Background: Acute interstitial nephritis (AIN) is one of the chief causes of acute kidney injury (AKI). AIN might be produced by drugs, infections, autoimmune diseases, or can be idiopathic. Among these etiologies, drug-induced AIN (DI-AIN) is the dominant one in many countries. Even when DI-AIN is suspected, identification of the putative drug is challenging. Summary: DI-AIN is an increasingly common cause of AKI. Diagnosis continues to pose a challenge for physicians due to nonspecific clinical symptoms, and the fact that it can be triggered by a wide variety of medications. Furthermore, the gold standard for the diagnosis is kidney biopsy. All these aspects render the diagnosis more difficult. The withdrawal of the causative drug of DI-AIN is the centerpiece of the treatment, and if early restoration of original kidney function is not obtained, several studies support the treatment with steroids especially when they are started quickly. Key Messages: Almost all drugs have the potential to produce drug-induced acute interstitial nephritis (DI-AIN); however, antibiotics, nonsteroidal anti-inflammatory agents, and proton pump inhibitors account for the majority of the reported cases. DI-AIN is produced by an idiosyncratic delayed type IV hypersensitivity reaction, but the precise pathophysiological mechanism remains to be elucidated. DI-AIN symptoms are nonspecific, and most of the patients will present mild symptoms including malaise, nausea, and vomiting. The classical triad, associating fever, rash, and eosinophilia, is seldom present. Nonoliguric acute kidney injury is the main renal manifestation of DI-AIN. Tubular nonnephrotic range proteinuria is usually present. Diagnosis of DI-AIN relies on maintaining a high index of suspicion in those patients at greater risk, but kidney biopsy is required to confirm diagnosis. Histologically, AIN is characterized by the presence of an extensive interstitial infiltrate, mainly composed of lymphocytes and monocytes, but eosinophils, plasma cells, histiocytes, and polymorphonuclear cells can also be found. The withdrawal of the presumed causative drug of DI-AIN is the mainstay of the treatment. When there is no evidence of kidney function recovery after an interval of 5–7 days since interrupting the treatment with the suspected drug, several studies support the treatment with steroids, especially when they are promptly started. Early corticosteroids would decrease the inflammatory infiltrates of the kidney interstitium, thus preventing the risk of subsequent fibrosis.

© 2022 S. Karger AG, Basel

Introduction

Acute interstitial nephritis (AIN) is one of the leading causes of acute kidney injury (AKI), and probably, the real frequency is underestimated since a significant number of AKI of unclear etiology would correspond to AIN if a noninvasive diagnostic tool rather than a kidney biopsy was available. Histologically, it is characterized by inflammatory infiltrates in the kidney interstitium, with usually preserved glomerulus and vessels. This disease was first described in 1898, mainly associated with infectious diseases [1]. However, this changed during the XX century when drug-induced AIN (DI-AIN) started to be relevant, currently assuming 85% of AIN cases. The first cases of DI-AIN were 7 patients with penicillin-induced AIN. Since then, a considerable number of cases and multiple drugs have been involved.

Currently, DI-AIN roughly represents the leading cause of AIN in many countries, but infectious AIN is still an important cause of AIN in less developed countries. It has been estimated that DI-AIN is the third most common cause of AKI in hospitalized patients [2].

It is difficult to assess the exact prevalence of DI-AIN since studies are based on retrospective analysis and most notably because the gold standard for the diagnosis is kidney biopsy. According to biopsy data registries, it is possible to extrapolate an approximate prevalence of the disease. Globally, it has been estimated that on average between 0.5% and 5% of all kidney biopsies performed have the disease [3]. Nevertheless, this percentage increases up to 27% if the biopsy is performed in the setting of an AKI [2]. In the Spanish Registry of Glomerulonephritis, the percentage of biopsies meeting this diagnosis has increased from 3.6% to 10.5% during the past decade, with a more pronounced increase in the elderly population (from 1.6% to 12.3%). Such an increase is apparently linked to the generalized use of antibiotics and other over-the-counter medications such as proton pump inhibitors (PPIs) and nonsteroidal anti-inflammatory agents (NSAIDs) in most countries. DI-AIN is highly prevalent among the elderly with an estimated prevalence of19% [4]. These differences could be explained by several points: chronic medication and over-the-counter medication are more common among older patients, they tend to present some degree of renal impairment, and to the fact that kidney biopsies are nowadays performed more frequently in this group.

