TY - JOUR T1 - PET Imaging to Determine HGF Levels in Tumors for Enhanced Patient Selection JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 56 LP - 56 VL - 57 IS - supplement 2 AU - Eric Price AU - Kimberly Edwards AU - Kathryn Carnazza AU - Sean Carlin AU - Kuntal Sevak AU - Yelena Janjigian AU - Jason Lewis Y1 - 2016/05/01 UR - http://jnm.snmjournals.org/content/57/supplement_2/56.abstract N2 - 56Objectives The antibody rilotumumab (AMG102) has been in clinical trials for treatment of several cancers, including gastric, lung, and brain. AMG102 is a fully human antibody that binds hepatocyte growth factor (HGF), preventing it from binding its receptor (MET) and therefore providing therapeutic benefit. It may be that a patient must have elevated local levels of HGF in tumors to benefit from HGF-targeted treatment, and clinical trials to date have had mixed results, perhaps due to poor patient selection due to only using serum HGF and/or tumor MET expression. As a tool for patient selection in precision medicine, we have created a diagnostic from AMG102 by conjugating the chelator desferrioxamine (DFO) and radiolabeling with the PET imaging isotope 89Zr (t1/2 = ~3.3 d). 89Zr-DFO-AMG102 will be tested in U87MG (HGF+) and MKN45 (HGF-) xenografts, and then in MET+ patient derived xenograft (PDX) models with unknown levels of HGF expression.Methods The chelator p-SCN-Bn-DFO was conjugated to AMG102 and purified. DFO-AMG102 was then radiolabeled with 89Zr with high radiochemical yields (>98%) and purity (>99%). The serum stability was > 98% stable after 7 days (37°C), and HGF binding affinity by ELISA was similar to unmodified AMG102. The brain cancer cell line U87MG was used for proof-of-principle as it has been previously used to evaluate AMG102 therapeutically and is known to express high levels of HGF and MET (autocrine production). Mice were xenografted with U87MG and MKN45 cells (n = 5 per group), and 89Zr-DFO-AMG102 was injected (~150 μCi, ~35 μg antibody) and imaged by PET from 24 h to 120 h p.i. Biodistribution experiments were performed with 89Zr-DFO-AMG102 (~ 30 μCi, ~5 μg) in both U87MG (MET+, HGF+) and MKN45 (control, MET+, HGF-) xenograft models, and as an additional control the non-specific antibody 89Zr-DFO-IgG was used. PDX mouse models (DY, DC, EK, DF) were then injected with 89Zr-DFO-AMG102), and then imaged by PET, with biodistributions being performed at 120 h. Autoradiography and histology was performed on tumor samples.Results 89Zr-DFO-AMG102 substantially accumulated in U87MG xenografts (high autocrine HGF expression), with biodistribution data showing high tumor uptake of 36.8 ± 7.8 %ID/g at 120 h p.i. (T/blood = 2.4 ± 0.6, T/liver = 6.6 ± 1.5 %ID/g). These studies demonstrated high uptake in U87MG xenografts (MET+, HGF+), which was attenuated by a blocking dose of ”cold” AMG102 (500 μg). A control of 89Zr-DFO-IgG in U87MG xenografts was substantially lower (11.5 ± 3.3 %ID/g at 120 h), and 89Zr-DFO-AMG102 in MKN45 xenografts (MET+, HGF-) was also much lower (4.5 ± 1.0 %ID/g at 120 h), showing selective uptake in tumors with high HGF levels. A dramatic therapeutic effect in U87MG xenografts was observed, with a blocking dose of cold AMG102 (500 μg) shrinking tumors (0.06 ± 0.03 g), a ~5 μg injection of 89Zr-DFO-AMG102 slowing growth (0.21 ± 0.05 g), and 89Zr-DFO-IgG having no noticeable effect on growth (0.41 ± 0.11 g) (also no effect on growth in HGF- MKN45 tumors). Uptake in mice bearing 4 different MET+ gastric PDX (DY, DC, EK, DF) was between 4-7 % ID/g (low, similar to controls), suggesting little or no HGF present in these tumors and that these patients might be poor candidates for HGF targeted therapies.Conclusions We have successful synthesized 89Zr-DFO-AMG102 and shown selective uptake in U87MG xenografts with high local HGF levels, and low uptake in MKN45 (MET+, HGF-) xenografts, demonstrating the potential of this PET agent for non-invasively determining local HGF levels in tumors. PET imaging in 4 different gastric PDX models had low tracer uptake, revealing little or no HGF present in these tumors, suggesting that these PDX tumors would be poor candidates for HGF-targeted therapies. This work shows the potential and utility of 89Zr-DFO-AMG102 to provide valuable clinical information for patient selection of MET/HGF positive tumors for HGF targeted therapies, and to non-invasively identify tumors with autocrine HGF/MET loops and high local levels of HGF. $$graphic_AD064231-24F9-48DF-AEB6-124CC09F4B74$$ ER -