Drug-Induced Lupus

Abstract

Drug-induced lupus erythematosus (DILE) is an important differential to consider in a clinical setting of mild lupus. Numerous drugs ranging from the classically described procainamide and hydralazine to novel biological agents such as anti-tumor necrosis factors and immune checkpoint inhibitors are implicated in causing DILE. Various pathophysiologic mechanisms such as decreased central tolerance, molecular cross-reactivity, and epigenetic modifications of immune cells are known to play a role in the precipitation of drug-induced autoimmunity. Early recognition and prompt withdrawal of the offending agent is often sufficient to treat these autoimmune manifestations. A thorough knowledge of DILE is essential to ensure better patient care and better understand the role of xenobiotics in precipitating autoimmunity.

Introduction

Systemic lupus erythematosus (SLE) is possibly the most heterogeneous disease faced by a physician. It is important to exclude various lupus mimics and possible precipitants in a patient with a constellation of features suggestive of lupus. Drugs have been increasingly identified as important causes of lupus-like syndromes ever since the first description of sulfadiazine-induced lupus-like symptoms by Hoffman in 1945.[¹] The incidence of drug-induced lupus erythematosus (DILE) has been widely reported ever since.

While over a 100 drugs across ten drug categories have been implicated in DILE,[²] majority of the cases have been attributed to either procainamide (20% incidence over 1 year of treatment) or hydralazine (5%-8% incidence over a year of treatment).[³]

With changing times and advances in understanding neoplastic and autoimmune diseases, there have been drastic changes in the pattern of drug usage. Based on this, the era of DILE can be categorized into two periods: before and after the introduction of biological agents.[⁴] A recent report of DILE, assimilated from the VigiBase which is the WHO global deduplicated individual case study report (ICSR) database, delineated the two timeframes. These data originating from over 130 countries show that most of the DILEs reported prior to 2006 were due to nonbiological agents and those later were mainly attributed to biological agents.[⁵] A timeline on the important reports of DILE and the implicated drugs is presented in [Figure 1].

Figure 1: Timeline and details of the important drug-induced lupus cases. CTLA: Cytotoxic T-lymphocyte protein, DILE: Drug-induced lupus erythematous, IFN: Interferon, LE: Lupus erythematosus, PD1: Programmed cell death protein 1, RCT: Randomized controlled trial, SSA: Sjogren's syndrome antigen A, TNF: Tumor necrosis factor

Distinguishing DILE from SLE is of utmost importance to a clinician as the former can be treated simply by withdrawing the offending drug with minimal immunosuppression in a majority of cases. In clinical practice, DILE is often overlooked because of a variety of reasons, including a slow and insidious onset of symptoms after prolonged periods of drug exposure and an ever-increasing list of new drugs. Besides the clinical implication, understanding DILE can give us insights into how xenobiotic (a chemical substance found within an organism that is not naturally produced or expected to be present within the organism) stress influences the development of autoimmune diseases.

This review attempts to give a broad overview on how to identify and treat DILE while also briefly describing the pathophysiological processes involved in the precipitation of these drug-induced features.

Drugs Implicated in Causing Drug-Induced Lupus Erythematosus

Several drugs have been implicated in causing DILE, and most of the data are derived from case reports or case series. There are no prospective studies looking at the incidence of DILE with a particular drug. This could be a limiting factor in interpreting the existing literature as there is an inherent bias in reporting what is already known and in not publishing extensively reported cases. The drugs causing DILE are traditionally classified as follows:[⁵]

1. High risk - ≥10% year risk of developing DILE, for example, procainamide and hydralazine
2. Moderate risk - <10% but ≥1%/year risk of developing DILE, for example, quinidine
3. Low risk - <1% but ≥ 0.1%/year risk of developing DILE
4. Very low - <0.1%/year risk of developing DILE.

An extensive list of drugs causing DILE is summarized in [Table 1].[²⁶⁷⁸]

Table 1: Commonly used drugs classified based on the risk of developing drug-induced lupus*

Drug-Induced Lupus: Definitions, Clinical Features, Differences, and Similarities With Idiopathic Systemic Lupus Erythematosus

Several clinical features of DILE are similar to those seen in idiopathic SLE (iSLE). This poses a challenge to the clinicians to convincingly distinguish between the two. Several attempts to classify and define DILE have been made.

