Amyloid Imaging Update: How the Amyloid Landscape Is Changing in Light of the Recent Food and Drug Administration Approval of Antiamyloid Therapeutics

MECHANISM OF ACTION, PHARMACODYNAMICS, AND PHARMACOKINETICS

The first FDA-approved amyloid tracer was ¹⁸F-florbetapir (Amyvid; Eli Lilly), the second was ¹⁸F-flutemetamol (Vizamyl; GE HealthCare), and the third was ¹⁸F-florbetaben (Neuraceq; Life Molecular Imaging). An easy way to remember the branded names for the two tracers that have a similar generic name—florbetapir and florbetaben—is that florbetaben ends with an “n” and Neuraceq starts with an “n”; by default then, florbetapir is Amyvid.

The mechanism of action for all 3 tracers is the same; they bind to amyloid plaques. They are highly specific and do not bind to tau or α-synuclein in tissue from patients with AD (1–3).

From a pharmacodynamics perspective, the 3 tracers also have similar pharmacodynamics in that they all cross the blood–brain barrier. With ¹⁸F-florbetapir, the time–activity curves in the brains of subjects with positive scans show continual signal increases from time zero through 30 min after injection, with stable values thereafter up to at least 90 min after injection. With ¹⁸F-flutemetamol, the time–activity curves in the brains of subjects with positive scans show continual signal increases from time zero through 30 min after injection, with stable values thereafter up to at least 120 min after injection. There are no time–activity curve data for ¹⁸F-florbetaben in the pharmacodynamics section of the package insert (1–3).

From a pharmacokinetics perspective, less than 5% of the radioactivity from ¹⁸F-florbetapir is present in the blood by 20 min after injection, and less than 2% is present by 45 min. The residual ¹⁸F in circulation during the 30- to 90-min imaging window is principally in the form of polar ¹⁸F metabolites. For both ¹⁸F-flutemetamol and ¹⁸F-florbetaben, the plasma concentrations decline by approximately 75% at 20 min after injection. For ¹⁸F-flutemetamol, plasma concentrations decline approximately 90% in the first 180 min after injection, and for ¹⁸F-florbetaben, plasma concentrations decline by approximately 90% at 50 min (1–3).

Excretion of ¹⁸F-florbetapir is predominantly through the biliary and gastrointestinal tracts, with lower radioactivity detected in the bladder. For ¹⁸F-flutemetamol, approximately 37% is renal excretion and 52% is hepatobiliary excretion. For ¹⁸F-florbetaben, excretion is via the hepatobiliary route, and at 12 h after injection, approximately 30% of the injected radioactivity is excreted in urine (1–3).

APPROPRIATE USE CRITERIA (AUC) FOR AMYLOID PET

A report of the Amyloid Imaging Task Force, the Society of Nuclear Medicine and Molecular Imaging, and the Alzheimer’s Association on the AUC for amyloid PET was first published in The Journal of Nuclear Medicine in March 2013 (4). In June 2023, in a lecture at the annual meeting of the Society of Nuclear Medicine and Molecular Imaging meeting in Chicago, a member of the task force presented preliminary updated AUC for amyloid imaging. These updated AUC have not been approved or finalized by the task force. A draft was released by the Society of Nuclear Medicine and Molecular Imaging and the task force for public comments in February 2024, and a finalized version was expected to be approved and published sometime in the latter half of 2024 (5).

The AUC for amyloid PET were created to help neurologists, dementia experts, and other physicians know when to order an amyloid PET scan. The working group of experts deemed the following reasons to be appropriate: the patient has persistent or progressive unexplained mild cognitive impairment (MCI), the patient satisfies the core clinical criteria for possible AD but has an unclear clinical presentation (either an atypical clinical course or an etiologically mixed presentation), or the patient has progressive dementia with an atypically early age of onset (usually defined as ≤65 y old) (4). The June 2023 lecture in Chicago indicated that the updated AUC (5) included 2 additional appropriate reasons for ordering an amyloid PET scan for patients requiring documentation of amyloid plaques before being started on AAT, but the details of the 2 new reasons were not specified.

