Abstract
Background
[68Ga]Ga-FAPI-04 targets FAP overexpressed in pancreatic cancer microenvironment. This study compared diagnostic efficacy and clinical impact with [18F]FDG PET/CT in pancreatic cancer.
Results
Of 97 suspected pancreatic cancer patients (58 males; mean age 63.0 ± 10.1 years), 88 were confirmed as pancreatic cancer (57 by pathology, 31 by imaging follow‑up). [68Ga]Ga-FAPI-04 PET/CT showed significantly higher SUVmax than [18F]FDG PET/CT in primary lesions (10.2 ± 3.5, 95%CI: 9.320-11.126 vs. 6.1 ± 3.8, 95%CI: 5.092–7.152; p = 0.001) and lymph node metastases (4.1 ± 2.4, 95%CI: 3.718–4.559 vs. 3.3 ± 2.3, 95%CI: 2.845–3.714; p = 0.001), with superior AUCs for primary tumors and lymph nodes(0.851, 95% CI: 0.735–0.981 vs. 0.802, 95% CI: 0.354–0.803; p = 0.001) (0.867, 95% CI: 0.827–0.907 vs. 0.701, 95% CI: 0.641–0.761; p = 0.001). The T/B ratio of bone and visceral metastases was higher in [68Ga]Ga-FAPI-04 PET/CT than [18F]FDG PET/CT(4.7 ± 3.1, 95%CI: 4.108–5.245 vs. 2.5 ± 1.6, 95%CI: 2.231–2.835; p = 0.001) with superior AUC(0.938, 95% CI: 0.912–0.964 vs. 0.919, 95% CI: 0.886–0.951; p = 0.040), as were liver (5.0 ± 4.0, 95%CI: 3.706–6.284 vs. 2.7 ± 1.3, 95%CI: 2.211–3.131; p = 0.001), peritoneal (5.4 ± 3.3, 95%CI: 4.318–6.482 vs. 3.0 ± 2.1. 95%CI: 2.211–3.745; p = 0.001) and other organ metastases (4.4 ± 2.5, 95%CI: 3.756–5.035 vs. 2.7 ± 1.7, 95%CI: 1.379–2.254; p = 0.001). [68Ga]Ga‑FAPI‑04 SUVmax and TBR were positive predictors for primary tumors (SUVmax: OR = 1.951, 95%CI: 1.286–2.960, p = 0.004) and nodal metastases (SUVmax: OR = 4.199, 95%CI: 2.970–5.939, p = 0.001; TBR: OR = 14.502, 95%CI: 7.228–29.095, p = 0.001). For bone and visceral metastases, both tracers were positive predictors, but [68Ga]Ga‑FAPI‑04 SUVmax (OR = 2.925, 95%CI: 2.048–4.177, p = 0.003) and T/B ratio (OR = 3.520, 95%CI: 2.311–5.362, p = 0.001) outperformed [18F]FDG (SUVmax: OR = 1.901, 95%CI: 1.308–2.761, p = 0.003; T/B ratio: OR = 2.480, 95%CI: 1.488–4.136, p = 0.001). Furthermore, [68Ga]Ga-FAPI-04 achieved higher accuracy (93.2% vs. 72.7%, p = 0.007), revised the staging of 7 patients and altered the treatment (surgical resection to unresectable) in 6 patients. Limitations include single‑center, short follow‑up, small subgroups (e.g., recurrent lesions) and incomplete pathologically confirmation.
Conclusion
[68Ga]Ga-FAPI-04 PET/CT suggested potentially better diagnostic performance to [18F]FDG PET/CT and improves staging and clinical decision-making in pancreatic cancer, suggesting potential clinical value.
Supplementary Information
The online version contains supplementary material available at 10.1186/s13550-026-01456-2.
Keywords: Pancreatic cancer, Fibroblasts, Desmoplasia, Molecular imaging, Deoxyglucose, Positron emission tomography, Computed tomography
Introduction
Pancreatic cancer (PC) is one of the most aggressive and lethal cancers with a gradual rise in incidence. Worldwide, pancreatic cancer is the seventh leading cause of death among all cancers [1]. According to projections, by 2050 the global incidence of pancreatic cancer is expected to rise to 18.6 cases per 100,000 people, with an average annual growth rate of 1.1%, meaning that pancreatic cancer will pose a significant public health burden [2]. Due to its high malignancy and insidious onset, most patients are already in the advanced stage at the first diagnosis, losing the opportunity for radical resection. Furthermore, pancreatic cancer has a low surgical resection rates and is prone to recurrence and metastasis after surgery. Despite decades of research, the improvement of overall 5-year survival of pancreatic cancer is still not significant [3]. Early diagnosis and accurate staging are crucial for improving the overall survival rates and the prognosis in pancreatic cancer patients.
Pancreatic cancer is characterized by a strong desmoplastic response around the cancer cells [4–5]. Pancreatic stellate cells are the primary cellular source of fibrosis. They are activated by injury, inflammation, or tumor signals and can further differentiate into various subpopulations of cancer-associated fibroblasts (CAFs) with distinct functions, such as α-SMA-positive myofibroblasts, which secrete large amounts of extracellular matrix components such as collagen and fibronectin. Fibroblast activation protein (FAP) is primarily expressed in CAFs, participates in extracellular matrix degradation and remodeling, plays a functional driving role in the development of fibrosis, and serves as a specific marker for activated CAFs [6–8]. [18F]deoxyglucose (fluorodeoxyglucose, FDG) PET/CT is now widely used in the clinical diagnosis and staging of pancreatic cancer. However, there exist some limitations, such as interference from blood glucose levels, inflammation or infected tissue [9].Previous studies have demonstrated that [68Ga]Ga-FAP-04 inhibitors (FAP inhibitors (FAPI)) has higher sensitivity and accuracy in the detection of lymph node and distant metastases compared to [18F]FDG [10]. However, a comprehensive, head-to-head comparison is still lacking, especially large-scale prospective studies that systematically evaluate their combined effects on diagnostic performance, accurate staging, and clinical management. This study aims to explore the value of [68Ga]Ga-FAPI-04 PET/CT in the diagnosis, staging and impact on clinical managementof pancreatic cancer compared with [18F]FDG PET/CT.
Materials and methods
Study population
This study has been approved by the Institutional Review Board of The Fourth Hospital of Hebei Medical University (Approval No.: 2021069), in accordance with the Declaration of Helsinki. And all patients provided written informed consent prior to participation.
Diagnostic reference standards
According to the eighth edition of the American Joint Committee on Cancer (AJCC) guidelines, TNM staging was based on the patient’s pathologic findings (via surgical and/or puncture biopsy) [11]. For suspicious lesions without pathological confirmation, the final diagnosis was based on clinical, laboratory, and imaging follow‑up of at least 3 months (CT or MRI). Lesion status was determined by changes in size, number, or morphology on serial imaging, combined with tumor marker trends when available.
Patients’ selection
The inclusion criteria were as follows: (1) PC patients who were confirmed by pathological examination or 3 months of sufficient follow-up; (2) Patients who underwent [18F]FDG PET/CT and [68Ga]Ga-FAPI-04 PET/CT imaging to distinguish pancreatic mass lesions and tumor staging, with an interval of less than one week between the two scans and no treatment during this period; (3) Signed informed consent forms. Exclusion criteria: (1) Pregnancy; (2) Treatment between the two scans; (3) Patients with other primary malignant tumors.
Synthesis of radiopharmaceuticals
[68Ga]Ga-labeled 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA)-FAPI-04 (referred to as [68Ga]Ga-FAPI-04) was prepared for the study using a full procedure on the ITM [68Ge]Ge/[68Ga]Ga generator (Munich, Germany). TZ-Bio (Nanchang, Jiangxi, China) provided the fibroblast activation protein inhibitor (FAPI) precursor (DOTA-FAPI-04), which was radiolabelled by first mixing 2 ml [68Ga]Ga solution (370 MBq) with 195 µl sodium acetate (1.0 M), followed by addition of 40 µg precursor DOTA-FAPI-04 to give a final pH of 4.0–4.5. The reaction was heated at 95 °C for 15 min and left at room temperature naturally. Reversed-phase high-pressure liquid chromatography (RP-HPLC) was used for quality control of the radiosynthesis. The decay-corrected radiochemical yield of [68Ga]Ga-FAPI-04 was (71.4 ± 2.5)% and the radiochemical purity was more than 95%. The radioactive agent exhibited excellent in vitro stability: its radiochemical purity remained above 95% after 180 min of storage at room temperature. [18F]FDG was provided by Hebei Andik Technology Co., Ltd., with its radiochemical purity above 98.0%. Quality control parameters met all specifications: pH 6.5–7.5, endotoxin content < 2.0 EU/mL, and pyrogen-free and sterile.
