BACKGROUND:Small bowel adenocarcinoma (SBA) is molecularly distinct from colorectal and gastric cancers, yet treatment typically parallels colorectal cancer. We evaluated the activity of taxane-based therapy in the largest SBA cohort to date. METHODS:We retrospectively reviewed SBA patients treated with taxane-based chemotherapy at MD Anderson Cancer Center from 1994 to 2024. Eligible patients had pathologic confirmation, received >1 treatment cycle, and had tumor response evaluation. Survival analyses were analyzed using Kaplan-Meier and Cox proportional hazard models. RESULTS:Seventy patients were identified. Median age was 57, and 59% were male. Primary sites were duodenum (44%), jejunum (34%), and ileum (16%). Metastatic sites included peritoneum (39%) and liver (31%). Common mutations were TP53 (63%), KRAS (47%), SMAD4 (24%), and APC (15%). Taxanes were administered as single agents (29%) or in combination (71%), most often in second- (40%) or third-line (33%) settings. Overall response rate was 24%. Median time to progression (mTTP) was 3.1 months (95% CI: 2.0-4.2) and median overall survival (mOS) was 8.7 months (95% CI: 7.4-10.1). Efficacy did not differ by treatment line, regimen type, or tumor site, but was significantly associated with TP53 status; response rate was 20% in TP53-mutated vs 45% in wild-type (P = .009), with mTTP 2.5 vs 4.9 months (P = .009) and mOS 7.3 vs 10.6 months (P = .002). On multivariable analysis, TP53 mutation predicted worse outcomes. CONCLUSION:Taxane-based therapy demonstrated activity in metastatic SBA, with 24% response and 3.1-month mTTP. TP53 mutation may be a negative predictive marker for taxane efficacy. These findings support prospective investigation in metastatic SBA.

Abstract Background: Pancreatic ductal adenocarcinoma (PDAC) remains highly lethal even in localized stages, with frequent recurrence despite curative-intent resection and perioperative chemotherapy. Personalized neoantigen vaccination offers a promising strategy to elicit tumor-specific T cell responses against micrometastatic disease in the adjuvant setting. Our personalized peptide-based vaccine platform - termed NeoAg-VAX - targets up to 10 neoantigens per patient and has demonstrated favorable safety and immunogenicity across solid tumors, including lung and colorectal cancer (Li F et al, JITC 2021; Haldar SD et al, AACR 2025). However, PDAC has a profoundly “cold,” immunosuppressive microenvironment that can limit effective priming and durability of vaccine-induced antitumor immunity, supporting rational combinations with PD-1 blockade and CD40 agonism to enhance antigen presentation and reinvigorate effector function. Methods: This study is a single-center, non-randomized, open-label, investigator-initiated phase 1 trial of NeoAg-VAX in combination with anti-PD-1 pembrolizumab +/- CD40 agonist APX005M in patients with surgically resected PDAC. Whole exome/RNA sequencing of resected tumor tissue is performed with HLA binding prediction to identify patient-specific neoantigen peptides for vaccine formulation. Key eligibility criteria include complete R0/R1 resection, receipt of >1 line of standard chemotherapy, adequate organ/marrow function, and ECOG 0-1. The co-primary endpoints are safety and feasibility. Secondary endpoints include immunogenicity, ctDNA clearance, RFS, and OS. Vaccines are given subcutaneously (SQ) with topical imiquimod adjuvant and APX005M (0.3 mg/kg IV + 250 µg SQ) on weeks 0, 1, 3, 4, 6, 9, 12, 15, 18, 21, and 24. Pembrolizumab 200 mg IV is given every 3 weeks during weeks 3-24. Although drug supply for APX005M was terminated in August 2023, accrual for vaccine plus anti-PD-1 therapy has continued thereafter. Translational studies evaluating vaccine-induced ctDNA dynamics and functional neoantigen-specific T cell responses are in progress. As of January 2026, 6 patients have been vaccinated in this cohort with follow-up and analyses ongoing (NCT02600949). Citation Format: S. Daniel Haldar, Jason Willis, Arjun Katailiha, Amjad Talukder, Brandon Smaglo, Nicole Balmaceda, Camila Braganca Xavier, Dan Zhao, M. Pia Morelli, Ryan Huey, Chandrikha Chandrasekharan, Florencia McAllister, Shubham Pant, Jane Thomas, Anirban Maitra, Scott Kopetz, Greg Lizee, Michael J. Overman. Trial in progress: A phase I study of personalized neoantigen vaccination combined with PD-1 blockade and CD40 agonism in resected pancreatic ductal adenocarcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(8_Suppl):Abstract nr CT091.

