Iqbal, Syma

4107 Background: Aberrant Wnt/β-catenin activation has been implicated in tumor formation and progression in BTC. CTNNB1 is a key transcriptional co-activator in Wnt signaling. The impact of CTNNB1 alterations on outcomes in intrahepatic (IHCC) vs extrahepatic (EHCC) vs gallbladder (GB) tumors and patterns of gene co-expression is unclear. We examined the molecular correlates and predictive and prognostic significance of CTNNB1 expression in a real-world cohort of pts with BTC. Methods: 7450 BTC tumors were analyzed using Next Generation Sequencing (NextSeq), Whole Exome and Whole Transcriptome Sequencing (NovaSeq) at Caris Life Sciences (Phoenix, AZ). Top quartile transcripts per million (TPM) expression was considered high expressors (Q4), bottom quartile was considered low expressors (Q1) within each subtype. Real-world overall survival (OS) data was obtained from insurance claims. Hazard ratio (HR) was calculated using Cox proportional hazards model, P values were calculated using log-rank test. Significance was determined to be p <0.05. Chi-square and Mann-Whitney tests determined molecular differences between subgroups and adjusted for multiple comparisons (q<0.05). Results: The cohort was comprised of 61% IHCC, 14% EHCC, 23% GB, 2% NOS. CTNNB1 mutations are rare in BTC (1.3%). Pts with CTNNB1-wt tumors had significantly better OS vs CTNNB1-mt (13.6 vs 10.2 mo, HR 0.74, p=0.008). This association remained significant in IHCC (13.2 vs 4.5 mo, HR 0.64, p=0.031) but not EHCC (18.0 vs 11.8 mo, HR 0.87, p=0.64) or GB (12.9 vs 11.0 mo, HR 0.75, p=0.071). CTNNB1-mt tumors were more frequently TP53-mt (63% vs 41%), ERBB2-mt (9% vs 2%) and ATM-mt (9% vs 3%). CTNNB1-wt tumors were more likely BAP-mt (9% vs 0%) and IDH1-mt (10% vs 0%) (all q<0.05). CTNNB1-mt tumors had higher median CTNNB1 expression vs wt (154.9 vs 113.4, q<0.009). Pts with CTNNB1 Q1 expression had significantly better OS vs Q4 (14.5 vs 12.4 mo, HR 0.88, p=0.007). This association held in IHCC (15.0 vs 10.6 mo, HR 0.77, p<0.00001) but not EHCC (17.5 vs 19.9 mo, HR 1.05, p=0.70) or GB (12.4 vs 12.8 mo, HR 1.05, p=0.63). CTNNB1 Q4 expressing tumors were more likely to be TP53-mt (55% vs 34%), PIK3CA-mt (9% vs 5%), KRAS-mt (23% vs 12%) and less likely IDH1-mt (4% vs 14%) compared to Q1 (all q<0.05). CTNNB1 Q4 pts had higher ARID1A, CTLA4, LAG3, HIF1A, TGFB1/2/3, EPHB4, EPHA2, VEGFA expression (FC: 2.1-4.2, all q<0.05) and higher T-cell inflamed score vs Q1. There was a trend towards improved OS in pts with IHCC CTNNB1 Q1 tumors receiving gemcitabine or fluorouracil (17.5 vs 15.0 mo, HR 0.86, p=0.064); no difference in outcomes by CTNNB1 expression in pts receiving immunotherapy. Conclusions: CTNNB1 mutations and mRNA expression levels impact survival in BTC, especially IHCC, and may be associated with benefit from chemotherapy. CTNBB1 alterations are associated with immunogenic, DNA repair and angiogenic pathways and may help identify novel therapeutic strategies for BTC.