To date, three major drug classes (antibiotics, NSAIDs, and PPIs) accounted for 80–90% of the reported cases of DI-AIN. However, a new class of medications, immune checkpoint inhibitors (ICIs), has increasingly gained attention as a cause of AIN [5]. These new drugs pose the same diagnostic or therapeutic challenges as other DI-AIN, but they add an important ethical question since in some patients there is not an alternative treatment, and therefore, the treatment with ICIs cannot be discontinued.

Pathogenesis

It has been suggested that the underlying mechanism of DI-AIN would be a type IV idiosyncratic delayed hypersensitivity reaction mediated by T cells. This hypothesis is supported by several observations: eosinophils are commonly found in kidney biopsy, T cells are found to be the dominant lymphocytes forming sometimes granulomas, the symptoms of AIN remind those of a systemic hypersensitivity reaction, the time course of disease, and the negative immunofluorescence (IF) staining for immunoglobulins, and complement favor all together a type IV hypersensitivity reaction theory [6, 7].

Several mechanisms have been proposed to explain how drugs can exert hypersensitivity responses, but the precise physiopathological mechanism is still unknown (Fig. 1). It is possible that more than one pathological mechanism contributes to the final renal damage, which depends on the putative drug [6, 8].

Fig. 1.

DI-AIN pathogenesis. The pathogenesis of DI-AIN is driven by different mechanisms. Haptens: drugs might have the ability to form protein complex termed “haptens” which elicit renal damage after recognition by DCs and subsequent T-cell-mediated toxicity. P-i concept: Some drugs with some special structural features might stimulate T cells allowing a noncovalent bind to the major histocompatibility peptide complex. This method has some similarities to superantigen stimulations. Direct damage: drugs might cause direct tubular damage. Antibodies: even if cytotoxic T-cell injury is most frequently involved in pathogenesis of AIN, there are some cases related to antibody-mediated damage.

One of the most widespread theories to explain the pathogenesis of the disease is the formation of haptens. Some drugs might have the ability to become immunogenic by binding to proteins through covalent links [8-11]. This new complex, termed hapten, has the capacity to act as an immunogenic antigen, and thus, it would be able to stimulate an immune response. This process can occur locally in the kidney, where drugs will bind to renal-specific tubulointerstitial proteins or in the peripheral circulation, that is, beta-lactam antibiotics. If haptenization occurs peripherally, the haptens will be carried out to the kidneys, where they will become “trapped”, and subsequently, filtered and endocytosed by the tubular epithelial cells of the kidney, starting the first phase of the immunogenic process, the antigen-recognition and presentation phase [10].

Tubular epithelial cells will process and metabolize the antigens, and then they will be presented to dendritic cells (DCs). DCs are located in the kidney interstitium, but they are in tight contact with tubular epithelial cells through their long-foot processes. This allows them to detect any problem and to further propagate cell injury [12]. When DCs are exposed to antigens, they become active by expressing the antigenic compounds as peptides on their surface major histocompatibility complex II molecules. Thereafter, DCs migrate from the kidneys to the regional lymph nodes to present the antigen to naïve T cells, which are then activated, and migrate to the antigenic source of injury.

Damaged cells will become activated, and they will generate proinflammatory chemokines and cytokines that will induce the recruitment of macrophages and the proliferation of fibroblasts [13]. Due to the massive proliferation of fibroblasts, there will be an increased amount of extracellular matrix by epithelial-mesenchymal transition process. This reaction is associated with tubular cell damage and kidney damage, which in the long term could lead to fibrosis and chronic kidney disease (CKD) [10, 14, 15].

DI-AIN is mainly a cell-mediated process. This hypothesis is supported by the observation that T cells are the predominant lymphocytes in DI-AIN renal biopsies [8]. Another fundament is the observation of drug-specific T cells in the peripheral blood of patients with biopsy-proven DI-AIN with simultaneous demonstration by immunohistochemistry of these T lymphocytes (either CD4+ or CD8+) in the kidney interstitium on kidney biopsy specimens [9, 16].

Alternatively, it has been described that drugs might directly interact with the immune receptors such as the T-cell receptor or the major histocompatibility complex. This process has been designated as “pharmacological interaction with immune receptors” or the p-i concept. This “pharmacologic” interaction can lead to T-cell activation and expansion, thereby producing an immunogenic response [11, 17].

However, it seems that there might be other relevant pathophysiological mechanisms that contribute to triggering this process. Therefore, several hypotheses have been proposed to explain the pathogenesis of the disease.

