Real time transferrin-based PET detects MYC-positive prostate cancer

Abstract

Non-invasive biomarkers that detect the activity of important oncogenic drivers could significantly improve cancer diagnosis and management of treatment. The goal of this study was to determine if ⁶⁸Ga-citrate (which avidly binds to circulating transferrin) labeled transferrin (Tf) can detect MYC-positive prostate cancer tumors, since the transferrin receptor is a direct MYC target gene. Positron emission tomography (PET) imaging paired with ⁶⁸Ga-citrate and molecular analysis of preclinical models, human cell-free DNA (cfDNA) and clinical biopsies were conducted to determine whether ⁶⁸Ga-citrate can detect MYC-positive prostate cancer. Importantly, ⁶⁸Ga-citrate detected human prostate cancer models in a MYC-dependent fashion. In patients with castration resistant prostate cancer (CRPC), analysis of cfDNA revealed that all patients with ⁶⁸Ga-citrate avid tumors had a gain of at least one MYC copy number. Moreover, biopsy of two PET avid metastases showed molecular or histological features characteristic of MYC hyperactivity. These data demonstrate that ⁶⁸Ga-citrate targets prostate cancer tumors with MYC hyperactivity. A larger prospective study is ongoing to demonstrate the specificity of ⁶⁸Ga-citrate for tumors with hyperactive MYC.

INTRODUCTION

The MYC oncogene is an important mediator of tumor initiation and progression in prostate cancer(1–3). In patients with metastatic castration resistant prostate cancer (mCRPC), recent data implicate MYC (and closely related oncogene MYCN) in the emergence of treatment-associated small cell/neuroendocrine prostate cancer (tSCNC). t-SCNC is a lethal disease subset that is increasing in prevalence in response to androgen ablasting therapy (4, 5). Preclinical studies with prostate cancer models demonstrating that MYC hyperactivation in concert with PI3K/Akt/mTOR signaling can drive neuroendocrine differentiation and visceral metastasis formation provide support for a potential causal role of MYC in promoting tumor progression to an aggressive mCRPC phenotype(6).

MYC was previously thought to be ‘undruggable’; however, multiple classes of therapies are now aimed at indirectly inhibiting MYC or its downstream mediators. Foremost among them are bromodomain extra terminal (BET) inhibitors, which have demonstrated the ability to directly downregulate MYC expression. These inhibitors also display significant antitumor activity in a variety of androgen-independent, MYC-high prostate cancer cell lines and xenograft models(7–9). Multiple early-phase clinical trials of BET inhibitors in abiraterone/enzalutamide-resistant mCRPC are underway (e.g. NCT02705469, NCT02607228). Additional drugs targeting effectors of MYC signaling, including cyclin-dependent kinase (CDK) and PIM kinases, are also in pre-clinical and clinical development in prostate cancer and other solid tumor malignancies(10, 11).

Similar to how ¹⁸F-dihydrotestosterone (DHT) PET fostered the development of the androgen receptor (AR) antagonists enzalutamide and apalutamide(12, 13), the burgeoning number of therapies aimed at inhibiting MYC underscores the need to develop a companion imaging biomarker capable of monitoring MYC transcriptional activity in real-time, both to identify treatment naïve patients whose tumors harbor hyperactive MYC, and to enable longitudinal assessment of MYC pathway modulation and therapeutic response.

MYC transcribes the transferrin receptor (TFRC)(14). TFRC-targeting radiolabeled analogues may therefore serve as a non-invasive quantitative measurement of MYC transcriptional activity. Consistent with this hypothesis, we have previously shown that ⁸⁹Zr coupled to transferrin via the chelator desferrioxamine is retained in cancer models in a MYC dependent fashion(15–18). The goal of this study was to determine if ⁶⁸Ga-citrate, a human ready radiotracer that measures TFRC expression levels in vivo(19), could be used to detect MYC positive prostate cancer models and metastases (⁸⁹Zr-transferrin is not yet cleared for human use). Building on a pilot clinical study in which we showed that ⁶⁸Ga-citrate is taken up in human prostate cancer metastases(20), we conducted the first preclinical tumor imaging studies with ⁶⁸Ga-citrate, as well as the first patient studies of ⁶⁸Ga-citrate PET imaging coupled with analysis of cell free DNA (cfDNA) and paired metastatic tumor biopsies to investigate whether ⁶⁸Ga-citrate can detect CRPC with MYC hyperactivity and neuroendocrine differentiation.

