Genomic landscape of 5’methylthioadenosine phosphorylase ( <i>MTAP)</i> deleted ( <i>MTAP</i> loss) non-squamous carcinoma of unknown primary site (nsCUP).

Abstract

3081 Background: MTAP , a key enzyme in the polyamine pathway breaks down 5’Deoxy-5-Methylthioadenosine ( MTA ) into methionine and adenine. MTAP loss reduces adenine and accumulates MTA, which inhibits protein arginine methyltransferase 5 ( PRMT5 ). This suggests MTAP loss cancers may respond to PRMT5 inhibition. Methionine adenosyl transferase 2α ( MAT-2A ) is a primary producer of donor S-adenosylmethionine ( SAM ) and the depletion of MAT-2A has antiproliferative effect in cancers with MTAP loss. Based on the synthetic lethality concept, MTAP loss is being used as a biomarker for accrual in multiple trials with PRMT5 and MAT-2A inhibitors. We queried the genomic landscape of MTAP loss in patients with nsCUP. Methods: DNA extracted from formalin-fixed paraffin-embedded (FFPE) tissue of 7,440 nsCUP cases from 2020 to 2024 underwent hybrid capture-based comprehensive genomic profiling (CGP) to assess all classes of genomic alterations (GA). All cases underwent central pathology review to confirm that at the time of sequencing, a primary site for the cases was not established. Microsatellite instability (MSI) status and tumor mutational burden (TMB) were derived from the CGP data. Programmed death-ligand 1 (PD-L1) was determined by immunohistochemistry (IHC) using the DAKO 22C3 system. Results: 853 (11.5%) of nsCUP cases had either complete or partial MTAP loss with 0.7% 1 exon, 1.2% 2 exons, 2.9% 3 exons, 5.1% 4 exons, 0.5% 5 exons, 2.5% 6 exons, 32.8% 7 exons and 54.3% 8 exons lost. The median age of the MTAP loss patients was higher (68 vs 65; p&lt;.0001) and the gender distributions were similar (52% to 54% female; not significant (NS)). Cyclin-dependent kinase inhibitor 2A (CDKN2A) loss co-occurred in 99.8% in patients with MTAP loss. MSI-high status was uncommon in both MTAP loss vs MTAP wildtype (0.4% vs 0.7%; NS). The MTAP wildtype group had higher tumor mutational burden (TMB) &gt; 10 mutations/mb (15.7% vs 11.1%; p=.0004) and TMB &gt;20 mutations/Mb (5.4% vs 3.5%; p=.017) rates. MTAP loss nsCUPs had higher frequencies of KRAS GA and KRAS G12C, whereas MTAP wildtype cases had greater frequencies of ERBB2 , PTEN , MET and EGFR GA (Table). GA in BRCA1 / 2 and FGFR2 were similar in both groups. GA in ALK , RET , ROS1 , RET and TRK were extremely uncommon in both groups (all less than 1%). Conclusions: At 11.5%, nsCUP features a relatively high frequency of MTAP loss, with the vast majority involving either all (8 of 8) or nearly all (7 of 8) exons. MTAP loss patients are slightly older and have reduced TMB levels which may impact their responsiveness to immunotherapy-based combination regimens with PRMT5/MAT-2A inhibitors. Clinical trials for the development of targeted therapies to use PRMT5 inhibition and MAT-2A in nsCUP are warranted. nsCUP MTAP Loss (N=853) nsCUP MTAP wildtype ( N=6,587) P value KRAS all/G12C 45.9%/7.6% 31.2%/4.2% &lt;.0001/&lt;.0001 ERBB2 all/amp only 6.7%/4.3% 10.9%/8.1% &lt;.0001/&lt;.0001 PIK3CA 6.2% 7.1% NS BRAF 6.0% 5.0% NS FGFR2 4.3% 4.0% NS PTEN 4.1% 6.1% .02 BRCA1 / 2 1.8%/2.3% 2.1%/2.4% NS/NS MET 2.1% 4.6% .0004 EGFR 2.6% 4.1% .031