PET/CT with 18F-choline or 18F-FDG in Hepatocellular Carcinoma Submitted to 90Y-TARE: A Real-World Study

Our aim was to assess the role of positron emission computed tomography (PET/CT) with 18F-choline (18F-FCH) or 18F-fluorodeoxyglucose (18F-FDG) in hepatocellular carcinoma (HCC) submitted to 90Y-radioembolization (90Y-TARE). We retrospectively analyzed clinical records of 21 HCC patients submitted to PET/CT with 18F-fluorocholine (18F-FCH) or 18F-fluodeoxyglucose (18F-FDG) before and 8 weeks after 90Y-TARE. On pre-treatment PET/CT, 13 subjects (61.9%) were 18F-FCH-positive, while 8 (38.1%) resulted 18F-FCH-negative and 18F-FDG-positive. At 8-weeks post 90Y-TARE PET/CT, 13 subjects showed partial metabolic response and 8 resulted non-responders, with a higher response rate among 18F-FCH-positive with respect to 18F-FDG-positive patients (i.e., 76.9% vs. 37.5%, p = 0.46). Post-treatment PET/CT influenced patients’ clinical management in 10 cases (47.6%); in 8 subjects it provided indication for a second 90Y-TARE targeting metabolically active HCC remnant, while in 2 patients it led to a PET-guided radiotherapy on metastatic nodes. By Kaplan–Meier analysis, patients’ age (≤69 y) and post 90Y-TARE PET/CT’s impact on clinical management significantly correlated with overall survival (OS). In Cox multivariate analysis, PET/CT’s impact on clinical management remained the only predictor of patients’ OS (p < 0.001). In our real-world study, PET/CT with 18F-FCH or 18F-FDG influenced clinical management and affected the final outcome for HCC patients treated with 90Y-TARE.