MATERIALS AND METHODS

General methods

PC3 and 22Rv1 cells were obtained from ATCC and subcultured according to manufacturer’s recommendations. (+)-JQ1 was a generous gift from Dr. James Bradner. iBET-151 was purchased from Selleckchem and used without further purification. Human holo-transferrin was purchased from Sigma Aldrich and succinimidyl-DFO was obtained from Macrocyclics (Dallas, TX). Zirconium-89 was purchased from 3D Imaging, LLC (Maumelle, AR). ⁶⁸Ga-citrate was produced by the cyclotron core at UCSF and administered in buffered PBS. Antibodies against TFRC (Santa Cruz Biotechnology), MYC (Abcam), and Actin (Sigma) were utilized for Western blot analyses.

Radioiodination of transferrin and in vitro uptake assays

Iodination with iodine-125 was done in pre-coated iodination tubes (Pierce). 100 mg of transferrin was dispersed in 200 μL of PBS solution and added to the pre-coated iodination tubes. In a separate eppendorf, 1 μL of HCl (0.2 M), 2.5 μL of phosphate buffer (0.5 M, pH = 8), 10 μL of potassium iodide solution (1 mg/ml) was prepared. 3 mCi of iodine-125 (Perkin Elmer) was added into the tubes and the previous solution was then mixed in the iodination tubes. After 15 min of reaction at room temperature the solution was purified via PD10 column pre equilibrated with 20 mL of PBS solution. The purity was assessed via iTLC, and ¹²⁵I-Tf was always > 98% pure.

PC3 and 22Rv1 cells were counted and plated at fixed cell concentrations between treatment arms. Cells were treated with vehicle or the bromodomain inhibitors iBET-151 or JQ1 (1 μM) for 48 hours, whereupon they were washed and incubated with 10 μCi ¹²⁵I-Tf for 30 min at 37° C. After washing twice with PBS, the cell associated activity was harvested in 1M NaOH (aq.). The cell associated activity was expressed as a % of total activity to which the cells were exposed. This value was further normalized to cell number to correct for treatment-induced changes in cell viability.

In vitro studies of MYC expression

PC3 and 22Rv1 cells (4 × 10⁵) were incubated with vehicle, (+)-JQ-1 (1 μM) and IBET-151 (1 μM) at 37 degrees Celsius for 48 hours. Cells were lysed, and the mRNA was extracted using the Quick-StartProtocol of the Qiagen. The mRNA was converted to cDNA using High Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Real-time PCR analysis was performed using PikoReal 96 Real-Time PCR System. MYC and GADPH expression (serving as housekeeper gene) levels were probed using validated seqeunces, and the relative ratio of the expression of the two genes was measured.

Animal studies

All animal studies were conducted in compliance with the Institutional Animal Care and Use Committee at UCSF. ⁸⁹Zr-Tf was prepared as previously reported(16). Three to five week old male nu/nu mice and C57BL/6 mice were obtained from Charles River. Nu/nu mice were inoculated with 1 × 10⁷ 22Rv1 or PC3 cells subcutaneously into one flank in a 1:1 mixture (v/v) of media and Matrigel (Corning). Tumors were palpable within 14–21 days after injection. The drugs were suspended in HPMT solution (0.5% w/v hydroxypropyl-methylcellulose dissolved in water plus 0.2% v/v Tween 80). Tumor bearing mice were treated once daily via oral gavage with iBET-151 (30 mg/kg/d), (+)-JQ1 (50 mg/kg, BID), and vehicle for 3 days prior to radiotracer injection. Drug treatment continued during the 48 hours that ⁸⁹Zr-Tf distributed in the body. C57BL/6 mice were treated with 50 μL of neat turpentine in the right hindlimb and injected 24 hours later with ⁶⁸Ga-citrate.

Small-animal PET and biodistribution studies

Tumor bearing mice (n=5 per treatment arm) received ~400 μCi of ⁸⁹Zr-Tf or ⁶⁸Ga-citrate in 100 μL volume intravenously using a custom mouse tail vein catheter with a 28-gauge needle and a 100–150 mm long polyethylene microtubing (0.28 mm I.D. × 0.64 mm O.D., Scientific Commodities, Inc., Lake Havasu City, AZ). Approximately 32 μg of ⁸⁹Zr-Tf at a specific activity of 0.4 mCi/nmol was administered per mouse. The mice were imaged on a dedicated small animal PET/CT scanner (Inveon, Siemens Healthcare, Malvern, PA). Mice were imaged at 4 or 48 hours post injection. Animals were scanned for 20–40 minutes for PET, and the CT acquisition was 10 minutes. The coregistration between PET and CT images was obtained using the rigid transformation matrix from the manufacturer-provided scanner calibration procedure since the geometry between PET and CT remained constant for each of PET/CT scans using the combined PET/CT scanner. Animals were anesthetized with gas isoflurane at 2% concentration mixed with medical grade oxygen. PET data were framed dynamically for the first two time points. The durations of the 0.5-h PET data were: 10×10s, 5×40s, 1×300s, and 5×600s. The 3–5 h PET data were also divided to two 1800s frames. The in vivo CT parameters were 120 projections of continuous rotations to cover 220° with an x-ray tube operated at 80 kVp, 0.5 mA, and 175 ms exposure time.