Inappropriate use criteria must also be reviewed to make sure neurologists, dementia experts, and other physicians are not ordering amyloid PET scans for patients inappropriately, subjecting them to unnecessary radiation exposure and increased out-of-pocket costs, as well as incurring unnecessary costs to their insurance company. Inappropriate reasons for performing amyloid PET are evaluation of patients with the core clinical criteria for probable AD and a typical age of onset; determination of dementia severity; evaluation of asymptomatic individuals either with a positive family history of AD or with the ε4 allele of apolipoprotein E; evaluation of patients with a cognitive complaint that is unconfirmed on clinical examination; evaluation of suspected autosomal dominant mutation carriers, in lieu of genotyping; evaluation of asymptomatic individuals; and nonmedical uses (e.g., legal, insurance coverage, or employment screening) (5). Table 1 outlines appropriate and inappropriate reasons for amyloid PET, as well as 17 clinical scenarios and their scores.

CLINICAL INDICATION AND LIMITATIONS OF USE

All 3 FDA-approved amyloid tracers have the same clinical indication described in their package inserts (1–3). ¹⁸F-florbetapir, ¹⁸F-flutemetamol, and ¹⁸F-florbetaben are radioactive diagnostic agents indicated for PET of the brain to estimate amyloid neuritic plaque density in adults with cognitive impairment who are being evaluated for Alzheimer disease (AD) and other causes of cognitive decline. A negative scan indicates sparse to no neuritic plaques and is inconsistent with a neuropathologic diagnosis of AD at the time of image acquisition; a negative scan reduces the likelihood that a patient’s cognitive impairment is due to AD. A positive scan indicates moderate to frequent amyloid neuritic plaques; neuropathologic examination has shown that this amount is present in AD but may also be present in other types of neurologic conditions and in people with normal cognition. Amyloid imaging is an adjunct to other diagnostic evaluations (1–3).

Regarding the limitations of use for all 3 tracers, a positive scan does not establish the diagnosis of AD or any other cognitive disorder (1–3); it just indicates that the patient has amyloid plaques. The safety and effectiveness of the 3 tracers has not been established for predicting the development of dementia or other neurologic conditions or for monitoring responses to therapies. Regarding monitoring therapeutic response, all 3 package inserts were originally written in the mid-2000s, when there were no disease-modifying therapies; there were only drugs, such as donepezil (Aricept; Eisai Co., Ltd.), that helped manage symptoms. Therefore, all 3 package inserts currently state that the tracers do not monitor therapies. Now that there are FDA-approved disease-modifying therapies, the clinical indication for all 3 tracers could be changed, but it is not known whether that will happen.

Another important topic is AQ when monitoring a patient’s response to AAT. Currently, the approved tracers do not have quantification in their package inserts, which state that uptake is interpreted visually (1–3). Once patients on AAT have undergone baseline and follow-up PET, dementia experts will want quantifiable data documenting a successful response. It will be important to have measurable numbers for each area of the brain and not just a visual assessment. Images will need to be compared both visually and quantitatively at baseline and follow-up to see whether the patient is responding to the therapy and whether disease progression is slowing. In this context, AQ and the term centiloids are important for nuclear medicine technologists and interpreting physicians to understand. Currently, AQ and centiloids are used mostly for research purposes, but this article will discuss them because of their importance in measuring a patient’s response to therapy.

PATIENT PREPARATION

Amyloid imaging is probably one of the easiest PET scans a nuclear medicine technologist will ever do. There is no patient preparation, glucose monitoring, or fasting, but the patient must be well hydrated before being injected.

There are no contraindications before performing an amyloid PET scan (1–3); however, knowing whether other radiopharmaceutical-dependent nuclear medicine studies have recently been performed can be helpful to ensure the highest-quality amyloid PET scan with no interference from any other radiopharmaceutical.

DOSE AND ADMINISTRATION

Each tracer has a different recommended dose—370 MBq (10 mCi) for ¹⁸F-florbetapir, 185 MBq (5 mCi) for ¹⁸F-flutemetamol, and 300 MBq (8.1 mCi) for ¹⁸F-florbetaben (1–3)—and each is intravenously injected according to package insert guidelines (Table 2). The dose is not diluted before injection for any of the tracers (1–3). A 0.9% sodium chloride flush is recommended for all 3 tracers to ensure delivery of the proper dose (1–3). After injection, the syringe is rinsed to flush any residual radioactivity into the patient; a 10-mL flush is adequate. Per the package inserts, the 3 tracers have very different injection-to-scan times—vital information if the technologist is to achieve the highest-quality scan for each tracer. The imaging department should pick a scan time after injection that not only meets the package insert guidelines but also best fits the department’s workflow. For example, if the department performs most scans at 45–60 min after injection, amyloid scans can also be performed at 45–60 min after injection as long as the guidelines concur.