Introduction
Hepatocellular carcinoma (HCC) represents the fifth most common malignancy and a leading cause of cancer-related death worldwide [1]. Surgery is the therapy of choice for localized HCC, but the condition of some patients is not amenable to surgery, due to several reasons, such as large or multifocal lesions, portal vein invasion (PVI), extrahepatic spreading, poor liver function, etc. Advanced HCC has limited therapeutic options: tyrosine kinase inhibitors (TKIs), for example, proved effective in delaying disease progression; additionally, prolonging overall survival and immunotherapy with checkpoint inhibitors was found to exert strong anti-tumor activity in a subset of HCC patients [2,3].
Transarterial therapies, a group of treatments based on the intra-arterial administration of embolic or cytotoxic agents directed into the target lesions through their arterial feeders, play a crucial role in HCC therapeutic workflow [4]. In particular, the intra-arterial administration of glass or resin microspheres labeled with the radionuclide yttrium-90 ( 90 Y), also known as selective internal radiation therapy (SIRT) or 90 Y-transaterial radioembolization ( 90 Y-TARE), has gained an ever-increasing importance for the management of liver cancer [5,6], demonstrating a satisfying response rate and relevant impact on patients' quality of life [7]. In a meta-analysis including 21 studies investigating TARE's impact on In this retrospective analysis, we included all the consecutive HCC patients who were examined on our PET/CT with 18 F-FCH or, in case of 18 F-FCH-negative tumors, with 18 F-FDG before and 8 weeks after 90 Y-TARE between 01/2018 and 03/2019. Previous medical history, including presence of cirrhosis, previously performed loco-regional or systemic therapies, the results of performed diagnostic imaging (contrast-enhanced CT, MRI, liver ultrasonography), and laboratory tests (e.g., alphafetoprotein, hepatic enzymes, bilirubin, albumin) were recorded. All of the patients included had to present a complete and detailed available clinical history.
Selected patients were then reviewed on a case-by-case basis and were identified as those belonging to 1 of these 2 possible clinical settings: (a) HCC patients showing a pre-treatment 18 F-FCH-positive PET/CT scan who were treated with 90 Y-TARE and monitored with an 8-week post 18 F-FCH PET/CT procedure; (b) HCC patients with a pre-treatment 18 F-FCH-negative and a 18 F-FDG-positive PET/CT scan, who were submitted to 90 Y-TARE and then followed-up with an 8-week 18 F-FDG PET/CT scan.
The primary endpoint of the study was to define if an 8-week post-treatment response assessed by PET/CT with 18 F-FCH or 18 F-FDG influenced patients' clinical management. PET/CT's impact was scored as significant if (1) it provided an indication for a further 90 Y-TARE procedure selectively targeting metabolically active HCC remnant detected on PET/CT imaging; (2) it entailed the implementation of PET-directed RT on isolated metastatic localizations. The secondary endpoint was to determine whether PET-directed therapies affected patients' final outcome (i.e., OS).
This was a retrospective study on data available for clinical practice in which clinical records of all patients in follow-up for HCC submitted to 90 Y-TARE were reviewed. Data were anonymously collected and cumulatively gathered in an electronic database for analysis. Patients were not required to give informed consent to the study because the analysis used anonymous data that were obtained after each patient agreed to be followedup and to collect clinical records by institutions. No experimental procedures, novel devices, or experimental drugs were used, and no founds were received. The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki.
All patients provided written informed consent prior to procedure and associated risk. Pre-procedural evaluation included baseline imaging studies (liver sonography, clinical and laboratory examination, ce-CT, and PET/CT).
Angiography with selective visceral catheterization was performed in order to evaluate the vascular and tumor anatomy and blood-flow dynamics. A 99m Tc-macroaggregated albumin scan was carried out to test gastrointestinal flow and to estimate the percent of injected activity shunted to the lungs. After 7-10 days, the patients returned to our department for a treatment session performed by selective catheterization of the main hepatic artery by the transfemoral approach, embolization of gastroduodenal and gastric artery. After selective catheterization of the right/left hepatic artery, the patient, without sedation, was administered with a slow, manually controlled injection lasting about 30 min, under intermittent fluoroscopic guidance, alternating the 90 Y-microspheres suspended in 5% glucose solution with contrast medium for assessing persevered anterograde arterial flow. In all cases resin spheres (SIR-Spheres; Sirtex Medical, Sydney, Australia) were administered. In the case of bilobar lesions, each hepatic lobe was sequentially administered with 90 Y-microspheres in a separate session within an interval of 6-8 weeks to reduce the risk of radioembolization-induced liver disease (RIELD).
The prescribed 90 Y activity was determined as the patient-specific activity according to the body surface area (BSA) formula. After 90 Y-TARE procedure, all subjects underwent a 90 Y-PET/CT scan to assess the microsphere distribution pattern [15].

Imaging
All patients underwent a PET/CT scan 20 min after the intravenous (i.v.) administration of 3.7 KBq/kg of 18 F-methyl-choline (IASOcholine/Pcolina ® , Iason GmbH, Graz Seiersberg, Austria) or 60 min after i.v. administration of 3.7 KBq/kg of 18 F-FDG (Gluscan ® , Advanced Accelerator Applications, Venafro, Italy) according to the International Guidelines [16]. For both radiopharmaceuticals, the PET/CT device was a Discovery ST (General Electric, GE, Milwaukee, WI, USA) with bismuth germanate crystal units arranged to form 24 rings combined with a 16-slice Light Speed Plus CT scanner. The average FWHM axial resolution of PET (full width at half maximum) is 5.2 mm and system sensitivity 9.3 cps/KBq for 3D acquisition mode. Scanning was performed from the neck to the proximal tight in 3D modality, with an acquisition time of 3 min per table position. Images were reconstructed by using an ordered subset expectation maximization iterative algorithm (OSEM-SV, VUE Point HD, GE, 2 iterations, 15 subsets). The CT was performed immediately before PET in the identical axial field of view using a standardized protocol consisting of automatic tube current modulation with auto mA-tube rotation time of 0.5 s/rotation, slice thickness of 3.75 mm. The CT data were resized from 512 × 512 to a 256 × 256 matrix to match the PET data. The data were transmitted to a nuclear medicine database, fused, and displayed using dedicated software (Advantage, GE).

