Background:The incidence of early-onset colorectal cancer (EO-CRC), defined as a diagnosis before 50 years of age, is increasing worldwide. However, its clinical characteristics and outcomes compared to average-onset colorectal cancer (AO-CRC) remain under debate, especially in the setting of locally advanced rectal cancer (LARC). Objectives:This study aimed to compare clinical characteristics, treatment responses, and survival outcomes between patients with early-onset and average-onset locally advanced rectal cancer. Design:A multicenter retrospective cohort study. Methods:We retrospectively analysed 305 patients with stage II-III rectal cancer treated between 2012 and 2022 across three Italian oncology centrers. Patients were categorised as EO-RC (⩽50 years) or AO-RC (>50 years). All patients underwent neoadjuvant chemoradiotherapy followed by total mesorectal excision. Pathological and radiological responses were evaluated, and survival outcomes were assessed through Kaplan-Meier methods. Results:Early-onset patients accounted for 10.5% of the cohort. Clinical and pathological characteristics were broadly similar between groups, although EO-RC patients had a higher prevalence of proficient mismatch repair status. Radiological and pathological response rates were comparable. After a median follow-up of 120 months, the 10-year overall survival (OS) was 73.3% in EO-RC and 91.9% in AO-RC (HR 7.60, 95% CI 2.22-26.06; p = 0.0012). Disease-free survival (DFS) at 10 years was 57.1% in EO-RC and 70.9% in AO-RC (HR 1.80, 95% CI 0.86-3.78; p = 0.1177). Conclusion:Early-onset rectal cancer patients exhibit similar response rates and DFS compared to older patients, but appear to have worse OS. Further studies are needed to explore biological factors and post-recurrence treatment strategies that may influence these outcomes.

Aspirin is ubiquitous, inexpensive, and mechanistically compelling in colorectal cancer (CRC), yet its clinical role has become more ambiguous with—notwithstanding—new data. Neutral phase III results in unselected disease, biomarker-restricted signals of benefit, and guideline shifts driven by bleeding risk have left clinicians without a consistent, setting-specific approach. This review fills that gap by translating the post-precision aspirin literature into a selection-first framework that specifies who may benefit, when off-trial use is reasonable versus discouraged, and why trial enrolment remains the preferred option in key areas of uncertainty. In prevention, aspirin produces modest reductions in adenoma recurrence in selected high-risk surveillance cohorts and shows delayed cancer-endpoint benefit in Lynch syndrome, highlighting that baseline CRC risk and time horizon are decisive. In older adults and lower-risk settings, the time-to-benefit often does not align with a higher, earlier bleeding liability. After curative-intent treatment, randomised trials now define clear boundaries: routine adjuvant aspirin in unselected resected CRC is not supported, whereas recurrence reduction appears confined to PI3K-pathway / PIK3CA-altered localised disease, supporting biomarker-restricted consideration and rigorous prospective validation. In metastatic CRC, existing signals are largely non-randomised and confounded, precluding anticancer-motivated prescribing outside prospective studies. Aspirin in CRC should be targeted rather than routine. The proposed clinical algorithm integrates setting, absolute risk, PI3K/PIK3CA status, bleeding/antithrombotic context, and expected time horizon to maximise net benefit while reducing avoidable harm, and to prioritise trial enrolment where equipoise persists.