Because amyloid tracers look at amyloid plaque burden and not at glucose metabolism, light and sound do not affect uptake; thus, patients undergoing amyloid PET do not need to sit in a quiet, dark room as is necessary with ¹⁸F-FDG PET, nor does physical activity need to be avoided before injection as is necessary with ¹⁸F-FDG. As with all PET procedures, the patient must stay hydrated and void frequently to help decrease radiation exposure after receiving the tracer (1–3). Sedation can be given before imaging if needed, provided the patient has a ride home.

ACQUISITION AND PROCESSING

The acquisition and processing for amyloid PET are fairly simple. First, the existing ¹⁸F-FDG or amyloid brain protocol is cloned, adding the amyloid tracer to the isotope inventory and changing the acquisition time to that required for the tracer (Table 2). The imaging protocol is then saved as “¹⁸F-florbetapir PET,” “¹⁸F-flutemetamol PET,” or “¹⁸F-florbetaben PET.” To acquire adequate counts and ensure a good-quality scan on an older scanner, the patient must be scanned for the longest time suggested in the package insert. For newer scanners, the shorter acquisition time outlined in the package insert is fine. Scanning outside the time parameters of the package insert is not recommended; however, it is ultimately the choice of the interpreting physician, who is responsible for ensuring that the images are of the highest quality. For PET/CT protocols, the CT acquisition parameters are taken from the manufacturer’s recommendations (1–3). Table 3 outlines PET scanner type, energy peak, energy window, and preferred patient positioning for each of the 3 tracers.

Other important acquisition steps in amyloid PET, regardless of tracer, are to have the patient void before scanning; to have the patient remove hearing aids, earrings, and hair clips; and to position the patient supine with the head properly positioned in an off-table head holder. Any gaps between the patient’s head and the head holder should be filled with a blue Chux industrial towel (Chicopee), a rolled-up towel, or a taper positioning cushion; feeling something against their head reminds patients to hold still. Chin and forehead straps also help eliminate head motion, as does immobilization using Coban wrap (3M) or adhesive tape. Any head motion may affect scan quality, as the scan compares white matter to adjacent gray matter, with the latter normally being only 2–4 mm thick (6). Figures 1 and 2 illustrate proper use of the head holder, chin straps, and head straps. The last step before starting the scan is to make sure the entire brain, including the cerebellum, is centered in the field of view (1–3). The table is adjusted to the proper height by making sure the side laser light of the PET scanner is at the level of the patient’s ear, and the patient’s head is straightened by centering it on the second laser light (Fig. 2B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Use of off-table head holder. (A) Canthomeatal line should be vertical and perpendicular to imaging table. (B) Even small deflections of canthomeatal line from vertical position can result in abnormal head tilt. (Reprinted from (22).)

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Patient positioning. (A) Body strap, knee cushion, and blanket applied for patient comfort. (B) Chin strap, head strap, and laser lights applied to position and secure head. Long arrow indicates laser light at level of patient’s ear, assisting in fixing table at proper height. Short arrow indicates laser light centered on patient’s head to keep it straight. (Courtesy of Barbara Grabher and Virtua Health System, Vorhees, NJ.)

Once the scan is complete, the technologist reviews the raw data for motion—a simple but essential step. Patient motion cannot be corrected on a PET scanner as it can on a SPECT scanner and should be evaluated in both a negative and a positive amyloid scan. If a thin, well-delineated scalp line can clearly be seen along with an adjacent ring (the skull and cerebrospinal fluid), there is no patient motion. If the scalp line is thick or abuts the cerebral cortex, there is patient motion (the cerebral cortex would never abut the scalp because they are separated by the skull and cerebrospinal fluid). If motion is suspected, the interpreting nuclear medicine physician is asked to decide whether the scan can be read or needs to be repeated. Figure 3 illustrates examples of patient motion and a positive and negative scan with no patient motion.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Examples of patient motion on static (A–C) and dynamic (D and E) images. (D) No motion. (E) Patient motion indicated by arrows. (F) Gold standard of motion-free negative (left) and positive images (right). Yellow arrows in A and B indicate thick scalp. Red arrow in C illustrates that tracer uptake does not reach the scalp. Blue arrow in C shows activity shifts (bright/dark) on same structure. Yellow arrows in F indicate scalp, and blue arrow indicates skull and cerebral spinal fluid. (A–C and F courtesy of Life Molecular Imaging; D and E reprinted from (23).)