Pre-Treatment Image Evaluation
Before 90 Y-TARE, all patients underwent an 18 F-FCH PET/CT scan as the first line PET diagnostic modality. Each 18 F-FCH PET/CT scan was reviewed jointly by 2 boardcertified nuclear medicine physicians (L.F. and O.B., both with >15 years of experience); images were visually evaluated for pathological tracer uptake, defined as a focally increased radiopharmaceutical's incorporation within the hepatic lesions greater than that of the neighboring parenchyma, and were classified as positive or negative. In the case of negative 18 F-FCH PET/CT scans, patients were submitted to 18 F-FDG PET/CT scan that was carried out within 1 week from previously performed 18 F-FCH imaging. HCCs were considered as 18 F-FDG-positive if they showed increased tracer uptake greater than adjacent normal liver.

Post-Treatment Image Assessment
In 18 F-FCH-positive PET/CT scans, standardized uptake values (SUVs) were calculated using regions of interest (ROI). In each patient, up to 3 of the most 18 F-FCH-avid hepatic localizations were selected as the target lesions and the normal adjacent parenchyma as the background control. In order to normalize tumor SUVs, the ratio of SUVmax of the lesions to the mean SUV of the normal adjacent parenchyma (SUVmean), the tumorto-normal liver ratio (TNR), was gauged. In order to minimize potential partial volume effects, the reference ROI in the normal hepatic parenchyma was drawn with a diameter of 2 cm. Eight weeks after the 90 Y-TARE, patients underwent a further 18 F-FCH PET/CT scan to assess metabolic response to 90 Y-microspheres.
Since no standard quantitative criteria have been established to define metabolic response on 18 F-FCH PET/CT, the authors adopted the following: post-treatment PET/CT scans were compared with the pre-treatment ones and the relative change in TNR ratio (∆TNR) was determined. Metabolic response was defined as a reduction of ≥50% in ∆TNR, while subjects were classified as non-responders in case of ∆TNR reduction < 50% or if new lesions were evident on the post-treatment PET/CT scan.
In 18 F-FDG-positive HCCs, SUV measurement was determined on the pre-treatment and 8-weeks post-treatment PET/CT scan using PET VCAR (GE Healthcare, Milwaukee, WI, USA). To assess metabolic response to 90 Y-TARE, the follow-up PET/CT was compared to the pre-treatment scan according to the PET Response Criteria in Solid Tumors (PERCIST) [17].

Toxicity and Follow-Up
Toxicity was assessed according to the Common Terminology Criteria for Adverse Events (CTCAE), version 5.0, on the basis of laboratory tests, CT or PET/CT imaging, and clinical examinations. Before the procedure, all patients were submitted to laboratory tests, including total bilirubin, alanine transaminase, aspartate transaminase, alkaline phosphatase, and γ-glutamyl transpeptidase. After 90 Y-TARE, all patients resumed a routine schedule of laboratory tests that were carried out at 2 and 4 weeks after the procedure and repeated at a 3-month interval. Clinical toxicities, including pain, fever, fatigue, and gastrointestinal adverse events, were evaluated at the regular follow-up visits.

Statistics
The normality of the distribution of the continuous variables was evaluated with the Shapiro-Wilk test. In the case of symmetric distribution, the variables are expressed with Biomedicines 2022, 10, 2996 5 of 16 median, mean, and standard deviation (SD), while categorical data are represented as numbers and percentages.
Free survival (PFS) and overall survival (OS) were calculated by the Kaplan-Meier method (MedCalc 11.3.8.0; MedCalc Software, Mariakerke, Belgium), defined as the time from first 90 Y-TARE to disease progression and to patient death, respectively. Fisher's exact test was applied to examine differences in response to 90 Y-TARE and PET/CT's impact on clinical management among 18 F-FCH-positive and 18 F-FCH-negative/ 18 F-FDG-positive patients. The Kaplan-Meier method was used to analyze differences in OS, and Cox regression analysis was applied to identify prognostic factors. Significance was established at two-tailed p < 0.05 level.