Image acquisition and processing
[18F]FDG PET/CT and [68Ga]Ga-FAPI-04 PET/CT examinations were performed within one week of each other.Patients were required to fast for at least 6 h before the [18F]FDG PET/CT scan. And their random peripheral blood glucose level should be less than 11.1 mmol/L. No specific preparation was required for the [68Ga]Ga-FAPI-04 PET/ CT examination. The intravenous injection doses of [18F]F-FDG and [68Ga]Ga-FAPI-04 were 3.7-5.5MBq/kg and 1.5-1.8MBq/kg, respectively. 60 ± 5 min after the intravenous injection of [18F]FDG and approximately 20 min after the intravenous injection of [68Ga]Ga-FAPI-04, participants underwent scanning using a Philips VEREOS PET/CT machine in the Netherlands, with scanning ranging from the skull base to upper thigh (for [18F]FDG, the head scan was performed separately among patients with suspected brain metastases) or from head to the upper thighs (for [68Ga]Ga-FAPI-04). CT scanning parameters were set as follows: tube voltage of 120 kV, current of 120 mA for fluorodeoxyglucose, and 60 mA for FAPI; slice thickness 4.0 mm. [18F]FDG and [68Ga]Ga-FAPI-04 PET scanning were performed in 3D acquisition mode immediately after CT scanning with 8–10 beds, 1.5 and 2.5 min/position, and PET images were reconstructed using the ordered subset maximum expectation value iteration method. All acquired data were transferred to the Postprocessing Workstation (IntelliSpace Portal, Philips Medical Systems, Inc.). Attenuation correction was performed using CT data, and the corrected PET images were fused with the CT images.
Image interpretation
[18F]FDG PET/CT and [68Ga]Ga-FAPI-04 PET/CT images were interpreted in random order by two nuclear medicine physicians. [18F]FDG PET/CT and the interpreters were unaware of all clinical data, including the pathology results and clinical follow-up information. Disagreements were resolved by consensus with a third blinded reader. The intraclass correlation coefficient (ICC) was calculated for the measurements taken by two blinded readers. The ICC for all parameters demonstrated excellent interobserver agreement (all ICCs > 0.93). Detailed results are presented in Supplemental Table 1. (1) Visual assessment. Lesions were considered positive when tracer uptake was higher than the surrounding tissue background (excluding physiologic tracer uptake and definite benign disease). (2) Semi-quantitative analysis. The ROI of positive lesions was manually outlined in the cross-sectional position of PET/CT images, and the maximum standardized uptake value (SUVmax) of the lesion was automatically calculated. To ensure that SUVmax was relatively comparable. For lymph node, lung, and peritoneal lesions, the tumor-to-mediastinal blood pool ratio (TBR) (lesion SUVmax/descending aorta SUVmean) was calculated. For liver and bone lesions, calculate the tumor-to-normal liver parenchyma ratio (TLR) (lesions SUVmax/liver SUVmean). The PET/CT evaluation results were compared with biopsy or surgical pathology results, and semiquantitative parameters associated with [68Ga]Ga-FAPI PET/CT and [18F]FDG PET/CT images were comparatively analyzed.
Statistical analysis
Statistical analysis was performed using IBM SPSS Statics 26.0 software. Paired t-test was used for comparison between groups for quantitative data. The McNemar’s test was used to compare the difference in detection rates of pancreatic cancer lesions between [18F]FDG PET/CT and [68Ga]Ga-FAPI-04 PET/CT. ROC curves were constructed to compare the diagnostic performance of the two imaging agents, with optimal cut-off values determined by the Youden index. The DeLong test was used to compare the areas under the ROC curves (AUC). Logistic regression combined with the DeLong test was performed to evaluate the diagnostic and predictive value of the two examinations for lesions. The chi-square test was used to compare the accuracy for staging between the two methods. To control the family‑wise error rate due to multiple hypothesis tests (including comparisons of SUVmax, T/B ratio, AUCs from DeLong tests, and diagnostic accuracy indices across different lesion types), the Holm‑Bonferroni sequential correction was applied. Finally, the results of the McNemar test, DeLong test, and logistic regression were adjusted. Adjusted p < 0.05 was considered statistically significant.
Sensitivity power analysis
Statistical analyses were performed using R software (version 4.5.3). Paired McNemar’s test was used to compare lesion detection rates between [68Ga]Ga‑FAPI‑04 and [18F]F‑FDG PET/CT. A sensitivity power analysis was conducted to assess the adequacy of the sample size. With a two‑sided α of 0.05 and a target power of 0.80, the minimum detectable odds ratio (OR) was calculated based on the number of discordant pairs. Among 332 lesions, the observed OR for discordant pairs was 7.0, which exceeded the minimum detectable OR of 3.67. Therefore, the sample size provided sufficient power to detect the difference between the two tracers, supporting the robustness of the lesion‑level comparative results.
Subgroup analysis according to prior treatment
Patients were stratified into initial diagnosed (n = 59) and previously treated (n = 29) groups. Within each subgroup, paired t‑tests, McNemar’s tests, and DeLong’s tests were used to compare SUVmax, T/B ratio, and AUC, respectively, between [68Ga]Ga-FAPI-04 and [18F]FDG. Delta values (Δ = FAPI-FDG) were compared between subgroups using two‑sample t‑tests. A mixed‑effects model with a tracer‑by‑treatment interaction term was fitted to assess effect modification. p values were adjusted using the Holm‑Bonferroni method where applicable.
Results
Patients’ characteristics
A total of 97 suspected PC patients (58 males, 39 females, 63.0 ± 10.1 years old) were enrolled in this study from January 2022 to January 2026 prospectively. 88 patients were confirmed as pancreatic cancer by pathology (n = 57) or imaging follow-up(n = 31). The 57 patients included 49 with pancreatic ductal adenocarcinoma, 1 with mucinous adenocarcinoma, and 1 with adenocarcinoma. 9 patients that were confirmed to have benign lesions by follow-up or pathological findings, including 3 with pancreatitis and 6 with pancreatic serous cystadenoma. Patients underwent both [18F]FDG and [68Ga]Ga-FAPI-04 PET/CT for initial staging (n = 59) and for the detection of recurrence (n = 29). Table 1 demonstrates the characteristics of patients.
Table 1.