In the processing of an amyloid PET scan, the PET scanner manufacturer’s settings are followed unless the interpreting physician wishes the iterations and subsets changed to produce smoother or higher-contrast images. However, once the interpreting physician agrees on the filter settings, all future amyloid PET scans should be processed using those settings. Nuclear medicine technologists should not change the filters, because each technologist’s eye is different and the interpreting physician must receive consistent images each time. Any questions about reconstruction parameters for a particular scanner are directed to the scanner manufacturer.

TRACER DISTRIBUTION PATTERNS

Normal and abnormal distribution patterns for the 3 tracers are similar (Fig. 4). A normal pattern is one in which the white matter and the gray matter can be differentiated; an abnormal pattern is one in which they cannot be differentiated. ¹⁸F-flutemetamol is the only FDA-approved tracer that requires assessment of striatal binding. Figure 4F illustrates the striatal gap and striatal bridge.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Normal and abnormal distribution patterns of uptake of ¹⁸F-florbetapir (A and B), ¹⁸F-flutemetamol (C–E), and ¹⁸F-florbetaben (G and H). Normal patterns of tracer uptake are similar where you can differentiate white matter from gray matter. In abnormal pattern of tracer uptake, you cannot differentiate white matter from gray matter. (F) Assessment of striatal binding, as required by FDA for Vizamyl only. White arrows denote level of transverse slice. Yellow dashed lines and arrowheads in G and H denote areas taken directly from florbetaben package insert. Cerebellum: contrast between white matter (arrows) and gray matter is seen in both negative and positive scans. Extracerebral tracer uptake in scalp and in posterior sagittal sinus (arrowhead) can be seen. Lateral temporal lobes: Spiculated or “mountainous” appearance of white matter (arrows) is seen in negative scan, and radioactive signal does not reach outer rim of brain (dashed line) because of lower tracer uptake in gray matter. Positive scan shows “plumped”, smooth appearance of outer border of brain parenchyma (dashed line) due to tracer uptake in gray matter. Frontal lobes: Spiculated appearance of white matter in frontal lobes (arrows) is seen in negative scan. Positive scan shows that tracer uptake in these regions has “plumped”, smooth appearance due to increased gray matter signal (dashed line). Posterior cingulate/precuneus: Adjacent and posterior to splenium (arrow), these regions appear as hypointense “hole” (circle) in negative scan, whereas this hole is “filled-up” (circle) in positive scan. Parietal lobes: In negative scan, midline between parietal lobes can be easily identified (long arrow); white matter has spiculated appearance (short arrow) with low signal near outer rim of brain (dashed line). In positive scan, midline between parietal lobes is much thinner. Cortical areas are “filled-up” and are smooth in appearance as tracer uptake extends to outer rim of brain. f = frontal; lt = lateral temporal; s = striatal region; pc = posterior cingulate region; cc = callosum; p = pons; ip = inferior parietal. (A and B reprinted from (1), C–E reprinted from (2), F reprinted with permission of (24), and G and H reprinted from (3).)

IMAGE INTERPRETATION

Although training is not required to read amyloid images, it is highly recommended. Interested physicians can register for the training through the tracer sales representative or website (Table 4). Training is online, self-paced, and accessible anytime.

Amyloid PET scans are interpreted as either amyloid-negative or amyloid-positive, based on visual assessment of uptake in the white matter and gray matter (1–3). If the two can be differentiated, the scan is negative (1–3); if they cannot, the scan is positive. The package insert interpretation guidelines must be followed, as different tracers require a different number of regions to be positive for a scan to be considered positive (1–3). Figure 4 illustrates positive and negative images for the 3 tracers. Both false positivity and false negativity can occur, as discussed in “Benefits and Limitations of Amyloid PET” below.

