Strong irritants masquerading as skin allergens: the case of benzalkonium chloride

Authors


Dr David A Basketter
Safety and Environmental Assurance Centre
Unilever Colworth Laboratory
Sharnbrook
Bedford
UK
Tel: +44 01234264776
Fax: +44 01234264744
e-mail: david.basketter@unilever.com

Abstract

Chemicals may possess a number of hazards to human health including the ability to cause skin irritation and contact allergy. Identification and characterization of these properties should fall within predictive toxicology, but information derived from human exposure, including clinical experience, is also of importance. In this context, it is of interest to review the case of benzalkonium chloride, a cationic surfactant. This chemical is a well-known skin irritant, but on occasions it has also been reported to have allergenic properties, typically on the basis of positive diagnostic patch test data. Because the accumulated knowledge concerning the properties of a chemical is employed as the basis for its regulatory classification (e.g. in Europe), as well as for informing the clinical community with respect to the diagnosis of irritant versus allergic contact dermatitis (ACD), it is important to distinguish properly which chemicals are simply irritants from those which are both irritant and allergenic on skin. A review of the information on benzalkonium chloride confirms that it is a significant skin irritant. However, both predictive test results and clinical data lead to the conclusion that benzalkonium chloride is, at most, an extremely rare allergen, except perhaps in the eye, but with many supposed cases of ACD being likely to arise from the misinterpretation of patch test data. As a consequence, this substance should not normally be regarded as, or classified as, a significant skin sensitizer.

Benzalkonium chloride is a quaternary ammonium cationic surfactant, widely used in cosmetics, skin disinfectants and ophthalmic preparations and various other products (1, 2). It is generally considered to be a significant skin irritant and as such is widely used as a model irritant in the investigation of skin irritation, typically at a concentration of 0.5% aq. or above (3). However, on occasions, benzalkonium chloride has been reported as a cause of allergic contact dermatitis (ACD) (1, 4) and has been considered to produce positive responses in some sensitization assays (5). However, irritant and ACD have a generally very similar appearance in skin, both clinically and experimentally. This review attempts to establish the skin irritancy and sensitization potential of benzalkonium chloride in predictive and experimental testing procedures, as well as by reference to reported case studies, to elucidate the true characteristics of this chemical. In our opinion, the prominence of the irritancy of benzalkonium chloride suggests that at least a proportion of the reported evidence of allergy are probably false-positive irritant reactions. This is of importance both for ensuring correct clinical interpretation of diagnostic patch testing and for substance classification according to European regulations (6). In this latter case, the relatively limited evidence of an association between benzalkonium chloride and allergy has led one regulatory group to a premature conclusion (7).

Benzalkonium Chloride: Evidence for Irritancy

Experimental induction of irritant contact dermatitis is widely performed both to provide positive controls for diagnostic patch testing and as a material for histological and biochemical studies of irritant contact dermatitis (3, 8, 9), and benzalkonium chloride is generally recommended for use in such studies at a concentration of 0.5% aq. Willis et al. (3) investigated 7 different chemicals, including benzalkonium chloride, to determine the optimum irritant concentrations in order to maximize their routine experimental procedures. They tested benzalkonium chloride at 0.5 and 1% aq. and found considerable variability in the reactions to benzalkonium chloride, with a significant proportion resulting in blistering at 1% aq. thus recommending the use of this chemical at 0.5% for diagnostic patch testing. This wide interindividual reactivity to benzalkonium chloride is in accord with our experience (data not shown). Benzalkonium chloride has also been used by Berardesca and colleagues (10) in the investigation of the irritancy of surfactants using the plastic occlusion stress test. Benzalkonium chloride 1% aq. was also used to investigate the influence of pre-existing skin disease on the irritant response to surfactants (3). In fact, benzalkonium chloride has been shown to cause similar modalities of skin effects to those caused by sodium lauryl sulphate (SLS), a 1% aq. solution of benzalkonium chloride being shown to alter the permeability of the skin and to cause similar damage to SLS. This was further supported by Bjornberg (11) who showed that SLS and benzalkonium chloride had similar irritant effects when visually assessed (although occasionally benzalkonium chloride can cause very strong irritant reactions of a degree not seen with SLS). However, at a more subtle level, the mechanisms of skin damage have been shown to differ between the 2 materials (12).

