Goblet cell appendiceal adenocarcinoma is molecularly distinct from other histologic subtypes: Insights from large-scale genomic, transcriptomic, and organoid modeling

Background: Appendiceal Adenocarcinoma (AA) is a rare cancer, though recent data indicate a rising incidence, now exceeding 1 per 100,000 individuals annually. Goblet cell adenocarcinoma (GCA) of the appendix, which accounts for ~15% of all AA, is associated with worse survival than non-goblet AA. To elucidate the unique tumor biology of GCA, we systematically compared the genomic and transcriptomic landscapes of GCA versus non-GCA AA, identifying GCA-specific driver mutations and dysregulated pathways that distinguish GCA from other AA histologic subtypes. Methods: Whole-exome sequencing (WES) data from 1,202 AA tumors were generated as part of the standard workflow for personalized, tumor informed circulating tumor DNA testing (Signatera TM ), including 385 patients with GCA, 450 with mucinous, 206 with enteric-type, and 161 with signet ring cell histology. Transcriptomic profiling was available for 63 tumors (n=25 GCA, 38 non-GCA) using Altera TM tumor genomic profiling test (Natera, Inc.). Isogenic organoids with gene knock-out (KO) were generated using CRISPR/Cas9 in a KRAS/GNAS co-mutant murine tumor. Results: There were striking differences in the somatic mutation landscapes between GCA and non-GCA tumors, with frequent mutations in KRAS (73.4%), TP53 (39.5%), and GNAS (36%) in non-GCA tumors. In contrast, GCA did not harbor a single gene mutation in more than 20% of the cohort. In GCA, the most commonly mutated gene was ZFP36L2 (18%), a zinc finger protein that binds and stabilizes mRNA transcripts, and has previously been reported to inhibit cell proliferation. Other commonly mutated genes in GCA included TP53 (17%) and SMAD4 (10%), with only 10% of tumors harboring KRAS mutations. Transcriptional regulators such as WTAP (9%), SOX9 (8%), ARID1A (7%), MLLT4 (7%), KMT2D (7%), and KMT2B (4%) were also frequently altered. Performing Gene Set Enrichment Analysis (GSEA) on the differentially expressed genes between GCA and non-GCA tumors revealed that the bile acid and adipogenesis pathways were upregulated in GCA. Survival analysis showed worse OS in GCA vs non-GCA AA (94.7 months vs not reached; HR 3.51, 95% CI 2.67-4.63; p < 0.0001). To explore the potential significance of ZFP36L2, organoids derived from ZFP36L2 KO showed decreased doubling time and allowed propagation of the organoid culture without doxycycline-induced expression of GNAS R201C , a known oncogenic driver in AA. Transcriptomic analysis comparing ZFP36L2 KO to parental lines showed enrichment for androgen response, EMT, TNFalpha, Wnt, Hedgehog, KRAS, and TGFbeta signaling. Conclusion: Clear molecular differences between GCA and the more common mucinous AA were observed. It is clear that strategies for developing new drugs for AA must treat GCA as a distinct entity and not merge these patients with other histological subtypes, as has frequently been done previously.