Visual Abstract
Abstract
Fibroblast activation protein is overexpressed in the stroma of several cancer types. 18F-fibroblast activation protein inhibitor (FAPI)–74 is a PET tracer with high selectivity for fibroblast activation protein and has shown high accumulation in human tumors in clinical studies. However, the use of 18F-FAPI-74 for PET imaging of gastrointestinal cancer has not been systematically investigated. Herein, we investigated the diagnostic accuracy of 18F-FAPI-74 (18F-LNC1005) PET/CT in gastric, liver, and pancreatic cancers and compared the results with those of 18F-FDG PET/CT. Methods: This prospective study analyzed patients with confirmed gastric, liver, or pancreatic malignancies who underwent concurrent 18F-FDG and 18F-FAPI-74 PET/CT between June 2022 and December 2022. PET/CT findings were confirmed by histopathology or radiographic follow-up. 18F-FDG and 18F-FAPI-74 uptake and tumor-to-background ratios were compared using the Wilcoxon signed-rank test. The McNemar test was used to compare the diagnostic accuracy of the 2 scans. Results: Our cohort consisted of 112 patients: 49 with gastric cancer, 39 with liver cancer, and 24 with pancreatic cancer. Among them, 69 patients underwent PET/CT for initial staging and 43 for recurrence detection. Regarding lesion-based diagnostic accuracy, 18F-FAPI-74 PET/CT showed higher sensitivity than did 18F-FDG in the detection of primary tumors (gastric cancer, 88% [22/25] vs. 60% [15/25], P = 0.016; liver cancer, 100% [22/22] vs. 82% [18/22], P = 0.125; pancreatic cancer, 100% [22/22] vs. 86% [19/22], P = 0.250), local recurrence (92% [23/25] vs. 56% [14/25]; P = 0.021), involved lymph nodes (71% [41/58] vs. 40% [23/58]; P < 0.001), and bone and visceral metastases (98% [350/358] vs. 47% [168/358]; P < 0.001). Compared with 18F-FDG, 18F-FAPI-74 PET/CT upstaged 17 patients’ TNM staging among all treatment-naïve patients (17/69, 25%) and changed the clinical management of 4 patients (4/43, 9%) in whom recurrence or metastases were detected. Conclusion: 18F-FAPI-74 PET/CT is superior to 18F-FDG PET/CT in detecting primary tumors, local recurrence, lymph node involvement, and bone and visceral metastases in gastric, pancreatic, and liver cancers, with higher uptake in most primary and metastatic lesions.
According to the latest global cancer statistics, the incidence and mortality rates of gastrointestinal tumors, including gastric, hepatic, and pancreatic cancers, are increasing annually and account for a considerable proportion of other cancers, especially in eastern Asia (1). Therefore, early and accurate diagnosis is essential for their prognosis.
18F-FDG PET/CT has been introduced as a supplement to endoscopic ultrasonography, CT, and MRI for the diagnosis and staging of gastrointestinal tumors. Although 18F-FDG PET/CT is a valuable imaging technique for tumor staging and recurrence detection (2), some limitations exist for gastrointestinal malignancies. First, the relatively low sensitivity of 18F-FDG PET/CT for lymph node (LN), liver, and peritoneal metastases from gastrointestinal cancer affects the accuracy of tumor staging (3–5). Second, physiologic uptake in the gastrointestinal tract interferes with the detection of metastatic lesions, particularly for peritoneal metastases. Third, low 18F-FDG uptake is commonly observed in specific cancer types, including gastric signet ring cell carcinomas, mucinous adenocarcinomas, and hepatocellular carcinomas (HCCs) (6). Therefore, the development of a valid PET tracer is expected to improve the diagnosis of gastrointestinal cancer and contribute to individualized patient care.
Fibroblast activation protein (FAP), overexpressed by cancer-associated fibroblasts (CAFs) in more than 90% of epithelial tumors, is a membrane-anchored serine protease with dipeptidyl peptidase and endopeptidase activities (7). FAP inhibitor (FAPI), a quinoline-based ligand targeting FAP, can be used as a PET imaging tracer to visualize the tumor stroma (8,9). In previous studies, 68Ga-FAPIs have been shown to be superior to 18F-FDG in various cancers (8,10,11). 18F (half-life, 110 min) is an ideal radionuclide for PET imaging featuring a high positron yield of 97%, a low mean positron range of 0.5 mm, a longer half-life than 68Ga (110 vs. 68 min), and no simultaneous γ-ray emission (12). In addition, with the popularization of medical cyclotrons, 18F may be more suitable for promoting this imaging modality to benefit patients. A new PET imaging tracer, 18F-FAPI-74, has been successfully used to image patients with various cancers (13,14). However, the use of 18F-FAPI-74 in gastrointestinal tumors remains limited. In this single-center prospective study, we aimed to investigate the diagnostic accuracy of 18F-FAPI-74 (18F-LNC1005) PET/CT for gastric, liver, and pancreatic cancers and compare the results with those of 18F-FDG PET/CT.
