Comparison of [68 Ga]Ga-FAPI-04 and [18F]-FDG for the detection of primary and metastatic lesions in patients with gastric cancer: a bicentric retrospective study

The low sensitivity of [18F]-fluorodeoxyglucose ([18F]-FDG) for the diagnosis of gastric cancer limits its application. In this study, we aimed to investigate the potential advantage of [68 Ga]Ga-FAPI-04 over [18F]-FDG in the evaluation of gastric cancer. This was a bicentric retrospective analysis of a prospective parent study (clinical trial: HS-KY-2020–826 (Huashan Hospital) and DF-2020–102 (Shanghai East Hospital)). Thirty-eight patients with gastric cancer (31 with adenocarcinoma and 7 with signet ring cell carcinoma) were included in this study. All of the participants underwent [68 Ga]Ga-FAPI-04 and [18F]-FDG imaging by positron emission tomography (PET)/computed tomography (CT) or PET/magnetic resonance (MR). The scans were interpreted by two experienced nuclear medicine physicians, and the maximum standardized uptake value (SUVmax) was calculated. Histopathological findings obtained from biopsy or resected surgical specimens were used as a reference for the final diagnosis. For the detection of primary gastric cancer, the sensitivities of [68 Ga]Ga-FAPI-04 PET and [18F]-FDG PET were 100% (38/38) and 82% (31/38), respectively (P = 0.016). Four cases of adenocarcinoma and three cases of signet ring cell carcinoma were missed by [18F]-FDG PET. The mean SUVmax of [68 Ga]Ga-FAPI-04 in tumours greater than 4 cm (11.0 ± 4.5) was higher than that in tumours less than 4 cm (4.5 ± 3.2) (P = 0.0015). The mean SUVmax of [68 Ga]Ga-FAPI-04 was higher in T2–4 tumours (9.7 ± 4.4) than in T1 tumours (3.1 ± 1.5) (P = 0.0002). For the detection of metastatic lesions, the sensitivities of [68 Ga]Ga-FAPI-04 PET and [18F]-FDG PET in 10 patients with regional lymph node metastasis and distant metastasis were 6/10 and 5/10, respectively. In this selected cohort, [68 Ga]Ga-FAPI-04 PET had a superior detection rate than [18F]-FDG PET for primary gastric cancer. [68 Ga]Ga-FAPI-04 PET could provide better performance with regard to gastric cancer diagnosis and staging. Prospective clinical trials are warranted.


