Estimating amyloid-β deposition using brain perfusion SPECT in Alzheimer's disease

Alzheimer's disease is a progressive condition that slowly erodes memory and thinking skills, eventually becoming the most common form of dementia worldwide. At its core, the disease is defined by two specific types of damage that accumulate inside the brain: sticky clumps of a protein called amyloid-beta and twisted strands of a protein called tau. These clumps and tangles disrupt the brain's ability to communicate, leading to the symptoms of the disease. For decades, doctors could only confirm these changes after a patient passed away by examining brain tissue. Today, however, we can detect these proteins while a person is still alive. The most accurate way to do this involves injecting a special radioactive tracer into the body and using a scanner to see where the proteins have settled, a process known as a PET scan. While highly effective, these scans are expensive and not available in every hospital. Another method involves taking a sample of the fluid that surrounds the brain and spinal cord, known as cerebrospinal fluid, to measure the levels of these proteins. This is also accurate but requires a needle inserted into the lower back, a procedure that many people find uncomfortable or frightening. Because of these barriers, researchers are constantly looking for simpler, more accessible ways to identify who might have the disease before resorting to these more invasive or costly tests.

A team of researchers in Japan set out to see if a much older and more common imaging tool could help fill this gap. They focused on a technique called SPECT, which stands for single photon emission computed tomography. Unlike the PET scan that looks directly for the sticky protein clumps, a SPECT scan does not see the proteins themselves. Instead, it measures how much blood is flowing to different parts of the brain. The logic behind this approach is that when brain cells are damaged or dying, they need less energy and therefore draw less blood. In Alzheimer's disease, specific areas of the brain, particularly those involved in memory and spatial awareness, tend to show reduced blood flow long before severe symptoms appear. The researchers wanted to know if the pattern of this reduced blood flow could act as a reliable clue to predict whether a patient actually had the amyloid protein buildup that defines the disease.

To investigate this, the team looked back at the medical records of 88 patients who had visited a general medical center in Hyogo, Japan, for evaluation of memory problems between 2014 and 2024. All of these patients had undergone a standard set of tests, including a cognitive exam to measure their memory and thinking skills, a magnetic resonance imaging (MRI) scan to look at the structure of their brains, a SPECT scan to measure blood flow, and a lumbar puncture to collect cerebrospinal fluid for analysis. The researchers used the fluid samples as the "truth" against which to test the SPECT scans. They analyzed the fluid to see if it contained high levels of the amyloid protein and the tau protein, which would confirm the presence of Alzheimer's pathology. They then compared these results with the blood flow maps from the SPECT scans to see if there was a connection.

The analysis revealed a clear pattern. The researchers found that in patients who had the amyloid protein buildup, there was indeed a specific reduction in blood flow in the back and side parts of the brain, particularly in an area called the posterior cingulate and the parietal regions. This matched what scientists already knew about how Alzheimer's affects the brain. When they looked closer, they found that the level of amyloid protein in the fluid was positively linked to blood flow in these areas, meaning that less protein buildup was associated with better blood flow. Conversely, the levels of tau protein in the fluid were negatively linked to blood flow, meaning that higher amounts of tau were associated with lower blood flow in the same regions. This confirmed that the blood flow changes seen on the SPECT scans were indeed related to the biological changes of the disease.

However, when the researchers tried to use the blood flow data alone to predict whether a patient had the disease, the results were only moderately successful. They built a mathematical model to see if they could guess the presence of amyloid based on the blood flow numbers. The best result came from looking at the blood flow in the left parietal region of the brain. This single measurement could distinguish between patients with and without the disease with a level of accuracy that was better than random guessing, but it was not perfect. The researchers compared this performance to the accuracy of the fluid tests and found that the fluid tests were significantly better at making the correct call. While the blood flow scan could suggest a likelihood, it could not replace the definitive answer provided by the fluid analysis or the specialized PET scan.

The study also examined other factors, such as the patients' ages, their scores on memory tests, and how much time had passed between their different scans. None of these factors performed as well as the blood flow measurements in the specific brain regions. The researchers concluded that while a SPECT scan cannot substitute for the more direct methods of detecting amyloid, it holds value as a supportive tool. In a busy clinical setting, a SPECT scan is far more affordable and widely available than a PET scan. It could serve as a useful first step to help doctors decide which patients are most likely to benefit from the more expensive or invasive confirmatory tests. By identifying patients who show the characteristic drop in blood flow, doctors might be able to triage patients more effectively, ensuring that those who need the advanced testing get it sooner, while sparing others from unnecessary procedures.

The researchers were careful to note the limits of their work. The study was retrospective, meaning they looked at past data rather than following new patients forward in time, and the group of 88 people was relatively small. Additionally, the fluid tests were performed over a ten-year period using different machines, which could have introduced some variation in the results. Despite these limitations, the findings offer a practical perspective on how existing technology can be repurposed. The study suggests that the brain's blood flow patterns, visible on a standard SPECT scan, carry a quiet but meaningful signal about the underlying biology of Alzheimer's. While it is not a standalone solution, it represents a potential bridge between the everyday tools available in most hospitals and the high-tech diagnostics needed to confirm the disease, offering a way to make early detection more accessible to those who need it.