated with exposure levels; that is, serum concentrations of specific IgG were significantly higher in the high exposure group when compared to the low exposure group (Halpin et al., 1994).

In latex-exposed persons, inhalation exposure to a concentration of ≥0.6 ng/m^sup 3^ latex resulted in sensitization of 18% of people exposed. Lower concentrations were not associated with IgE- mediated sensitization (Bauret al., 1998a).

In clinical studies it was found that the frequency of latex sensitization in spina bifida patients was strongly related to the number of surgeries undergone (Michael et al., 1996; Chenet al., 1997a, 1997b; Porri et al, 1997).

In young asthmatic children, a dose-response relationship was found between cat exposures, measured as either reported degree of cat exposure or cat allergen (FeI d 1) levels in dust samples, and sensitization as measured by the amount of IgE antibodies. High levels of cat allergen were considered to be >8 g/g dust. No such relationship was found between exposure and sensitization to dog (Can f 1) allergens (Lindfors et al., 1999). In another study, it was found that many children exposed to greater than 20 g FeI d l/g dust at home produced an IgGj and IgG4 antibody response to FeI d 1 without IgE antibody. This response was not associated with symptoms and, according to the authors, should be regarded as a form of immunological tolerance (Platts-Mills et al., 2001). These studies show that the dose-response relationship between cat allergen exposure and sensitization is bell-shaped (Murray et al., 2001) and it was suggested that high exposure to cat allergen can modify the Th2 response to suppress IgE production while maintaining or increasing IgG^sub 4^ and IgG^sub 1^ antibody production (Erwin et al., 2005).

On the basis of data from several studies, a significant correlation between cumulative exposure to house dust-mite allergen and the risk of sensitization was reported (Anonymous, 1993). The dose-response relationship appears to have a linear relationship (Murray et al., 2001). A threshold level was proposed: Exposure to more than 2 g Group 1 mite allergen/g dust should be regarded as a risk factor for the development of IgE antibody and asthma in susceptible children (Custovic and Chapman, 1998; Custovic et al., 1998). A linear dose-response relationship between exposure and sensitisation was also reported for cockroach allergen (Murray et al., 2001). Also, the frequency and extent of exposure to Chironomidae insect allergens (Chi t 1-9) was associated with IgE- mediated sensitization (liebers et al., 1993).

b. Animal Studies (Table 3).

Respiratory sensitization measured by antibody levels. Guinea pigs were exposed by intratracheal administration to various concentrations of the enzyme protein Alcalase in a detergent base once a week for 10 weeks. Other groups of guinea pigs were exposed by inhalation (6 h/day, 4 days/week for 10 weeks) to 1 mg/m^sup 3^ of the aerosolized detergent base containing various concentrations of Alcalase protein. The antibody levels to Alcalase increased dose- dependently by both the inhalation and intratracheal routes of exposure. No-effect levels were not detected (Ritzetal., I993).

A concentration-related antigen-specific antibody response was observed in guinea pigs that had been exposed for 15 min/day during 5 consecutive days to various c\oncentrations of subtilisin up to 1.9 mg/m^sup 3^. Antigen-specific antibody levels were not increased further upon exposure to higher subtilisin concentrations (up to 15 mg/m^sup 3^). A separate, single 20-min exposure to a concentration of 1.9 mg/m^sup 3^ resulted in the same antibody response as that produced following 5 days of exposure to the same concentration. Long-term exposure to low levels of the enzyme (11 weeks to a concentration of 0.00068 mg/m^sup 3^ followed by 6 weeks to a concentration of 0.0051 mg/m^sup 3^; i.e., 1.12 mg/m^sup 3^ in total) resulted in 36% (9 of 25) of the animals producing significant levels of specific IgG or IgM antibodies (Hillebrand et al., 1987).

BN rats were exposed 30 min/wk for 6 weeks to respirable OA aerosols at concentrations of <1, 3.3, 15.4, or 64.1 mg/m^sup 3^. OAspecific IgE, IgG, and IgA were measured throughout the study period. Rats were sacrificed 1 day after the last exposure. Serum concentrations of OA-specific antibodies increased with both exposure concentration and number of exposures. The number of rats with measurable titers also increased with both dose and time. A no- effect level for antibody production was not present (Siegel et al., 2000).