Manufacturer-provided ordered subsets expectation maximization (OS-EM) algorithm was used for PET reconstruction that resulted in 128×128×159 matrices with a voxel size of 0.776×0.776×0.796 mm³. The CT image was created using a conebeam Feldkamp reconstruction algorithm (COBRA) provided by Exxim Computing Corporation (Pleasanton, CA). The matrix size of the reconstructed CT images was 512×512×662 with an isotropic voxel size of 0.191×0.191×0.191 mm³. The photon attenuation correction was performed for PET reconstruction using the coregistered CT-based attenuation map to ensure the quantitative accuracy of the reconstructed PET data.

To evaluate the uptake of ⁶⁸Ga and ⁸⁹Zr-Tf radiotracers in human xenografts, biodistribution studies were conducted following imaging. Animals were euthanized by CO₂ asphyxiation after scans were completed. Sixteen tissues, including the tumor, were harvested immediately following sacrifice. The tissues were weighed and counted using a Wizard3 gamma counter (Perkin Elmer) to assess ⁶⁸Ga or ⁸⁹Zr concentration. Calibration with known amounts of ⁶⁸Ga and ⁸⁹Zr was performed to determine the amount of activity in each organ. This activity was then decay corrected and the percentage of the injected dose per gram (%ID/g) of tissue was calculated and reported.

Patient selection and clinical study design

The clinical study was approved by the UCSF institutional review board, and all patients provided written informed consent before study enrollment. All patients had histologically confirmed prostate cancer, and had developed progressive metastatic castration resistant disease by PCWG2 criteria. All had progressive disease following treatment with abiraterone and/or enzalutamide(21). Patients were required to undergo conventional scans including cross-sectional imaging of the chest/abdomen/pelvis and whole body radionuclide bone scan (either ^99mTc- HDP or ¹⁸F-NaF) within 12 weeks prior to ⁶⁸Ga-citrate PET imaging. All included patients had imaging evidence of at least three metastatic lesions by conventional imaging. On the day of ⁶⁸Ga-citrate PET imaging, prior to tracer injection, patients had peripheral blood collected for determination of cell free tumor DNA MYC copy number status. A subset of patients underwent CT-guided core needle biopsy of an accessible metastatic lesion performed within 4 weeks following completion of ⁶⁸Ga-citrate PET imaging. Patients underwent an optional second ⁶⁸Ga-transferrin PET scan within 12 weeks of first scan.

PET imaging protocol

Patients were injected with up to 15 mCi (555 MBq) (average 7.42 mCi [274.6 MBq], range 3.7 to 11.9 mCi [136.9 to 438.5 MBq]) ⁶⁸Ga-citrate intravenously. PET acquisition was acquired between 120 and 263 minutes after injection (average 210 minutes). Images were acquired on either a PET/CT or PET/MR. PET/CT examinations were performed on either a Biograph 16 (Hi-Rez) PET/CT scanner (Siemens AG, Erlangen, Germany) with an integrated PET and 16-MDCT scanner or a Discovery VCT PET/CT scanner (GE Medical Systems, Milwaukee, WI) with an integrated PET and 64-MDCT scanner. A low-dose CT was acquired for PET attenuation correction. PET/MR images were performed on a SIGNA PET/MR (GE Medical Systems, Milwaukee, WI). Attenuation correction for PET reconstruction was performed using a MR-based attenuation correction (MRAC) technique provided by the scanner manufacturer.

Image analysis

Maximum intensity projection (MIP), axial, coronal and sagittal reconstructions and PET/MR fused images were reviewed on an Advantage Windows Workstation (AW, Waukesha, WI). PET images were evaluated by trained nuclear medicine physician blinded to the results of conventional imaging scans as well as clinical/genomic features of the case and scored for the presence of PET avid lesions. Lesions were considered PET positive if uptake was focal, greater than the adjacent background soft tissue and not in an expected physiologic structure such as the urinary bladder, vessels or salivary glands.

For semi-quantitative analysis, a volume of interest (VOI) was manually drawn around PET-avid lesions and SUV_max were recorded. The location of abnormal radiotracer uptake was compared to CT and nuclear medicine bone-scans. Additionally, SUV_mean values were recorded in the liver, paraspinous soft tissues, bone (right sacrum), and mediastinal blood pool for determination of normal structures.

With conventional imaging, soft tissue metastases were considered positive if greater than 1 cm in long axis, except for lymph nodes that were considered positive if greater than 1.5 cm in short axis. Bone lesions on radionuclide scan were considered positive if uptake was focal and not in a pattern consistent with arthritis or antecedent trauma/fracture.

