The development of drug-induced acute interstitial nephritis (AIN) is usually managed by withdrawal of the precipitating agent; however, renal function does not always return to normal afterwards.OBJECTIVE
To determine the effect of steroid treatment on long-term renal function in patients with drug-induced AIN.DESIGN AND INTERVENTION
The medical records of patients with biopsyproven drug-induced AIN who had been admitted to 1 of 10 hospitals in Madrid, Spain during the period 1975-2006 were retrospectively reviewed Suspected systemic disease was an exclusion criterion. The presence of urinary tract obstruction or infection was excluded by radiological examination and urinalysis.OUTCOME MEASURE
The outcome measure was serum creatinine level.RESULTS
The 61 patients included in the analysis had a mean serum creatinine level of 1.1 ± 0.39 mg/dl (97 ± 34 μmol/l; range 0.4-2.3 mg/dl [35-203 μmol/l])at baseline (0.5-16 months before the onset of AIN).The causative agent (an antibiotic in 34 [56%] patients; an NSAID in 23 [37%rsqb; patients; other drugs in 4 [7%] patients) was withdrawn in all cases, and steroid treatment was initiated a mean of 23 ± 17 days (range 2-68 days) afterwards, in 52 patients. The 52 patients who received steroids and the 9 who did not were similar in terms of demographic characteristics, baseline serum creatinine level and cause of interstitial nephritis (i.e. drug type). The most commonly used steroid regimen comprised intravenous pulse methyl-prednisolone 250-500 mg/day for 3-4 days, followed by oral prednisone (starting dose 1 mg/kg/day) for 8-12 weeks. Six months after withdrawal of the causative agent, the steroid-treated group had a significantly lower mean serum creatinine level than the untreated group (2.1 ± 2.1 mg/dl [186 ± 186 μmol/l ] vs 3.7 ± 2.9 mg/dl [327 ± 256 μmol/l ]; P <0.05), and fewer of the steroid-treated patients required chronic dialysis (2 [3.8%] vs 4 [44.4%]; P <0.001).The steroid-treated group was then subdivided into patients who experienced complete renal recovery (n =28)and those who experienced incomplete renal recovery (i.e. serum creatinine level remained ≥25% higher than baseline at 6 months after withdrawal of the causative agent; n =24).There was no significant difference in the duration of steroid treatment between the subgroups, but the patients who experienced incomplete renal recovery had a significantly longer interval between withdrawal of the causative agent and start of steroid treatment (34 ± 17 days vs 13 ± 10 days; P <0.0001). Multivariate regression analysis revealed that an interval of >7 days was associated with a more than 6 times increased risk of experiencing incomplete recovery of renal function (odds ratio 6.6, 95% CI 1.3-33.6; P =0.02). No steroid-related adverse events occurred.CONCLUSION
Steroid treatment seems to be associated with improved renal outcome in patients with drug-induced AIN, especially when initiated soon after withdrawal of the offending agent.