Editorial Desensitization to Hydroxychloroquine: Alternative Interpretations Considering the increasingly important role of the amino- Psoriasis (exacerbation) quinoline antimalarials (AA) in the management of systemic Pustular eruption lupus erythematosus (SLE) and rheumatoid arthritis (RA), Rash [sic] Mates and coworkers are to be congratulated for addressing Stevens-Johnson syndrome an important clinical issue in the management of antimalar- Toxic epidermal necrolysis ial cutaneous adverse reactions1. However, their report rais- Urticaria es several important issues not commented upon by these Vasculitis investigators. We wish to offer an alternative interpretation to the results of their proposed desensitization protocol and An exanthem is the most commonly noted cutaneous provide a broader perspective on clinical issues related to adverse reaction pattern to all drugs, ranging between 51% AA cutaneous adverse drug reactions. and 95% in various series5,6; it is also the most common A pruritic maculopapular eruption that was presumably cutaneous adverse reaction to AA7. symmetrically distributed on the trunk and extremities Other than the suggestion that hydroxychloroquine- occurring 1–2 weeks after starting hydroxychloroquine (pre- induced exanthems might be more common in dermato- sumably hydroxychloroquine sulfate; HCQ) was reported in myositis than in other disease settings7, drug-induced exan- all 4 cases described by Mates, et al. Although skin biopsy thems produced by AA are not significantly different in results were not presented, this pattern of cutaneous hyper- clinical features or prognosis from drug-induced exanthems sensitivity would appear to be best classified as a simple produced by a host of other more commonly employed drug-induced exanthem (synonymous with exanthematous agents including antibiotics, anticonvulsants, antiinflamma- drug rash/eruption, maculopapular drug rash/eruption, mor- tories, and allopurinol. In addition, a number of viral infec- billiform drug rash/eruption, rubeliform drug rash/eruption, tions can produce cutaneous exanthems that closely simu- 2 scarlatiniform drug rash/eruption) . late or are identical to drug-induced exanthems. It has been The following list presents a more complete menu of the speculated that concurrent infections might actually repre- clinicopathological patterns of cutaneous adverse drug reac- 3 sent a priming factor for the development of drug-induced tions that have been associated with AA . exanthems. The frequency of amino-penicillin (ampicillin, amoxicillin) induced exanthematous eruptions in infectious Patterns of cutaneous hypersensitivity reactions seen with 8 3 mononucleosis approaches 100% . aminoquinoline antimalarials . It is typical for drug-induced exanthems to resolve spon- taneously even if the offending drug is not discontinued. Acute generalized exanthematous pustulosis Thus if necessary, one can “treat through” drug-induced Angioedema exanthems, including AA-induced exanthems8. However, Bullous eruptions this should be done with vigilance, as extension of the cuta- Erythema annulare centrifugum neous injury pattern to a potentially more serious one is Erythema multiforme possible, such as exfoliative erythroderma, delayed drug Erythema nodosum hypersensitivity reaction (synonomous with drug rash with Erythroderma eosinophilia and systemic symptoms, drug-induced delayed Exanthems multiorgan hypersensitivity syndrome), Stevens-Johnson Exfoliative dermatitis syndrome, or toxic epidermal necrolysis. However, such Fixed drug eruption extension appears to be quite rare. Treating through other Lichenoid eruption* patterns of AA cutaneous adverse drug reactions (Table 1) Photosensitivity such as lichenoid eruptions, vasculitis, and exfoliative der- Polymorphous light eruption matitis has generally been felt to carry too much risk to be * This pattern of cutaneous adverse drug reaction can be a recommended. harbinger of severe AA-induced bone marrow toxicity4. Drug-induced exanthems are also characterized by their Personal non-commercial use only. The Journal of Rheumatology Copyright © 2007. All rights reserved. Sontheimer: Editorial 253 Downloaded on September 24, 2021 from www.jrheum.org frequent failure to reappear following oral challenge with clinical expression of a drug-induced exanthem. As previ- the offending drug. In a recent study, 784 patients having ously discussed, concurrent viral infections that activate the prior cutaneous adverse drug reactions of various types (pre- cutaneous innate immune response might represent such a dominately drug-induced exanthem, fixed drug eruption, danger signal. and urticaria) underwent 1001 oral challenges with the There are several other management options that can be offending agent