Borchers et al. in 2007 proposed a set of prerequisites to diagnose DILE, which included any patient who had a suspected drug exposure for a sufficient time period developing at least one lupus feature which resolved within weeks of discontinuation of the suspected drug.[⁹] The criterion was further improvised by Xiao and Chang who proposed an algorithm for the diagnosis of DILE.[⁷] The authors emphasized on a detailed history of autoimmune features in self or in family members, assessment of auto-antibody profile, and recurrence of symptoms on re-challenging for a confirmed diagnosis.

DILE usually presents in the older age group with predominant systemic features such as fever. Besides this, the common clinical manifestations in these patients include arthralgia, myalgia, serositis, and skin lesions that are unlike the ones seen in iSLE. Major organ involvement such as lupus nephritis and neuropsychiatric manifestations, though reported, is extremely rare.[⁶]

Based on these presenting features, DILE can be broadly classified into the following:

1. Systemic DILE
2. Cutaneous DILE.
1. Drug-induced subacute cutaneous lupus erythematosus (DISCLE), for example, annular polycyclic or papulosquamous, vesiculobullous lesions and Erythema multiforme-like lesion, and necrotizing vasculitis
2. Chronic cutaneous DILE (CCDILE), for example, discoid lesions and lupus tumidus-like lesions
3. Nonspecific cutaneous lesions, for example, purpura, erythema nodosum, photosensitivity, urticarial, and urticarial vasculitis.

Over the years, several investigators have attempted to identify pointers to differentiate DILE from iSLE, which are summarized in [Table 2].[¹⁰¹¹]

Table 2: Differentiating features between drug-induced and idiopathic systemic lupus erythematosus

The common cutaneous DILE are subacute lesions such as leukocytoclastic small-vessel vasculitis, erythema nodosum, photosensitivity, and urticarial and necrotizing vasculitis. Drug-induced SCLE may be grossly similar to idiopathic SCLE in morphology, but the distribution is more widespread with frequent involvement of lower limbs in the former. Subtle differences in morphology such as a vesiculobullous change along the margin of the lesion are more suggestive of a drug-induced lesion as compared to its idiopathic counterpart.

Erythema multiforme and erythema nodosum are more common in DILE as compared to iSLE once infections are ruled out.[¹²]

A recently published multicentric study conducted across Europe studied the difference between iSLE and DISCLE. The study found that an integrated clinical and immunopathologic evaluation was useful to differentiate idiopathic SCLE from DISCLE. Older age at onset and more frequent systemic symptoms were often encountered in DISCLE as was understood in the past. On histopathological examination of the skin lesions, they reported a higher amount of mucin deposition at the dermo-epidermal junction in cases of iSLE. On direct immunofluorescence, though both the subgroups were similar, there was a higher immunoglobulin (Ig) IgM+C3c deposition along the basement membrane in the idiopathic variant. The eosinophil infiltration was similar in both the groups, which may help the clinician further differentiate these lesions from drug hypersensitivity. Besides this, DISCLE had a higher number of leukocytoclastic vasculitis when compared to its idiopathic counterparts.[¹³]

Drug-induced chronic cutaneous lupus erythematosus (LE) though rare is an important differential especially when fluorouracil agents, such as tegafur and anti-tumor necrosis factor (TNF) alpha agents, are used. Discoid LE is the most commonly reported CCDILE, with only a few isolated case reports of lupus tumidus associated with anti-TNF agents.

Antihistone antibodies are lesser prevalent in DISCLE (30%) than in DILE,[¹⁴] and anti-double-stranded deoxyribonucleic acid (dsDNA) is rare. Associations of positive anti-SSA/Ro with DISCLE are described variably from 40% to 80% positivity in literature.

DILE caused by TNF-α inhibitors or the anti-TNF-induced lupus (ATIL) has frequent involvement of skin. This is in contrast to DILE reported with the “traditional” agents such as procainamide, where arthritis or arthralgia and systemic symptoms are the predominant findings. ATIL rarely presents with hematologic, renal, or central nervous system manifestations. Anti-nuclear antibodies (ANAs) are invariably positive, and unlike in classical DILE, antihistone antibodies are usually absent but anti-dsDNA is more common. Hypocomplementemia may also be encountered in cases of ATIL.

