NSAIDs and Nephrotic Syndrome

The association between nonsteroidal anti-inflammatory drugs (NSAIDs) and nephrotic syndrome has long been recognized. Minimal change disease and membranous nephropathy have been the most common findings in those patients in whom a kidney biopsy was performed (1–6). Regarding NSAIDs-related minimal change disease, it is a peculiar type of nephrotic syndrome in which most of reported patients present with a severe AKI accompanying nephrotic syndrome manifestations (edema, proteinuria >3.5 g/d, and hypoalbuminemia). Kidney biopsies typically show the characteristic pattern of drug-induced acute interstitial nephritis (AIN): a diffuse interstitial infiltrate composed predominantly of T lymphocytes, although eosinophils, macrophages, and plasma cells can also be observed (7). The glomeruli are normal in light microscopy, but a diffuse effacement of podocyte foot processes is observed in electron microscopy (1–4).

NSAIDs-related minimal change disease could be, therefore, categorized as a complication of NSAIDs-induced AIN. However, a “pure” nephrotic syndrome has been described in some patients taking NSAIDs, with diffuse fusion of foot processes but no accompanying interstitial infiltrates. In some patients, tubular necrosis without interstitial infiltrates has been the most salient histopathologic finding. Patients with AIN accompanied by minimal change disease have been reported with a variety of conventional NSAIDs and selective COX-2 inhibitors. As usual in NSAIDs-induced AIN, the extrarenal manifestations (low-grade fever, skin rash, and eosinophilia) that frequently accompany AIN induced by antibiotics and other drugs are characteristically absent (7).

Although NSAIDs are one of the most frequent causes of drug-induced AIN, the incidence of nephrotic syndrome seems to be low. In two patient series of drug-induced AIN, only three patients of a total of 121 showed nephrotic-range proteinuria, although NSAIDs were the offending drug in 40% of the patients (7). However, patients with NSAIDs-related AIN tend to present higher proteinuria values than other types of drug-induced AIN, although they rarely reach the nephrotic range (7,8). Notably, nephrotic syndrome and nephrotic-range proteinuria seem to be a specific complication of NSAIDs, because they have been only exceptionally reported with other types of drug-induced AIN. To explain this susceptibility for proteinuria, it has been suggested that a decrease in the synthesis of prostaglandins induced by NSAIDs could result in an increased conversion of arachidonic acid to leukotrienes, which could activate T-helper cells and induce a diffuse podocyte damage. However, no studies have confirmed this hypothesis.

In most patients, NSAIDs-related minimal change disease resolves after drug discontinuation, which is accompanied in some patients by a short course of corticosteroids (1–4). It has been reported that early treatment with corticosteroids can induce a more rapid and efficient recovery of kidney function in NSAIDs-induced AIN accompanied or not by nephrotic proteinuria (8).

Paradoxically, NSAIDs induce a marked reduction of proteinuria in different types of nondiabetic glomerular nephropathies and can potentiate the antiproteinuric effect of renin-angiotensin blockers (9). Changes in glomerular hemodynamic through a preglomerular vasoconstriction and an improvement of the glomerular protein permselectivity have been invoked to explain this effect. However, the frequent side effects of NSAIDs have prevented their clinical use for the treatment of glomerular diseases.

Nephrotic syndrome with a histopathologic pattern of membranous nephropathy, a paradigmatic type of immune complex glomerular disease, is another side effect of NSAIDs, and different types of conventional and selective NSAIDs have been reported as causatives of this complication (1,5,6). The presence of electron-dense subepithelial deposits along the glomerular capillary loops and a complete effacement of foot processes were the most characteristic findings in kidney biopsies (5,6). Unlike NSAIDs-related minimal change disease, inflammatory interstitial infiltrates were absent in most reported patients. This histologic difference explains why most of the patients with NSAIDs-induced minimal change disease present with nephrotic syndrome accompanied by severe AKI, whereas nephrotic syndrome with normal kidney function is the most frequent presentation in NSAIDs-induced membranous nephropathy.

Immunostaining for the different subclasses of IgG can help to differentiate NSAIDs-induced membranous nephropathy from primary forms of the disease. As in other types of secondary membranous nephropathy, deposition of IgG1 has been reported in patients with cases associated with NSAIDs (5), whereas deposition of IgG4 is characteristic of primary forms. The pathogenesis of membranous nephropathy associated with NSAIDs is unknown. Glomerular deposition of antigens bound to NSAIDs could elicit an immune response, although these drugs could also exacerbate or trigger autoimmune reactions against podocyte antigens. Intriguingly, positive immunostaining for PLA2R has been reported in a patient with membranous nephropathy precipitated by piroxicam, although serum anti-PLA2R was not available (5).

The incidence of NSAIDs-related membranous nephropathy could be greater than suspected according with the study of Radford et al. (6). They reported 125 patients with early membranous nephropathy (stages 1 and 2 membranous nephropathy with small subepithelial deposits). Twenty-nine of them were taking NSAIDs at the time of diagnosis, of which 13 (10%) fulfilled the criteria defined by the authors to establish the diagnosis of NSAIDs-related membranous nephropathy: onset of the nephrotic syndrome while taking NSAIDs, exclusion of other causes of secondary membranous nephropathy, and rapid disappearance of proteinuria after NSAIDs withdrawal (6).

