Proportional Hazards Model Research Articles

Survival disparities in small intestine carcinoid tumors: A SEER-based analysis.

653 Background: Small intestine carcinoid tumors (SI-carcinoids) are rare neuroendocrine neoplasms with variable survival outcomes. This study examines the impact of age, sex, race, income, and geographic location on overall survival (OS) in patients with SI-carcinoid tumors. Methods: A retrospective cohort study was conducted using the Surveillance, Epidemiology, and End Results (SEER) registry 22 to identify patients aged 20 and above with SI-carcinoid tumors from 2000 to 2021. Kaplan-Meier survival analysis and log-rank tests were used to assess differences in survival. Cox proportional hazards models estimated hazard ratios (HR) for Overall Survival (OS). Results: 20,777 patients were included, with 7,029 deaths recorded. Median OS varied significantly by age, sex, race, income, and tumor site. Patients aged 60-85 had a median OS of 10 months (95% CI: 9.75-10.42), while those over 85 had 2.92 months (95% CI: 2.67-3.33). Females had better OS (13.92 months; 95% CI: 13.42-14.42) than males (12.83 months; 95% CI: 12.42-13.25). Non-Hispanic Black patients had worse OS (12.33 months; 95% CI: 11.50-13.33) compared to Non-Hispanic Whites (13.25 months; 95% CI: 12.92-13.67). Patients earning less than $50,000 had worse OS (11.42 months; 95% CI: 10.67-12.42) than those earning $100,000–$120,000 (15.42 months; 95% CI: 14.25-17.83). Geographic disparities were noted, with nonmetropolitan areas showing worse OS (11.42 months; 95% CI: 10.83-12.92) compared to large metropolitan areas (14.00 months; 95% CI: 13.50-14.42). Cox models confirmed age (HR for 85+ years: 21.99; 95% CI: 17.40-27.81), male sex (HR: 1.12; 95% CI: 1.07-1.17), and Black race (HR: 1.29; 95% CI: 1.18-1.42) were linked to worse OS. Tumor site also affected OS, with ileal tumors showing better survival compared to duodenal tumors (HR: 0.89; 95% CI: 0.84-0.94). Conclusions: Age, sex, race, income, and geographic location significantly affect OS in patients with small intestine carcinoid tumors, highlighting the need for tailored interventions for vulnerable groups.

Real-world microscopic residual disease (MRD) monitoring with circulating tumor DNA (ctDNA) in pancreatic adenocarcinoma (PDAC).

696 Background: ctDNA is a powerful tool that can detect MRD with limited but emerging data in patients (pts) with PDAC. We describe the clinical outcomes of pts with PDAC who underwent real-world MRD testing at our institutions. Methods: We retrospectively analyzed pts at 2 institutions with ≥1 tumor-informed ctDNA test (Natera, Inc) ordered for MRD testing post-operation (post-op) between June 2020-August 2024. Descriptive statistics were used to characterize clinicopathologic factors, ctDNA positivity (ctDNA+), and relapse rates (RR). Median relapse-free survival (mRFS) and overall survival (mOS) were calculated using Kaplan-Meier methods, comparisons between groups by log-rank test, and hazard ratios (HR) by Cox proportional hazard models. Associations between clinicopathologic factors and outcomes were assessed by Fisher’s exact and Chi-square tests. Results: Of the 54 pts who underwent MRD testing, 48 pts (88.9%) had successful tests. Failures were due to insufficient tissue. Patients with successful tests were staged as follows: 31.2% stage 1, 50.0% stage 2, and 18.8% stage 3, with 12.5% having R1/2 resections and 46.8% being node-positive. With a median follow-up of 21.0 months (mo), 22 (45.8%) pts had relapsed and 22 (45.8%) were anytime-ctDNA+. Sensitivity and specificity of ctDNA for relapse was 77.3% and 86.4%, respectively. Between anytime- and never-ctDNA+ pts, RR were 77.0% vs. 23.1%, mRFS 14.2 mo vs. not reached (NR) (HR 4.9, 95% CI 1.9-12.9), and mOS 31.7 mo vs. NR (HR 3.3, 95% CI 1.0-10.9), respectively. For 27 pts tested in the MRD window (2-12 weeks post-op), mRFS was 6.6 vs. 25.0 mo (HR 3.1, 95% CI 1.0-9.4) and mOS was 25.5 mo vs. NR (HR 2.5, 95% CI 0.6-10.2) in ctDNA+ vs. negative pts. ctDNA+ preceded radiographic relapse in 54.5% of patients with a median lead time of 166 days (IQR 70-332). In anytime-ctDNA+ pts, there were non-statistically significant trends for improved RFS and OS based on ctDNA decrease and clearance (table). No clinicopathologic factors predicted ctDNA+. Post-op ctDNA+ (p <0.001) and CA 19-9 (p=0.03) were the only factors that had a significant association with RFS. Conclusions: Post-op ctDNA is a strong prognostic marker for survival in PDAC, predicting relapse about five months before imaging. ctDNA+ status is independent of clinicopathologic factors and the strongest predictor of recurrence. These findings highlight a window where novel therapies could be tested to eradicate MRD. Longer follow up and prospective studies are needed to validate its predictive role with therapies in the MRD setting to improve outcomes in high-risk pts. RFS events/total mRFS (mo) HR (95% CI) p OS events/total mOS (mo) HR (95% CI) p ctDNA clearance Yes 6/9 17.0 0.61 (0.2-1.7) 0.35 2/9 NR 0.37 (0.08-1.8) 0.20 No 11/13 10.3 8/13 31.7 ctDNA MTM/mL reduction ≥50% 8/11 15.5 0.93 (0.3-2.9) 0.90 3/11 NR 0.54 (0.1-2.4) 0.42 <50% 6/8 13.6 4/8 31.7

