Human epidermal growth factor receptor 2 (HER2)-targeted therapies have been investigated for therapeutic benefit in RAS/BRAF wild-type/HER2+ metastatic colorectal cancer (mCRC). Unlike HER2+ breast and gastric cancer, there are no regulatory criteria for determining HER2 overexpression in patients with mCRC. This systematic literature review describes unmet needs for patients with HER2+ mCRC in relation to testing and treatment, highlights the importance of early HER2 testing at mCRC diagnosis, and discusses the evolving treatment landscape. We utilised PubMed and EMBASE databases up to November 2023 to identify journal articles and published congress abstracts relating to the HER2+ disease and treatment landscape, HER2 testing in mCRC, and HER2-targeted treatments in mCRC. Many studies have demonstrated the utility of immunohistochemistry, in situ hybridisation, and next-generation sequencing (tissue- and circulating tumour DNA-based) for detecting HER2 overexpression/amplification in mCRC and have attempted to establish consolidated criteria like those used for breast and gastric cancer. The value of HER2-targeted treatments in patients with HER2+ mCRC has been evidenced by clinical trials and meta-analyses, with strong evidence from MOUNTAINEER and DESTINY-CRC01 which supports the use of tucatinib + trastuzumab or trastuzumab deruxtecan, respectively, among this patient population. However, real-world analyses have confirmed that patients with HER2+ mCRC are not routinely tested for HER2 overexpression. Strong evidence and clinical guidelines support the value of HER2-targeted treatment among patients with HER2+ mCRC. There is a need for increased awareness and earlier uptake of HER2 testing among patients with mCRC to broaden treatment options and optimise outcomes for this patient population.

Introduction: Pancreatic ductal adenocarcinoma (PDAC) remains a leading cause of cancer-related mortality. Radiological distinctions between borderline resectable (BR) and locally advanced disease (LA) are increasingly recognised as imperfect when considered without dynamic assessment. Neoadjuvant therapy (NAT) improves outcomes through tumour downstaging and early treatment of occult metastatic disease, but the optimal NAT strategy, particularly in BR disease, remains uncertain. Published data evaluating combined systemic anti-cancer therapies (SACT) with or without chemoradiation (CRT) are limited and heterogeneous. Methods: This is a single-centre retrospective analysis of 44 patients with BR PDAC and a comparator cohort of 121 patients with LA PDAC treated with a total neoadjuvant approach of SACT with or without CRT and surgical resection between June 2017 and September 2022. Results: Median overall survival (OS) did not differ significantly between BR and LA disease (18 vs. 16 months, p = 0.14). Following NAT, 47.7% of BR and 18.1% of LA patients were anatomically suitable for surgical resection. Among unresected BR and LA patients, those treated with CRT in addition to SACT had a median OS of 18 and 21 months respectively. In the resected subgroup, resection margin status was the primary factor associated with survival; with R0 resection conferring a substantial OS advantage over R1, irrespective of initial BR/LA classification as diagnosis (47 vs. 22 months, p < 0.001). Conclusions: Despite anatomical differences at diagnosis, BR and LA PDAC demonstrated comparable survival outcomes when treated with total neoadjuvant strategies in this cohort. These findings challenge traditional radiological staging-based treatment paradigms and confirm that a margin-negative surgical resection offered the greatest opportunity for long-term survival for BR/LA PDAC patients.