Characteristics of patients
| Characteristic | Overall value* |
|---|---|
| Age | |
| Mean ± SD | 63.0 ± 10.1 |
| Sex | |
| Male | 58(59.8) |
| Female | 39(40.2) |
| Pathological Type for PC patients | |
| Pancreatic ductal adenocarcinoma | 55(96.5) |
| Mucinous adenocarcinoma | 1(1.75) |
| Adenocarcinoma | 1(1.75) |
| Role of PET/CT for PC patients | |
| Initial staging | 59(67.0) |
| Restaging | 29(33.0) |
| Therapy before PET/CT | |
| Any therapy | 29(33.0) |
| Chemotherapy | 2(2.3) |
| Resection | 14(15.9) |
| Chemo/radio therapy after resection | 13(14.8) |
| None | 59(67.0) |
| T stage | |
| T1 | 3(4.8) |
| T2 | 22(35.5) |
| T3 | 29(46.8) |
| T4 | 8(12.9) |
|
N stage N0 N1 N2 M stage M0 M1 Clinical stage Ⅰ Ⅱ Ⅲ Ⅳ |
53(60.2) 20(22.7) 15(17.1) 51(58.0) 37(42.0) 8(12.9) 11(17.7) 11(17.7) 32(51.6) |
*Data are reported as numbers of patients, with percentages in parentheses, unless otherwise indicated
Only N-staging and M-staging were performed on patients who had undergone surgery for pancreatic cancer
Evaluation of the primary tumors and recurrence
In the comparison of 88 patients between [68Ga]Ga-FAPI-04 and [18F]FDG PET/CT, 26 primary lesions were pathologically diagnosed via puncture biopsy or surgery. As is shown in Table 2; Figs. 1A, the SUVmax of pancreatic cancer primary lesions in [68Ga]Ga-FAPI-04 PET/CT was significantly higher than that of [18F]FDG PET/CT (10.2 ± 3.5, 95%CI: 9.320-11.126 vs. 6.1 ± 3.8, 95%CI: 5.092–7.152; p = 0.001). The SUVmax of recurrent lesions did not significantly differ between [68Ga]Ga-FAPI-04 and [18F]FDG PET/CT (7.4 ± 4.6, 95%CI: 3.930-10.959 vs. 6.5 ± 5.4, 95%CI: 0.834–12.066; p = 1.000). Analysis based on patients in Table 3 showed that the sensitivity (98.3% vs. 89.8%, p = 0.227), specificity (88.9% vs. 77.8%, p = 1.000) and accuracy (97.1% vs. 88.2%, p = 0.250) of [68Ga]Ga-FAPI-04 for detecting primary lesions were higher that of [18F]FDG but there were no statistically significant difference between the two. Analysis based on the lesions in Table 4 showed the diagnostic efficacy of [68Ga]Ga-FAPI-04 and [18F]FDG. The sensitivity(98.4% vs. 90.2%, p = 0.378), specificity(88.9% vs. 77.8%, p = 1.000) and accuracy(97.1% vs. 87.1%, p = 1.000) of [68Ga]Ga-FAPI-04 for detecting primary lesions were the same at both the patient level and the lesion level, which tended to be higher than that of [18F]FDG. The diagnostic sensitivitiy (100% vs. 66.7%, p = -) and accuracy (100% vs. 89.7%, p = 0.750) for recurrent lesions in [68Ga]Ga-FAPI-04 tended to be higher than that of [18F]FDG. The specificity in [68Ga]Ga-FAPI-04 PET/CT and [18F]FDG PET/CT both had 100%. The results of the ROC curve analysis and the DeLong test are shown in Table 5. The AUC of SUVmax of primary lesions in [68Ga]Ga-FAPI-04 PET/CT was higher than that on [18F]FDG PET/CT (0.851, 95% CI: 0.735–0.981 vs. 0.802, 95% CI: 0.354–0.803; p = 0.001). No statistical significance was observed in DeLong test for recurrent lesions. In Table 6 and the forest plot in Fig. 2, the results of logistic regression analysis showed that [68Ga]Ga‑FAPI‑04 SUVmax values were positive predictors of primary tumors (OR = 1.951, 95%CI: 1.286–2.960, p = 0.004). Comprehensive comparative results are illustrated in Figs. 3 and 4.
Table 2.
Comparison of [18F]FDG and [68Ga]Ga-FAPI-04 uptake in primary and metastatic lesions in pancreatic cancer
| [18F]FDG | [68Ga]Ga-FAPI-04 | p(raw) | p(corrected)* | t | Cohen’s d | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| No. of patients | No. of leisions | Mean ± SD | 95%CI | No. of leisions | Mean ± SD | 95%CI | |||||
| Primary tumors | |||||||||||
| SUVmax | 59 | 55 | 6.1 ± 3.8 | 5.092–7.152 | 60 | 10.2 ± 3.5 | 9.320-11.126 | < 0.001 | 0.001 | 6.921 | 0.933 |
| Recurrent lesions | |||||||||||
| SUVmax | 9 | 6 | 6.5 ± 5.4 | 0.834–12.066 | 9 | 7.4 ± 4.6 | 3.930-10.959 | 0.579 | 1.000 | 0.593 | 0.242 |
| Lymph nodes | |||||||||||
| SUVmax | 32 | 108 | 3.3 ± 2.3 | 2.845–3.714 | 123 | 4.1 ± 2.4 | 3.718–4.559 | < 0.001 | 0.001 | 3.963 | 0.387 |
| TBR | 108 | 2.4 ± 1.8 | 2.097–2.792 | 123 | 1.6 ± 1.0 | 1.444–1.783 | < 0.001 | 0.001 | -4.314 | -0.421 | |
| Bone and* visceral metastases | |||||||||||
| SUVmax | 53 | 125 | 4.9 ± 3.1 | 4.317–5.496 | 139 | 4.8 ± 2.0 | 4.390–5.134 | 0.780 | 1.000 | -0.280 | -0.027 |
| T/B Ratio | 125 | 2.5 ± 1.6 | 2.231–2.835 | 139 | 4.7 ± 3.1 | 4.108–5.245 | < 0.001 | 0.001 | 7.388 | 0.721 | |
| Liver | |||||||||||
| SUVmax | 23 | 35 | 5.6 ± 2.6 | 4.658–6.451 | 39 | 4.7 ± 1.8 | 4.200-5.339 | 0.090 | 0.360 | -1.744 | -0.299 |
| TLR | 35 | 2.7 ± 1.3 | 2.211–3.131 | 39 | 5.0 ± 4.0 | 3.706–6.284 | < 0.001 | 0.001 | 4.438 | 0.761 | |
| Peritoneal | |||||||||||
| SUVmax | 19 | 32 | 5.1 ± 4.4 | 3.505–6.658 | 38 | 4.6 ± 2.1 | 3.854–5.251 | 0.256 | 0.768 | -1.157 | -0.205 |
| TBR | 32 | 3.0 ± 2.1 | 2.211–3.745 | 38 | 5.4 ± 3.3 | 4.318–6.482 | < 0.001 | 0.001 | 3.967 | 0.701 | |
| Metastases in other organs | |||||||||||
| SUVmax | 27 | 58 | 4.7 ± 2.7 | 3.996–5.391 | 62 | 5.2 ± 2.2 | 4.658–5.783 | 0.079 | 0.395 | 1.787 | 0.235 |
| TBR | 58 | 2.7 ± 1.7 | 1.379–2.254 | 62 | 4.4 ± 2.5 | 3.756–5.035 | < 0.001 | 0.001 | 4.676 | 0.614 | |
*p values were adjusted for multiple comparisons using the Holm-Bonferroni method (10 comparisons)
Fig. 1.
A SUVmax(left) and T/B ratio(right) of primary and metastatic lesions in [18F]FDG(green) and [68Ga]Ga-FAPI-04 PET/CT(orange) B Comparison of TNM staging, T staging, N staging, and M staging between [18F]FDG(left) and [68Ga]Ga-FAPI-04 PET/CT(right). Green: concordant staging; yellow: underestimated staging; pink: overestimated staging
Table 3.