Cell free DNA analysis of MYC amplification

Blood (20 ml) was collected from mCRPC patients in purple top EDTA tubes and was fixed using 10% neutral buffered solution containing 4% paraformaldehyde within 2 hrs of receiving the sample. Plasma was isolated immediately from the fixed blood by centrifuging the samples for 10 min at 200 × g at room temperature followed by a second spin at 1600 × g for 10 min. Avoiding the buffy coat, plasma was carefully transferred to new 15 ml tubes and centrifuged at 1600 × g for an additional 10 min. The plasma was transferred to 50 ml tubes and stored at −80°C. cfDNA was extracted from plasma using the QiAamp Circulating Nucleic Acids kit (Qiagen) according to the manufacturer’s protocol. The extraction was carried out within a week of the plasma extraction. The extracted cfDNA was eluted in 20 μl of AE (Qiagen) buffer and assayed for MYC copy number using Digital PCR (QuantStudio, Life Technologies). 1X QuantStudio 3D Digital PCR Master Mix V2 (Applied Biosystems), TaqMan Copy Number Assay Hs02045885_cn (MYC) and 1X RNAseP TaqMan copy number reference assay (Applied Biosystems) were used in the reaction. DNA extracted from PC3 cell lines and PC3 DNA spiked into healthy donor blood was used as positive (MYC amplified) controls while male normal germline DNA was used as negative control (MYC wild type). 6μl of cfDNA was combined with the PCR reaction mix and loaded onto a QuantStudio 3D Digital PCR 20K chip. The QuantStudio 3D Analysis Suite Software was used to evaluate MYC copy number for each patient sample based on the reference gene (RNAseP) copy number (i.e. 2 copies).

Analysis of Histologic Evidence of Small Cell Neuroendocrine Carcinoma in Metastatic Tumor Biopsies

FFPE tissue was histologically evaluated in a central CLIA-certified lab for evidence of neuroendocrine differentiation by morphologic and immunohistochemical criteria(22).

Statistical Analysis

The percentage of injected activity taken up in prostate cancer tumors versus other organs in mice injected with either ⁶⁸Ga-citrate or ⁸⁹Zr-Tf analyzed using the unpaired, two-tailed Student’s t-test. Differences at the 99% confidence level (P < 0.01) were considered to be statistically significant. The imaging data from patient studies were summarized using descriptive statistics. Fisher’s exact test was used to compare frequency of PET-avid lesions across organ site.

RESULTS

⁶⁸Ga-citrate shows MYC dependent uptake in human prostate cancer models

⁶⁸Ga-citrate has not been previously evaluated in tumor bearing mice. Therefore, we first conducted a time course study in nu/nu mice bearing subcutaneous PC3 xenografts, a MYC positive model of human PCa, to define the optimal time point post injection to study ⁶⁸Ga-citrate. Mice received 400–500 μCi of ⁶⁸Ga-citrate, and biodistribution studies were conducted at 2, 4, and 6 hours post injection (Figure 1A and Supplemental Figure 1). Peak tumor uptake was observed at 4 hours post injection. Moreover, blood pool activity declined from 2 – 4 hours post injection, and remained stable from 4 – 6 hours post injection. Radiotracer uptake in normal tissues (liver, spleen, kidney, and muscle) saturated within 2 hours post injection, and the level of uptake was lower than what was observed in the tumor. The highest tumor to blood and tumor to muscle ratios were found to be at 4 and 6 hours post injection (Supplementary Table 1). A follow-up study in mice bearing subcutaneous 22Rv1 xenografts, an androgen receptor and MYC positive model of human prostate cancer, also showed high ⁶⁸Ga-citrate uptake 4 hours post injection, with similar uptake in normal mouse tissues and compartments (Figure 1B and Supplemental Figure 2).

Figure 1.

A. A biodistribution study showing the accumulation of ⁶⁸Ga-citrate in normal mouse tissues and subcutaneous PC3 tumors at 2, 4, and 6 hours post injection. Peak radiotracer uptake was observed in the tumors at 4 hours post injection. B. A biodistribution study showing the relative distribution of ⁶⁸Ga-citrate in normal mouse tissues and subcutaneous 22Rv1 tumors at 4 hours post injection. C. Biodistribution data from mice bearing subcutaneous 22Rv1 tumors showing that the amount of ⁸⁹Zr-Tf uptake in the tumors is equivalent at 48 hours post injection is equivalent to the amount of ⁶⁸Ga-citrate uptake in tumors at 4 hours post injection. Importantly, the tumor uptake of ⁸⁹Zr-Tf exceeded blood pool activity, showing that ~7% ID/g is a value that can represent specific binding.

