⁶⁸Ga-PSMA PET/CT Versus ¹⁸F-FDG PET/CT for Imaging of Hepatocellular Carcinoma

Abstract

Objectives:

This study aimed to compare the metabolic parameters obtained from ¹⁸fluorine-fluorodeoxyglucose (¹⁸F-FDG) positron emission tomography/computed tomography (PET/CT) and gallium-68 (⁶⁸Ga)-prostate-specific membrane antigen (PSMA) PET/CT and investigate the relationship between serum alpha-fetoprotein and PET scan parameters in patients with hepatocellular carcinoma.

Methods:

Fourteen patients were recruited after dynamic magnetic resonance imaging (MRI) of the upper abdomen, and ¹⁸F-FDG and ⁶⁸Ga-PSMA PET/CT imaging studies were conducted. Regions of interest (ROIs) were drawn from lesion-free liver tissue, abdominal aorta (A), and right medial gluteal muscle (G) for the background activity. Maximum standard uptake value (SUV_max) of these regions were compared with the SUV_max of primary tumor (T).

Results:

On visual assessment, five patients (36%) experienced low ¹⁸F-FDG uptake in the primary lesion, three patients (21%) experienced moderate uptake, and six patients (43%) experienced high uptake. However, only one patient (7%) showed low ⁶⁸Ga-PSMA uptake, two patients (14%) showed moderate uptake, and 11 patients (79%) showed high uptake. Four patients with a low ¹⁸F-FDG uptake showed high ⁶⁸Ga-PSMA uptake, while one patient exhibited low uptake with both ¹⁸F-FDG and ⁶⁸Ga-PSMA. The number of lesions on ⁶⁸Ga-PSMA PET/CT and MRI was significantly higher than ¹⁸F-FDG PET/CT (p=0.042 and 0.026, respectively). T/A and T/G values were significantly higher in ⁶⁸Ga-PSMA than ¹⁸F-FDG (p=0.002 and 0.002, respectively).

Conclusion:

⁶⁸Ga-PSMA PET/CT is superior to ¹⁸F-FDG PET/CT in the staging of hepatocellular carcinoma. High ⁶⁸Ga-PSMA uptake could be promising for PSMA-targeted radionuclide treatments.

Introduction

Liver cancer is the 6^th most frequent malignancy and the 4^th most common cause of cancer-related deaths worldwide. Hepatocellular carcinoma (HCC), which develops due to major risk factors such as hepatitis B virus, hepatitis C virus, aflatoxin-containing foods, and non-alcoholic steatohepatitis, accounts for 75%-85% of primary liver cancers (1). Conventional dynamic contrast-enhanced imaging methods, including computed tomography (CT) and magnetic resonance imaging (MRI), are routinely used in the diagnosis of HCC, with 62%-82% sensitivity and over 90% specificity. Nodules larger than 1 cm show high contrast enhancement in the arterial phase of CT and MRI and wash out in venous and late phases (2). Alpha-fetoprotein (AFP), a serum biomarker, is one of the most commonly used markers in HCC screening and diagnosis. However, its sensitivity and specificity are unsatisfactory, especially in early-stage HCCs (3). The histopathological examination of the tumor tissue is the gold standard in the definitive diagnosis of HCC, but it may cause tissue damage and seeding along the biopsy tract (4).

¹⁸Flourine-fluorodeoxyglucose (¹⁸F-FDG) positron emission tomography(PET)/CT is associated with the aggressiveness of HCC; moderately and well-differentiated HCCs exhibit a low ¹⁸F-FDG uptake, while poorly differentiated HCCs show high uptake (5,6). ¹⁸F-FDG PET/CT shows a high sensitivity in detecting lymph nodes and extrahepatic metastases, which are poor prognostic factors for HCC, but show a low sensitivity in detecting primary HCC lesions (7,8).

Prostate-specific membrane antigen (PSMA) is a type 2 transmembrane protein and is overexpressed in prostate cancer (PCa). ⁶⁸Ga-PSMA PET/CT is widely used in staging, evaluating the treatment response, and assessing relapse in PCa (9). However, in many solid tumors, including HCCs, high PSMA uptake indicates neoangiogenesis (10,11).

In this prospective study, we compared the metabolic parameters obtained from ¹⁸F-FDG PET/CT and ⁶⁸Ga-PSMA-11 PET/CT and investigated the association between PET parameters and serum AFP in patients with HCC.