It has been hypothesized that drugs might produce a direct toxic damage to interstitial kidney structures that would render them immunogenic. Injured cells will secrete proinflammatory chemokines and cytokines that will attract inflammatory cells and induce the proliferation of fibroblasts that ultimately cause interstitial fibrosis [6, 18]. A second theory proposes that drugs could elicit an immune response by mimicking self-peptides due to structural similarities, thereby producing a cross-reactivity reaction between the tubular basement membrane (TBM) and the interstitial antigen [6, 19].

Some human leukocyte antigen (HLA) genotypes have been found to be associated with AIN. Cheng et al. [20, 21] demonstrated that HLA-DR was expressed in the majority of the patients with DI-AIN included in their study. Moreover, HLA-DR expression correlated to the severity of tubular atrophy and interstitial fibrosis. In addition, there was a positive correlation between the number of DR-positive cells in the interstitium and tubules [20, 21]. Accordingly, other studies have studied the susceptibility to the disease among other HLA genotypes. HLA-DQ seems to be implicated in pathogenesis. According to a single-center prospective cohort, HLA-DQA1, -DQB1, and -DRB1 are risk alleles both for DI-AIN and TINU syndrome. Histologically, this study showed that there were positive correlations between the number of interstitial CD4+ T lymphocytes and the tubular expression of HLA-DR and HLA-DQ. In the same way, the infiltration of monocytes/macrophages into the interstitial compartment and into the renal tubular wall was closely correlated with the tubular expression of HLA-DR and HLA-DQ [22].

Even if the injury in DI-AIN is mainly cell-mediated, the immune reaction in AIN can also be driven by the antibody-mediated immunity in a minor proportion of cases. It has been described in some cases that antigen-antibody complexes can deposit in the kidney, leading to kidney injury [14]. Even so, this is a rare finding, and it is often associated with tubulointerstitial injury. In these patients, IF staining demonstrates linear IgG deposition along the TBM. For example, methicillin is known to induce DI-AIN by the deposition of a granular immune complex along the TBM [23]. Some reports of methicillin-induced AIN have identified antibodies targeting a specific hapten in the TBM. Other drugs such as NSAIDs, phenylhydantoin, and allopurinol might also cause positive IF staining. Rifampicin-induced AIN is associated with the production of anti-rifampicin antibodies. It is likely that in DI-AIN, the same pathological pattern is the product of different triggers, and it is induced by distinct mechanisms as it happens in other kidney disorders.

The duration and severity of the effector phase will depend on the duration of the renal insult. In case of an intense and long exposition to the antigen, tubular epithelial cells will produce inflammatory chemokines and cytokines (TGF-b, IL1, IL4, lipid peroxidation products, ET-1, IGF-1, PDGF-BB) that will elicit activation of endothelial cells, the recruitment of fibroblasts, and the irreversible process of tubulointerstitial fibrosis [15].

Etiology

The most common causes of AIN can be clustered in drug-induced, infections-related, autoimmune diseases and idiopathic forms, being the first one responsible for more than two-thirds of the cases, followed by infection-related AIN in 15% of cases. Etiology of AIN changes with aging since the elderly tend to have more frequent DI-AIN, especially due to PPIs and penicillin, and a lower incidence of AIN due to autoimmune diseases or systemic causes [4].

A large number of drugs can be implicated in the pathogenesis of the disease, and at least from a theoretical point of view, any drug can induce AIN (Table 1). However, NSAIDs, PPIs, and antibiotics continue to be the most common drugs involved in AIN [24].

Table 1.

In the study conducted by Murithi et al. [25], antibiotics were the most prevalent cause (49%), followed by PPIs (14%) and NSAIDs (11%). Valluri et al. [26] reported that antibiotics and PPIs were the most prevalent causes of DI-AIN in their cohort (35%) followed by NSAIDs (20%). The multicenter retrospective study by González et al. [27] showed that antibiotics and NSAIDs were the two primary offending classes of drugs, comprising 93% of the patients. It is possible that the pattern is changing, and recently, Fernández-Juárez et al. [28] reported that the most frequently offending drugs were NSAIDs (27%), followed by antibiotics (22%) and PPIs (4%).

With regard to antimicrobial agents, beta-lactam antibiotics are the most frequently involved drugs, and methicillin is the most well-documented drug. Methicillin is reported to induce AIN in up to 17% of the patients who have received at least 10 days of treatment [29]. It is characteristic that kidney injury appears days after drug intake. Extrarenal manifestations are characterized by hypersensitivity reaction features (skin rash, fever, and eosinophilia), with fever being the most common manifestation, whereas rash and/or arthralgia are less common.

