Drug Induced Lupus
IntroductionSystemic lupus erythematosus (SLE) is a genetically complex disease upon which exogenous agents, e.g. UV light, viral infection, can have significant influence in terms of exacerbating or precipitating clinical symptoms. Therapeutic drugs are one other exogenous influence and, given that most individuals receiving one of the, approximately 100, drugs associated with lupus inducing properties (LIDs; Table 1) do not develop a lupus-like illness, suggests genetic background is a critical player in this potentially serious side-effect. Those drugs that have reported associations with drug-induced lupus (DIL) do so to differing degrees, from relatively rigorous clinical and mechanistic investigations to more limited case reports. While LIDs can induce a syndrome similar to SLE, this often has less severe, or no, involvement of selected organs, e.g. kidneys, central nervous system. Some LIDs induce a condition more akin to subacute cutaneous lupus erythematosus (SCLE) consequently, unless stated otherwise, reference to DIL in this chapter means the term to encompass drug-induced SLE, SCLE and CCLE (chronic cutaneous lupus erythematosus). The use of the abbreviation ‘SLE’ means idiopathic SLE throughout this chapter.
Incidence and demographics of DILIt has been estimated that, globally, the incidence of DIL caused by all drugs represents approximately 5 to 10% of the total number of patients with SLE. There are differences between DIL and SLE, DIL patients are generally older (50- 70 years, a reflection of the increased use of some therapeutic drugs in older populations) than patients with SLE (average age at diagnosis, 29), the gender bias for SLE (9:1 in favour of females in the reproductive years) is not seen in DIL. White Caucasians may be affected by DIL up to six times more frequently than blacks, where the incidence of SLE is greater. The risk of developing DIL is much greater for doses of drug greater than 200mg/day and individuals often need to be taking these doses for several months to years before symptoms develop. The development of symptoms can be abrupt, initially presenting in a mild or limited manner but gradually worsening the longer the patient is maintained on the drug.
Clinical features of DILAccurate diagnosis of DIL is difficult. Ideally, to be considered a LID, a patient receiving a therapeutic compound should satisfy 4 out of 11 diagnostic criteria for SLE and the symptoms should disappear on withdrawal of the drug. However, the range of symptoms that are positive and could be usefully employed as diagnostic criteria for DIL are limited (arthralgia/myalgia/arthritis, serositis, anti-nuclear antibodies and anti-histone/anti-single stranded DNA antibodies).While the specific criteria used to diagnose DIL have not been formalised, there are several points a clinician should look out for if DIL is suspected in a patient:
• One or more clinical symptoms of SLE are present.
• Presence of antinuclear antibodies, particularly anti-histone antibodies in the absence of other anti-nuclear specificities such as anti-dsDNA.
• The patient had no history of SLE prior to taking the suspect drug.
• The suspect drug was taken continuously anytime from 1 month to 2 years prior to the onset of symptoms.
• Symptoms improve within days to weeks when the drug is discontinued, with antinuclear antibodies and other serologic markers slowly returning to more normal levels.
Arthralgia is the most common, often the only, clinical symptom in 90% of patients with DIL, with myalgia also common and present in 50% of patients. Fever, pleurisy and pericarditis are other characteristic symptoms, with cardiac tamponade associated with selected LIDs. Certain cutaneous symptoms are more evident in DIL compared to SLE (purpura, erythema nodosum, erythematous papules) although, generally, skin involvement is not as common and the cutaneous manifestations more frequently observed in SLE (malar rash, alopecia, discoid lesions and photosensitivity) are largely infrequent in DIL. Mucosal ulcers, lymphadenopathy, Raynaud’s phenomenon, anaemia, leukopenia and thrombocytopenia are generally rarer in DIL. It is important to note, however, that the relative frequency of various symptoms can depend on the identity of the LID. As mentioned earlier, neuropsychiatric symptoms and glomerulonephritis are very rare in DIL. Although there are marked differences in clinical symptoms between DIL and SLE, the differences between drug-induced SCLE and idiopathic SCLE tend to be less marked, for example, the serological profiles between the two diseases are generally similar and DIL is also more common than SCLE.
SerologyAlthough there are similarities with SLE, serological changes are not identical to DIL and in many cases there can be significant elements of drug-specificity in the serological changes. These and other findings are noted in Table 2.
Drug-induced lupus: mechanistic considerationsAn understanding of how LIDs induce lupus is still not completely clear, although notable advances have been made during the past few years. It seems that no single pathogenic mechanism can explain how all LIDs induce DIL, but of the possible mechanisms there are some notable features, which interestingly seem to tie in with aspects of what is known about the pathology of SLE. While some LIDs may have lupus-inducing properties without metabolic transformation, e.g. anti-TNFa therapies, others, such as procainamide,may exert their lupus inducing effects following oxidative metabolism. A point worth re-iterating is that since most individuals do not react to the intake of potential LIDs by developing DIL, individual genetic constitution is key to determining adverse outcome, although the identities of the genetic pre-disposing factors are still largely unknown. Also, besides being of ‘scientific interest’, understanding the mechanism of action of LIDs is considered useful to aid understanding SLE, perhaps with a view to designing novel therapies.
