Appendiceal cancer is a rare, orphan disease with no therapies currently approved by the FDA for its treatment. Given the limited data regarding drug efficacy, these tumors have historically been treated with chemotherapy designed for colon cancer. However, an overwhelming body of molecular data has demonstrated that appendiceal adenocarcinoma is a distinct entity with key molecular differences from colon cancer, notably rare APC mutation. Recognizing that APC loss-of-function is thought to contribute to taxane resistance, and that taxanes are effective in the treatment of other gastrointestinal tumors including gastric, esophageal, and small bowel adenocarcinoma, we completed a single-center retrospective study to assess efficacy. In a cohort of 13 patients with metastatic appendiceal adenocarcinoma, treated with taxane chemotherapy the median overall survival was 8.3 months. Of 10 evaluable patients we observed 3 responses, 4 patients with stable disease, and 3 with progression (30% response rate, 70% disease control rate). The results of this study showing activity of taxane-based chemotherapy in appendiceal adenocarcinoma support further clinical investigation of taxane therapy in this orphan disease.

Background: The GNASR201 mutation is the single most frequent cancer-causing mutation across all heterotrimeric G proteins. This gain of function mutation in GNAS drives oncogenesis in appendiceal, colorectal, and gastric adenocarcinoma, as well as Intraductal Papillary Mucinous Neoplasms (IPMN) and Small Cell Lung Cancer (SCLC). In this study, we investigated the role of GNAS in tumor growth using peritoneal models of colorectal cancer (CRC). Methods: GNAS was knocked out in multiple GNASR201C/H mutant colon cell lines (KM12, SNU175, and SKCO1) and overexpressed in GNASWT LS174T cells. Isogenic pairs of KM12 and LS174T cells were injected into the peritoneum of NSG mice to study the role of GNAS in cell line derived xenograft (CDX) models of peritoneal metastasis. Cell lines and CDX were profiled with RNAseq, reverse phase protein assay (RPPA), and immunohistochemistry (IHC) to identify potential pathways regulated by mutant GNAS. Identified mediators of GNAS signaling were then validated using chemical inhibitors of PKA (H-89) and β-catenin (LF3). Results: GNAS knockout significantly decreased 2D colony formation by KM12 (68%), SNU175 (76%), and SKCO1 (85%) cells (all p < 0.0001), and decreased 3D organoid area of KM12 cells by 62% (p = 0.043). There was significant increase in colony formation by LS174T cells overexpressing GNASR201C (193%, p = 0.016) and GNASR201H (170%, p =0.0037). Mice injected with KM12 GNAS-knockout tumors exhibited a significant 68% reduction in tumor growth (n = 6, p = 0.016) and were more likely to survive at 7 weeks (0% vs. 100%, p = 0.0007) relative to those with parent KM12-GNASR201H tumors. Likewise, mice injected with GNASR201H LS174T cells showed significant increase in tumor growth (934% vs. 100%, n = 3, p = 0.042). Histology of GNAS-knockout tumors showed a marked decrease in mucinous stroma and increased lytic necrosis, suggesting an interaction between oncogenic GNAS signaling and the peritoneal environment. GNAS is known to stimulate adenylate cyclase and GNAS knockout decreased levels of cyclic AMP in KM12 cells, confirming an on-target effect. RPPA and RNAseq profiling of KM12 and LS174T PDX tumors identified phosphorylation of β-catenin and activation of Wnt/β-catenin targets (NES = 1.25, 1.24) as critical downstream effects of mutant GNAS signaling, confirmed by a 4.9-fold (p = 0.001) increase in nuclear β-catenin intensity in LS174T GNASR201H tumors (19% vs. 99% nuclei stained positive, p = 0.0002). Chemical inhibition of both PKA and β-catenin reduced growth of GNAS mutant organoids by 79% (p = 0.004) and 67% (p = 0.007) respectively, supporting the involvement of the cAMP/PKA and β-catenin pathways in mutant GNAS signaling. Conclusions: Our findings demonstrate oncogene addiction to GNAS in peritoneal models of GNASR201C/H tumors, which signal through the cAMP/PKA and Wnt/β-catenin pathways. Thus, GNAS and its downstream mediators are promising therapeutic targets for GNAS mutant tumors. Citation Format: Aditya More, Valsala Haridas, Ichiaki Ito, Saikat Chowdhury, Yue Gu, Natalie W. Fowlkes, John P. Shen. Oncogene addiction to GNAS in GNASR201 mutant tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 822.