Although NSAIDs can induce kidney injury by several mechanisms, these drugs are well-known causes of DI-AIN and have become one the most frequent causes. Most individuals are older than 50 years and are chronic users. Elderly patients are at a particularly high risk as they commonly take NSAIDs and have underlying comorbidities that increase their overall risk. Although all NSAIDs can be pathogenic, some nonselective NSAIDs are less toxic than others as they produce fewer effects on kidney prostaglandins, that is, low-dose aspirin, low-dose ibuprofen, and sulindac [14, 30, 31]. Clinically, NSAID-induced AIN is characterized by fewer extrarenal symptoms and longer latency periods.

Current use of PPIs is associated with a significantly increased risk of AIN. According to a nested case-control study, the crude incidence rate of PPI-related DI-AIN is estimated in 11.98 (95% CI 9.11–15.47) per 100,000 person-years [32]. PPIs are one of the more widely prescribed medications worldwide. These drugs are often perceived as safe, but the use of PPIs beyond their established indications is dangerous because it can produce a substantial kidney impact. According to some reports, PPIs are the second most frequent cause of DI-AIN, accounting for 14–64% of the cases [33, 34]. Its clinical presentation is nonspecific, and the interval within the drug use and the onset of AIN ranges between 1 week and 9 months. The severity of AKI in PPI-induced AIN may be lower compared to other drugs like antibiotics, but the probability of recovery is also lower [25].

This could be explained by the longer duration of treatment and the consequent greater exposure to the drug. These characteristics draw a kidney injury that is especially challenging in its diagnosis and whose clinical expression would be more compatible with CKD than with AKI episodes. In fact, recent studies have focused on analyzing the effect of PPIs on long-term kidney function, and the results suggested that omeprazole might be associated with an increased risk of CKD [33, 34].

In a large northern European cohort analysis comparing new users of PPIs and histamine H2-receptor antagonists, PPIs users were at increased risk of suffering worse kidney outcomes. Indeed, they were at higher risk of doubling creatinine levels (hazard ratio [HR]: 1.26, 95% CI 1.05–1.51), of decreasing eGFR more than 30% (HR: 1.26; 95% CI 1.16–1.36), and for the development of end-stage kidney disease (HR: 2.4; 95% CI: 0.76–7.58) and AKI (HR: 1.3; 95% CI: 1–1.69). However, the results should be treated with caution as the results have weaknesses that limit conclusions about causality because the strength of the association was modest (HR: 1.26 for both), and there is considerable potential for residual confusion [35, 36].

In the last decade, drugs such as 5-aminosalicylic acid or new therapies against cancer like ICIs have started to play an important role in the pathogenesis of this disease. ICIs, such as nivolumab, ipilimumab, or pembrolizumab, have recently been described as causes of AIN with similar clinical and histologic features to the rest of DI-AIN but with longer latency periods. However, the etiopathogenesis can be due to a change in the immune system that can cause a loss of tolerance against endogenous antigens or that ICI reduces tolerance to other coexisting drugs implicated in AIN. Some articles have shown a close relationship with concomitant use of PPIs [37].

Cortazar et al. [38-40], reported that concomitant use of PPIs was a risk factor for DI-AIN together with dual ICI therapy and baseline eGFR <30 mL/min/1.73 m2. Accordingly, combination treatment with more than one ICI increases the probability to develop DI-AIN, and also the severity of the eventual episode [41-43].

Clinical Manifestations

Symptoms are highly variable, and most of the patients will present mild symptoms. Therefore, diagnosis of DI-AIN relies on maintaining a high index of suspicion in those patients at higher risk for this disease, but, in the end, kidney biopsy is still needed to confirm AIN diagnosis.

In the majority of cases, AIN-related symptoms are nonspecific, including malaise, nausea, and vomiting. However, many patients are asymptomatic. The classical triad, composed of fever, rash, and eosinophilia, is seldom present accounting for 10% of the cases [44]. However, when this entity was initially reported, this triad was more frequently found. Some agents, such as NSAIDs and PPIs, are less commonly associated with fever, rash, or eosinophilia compared with other agents.

Skin eruptions are usually maculopapular or morbilliform, although diffuse erythroderma versus toxic epidermal necrolysis may also develop [2, 45]. On the other hand, fever can be low grade or intermittent but it can be absent [45].

Regarding the interval of time between the start of the drug and the episode of DI-AIN, beta-lactams and cephalosporins usually have a close correlation between drug initiation and the development of AIN. However, PPIs and NSAID-induced kidney injury occur with longer drug exposure [4, 34]. When we analyzed ICI-related DI-AIN, there is a longer latency period from drug administration to diagnosis of DI-AIN compared to other drugs and a slower decrease in creatinine after treatment [3]. Rifampin is unique in that the interstitial nephritis generally occurs when the antibiotic is reintroduced after an interval, and the latent period may be as short as 1 day.

