Utility of [68Ga]FAPI-04 and [18F]FDG dual-tracer PET/CT in the initial evaluation of gastric cancer

Objectives We aimed to investigate the role of [68Ga]FAPI-04 and [18F]FDG dual-tracer PET/CT for the initial assessment of gastric cancer and to explore the factors associated with their uptake. Methods This study enrolled 62 patients with histopathologically confirmed gastric cancer. We compared the diagnostic performance of [68Ga]FAPI-04, [18F]FDG, and combined dual-tracer PET/CT. The standardized uptake value (SUV) and tumor-to-background ratio (TBR) were also measured, and the factors that influence tracer uptake were analyzed. Results [68Ga]FAPI-04 PET/CT detected more primary lesions (90.3% vs 77.4%, p = 0.008) and peritoneal metastases (91.7% vs 41.7%, p = 0.031) and demonstrated higher SUVmax and TBR values (p < 0.001) of primary lesions compared to [18F]FDG PET/CT. Dual-tracer PET/CT significantly improved the diagnostic sensitivity for the detection of distant metastases, compared with stand-alone [18F]FDG (97.1% vs 73.5%, p = 0.008) or [68Ga]FAPI-04 (97.1% vs 76.5%, p = 0.016) PET/CT. Subsequently, treatment strategies were changed in nine patients following [68Ga]FAPI-04 and [18F]FDG dual-tracer PET/CT. Nevertheless, [68Ga]FAPI-04 uptake was primarily influenced by the size and invasion depth of the tumor. Both [68Ga]FAPI-04 and [18F]FDG PET/CT showed limited sensitivity for detecting early gastric cancer (EGC) (37.5% vs 25.0%, p > 0.05). Conclusions In this initial study, [68Ga]FAPI-04 and [18F]FDG dual-tracer PET/CT were complementary and improved sensitivity for the detection of distant metastases pre-treatment in gastric cancer and could improve treatment stratification in the future. [68Ga]FAPI-04 had limited efficacy in detecting EGC. Key Points • [68Ga]FAPI-04 and [18F]FDG dual-tracer PET/CT are complementary to each other for improving diagnostic sensitivity in the initial evaluation of distant metastases from gastric cancer. • [68Ga]FAPI-04 PET/CT showed limited sensitivity in detecting EGC. • Need for further validation in a larger multi-centre prospective study. Supplementary Information The online version contains supplementary material available at 10.1007/s00330-022-09321-1.


Introduction
Gastric cancer ranks as the fifth and fourth in cancer incidence and cancer-related deaths globally, respectively [1]. Patients are frequently diagnosed with advanced gastric cancer (AGC) due to the insidious early symptoms. Treatment of gastric cancer is currently based on multidisciplinary management, including surgery, systemic chemotherapy, radiotherapy, immunotherapy, and targeted therapy [2]. Accurate evaluation of disease extent is paramount for selecting the appropriate treatment method. [ 18 F]FDG PET/CT imaging for gastric cancer can sometimes be suboptimal, particularly in individuals with non-intestinal-type gastric cancers or individuals with signet ring cell carcinomas (SRCC) or mucinous adenocarcinomas (MAC) [3,4].
Fibroblast activation protein (FAP) is commonly overexpressed in cancer-associated fibroblasts, which are known to be the primary components of stromal cells that contribute up to 90% of the tumor mass [5,6]. Recently, 68 Ga-labeled quinoline-based FAP inhibitor (FAPI) has allowed for the imaging of tumor stroma by targeting FAP, among which [ 68 Ga]FAPI-04 has exhibited favorable tumor-to-background ratio (TBR) and kinetics [7,8]. [ 68 Ga]FAPI-04 PET/CT reportedly outperformed [ 18 F]FDG PET/CT, especially in cancers of unknown primary origin, breast cancer, and several digestive system tumors, including gastric cancer; thus, it may be an alternative to [ 18 F]FDG PET/CT in the detection of these tumors [9,10]. However, the number of SRCC patients enrolled in previous studies on gastric cancer was limited. Additionally, elevated FAP expression has also been observed during wound healing and matrix remodeling, including chronic inflammation, atherosclerosis, and liver and lung fibrosis [6]. Whether [ 68 Ga]FAPI-04 PET/CT could replace or supplement [ 18 F]FDG PET/CT in the initial evaluation of gastric cancer needs to be further investigated.
Based on the comparison of [ 68 Ga]FAPI-04 and [ 18 F]FDG PET/CT in a larger cohort, our research further explored the role of combined dual-tracer PET/CT in the initial assessment of gastric cancer and analyzed the clinicopathological factors that influence tracer uptake.

