Y-90 PET/CT After Radioembolization

Introduction

Y-90 is a near-pure beta emitter with no useful gamma ray, yet a vanishingly small branch of its decay lets a modern PET/CT scanner image exactly where the microspheres landed — with resolution and quantitative accuracy that bremsstrahlung SPECT cannot match. Post-radioembolization Y-90 PET/CT is how a program turns a therapy delivery into a verified, dose-mapped result.

Transarterial radioembolization (also called selective internal radiation therapy, or SIRT) delivers millions of Y-90-loaded microspheres into the hepatic arterial supply of liver tumors. The therapeutic work is done by short-range beta particles. But the moment the delivery is over, two questions matter: did the microspheres go where they were supposed to, and what absorbed dose did the tumor and the normal liver actually receive? Answering those questions requires imaging a radionuclide that, on paper, has almost nothing for an imaging system to detect. ¹²

This guide explains the decay physics that make Y-90 PET possible, why the images are inherently count-starved, how PET compares with bremsstrahlung SPECT, how voxel dosimetry is done with the local deposition method, and the quality control and regulatory framework that make post-therapy dose numbers defensible. DRPS supports these programs through its PET/CT and nuclear medicine physics and medical physics consulting services across Florida, Maryland, Virginia, Washington DC, California, Nevada, Pennsylvania, New York, New Jersey, and Delaware.

Topic Explanation

The decay physics: how a beta emitter makes a PET signal

Y-90 decays to zirconium-90 (Zr-90) with a half-life of about 64.05 hours (2.67 days) and a beta endpoint energy of 2.28 MeV, with a mean beta energy near 0.934 MeV. Almost all of that decay energy is carried by beta particles that stop within a few millimeters of tissue — ideal for therapy, useless for external imaging because there is no characteristic gamma ray. ³

The imaging opportunity comes from a tiny side channel. A very small fraction of Y-90 decays populate the 0⁺ first excited state of Zr-90 at about 1.76 MeV. A 0⁺-to-0⁺ transition cannot emit a single gamma ray, so that excited state instead de-excites by internal pair production: it creates an electron-positron pair. The positron then annihilates like any other, producing the two back-to-back 511 keV photons that a PET scanner is built to detect. The catch is abundance: only about 32 positron-producing decays per million occur — roughly $3.2 \times 1 0^{- 5}$ per decay. ²⁷

That single number explains almost everything about Y-90 PET. It is why the images are count-starved, why acquisitions run long, and why reconstruction and scanner performance matter so much. For a broader view of how positron range and detector design shape PET, see our note on common PET and RPT isotopes.

Two ways to image Y-90 — and why PET usually wins for quantification

Y-90 can be imaged two ways after therapy:

Bremsstrahlung SPECT/CT. As beta particles decelerate in tissue, they emit a continuous bremsstrahlung X-ray spectrum. A gamma camera can form an image from these photons, but the broad, continuous spectrum with no photopeak makes scatter and energy-window choices difficult and quantification poor. With optimized protocols it is clinically useful for confirming gross distribution. ²
Y-90 PET/CT. By exploiting the internal-pair-production positrons, PET produces higher-resolution images and, importantly, images that can be quantified in becquerels per milliliter and converted to absorbed dose. Multiple studies report PET quantification superior to bremsstrahlung SPECT. ²⁴⁶

Our companion guides on Y-90 bremsstrahlung SPECT imaging and Y-90 radioembolization dosimetry go deeper on each; this article focuses on the PET pathway and its dosimetry.

Key Technical Principles

The count-starved regime

Because only about 32 in a million decays yield a positron, the coincidence signal is orders of magnitude weaker than a routine F-18 FDG study at comparable activity. Several consequences follow:

Long acquisitions. Bed times are far longer than for FDG to accumulate adequate counts.
Reconstruction sensitivity. Iteration/subset choices and post-reconstruction filtering trade noise against quantitative accuracy, and the optimal choice depends on object size. Phantom work with digital photon-counting PET found that a small post-reconstruction Gaussian filter preserved dosimetric accuracy while reducing noise, and that acquisition beyond about 10 minutes per bed reduced noise without materially changing quantification. ⁷
Time-of-flight (TOF) helps. TOF improves quantitative accuracy at lower activity concentrations by localizing each annihilation along the line of response, which is exactly the regime Y-90 lives in. ⁷
Scanner background matters. LYSO-based PET detectors contain intrinsic radioactivity (Lu-176) that adds a background; its effect is negligible at the high activity concentrations of radioembolization but is a consideration for accurate low-concentration quantification. ⁷