Renal Manifestations

Nonoliguric AKI is present in every DI-AIN case. Hyperkalemic, hyperchloremic metabolic acidosis, inconsistent with the degree of kidney failure, may be present and provides a clue that the tubulointerstitium is the site of kidney injury [46].

Tubular nonnephrotic range proteinuria is usually present in the setting of DI-AIN (90% of the patients). Proteinuria range usually varies from 0.7 to 1.5 g/24 h. Older individuals might be more likely to have a higher degree of proteinuria [47]. The incidence of nephrotic syndrome seems to be low and accounts for around 2.5–3% of the patients [2, 28, 48]. Patients with NSAIDs related to AIN tend to present higher proteinuria levels than other types of drugs inducing AIN, and nephrotic-range proteinuria seems to be a specific complication of NSAIDs. Nephrotic syndrome has been only exceptionally reported with other types of drugs such as methicillin [49]. In this setting, nephrotic proteinuria has been reported in 80% of the patients, more often associated with fenoprofen, naproxen, and ibuprofen or selective COX-2 inhibitors [50]. Nevertheless, in some patients, nephrotic proteinuria might be due to an underlying disease, so its presence might raise the suspicion for a concomitant diabetic kidney disease or another underlying glomerular disease [51]. In these cases, it is highly recommended evaluating previous blood and urine tests.

Despite the foregoing, the absence of nephrotic range proteinuria does not exclude the possibility of NSAID-related AIN because in some published series, a much lower prevalence has been documented. To explain the susceptibility to proteinuria caused by NSAIDs, it has been hypothesized that the underlying pathogenic mechanism would be a decrease in the synthesis of prostaglandins induced by NSAIDs, which could lead in an increased conversion of arachidonic acid to leukotrienes, producing an activation of T-helper cells inducing diffuse podocytes damage [52]. Extrarenal manifestations (low-grade fever, skin rash, and eosinophilia) that are frequently observed in AIN associated with antibiotics and other drugs are characteristically absent in NSAID-induced AIN [52, 53]. Urinalysis will present diverse abnormalities such as sterile pyuria in up to two-thirds of the cases; leukocyte casts (3–14%), and hematuria [40, 54].

In early clinical case reports, in which methicillin was the predominant putative agent, macroscopic hematuria appeared as a common finding of this entity; however, it is currently a very rare initial presentation [55]. While microscopic hematuria is more common, and it is found in nearly two-thirds of the patients, the presence of hematuria is known to be associated with a more severe disease and worse kidney outcomes [56]. The presence of red-blood-cell casts was once thought to be specific for glomerular disease; however, up to one-third of AIN patients exhibit red-blood-cell casts, probably related to disruption of interstitial blood vessels and leakage of erythrocytes to the tubular lumen [57]. Acute tubulointerstitial injury can induce tubular dysfunction, clinically detected as glycosuria. In a study of our group, glycosuria was present in the 21% of biopsy-proven AIN cases, compared to the 4% in those AKI due to other etiologies. The specificity for glycosuria and AIN was 99% (95% CI: 97–99%). Further studies are needed to confirm this finding [58].

Eosinophiluria

The presence of urinary eosinophils may be explained by the release of eosinophil granules after activation, with the potential for leaking into the tubules and being excreted into the urine [59]. Even if eosinophiluria was traditionally considered a classical biomarker in AIN, the presence of eosinophiluria in the urinary sediment is a nonspecific, and unreliable finding since it can be seen in many other kidney diseases such as renal abnormalities, urinary tract infections, acute tubular necrosis, and glomerulonephritis. Muriithi et al. [60] identified 566 patients with CKD caused by different etiologies, including 91 patients with AIN, 80% of them with DI-AIN. They found eosinophiluria in a variety of diagnoses besides AIN. Using a >1% urinary eosinophils cut-off, only 31% of AIN patients were identified, with a similar rate in acute tubular necrosis (29%). When increasing the cut-off up to 5%, sensitivity declined, but specificity improved. However, eosinophiluria was a poor test to discriminate against AIN and other kidney diseases [60]. In another study conducted in 1994 comprising 51 patients with suspected AIN, the sensitivity of eosinophiluria was 40% for the detection of AIN, and the positive predictive value was 38% [24]. Hansel stain has been considered as a better method to visualize urinary eosinophils than Wright’s stain [59].

