[18F] AlF-NOTA-FAPI-04 PET/CT can predict treatment response and survival in patients receiving chemotherapy for inoperable pancreatic ductal adenocarcinoma

Abstract

Purpose: We investigated whether uptake of [¹⁸F] AlF-NOTA-FAPI-04 on positron emission tomography/computed tomography (PET/CT) could predict treatment response and survival in patients with pancreatic ductal adenocarcinoma (PDAC).

Methods: We prospectively evaluated 47 patients with histopathologically confirmed primary PDAC who provided pretreatment [¹⁸F] AlF-NOTA-FAPI-04 scans to detect fibroblast activation protein (FAP) on the tumor surface by uptake of [¹⁸F] AlF-NOTA-FAPI-04. PDAC specimens were immunohistochemically stained with cancer-associated fibroblast (CAF) markers. We obtained a second PET scan after one cycle of chemotherapy to study changes in FAPI uptake variables from before to during treatment. Correlations between baseline PET variables and CAF-related immunohistochemical markers were assessed with Spearman's rank test. Cox regression and Kaplan-Meier methods were used to assess relationships between disease progression and potential predictors. Receiver operating characteristic (ROC) curve analysis was used to define the optimal cut-off points for distinguishing patients according to good response vs. poor response per RECIST v.1.1.

Results: The FAPI PET variables maximum and mean standardized uptake values (SUV_max, SUV_mean), metabolic tumor volume (MTV), and total lesion FAP expression (TLF) were positively correlated with CAF markers (FAP, α-smooth muscle actin, vimentin, S100A4, and platelet-derived growth factor receptor α/β, all P < 0.05). MTV was associated with survival in patients with inoperable PDAC (all P < 0.05). Cox multivariate regression showed that MTV was associated with overall survival (MTV hazard ratio [HR] = 1.016, P = 0.016). Greater changes from before to during chemotherapy in SUV_max, MTV, and TLF were associated with good treatment response (all P < 0.05). ΔMTV, ΔTLF, and ΔSUV_max had larger areas under the curve than ΔCA19-9 for predicting treatment response. Kaplan-Meier analysis showed that the extent of change in MTV and TLF from before to after treatment predicted progression-free survival, with cut-off values (based on medians) of - 4.95 for ΔMTV (HR = 8.09, P = 0.013) and - 77.83 for ΔTLF (HR = 4.62, P = 0.012).

Conclusions: A higher baseline MTV on [¹⁸F] AlF-NOTA-FAPI-04 scans was associated with poorer survival in patients with inoperable PDAC. ΔMTV was more sensitive for predicting response than ΔCA19-9. These results are clinically meaningful for identifying patients with PDAC who are at high risk of disease progression.

Keywords: Fibroblast activation protein; Metabolic tumor volume; Pancreatic adenocarcinoma; Total lesion FAP expression.