223 Background: Goblet Cell tumors (GCTs) of the appendix are a rare, distinct, and under studied malignancy. Since 2019 the preferred World Health Organization (WHO) terminology is Goblet Cell Adenocarcinoma (GCA), but previously many terms have been used to describe these tumors including Goblet Cell Carcinoid and Adenocarcinoma ex goblet cell carcinoid as these tumors have a histological appearance that blends neuroendocrine and adenocarcinoma features. Historically goblet cell tumors have been considered one of the more aggressive subtypes of appendiceal cancer, but limited data exists and is mostly in the form of case reports. Here we present the retrospective analysis of a large single institution cohort. Methods: The internal database of the University of Texas MD Anderson Cancer Center (MDACC) was queried to identify all patients diagnosed with goblet cell appendiceal tumor. Patients were classified to two different histopathological groups, GCA (n=220) and GCA with signet ring adenocarcinoma (SRA) (n=146). Clinical, histopathological, and molecular data were extracted from the database in semi-automated fashion. Survival analysis were performed using Kaplan Meier methodology. Results: 366 patients with GCTs were identified from 1986 to 2022. 132 (36%) patients were seen during the last five years, with an average of 26 patients per year. Median follow up time was 54 months, while median age at diagnosis was 57 years. Tumor grade data was available for 294 patients. 95% of the patients had high grade tumors (moderately, moderately to poorly and poorly-differentiated) (n= 278), and 5% had low grade tumors (well and well to moderately-differentiated (n=16). The median overall survival was 85 months, and significantly different between the two groups, 118 months for the GCA group and 57 months for the GCA with SRA (p= 0.003). Lymph node (LN) status was known for 168 patients, rate of LN involvement was 53% (n=89) and significantly different between the two groups with 41% (n=39) for GCA and 68% (n=50) for GCA with SRA (p= <0.0006). The internal database of MDACC was queried for LN status of Mucinous adenocarcinoma (MA) (n=242) and SRA (n=104) for comparison purposes, rate of LN positivity was 13% in MA and 76% in SRA. Node positive patients had significantly worse overall survival with median overall survival of 51 months vs 85 months for node negative patients (p<0.004). By multivariate analysis, both LN status and SRA component were independent predictors of overall survival. 107 patients had gene mutation analysis tested, TP53, SMAD4, GNAS and KRAS were the most commonly mutated with 13%, 9%, 4%, and 3% respectively. Conclusions: This study highlights the heterogenicity of GCTs of the appendix and the importance of the histopathological classification in this distinct entity. GCT are much more likely to spread to LN and have a distinct somatic mutation profile relative to MA.

221 Background: Due to the rarity of appendiceal adenocarcinomas (AA), systematic study of these tumors has been limited. Thus, guidelines for the diagnosis and treatment of AA are often based on other related tumor types such as colorectal cancer. However, given that AA has been shown to be molecularly and functionally distinct, there is a need for focused clinical data to guide disease management. In AA, tumor marker levels are used by some practitioners to monitor response to treatment and aid in diagnosis. This study evaluates the association of elevated tumor marker levels with survival outcomes. Methods: The MDACC database was queried to identify patients with AA between 1997 to 2022. Patients with reported values for the tumor markers CA-125 (n=1076), CA 19-9 (n=1060), and CEA (n=1249) were then selected for analysis. Elevation of tumor markers was defined as above the laboratory upper limit of normal (CA-125 > 37 U/mL, CA 19-9 > 37 U/mL, and CEA > 3 ng/mL and survival outcomes were compared with a log-rank (Mantel-Cox) test. This analysis was repeated while controlling for tumor grade, which was defined by low-grade: well, well to moderately differentiated and high-grade: moderate, moderate to poor, and poorly differentiated. Results: Elevated CA-125 was predictive of overall survival in all patients with median survival not-yet-reached for those with normal CA-125 vs. 87.4 months for those with elevated CA-125 (HR: 5.8, p < 0.0001). Similarly, elevated CA 19-9 and CEA were also predictive of overall survival (HR: 2.8, 4.6, respectively, p < 0.0001 for each). Given that tumor grade is the primary driver of prognosis in AA, survival analysis was repeated while controlling for tumor grade. While elevated levels of all tumor markers were predictive of overall survival for both low-grade and high-grade tumors, elevated CA-125 was an especially strong predictor of survival in patients with high-grade tumors (OS: 69.8 months vs. not-yet-reached, HR: 14.3, p < 0.0001). Moreover, high-grade patients with elevated CA-125 had a reduced 5-year survival rate of 56% vs. 91%. Elevated CA-125 also stratified 5-year survival rate in low-grade patients (83% vs. 99%). Elevated levels of CA 19-9 and CEA were strongly predictive of overall survival for patients with low-grade tumors (HR: 10.6, 26.8, respectively, p < 0.0001 for each). Notably, patients with low-grade tumors expressing normal levels of CA 19-9 or CEA had excellent 5-year survival rates of 99% and 100%, respectively. Conclusions: These results highlight the utility of using tumor marker levels in conjunction with tumor grade to more accurately predict prognosis in AA. CA 19-9 and CEA were particularly useful indicators of outcome in patients with low-grade AA, while CA-125 had greatest prognostic value in high-grade tumors.