Development and validation of a prognostic model for advanced gastric and oesophageal carcinoma (AGOC) using individual patient data from two clinical trials.

499 Background: Estimating survival in advanced gastric and oesophageal carcinoma (AGOC) remains a significant challenge. We developed a prognostic model incorporating readily accessible patient and clinical data and specific patient-reported outcomes (PROs), utilizing individual participant data from two randomized trials. Methods: We used data from 2 trials comparing regorafenib to placebo; AGITG INTEGRATE IIa (n=251) for model development and AGITG INTEGRATE (n=152) for validation. Both trials enrolled metastatic or AGOC following failure of 1 or more prior lines of systemic therapy. Candidate variables were chosen from systematic literature review and expert consultation. Significant prognostic factors for inclusion in the multivariable model were identified using univariable Cox proportional hazards models with p-value threshold of 0.1. Multivariable models were developed via Lasso regression. An initial model was created using clinical variables alone, followed by the integration of PROs. C statistics were used to discriminate the model’s efficiency. Results: Univariable analysis identified 9 clinical variables and 4 PRO domains as significant, including body mass index (BMI)(p=0.08), ECOG PS (0.02), extent of cancer (p<0.001), liver involvement (p=0.04), treatment with regorafenib (p=0.005), neutrophil-lymphocyte ratio (NLR) (p<0.001), LDH (p<0.001), albumin (p<0.001), and CA 19-9 (p=0.007). Significant baseline PROs included appetite loss (p<0.001), constipation (p<0.001), fatigue (p<0.001), and pain (p<0.001). Gastrectomy violated the Cox model proportional hazards assumption necessitating that the analysis be stratification by gastrectomy. The primary model (M1) incorporated region (Asia vs non-Asia, p=0.25, was included because it was significant in previous analysis (INTEGRATE), ECOG PS status, cancer extent, treatment with regorafenib, NLR, BMI, LDH, CA 19-9, and albumin. The final model (M2), which included PROs, demonstrated superior discriminative ability with the highest c-statistic values in both the gastrectomy and non-gastrectomy strata. The c-statistic increased from 0.695 for M1 to 0.723 for M2, indicating the added prognostic value of including PROs. Conclusions: The prognostic model that integrated both clinical data and PROs, proved robust for predicting survival in AGOC. The addition of PROs significantly enhanced prognostic accuracy, highlighting their importance in survival models. Further validation and refinement of the model is ongoing.

Impact of insurance status on mortality following surgical treatment of colorectal cancers.