Established predictive irritancy tests are widely used to classify the irritant potential of materials and include tests such as the Draize test, human 4-h patch test and murine ear-swelling assay. When tested under these experimental conditions, benzalkonium chloride is widely reported to be a moderate irritant (8, 13). These tests involve application of the materials to an area of skin either on a human or an animal and recording the level of visible irritation elicited which could either be erythema, as in the human 4-h patch test, or thickness of the ear in the murine ear-swelling assay. In the human 4-h patch test, 19 of 56 reacted to application of 7.5% aq. benzalkonium chloride compared to 32 of 56 reacting to 20% aq. SLS (14).

Benzalkonium chloride and irritancy in reported case studies

Generally, clinical irritant reactions to benzalkonium chloride are not widely reported. However, Ling and Highet (15) reported 7 case studies of an irritant reaction to an antiseptic bath emollient containing 6% benzalkonium chloride. The reactions were determined to be irritant rather than allergic because of the lack of itch, prominence of desquamation and the successful subsequent use in most of the patients at a lower concentration.

In addition to the above, it is the general impression from standard texts (1, 16, 17) that benzalkonium chloride is widely recognized to be irritant in nature.

Benzalkonium Chloride: Evidence for Allergy

Benzalkonium chloride in sensitization assays

As well as producing an irritant response in predictive testing, benzalkonium chloride has also been shown to induce responses that would classify it as an allergen in sensitization assays. There are a variety of predictive test procedures used for determining the sensitization potential of chemicals, and benzalkonium chloride has been reported to elicit a positive response in certain of these, although not by all authors. It is not uncommon for strong irritants to produce false-positive reactions in sensitization assays, e.g. SLS, a widely used and well-known irritant, has been reported to give false-positive results in the local lymph node assay (LLNA) (13, 18, 19). In addition, searching questions have been asked concerning the interpretation of results from guinea-pig predictive tests, notably the maximization test (20, 21). Consequently, it is questionable whether these results truly demonstrate that it is a skin sensitizer.

In many cases, benzalkonium chloride elicits few positive responses indicating that it may have sensitization potential but not offering clear evidence that this is the case. In a guinea-pig optimization test reported by Maurer (5), arguably the most sensitive (but not necessarily specific) method ever developed, benzalkonium chloride elicited 11 of 20 positive responses by intradermal challenge at a concentration of 0.1% and 4 of 19 positive responses via epicutaneous challenge at a concentration of 10%. Thus, although apparent positive responses were elicited suggesting a possible sensitization potential, those elicited by epicutaneous application, the correct route of administration, were still too low to classify this material as a skin sensitizer according to the European criteria for the classification of substances (6). Goh (22) reported benzalkonium chloride as being a moderate sensitizer using a modified Beuhler technique, with 20% of guinea pigs showing a response at the challenge phase. Woolhiser et al. (13) investigated the sensitization and irritancy potential of potent chemical sensitizers and irritating agents including benzalkonium chloride at 0.5, 1, 2, 3 and 5% in both a murine LLNA and a murine ear-swelling irritancy assay. In the murine ear-swelling assay, benzalkonium chloride elicited an irritation dose–response that peaked at 5%, while at 2.0% lymph node proliferation in these animals peaked, which the authors suggested could indicate the possible sensitization potential of benzalkonium chloride. Taken together, these tests thus suggest a possible sensitization potential but lack any absolute proof that this chemical is an allergen. In fact, when examining benzalkonium chloride performance in the LLNA test, the results are conclusive, Basketter et al. (19) reported benzalkonium chloride as giving a negative response. It should be noted, however, that a non-standard version of this assay gave contrasting results, showing a marginally positive result (23).

Many assays are being developed to investigate the nature of a chemical and are designed to discriminate between allergic and irritant responses. Recent work has shown that it is possible to discriminate between allergens and irritants by investigating the expression of interleukin-1α (IL-1α) and IL-8. Reconstructed human epidermis was used as an in vitro model to discriminate between 5 skin sensitizers and 3 skin irritants that included benzalkonium chloride (24). From these results, it was proposed that chemicals with a sensitization potential elicit an increase in IL-8 that is higher than that in extracellular IL-1α elicited, whereas skin irritants are chemicals which elicit an increase in extracellular IL-1α which is equal to or greater than the that in IL-8. Benzalkonium chloride fitted this profile and under this particular test regime/end point would be classified as an irritant. This work reflects that completed by others (25, 26) who showed that IL-1α was upregulated by both allergens and irritants.

Gerberick et al. (23) investigated the use of B220+ marker to discriminate between allergens and irritants in the LLNA. Benzalkonium chloride is often reported to produce proliferation in the LLNA, thus resulting in possible classification as an allergen. However, when using the B220+ marker to discriminate further between irritants and allergens, benzalkonium chloride was classified as an irritant in 12 of 14 separate experiments.

