Abstract
2486
Introduction: PARP inhibitors (PARPi) are approved for maintenance therapy and treatment of recurrent ovarian cancer. Patient selection for these agents leverage known associations with DNA repair defects or platinum sensitivity. However, despite stringent selection criteria, response to these agents remain variable, highlighting the need for better biomarkers that predict response to PARPi. To address this need, the PET agent [18F]FluorThanatrace ([18F]FTT), a measure of PARP-1 expression (the pharmacologic target of PARPi) and drug-target engagement, is being evaluated as a PET biomarker to predict response to PARPi therapy in ovarian cancer.
Methods: This single-center, prospective, cohort, IRB-approved study enrolled women with known or suspected epithelial ovarian, fallopian tube, or primary peritoneal cancer. Subjects enrolled in two concomitant trials studying the PARPi, olaparib, in subjects with ovarian cancer were specifically targeted for recruitment. Off-trial patients treated with other PARPi were also included. [18F]FTT PET/CT imaging was obtained from the skull base to the proximal thighs on an Ingenuity TF scanner (Philips Healthcare, Andover, MA) 60-90 minutes after intravenous injection of 8-12 mCi [18F]FTT. Imaging was obtained at baseline prior to the initiation of PARPi therapy, and ~1 week after the initiation of PARPi therapy. Target lesions (primary tumors and/or metastases) were identified at the time of the baseline imaging on correlative anatomic imaging using RECIST 1.1. Maximum standardized uptake value (SUVmax) data, normalized by body weight, was collected for target lesions. Progression-free survival (PFS) was calculated as the time from starting a PARPi to the time treatment stopped secondary to progression or death, or rise in CA-125 determined as progression by the provider.
Results: 12 subjects had a baseline [18F]FTT-PET imaging; 9 of these subjects had a second [18F]FTT-PET a median of 7 days (range 6-14 days) after starting a PARPi. Measurable target lesions (median 1.5; range 1-4) were identified per patient on anatomic imaging by CT per RECIST 1.1. First anatomic response follow-up assessment by CT was at a median 97 days (range 57-158 days) from baseline [18F]FTT-PET. By RECIST, of 12 evaluable patients, one patient achieved a Complete Response and three patients a Partial Response prior to disease progression with overall response rate of 33%. PFS had a median of 5.3 months (range 3-24 months). In subjects with paired [18F]FTT-PET scans, individual lesions that decreased in size at first CT assessment had at least a 40% decrease in SUVmax between PET scans (n=6), although not all lesions that had a >40% decrease on PET decreased in size (6 lesions decreased in size, 2 had no change, 3 increased in size). On a subject level, subjects who achieved a greater than 50% decrease in SUVmax on the second scan after PARPi initiation had a PFS >6 months (n=4), while those with less than a 50% decrease had a PFS < 6 months (n=5). All patients with a >50% decrease in SUVmax (n=4) also achieved a greater than 50% decrease in CA-125; all patients who did not achieve a 50% decrease in SUVmax (n=3) did not achieve such a CA-125 response. Utilizing baseline FTT imaging for PARP-1 target expression only did not prove useful in predicting decrease in lesion size by CT, PFS, or decrease in CA-125.
Conclusions: Our study shows that the percent decrease in [18F]FTT SUVmax after PARPi, a measure of drug target engagement, is associated with response to PARPi in a limited sample size warranting further study. Larger studies are needed to corroborate these results.