Banking on biomarkers
Risk scores for assessing the likelihood of people developing incident heart failure date back more than a decade. More recent efforts have focused on patients with type 2 diabetes, starting with scores that relied entirely on clinical markers of risk such as prior heart failure, established coronary artery disease, and chronic kidney disease. Reports of two of these validated scores appeared in 2019, one from a team led by Dr. Berg and associates in 2019, and a second score developed by Dr. Pandey and associates.
More recently, both research teams behind these two scores validated newer versions that further refined assessment of patients with diabetes by including biomarkers of incipient heart failure, such as N-terminal of the prohormone brain natriuretic peptide (NT-proBNP). The UT Southwestern group’s biomarker-based score relies on levels of NT-proBNP as well as on levels of high sensitivity troponin T (hsTnT) and C-reactive protein, plus ECG-based assessment of left ventricular hypertrophy to assess risk for incident heart failure. Developers reported in 2021 that this biomarker score could account for 74% (C-statistic) of the 5-year risk for heart failure among patients with diabetes.
The biomarker-based score devised by Dr. Berg and associates, relies on NT-proBNP, hsTnT, and a history of heart failure to predict the risk for a future hospitalization for heart failure. They reported in Diabetes Care that in validation testing this score accounted for 84% of the risk.
“I’m hopeful that both our original clinically-based risk score and our new biomarker-based score will be endorsed by professional society guidelines. The intent of the biomarker-based score is not to replace the clinical one,” Dr. Berg stressed in an interview. But he acknowledged that it uses biomarker values that currently are not routinely collected in U.S. practice. Biomarkers like NT-proBNP “are highly associated with future heart failure risk, but are not yet routinely assessed,” he said. Because of this, “widespread adoption of the [biomarker] risk tool will require some education.”
It may also require some sort of preliminary screening to determine the appropriateness of using it in a specific patient because of the relative expense of a test for NT-proBNP.
A Texas two-step process
“We can’t perform a [NT-proBNP] test on every patient with type 2 diabetes because cost is a huge barrier,” with a U.K. price of roughly £28 (about $40) per test, commented Naveed Sattar, MD, PhD, professor of metabolic medicine at the University of Glasgow. “NT-proBNP is the best biomarker by far to predict risk” for heart failure,” but “it’s too expensive. It’s not going to happen in everyone,” he said in an interview. He suggested taking a two-step approach to identify patients to test for NT-proBNP based on clinical measures like blood pressure, weight and height, lipid levels and renal function and the presence of suggestive symptoms like dyspnea, fatigue, and peripheral edema, an argument he recently spelled out in detail in an editorial he coauthored.
“More work is needed to define which patients would usefully have cardiac biomarkers measured,” Dr. Sattar wrote with his associate.
Two-step is the approach used in routine practice by clinicians at UT Southwestern Medical Center. “We screen all patients with type 2 diabetes and no diagnosed heart failure who are not already on an SGLT2 inhibitor” using their 2019 screening tool, called the WATCH-DM Risk score, said Dr. Pandey. Patients flagged at high risk by their clinical score receive an SGLT2 inhibitor (presuming no contraindications). The remaining patients with low or intermediate risk may then undergo biomarker-based assessment to find additional patients who warrant SGLT2 inhibitor treatment, he said in an interview.
Often, a record of the most important biomarker, NT-proBNP, is already in the patient’s record and less than a year old, in which case clinicians use that value. An NT-proBNP level of at least 125 pg/mL indicates increased risk in people with a body mass index of less than 30 kg/m2, while for those with higher body mass indexes clinicians at Southwestern apply a threshold for higher risk of at least 100 pg/mL.
In addition to starting those patients on an SGLT2 inhibitor, the Southwestern protocol calls for intensified efforts at weight loss and improved fitness to further lower incident heart failure risk, and they are also considering targeting treatment with a glucagonlike peptide–1 receptor agonist to these patients as well. They have a research protocol in place, called WATCH-DM, that will prospectively assess the efficacy of this strategy.
Despite the cost, others also believe that the time is right for biomarker-based tests to boost access to the benefits that treatment with SGLT2 inhibitors can give patients with type 2 diabetes.
Dr. Vanita R. Aroda
“In theory it’s reasonable” to use a risk score like the recent one reported by Dr. Berg and coauthors, said Vanita R. Aroda, MD, an endocrinologist and director of diabetes clinical research at Brigham and Women’s Hospital in Boston. “We need to pay attention to heart failure as an outcome and use risk stratification” to decide which patients with type 2 diabetes but without established cardiovascular disease warrant treatment with an SGLT2 inhibitor, she said in an interview. “Given the data, we need more concrete recommendations” from medical societies on how to reasonably use biomarkers and imaging to identify patients with type 2 diabetes who are at increased risk for heart failure and hence would benefit from treatment. “This should be of high interest to guidelines committees,” she added.
The earlier version of Dr. Berg’s score, based exclusively on clinical observations and conventional measures like estimated glomerular filtration rate and urinary creatinine to albumin ratio, had overlap with established criteria for starting treatment with an SGLT2 inhibitor, such as the presence of chronic kidney disease, she noted. “A biomarker-based score may provide the additional level of discrimination needed to characterize risk and potential benefit.”