40 Background: Healthcare disparities in colorectal cancer, driven by insurance status and socioeconomic factors, lead to delayed diagnoses and poorer surgical outcomes. We sought to examine the impact of insurance status on presentation, treatment, and in-hospital mortality in CRC patients undergoing colectomy using the Nationwide Inpatient Sample (NIS) database. Methods: We included patients aged 18-65 years diagnosed with colon cancer and undergoing colectomy, as identified by ICD-9-CM codes, who had private insurance, Medicaid, or no insurance from January 1, 2005, through December 31, 2014. The primary variable was insurance status; patient characteristics such as age, sex, race, income, and tumor type were analyzed. We used the Elixhauser comorbidity index to control for confounding and performed a subset analysis on patients without major comorbidities. The primary outcome was in-hospital postoperative death. Associations between this outcome and insurance status were analyzed using the Cox proportional hazard model, both in the full cohort and in a subset of patients without comorbidities, with models stratified by hospitals to account for clustering effects from variations in access to care. Results: The study cohort included 301,304 patients, of whom 238,158 (79.0%) were privately insured, 40,417 (13.4%) on Medicaid, and 22,729 (7.6%) were uninsured. Most patients were White (71.6%), followed by African American (12.6%), Hispanic (8.4%), Asian/Pacific Islander (3.8%), and Native American (0.5%). A total of 55.4% of cases took place in teaching hospitals. In the unadjusted analysis, the mortality rate for privately insured patients was 0.7% (95% CI, 0.6%-0.7%) compared with 2.1% for uninsured patients (95% CI, 1.7%-2.5%) and 1.5% for Medicaid recipients (95% CI, 1.2%-1.8%; p=0.001). After adjusting for patient characteristics and stratifying by hospital in patients with no comorbidity, uninsured patients still had a higher risk of experiencing in-hospital death (HR, 1.60; 95% CI, 1.24-2.07) compared with privately insured patients, while no significant disparity was found in Medicaid recipients (HR, 0.95; 95% CI, 0.75-1.22) (Table). Conclusions: Uninsured patients undergoing colectomy for colon cancer experienced the highest in-hospital mortality, a disparity not fully explained by overall health differences. These findings underscore the critical role of insurance coverage in improving surgical outcomes and highlight the need for policy interventions to reduce mortality disparities. In-hospital mortality after surgery for colorectal cancer. Covariate Full sample Patients With No Comorbidity Patients With No Comorbidity in Urban Teaching Hospitals Hazard Ratio (95% Cl) Insurance Private Reference Reference Reference Medicaid 0.95 (0.75-1.22) 1.38 (0.91-2.11) 1.66 (0.91-3.02) Uninsured 1.60 (1.24-2.07) 2.12 (1.42-3.15) 2.28 (1.25-4.16)

Development and validation of the SPEAR scoring model and predicting overall survival in unresectable hepatocellular carcinoma patients treated with TACE, molecular targeted therapies, and immune checkpoint inhibitors.

533 Background: To develop and validate a predictive model to identify hepatocellular carcinoma (HCC) patients likely to benefit from combined treatment of transarterial chemoembolization (TACE), molecular targeted therapies (MTTs), and immune checkpoint inhibitors (ICIs). Methods: Data from 500 patients with unresectable HCC treated with TACE, MTTs, and ICIs (2019-2023) were analyzed. Patients were divided into a training set (241 from Nanfang Hospital) and a validation set (259 from various hospitals). Key prognostic variables were identified using random forest survival analysis, and the top five were used to construct a Cox proportional hazards model. Model performance was assessed using time-dependent AUC values, calibration curves, clinical decision curves (DCA), and net reclassification improvement (NRI) indices. Risk scores and optimal cut-off values for risk stratification were determined and validated. Results: Five key survival variables were identified: serum alpha-fetoprotein (AFP), red cell distribution width coefficient of variation (RDW-CV), aspartate aminotransferase (AST), platelet-to-lymphocyte ratio (PLR), and extrahepatic metastasis (EHM). The Cox model had a C-index of 0.694 in the training and 0.691 in the validation set. The training set's 1-year, 2-year, 2.5-year, and 3-year AUC values were 0.77, 0.73, 0.73, and 0.72, respectively, and 0.74, 0.74, 0.76, and 0.66 in the validation set. The ARAPE model demonstrated a more significant net benefit than existing models. Median overall survival (mOS) was 33.9 months for low-risk, 20.4 months for intermediate-risk, and 14.2 months for high-risk groups in the training set; in the validation set, mOS was 30.8, 18.7, and 10.5 months, respectively. Conclusions: The SPEAR model (Serum AFP, PLR, EHM, AST, and RDW-CV) model effectively stratifies risk for HCC patients receiving TACE combined with MTTs and ICIs, aiding in personalized treatment decisions.

