Prostate-specific membrane antigen (PSMA) expression in patients with metastatic triple negative breast cancer: Initial results of the PRISMA study

1025 Background: Radioligand therapy (RLT) targeting prostate-specific membrane antigen (PSMA) prolongs survival of patients (pts) with prostate cancer. Preliminary immunohistochemistry (IHC)-based studies suggest that PSMA is highly expressed in triple negative breast cancer (TNBC). The PRISMA study aims to assess PSMA expression in pts with metastatic TNBC (mTNBC) via positron-emission tomography/computed tomography (PET/CT) to evaluate the feasibility of PSMA-targeted RLT for TNBC. Methods: PRISMA is a prospective, single-center study that enrolled pts with progressive mTNBC and measurable disease on 18 F-FDG PET/CT to undergo 68 Ga-PSMA-11 PET/CT. TNBC definition (ASCO/CAP) was based on the IHC of the primary breast cancer (BC) or metastasis. Target lesions (TL) were defined as metastatic lesions ≥1.5 cm with significant uptake in 18 FDG-PET/CT. The same TL regions were propagated to the 68 Ga-PSMA-PET/CT. 68 Ga-PSMA-11 maximum standard uptake value (SUV max ) was measured for each TL. PSMA patterns A and B were defined according to presence of a majority (pattern A) or a minority (pattern B) of positive TL on 68 Ga-PSMA (ie. uptake greater than that of normal liver parenchyma). Mann-Whitney U test was used to assess differences in median SUV max (mSUV max ) between subgroups. Results: Ten pts with progressive mTNBC were enrolled. One pt was excluded due to lack of TL on 18 FDG-PET/CT. Median age was 48 years (range 36-68), median number of TL per patient was 8 (5-18). Five pts had a TNBC IHC since the initial BC diagnosis, while 4 had a switch to TNBC (4 with initial estrogen receptor [ER]-positive and 1 with HER2-positive BC). All evaluable pts had inter- and intra-organ heterogeneity in PSMA uptake in TL and 2 pts had some mismatch with FDG-positive and PSMA-negative TL. mSUV max was 2.99 (range 1.7-6.3) in the overall population and was higher in pts with pattern A (5.1 vs 2.6 in patterns A and B, respectively, p = 0.03). mSUVmax was numerically higher in younger pts (5.0 vs 2.6 in pts <50y and ≥50y, p = 0.28), PD-L1-positive (5.0 vs 2.7 in PD-L1-positive and -negative, p = 0.14), de novo metastatic disease (5.0 vs 2.9 in de novo and recurrent BC, p = 0.38) and those with ER-negative primary breast tumor (3.4 vs 3.0 in ER-negative and ER-positive primary BC, p = 0.9). Conclusions: A significant PSMA uptake was observed in a proportion of pts with mTNBC. Inclusion of additional pts and translational analyses will allow the identification of PSMA expression determinants. PSMA-targeted RLT may be considered as an innovative therapeutic strategy to be explored in selected pts with mTNBC. [Table: see text]