Skip to main content
Premium Trial:

Request an Annual Quote

Columbia Team IDs Two Protein Markers to Improve Prediction of Acute Kidney Injury, Adverse Effects


By Molika Ashford

A combination of protein biomarkers could improve the prediction of acute kidney injury and could also serve as prognostic markers for the risk of severe injury and death in AKI patients, according to an international, multi-center study led by researchers at Columbia University.

The group assessed five previously identified proteins in a prospective observational study, and found that two of the urinary markers— urinary neutrophil gelatinase-associated lipocalin, or uNGAL, and kidney injury molecule-1, or uKIM-1 — were the most useful in diagnosing AKI and were also predictive of the severity and duration of complications. They published their results this week in the online edition of the Journal of the American College of Cardiology.

The Columbia team discovered uNGAL in earlier research and has since licensed the protein to Abbott Laboratories. The researchers also used the company's Architect-NGAL assay, available commercially outside the US, to analyze samples in the study.

KIM-1 is a discovery of another group from Harvard, which licensed it to Johnson and Johnson and Biogen Idec.

While both of these markers, as well as three other proteins that the group assessed, have already been evaluated in smaller single-center trials, the Columbia group wanted to test them side-by-side to establish which, if any, were the most predictive, said Jonathan Barasch, the study's senior author.

Barasch told ProteMonitor that the standard method of measuring serum creatinine and urinary output to diagnose AKI is not an ideal method because AKI patients may not show an increase in creatinine upon admission, or may show low levels despite injury to the nephrons.

"Creatinine is measured in the blood and is the mainstay of understanding how the kidney works," he said. "But the issue is that it comes from muscle, and if you have a decrease in function it could take many hours typically, or even a few days, to accumulate in the blood to a level that would cause alarm bells to go off."

Because of this, creatinine can be normal despite kidney dysfunction, and high levels can also be present without nephron damage, Barasch said, due to altered hemodynamics or chronic kidney disease.

"How a physician takes care of these different patients is different, so you really want to be able to distinguish." For this reason, urinary biomarkers have been viewed as a promising alternative.

Barasch and his colleagues set out in their latest study to measure how five of the most promising markers — uNGAL, uKIM-1, uIL-18, uL-FABP, and uCysC — behaved in a large heterogeneous population from three separate hospital sites in the US and Germany, and to establish which marker or markers were best at discriminating AKI from other states.

The group identified uNGAL in earlier cell growth research and found through subsequent studies that the protein was elevated after kidney injury.

"The long and short of it is that the protein is normally at low expression – when a patient or animal is healthy or simply volume depleted … but [in patients] who have more long-lasting damage to the kidney due to ischemia, hypoxia, or in the setting of sepsis or certain medications, then this protein increases substantially," Barasch explained.

"On an RNA level, in the kidney, you can have a 1,000-fold increase. On a protein level in the urinary system, [it] can advance as much as 100-fold," he said.

In addition, the researchers discovered that uNGAL was both dose dependent – rising in concentration with greater kidney damage — and appeared early on in the injury, as soon as three to six hours after a damaging stimulus, according to Barasch.

In the study published this week, the team enrolled 1,635 adult patients, of which 1,234 were assignable to one of four diagnostic categories: normal kidney function, stable chronic kidney disease, pre-renal AKI (altered hemodynamics or volume depletion), and intrinsic AKI. The team categorized patients into these four diagnostic categories, and then un-blinded the data to compare these diagnoses with the biomarker measurements.

The researchers measured four of the five proteins using the Abbott Architect, and one, uL-FABP, using an ELISA.

Overall, uNGAL was the most accurate of the five markers in predicting AKI, Barasch said. The team found that uNGAL discriminated AKI with an area under the receiver operating curve of 0.81, which was significantly greater than that of the other four markers.

KIM-1 and uL-FABP had AUCs of 0.71 and 0.70, respectively, and were classified as "fair" distinguishers in the study. The other two markers, uCysC and uIL-18, showed relatively poor accuracy, the researchers reported. Other measures of how the biomarker levels paralleled severity and duration of AKI also indicated that uNGAL was the "most powerful indicator," the group wrote.

Barasch said that, based on the study results, "all the biomarkers could make a potential contribution to an immediate spot diagnosis when patients present in the emergency room, but the best was NGAL in terms of accuracy, and also … in terms of dose-responsiveness."

"It also was very good at distinguishing whether the patients' renal dysfunction would last up to 72 hours or beyond, much better than the other markers," he said.

Barasch said that one other benefit of NGAL is its relative abundance compared to the other proteins the team interrogated — in the “high-nanogram per ml range” compared to the picogram range for the others.

“In the case of NGAL, the most important assay is just a Western blot, and the reason is we can actually see the protein and can be sure we are measuring the protein we say we are.

The group partnered with Abbott for the trial and the company performed analysis of several of the proteins using its Architect platform. Abbott markets an uNGAL assay outside the US, which the group measured against Western blot analysis. The Abbott assay closely mirrored Western blot in accuracy, the researchers reported.

In predicting prognosis, both uNGAL and uKIM-1, added to a baseline predictive model, independently increased the ability to predict clinical outcomes of death or dialysis during hospitalization. However, there was no synergy between the two; using both together increased the prediction only slightly, the team reported.

"If either of [the two markers] were elevated, the patient would have a higher likelihood of needing dialysis or dying in the seven days after admission," Barasch said. "But you'll see in the paper that when we added [them] together, while you'd expect them to synergize their power of diagnosis because they derive from different parts of the kidney, we didn't see that."

Further analysis by one of the study’s authors, Kai Schmidt-Ott of the Max Delbruck Center for Molecular Medicine in Berlin, showed that approximately 15 percent of the study population had low serum creatinine levels but high biomarker levels, "placing them at low risk by conventional stratification, but at an increased risk on application of biomarker-aided stratification," the authors wrote.

This highlighted the potential utility of uNGAL and uKIM-1 in identifying patients “invisible” to standard measures in the emergency setting, Barasch said.

“Kai demonstrated that if KIM-1 or uNGAL were both turned on and creatinine was high, risk was up to 15 percent,” he said.

“In a busy ER, these patients could be otherwise silent to physicians, but the combination [of measurements] picks out a really robust subset that needs to be treated.”

What this treatment would entail could depend on which biomarker was elevated, according to Barasch. “We know with uNGAL that sepsis, low blood flow to the kidney, or obstruction downstream are the key elements that cause it to go way up, so if you spotted those patients sitting in triage you would act immediately on those three things.”

However, he said, establishing treatment protocols would require research beyond the team’s observational trial to track the impact of proteomically informed intervention on patient outcome.

“What we're talking about requires an intervention study,” and the team is hoping to set up such a trial, evaluating the influence of a “rapid response team” on the outcome of patients with high uNGAL, Barasch said.

“This is 100 percent the most important next step clinically,” he said.

Have topics you'd like to see covered in ProteoMonitor? Contact the editor at mashford [at] genomeweb [.] com.

The Scan

Harvard Team Report One-Time Base Editing Treatment for Motor Neuron Disease in Mice

A base-editing approach restored SMN levels and improved motor function in a mouse model of spinal muscular atrophy, a new Science paper reports.

International Team Examines History of North American Horses

Genetic and other analyses presented in Science find that horses spread to the northern Rockies and Great Plains by the first half of the 17th century.

New Study Examines Genetic Dominance Within UK Biobank

Researchers analyze instances of genetic dominance within UK Biobank data, as they report in Science.

Cell Signaling Pathway Identified as Metastasis Suppressor

A new study in Nature homes in on the STING pathway as a suppressor of metastasis in a mouse model of lung cancer.