Introduction
Gastric cancer is one of the most common malignant tumours, and it is associated with a poor prognosis in East Asia and the USA [1,2]. Although the diagnosis and treatment of gastric cancer have improved, the prognostic outcome remains suboptimal. Early diagnosis, accurate staging and quantitative evaluation are important for the treatment, management, and prognosis of gastric cancer. In clinical practice, gastroscopy and imaging examinations are the main diagnostic methods for gastric cancer. Computed tomography (CT) and magnetic resonance imaging (MRI) have been used for the primary staging of gastric cancer. However, CT and MRI are anatomy-based imaging techniques that have been shown to be inadequate for staging, particularly with regard to the evaluation of lymph node involvement and distant metastasis [3].
Positron emission tomography (PET) using [ 18 F]-fluorodeoxyglucose ([ 18 F]-FDG) has been used for the diagnosis of various cancers, including gastric cancer [4,5]. However, reports indicate that up to 53% of primary gastric cancers are not avid for [ 18 F]-FDG; FDG uptake is strongly associated with the histological type and size of the tumour [6,7]. Furthermore, the low sensitivity for the detection of lymph node metastasis and peritoneal metastasis limits the use of [ 18 F]-FDG for determining the clinical stage of gastric cancer [7]. Signet ring cell carcinoma and mucinous carcinoma also show lower [ 18 F]-FDG uptake and sensitivity than conventional adenocarcinoma [8][9][10].
Currently, fibroblast activation protein (FAP), which is highly expressed in the cancer-associated fibroblasts (CAFs) of many malignant tumours, has been identified as a new target for tumour tracer development because CAFs are one of the most abundant stromal components in the tumour microenvironment [11]. Several tracers targeting FAP have been developed, among which [ 68 Ga]Ga-FAPI-04 is the most promising, with a high affinity for FAP and suitable kinetics [12][13][14]. [ 68 Ga]Ga-FAPI-04 PET has achieved good outcomes with regard to the diagnosis and staging of several tumour types [11,[15][16][17]. Moreover, previous studies have shown that FAP is overexpressed in the CAFs of gastric cancer and plays an important role in the invasion and migration of gastric carcinomas [18,19]. [ 68 Ga] Ga-FAPI-04 PET was used to image primary gastric carcinoma and metastatic lesions in one patient [20]. However, no studies have reported the potential advantage of [ 68 Ga] Ga-FAPI-04 with regard to diagnosing gastric cancer and the difference between [ 68 Ga]Ga-FAPI-04 and [ 18 F]-FDG.
The aim of this study was to assess the potential advantage of [ 68 Ga]Ga-FAPI-04 PET over [ 18 F]-FDG for the diagnosis of gastric cancer and to investigate the performance of [ 68 Ga]Ga-FAPI-04 with regard to the detection of lymph node metastasis in patients with gastric cancer confirmed by histopathology from biopsy or resected surgical specimens.  [ 18 F]-FDG and [ 68 Ga]Ga-FAPI-04 were prepared as described previously [13,15,21], and both tracers were synthesized in the Radiochemistry Facility of the PET Center, Huashan Hospital, Fudan University, in a GMP environment.

Radiopharmaceuticals and imaging protocols
PET/MR (uPMR790 TOF, United Imaging, China) or PET/CT (Biograph mCT, Siemens Healthineers, Germany; Ingenuity TF, Philips Healthcare, USA; uMI510, United Imaging, China) scanners were used in this study. Since different scanners were used in this study, standardized uptake value (SUV) measurements generated by different devices were normalized after data collection. A NEMA IEC body phantom (Data Spectrum Corporation, Durham, NC, USA) with 6 simulated lesion spheres (diameters: 10 mm, 13 mm, 17 mm, 22 mm, 28 mm, and 37 mm) was used for SUV normalization with 2, 4, 8, and 16 times the background activity (background activity concentration equal to 2 kBq/ ml). The CT template of the NEMA IEC body phantom was prepared to correct the attenuation of PET/MR. Images of the phantom were acquired in each scanner with routine protocols. Correlation coefficients for the SUV were obtained through this phantom study and used to standardize the SUV measurements with those obtained with the Siemens Biograph mCT.

Imaging analysis
To calculate the maximum SUV (SUV max ), circular regions of interest were drawn around the tumours on transaxial slices and automatically adapted to a three-dimensional VOI with Syngo.via software (Siemens Molecular Imaging, Hoffman Estates, Illinois, USA) at a 60% isocontour. For the definition of the background, 10-mm-diameter ROIs were drawn in the mediastinum (blood pool), and the SUV max was recorded. The gastric wall was chosen in an area with a thin stomach wall (CT or MRI) and uniform [ 68 Ga]Ga-FAPI-04 or [ 18 F]-FDG uptake. Circular ROIs with a diameter of 10 mm were placed in the chosen areas that were the most active (the highest uptake) and free from disease. Tumourto-background uptake ratios and tumour-to-normal gastric wall background ratios were calculated as the SUVmax of the hottest tumour lesion/SUVmax of respective background or normal gastric wall. The results of the [ 68 Ga]Ga-FAPI-04 PET and [ 18 F]-FDG PET scans were independently evaluated by 2 experienced nuclear medicine physicians (JZ and FH) who were blinded to the clinical data and pathologic findings. Any difference of opinion between these two physicians was resolved by a consensus. For primary tumours and metastatic lesions, positive uptake was identified as areas of focal increase compared to the surrounding normal tissue based on a visual assessment supported by the ratio of the suspicious lesion to the surrounding normal tissue [15,24].