Because blood pool activity was approximately equivalent with tumor uptake of ⁶⁸Ga-citrate in both mouse cohorts, we next tested if tumor uptake of ⁶⁸Ga-citrate was equivalent with that of ⁸⁹Zr-Tf at time points we previously showed to reflect receptor mediated binding to tumors(15–17, 23, 24). We reasoned that equivalent levels of radiotracer uptake in tumor would be strong evidence of specific receptor binding by ⁶⁸Ga-citrate. Nu/nu mice bearing subcutaneous 22Rv1 xenografts were treated with ~400 μCi of ⁸⁹Zr-Tf and radiotracer biodistribution was studied 48 hours post injection. Blood pool associated activity was significantly lower than that of tumor, as expected (Figure 1C). Moreover, tumor uptake of ⁸⁹Zr-Tf was equivalent to the uptake of ⁶⁸Ga-citrate.

To further show evidence of specific receptor binding by ⁶⁸Ga-citrate at 4 hours post injection, a cohort of nu/nu mice bearing subcutaneous PC3 or 22Rv1 tumors were pretreated with vehicle or the BET bromodomain inhibitors (+)-JQ1 or iBET-151. Both JQ1 and iBET-151 therapies down-regulated MYC mRNA and Tf uptake in vitro (Supplemental Figure 3). Tumors in the mice receiving 5 days of treatment had significantly less ⁶⁸Ga-citrate uptake (Figure 2A and 2B, see Supplemental Figures 4 and 5). Moreover, the percent change in tumor uptake observed in BET inhibitor-treated mice was similar between ⁶⁸Ga-citrate and ⁸⁹Zr-Tf (⁸⁹Zr-Tf biodistribution was studied 48 hours post injection.

Figure 2.

A. Biodistribution data from nu/nu mice bearing subcutaneous 22Rv1 xenografts shows reduction of tumor uptake of ⁸⁹Zr-Tf and ⁶⁸Ga-citrate after treatment with (+)-JQ1 or iBET-151. Mice were treated for 5 days prior to radiotracer administration. Biodistribution studies were conducted 4 hours post injection of ⁶⁸Ga-citrate, and 48 hours post injection of ⁸⁹Zr-Tf. Drug treatment continued while ⁸⁹Zr-Tf equilibrated into peripheral tissues. *p < 0.01, ** p< 0.05 B. A plot showing the equivalent percent reduction in tumor uptake of ⁶⁸Ga-citrate or ⁸⁹Zr-Tf due to MYC and TFRC suppression by (+)-JQ1 or iBET-151. C. A plot of biodistribution values for ⁶⁸Ga-citrate in normal muscle (N.M.), an inflamed muscle (I.M.) due to turpentine treatment, and the two prostate cancer tumors shows that radiotracer uptake is higher in tumor.

⁶⁸Ga-citrate has been used in tumor naïve humans and rodents to detect the focal accumulation of activated leukocytes due to inflammation or infection(25). On this basis, we tested whether ⁶⁸Ga-citrate could quantitatively distinguish tumor from an inflammatory abscess. A cohort of tumor naïve mice were injected with ~50 μL of turpentine in the right hindlimb to induce an acute phase response, and the mice were treated with ⁶⁸Ga-citrate 24 hours after injection. PET and biodistribution studies conducted 4 hours post radiotracer injection showed detectable accumulation of ⁶⁸Ga-citrate in the inflamed muscle above the contralateral untreated hindlimb (Figure 2C). The magnitude of ⁶⁸Ga-citrate accumulation in the inflamed muscle was statistically lower than what was observed in either prostate cancer tumor using the same imaging conditions.

Patient Imaging Results

20 patients were enrolled on study between May 2015 and December 2016. Baseline characteristics are shown in Supplementary Table 2. All patients had mCRPC with prior progression on abiraterone and/or enzalutamide.

A total of 326 lesions were detected on conventional imaging (CT, ^99mTc-HDP), of which 53% were avid for ⁶⁸Ga-citrate. Roughly two thirds (63.8%) of osseous lesions were avid for radiotracer, while ~20% of lymph node and ~5% of visceral organ lesions were avid. Significant qualitative and quantitative inter-tumoral heterogeneity was observed with respect to ⁶⁸Ga-citrate uptake (Figure 3). 174 metastatic lesions (53.3%) were PET avid and 152 (46.7%) were negative on ⁶⁸Ga-citrate PET imaging. The average SUV_max of ⁶⁸Ga-citrate positive lesions was 5.74 (SD 2.89, range 1.8 to 19.4). The median percentage of PET-positive lesions per patient was 50% (range 0–93%). There was no relationship between dose of ⁶⁸Ga-citrate or tumor uptake time with the percentage of PET avid lesions or intra-tumoral SUV_max. PET avid lesions were more likely to be detected in the bone versus soft tissue (63.8% vs. 12.1%; p < 0.0001) (Table 1). Figure 4 highlights a particular example of a patient with heterogeneous ⁶⁸Ga-citrate uptake within metastatic lesions in the bony pelvis, with adjacent lesions demonstrating widely varying degree of intra-tumoral uptake.