WARNINGS, PRECAUTIONS, AND ADVERSE REACTIONS

For any radiopharmaceutical being injected into a patient, the warnings, precautions, and adverse reactions in the package insert should be reviewed carefully. Each amyloid tracer has different warnings (1–3), but several points are mentioned in the package inserts for all 3 tracers. The first is the possible risk for image misinterpretation and other errors due mostly to atrophy, which limits the ability to distinguish gray and white matter on an amyloid scan, and patient motion artifacts that can distort the amyloid images. The second is that image interpretation should be performed independently of the patient’s clinical information. Finally, like other radiopharmaceuticals, all injected amyloid tracers contribute to the patient’s overall long-term cumulative radiation exposure, which is associated with an increased risk of cancer. It is crucial to ensure safe handling of the tracer to protect patients and health care workers from unintentional radiation exposure (1–3).

Most adverse events are relatively minor for all 3 tracers, and there were no serious adverse events in the clinical studies for ¹⁸F-florbetapir (555 patients) or ¹⁸F-florbetaben (872 patients with >1,090 injections). Adverse events reported for ¹⁸F-florbetapir and ¹⁸F-florbetaben were predominantly mild to moderate and included headache (1.8%), musculoskeletal pain (0.7%), and increased blood pressure (0.7%) for ¹⁸F-florbetapir and injection site–related erythema (1.7%), irritation (1.1%), and pain (3.4%) for ¹⁸F-florbetaben (1,3).

According to the prescribing information for ¹⁸F-flutemetamol, the most commonly reported adverse events in clinical trials (761 patients) were flushing (2%), headache (1%), increased blood pressure (2%), nausea (1%), and dizziness (1%). Also important is that the warning and precautions section (section 5.1) of the ¹⁸F-flutemetamol package insert states that “Hypersensitivity reactions may occur: Ask patients about prior reactions to Vizamyl™. Observe for hypersensitivity signs and symptoms following Vizamyl™ administration. Have resuscitation equipment and trained personnel available at time of Vizamyl™ administration.” (2).

USE IN SPECIAL PATIENT POPULATIONS

Amyloid tracer use in specific patient populations is also important to review (section 8 of each package insert) (1–3).

AQ AND CENTILOIDS

AQ is not currently FDA-approved and is not mentioned in the package inserts for any of the 3 approved tracers. With the FDA approval of AATs, AQ may play an important role in monitoring a patient’s response to therapy. Currently, AQ is used primarily in a research setting and seldom in daily clinical practice, with many interpreting physicians and nuclear medicine technologists not understanding what AQ is or how to use it. Some have never even used AQ software. Many also do not know what centiloids are or their importance in baseline and follow-up amyloid PET scans monitoring patients on AATs.

AQ uses a concept similar to that in myocardial perfusion imaging quantification, in which patient images are compared with a reference database. Each amyloid tracer has its own reference database. SUVs and SUV ratios (SUVRs) are created for a patient’s scan and then compared with the reference database. When AQ is used to determine response to AAT, baseline and follow-up PET scans are compared. A visual assessment of uptake is not sufficient.

Most AQ requires MRI. MR images provide important anatomic information, but some programs have been developed without MRI. The basic AQ algorithm consists of several steps: coregistration of the PET and MR images, segmentation of the MR image, voxel-to-voxel intersection of the segmented MR image to the predefined atlas to obtain individual regions of interest, application of the individual regions of interest onto the coregistered PET image, and extraction of regional values (Fig. 5) (7).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

AQ steps. (1) Coregistration of PET and MR images. (2) Segmentation of MR image. (3) Voxel-to-voxel intersection of segmented MR to predefined atlas to obtain individual regions of interest. (4) Application of individual regions of interest onto coregistered PET image and extraction of regional values. (Reprinted with permission of (7).)

There are several ways to quantify amyloid burden in the brain, the most widely used being SUVs and SUVRs (8). SUVRs provide information on uptake within a particular region of the brain. This method involves computing the ratio of uptake between the target and a predefined reference region. To be considered a reference, a region must display specific characteristics: no specific tracer binding, tissue characteristics and kinetic properties similar to those of the target region, and uptake unaffected by the disease state. The cerebellar cortex is, in most cases, a suitable reference region for amyloid tracers (8), but each tracer has its own defined reference region. That for ¹⁸F-florbetapir and ¹⁸F-florbetaben is the cerebellum (1,3), and that for ¹⁸F-flutemetamol is the pons set to 90% of the color scale for the images (2). To correctly adjust the display scales when reading ¹⁸F-flutemetamol images, the ¹⁸F-flutemetamol package insert must be reviewed. Figure 6 illustrates how SUVs and SUVRs are calculated, where the reference regions are, and where the corresponding target regions are.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

(A) How SUV and SUVRs are calculated for particular brain regions. (B) Defined reference regions and 7 defined target regions. (A courtesy of Life Molecular Imaging Medical Affairs; B reprinted with permission of (9).)

