Allergic Contact Dermatitis Following Ocular Use of Corticosteroids

Introduction

Allergic contact dermatitis (ACD) is a delayed hypersensitive inflammatory reaction due to direct contact of the skin with environmental chemicals, for example, topical drugs. ¹ The immunological reaction, in which T-lymphocytes are activated in response to specific allergens, has 2 main phases: sensitization and elicitation. During the sensitization phase, the individual acquires a specific immunological memory of the contact sensitizer at the T-cell level. Upon renewed contact with the sensitizing substance, irrespective of where on the skin this contact takes place, the often lifelong memory will mount an immune response, resulting in eczema. The immune response is a delayed-type hypersensitivity reaction as classified by Coombs and Gell, ² since it usually takes more than 12 h to become manifest and peaks at about 24–72 h. The molecular mechanisms that decide whether an individual will become sensitized or not are still largely unknown. ³

ACD is said to occur in 46% to 74% of the patients suffering from conjunctivitis and/or eyelid dermatitis,^4–6 the most important allergens being topical pharmaceutical products, that is, active principles, vehicle components, and preservative agents present in ophthalmic preparations.^4,7–12

Topical corticosteroids (CSs), potent anti-inflammatory and immuno-modulator agents, are important pharmaceutical products for treating individuals with ACD since they reduce immunological reactions. However, in some individuals, CSs could become an antigen that triggers CS allergy. This manifests as an ACD that is unresponsive to CS or provokes a worsening of a pre-existing condition. ¹³

Despite the wide ocular use of CSs, only few cases of patients experiencing allergic reactions have been reported.^{4,7–11,14–17} For example, Alani in 1976 ¹⁴ was the first to report 2 patients suspected of allergic dermato-conjunctivitis due to CSs: they were both tested with the constituents of the ointments applied and showed positive patch test reactions to betamethasone valerate and to hydrocortisone, respectively. Mathias et al.^15,16 described 3 patients, suffering from chronic iridocyclitis, who developed severe conjunctival inflammation following retrobulbar injections with a methylprednisolone-acetate suspension. Lewis ¹⁷ also reported multiple contact sensitivities to eyedrop ingredients, including the drugs prednisolone. Most authors of clinical reviews mention intraocular CSs as potential causes of ACD.^{4,7,8,10–12}

However, when problems arise on the eyes and/or eyelids, the significance of contact allergy to CSs cannot easily be assessed, since ocular preparations are not often suspected as potential causes and, hence, have rarely been patch tested.

We wanted to study the clinical features and to identify the molecules responsible for contact-allergic reactions following ocular use of CS preparations.

Methods

We reviewed the clinical data and the results of patch testing of an extended CS series in the Dermatology Department at the K.U. Leuven University Hospital during an 18-year study period, that is, from January 1, 1990, to June 30, 2008. ¹⁸ All patients attending the Contact Allergy Unit were examined clinically, and routinely patch tested with a baseline (standard) series containing (1) tixocortol pivalate 0.1% pet., a marker that detects allergic reactions to CS molecules such as hydrocortisone, methylprednisolone, and prednisolone; (2) budesonide 0.01% pet., a marker for CS such as triamcinolone acetonide and amcinonide (19) and (3) hydrocortisone 17-butyrate and prednisolone caproate both diluted in 0.1% ethanol. Moreover, those patients having been exposed to CSs were, in most cases, also tested with the respective molecules used. Patients with a proven contact allergy to a CS and some presumed to have had such a reaction, according to their history, were later tested with an extended CS series to study simultaneous positive (and cross) reactions.

Sixty-six CSs, including those from the baseline series, as well as 2 sexual hormones, that is, progesterone and testosterone, were thus tested (Table 1). All CSs (except for the 2 screening agents officially added to the baseline series ¹⁹ ) were made up in the Contact-allergy Unit by a laboratory technician; the substances used for these preparations had previously been obtained from the pharmaceutical companies marketing them. Dexamethasone Na-phosphate (introduced in November 1990) and difluprednate (Sicor, Milan, Italy; marketed in 2008 to treat ocular pain ²⁰ ) have been included in the base line series since October 1994. The test solutions were freshly prepared every 6 to 12 months.

Ethanol was the vehicle used to test all CS molecules except for tixocortol pivalate and budesonide where petrolatum worked well.^21,22 Due to their weak trans-epidermal penetration, hydrocortisone and cortisone acetate were diluted in an equal mixture of ethanol and dimethyl sulfoxide (DMSO). ²¹ Both ethanol and ethanol/DMSO were included as controls in the extended series.

The concentrations of the test solutions were defined according to the methods of Isaksson et al.^23,24 (Table 1). Particularly for budesonide, lower concentrations were also tested, that is, 0.01% pet. and 0.002% eth.

The patch test materials used were van der Bend chambers (van der Bend, Briele, The Netherlands) on Mepore^® (3M), covered on the upper back with Mefix^®. The patches were removed after 2 days and the patch test reactions were evaluated at day 2, day 4, and in many cases also after 7 days, in accordance with the International Contact Dermatitis Recommendation Group (ICDRG) criteria. All clearly positive tests were taken into account in this study. Although, in general, dubious reactions were excluded from our analysis, weak irritant reactions to the DMSO/eth (even in the presence of hydrocortisone) could not always be ruled out with certainty.