Figure 3.

Distribution of intra-tumoral ⁶⁸Ga-citrate PET uptake by site of metastasis in the patient cohort. SUVmax = maximum standardized uptake value on PET per lesion

Table 1.

Distribution of PET-avid metastatic lesions by body site

Organ Site	Number of Lesions on Conventional Imaging	Number of PET-avid lesions on 68Ga-Tf PET (%)	Average SUVmax (sd)
Bone	260	166 (63.8%)	5.63 (2.76) Range: 1.8–18.8
Lymph Node	29*	6 (20.7%)	7.04 (2.43) Range: 4.4–10.8
Visceral Organ	37	2 (5.4%)	5.85 (0.07) Range: 5.8–5.9

Figure 4. ⁶⁸Ga-citrate uptake is heterogeneous among lesions detected with conventional imaging in the same patient.

A. Multiple regions of uptake on ^99mTc-HDP bone scan (left) including intense uptake in the right femur (blue arrow). While the ⁶⁸Ga-citrate PET MIP image (right) shows matching uptake in the right femur (blue arrow) it also reveals discordant uptake in the right ischium (orange arrow). B. Axial PET, CT, and fused ⁶⁸Ga-citrate PET/CT through the right ischium show that the CT is normal in region of ⁶⁸Ga-citrate uptake in the right ischium. CT also showing sclerosis in the region of the abnormal bone scan and ⁶⁸Ga-citrate uptake (blue arrow).

⁶⁸Ga-citrate PET Uptake in Tumor and Normal Organs is Reproducible Upon Serial Imaging

Four patients underwent paired imaging with ⁶⁸Ga-citrate PET imaging, with a median interval of 37 days (range 30 – 39 days) between scans (Supplementary Table 3). During this interval, patients were clinically stable and had no significant change in serum PSA level to indicate disease progression. The average change in SUV_mean between patient scans for blood pool, liver, paraspinus muscles and bones was 0.38 (range = −0.4 – 0.4; SD =0.05), 0.25 (range = −0.6 to 0.1; SD = 0.24), 0.28 (range = −0.4 – 0.4; SD = 0.15), and 0.3 (range = −0.4 to 0.3; SD = 0.08). Across all PET avid metastatic lesions for these four patients (N = 10; all osseous), there was likewise no significant difference in ⁶⁸Ga-citrate uptake between pre- to post-scans (mean change from baseline in SUV_max = −0.29 (range −5.3 – 1.4; SD = 1.89).

Pronounced ⁶⁸Ga-citrate PET uptake is observed in a subset of mCRPC patients with high tumor MYC amplification

Eighteen of twenty patients (90%) were evaluable for MYC copy number gain by cell free DNA (cfDNA) analysis. We observed a gain of at least one copy of MYC in 16 out of 18 (89%) evaluable samples (Table 2). This is generally consistent with prior reports demonstrating high prevalence of 8q chromosome gain in prostate cancer (26, 27). All 16 patients with evidence of MYC copy gain on cfDNA analysis had at least one ⁶⁸Ga-citrate-avid metastatic lesion. Higher level focal MYC amplification (> 2 copy number gain) was observed in six patients (33%), also consistent with prior molecular analyses of mCRPC tumors. Though the limited sample size precludes definitive assessment of an association between MYC amplification in cfDNA with ⁶⁸Ga-citrate PET, the subset of 6 patients with high level MYC amplification on cfDNA demonstrated a higher percentage of ⁶⁸Ga-citrate-avid lesions compared to those without MYC amplification (68.8% vs. 46.6%). Figure 5 illustrates one such patient (Patient-004) with MYC amplification on cfDNA analysis (estimated 6 copies of MYC gained) with striking PET avidity on ⁶⁸Ga-citrate PET, with 29 out of 32 total osseous metastases (90%) positive for uptake (average SUV_max = 6.3/lesion).

Table 2.

MYC Copy Number Status in Cell Free DNA by Digital PCR Analysis

Samples	# of MYC copies gained	# of MYC copies gained, rounded
PC3 stock (+MYC) control	2.74	3
PC3 cfDNA (+MYC) control	1.58	2
Male germline DNA (WT MYC)	0	0
Patient-001	1.16	1
Patient-003	3.48	3
Patient-004	6.28	6
Patient-005	0.66	1
Patient-006	0.68	1
Patient-007	1.3	1
Patient-008	0.32	0
Patient-010	0.58	1
Patient-011	0.78	1
Patient-012	3.5	4
Patient-013	0.74	1
Patient-014	1.84	2
Patient-015	1.1	1
Patient-016	2.76	3
Patient-017	0.28	0
Patient-018	4.34	4
Patient-019	5.78	6
Patient-021	0.82	1

Figure 5.

