**1. Introduction**

**\***

Selective internal radiation therapy (SIRT) is a treatment option for patients with tumours in the liver that cannot be surgically resected. SIRT is most commonly used to treat liver metastases from colorectal cancer and hepatocellular carcinoma, and involves injection of 90Y- or 166Ho-microspheres into the hepatic arterial vasculature [1,2]. Although SIRT has been in clinical use for over 15 years, the benefits of this treatment remain unclear. A review of SIRT in randomized controlled trials found a lack of evidence for improved survival or quality of life for colorectal cancer patients with metastatic disease in the liver [3]. Investigations into the efficacy of SIRT have demonstrated the existence of a strong dose– response relationship [4–8]. Hence, a personalised, optimised approach ensuring adequate tumour-absorbed dose may be crucial to increase the efficacy of the technique and to allow demonstrable and significant treatment benefits.

SIRT dose-response studies generally use 90Y PET imaging to provide estimates of tumour absorbed doses. Optimisation of the treatment, and dose verification, are hence reliant on accurate image-derived dosimetry [9]. Respiratory motion during the PET acquisition degrades the image quality. If left uncorrected, this leads to images with a blurring of the radioactivity distribution which will produce errors in dosimetry. In the case of SIRT, it is the motion of the liver that is of interest. This organ typically moves 10–26 mm cranio-caudally during normal respiration [10,11]. The purpose of the current investigation was to assess the effect of respiratory motion on tumour absorbed doses calculated from 90Y PET images following SIRT, and to evaluate the benefits of respiratory motion correction for this application.

The probability of positron production during 90Y decay is extremely low, making 90Y a challenging radionuclide to image using PET [12]. Despite relatively high activities in the

**Citation:** Walker, M.D.; Gear, J.I.; Craig, A.J.; McGowan, D.R. Effects of Respiratory Motion on Y-90 PET Dosimetry for SIRT. *Diagnostics* **2022**, *12*, 194. https://doi.org/10.3390/ diagnostics12010194

Academic Editors: Lioe-Fee de Geus-Oei and F.H.P. van Velden

Received: 29 November 2021 Accepted: 10 January 2022 Published: 14 January 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

scanner's field of view (e.g., 0.5–3 GBq), the coincidence count-rate is low, hampering quantification. The quantitative accuracy of 90Y PET in the absence of motion has been evaluated by several groups [13–16] including a multi-centre phantom evaluation [17]. The consistent findings are that modern time-of-flight (TOF) PET scanners provide accurate quantification of radioactivity concentrations in phantoms, subject to the expected errors arising from a limited spatial resolution. Simulations performed by Ausland et al. [18] predicted that the errors in tumour dose quantification caused by respiratory motion could be substantial. The effects of image noise, respiratory motion, and motion compensations for 90Y SIRT have been considered by Siman et al. [19] who used short-duration images from the National Electrical Manufacturers Association (NEMA) International Electrotechnical Commission (IEC) body phantom, filled with 18F, to mimic noisy 90Y acquisitions. With no motion the quantification was found to be accurate, but large errors occurred when motion was applied (>50% for some dose measures), with these errors ameliorated by motion compensation (quiescent-period gating). Effects of respiratory motion have also been studied via simulation for pre-SIRT 99mTc MAA SPECT [20]. Siman et al. [19] suggested that further studies, using anthropomorphic phantoms, with 90Y PET-CT studies, were needed to confirm the effectiveness of the motion compensation method. Our investigation addresses this by using clinically relevant activities of 90Y within an anthropomorphic phantom specifically designed for evaluation of SIRT dosimetry [21]. The phantom was filled and imaged several times to provide a range of activities and contrast ratios. Furthermore, we acquired data on two time-of-flight PET scanners from different vendors, allowing interpretation of the results in a wider context as opposed to being scanner-/vendor-specific. Respiratory gating signals were obtained using external devices, as data driven respiratory gating in clinical use for 18F and 68Ga based PET [22,23] has ye<sup>t</sup> to be robustly implemented for 90Y PET [24].

#### **2. Materials and Methods**