Histopathology
The histopathological analysis of tissue obtained from biopsy or resected surgical specimens was used as a reference for the final diagnosis based on the pathology reports.

Immunohistochemistry of FAP expression
Immunohistochemical staining of FAP was performed on the tumour tissue obtained after surgical resection in one patient with adenocarcinoma. An antibody against FAP (ab207178, Abcam) was used.

Statistical analysis
The primary outcome of the study was the comparison of the sensitivity, specificity, accuracy, positive predictive value (PPV), and negative predictive value (NPV) for the detection of gastric cancer of [ 68 Ga]Ga-FAPI-04 PET and [ 18 F]-FDG PET, with histopathology used as the reference standard. The secondary outcome was the comparison of tumour SUVs and tumour/blood pool or normal gastric wall background ratios (TBRs). Statistical analysis was performed using Prism 7 (GraphPad Software, CA, USA) and STATA (version 15.1 StataCorp LLC) software. The comparisons of sensitivity, specificity, accuracy, PPV, and NPV for the detection of gastric cancer between [ 18 F]-FDG and [ 68 Ga] Ga-FAPI-04 were performed with McNemar's test. Categorical variables are described as frequencies and percentages. Continuous variables are described as the means ± standard deviations (SDs). The Mann-Whitney U test was used to compare the mean SUV max between different categorized groups and the TBRs between [ 68 Ga]Ga-FAPI-04 and [ 18 F]-FDG in the primary tumour. A P value less than 0.05 was considered statistically significant.

Patients
The clinical characteristics of the 38 patients are shown in Table 1 and Supplementary Table 1. Among the 38 patients with gastric cancer (29 men and 9 women; age 63.7 ± 15.3 years; age range, 25-86 years) with diagnosis was confirmed by pathological biopsy obtained during gastroscopy, 31 had adenocarcinoma, and 7 had signet ring cell carcinoma. After the PET evaluation, 24 patients (19 with adenocarcinoma and 5 with signet ring cell carcinoma) received surgical treatment, while the other patients received chemotherapy. Among the 24 patients treated with surgery, 10 (9 with adenocarcinoma and 1 with signet ring cell carcinoma) had pathologically confirmed regional lymph node metastasis after surgical resection. According to the depth of tumour invasion based on examination of the surgical specimen after resection, 12 patients were classified as having T1 disease (tumour invading the lamina propria, muscularis mucosae, or submucosa), 3 patients were classified as having T2 disease (tumour invading the muscularis propria), and 9 patients were classified as having T4 disease (tumour invading the visceral peritoneum or adjacent structures). According to the maximum tumour diameter, 17 patients had a small tumour (≤ 4 cm in diameter), and 7 patients had a large tumour (˃ 4 cm in diameter). According to the histologic grade, 1 patient had a well-differentiated tumour, 15 patients had a moderately differentiated tumour, and 8 patients had a poorly differentiated tumour (4 adenocarcinomas and 4 signet ring cell carcinomas).

Immunohistochemistry
One primary tumour sample obtained from a patient with gastric adenocarcinoma was assessed for FAP expression by immunohistochemistry. Stromal cells around the tumour had prominent FAP expression, while tumour cells had weak FAP expression (Fig. 3). [ 68 Ga]Ga-FAPI-04 PET showed high uptake (SUV max = 10.2) in the primary gastric tumour of this patient.