MATERIALS AND METHODS
Patients
This study was approved by the institutional review board and was registered at ClinicalTrials.gov (NCT05430841). All participants provided written informed consent. With agreement from the oncologists and upon the determination of eligibility, patients were consecutively recruited for enrollment in the study from June 2022 and December 2022 at the First Affiliated Hospital of Xiamen University. Inclusion criteria were patients who had been diagnosed with or who were suspected of having gastric, liver, and pancreatic cancers before treatment; patients with suspected recurrence or metastases as a result of clinical findings or other diagnostic imaging, such as CT, MRI, or ultrasound, after treatment; agreement to perform both 18F-FDG and 18F-FAPI-74 PET/CT; and performance of the 2 scans with an interval of less than 1 wk. Exclusion criteria were patients with nonmalignant disease and pathologically proven nongastrointestinal tumors, patients who received any other treatment 4 wk before PET imaging, and inability or unwillingness of the research participant, parent, or legal representative to provide written informed consent.
Synthesis of 18F-FDG and 18F-FAPI-74
18F-FDG was manufactured using the coincidence 18F-FDG synthesizer (Tracer Lab FxFN; GE Healthcare) according to methods described by our laboratory (15). The precursor for 18F-FAPI-74 (18F-LNC1005) was obtained from Yantai Lannacheng Biotechnology Co. Ltd. for research and development purposes. Radiolabeling with aluminum-18F-fluoride was realized according to a previously reported protocol (14). Radiochemical purity and sterility tests were conducted in-house at an institutional radiochemistry facility. Radiochemical purities of 18F-FDG and 18F-FAPI-74 exceeded 95%, both meeting sterility standards and injection criteria for human administration.
Acquisition of PET/CT Images
For 18F-FDG PET/CT, patients fasted for more than 6 h, and blood glucose levels were monitored before they received an injection of 18F-FDG (5.55 MBq/kg). PET/CT images (1.5 min per bed position) were acquired with an uptake time of 63.0 ± 4.9 min (range, 50–75 min). For 18F-FAPI-74 PET/CT, imaging was performed (1.5 min per bed position) with an uptake time of 63.3 ± 4.0 min (range, 50–80 min) after receiving an injection of 210.9 ± 48.1 MBq (range, 129.5–333 MBq). Patients were scanned using a hybrid PET/CT scanner (Discovery MI; GE Healthcare) with the following CT parameters: tube voltage with a 110-kV current, 80 mA, and slice thickness of 3.75 mm. All obtained data were transferred to Advantage Workstation (version AW 4.7; GE Healthcare). Data reconstruction was performed using the Bayesian penalized likelihood reconstruction algorithm (Q.clear; GE Healthcare) with a penalization factor of 500. Reconstructed images were coregistered and displayed.
Imaging Review and Analysis
Advantage Workstation was used to analyze 18F-FDG and 18F-FAPI-74 PET/CT images. To prevent bias, images were separated into 2 sets according to the radiotracer and interpreted independently by 2 teams of 4 nuclear medicine physicians. 18F-FAPI-74 PET/CT images were analyzed by experts with more than 10 y of experience in nuclear oncology, whereas 18F-FDG PET/CT images were analyzed by experts with 5 y of experience in nuclear oncology. The reviews were conducted without considering other imaging modalities, such as CT, MRI, and pathologic results, as well as the information from the other PET/CT scan, and reviewers were provided with the patient’s medical history and indication for PET/CT. Differences in opinion were resolved by consensus.
Semiquantitative analysis was performed by manually drawing regions of interest around tumor lesions on transaxial images. For 18F-FAPI-74 and 18F-FDG, lesions were considered positive in cases with nonphysiologic foci of increased radiotracer uptake on PET images. Lesions were considered negative in cases with no pathologic uptake on PET images. SUVmax was automatically calculated to quantify tracer uptake in primary tumors, involved LNs, and bone and visceral metastases. The tumor-to-background ratio (TBR) was computed using the following formula: TBR = SUVmax-t / SUVmean-b, where SUVmax-t refers to the SUVmax of the primary or metastatic tumors and SUVmean-b refers to the SUVmean of the healthy background tissue surrounding each lesion.
Reference Standards
Surgery or biopsy results were used as reference standards for the final diagnosis. Radical excision and LN dissection were performed in 14 patients, and biopsies for suggestive lesions were performed in 84 patients (1 biopsy per patient). To evaluate the diagnostic accuracy of PET/CT for LN staging, images were reviewed by the surgeon to identify corresponding resected LNs in the PET images. Serial tumor marker follow-up (minimum of 6 mo), clinical examination results, and standard-of-care imaging (including ultrasound, contrast-enhanced CT, bone scan, or MRI) were used as reference standards to validate PET/CT findings when tissue biopsy was not applicable. Lesions were considered malignant if they met any of the following follow-up criteria: typical malignant features confirmed by multimodal medical imaging, significant progression on follow-up imaging, and significant posttreatment tumor size decrease. The minimum follow-up time was 6 mo.
