Introduction:

Molecular-based PET imaging for prostate cancer is transforming our ability to detect and target previously unknown sites of disease. While ¹⁸F-fluciclovine (Axumin) PET-CT is approved for use in the United States and recommended by the NCCN, prostate-specific membrane antigen (PSMA) PET-CT is more widely used in Europe/Australia and is recommended by the European Association of Urology. Less is known about the distinct biology of lesions detected by either modality or the optimal clinical setting for their use. While Axumin PET relies on radiotracer uptake by amino acid transporters (LAT1-4 and ASCT1-2), PSMA PET is dependent on surface expression of PSMA. Our objective was to compare relative expression of PSMA and Axumin transporter genes in a large cohort of radical prostatectomy (RP) samples to determine their distribution across molecular subtypes and correlation with long term clinical outcomes.

Methods:

Gene expression data of 17,967 prospective RP samples generated from Decipher testing (GenomeDx Biosciences, San Diego, CA) and 1,135 retrospective samples with long term follow up were used in these analyses. All samples were from formalin-fixed paraffin-embedded tissues and the Affymetrix Human Exon 1.0 ST microarray (Affymetrix, Santa Clara, CA) was used for gene expression profiling. Associations between expression of PSMA and Axumin transporter genes (LAT1-4 and ASCT1-2) and pathologic variables, PAM50 molecular subtypes, and clinical outcomes were conducted. Multivariable analysis was performed to compare metastasis free survival and lymph node involvement.

Results:

In RP specimens (n=17,967), all Axumin transporter genes (LAT 1-4, ASCT1-2) were expressed at lower levels when compared with PSMA (IQR -0.1-0.8 versus IQR 1.4-2.6, p <0.0001). LAT2 and ASCT2 were more highly expressed than other Axumin transporters (p<0.0001). While PSMA expression was positively correlated with metastatic genomic risk (Decipher), and pathologic Gleason score (GS) (p<0.0001 for both), LAT2, LAT3 and ASCT2 were inversely correlated with metastatic genomic risk in primary tumors (p<0.0001) and less expressed in GS 9-10 tumors (p<0.0001 for all). In one retrospective cohort of 780 men with >10 years of median follow up, higher PSMA expression was associated with poorer metastasis-free survival (MFS) (HR 1.45, p=0.001) and lymph node involvement (LNI, HR 2.14, p<0.0001), whereas high expression of LAT2, LAT3, ASCT2 expression was associated with better MFS (HR 0.85, p=0.04; 0.63, p<0.0001; 0.74, p<0.0001 respectively). In multivariable analysis adjusting for Gleason grade, stage, PSA at diagnosis, and LNI, high PSMA expression remained independently prognostic of MFS (HR 1.3, p=0.028) but did not reach significance for LNI (HR 1.3, p=0.09).  After multivariable analysis, only LAT3 expression was prognostic of better MFS (HR 0.66, p<0.0001). With regard to molecular subtypes, Luminal B subtype was notable for overexpression of PSMA, while Luminal A was enriched in ASCT2 and LAT2 (p<0.0001). Basal subtype was notable for lowest expression of ASCT1 and LAT3 (p<0.0001). While PSMA expression did not correlate with ERG positive or negative prostate cancers, LAT2, ASCT1, and ASCT2 were overexpressed in ERG negative tumors (p<0.0001 for all).

Conclusion:

In RP specimens, PSMA expression is positively correlated with genomic risk scores and predictive of poorer metastasis-free survival. After multivariate analysis, LAT3 expression was inversely correlated to metastatic risk. Molecular subtypes of prostate cancer variably express PSMA and Axumin transporter genes. While these findings suggest that gene expression analysis may have utility in selecting prostate cancer patients for different imaging modalities, further studies are necessary to validate the prognostic value of these imaging targets. Additional studies into the role of these targets in underlying tumor biology and metastatic disease are also warranted.

Funding: Prostate Cancer Foundation, NIH R01 CA235741, Benioff Initiative for Prostate Cancer Research