Comparison of diagnostic efficacy between [18F]FDG and [68Ga]Ga-FAPI-04 PET/CT based on patients
| Location of lesions | Sen (%) | Spe (%) | Acc(%) | PPV(%) | NPV(%) | |
|---|---|---|---|---|---|---|
| Primary tumors | ||||||
| [68Ga]Ga-FAPI-04 | 98.3(58/59) | 88.9(8/9) | 97.1(66/68) | 98.3(58/59) | 88.9(8/9) | |
| [18F]FDG | 89.8(53/59) | 77.8(7/9) | 88.2(60/68) | 96.4(53/55) | 53.9(7/13) | |
| κ | 0.264 | 0.609 | 0.719 | χ2 | 0.419 | - |
| p(raw) | 0.063 | 1.000 | 0.125 | p | 0.517 | 0.101 |
| p(corrected)* | 0.227 | 1.000 | 0.250 | |||
| Lymph nodes | ||||||
| [68Ga]Ga-FAPI-04 | 96.9(31/32) | 94.4 (17/18) | 96.0(48/50) | 96.9(31/32) | 94.4(17/18) | |
| [18F]FDG | 87.5(28/32) | 94.4 (17/18) | 90.0(45/50) | 96.6(28/29) | 81.0(17/21) | |
| κ | 0.549 | 1.000 | 0.509 | χ2 | 0.509 | 1.564 |
| p(raw) | 0.248 | 1.000 | 0.250 | p | 1.000 | 0.211 |
| p(corrected)* | 0.750 | 1.000 | 0.500 | |||
| Bone and visceral metastasis | ||||||
| [68Ga]Ga-FAPI-04 | 92.5(49/53) | 91.7(11/12) | 92.3(60/65) | 98.0(49/50) | 73.3(11/15) | |
| [18F]FDG | 84.9(45/53) | 91.7(11/12) | 86.2 (56/65) | 97.8(45/46) | 57.9(11/19) | |
| κ | 0.444 | 1.000 | 0.525 | χ2 | 0.695 | 0.330 |
| p(raw) | 0.219 | 1.000 | 0.219 | p | 0.875 | 0.566 |
| p(corrected)* | 0.657 | 1.000 | 0.657 | |||
| Recurrent leisions | ||||||
| [68Ga]Ga-FAPI-04 | 100(9/9) | 100(20/20) | 100(29/29) | 100(9/9) | 100(20/20) | |
| [18F]FDG | 66.7(6/9) | 100(20/20) | 89.7(26/29) | 100(6/6) | 86.7(20/23) | |
| κ | - | - | 0.734 | χ2 | - | - |
| p(raw) | - | - | 0.250 | p | - | 0.236 |
| p(corrected)* | - | - | 0.750 | |||
*p values were adjusted for multiple comparisons using the Holm-Bonferroni method (10 comparisons)
Table 4.
Comparison of diagnostic efficacy between [18F]FDG and [68Ga]Ga-FAPI-04 PET/CT based on lesions
| Location of lesions | Sen (%) | Spe (%) | Acc(%) | PPV(%) | NPV(%) | |
|---|---|---|---|---|---|---|
| Primary tumors | ||||||
| [68Ga]Ga-FAPI-04 | 98.4(60/61) | 88.9(8/9) | 97.1(68/70) | 98.4(60/61) | 88.9(8/9) | |
| [18F]FDG | 90.2(55/61) | 77.8(7/9) | 87.1(62/70) | 96.5(55/57) | 53.9(7/13) | |
| κ | 0.265 | 0.609 | 0.678 | χ2 | 0.416 | - |
| p(raw) | 0.063 | 1.000 | 0.219 | p | 0.953 | 0.101 |
| p(corrected)* | 0.378 | 1.000 | 1.000 | |||
| Lymph nodes | ||||||
| [68Ga]Ga-FAPI-04 | 93.9(123/131) | 97.5 (196/201) | 96.1(310/332) | 96.1(123/128) | 96.1(196/204) | |
| [18F]FDG | 82.4(108/131) | 99.0 (199/201) | 92.5(307/332) | 98.2(108/110) | 89.6(199/222) | |
| κ | 0.255 | 0.565 | 0.843 | χ2 | 0.936 | 5.173 |
| p(raw) | 0.001 | 0.250 | <0.001 | p | 0.333 | 0.023 |
| p(corrected)* | 0.008 | 1.000 | 0.005 | |||
| Bone and visceral metastasis | ||||||
| [68Ga]Ga-FAPI-04 | 92.7(139/150) | 92.9(118/127) | 92.8(257/277) | 96.2(139/148) | 91.5(118/129) | |
| [18F]FDG | 83.3(125/150) | 96.1(122/127) | 89.2 (243/277) | 93.9(125/130) | 83.0(122/147) | |
| κ | 0.443 | 0.587 | 0.855 | χ2 | 0.695 | 4.691 |
| p(raw) | 0.001 | 0.348 | < 0.001 | p | 0.404 | |
| p(corrected)* | 0.007 | 0.696 | 0.005 | |||
| Recurrent leisions | ||||||
| [68Ga]Ga-FAPI-04 | 100(9/9) | 100(20/20) | 100(29/29) | 100(9/9) | 100(20/20) | |
| [18F]FDG | 66.7(6/9) | 100(20/20) | 89.7(26/29) | 86.7(20/23) | ||
| κ | - | - | 0.734 | χ2 | - | - |
| p(raw) | - | - | 0.250 | p | - | 0.236 |
| p(corrected)* | - | - | 0.750 | |||
*p values were adjusted for multiple comparisons using the Holm-Bonferroni method (10 comparisons)
Table 5.
Comparison of ROC curves between [18F]FDG and [68Ga]Ga-FAPI-04 PET/CT
| Location of lesions | AUC(95%CI) | Cut-off | Sen (%) | Spe (%) | Acc(%) | PPV(%) | NPV(%) | DeLong test p(raw) | DeLong test p(corrected)* | |
|---|---|---|---|---|---|---|---|---|---|---|
| Primary tumors | ||||||||||
| SUVmax | FAPI | 0.851(0.735–0.981) | 8.20 | 80.0 | 88.9 | 81.2 | 97.9 | 77.8 | < 0.001 | 0.001 |
| FDG | 0.802(0.354–0.803) | 7.80 | 76.7 | 77.8 | 76.8 | 95.8 | 33.3 | |||
| Lymph nodes | ||||||||||
| SUVmax | FAPI | 0.867(0.827–0.907) | 2.15 | 81.3 | 83.3 | 80.1 | 71.7 | 81.1 | < 0.001 | 0.001 |
| FDG | 0.701(0.641–0.761) | 2.55 | 47.3 | 87.4 | 69.3 | 67.5 | 87.1 | |||
| TBR | FAPI | 0.754(0.699–0.808) | 1.32 | 55.5 | 85.3 | 72.3 | 70.4 | 85.5 | 0.816 | 1.000 |
| FDG | 0.750(0.696–0.804) | 1.44 | 70.0 | 68.9 | 69.6 | 60.3 | 71.1 | |||
| Bone and visceral metastases | ||||||||||
| SUVmax | FAPI | 0.958(0.934–0.983) | 2.35 | 89.2 | 94.6 | 88.8 | 92.8 | 92.1 | 0.128 | 0.768 |
| FDG | 0.923(0.893–0.953) | 2.35 | 86.9 | 83.7 | 85.9 | 90.5 | 89.8 | |||
| T/B Ratio | FAPI | 0.938(0.912–0.964) | 1.76 | 89.9 | 83.7 | 84.1 | 83.5 | 79.5 | 0.040 | 0.040 |
| FDG | 0.919(0.886–0.951) | 1.19 | 89.2 | 83.0 | 83.4 | 84.7 | 81.9 | |||
| Recurrent lesions | ||||||||||
| SUVmax | FAPI | 0.676(0.441–0.910) | 8.55 | 44.4 | 94.4 | 77.8 | 44.4 | 77.2 | 0.446 | 1.000 |
| FDG | 0.574(0.344–0.804) | 3.25 | 66.7 | 55.6 | 63.0 | 83.3 | 92.3 | |||
*p values were adjusted for multiple comparisons using the Holm-Bonferroni method (10 comparisons)
Table 6.
Comparison of OR value between [18F]FDG and [68Ga]Ga-FAPI-04 PET/CT
| [18F]FDG | p(raw) | p(corrected)* | [68Ga]Ga-FAPI-04 | p(raw) | p(corrected)* | |
|---|---|---|---|---|---|---|
| OR(95%CI) | OR(95%CI) | |||||
| Primary tumor | ||||||
| SUVmax | 0.646(0.486–0.859) | 0.003 | 0.003 | 1.951(1.286–2.960) | 0.002 | 0.004 |
| Lymph nodes | ||||||
| SUVmax | 0.703(0.613–0.807) | < 0.001 | 0.001 | 4.199(2.970–5.939) | < 0.001 | 0.001 |
| TBR | 2.979(2.174–4.081) | < 0.001 | 0.001 | 14.502(7.228–29.095) | < 0.001 | 0.001 |
| Bone and visceral metastases | ||||||
| SUVmax | 1.901(1.308–2.761) | 0.001 | 0.003 | 2.925(2.048–4.177) | < 0.001 | 0.003 |
| T/B Ratio | 2.480(1.488–4.136) | < 0.001 | 0.001 | 3.520(2.311–5.362) | < 0.001 | 0.001 |
*p values were adjusted for multiple comparisons using the Holm-Bonferroni method (10 comparisons)
Fig. 2.
