NANETS2022 15th Annual Multidisciplinary NET Medical Symposium NANETS 2022 Other (12 abstracts)
1Masonic Cancer Center, University of Minnesota, Minneapolis, MN, USA; 2Caris Life Sciences, Phoenix, AZ, USA; 3Comprehensive Cancer Center Innsbruck, Medical University of Innsbruck, Innsbruck, Austria; 4University of Texas Southwestern, Dallas, TX, USA; 5Hoag Memorial Hospital Presbyterian, Newport Beach, CA, USA; 6Georgetown University, Washington, DC, USA; 7Huntsman Cancer Institute, University of Utah, Salt Lake City, UT, USA.
Background: SSTR substypes are collectively expressed in the majority of NENs. However, SSTR subtype expression is not routinely assessed for clinical decision-making, including patients eligible for targeted radionucleotide therapy. Elucidating the landscape of SSTR subtypes in context of molecular profiles for low-grade (LG-) and high-grade NENs (HG-NENs) provides an opportunity to better tailor targeted therapy. Here, we leverage the ability of transcriptomics to predict NEN grade, while identifying molecular landscapes associated with SSTR subtype expression in Pancreatic NENs (PanNENs).
Methods: 1768 cases of NENs were analyzed using Next Generation Sequencing(NextSeq), Whole Exome or Whole Transcriptome Sequencing(WTS, NovaSeq) at Caris Life Sciences(Phoenix, AZ). Significance was determined using chi-square, Fisher-Exact or Mann-Whitney U and p-adjusted for multiple comparisons (q<0.05).
Results: Using Receiver-Operating Characteristic analysis on 318 cases with histological grade annotation (hga), we identified a threshold of MKI67 expression which differentiated LG- from HG-NENs, with a true positive rate of 86.84% and false positive rate of 11.9%(AUC=95%). This threshold was applied to the entire cohort to infer HG/LG. The differences between the mutational landscapes of HG- and LG-NENs were faithfully recapitulated in hga- and MKI67-based cohorts, including TP53(delta=58.2%, 42.8%), RB1 (delta=46.6%, 35.2%), KRAS (delta=14.8%, 10%), and MEN1 (delta=-18.4%, -10.8%). Further, the expression of SSTR-1, -2 and -3 were lower, while -4 was higher in HG- vs LG-PanNENs with a similar trend in TP53, RB1, KRAS and MEN1 alterations (all q<0.05) as mentioned above. For each SSTR subtype, we established high and low cohorts based on their median mRNA expression. Among SSTR-1,-2-high vs low HG-PanNENs, the mutational prevalence of MEN1 (delta=29.5%, 34.4%), ATRX (delta=16.5%, 30.4%), and TSC2 (delta=16.5%, 30.4%) were increased, while KRAS (delta=-35%, -37%) and RB1 (delta=-35%, -41%), were decreased (all at least P<0.05). Similar, but less pronounced differences were observed in LG-PanNENs. Gene Set Enrichment Analysis revealed increased adipogenesis, hedgehog and IL-2/STAT5 signaling in HG-PanNENs and increased DNA damage repair and PI3K/AKT/mTOR pathways in LG-PanNENs in the SSTR-1,-2-high cohorts. Finally, only patients with SSTR-5-high LG-PanNENs had significantly better prognosis (HR=0.248, P=0.01).
Conclusion: Here, we provide evidence that WTS can be effectively leveraged to predict NEN grade and lay the foundation for defining characteristic differences in the molecular landscapes associated with specific SSTR subtypes in HG- and LG-PanNENs. Incorporating molecular profiling in this manner can assist in tailoring treatment for patients with PanNENs.
Abstract ID 21474