This study was part of a prospective, single-site phase 0 clinical trial to demonstrate the feasibility of imaging PARP1 expression with FTT-PET/CT in cancer patients (NCT02469129, IND 124,116). The study was approved by all institutional regulatory committees and performed according to the Declaration of Helsinki and the rules of Good Clinical Practice guidelines. All subjects gave written informed consent for participation. Trial eligibility for this investigational imaging study was broad; patients with any type of cancer planned to undergo surgical resection/biopsy, or who were planned to receive systemic therapy, were eligible. Subjects were required to have measurable disease per RECIST 1.1 on CT or other standard radiological imaging. Measurable disease was defined as one or more tumor sites measuring > 1 cm in the shortest transaxial diameter. Subjects with disease < 1 cm were ineligible. Patients who had claustrophobia or who were unable to lie in the PET scanner were also excluded. Vital signs and laboratory evaluations (CBC, CMP) were performed before and after completion of FTT-PET/CT for assessment of any clinically significant change that may have been related to FTT administration. Subjects also were closely monitored during the study for any symptoms that may have been related to FTT injection.

For the present study, patients were consented under the NCT02469129 protocol and were recruited from advanced PC patients in our clinical practice who were either treatment naïve or were scheduled to start a new treatment regimen (Fig. 1). Four patients had metastatic hormone-sensitive prostate cancer (mHSPC); two patients had de novo metastatic disease, and two had recurrent metastatic disease. The remaining 5 patients had metastatic castration-resistant disease (mCRPC) (Table 1). All patients underwent standard of care bone scintigraphy and computed tomography (CT) at baseline prior to enrollment for measurable disease assessment prior study enrollment. All patients also underwent [18F]fluorodeoxyglucose (FDG)-PET/CT) before or after FTT-PET/CT (mean interval 3.6 days). Subjects were encouraged to undergo follow-up imaging after therapy (minimum of 3 months after therapy), but in our cohort, only three patients (#3, 6 and 8) underwent FTT-PET/CT after therapy (Table 1).

Study schema. mHSPC = metastatic hormone sensitive prostate cancer, mCRPC = metastatic castration resistant prostate cancer.

Baseline metastatic biopsy was not required for study entry, and 7 of 9 patients had metastatic tissue available for genomic analysis. Standard of care next-generation sequencing (NGS) was performed utilizing the Tempus xT 648 platform with matched normal blood analysis for germline testing (Tempus Labs; Chicago, IL, USA, https://www.tempus.com; Table 1). Two of the 9 patients did not have metastatic tissue available and thus underwent peripheral blood analysis for germline genetic analysis with the Ambry and Myriad platforms, respectively. One patient also underwent additional ctDNA testing with Guardant 360 in addition to tissue NGS at a separate timepoint prior to receiving care at our institution (Table 1). Four of the 9 patients had sufficient tissue for PARP1 expression analysis via immunohistochemistry (IHC). For this study, we classified BRCA1/2 and ATM mutations, the genomic aberrations most associated with response to PARPi, as “MUT,” and all other mutations as “WT”.

FTT was prepared as previously described in our cyclotron GMP facility using the GE FXNPro module under current good manufacturing practice (cGMP) guidelines [14]. Briefly, FTT was synthesized under cGMP conditions using a GE FXN Pro reaction module via nucleophilic substitution of a tosylate precursor with a solution of pre-dried [18F]KF/K2CO3/Kryptofix2.2.2 in DMSO at 105 °C. The crude FTT was then purified via a reversed phase HPLC (Zorbax SB-C18, 250 × 9.4 mm, mobile phase 15% acetonitrile (0.1% TFA) in water). The isolated peak was diluted in 40 mL of sterile water for injection (SWI), trapped on an HLB cartridge, eluted from the cartridge using 0.9 mL of ethanol (200 Proof), and formulated in normal saline to provide a 10% ethanol/saline solution. This was sterile filtered through a 0.2-micron Millex-GV filter directly into a sterile final product vial. The final product was analyzed by HPLC, GC, pH, filter membrane integrity test, appearance, color, radiochemical identity, bacterial endotoxin test, and sterility was confirmed as a post release test. All release criteria were met for all injections [14, 16]. The final product meets the releasing criteria approved by FDA.

Patients underwent a 60-min dynamic study immediately following intravenous administration of FTT (median 370 MBq, range 336.7–388.5) over the largest and/or the previously biopsied lesion. Emission imaging were then obtained extending from the skull base to the upper thighs. FTT-PET images were evaluated for areas of abnormal tracer uptake in comparison with anatomical, bone scintigraphy and FDG-PET/CT images using tumor maximum standardized uptake value (SUVmax). For each of the 9 patients, the SUVs of 5 lesions (typically, the largest) were determined, and a total of 45 different disease sites were evaluated.

Biopsy samples were analyzed for PARP1 using an immunohistochemistry (IHC) staining protocol previously described by our group with modifications [17]. Briefly, formalin-fixed paraffin-embedded tissue Sects. (5 µm) were deparaffinized, rehydrated, and subjected to antigen retrieval (citrate buffer, pH 6.0). The sections were blocked (10% normal goat serum), and immunostained with PARP1 (46D11) rabbit antibody (1:300; #9532, Cell Signaling Technologies, USA) followed by ImmPRESS goat anti-rabbit secondary antibody (Vector Laboratories, USA). Diaminobenzidine (Agilent Dako, USA) was used as a chromogen, and sections were counterstained with hematoxylin, dehydrated, cleared, and mounted with Cytoseal XYL (Thermo Scientific, USA). Images were captured using a BX51 microscope (Olympus, Japan). Scoring was performed by an experienced genitourinary pathologist blinded to the clinicopathologic data of the patients. Scoring was adapted from the Allred model of breast cancer estrogen receptors/progesterone receptor staining, whereby an IHC score is calculated as the sum of the positive proportion (0 = 0% positive tumor cells; 1 =  < 1%; 2 = 1–10%; 3 = 11–33%; 4 = 34–66%; 5 = 67–100%) and the staining intensity (0 = no staining; 1 = weak; 2 = moderate; 3 = strong) for a possible total score of 8. PARP1 was considered positive if the IHC score was greater than 4.

Demographic and clinical characteristics were summarized by descriptive statistics. The linear mixed effects model for repeated measures was applied, separately to FTT SUVmax in raw and log scales, to model the effect of gene mutation while accounting for the repeated measures of tumor SUVmax for the same patient. The model residuals were examined for goodness of fit, which showed better model fitting using the SUVmax in log scale. The least square mean for MUT and WT was reported with associated standard error and Wald test P value testing the estimated mean against 0. The least square mean difference between MUT (n = 5) and WT (n = 4) patients was reported with associated standard error, the Wald test p value testing the mean difference against 0, and the type III test p value comparing the model with mutation versus the null model without.