Forest plot comparing the diagnostic value of [68Ga]Ga-FAPI-04 and [18F]FDG PET/CT parameters in pancreatic cancer.red: SUVmax of primary tumors; blue: SUVmax and T/B ratio of lymph nodes; yellow: SUVmax and TBR of bone and visceral metastases
Fig. 3.
Comparison of [18F]FDG and [68Ga]Ga-FAPI-04 PET/CT imaging of a patient with pancreatic cancer (male, 63 years old) (arrow indicates the lesion). (a) [18F]FDG PET/CT maximum intensity projection (MIP) and transverse images showed that the primary tumor lesion and the lesion in lung was negative for [18F]FDG uptake; (b) In [68Ga]Ga-FAPI-04 PET/CT, the primary tumor (SUVmax = 13.2) and the lesion in lung (SUVmax = 2.9) showed significantly increased [68Ga]Ga-FAPI-04 uptake, and the tumor outline was clearly displayed. Postoperative pathology confirmed pancreatic ductal adenocarcinoma
Fig. 4.
Comparison of [18F]FDG and [68Ga]Ga-FAPI-04 PET/CT imaging of a patient with pancreatic cancer (male, 59 years old) (arrow indicates the lesion). a.[18F]FDG PET/CT maximum intensity projection (MIP) and transverse images showed the uptake of primary tumor lesion (blue arrow, SUVmax = 3.6) and the leision in liver (black arrow) was negative in [18F]FDG; b.In [68Ga]Ga-FAPI-04 PET/CT, the primary tumor (blue arrow, SUVmax = 9.4) and the lesion in liver (black arrow, SUVmax = 5.2) showed significantly increased [68Ga]Ga-FAPI-04 uptake, and the tumor outline was clearly displayed. Postoperative pathology confirmed pancreatic ductal adenocarcinoma
Evaluation of lymph node metastasis
Tables 2, 3 and 4 illustrates the analysis of 131 lymph nodes in 32 patients with pathologically confirmed diagnosis (3/32) or imaging follow-up suggestive (29/32). [68Ga]Ga-FAPI-04 PET/CT showed significantly higher SUVmax of lymph node lesions than [18F]FDG (SUVmax, 4.1 ± 2.4, 95%CI: 3.718–4.559 vs. 3.3 ± 2.3, 95%CI: 2.845–3.714; p = 0.001). For diagnostic efficacy based on patients, the sensitivity (96.9% vs. 87.5%, p = 0.750) and accuracy (96.0% vs. 90.0%, p = 0.500) of [68Ga]Ga-FAPI-04 were higher that of [18F]FDG but there were no statistically significant difference between the two. For diagnostic efficacy based on lesions, the sensitivity and accuracy of lymph node metastases in [68Ga]Ga-FAPI-04 PET/CT were significantly higher than [18F]FDG PET/CT (93.9% vs. 82.4%, p = 0.008) (96.1% vs. 92.5%, p = 0.005). The AUC of SUVmax of lymph node metastases in [68Ga]Ga-FAPI-04 PET/CT was higher than that in [18F]FDG PET/CT(0.867, 95% CI: 0.827–0.907 vs. 0.701, 95% CI: 0.641–0.761; p = 0.001), with optimal SUVmax cut‑off values of 2.15 and 2.55, respectively. Besides, [68Ga]Ga‑FAPI‑04 SUVmax and TBR were positive predictors of lymph node metastases (SUVmax: OR = 4.199, 95%CI: 2.970–5.939, p = 0.001;TBR: OR = 14.502, 95%CI: 7.228–29.095, p = 0.001).
Evaluation of distant metastasis
In 53 PC patients with bone and visceral metastases confirmed by pathology (15/53) or imaging following-up (38/53). Bone and visceral metastases include subtypes of liver, peritoneal and other organ metastases. Of the 23 patients with liver metastases, 8 were confirmed by needle biopsy or postoperative pathology, and 15 were confirmed by imaging follow-up. Of the 19 patients with peritoneal metastases, 5 were confirmed by postoperative pathology, and 14 were confirmed by imaging follow-up. [68Ga]Ga-FAPI-04 PET/CT showed higher T/B ratio than that of [18F]FDG PET/CT in bone and visceral metastases (4.7 ± 3.1, 95%CI: 4.108–5.245 vs. 2.5 ± 1.6, 95%CI: 2.231–2.835; p = 0.001), as were liver (5.0 ± 4.0, 95%CI: 3.706–6.284 vs. 2.7 ± 1.3, 95%CI: 2.211–3.131; p = 0.001), peritoneal (5.4 ± 3.3, 95%CI: 4.318–6.482 vs. 3.0 ± 2.1. 95%CI: 2.211–3.745; p = 0.001) and other organ metastases (4.4 ± 2.5, 95%CI: 3.756–5.035 vs. 2.7 ± 1.7, 95%CI: 1.379–2.254; p = 0.001). For diagnostic efficacy based on patients, the sensitivity (92.5% vs. 84.9%, p = 0.657) and accuracy (92.3% vs. 86.2%, p = 0.657) of [68Ga]Ga-FAPI-04 for detecting distant metastases were higher that of [18F]FDG but there were no statistically significant difference between the two. In addition, the sensitivity and accuracy of [68Ga]Ga-FAPI-04 for detecting distant metastases based on the lesions were higher than those of [18F]FDG. (92.7% vs. 83.3%, p = 0.007)(92.8% vs. 89.2%, p = 0.005) Furthermore, the AUC of T/B ratio of bone and visceral metastases in [68Ga]Ga-FAPI-04 PET/CT was higher than that in [18F]FDG PET/CT(0.938, 95% CI: 0.912–0.964 vs. 0.919, 95% CI: 0.886–0.951; p = 0.040), with optimal T/B ratio cut‑off values of 1.76 and 1.19, respectively. For bone and visceral metastases, both tracers were positive predictors, but [68Ga]Ga‑FAPI‑04 SUVmax (OR = 2.925, 95%CI: 2.048–4.177, p < 0.001) and T/B ratio (OR = 3.520, 95%CI: 2.311–5.362, p < 0.001) outperformed [18F]FDG (SUVmax: OR = 1.901, 95%CI: 1.308–2.761, p = 0.003; T/B ratio: OR = 2.480, 95%CI: 1.488–4.136, p = 0.001). Comprehensive comparative results are illustrated in Figs. 1A, 3 and 4.
Comparison of TNM staging based on[18F]FDG and [68Ga]Ga-FAPI-04 PET/CT
[68Ga]Ga-FAPI-04 achieved higher accuracy (93.2% vs. 72.7%, p = 0.007). [68Ga]Ga-FAPI-04 PET/CT could detect more lymph node metastases, and the N stage changed from N0 to N1 in 3 patients and from N1 to N2 in 2 patients compared with [18F]FDG PET/CT. [68Ga]Ga-FAPI PET/CT overestimated the stage of one patient with N1 as N2. Both [18F]FDG PET/CT and [68Ga]Ga-FAPI-04 PET/CT failed to detect one patient with N1-stage disease. In addition, 3 patients with distant metastases missed by [18F]FDG PET/CT were changed from M0 to M1 after [68Ga]Ga-FAPI-04 PET/CT. Both [18F]FDG PET/CT and [68Ga]Ga-FAPI-04 PET/CT failed to detect one patient with M1-stage disease.Compared with [18F]FDG PET/CT, [68Ga]Ga-FAPI-04 PET/CT upgraded the clinical staging of 7 patients (2 from stage I to stage II, 1 from stage I to stage III, 2 from stage I to stage IV, and 2 from stage II to stage IV). Comprehensive comparative results are illustrated in Fig. 1B.Six patients ultimately had a change in treatment (surgical resection to unresectable). Comprehensive comparative results are illustrated in Fig. 1B.