Statistical Analysis
Statistical analyses were performed using SPSS software (version 26.0; IBM Inc.). The number of positive lesions shown using 18F-FAPI-74 and 18F-FDG PET/CT was compared using the McNemar test, and the Wilcoxon signed-rank test was used to assess differences in SUVmax and TBR between the 2 scans. The McNemar test was used to evaluate the diagnostic accuracy of 18F-FAPI-74 and 18F-FDG PET/CT. The mean and SD were calculated for normally distributed measurements, and the median and range were calculated for nonnormal measurements. The 95% CIs for the diagnostic accuracy data were calculated using the Wilson score method. A statistically significant difference between the comparison groups was indicated when a 2-tailed P value of less than 0.05 was considered. To evaluate the correlation between continuous variables with nonnormal distribution, the Spearman rank correlation coefficient was calculated. A P value of less than 0.05 (2-tailed) was used as the significance level for all statistical analyses.
RESULTS
Patient Characteristics
The study cohort consisted of 112 patients (76 men, 36 women): 49 with gastric cancer (31/49 men; median age, 62 y), 39 with liver cancer (31/39 men; median age, 62 y), and 24 with pancreatic cancer (14/24 men; median age, 60.5 y). Sixty-nine patients (62%) underwent PET/CT for initial tumor staging, and 43 patients (38%) underwent PET/CT to detect tumor recurrence or metastasis. 18F-FDG and 18F-FAPI-74 PET/CT scans were performed with a median interval of 2 d between examinations (range, 1–6 d). The patients’ clinical characteristics are listed in Table 1, and the research flowchart is shown in Figure 1.
Patient Characteristics
Flow diagram shows patient selection details. GI = gastrointestinal.
All patients tolerated the examination well. No drug-related pharmacologic effects or physiologic responses occurred. All observed parameters (e.g., blood pressure, heart rate, and body temperature) remained normal and unchanged during and after the examination. No patients reported subjective symptoms. Compared with 18F-FDG, 18F-FAPI-74 PET/CT showed excellent image contrast with low background activity throughout the body. Representative images from patients with gastric, liver, and pancreatic cancers are shown in Figure 2.
Nine representative patients who underwent 18F-FDG and 18F-FAPI-74 PET imaging. 18F-FAPI-74 PET outperforms 18F-FDG PET in detecting primary tumors (patients 11, 39, 50, 58, 79, and 101; solid black arrows), local recurrences (patient 4; blue arrows), abdomen LN metastases (patients 4 and 50; green arrows), intrahepatic metastases (patient 50; red arrows), bone metastases (patient 85; arrowheads), and peritoneal metastases (patients 4, 11, 85, 97, and 101; dotted arrows).
Comparison of 18F-FAPI-74 and 18F-FDG Uptake in Gastric Cancer
Patients with gastric cancer (n = 49) were diagnosed with adenocarcinoma (n = 44) and signet ring cell carcinoma (n = 5). Among 25 patients for primary staging, 18F-FAPI-74 PET/CT visualized most primary tumors (22/25, 88%), whereas 18F-FDG PET/CT missed 9 primary tumors (16/25, 64%; Supplemental Table 1 [supplemental materials are available at http://jnm.snmjournals.org]). A greater number of suggestive LNs were visualized using 18F-FAPI-74 PET/CT than using 18F-FDG PET/CT (88% [104/118] vs. 55% [65/118]; P < 0.001), with higher radiotracer uptake (median SUVmax, 6.2 vs. 2.9; P < 0.001) and TBR in abdominal regions (median, 4.5 vs. 1.7; P < 0.001) and supradiaphragmatic regions (median, 4.4 vs. 2.6; P = 0.010). Regarding bone and visceral metastases, 18F-FAPI-74 PET/CT demonstrated a greater number of peritoneal metastases than did 18F-FDG (100% [145/145] vs. 64% [38/145]; P < 0.001), with higher radiotracer uptake (median SUVmax, 6.9 vs. 2.5; P < 0.001) and TBR (median, 5.8. vs. 1.4; P < 0.001). Compared with 18F-FDG, 18F-FAPI-74 PET/CT showed similar lesion detectability and tracer uptake in local recurrence and bone metastases (Supplemental Table 1). Representative cases demonstrating the superiority of 18F-FAPI-74 over 18F-FDG in gastric cancer are shown in Figure 3.
