Lupus Nephritis: "Ur-ine" Trouble

I spent my afternoon wearing amber coloured glasses in the lupus nephritis abstract session. Michelle Petri presented data from her Hopkins lupus cohort (Abstract #963) bringing the topic of homocysteine back to the forefront. She showed that despite meta-analyses suggesting limited support for an association between homocysteine levels and cardiovascular events in the lupus and general populations, there are important pro-inflammatory roles of homocysteine in areas such as the kidney that must be considered. She also reminded the audience that several studies have shown a link between homocysteine and venous thrombosis and stroke.

To that end, she described a study of 829 Hopkins Lupus cohort patients that investigated associations between homocysteine and the urine protein:creatinine ratio using serial tests of homocysteine. She found that in males and those of African American descent, homocysteine was strongly associated with nephritis and renal damage. She also demonstrated that a 10 micromol/L decrease in homocysteine was associated with lower urine protein:creatinine. While she acknowledged that there is no coverage for medications (e.g. folic acid, etc) to treat homocysteine in the United States, she felt justified to conduct the study and provide this medication less for the kidneys and more for the additional benefit of reducing thrombosis risk in her lupus cohort. In the end, this study was "interesting" but not strong enough to change management in our lupus nephritis patients. Moreover, longitudinal data is lacking on the long-term outcome of measuring and treating homocysteine levels in lupus patients. I suspect this will be forthcoming with the Hopkins cohort.

Moving from blood to urine, there was a fascinating presentation describing 6 candidate biomarkers to predict treatment response in lupus nephritis in childhood onset SLE (Gulati, Abstract #966). These biomarkers (ceruloplasmin, neutrophil gelatinase associated lipocalcin, kidney injury molecule 1, adiponectin, monocytes chemotactic protein 1, and hemopexin) were measured serially while patients received standard immunosuppressive therapies. Treatment non-responders and responders were evaluated at 6 months. Biomarkers were lower for the responder group and this was significant at 3 months, which predicted response at 6 months . However, at 12 months, only the biomarkers NGAL and HPX remained significant. The area under the curve (ROC) (closer to 1 is good for a test) were impressive for the 6 biomarkers (ROC 0.72 at 0 months, 0.92 at 3 months and 0.84 at 6 months).

These two abstract presentations reflect the current "raison d'etre" in rheumatology – to find a cheap, easy biomarker which may have a role in predicting disease prognosis and/or treatment response. Multiple other candidate biomarkers are also in development, and it will be interesting to see which ones make the cut.

More reports from ACR 2016