256 Background: Immunotherapy has shown the ability to induce a high rate of pathologic and clinical complete response in patients with deficient mismatch repair (dMMR)/microsatellite instability-high (MSI-H) colorectal cancer (CRC). The role of circulating tumor DNA (ctDNA) as a prognostic biomarker for survival in patients who receive non-operative management (NOM) is unknown. Among patients enrolled in a phase II trial (NCT04082572) that evaluated the efficacy of pembrolizumab in patients with locally advanced dMMR/MSI-H solid tumors, we evaluated the impact of ctDNA on progression-free survival (PFS) among 12 patients with locally advanced dMMR/MSI-H CRC who received NOM. Methods: ctDNA testing was performed in the CLIA-certified laboratory at MD Anderson using a 70-gene liquid biopsy panel (LBP-70), which used digital sequencing of cell-free DNA. Formalin-fixed paraffin-embedded tumor samples and germline peripheral blood mononuclear cells were profiled using next-generation sequencing. Mutations in ctDNA were considered tumor-specific if they were confirmed with matched tumor tissue profiling. The lower limit of detection of the LBP-70 assay was a mutational variant allele frequency of <0.3%. "ctDNA clearance" and “ctDNA(-)” status were defined as the absence of any detectable tumor-specific mutations in ctDNA. Two-year progression-free survival (PFS) was estimated using Kaplan-Meier techniques and compared with log-rank tests. Results: The median follow-up between the first cycle of pembrolizumab and last radiographic assessment was approximately three years (36.7 mo; range (range 1.2 – 51.0 mo). Forty-nine LBP-70 assays were performed among the 12 patients over the course of treatment with pembrolizumab. The median number of cycles was 16 (range 1 – 16 cycles). Six patients were ctDNA(-) prior to pembrolizumab, all of whom remained ctDNA(-) after. None of these six patients experienced disease progression (PD) at their last follow-up. The other six patients were ctDNA(+) prior to pembrolizumab (median baseline VAF 0.4%, range 0.3%–4.0%). Of these six patients, three experienced ctDNA clearance while receiving pembrolizumab. None of the three patients who cleared ctDNA experienced PD as of their last follow-up. Of the three patients who did not clear ctDNA, two experienced PD in their primary tumor (at two and six months after their first cycle, respectively), while the third patient remains without PD at their last follow-up. The two-year PFS rate was 100% among the nine patients who were ctDNA(-) after pembrolizumab compared to 33% among the three patients who were ctDNA(+) after pembrolizumab (p = 0.03). Conclusions: ctDNA may be a valuable tool for assessing treatment response and improving the ability to tailor organ-sparing, curative strategies to patients with locally advanced dMMR/MSI-H CRC.