Another important AQ-related term is centiloids. The centiloid method of AQ enables quantitative values from amyloid imaging to be expressed in a universal unit (8). Centiloids provide a way to normalize amyloid levels across all amyloid tracers, allowing them to be compared with each other (i.e., comparing apples to apples instead of apples to oranges). Each tracer being converted to centiloids must be compared with ¹¹C-Pittsburgh compound B first (8), and then centiloids are calculated for each tracer’s SUVR using a specific equation (9). Table 5 outlines the centiloid equations for ¹¹C-Pittsburgh compound B, ¹⁸F-florbetapir, ¹⁸F-flutemetamol, and ¹⁸F-florbetaben, and Figures 7A and 7B illustrate how the Clarity AD study used centiloids to document one patient’s response to placebo at baseline and 18 mo and another patient’s response to lecanemab at baseline and 18 mo. In the placebo subject, centiloids went up, whereas in the subject receiving lecanemab, centiloids went down, documenting a positive response and a decrease in amyloid plaques. Centiloid averages range from 0 for young, healthy people to 100 for mild AD (9). On the centiloid scale, there is a range—the gray zone—at which the values are not totally normal but are also not totally abnormal (Fig. 7C). The higher end of the gray zone (20–35 centiloids) can help detect emerging amyloid plaques (10).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

Clarity AD study using centiloids (CLs) to document patient’s response to therapy. (A) Placebo subject images showing increase in CLs both at baseline and at 18 mo. (B) Lecanemab subject images showing decrease in CLs between baseline and 18 mo, indicating positive response to therapy. (C) Explanation of CL scale, (−20) to (+20), on scale indicative of amyloid-negative scan. CL levels above 35 indicate amyloid-positive scan. Gray zone of CL scale is where amyloid scan is not totally negative but also not totally positive. Gray zone is from (+21) to (+35). Upper levels of gray zone can help detect emerging amyloid plaques (arrow). (C reprinted with permission of (10); A and B reprinted with permission of Life Molecular Imaging.)

There are many benefits to AQ. It allows for harmonization of the 3 tracers; they can be compared with each other using a normalized unit expressed in centiloids. AQ can also compare different brain regions derived from SUVRs, objectively assessing amyloid accumulation over time either as a percentage change or as the overall centiloid increase. AQ can be further applied to the whole brain or to specific regions of interest (11). Earlier detection of emerging amyloid plaques may be possible by looking at centiloid scores in the gray zone. AQ can be used as an adjunct to visual assessment, allowing for objective confirmation. Inexperienced readers may find AQ helpful when they are not confident about their visual assessment. AQ can also help when amyloid levels are close to pathologic thresholds—levels at which visual assessment alone can be difficult. All amyloid tracers require detection of a threshold level of amyloid on visual assessment; that threshold can be quantified using centiloids. Lower levels of amyloid can be visually assessed by some experienced readers, though there is always some level of risk when the levels are too low (9,10,12). AQ may also be helpful in challenging reads such as in a patient with atrophy. Outside of a binary visual interpretation of a scan as just positive or negative, AQ can provide longitudinal information and objectively assess changes over time. Last and most important, AQ can help evaluate whether a patient is responding to AATs over time. All these AQ benefits are off-label for the 3 tracers and mostly used in research only; nonetheless, AQ may play an important future role in amyloid PET, especially when evaluating response to AATs.

There are various approved (FDA 510k and Conformité Européenne mark) software packages for AQ, but because it is used mainly in research, the scanners in many PET departments do not have AQ software installed. Departments interested in performing amyloid PET scans on AAT patients may wish to install such software so that therapeutic response can be quantified. PET scanner manufacturers can provide pricing for installation of a neuro package with a reference database for the amyloid tracer of interest. Table 6 outlines some of these packages; others, available for research settings, include PMOD (https://www.pmod.com/web/), MIAKAT (https://nmmitools.org/2019/01/01/miakat/), CapAIBL (https://aehrc.csiro.au/), ELBA (Evaluation of Brain Amyloidosis), and NiftyPET (https://github.com/NiftyPET/NiftyPET) (9).