370 Background: For patients (pts) with resectable gastric or gastroesophageal junction adenocarcinoma who are candidates for intensive therapy, the FLOT regimen (fluorouracil, oxaliplatin, docetaxel) represents the standard backbone chemotherapy (CT) in perioperative setting. The Italian RealFLOT multicenter observational study investigated the feasibility, safety, and efficacy of perioperative FLOT in a real-world setting. Methods: In this updated analysis of the RealFLOT trial, the long-term feasibility, safety, and efficacy of FLOT were investigated, including endpoints such as disease-free survival (DFS), overall survival (OS), pathological complete response (pCR), and major pathological response (MPR). Survival probabilities were estimated by the Kaplan–Meier method and compared using the log-rank test. Hazard ratios (HRs) with 95% confidence intervals (CIs) were calculated by Cox regression analysis. Results: A total of 307 pts were enrolled, most with stage III disease (65.1%), gastric primary (57.0%), and non-intestinal histology (46.5%). High microsatellite instability (MSI-H) was detected in 13 of 148 pts (8.8% ) and HER2 positivity in 20 of 147 (13.6% ) of tested cases. Among pts receiving perioperative FLOT, 171 (55.7%) completed at least four full-dose cycles. Surgery was performed in 283 cases (92.2%), with 222 (72.3%) subsequently proceeding to postoperative treatment. During the neoadjuvant phase, CT delays occurred in 216 pts (70.6%) and treatment discontinuation in 15 pts (4.9%). Gastrointestinal toxicity represented the most frequent adverse event (65.9% in the preoperative and 67.6% in the postoperative setting, with grade 3–4 events occurred in 24 pts (11.3%). In the postoperative phase, FLOT was administered to 71.1% of pts although delays were reported in 71 pts (32.7%) and treatment discontinuations in 31 pts(14.2%). Among those who underwent surgery, an R0 resection was achieved in 261 of 280 evaluable cases (93.2%). With a median follow-up of 31.2 months (95% CI, 25.5–37.8), median DFS was 25.5 months (95% CI, 17.8–33.1) and median OS was 73.5 months (95% CI, 36.5–not estimable). Pathological CR was achieved in 9.0% of patients, and MPR in 28.2%. Both DFS and OS were significantly improved in pts achieving pCR (HR 0.22; 95% CI, p=0.001 and HR 0.16; 95% CI, p=0.004, respectively) or MPR (HR 0.18; 95% CI, p<0.001 and HR 0.14; 95% CI, p<0.001, respectively). Conversely, pts with advanced postoperative stage achieved shorter mDFS (12.6 vs 73.4 months; HR 3.22; 95% CI, p<0.001) and mOS (23.9 vs 85.2 months; HR 2.94; 95% CI, p<0.001). Notably, OS was significantly prolonged in patients with MSI-H tumors (HR 0.00 [0.01–NE]; 95% CI, p=0.029). Conclusions: Updated real-world evidence confirms the feasibility and efficacy of perioperative FLOT in routine clinical practice. Prognostic outcomes are influenced by pCR, MPR, postoperative stage, and MSI-H status.

Liquid biopsy has emerged as a valuable tool for the detection and monitoring of colorectal cancer (CRC), providing minimally invasive insights into tumor biology through circulating biomarkers such as circulating tumor DNA (ctDNA), circulating tumor cells (CTCs), microRNAs (miRNAs), long non-coding RNAs (lncRNAs), and circular RNAs (circRNAs). Additional biomarkers, including tumor-educated platelets (TEPs) and exosomal RNAs, offer further potential for early detection and prognostic role, although ongoing clinical validation is still needed. This review summarizes the current evidence on the diagnostic, prognostic, and predictive capabilities of liquid biopsy in both metastatic and non-metastatic CRC. In the non-metastatic setting, liquid biopsy is gaining traction in early detection through screening and in identifying minimal residual disease (MRD), potentially guiding adjuvant treatment and reducing overtreatment. In contrast, liquid biopsy is more established in metastatic CRC for monitoring treatment responses, clonal evolution, and mechanisms of resistance. The integration of ctDNA-guided treatment algorithms into clinical practice could optimize therapeutic strategies and minimize unnecessary interventions. Despite promising advances, challenges remain in assay standardization, early-stage sensitivity, and the integration of multi-omic data for comprehensive tumor profiling. Future efforts should focus on enhancing the sensitivity of liquid biopsy platforms, validating emerging biomarkers, and expanding multi-omic approaches to support more targeted and personalized treatment strategies across CRC stages.