e16067 Background: The incidence of brain metastasis in gastrointestinal (GI) malignancies is rare, estimated at less than 1%. However, its occurrence is associated with devastating outcomes. With the advent and prolonged use of Anti-HER2 therapy in patients expressing HER2 in their primary malignancy, there appears to be an increasing trend in brain metastasis, though the reason is unknown. Methods: This study included patients with a confirmed diagnosis of primary GI cancer treated at MD Anderson Cancer Center between 2000 and 2024. Data was extracted from electronic health records (EHR) and the tumor registry using the Palantir Foundry platform (Palantir Technologies, Denver, CO). Patients were included if their primary malignancy exhibited positive HER2 expression (IHC 3+) or ERBB2 amplification. The cohort was divided into two groups based on treatment received: (1) those who received Anti-HER2 therapy and (2) those who did not receive Anti-HER2 therapy. The Institutional Review Board (IRB) of MD Anderson Cancer Center approved the study. Results: Out of 83,329 patients with gastrointestinal cancers, HER2 testing was performed on 13,153 individuals, identifying HER2 positivity or ERBB2 amplification in 571 patients: 49% with gastroesophageal cancer, 37% with colorectal cancer, 11.5% with Biliary tract cancers, 1% with appendiceal cancer, 0.8% with small intestinal cancer, and 0.7% with anal cancer The cohort of 571 patients comprised 68% males (387 patients) and 32% females (184 patients), with a median age at diagnosis of 58 years (range 21–89 years).Among the 177 patients receiving anti-HER2 therapy, (trastuzumab and pertuzumab or trastuzumab deruxtecan) 33 (18%) developed brain metastasis, while 20 (5%) of the 396 patients not receiving anti-HER2 therapy developed brain metastasis, p < 0.001 (RR = 3.7). Of the 33 patients with brain metastases in the anti-HER2 therapy group, 31 (94%) had metastases to other sites, while among the 20 patients with brain metastases in the non-anti-HER2 therapy group, 15 (75%) had metastases to other sites, p = 0.08. The median time between disease diagnosis and the development of brain metastases was 19 months for the group receiving anti-HER2 therapy and 13 months for the group not receiving anti-HER2 therapy, with a statistically non-significant p-value of 0.7. The median duration from the start of anti-HER2 therapy to the development of brain metastasis was 14 months (range 1 to 86). Conclusions: These findings suggest a significantly higher incidence of brain metastasis in patients treated with anti-HER2 therapy (18%) compared to those who did not receive this treatment (5%). These findings underscore the need for further research to understand the mechanisms driving this trend and to explore strategies for early detection and prevention of brain metastasis in this population.

Novel immune strategies are needed to improve outcomes in patients (pts) with microsatellite-stable metastatic colorectal cancer (MSS mCRC). Our personalized peptide vaccine platform - termed NeoAg-VAX - targets up to 10 tumor-derived neoantigens per pt. We hypothesize that this vaccine combined with anti-PD-1 can enhance antitumor immunity and revert the cold status of the MSS CRC tumor microenvironment (TME). This is a single-center, non-randomized, open-label, investigator-led pilot study of NeoAg-VAX +/- pembrolizumab (pembro) in MSS mCRC pts. Tumor antigens were selected based on predicted HLA-binding affinity using whole-exome/RNA sequencing and tandem mass spectrometry with bioinformatics tools. Key eligibility criteria: biopsiable disease, > 1 line of standard therapy, ECOG 0-1. Cohort A: NeoAg-VAX alone. Cohort B: NeoAg-VAX + pembro 200 mg IV Q3W. Vaccines were given subcutaneously with topical imiquimod adjuvant on weeks 0, 1, 3-6, 9, 12, 15, 18, 21, 27, 30, 39, and 51. Primary endpoints: feasibility, toxicity. Secondary endpoints: ORR, PFS, 12-week PFS rate, T cell response by IFNγ ELISPOT. Spatial proteomic analysis of TME immune features was performed using imaging mass cytometry (IMC) of pre-treatment tumors. A total of 28 pts were treated with NeoAg-VAX +/- pembro (Cohort A: n=13, Cohort B: n=15). Of note, 8 pts from Cohort A were sequentially retreated in Cohort B. In 24 of 28 pts, NeoAg-VAX was manufactured within 12 weeks of enrollment with at least 1 dose given, indicating a feasibility rate of 85.7%. Most treatment-emergent adverse events were grade 1-2 in severity, including pain (50.0%), immune-related hypothyroidism (17.8%), and arthralgias (10.7%). Two pts achieved a partial response per RECIST (ORR 7.1%), while 15 pts experienced a best response of stable disease. Median PFS was 2.9 months with a 12-week PFS rate of 57.1%. IFNγ ELISPOT of peripheral blood showed robust induction of neoantigen-specific T cell responses with a median 5.9-fold increase (range: 1.3-18.1) compared to baseline. IMC of pre-treatment tumors (n=9) revealed heterogeneity in immune cell composition across different metastatic sites (6 lung, 2 liver, 1 ovary). Neighborhood analysis revealed shorter distances between T cells and tumor cells as well as longer distances between myeloid cells and tumor cells in lung vs liver mets. Lastly, there was a positive correlation between the magnitude of vaccine-induced T cell response and density of intratumor T cells (r=0.72, p=0.043), B cells (r=0.88, p=0.0043) and macrophages (r=0.72, p=0.044) at baseline. Personalized immunotherapy with NeoAg-VAX +/- anti-PD-1 was feasible, safe, and immunogenic in MSS mCRC pts. Although clinical responses were modest, spatial proteomics revealed key features of tumor immune contexture that may inform future therapeutic approaches. Saurav Daniel Haldar, Gayathri Anandappa, Arjun Katailiha, Amjad Talukder, Sarah Shin, Jae Lee, Erin Coyne, Alens Valentin, Sapna Parshottam, Chantale Bernatchez, Jane Thomas, Jason Roszik, Anirban Maitra, Scott Kopetz, Elizabeth Jaffee, Won Jin Ho, Gregory Lizee, Michael Overman. Personalized neoantigen vaccine with or without pembrolizumab in patients with microsatellite-stable metastatic colorectal cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_2):Abstract nr CT012.