Material and Methods

Patient Characteristics

Fourteen patients [13 males; 1 female; mean age: 63.8±6.0 years (58-76)] were included in this study. Twelve patients had a Child-Pugh (CP) score “A” cirrhosis, and two patients had a CP-B cirrhosis. Twelve patients had a newly diagnosed HCC, one patient had a history of transarterial chemoembolization (TACE), and one had radiofrequency ablation + TACE for HCC. While six patients had a histopathological confirmation, eight patients were diagnosed with HCC based on radiological findings and serum AFP levels. Patients were recruited after dynamic MRI of the upper abdomen, and ¹⁸F-FDG and ⁶⁸Ga-PSMA-11 PET/CT imaging studies were conducted in the same week. AFP and routine laboratory tests and ¹⁸F-FDG PET/CT were performed for all patients on the same day. Patients who previously received chemotherapy or had a history of hepatic tumor surgery were excluded from the study. This study was conducted in concordance with the local good clinical practice guidelines and current laws. The Local Ethics Committee of İstanbul Training and Research Hospital approved this study under the decision number: 2018/1297. Written informed consent was obtained from all patients.

PET/CT Scan and Evaluation

Whole-body PET/CT imaging was performed 60 min after intravenous injection of ¹⁸F-FDG (3.5-5.5 MBq/kg) and ⁶⁸Ga-PSMA-11 (2-2.5 MBq/kg) in a PET/CT scanner [mCT 20 ultra HD LSO PET/CT (Siemens molecular imaging, Hoffmann Estates, Illinois, USA)] on different days. CT imaging was performed in the craniocaudal direction with a 5 mm slice thickness and rotation time of 0.5 sec [80-140 kV, 20-266 mAs, 0.8 pitch, and 512x512 matrix (personalized settings determined by the automatic exposure control system; automatically defined by the software used by the manufacturer, depending on the patient)]. Then, PET imaging was performed in the same range through the craniocaudal direction for 2 min for each PET bed; ultra HD images were acquired using the time of flight + true X algorithm at iteration two and subset 16 values for reconstruction.

¹⁸F-FDG and ⁶⁸Ga-PSMA-11 PET/CT images were both evaluated by two nuclear medicine physicians with at least 10 years of experience in PET/CT, and decisions were made with consensus. Both PET/CT studies were scored visually. 1: Low uptake (equal or less than liver), 2: Moderate uptake (slightly higher than liver), 3: High uptake (markedly higher than liver). SUV_max of primary lesions were acquired by drawing a volume of interest to include the lesion in all three planes in ¹⁸F-FDG and ⁶⁸Ga-PSMA-11 PET/CT. Moreover, regions of interest of 1 cm diameter were drawn from lesion-free liver tissue (L), abdominal aorta (A), and right medial gluteal muscle (M) for background maximum standard uptake value (SUV_max). Using these three background SUV_max, tumor to normal liver parenchyma (T/L), tumor to abdominal aorta (T/A), and tumor to gluteal muscle (T/G) parameters were calculated separately.

Statistical Analysis

SPSS version 21.0 software (IBM Corporation, Armonk, New York, USA) was used for statistical analyses of the variables. The normality of one-variable data was tested with the Shapiro-Francia test, while variance homogeneity was evaluated using Levene’s test. Mann-Whitney U test was used to compare independent and non-normally distributed variables, while the Wilcoxon signed-rank test was used to compare dependent and normally distributed variables. Pearson and Spearman’s rho tests were used to analyze the correlation of variables. The variables had a 95% confidence interval, and a p value less than 0.05 was considered significant.

Results

Six patients exhibited a bilobar involvement, while eight patients had a lobar involvement. On MRI, nine patients showed a mosaic enhancement pattern, and five patients showed a homogeneous enhancement pattern. The median size of primary tumors on MRI is 80.5 mm (20 mm-140 mm). The smallest lesion detected on ⁶⁸Ga-PSMA PET had a diameter of 8 mm, and the lesion had been described on MRI.