Results
The interrogation of our database identified 21 HCC patients fulfilling inclusion criteria. All the included subjects were submitted between January 2018 and March 2019 to PET/CT with 18 F-FCH or, in case of 18 F-FCH-negative tumors, with 18 F-FDG before and 8 weeks after 90 Y-TARE, as shown by Figure 1. Clinical-demographic characteristics of the patients and the values of clinical and PET-derived quantitative variables are summarized in Table 1.
All patients had preserved Eastern Cooperative Oncology Group Performance Status (ECOG ≤ 1) and hepatic function (Child-Pugh score ≤ 6). Only two patients had extrahepatic metastases before 90 Y-TARE enrollment: one subject showed a small lung nodule (1 cm diameter) stable in several repeated CT controls and one presented a small peritoneal localization to the anterior abdominal wall (1.5 cm diameter).
routine schedule of laboratory tests that were carried out at 2 and 4 weeks after the cedure and repeated at a 3-month interval. Clinical toxicities, including pain, feve tigue, and gastrointestinal adverse events, were evaluated at the regular follow-up v

Statistics
The normality of the distribution of the continuous variables was evaluated wit Shapiro-Wilk test. In the case of symmetric distribution, the variables are expressed median, mean, and standard deviation (SD), while categorical data are represente numbers and percentages.
Free survival (PFS) and overall survival (OS) were calculated by the Kaplan-M method (MedCalc 11.3.8.0; MedCalc Software, Mariakerke, Belgium), defined as the from first 90 Y-TARE to disease progression and to patient death, respectively. Fis exact test was applied to examine differences in response to 90 Y-TARE and PET/ impact on clinical management among 18 F-FCH-positive 18 F-FCH-negative/ 18 F-FDG-positive patients. The Kaplan-Meier method was used t alyze differences in OS, and Cox regression analysis was applied to identify progn factors. Significance was established at two-tailed p < 0.05 level.

Results
The interrogation of our database identified 21 HCC patients fulfilling inclu criteria. All the included subjects were submitted between January 2018 and March to PET/CT with 18 F-FCH or, in case of 18 F-FCH-negative tumors, with 18 F-FDG before 8 weeks after 90 Y-TARE, as shown by Figure 1. Clinical-demographic characteristi the patients and the values of clinical and PET-derived quantitative variables are marized in Table 1.

Pre-TARE PET/CT Imaging
Thirteen of 21 patients (61.9%) presented 18 F-FCH-positive HCC tumors on the pretreatment PET/CT scan, while eight (38.1%) subjects were negative. 18 F-FCH-positive PET/CT scans showed a median SUVmax of 16.5 and a mean SUmax of 15.5 ± 3.6, while the median and mean tumor-to-normal liver ratio (TNR) resulted in 2.2 and 2.2 ± 0.4, respectively. Notably, the two patients with known extrahepatic localizations before 90 Y-TARE enrollment showed 18 F-FCH incorporation in both HCC and metastases.
In the eight patients with pre-treatment 18 F-FCH-negative PET/CT scans, 18 F-FDG PET/CT resulted positive in all cases, with a median SUVmax of 15.5 and a mean SUVmax of 14.3 ± 7.1. Table 2 shows the correlation between PET/CT's results and histological findings in nine patients, who had been submitted to surgery (n = 4) or biopsy (n = 5), classified according to World Health Organization (WHO) [18].

90 Y-TARE Procedure
An overall number of 21 90 Y-TARE procedures were performed, 17 were carried out according to a lobar 90 Y-microsphere administration (i.e., 14 to the right hepatic lobe and 3 to the left lobe) and 4 following a sequential lobar approach. The mean administered activity was 1.5 ± 0.18. In 8 patients, a second TARE procedure was carried out on the basis of post-treatment PET/CT results.

