Short title: Viral infections and drug hypersensitivity
Werner J. Pichler 1 & Marie-Charlotte Brüggen2, 3, 4
1 ADR-AC GmbH, Holligenstr 91, 3008 Bern
2 Allergy Unit, Department of Dermatology, University Hospital Zürich
3 Faculty of Medicine, University Zürich
4 Christine Kühne – Center for Allergy Research and Education (CK-CARE)
ORCID IDs: Werner Pichler: 0000-0002-8117-359X Marie-Charlotte Brüggen: 0000-0003-3215-3370
Corresponding author: Werner J Pichler, MD ADR-AC GmbH Holligenstr 91 CH 3008 Bern Switzerland e-mail: werner.pichler@adr-ac.ch
Acknowledgments: the authors thank Daniel Yerly and Oliver Hausmann for inspiring discussions and Katja Martin as well as Clara Clemens for help in preparing the figures and text.
Abstract:
Virus infections and T cell-mediated drug hypersensitivity reactions (DHR) can influence each other. In most instances, systemic virus infections appear first. They may prime the reactivity to drugs intwo ways: First, by virus-induced second signals: certain drugs like β-lactam antibiotics are haptens and covalently bind to various soluble and tissue proteins, thereby forming novel antigens. Under homeostatic conditions, these neo-antigens do not induce an immune reaction, probably because co-stimulation is missing. During a virus infection, the hapten-modified peptides are presented in an immune-stimulatory environment with co-stimulation. A drug-specific immune reaction may develop and manifest as exanthema. Second, by increased pharmacological interactions with immune receptors (p-i) : drugs tend to bind to proteins and may even bind to immune receptors. In the absence of viral infections, this low affine binding may be insufficient to elicit T cell activation. During a viral infection immune receptors are more abundantly expressed and allow more interactions to occur. This increases the overall avidity of p-i reactions and may even be sufficient for T cell activation and symptoms. There is a situation, where the virus-DHR sequence of events is inversed: in drug reaction with eosinophilia and systemic symptoms (DRESS), a severe DHR can precede reactivation and viremia of various herpes viruses. One could explain this phenomenon by the massive p-i mediated immune stimulation during acute DRESS, which coincidentally acvitates many herpes virus-specific T cells. Through p-i stimulation, they develop a cytotoxic activity with killing of herpes peptide-expressing cells and release of herpes viruses. These concepts could explain the often transient nature of DHR occurring during viral infections and the often asymptomatic herpes-virus viraemia after DRESS.
Keywords: exanthems, virus infection, drug hypersensitivity, hapten, p-i concept, DRESS, herpes virus, viraemia, SARS-Cov2
Abbreviations: antigen-presenting cell: APC; drug hypersensitivity reactions: DHR; p-i: pharmacological interactions with immune recepto; human leukocyte antigens: HLA; T cell receptor: TCR; maculopapular exanthema: MPE, Stevens-Johnson syndrome and toxic epidermal necrolysis: SJS/TEN; drug reaction with eosinophilia and systemic symptoms: DRESS; Epstein-Barr virus: EBV; human herpes virus 6: HHV6; cytomegalovirus: CMV;
Introduction
Viral infections are common and can cause a multitude of symptoms. The most frequently observed are respiratory , fever, and exanthems. Due to the similarities between the symptoms presented during the early stages of viral and bacterial infections (e.g., fever and tonsillitis) and/or because a bacterial superinfection is considered, many patients with viral infections are treated with antibiotics and antipyretics. Shortly after these drugs were used in practice, a high incidence of exanthems was observed, which was later attributed to viral infection. Such connections between viral infections and drug hypersensitivity (DH) were particularly apparent in young children experiencing their first encounter with various respiratory viruses, and in adult patients with Epstein-Barr virus (EBV) and human immunodeficiency virus (HIV) infections [1-5].
While the clinical manifestation of the interplay between viral infections and DH has been widely acknowledged for many years, an explanation for this increased incidence is missing. Indeed, the topic is complex due to virus diversity, the number of causative drugs, and the varied clinical manifestations involved.
The mechanisms underlying DH have been explored intensively in the last 25 years. A substantial fraction of DHR involves the adaptive immune system (IgE, T cells), and drug-reactive T cells and antibodies have been identified. Regarding T cell stimulations, two predominant mechanisms have been elaboratedussed; the hapten and p-i concepts.