Subgroup analysis according to prior treatment
Results stratified by prior treatment are summarized in Supplemental Table 2. In initial diagnosed (n = 59) and previously treated (n = 29) subgroups, [68Ga]Ga-FAPI-04 PET/CT showed significantly higher SUVmax than [18F]FDG PET/CT for lymph node metastases (p < 0.001 and p = 0.026, respectively), and higher T/B ratio for bone and visceral metastases (p < 0.001). No significant between‑subgroup differences were observed for ΔSUVmax or ΔTBR (all p > 0.05). AUCs for [68Ga]Ga-FAPI-04 PET/CT were superior to [18F]FDG PET/CT in both subgroups for lymph nodes SUVmax (p < 0.001) and for bone and visceral T/B ratio (p = 0.042 and p = 0.015). Mixed‑effects analysis revealed no significant tracer‑by‑treatment interaction (for FAPI lymph node SUVmax, p = 0.121; for FAPI bone and visceral metastases SUVmax, p = 0.069; for FAPI lymph node TBR, p = 0.964; for FAPI bone and visceral metastases T/B, p = 0.513) (for FDG lymph node SUVmax, p = 0.070; for FDG bone and visceral metastases SUVmax, p = 0.256; for FDG lymph node TBR, p = 0.780; for FDG bone and visceral metastases T/B, p = 0.870), indicating that prior treatment did not modify the relative diagnostic advantage of [68Ga]Ga-FAPI-04 PET/CT over [18F]FDG PET/CT.
Discussion
This study systematically compared the diagnostic performance, staging accuracy, and impact on clinical decision-making of [68Ga]Ga-FAPI-04 PET/CT versus [18F]FDG PET/CT in patients with pancreatic cancer through a prospective, head-to-head cohort study. Overall, compared with [18F]FDG PET/CT, [68Ga]Ga-FAPI-04 PET/CT demonstrates potential advantages in detecting primary tumors, recurrent tumors, lymph node metastases, and bone and visceral metastases, and provides more accurate TNM staging.
Accurate diagnosis and staging are essential for selecting treatment options and improving prognosis in patients with PC. [18F]FDG PET/CT has been widely applied in the evaluation of malignant tumors. However, [18F]FDG PET/CT has limited application value for detection and staging of pancreatic cancer lesions. [18F]FDG accumulates not only in tumor cells but also in inflammatory areas and metabolically active normal tissues, reducing the specificity of imaging. [18F]FDG PET/CT may also fail to detect early tumor lesions and small metastases [12]. Furthermore, diabetes is a common complication of pancreatic cancer [13]. Elevated serum glucose levels in patients can reduce FDG uptake by pancreatic tumor cells, leading to false-negative results. In this study, the detection rate of [68Ga]Ga-FAPI PET/CT was higher than [18F]FDG PET/CT for PC primary and metastatses which is consistent with the findings of previous studies. This superior performance may be attributed to the prominent and bulky pro-fibrotic and proliferative stroma of PC tumors, in which cancer-associated fibroblasts account for more than 90% of the tumor volume [4].The sensitivity of [68Ga]Ga-FAPI-04 PET/CT for diagnosing primary lesions was similar to that reported by Pang et al.(98.4% vs.100%), but the specificity of [68Ga]Ga-FAPI-04 PET/CT was higher in this study(88.9% vs. 30.0%). This may be because the benign lesions in this study were predominantly serous cystic adenomas, which likely exhibit low FAP expression; in contrast, Pang’s study included IgG4-related disease and active pancreatitis, which are characterized by high FAP expression [14]. [68Ga]Ga-FAPI-04 PET/CT in pancreatic cancer exhibits high SUVmax, TBR, and clear tumor boundaries. These features not only facilitate the early diagnosis of pancreatic cancer but also provide valuable information for target volume delineation, thereby enhancing the accuracy of total tumor volume outlining in radiation therapy planning [15]. However, [68Ga]Ga-FAPI-04 PET/CT did not show superior specificity to [18F]FDG PET/CT. Normal pancreatic tissues and benign lesions can show [68Ga]Ga-FAPI-04 uptake, with SUVmax ranging from 0.94 to 8.26 for normal pancreatic tissues, and SUVmax ranging from 3.1 to 9.1 for benign lesions, such as pancreatic pseudocysts, pancreatitis, IgG4-associated diseases, and plasmacytoid cystadenomas [16–19]. A significant overlap in SUVmax values between pancreatic malignant and benign lesions may reduce the diagnostic specificity of [68Ga]Ga-FAPI-04 PET/CT for pancreatic cancer. It has been suggested that multi-temporal dynamic observation with [68Ga]Ga-FAPI-04 PET/CT helps to distinguish pancreatic cancer from benign pancreatic lesions.And pancreatic cancer shows an increase in uptake of lesions with longer acquisition time, while the opposite is true for pancreatic inflammatory lesions [20]. Furthermore, due to the limited sample size, no significant diagnostic advantage for [68Ga]Ga-FAPI-04 PET/CT was observed in pancreatic cancer recurrence lesions and the logistic regression model for recurrence sites showed no statistical significance.This finding still requires validation in studies with larger patient cohorts.
Lymph node metastasis is the primary metastatic pattern of pancreatic cancer, and the number of involved lymph nodes and lymph node ratio (LNR) are closely related to the clinical management of pancreatic cancer patients [21]. [18F]FDG PET/CT has low to moderate sensitivity in assessing lymph node metastasis [22]. In this study, [68Ga]Ga-FAPI-04 PET/CT detected more involved lymph nodes than [18F]FDG PET/CT, leading to N-stage adjustments in five patients and demonstrating higher sensitivity that is superior to the study by Koerber et al. (93.9% vs. 89%) [23]. Compared with previous studies, [68Ga]Ga-FAPI-04 PET/CT demonstrated higher sensitivity and specificity. This may be related to the differences in inflammatory/fibrotic activity among benign lesions. Besides, this study determined the optimal SUVmax cutoff value of 2.15 using ROC curves. By combining visual interpretation with CT morphology and this threshold, subjective errors were reduced, thereby maintaining high sensitivity and eliminating false-positive lesions. Unlike the study by Ding et al., this study focused more on diagnostic performance and staging accuracy through using a logistic regression model and directly compared the ORs of SUVmax and TBR between the two tracers as independent predictive factors [24]. The results showed that for lymph node metastasis, both SUVmax (OR = 4.199, 95%CI: 2.970–5.939) and TBR (OR = 14.502, 95%CI: 7.228–29.095) in [68Ga]Ga-FAPI-04 PET/CT were significantly higher than those of [18F]FDG PET/CT, suggesting that [68Ga]Ga-FAPI-04 PET/CT parameters are stronger positive predictors. Although [68Ga]Ga‑FAPI‑04 PET/CT showed significantly higher SUVmax in lymph node metastases than [18F]FDG PET/CT, its TBR was lower (TBR, 1.6 ± 1.0, 95%CI: 1.444–1.783 vs. 2.4 ± 1.8, 95%CI: 2.097–2.792; p = 0.001). This may be explained by the rapid clearance of [68Ga]Ga‑FAPI‑04 from the blood pool, leading to a lower mediastinal background. [18F]FDG often exhibits higher background uptake in inflammatory or reactive lymph nodes, which elevates its TBR without necessarily indicating true metastatic involvement [25].By overcoming the current limitations in the accurate assessment of N-stage, [68Ga]Ga-FAPI-04 PET/CT can optimize surgical planning and ultimately improve the prognosis of pancreatic cancer patients.
