Drugs are an important yet underrecognized cause of glomerular disease. The mechanisms of drug-induced glomerular disease are diverse, ranging from direct podocyte toxicity to induction of autoantibody formation. Differentiation of drug-induced glomerular lesions from idiopathic disease is critical because withdrawal of the offending agent is essential for clinical improvement. Moreover, identification of drug-induced disease often allows for immunosuppressive therapy to be avoided or significantly abridged. The sections in this chapter review the most common drug-induced glomerular diseases encountered in clinical practice: membranous nephropathy (MN), antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis, and the podocytopathies (minimal change disease [MCD] and focal segmental glomerulosclerosis [FSGS]). Glomerular diseases caused by chemotherapeutic agents are reviewed in Chapter 19.

Drugs, along with autoimmune diseases, malignancies, and infections, are a cause of secondary MN (Table 18.1). The attribution of causality in drug-induced MN is particularly challenging given the potential for spontaneous remission, which may confound the ostensible association of drug withdrawal with clinical improvement. Certain pathologic features can suggest a secondary etiology. For example, deposits in primary MN are characteristically immunoglobulin (Ig) G4 dominant.¹ The predominance of an alternative IgG subclass (eg, IgG1) raises suspicion for a
secondary etiology and should prompt a review of potential drug culprits. Likewise, the presence of concomitant mesangial and/or subendothelial deposits can be characteristic of secondary MN.² These findings, however, are not sensitive for, nor are they specific for, drug-induced disease, and there may be no distinguishing pathologic features. Moreover, reports of drug-induced MN have largely predated the discovery of the currently identified target antigens in primary MN, including the phospholipase A2 receptor, thrombospondin type-1 domain-containing 7A, and neural epidermal growth factor-like 1 protein.³,⁴,⁵ It is likely that a proportion of patients historically identified as having drug-induced MN actually had primary MN despite exposure to a suspected drug culprit. Available data on drug-induced MN must be considered in the context of these limitations.

Initial treatment of suspected drug-induced MN is withdrawal of the offending agent. It should be noted that clinical resolution of immune complex disease can be protracted, and proteinuria may not reach a nadir until 12 months or longer following cessation of the implicated drug. Therapy with an inhibitor of the renin-angiotensin system should be instituted in all patients. Patients with severe decompensated nephrotic syndrome or those with high-risk features who fail to improve after drug withdrawal should be considered for an immunosuppressive regimen used to treat primary MN. Failure to improve after discontinuation of the implicated drug should prompt reconsideration of a diagnosis of drug-induced MN.

Exposure to a variety of drugs has been associated with the development of ANCA-associated vasculitis (AAV) (Table 18.2). The drugs with the strongest causal association are hydralazine, propylthiouracil (PTU), methimazole, minocycline, and levamisole (the latter frequently as part of adulterated cocaine in the United States).¹⁵ Other drugs, including allopurinol and sulfasalazine, have been proposed as potential culprits, but data are lacking to definitively infer causality. Essentially, all cases of drug-induced AAV are associated with myeloperoxidase (MPO)-ANCA. A proportion of drug-induced cases have dual positivity for both MPO- and proteinase 3 (PR3)-ANCA, but isolated PR3-ANCA seropositivity largely excludes a diagnosis of drug-induced AAV.¹⁶

The mechanism by which culprit medications induce ANCA production has not been completely elucidated, but mounting evidence suggests that the promotion of neutrophil extracellular trap (NET) formation is a key pathogenic step in many cases. NETosis, a unique pathway of neutrophil death, results in weblike structures comprising DNA and neutrophil-derived cytosolic proteins that can bind and kill extracellular pathogens.¹⁷ Multiple drugs associated with the development of ANCA, including PTU, hydralazine, and levamisole, have been shown to induce NET formation, thereby exposing potential intracellular autoantigens, including MPO and PR3.¹⁸,¹⁹ Other drug-specific mechanisms are also likely operative in inducing ANCA production. For example, in addition to acting as a vasodilator, hydralazine is also an inhibitor of DNA methylation. This may result in decreased epigenetic silencing of the genes encoding MPO, thereby leading to increased expression of the target antigen.¹⁵

The evaluation of any patient with a presentation compatible with AAV, particularly among those found to have MPO-ANCA, should include a detailed review of the patient’s medications and exposure to recreational drugs. Certain features heighten the suspicion of drug-induced disease. Patients with drug-induced AAV
often have markedly elevated MPO-ANCA titers when compared to their counterparts with idiopathic disease²⁰; Visual Abstract 18.1. In addition, dual positivity for both MPO- and PR3-ANCA is highly suggestive of drug-induced disease. Finally, many patients with drug-induced AAV have coexisting antinuclear antibodies (ANAs), including antihistone antibodies. A considerable proportion of patients historically diagnosed with drug-induced lupus, particularly those with aggressive kidney involvement, likely had drug-induced AAV.

The management of drug-induced AAV is dependent on the severity of presentation. Among patients without organ-threatening disease (eg, patients with only fatigue and myalgias), discontinuation of the offending drug with or without a short course of glucocorticoids is often sufficient. Conversely, patients with organ-threatening disease (eg, glomerulonephritis [GN]) should be treated with an induction regimen used for idiopathic disease in addition to discontinuation of the implicated drug. Unlike patients with idiopathic disease, most patients with drug-induced disease do not require prolonged maintenance therapy to prevent relapse. Nonetheless, ongoing monitoring is warranted because the possibility exists that an ostensible case of drug-induced AAV is, in fact, idiopathic disease with coincidental exposure to a potential drug culprit.