BENEFITS AND LIMITATIONS OF AMYLOID PET

There are many benefits to amyloid imaging regardless of the choice of tracer; all 3 tracers are sensitive, specific, and accurate for detecting amyloid plaques. Amyloid imaging is fairly noninvasive compared with a lumbar puncture for documenting the presence of amyloid plaques, which can be detected 10–15 y before the onset of cognitive issues. According to the 2024 Alzheimer’s Disease Facts and Figures, a patient can start to have brain changes 20 y or more before showing any signs of a cognitive issue (13–15).

Although amyloid imaging with all 3 tracers has many benefits, it does have some limitations. The package inserts for all 3 tracers list the same limitations of use: a positive scan does not establish a diagnosis of AD or other cognitive disorders, and safety and effectiveness have not been established for predicting the development of dementia or other neurologic conditions or for monitoring therapeutic response (1–3).

Although approximately 80% of the amyloid scans that physicians read will be fairly easy to interpret as positive or negative, in approximately 15%–20% of cases there will be challenges such as patient motion, a head tilt, or brain atrophy making it difficult to differentiate white matter from gray matter (16). Other image artifacts that can challenge amyloid scan interpretation include patients with meningitis or meningiomas, which avidly take up tracer. Rapid metabolism in vascular structures such as the sagittal sinuses, a prominent scalp, and facial activity may also be seen (17).

Both false-negative and false-positive scans can occur. Binding of amyloid tracers does not always indicate the presence of amyloid plaques. For instance, binding may also occur in individuals with cerebral amyloid angiopathy rather than AD (18). A patient may have a positive amyloid scan with moderate to frequent amyloid plaques but not have AD; older individuals with normal cognition may have amyloid plaques but not AD—their plaques are because of their older age. False positives may also occur in people with normal-pressure hydrocephalus, which can lead to mildly to moderately positive scans. Patients presenting with MCI for other reasons (e.g., depression or low vitamin B₁₂) and are older could have scans positive for amyloid plaques but the plaques are incidental. Patients with severe atrophy can have false-negative amyloid scans, with fingerlike projections of gray matter mimicking normal white matter activity, but this is usually a problem only in very advanced disease (17). Overprocessing of images using the wrong number of iterations and subsets, with too much information being removed from the scan, can also cause a false-negative. False-positive scans can happen if the images are too smooth as well. The PET scanner must initially be set up using the manufacturer’s recommend reconstruction parameters and then can be tweaked on the basis of the interpreting physician’s feedback. However, the manufacturer’s settings are usually fine, with no changes needed.

The sensitivity and specificity of the 3 amyloid tracers are fairly high per their respective package inserts. ¹⁸F-florbetapir has a sensitivity of 92% and a specificity of 100% (1), ¹⁸F-flutemetamol has a sensitivity of 91% and a specificity of 90% (2), and ¹⁸F-florbetaben has a sensitivity of 98% and a specificity of 80% (3).

MARKET ACCESS AND REIMBURSEMENT SUPPORT

The amyloid landscape has changed, passthrough status for all 3 tracers has long expired, and Medicare coverage decisions for amyloid imaging have also changed. Imaging facilities can find it challenging to be paid properly for the amyloid PET scans they perform. If billing as a hospital, one needs to understand and navigate additional challenges such as bundled payment of the radiopharmaceutical, with the cost of the radiopharmaceutical being bundled into the cost of the procedure and the radiopharmaceutical not being paid separately as it is in the outpatient imaging setting. In the new era of amyloid imaging, in which we are documenting whether a patient has amyloid plaques before being started on AATs, many nuclear medicine departments are performing significantly more amyloid scans than previously. When first imaging with any of the amyloid tracers, one must confirm that the imaging center and billing department have the proper billing and coding information, any prior authorization information, and any other information needed for insurance companies to process the claim, so the imaging center gets paid for the scan.

