Monitoring concentrations of a protein marker associated with chronic heart failure in order to guide treatment decisions can result in markedly better outcomes for patients than standard monitoring and treatment, according to a study by researchers at Massachusetts General Hospital.
The group, led by James Januzzi, reported the results of the PROTECT clinical trial, which ended last year, in the Journal of the American College of Cardiology this month. In the study, they confirm that using concentrations of a protein called N-terminal pro-B-type natriuretic peptide to guide heart failure medication dosing resulted in much better patient outcomes than standard practices. The benefit to patients was found to be so great that the study was ended early so that all patients could receive NT-proBNP-guided treatment, the authors wrote.
"We saw a remarkable reduction in the number of adverse events patients had, [as well as an improvement in] the time to first event, so not only did they have fewer [problems], but they did better for longer," Januzzi told PGx Reporter.
"Quality of life also improved in the NT-proBNP guided arm," he said.
According to Januzzi, earlier studies by his group and others had shown that concentrations of biomarkers like NT-proBNP are very strongly associated with the presence and severity of heart failure, and fall when medications are increased. His team had hoped with the PROTECT (short for Pro-BNP Outpatient Tailored Chronic HF Therapy) trial to definitively demonstrate that measuring levels of the protein could be used to guide dosage decisions and improve patient outcome.
"Heart failure can be a challenging diagnosis not just to identify, but also to recognize whether a patient on treatment is adequately managed," he said.
"We use a number of different classes of drugs to reduce the risk of hospitalization and death," he said. "And each of these has a specific dosing range we are instructed to use such that we achieve maximum tolerated doses."
Physicians depend on symptoms like shortness of breath, blood pressure, and heart rate to decide if they can adjust dosage. "These are fairly crude tools," Januzzi said, so it is hard to know if a patient is really getting their maximum dose of medication.
"Having an objective biological measure that’s not only useful for identifying risk in heart failure, but also modulates in response to therapy, would be a great advance."
In the trial, the researchers limited their cohort to patients with left ventricle systolic dysfunction, randomizing the group into either a standard heart failure management arm or an arm where the goal was to adjust dosages to lower NT-proBNP below 1,000 pg/ml, a threshold the researchers previously identified as a reasonable target, Januzzi said.
They followed subjects for approximately 10 months on average and assessed quality of life and health factors. They also measured heart performance using echocardiography at the beginning and end of the trial.
"Through the ten months of follow-up we were able to reduce the NT-proBNP concentration by 50 percent in the biomarker-guided arm," Januzzi said, "while in the standard arm, illustrating the challenges that standard management has, the NT-proBNP didn't fall at all."
Meanwhile, he said, "at the 50 percent mark through the trial, we did a blinded assessment of outcomes and found such a dramatic reduction in adverse outcome in the BNP arm, we suspended enrollment in the trial."
He added that the researchers saw a 50 percent reduction in heart failure hospitalizations in the biomarker arm.
While analysis of the patient outcomes was blinded, patients and treating doctors were aware of which arm they were in, he explained. This study arrangement is called a prospective randomized open label blinded end point analysis, or PROBE.
"PROBE studies are more and more common in cardiovascular clinical trails because it's almost impossible to implement a therapeutic strategy where patients aren’t aware of what they're getting," he said.
While the design comes with obvious limitations, there are also benefits, according to Januzzi. In the PROTECT trial, "despite the fact that both doctors and patients knew which arm they were in, there was actually a substantial degree of up-titration of therapy in the standard arm – implying that physicians didn't undertreat patients just because they knew they weren’t in the active BNP arm," Januzzi said.
"Also, it's very difficult to underestimate the importance of the patient knowing what their NT-proBNP value is with respect to their likelihood of attentive self-care," he said. "If a patient knows their level is rising, it's very likely they're going to be doubly careful with their diet, salt and water restrictions, and taking medications."
Beyond measures of improvement in symptoms and general health, the group also found that there were biological effects associated with the guided treatment. "We were able to obtain a baseline and follow-up measure of cardiac structure and function in our patients — a very detailed phenotyping," Januzzi said.
"NT-proBNP-guided care was associated with a substantially better improvement in cardiac structure and function compared to standard treatment, so really we found a biological measurement that was associated with biological benefits," he said.
"It wasn't just that patients got on higher doses of their meds," he added.
Januzzi has received research grants and consultancy fees from Roche Diagnostics, which manufactures the NT-proBNP test used by the researchers and which funded the PROTECT trial.
According to Januzzi, a number of different companies also have NT-proBNP or B-type natriuretic peptide assays, and he has worked with many of them as well.
He said his research group is now planning another trial looking at BNP, another heart failure-associated protein marker. This trial, a home-based finger stick study called HABIT-II, will investigate BNP and the impact of more constant home surveillance, Januzzi said.
Additionally, the researchers are also hoping to conduct a multi-center trial furthering the NT-proBNP research, which is intended to include 1,100 patients and estimated to start in 2012.
Have topics you'd like to see covered in Pharmacogenomics Reporter? Contact the editor at mashford [at] genomeweb [.] com.