Representative case of gastric cancer detected using 18F-FAPI-74 PET/CT but missed using 18F-FDG PET/CT. (A) 38-y-old woman with pathologically confirmed gastric signet ring cell carcinoma underwent initial staging using 18F-FDG PET/CT, which shows involved periintestinal LNs (arrows) only in CT image. (B) 18F-FAPI-74 PET/CT shows intense radiotracer uptake in primary lesion (both black and white arrows) and periintestinal LN (green arrows). Subsequent percutaneous biopsy in this LN confirmed metastatic disease (focal activity above right kidney indicates physiologic gallbladder uptake).
Comparison of 18F-FAPI-74 and 18F-FDG Uptake in Liver Cancer
Patients with liver cancer (n = 39) were diagnosed with HCC (n = 31) and intrahepatic cholangiocarcinoma (n = 8). 18F-FAPI-74 PET/CT visualized all primary tumors (22/22, 100%), whereas 18F-FDG PET/CT missed 4 primary tumors (18/22, 82%). 18F-FAPI-74 PET/CT showed superiority over 18F-FDG in detecting intrahepatic metastases (90% [46/51] vs. 53% [27/51]; P < 0.001), peritoneal metastases (100% [8/8] vs. 25% [2/8]; P = 0.031), local recurrence (92% [12/13] vs. 38% [5/13]; P = 0.016), and suggestive LNs (97% [36/37] vs. 43% [16/37]; P < 0.001), with significantly higher tracer uptake and TBR in most lesions (Supplemental Table 1). Representative cases of liver cancer are presented in Figure 4. 18F-FDG showed more metastatic lesions than did 18F-FAPI-74 in 2 patients with HCC. Regarding the 2 patients, 13 lesions (4 LN metastases and 9 intrahepatic metastases) showed low to mild uptake on 18F-FAPI-74 PET/CT. The patient with more lesions seen on 18F-FDG PET/CT is presented in Figure 5.
Representative case of metastatic liver cancer (intrahepatic cholangiocarcinoma). (A) 69-y-old woman with pathologically confirmed intrahepatic cholangiocarcinoma underwent initial staging using 18F-FDG PET/CT, which shows low to moderate radiouptake in primary tumor (arrows). (B) 18F-FAPI-74 PET/CT shows intense radiotracer uptake in primary lesion (arrows) and more intrahepatic metastases with intense uptake.
(A) 46-y-old man underwent 18F-FDG PET/CT for recurrence detection with subsequent biopsy-confirmed abdominal LN metastases. It shows high radiotracer uptake in abdominal LN metastases (arrows). 18F-FAPI-74 PET/CT shows lower radiotracer uptake and fewer abdominal LN metastases. Focal activity in liver indicates physiologic uptake in gallbladder. (B) 62-y-old man with pathologically confirmed liver cancer underwent initial staging with 18F-FDG PET/CT, which shows more intrahepatic metastases with higher tracer uptake (arrows) than did 18F-FAPI-74 PET/CT.
Comparison of 18F-FAPI-74 and 18F-FDG Uptake in Pancreatic Cancer
Patients with pancreatic cancer (n = 24) were diagnosed with pancreatic duct adenocarcinoma (n = 21), squamous cell carcinoma (n = 1), mucinous adenocarcinoma (n = 1), and undifferentiated carcinoma (n = 1). Positive 18F-FAPI-74 and 18F-FDG uptake (median SUVmax, 16.5 vs. 7.3) in primary tumors was observed in 22 and 19 patients, respectively. 18F-FAPI-74 PET/CT showed more metastatic lesions than did 18F-FDG, especially in liver metastases (100% [51/51] vs. 55% [28/51]; P < 0.001) and peritoneal metastases (100% [41/41] vs. 44% [18/41]; P < 0.001). In addition, greater numbers of suggestive LNs (97% [36/37] vs. 41% [15/37]; P < 0.001) were visualized using 18F-FAPI-74 PET/CT, with higher radiotracer uptake in abdominal regions (median SUVmax, 9.6 vs. 2.5; P < 0.001) and supradiaphragmatic regions (median SUVmax, 11.6 vs. 3.8; P = 0.028). In the semiquantitative analysis, SUVmax with 18F-FAPI-74 was higher than with 18F-FDG in primary tumors (median SUVmax, 16.5 vs. 7.3; P < 0.001), liver metastases (median SUVmax, 7.5 vs. 4.0; P < 0.001), peritoneal metastases (median SUVmax, 8.2 vs. 2.8; P < 0.001), and bone metastases (median SUVmax, 5.3 vs. 3.1; P = 0.009). Representative cases of pancreatic cancer are shown in Figure 6.