Local deposition dosimetry and a worked dose calculation

Because the Y-90 beta range is short relative to typical PET voxels, the local deposition method (LDM) is a natural and accurate model: assume each voxel absorbs all the energy emitted by the activity it contains. The absorbed dose to a region then follows directly from the mean energy per decay and the cumulated activity.

Start from the mean absorbed dose to a uniformly treated mass $m$ :

\overset{ˉ}{D} = \frac{A ~ E ˉ}{m}

where $\tilde{A}$ is the cumulated activity (total number of decays) and $\overset{ˉ}{E}$ is the mean energy deposited per decay. For complete physical decay of an administered activity $A_{0}$ , the cumulated activity is $\tilde{A} = A_{0} / λ = A_{0} T_{1/2} / ln 2$ .

For Y-90, $\overset{ˉ}{E} = 0.934 MeV = 1.496 \times 1 0^{- 13} J$ and $T_{1/2} = 64.05 h$ , so the mean life is $T_{1/2} / ln 2 = 92.4 h = 3.326 \times 1 0^{5} s$ . The energy released per 1 GBq that fully decays and deposits locally is:

E = (1 \times 1 0^{9} s^{- 1}) (3.326 \times 1 0^{5} s) (1.496 \times 1 0^{- 13} J) \approx 49.8 J

Dividing by 1 kg gives the well-known Y-90 conversion factor:

\overset{ˉ}{D} (Gy) \approx 49.8 \frac{A _{0} ( GBq )}{m ( kg )}

Worked example. If 1.5 GBq of Y-90 is retained in a 1.8 kg perfused liver volume, the mean absorbed dose is $49.8 \times 1.5/1.8 \approx 41.5$ Gy. In practice the activity is not uniform, which is the whole point of PET: it maps the real distribution voxel by voxel, and the local deposition method converts each voxel's activity to dose. Anchoring the PET activity map to the known administered activity ("LDM with known activity") is reported as the preferred approach because it sidesteps the hardest parts of absolute PET quantification. ⁵

Getting the activity right: motion and co-registration

Absolute Y-90 PET quantification is challenged by respiratory motion and PET/CT co-registration, which blur and displace the liver. A 35-patient study found that summing image-based activity over the perfused volume underestimated the administered activity by roughly 23%, and that only after expanding the volume of interest by about 2 cm to capture motion-blurred counts did the image-based activity match the administered activity. ⁵ The practical implication: dose metrics are sensitive to how the volume of interest is drawn, and a program needs a consistent, validated method.

PET-versus-SPECT dosimetry: how much does the modality change the numbers?

When the same patients are imaged both ways, the dose distributions correlate strongly but are not identical. A 35-patient comparison of Y-90 PET/CT and Y-90 SPECT/CT voxel dosimetry found strong correlation for perfused-volume mean dose (with essentially zero bias), while PET yielded, on average, about 14% higher mean dose to tumors and about 9% lower mean dose to perfused normal liver than SPECT. ⁴ Those differences are large enough to matter when comparing against dose-response thresholds, which is why the imaging modality and dosimetry method must be stated whenever a dose number is reported.

A comparison of the two post-therapy imaging routes

Feature	Bremsstrahlung SPECT/CT	Y-90 PET/CT
Signal detected	Continuous bremsstrahlung X-rays (no photopeak)	511 keV annihilation pairs from internal pair production
Fraction of decays imaged	Broad, low-contrast continuum	≈ 32 positrons per million decays ²⁷
Spatial resolution	Lower	Higher ²
Quantitative accuracy	Limited (scatter, no photopeak)	Better; supports voxel dosimetry ²⁴
Benefit of time-of-flight	Not applicable	Improves accuracy at low activity concentration ⁷
Typical role	Confirm gross distribution when PET unavailable	Quantitative distribution and absorbed-dose mapping ²⁴