A maximum intensity projection (MIP) image of Patient-004 demonstrating widepsread uptake within multilevel osseous metastases within the spine, with concurrent cell free DNA evidence of focal, high MYC copy gain

⁶⁸Ga-citrate PET detects tumors with histologic and serologic evidence of neuroendocrine differentiation

11 patients (55%) underwent CT-guided metastatic tumor biopsy following ⁶⁸Ga-citrate PET scan. The median time interval between PET scan and subsequent image-guided tumor biopsy was 6 days (range 4 – 38). Of the 7 patients with metastatic tissue evaluable for histologic analysis, one demonstrated pure small cell neuroendocrine differentiation on pathology review (patient-011), and the other six tumors demonstrated adenocarcinoma differentiation (Supplementary Table 4). The tumor with pure SCNC histology was located in the R femur of this patient (Figure 6). On coupled ⁶⁸Ga-citrate PET imaging, this lesion demonstrated high avidity for the tracer, above the median for the patient cohort (SUV_max = 6.9), and significantly higher than adjacent metastastic lesions (Figure 6). Genomic interrogation of the metastatic tumor biopsy revealed high-level MYC amplification (estimated 4 copy number gain) consistent with prior reports of treatment-emergent small cell neuroendocrine prostate cancer. An additional patient (patient-004) had serologic evidence of neuroendocrine differentiation (serum chromogranin = 66 ng/mL; upper limit of normal = 15 ng/mL) with a concordant markedly positive PET scan with 29 out of 32 metastatic lesions positive for uptake, as previously described above (Figure 5).

Figure 6.

A. Axial PET, CT, and fused ⁶⁸Ga-citrate PET/Ct showing a ⁶⁸Ga-citrate avid lesion within the right femur with no CT correlate (blue arrow). B. Representative pathology image of the metastasis from the patient’s biopsy demonstrateing small cell neuroendocrine prostate cancer.

In the two patients with genomically evaluable paired cfDNA and metastatic tumor biopsies, it is worth noting that patient-011 demonstrating only 1 copy gain of MYC in analysis of cfDNA vs. 4 copies gained in metastatic biopsy. Patient-003, in contrast, demonstrated 3 copies of MYC gained in both metastatic biopsy and cfDNA. These results support the significant interlesional and interpatient heterogeneity observed on PET imaging and highlight the difference between lesion-specific genomic assessment and aggregated results on cfDNA analysis.

⁶⁸Ga-citrate uptake is recapitulated in patient-derived xenograft of a liver metastasis

We isolated tumor tissue from the liver metastasis (patient-003) and propagated it in a NOD SCID mouse model (Supplemental Figure 7). The patient-derived xenograft (PDX) tumor was histologically similar to the tumor biopsy (Supplementary Figure 8), and demonstrates similar mRNA transcript levels (log-scale) of MYC (liver metastasis = 10.9; PDX = 10.6) and TFRC (liver metastasis = 13.2; PDX = 12.3). Upon imaging of the PDX tumor with ⁶⁸Ga-citrate PET, we observed comparable intra-tumoral uptake as that observed in the liver metastasis from the patient PET scan (% ID/g = 2.3 ± 0.9 in PDX tumor; SUV_mean = 4.9 in liver metastasis, see Supplemental Figure 9).

DISCUSSION

In this report, we show that both human prostate cancer models and clinical disease bearing the molecular and/or histological features of MYC hyperactivity are detectable with ⁶⁸Ga-citrate PET. Our preclinical experiments show specific tumor binding of ⁶⁸Ga-citrate in MYC-high prostate cancer cell lines and xenograft models, and suggest a novel tool to permit functional read-out of pharmacologic inhibition of the MYC signaling pathway by BET inhibitors and perhaps other emerging therapies. In patients with mCRPC, we show that a subset of patients harbored markedly ⁶⁸Ga-citrate-avid tumors with concordant molecular evidence of MYC activation detectable in paired analysis of cfDNA. With biopsy, two examples are shown of patients with histologic or serologic evidence of SCNC, and they also have tumors markedly avid for ⁶⁸Ga-citrate. These results provide robust justification to further investigate the clinical application of ⁶⁸Ga-citrate PET as a real-time, noninvasive monitoring tool that can potentially capture MYC-driven prostate cancer, including tumors with SCNC histology.

MYC transcriptional activity is associated with SCNC, a subset of advanced prostate cancer that is highly lethal, and that can be difficult to identify without invasive and sometimes technically challenging metastatic biopsies. Although MYC inhibitors are under development, pharmacodynamic markers of MYC inhibition are not readily available. Therefore, these data are timely and significant, as the successful outcome of preclinical and/or clinical trials with indirect MYC inhibitors will likely depend on biomarkers like ⁶⁸Ga-citrate that can identify treatment naïve tumors with MYC hyperactivity (which therefore may be most likely to respond to an anti-MYC inhibitor), as well as monitor the down-regulation of MYC activity longitudinally that might predict a clinical response to treatment.