Patients
The Ruijin Hospital Ethics Committee of Shanghai Jiao Tong University School of Medicine approved this prospective clinical study (2020 CER No.172). This study enrolled 62 patients pathologically diagnosed with gastric cancer by gastroscopy biopsy for initial staging. All patients signed written informed consent prior to PET/CT imaging. Subsequently, [ 68 Ga]FAPI-04 PET/CT and [ 18 F]FDG PET/CT imaging were carried out before treatment. Following comprehensive imaging results, clinical evaluations, and patients' willingness, 20 patients underwent primary surgery, 25 patients underwent chemotherapy followed by surgery (including 19 patients who received neoadjuvant chemotherapy and 6 patients who received conversion therapy), and 17 patients underwent antitumor treatment without surgery. Figure 1 shows the study flowchart. Table 1 summarizes the clinicopathological characteristics of the 62 patients. TNM staging was classified according to the eighth edition of the American Joint Committee on Cancer TNM system.

Radiopharmaceuticals
[ 68 Ga]FAPI-04 was prepared following the prior approach [7]. Briefly, radioactive gallium ( 68 Ga) was eluted from a 68 Ge-/ 68 Ga generator and added to a reactor vial containing 20 ug of DOTA-FAPI-04 (CSBio), then mixed with NaOAc (1 mol/L, 1 mL), which resulted in a pH of 4. The mixture was further reacted at 100°C for 10 minutes using an automatic synthesis module (Trasis contraindications, 20 mg of hyoscine butylbromide was injected intravenously before scanning, followed by drinking approximately 500 mL of water to achieve gastric distension [11,12]. Diagnostic non-contrast-enhanced CT (non-CECT) scans were performed using the CARE Dose 4D technique (120 kV, automatic mA-modulation). PET images were obtained in 3D mode and reconstructed in a 440 × 440 matrix size (iteration: 4, subset: 5) using the TrueX + TOF (ultraHD-PET) method. The interval between the two PET/CT scans was within 9 days.

Image analysis
Two experienced nuclear medicine physicians (G.R. and H.X.Y., with 12 and 5 years of experience in nuclear oncology, respectively) independently analyzed the [ 68 Ga]FAPI-04 PET/ CT and [ 18 F]FDG PET/CT images. A positive dual-tracer PET/ CT was defined as [ 68 Ga]FAPI-04 PET/CT-positive or [ 18 F]FDG PET/CT-positive. For semiquantitative analysis, a spherical region of interest was delineated around the tumor lesions, which was automatically adjusted to a 3D volume of interest (VOI) at a 60% isocontour using syngo.via software (Siemens Healthineers), and the maximum standardized uptake value (SUV max ) was recorded. Additionally, a 10-mm diameter VOI was placed over the non-lesional gastric wall to obtain the SUV max of the normal gastric wall background, a 10-mm diameter VOI was drawn on the descending aorta to acquire the mean standardized uptake value (SUV mean ) of the mediastinal blood pool background, and a 20-mm diameter VOI was set on the non-lesional right liver lobe to obtain the SUV mean of liver blood pool background [13]. The TBR was displayed as TBR-G, TBR-A, and TBR-L, which were calculated by dividing the SUV max of the gastric tumors with the background of the normal gastric wall, mediastinal blood pool, and liver blood pool, respectively. Histopathological findings, laparoscopic exploration, and contemporaneous and follow-up imaging were the reference standards for the final diagnosis. Progression of metastatic lesions or reduction in the size/number of lesions after chemotherapy on follow-up imaging was considered a malignant feature [14].