Interestingly and reinforcing the causal association between NSAIDs and nephrotic syndrome, relapses of proteinuria have been reported after re-exposure to NSAIDs, and in some patients, re-exposure was to a type of NSAIDs different from the one causing the first episode of nephrotic syndrome (5).

In this issue of CJASN, Bakhriansyah et al. (10) describe the results of a systematic observational matched patient-control study about the association between NSAIDs and nephrotic syndrome. The study used data from the Clinical Practice Research Datalink, a general practitioner database of the United Kingdom National Health Service. Patients had a diagnosis of nephrotic syndrome established in the period 1989–2017, and controls were patients without nephrotic syndrome before and at the date of diagnosis of nephrotic syndrome in matching cases. Exposure to NSAIDs was divided into current use (NSAIDs prescription within the last month before the date of diagnosis), recent use (prescription within 1–2 months before), or past use (prescription >2 months before). According to the duration of NSAIDs exposure, current use was divided into use of 1–14, 15–28, or >28 days. Patients with past use were divided into those in whom NSAIDs had been discontinued between 2 months and 2 years before the date of diagnosis and those with an NSAIDs discontinuation >2 years. NSAIDs were classified as acetic acid derivatives (such as indomethacin, ketorolac, and diclofenac), propionic acid derivatives (such as naproxen, ibuprofen, and ketoprofen), selective COX-2 inhibitors (coxibs), fenamates, and oxicams, and other NSAIDs.

A total of 2620 patients and 10,454 matched controls were included in the analysis. The results show that current use for >2 weeks, recent use, and past use (discontinuation >2 months to 2 years) of conventional NSAIDs (acetic acid and propionic acid derivatives) were associated with a significantly higher risk of nephrotic syndrome (adjusted odds ratio, 1.34; 95% confidence interval, 1.06 to 1.70; adjusted odds ratio, 1.55; 95% confidence interval, 1.11 to 2.15; and adjusted odds ratio, 1.24; 95% confidence interval, 1.07 to 1.43, respectively) compared with nonuse and that the risk disappeared after 2 years of discontinuation. The use of selective COX-2 inhibitors was not associated with a higher risk of nephrotic syndrome, although there was not a statistically significant trend among patients with a past use (>2 months to 2 years).

This study is the first systematic analysis of the association between NSAIDs consumption and the risk of developing a nephrotic syndrome. Its conclusions are clinically relevant, although several important caveats should be taken into account. One of them is the inevitable chronological imprecision when analyzing drugs, like conventional NSAIDs, that are frequently available over the counter and consumed in a discontinuous manner by a large number of normal subjects. Additionally, the accuracy of the diagnosis of nephrotic syndrome using pre-established codes should be taken with caution in studies like this that involve a large number of patients over a long period of time. The criteria used to establish the diagnosis of nephrotic syndrome are not reported, and therefore, we do not know if all included patients presented a proteinuria >3.5 g/d accompanied by hypoalbuminemia.

A major constraint of the study is the low percentage of patients with histologic confirmation of kidney disease. Although patients 18 years old or younger were excluded (another limitation of the study), a kidney biopsy was performed in only 11% of the patients, despite the fact that kidney biopsy is considered a central tool in the diagnostic workup of most adult patients presenting with nephrotic syndrome. The list of histologic diagnoses is somewhat surprising, because only a minority of patients (15 of 167) taking NSAIDs and presenting a nephrotic syndrome received the diagnosis of minimal change disease or AIN, and other diagnoses (for instance, crescentic or mesangiocapillary GN) have no apparent pathogenic relationship with these drugs.

The study of Bakhriansyah et al. (10) raises questions of great clinical importance for the clinician. Are we overlooking patients with NSAIDs-induced nephrotic syndrome? If so, are we prescribing long courses of immunosuppressive treatments to patients with minimal change disease or membranous nephropathy caused by NSAIDs that would have resolved with the discontinuation of the drug accompanied perhaps by a short course of corticosteroids? Considering that the existing literature on this topic is in general old and scarce, with only patient reports or short series of patients published, data from this study should encourage nephrologists to, on one hand, carefully review the current or recent intake of NSAIDs in any patient with nephrotic syndrome, particularly when minimal change disease or membranous nephropathy is found in kidney biopsies, and on the other hand, perform collaborative studies to collect large series of patients with unequivocal diagnosis of NSAIDs-induced nephrotic syndrome to identify their differential clinical and histopathologic characteristics and delineate their more efficient treatment. Basic research about the pathogenic mechanisms through which NSAIDs cause glomerular damage is also needed.

Disclosures

Dr. Praga has received personal fees for lectures from Alexion, Fresenius, Otsuka, and Retrophin and grant support and personal fees from Alexion, outside of the submitted work. Dr. Mérida has nothing to disclose.