Clinical Impact

Post-therapy Y-90 imaging is not a formality — it changes what a program knows and does:

Verifying delivery and catching non-target deposition. PET/CT confirms the microspheres are in the intended tumor-bearing segments and helps detect unintended extrahepatic deposition, informing follow-up.
Building dose-response evidence. Y-90 PET voxel dosimetry has established radiobiologically meaningful relationships. In hepatocellular carcinoma treated with resin microspheres, complete responses were generally achieved at a minimum dose to 70% of tumor volume (D70) above about 100 Gy, with incomplete responses generally below that, and smaller tumors reaching the threshold more readily. ⁶
Personalizing future treatment. Because PET quantification is accurate enough to correlate intended and delivered tumor doses closely under good conditions, it supports moving from fixed prescriptions toward personalized dosimetry. ⁶

The quantitative-imaging QC that underlies all of this — calibration, cross-calibration with the dose calibrator, and NEMA-based performance verification — is the same discipline described in our PET/CT NEMA NU-2 performance testing guide.

Practical Optimization Tips

1. Treat the PET scanner's quantitative calibration as sacred

Y-90 dosimetry is only as good as the scanner's activity calibration and its cross-calibration to the dose calibrator and clock. Maintain the well-counter/dose-calibrator/scanner calibration chain and verify it on the schedule your program's QC plan defines. ¹

2. Use time-of-flight and validated reconstruction settings

Because Y-90 is count-starved, enable TOF where available and lock down reconstruction parameters (iterations, subsets, filter) validated against a phantom for your object sizes rather than borrowing FDG settings. ⁷

3. Standardize the volume-of-interest method

Given the roughly 23% activity underestimation from motion and co-registration before volume expansion, define and document a consistent VOI approach (including any expansion to capture motion-blurred counts) so dose numbers are reproducible. ⁵

4. Anchor to the administered activity

Prefer the local deposition method anchored to the known administered activity. Determine the delivered activity from pre- and post-procedure vial/line measurements, and use that as the ground truth the PET map is scaled to. ¹⁵

5. Always state modality and method with any dose number

Because PET and SPECT dosimetry can differ by roughly 10–15% for tumor and normal-liver means, a reported dose is meaningless without stating the imaging modality and dosimetry method used. ⁴

Common pitfalls to avoid

Borrowing FDG protocols. Y-90's count statistics are nothing like FDG's; acquisition and reconstruction must be tuned for the low-count regime. ⁷
Ignoring respiratory motion. Uncorrected motion and tight VOIs systematically underestimate activity and dose. ⁵
Comparing PET doses to SPECT-derived thresholds (or vice versa). Thresholds are method-specific. ⁴⁶
Treating SPECT and PET as interchangeable for quantification. SPECT confirms distribution; PET is generally the quantitative tool. ²
Skipping the calibration chain. An uncalibrated or drifting scanner produces confident but wrong dose maps.

Regulatory Considerations

Y-90 microsphere radioembolization is medical use of byproduct material, regulated by the NRC or an Agreement State under 10 CFR Part 35 — generally as an other-medical-use procedure under 35.1000 with device- and manufacturer-specific licensing and training conditions. Establishing a program is a substantial undertaking.

Key frameworks:

AAPM Medical Physics Practice Guideline 14.a, "Yttrium-90 microsphere radioembolization." Defines minimum practice for establishing and supporting a program: pre-treatment mapping (Tc-99m lung-shunt study), dose-calibrator assay traceable to a national standard, treatment-suite posting and staff dosimetry, post-procedure surveys, and determination of delivered dose from pre/residual exposure-rate measurements — plus the license amendment and training required. ¹
10 CFR Part 35 (including 35.1000). Governs authorized users, written directives, radiation safety, and recordkeeping for the medical use, with Y-90 microspheres handled under the other-medical-use pathway and manufacturer training. ⁹
NRC licensing guidance for Y-90 microsphere brachytherapy sources and devices. Provides the device-specific expectations facilities must meet. ¹⁰
NEMA NU 2-2024 (the current revision, superseding NU 2-2018) defines the standardized PET performance measurements — resolution, sensitivity, count-rate/NECR, image quality — that underpin quantitative confidence. ⁸
ICRP Publication 107 is the authoritative nuclear-decay-data source for the Y-90 half-life, beta energies, and the internal-pair-production branch used in dose calculations. ³

