Improving Chronic Kidney Disease Risk Prediction
Biomarkers Highlight Link to Cardiovascular Disease
By Bill Malone
Chronic kidney disease (CKD) sharply raises the risk for cardiovascular disease (CVD), with annual CVD-related mortality increased up to 100-fold in individuals with kidney failure. However, most primary care practices screen fewer than 20% of their diabetic Medicare patients for CKD, and patients are often referred too late to a nephrologist. Underscoring the connection between CKD and CVD, a recent observational study evaluated a panel of CVD-related biomarkers that were able to predict CKD up to 10 years before clinical onset. This issue of Strategies examines those findings.
Although millions of Americans have undiagnosed CKD, early detection remains challenging because the primary biomarker, estimated glomerular filtration rate (eGFR), is only reliable at <60 ml/min/1.73 m2, when most individuals have already lost nearly half their kidney function. CKD’s link to CVD makes for a dangerous combination, with the risk of CVD increased up to 2 times when creatinine levels are >1.4 - 1.5 mg/dL and up to 3.5 times with microalbuminuria.
A team of researchers from the National Heart, Lung, and Blood Institute’s Framingham Heart Study sought to elucidate this relationship. The investigators compared a multi-marker panel of seven circulating biomarkers to incidence of CKD and microalbuminuria in 2,345 participants of the Framingham Offspring Study, a long-term follow-up study of heart disease risk factors and outcomes (J Am Soc Nephrol 2010;21:2017–2019). Among C-reactive protein, renin, plasminogen-activator inhibitor type 1, fibrinogen, B-type natriuretic peptide (BNP), aldosterone, and homocysteine, the latter two biomarkers were significantly associated with CKD incidence. Log-transformed aldosterone, BNP, and homocysteine were also significantly associated with incident microalbuminuria. “This study is a proof of principle that certain biomarkers which are abnormal earlier in the disease process can help target individuals at risk for CKD,” said lead author Caroline Fox, MD, MPH, medical officer for the Framingham Heart Study, and assistant clinical professor of medicine at Harvard Medical School.
All participants in the study had normal kidney function when blood samples were taken 1995–1998. An average of 9.5 years later, 9% had developed CKD and 8% had microalbuminuria at follow-up. The researchers tested the stored blood to see if the biomarker panel could predict which patients were most likely to develop CKD. The initial analytic step was to evaluate the entire panel of biomarkers in association with CKD, then eliminate them backwards to see which remained significant.
Though BNP, aldosterone, and homocysteine are fairly common and have been evaluated in association with various disease states, this is the first time they have been shown to possess incremental predictive utility for CKD, Fox noted. “Homocysteine has been looked at in multiple settings, including reducing homocysteine as a means of reducing risk, but it hasn’t been borne out in clinical trials. However, homocysteine is very strong in our data, so it may be that it is just an alternative marker of filtration and not the causal pathway,” she said. Aldosterone is associated with inflammation and fibrosis, the authors noted, while BNP may have a unique association in the renal system: “The natriuretic peptides have been implicated in renal disease in several experimental studies where they seem to protect against mesangial fibrosis, glomerular hypertrophy, and glomerular hyperfiltration. Therefore, higher BNP levels may also reflect compensatory responses to limit subclinical renal disease.”
These biomarkers may not yet be ready for clinical use, but the study represents a step forward in connecting what’s known about CVD risk to the progression of CKD, all the more notable in that the results come out of the landmark Framingham Heart Study, according to Glen Hortin, MD, PhD, medical director of Quest Diagnostics. The findings make sense in light of how vascular injury is understood to affect the kidneys. “Microalbuminuria has often been considered to reflect low-level endothelial damage, so it’s not surprising that any process that damages the blood vessels of the kidney would lead to chronic kidney disease. There are other independent factors involved in kidney disease as well, but I think the vascular component is certainly a prime factor,” he said. Hortin also serves on the National Kidney Disease Education Program’s Laboratory Working Group.
Hortin predicts that improvements in the eGFR equation will bear fruit long before researchers validate newer biomarkers for clinical use. For example, he believes aldosterone would be difficult to use widely for screening due to its sensitivity to a patient’s diet. “Instead of focusing on prediction, I think earlier detection will become clinically important first. With an improved equation, we could identify individuals once eGFR starts falling below 90 ml/min/1.73 m2 rather than 60 ml/min/1.73 m2 and start implementing treatment earlier,” he said. “We have made a lot of progress in the last ten years through a variety of efforts, including standardizing creatinine, validating the equations, and understanding the usefulness of these measures in detecting CKD.”
Before such a biomarker panel could be used as a screening tool, it would need to be replicated externally in participants of varying ethnicities, as the sample in the current study was all white, emphasized Fox, especially since the risk of kidney disease is not uniform. African Americans carry almost 4 times the risk of whites; Native Americans 2 times, and Asians 1.3 times. The panel would also need to be tested for its positive and negative predictive value in a clinical trial setting and its cost-effectiveness studied.
Hortin expects the eGFR equation developed by investigators in the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI), a research group established by the National Institute of Diabetes and Digestive and Kidney, to become the next step in early detection of CKD (Ann Intern Med 2009;150:604-61). According to the CKD-EPI researchers, despite their study being limited by having few subjects older than age 70 or racial or ethnic minorities, their equation is more accurate than previous estimates for GFR values above 60, and could replace the current Modification of Diet in Renal Disease (MDRD) Study equation.
Join us for a discussion of this article, and interact with your peers, on LinkedIn.