TNFR1, ACR ratios predict which diabetics are at high risk for renal failure

In patients with type 1 and type 2 diabetes, simple measurements of tumor necrosis factor receptor 1 (TNFR1) levels and albumin-to-creatinine ratios (ACR) may help clinicians better assess the risk of end-stage renal disease (ESRD), according to results from researchers from the Joslin Diabetes Center, published online in the journal Kidney International.

Traditionally, the design of phase 3 trials for diabetic nephropathy requires enrollment of patients at a high risk of progression, defined as within 3 years of a hard end point, such as end-stage renal disease, 40% loss of estimated glomerular filtration rate, or death. Clinicians typically used two biomarkers to identify patients at a higher risk of kidney failure as well as to select those to enroll in clinical trials—urinary ACR and estimated glomerular filtration rate. But, according to these criteria, many high-risk patients were missed. In addition, these biomarkers had failed to accurately predict the time of onset of ESRD.

“Overall efficiency and cost effectiveness of clinical trials depends on the diagnostic tools used to enroll study patients,” says senior study author Andrzej S. Krolewski, MD, PhD, Head, Section on Genetics & Epidemiology, Joslin Diabetes Center, and professor of medicine, Harvard Medical School, Boston, MA. “If you recruit people who are not at risk of progressing to ESRD during the clinical trial period, statistical power declines and you can’t prove anything.”

In 2012, Dr. Krolewski and colleagues found a link between TNFR1 levels and declining renal function in individuals with both type 1 and type 2 diabetes. In this study, they sought to turn their discovery into a practical, reliable test to assess care and enroll patients in clinical trials. Using natural history data from the Joslin Kidney Studies of chronic kidney disease in patients with diabetes, they enrolled 279 patients with type 1 diabetes, and 134 end points within 3 years, and a validation cohort of 221 patients with type 2 diabetes and 88 end points. They applied classification and regression trees analysis from among all clinical characteristics and markers to select the best prognostic criterion that divided patients with type 1 diabetes according to risk.

The optimal criterion to select patients according to risk, they discovered, was a TNFR1 level over 4.3 ng/ml alone or 2.9-4.3 ng/ml with an ACR over 1,900 mg/g. This criterion produced similar results in type 1 and type 2 patients, and had high sensitivity and prognostic value (72% and 81%, respectively).

“Remarkably, when we used the TNF receptor to analyze risk of ESRD, the risk was almost identical for both type 1 and type 2 diabetes. This implies that the etiologies are similar,” explained Dr. Krolewski. “This is a very important observation because in the medical community, the impression is that the progression to ESRD in type 1 is somehow different from type 2. As a result, many clinical trials do not include patients with type 1.”

Dr. Krolewski and fellow researchers also applied this criterion to a hypothetical 3-year clinical trial in order to determine its impact on sample size and statistical power.

“Currently, about 80% of patients in these clinical trials provide no useful information. If our criterion is used in the recruitment of patients, you will not need 2,000 or 3,000 patients for a clinical trial, you will only need 400 patients,” concluded Dr. Krolewski.