With the approval of ¹⁸F-florbetapir, ¹⁸F-flutemetamol, and ¹⁸F-florbetaben by the FDA in 2012, 2013, and 2014, respectively, and the expiration of each tracer’s passthrough status 3 y later, the issue was what insurance companies would establish in terms of coverage for these tracers. In September 2013, CMS established a National Coverage Determination outlining coverage for amyloid PET in dementia and neurodegenerative disease for those covered by traditional Medicare. The coverage was restrictive; beneficiaries were limited to a single amyloid PET scan per lifetime, had to have a diagnosis from a specific listing, and had to be enrolled in a “coverage with evidence development” clinical research study that would assist in the development of potential new therapies. Two such research studies were Imaging Dementia—Evidence for Amyloid Scanning (IDEAS study), which enrolled patients from February 2016 to December 2018, and the New IDEAS study, which enrolled patients from December 2021 to March 2024. Data were collected through these research programs, which were supported by the American College of Radiology. Under that support, a patient’s out-of-pocket costs for a scan, such as copays and deductibles, would be paid by American College of Radiology. The American College of Radiology also continued to advocate for individuals covered by Medicare Advantage plans to have access to testing, since these plans were obliged to follow the CMS guidelines. This process continued for nearly a decade, providing confirmation of an AD diagnosis for hundreds of thousands of individuals.

It is estimated that Medicare-aged Americans number almost 66 million, with almost 50% being covered by a Medicare Advantage plan (19). This means that half the population over 65 y old would be covered on the basis of guidelines established by CMS and enforced by regional Medicare administrative contractors. The other half would be at the mercy of the policy adopted by their private insurance company. These private carriers often look to CMS for a general guide before establishing more restrictive coverage policies for their enrollees. In the absence of an established policy, these plans are instructed to follow CMS guidelines. At the outset, it seemed there was no clear path forward for almost 50% of those who needed access to imaging to document amyloid plaques. Discussions between CMS and private insurance companies began, and meetings occurred with medical directors to establish that a physician had to prove medical necessity before these scans would be covered by private insurers.

With the accelerated approval of new monoclonal antibody therapies (20), CMS established a “coverage with evidence development” ruling for them in April 2022. Increased access to amyloid imaging would clearly be needed, as it was one of the diagnostic tests available to determine a patient’s candidacy for treatment. Under the established ruling, a patient would need a positive amyloid scan or lumbar puncture to be enrolled in a national registry that granted coverage for the therapy. Private insurers also began establishing coverage for not only the testing but also the therapy itself.

In October 2023, CMS decided to end the national coverage decision for amyloid imaging. Private insurance companies were now responsible for establishing and enforcing their own coverage criteria for amyloid PET, and direction and guidance for traditional Medicare enrollees were provided. CMS released what is known as a transmittal, or instructions on how each Medicare administrative contractor should cover claims. Retiring the national coverage decision removed the limit of a single amyloid scan per lifetime, the restriction of a specific set of diagnostic codes, and mandatory enrollment into a registry. Traditional Medicare enrollees could now undergo amyloid PET simply to confirm a diagnosis of AD even if they decided not to proceed with the newly approved therapies; their physician needed only to document that the scan was medically necessary (21).

Although traditional Medicare is paying for medically necessary amyloid scans, private insurance companies are still setting up guidelines and rulings regarding approval, and reimbursement has yet to be seen. Some of the larger insurance companies, such as UnitedHealthcare, Aetna, and Cigna, have established policies, whereas other companies are still working to recognize the benefit of an early diagnosis, along with treatment. An argument has been made that early treatment can reduce the cost of care for an enrollee who must live in a facility for those with cognitive issues. At the same time, the cost of the therapies seems to be at the forefront of the minds of those who write medical policies for coverage. As physicians continue to put their patients on AATs and order amyloid PET scans to document plaques, there is no doubt that reimbursement and insurance policies are a fluid situation that will continue to evolve as more data become available.

To help make sure imaging centers get the support they need before, during, and after performing an amyloid PET scan, Eli Lilly, GE HealthCare, and Life Molecular Imaging have created market access and reimbursement services for their respective radiopharmaceuticals (Table 7). Reimbursement services help imaging centers effectively submit proper insurance documentation to successfully get their insurance claims processed and paid. Imaging support with peer-to-peer image interpretation assistance helps interpreting physicians new to amyloid imaging obtain answers to their imaging interpretation questions.

DISCLOSURE

No potential conflict of interest relevant to this article was reported.