Abstract Immune checkpoint inhibitors improve survival in patients with mismatch repair deficiency/microsatellite instability-high (MSI-H) colorectal cancer. The recurrence outcomes following discontinuation of immunotherapy after prolonged disease control have not been definitively reported in large series. Records from patients with advanced MSI-H colorectal cancer from The University of Texas – MD Anderson Cancer Center who received immunotherapy between 2014 and 2022 and stopped after prolonged clinical benefit were reviewed. Median progression-free and overall survival were estimated. Associations between the event of recurrence and coexisting mutations (KRAS/NRAS, BRAFV600E), metastatic organ involvement (lung, liver, lymph node, or peritoneum), metastatic timing (synchronous vs. metachronous), prior immunotherapy [anti-PD-(L)1 alone or in combination with anti-CTLA antibodies], etiology of MSI status (sporadic vs. hereditary non-polyposis colorectal cancer), and duration of immunotherapy were assessed. Sixty-four patients with MSI-H colorectal cancer without progression on immunotherapy were reviewed. Of these 48 and 16 received anti-PD(L)1 antibody alone or in combination with anti-CTLA-4 antibody, respectively. Median exposure to immunotherapy was 17.6 months (range, 1.3–51.9). After a median follow-up of 22.6 months (range, 0.3–71.7) after stopping immunotherapy, 56 of 64 patients (88%) remained without disease progression. Lung metastases were associated with recurrence/progression (OR, 6.1; P = 0.04), but coexisting mutation, primary tumor sidedness, and immunotherapy were not. These data provide a retrospective, single-institution analysis that showed that most patients with advanced MSI-H colorectal cancer do not recur after treatment cessation, regardless of the reason for stopping treatment or a variety of patient and disease features, supporting an optimistic prognosis of sustained disease control. Significance: Outcomes for patients with MSI-H colorectal cancer stopping immunotherapy after disease control remain unknown. Sixty-four patients with MSI-H colorectal cancer from our institution stopping treatment for sustained benefit or toxicity were retrospectively assessed. After median follow up of 22 months and median immunotherapy exposure of 18 months, 88% patients remained without progression. All patients who recurred or progressed and were rechallenged with immunotherapy have continued to experience disease control.