On the visual evaluation, five patients (36%) showed a low ¹⁸F-FDG uptake in the primary lesion, three patients (21%) showed a moderate ¹⁸F-FDG uptake, and six patients (43%) showed a high ¹⁸F-FDG uptake. In contrast, one patient (7%) showed low ⁶⁸Ga-PSMA uptake, two patients (14%) showed moderate ⁶⁸Ga-PSMA uptake, and 11 patients (79%) showed high ⁶⁸Ga-PSMA uptake. Four patients with low ¹⁸F-FDG uptake showed high ⁶⁸Ga-PSMA uptake (Figure 1), while one patient exhibited low uptake with both ¹⁸F-FDG and ⁶⁸Ga-PSMA (Figure 2). Two patients with moderate ¹⁸F-FDG uptake showed higher ⁶⁸Ga-PSMA uptake (Figure 3). In contrast, one patient with moderate ⁶⁸Ga-PSMA uptake showed higher ¹⁸F-FDG uptake (Table 1).

Figure 1.

A 69-year-old male diagnosed with hepatocellular carcinoma (AFP: 4.5 ng/mL) by magnetic resonance imaging. The left lobe mass showed (A) an intense ⁶⁸Ga-PSMA (T/L: 4.49), while it showed no significant ¹⁸F-FDG (B) uptake (T/L: 1.01)

AFP: Alpha-fetoprotein, ⁶⁸Ga: Gallium-68, ¹⁸F-FDG: ¹⁸Fluorine-fluorodeoxyglucose, T: Tumor uptake, L: Normal liver parenchyma uptake, PSMA: Prostate-specific membrane antigen

Figure 2.

A 62-year-old male with alcohol-induced cirrhosis. Magnetic resonance imaging revealed a mass compatible with hepatocellular carcinoma, showing a typical contrast enhancement pattern (not shown). AFP: 1.648 ng/mL. ⁶⁸Ga-PSMA: (A) SUV_max: 9.3; T/L 0.93. ¹⁸F-FDG: (B) SUV_max: 4.5; T/L 0.85

AFP: Alpha-fetoprotein, ¹⁸F-FDG: ¹⁸Fluorine-fluorodeoxyglucose, SUV_max: Maximum standard uptake value, PSMA: Prostate-specific membrane antigen, T: Tumor uptake, L: Normal liver parenchyma uptake

Figure 3.

A 67-year-old cirrhotic male with hepatocellular carcinoma in the left lobe of the liver. ⁶⁸Ga-PSMA: (A) SUV_max: 28.1; T/L: 3.95; T/G: 31.2. ¹⁸F-FDG: (B) SUV_max: 10.1; T/L: 2.76; T/G: 7. In arterial phase CT images (C), it is seen that PSMA uptake is higher in the hyperenhancement areas and ¹⁸F-FDG uptake is higher in the non-enhancing areas of the tumor

T: Tumor uptake, L: Normal liver parenchyma uptake, G: Gluteus medius muscle uptake, ⁶⁸Ga: Gallium-68, ¹⁸F-FDG: ¹⁸Fluorine-fluorodeoxyglucose, PSMA: Prostate-specific membrane antigen, SUV_max: Maximum standard uptake value, CT: Computed tomography

Table 1. ¹⁸F-FDG and ⁶⁸Ga-PSMA PET/CT findings of patients.

The total number of liver lesions on ⁶⁸Ga-PSMA PET/CT, MRI, and ¹⁸F-FDG PET/CT are 61, 57, and 30, respectively. The number of liver lesions on ⁶⁸Ga-PSMA PET/CT and MRI were significantly higher than ¹⁸F-FDG PET/CT (p=0.042 and 0.026, respectively). There was no statistically significant difference between the number of liver lesions on MRI and ⁶⁸Ga-PSMA PET/CT (p=0.593) (Table 2).

Table 2. Comparison of ¹⁸F-FDG and ⁶⁸Ga-PSMA PET/CT uptake ratios (SUV_max).

⁶⁸Ga-PSMA PET/CT revealed a pathologically increased radiotracer uptake in the abdominal lymph nodes of four patients. Of these four patients, one patient had no ¹⁸F-FDG uptake. Two patients had ¹⁸F-FDG and ⁶⁸Ga-PSMA-positive mediastinal lymph nodes, while the remaining patient, who was later histopathologically diagnosed with anthracosis, had only ¹⁸F-FDG uptake. One patient had focally increased radiotracer accumulation in the prostate gland on both ¹⁸F-FDG and ⁶⁸Ga-PSMA PET/CT, consistent with a synchronous tumor in the prostate.