Anti-angiogenic drugs represent a milestone of metastatic colorectal cancer (mCRC) pts) treatment. To date, no validated prognostic biomarkers are available. Recently, hematological parameters became of particular interest in solid tumors, including mcRC. Here, we present the results of our research in order to identify potential prognostic factors among clinical and laboratory parameters in mCRC pts receiving anti-angiogenic agents at our Centre. We retrospectively collected clinical and laboratory data of mCRC pts treated with anti-angiogenic drugs at the Medical Oncology Unit of Cagliari University Hospital (2018-04/2024) in order to identify a potential prognostic tool. Statistical analysis was performed with MedCalc (survival distribution: Kaplan-Meier; survival comparison: log-rank test; cut-off: ROC curves). Globally, 42 mCRC pts were included in our study. 25 were male, 17 female; 17 had a RAS wild-type tumor, 30 had a left-sided primary). 14 pts received anti-angiogenic agents in the the first-line, 15 in the second-line (4 bevacizumab and 11 aflibercept) and 13 in the first and further lines. Median OS was 31.4 months (95% CI: 18.3 -39.5). When assessing the correlation between clinical and laboratory parameters and prognosis of the study population, a statistically significant improvement in overall survival (OS) was found in patients with platelet to lymphocyte ratio (PLR) equal or lower than 179.23 (31.4 months [95% CI: 20.3-47.1] versus 10.5 months [95%: CI 6.8 - 39.5], p=0.0315, HR= 0.27) and platelet count equal or lower than 360/μl (31.4 months [95% CI: 20.3-41.8] versus 10.3 months [95%: CI 6.8-10.5], p<0.0001, HR=0.0002). Then, according to these findings, we separated pts in two prognostic groups: good prognostic group (no unfavorable variables) and poor prognostic group (≥1 unfavorable variables). We observed a trend for longer OS in patients belonging to the good prognostic group (31.4 months [95% CI: 20.3-41.8] versus 10.5 months [95% CI: 6.8 - 39.5], p=0.0521, HR=0.3166). Our research showed a promising prognostic role of baseline platelet count and PLR in mCRC patients receiving with anti-angiogenic drugs in a limited population at our center. Platelet to lymphocyte ratio and platelet count: new frontiers as prognostic factors in metastatic colorectal cancer patients receiving anti-angiogenic drugs. Platelet to lymphocyte ratio and platelet count: new frontiers as prognostic factors in metastatic colorectal cancer patients receiving anti-angiogenic drugs [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 4653.

e16407 Background: Recently, there has been an increase in the incidence of early-onset PDAC (EO-PDAC), which is conventionally defined as cancer occurring in adults younger than 50 years. However, clinical and molecular data remain limited and inconsistent. Our study aimed to examine the differences in outcomes and molecular characteristics between EO-PDAC and late-onset pancreatic ductal adenocarcinoma (LO-PDAC) patients. Methods: We conducted a retrospective study involving 486 patients with metastatic pancreatic ductal adenocarcinoma across six different institutions in Italy. The study's primary objective was to evaluate the median overall survival of EO-PDAC patients compared to those with late-onset PDAC. Additionally, we aimed to assess median progression-free survival (mPFS), molecular profiles, and treatment sequences in both subgroups. Results: Of the patients, 80 (16.5%) were early onset and 406 (83.5%) were late onset. The median ages were 46 (±5) and 67 (±8) years, respectively. In the overall population, the median mOS was significantly lower in EO-PDAC patients: 10.0 versus 14.0 months ( p = 0.0103 ). Furthermore, the mPFS was also significantly lower in EO-PDAC patients: 4.0 versus 6.0 months ( p = 0.0132 ). Patients with EO-PDAC exhibited a higher prevalence of KRAS G12D and KRAS G12V mutations, at 31% and 24% respectively, compared to those with LO-PDAC. Finally, a higher percentage of LO-CRC patients continued to a second line (50.7% versus 48.6%). A percentage of 4.18% of LO-CRC patients started a fourth line, compared to 1.25% of EO-CRC. Conclusions: The results of our work showed a worse prognosis for EO-PDAC patients compared to LO-PDAC patients. This is associated with a different molecular profile and a lower percentage of patients reaching subsequent lines. Further research is necessary to better understand this expanding group of patients from molecular and therapeutic perspectives.

BACKGROUND:The FLOT regimen, a combination of fluorouracil, leucovorin, oxaliplatin, and docetaxel, is a standard treatment for gastric and esophagogastric junction cancers, but concerns exist about its potential renal toxicity. The exact prevalence and severity of renal toxicity need to be well documented, and this knowledge gap could impact the optimal use of the FLOT regimen in clinical practice. We aimed to evaluate the renal toxicity profile of the FLOT regimen with a specific focus on acute kidney disease (AKD) onset in a real-life setting and explore associated risk factors. METHODS:We conducted a multicenter retrospective study involving 96 patients treated with preoperatory FLOT. The incidence of AKD and potential risk factors were identified. RESULTS:AKD occurred in 3 patients (incidence rate of 0.0122 cases/month of follow-up). Oral antidiabetic agents and prostate hypertrophy emerged as significant predictors of AKD. CONCLUSION:AKD is uncommon in patients treated with the preoperatory FLOT regimen. Our findings highlight the importance of diligent renal function monitoring and appropriate preventive strategies in patients undergoing FLOT treatment. These results encourage the conduction of further studies and clinical experience in larger populations and patients with lower baseline estimated glomerular filtration rate.