BRAFV600E-mutant metastatic colorectal cancer (mCRC) is an aggressive molecular subtype associated with poor prognosis and relative chemoresistance. Outcomes following progression on chemotherapy and MAPK-targeted therapy with encorafenib plus cetuximab remain poor, highlighting an unmet need for effective later-line treatments. Immune checkpoint inhibitors have limited activity in microsatellite-stable (MSS) mCRC, and predictive biomarkers beyond mismatch repair deficiency remain poorly defined. We report a patient with BRAFV600E-mutant MSS mCRC who achieved a prolonged response to PD-1 blockade with pembrolizumab following a dynamic increase in tumor mutational burden (TMB). A 58-year-old woman initially responded to first-line chemotherapy but subsequently progressed on multiple treatment lines, including encorafenib plus cetuximab. Longitudinal genomic profiling using tumor tissue and circulating tumor DNA (ctDNA) revealed persistent BRAFV600E signaling with molecular evolution, including acquisition of a PIK3R1 variant associated with resistance to MAPK inhibition and emergence of TMB-high status. Fifth-line pembrolizumab produced a dramatic radiological response, with ongoing disease control more than 2 years later. We discuss the mechanistic framework by which a subset of BRAFV600E-mutant MSS mCRC may respond to immune checkpoint inhibition through an inflamed immune microenvironment driven by constitutive MAPK signaling. This case illustrates the interplay between tumor-agnostic biomarkers such as TMB-high and tumor-specific context, and highlights the value of longitudinal genomic profiling, including ctDNA, to identify resistance mechanisms and guide treatment selection.

Colorectal cancer predominantly affects older adults (ie, those aged 70 years or older). Along with shifting demographic trends, the number of patients with colorectal cancer in this age group is estimated to sharply increase globally over the next few decades. During the past decade, new treatment modalities and systemic therapeutic options have been introduced through well conducted clinical trials that shaped and fundamentally redefined the treatment landscape. However, older adults with reduced physiological reserves, functional impairment, comorbidities, and geriatric syndromes are under-represented in such studies. Yet, for this population, growing evidence for geriatric assessment and management shows patient-centred benefits with the potential to further improve outcomes, parallel to the broadening armamentarium of medical innovations. The purpose of this Review is to summarise the evidence base of the current therapeutic landscape of colorectal cancer and guide management of older adults in daily clinical practice through pragmatic decision making.

The role of perioperative immunotherapy as a chemotherapy-free option for patients with resectable mismatch repair-deficient/microsatellite instability-high (dMMR/MSI-H) colon cancer is evolving, with early-phase neoadjuvant studies reporting favorable long-term outcomes. AZUR-2 is an ongoing global, Phase III, open-label, randomized study evaluating the efficacy of perioperative dostarlimab monotherapy compared with standard of care (SoC) (adjuvant chemotherapy or surveillance) in adult patients with previously untreated, pathologically confirmed, radiologically evaluable T4N0 or Stage III resectable dMMR/MSIH colon adenocarcinoma. Patients will be randomized 2:1 to receive neoadjuvant dostarlimab 500 mg every 3 weeks (4 cycles), followed by surgery, then adjuvant dostarlimab 1000 mg every 6 weeks (6 cycles), or to receive immediate surgery followed by SoC. The primary endpoint is event-free survival assessed by blinded independent central review. The key secondary endpoint is overall survival; additional secondary endpoints include pathological response assessed by residual viable tumor determined by local assessment, safety, and tolerability.Clinical trial registration: NCT05855200 (www.clinicaltrials.gov).

3556 Background: Liver resection is routinely performed for oligometastatic colorectal cancer (CRC). ctDNA analysis is becoming important in the management of primary CRC following surgery, but in patients with liver metastases, its role is less clear. This study aimed to evaluate the performance of a plasma ctDNA assay for early detection of recurrence in patients with liver only CRC metastases post surgical resection. The cohort included patients who had inoperable disease prior to induction chemotherapy and those having repeat resections. Patients with the primary cancer in situ were included if a synchronous resection was performed. Methods: 61 patients (42 male, median age 68 [range 41-85]) were studied. All patients had CRC liver metastases as the sole metastatic site on evaluation by CT and MRI. Patients were followed up clinically with regular CT imaging. Plasma samples were collected at baseline (pre-surgery, n = 55) and 5 days to 1+ year after surgery, and were tested retrospectively on a blood-only, tumor-agnostic, targeted methylation assay that interrogates informative methylation regions for signal indicative of ctDNA using a machine learning classifier (GRAIL, Inc., Menlo Park, CA). Results: ctDNA was detected in 94.5% (52/55) of patients at the pre-surgery timepoint. During clinical surveillance, 46 developed recurrence, 5 had no recurrence for at least 2 years after surgery, and 10 had follow-up for less than 2 years. Among patients with recurrence, ctDNA was detected in 80.4% (37/46) of patients before recurrence, with a median lead time of 245 days [IQR: 65 - 393 days]. All five patients who remained recurrence-free for at least 2 years had undetectable ctDNA following surgery. ctDNA was detected in 57.8% (26/45) of patients who recurred and had plasma collected within 3 months following surgery. Recurrence-free survival (RFS) was significantly lower in patients with ctDNA detected within 3 months after surgery (median RFS 272 vs. 593 days, log-rank p-value < 0.01). Conclusions: ctDNA analysis using a methylation-based assay allowed the detection of minimal residual disease and the prediction of recurrence in many patients following resection of CRC liver metastases. These data support the potential use of this biomarker for patient management pending confirmation in prospective clinical studies.

