How Alzheimer’s Is Diagnosed Now: Dr. Marwan Sabbagh on Blood Tests, PET Scans, and What Comes First

Blood-based biomarkers are reshaping the Alzheimer’s diagnostic pathway. Alongside established tools like amyloid PET and CSF analysis, these assays offer a less invasive, more accessible entry point for detecting underlying pathology. But their arrival raises practical questions for clinicians: where does a blood test fit, when should it be followed by PET or CSF, and how should results guide subsequent workup?

Dr. Marwan Sabbagh, a behavioral neurologist and the Moreno Family Chair for Alzheimer’s Research in the Alzheimer’s and Memory Disorders Program at Barrow Neurological Institute in Phoenix, has built his career on the diagnosis and management of Alzheimer’s and other memory disorders.

In a conversation with Being Patient founder Deborah Kan, recorded at the Alzheimer’s Association International Conference, Sabbagh outlined his stepwise diagnostic algorithm, initiating with a blood-based biomarker as a screening tool, then proceeding to PET or CSF for confirmation. He discussed why he now assesses amyloid burden quantitatively rather than as a binary positive/negative finding, how he communicates results to patients, and where interpretive uncertainty persists. He also noted how biomarkers have advanced diagnostic accuracy from roughly 67 percent into the 90 percent range within just a few years, and why aligning therapeutic options with this diagnostic precision is the next imperative.

Being Patient: Marwan, we’re at the AAIC. A lot of people are talking about blood-based biomarkers and how you implement them into practice. You and I have had discussions about that, but how do you use amyloid PET now that we have blood tests?

Dr. Marwan Sabbagh: The discussion is about whether I use blood-based biomarkers as sufficient evidence for a diagnostic. A lot of people think that’s the direction we’re going. A lot of people think that we will get there. I still think that some of us, myself included, would say a blood-based biomarker is an important screening tool. I understand the screening versus confirmatory, and that if they’re abnormal on their blood test, then I go to use PET as a secondary mechanism. I’m not replacing the PET yet. And I find that that stepped approach seems to be working pretty well.

Being Patient: So that’s for the diagnostic portion of it. But now I think there’s confusion in the process about the order of what you use in terms of diagnostics and really analyzing, right? What do you get from a PET scan? And what’s the order — when you get the report from a radiologist, then what happens next?

Sabbagh: One of the things that we’re seeing in amyloid PET is that we’ve now moved to want to see the quantification, not just the interpretation. Because I will tell you that if it’s read as positive, negative, that’s informative, but having the number of Centiloids has really helped me a lot. Would I give a patient with very few, barely in the abnormal range — would I give them a monoclonal antibody or not? So I’m getting a lot of information about amyloid PET that I had not been looking for before.

Being Patient: What about in terms of the actual scan — color versus black and white? How do you feel about that?

Sabbagh: I was trained when people are looking at amyloid PET, it was originally grayscale. You had to look at it grayscale, but some of the newer tracers are using the color scale. I think color scale is easy to look at. I think grayscale is very fuzzy, but I will use the color scale if I have a chance to do so.

Being Patient: And in terms of the color scale and the placement of amyloid, are you looking at the location of where the amyloid is?

Sabbagh: No, it has to be easily identified. I actually am one of those people that will call my radiologist saying, “Are you sure this is positive?” Because I’m looking at it and I’m not sure. Because at the end of the day, these things will have an impact on my decision to treat a patient with an anti-amyloid targeted therapy or not. If it’s a lot and it’s easy to identify, that’s an easy decision. If it’s barely abnormal, do I want to incur the risk — do the risk-benefit analysis?

Being Patient: So take me through the diagnostic pathway step by step. So someone’s coming to you saying, “I’m experiencing memory problems.”

Sabbagh: I will say to you that this is an iterative process. I think these discussions will change as we see our processes reflected. When a patient presents me with a cognitive complaint, I will typically order — historically we do a B12, a TSH, and an MRI, but I now add a p-tau217, p-tau181. I will often get neuropsych testing. I still get my MRI, but in the MRI, I’m asking for a lot of things, not just the macro exclusionary pathology. I want them to tell me about white matter scoring, Fazekas score, volumetric changes. So there’s a lot of things I’m asking for that I didn’t ask for before, but I still also like having neuropsych testing. So that’s round one. If it’s abnormal — let’s say my p-tau level is like 0.4 — then I’ll go into round two, which is the confirmatory testing with the PET or CSF.

Being Patient: And not until you have that will you deliver a diagnosis?

Sabbagh: I deliver a diagnosis at the first consultation. I say, here’s what I’m worried about: mild cognitive impairment, mild dementia. I will actually explain what is mild cognitive impairment, what is dementia, and then I will talk about: these are categorical definitions, they’re not etiological definitions, and our goal, our task, our charge is to figure out what is the cause — mild cognitive impairment or dementia due to Alzheimer’s, Parkinson’s, Lewy body, et cetera. And then I’ll say, this is what I think you don’t have: Parkinson’s, stroke, Lewy body, et cetera, and this is the point of the tests. So when I’m communicating every time, I’ve already laid the groundwork to say, okay, I’m worried about an Alzheimer’s process, here’s why, and this is what we’re finding.

Being Patient: How do you communicate the results of a PET scan?

Sabbagh: I will actually pull up the slides of my negative PET scan and positive PET scan and say, here’s what a negative PET scan should look like, here’s what a positive PET scan would look like. We hope your scan looks like this. We don’t want your scan to look like that. And then I’ll pop in their scan and show them side by side.

Being Patient: So because of the new technologies, the whole diagnostic pathway, as we talked about, is really changing. Where do you think the confusion lies right now?

Sabbagh: The more we get tests, and the better we think we’re getting, the more confusion we add, right? One of the biggest confusions is how a p-tau blood test is a marker of amyloid, not tau. I have a heck of a time trying to explain that to my patients. Do the p-tau tests, do the blood tests have predictive value, much like we did with the APOE? And all these discussions — questions that have been answered before that are now being resurfaced. Am I going to get worse? How likely? How soon? Things like that are stuff that we still have to use data to drive those answers. And I think we’ve made some progress. I think there’s some more progress to be made.

Being Patient: And just to sum it up, when your colleagues say, “Marwan, how do I diagnose someone today?”

Sabbagh: It’s way better than it was. What has changed is the diagnosis of exclusion, which was at 67 percent accuracy. The biomarkers — what you’re talking about, PET, CSF, or blood — have soundly and roundly moved that well into the 90s. I think we’ve solved a huge problem in just less than 10 years, five years. So I feel good about that. Now that we have diagnostic accuracy, we need to bring the therapies along.