over a 25 year period. The prior cutaneous considered when faced with patients who experience AA adverse drug reaction in 51% of these 784 patients was cutaneous adverse drug reactions, including oral challenge thought to be a drug-induced exanthem. Skin test-positive with alternative AA structures. Few published data exist patients and patients having a history of serious cutaneous concerning oral challenge with chloroquine in patients who adverse drug reactions such as Stevens-Johnson syndrome, have demonstrated a HCQ sulfate-associated drug-induced toxic epidermal necrolysis, or severe angioedema were not exanthem. Pelle and Callen reported that only one of 3 challenged. Only 13% of the patients challenged developed patients with dermatomyositis who had exhibited HCQ sul- 5 a positive challenge reaction . Seventy-one percent of the fate-associated exanthems re-expressed the exanthem after positive challenge reactions were drug-induced exanthems. oral challenge with chloroquine phosphate7. It has been the It is possible, therefore, that delivery of increasing doses personal experience of members of the North American of HCQ suspension over time in the desensitization protocol Rheumatic Skin Disease Study Group Organizing described by Mates, et al was a null event rather than one Committee that chloroquine phosphate can be adminis- that induced immunological tolerance to HCQ. It would tered to patients who have experienced prior HCQ sulfate- seem premature for the authors to speculate about tolerance- associated exanthems with a low risk of re-expression of inducing capabilities of this desensitization regimen based the exanthem or appearance of other clinical forms of on the data presented. cutaneous hypersensitivity. This might be explained in The histopathology of drug-induced exanthems is several ways: (1) the above noted tendency of some cuta- marked by the presence of skin-homing CLA+, CD4+ T neous adverse drug reactions such as drug-induced exan- cells bearing activation surface markers [interleukin 2 (IL-2) thems to fail to reappear upon challenge; (2) lack of receptor, HLA-DR, lymphocyte function-associated anti- immunological crossreactivity between the 2 closely gen-1, L-selectin) that are focused in a perivascular distri- related base structures of HCQ and chloroquine; (3) anti- 2 bution (primary data summarized in Lerch, et al , genic differences relating to the different salt moieties of 9 Yawalkar ). The associated dermal microvascular endothe- HCQ sulfate and chloroquine phosphate; and/or (4) dif- lial cells display activation markers including E-selectin and fering excipient profiles of tablets of the same drug from P-selectin. Eosinophils are also frequently seen in associa- different manufacturers. tion with activated dermal T cells thought to be the result of Multiple excipients have been implicated as the cause of eosinophil chemotaxis from local production of IL-5, eotax- cutaneous adverse drug reaction10. Tablets from different in, and RANTES. In addition, activated CD4+ T cells and manufacturers containing the same AA base molecule and CD8+ T cells expressing cytotoxicity markers (perforin, salt can differ in the excipients they contain10. Also, tablets granzyme B) are seen at the dermal-epidermal junction in an containing different salt forms of the same AA base can have interface dermatitis pattern that includes vacuolar degenera- different excipient profiles10. There are currently at least 2 9 tion of activated keratinocytes in the epidermal basal layer . tablet forms of HCQ sulfate on the US market, one branded It has been suggested that activated keratinocytes expressing product [Plaquenil (Sanofi-Aventis)] and one generic prod- HLA-DR and various adhesion molecules might be capable uct. Mates, et al did not indicate whether the specific tablet of presenting drug-related antigens to previously sensitized form of HCQ used in their oral desensitization protocol was infiltrating T cells. identical (i.e., from the same manufacturer and lot number) These observations are consistent with the hypothesis to the one implicated in the original cutaneous adverse drug that drug-induced exanthems result from cytotoxic cellular reaction experienced by their 4 patients. Thus, we cannot be injury within the upper dermis and epidermal keratinocyte certain that the commercial form of HCQ tablets that were basal layer resulting from a conventional T cell-mediated used to prepare the desensitization solutions was the same as delayed hypersensitivity mechanism involving immunolog- that which produced the original cutaneous adverse drug ic memory. This is supported by the observation that 70% of reactions. Without this knowledge, it is possible that the patients with drug-induced exanthems display positive