The field of nephrology can benefit from advances in noninvasive diagnostic tools (1). Although the use of careful history, physical examination, and urine microscopy is essential in the evaluation of patients with acute kidney injury (AKI) (2), we often need a kidney biopsy to confirm the diagnosis of entities such as acute interstitial nephritis (AIN). This diagnosis of AIN can often be challenging, with available, noninvasive tests suffering from poor accuracy. To improve the diagnostic dilemma, Moledina et al. (3) provided a potential key step forward in establishing a diagnostic test for AIN. Using aptamer-based urine proteomics, they demonstrated that a urine measurement of chemokine C-X-C motif ligand 9 (CXCL9) can be used to identify patients with AIN. CXCL9 is a chemokine induced mainly by interferon production, resulting in predominantly lymphocyte infiltration to the local tissue (4). Using the discovery and the validation cohorts composed of patients with AKI from AIN and non-AIN, Moledina and colleagues (3) showed that CXCL9 is uniquely elevated in the urine of patients with tubulointerstitial inflammation rather than glomerular involvement, making it a potentially valuable biomarker for AIN (5).
The study offered specific findings regarding CXCL9 (Figure 1). First, urinary CXCL9 offered the best accuracy among 180 candidate proteins identified using an unbiased proteomic analysis of the urine, with levels 7.6 times higher in AIN compared with the control group. Second, 31 out of 204 consecutive biopsies done for AKI were found to have AIN. Urinary CXCL9, as part of a sandwich immunoassay, was discriminatory, with levels eight times higher compared with acute tubular injury and 5.5-fold higher in those who had other causes of AKI. Third, the association of urinary CXCL9 and AIN was tested using a logistic regression model. The highest quartile was not only six times higher in AIN but also affirmed the previously validated diagnostic model for AIN. Fourth, urinary CXCL9 was not only superior to previously identified urinary biomarkers tumor necrosis factor α (TNF-α) and interleukin 9 (IL-9) but when combined, improved the predictability of AIN. Fifth, elevated urinary CXCL9 was associated with increased CXCL9 mRNA expression in kidney biopsy tissue. Sixth, the association between urinary CXCL9 and AIN remained consistent in two external cohorts. The study suggested that a urinary CXCL9:urinary creatinine ratio below 14.2 ng/g could be used to rule out AIN, a urinary CXCL9:urinary creatinine ratio above 58.9 ng/g could be used to rule in AIN, and biopsy could be considered between 14.2 and 58.9 ng/g.
The findings by Moledina et al. (3) could help to facilitate the use of a noninvasive urine test measuring CXCL9. Advantages of this new test would also include avoiding steroids and related side-effects as well as preventing discontinuation of critical medications when AIN is ruled out. If the sensitivity of this test is confirmed by future studies, a kidney biopsy may not be necessary for diagnosis of AIN.
The findings by Moledina et al. could help to facilitate the use of a noninvasive urine test measuring CXCL9.
Studies with a larger sample size are needed now to confirm the diagnostic capacity of CXCL9 in AIN in general and in the setting of various causes of AIN.