The results were collected from a computer database containing the following information: sex, date of birth, occupation, personal and familial atopic history, localization and lesion type, use of CSs, and patch test results. The relevance of the reactions was considered on the basis of a history of known or likely contact of the affected skin area with products containing the allergen, as well as according to the follow-up of the patients in whom the dermatitis got better, for example, in the case of an underlying disease such as atopic or seborrheic dermatitis, or disappeared on discontinuing the use of the incriminated CS.

To categorize the multiple positive reactions often observed with CSs, which may be due to simultaneous or subsequent sensitization or to cross-reactivity, ²⁵ the CSs were classified into 4 reacting groups: groups A, B, C, and D. ²⁶ Group D has been further sub-classified into 2 subgroups: D1 and D2^18,27,28 (Table 2).

Results

Eighteen (5.7%) out of 315 CS allergic patients (Table 3) presented with facial, periocular, or eyelid eczema, and/or conjunctivitis, following ocular application of CSs. Their symptoms responded to treatment either poorly or not at all; decreased very rapidly when the treatment was stopped; or got worse when CS-based collyria or ointments were reapplied (Figs. 1–2 provide examples of ACD following the ocular use of dexamethasone Na-phosphate).

OPEN IN VIEWER

FIG. 1.

Allergic contact dermatitis from corticosteroids, in this case following the ocular use of dexamethasone (ointment) and dexamethasone Na-phosphate (collyrium), respectively: periocular edema and eczema, conjunctivitis, and worsening of the ocular pathology for which the corticosteroid was prescribed.

OPEN IN VIEWER

FIG. 2.

Patch test results. Positive patch tests of dexamethasone containing ointment and Na-Phosphate containing collyrium, respectively (Deicol^®; Meda Pharma, Brussels, Belgium).

According to the clinical history, the patients had used ophthalmic preparations containing hydrocortisone, prednisolone, dexamethasone, fluorometholone, and/or triamcinolone actonide. All 18 patients tested positively to the group A CSs, that is, hydrocortisone, prednisone, prednisolone, methylprednisolone, and/or particularly to tixocortol pivalate, the latter being the contact allergy marker of this group.

Eight patients, all of whom had used hydrocortisone ophthalmic ointment [numbers (nos.) 1, 3, 5, 8, 11, 13, 17, and 18; Table 3], reacted exclusively to group A, particularly to hydrocortisone, and presented with cross-reactions with other CSs from the same group.

Some patients (nos. 2, 3, 4, 5, 6, 9, 10, 12, 16, and 17; Table 3) also tested positively to other allergens present in opthalmic products, such as benzalkonium chloride and thiomersal (preservative agent), polymixine B, tetracycline, neomycine, chloramphenicol, and rifamycine (local antibiotics), and wool alcohols (an excipient often used in ophtalmic ointments).

Discussion and Conclusion

Not many cases of allergic reactions due to ophthalmologic use of CSs have been reported in the literature. In this study, out of 315 patients with a CS allergy, 6% presented with ACD and/or conjunctivitis following ocular administration.

Recognizing CS allergy can be difficult as the clinical signs are neither specific nor spectacular, or display a completely atypical chronology, which is due to the anti-inflammatory properties of the CS. As mentioned by Le Coz, ³¹ the clinical manifestation of such reaction depends on 2 competing effects that are of variable intensity and offset in time, namely, the immunological allergic response and the pharmacological antiallergic effect.

Facial and periocular edema and/or eczema, conjunctivitis, a dry sensation, burning, itching (pruritus), or tearing are symptoms that may indicate sensitization. Worsening, rapid increase, and/or absence of response of the ocular pathology for which the CSs was prescribed should alert the clinician to suspect a contact allergy.

All patients presented with multiple positive reactions to CS, which may be due to concomitant but certainly also to cross-sensitivity.^25–29,32 The potential of CSs to cross-react within and between groups could be related to structural homology and biometabolization of the CS. Foti et al. ³³ suggested that the cross reactivity of CS may occur at T-lymphocyte level during antigen presentation. Molecules from group A were mainly implicated, in particular hydrocortisone, the most prescribed ophthalmic CS. Group A CSs classically cross-react with molecules from group D2, which are rapidly metabolized in the skin to group A molecules.^{18,26–29,34} One example of intracutaneous biometabolization is the conversion of hydrocortisone 17-butyrate (group D2) to hydrocortisone 21-butyrate, followed by an enzymatic transformation into hydrocortisone (group A). This illustrates the typical cross-reaction pattern observed between group A and group D2.

Further, besides CSs, multiple positive tests were also observed for other potential allergens present in ocular preparations, that is, topical antibiotics, antiseptics, and even vehicle components.

Conventional patch testing of ocular products as such (as is), in particular beta-blocking agents, is often unsuccessful in diagnosing ACD and the baseline standard series does not seem very useful either.^35,36 The anatomological and physiological properties of eyelid skin, sometimes previously damaged by dermatitis, may be the cause of a lower threshold to contact allergy in comparison with the thick healthy skin of the back. ³⁶ Repeated open application tests on the forearm (1 application 2 or 3 times a day during several days) is sometimes necessary to confirm sensitization.

Extensive patch testing (and sometimes additional tests) with all ingredients (including CSs) of the preparations used, are thus required to detect all relevant allergens, to advise patients about which products should be avoided, ³⁷ as well as with regard to the use of future safe alternatives.