Statistical analysis
IBM SPSS Statistics 26.0 was used for statistical analysis. Continuous variables were presented as medians and interquartile range (IQR), whereas categorical variables were presented as numbers and percentages. The diagnostic performance, including sensitivity, specificity, accuracy, positive predictive value, and negative predictive value, was analyzed. The comparison of SUV max or TBR between [ 68 Ga]FAPI-04 and [ 18 F]FDG PET/CT was conducted using the Wilcoxon signed-rank test. The Mann-Whitney U test was used to compare SUV max within According to the clinical staging, neoadjuvant chemotherapy or conversion therapy was performed in patients with advanced gastric cancer. groups. The comparison of diagnostic performance between and within groups was performed using the McNemar test, χ 2 test, or Fisher's exact test. All statistical tests were two-sided, and a value of p < 0.05 was considered statistically significant.  Performance of [ 68 Ga]FAPI-04, [ 18 F]FDG, and dualtracer PET/CT in diagnosing regional lymph node metastases Table 3 summarizes the performance of [ 68 Ga]FAPI-04, [ 18 F]FDG, and dual-tracer PET/CT in diagnosing regional lymph node metastases. A patient-based analysis was conducted in 20 patients who underwent surgery without preoperative antitumor treatment. Of these, 11 (55.0%) were pathologically confirmed as having regional nodal metastases.

Results
[ 68 Ga]FAPI-04 and [ 18 F]FDG PET/CT missed to detect nodal metastases in four (36.4%) and five (45.5%) patients, respectively, whereas the false-negative patients missed by each were slightly different. In detecting regional nodal metastases, the sensitivity, specificity, and accuracy of [ 68 Ga]FAPI-04     Table 5 shows the respective results. Both the median SUV max of [ 68 Ga]FAPI-04 and that of [ 18 F]FDG were markedly higher in AGC compared to EGC and were also higher in tumors > 3 cm than in tumors ≤ 3 cm. Additionally, the median SUV max of [ 18 F]FDG was evidently lower in the subgroup of PCC (including SRCC) than that of non-PCC and was also lower in the subgroup of the non-intestinal type than that of the intestinal type. In contrast, the median SUV max of [ 68 Ga]FAPI-04 did not differ significantly between the subgroups according to histological type, Lauren classification, or degree of differentiation.