702 Background: In the last decade several studies have shown an increase in the incidence of early-onset PDAC (EO-PDAC), conventionally defined as cancer that occurs in adults between the ages of 18 and 49. Clinical and prognostic data on this setting are limited and conflicting. The aim of our study was to evaluate the clinical, and prognostic differences in a large group of patients with early onset mPDAC. Methods: We retrospectively collected data from 368 patients affected with metastatic pancreatic ductal adenocarcinoma, from 3 different Italian Institutions. All patients had one or more metastatic sites, and received first-line chemotherapy. The main objective of the study was to evaluate the median overall survival in EO-PDAC patients compared to late onset PDAC (LO-PDAC), while secondary endpoint was evaluations of mPFS. Statistical analysis was performed with the MedCalc package. Results: 30 (8,1%) patients were early onset and 338 (91,9%) were late onset; median age was 46 (±5) and 68 (±8), respectively. M/F ratios were 1:1 in both group. In the overall population mOS was significantly lower in EO-PDAC patients: 10,0 versus 15,0 months ( p = 0,03). Furthermore, mPFS was significantly lower in EO-PDAC patients: 5,0 versus 8,0 months ( p = 0,04). ORR obtained from 244 pts were: 11% in EO-PDAC and 24,7% in LO-PDAC. Conclusions: The results of our work, although limited by the retrospective nature of the study, showed a worse prognosis for patients with early onset PDAC compared to late onsets. Further investigations will be needed to better understand this growing group of patients from a molecular and therapeutic point of view.

206 Background: Anti-angiogenic drugs represent a cornerstone of metastatic colorectal cancer (mCRC) patients (pts) treatment. Currently, no prognostic/predictive biomarkers were validated to identify who is more likely to benefit from these agents. We performed at our Centre a retrospective research to assess potential prognostic/predictive factors in this population. Methods: We retrospectively collected laboratory, radiological and clinical data of mCRC pts receiving anti-angiogenic agents at the Medical Oncology Unit of Cagliari University Hospital (2018- 09/2023) in order to assess their correlation with overall survival (OS) and progression free survival (PFS) from the treatment start. Statistical analysis was performed with MedCalc (survival distribution: Kaplan-Meier; survival comparison: log-rank test; cut-off: ROC curves). Results: Globally, 32 mCRC pts were included in our research (19 male, 13 female; 12 RAS wild-type, 21 left-sided primary). 10 received anti-angiogenic drugs in the 1st-line, 12 in the 2nd-line (bevacizumab and aflibercept) and 10 in 1st-2nd line. Median OS was 34.8 months (m) (95%CI:24.7-41.6). We observed a statistically significant improve in OS (34.8 months [m] versus [vs] 8.2 m) in pts with higher monocyte count (>0.3x103/μL; 95%CI 26.5-39.5 vs 95%CI 8.2-24.7, p=0.0103, HR 0.01), lower alcalin phosphatase (ALP; <136 U/l; p=0.0001, HR=0.0001; 95%CI 12-39.5 vs 95% CI 8.2-10.4), lactate dehydrogenasis (LDH; <365 U/l; p<0,0001, HR=0.0000006; 95%CI 24.7-39.5 vs 95% CI 8.2-10.4) and platelet to lymphocyte ratio (PLR; ≤253; p<0.0001, HR= 0.00000062; 95%CI 24.7-39.5 vs 95% CI 8.2-10.4). Median PFS was 14.9 m (95% CI 9.5-18.5). The same factors were significantly related to PFS, which was 18.3 m vs 2.7 m for higher monocyte count (95%CI 11.4-22.1 vs 95%CI 2.7-14, p=0.0386, HR=0.05), lower LDH (95%CI 11.4-22.1 vs 95%CI 2.7-8.9, p=0.0024, HR=0.004), lower PLR (95%CI 11.4-22.1 vs 95%CI 2.7-8.9, p=0.0024, HR=0.004) and 18.3 m vs 8.9 m for lower ALP (95%CI 11.4-23.3, p=0.0369, HR=0.04). Furthermore pts not developing bleeding during treatment showed a significant benefit in OS (39.5 m [95%CI 26.5-41.6] vs 8.2 m [95%CI 8.2-12], p< 0.0001, HR 0.000001761) and improved PFS (17.9 m [95%CI 9.9-18.4] vs 2.4 m [95%CI 2.4-2.7]; p<0.0001). Conclusions: Our findings showed a promising prognostic role of baseline monocytes, ALP, LDH and PLR and absence of bleeding occurrence in mCRC patients receiving anti-angiogenic drugs, even if in a limited population. Further larger prospective studies are encouraged for confirmation.