Immune checkpoint inhibitors (CPIs) are exceedingly being used in cancer immunotherapy. These drugs act by blocking the normal regulatory mechanisms of immune activation. Although numerous rheumatological adverse events have been reported with these drugs, DILE seems to be rare. The reporting of these cases also seems to be low as a recent analysis of Food and Drug Administration Adverse Event Reporting System revealed 18 cases of SLE that were not reported in literature. The severity of DILE reported in cases [Figure 1] varies from skin lesions to full-blown lupus nephritis.[⁸¹⁵] The treatment usually involves stopping the CPI and addition of glucocorticoids, rarely requiring the addition of a second-line agent. Administration of CPIs is considered safe in preexisting lupus or any autoimmune disease, but needs observation for flare of disease.

Pathophysiology of Drug-Induced Lupus Erythematosus

Understanding the pathophysiology of DILE provides us with insights on how certain xenobiotic stress leads to autoimmunity. Like the conventional concept of autoimmunity, DILE also develops in genetically predisposed individuals. Although the number of studies are small and supporting evidence is limited, some investigators have found compelling associations and others have provided strong hypothesis and proof of concept for the role played by genetic factors in the development of DILE. A comprehensive view of the pathophysiological process is depicted in [Figure 2].

Figure 2: Pathophysiological process involved in drug-induced lupus erythematosus. LFA-1: Lymphocyte function-associated antigen 1, MPO: Myeloperoxidase, MR: Muscarinic receptor, NET: Neutrophil extracellular traps, PAHA: Procainamide hydroxylamine, ROS: Reactive oxygen species

Genetic Predisposition

Human leukocyte antigen

The immune response seen in patients with DILE is predominantly the IgG type. This finding implicates the adaptive immune system as the major component in causing DILE. The antigen presentation to T cells is the most important step in activating adaptive immunity, and the Class II major histocompatibility complex molecules are of paramount importance in this step. Studies have shown association of human leukocyte antigen (HLA) DR2, DR3, and DR4, with increased risk for DILE, especially with minocycline, terbinafine, and hydralazine.[¹⁶]

Besides the Class II HLA, the components of Class III, namely the C4A and C4B null complement alleles, are associated with DILE.[¹⁷] Although the complement allele polymorphisms are known to be associated with iSLE, Sim et al. demonstrated that hydralazine, penicillamine, isoniazid, and the metabolic products of procainamide inhibited complement component C4 binding.[¹⁸] This inhibits the activation of C3, which is necessary for the clearance of immune complexes. Thus, these drugs indirectly decrease immune complex clearance, ultimately leading to a chronic and persistent inflammatory state.

Acetylation polymorphism

During the biotransformation of drugs, acetylation in the liver plays an important role. Acetylation is carried out by the two isoenzymes of N-acetyltransferase, namely NAT1 and NAT2. The process involves the transfer of an acetyl group from acetyl coenzyme A to the amino or hydrazino group of the drug. Polymorphisms involving the genes coding for these isoenzymes were proposed to alter the duration of action, occurrence of adverse events, and the amount of intermediate metabolite formed in vivo.[¹⁹] Procainamide, hydralazine, and isoniazid, by the virtue of their structure containing aromatic amines or hydralazines, are predominantly metabolized by acetylation.

Initial evidence suggested that slow acetylators treated with procainamide needed lesser time before seroconversion to positive ANA.[²⁰] This, however, did not translate to developing DILE, and there was no association found between the genotype and incidence of drug-induced lupus.[²¹] Over the years, the theory on the role of acetylator phenotype in causing DILE has fallen out of favor as the gene expression and function of the enzymes are confounded by various variables such as comorbid conditions, physical state, effect of microbiota, gender, and race of individuals.[²²] However, the role played by these genes cannot completely be overlooked.

Oxidation polymorphism

During the first-pass metabolism, most of the drugs are subjected to oxidation. Oxidation is commonly carried out by the P-450 isoenzymes. During this transformation, cytochrome P-450 introduces a functional hydroxyl (-OH) group to several drugs including antidepressants, antipsychotics, beta-blockers, and phenytoin. The oxidation reaction may result in the formation of various drug metabolites such as procainamide hydroxylamine (PAHA), which have marked effects on the immune system (discussed below). Few studies investigating the effect of oxidation on hydralazine have reported a formation of Z-DNA. Z-DNAs are unfavorable forms which have a left-handed helix and have no difference in the structure along the major and minor grooves. The Z-DNA results in elevated type-I interferon (IFN) production, which is widely known to play an important role in lupus pathophysiology. Although studying these genes was promising, the enthusiasm has died down. Similar to the discrepancy between the acetylation genotype and phenotype due to numerous confounders, genes involved in oxidation also have significant interactions with numerous factors.[²³]

Even though theories trying to explain the onset of DILE as a result of these polymorphisms have not gained much ground, it is still possible that they play roles that are yet to be identified in predisposing individuals to drug-induced autoimmunity.