4106 Background: The WNT/beta catenin (CTNNB1) pathway plays an integral role in the development of HCC. CTNNB1 has been implicated in HCC progression, metastasis, and drug resistance. We examined associations between CTNNB1 mutations and mRNA expression and clinical outcomes in a real-world cohort of patients with HCC. Methods: 1652 HCC tumors were tested at Caris Life Sciences (Phoenix, AZ) and analyzed with Whole Transcriptome Sequencing (WTS; Illumina Novaseq), Whole Exome Sequencing (NovaSeq,WES) and NextGen DNA sequencing (NextSeq, 592 genes). mRNA expression (transcripts per million) was further stratified into top (Q4) and bottom quartiles (Q1). Kaplan Meier estimates were calculated for overall survival (OS) in the molecularly defined cohorts and estimated from time of tissue collection to last contact. Significance was determined to be p <0.05. Chi-square and Mann-Whitney tests determined molecular differences between subgroups and adjusted for multiple comparisons (q<0.05). Results: Pathogenic CTNNB1 mutations were present in 32% of HCC and mutation status was not associated with CTNNB1 expression level. Pts whose tumors had lower CTNNB1 expression had significantly improved OS, 17.9 vs 12.4 months, Q1 vs Q4, HR 0.72, CI: 0.58-0.89 p=0.002. This association remained significant in patients who received first-line IO (Q1 vs Q4: OS 16.7 vs 10.6 mo, HR 0.54, CI: 0.32-0.94, p=0.025) or TKI (Q1 vs Q4: OS 27.1 vs 17.6 mo, HR 0.60, CI: 0.38-0.94, p=0.025). Low CTNNB1-expressing tumors (Q1) had more frequent ARID1A mutations (15% vs 8%); less frequent TP53 mutations (31% vs 42%); lower VEGFA, EPHB4, EPHA2, HIF1A, TGFB1/2/3 expression, lower MAPK activation and lower T-cell inflamed scores vs Q4 tumors (all q < 0.05). All tumors were MSS. CTNNB1 mutation status (MT vs WT) did not impact survival (OS 17.6 vs 15.3 mo, CTNNB1-MT vs CTNNB1-WT, HR 1.05, CI: 0.91-1.20, p=0.55). Similarly, CTNNB1 mutation status (MT vs WT) did not impact OS in pts treated with IO (19.8 vs 19.6 mo, HR 0.99, p=0.94) or TKI (23.0 vs 22.0 mo, HR 0.86, p=0.33). Conclusions: CTNNB1 mRNA expression, but not CTNNB1 mutation status, is associated with survival in HCC. Patients whose tumors had lower CTNNB1 expression appeared to derive more benefit from immune checkpoint inhibitors and TKI therapy in first line. CTNNB1 expression is associated with DNA repair, immune, neuronal and angiogenic pathways which may pave the way for potential therapeutic opportunities. Further studies are needed to prospectively evaluate CTNNB1 as a biomarker for treatment selection in HCC.

4027 Background: DKN-01 is an anti-DKK1 mAb which has demonstrated anti-tumor activity in patients with advanced GEA with low tumor PD-L1 expression, a subset with very limited therapeutic options. DKN-01 has immunomodulatory activity, stimulates a pro-inflammatory tumor microenvironment, and upregulates PD-L1 levels. Here we present 2-year survival data for 1L advanced GEA patients who received combination treatment with DKN-01, the Fc-optimized anti-PD-1 tislelizumab, and CAPOX chemotherapy. The ORR was previously reported (68% in ITT and 79% in PD-L1-low populations). Methods: This Phase 2, multi-center, single arm Part A of the DisTinGuish (NCT04363801) study investigated DKN-01 + tislelizumab + CAPOX as 1L therapy in advanced HER2(-) GEA, regardless of DKK1 and PD-L1 expression levels. Tumor DKK1 and PD-L1 were assessed by central laboratories. The primary endpoint was ORR in a modified intent to treat population (>1 dose DKN-01); secondary endpoints included PFS and OS in the ITT population. Results: Twenty-five patients were enrolled from September 2020 to April 2021. As of 23 January 2023: Median age was 61 years (22, 80); 17 patients had GEJ adenocarcinoma; 8 had gastric cancer. Twenty-one patients had tumors with evaluable DKK1 