Pancreatic cancer is obviously invasive and metastatic, and can metastasize to distant organs such as lungs, liver, and bones in the early stage of cancer [21]. [18F]FDG PET/CT has some limitations in detecting liver metastases, including higher physiological liver uptake and poorer visualization of small metastases [26]. Physiologic uptake of [68Ga]Ga-FAPI-04 does not exist in the normal liver, resulting in a high target-to-background ratio of [68Ga]Ga-FAPI-04 PET/CT [27–28]. In this study, [68Ga]Ga-FAPI PET/CT identified liver metastatic lesions in six patients, which were missed by [18F]FDG PET/CT. In addition, two patients with liver metastases were found to be missed by both [68Ga]Ga-FAPI-04 PET/CT and [18F]FDG PET/CT confirmed by follow-up which is consistent with the characteristics of defibrillated metastases described by Özdemir et al. [28].One of these patients had no abnormalities in the density of liver metastases or in the uptake of the contrast agent, which may be related to the small size of the metastases, the undamaged structure, and the de-interstitialized nature of the metastases [18, 28, 29]. The other patient had liver metastases with no significant abnormal uptake of contrast, which may be related to the metastases being rich in mucus or cystic degeneration, low proliferative activity or low mesenchymal response, and small lesion size [19, 30, 31]. [68Ga]Ga-FAPI-04 PET/CT imaging also demonstrated higher T/B ratio of peritoneal, bone, and lung metastatic lesions. According to a meta-analysis by Zhao et al., in gastric, colorectal, and liver cancers, [68Ga]Ga-FAPI-04 PET/CT also demonstrated superior lesion detection rates compared to [18F]FDG PET/CT due to the abundance of cancer-associated fibroblasts, particularly in cases of peritoneal and liver metastases. However, pancreatic cancer exhibits a more pronounced and diffuse degree of fibrosis compared to the aforementioned tumors, which explains why [68Ga]Ga-FAPI-04 PET/CT SUVmax was higher than that of [18F]FDG PET/CT in primary pancreatic lesions in this study (10.2 ± 3.5, 95% CI: 9.320–11.126 vs. 6.1 ± 3.8, 95% CI: 5.092–7.152; p = 0.001) and served as a positive predictor of the primary tumor (SUVmax: OR = 1.951, 95% CI: 1.286–2.960, p = 0.004) [23]. This is attributed to the intense fibrotic response (resulting in increased tumor uptake) and the absence of interfering physiological uptake (leading to low background uptake) in metastatic lesions. The critical role of [68Ga]Ga-FAPI-04 PET/CT in the early detection and precise localization of metastatic lesions facilitates timely intervention and treatment planning. However, postoperative [68Ga]Ga-FAPI-04 uptake in benign lesions such as fibrosis, inflammatory reactions, tuberculous peritonitis, and fibrous tissue hyperplasia of the greater omentum is often misdiagnosed as peritoneal metastase [32–34]. Overall, [68Ga]Ga-FAPI-04 PET/CT can be a valuable complement to [18F]FDG PET/CT.
Compared with [18F]FDG PET/CT, [68Ga]Ga-FAPI-04 PET/CT has a higher diagnostic accuracy for TNM staging in patients with pancreatic cancer.[68Ga]Ga-FAPI-04 PET/CT ultimately upgraded the clinical staging of seven patients and changed the treatment plan of six patients due to its ability to show more metastatic lesions, especially in identifying lymph node, visceral, and bone metastases. Thus, [68Ga]Ga-FAPI-04 PET/CT demonstrated superior diagnostic efficacy in the staging of pancreatic cancer patients and can be used in clinical applications to evaluate primary tumor and recurrent metastases in pancreatic cancer patients.
Subgroup analysis by prior treatment status demonstrated that the diagnostic superiority of [68Ga]Ga-FAPI-04 over [18F]FDG was consistent across both initial consultation and previously treated patients, with no significant interaction between tracer and treatment status. This suggests that prior therapy does not substantially confound the comparative performance of the two imaging modalities.
This study has several limitations. First, this was a single-center prospective analysis which may have introduced selection bias. Second, certain subgroups (e.g., recurrent lesions) had small sample sizes.Histopathological grading was not available for all patients, and unable to stratify by tumor differentiation which may limit the generalizability of our findings. Third, not all lesions were pathologically confirmed, and the follow-up period was relatively short(3 months).
Conclusion
In summary, compared with [18F]FDG PET/CT, [68Ga]Ga-FAPI-04 PET/CT showed potential advantages in the tracer uptake and detection of pancreatic cancer tumor, suggesting its clinical value. Future studies including randomized controlled trials with longer follow-up periods and larger, multicenter cohorts are warranted to validate the diagnostic superiority of [68Ga]Ga-FAPI-04 PET/CT versus [18F]FDG PET/CT in pancreatic cancer.
Electronic Supplementary Material
Below is the link to the electronic supplementary material.
Acknowledgements
We thank the patients and their families for participating in this study.
Abbreviations
- PC
Pancreatic cancer
- CAFs
Cancer-associated fibroblasts
- FAP
Fibroblast activating protein
- FAPI
Fibroblast activating protein inhibitors
- [68Ga]Ga-FAPI-04
[68Ga]Ga- fibroblast activating protein inhibitors 04
- [18F]FDG
[18F]fluorodeoxyglucose
- AJCC
American Joint Committee on Cancer
- DOTA
1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid
- RP-HPLC
Reversed-phase high-pressure liquid chromatography
- PET/CT
Positron emission tomography computed tomography
- SUVmax
Maximum standardized uptake value
- SUVmean
Mean standardized uptake value
- TBR
Tumor-to-mediastinal blood pool ratio
- TLR
Tumour-to-normal liver parenchyma ratio
- LNR
Lymph node ratio
- ROC
Receiver Operating Characteristic curve
- AUC
Area Under the Curve
- CI
Confidence Interval
- Sen
Sensitivity
- Spe
Specificity
- Acc
Accuracy
- OR
Odds Ratio
- ICC
Intraclass Correlation Coefficient
- T/B Ratio
Tumor-to-Background ratio
Author contributions
All authors contributed to the study conception and design. Material preparation and data collection were performed by RXY, YHS, RA, MJW, FYS, BG. Data analysis was performed by RXY, XLC, XP. RXY, JMZ, JFW, JYH, TM were responsible for the concept and design of the study. The first draft of the manuscript was written by RXY and all authors commented on previous versions of the manuscript. Ruoxi Yang revised it and XMZ supervised the study. All authors read and approved the final manuscript.
Funding
This work was funded by the Hebei Province Government Funded clinical Medicine Excellent Talent Training Project (No. ZF2024098). No other potential conflict of interest relevant to this article was reported.