Representative case of widespread peritoneal metastatic pancreatic cancer (pancreatic ductal adenocarcinoma). (A) 55-y-old woman with biopsy-confirmed bone metastases underwent 18F-FDG PET/CT for detection of primary tumor; low uptake is shown in primary lesion (solid arrows), peritoneal metastases (dotted arrows), and bone metastases (arrowheads). 18F-FAPI-74 PET/CT shows intense uptake in these lesions. (B) In addition, 18F-FAPI-74 PET/CT shows larger extent of peritoneal metastases than does 18F-FDG.
Diagnostic Accuracy of 18F-FAPI-74 and 18F-FDG PET/CT
Primary tumors from 69 patients were confirmed by biopsy (n = 57) and surgery (n = 12). 18F-FAPI-74 PET/CT had significantly higher sensitivity than did 18F-FDG PET/CT for primary tumor detection (96% [66/69] vs. 75% [52/69]; P < 0.001), especially for diagnosis of gastric cancer (88% [22/25] vs. 60% [15/25]; P = 0.016). We were unable to obtain false-positive PET data for primary tumors because of restrictions on the study’s inclusion criteria. Therefore, the specificity and accuracy of 18F-FAPI-74 and 18F-FDG PET/CT for primary tumor assessment could not be established.
To assess the diagnostic accuracy of local recurrence, 43 patients were evaluated by radiographic follow-up (n = 14), biopsy (n = 27), and surgery (n = 2). Local recurrence was confirmed in 25 patients. Lesion-based sensitivity, specificity, and accuracy for 18F-FAPI-74 PET/CT were 92% (23/25), 94% (17/18), and 93% (40/43) and for 18F-FDG PET/CT were 56% (14/25), 89% (16/18), and 70% (30/43), respectively. Thus, 18F-FAPI-74 PET/CT had significantly greater sensitivity (92% [23/25] vs. 56% [14/25]; P = 0.021) and accuracy (93% [40/43] vs. 70% [30/43]; P = 0.039) than did 18F-FDG for the diagnosis of local recurrence (Table 2).
Diagnostic Accuracy of 18F-FAPI-74 and 18F-FDG PET/CT for Detection of Primary and Metastatic Tumors Using Lesion-Based Analysis
To evaluate diagnostic performance for LN metastases, 392 LNs in 14 patients were evaluated by histology (extended LN dissection was performed, along with radical excision). Among these, 58 metastatic LNs were confirmed in 8 patients. Nodular-based sensitivity, specificity, and accuracy were 71% (41/58), 95% (318/334), and 92% (359/392) for 18F-FAPI-74 PET/CT and 40% (23/58), 96% (321/334), and 88% (344/392) for 18F-FDG PET/CT, respectively. Thus, 18F-FAPI-74 PET/CT had significantly greater sensitivity (71% [41/58] vs. 40% [23/58]; P < 0.001) and accuracy (92% [359/392] vs. 88% [344/392]; P = 0.001) than did 18F-FDG PET/CT for the diagnosis of LN metastases (Table 2).
In all pathologically proven LNs, 18F-FDG–derived SUVmax in metastatic lesions demonstrated a significant correlation with LN size (ratio SUVmax [correlation coefficient], 0.440; P = 0.036). However, SUVmax derived from 18F-FAPI-74 PET/CT did not exhibit a significant association with LN size (P = 0.219). In the subgroup analysis specifically focused on gastric cancer, both SUVmax (ratio SUVmax [correlation coefficient], 0.729; P < 0.001) and TBR (ratio TBR [correlation coefficient], 0.588; P = 0.001) derived from 18F-FAPI-74 PET/CT were found to be significantly correlated with LN size. 18F-FDG PET/CT missed 35 LNs with an average size of 7.3 mm (range, 3–14 mm), whereas 18F-FAPI-74 PET/CT missed 17 LNs with an average size of 6.5 mm (range, 3–13 mm). Therefore, 18F-FAPI-74 PET/CT may have lower sensitivity in the detection of LNs that are less than 7 mm.
To assess the diagnostic accuracy of bone and visceral metastases, 432 suspected metastatic lesions from 89 patients were demonstrated using various imaging modalities. The lesions’ evaluation entailed multimodal imaging findings (n = 21), radiographic follow-up (n = 264), and biopsy (n = 147). Among these, 358 lesions were confirmed as metastatic in 62 patients, with 144 lesions identified by percutaneous biopsy, 21 confirmed using multimodal imaging, and 193 validated through radiographic follow-up. 18F-FAPI-74 PET/CT had significantly greater sensitivity (98% [350/358] vs. 47% [168/358]; P < 0.001) than did 18F-FDG PET/CT for diagnosis of bone and visceral metastases, especially for detection of peritoneal metastases (100% [194/194] vs. 30% [58/194]; P < 0.001) and liver metastases (95% [97/102] vs. 54% [55/102]; P < 0.001).