Our pre-clinical experiments indicate a high degree of tumor specificity of ⁶⁸Ga-citrate for MYC-high prostate cancer cell lines and xenograft models, and suggest a novel tool to permit functional read-out of pharmacologic inhibition of the MYC signaling pathway by BET inhibitors and perhaps other emerging therapies. In patients with mCRPC, we show that the subset of patients with focal MYC amplification detectable in cfDNA harbored a higher proportion of tumors with ⁶⁸Ga-citrate-avid avidity. We further demonstrate that ⁶⁸Ga-citrate PET intra-tumoral uptake is reproducible but with appreciable inter-tumoral and inter-patient heterogeneity that may reflect divergent clonal evolution with varying degree of MYC hyperactivation in the treatment-resistant mCRPC setting (28). Whole body ⁶⁸Ga-citrate PET imaging may therefore represent a valuable novel tool to annotate the degree of inter-tumoral heterogeneity with respect to MYC/neuroendocrine signaling in a non-invasive, real-time fashion.

The underlying reason for the preferential ⁶⁸Ga-citrate PET avidity in bone lesions relative to soft tissue metastases is not readily apparent. There are no significant differences in TFRC expression levels by anatomic site of disease in mCRPC in ours and other publicly available datasets. One prior report has suggested that ⁶⁸Ga-citrate accumulation is due to regional variation in perfusion and non-receptor mediated accumulation in regions of leaky vasculature(27). The consistency in ⁶⁸Ga-citrate uptake between our PDX subcutaneous model and in situ liver metastasis, however, supports tumor-specific uptake as opposed to regional vascular permeability as the predominant determinant of radiotracer accumulation. Additional studies in CRPC patients harboring soft tissue metastases will be necessary to fully characterize the expected pattern of ⁶⁸Ga-citrate PET uptake in non-osseous sites of metastases.

Asangani et al. previously did not observe any down regulation of MYC in PC3 cells when treated with JQ1(8). Reviewing each study carefully, we hypothesize that methodological differences in the experiments may account for the apparent discrepancy with our data. Asangani et al treated PC3 cells for 24 hours with JQ1 and dosed at concentration of 500 nM, which did not reduce MYC protein levels by Western blot. In our study, we treated PC3 cells with JQ1 at a dose of 1 μM for 48 hours, which did reduce MYC mRNA. Our higher dose and longer drug exposure may account for why we observe anti-MYC effects that the previous authors did not report. Importantly, we also show anti-MYC effects in PC3 with a structurally discrete BET bromodomain inhibitor.

MYC hyperactivity in other cancers has also been associated with specific imaging patterns. Palaskas et al. reported that ¹⁸F-FDG specifically demarcated basal like breast cancer with MYC hyperactivity(28). It is unclear if there is a relationship between MYC and ¹⁸F-FDG avidity in prostate cancer, particularly given that CRPC is not avid for ¹⁸F-FDG in most circumstances.

The case examples discussed demonstrate the potential utility of ⁶⁸Ga-citrate PET as a detection and treatment response monitoring tool for therapies applied against the MYC signaling pathway in SCNC. Given the likely growing prevalence of SCNC as a resistance mechanism to potent androgen signaling blockade, and the lack of validated methods to identify SCNC differentiation other than metastatic tumor biopsy, there is a clear need to develop non-invasive biomarkers to detect and monitor treatment response in this highly aggressive, lethal disease subset. Our studies with ⁶⁸Ga-citrate come at an opportune time given the recent FDA approval of ⁶⁸Ga-DOTA-TATE for the detection of neuroendocrine tumors. Preliminary reports with DOTA-TATE imaging in CRPC indicate potential utility of detecting lesions with neuroendocrine differentiation(29). However the range of uptake is modest, and it is unclear whether DOTA-TATE avidity will be observed in higher grade tumors, analogous to the inverse correlation between tracer uptake and tumor grade observed with neuroendocrine tumors of GI origin. Further patient studies with paired ⁶⁸Ga-citrate PET scan coupled with tumor biopsies to explore the relationship between SCNC differentiation and ⁶⁸Ga-citrate uptake on PET scan are underway.

Implications.

Non-invasive measurement of MYC activity with quantitative imaging modalities could substantially increase our understanding of the role of MYC signaling in clinical settings for which invasive techniques are challenging to implement or do not characterize the biology of all tumors in a patient. Moreover, measuring MYC activity non-invasively opens the opportunity to study changes in MYC signaling in patients under targeted therapeutic conditions thought to indirectly inhibit MYC.