BACKGROUND:PLATFORM is an adaptive phase II trial assessing maintenance therapies in advanced oesophagogastric adenocarcinoma (OGA). We evaluated maintenance capecitabine in patients with disease control after first-line chemotherapy. METHODS:HER2-negative patients with advanced OGA who had response or stable disease after 18 weeks of first-line chemotherapy were randomised (1:1) to surveillance or capecitabine. The primary endpoint was progression-free survival (PFS); secondary endpoints included overall survival (OS) and safety. RESULTS:Between May 2015 and May 2024, 266 patients were randomised (129 surveillance, 137 capecitabine). Median follow up was 70.7 months. Capecitabine significantly improved PFS (HR 0.69; 95% CI 0.54-0.89; p = 0.002), with median PFS of 5.0 vs 2.8 months. One-year PFS rates were 19.9% vs 6.8%; and two-year rates 8.1% vs 4.3%. No OS difference was observed (median OS: 10.5 vs 10.0 months; HR 0.87; 95% CI 0.67-1.12; p = 0.143). One and two-year OS rates were similar (1-year: 44.1% vs 45.7%; 2-year: 18.8% vs 16.8%). Grade ≥3 adverse events were more frequent with capecitabine (46% vs 29%), with 21% experiencing grade 3 treatment related events. DISCUSSION:Maintenance capecitabine significantly prolonged PFS compared to surveillance, meeting the primary endpoint and supporting its use to extend disease control in advanced OGA.

Introduction: Increasingly, identification of BRAF mutation in colorectal cancer is used to guide management and predict cancer behaviour. There is, however, still significant diversity within this cohort of patients, both in terms of clinical phenotype and treatment outcomes. This may be explained, at least in part, by differences between classes of BRAF mutations and the presence of concomitant mutations. Methods: We present a retrospective cohort study of sequential patients diagnosed with BRAF-mutated (V600 and non-V600) colorectal cancer between 2014 and 2022. Information regarding presentation, treatment outcomes and molecular subtype was identified using the electronic medical record. Results: This study included 406 patients with BRAF-mutated colorectal cancer, 253 (228 V600BRAF) of whom had localised disease and 153 (137 V600BRAF) with metastatic disease at the time of diagnosis. In patients with localised disease at diagnosis, the V600BRAF mutation was associated with older median age (73 vs. 63 years, p = 0.04) and a higher prevalence of right-sided primary (73% vs. 40%, p < 0.01), mismatch repair deficiency (56% vs. 8%, p < 0.01), and faster time to disease relapse (p = 0.006). In the metastatic setting, non-V600BRAF mutation was associated with a higher prevalence of KRAS mutation (27% vs. 1%, p < 0.01), NRAS mutation (14% vs. 3%, p = 0.04) and PIK3CA mutation (33% vs. 8%, p = 0.02). Mismatch repair deficiency was more common in patients with V600BRAF mutations than in those with non-V600BRAF mutations (20% vs. 0%, p = 0.01). The median survival of patients with the V600BRAF mutation was 14 months, and 34 months in those with non-V600BRAF mutations. Concomitant RNF43 mutation in metastatic disease, was associated with a significantly higher incidence of disease control from combined BRAF and EGFR inhibition, when compared to those without an RNF43 mutation (100% vs. 54%, p = 0.02). Conclusions: Presentation and outcomes of BRAF-mutated colorectal cancer are heterogenous. The type of BRAF mutation, and the presence of concomitant RNF43 mutation, may explain some of the differences in cancer behaviour. Routine reporting of RNF43 mutations would assist clinicians to give more personalised treatment recommendations.