Characteristics of postimmunotherapy recurrence/progression

Background Neoadjuvant immune checkpoint blockade (IO) is emerging as a therapeutic option for patients with deficient mismatch repair (dMMR) colorectal cancer (CRC) given high pathological response rates. The aim of the study was to characterise imaging and endoscopic response to IO.Methods A retrospective analysis of patients with localised dMMR CRC that received at least one cycle of neoadjuvant anti-PD-1 therapy was conducted. Endoscopy, imaging, and pathological outcomes were reviewed to determine response to treatment according to standardised criteria.Results Thirty-eight patients had received IO for the treatment of localised CRC (median eight cycles). Among evaluable cases (n = 31 for endoscopy and n = 34 for imaging), the best endoscopic response was complete response (CR) in 45% of cases, and the best radiographic response was CR in 23% of cases. Imaging CR rate after <= 4 cycles of IO (n = 1) was 6% compared to 44% after >4 IO cycles (n = 7). Among 28 patients with imaging and endoscopy available, a discrepancy in best response was noted in 15 (54%) cases. At a median follow-up of 28.2 months from IO start, 18 patients underwent surgical resection of which 11 (61%) had pathological CR (pCR). Despite pCR or no evidence of progression >= 6 months after completion of IO among non-operatively managed patients, 72% and 42% of patients had non-CR on imaging and endoscopy, respectively.Conclusions Discrepancies between imaging and endoscopy are prevalent, and irregularities identified on these modalities can be identified despite pathological remission. Improved clinical response criteria are warranted.

KM curves for stopping immunotherapy due to disease control or toxicity

PURPOSE Pembrolizumab significantly improves clinical outcomes in advanced/metastatic microsatellite instability high (MSI-H)/deficient mismatch repair (dMMR) solid tumors but is not well studied in the neoadjuvant space. METHODS This is a phase II open-label, single-center trial of localized unresectable or high-risk resectable MSI-H/dMMR tumors. Treatment is pembrolizumab 200 mg once every 3 weeks for 6 months followed by surgical resection with an option to continue therapy for 1 year followed by observation. To continue on study, patients are required to have radiographic or clinical benefit. The coprimary end points are safety and pathologic complete response. Key secondary end points are response rate and organ-sparing at one year for patients who declined surgery. Exploratory analyses include interrogation of the tumor immune microenvironment using imaging mass cytometry. RESULTS A total of 35 patients were enrolled, including 27 patients with colorectal cancer and eight patients with noncolorectal cancer. Among 33 evaluable patients, best overall response rate was 82%. Among 17 (49%) patients who underwent surgery, the pathologic complete response rate was 65%. Ten patients elected to receive one year of pembrolizumab followed by surveillance without surgical resection (median follow-up of 23 weeks [range, 0-54 weeks]). An additional eight did not undergo surgical resection and received less than 1 year of pembrolizumab. During the study course of the trial and subsequent follow-up, progression events were seen in six patients (four of whom underwent salvage surgery). There were no new safety signals. Spatial immune profiling with imaging mass cytometry noted a significantly closer proximity between granulocytic cells and cytotoxic T cells in patients with progressive events compared with those without progression. CONCLUSION Neoadjuvant pembrolizumab in dMMR/MSI-H cancers is safe and resulted in high rates of pathologic, radiographic, and endoscopic response, which has implications for organ-sparing strategies.

