Monte Minute‎ > ‎Monte Minute‎ > ‎

Acute Interstitial Nephritis - 1/30/2013

posted Jan 30, 2013, 7:59 AM by Rohit Das   [ updated Jan 30, 2013, 10:21 AM by Purnema Madahar ]
Acute interstitial nephritis (AIN), a very under-appreciated cause of acute renal failure, was the topic of the day. It is difficult to get a good incidence number on AIN, mainly because it is underdiagnosed, but estimates are anywhere between 5 to 20% of cases of AKI. Importantly, AIN is a potentially treatable cause of renal failure (…though the efficacy of that treatment is questionable, which we’ll review), making it that much more important to consider.

Today, we were presented a case of an older gentleman who presented with severe sepsis, and was found to have Staphylococcal auerus endocarditis, sensitive to Methicillin. In the context of being treated with a course of Oxacillin for about 2 weeks, he subsequently developed acute renal failure, with a creatinine rise of 0.6 to 4.9 over 48 hours. This occurred in the absence of previous chronic kidney disease or signs of glomerular pathology. A kidney gallium scan showed increased bilateral uptake, and thus, AIN popped its head in the chart, re-invigorating my clinical curiosity…

·         What’s the pathology behind AIN, and what are the most common etiologies?

·         How does it present clinically, and what are the frequency of important symptoms and signs? How may one make this diagnosis?

·         What’s the prognosis and treatment of AIN?

What’s the pathology behind AIN, and what are the most common etiologies?

·         AIN is believed to be, and fairly confidently among renal gurus, an immune-mediated process reacting to exogenous antigens (or endogenous antigens in idiopathic cases, which I won’t go into) present in the tubular interstitium. The source of these antigens is most commonly various drugs, but they also may come from specific infectious pathogens or systemic autoimmune diseases. It’s thought that T-suppressor function normally quiets this response, and those who develop AIN seem to have an unclear, genetic predisposition hindering the effectiveness of this protective, the end result – an interstitial, inflammatory infiltrate filled with gunk that makes our kidneys hurt.

·         Table 1 gives a nice synopsis of the most common etiologies. By far, the majority of cases are due to drugs (>75%), and specifically, antibiotics (and even more specifically, B-lactam antibiotics). In fact, initial descriptions of AIN in the literature occurred with methicillin, which was VERY common, occurring in nearly 20% patients. As mentioned, particular infections (5-10%) as well as systemic autoimmune disease (10-15%) also cause AIN – for the inquisitive folks, please give Table 1 a glance.


How does it present clinically, and what are the frequency of important symptoms and signs? How may one make this diagnosis?

·         Generally, second exposures to a drug lead to a shorter latent period before onset of AIN (a mean of 3-5 days), but first exposures can lead to a longer period, with a median of 8-10 days. This is a key clinical point in trying to tease out which drug may be the offending agent when patients are on severable potential conspirators…

·         The classic triad of drug-induced AIN is fever, maculopapular rash and peripheral eosinophilia…but it is actually not so classic, as the three together occur in <10% of patients. Taken individually, these symptoms and signs are STILL not helpful – fever occurs in 36%, rash occurs in 22% and eosinophilia occurs in 35%. So, the MAJORITY of patients will actually NOT have the classic symptoms and signs we commonly associate with AIN.

·         Other important numbers include a lack of gross hematuria (5%) and nephrotic range proteinuria (2%). An important positive finding is leukocyuria (i.e., white cells or, more rarely, WBC casts), which occurs in nearly 80% of patients. Eosinophiluria is of limited utility, with a sensitivity of around 60-70%, though better specificity of around 80-85%.

·         An interesting modality for diagnosis is gallium scanning; Ga3+ localizes to areas of inflammation and/or infection, and initial reports 2-3 decades ago showed that AIN leads to marked uptake of gallium and positive scans. Subsequent studies have shown that this test has decent sensitivity (70-80%), but a low specificity, since positive scans can result from other processes, like GN, pyelonephritis…etc…overall, not a great test.

·         All this taken together – short of a biopsy, AIN is a very difficult diagnose to make. The ideal circumstance is to have the characteristic urinalysis with an appropriate offending agent, but very often, the ideal world eludes us. Most of the time, the diagnosis is made in retrospect, when renal failure improves after stopping the offending agent. Biopsy can be considered when the urinalysis doesn’t quite match the clinical picture (i.e., the drug isn’t known to cause AIN), for patients who have severe AKI and the idea of glucocorticoids are thrown onto the table, and when presumed AIN does not improve after cessation of the thought-to-be culprit drug.


What’s the prognosis and treatment of AIN?

·         In initial reports with methicillin, AIN was thought to be a benign condition, with kidney function returning to normal after discontinuation of methicillin in the large majority of cases. However, further studies have clearly shown that AIN is far from a benign process, and 40-50% of cases have an elevated creatinine after 1-2 months of discontinuing the drug. The risk of acquiring “chronic” interstitial nephritis seems to be most closely related to the length of initial acute renal failure prior to diagnosis, with a 3 week cutoff being particularly predictive. Other predictive factors include AIN associated with NSAID use and certain histologic findings on biopsy.

·         Given its immune-related etiology, several series (all small, non-randomized, retrospective studies) have been published looking at the utility of glucocorticoids in preventing progression to chronic kidney disease. The data for the efficacy of this is extremely controversial. In the largest retrospective study I came across (60 patients), >90% of cases were drug-related and there was no difference in creatinine at 1, 6 and 12 months of follow up between patients who received and did not receive glucocorticoids. However, steroids were started only after renal biopsy, leading to longer intervals between onset of renal failure and initiation of treatment. Other studies have shown that steroid initiation at a sooner interval (ideally <7 days) may significantly improve outcomes…

·         The bottom line…who knows? The data seems to collectively suggest that starting steroids sooner rather than later is ideal, but given the nature of these studies, it is impossible to get an accurate assessment of its true efficacy. Generally, prednisone is given at a dose of 1mg/kg per day for 1-2 weeks, followed by a taper.


Couple of solid review articles attached, along with the above mentioned retrospective study. I hope this edition has somewhat satiated those who have remained anxious for the daily the last month or so…certainly more to come… 

Acute Interstitial Nephritis
Praga et. al., Kidney Int 2010, Volume 77: 956-961

Rossert et. al., Kidney Int 2001, Volume 60: 804-817

Clarkson et. al., Neph Dial Trans 2004, Volume 19: 2778-83