Drug Metabolism and Biotransformation

Independent of the genetic polymorphisms, drug metabolites resulting from in-vivo modifications may result in chemical end products which have similar reactivity. Studies on nonbiological drugs have shown that they are substrates for neutrophil myeloperoxidase (MPO)-mediated oxidation and cytolysis.

The most investigated drug is procainamide. Procainamide is taken up and oxidized by activated neutrophils. This oxidative burst leads to the production of a toxic metabolite called PAHA. The combination of PAHA, MPO, and reactive oxygen species leads to increased cytotoxicity of neutrophils and the surrounding cells. Anti-MPO antibodies are also seen in the sera of patients with DILE, which may be responsible for various manifestations of DILE.[²⁴] A similar process of biotransformation has also been described with other drugs including hydralazine, quinidine, phenytoin, sulfone, penicillamine, chlorpromazine, and isoniazid.

Biotransformation of drugs can also lead to smaller molecules (<1000 Da) that are by themselves incapable of causing any kind of immune response. Some such small molecules bind to self-proteins. These form haptens and thus can cause immune activation or dysregulation.[²⁵]

Cross-Reactivity

Drug metabolites, mainly those of minocycline, cause DILE among genetically susceptible individuals. These metabolites lead to antibody generation against microsomal cytochromes, resulting in nuclear and chromatin degradation. This, in turn, leads to autoimmunity with a predominant anti-dsDNA antibody-positive phenotype.

Effect on the Adaptive Immune System

Culprit drugs and their metabolites can alter the adaptive immune response mainly by three pathways, as follows.

Nonspecific activation of lymphocytes

Several drugs or their haptenized forms can act as neoantigens leading to proliferation of CD4+ T lymphocytes. Besides this, various drug metabolites bind to antigen-presenting cells and polarize them to an activated phenotype. This, in turn, will lead to nonspecific activation and proliferation of the T lymphocytes, ultimately cascading into a drug-induced inflammatory state.

Direct action on thymus

PAHA and other similar metabolites lead to loss of self-tolerance by directly acting on the thymocytes. These metabolites hamper positive selection in the thymus, resulting in a lack of tolerance to low-affinity self-antigens, resulting in self-reactive T cells.[²⁶]

Epigenetic dysregulation

Several drugs including hydralazine and procainamide are known to inhibit T cell DNA methylation by either competitive DNA methyltransferase inhibition or by preventing.

DNA methyltransferase induction via the ERK signaling pathway.[²⁷]

Hypomethylation of T cell DNA causes elevated lymphocyte function-associated antigen1 expression, leading to autoimmunity.[²⁸]

Histone acetylation, though less explored in DILE, may lead to drug-induced autoimmunity. The acetylation at the tail of histone-containing lysine leads to opening up of chromatin, which enhances transcription. Drugs with aromatic hydrocarbons are the predominant acetylators and may play a role in disease precipitation.[²⁹]

Effect on Innate Immune System

The role of innate immune system in causing DILE is found by the evidence of activation of neutrophils and monocytes. The monocytes and macrophages are predominantly involved in D-penicillamine-, isoniazid-, and hydralazine-induced DILE. These drugs have a propensity to bind with the aldehydes in macrophages, forming hydrazones. This covalent bonding leads to prolonged interaction of macrophage/monocyte with T cells, thus resulting in autoreactivity.[³⁰]

The role of neutrophils, particularly the formation of neutrophil extracellular traps (NETs) as a mechanism for drug-induced autoimmunity, has been recently described. Some of the culprit drugs known to cause DILE promote the formation of NETs. The drug metabolites act via the muscarinic receptors on neutrophils, which when activated lead to intracellular calcium flux, along with the oxidative burst by PAHA, which leads to NET formation. NETs are a rich source of autoantigens and chromatin materials which cause increased IFN secretion by dendritic cells and lead to autoimmunity.[³¹]

Effect on Apoptosis

Anti-TNF antibodies and minocycline lead to defective apoptosis and defective clearance of apoptotic bodies. Besides, this defective apoptosis leads to decreased clearance of autoreactive T and B cells or in unregulated and random cell death. This leads to excess cell debris, which will result in autoimmunity and inflammation.[²⁸]

Other Mechanisms

Role of interferon

Several case reports of IFN-alpha-induced SLE are reported. IFN results in the upregulation of various facets of adaptive immunity. They decrease Treg cell functions, enhance antigen presentation by the dendritic cells, increase the activation of T-helper 1 (Th1) CD4 + T cells, and also lead to increased Bcell survival by increasing the levels of B cell-activating factor.[³²]

Role of checkpoint inhibition

Immune checkpoints, mainly the cytotoxic T-lymphocyte protein 4 (CTLA-4) and programmed cell death protein 1 pathway (PD-1/PD-L1) inhibitors, are increasingly being used in cancer chemotherapy.