ApoptosisA central player in the pathogenesis of SLE, and probably DIL, would appear to be apoptosis, where dysfunctional cells, or those past their biological usefulness, are removed by an ordered destruction process and the contents of the cell are engulfed by phagocytosis. Apoptosis has an important role in immune function such as the deletion of autoreactive T- and B-cells and the removal of intracellular contents via phagocytosis to prevent the exposure of, otherwise immunologically concealed, intracellular contents, e.g. nucleosomal material, to the immune system. Evidence strongly suggests that apoptotic cells are a significant source of the autoantigens found in SLE. Interference with apoptosis could have profound implications for the development of DIL, through a failure to remove autoreactive T- and B-cells and/or inducing apoptosis, which, on the correct genetic background, where relatively minor perturbations in the load of apoptotic material may lead to the exposure of the immune system to ordinarily sequestered intracellular components, as these cannot be cleared effectively. This latter point is also important if the immune system is dysfunctional and is already pre-disposed to react to self-antigens. In SLE several factors that may conspire to increase apoptotic load, including defects in complement and the ability of macrophages to clear apoptotic cells and debris, along with defects in the enzyme DNAse 1 and serum amyloid P component, which can degrade/mask autoantigens. Several DIL may interfere with apoptosis:
Quinidine and procainamide: At therapeutic concentrations, inhibit macrophage scavenging of apoptotic debris.
Chlorpromazine: Induces apoptosis in activated lymphocytes.
Captopril and lisinopril: These angiotensin II converting enzyme (ACE) inhibitors are widely prescribed for the control of hypertension, yet the apparent expression of ACE on the surface of selected immune cell populations such as Tlymphocytes and monocytes suggests it may have functions in the biological activity of these cells. The effect of captopril may be to block activation-induced apoptosis, perhaps inhibiting clonal deletion and acquisition of self-tolerance.
Statins: These widely prescribed drugs (estimated at least ~25 million people currently) used to control hypercholesterolemia may exacerbate or trigger apoptosis. Also statins may directly modify T-cell function, possibly via inhibition of cholesterol synthesis in these cells leading to plasma membrane perturbations. Statins can promote a shift from Th1 to Th2 immune response, promoting a bias in the latter which is associated with B-cell activation/reactivity.
Anti-TNFa therapies: These agents such as infliximab, etanercept and adalimumab employ various strategies to inhibit TNFa. Infliximab, for example, has been shown to induce apoptosis in activated T-cells, induce increases in circulating nucleosomes and, therefore, may contribute to the burden of apoptotic material in susceptible individuals. Anti-TNF alpha therapy may down regulate control mechanisms for limiting B-cell hyperactivity (a feature of SLE) as TNFa is also an immunoregulatory molecule, which may be involved in the deletion of autoreactive T-cells, so the presence of elevated levels of TNFa may be beneficial in some circumstances. Serum TNFa levels can be elevated in SLE and mouse models with mild SLE may develop a much more severe disease when TNFa deficiency is induced.
Modulation of the immune system using other ‘biological therapies’ such as IL-2 and Type 1 interferons (IFN-a, IFN-ß) can have impacts for DIL -
IL-2 is a pro-inflammatory lymphokine that has been used therapeutically, but elevated IL-2 has been noted in association with a number of autoimmune phenomena and its hypersecretion has been associated with active disease in SLE.
IFN-a and IFN-ß have been associated with the clinical and serological manifestations of SLE thus, on the correct genetic background, use of these compounds therapeutically may lead to DIL in some individuals.
Epigenetic alterationsThe control of gene expression in all cells is exercised at several levels, one process involves epigenetic phenomena (such process can lead to changes in gene expression that do not involve changing the inherent DNA sequence). The methylation of selected cytosines in a promoter region or within a gene can affect the expression of that gene and is an epigenetic process. The consequences of inappropriately modifying DNA methylation depends on the cell in question and where the methylation changes have occurred.The lupus-inducing actions of two well established LIDs, procainamide and hydralazine, either as their parent compounds or oxidised metabolites, may be partly explained by an interference with DNA methylation. In experimental animals, interference with T-cell methylation patterns using agents such as 5-aza-cytidine can induce a lupus-like disease, suggesting the correct maintenance of DNA methylation patterns is important to avoid autoimmunity. Both procainamide and hydralazine can similarly interfere with T-cell methylation patterns and, in so doing, alter gene expression and impact on T-cell function leading to autoimmune responses, the exact mechanism by which procainamide and hydralazine inhibit DNA methylation is slightly different for each compound but the outcome is similar, the promotion of autoimmunity.
Again, comparisons can be drawn between DIL and SLE in this context, hypomethylation of lymphocyte DNA is a biochemical abnormality detected in SLE, as are defects in the same cell signalling pathway in lymphocytes affected by hydralazine. Impaired T-cell methylation leads to overexpression of selected cell surface molecules leading to T-cell hyperresponsiveness, thus signals that do not normally trigger T-cell activation do so, this enhanced Tcell function helps drive autoantibody production by B-cells.