Changes in Initial Assessment and Recurrence Detection After 18F-FAPI-74 PET/CT
Among all treatment-naïve patients, 18F-FAPI-74 PET/CT led to upstaging of the clinical TNM stage in 17 patients (25%, 17/69) compared with the stage resulting from 18F-FDG PET/CT (TNM stage upstaged in 7/25 gastric cancer, 6/22 liver cancer, and 4/22 pancreatic cancer patients). Of 69 treatment-naïve patients, 18F-FAPI-74 PET/CT showed superior primary tumor detection in 10 patients (visualized using 18F-FAPI-74 but missed using 18F-FDG), demonstrated larger disease extent of peritoneal metastases in 9 patients, and visualized more LN, visceral, and bone metastases in 28 patients (Table 3). Regarding 43 patients with detected recurrence or metastases, true-positive rates (patient-based) for 18F-FDG and 18F-FAPI-74 PET/CT were 56% (14/25) and 92% (23/25), respectively, per patient. Of the 43 restaging patients, 18F-FAPI-74 PET/CT showed superior local recurrence detection in 4 patients (visualized using 18F-FAPI but missed using 18F-FDG), demonstrated larger disease extent of peritoneal metastases in 18 patients, and visualized a greater number of LN, visceral, and bone metastases in 14 patients. Finally, 18F-FAPI-74 PET/CT changed the clinical management of 4 patients with detected recurrence or metastases (9%, 4/43; Table 4).
Comparative Results for Initial Staging in Gastrointestinal Tumors Using Patient-Based Analysis
Comparative Results for Recurrence Detection in Gastrointestinal Tumors Using Patient-Based Analysis
DISCUSSION
This study investigated the clinical utility of 18F-FAPI-74 PET/CT for primary staging and recurrence or metastasis detection in patients with gastrointestinal cancer. We found that 18F-FAPI-74 is a promising PET tracer that can be used for imaging gastric, liver, and pancreatic cancers, with intense radiotracer uptake and clear tumor delineation in most primary and metastatic lesions. 18F-FAPI-74 PET/CT was better at detecting primary tumors and revealed more metastatic lesions than did 18F-FDG, leading to TNM stage upgrading.
Previous studies have shown the limitations of 18F-FDG PET/CT in detecting gastrointestinal malignancies, especially in mucinous adenocarcinoma, signet ring cell carcinoma (16), and HCC (10). Recent studies using 68Ga-FAPI (68Ga-FAPI-04/46) have demonstrated its superiority over 18F-FDG for the diagnosis of gastrointestinal cancer (17). However, the use of 18F-FAPI for diagnosing gastrointestinal cancer has not yet been systematically established. Consistent with a previous 68Ga-FAPI study (18), 18F-FAPI-74 PET/CT outperformed 18F-FDG PET/CT in detecting primary and metastatic lesions in gastrointestinal cancer, particularly for gastric cancer. In this study, 18F-FAPI-74 PET/CT detected primary tumors that were missed by 18F-FDG PET/CT in 28% of the patients, including 3 patients with gastric signet ring cell carcinomas. This emphasizes the advantage of using 18F-FAPI-74 PET/CT in this specific histologic type, especially for tumor diagnosis in early stages. However, unlike previous studies, we found no evidence of differences between 18F-FDG and 18F-FAPI-74 PET/CT for the detection of primary tumors in liver and pancreatic cancers. This may result from the advanced tumor stage of enrolled patients and the limited number of participants. Thus, TNM stages were upgraded after 18F-FAPI-74 PET/CT in 25% of patients with initial staging, and clinical management was modified in 9% of those with recurrence or progressive disease.
Nodal staging of patients with gastrointestinal cancer is crucial for its treatment and prognosis (19). However, 18F-FDG PET/CT has low to moderate sensitivity for LN staging in patients with gastrointestinal cancer, which may underestimate tumor stage and affect subsequent treatment regimens (20,21). Consistent with previous studies, the detection sensitivity of LNs shown using 18F-FAPI was higher than that of 18F-FDG in gastrointestinal cancer (79%–82% vs. 54%–59%, lesion-based sensitivity) (16,22). Thus, 18F-FAPI-74 PET/CT demonstrated nodal staging superior to that of 18F-FDG. This may overcome existing problems in accurately assessing LNs and aid in guiding surgical resection.
There is a significant correlation between both SUVmax and TBR derived from 18F-FAPI-74 and size of LNs (especially in gastric cancer), in agreement with previous research (23). Previous studies have demonstrated a close association between the expression of FAP and the depth of tumor invasion, TNM staging, LN metastasis, and distant metastasis (24). In addition, it has been reported that the degree of hypoxia is positively correlated with tumor size and is a strong inducer of FAP expression in CAFs (25,26). As a result, small LNs may exhibit mild hypoxia, resulting in lower uptake of 18F-FAPI-74 in these lesions. This hypothesis has the potential to elucidate the connection between both SUVmax and TBR derived from 18F-FAPI-74 and LN size.
