key: cord-1019266-npihsoet
authors: Siripongsatian, Dheeratama; Promteangtrong, Chetsadaporn; Kunawudhi, Anchisa; Kiatkittikul, Peerapon; Boonkawin, Natphimol; Chinnanthachai, Chatchawarin; Jantarato, Attapon; Chotipanich, Chanisa
title: Comparisons of Quantitative Parameters of Ga-68-Labelled Fibroblast Activating Protein Inhibitor (FAPI) PET/CT and [(18)F]F-FDG PET/CT in Patients with Liver Malignancies
date: 2022-04-29
journal: Mol Imaging Biol
DOI: 10.1007/s11307-022-01732-2
sha: 725e820ea06e3fbe6e99ff4487d1ae4dd47cbc5c
doc_id: 1019266
cord_uid: npihsoet

PURPOSE: To compare quantitative parameters and tumour detection rates of [(68) Ga]Ga-FAPI PET/CT with those of dedicated liver PET/MRI and (18)F-FDG PET in patients with liver malignancies. PROCEDURES: Twenty-seven patients (29 imaging studies) with diagnosed or suspected liver malignancies who underwent [(68) Ga]Ga-FAPI-46 PET/CT, liver PET/MRI, and [(18)F]FDG PET/CT between September 2020 and June 2021 were retrospectively analysed. MRI findings were used as the reference standard for diagnosis. RESULTS: The 27 patients had a median age of 68 years (interquartile range: 60–74 years; 21 men). Primary intrahepatic tumours were reported in 13 patients (15 imaging studies) with cholangiocarcinoma (CCA) and in 14 patients with hepatocellular carcinoma (HCC). All intrahepatic lesions detectable on MRI were also detected on [(68) Ga]Ga-FAPI-46 PET/CT giving a sensitivity of 100% (19/19), whereas the sensitivity of [(18)F]FDG PET/CT was 58% (11/19). All intrahepatic lesions were detected on [(68) Ga]Ga-FAPI-46 PET/CT, on which they showed higher activity (median SUVmax: 15.61 vs. 5.17; P < .001) and higher target-to-background ratio (TBR; median, 15.90 vs. 1.69, P < .001) than on [(18)F]FDG, especially in patients with CCA (median TBR, 21.08 vs. 1.47, respectively; P < .001). The uptake positivity rate in regional node metastasis was 100% (12/12) on [(68) Ga]Ga-FAPI-46 PET/CT compared with 58% (7/12) on [(18)F]FDG PET/CT. All patients with distant metastasis (100%, 14/14) were detected on both [(18)F]FDG and [(68) Ga]Ga-FAPI-46 PET/CT imaging, although more distant metastatic lesions were detected on [(68) Ga]Ga-FAPI-46 PET/CT than on [(18)F]FDG (96% (42/44) vs. 89% (39/44), respectively). CONCLUSION: [(68) Ga]Ga-FAPI PET/CT with dedicated liver PET/MRI shows potential for superior detection of hepatic malignancy compared with [(18)F]FDG PET/CT or MRI alone. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s11307-022-01732-2.

Liver cancer is the third most common cause of cancer-related death worldwide [1] . Primary liver cancer consists of two major histological subtypes: hepatocellular carcinoma (HCC) and cholangiocarcinoma (CCA). In Thailand, primary liver cancer is the leading cancer in men, but in contrast to other countries, cholangiocarcinoma is the predominant subtype of liver cancer, rather than HCC [2] . The high mortality rate of liver cancer indicates the inefficiency of current strategies for evaluating and treating it [3] . Imaging plays an important role in diagnosis, initial staging, evaluation of treatment response, and detection of recurrence. While magnetic resonance imaging (MRI) has significantly better sensitivity for detecting HCC than contrastenhanced computed tomography, there is no significant difference in specificity between the two modalities [4] . Currently, fluorine-18 fluoro-2-deoxyglucose ([ 18 F]FDG) positron emission tomography (PET)/CT is increasingly being recognised as an effective diagnostic imaging tool for cancer. Variable FDG uptake in HCC is related to the differentiation of the tumour. The sensitivity of [ 18 F]FDG PET for moderately or poorly differentiated HCC is greater than that for well-differentiated or low-grade tumours [5, 6] . The sensitivity of [ 18 F]FDG PET/ CT ranges from 50 to 70% for HCC, similar to that of CT and MRI [5] [6] [7] . [ 18 F]FDG PET has comparable diagnostic accuracy to MRI with respect to the tumour T stage. The reported pooled sensitivity and specificity of FDG PET/CT for tumour T stage are 91% and 85%, respectively, values that compare very closely with those of MRI, at 90% and 84%, respectively. However, data regarding regional node metastasis and N stage are limited owing to the unsatisfactory diagnostic accuracy of both imaging modalities. Furthermore, [ 18 F]FDG PET/CT has low sensitivity but high specificity for detecting distant metastasis, with reported pooled sensitivity and specificity of 56% and 95%, respectively [8] . Gallium (Ga)68-labelled fibroblast activating protein inhibitor (FAPI) was developed to evaluate fibroblast activation in oncological imaging, with fibroblast activating proteins being highly expressed in cancer-associated fibroblasts in various epithelial carcinomas. The significantly lower hepatic background uptake of [ 68 Ga]Ga-FAPI in comparison with that of [ 18 F]FDG is also an important factor improving the sensitivity of [ 68 Ga]Ga-FAPI for liver tumour detection in comparison with [ 18 F]FDG PET [9, 10] . Previous studies showed that [ 68 Ga]Ga-FAPI PET/CT has better sensitivity than [ 18 F]FDG PET/CT for detecting liver malignancies [11, 12] .

This study aimed to compare the quantitative parameters and tumour detection rates of dedicated liver PET/MRI and [ 18 F] FDG PET/CT in patients with liver malignancies. 

All PET images were analysed using a Syngovia workstation (Siemens Healthineers) and were displayed in coronal, axial, and sagittal planes with rotating 3D images. The MRI dataset was analysed using the MR oncology workflow program of the Syngovia workstation. The acquired PET data were reconstructed using ordered subset expectation maximisation and a point spread function.

All PET/CT scans were interpreted independently by two experienced board-certified nuclear medicine physicians (with 5 and 10 years of experience) who were blinded to the patients' clinical information. On visual analysis, positive lesions were defined as those showing non-physiological focal tracer uptake on PET images. Three-dimensional regions of interest were drawn around the region of abnormal uptake on axial slices to obtain the quantitative parameters. The maximum standardised uptake value (SUVmax) was automatically calculated to evaluate tracer uptake in primary tumours, lymph nodes, and distant metastases. The target-to-background ratio (TBR) of each primary tumour was calculated by dividing the SUVmax of the lesion by the SUVmean of normal background liver. MRI images from the dedicated liver PET/MRI were used as the reference standard, and PET/ MRI images were used for coregistration of intrahepatic lesions. On MRI, positive lesions were considered according to their abnormal enhancing pattern and fluid restriction pattern.

Semiquantitative parameters are described as median and interquartile range, and were compared using paired and unpaired t-tests, depending on the data characteristics. The MRI findings were used as the reference standard in the diagnostic performance calculations. Statistical significance was set at P < 0.05, and STATA software version 11 (StataCorp, College Station, TX, USA) was used for all analyses. 

The uptake-positive rates of [ 18 

The histopathological results and typical radiographic findings on MRI were collected for the primary liver tumours of the treatment-naïve patients. Five out of six initially treatment-naïve patients had pathological confirmation. Another patient had the typical pattern of HCC following radiofrequency ablation treatment. Nonetheless, a decreased alpha fetoprotein (AFP) level was noted in clinical follow-up. In these treatment-naïve patients, the uptake-positive rates of [ 18 F]FDG and [ 68 Ga]Ga-FAPI-46 in the primary lesions on PET/CT imaging were 67% (4/6) and 100% (6/6), respectively. The SUVmax, SUVmean, SUVpeak, and TBR of [ 68 Ga]Ga-FAPI-46 were significantly higher in patients with CCA than in those with HCC (P < 0.005).

Pathological confirmation was available for 2 of the 21 posttreatment patients. Because of the invasiveness of procedures required for pathological diagnosis, follow-up imaging and clinical results were used as the reference standard for the other patients. One post-treatment HCC patient died from respiratory failure due to COVID-19. The uptake-positive rates in local recurrence and residual tumour lesions were 46.7% The typical pattern of FAPI uptake in post-surgical/posttreatment inflammation was diffuse uptake, especially in inflammation from radiotherapy or surgery, but areas of inflammation did not show as abnormal on MRI. Except for two patients with HCC who underwent transarterial chemoembolisation, the images showed abnormal focal (but not diffuse) FAPI uptake in the intrahepatic lesions, which MRI and follow-up CT confirmed as post-treatment change. According to the MRI findings, there were two false-positive lesions on [ 68 Ga]Ga-FAPI-46 PET/CT in patients with HCC. One false-positive lesion in a patient with HCC was false-positive on both [ 68 Ga]Ga-FAPI-46 PET/CT and [ 18 F]FDG PET/CT. One lesion was 6.8 cm, with 2-month follow-up CECT showing a partial lipiodolstained mass with peripheral rim uptake and stable size. The other lesion was 1.5 cm at 7-month follow-up and showed no progression according to imaging or tumour markers. Both were negative according to the enhancing pattern on MRI. All lesions identified on MRI were also identified on [ 68 Ga]Ga-FAPI-46 PET/CT (100%, 41/41), whereas a much lower number of lesions (detection rate: 39% [16/41]) were detected on [ 18 F]FDG PET/CT. Figure 3 presents a representative case.

The number of detected regional node metastases was higher in patients with CCA (n = 10 patients, 29 nodes) than in those with HCC (n = 2 patients, 2 nodes). The uptake-positive rates of regional node metastases were 100% (12/12) on [ 68 Ga] Ga-FAPI-46 PET/CT and 58% (7/12) on [ 18 Figure 5 presents a representative case. According to previous studies, MRI has favourable diagnostic performance for intrahepatic tumour detection (100% accuracy) [11] . Thus, we used MRI as the reference standard for the diagnostic performance calculations. According to the MRI findings, there were two false-positive patients with HCC (post-treatment inflammatory processes) and no false positives among the CCA patients on [ 68 Ga] Ga-FAPI-46 PET/CT. Moreover, we noticed that the two false-positive patients with HCC underwent transarterial chemoembolisation. All the intrahepatic lesions detectable on MRI were also detected on [ 68 Ga]Ga-FAPI-46 PET/ CT (100%, 19/19) , whereas the sensitivity of [ 18 F]FDG PET/CT was only 57.9% (11/19) . Data from this study and previous ones show intense [ 68 Ga]Ga-FAPI-46 uptake caused by procedural-induced inflammation, which may be mistaken for local recurrence or residual viable tumour. Inflammation-induced nonspecific fibrosis can also lead to false-positive results; in this situation, morphological characteristics on MRI may help differentiate inflammatory lesions from true malignancy. Therefore, liver MRI is an important imaging modality with high specificity for the detection and characterisation of intrahepatic lesions. Our positivity rates and sensitivities of 100% for detecting intrahepatic lesions in patients with CCA were equal to those of Shi et al. [12] and Guo et al. [11] . In contrast, the positivity rates and sensitivities for detecting intrahepatic lesions in patients with HCC were slightly higher in our study and that of Shi et al. than in the study of Guo et al. (100%, 100%, and 94%, respectively). Moreover, our median TBR and median SUVmax values, as well as those of Shi et al., were higher than those of Guo et al. These differences might be related to different scanner technology. Cross calibration between the PET/CT and PET/MRI scanners is a key process for improving quantitative and semiquantitative parameters; it improves accuracy and provides external validity.

Metabolic tumour volume (MTV) and estimated stroma volume could not be evaluated in this study because almost half of the intrahepatic lesions showed no increased FDG uptake compared with the hepatic background, and the markedly high SUVmax of [ 68 Ga]Ga-FAPI-46 PET in the intrahepatic lesions affected the estimated stroma volume [16, 17] . Interestingly, we noted that [ 68 Ga]Ga-FAPI-46 uptake was higher in most CCA primary lesions than in HCC lesions, a finding that could be attributed to dense desmoplastic stroma and cancer-associated fibroblasts, which are considered hallmark histological features of cholangiocarcinoma [18] [19] [20] , high expression of glucose-6-phosphatase [21] , and histopathological grade, especially in HCC [5, 6, 22, 23] . One case of HCC showed a radiopharmaceutical uptake pattern that differed from the others, with the [ 18 F]FDG uptake being greater than [ 68 Ga]Ga-FAPI-46 uptake, especially in bone metastases, which were false-negative on [ 68 Ga]Ga-FAPI-46 PET. We hypothesise that these results were related to the heterogeneity of FAP in the metastatic lesions and the histopathology of the HCC lesions.

The F]FDG PET/CT were stronger in CCA than in HCC, which is possibly due to the characteristics of the primary tumours. In other words, the correlation between cancer-associated fibroblasts and glucose metabolism is stronger in CCA than in HCC.

CCA is associated with a "desmoplastic response", or the deposition of fibrotic or connective tissue, which is usually regarded as a response to malignant cells. Formation of stroma in CCA presents as a secondary insult in response to malignant growth [24] . Cancer-associated fibroblasts promote tumour progression [20] proteins and alter the cellular architecture through overproduction of extracellular matrix [25] . [ 18 F]FDG uptake reflects glycolytic activity which could be a prognostic indicator of aggressiveness. This knowledge suggests that FAPI PET and FDG PET might be a promising prognostic imaging workup for CCA, with their being a correlation between the tracers. However, the use of [ 18 F]FDG PET as a prognostic imaging tool is limited by its sensitivity, especially in small infiltrative mucinous cholangiocarcinoma and small duct cholangiocarcinoma [26] [27] [28] [29] .

Conversely, in hepatocellular carcinogenesis, cirrhosis and fibrosis generate a microenvironment that promotes tumourigenesis [24] . Most HCCs develop in fibrotic or cirrhotic livers, thereby proposing an important role of liver fibrosis in the premalignant environment of the liver [30] . Previous studies demonstrated that the gene signature of adjacent non-cancerous tissue contains key molecular information on recurrence and prognosis, illustrating that molecular prognostic determinants depend on both cancerous and non-cancerous tissue [31] . In our study, a weak correlation between cancer-related fibroblasts and glucose metabolism in HCC supports the knowledge that there is rare expression of GLUT-1 but increasing hexokinase-II expression in HCC [32, 33] . Nonetheless, prognostic imaging in HCC remains a challenge because of the limitations described above and histopathological variation.

DWI can be quantified with apparent diffusion coefficient (ADC) values and qualitatively assessed using visual analysis. Our mean ADC value for HCC was similar to that in a previous study [34] . The mean ADC values of hybrid PET/MR showed poor correlation with both desmoplastic response and glucose metabolism. This may imply that diffusion-weighted images cannot be used as a substitute for PET imaging in the specific molecular characterisation of HCC and CCA. Finally, multimodality hybrid imaging with novel radiopharmaceuticals can provide information on various aspects of cancer, resulting in a more complete biological profile.

This study has several limitations. First, this was an analysis of a heterogeneous population (patients with suspected, newly diagnosed, or previously treated liver cancer, including HCC and CCA) with a small sample size. There is a need for large-scale studies of specific cancer types. Second, pathological confirmation of nodal metastasis, distant metastasis, and local recurrent/residual tumour was not available.

[ 68 Ga]Ga-FAPI PET/CT with dedicated liver PET/MRI has superior potential for detecting hepatic malignancy than [ 18 F] FDG PET/CT or MRI alone.

Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries

Cholangiocarcinoma: lessons from Thailand

Projecting cancer incidence and deaths to 2030: the unexpected burden of thyroid, liver, and pancreas cancers in the United States

The diagnostic performance of gadoxetic acid disodium-enhanced magnetic resonance imaging and contrastenhanced multi-detector computed tomography in detecting hepatocellular carcinoma: a meta-analysis of eight prospective studies

Positron emission tomography scanning in the evaluation of hepatocellular carcinoma

Fluorine-18 FDG positron emission tomography for imaging of hepatocellular carcinoma

Clinical usefulness of positron emission tomography with fluorine-18-fluorodeoxyglucose in the diagnosis of liver tumors

Comparison of magnetic resonance imaging and 18-fludeoxyglucose positron emission tomography/ computed tomography in the diagnostic accuracy of staging in patients with cholangiocarcinoma: a meta-analysis. Medicine (Baltimore) 99:e20932

) 68Ga-FAPI PET/CT: tracer uptake in 28 different kinds of cancer

Comparison of [68Ga]Ga-DOTA-FAPI-04 and [18F] FDG PET/CT for the diagnosis of primary and metastatic lesions in patients with various types of cancer

Imaging fibroblast activation protein in liver cancer: a single-center post hoc retrospective analysis to compare [68Ga]Ga-FAPI-04 PET/CT versus MRI and [18F]-FDG PET/ CT

Comparison of PET imaging of activated fibroblasts and 18F-FDG for diagnosis of primary hepatic tumours: a prospective pilot study

Development of quinoline-based theranostic ligands for the targeting of fibroblast activation protein

A tumor-imaging method targeting cancer-associated fibroblasts

The first radiolabeled 68Ga-FAPI-46 for clinical PET applications using a fully automated iQS-TS synthesis system in Thailand

Diagnostic utility of Glut-1 and CA 15-3 in discriminating adenocarcinoma from hepatocellular carcinoma in liver tumors biopsied by fine-needle aspiration

Different glucose uptake and glycolytic mechanisms between hepatocellular carcinoma and intrahepatic mass-forming cholangiocarcinoma with increased (18) F-FDG uptake

Targeting cholangiocarcinoma

Stellate cells and the development of liver cancer: therapeutic potential of targeting the stroma

The role of cancer-associated myofibroblasts in intrahepatic cholangiocarcinoma

In vivo assessment of glucose metabolism in hepatocellular carcinoma with FDG-PET

Preoperative staging of cholangiocarcinoma and biliary carcinoma using 18F-fluorodeoxyglucose positron emission tomography: a meta-analysis

The role of F-18 FDG PET/CT in intrahepatic cholangiocarcinoma

Role of desmoplasia in cholangiocarcinoma and hepatocellular carcinoma

Extracellular matrix proteins: a positive feedback loop in lung fibrosis?

Fluorodeoxyglucose PET imaging in the evaluation of gallbladder carcinoma and cholangiocarcinoma

FDG PET in the diagnosis of hilar cholangiocarcinoma

Impact of integrated positron emission tomography and computed tomography on staging and management of gallbladder cancer and cholangiocarcinoma

Differences in 18F-FDG uptake and expression of glucose transporter between 2 distinct subtypes of mass-forming intrahepatic cholangiocarcinomas

The role of cancer-associated fibroblasts and fibrosis in liver cancer

Gene expression in fixed tissues and outcome in hepatocellular carcinoma

Assessment of Glut-1 expression in cholangiocarcinoma, benign biliary lesions and hepatocellular carcinoma

Glucose catabolism in cancer cells. Isolation, sequence, and activity of the promoter for type II hexokinase

Role of diffusion weighted imaging (DWI) for hepatocellular carcinoma (HCC) detection and its grading on 3T MRI: a prospective study

Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations

Acknowledgements We thank SOFIE for providing the [ 68 Ga]Ga-FAPI-46 precursor free of charge. We thank Edanz (www. edanz. com/ ac) for editing a draft of this manuscript. 

The online version contains supplementary material available at https:// doi. org/ 10. 1007/ s11307-022-01732-2.