Data availability
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
Ethical approval was obtained Institutional Ethics Committee of the Fourth Hospital of Hebei Medical University. (Ethics committee approval No. 2021069)Informed consent was obtained from the patient included in this study.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Partyka O, Pajewska M, Kwaśniewska D, et al. Overview of Pancreatic Cancer Epidemiology in Europe and Recommendations for Screening in High-Risk Populations. Cancers (Basel). 2023;15(14). 10.3390/cancers15143634. [DOI] [PMC free article] [PubMed]
- 2.Hu JX, Zhao CF, Chen WB, et al. Pancreatic cancer: A review of epidemiology, trend, and risk factors. World J Gastroenterol. 2021;27(27):4298–321. 10.3748/wjg.v27.i27.4298. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Stoffel EM, Brand RE, Goggins M. Pancreatic Cancer: Changing Epidemiology and New Approaches to Risk Assessment, Early Detection, and Prevention. Gastroenterology. 2023;164:752–65. 10.1053/j.gastro.2023.02.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Lindholm H, Brolin F, Jonsson C, Jacobsson H. The relation between the blood glucose level and the FDG uptake of tissues at normal PET examinations. EJNMMI Res. 2013;3:50. 10.1186/2191-219X-3-50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kiani M, Jokar S, Hassanzadeh L, et al. Recent Clinical Implications of FAPI: Imaging and Therapy. Clin Nucl Med. 2024;49:e538–56. 10.1097/RLU.0000000000005348. [DOI] [PubMed] [Google Scholar]
- 6.Cannon A, Thompson CM, Bhatia R, et al. Molecular mechanisms of pancreatic myofibroblast activation in chronic pancreatitis and pancreatic ductal adenocarcinoma. J Gastroenterol. 2021;56(8):689–703. 10.1007/s00535-021-01800-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Shen J, Wu R, Yin T, et al. Harnessing the power of cancer-associated fibroblasts to revolutionize pancreatic cancer treatment. Cancer Biol Med. 2025;22(12):1473–92. 10.20892/j.issn.2095-3941.2025.0288. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Li H, Liu D, Li K, et al. Pancreatic stellate cells and the interleukin family: Linking fibrosis and immunity to pancreatic ductal adenocarcinoma (Review). Mol Med Rep. 2024;30(3). 10.3892/mmr.2024.13283. [DOI] [PMC free article] [PubMed]
- 9.Pijl JP, Nienhuis PH, Kwee TC, et al. Limitations and Pitfalls of FDG-PET/CT in Infection and Inflammation. Semin Nucl Med. 2021;51(6):633–45. 10.1053/j.semnuclmed.2021.06.008. [DOI] [PubMed] [Google Scholar]
- 10.Hosein AN, Brekken RA, Maitra A. Pancreatic cancer stroma: an update on therapeutic targeting strategies. Nat Rev Gastroenterol Hepatol. 2020;17:487–505. 10.1038/s41575-020-0300-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Chun YS, Pawlik TM, Vauthey JN. 8th Edition of the AJCC Cancer Staging Manual: Pancreas and Hepatobiliary Cancers. Ann Surg Oncol. 2018;25:845–7. 10.1245/s10434-017-6025-x. [DOI] [PubMed] [Google Scholar]
- 12.Matsumoto I, Shirakawa S, Shinzeki M, et al. 18-Fluorodeoxyglucose positron emission tomography does not aid in diagnosis of pancreatic ductal adenocarcinoma. Clin Gastroenterol Hepatol. 2013;11:712–8. 10.1016/j.cgh.2012.12.033. [DOI] [PubMed] [Google Scholar]
- 13.Andersen DK, Korc M, Petersen GM, et al. Diabetes, Pancreatogenic Diabetes, and Pancreatic Cancer. Diabetes. 2017;66:1103–10. 10.2337/db16-1477. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Pang Y, Zhao L, Shang Q, et al. Positron emission tomography and computed tomography with [68Ga]Ga-fibroblast activation protein inhibitors improves tumor detection and staging in patients with pancreatic cancer. Eur J Nucl Med Mol Imaging. 2022;49:1322–37. 10.1007/s00259-021-05576-w. [DOI] [PubMed] [Google Scholar]
- 15.Erkan M, Hausmann S, Michalski CW, et al. The role of stroma in pancreatic cancer: diagnostic and therapeutic implications. Nat Rev Gastroenterol Hepatol. 2012;9:454–67. 10.1038/nrgastro.2012.115. [DOI] [PubMed] [Google Scholar]
- 16.Meyer C, Dahlbom M, Lindner T, et al. Radiation Dosimetry and Biodistribution of 68Ga-FAPI-46 PET Imaging in Cancer Patients. J Nucl Med. 2020;61:1171–7. 10.2967/jnumed.119.236786. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Zboralski D, Hoehne A, Bredenbeck A, et al. Preclinical evaluation of FAP-2286 for fibroblast activation protein targeted radionuclide imaging and therapy. Eur J Nucl Med Mol Imaging. 2022;49:3651–67. 10.1007/s00259-022-05842-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Siveke JT. Fibroblast-Activating Protein: Targeting the Roots of the Tumor Microenvironment. J Nucl Med. 2018;59:1412–4. 10.2967/jnumed.118.214361. [DOI] [PubMed] [Google Scholar]
- 19.Kratochwil C, Flechsig P, Lindner T, et al. 68Ga-FAPI PET/CT: Tracer Uptake in 28 Different Kinds of Cancer. J Nucl Med. 2019;60:801–5. 10.2967/jnumed.119.227967. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Chen H, Pang Y, Wu J, et al. Comparison of [68Ga]Ga-DOTA-FAPI-04 and [18F] FDG PET/CT for the diagnosis of primary and metastatic lesions in patients with various types of cancer. Eur J Nucl Med Mol Imaging. 2020;47:1820–32. 10.1007/s00259-020-04769-z. [DOI] [PubMed] [Google Scholar]
- 21.Chen H, Zhao L, Ruan D, et al. Usefulness of [68Ga]Ga-DOTA-FAPI-04 PET/CT in patients presenting with inconclusive [18F]FDG PET/CT findings. Eur J Nucl Med Mol Imaging. 2021;48(1):73–86. 10.1007/s00259-020-04940-6. [DOI] [PubMed] [Google Scholar]
- 22.Sho M, Murakami Y, Motoi F, et al. Postoperative prognosis of pancreatic cancer with para-aortic lymph node metastasis: a multicenter study on 822 patients. J Gastroenterol. 2015;50:694–702. 10.1007/s00535-014-1005-4. [DOI] [PubMed] [Google Scholar]
- 23.Zhao Y, Wang S, Du B, et al. Head-to-head Comparison of [(68)Ga]Ga-FAPI-04 PET and [(18)F]FDG PET in the Evaluation of Primary Gastrointestinal Tumors and Metastases: A Systematic Review and Meta-analysis. Acad Radiol. 2025;32(11):6857–72. 10.1016/j.acra.2025.05.053. [DOI] [PubMed] [Google Scholar]
- 24.Ding J, Qiu J, Hao Z, et al. Comparing the clinical value of baseline [68Ga]Ga-FAPI-04 PET/CT and [18F]F-FDG PET/CT in pancreatic ductal adenocarcinoma: additional prognostic value of the distal pancreatitis. Eur J Nucl Med Mol Imaging. 2023;50:4036–50. 10.1007/s00259-023-06297-y. [DOI] [PubMed] [Google Scholar]
- 25.Chen R, Yang X, Yu X, et al. Tumor-to-blood ratio for assessment of fibroblast activation protein receptor density in pancreatic cancer using [68Ga]Ga-FAPI-04. Eur J Nucl Med Mol Imaging. 2023;50(3):929–36. 10.1007/s00259-022-06010-5. [DOI] [PubMed] [Google Scholar]
- 26.Sahani DV, Bonaffini PA, Catalano OA, Guimaraes AR, Blake MA. State-of-the-art PET/CT of the pancreas: current role and emerging indications. Radiographics. 2012;32:1133–60. 10.1148/rg.324115143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Kato K, Nihashi T, Ikeda M, et al. Limited efficacy of (18)F-FDG PET/CT for differentiation between metastasis-free pancreatic cancer and mass-forming pancreatitis. Clin Nucl Med. 2013;38:417–21. 10.1097/RLU.0b013e3182817d9d. [DOI] [PubMed] [Google Scholar]
- 28.Özdemir BC, Pentcheva-Hoang T, Carstens JL, et al. Depletion of carcinoma-associated fibroblasts and fibrosis induces immunosuppression and accelerates pancreas cancer with reduced survival. Cancer Cell. 2014;25:719–34. 10.1016/j.ccr.2014.04.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Low G, Panu A, Millo N, Leen E. Multimodality imaging of neoplastic and nonneoplastic solid lesions of the pancreas. Radiographics. 2011;31:993–1015. 10.1148/rg.314105731. [DOI] [PubMed] [Google Scholar]
- 30.Liu Q, Shi S, Liu S, et al. The added value of [68Ga]Ga-DOTA-FAPI-04 PET/CT in pancreatic cancer: a comparison to [18F]F-FDG. Eur Radiol. 2023;33:5007–16. 10.1007/s00330-023-09445-y. [DOI] [PubMed] [Google Scholar]
- 31.Koerber SA, Röhrich M, Walkenbach L, et al. Impact of 68Ga-FAPI PET/CT on Staging and Oncologic Management in a Cohort of 226 Patients with Various Cancers. J Nucl Med. 2023;64:1712–20. 10.2967/jnumed.123.266046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Loktev A, Lindner T, Mier W, et al. A Tumor-Imaging Method Targeting Cancer-Associated Fibroblasts. J Nucl Med. 2018;59:1423–9. 10.2967/jnumed.118.210435. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Zhao L, Pang Y, Wei J, Hao B, Chen H. Use of 68Ga-FAPI PET/CT for Evaluation of Peritoneal Carcinomatosis Before and After Cytoreductive Surgery. Clin Nucl Med. 2021;46:491–3. 10.1097/RLU.0000000000003611. [DOI] [PubMed] [Google Scholar]
- 34.Alçın G, Tatar G, Şahin R, Baloğlu MC, Çermik TF. Peritoneal Tuberculosis Mimicking Peritoneal Carcinomatosis on 68 Ga-FAPI-04 and 18 F-FDG PET/CT. Clin Nucl Med. 2022;47:e557–8. 10.1097/RLU.0000000000004174. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.