Despite a reduction of both incidence and mortality from CRC, recent studies have shown an increase in the incidence of early-onset CRC (EO-CRC). Data on this setting are limited. The aim of our study was to evaluate the clinical and molecular profiles of metastatic EO-CRC patients in order to identify differences compared to a late-onset CRC (LO-CRC) control group. We retrospectively collected data from 1272 metastatic colorectal cancers from 5 different Italian Institutions. The main objective was to the evaluate clinical outcome for EO-CRC patients in comparison to patients included in the control group. In the overall population, mOS was 34,7 in EO-CRC pts vs 43,0 months (mo) (p < 0,0001). In the RAS/BRAF mutated subgroup mOS in EO-CRC pts was 30,3 vs 34,0 mo (p = 0,0156). In RAS/BRAF wild-type EO-CRC mOS was 43,0 vs 50,0 mo (p = 0,0290). mPFS was 11,0 in EO-CRC pts vs 14,0 mo (p < 0,0001). Findings indicate a general worse prognosis for patients with early-onset colorectal cancer compared to late-onset patients. Interestingly this seems to occur regardless of the molecular status. These observations might have a considerable impact on clinical practice and research.

70 Background: Despite a reduction in both the incidence and mortality of CRC in the elderly population, recent studies have highlighted an increase in the incidence of early-onset CRC (EO-CRC). Clinical and prognostic data in this context are limited and conflicting. The aim of our study was to evaluate the clinical differences and outcomes of patients with early onset locally advanced rectal cancer. Methods: We retrospectively collected data from 305 patients affected by LARC treated in Italy at the Medical Oncology Units of the University Hospital of Cagliari, Istituto Nazionale dei Tumori Milan, and AOU Ospedali Riuniti Ancona. All patients underwent neoadjuvant chemoradiotherapy. The primary objective was overall survival (OS) while while secondary objectives were overall response rate (ORR) and major TRG. Results: Twenty five (8,2%) pts were EO-RC and 280 (91,8%) were LO-RC. In EO-RC the locations were distributed as follows: 9 (36%) lower rectum, 13 (52%) medium rectum, and 3 (12%) upper rectum. Pathologic responses were as follows: 3 (12%) TRG-0, 6 (24%) TRG-1, 10 (40%) TRG-2, and 6 (24%) TRG-3. While the radiological responses were as follows: 2 (8%) CR, 15 (60%) PR, 7 (28%) SD, and 1 (4%) PD. In LO-RC the locations were: 78 (27,8%) lower rectum, 153 (54,6%) medium rectum, and 49 (17,6%) upper rectum. Pathologic responses were as follows: 29 (10,4%) TRG-0, 60 (21,4%) TRG-1, 159 (56,8%) TRG-2, and 32 (11,4%) TRG-3. Radiological responses were as follows: 25 (8,9%) CR, 155 (61,4%) PR, 64 (22,9%) SD, and 19 (6,8%) PD. The 10 year overall survival was significantly higher in LO-RC patients compared to EO-CRC patients: 92.54% versus 70.83% (p = 0.0005). Conclusions: Compared to LO-RCs, EO-RC patients had more frequent low rectal tumors and higher rates of major pathological responses. Despite this, 10-year OS was found to be inferior in EO-RC. Further studies and insights will be necessary to better understand the biological and clinical characteristics of early-onset tumors.