Accurate diagnosis of peritoneal metastasis involvement is crucial in determining the resectability of a tumor and plays a significant role in the prognosis of patients with gastrointestinal disorders (27). However, 18F-FDG has a low detection rate of peritoneal metastasis, limiting its utility in staging and surgical planning (6,28). In our study, 18F-FAPI-74 PET/CT showed a higher detection rate and larger disease extent of peritoneal metastases than did 18F-FDG PET/CT. This may be explained by the strong tumor-induced fibrotic response that occurs when the tumor invades peritoneal tissue and by low tracer uptake within the digestive tract (14,29). Accurate evaluation of intrahepatic metastasis will affect the choice of treatment methods and median overall survival, especially in patients with hepatic oligometastasis (30). Although 18F-FAPI-74 PET/CT shows physiologic uptake in the biliary system (31), it does not affect observation of intrahepatic lesions because of low tracer uptake in normal liver parenchyma. Consistent with previous 18F-FAPI-based studies (85%–97% vs. 52%–80%, lesion-based sensitivity) (10,32), 18F-FAPI-74 PET/CT showed a higher detection rate of intrahepatic metastases than did 18F-FDG PET/CT.
We observed that 18F-FAPI-74 PET/CT resulted in changes in TNM staging in 28% and 18% of patients with gastric and pancreatic cancers, respectively. The impact on TNM staging was particularly pronounced in cases of progressive disease and tumor recurrence, which is in line with our previous findings (16,22). Moreover, changes of 22% and 17% in patient management were observed after the introduction of 18F-FAPI-74 PET/CT in gastric and pancreatic cancers, respectively, demonstrating the power of this imaging modality to alter patient care. 18F-FAPI PET/CT appears to be superior to other modalities in the detection of peritoneal metastases, which is a challenging indication with conventional imaging and 18F-FDG PET/CT because of their relatively low sensitivity. The data from this study suggest that PET/CT is a promising diagnostic approach for peritoneal metastases and that it is more sensitive than conventional imaging and 18F-FDG PET/CT. In summary, in patients with gastric and pancreatic cancers in which 18F-FDG PET/CT shows low performance, 18F-FAPI-74 PET/CT demonstrates great potential in terms of higher uptake and superior lesion detectability, which becomes particularly noticeable with bigger tumors, higher stages, and the signet ring cell carcinoma histologic type. Moreover, there are some encouraging preliminary results regarding therapy response evaluation by 18F-FAPI PET (33) to be further investigated.
MRI has high sensitivity in the detection of liver cancer, whereas a previous study demonstrated that 68Ga-FAPI PET/CT and MRI have similar sensitivity in the identification of primary liver tumors and intrahepatic metastatic lesions (10). In our study, 18F-FAPI-74 PET/CT was able to detect extrahepatic metastases in 36% of patients with liver cancer, emphasizing the potential added value of 18F-FAPI-74 PET/CT in the evaluation of extrahepatic lesions. Therefore, 18F-FAPI-74 PET/CT may play a complementary role to MRI in liver cancer, making it a promising candidate as the future modality of choice in tumor staging, molecular image–guided radiotherapy, and perhaps even molecular imaging–guided discrimination among liver tumors.
However, our result was inconsistent with a previous study (11), in which 18F-FDG detected more liver metastases (181 vs. 104; P < 0.001) and yielded higher tracer uptake (mean, 8.6 vs. 6.0; P = 0.001) than did 18F-FAPI in pancreatic cancer. This difference may be caused by the difference in the origin of CAFs, which leads to different phenotypes and functions of CAFs (34). This may cause uneven distribution of CAF markers such as FAP-α (35,36). A similar finding was observed in patients with HCC. In this study, a patient with recurrent HCC showed increased 18F-FDG but low 18F-FAPI-74 uptake in LN metastases, and histopathologic examination confirmed a well-differentiated HCC in the metastatic LNs. A similar observation was made in another patient with multiple intrahepatic metastases, and histopathology confirmed the presence of a moderately differentiated HCC. Previous research has indicated that most well- to moderately differentiated HCCs presented relatively low FAP expression, along with low 18F-FAPI uptake, particularly in comparison to poorly differentiated HCCs (37). In addition to the degree of tumor differentiation, the relatively low 18F-FAPI uptake in the HCC lesions may be related to the considerable intratumor spatial heterogeneity of the histomorphologic profile, along with the complexity of tumor biology.
In our study, there was an overlap in uptake intensity in the pancreatic primary tumor and in the tumor-induced obstructive pancreatitis of the pancreatic parenchyma (50% of patients showed intense uptake in the entire pancreas). Studies using 68Ga-FAPI PET/CT have observed a similar phenomenon, confirming increased pancreatic uptake in cases of inflammation-induced fibrosis (38). Delayed imaging has been previously investigated as a solution to distinguish between fibrosis and tumors, with positive results observed in some patients (22). However, larger sample sizes are needed to verify the feasibility of using the delayed PET scans.