expression. Twenty-two patients had tumors with evaluable PD-L1 expression; the majority (73%) were low expressors (vCPS <5%). Median (m) duration on treatment is 11.3 months (mo). Seven patients remain on study, with 4 on-treatment beyond 2 years. The mPFS is 11.3 mo in the overall ITT population and 10.7 mo in patients with low tumor PD-L1 expression. The mOS is 19.5 mo in the overall ITT population and 18.7 mo in the patients with low tumor PD-L1 expression. Treatment related adverse events (TRAEs) were mild with most G1/2. The most common AEs related to the study drug regimen were nausea (72%), diarrhea (64%) and fatigue (60%). Five patients experienced G3 DKN-01-TRAEs including decreased neutrophil count (1), diarrhea (1), vomiting (1), hypophosphatemia (2), and pulmonary embolism (1). Conclusions: At 2-years of follow-up, 1L treatment of advanced GEA patients with DKN-01 in combination with tislelizumab and CAPOX resulted in a prolongation of PFS and OS compared with the modern SOC regimen of nivolumab plus chemotherapy, both in the overall population (mPFS 11.3 vs 7.7 mo; mOS 19.5 vs 13.8 mo) and in the PD-L1 low subgroup (mPFS 10.7 vs 7.5 mo; mOS 18.7 vs 12.4 mo), alongside a manageable safety profile. The prolonged PFS and OS observed in the current study are notable, especially in our trial population dominated by patients with tumors of low PD-L1 expression, where the known benefit of anti-PD-1 therapy is limited. The ongoing Part C of this study is evaluating mFOLFOX6/CAPOX plus tislelizumab with or without DKN-01 in the same 1L GEA patient population. Further evaluation of biomarkers is also ongoing. Clinical trial information: NCT04363801 .

TPS4135 Background: Gastric cancer is a prevalent and morbid illness for which new treatment strategies are needed. Peri-operative therapy is a standard approach that is associated with only a 15% improvement in overall survival. Early FDG-PET scanning, performed prior to cycle 2, can distinguish patients (pts) responding to chemotherapy from those who are not. FDG-PET non-responding pts have poorer survival. This study addresses the question of salvage therapy in pts who are PET non-responders. Specifically, does salvage chemotherapy with docetaxel /irinotecan improve pt survival in FDG-PET-nonresponding pts with locally advanced gastric cancer. Methods: This is a multicenter, cooperative group, NCI supported, randomized phase II study of salvage chemotherapy + surgery versus surgery and post-operative chemoradiotherapy in pts who are FDG-PET non-responders (defined as a decrease in SUVmaxof the primary tumor of less than 35%). A total of 176 pts with locally advanced, resectable gastric cancer who were FDG-PET non-responders to cycle 1 of platinum/capecitabine based chemotherapy will be randomized in a 1:1 manner to receive either (Arm A) surgery followed by fluoropyrimidine-sensitized radiotherapy (4500 cGy), or (Arm B) salvage chemotherapy with docetaxel / irinotecan (DI) for 2 cycles followed by standard resection and 3 cycles DI post-operatively. DI is administered as D 30 mg/m2, I 50 mg/m2 given on D1, D8 of a 21 day cycle. 416 pts will be screened to yield 162 non-responders (81 FDG-PET non-responders per treatment group). With expected 120 events, the study will have 80% power to detect a hazard ratio of 0.625 for improved survival at the one-sided significant level of 0.15. Pts will be required to provide tissue at the time of resection, as well as whole blood prior to resection for correlative studies associated with platinum sensitivity, FDG avidity, and prognostic markers. This study is available through all cooperative groups (SWOG, ACRIN/ECOG, Alliance, and NRG) and the National Clinical Trials Network. Enrollment began in November 2015. Support: U10CA180821, U10CA180882; Clinical Trial Information: NCT02485834 .