PLATFORM is an adaptive phase II study assessing maintenance therapies in advanced esophagogastric adenocarcinoma (OGA). We evaluated the role of capecitabine plus a vascular endothelial growth factor receptor 2 inhibitor ramucirumab (cape-ram) in these patients. Human epidermal growth factor receptor 2 (HER2)-negative patients with advanced OGA with stable or responding disease after 18 weeks of induction platinum-based chemotherapy were randomly assigned 1:1 to surveillance or cape-ram. The primary end point was progression-free survival (PFS), and key secondary end points were overall survival (OS) and safety. Recruitment to the cape-ram arm closed prematurely because of industry support withdrawal. A one-sided log-rank test with a 2.5% significance level was considered significant. Between April 2019 and November 2022, 25 surveillance and 22 cape-ram patients were contemporaneously randomly assigned. Median follow-up was 24.4 months. Compared with surveillance, cape-ram significantly prolonged PFS (hazard ratio [HR], 0.33 [95% CI, 0.17 to 0.63], P < .001; median PFS: 2.5 months with surveillance versus 5.5 months with cape-ram; 6-month PFS rate: 4% [95% CI, 0.3% to 17.0%] v 42.9% [95% CI, 21.9% to 62.3%], respectively) and OS (HR, 0.51 [95% CI, 0.26 to 1.00], P = .023; median OS: 7.1 months with surveillance v 14.4 months with cape-ram; median OS from start of induction chemotherapy was 12.1 months v 19.5 months, respectively). Of 10 cape-ram patients with measurable disease, 1 had an incremental partial response. Grade ≥3 adverse events (AEs) were seen in 32% surveillance and 57% cape-ram patients. Six cape-ram patients had grade 3 treatment-related AEs, and no new safety signals were identified. Maintenance cape-ram after induction chemotherapy for patients with HER2-negative OGA significantly improved survival compared with surveillance. To our knowledge, this is the first randomized maintenance study demonstrating survival benefit and provides support for maintenance treatment.

Objectives The ‘tumour, node, metastasis’ (TNM) classification of colorectal cancer (CRC) predicts prognosis and so is vital to consider in analyses of patterns and outcomes of care when using electronic health records. Unfortunately, it is often only available in free-text reports. This study aimed to develop regex-based text-processing algorithms that identify the reports describing CRC and extract the TNM staging at a low computational cost.Methods The CRC and TNM extraction algorithms were iteratively developed using 58 634 imaging and pathology reports of patients with CRC from the Oxford University Hospitals (OUH) and Royal Marsden (RMH) NHS Foundation Trusts (FT), with additional input from Imperial College Healthcare and Christie NHS FTs. The algorithms were evaluated on a stratified random sample of 400 OUH development data reports and 400 newer ‘unseen’ OUH reports. The reports were annotated with the help of two clinicians.Results The CRC algorithm achieved at least 93.0% positive predictive value (PPV), 72.1% sensitivity, 64.0% negative predictive value (NPV) and 90.1% specificity for primary CRC on pathology reports. On imaging reports, it demonstrated at least 78.0% PPV, 91.8% sensitivity, 93.0% NPV and 80.9% specificity. For the main T/N/M categories, the TNM algorithm achieved PPVs of at least 93.9% (T), 97.7% (N) and 97.2% (M), and sensitivities of 63.6% (T), 89.6% (N) and 64.8% (M). NPVs were at least 45.0% (T), 91.1% (N), 88.4% (M), and specificities 95.7% (T), 98.1% (N), 99.3% (M). Reductions in performance were mostly due to implicit staging. For extracting explicit TNM stages, current or historical, the algorithm made no errors on 400 pathology reports and six errors on 400 imaging reports.Conclusion The TNM algorithm accurately extracts explicit TNM staging, but other methods are needed for retrieving implicit stages. The CRC algorithm is accurate on non-supplementary reports, but outputs need additional review if higher precision is required.