Agreement State versus direct-NRC jurisdiction matters. Of the states DRPS serves, Florida, Maryland, Virginia, California, Nevada, Pennsylvania, New York, and New Jersey are NRC Agreement States that license medical use under their own radiation-control programs, while Washington, DC and Delaware are regulated directly by the NRC for radioactive material. The radiation-safety side of a program — treatment-suite controls, surveys, waste, and patient-release considerations — is covered in our Y-90 radioembolization radiation safety guide, and is coordinated with radioactive material license support and radiation safety officer services.

Frequently Asked Questions (FAQs)

Can Y-90 be imaged with PET if it is a beta emitter?

Yes. Y-90 is predominantly a beta emitter, but a very small fraction of decays populate an excited state of zirconium-90 that de-excites by internal pair production, releasing a positron-electron pair. That positron annihilates and produces the 511 keV photon pair that a PET scanner detects, so PET can image the Y-90 microsphere distribution directly.

Why is Y-90 PET count-starved?

Only about 32 decays per million produce a positron, so the available PET signal is orders of magnitude lower than a conventional F-18 study. This low branching ratio makes Y-90 PET count-starved, which is why acquisition times are long, reconstruction settings matter, and time-of-flight capability helps at lower activity concentrations.

Is Y-90 PET better than bremsstrahlung SPECT?

For quantitative post-therapy dosimetry, PET/CT generally outperforms bremsstrahlung SPECT. Bremsstrahlung imaging suffers from a continuous, low-contrast photon spectrum that is hard to quantify, while PET gives higher spatial resolution and more accurate quantification. SPECT/CT with optimized protocols remains adequate for confirming distribution when PET is unavailable.

What is the local deposition method?

The local deposition method assumes each voxel absorbs all the energy of the Y-90 decays it contains, which is a good approximation because the Y-90 beta range is short relative to typical voxel sizes. Absorbed dose is then obtained directly from the activity map, and anchoring the map to the known administered activity is the most accurate approach for Y-90 PET/CT dosimetry.

Does time-of-flight matter for Y-90 PET?

Yes, especially at the low count rates typical of Y-90. Time-of-flight reconstruction improves quantitative accuracy at lower activity concentrations by better localizing each annihilation event, which helps in the count-starved regime that characterizes Y-90 imaging.

What absorbed doses correlate with tumor response?

In hepatocellular carcinoma treated with resin microspheres, published Y-90 PET dosimetry found that complete responses were generally achieved when the minimum dose to 70 percent of the tumor exceeded about 100 Gy, while lower doses were associated with incomplete response. Absorbed-dose thresholds vary by tumor type, microsphere product, and method, so program-specific analysis is essential.

What regulatory requirements apply to Y-90 microsphere therapy?

Y-90 microsphere radioembolization is regulated by the NRC or an Agreement State under 10 CFR Part 35, generally as an other-medical-use procedure (35.1000) with manufacturer training and licensing conditions. Establishing a program requires a radioactive material license amendment, authorized-user and staff training, dose-calibrator assay traceable to a national standard, and documented survey and dose-determination procedures.

How is the delivered activity determined after treatment?

Delivered activity is typically determined from the difference between the assayed activity in the delivery vial before treatment and the residual activity in the vial and administration set afterward, using pre- and post-procedure exposure-rate measurements. Y-90 PET/CT then provides the spatial distribution of the activity that reached the patient.