Subversion of T- and B-cell toleranceA critical feature of a normally functioning immune system is that it does not recognise the body’s own cells or their components as foreign. In the case of the contents of the cell, these are usually hidden from the immune system, even when the cell dies, as they are removed by apoptosis. However, during B- and T-cell development, autoreactive cells do arise and these need to be kept in check so they do not enter the periphery and participate in autoimmune reactions. Mechanisms to control this include deletion of autoreactive cells by apoptosis, anergy (where the cells are not deleted but are rendered non-responsive) and an additional mechanism in B-cells, antigen receptor editing, whereby those cells that are initially reactive to self-antigens are rendered un-reactive via modification of cell surface receptors. Again, primarily studied experimentally with procainamide and hydralazine, these agents may subvert tolerance to selfantigens in both T-cells and B-cells acting either as the parent compound or as a metabolite (see next section). Intra-thymic injection of procainamide hydroxylamine, an oxidised metabolite of procainamide, in experimental animals induces an autoimmune syndrome with remarkable similarities to DIL, including elements of the serological profile of DIL. Procainamide hydroxylamine appears to interfere with the establishment of self-tolerance by preventing the induction of anergy. Ideas that the thymus may largely be non-functional in adults would not seem to be completely accurate thus this mechanism may have relevance. In B-cells, hydralazine has a negative impact on antigen receptor editing, perhaps by targeting the same cell signalling system that leads to disruption of DNA methylation by this LID.
Metabolic activation of LIDs, haptenisation and oxidative processesAll compounds foreign to the body undergo xenobiotic metabolism to reduce exposure, accumulation and facilitate excretion.Therapeutic drugs are perceived as foreign and, therefore, also subject to xenobiotic metabolism. Some LIDs may need to be metabolised to induce DIL and how an individual handles metabolism can be an important determinant, for example, selected LIDs may be shunted down a particular metabolic route, if a route that leads to generation of a nontoxic metabolite is slow or impaired. An example may be in the case of slow acetylators, where LIDs may end up accumulating and being oxidised rather than being acetylated. The oxidation products of certain LIDs may be the ultimate species that interferes with immune cell function. Activated neutrophils, via the action of myeloperoxidase, may be one such route for oxidative metabolism of LIDs, and multiple examples of LIDs have been shown to be oxidised by neutrophils to generate cytotoxic metabolites (e.g. hydralazine, procainamide, propylthiouracil, isoniazid, quinidine, clozapine, chlorpromazine). Once generated, these metabolites may subsequently have several effects: induce apoptosis; bind to cell surfaces making such cells targets for immune-mediated destruction; react with and modify endogenous macromolecules (haptenisation) which are recognised as foreign and processed by the immune system; modify cell surface proteins on immune cells and modify their biological function.
Oxidative stress, that is an imbalance of cellular oxidants and anti-oxidants in favour of the former, may have some role to play in SLE and DIL. Induction of excessive production of reactive oxygen species in target cells could be one way LID metabolites induce cell death. Additionally, DNA that has been modified by oxidants is more antigenic for the anti-dsDNA antibodies found in SLE and antioxidants, such as vitamin E,may positively modulate lupus-like disease in some animal models. Recent work in a mouse model has shown that deletion of a particular gene which controls the expression of a battery of genes involved in xenobiotic metabolism, detoxification reactions and protection against oxidative stress intriguingly leads to the development (predominantly in female mice) of a pathologic syndrome bearing striking similarities to SLE, with many of the clinical and serological findings. Notably, oxidative modification of biomolecules and markers of DNA and lipid oxidation are elevated in these animals and precede the development of lupus-like pathology. This may suggest that the gene itself or one or more of the genes under its control may be worthy of exploration in terms of individual susceptibility to DIL.
# - Differentiation between induction of cutaneous LE and SLE is not made in this table.
+ - Weak association in this instance means there are currently insufficient reports of cases to enable a firm decision to be reached, this may be because the lupus-inducing properties of some of these drugs has only recently been appreciated.
Table 1 – Identities of lupus-inducing drugsProven Association
Chlorpromazine. Hydralazine. Isoniazid. a-Methyldopa. Minocycline. Procainamide. Quinidine.
Table 2+: Serological and other findings in DIL and SLE
Anti-native DNA Ab
Anti-denatured DNA Ab
Rheumatoid factor Ab
*Based on two well established LIDs, hydralazine and procainamide.
#Ab = antibodies.
LID = lupus-inducing drug; ox-LID = oxidised LID; + = promotion; - = inhibition.
1. Induction of apoptosis.
2. Interference with disposal of apoptotic debris.
3. Alteration of DNA methylation in T-cells.
4. Interference with establishing tolerance to self-antigens.
5. Metabolism of LID by oxidative metabolism to yield biologically active LID metabolites.
Dept Cancer Studies & Molecular Medicine
University of Leicester
RKCSB Leicester Royal Infirmary
University Hospitals of Leicester NHS Trust
Leicester LE2 7LX
Leicester Royal Infirmary
University Hospitals of Leicester NHS Trust
Leicester LE2 5WW