Imaging HER2-positive metastatic esophagogastric cancer with 89Zr-trastuzumab PET and 18F-FDG PET.

Abstract

4068 Background: Variations in HER2 expression between primary tumor and metastases may contribute to drug resistance in HER2-positive mEG cancer. Whole body imaging with 89Zr-trastuzumab PET has a potential advantage over single site biopsies as it can non-invasively assess variations in HER2 expression and target engagement. 89Zr-trastuzumab PK, biodistribution and dosimetry in mEG cancer were previously published by our group (O’Donoghue, JNM 2018). We now present lesion level analysis of baseline 18F-FDG-PET and CT in comparison with baseline 89Zr-trastuzumab imaging. Methods: Patients with metastatic HER2-positive (IHC 3+, IHC 2+/FISH > 2.0) mEG cancer and RECIST 1.1 measurable disease were consented and imaged with 89Zr-trastuzumab PET, 18F-FDG-PET and CT. All visualized lesions (maximum 5 per organ) were annotated in detail using RECIST 1.1 measurements (CT) and maximum standard uptake values (SUVmax) on 89Zr-trastuzumab and 18F-FDG PET scans. Correlation of visualized disease burden between imaging modalities with clinical and pathologic characteristics was performed. Results: 33 patients with mEG adenocarcinoma were imaged with 89Zr-trastuzumab PET, 18F-FDG-PET and CT (12% esophageal, 64% GEJ, 24% gastric). HER2 status (IHC 3+ 70%; IHC2+/FISH+ 27%, NGS 3%) was assessed from biopsy of primary (66.7%) or metastasis (33.3%). Median time from diagnosis to 89Zr-trastuzumab PET was 12.6 months. At the time of 89Zr-trastuzumab PET, 39% were treatment naive, while 61% had received prior therapy including trastuzumab (58%) and afatinib (median lines 2, range 0-6). 82% of patients had the primary tumor in place and all patients had metastatic disease at the time of enrollment with metastases to lymph nodes (70%) peritoneum (24%), liver (58%), lung (33%), and/or bone (9%). Median number of RECIST 1.1 lesions measured on baseline CT was 6 (range 1-15). PET analysis is complete in 18 of 33 patients. Median number of 89Zr-trastuzumab PET positive (SUV > 3) lesions was 5.5 (range 1-11) with 5 patients having at least 1 (max 3) very intense lesion (SUV > 15). Median number of 18F-FDG-PET-positive lesions was 8 (range 1-14). Three patients had at least 1 lesion positive on 89Zr-trastuzumab-PET and negative on 8F-FDG-PET (range 0-5), and 8 patients had at least 1 lesion positive on 18F-FDG-PET and negative on 89Zr-trastuzumab-PET (range 0-7). Of the total lesions identified on 89Zr-PET but not CT, 40% were in bone. Of those identified on 89Zr but not 8F-FDG-PET, 40% were in bone, and of those on 18F-FDG- but not 89Zr-PET, 6% were in bone. Conclusions: 89Zr-trastuzumab PET effectively identifies heterogeneity of HER2 expression and allows assessment of lesions throughout the whole body. Bone lesions were better identified and appeared earlier on 89Zr-trastuzumab PET compared with CT. Additional analyses including correlation of baseline HER2 heterogeneity with overall and disease-free survival will be presented. Clinical trial information: NCT02023996.