Sterile Pyuria

Sterile pyuria is a common nonspecific finding (50–75% of the patients), especially in the setting of antibiotics related to AIN.

Urinary Sodium

It is a common belief that urine sodium concentration and fractional excretion of sodium would be low in these patients, but the reality is that it is unhelpful in the diagnosis of AIN as both can be either elevated or depressed [51].

Histology

The definitive diagnosis of AIN is given by the kidney biopsy. However, as previously mentioned, kidney biopsy is not always performed because on many occasions, we are faced with fragile patients, mild increase of serum creatinine, or at high risk of complications of the technique, assuming the diagnosis of this entity.

Histologically, it is characterized by the presence of an extensive interstitial infiltrate, mainly composed of lymphocytes and monocytes, but eosinophils, plasma cells, histiocytes, and polymorphonuclear cells can also be found. When immunohistochemical techniques are performed, the mononuclear component in drug-associated cases is mostly T lymphocytes.

It can also associate images of tubulitis and, accompanying this, typically, there are tubular degeneration signs such as loss of brush border, flattening of the epithelium, and desquamation of cells toward the tubular lumen. Regenerative changes such as mitosis and enlarged nuclei with prominent nucleoli can be found.

Initially, the inflammatory infiltrate can be associated with interstitial edema, but with time, it can develop fibrosis and tubular atrophy, these findings being characteristic of chronic interstitial nephritis. Elderly patients tend to present with a higher percentage of sclerotic glomeruli and more interstitial fibrosis and tubular atrophy [4]. Both the glomerular and vascular compartments are usually normal except when they have some other associated pathology, something that is always important to rule out.

IF does not usually show immune deposits, and fibrinogen deposits can be detected in the interstitium. In those associated with penicillins and cephalosporins, sporadic cases have been described with the presence of linear IgG and C3 in tubular bases. At the electron microscope level, no specific findings are detected.

Granulomatous interstitial nephritis is a different, but less common, form of presentation. It is associated with drugs such as NSAIDs, antimicrobials like vancomycin or ciprofloxacin, and also granulomatous diseases like sarcoidosis or TINU syndrome. Granulomas are mainly composed of macrophages and multinucleated giant cells.

Treatment

There is general agreement that withdrawal of the presumed causative drug is the mainstay of the treatment of DI-AIN. But even when DI-AIN is suspected, identification of the putative drug is challenging for several reasons among others due to the large list of possible putative drugs implicated in DI-AIN, the time from starting the drug to AKI is quite ranged, and the lack of precise screening methods to discriminate the offending drug [3].

This situation is even more complicated with the ever more widespread use of over-the-counter medicines such as NSAIDs or PPIs. Patients usually perceive these drugs as secure with low side-effect profile, and patients usually take them for chronic conditions during months or even years increasing the risk for side adverse effects [34, 61-63]. In these situations, only an in-depth focused interview can uncover the uptake of this kind of drugs. Another difficult situation are those patients treated with polypharmacy, especially those with chronic diseases and the elderly, which are at a particularly high risk for DI-AIN for their inherent fragility, eventual impaired kidney function and other comorbidities [64].

Despite all the efforts to identify the responsible drug, sometimes the culprit drug cannot be withdrawn or substituted due to its inherent benefits. This situation occurs frequently in AIN induced by chemotherapeutic drugs, specifically by ICI.

When there is no evidence of kidney function recovery after a 5–7-day period since interrupting the treatment with the suspected drug, many studies support the treatment with steroids, especially when they are promptly started (Fig. 2). The rationale for this approach is that early corticosteroids would reduce the inflammatory infiltrates of the kidney interstitium, thus preventing the risk of subsequent fibrosis. In moderate-to-severe cases of AIN, the period from the withdrawal of the drug to the start of corticosteroids treatment could even be shorter.

Fig. 2.

Proposed treatment algorithm for DI-AIN. The mainstay for the treatment of DI-AIN is the withdrawal of the suspected medication. We recommend reevaluating serum creatinine after a week. In case there is not a significant improvement, we suggest starting treatment with oral prednisolone 0.8–1 mg/kg during 2 weeks. Once there is a complete renal response, we advocate to initiate steroid tapering, whereas in case of an incomplete response, we suggest to maintain the same dose of prednisone 2 more weeks and then to initiate tapering, which should last 5–7 weeks.