Discussion
Gastric cancer is one of the most common cancers and the third leading cause of cancer-related death worldwide [25]. Achieving an early and accurate diagnosis is extremely important for the treatment and outcome. [ 68 Ga]Ga-FAPI-04 PET has achieved good outcomes with regard to the diagnosis of various tumours [11,[15][16][17]. These include high expression across a wide range of cancer types, including several with typically low FDG avidity and low uptake in almost all normal tissues, where high physiological uptake can obscure primary or metastatic disease [26] (Fig. 3). Furthermore, in another patient with gastric adenocarcinoma, [ 68 Ga]Ga-FAPI-04 displayed focal uptake in the primary tumour, whereas [ 18 F]-FDG showed slightly and diffusely elevated uptake in the gastric wall (Fig. 4). As a result, [ 68 Ga]Ga-FAPI-04 could provide better information that could be used when making surgical decisions regarding gastric cancer. Moreover, the background activity of [ 68 Ga]Ga-FAPI-04 was low, resulting in a higher tumour-to-background contrast than [ 18 F]-FDG.
Previous studies have shown that the sensitivity of [ 18 F]-FDG PET for the detection of gastric cancer ranges from 47 to 96% [7]. One of the reasons for this disparity may be the different histological types of gastric cancer. [ 18 F]-FDG showed lower sensitivity and uptake in signet ring cell carcinoma and mucinous carcinoma than in conventional adenocarcinoma [8][9][10]. This is due to the relatively low expression level of glucose transporter 1 (GLUT-1) in signet ring cell and mucinous carcinomas [27]. These results are consistent with our finding that signet ring cell carcinoma had lower [ 18 F]-FDG uptake than adenocarcinoma, although the difference was not statistically significant. In our study, [ 68 Ga]Ga-FAPI-04 detected 3 cases of signet ring cell carcinoma that were not detected by [ 18 F]-FDG, resulting in a sensitivity of 100% (7/7) for the detection of signet ring cell carcinoma, outperforming [ 18 F]-FDG (57%, 4/7). Therefore, [ 68 Ga]Ga-FAPI-04 PET has an obvious advantage for the detection of signet ring cell carcinoma, especially when the  Tumour size is also an important factor influencing the detection rate of gastric cancer on PET scans. Nakajo et al. [28] examined the relationship between the SUV max of primary gastric cancers and the size of visible tumours and found that the [ 18 F]-FDG uptake of primary tumours was significantly associated with tumour size. In another study, small gastric cancers were reported to be difficult to detect with [ 18 F]-FDG PET [29]. Interestingly, for the detection of tumours less than 4 cm, [ 68 Ga]Ga-FAPI-04 PET also yielded a better sensitivity than [ 18 F]-FDG (100% vs. 71%, P = 0.062), although the difference was not significant. All 17 tumours ≤ 4 cm in size, including 5 with negative [ 18 F]-FDG uptake, were detected by [ 68 Ga]Ga-FAPI-04 PET in our study. The uptake of [ 68 Ga]Ga-FAPI-04 in small gastric cancers (≤ 4 cm in diameter) was lower than that in large gastric cancers (> 4 cm in diameter), and further confirmation of whether [ 68 Ga]Ga-FAPI-04 PET has advantages for the detection of small gastric cancers is needed. The depth of invasion in primary gastric cancer is an essential determinant of prognosis and therapy. The results from previous studies have shown that the SUV max of [ 18 F]-FDG does not correlate with the degree of infiltration [28,30,31]. In addition, other tracers used to detect gastric cancer, such as [ 18 F]-FLT, are not suitable for evaluating the degree of infiltration in gastric cancers [28]. In our study, the mean SUV max of [ 68 Ga]Ga-FAPI-04 in pathological T2-4 tumours was significantly higher than that in T1 tumours (9.7 ± 4.4 vs 3.1 ± 1.5, P = 0.0002). Thus, [ 68 Ga] Ga-FAPI-04 PET could provide insight into the degree of tumour invasion in gastric cancer.