The ‘immune pattern’ of the esophagogastric junction and gastric cancer (GC) has been related to tumour progression and/or response to therapies. GC can be classified as immunogenic or immune-resistant based on the infiltration of the tumour microenvironment (TME) from different immune cells (ICs), the most significant of which are activated lymphocytes (a-Ly) CD8+ and CD4+. From the amount of a-Ly infiltration in TME, we can identify tumours with high Immunoscore (IS) which correlates with better prognosis in terms of disease-free survival and overall survival (OS). IC activation and consequent immunogenic TME requires high nutrient absorption and synthesis and accumulation of protein, lipid, and nucleotides. However, tumour cells (TCs) promote their own proliferation by stealing micronutrients to ICs, therefore leading to an immunosuppressive TME phenotype. This proof-of-concept protocol aims to assess whether a controlled enteral immune nutrition (EIN) supplementation can revert the TME in favour of ICs over TCs, in both auxotrophic and non-auxotrophic malignancies, witnessed by an increase in IS in surgical specimen over biopsy controls. Secondary endpoints are: (i) tumour regression grade assessment compared to historical data from FLOT4/AIO; (ii) combined proportion score ratio; (iii) relapse-free survival at 3 years and OS; and (iv) quality of life by the European Organisation for Research and Treatment of Cancer quality-of-life questionnaire (EORTC QLQ)-30.

e15575 Background: Even if anti-angiogenic agents are crucial for metastatic colorectal cancer (mCRC) patients (pts) treatment, to date, no validated prognostic biomarkers are available. We conducted at our Centre a retrospective research to identify a prognostic tool to be applied in clinical practice in this population. Methods: We retrospectively collected laboratory, radiological and clinical data of mCRC pts treated with anti-angiogenic agents at the Medical Oncology Unit of Cagliari University Hospital (2018- 02/2024) in order to identify a potential prognostic tool. Statistical analysis was performed with MedCalc (survival distribution: Kaplan-Meier; survival comparison: log-rank test; cut-off: ROC curves; differences among variables: Chi-square test). Results: Globally, 33 mCRC pts were evaluated in our study (19 male, 14 female; 13 RAS wild-type, 22 left-sided primary). 10 were treated with anti-angiogenic drugs in the 1st-line, 13 in the 2nd-line (bevacizumab and aflibercept) and 10 in 1st-2nd line. Median OS was 39.5 months (m) (95%CI:28.2-41.3), median Progression Free Survival (PFS) was 14.9 m (95%CI:9.5-18.5). Pts with lower platelet to lymphocyte ratio (P; ≤253; p < 0.0001, HR = 0,00000048; 95%CI 24.7-39.5 versus [vs] 95% CI 8.2-10.5), alkaline phosphatase (A; < 136 U/l; p = 0.0001, HR = 0.0001; 95%CI 12-39.5 vs 95% CI 8.2-10.5) and lactate dehydrogenasis (L; < 365 U/l; p < 0.0001, HR = 0.0000004; 95%CI 24.7-39.5 vs 95% CI 8.2-10.5) and higher monocyte count (M; > 0.3x103/μL; p = 0.0093, HR 0.01; 95%CI 26.5-39.5 vs 95%CI 8.2-24.7), showed a statistically improved OS. We created the “PALM score” and separated pts in two prognostic groups: good prognostic group (0 unfavorable variables: PALM = 0) and poor prognostic group (≥1 unfavorable variable, PALM = 1). OS was significantly improved in the good prognostic group (PALM = 0): 31.5 m (95%CI:24.7-39.5) vs 10.5 m (PALM = 1) (p = 0.0031, HR = 0.00006). Chi-square test revealed also a statistically significant correlation of upfront resection of primary tumor with the PALM score (73.9% of patients belonging to the PALM = 0 group had undergone surgery for primary tumor and all patients with resected primary belonged to the PALM = 0 group (p = 0.0352). Conclusions: Our research showed a promising prognostic role of easy-to-use PALM score tool in mCRC patients treated with anti-angiogenic drugs in a limited population at our center. Further prospective studies with larger sample size are needed to confirm our findings.

The aggressive nature of gastric cancer often leads to late diagnosis and poor prognosis. Chemotherapy and the more recently added immunotherapy remain key treatments for this disease. Several studies have focused on identifying tissue biomarkers with prognostic and/or predictive roles and therefore the therapeutic options are rapidly growing. In this narrative review, we summarize the major tissue biomarkers routinely assessed in clinical practice. In addition, we focus on new evidence about emerging tissue biomarkers that could have a predictive role in future therapeutic approaches and also on the potential role of liquid biopsy in this neoplasm.