Clarkson and colleagues [48] published a retrospective study on 60 patients diagnosed with AIN (92% DI-AIN), although clinical follow-up data were only available in 42 patients. They found no difference in kidney function at 6 and 12 months of follow-up between patients treated with corticosteroids (60%) and those who merely received supportive treatment (40%) [48]. The Madrid Interstitial Nephritis Group studied 61 patients with kidney-biopsy-proven DI-AIN. Most patients received corticosteroids (85%). After a mean follow-up of 19 months, kidney function showed a greater improvement in those treated (final serum creatinine 2.1 mg/dL (range 0.7–12.7) compared to those not treated (final serum creatinine 3.7 mg/dL (range 0.7–8.9) (p < 0.05), and the number of patients who required kidney replacement therapy was significantly lower (3.8% vs. 44.4%, p value <0.001) [27]. Interestingly, in this study, there was a significant correlation between the delay in the onset of corticosteroid treatment after drug withdrawal and the final kidney function (r = 0.45, p < 0.005). Specifically, the benefit was greatest for patients, who received early treatment with steroids, in particular during the first week after stopping the causative drug.

Later studies showed similar findings. In the study conducted by Muriithi et al. [25], 86% of the patients received steroids, and the rest of patients received conservative treatment. After 6 months, serum creatinine in both groups was similar (1.4 vs. 1.5 mg/dL). An interesting observation in this study is the fact that patients achieving partial or complete remission started corticosteroid treatment earlier (at 8 and 11 days, respectively), whereas patients not attaining remission commenced treatment considerably later (at 35 days) (p value = 0.05) [25].

In view of the published results, the following studies have focused on the question of when it is optimal to introduce corticosteroids. This controversy has been addressed in a study including 182 patients with biopsy-proven cases of DI-AIN treated with corticosteroids. Those patients starting treatment with corticosteroids during the first 2 weeks showed better kidney outcomes. In contrast, treatment started beyond 4 weeks exhibited little benefit on the final kidney function. More in detail, the mean time to start corticosteroid treatment was 9 days in the group that had complete recovery, 12 days in the group that had partial recovery, and 29 days in the group that had no recovery (p value = 0.008) [28].

On the other hand, the length of the therapy is not a trivial matter because the use of corticosteroids might produce a wide range of adverse effects. Therefore, a strict balance of the dose and the duration of the treatment must be considered to avoid administering more doses for longer than necessary. Despite its importance, the duration of the treatment with corticosteroids has not been discussed in depth. The study conducted by Fernandez-Juarez et al. [28], discussed the length of the duration of the treatment, concluding that the greatest kidney function recovery was obtained when the entire treatment lasted 8 weeks at maximum, out of which, 3 weeks would correspond to high-dose corticosteroids therapy followed by a tapering period of up to 5–6 weeks.

In conclusion, treatment of AIN remains a challenge since evidence in the literature comes mainly from retrospective studies. The first step in the treatment of DI-AIN is to remove the causative agent of the disease; however, there is evidence on the use of corticosteroids for the treatment of this entity when started rapidly. Therapy should not exceed a total duration of 8 weeks maximum, of which 3 weeks would correspond to high doses of corticosteroids followed by a tapering period of up to 5–6 weeks (Fig. 3).

Fig. 3.

Relapses. Some of the patients who have suffered an episode of DI-AIN must face a re-exposure to the drug that had previously been suspected as a culprit of the disease. If the patient suffers a new episode of DI-AN, the drug should be withdrawn but if there is no new episode, it is necessary to test other drugs.

The treatment alternatives for these patients are not yet standardized. In fact, there is not enough evidence in the literature that encourages the use of other immunosuppressive agents. Clinical cases treated with other immunosuppressive agents, such as cyclophosphamide, cyclosporine, or mycophenolate, are scarce in the literature [65, 66]. In this setting, a recent report of 8 patients with corticosteroid dependency and recurrent AIN documented an interesting beneficial effect of mycophenolate mofetil.

Prognosis

The initial reports of D-AIN related to methicillin suggested that this entity was a relatively benign disease with a favorable prognosis. Nevertheless, this is far from reality because almost 40% of the patients suffering DI-AIN will experience an incomplete renal recovery with some degree of CKD [15], and around one-third of the patients will require dialysis [45]. The elderly, in addition to being at greater risk for the disease, also have a worse prognosis and usually have more severe kidney failure even requiring hemodialysis [4].

Approximately 2% of all CKD is considered to be due to AIN, which is equivalent to 10 million prevalent worldwide cases. Furthermore, AIN is the primary cause of ESRD in 3–4% of all the cases [67]. It is one of the few potentially treatable causes of AKI if identified and treated early [68]. However, delayed or missed AIN diagnosis leads to ongoing inflammation, resulting in interstitial fibrosis, tubular atrophy, and permanent kidney damage, which may be the explanation for CKD occurring in 40–60% of patients after an episode of AIN [25, 61].