CTLA-4 attenuates activation of naïve and memory T cells by competitively inhibiting binding of the co-stimulatory molecule CD28 at the immunological synapse. Blocking CTLA-4 results in uninterrupted activation of the T cells and leads to a hyperactive adaptive immune response.

PD-1 is overexpressed on exhausted lymphocytes. The interaction between PD-1 and its ligands (PD-L1 and PD-L2) downregulates T cell function and maintains a balance between T cell activation, tolerance, and tissue damage. Blocking this pathway leads to maintenance of the activated state among T cells, thus perpetuating immune response. Thus, when the two important mechanisms that are naturally present to regulate autoimmunity are disrupted, chances of developing autoimmune diseases are high.[³³]

Immunosuppressive drugs including anti-TNF increase the risk of infection. Repeated infectious trigger and bacterial DNA can stimulate poly- to oligoclonal proliferation of B cells leading to autoantibody formation. Anti-TNF also plays a role in skewing immune response to a T-helper 2 phenotype by suppressing Th1 response, which also results in excess autoantibody generation.[⁴]

Treatment and Outcome of Drug-Induced Lupus Erythematosus

The prognosis of DILE is favorable as most of the cases are self-limiting once the culprit drug is withdrawn. Hence, the first step to effective treatment is distinguishing DILE from iSLE. Once the drug causing DILE is identified, it should be immediately withdrawn and the symptoms are expected to resolve over 3-6 weeks. The offending drug should not be reintroduced. However, ATIL is commonly reported with infliximab and etanercept. Studies have shown that switching to an alternative anti-TNF may not induce similar adverse events, and this may be a viable option.[³⁴]

As most of the features are mild, they can be managed with short course of low-dose steroids. Higher doses of corticosteroids and antimalarial may be needed in manifestations such as pericarditis with tamponade, inflammatory pleural effusions, debilitating polyarthritis, or glomerulonephritis. A word of caution is essential in cases of certain drugs such as minocycline and statins which cause a more severe phenotype of DILE. These drugs have sinister manifestations such as autoimmune hepatitis and may also result in death. Hence, optimum supportive care with appropriate immunosuppression is essential in select cases.

The autoantibodies in DILE, however, take a longer time to resolve. Hence, the disappearance of these antibodies should not be a marker of improvement while treating a case of DILE.

Differentials of Drug-Induced Lupus Erythematosus

While several drugs cause distinct autoimmune manifestations mimicking drug-induced lupus, some drugs can potentially uncover or exacerbate iSLE. It is important for clinicians to identify these entities. Patients who have a predisposition to developing SLE have a higher incidence of drug hypersensitivity.[⁴] As the hypersensitivity skin reaction and introduction of a new drug have temporal relation to the appearance of lupus features, it should not be passed off as a case of mere DILE. These patients may in fact benefit from prolonged follow-up after the withdrawal of the offending drug to ensure that features of iSLE do not go unnoticed.

Photosensitivity is another important side effect of numerous drugs and is also a major feature of lupus rash. Drug hypersensitivity in the form of skin rash that is triggered by exposure to ultraviolet rays is reported with sulphonamides, tetracyclines, griseofulvin, piroxicam, and benoxaprofen. Use of such photoactive medications may result in uncovering the underlying lupus or may mimic DILE and will need to be carefully evaluated.[⁶]

Conclusion

As drug-induced lupus accounts to about 10% of reported cases of SLE, it is of paramount importance that these manifestations are rightly identified. With newer drugs altering various biologic pathways entering clinical practice, more such adverse events are inevitable. Proper understanding of the differences and the similarities DILE shares with iSLE and their possible pathophysiological process will enable clinicians to provide better care to the patients. A better understanding of the pathophysiological process will also allow us to decipher the triggers for autoimmunity with greater clarity.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.