Identification of the Causative Drug in Drug-induced Acute Interstitial Nephritis

Acute interstitial nephritis (AIN) is a nosological entity characterized by the presence of an inflammatory infiltrate and edema in the renal interstitium responsible for acute renal failure. AIN represent 1-3% of all renal biopsies in some studies and up to 20% of diagnosis in acute renal failure. There are several etiologies of AIN: drug-induced (DAIN), immunological, infectious and idiopathic. In a North American single-center retrospective study of 133 patients with histologically proven AIN between 1993 and 2001, the most common cause was drug-induced (70%). The involved drugs were mainly antibiotics (49%), proton pump inhibitors (14%) and non-steroidal anti-inflammatory drugs (11%). More than 250 drugs can cause DAIN and it is established that any drug can be responsible for DAIN.

The clinical presentation of DAIN is aspecific and variable depending on the drug class involved. It usually occurs 7 to 10 days after the introduction of the causative drug, but this delay may be one day with some antibiotics or several months with non-steroidal anti-inflammatory drugs or proton pump inhibitors. The only consistent manifestation is acute or subacute renal failure, requiring extra-renal epuration in 40% of cases. Clinically, patients may be asymptomatic or present aspecific clinical manifestations such as arthralgias (45%), fever (36%), and skin rash (22%). Biological examinations systematically reveal elevated creatinine levels and, to varying degrees, eosinophilia (35%), microhematuria (67%), leukocyturia (82%) and non-nephrotic proteinuria (93%). The classic triad of fever, rash and eosinophilia is highly suggestive of DAIN and type IV hypersensitivity reaction but is present in only 10-15% of cases. The diagnosis of DAIN should therefore be systematically evoked in the presence of unexplained acute renal failure associated with recent drug exposure.

The pathophysiology of DAIN remains poorly understood. It would involve a drug hypersensitivity reaction. This pathophysiological hypothesis is based on several observations:

(1) DAIN is an idiosyncratic and non-dose-dependent pathology, (2) the presence of eosinophils in the inflammatory infiltrate is frequent (3) it may be accompanied by a systemic hypersensitivity reaction (skin rash, liver damage, eosinophilia), (4) there is usually a 7-10 day interval between the onset of drug exposure and the onset of acute renal failure. Delayed hypersensitivity (Gell and Combs type IV) in which T cells play a central role is the current hypothesis in the pathophysiology of DAIN. Moreover, the preponderance of T lymphocytes in the interstitial inflammatory infiltrate, the absence of immunoglobulin and complement deposits in direct immunofluorescence in the majority of cases also pleads in favour of a delayed hypersensitivity of type IV mediated by T lymphocytes.

The kidney is a target organ for delayed hypersensitivity for 2 main reasons. First, blood flow is high in the kidneys where potential antigens are filtered, secreted, metabolized and/or concentrated. On the other hand, the kidney plays a central role in the excretion of most drugs. These 2 mechanisms explain the particular susceptibility of the kidneys to delayed hypersensitivity due to the exposure of the renal parenchyma to numerous antigens from pharmacological agents and their metabolites.

The diagnostic and therapeutic management of DAIN is currently not codified. It is based first of all on the identification of the causal drug. Discontinuation of the causative drug and its definitive contraindication are essential. If the causative drug is continued or reintroduced, patients with DAIN are at high risk for severe hypersensitivity events that can lead to end-stage renal disease and death. The precise identification of the causative drug still remains a real challenge since in the majority of cases several drugs are suspected (often elderly patients with polymedication). However, this identification is essential in order to stop the allergic process by avoiding the allergen and to counter-indicate the causal drug for life in order to avoid reintroduction, which would expose the patient to a potentially lethal severe hypersensitivity reaction. Polymedication in patients with DAIN leads in about 30% of cases to the discontinuation and contraindication of several drugs, at the risk of having a negative impact on the management of pathologies or co-morbidities.

In vitro allergological tests have been developed to identify the causal drug in immuno-allergic drug toxidermia related to type IV delayed hypersensitivity. This is mainly the lymphocyte transformation test (LTT), which quantifies lymphocyte proliferation after exposure to the allergen in comparison with a control. The LTT has the advantage of being able to be performed at a distance from the allergic episode due to the persistence of memory T cells specific to the allergenic drug, reported for example up to 12 years after an episode of severe toxidermia. The ELISpot is also an in vitro test based on an ELISA-type enzyme-linked immunosorbent assay that measures the number of cytokine-secreting cells (interferon gamma, IL-4, IL-5) among T lymphocytes exposed to allergens. Finally, the measurement of CD154 expression, a marker expressed early on the membrane surface of activated CD4+ T lymphocytes, can also provide interesting information on lymphocyte activation induced by an allergen. The sensitivity of in vitro tests for the identification of the causative drug during DAIN is not known. Several publications emphasize the value of in vitro tests to identify the causal drug among several treatments. In the rare observations of DAIN with identification of the drug by in vitro tests, these were performed in the weeks up to 12 years after the initial episode. In vitro tests have the double advantage of being performed on a simple venous sample during the follow-up of patients without the need for a specialized allergology consultation and do not present any other risk (notably no risk of recurrence) than that associated with a blood sample.