Experimental 10-Protein Blood Test Detects ALS With 98 Percent Accuracy

For a patient presenting with unexplained muscle weakness, physicians often face a frustrating waiting game. They must rule out other conditions one by one, a process that can delay a definitive diagnosis of amyotrophic lateral sclerosis by nearly a year.[2][4]

Now, researchers at Emory University have developed a 10-protein blood signature that could fundamentally change this timeline. Published this week in Nature Communications, the experimental panel distinguished people with ALS from healthy individuals with 98.3 percent balanced accuracy.[2][5]

The new signature substantially outperformed the current leading biomarker, neurofilament light chain, which achieved 89.9 percent accuracy in the same patient group. The findings offer a highly sensitive new approach to detecting the neurodegenerative disease.[3][4][5]

By measuring a broader spectrum of nerve damage and inflammation, the multi-protein approach reduces false positives and catches cases that a single marker misses. This could eventually allow doctors to identify the disease months earlier than current methods permit.[3][5]

The research team emphasized that the test remains in the experimental phase and requires further validation. However, the results represent a major step toward a simple blood test that could expedite access to specialized care and clinical trials.[4][5]

Moving beyond a single biomarker

ALS is a complex, progressive neurological disease that destroys motor neurons, the nerve cells responsible for voluntary movement. Because early symptoms like cramping and difficulty walking mimic other disorders, there is currently no single definitive test to confirm the disease.[2][4]

Historically, researchers have focused on neurofilament light chain, a protein released when nerve fibers are damaged, as a primary indicator of ALS. However, levels can also spike in patients with other neurological conditions, leading to diagnostic confusion and potential false alarms.[2][4]

"ALS is a complex disease, and relying on a single biomarker doesn't capture the full picture," said Dr. Eleanor Varian Thomas, assistant professor of neurology at Emory University School of Medicine and co-principal investigator of the study.[2][4]

Thomas explained that combining proteins associated with different aspects of the disease, including nerve cell injury and inflammation, allowed the team to identify ALS with much greater accuracy. The broader panel captures the biological reality of the condition more effectively.[2]

A reliable blood screening tool would not replace a full neurological workup, but it would serve as a crucial early indicator. It could quickly flag patients who need immediate evaluation at a specialized ALS center, drastically reducing diagnostic delays.[2][5]

How the 10-protein panel was built

To build the new diagnostic model, the Emory team started by analyzing blood plasma from an initial cohort of 632 people with ALS and 160 healthy controls. They needed to find a consistent pattern hidden within the complex chemistry of the blood.[3][4][5]

They utilized a highly sensitive platform called NULISA, developed by Alamar Biosciences, which can measure extremely small amounts of proteins. The technology allowed the researchers to simultaneously evaluate more than 300 neurological and inflammatory proteins in a single blood sample.[2][3]

A machine-learning algorithm sorted through the resulting data to identify the proteins that most strongly correlated with the presence of ALS. The algorithm initially flagged 84 proteins with statistically significant differences between the patients and the healthy controls.[3][5]

The researchers eventually narrowed those candidates down to a shortlist of 12, before finalizing the panel at 10 specific proteins. Four of the selected proteins, including neurofilament light chain and phosphorylated tau, already had established links to ALS biology.[3][4][5]

The remaining six proteins had never previously been reported as blood biomarkers for the disease. Their inclusion highlights how machine learning can uncover hidden biological signals that traditional research methods might overlook when searching for a single definitive marker.[3][4][5]

Uncovering new disease biology

The discovery of six new proteins associated with ALS does more than just improve diagnostic accuracy; it opens new avenues for understanding the disease itself. Researchers believe these markers could provide fresh insights into how motor neurons degrade over time.[4][5]

"Several of these new proteins may also tell us something about ALS biology itself, which we plan to explore," Thomas noted in a statement. Understanding the specific inflammatory pathways these proteins represent could eventually lead to novel therapeutic targets.[4]

Currently, treatments for ALS are severely limited, and most clinical trials require patients to have a confirmed diagnosis before they can enroll. Because the diagnostic process takes so long, many patients are already in advanced stages of the disease by the time they qualify.[2][5]

An accurate early-detection blood test would allow patients to enter clinical trials months earlier, when their motor neurons are less damaged. Testing new drugs on patients in the earliest stages of ALS is widely considered the best chance for finding a treatment.[2][5]

Validation and reducing false positives

The critical test for the new signature came when the researchers applied it to a completely independent validation cohort. This separate group consisted of 144 ALS patients and 37 healthy controls whose blood had not been used to train the algorithm.[3][5]

In this unseen group, the 10-protein panel achieved its 98.3 percent balanced accuracy, proving that the model could successfully evaluate new samples. The high score confirms that the signature relies on genuine biological markers rather than statistical noise from the initial data.[3][4][5]

The inclusion of the newly identified proteins proved essential for distinguishing ALS from other conditions that cause nerve damage. One protein in particular played a crucial role in correctly classifying individuals who did not have ALS but still presented with elevated neurofilament levels.[3][5]

By catching these nuances, the multi-protein signature significantly reduced false-positive results compared to relying on a single marker alone. This precision is vital for any future clinical test, as a false ALS diagnosis carries immense psychological weight for patients.[3][5]

The panel also successfully identified some ALS patients who had relatively little disability and whose neurofilament levels were not high enough to trigger a standard positive result. Catching the disease in these early stages is exactly when experimental treatments are most likely to be effective.[3]

AI

The path to clinical application

While the results are highly promising, the 10-protein signature remains a research finding rather than a clinically validated diagnostic test. The current study compared ALS patients exclusively against healthy controls, which represents a controlled experimental environment.[4][5]

In a real-world clinical setting, a diagnostic blood test must reliably distinguish ALS from other neurodegenerative diseases that present with similar early symptoms. The panel must prove it can differentiate ALS from conditions like multiple sclerosis or Parkinson's disease.[4][5]

Researchers plan to test the protein signature in larger, more diverse patient groups across multiple medical centers to confirm its broader applicability. These upcoming trials will determine whether the test can perform consistently across different demographics and clinical environments.[4][5]

"ALS patients and their families often wait far too long for answers," said Dr. Yuling Luo, founder and CEO of Alamar Biosciences. He noted that the technology was specifically built to measure these elusive proteins and accelerate the diagnostic timeline.[3]

If validated in broader trials, this blood-based screening tool could fundamentally alter how neurologists approach early-stage muscle weakness. By providing a clear biological signal early in the disease process, it could finally give patients the timely answers they desperately need.[2][5]