Washington ? A highly sensitive new test could lead to a different way to diagnose people with Alzheimer’s disease, possibly helping find the illness in its early stages when there might be time for treatment.

While as many as 4 million Americans are thought to suffer from the memory-destroying illness, the only way to diagnose it definitively is by studying brain tissue during an autopsy.

“If you can’t diagnose it, you’re not going to have a therapy for it,” said Chad A. Mirkin of Northwestern University.

Many companies have experimental therapies, he said, “But those therapeutics aren’t very good if you can’t definitively diagnose and follow a disease,” explained Mirkin, a lead researcher — along with William L. Klein — on a team that developed the new test, which can detect small amounts of proteins in spinal fluid.

The team’s findings are reported in Tuesday’s issue of Proceedings of the National Academy of Science.

The new test, called a bio-barcode assay, is 100,000 times to 1 million times more sensitive than other tests, Mirkin said in a telephone interview.

It was first used last year in testing for a marker for prostate cancer, and Mirkin said he invited other investigators to suggest subjects for further testing

Klein, also at Northwestern, had done research associating Alzheimer’s with a protein in the brain called amyloid-beta-derived diffusable ligand, or ADDL, Mirkin said.

So the research team set out to try and detect ADDL in spinal fluid.

They got samples of the spinal fluid of 30 people, 15 who had Alzheimer’s disease and 15 who did not.

The researchers found at least some ADDL in all the patients, which Mirkin said is an indication that everyone may have a baseline level of the protein.

“What was really encouraging,” he said, is that the concentration of ADDL increases as the disease gets worse, so the progression of the illness could be followed.

“Do we have a new diagnostic for Alzheimer’s?” Mirkin said. “That’s a bit premature.”

The method needs to be repeated and tested on more patients, he said.

Dr. Samuel Gandy, who was not part of the research team, said the report is impressive but needs to be repeated with larger numbers of subjects.

If the test can, in fact, correlate the presence of ADDLs with brain function, “this is good news indeed for identifying who is at risk for Alzheimer’s and potentially for following the effectiveness of many new anti-amyloid medicines that are now in clinical trials,” said Gandy, vice chair of the National Medical and Scientific Advisory Council of the Alzheimer’s Association and director of the Farber Institute for Neurosciences at Thomas Jefferson University in Philadelphia.