Abstract
1419
Objectives: Prostate Specific Membrane Antigen (PSMA) is known to be overexpressed on non-prostate malignancies, thereby giving a potential target for imaging using PET/CT tracers and therapy using Lu-177 chelates. Our aim was to demonstrate both the immunohistochemistry (IHC) and PET/CT expression of PSMA on Hepatocellular carcinoma (HCC) and Cholangiocarcinoma cells from Asian patients in preparation for an in human prospective study.
Methods: Relevant ethics board approvals were obtained. We performed IHC studies on tumor microarrays of HCCs and cholangiocarcinomas to verify PSMA overexpression. HCC tumors were previously used to create xenograft lines. Two PSMA-positive models and two PSMA-negative models were selected to establish tumors in male SCID mice aged 9 to 10 weeks. Experiments started when the tumors reached the size of approximately 170-250 mm3. Mice bearing indicated tumors were injected in the ophthalmic vein with Ga-68 PSMA-11. Scanning was done using a Vector4CT scanner 60 minutes post injection for a total of 40 minutes. The animals were then sacrificed and digital autoradiography was performed on the explants.
Results: 25 out of 25 HCCs were found to express PSMA, either in the canaliculi or sinusoids. 15 out of 26 cholangiocarcinomas expressed PSMA in their endothelium. However none of the tumours were found to show uptake more than background on in vivo PET/CT imaging or digital autoradiography. Physiological salivary gland, kidney and urinary bladder activity were observed confirming successful injection of tracer.
Conclusions: We confirm that a high proportion of HCCs show PSMA expression in vitro, but the proportion of positivity of PSMA in HCC lines decreases in IHC of xenograft models, and in vivo imaging is unrewarding despite IHC positivity. About half of cholangiocarcinomas show IHC positivity, and this should be kept in mind when reporting incidental hepatobiliary lesions on PSMA PET/CT. This study adds evidence that murine model HCC imaging with PSMA PET/CT agents is not ideal for studying the dosimetry and distribution of tracer prior to in human studies, and future pre-clinical studies will have to consider using other non-murine models for this purpose.