NEW YORK – A European research team this month kicked off a project to develop biomarkers for diagnosing different types of hypertension (HT) and guiding targeted treatment selection.
Called HT-Advance, the effort is backed with €8.6 million ($9.3 million) in EU funding, plus an additional €4.8 million from the UK and €1.5 million from Switzerland. The project commenced on March 1 and will run through February 2029. Investigators from the French National Institute of Health and Medical Research (Inserm), and Radboud University Medical Center in the Netherlands are spearheading the endeavor.
Other participating institutions include the University of Torino and the University of Padua in Italy; Technical University of Dresden and Helmholtz Munich in Germany; Saint-Luc University Hospital, Assistance Publique–Hôpitaux de Paris, and Paul Sabatier University in France; the University of Dundee, the University of Glasgow, and the University of Birmingham in the UK; and the University of Zurich in Switzerland.
According to Maria-Christina Zennaro, research director at Inserm's Paris Cardiovascular Research Center, HT-Advance aims to further develop and validate classifiers discovered during a previous EU-funded effort called the European Network for the Study of Adrenal Tumors–Hypertension, or ENSAT-HT, project.
ENSAT-HT ran from 2015 through 2021 and developed multiomics-based biomarkers to distinguish patients with endocrine hypertension (EHT) from those with primary hypertension (PHT), as well as to subtype them.
These biomarkers, Zennaro noted, include a mix of genomic and metabolomic measurements made in blood and urine samples using mass spectrometry and quantitative PCR. In a study published in EBiomedicine last year, they showed that their multiomics classifiers could distinguish different EHT subtypes from PHT more effectively than any individual omics methods.
HT-Advance will also explore whether these classifiers could help select antihypertensive therapy, potentially reducing the burden on clinicians who still rely on a trial-and-error approach when prescribing therapies.
"The problem is that hypertension is very much undertreated, whereas it is the most frequent cause of death through its cardiovascular complications," said Jaap Deinum, an associate professor at Radboud and a coordinator of HT-Advance. While some patients may not respond to a drug, others may develop side effects, and clinicians have no way to determine in advance which drugs are effective for an individual. "If we could get better treatments to people, that could help us successfully prevent cardiovascular disease," he said.
HT-Advance will run three clinical trials called HT-Endo, HT-Predict, and HT-Treat. Investigators will use machine-learning techniques for these to integrate the genomic and metabolomic measurements, as well as to provide diagnostic and therapeutic information for clinicians. They will also include an economic assessment of the markers and produce legal and ethical recommendations for clinical decision making.
HT-Endo will be a randomized controlled trial, involving about 250 patients, to validate the markers developed in ENSAT-HT. According to Deinum, patients will be diagnosed according to conventional methods or the composite biomarkers.
"If we can prove that the final clinical results are identical or better, this could be something that could simplify clinical practice and also make it cheaper," he said.
HT-Predict, involving about 100 patients, will use the same approach developed in ENSAT-HT to identify new markers with the goal of predicting the response of patients with PHT to different therapies. Patients will be genotyped using PCR-based panels, and genomic and metabolomics data will be gathered.
Zennaro said that HT-Advance will use custom SNP panels covering markers associated with blood pressure and treatment response and resistance. Blood pressure in response to therapies will be recorded and associated with the omics data.
Signatures developed during HT-Predict will then be deployed, together with those used in HT-Endo, in a third trial, called HT-Treat, to identify patients with EHT for targeted treatment and to guide treatment decisions in patients with PHT. The number of patients for HT-Treat depends on the results from the HT-Predict trial, Deinum said.
Machine-learning tools are a component of HT-Advance. Christian Cole's group at the University of Dundee will develop a clinical tool, to be used as part of the three trials, for the automated prediction of hypertension type as well as response to treatment based on patients' genomic and metabolomic data. This will be accomplished by integrating results from the laboratories taking part in HT-Advance and extracting dozens of features from that data to inform an artificial intelligence-based predictor, according to Cole, a senior lecturer in health informatics at Dundee.
"The data platform will bring clinicians closer together with modern machine-learning tools to aid their diagnosis of patients," he explained in an email. "The aim is to provide understandable results and speed up decision making to provide the most appropriate care."
The ultimate goal is to change the way hypertension is diagnosed and treated, Zennaro noted.
"Ideally, you would go to the doctor, get our test, and the doctor would be able to say what type of hypertension you have," she said. "If you have endocrine hypertension, he would direct you to a specialized center, and if you have the primary form, then he would give you a targeted treatment, without side effects."
Zennaro said the investigators have discussed licensing the IP resulting from ENSAT-HT and HT-Advance to a biotechnology company for commercialization in a chip format.
It is unclear if any competitive approaches are available or under development at this time. Diagnostiki Athinon, a Greek clinical and research laboratory, offers a test called Hypertension.Genomix, which focuses on markers that predispose patients to developing primary hypertension.
Should HT-Advance succeed in developing biomarkers to the point that they can be commercialized by the end of this decade, clinicians would need to change their behavior, which Zennaro said would be another struggle in itself.
"When people are used to given procedures, even though they think they are imperfect, they are used to them," she remarked. "To implement new procedures, you need to have care providers who trust in the approach. That is something we would like to work on, and from the very beginning."