The median SUV_max of primary lesions in ¹⁸F-FDG and ⁶⁸Ga-PSMA PET/CT were 6.45 (range: 3.7-21.3) and 16.7 (range: 9.3-48.9), respectively. When median T/L, T/A, and T/G ratios were compared, T/L ratio had no statistically significant difference between ¹⁸F-FDG and ⁶⁸Ga-PSMA (p=0.331), whereas T/A and T/G were significantly higher in ⁶⁸Ga-PSMA than ¹⁸F-FDG (p=0.002 and 0.002, respectively).

Of our six patients with histopathology results, one was reported as having poorly differentiated HCC, two, as well-differentiated HCC, and three, as HCC without specifying differentiation levels. ¹⁸F-FDG (T/B: 1.48) and PSMA (T/B: 1.44) uptake in the patient with poorly differentiated HCC were similar, and a moderate radiopharmaceutical uptake was observed in both studies. The primary tumors in the two patients with well-differentiated HCC showed a low level of radiopharmaceutical uptake (T/B: 1.11 and 1.12) in ¹⁸F-FDG PET/CT and intense radiopharmaceutical uptake (T/B: 2.74 and 3.96) in PSMA PET/CT. Two of the three patients whose differentiation level was not specified showed intense ¹⁸F-FDG (T/B: 3.86 to 2.76) and PSMA (T/B:10.1 to 3.98) uptake, while one patient showed low ¹⁸F-FDG (T/B: 0.86) and intense PSMA (T/B: 2.1) uptake.

When the relationship between the laboratory results and PET parameters was examined, serum AFP levels showed a statistically significant positive correlation with ¹⁸F-FDG T/A ratio only (r=0.641 p=0.007). However, there was no correlation between ⁶⁸Ga-PSMA parameters and serum AFP (p>0.05).

Discussion

In this prospective study, we investigated the contribution of ⁶⁸Ga-PSMA PET/CT to the evaluation of HCCs. The most important findings in this study were primary tumors showing higher ⁶⁸Ga-PSMA uptake on visual assessment and the ability of ⁶⁸Ga-PSMA to detect more primary and metastatic lesions compared with ¹⁸F-FDG. Conventional imaging methods, such as MRI, are routinely used as the first choice in the diagnosis of HCC due to the typical enhancement pattern with hyperenhancement in the arterial phase and wash out in the portal and late venous phases (12). Although MRI is generally sufficient for diagnosis, it cannot provide information about the biological behavior of HCC. PET radiopharmaceuticals, especially ¹⁸F-FDG, are helpful in this context. ¹⁸F-FDG PET/CT appears as an important non-invasive diagnostic tool, especially in terms of detecting metastatic lesions in HCC. It is known that ¹⁸F-FDG PET/CT findings constitute a stronger prognostic factor than the nodule’s size and number, as described in Milan criteria. Because ¹⁸F-FDG avidity may predict the risk of relapse in patients who are planned to undergo liver transplantation, resection, or ablation, it may have a direct effect on the transplantation and ablation outcome (13,14). However, ¹⁸F-FDG PET/CT has a low sensitivity in HCC due to overexpression of multidrug resistance protein and increased glucose-6-phosphatase activity in HCC cells, and its use in routine clinical practice is limited (15,16). Therefore, different radiopharmaceuticals have been investigated for the evaluation of primary and extrahepatic metastases of HCCs. Agents with high sensitivity, including ¹⁸F-fluorocholine and ¹¹C-acetate, have a relatively poorer availability and, therefore, a limited use (17,18).

Non-prostate solid tumors may exhibit a wide endothelial PSMA expression, associated with neoangiogenesis and vascular growth factor regulation (19,20,21). Recent studies have shown that HCC shows a higher ⁶⁸Ga-PSMA uptake compared with ¹⁸F-FDG (11,22). In our study, ⁶⁸Ga-PSMA PET/CT revealed more lesions than ¹⁸F-FDG PET/CT, and the lesions showed a higher PSMA uptake compared with ¹⁸F-FDG.

A recent study has reported that peritumoral/vascular expression of PSMA is greatly associated with grade 3 HCC (5/6, 83.3%) but can also be observed in grade 2 HCC (10/15, 66.7%). This was associated with the clinicopathological characteristics of HCC. Fibrolamellar HCC, normal hepatic tissue, and non-neoplastic cirrhotic tissue are reported to not overexpress PSMA. HCCs, arising in the setting of cirrhosis (9/10, 90.0%), show a significantly increased peritumoral/vascular PSMA expression compared with non-cirrhotic HCCs (6/12, 50%) (p<0.05) (23).

