(P 180) TUMOR VASCULAR SIGN IN THE RESPONSE ASSESSMENT OF GASTROINTESTINAL STROMAL TUMOR TREATED WITH RIPRETINIB: A MULTICENTER STUDY

Objective: To explore the occurrence and distribution of tumor angiogenesis sign (TAS) in metastatic gastrointestinal stromal tumor (GIST), and the role of TAS in tumor response assessment in GIST with ripretinib treatment.

Methods: This study followed a lesion-by-lesion-based design. 81 patients with advanced GIST from 8 medical centers were retrospectively and continuously included. All patients received ripretinib as later-line treatment after failure of imatinib, sunitinib or regorafenib. Abdominopelvic enhanced CT were analyzed at baseline and the first follow-up (2±1 months after treatment). All target lesions (the longest transverse diameter≥10mm) were observed and recorded. Lesions were classified into early progression disease(ePD) group and non-ePD group based on size (RECIST) and density (Choi) changes at 6 months after treatment. TAS were categorized into intra-tumor angiogenesis sign (iTAS) and peri-tumor angiogenesis sign (pTAS). Multivariate logistic regression analysis was employed to identify the independent predictor of poor responder, and prediction models were established. The performance of the model was evaluated by calculating the area under the receiver operating characteristic (ROC) curve (AUC).

Results: 530 metastatic lesions were analyzed. In the metastatic sites, hepatic metastases were prone to developing more iTAS than pTAS (P=0.14), while peritoneal metastases were more likely to exhibit pTAS (P=0.03). At baseline, the positive rate of iTAS was higher in ePD group than that in non-ePD group (50.4% vs. 38.4%, P=0.02), however, the positive rate of pTAS did not show significant difference between the two groups (78.8% vs. 71.2%, P=0.09). Progressive iTAS was the independent risk factor of ePD. A combined model 1 was established using iTAS and size change (formula 1 = iTAS progression ×1.39+RECIST progression × 2.17), which showed good efficacy in predicting progression at 6 months (AUC=0.78), superior to RECIST alone(AUC=0.74,P=0.003)or iTAS alone(AUC=0.65,P＜0.001). A joint model 2 was established using iTAS and density (formula 2 = iTAS progression × 1.12+ Choi progression × 2.61), presenting good efficacy in predicting 6-month progression (AUC=0.82), superior to Choi alone (AUC=0.79, P=0.006).

Conclusion:iTAS and pTAS were commonly observed in metastatic GIST. The presence of iTAS and its changes after ripretinib were associated with tumor response. The combination of iTAS with size or density may aid in the identification of progressive lesions.