The increased expression of IL-1β by epidermal Langerhans' cells (LCs) following exposure of mice to contact allergens has also recently been investigated as a strategy to identify skin sensitizers. It has been demonstrated that culture of human blood-derived LC-like cells with selected allergens stimulates the expression of IL-1β, while under the same conditions of exposure, application of skin irritants such as SLS and benzalkonium chloride has been shown to have no effect on IL-1β expression (27–29).

Thus, this combined in vivo and in vitro evidence clearly identifies benzalkonium chloride as an irritant. Indeed, under the EU scheme (6), higher concentrations are classified as corrosive to skin.

Benzalkonium chloride and allergy in reported case studies

Sensitization to benzalkonium chloride has been reported as a result of occupational exposure (30) in physicians, nurses, dentists and veterinarians, and allergic reactions to a variety of different products containing this chemical have been reported, including detergents/antiseptics (30–34), antifungal creams (35, 36), ophthalmic preparations (4, 37–39), plaster of Paris (40, 41), as a denaturant of ethanol (42) and in toothpastes (43).

Contact sensitivity to benzalkonium chloride appears to be reported most commonly for ophthalmic preparations. Herbst and Maibach (39) reviewed the literature to identify an ophthalmic tray for testing contact allergy suspected from ophthalmic preparations and reported a total of 16 cases of positive patch tests to benzalkonium chloride taken from previously reported cases studies (6 different papers) at concentrations ranging from 0.005 to 0.13%. As such, they recommend routinely using benzalkonium chloride in patch testing for ophthalmic product-related allergy at 0.1%. Both Klein et al. (38) and Cox (37) report cases of ACD associated with benzalkonium chloride-containing eye drops.

When used for diagnostic patch testing, benzalkonium chloride is generally used at 0.1% aq. However, this concentration is often reported to elicit irritant responses and, when Brasch et al. (44) tested 1775 patients with 0.1% benzalkonium chloride, they found that this concentration yielded more irritant reactions than allergic responses. Therefore, it has been suggested that true allergic reactions can be obtained only by testing with a concentration of 0.01% aq. (1, 31, 44), and it is possible that many of the reported allergic responses to benzalkonium chloride are in fact irritant responses that have been misinterpreted. It is accepted that patch test reactions are often misinterpreted as being allergic, particularly when a weak reaction is elicited. Ale and Maibach (45) recommend repeating the patch test to rule out false-positive reactions and also to perform a use test, as also suggested by others (38). Ale and Maibach detailed many different measures which help to determine the clinical relevance of patch test reactions. These measures should help to confirm the true nature of the reaction to the material in question. When examining the case reports detailed in this review, it is clear that many of the documented allergic responses should have been investigated further.

Discussion

Chemicals may be irritating to skin, may cause skin contact allergy or indeed may do both. There is reliance on predictive tests for irritancy and allergy to identify these properties, so that the risks arising from human exposure can be assessed and managed. However, such predictive tests are rarely 100% accurate. For example, the largest assessment to date shows that in vivo sensitization tests have an accuracy of about 90% (46). Predictive tests for skin irritancy fare less well, with the exception that involving humans can be regarded as highly accurate, at least where the exposure protocol involved is adequate (14). For both irritation and sensitization endpoints, the reality check arises from human experience.

Concerning benzalkonium chloride, the overwhelming evidence from predictive tests is that this material is a skin irritant but is not a skin sensitizer. At first sight, this appears to conflict with part of the human experience; whilst benzalkonium chloride is known to fulfil its potential to cause skin irritation, it has also been reported with some frequency to cause allergy. Does this mean that the predictive skin sensitization assays have given a false-negative result? Alternatively, might it be that, on occasion, skin allergy and skin irritation have been confused clinically? Sometimes, the above questions have surfaced in the literature (47, 48), but there is no conclusive proof either way from the material reviewed above. 2 points may be made which bear on this: first, there is a much more persuasive argument from the clinical evidence that benzalkonium chloride does have ocular effects in contrast to those attributed to skin allergy; second, there is evidence from predictive skin sensitization testing that benzalkonium chloride causes false-positive results. Taken together, we conclude that benzalkonium chloride is irritant to skin, but that the reports of skin allergy are exaggerated, tending to arise from misinterpretation of diagnostic patch tests. As a consequence, this substance should not normally be regarded as, or classified as, a skin sensitizer.

Ancillary