Although 18F-FAPI-74 PET/CT showed higher tracer uptake and sensitivity than did 18F-FDG PET/CT for most bone and visceral metastases, it was found to yield more false-positive lesions caused by nonspecific fibrosis. In our study, false-positive uptake of 18F-FAPI was observed in bone (periarthritis, fibroosseous lesions, degenerative osteophyte, ischemic necrosis, and fracture), lung (pulmonary tuberculosis), thyroid (adenoma), uterus (uterine fibroids), and breast (mammary hyperplasia; Supplemental Table 2). It is inappropriate to characterize a lesion as benign or malignant referencing solely the 18F-FAPI uptake. The comprehensive combination of other imaging findings and clinical information is needed.
In contrast to previous clinical investigations on 18F-FAPI, our study brings forth some important nuances and enhancements. We presented the comparative results and a subgroup analysis of the diagnostic accuracy of 18F-FAPI-74 versus 18F-FDG PET/CT in gastric, pancreatic, and liver cancers, which were not extensively covered in previous work. In addition, we calculated and compared the SUVmax and TBRs derived from these 2 PET scans, providing more depth to our results. Our study on 18F-FAPI-74 PET/CT in gastrointestinal cancer involves a larger patient population than that of a recent study conducted by Watabe et al. (39), encompassing 112 as opposed to 31 patients. Moreover, pathologic evaluation via surgery or biopsy was performed in 88% (98/112) of the patients in this study, adding robustness to the 18F-FAPI-74 PET/CT findings. We also provide observations on the diagnostic accuracy for LN metastases.
Our study has several limitations. First, this was a single-center study, which may have caused selection bias. Second, histopathologic results were unavailable for a subset of lesions, because not all PET-positive findings were biopsied. Third, more than half of the enrolled patients (55%) in this study were at an advanced TNM stage (III and IV). The role of 18F-FAPI-74 PET/CT in detecting early disease requires further investigation.
CONCLUSION
Our results suggest that the diagnostic accuracy of 18F-FAPI-74 in gastric, pancreatic, and liver cancers was higher than that of 18F-FDG. However, in terms of specificity, 18F-FAPI-74 did not exhibit a significant advantage over 18F-FDG. This nuanced understanding of the results positions our work as a contribution to ongoing research in this field.
Compared with 18F-FDG, 18F-FAPI-74 PET/CT requires no fasting or blood glucose level monitoring, resulting in a shorter preparation time and improved patient flow. In addition, the physiologic uptake of 18F-FAPI-74 PET/CT was lower than that of 18F-FDG PET/CT in the gastrointestinal tract, enhancing image contrast. Therefore, in addition to higher lesion detectability, 18F-FAPI-74 PET/CT offers greater potential value during the examination process and during image acquisition, ultimately benefiting patients.
KEY POINTS
QUESTION: Is 18F-FAPI-74 PET/CT imaging useful for the evaluation of gastrointestinal cancer?
PERTINENT FINDINGS: This prospective study analyzed patients with confirmed gastric, liver, or pancreatic malignancies who underwent concurrent 18F-FDG and 18F-FAPI-74 PET/CT. The findings of the study showed that 18F-FAPI-74 PET/CT is superior to 18F-FDG in detecting primary tumors and metastatic lesions in gastric, pancreatic, and liver cancers, with higher radiotracer uptake in most primary and metastatic lesions.
IMPLICATIONS FOR PATIENT CARE: The use of 18F-FAPI-74 PET/CT is expected to improve the diagnosis of gastrointestinal cancer and contribute to individualized patient care.
DISCLOSURE
This work was funded by the National Natural Science Foundation of China (82071961 and 82272037), Key Scientific Research Program for Young Scholars in Fujian (2021ZQNZD016), Fujian Natural Science Foundation for Distinguished Young Scholars (2022D005), Fujian Research and Training Grants for Young and Middle-aged Leaders in Healthcare, The National University of Singapore (NUHSRO/2020/133/Startup/08, NUHSRO/2023/008/NUSMed/TCE/LOA, NUHSRO/2021/034/TRP/09/Nanomedicine), National Medical Research Council (MOH-001388-00, CG21APR1005), Singapore Ministry of Education (MOE-000387-00), and National Research Foundation (NRF-000352-00). Xiaoyuan Chen is co-founder of the Yantai LNC Biotechnology. No other potential conflicts of interest relevant to this article was reported.
Footnotes
↵* Contributed equally to this work.
Published online Oct. 26, 2023.
- © 2024 by the Society of Nuclear Medicine and Molecular Imaging.
REFERENCES
- Received for publication July 12, 2023.
- Revision received September 27, 2023.