PET/CT Post-Treatment Assessment of Response
Thirteen patients (61.9%) showed metabolic response to 90 Y-TARE. Of the 13 patients with pre-TARE 18 F-FCH-positive PET/CT, ten (76.9%) exhibited metabolic response to 90 Y-TARE with a mean ∆TNR of 68 ± 5.3. The two patients with 18 F-FCH-positive extrahepatic localizations on pre-treatment PET/CT were both responders to 90 Y-TARE with metastases' stability on follow-up PET/CT scan. Three subjects with 18 F-FCH-avid HCCs were non-responders: one patient showed ∆TNR < 50% and was categorized as stable metabolic disease, while other two subjects presented new-onset hepatic lesions and were therefore classified as progressive metabolic disease.
Among the eight patients with pre-treatment 18 F-FCH-negative/ 18 F-FDG-positive HCCs, three (37.5%) subjects showed metabolic response to 90 Y-TARE (partial metabolic response). Five patients were classified as non-responders: two subjects had stable metabolic disease, while three patients showed extrahepatic spreading with localizations to abdominal lymph nodes, as shown in Figure 2.
Although the percentage of responders was higher in the 18 F-FCH-positive patients with respect to 18 F-FDG-positive ones (i.e., 76.9% vs. 37.5%), this difference did not reach the threshold of statistical significance (i.e., p = 0.46).

Post-Treatment PET/CT's Impact on Patients' Clinical Management
The results of post-treatment PET/CT evaluation at 8 weeks were discussed and analyzed by the multidisciplinary disease management team (MDMT), including nuclear medicine physicians and interventional radiologists together with each referring physician, who jointly defined the most appropriate therapeutic pathways, according to patients' clinical status and eventual 90 Y-TARE-related toxicity.
Post-TARE PET/CT affected patients' clinical management in 10 out of 21 cases (47.6%). In particular, a second 90 Y-TARE treatment was performed in eight patients with evidence of metabolically active HCC remnant (n = 7) or new-onset lesion (n = 1). In such cases, a further angiography was performed before the second 90 Y-TARE, and vascular imaging was accurately examined in order to identify and selectively catheterize the arterial branch supplying the metabolically active tissue disclosed by 18 F-FCH (n = 7) or 18 F-FDG (n = 1) PET-imaging. In all these patients, complete metabolic response was registered after PET-guided 90 Y-TARE (Figure 3).
PET/CT meaningfully affected two out of three patients with evidence of extrahepatic progression on post-treatment 18 F-FDG PET/CT. In such cases, stereotactic RT was carried out on metastatic localizations to celiac lymph nodes: post-treatment PET/CT images were utilized to draw biological target volume (BTV) that was incorporated into radiation therapy (RT) planning with optimal clinical and imaging response in both cases, as shown in Figure 4.

Post-Treatment PET/CT's Impact on Patients' Clinical Management
The results of post-treatment PET/CT evaluation at 8 weeks were discussed alyzed by the multidisciplinary disease management team (MDMT), including medicine physicians and interventional radiologists together with each referrin cian, who jointly defined the most appropriate therapeutic pathways, accordin In 11 patients, post-treatment PET/CT impact was scored as non-relevant. In particular, monitoring patients through periodic examinations until the evidence of progressive disease was the clinical decision in six patients with partial metabolic response ( 18 F-FCH-positive, n = 4 and 18 F-FDG-positive PET/CT, n = 2) and in three patients with stable metabolic disease on post-TARE ( 18 F-FDG-avid, n = 2 and 18 F-FCH-avid, n = 1). In such cases, a further 90 Y-TARE was not carried out, in spite of metabolically active HCC remnant detected on post-treatment PET/CT, due to increased value of bilirubin (grade II, n = 1) or hepatic enzymes (grade II, n = 2), ascites (grade I, n = 1), or since subjects (n = 5, in all cases aged > 69 years) refused a repeated 90 Y-microsphere administration due to post-embolization syndrome (PES), mainly consisting of nausea and vomiting, occurred during the first 72 h following the first 90 Y-TARE procedure.
Biomedicines 2022, 10, x FOR PEER REVIEW 9 of 16 or 18 F-FDG (n = 1) PET-imaging. In all these patients, complete metabolic response was registered after PET-guided 90 Y-TARE (Figure 3). PET/CT meaningfully affected two out of three patients with evidence of extrahepatic progression on post-treatment 18 F-FDG PET/CT. In such cases, stereotactic RT was carried out on metastatic localizations to celiac lymph nodes: post-treatment PET/CT images were utilized to draw biological target volume (BTV) that was incorporated into radiation therapy (RT) planning with optimal clinical and imaging response in both cas-  In 11 patients, post-treatment PET/CT impact was scored as non-relevan ular, monitoring patients through periodic examinations until the evidence sive disease was the clinical decision in six patients with partial metabo 18 18 In two subjects with progressive metabolic disease, the therapeutic decision was the implementation of tyrosine-kinase therapy.