Changes in TNM staging and treatment strategies following [ 68 Ga]FAPI-04 and [ 18 F]FDG PET/CT
Overall, 57 of 62 patients underwent concurrent CECT for preoperative staging. Supplementary Fig. 1 shows the staging changes following PET/CT scans. In terms of N staging and compared with CECT, two patients were upstaged and    [14][15][16][17]. This may be attributed to the difference in stage distribution and tumor size of the enrolled patients as well as differences among the observers' interpretations based on visual assessments. Moreover, the sensitivity of dual-tracer PET/ CT in detecting primary lesions of gastric cancer was equivalent to that of [ 68 Ga]FAPI-04 and higher than that of [ 18 F]FDG.
In the diagnosis of regional nodal metastases of gastric cancer, our present patient-based analysis indicated that the sensitivity of [ 68 Ga]FAPI-04 PET/CT was not significantly different from that of [ 18 [18]. The primary reasons for the limited sensitivity of [ 68 Ga]FAPI-04 PET/CT in detecting regional nodal metastases in our study may be attributed to three factors: First, regional and distant nodal metastases were separately analyzed in our study. The diagnosis of regional nodal metastases was based on the postoperative pathology from lymph node dissection, which could potentially increase the number of false-negative lymph nodes compared with distant lymph node analysis. Second, the patients included in the regional lymph node analysis were at a relatively early stage of the disease, and the metastatic lymph nodes might be small and insidious. Additionally, the uptake of small perigastric lymph nodes might be obscured by the radioactive volume effect of the primary gastric tumor and stomach motility. Dual-tracer PET/CT did not significantly improve diagnostic performance in regional nodal metastases compared with either single-tracer PET/CT.
For the detection of distant metastases from gastric cancer, the sensitivity of [ 18 F]FDG PET/CT in our study was 73.5%, which was higher than that of the Multicenter Prospective Dutch Cohort Study (PLASTIC) that showed a sensitivity of only 33% [19]. The main reason for this discrepancy would be the different TNM stages of the enrolled patients: the PLASTIC study was restricted to those with locally advanced (≥ cT3 and/or N+, M0) and surgically resectable (< cT4b) gastric cancer after primary staging with CT, whereas advanced patients with distant metastases were also included in our study. Moreover, the lack of follow-up in most patients and a higher proportion of patients with peritoneal metastases in the PLASTIC study also contributed to this discrepancy. As [ 18 F]FDG PET/CT was sub-optimal in detecting peritoneal metastases of gastric cancer due to the physiological or inflammatory interference in the intestines and low avidity of [ 18 F]FDG in SRCC/MAC [20,21]. Our present work demonstrated that [ 68 Ga]FAPI-04 PET/CT was more sensitive than [ 18 F]FDG for detecting peritoneal seeding as it confirmed peritoneal metastases in six additional patients. This superiority was attributed to the lack of physiological accumulation of [ 68 Ga]FAPI-04 in the intestines, resulting in a low background uptake in the peritoneal cavity. Additionally, tumor lesions that exceed 2 mm require a supporting stroma, which can be greater in volume than the tumor cells themselves [22]. Therefore, [ 68 Ga]FAPI-04 may be more sensitive than [ 18 F]FDG even in small lesions, assuming there is sufficient FAP-expressing stroma. Our results were in line with the findings reported by previous studies [15,17,23]. However, both   [15]. In the diagnosis of ovarian metastases, [ 68 Ga]FAPI-04 detected one additional patient with PCC. However, as a hormone-responsive organ, the physiological uptake of both tracers in the ovaries of premenopausal women may potentially increase the uncertainty in the interpretation of ovarian lesions.
With respect to liver, lung, and bone metastases, [ 18 F]FDG PET reportedly performed well, with a sensitivity of 95.2% and a specificity of 100% [24]. 68 Ga-FAPI PET/CT was found to outperform [ 18 F]FDG PET/CT in detecting liver metastases from gastrointestinal cancer [25]. In our research, however, [ 18 F]FDG PET/CT recognized three additional liver metastases, which were all missed by [ 68 Ga]FAPI-04 PET/CT, whereas one of the liver metastases detected by [ 68 Ga]FAPI-04 PET/CT was a false-positive uptake. Figure 6 shows a typical case of our findings. A similar result was obtained by Zhang et al, who found that more liver metastases from pancreatic cancer were detected by [ 18  In the subgroup analysis, large tumor size, AGC, intestinal subtype, and non-PCC histological type were predictors of higher avidity of [ 18 F]FDG, which is consistent  [29]. Several limitations exist in the present study. First, in some patients, pathological information such as Lauren classification and degree of differentiation were missing, resulting in a reduced sample size available for analysis. Second, the patients included mainly had AGC; thus, not each suspected metastatic lesion was pathologically verified; the diagnosis of distant metastases depends on our reference standard of comprehensive clinical information. Third, the number of patients with EGC was limited.
In conclusion, our initial study showed that [ 68 Ga]FAPI-04 and [ 18 F]FDG dual-tracer PET/CT were complementary and improved the sensitivity of detecting pre-treatment distant metastases in gastric cancer, thus helping to improve treatment stratification for gastric patients. Additionally, it should be noted that [ 68 Ga]FAPI-04 had limited efficacy in detecting EGC.

Declarations
Guarantor The scientific guarantor of this publication is Biao Li, MD, Ph.D.

Conflict of interest
The authors of this manuscript declare no relationships with any companies whose products or services may be related to the subject matter of the article.
Statistics and biometry Prof. Jian Li kindly provided statistical advice for this manuscript.
Informed consent Written informed consent was obtained from all subjects (patients) in this study.

Methodology
• prospective • diagnostic or prognostic study • performed at one institution Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.