In our patient-based analysis, the sensitivity of [ 18 F]-FDG PET for the detection of regional metastatic lymph nodes was 50% (5/10), which is in line with previous studies of the use of [ 18 F]-FDG PET for the detection of lymph node metastasis (mean sensitivity: 45%) [7]. The sensitivity of [ 68 Ga]Ga-FAPI-04 PET for the detection of regional metastatic lymph nodes was 60% (6/10). As shown in Fig. 5, lymph node metastasis at the lesser curvature of the stomach in one patient with moderately differentiated adenocarcinoma presented as elevated uptake of [ 68 Ga]Ga-FAPI-04 but negative uptake of [ 18 F]-FDG. Chen et al. [15] reported 12 cases of gastric cancer (4 signet ring cell carcinomas and 8 adenocarcinomas) and found that [ 68 Ga]Ga-FAPI-04 PET/CT was more sensitive than [ 18 F]-FDG PET/CT for the detection of lymph node metastases of gastric cancer, according to a lesion-based analysis.
In addition, in contrast to [ 18 F]-FDG PET, [ 68 Ga]Ga-FAPI-04 PET has an advantage for the detection of distant metastasis in gastric cancer. Although the pathological analysis of a biopsy is the gold standard for the diagnosis of metastases in gastric cancer, noninvasive imaging has become a standard modality for staging before treatment. In this study, distant lymph nodes in 3 patients showed high at the lesser curvature of the stomach (arrow). d-f [ 18 F]-FDG PET/ MR showed high uptake in the primary tumour (SUV max = 6.9) but no uptake in the regional lymph nodes [ 68 Ga]Ga-FAPI-04 uptake but negative [ 18 F]-FDG uptake. These distant lymph nodes, which included the posterior peritoneum lymph nodes and supraclavicular lymph nodes, are prevalent metastatic sites of gastric cancer. As shown in Fig. 6, one patient displayed discernible [ 68 Ga]Ga-FAPI-04 uptake in the peritoneum that was not detected by [ 18 F]-FDG PET. Distant metastasis has an important impact on treatment and prognosis. In gastric cancer, peritoneal metastasis is associated with a poor prognosis [32]. According to previous studies, the sensitivity of [ 18 F]-FDG PET for the detection of metastatic peritoneal disease is low [33,34]. Overall, [ 68 Ga]Ga-FAPI-04 PET may play an important role in the detection of metastases of gastric cancer with or without confirmation by pathological biopsy.
There were several limitations of our study. The small sample size limited the power of the analysis, and not all histological types of gastric cancer were analysed. Moreover, as there was no histological verification available for some cases of highly suspicious distant metastases, latent bias may be present. We also intended to investigate the advantage of PET/MR for the detection of gastric cancer metastases. However, the small sample size limited the power of the analysis. Twenty-nine patients received both [ 18 F]-FDG and [ 68 Ga]Ga-FAPI-04 PET/MR, and nine patients received PET/CT. This investigation was not feasible due to the lack of contrast-enhanced MRI data in this study. Therefore, the usefulness of [ 68 Ga]Ga-FAPI-04 PET for the detection of distant metastases of gastric cancer needs to be further investigated in a larger cohort.

Conclusion
In summary, [ 68 Ga]Ga-FAPI-04 PET is superior to [ 18 F]-FDG PET for the detection of primary gastric cancers. For tumours less than 4 cm in size, [ 68 Ga]Ga-FAPI-04 PET is a more sensitive method of detection. The SUV max of [ 68 Ga]Ga-FAPI-04 was higher in T2-4 gastric cancers than in T1 gastric cancers, indicating its potential role in the assessment of the degree of gastric cancer invasion. [ 68 Ga]Ga-FAPI-04 PET could be a better method of diagnosing and staging gastric cancer. Nonetheless, it is still necessary to evaluate the N (metastatic lymph nodes) and M (distant metastases) staging value of [ 68 Ga] Ga-FAPI-04 in gastric cancer in more cases of signet ring cell carcinoma and in a larger cohort.