Key Takeaways

A tiny decay branch makes Y-90 PET possible. Internal pair production yields about 32 positrons per million decays — enough to image, but count-starved. ²⁷
PET generally beats bremsstrahlung SPECT for quantification. Higher resolution and true activity quantification support voxel dosimetry; SPECT confirms distribution. ²⁴
The local deposition method is the natural dosimetry model. Short beta range means each voxel absorbs its own energy; anchoring to administered activity is preferred. ⁵
The Y-90 dose conversion is about 49.8 Gy·kg/GBq. For uniform deposition, mean dose ≈ 49.8 × A/m. ³
Method and modality change the numbers. PET and SPECT tumor/normal-liver mean doses can differ by roughly 10–15%, so always report both. ⁴
Regulation is substantial. A Part 35 (35.1000) program needs a license amendment, training, a traceable dose-calibrator assay, and documented surveys. ¹⁹

Conclusion

Y-90 radioembolization is one of the few places in nuclear medicine where the imaging physics feels almost paradoxical: a therapy radionuclide with no imaging gamma ray, imaged anyway, because roughly one decay in thirty thousand throws off a positron. That thin signal, captured on a well-calibrated PET/CT and interpreted with the local deposition method, turns a delivery into a verified, dose-mapped result — and increasingly into the evidence base for personalized dosimetry.

The physics rewards discipline. Long, count-aware acquisitions; time-of-flight; validated reconstruction; a consistent volume-of-interest method that respects respiratory motion; and a scrupulously maintained calibration chain are what separate a defensible tumor-dose number from a confident-looking artifact. A program that treats Y-90 PET quantification as a rigorous measurement — and that states its modality and method every time it reports a dose — is a program whose dosimetry can be trusted.

How DRPS Can Help

Diagnostic Radiation Physics Services supports radioembolization programs across the full arc: PET/CT quantitative calibration and NEMA-based performance verification, reconstruction-protocol validation for the Y-90 low-count regime, voxel-dosimetry method review, and the radiation-safety and licensing framework the therapy requires. This work is delivered through PET/CT and nuclear medicine physics, medical physics consulting, and radioactive material license support.

DRPS supports facilities across our service locations, including Florida, Maryland, Virginia, Washington DC, California, Nevada, New York, Pennsylvania, New Jersey, and Delaware.

A verified dose map is the difference between hoping the microspheres went where they should and knowing what dose the tumor received.

Related Resources

References

Busse NC, Al-Ghazi MSAL, Abi-Jaoudeh N, et al. AAPM Medical Physics Practice Guideline 14.a: Yttrium-90 microsphere radioembolization. J Appl Clin Med Phys. 2024;25(2):e14157. doi:10.1002/acm2.14157. PubMed
Rice M, Krosin M, Haste P. Post Yttrium-90 Imaging. Semin Intervent Radiol. 2021;38(4):460-465. doi:10.1055/s-0041-1735569. PubMed
International Commission on Radiological Protection. ICRP Publication 107: Nuclear Decay Data for Dosimetric Calculations. Ann ICRP. 2008;38(3). icrp.org
Kappadath SC, Henry EC, Lopez BP, Mahvash A. Quantitative evaluation of Y-90-PET/CT and Y-90-SPECT/CT-based dosimetry following yttrium-90 radioembolization. Med Phys. 2024;51(9):6061-6074. doi:10.1002/mp.17175. PubMed
Henry EC, Mahvash A, Lopez BP, Kappadath SC. A comparison of methods for in vivo activity and absorbed dose quantification with PET/CT following yttrium-90 radioembolization. Med Phys. 2024;51(9):6034-6045. doi:10.1002/mp.17174. PubMed
Kao YH, Steinberg JD, Tay YS, et al. Post-radioembolization yttrium-90 PET/CT — part 2: dose-response and tumor predictive dosimetry for resin microspheres. EJNMMI Res. 2013;3(1):57. doi:10.1186/2191-219X-3-57. PubMed
Labour J, Boissard P, Baudier T, et al. Yttrium-90 quantitative phantom study using digital photon counting PET. EJNMMI Phys. 2021;8(1):56. doi:10.1186/s40658-021-00402-6. PubMed
National Electrical Manufacturers Association. NEMA Standards Publication NU 2-2024: Performance Measurements of Positron Emission Tomographs (PET). Rosslyn, VA: NEMA. nema.org
U.S. Nuclear Regulatory Commission. 10 CFR Part 35: Medical Use of Byproduct Material. ecfr.gov
U.S. Nuclear Regulatory Commission. Yttrium-90 Microsphere Brachytherapy Sources and Devices Licensing Guidance. nrc.gov