3610 Background: Early data suggests that immunotherapy (IO) for localized colorectal cancer (CRC) is associated with high rates of pathologic complete response (pCR) and durable remission. This study aims to characterize endoscopic, imaging, and pathological outcomes among patients who received IO for localized CRC. Methods: This was a single-institution retrospective analysis of patients with MSI-H CRC who received at least one cycle of anti-PD-1 therapy for localized (stage I-III) MSI-H CRC. Patient and tumor characteristics were obtained from the clinical record. Endoscopy reports were assessed using standardized criteria to determine response and imaging was assessed using RECIST v1.1 where applicable. Endoscopic complete response (endoCR) was characterized by flat, white scar and/or telangiectasia without residual ulcer or nodularity. Categorical variables were compared using Fisher’s exact test. Results: 37 patients who received a median of 8 cycles of pembrolizumab (n = 36) or nivolumab (n = 1) for localized CRC were identified, of which 16 (43%) had a rectal primary. 10 patients (27%) had mucinous features on pathology. The best endoscopic response was complete response (CR) in 13 of 30 (43%) patients, which was achieved after a median of 6 cycles. The rate of endoCR after ≤ 4 cycles of IO was similar to the rate of endoCR after > 4 cycles (29% vs 50%, p = 0.284). The best imaging response was CR in 6 of 32 (19%) of patients. The rate of CR on imaging after ≤ 4 cycles and > 4 cycles of IO was 6% and 36% respectively (p = 0.064). In cases where imaging and endoscopy were available (n = 24), discrepancy in best response was noted in 14 (58%) cases. The most common discrepancy was partial response on imaging when endoCR was noted (n = 6 cases). 16 patients proceeded to surgical resection (after a median of 8 cycles) of which 11 (69%) had pCR despite non-CR on either endoscopy (n = 7) or imaging (n = 9). The endoscopic reports indicating residual disease prior to surgery included findings such as ulceration, scarring, strictures, and an obstructing mass. Of 15 patients who did not proceed to surgery and had follow-up more than 6 months after completion of IO, 13 (87%) had no evidence of progression at a median follow-up of 19.3 months from IO start. Conclusions: Discrepancies between endoscopic, imaging, and pathological outcomes were frequent, indicating a need for improved clinical response criteria. Nevertheless, immunotherapy was associated with deep and durable responses in most patients with localized MSI-H CRC.

BACKGROUND:This study aims to assess predictive markers for response to immunotherapy in dMMR/MSI-H metastatic colorectal cancer (mCRC) patients. MATERIALS AND METHODS:A study using two prospective cohorts from MD Anderson Cancer Center and Sheba Medical Center of consecutive patients with dMMR/MSI-H mCRC that were treated with immunotherapy between 2014-2022. Primary outcome was progression-free survival (PFS) and secondary outcome was overall response rate (ORR). Evaluated predictors included ECOG-PS score, RAS/BRAF status, single-agent versus doublet immunotherapy, metastatic sites, disease burden, and CEA levels prior to treatment initiation. Kaplan-Meier analysis and Cox proportional hazard regression model were used to analyze the effect of exposure variables on PFS. RESULTS:The study included 153 patients. Median follow-up time was 26 months (IQR 11-48). Median PFS was 51.6 months (95%CI 38.1-NR) and ORR was 58.1%. In a univariate analysis, male sex was associated with worse PFS with a HR of 1.67 (95% CI 1.00-2.79); Right-sided tumors were associated with improved PFS with a HR of 0.56 (95% CI 0.32-0.97); Liver or lung metastasis were associated with worse PFS with HRs of 2.35 (95%CI 1.43-3.88) and 2.30 (95%CI 1.31-4.04), respectively; ECOG-PS score ≥ 2, CEA levels ˃5 μg/L prior to treatment initiation and ≥ 3 metastatic sites were associated with worse PFS with HRs of 2.09 (95%CI 0.98-4.47), 2.23 (95%CI 1.30-3.81) and 3.11 (95%CI 1.61-6.03), respectively. Liver or lung metastasis remained significant in a multivariable model. CONCLUSIONS:Extent of disease (worse PFS with high CEA, poor ECOG-PS and ≥3 metastatic sites) and disease location (worse PFS with liver or lung metastasis and left sided tumor) were associated with immunotherapy outcome in dMMR/MSI-H mCRC.