Prognostic Factors on Patient Survival
The mean PFS and OS in all patients were 9.3 ± 2.1 months (95% confidence interval, 5.1-13.4 months; median 8 months) and 18.6 ± 2.1 months (95% confidence interval, 14.3-22.9 months; median 18 months), respectively. In order to perform Kaplan-Meier analysis for OS, continuous variables (i.e., bilirubin levels, age) were dichotomized by median value, while categorical data were dichotomized, as indicated in Table 4. Alphafetoprotein levels were not considered in the analysis, since they were found increased only in 11 patients.  By Kaplan-Meier analysis ( Figure 5), patients whose clinical management was influenced by post-treatment PET/CT had a significantly (p < 0.001) longer OS (26.3 ± 2.6 months) than those in which PET/CT's impact was scored as non-relevant (11.2 ± 1.5 months). Furthermore, subjects with age ≤ 69 years exhibited a significantly (p = 0.005) longer OS (23.1 ± 3.3 months) than older patients (13.7 ± 1.8 months).  . Panel a shows that patients with age ≤ 69 years (blue line, group 0) had significantly (p = 0.005) longer OS than those aged > 69 years (green line, group 1). Panel b demonstrates that patients whose clinical management was influenced by PET/CT (green line, group 1) had meaningfully (p < 0.001) longer OS than those in which PET/CT's impact was scored as non-relevant (blue line, group 0).

Discussion
Our real-world study assessed the clinical impact of post-treatment PET/CT with 18 F-FCH or 18 F-FDG in HCC patients submitted to 90 Y-TARE. We found that a PET/CT evaluation at 8-weeks post-treatment influenced clinical management in 47.6% of subjects, through the implementation of PET-directed therapies. In addition, post-TARE PET/CT's impact on clinical management resulted in a significant predictor of patients' final outcome both by Kaplan-Meier and Cox multivariate analysis. . Panel a shows that patients with age ≤ 69 years (blue line, group 0) had significantly (p = 0.005) longer OS than those aged > 69 years (green line, group 1). Panel b demonstrates that patients whose clinical management was influenced by PET/CT (green line, group 1) had meaningfully (p < 0.001) longer OS than those in which PET/CT's impact was scored as non-relevant (blue line, group 0).