Cluster of differentiation 44 (CD44) is a non-kinase cell surface glycoprotein. It is overexpressed in several cell types, including cancer stem cells (CSCs). Cells overexpressing CD44 exhibit several CSC traits, such as self-renewal, epithelial–mesenchymal transition (EMT) capability, and resistance to chemo- and radiotherapy. The role of CD44 in maintaining stemness and the CSC function in tumor progression is accomplished by binding to its main ligand, hyaluronan (HA). The HA-CD44 complex activates several signaling pathways that lead to cell proliferation, adhesion, migration, and invasion. The CD44 gene regularly undergoes alternative splicing, resulting in the standard (CD44s) and variant (CD44v) isoforms. The different functional roles of CD44s and specific CD44v isoforms still need to be fully understood. The clinicopathological impact of CD44 and its isoforms in promoting tumorigenesis suggests that CD44 could be a molecular target for cancer therapy. Furthermore, the recent association observed between CD44 and KRAS-dependent carcinomas and the potential correlations between CD44 and tumor mutational burden (TMB) and microsatellite instability (MSI) open new research scenarios for developing new strategies in cancer treatment. This review summarises current research regarding the different CD44 isoform structures, their roles, and functions in supporting tumorigenesis and discusses its therapeutic implications.

674 Background: The first-line treatment of metastatic pancreatic cancer involves different therapeutic regimens, among them, mFOLFIRINOX and gemcitabine-nabpaclitaxel. However, it remains to be clarified which sequence might correlate with better outcomes. The aim of our work was to evaluate which sequence leads to better overall survival in this setting. Methods: We retrospectively collected data from 358 patients affected by stage IV pancreatic ductal adenocarcinoma from 3 different Italian Institutions. Patients received mFOLFIRINOX or gemcitabine plus nabpaclitaxel as first-line chemotherapy, followed by a second-line non-cross-resistant chemotherapy (gemcitabine or fluoropyrimidine-based combinations). Survival distribution was assessed by Kaplan-Meier curves. The primary endpoint was median overall survival. Statistical analysis was performed with MedCalc package. Results: The median age was 66 (± 9), 195 (54,4%) were male and 163 (45,6%) were female. 292 (81,5%) patients start gemcitabine plus nabpaclitaxel as first line treatment, while 66 (18,5%) patients start mFOLFIRINOX first-line therapy. No statistically significant differences in terms of mOS were observed between the two mFOLFIRINOX and gemcitabine-nabpaclitaxel groups: 16 versus 15 months, respectively, p = 0,2). No significant differences were found in terms of mPFS either: 7 versus 8 months, respectively, p = 0,3). Conclusions: The results of our retrospective study showed no statistically significant survival advantage between the sequences starting with mFOLFIRINOX or with gemcitabine-nabpaclitaxel. Further studies will be necessary to establish the effectiveness of new therapeutic schemes to be introduced into clinical practice in order to define a better therapeutic algorithm.

3531 Background: Despite a reduction of both incidence and mortality from CRC in the elderly population, several studies published in the last decade have shown an increase in the incidence of early-onset CRC (EO-CRC), conventionally defined as cancer that occurs in adults between the ages of 18 and 49. Clinical and prognostic data on this setting are limited and conflicting. The aim of our study was to evaluate the clinical, prognostic, and molecular profiles of metastatic EO-CRC patients (age at diagnosis ≤ 50) in order to identify potentially relevant differences compared to a control group late-onset CRC (LO-CRC). Methods: We retrospectively collected data from 1272 metastatic colorectal cancers from 5 different Italian Institutions: 693 (54.5%) EO-CRC and 579 (45.5%) LO-CRC as control group. All patients had one or more metastatic sites, molecular profiling available (including RAS, BRAF, and MSI status), and underwent at least one line of treatment for metastatic disease. The main objective of the study was to the evaluate clinical outcome for the global population of EO-CRC patients in different clinical and molecular subgroups according to RAS and BRAF status and in comparison to patients included in the control group. Results: In the EO-CRC group median age was 42.8 (20.0-50.9) and 66.7 (51.0-86.2) in the control group. M/F ratios were 1:1 and 2:1, respectively. In the overall population, mOS was 34,7 in EO-CRC pts vs 43,0 months (mo) ( p < 0,0001 ) in the control group. In the RAS/BRAF mutated subgroup mOS in EO-CRC pts was 30,3 vs 34,0 mo in the control group ( p = 0,0156 ). In RAS/BRAF wild-type subgroup mOS in EO-CRC pts was 43,0 vs 50,0 mo ( p = 0,0290 ). Finally, in the BRAF V600E mutated subgroup EO-CRC pts showed a 16 mo mOS vs 26 mo ( p = 0,04 ). In the overall population, mPFS was 11,0 in EO-CRC pts vs 14,0 mo ( p < 0,0001 ) in the control group. Furthermore, the overall response rate (ORR) was 63% in EO-CRC and 67% in LO-CRC. Conclusions: Findings from a large population of EO-CRC patients indicate a general worse prognosis for patients with early-onset colorectal cancer compared to late-onset patients. Interestingly this seems to occur regardless of the molecular status. These observations might have a considerable impact on clinical practice and research. Subsequent investigations will be needed to further understand the specific clinical and molecular characteristics of this growing group of patients to better define the more appropriate treatment strategy.

