CHICAGO – Signals from the plasma proteome may help identify which transthyretin cardiac amyloidosis (ATTR-CM) patients will respond better to therapy, early-stage research presented Saturday at the American College of Cardiology's annual meeting suggests.
Researchers from Columbia University analyzed the plasma proteomes of patients with and without ATTR-CM, as well as with different types of the disease, to home in on signaling pathways that are dysregulated in the condition and may affect disease prognosis and treatment response. The distinct proteomic profiles they identified suggest that patients with hereditary ATTR-CM may respond better to targeted therapy and that patients belonging to a newly identified molecular subgroup of ATTR patients may have worse outcomes but also respond better to Pfizer subsidiary FoldRx's TTR stabilizer Vyndamax (tafamidis) than others.
In the US, about 100,000 people are estimated to have ATTR-CM, a severe and progressive disease marked by accumulation of abnormal transthyretin (TTR) proteins in the heart, though the condition is known to be underreported. Some cases of ATTR-CM are hereditary and are caused by mutations in the TTR gene that lead to misfolded proteins, while other cases involve patients who have wild-type TTR genes but develop the disease due to TTR proteins that misfold as they age. As the disease progresses, it leads to heart failure.
There are an increasing number of treatments for ATTR-CM. The US Food and Drug Administration approved Vyndamax in 2019 as an oral treatment for the disease, and in late 2024 approved BridgeBio Pharma's TTR stabilizer Attruby (acoramidis) and this month approved Alnylam Pharmaceuticals' small interfering RNA medicine Amvuttra (vutrisiran). Others are under development.
"It is important to predict prognosis and response to tafamidis in the era of multiple disease-modifying therapies," said Itsuki Osawa, a postdoctoral researcher at Columbia, during his presentation at the meeting.
Identifying the disease isn't always straightforward. However, Keitaro Akita, an associate research scientist in Yuichi Shimada's lab at Columbia, reported at the meeting that he and his colleagues were able to distinguish between ATTR-CM patients and patients with another heart condition by analyzing patients' plasma proteomes.
In an age- and sex-matched, case-controlled analysis, the researchers compared the plasma proteomes of 169 patients with ATTR-CM and those of 170 controls with hypertensive left ventricular hypertrophy, profiling 7,289 proteins. By using 70 percent of the data as the training set and 30 percent as the testing set, they developed a machine learning model that distinguished between the ATTR-CM cases and controls with a sensitivity of 0.96 and a specificity of 1.
In the proteins that differed significantly between the ATTR-CM and control cohorts, the researchers uncovered signaling pathways that are dysregulated in ATTR-CM, like the PI3K-AKT pathway and its up- and downstream pathways like the JAK-STAT and the complement and coagulation cascade pathways.
But there are also different types of ATTR-CM, as it can be either a hereditary or age-related disease. Following on their previous work, Akita and colleagues also compared the plasma proteomes of 31 patients with hereditary ATTR-CM and 138 patients with wild-type ATTR-CM.
They identified 394 proteins that varied between the two disease types, independent of age, sex, and race, which are known clinical parameters related to hereditary disease. A pathway analysis found that the RAS/MAPK and related pathways were dysregulated in hereditary ATTR-CM, as were pathways related to angiogenesis and phagocytosis or those involved in autophagy.
"This suggests that response to targeted therapy may differ between wild-type and variant ATTR," Akita said, noting that such therapy boosts macrophages and endocytosis.
Osawa, a postdoc in the same lab, noted in a separate session that ATTR-CM patients have exhibited a heterogenous responses to treatment with Vyndamax. He likewise used plasma proteome profiling to examine variability in prognosis and treatment response among ATTR-CM patients. "Our hypothesis here was [that] proteomics profiling could predict prognosis and treatment response to tafamidis accurately," Osawa said.
He and his colleagues analyzed more than 7,000 proteins in plasma samples from 142 patients with ATTR and compared their prognosis and treatment response. Through an unsupervised machine learning-based approach, they uncovered three distinct molecular subtypes of the disease, dubbed subtypes A, B, and C.
Patients with subtype A had the highest risk of mortality, with a nearly twofold higher risk of death compared to the other subgroups. At the same time, patients with subtype A also exhibited the best response to Vyndamax.
A pathway analysis further showed that patients with subtype A had increased dysregulation of the RAS/MAPK pathway, phagocytosis or autophagy pathways, and in metabolic or inflammatory pathways, compared to the other subtypes. The analysis suggested to Osawa that dysregulated biological defense mechanisms against transthyretin tetramers along with dysregulated responses may drive disease progression in patients with ATTR-CM. This mechanism, Osawa said, may also explain why patients in subtype A had the highest treatment response to Vyndamax, as it affects inflammation.
Osawa also noted a few possible explanations for the seeming disconnect between patients with subtype A having a high mortality rate but also the best treatment response, including that because these patients had the worst prognosis, they are also the ones that could benefit the most from treatment.
He cautioned, however, that it is unclear whether these changes in plasma protein levels reflect pathological changes in the myocardium. To confirm that, researchers would have to analyze the myocardial proteome. Still, Osawa said plasma was a pragmatic choice for analysis, since it is easily accessible from patients in the clinic through a blood draw.