The discovery of a DI-AIN biomarker would be a fundamental finding in order to establish an early and definite diagnosis. Several novel biomarkers have been evaluated as potential diagnostic and prognostic tools. Some studies have highlighted the role of MCP-1, KIM-1, IL-9, and sC5b-9 as possible biomarkers of the disease. Diagnostic techniques have also been developed, such as the lymphocyte transformation test, which measures the proliferation of lymphocytes in the presence of a drug. For example, in one case, a patient developed DI-AIN in the setting of the treatment with nivolumab and lansoprazole [44]. The LTT demonstrated reactivity against lansoprazole and not against nivolumab, indicating which was the culprit drug [24, 59, 60, 69].

Despite different attempts, much remained to be done in order to elucidate the prognostic factors of DI-AIN. So far, any clinical or histological factors have been shown to have an impact on the prognosis. Some authors have tried to correlate the extent of interstitial infiltrates, the severity of tubulitis, the extent of fibrosis, or the presence of granulomas with the outcome of the disease with despair results [6, 15]. The fact that fibrosis and chronic tubular damage is patchy may justify this lack of correlation [25, 70].

Relapses and Refractory Disease

Recurrent AIN is a rare condition, which is associated with worse prognosis and renal survival. In a Spanish study, 22 out of 205 patients diagnosed with DI-AIN presented a clinical relapse after the drug withdrawal. Twenty-seven percent of them had the causative drug reintroduced before the relapse, and 41% of these patients had an undiagnosed autoimmune disease underneath, with a previous wrong diagnosis of DI-AIN. In this study, only the presence of microhematuria, at the time of diagnosis, was more frequently seen in those with an underlying systemic disease.

The treatment landscape for patients with relapses is not standardized. In fact, there is not enough evidence in the literature supporting the use of other immunosuppressive agents than corticosteroids.

In case of refractory disease, it is highly recommended to identify if there is another putative cause, rather than the suspected drug, for example, an underlying disease (Fig. 3). It is also important to mention that when chronic kidney damage is already established and the acute infiltrate has already given way to chronic infiltrate and tubulointerstitial fibrosis, the damage is no longer reversible, and therefore, kidney recovery could not be possible. Thus, corticosteroids should be avoided in those with advanced fibrosis.

Faieta and colleagues [71], reported one case of refractory ICI-induced AIN successfully treated with mycophenolate mofetil. The patient was diagnosed with DI-AIN related to nivolumab, which is a PD-1 inhibitor, in the context of the treatment of a metastatic clear-cell renal-cell carcinoma. The patient was initially treated with corticosteroids for this condition but suffered three relapses. Finally, therapy with pulse 500 mg IV methylprednisolone and 1 daily gram of mycophenolate mofetil was initiated, and the patient recovered kidney function [71].

Conclusion

DI-AIN is an increasingly common entity in clinical practice and an increasing cause of AKI. Its diagnosis remains a challenge for clinicians, due to the nonspecific clinical manifestations and because it can be produced by a myriad of drugs. Certainly, this makes it extremely difficult to identify the causative agent due to polypharmacy in patients and in other cases because it is difficult to control the use of over-the-counter drugs.

So far, the optimal treatment of this entity is not well characterized because there are no clinical trials that provide the highest level of evidence. However, it is considered that the fundamental pillar of treatment consists in the withdrawal of the drug, and if an early recovery of the baseline kidney function is not obtained, treatment with corticosteroids might hasten the recovery of kidney function, and these drugs should be started as soon as possible.

Conflict of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

No funding was required to conduct this review.

Author Contributions

Conceptualization: Gema Fernandez-Juarez and Beatriz Sanchez-Alamo; methodology: Gema Fernandez-Juarez; literature review: Clara Cases-Corona and Beatriz Sanchez-Alamo; writing – original draft: Beatriz Sanchez-Alamo and Clara Cases-Corona; writing – review and editing: Beatriz Sanchez-Alamo and Gema Fernandez-Juarez; visualization: Beatriz Sanchez-Alamo and Gema Fernandez-Juarez; and supervision: Gema Fernandez-Juarez.

Data Availability Statement

All articles analyzed during this review are included in the references. Further inquiries can be directed to the corresponding author.

Copyright: All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher.
Drug Dosage: The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any changes in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug.
Disclaimer: The statements, opinions and data contained in this publication are solely those of the individual authors and contributors and not of the publishers and the editor(s). The appearance of advertisements or/and product references in the publication is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements.