68Ga-FAPI-04 Versus 18F-FDG PET/CT in the Detection of Hepatocellular Carcinoma

Background Fibroblast activation protein (FAP) is commonly expressed in activated stromal fibroblasts in various epithelial tumours. Recently, 68Ga-FAPI-04 has been used for tumour imaging in positron emission tomography/computed tomography (PET/CT). This study aimed to compare the diagnostic performances of 68Ga-FAPI-04 PET/CT and 18F-FDG PET/CT in hepatocellular carcinoma (HCC), and to assess factors associated with 68Ga-FAPI-04 uptake in HCC. Materials and Methods Twenty-nine patients with suspiciously HCC who received both 18F-FDG and 68Ga-FAPI-04 PET/CT were included in this retrospective study. The results were interpreted by two experienced nuclear medicine physicians independently. The maximum and mean standardized uptake values (SUVmax and SUVmean) were measured in the lesions and liver background, respectively. The tumour-to-background ratio (TBR) was then calculated as lesion’s SUVmax divided by background SUVmean. Results A total of 35 intrahepatic lesions in 25 patients with HCC were finally involved in the statistical analysis. 68Ga-FAPI-04 PET/CT showed a higher sensitivity than 18F-FDG PET/CT in detecting intrahepatic HCC lesions (85.7% vs. 57.1%, P = 0.002), including in small (≤ 2 cm in diameter; 68.8% vs. 18.8%, P = 0.008) and well- or moderately-differentiated (83.3% vs. 33.3%, P = 0.031) tumors. SUVmax was comparable between 68Ga-FAPI-04 and 18F-FDG (6.96 ± 5.01 vs. 5.89 ± 3.38, P > 0.05), but the TBR was significantly higher in the 68Ga-FAPI-04 group compared with the 18F-FDG group (11.90 ± 8.35 vs. 3.14 ± 1.59, P < 0.001). SUVmax and the TBR in 68Ga-FAPI-04 positive lesions were associated with tumour size (both P < 0.05), but not the remaining clinical and pathological features (all P > 0.05). Conclusions 68Ga-FAPI-04 PET/CT is more sensitive than 18F-FDG PET/CT in detecting HCC lesions, and 68Ga-FAPI-04 uptake is correlated mainly with tumour size.


INTRODUCTION
Hepatocellular carcinoma (HCC), the most frequent primary liver cancer, is the fourth most common cause of cancer-related death worldwide (1,2). The majority of HCCs occur in patients with underlying liver disease, mostly as a result of hepatitis B or C virus (HBV or HCV) infection or alcohol abuse (3). Unlike many other malignant tumours, HCC can be diagnosed by imaging based on non-invasive criteria without confirmatory pathology (4). Therefore, imaging plays a critical role in the detection and diagnosis of HCC. Conventional imaging modalities, including computed tomography (CT), magnetic resonance imaging (MRI) and ultrasound, are mainly utilized for anatomical evaluation, with limited value in the assessment of morphologically atypical lesions (5). In contrast, positron emission tomography/computed tomography (PET/CT) as one of the functional imaging approaches has the potential to provide relevant biological information in HCC and to improve response assessment (6). Furthermore, a subset of HCCs cannot be diagnosed non-invasively either because the patients do not have cirrhosis or the lesions do not follow established enhancement patterns on contrast-enhanced CT or MRI (1,3). Therefore, molecular imaging with PET/CT offers potential additional advantages to non-invasively confirm a diagnosis of HCC. However, the most widely available clinical PET tracer, 18 F-FDG, shows poor sensitivity for the detection of HCC ranging from 40% to 68%, mainly because of the relatively high glucose-6-phosphatase activity found in low-grade HCC (6). In addition, it is not useful for the detection of small HCC lesions (7,8). Therefore, several new tracers have been developed and applied for HCC detection, including 11 C-acetate, 11 Ccholine, 68 Ga-PSMA (9)(10)(11)(12).
Cancer-associated fibroblasts (CAFs) are among the most crucial components of the tumour microenvironment that creates a favourable microenvironment for tumour growth, invasion and metastasis (13,14). Fibroblast activation protein (FAP), a cell surface glycoprotein belonging to the serine protease family, is commonly expressed in activated stromal fibroblasts in various epithelial tumours (15,16). Recent investigations indicated that 68 Ga-labelled FAP inhibitor (FAPI) shows an equal or even improved tumour imaging with lower background uptake in the liver and the brain in comparison to 18 F-FDG in various cancers (17,18). 68 Ga-FAPI-04 PET/CT was also revealed to have high sensitivity in detecting hepatic malignancies (19)(20)(21). Therefore, 68 Ga-FAPI-04 may be a potential tracer for visualizing HCC by targeting CAFs that are abundant in the tumour microenvironment. This study aimed to comparatively assess the diagnostic performances of 68 Ga-FAPI-04 PET/CT and 18 F-FDG PET/CT in HCC and to assess factors associated with the uptake of 68 Ga-FAPI-04 in HCC.

Patients
This is a post-hoc analysis of a prior prospective study conducted at the Huashan Hospital of Fudan University. Twenty-nine patients with suspiciously incipient or recurrent HCC determined by clinical manifestations and conventional imaging techniques (CT, MRI and ultrasound) were included in this retrospective study. They underwent both 18 F-FDG and   68 Ga-FAPI-04 PET/CT examinations with an interval of one day before surgical treatment. In patients who underwent surgery or biopsy, the definitive diagnosis was confirmed by pathology. In patients who underwent transarterial chemoembolization (TACE), HCC diagnosis was based on a specific imaging pattern of hyperenhancement in the arterial phase and washout in the venous or delayed phase, on contrast-enhanced CT or MRI in the setting of liver cirrhosis (1,3). This study was approved by the institutional ethics committee, and written informed consent was obtained from all patients.

PET/CT Imaging
Whole-body static FDG PET/CT scans were obtained as a routine procedure on a dedicated PET/CT scanner (Biograph mCT Flow scanner, Siemens, Germany). Whole-body 68 Ga-FAPI-04 PET/CT scans were obtained on another PET/CT scanner (mMI510, Union imaging, Shanghai, China) within 60 min after intravenous injection of~185 MBq (~5 mCi) of 68 Ga-FAPI-04. Low-dose CT scans were obtained for attenuation correction and image fusion. PET images were acquired in the 3D mode, and reconstructed by the ordered subset expectation maximization 3D (OSEM 3D) method.
Because two different PETs were applied in this study, SUVs were normalized after data collection for PET/CT system performance harmonization. NEMA IEC body phantom (Data Spectrum Corporation, Durham, NC, USA) with 6 simulated lesion spheres (diameters of 10 mm, 13 mm, 17 mm, 22 mm, 28 mm and 37 mm, respectively) and 2, 4, 8, and 16 times of background activity (2 kBq/mL of background activity concentration) based on routine scan protocols was applied for SUV normalization.

18
F-FDG and 68 Ga-FAPI-04 PET/CT images were interpreted independently by two experienced nuclear medicine physicians blinded to other imaging and pathology results. The maximum standardized uptake value (SUV max ) was measured by delineating a spherical region of interest (ROI) for each lesion. The mean standardized uptake value (SUV mean ) of the liver background was measured by drawing a spherical ROI with 2 cm diameter in the non-tumour hepatic parenchyma of the right lobe in each patient. The tumour-to-background ratio (TBR) was calculated by dividing the lesion's SUV max with the background SUV mean . A lesion was considered to be positive on the basis of the visual judgment of elevated uptake in the tumour tissue by 2 experienced nuclear medicine physicians independently, supported by the calculation of the TBRs of 18 F-FDG and 68 Ga-FAPI-04, respectively. Any difference of opinion between these two physicians was settled by mutual agreement.

Statistical Analysis
All statistical analyses were conducted with the STATA statistical analysis software (StataCorp LLC, version 15.1). Categorical variables were presented as frequency and percentage, and continuous variables as mean ± standard deviation (SD). The McNemar's test and Fisher exact test were performed to compare categorical variables. Non-parametric tests were carried out for the comparison of continuous variables with non-normal distribution. The Spearman rank correlation coefficient was determined to assess the correlation between continuous variables with non-normal distribution. Two-tailed P < 0.05 was considered statistically significant.

Patient Characteristics
Twenty-nine patients were included in the current study, including 23 treated by hepatic surgery, 5 administered TACE, and one that underwent biopsy only. Except for 3 patients who were diagnosed with benign hepatic nodules, the remaining 26 patients were diagnosed with HCC. One recurrent HCC case who underwent surgical resection had extensive peritoneal dissemination but no intrahepatic lesions. Therefore, 25 HCC patients with 35 intrahepatic lesions were finally involved in the statistical analysis. The study flowchart is presented in Figure 1. According to microvascular invasion (MVI) number and distribution, 2, 8 and 10 patients were categorized into the M0 (no MVI), M1 (≤ 5 MVI in adjacent liver tissue ≤ 1 cm away from the HCC), and M2 (> 5 MVI or MVI in adjacent liver tissue > 1 cm away from the HCC) groups, respectively. According to the American Joint Committee on Cancer (AJCC) cancer staging system (8th Edition), 5, 10, 4 and 1 patients were categorized into stage I, II;, III and IV, respectively. The general characteristics of the 25 HCC patients are summarized in Table 1.

Comparison of 68 Ga-FAPI-04 With 18 F-FDG in Patient-Based Analysis
The results of the patient-based analysis of 18    Five HCC patients whose diagnosis was based on non-invasive criteria underwent TACE instead of hepatic surgery and, therefore, had no pathological data. HBsAg, hepatitis B surface antigen; Anti-HCV, anti-hepatitis C virus antibody; AFP, a-fetoprotein; MVI, microvascular invasion; AJCC TNM, American Joint Committee on Cancer tumournode-metastasis.
intrahepatic lesions compared with 18 F-FDG PET/CT (85.7% vs. 57.1%, P = 0.002). In subgroup analysis, 68 Ga-FAPI-04 PET/CT was more sensitive than 18 F-FDG PET/CT in the detection of intrahepatic lesions in patients with cirrhosis, low a-fetoprotein (AFP), multiple tumours, and non-serious MVI (M0 and M1) (all P < 0.05). Moreover, 68 Ga-FAPI-04 PET/CT detected 4 of the 5 lesions in patients with stage I disease, whereas 18 F-FDG PET/CT did not reveal any abnormal finding in these patients ( Figure 2). The sensitivity of 18 F-FDG PET/CT was associated with AJCC TNM stage (P = 0.016), while that of 68 Ga-FAPI-04 PET/CT was correlated with serum AFP levels (P = 0.045). These findings suggested that 68 Ga-FAPI-04 PET/CT was more sensitive than 18 F-FDG PET/CT in the detection of intrahepatic lesions, particularly in patients with cirrhosis, low AFP, multiple HCCs, and non-serious MVI.

Comparison of 68 Ga-FAPI-04 With 18 F-FDG in Lesion-Based Analysis
The results of lesion-based analysis of 18 F-FDG and 68 Ga-FAPI-04 PET/CT are summarized in Table 3.  Five HCC patients whose diagnosis was based on non-invasive criteria underwent TACE instead of hepatic surgery and, therefore, had no pathological data. *, statistically significant; AFP, a-fetoprotein; MVI, microvascular invasio; AJCC TNM, American Joint Committee on Cancer tumour-node-metastasis.
was more sensitive than 18 F-FDG PET/CT in detecting small HCCs (≤ 2 cm in diameter) (P = 0.008) and well-or moderatelydifferentiated HCCs (P = 0.031), but there were no significant sensitivity differences between the 2 tracers in the detection of HCCs > 2 cm in diameter (both P > 0.05) and poorlydifferentiated or undifferentiated HCCs (P > 0.05). The sensitivities of 18 F-FDG and 68 Ga-FAPI-04 PET/CT were significantly related to the size of intrahepatic lesions (both P < 0.05). These findings indicated that 68 Ga-FAPI-04 PET/CT was more sensitive than 18 Table 4). Although the lesion uptake (SUV max ) of 68 Ga-FAPI-04 was similar to that of 18 F-FDG (P > 0.05), its TBR was significantly higher than that of 18 F-FDG (P < 0.001) ( Figure  3A). Particularly, the background uptake (SUV mean ) of 68 Ga-FAPI-04 was much lower than that of 18 F-FDG in each patient. The SUV mean of 68 Ga-FAPI-04 in patients with cirrhosis was significantly higher than that of patients without cirrhosis (0.76 ± 0.39 vs. 0.40 ± 0.07, P < 0.001); however, no significant difference was obtained in SUV mean of 18 F-FDG between these two groups (1.82 ± 0.39 vs. 1.97 ± 0.44, P > 0.05) ( Figure 3B).
The SUV max of 18 F-FDG in positive lesions was associated with the degree of MVI (P = 0.048) and tumour differentiation (P = 0.045), while the TBR was only associated with tumour differentiation (P = 0.045). In contrast, the SUV max and TBR of 68 Ga-FAPI-04 in positive lesions were associated only with tumour size (both P < 0.05), but not with other clinical and  pathological features (all P > 0.05). In Spearman correlation analysis ( Figure 3C), the SUV max and TBR of 68 Ga-FAPI-04 in positive lesions were correlated with tumour size (rSUV max = 0.43, rTBR = 0.53, both P < 0.05). Further subgroup analysis revealed these correlations in poorly-differentiated or undifferentiated HCCs (rSUV max = 0.69, rTBR = 0.62, both P < 0.05), rather than well-or moderately-differentiated ones (both P > 0.05). For 18 F-FDG, SUV max , but not TBR, was correlated with tumour size (rSUV max = 0.48, P = 0.033). However, neither SUV max nor TBR exhibited a correlation with tumour size in subgroup analysis (all P > 0.05).

Characteristics of Extrahepatic Metastases and Other Benign Lesions
Lymph node metastasis in one patient with poorly-differentiated HCC showed strong uptake of 68 Ga-FAPI-04, but undetectable uptake of 18 F-FDG ( Figure 4). 68 Ga-FAPI-04 PET/CT detected a small metastatic lesion that was not revealed by 18

DISCUSSION
Nowadays, 18 F-FDG as the most widely available clinical PET tracer has been increasingly utilized for detecting extrahepatic metastases (22), TNM staging (23), selecting patients for liver transplantation (23), and predicting tumour progression or recurrence after treatments (24,25). However, 18 F-FDG has limited value in the early diagnosis of HCC because of its low sensitivity (6,26). Therefore, there remains an urgent need for highly sensitive tracers in the early diagnosis of HCC by PET/CT. Recently, 68 Ga-labelled FAPI was shown to be a novel tracer in PET/CT imaging of various cancers due to its high tumour-tobackground contrast (17,27). Furthermore, 68 Ga-FAPI-04 PET/ CT has high sensitivity in detecting hepatic malignancies, including HCC and ICC (19,21).
In line with previous studies reporting a range from 40% to 68%, the sensitivity of 18 F-FDG PET/CT in the detection of HCC was 57.1% in the present study (6). 68 Ga-FAPI-04 PET/CT had a better sensitivity (85.7%) than 18 F-FDG PET/CT in the detection of intrahepatic lesions in HCC patients. Of note, 68 Ga-FAPI-04 PET/CT was capable of detecting more than half of small HCC lesions (11 of 16, ≤ 2 cm in diameter) in the present cohort, whereas 18 F-FDG PET/CT detected only three of the 16 lesions, which is consistent with previous studies that consider 18 F-FDG an inappropriate tracer for visualizing small HCCs (7,8). Furthermore, 68 Ga-FAPI-04 PET/CT exhibited a relatively higher sensitivity in the detection of well-or moderatelydifferentiated HCCs (10 of 12, histologic grade I or II) compared with 18 F-FDG PET/CT (4 of 12). The poor sensitivity of 18 F-FDG PET/CT in detecting low-grade HCC is probably related to enhanced glucose-6-phosphatase activity causing the dephosphorylation of 18 F-FDG-6-PO 4 , which is therefore not trapped in HCC cells, resulting in false-negative results (6,28,29). In contrast, a CAF-targeting tracer can circumvent highly heterogeneous avidity exhibited by some tracers that target the tumour per se, because CAFs are among the most abundant stromal components in the tumour microenvironment of many solid tumours, and are found even at the early stages of tumorigenesis (30,31). The above correlation analysis of 68 Ga-FAPI-04 revealed that neither positive incidence nor uptake intensity was associated with tumour differentiation. Moreover, 68 Ga-FAPI-04 showed a high lesion-to-background contrast in the liver, which may partially increase sensitivity that is affected by the partial volume effect of PET/CT. As such, 68 Ga-FAPI-04 PET/CT can make up for the deficiencies of 18 F-FDG PET/CT in the detection of low-grade HCC. Additionally, a higher detection rate was observed with 68 Ga-FAPI-04 PET/CT in high-grade HCC in comparison with 18 F-FDG PET/CT, although this difference was not statistically significant. Therefore, 68 Ga-FAPI-04 PET/CT appears to be a promising new approach for the detection of intrahepatic HCC lesions with higher sensitivity compared with 18 F-FDG PET/CT. 68 Ga-FAPI-04 PET/CT could not detect 5 intrahepatic HCC lesions with the diameter within 2 cm in this cohort. These negative results may be due to the similar uptake intensity of 68 Ga-FAPI-04 between small HCC lesions and the liver background of cirrhosis. Hypoxia may be a reasonable explanation for the positive correlation between the uptake intensity of 68 Ga-FAPI-04 and tumour size. It has been reported that the degree of hypoxia correlates positively with tumour size (32) and that hypoxia is a potent factor inducing the expression of FAP in CAFs (33). Therefore, the degree of hypoxia is mild in small HCCs, leading to the low uptake of 68 GA-FAPI-04 in these lesions. In the cirrhotic liver, FAP is strongly expressed by activated hepatic stellate cells (34,35). In line with a previous study (19), this work also found that patients with cirrhosis presented elevated uptake of 68 Ga-FAPI-04 in the hepatic parenchyma compared with those without cirrhosis. Therefore, small lesions have relatively lower uptake of 68 Ga-FAPI-04 in comparison with large ones, which makes them susceptible to being masked by the background of cirrhosis.
Although 18 F-FDG PET/CT has a potential value in detecting extrahepatic metastases in HCC patients (7,8,36,37), heterogeneous uptake of 18 F-FDG in metastatic nodules remains a major reason limiting its wide application for tumour staging. CAFs play a critical role in constructing a microenvironment that favours tumour progression at the primary site, and are, moreover, responsible for creating a premetastatic niche in distal organs and triggering the subsequent metastatic events (30,31,38). Therefore, CAF-targeting tracers seem to be optimal candidates for PET/CT in the evaluation of extrahepatic metastases. In the present study, lymph node metastasis at the porta hepatis in one patient with poorlydifferentiated HCC presented a clear visualization in 68 Ga-FAPI-04 PET/CT, but an obscure image in 18 F-FDG PET/CT ( Figure 4). Separately, one small metastatic lesion in another recurrent HCC case with extensive peritoneal dissemination was only clearly visualised by 68 Ga-FAPI-04 PET/CT ( Figure 5). It appears that 68 Ga-FAPI-04 PET/CT may outperform 18 F-FDG PET/CT in detecting extrahepatic lesions in patients with advanced HCC. Nevertheless, the comparison of 68 Ga-FAPI-04 with 18 F-FDG in terms of applicability to the detection of extrahepatic metastasis of HCC needs to be clarified in future studies.
Despite the high sensitivity of 68 Ga-FAPI-04 PET/CT in the detection of malignancies, some benign lesions confirmed by pathological examinations in the current cohort presented positive results as well. Especially, AML, FNH, and one of two inflammatory nodules presented elevated uptake of 68 Ga-FAPI-04 in contrast to the hepatic background, perhaps because of the enhanced fibrosis around or within lesions. Moreover, increasing uptake of 68 Ga-FAPI-04 was observed in the postoperative area of the liver in one patient with recurrent HCC, which is consistent with a previous study that considered 68 Ga-FAPI-04 an inappropriate tracer for the discrimination between abnormal malignant progression and normal postoperative reaction (39). Nevertheless, negligible 68 Ga-FAPI-04 uptake has been observed in some other benign hepatic lesions such as adenoma (20,39), dysplastic nodule (20), granuloma (21), and haemangioma (21). Collectively, great caution should be exercised when regarding intrahepatic lesions with elevated uptake of 68 Ga-FAPI-04 as malignancy.
There were several limitations in the present study. First, it is unethical to biopsy all lesions as part of a research study, although that is perhaps not practical or needed. As a result, the lack of pathological data in 5 HCC patients whose confirmative diagnosis was based on non-invasive criteria may yield latent bias. Another limitation is that the present cohort only included a small number of patients with suspicious HCC who were willing to receive both 68 Ga-FAPI-04 and 18 F-FDG PET/CT examinations. Therefore, selection bias was inevitable. Finally, different scanners were used in this study for 68 Ga-FAPI-04 and 18 F-FDG imaging because of the different production places of PET tracers. For this reason, SUV normalization was applied after data collection for PET/CT system performance harmonization.

68
Ga-FAPI-04 PET/CT is more sensitive than 18 F-FDG PET/CT in detecting intrahepatic HCCs. This outperformance is more prominent in the detection of small and well-or moderatelydifferentiated HCCs. The uptake of 68 Ga-FAPI-04 was correlated mainly with tumour size in this study; therefore, 68 Ga-FAPI-04 PET/CT can be considered a very promising imaging modality in HCC diagnosis.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by ethics committee at Huashan Hospital of Fudan University. The patients/participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

AUTHOR CONTRIBUTIONS
LQ, FX, LL, WZ, and HW designed the study, interpreted the data and led the writing and review of the manuscript. HW, HJ, JC, WZ, and LL enrolled patients and collected clinical data. YG, JZ, QH, YK, and SR performed examinations. All authors contributed to the article and approved the submitted version.