Evaluation of FAPI PET imaging in gastric cancer: a systematic review and meta-analysis

Abstract

Purpose: Recent studies suggest that ⁶⁸Ga-FAPI PET/CT demonstrated superiority over ¹⁸F-FDG PET/CT in the evaluation of various cancer types, especially in gastric cancer (GC). By comprehensively reviewing and analysing the differences between ⁶⁸Ga-FAPI and ¹⁸F-FDG in GC, some evidence is provided to foster the broader clinical application of FAPI PET imaging. Methods: In this review, studies published up to July 3, 2023, that employed radionuclide labelled FAPI as a diagnostic radiotracer for PET in GC were analysed. These studies were sourced from both the PubMed and Web of Science databases. Our statistical analysis involved a bivariate meta-analysis of the diagnostic data and a meta-analysis of the quantitative metrics. These were performed using R language. Results: The meta-analysis included 14 studies, with 527 patients, of which 358 were diagnosed with GC. Overall, ⁶⁸Ga-FAPI showed higher pooled sensitivity (0.84 [95% CI 0.67-0.94] vs. 0.46 [95% CI 0.32-0.60]), specificity (0.91 [95% CI 0.76-0.98] vs. 0.88 [95% CI 0.74-0.96]) and area under the curve (AUC) (0.92 [95% CI 0.77-0.98] vs. 0.52 [95% CI 0.38-0.86]) than ¹⁸F-FDG. The evidence showed superior pooled sensitivities of ⁶⁸Ga-FAPI PET over ¹⁸F-FDG for primary tumours, local recurrence, lymph node metastases, distant metastases, and peritoneal metastases. Furthermore, ⁶⁸Ga-FAPI PET provided higher maximum standardized uptake value (SUVmax) and tumour-to-background ratios (TBR). For bone metastases, while ⁶⁸Ga-FAPI PET demonstrated slightly lower patient-based pooled sensitivity (0.93 vs. 1.00), it significantly outperformed ¹⁸F-FDG in the lesion-based analysis (0.95 vs. 0.65). However, SUVmax (mean difference [MD] 1.79 [95% CI -3.87-7.45]) and TBR (MD 5.01 [95% CI -0.78-10.80]) of bone metastases showed no significant difference between ⁶⁸Ga-FAPI PET/CT and ¹⁸F-FDG PET/CT. Conclusion: Compared with ¹⁸F-FDG, ⁶⁸Ga-FAPI PET imaging showed improved diagnostic accuracy in the evaluation of GC. It can be effectively applied to the early diagnosis, initial staging, and detection of recurrence/metastases of GC. ⁶⁸Ga-FAPI may have the potential of replacing ¹⁸F-FDG in GC in future applications.

Keywords: 18F-FDG; 68Ga-FAPI; PET/CT; fibroblast activation protein; gastric cancer.

Figure 1

Flowchart of the study inclusion process.

Figure 2

Summary assessment of risk of bias and applicability concerns of included studies.

Figure 3

Forest plots of diagnostic sensitivity and specificity of ⁶⁸Ga-FAPI (A) and ¹⁸F-FDG (B) (Data sets extracted from patient-based studies: ^aPrimary tumour, ^brecurrent tumour, ^cmetastatic lymph nodes; Data sets extracted from lesion-based studies: ^dmetastatic lymph nodes, ^edistant metastases).

Figure 4

Summary receiver-operation characteristic (SROC) curves for the overall performance assessment of ⁶⁸Ga-FAPI (A) and ¹⁸F-FDG (B) for GC (gastric cancer).

Figure 5

Forest plots comparing the uptake values of ⁶⁸Ga-FAPI and ¹⁸F-FDG (SUVmax) for primary and recurrent tumours (A), metastatic lymph nodes (B), distant metastases (C), bone metastases (D), and peritoneal metastases (E). There was significant heterogeneity when the SUVmax data sets for primary and recurrent tumours (I² = 74%), metastatic lymph nodes (I² = 96%), distant metastases (I² = 94%), and bone metastases (I² = 95%) were pooled, so a random-effects model was used. Heterogeneity was not apparent when SUVmax data sets were pooled for peritoneal metastases (I² = 24%), so a fixed-effects model was used.

Figure 6

Forest plot comparing the TBR of primary and recurrent tumours (A), metastatic lymph nodes (B), distant metastases (C), bone metastases (D), and peritoneal metastases (E) between ⁶⁸Ga-FAPI and ¹⁸F-FDG PET imaging. There was significant heterogeneity when the TBR data sets for metastatic lymph nodes (I² = 72%), distant metastases (I² = 64%), and bone metastases (I² = 91%) were pooled, so a random-effects model was used. There was no significant heterogeneity when the SUVmax data sets for primary and recurrent tumours (I² = 24%) and peritoneal metastases (I² = 0%) were pooled, so a fixed-effects model was used.

Figure 7

Funnel plots for SUVmax-based analysis (A) and TBR-based analysis (B).

Figure 8

Representative ¹⁸F-FDG and ⁶⁸Ga-FAPI PET/CT images of primary and metastatic gastric signet ring cell carcinoma (GSRCC). (A) A 55-year-old man with GSRCC underwent ¹⁸F-FDG-PET/CT for initial staging. ¹⁸F-FDG PET/CT showed normal findings, while ⁶⁸Ga-FAPI PET/CT revealed intense uptake along the gastric wall. (B) A 40-year-old woman with GSRCC underwent PET/CT for tumour staging. ⁶⁸Ga-FAP-2286 showed higher uptake in the primary tumour than ¹⁸F-FDG. (C) A 52-year-old man with widespread subcutaneous and bone metastases underwent ¹⁸F-FDG PET/CT for localizing the primary tumour. However, no intense ¹⁸F-FDG uptake that likely presenting the primary tumour was observed. ⁶⁸Ga-FAPI PET/CT revealed intense uptake in the lesser curvature of the stomach. A subsequent gastroscopic biopsy confirmed the diagnosis of GSRCC. (D) A 49-year-old man with prior gastrectomy for GSRCC presented with progressive abdominal pain. ⁶⁸Ga-FAPI PET/CT revealed higher radiotracer uptake and larger disease extent than ¹⁸F-FDG in peritoneal metastases. Source: From Pang, Yizhen et al. (2021), Pang, Yizhen et al. (2023), and Chen, Haojun et al. (2023) with modifications.