In a study that evaluates seven patients and 37 lesions, Kesler et al. (11) demonstrated ⁶⁸Ga-PSMA uptake to be much higher than the background hepatic activity in 36/37 lesions. Twenty-eight lesions with no ¹⁸F-FDG uptake showed high ⁶⁸Ga-PSMA uptake, while eight lesions showed both ¹⁸F-FDG and ⁶⁸Ga-PSMA uptake. In their study involving 19 patients, Kuyumcu et al. (22) reported that ⁶⁸Ga-PSMA uptake was higher than ¹⁸F-FDG uptake in nine patients. Four patients had a higher ¹⁸F-FDG uptake compared with ⁶⁸Ga-PSMA, while two patients showed no uptake (22). In our study, 13 patients had an increased ⁶⁸Ga-PSMA uptake, while nine patients had an increased ¹⁸F-FDG uptake. Four patients with no ¹⁸F-FDG uptake had a high ⁶⁸Ga-PSMA uptake, however, one patient showed neither ¹⁸F-FDG nor ⁶⁸Ga-PSMA uptake. One patient with moderate ⁶⁸Ga-PSMA uptake exhibited a higher ¹⁸F-FDG uptake. Because of these results, the staging and treatment strategy can be changed through using ⁶⁸Ga-PSMA PET/CT instead of ¹⁸F-FDG PET/CT for metabolic imaging in patients with HCC.

Kesler et al. (11) reported extrahepatic involvement in two of seven patients, while Kuyumcu et al. (22) reported extrahepatic involvement in one patient. In our study, four patients had extrahepatic involvement on ⁶⁸Ga-PSMA PET/CT, whereas ¹⁸F-FDG PET/CT failed to reveal the involvement in one of these patients. One patient with ⁶⁸Ga-PSMA-negative mediastinal lymph nodes, which was later evidenced to be anthracosis by histopathological examination, showed false positivity in ¹⁸F-FDG PET/CT. This supports the deduction that ⁶⁸Ga-PSMA PET/CT may provide more accurate staging than ¹⁸F-FDG PET/CT.

Kuyumcu et al. (22) found no statistically significant difference between the mean SUV_max of primary tumor in ¹⁸F-FDG and ⁶⁸Ga-PSMA PET/CT and T/L ratios. The researchers only evaluated T/L ratio but did not analyze T/A and T/G ratios (22). Because it has recently been reported that T/A ratios have a prognostic significance in rectal cancer (24), we analyzed T/A and T/G ratios in our study as well. We observed no statistical significance in terms of T/L ratios between ¹⁸F-FDG and ⁶⁸Ga-PSMA PET/CT, while we found significantly higher T/A and T/G ratios and SUV_max in ⁶⁸Ga-PSMA PET/CT.

Since patients with histopathology results are few in our study, it will be difficult to make a clear evaluation of the relationship between HCC differentiation and PSMA involvement. However, PSMA uptake was significantly higher than ¹⁸F-FDG uptake in our patients with well-differentiated HCC. In our patient with less differentiated HCC, ¹⁸F-FDG, and PSMA uptakes were found to be similar compared with the background activity. Low ¹⁸F-FDG uptake is an expected finding in patients with well-differentiated and moderately differentiated HCC, and our findings on the relationship between HCC differentiation and ¹⁸F-FDG involvement are consistent with the literature (25). Since there are no studies in the literature, no correlation could be made between PSMA PET/CT and HCC differentiation.

In this preliminary study, no relationship was found between ⁶⁸Ga-PSMA tumor uptake and serum AFP level, suggesting that tumor angiogenesis and AFP production are independent parameters in HCC.

Study Limitations

First limitation of the current study is the relatively small number of patients, although our population is similar to other prospective studies in the literature. Second, not all patients had a histopathologically confirmed diagnosis, and some patients were diagnosed according to typical radiological findings.

Conclusion

⁶⁸Ga-PSMA PET/CT is superior to ¹⁸F-FDG PET/CT in the diagnosis and staging of HCC. These preliminary findings show that ⁶⁸Ga-PSMA PET/CT has a supportive role for MRI in T staging, especially in demonstrating multicentric tumors, and it can be superior to MRI in demonstrating extrahepatic involvement. High PSMA uptake is promising for PSMA-targeted radionuclide treatments in metastatic HCC, which responds poorly to standard chemotherapy regimens. ⁶⁸Ga-PSMA PET/CT may also be helpful in evaluating treatment response, warranting further prospective studies in this area.