To evaluate the feasibility, safety, and efficacy of image-guided percutaneous cryoablation for the treatment of soft-tissue metastatic tumours in patients with oligoprogressive disease. Consecutive patients undergoing percutaneous cryoablation between March 2017 and December 2024 were identified from a prospectively maintained database. Patient demographics, disease characteristics, procedural details, and outcomes were recorded. The Kaplan–Meier method was used to calculate local tumour progression-free survival (LTPFS), progression-free survival (PFS), and overall survival (OS). Technical success, technique efficacy, complications, and oncologic outcomes were analysed. The primary endpoints were technical success, major complications, LTPFS, and PFS. The secondary endpoint was OS. Fifty-two metastatic tumours (median size: 20 mm [interquartile range, 12–36 mm]) were treated across 46 sessions in 38 patients. The technical success rate (defined as complete tumour coverage by the ablation zone with a 5-mm margin) was 100

391 Background: Although perioperative FLOT therapy is a standard for locally advanced oesophagogastric adenocarcinoma (LA-OGA), completing all cycles, especially in the adjuvant chemotherapy (ACT) phase, is often challenging. This study examined the survival impact of treatment delivery in LA-OGA patients receiving FLOT therapy. Methods: 423 patients who underwent radical resection at The Royal Marsden Hospital from 2017 to 2023 were screened. Major inclusion criteria included patients with LA-OGA (cT2≤ and Nany or Tany and N+), treated with FLOT (at least one cycle of neoadjuvant chemotherapy) and radical resection, with an ECOG performance status of 0–2. Patients were divided into therapy incomplete (Tx Incomp, less than eight cycles of FLOT) and therapy (Tx comp) groups. We performed univariate and multivariate analyses, as well as propensity score matching (PSM) analysis, to evaluate whether treatment delivery was associated with recurrence-free survival (RFS) and overall survival (OS). Kaplan-Meier curves and restricted mean survival times (RMST) up to 36 months were used to evaluate the survival time. The primary endpoint was the 3-year RFS and OS rate. Results: Of the screened patients, 210 met the inclusion criteria, with 79 (38%) in the Tx Incomp and 131 (62%) in the Tx comp group, and 50 (24%) patients did not receive ACT. The median follow-up time was 26.5 months. The 3-year RFS and OS rates in the Tx Incomp and Tx comp groups after PSM were 54% vs 59% (p=0.14) (RMST difference: 4.04 months, p=0.09) and 58% vs 78% (p=0.04) (RMST difference: 6.15 months, p<0.001), respectively. Multivariable analysis showed a significant impact of Tx comp on OS (HR 0.53, 95% CI: 0.32-0.88). In the subgroup of ypN-positive patients, a significant difference of RMST in OS rate was observed between those who started ACT and those who did not start ACT (RMST difference: 7.89 months, p=0.01), whereas no significant difference was found in the ypN-negative group (RMST difference: 0.65 months, p=0.75). Conclusions: This study suggests that completing FLOT therapy is associated with better OS. The impact of ACT might differ according to ypN stage.