95 Background: L1CAM is a cell adhesion molecule and stem cell marker belonging to the immunoglobulin superfamily of cell adhesion molecules (IgCAMs). L1CAM may be aberrantly expressed in several types of human tumors. The aim of the present study was to evaluate the correlation between L1CAM expression and outcomes in patients with metastatic colorectal cancer. Methods: We retrospectively collected data from 51 mCRC pts treated between 2017 and 2022 at the Medical Oncology Unit of the University Hospital of Cagliari. Tumor samples were retrospectively tested for L1CAM immunohistochemical (IHC) expression with the aim of evaluating the correlation with clinical outcome in terms of overall survival (OS) and progression-free survival (PFS). The primary endpoint was the mOS while secondary endpoint was mPFS. The aim of the present analysis was to evaluate the role of L1CAM expression in tumor cells in predicting the clinical outcome of CRC patients treated with standard chemotherapy. Statistical analysis was performed with the MedCalc Statistical Sofware Version 14.10.2. Results: Median age was 70 (±13), 58.8% were males and 41.2% were females. 29.4% of pts had carcinoma of the right colon, 39.2% left colon and 31.4% rectum. Negative or weak L1CAM score 0 was observed in 39.2% patients; 43.1% pts showed a score 1+; 11.8% showed score 2+, and 5.9% showed score 3+. Positivity for L1CAM correlated with a worse prognosis. The median OS for pts not expressing L1CAM was 74.0 months versus 32.0 for patients expressing the biomarker ( p = 0.0021). The mPFS for patients not expressing L1CAM was 21.0 months, vs 6.0 for patients expressing the biomarker ( p = 0.0066). The same negative correlation on outcomes was shown based on the degree of L1CAM expression. A different L1CAM score corresponded to a different OS and PFS. Pts with score 0 showed a mOS of 74 months vs 37 and 18 months for scores 1+ and 2+/3+, respectively ( p < 0.0001). Similar results were obtained with the mPFS: patients with score 0 showed a median of 21 months vs 8 and 2 months for scores 1+ and 2+/3+, respectively ( p = 0.0138). Conclusions: L1CAM expression in cancer cells correlates with a worse prognosis in mCRC patients. Immunohistochemical evaluation of this biomarker could allow the identification of a subgroup of patients capable of benefiting from a targeted therapy.

Colorectal cancer (CRC) is a leading tumor worldwide. In CRC, the angiogenic pathway plays a crucial role in cancer development and the process of metastasis. Thus, anti-angiogenic drugs represent a milestone for metastatic CRC (mCRC) treatment and lead to significant improvement of clinical outcomes. Nevertheless, not all patients respond to treatment and some develop resistance. Therefore, the identification of predictive factors able to predict response to angiogenesis pathway blockade is required in order to identify the best candidates to receive these agents. Unfortunately, no predictive biomarkers have been prospectively validated to date. Over the years, research has focused on biologic factors such as genetic polymorphisms, circulating biomarkers, circulating tumor cells (CTCs), circulating tumor DNA (ctDNA), and microRNA. Moreover, research efforts have evaluated the potential correlation of molecular biomarkers with imaging techniques used for tumor assessment as well as the application of imaging tools in clinical practice. In addition to functional imaging, radiomics, a relatively newer technique, shows real promise in the setting of correlating molecular medicine to radiological phenotypes.