← Back to Home

Ten blood proteins offer clues to identifying ALS, but clinical discrimination remains to be validated

An Emory research team used multiple protein signals to distinguish ALS patients from healthy individuals, achieving 98.3% balanced accuracy in an independent sample. Whether it can distinguish diseases with similar symptoms remains a key test as the research moves toward clinical use.

By SURL BioNews

For patients with amyotrophic lateral sclerosis (ALS, commonly known as Lou Gehrig’s disease), identifying the cause more quickly after symptoms appear is a diagnostic challenge. A research team at Emory University in the United States identified a signature comprising ten blood proteins that distinguished ALS patients from healthy individuals with 98.3% balanced accuracy in an independent validation sample. This offers a research direction for blood-based diagnostic support, but key questions remain before it can be used in clinical practice.

The study was published in Nature Communications on October 19. Alamar Biosciences subsequently announced the findings, which RTTNews also reported. The team first analyzed plasma from 632 ALS patients and 160 healthy controls, using Alamar’s NULISA platform to measure more than 300 proteins related to neurological and inflammatory processes, then applied machine learning methods to select a combination that could identify the disease.

The final model included ten proteins and was validated in a separate group of 144 ALS patients and 37 healthy controls. “Balanced accuracy” is the average of sensitivity in identifying patients and specificity in identifying people without the disease; therefore, 98.3% cannot be directly interpreted as an equally high probability that a member of the general population will receive a correct diagnosis after testing.

Interpreting multiple signals together is central to this study. According to Alamar’s announcement, neurofilament light chain alone achieved 89.9% balanced accuracy in the same validation group, lower than the ten-protein combination. Neurofilament light chain can reflect damage to nerve axons; the study’s combination also included proteins such as pTau-181, FOLR1 and TREM1, seeking to address the limitations of a single marker through different biological signals.

Both Alamar’s announcement and the RTTNews report stated that the combination identified some ALS patients who were not identified using neurofilament light chain alone, including those with relatively mild disability. However, milder disability does not mean that detection before symptoms appear has been demonstrated, nor is it sufficient to establish that the combination can shorten the time to clinical diagnosis.

The most important limitation at present is that the validation compared ALS patients with healthy individuals, rather than people who sought care for similar symptoms and were ultimately diagnosed with other diseases. Whether performance can be maintained in populations that more closely reflect clinical practice still requires direct testing. The research team also plans to expand the study to more diverse patient populations; until then, these ten proteins remain a candidate tool worth further validation, rather than a screening or diagnostic test that has completed clinical validation.

References

  1. Alamar Biosciences via BioSpace
  2. RTTNews