Discussion
Our real-world study assessed the clinical impact of post-treatment PET/CT with 18 F-FCH or 18 F-FDG in HCC patients submitted to 90 Y-TARE. We found that a PET/CT evaluation at 8-weeks post-treatment influenced clinical management in 47.6% of subjects, through the implementation of PET-directed therapies. In addition, post-TARE PET/CT's impact on clinical management resulted in a significant predictor of patients' final outcome both by Kaplan-Meier and Cox multivariate analysis.
PET/CT is a well-established imaging modality in oncology and plays an essential role for staging and monitoring response to treatment in many oncological conditions. Nevertheless, metabolic imaging is not routinely considered in HCC diagnostic work-flow. In a comparative study performed by Talbot et al. [19] in 81 patients with suspected liver nodules, PET/CT with 18 F-FCH showed a significantly higher sensitivity than that with 18 F-FDG (88% vs. 68%, p = 0.07) for HCC diagnosis, although 18 F-FDG had a higher detection rate for the less differentiated and more aggressive forms. 18 F-FDG incorporation into HCC is not only correlated with high histological grade, but also with the expression of genes strictly linked with cell survival, cell-to-cell adhesion, or cell spreading [20]. The combined use of 18 F-FCH and 18 F-FDG has been proposed for HCC imaging through PET technology according to the grade of differentiation: in a large cohort (n = 177) of HCC patients, dual tracer PET/CT substantially affected staging according to the Barcelona Clinical Liver Cancer (BCLC) classification and consequently changed subjects' management [21].
Few studies investigated the role of metabolic response assessed by PET/CT in HCC patients treated with 90 Y-microspheres. In a previous report from our group [22], we found decreased total lesion glycolysis (TLG), measured at 1 month after 90 Y-TARE, associated with a trend toward a longer OS in poorly differentiated HCC with portal vein invasion. Hartenbach and colleagues employed PET/CT with 18 F-fluoroethylcholine in 24 patients with locally advanced HCC and initially elevated AFP level; from semiquantitative analysis, increased SUV mean at diagnosis, decreased SUV max and in tumor-to-background ratio after 90 Y-therapy (i.e., ∆maximum SUV and ∆tumor-to-background ratio, respectively) showed the highest area under the curve to predict patient response [23]. The results reported by Hartenbach's group are substantially in agreement with those more recently described by Aujay and coworkers in nine HCC patients treated with 90 Y-TARE and monitored through 18 F-FCH PET/CT [24].
Reizine and colleagues applied dual tracer PET/CT with 18 F-FCH and 18 F-FDG in 37 HCC patients submitted to 90 Y-TARE for the assessment of early post-treatment response at 4-8 weeks after procedure [13]. All of the enrolled subjects were submitted to dual tracer PET/CT before 90 Y-microsphere administration: 28 patients resulted 18 F-FDG-positive; 9 were 18 F-FCH positive. Metabolic response detected on early post-treatment PET/CT showed 100% sensitivity and specificity for predicting 6-month radiological response assessed by mRECIST; furthermore, metabolic response was a significant predictor of OS. To the best of our knowledge, our report is the first study specifically highlighting the clinical usefulness of PET/CT with 18 F-FCH or 18 F-FDG after 90 Y-TARE in order to promptly identify patients with residual or progressive metabolically active disease, amenable to timely PET-guided retreatments. Of note, our results, indicating the feasibility of repeated 90 Y-TARE to treat recurrent or residual primary disease in similar hepatic arterial lobe or segments, are substantially in line with the terms of safety and efficacy for 90 Y-loaded glass microspheres, as published by Badar and colleagues [25].
The optimal time point to assess the radiobiological effects of 90 Y-microspheres has yet to be defined. It is expected that 90 Y-TARE might also have delayed effects, although this assumption is mainly based on the limitations of traditional imaging techniques in the early assessment of response to 90 Y-TARE. In a recently published in vitro study, in fact, 90 Y-microspheres were found to reduce colorectal cancer cell proliferation as early as within 96 h of observation [26]. In this perspective, a time interval of 8 weeks after procedure might represent a reasonable gap to assess 90 Y-microsphere effects on HCC.
As far as it concerns the role of patients' age in HCC treated with 90 Y-TARE, our data are not in line with those reported by the retrospective study conducted by The European Network on Radioembolization with Yttrium-90 resin microspheres study group (ENRY), including elderly (≥70 years, n = 128) and younger (<70 years, n = 197) subjects [27]. In the cited paper, 90 Y-TARE was equally tolerated in both cohorts with no significant differences in survival between the two groups. In our study, the effect of age on patients' final outcome might be explained by the higher compliance registered in younger subjects to be submitted to further PET-directed therapy (second 90 Y-microsphere administration or RT) for eradicating HCC remnant or new-onset metastases.
