Despite success in some cancer patients, a drug that neutralizes GDF-15 was clearly harmful in HF patients. The question is, why?
MINNEAPOLIS, MN—Neutralizing growth differentiation factor 15 (GDF-15), a stress-activated cytokine, appears harmful to patients with heart failure (HF).
That’s the bottom-line message from the phase II GARDEN-TIMI 74 trial.
The study was prematurely stopped when initial data showed that treatment with ponsegromab (Pfizer), a monoclonal antibody with prior success in addressing cachexia among cancer patients, was tied to an increase in worsening HF events—including hospitalizations for HF and urgent HF visits requiring IV diuretic therapy—when compared with placebo.
There were no improvements in quality-of-life scores and importantly no effect on cardiovascular or all-cause death rates.
“You often learn from when science goes the opposite direction from what you expect,” David D. Berg, MD (Brigham and Women’s Hospital, Boston, MA), who presented the findings at the Heart Failure Society of America 2025 Annual Scientific Meeting, told TCTMD. “The question to me is not so much: ‘Would I then go back and test this agent again in the heart failure population?’ I think that there’s legitimate safety concerns with doing that.”
Instead, the best questions will be those that seek to understand why neutralizing GDF-15 increased the risk among HF patients, he said.
Christopher O’Connor, MD (Inova Health System, Fairfax, VA), the discussant following the presentation, welcomed Berg to the club of investigators “who have contributed so many negative trials in the field of acute decompensated heart failure.”
Like Berg, though, O’Connor stressed that there is much to learn from negative studies, perhaps more so than with positive trials.
GARDEN-TIMI 74 Findings
In cancer patients with cachexia and elevated GDF-15, treatment with ponsegromab has been shown to help them gain weight, improve their physical activity levels, and decrease their cachexia symptoms. The idea behind GARDEN-TIMI 74 was that neutralizing GDF-15 might improve HF-related health status and boost physical activity levels.
Investigators enrolled 433 patients with HF (mean age 75 years; 25% female) from 115 sites throughout 11 countries who were randomized to treatment with placebo or ponsegromab (most received a 300-mg subcutaneous dose every 4 weeks, though some received 100- and 200-mg doses). Mean baseline LVEF was 35%, 44% were considered to have NYHA class III/IV symptoms, and 29% had cachexia. Mean baseline NT-proBNP was 2,156 pg/mL. Patients were well treated at baseline, with 78% on an SGLT2 inhibitor and 47% on four pillars of guideline-directed medical therapy.
From week 4 through 22, ponsegromab suppressed GDF-15 in treated patients by at least 96% compared with placebo. Researchers also saw an expected increase in weight among patients in the study arm, where the mean increase was 1.4 kg over the same time period.
They observed no significant treatment effect with ponsegromab on the KCCQ clinical summary score nor any of its individual components. There also was no effect on 6-minute walk distance.
You often learn from when science goes the opposite direction from what you expect. David D. Berg
The composite endpoint of worsening HF event or cardiovascular death was higher in the ponsegromab arm compared with placebo (19.8% vs 11.6%; HR 2.24; 95% CI 1.43-3.50), a finding driven by worsening HF events (19.3% vs 9.1%; HR 2.84; 95% CI 1.74-4.64). There was no significant difference in cardiovascular death (3.0% vs 2.5%; HR 0.86; 95% CI 0.31-2.36).
Notably, ponsegromab also significantly increased both NT-proBNP and high-sensitivity C-reactive protein compared with placebo by week 22.
‘Clearly Prognostically Important’
The GARDEN-TIMI 74 hypothesis was “good,” O’Connor said. “Tackling a marker, though, has been challenging in heart failure. As you know, we’ve tackled biomarkers in the past that haven’t resulted in positive clinical outcomes, so there had to be more.”
However, he pointed out that only 29% of patients had cachexia and NT-proBNP levels were “really not that elevated.” As such, “this is not the type of patient population that I think could have benefited from this therapy,” O’Connor said. A better population would have been those with “very advanced” heart failure with an average NT-proBNP level of 10,000 pg/mL.
There is enough prior evidence to suggest that GDF-15 is “clearly prognostically important [and] associated with adverse heart failure outcomes, physical impairment, greater heart failure symptom burden, and all the things that we worry about,” said Berg. “But there’s always this question about is it causally implicated in any way.”
While ponsegromab has shown positive outcomes among cancer patients with cachexia, the fact that it did not pan out in GARDEN-TIMI 74 illustrates “a lesson [that] cachexia may not be the same beast in different underlying disease states,” he suggested.
Additionally, it’s possible that “there may be a number of off-target effects” from GDF-15 neutralization, Berg said, including increasing systemic inflammation or potentially affecting the metabolic pathway.
“There’s mechanistic work to be done to sort that out as the next step,” he continued. “The good news is we collected a lot of samples and data in the study, and so I think we’re starting to dig into those questions.”
In the end, Berg noted that the GARDEN-TIMI 74 population had severely debilitating symptoms as well as physical limitations despite being “exceptionally well-treated” with contemporary therapy. “Our trial underscores that we must keep striving to help these patients, who are very much in need of new options,” he said. “GDF-15 neutralization may not be the answer, but I remain optimistic that we will continue to find new strategies to help this group of patients.”
Comments are closed.