Background Monotherapy with immune checkpoint blockade is ineffective for patients (pts) with microsatellite stable (MSS) metastatic colorectal cancer (mCRC). This study investigates whether the combination of trametinib (T) with durvalumab (D) can alter the immune tumor microenvironment (TME) by successfully priming and activating T-cells.Methods Open-label, single-center, phase II trial with primary endpoint of immune-related response rate for combination of T+D in refractory MSS mCRC pts (NCT03428126). T is 2 mg/day orally starting 1 week prior to D, which is given 1500 mg intravenously every 4 weeks. Simon 2-stage design used to enroll 29 pts into first stage, requiring a response in two or more pts to proceed to stage 2. Tumor biopsies were collected at baseline (BL) and early on-treatment (OT) at week 4.Results Twenty nine treated pts include 48% females, median age 48 years (range 28–75), and median prior therapies 2 (range 1–5). No grade (G) 4 or 5 treatment-related adverse events (TRAE). The most common TRAE of any grade was acneiform rash, 17% being G3. One of 29 pts had confirmed partial response (PR) lasting 9.3 months (mo) for an overall response rate of 3.4%. Seven pts had stable disease (SD) and five pts (1 PR, 4 SD) demonstrated decrease in total carcinoembryonic antigen ng/mL (best percentage reduction: 94%, 95%, 42%, 34%, and 22%, respectively). Median progression-free survival was 3.2 mo (range 1.1–9.3 months). Three pts with both liver and lung metastases demonstrated discrepant responses in which clinical benefit was present in the lung metastases but not liver metastases. Comparison of BL and 4-week OT tumor tissue flow cytometry demonstrated no changes in T-cell infiltration but upregulation expression of PD-1 and Tim3 on CD8 T cells. However, expression of PD-1 and Tim3 as single markers and as coexpressed markers was observed to increase OT relative to BL (p=0.03, p=0.06 and p=0.06, respectively).Conclusions T+D demonstrated acceptable tolerability in pts with refractory MSS mCRC. The response rate in the first stage of the study did not meet efficacy criteria to proceed to the second stage. Specific site of metastatic disease may impact outcomes in novel immunotherapy combination trials.Trial registration number NCT03428126.

Immune checkpoint inhibition (CPI) for metastatic colorectal cancer (mCRC) with deficient mismatch repair (dMMR) demonstrates high clinical activity that appears durable but the impact of CPI on pathological tumor response is unknown. In this retrospective analysis, our objective was to assess pathological response and clinical outcomes in dMMR mCRC patients treated with CPI prior to surgical resection of primary and/or metastatic tumor. Amongst 121 advanced dMMR mCRC patients treated with CPI at two institutions between November 2016 and December 2018, 14 underwent surgery. Pathologic complete response was noted in the resected specimens of 13 patients despite the presence of residual tumor on pre-operative imaging in 12 of those patients. With median follow-up of 9 months, no patients have had disease relapse or progression. For this small retrospective study, the data suggests that residual radiographic tumor may not require systematic resection following response to anti-PD1 based therapy. However, larger prospective studies are warranted.

794 Background: Duodenal adenocarcinoma (DA) is a rare cancer with limited data regarding the pattern of disease recurrence following resection. Methods: A retrospective review of 115 patients with Stage I-III DA from 3/1994 to 6/2018, at a single high-volume cancer center was conducted. Only patients (pts) who underwent a potentially curative surgical resection (R0/R1 margins) and had a postoperative follow-up radiographic evaluation were included. Periampullary adenocarcinomas were excluded. Clinicopathologic features and patterns of recurrence were compared across cohorts. Results: Of 76 patients who met inclusion criteria, 7 (9%) were stage I, 25 (33%) stage II, and 44 (57%) stage III. Histologic grade was moderate in 58% and poor in 38%. Median age was 63 years (range, 29-84), 38% were female, and R0 resection was 97%. Neoadjuvant therapy was given to 14% and adjuvant therapy to 61%. Radiation therapy (XRT) as either adjuvant/neoadjuvant therapy was used in 27%. Median follow-up was 44 (6-293) months. Median time to recurrence was 11mo, with 84% of recurrences occurring within 2 years. Median time to local recurrence (LR) vs. distant recurrence (DR) was 11mo vs. 12mo, respectively, p = 0.42. Stage impacted recurrence rate: 0% in stage 1 vs. 50% stage 2 vs. 71% stage 3 (p = 0.002). Median time to recurrence was 16mo for stage II and 11mo for stage III (p = 0.04). In total, 4 (5%) pts had LR only, 8 (10%) had LR concurrent with DR, and 32 (42%) had DR only. Recurrence distribution was similar across stage II (LR 8%, LR+DR 15%, DR 77%) and stage III (LR 10%, LR+DR 19%, DR 71%). LR was similar in patients that received XRT (10%) compared to those who did not (9%). Most common sites of DR were peritoneal (38%), liver (33%), distant lymph nodes (12%), and lung (10%). Conclusions: The recurrence pattern for resected DA is predominantly distant metastatic disease with the majority of recurrences occurring within the first two years. Future therapies should focus on improved systemic therapy, and surveillance should be most intensive in the first two years.