Some beta-lactams are haptens, which means that they bind by covalent bonds to proteins, and thus form a new and stable antigen [6]. But mostly they do not elicit immune reactions, presumably because no co-stimulation (second signal) is provided by the drug itself [7]. The second mechanism postulates a non-covalent (pharmacological, off-target) binding to immune receptors like human leukocyte antigen (HLA) or T cell receptors (TCR) [7, 8]. This binding to crucial components of the cellular immune system, can result in an unorthodox, sometimes severe immune stimulation in the absence of co-stimulation (pharmacological interaction with immune receptors, p-i concept). This concept has been summarized in previous review articlesemphasizing the similarity to an allo-and partly superantigen stimulation[9-11].
Here we try to incorporate these two concepts of drug-induced T cell stimulation to explain the association between viral infections and T cell-mediated DHR. We highlight that DHRs during viral infections are driven by co-stimulation of T cells, or simply by the increased expression of TCR/HLA on T cells. We believe, these concepts explain the transient nature and low reproducibility of most drug-induced exanthems originally appearing in the context of (transient) virus infections.
Exanthems in DH
The typical manifestations of virus-related DHR are so-called “rashes”. These are macular or maculopapular exanthems (MPE), some have vesicles. They are often mild in nature, and transient. However, some virus-linked DHRs which affect a large proportion of the body surface area, can last much longer, are more maculopapular, and can be associated with signs of liver involvement (transient elevation of ALAT/ASAT) [12, 13].
Systemic MPE itself represents a rather strong immune stimulation [14, 15]. Patients display an expansion of activated lymphocytes in peripheral blood, upregulation of HLA class II expression on CD8+ lymphocytes, expression of CD25 and CD69 on CD4+ cells, and increased concentrations of IFN-γ, TNFα, IL-5, IL-6, etc in the serum [16]. Histological analysis of MPE reveals CD4+, and, to a lesser extent, CD8+ T cells infiltrating the dermis and epidermis [17, 18]. This response has been demonstrated to be, in part, drug-specific [19]. Notably, in MPE, both the infiltrating CD8+ and CD4+ T cells are cytotoxic [17, 18, 20]. The target of this cytotoxic reaction are activated keratinocytes expressing HLA class II- and adhesion molecules (e.g. ICAM) [17, 18].
The extent of CD8+ cell infiltration seems to be variable [14, 15]. In most drug-induced MPE, CD4+ T cells predominate in the skin and CD8+ T cells predominate in the circulation. Milder exanthems are mediated by cytotoxic CD4+ T cells, while the more extensive and maculopapular forms are due to both CD4+ and CD8+ cell-mediated cytotoxicity. Bullous forms appear to be mediated mainly by CD8+ T cells [21]. A concomitant virus infection may also boost the CD8+ T cell involvement in the tissue. The recently described MPE during COVID-19 displayed massive polyclonal T cell activation in the circulation and a rather high percentage of infiltrating CD8+ T cells (CD8:CD4 ratio of 3:1) into the cutis/epidermis [22]. A similar but more severe response was also observed in some exanthemas following beta-lactam use during EBV infection [12]. The most powerful example of how a concomitant virus infection can impact DHR is seen during HIV infections. The incidence of SJS/TEN, which is a predominantly CD8+/NK cell-mediated reaction [21, 23, 24], is about 100-fold increased in HIV patients compared to non-infected individuals [25].
In normal MPE without virus involvement, around 50% of patients present with substantial eosinophilia [26, 27]. In contrast, eosinophilia is an unusual finding in viral exanthemas [27]. Indeed, some drug-specific T cells from patients with MPE excrete high amounts of IL-5, which may explain the typical eosinophilia in various forms of DHR [28][22].
Why some viral infections exacerbate DHR more frequently than other microorganisms is currently unknown. A partial explanation may lay in the similarly complex immune responses stimulated by viruses and drugs. They can provide soluble and cell-bound antigens, engage in peptide presentation via HLA class I and II and manipulate various components of the cellular immune system. This similarity is also reflected by the similar clinical and histological appearance of virus- or drug-induced exanthemas. Some drug-induced exanthemas are so similar to viral exanthems, that they were named accordingly (morbilliform or rubeola-like exanthems). Similar exanthems are also seen in graft-versus-host disease, where eosinophilia is also observed [29]. There are clinical similarities between virus, allo-allele and drug-induced exanthems and, as discussed previously, there may also be mechanistic similarities [9].
Viral infections enhance T cell hypersensitivity to drugs
A clue to better understanding the interplay between virus infection and DH lies in the two different ways, by which drugs can stimulate T cells. First by forming a new antigen (hapten -protein complex)[6] and second by the p-i mechanism, which results in an unorthodox, mostly cytotoxic T-cell stimulation [8, 9].