388 Background: Malnutrition and weight loss were reported as poor prognostic factors in operable OGA. This study examines the relationship between oncological outcomes and nutritional status in LA-OGA patients under the current standard of care, perioperative FLOT therapy. Methods: From 2017 to 2023, 423 patients who had radical resection at Royal Marsden Hospital were screened. Inclusion criteria included LA-OGA (cT2≤ and Nany or Tany and N+), at least one cycle of neoadjuvant chemotherapy (NAC), and ECOG PS 0–2. Nutritional status was assessed by body weight and prognostic nutritional index (PNI), with significant weight loss defined as a 5% loss within six months before diagnosis or during NAC. PNI was calculated as serum albumin (g/L) + 0.005 × lymphocyte count /μL, with categories of severe (<45), mild to moderate (45–49.9), and normal (≥50). The primary endpoint was 3-year recurrence-free survival (3y-RFS). Restricted mean survival time (RMST) at 36 months was calculated to analyze survival time. Multivariate analyses were performed to identify prognostic factors for RFS. The pathological response was assessed using tumour regression grade (TRG) (Mandard criteria). Results: Of the 423 patients, 210 met the inclusion criteria. Primary tumour sites were the oesophagus (52.4%), the oesophagogastric junction (18.1%), and the stomach (29.5%). The median follow-up was 26.5 months. Weight loss at baseline and after NAC, as well as PNI, did not significantly impact RFS (p=0.13, p=0.91, p=0.69, p=0.45). However, a PNI decrease from baseline was associated with significantly shorter 3y-RFS (46% vs. 69%, p<0.001), and the RMST difference was 5.46 months (p<0.001). Multivariable analyses showed ypN positivity (p<0.001), R1 resection (p<0.001), and PNI decrease (p=0.03) as negative prognostic factors for RFS. Pathological response did not differ regardless of PNI change (28.2% vs. 30.4%, p=0.75). Conclusions: A decrease of PNI is a significant poor prognostic factor for RFS in LA-OGA patients undergoing FLOT, suggesting that maintaining nutritional status during NAC could enhance survival outcomes.

353 Background: For first line (1L) treatment (tx) in HER2- gastric/gastroesophageal junction (G/GEJ) adenocarcinoma (AC), clinical guidelines recommended platinum-fluoropyrimidine (PF) chemotherapy (CT) in patients (pts) with a PD-L1 combined positive score (CPS) <5 and PF CT in combination with immunotherapy (IO) for CPS ≥5. There is little contemporary evidence into IO tx uptake in the real world. This real world study examined biomarker testing, 1L tx characteristics, and reasons for 1L tx choice in pts with HER2- G/GEJ AC to identify unmet needs and opportunities for optimal clinical practice. Methods: Data were derived from the Adelphi G/GEJ Cancer Disease Specific Programme: a systematic approach involving multi-national, multi-faceted data sources including cross-sectional surveys capturing data from physicians and their consulting pts. Physicians reported pt demographics, clinical characteristics, tx received and reasons for tx decisions, via chart review. Data were collected across US and Europe (EU: France, Germany, Spain, the UK) between Oct 2022 to Apr 2023. All pts were alive at data collection and analyses were descriptive. Results: Overall, 243 treating physicians (US n=60/EU n=183) provided data on 642 HER2- pts (US n=83/EU n=559) receiving 1L tx at data collection. For pts from US/EU, median age was 66/67 years, 84%/79% had an ECOG of 0-1 and 52%/75% had metastatic disease at data collection. In the US/EU, 87%/80% of pts were tested for PD-L1 status respectively, of which 63%/69% had a known CPS. For US/EU, 67%/45% of pts (n=30/n=139) had a CPS ≥5. Differences in biomarker testing rates were observed in US v EU pts respectively for MSI/dMMR (66% v 58%), NTRK-gene fusion (31% v 4%) and FGFR2 (27% v 3%). Of US/EU physicians, 70%/85% (n=60/n=183), reported the main reason for conducting biomarker testing was ‘to inform tx decisions’. Cost related concern (50%/39%) and time constraints (23%/15%) were the main perceived barriers to testing. Among pts with PD-L1 CPS <5, 27% of US pts and 2% of EU pts received IO + CT; in pts with CPS ≥5, 63% of US pts and 76% of EU pts received IO + CT. Whilst progression free survival/overall survival benefit were the most common reasons selected by physicians for 1L tx choice across the US/EU (75%/74% and 60%/61%), differences across regions were reported for tx choice based on maintaining and improving pt’s QoL (4%/28%) and pt’s performance status (2%/24%). Conclusions: PD-L1 testing was not performed by 19% of physicians and approximately one third of pts with CPS ≥5 across the US/EU were not receiving IO + CT in 1L. As informing 1L tx choice was a key driver for biomarker testing, there may be scope for improvement in 1L tx by increasing testing rates. Further studies are needed to address barriers to biomarker testing and uptake of IO in pts who can benefit from 1L IO tx.