152 Background: Monotherapy with immune checkpoint blockade (ICB) is ineffective for patients (pts) with MSS mCRC. Novel approaches to modulate the tumor microenvironment (TME) are needed. Here, we investigate whether the combination of trametinib (T) with durvalumab (D) can alter the immune TME by successfully priming and activating T-cells. Methods: An open-label, single center phase II trial with primary endpoint of immune-related response rate for T+D in refractory MSS mCRC pts (NCT03428126). T is 2mg/day orally starting 1 week prior to D, which is given 1500mg intravenously every 4 weeks. Dose de-escalation strategy performed to identify maximum tolerated dose (MTD). Simon 2-stage design utilized with plans to enroll 29 pts into the first stage, requiring response in 2 or more pts to proceed to stage 2 (n = 15). Results: Demographics for 29 treated pts: 48% female, median age 48 years (range 28-75), and median prior therapies was 2 (range 1-5). No grade (G) 4 treatment-related adverse events (TRAE). The most common G3 TRAE included autoimmune hepatitis (14%) and acneiform rash (10%). G1/2 TRAE included acneiform rash (69%), fatigue (24%) and anemia (21%). No fatal TRAE and 4 pts discontinued treatment due to TRAE. 1 of 29 pts had confirmed partial response (PR) lasting 9.3 months (mo) for an ORR of 3.4%. This pt had an ATM E221fs*14 mutation. 7 pts had stable disease (SD) with median time to progression (TTP) of 5.4 mo (range 3.9-9.3 mo). 1 pt remains on active therapy with SD ( > 10 mo). 5 pts (1 PR, 4 SD) demonstrated decrease in total CEA ng/mL (best percentage reduction: 94%, 95%, 42%, 34% and 21.6% respectively). Median TTP for the entire cohort was 3.2 mo (range 1.1-9.3 mo). Consensus molecular subtypes (CMS) were performed on the primary CRC in 23 pts: 12 CMS2, 2 CMS3, and 9 CMS4. 4 SD pts were CMS2, 1 SD pt was CMS4 and the CMS status of the pt with a PR was unknown. Conclusions: The combination of T+D did not meet efficacy criteria to proceed to the second stage of the study. Analysis of 15 paired on-treatment biopsies is ongoing and will be presented. Utilization of CMS characterization in mCRC clinical trials is feasible and may provide an improved biological understanding of treatment activity. Clinical trial information: NCT03428126.

Colorectal cancer (CRC) is the second most common cause of cancer-related death in the United States. Despite excellent prognosis for early stage disease, 5-year survival rates in metastatic disease remain low. A small subset of CRC is defined by a deficiency in mismatch repair (dMMR) resulting in high levels of microsatellite instability and are responsive to immunotherapy. Immune checkpoint inhibitors (ICIs) targeting the programmed death 1 (PD-1)/programmed death ligand 1 axis and cytotoxic T-lymphocyte antigen 4 have been explored and show robust clinical outcomes with prolonged progression-free survivals in nonrandomized single-arm clinical trials. On the basis of these data, single-agent therapy with pembrolizumab and nivolumab and combination therapy with nivolumab/ipilimumab have been approved by the US Food and Drug Administration for metastatic CRC that has progressed after treatment with fluoropyrimidine, oxaliplatin, and irinotecan. Ongoing clinical trials are exploring the use of these agents in earlier lines of therapy such as first-line metastatic therapy and adjuvant therapy for stage III CRC. However, resistance to ICIs does occur in a subset of patients and ongoing clinical trials are exploring novel approaches in these PD–1-refractory patients. The aim of this review is to outline the development and decision-making of ICIs in the treatment of dMMR CRC and to discuss ongoing clinical trials in this therapeutic space.