Respiratory sensitization measured by pulmonary reactions after inhalation challenge. Guinea pigs exposed to concentrations ranging from 0.15 to 15 mg subtilisin/m^sup 3^ for 5 consecutive days produced pulmonary reactivity upon an inhalation challenge at a concentration of 1.9 mg/m^sup 3^ for 20 min on day 10. Animals exposed to lower concentrations up to 0.041 mg/m^sup 3^ failed to demonstrate pulmonary reactivity upon a similar challenge. The single 20-min exposure to 1.9 mg/m^sup 3^ followed by a challenge at 1.9 mg/m^sup 3^ for 20 min on day 7 also resulted in pulmonary reactivity. In contrast, the 17-week exposure to a total of 1.12 mg/ m^sup 3^ did not result in pulmonary reactivity upon challenge, although the total cumulative exposure administered over a short period of time regularly induced pulmonary reactivity (Thorne et al., 1986).

B. Dose-Response Relationships and Thresholds for Elicitation Reactions in the Respiratory Tract

1. Low-Molecular-Weight Chemicals

a. Epidemiological or Clinical Studies (Table 4). Studies were carried out to assess the concentration-response association between exposure to acid anhydrides and airway symptoms. When the concentration of TCPA in air was between 0.21 and 0.39 mg/m^sup 3^, the prevalence of work-related respiratory symptoms was 27-39%. After installation of ventilation the TCPA-concentration dropped to 0.1 mg/m^sup 3^ and at the same time symptoms diminished considerably (Liss et al., 1993). For TMA, concentrations in air were reduced from 0.82-2.1 mg/m^sup 3^ to 0.01-0.03 mg/m^sup 3^ after the installation of ventilation. Subsequently, the number of workers with specific IgE antibodies and symptoms decreased (Bernstein et al., 1983).

In two condenser plants using epoxy resins containing MTHPA, 111 workers underwent a questionnaire survey and serology investigations. In plant A, air concentrations of MTHPA were higher than in plant B (geometric mean approximately 25-64 and 4.9-5.5 g/ m^sup 3^, respectively). In total, 24 workers (65%) in plant A and 38 workers (66%) in plant B had MTHPA-specific IgE. In sensitized workers in plant A, eyes, nose, and pharynx symptoms were observed at a higher incidence when compared to sensitized workers in plant B. In addition, work-related symptoms were often observed in 73% of the 26 symptomatic workers in plant A, and in only 15% of the 20 symptomatic workers in plant B. In plant B the minimum level that was associated with work-related symptoms was 15-22 g/m^sup 3^, indicating that levels above 15 g/m^sup 3^ should be avoided (Yokota et al., 1999).

TABLE 4

Summary of methods to determine airway elicitation dose-response relationships

In a previously mentioned cross-sectional study, the incidences of work-related symptoms (eyes, nasal, and lower airways) were concentration-related increased among HHPA and MHHPA exposed workers. An exposure level of <10 g/m^sup 3^ was not associated with a significant increase in symptoms (but did not prevent sensitization as measured by an increased incidence of workers with increased levels of specific IgE and IgG; Nielsen et al., 2001). Subjects who were IgE-sensitized against HHPA and who reported work- related nasal symptoms had a concentration-related increase in nasal symptoms and a concentration-related decrease in peak flow after increasing nasal challenge concentrations (1:100, 1:10 diluted or undiluted); a threshold was not found (Nielsen et al., 1994).

A case-control study was designed to compare isocyanate concentrations measured at 20 companies with 56 isocyanateinduced occupational asthma cases (total of 2256 workers) with isocyanate levels at 203 companies without occupational asthma cases (total of 4052 workers). Exposure was determined based on the highest level identified. Companies were categorized into two groups: always <0.005 ppm and ever ≥0.005 ppm. Although the small number of case companies limited the statistical power, cases were significantly more likely to be working at companies in the higher exposure category (50% vs. 25%; p < .05, OR = 3.1). The estimated incidence of occupational asthma in the 4-year study period was greater in high than in low exposure companies, 2.7% and 2.2%, respectively (Tarlo et al., 1997). It was also observed that a chronic decline in airflow occurs on spirometry in groups of workers with isocyanate exposure, in relation to the intensity of the exposure (Diem et al., 1982). In a laboratory challenge test in sensitized individuals, acute airflow limitation worsened with increased isocyanate exposure (Malo, 1990).

In a toluene diisocyanate (TDI) production unit, the annual incidence of asthma induced by TDI declined from 1.8% before 1980 to 0.7%

Source: Critical Reviews in Toxicology

More News in this Category