In our real-world study, we employed 18 F-FCH as the first line tracer for pre-treatment imaging of HCC with the aim of identifying the metabolically active tissue to be targeted with 90 Y-microsphere administration. It is worth mentioning that we did not systematically perform dual tracer PET/CT in all of the enrolled patients; only subjects with 18 F-FCHnegative tumors were submitted to 18 F-FDG as a second-line functional imaging modality, in order to limit the radiation burden delivered to patients. We cannot exclude that some of the 18 F-FCH-positive tumors might also express a variable grade of 18 F-FDG-avidity. 90 Y-microspheres are recommended and generally employed in subjects with advanced HCC, often heavily pre-treated, progressing after surgery, TACE, or systemic therapy [28]. In such cases, liver anatomy may be altered by the previously performed treatments, and identifying HCC-viable tissue might be challenging on conventional morphological imaging (ce-CT or MRI). In these patients, PET/CT with 18 F-FCH or 18 F-FDG may have an important role supporting clinicians both during 90 Y-TARE planning and in response assessment.
Our study has several limitations. First of all, the limited number of included patients and its retrospective nature may have introduced a selection bias in patient enrollment. However, it has to be underlined that our sample size (n = 21), although small, is not significantly different with respect to that included in the other cited papers focusing on 18 F-FCH PET/CT on HCC submitted to 90 Y-TARE (i.e., Hartenbach et al., n = 24; Reizine et al., n = 37) [13,23].
Furthermore, since our retrospective real-life world study was mainly aimed to provide information on the use of PET/CT with 18 F-FCH or 18 F-FDG in a specific HCC clinical setting, we did not perform a comparative assessment of 18 F-FCH/ 18 F-FDG PET/CT's impact on patient management with respect to that of more conventional diagnostic techniques, such as ce-CT and MRI [29]. In this regard, a prospective study performed by Barabasch and coworkers in 36 consecutive patients with liver metastases (20 colorectal, 14 breast cancer, 2 with other malignancies), submitted to 18 F-FDG PET/CT and diffusionweighted MRI (DWI-MRI) before and 4-6 weeks after 90 Y-TARE, showed that response based on DWI-MRI outperformed PET/CT for predicting final outcome [30]. Nevertheless, MRI-DWI's impact on clinical management in HCC subjects treated with 90 Y-microspheres, as compared to that of PET/CT, has not yet been assessed. This topic is worthy of future investigations.
In addition, we used the BSA method for the calculation of 90 Y-microsphere prescribed activity, while personalized provisional dosimetry is now recommended as the stateof-the-art to determine the dose delivered to tumor and non-tumor parenchyma [31]. In a prospective study, Ho et al. [32] employed dual tracer PET/CT with 18 F-FDG and 11 C-acetate, another surrogate imaging biomarker of phospholipid synthesis, to define the relationship between tumor dose (TD) and response, according to HCCs' grade of differentiation. In agreement with our findings, the authors found a higher response rate in more differentiated (i.e., 11 C-acetate-positive) than in aggressive ( 18 F-FDG-positive) HCCs (72.4% vs. 25%, respectively); furthermore, TD for response resulted meaningfully higher for poorly differentiated with respect to well/moderately differentiated HCCs (262 Gy vs. 152/174 Gy). The role of dual tracer PET/CT for the personalized dose prescription in HCC submitted to 90 Y-TARE will be topic of future investigations.

Conclusions
In our real-world study, post-treatment PET/CT with 18 F-FCH or 18 F-FDG, carried out at 8 weeks after 90 Y-TARE, influenced patients' clinical management by the implementation of PET-guided therapies and significantly affected final outcome. Further well-designed studies with larger cohorts, ideally prospective and entailing multicenter co-operations, are needed to better define the role of metabolic imaging in this specific clinical setting.

Institutional Review Board Statement:
This was a retrospective study on data available for clinical practice in which clinical records of all enrolled patients were reviewed. Data were anonymously collected and were cumulatively gathered in an electronic database for analysis. Patients were not required to give informed consent to the study because the analysis used anonymous data that were obtained after each patient agreed to be followed-up and to collect clinical records by institutions. No experimental procedures, novel devices, or experimental drugs were used, and no founds were received. The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki.
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.