Investigating disparity in labor epidural analgesia management in black vs. white women: a retrospective case-control study (2018-2022).

Disparities in labor epidural analgesia (LEA) management could reduce maternal satisfaction and increase risk. We compared times from the first administration of breakthrough pain medication (top-up) to LEA replacement to evaluate disparities across race. In this retrospective cohort study (01-01-2018 to 12-31-2022), all patients with LEA and maternal race/ethnicity of non-Hispanic White or Black were eligible. Patients with a scheduled cesarean delivery, previous back instrumentation, or maternal age<18 were excluded. We used a Cox Proportional Hazards model to evaluate our primary outcome. Predefined top-up medications given≥60minutes from initial LEA placement and before replacement were valid. Any replacement≥60minutes from the initial LEA was valid. We also studied secondary outcomes incidence of epidural replacement and postoperative patient satisfaction. There were 11,168 total patients receiving LEA, with 479 (5.5%) replacements in White patients and 127 (5.0%) in Black patients. There were 387 (3.5%) LEAs with a top-up followed by replacement within 24hours. After adjusting for confounders, no association was detected between race and LEA management (Hazard Ratio 0.82; 95% CI 0.63, 1.06; P=0.13). We failed to detect an association between patient race and the incidence of replacement (P=0.23). We found that race (P=0.02) and LEA replacement (P<0.001) were associated with increased odds of lower postpartum satisfaction. We didn't detect disparity in treatment at our institution with standardized LEA management protocols. However, Black women and those with LEA replacements were less satisfied. Future studies should evaluate the generalizability and explore interventions that improve patient satisfaction.

Joint effect of polysocial risk score, lifestyle and genetic susceptibility with the risk of dementia: A prospective cohort study

ObjectivesThe comprehensive impact of polysocial risk score (PsRS)—encompassing multiple social determinants of health (SDoHs) with genetic and lifestyle factors on dementia incidence remains to be elucidated. Study designThis study aimed to clear the associations between PsRS and dementia incidence and evaluated how genetic and lifestyle factors modified these associations in the UK Biobank cohort.Methods: The detailed prospective study involved over 500,000 participants when recruited in 2006–2010. The PsRS was calculated by 12 SDoHs across psychosocial factors, socioeconomic status, and neighborhood and living environment. A healthy lifestyle score was constructed from physical activities, alcohol consumption, smoking status, and diet. A genetic risk score (GRS) was computed via genotype data from UK Biobank. The Cox proportional hazards model was used to estimate hazard ratios (HRs) and 95 % confidence intervals (CIs) for the association between PsRS, lifestyle factors, GRS and dementia. ResultsResults showed the participants with intermediate (HR = 1.32, 95%CI: 1.20–1.45) and high PsRS (HR = 2.10, 95 % CI: 1.91–2.32) were significantly associated with an increased risk of dementia compared with those with a low PsRS. Then, compared with participants with low PsRS and favorable lifestyle/low GRS, high PsRS and unfavorable lifestyle/high GRS had the highest risk of dementia (HR = 3.11,95%CI: 2.63–3.68)/(HR = 3.56, 95%CI: 2.62–4.85). ConclusionsBoth high PsRS and GRS were significantly associated with higher dementia risk. A favorable lifestyle could reduce dementia incidence regardless of high PsRS or GRS. Additionally, focusing on the intervention of SDoHs would be positive in preventing dementia.

Impact of expression of metabolic genes on patient (pts) outcomes in metastatic colorectal cancer (mCRC): Data from CALGB/SWOG 80405 (Alliance).

224 Background: Metabolic pathways are reprogrammed during CRC development, leading to increased glycolysis, glutaminolysis, and fatty acid synthesis. Understanding the impact of metabolic reprogramming on treatment outcomes is paramount. Hence, we investigated whether the tumor gene expression of 4 selected major metabolic genes ( PKM, SLC2A1 , SLC16A1 , and CAV1 ) could affect treatment response in pts enrolled in the CALGB/SWOG 80405 trial. Methods: 433 mCRC pts treated with either bevacizumab (bev, n = 226) or cetuximab (cet, n = 207) in combination with first-line chemotherapy were analyzed. RNA was isolated from FFPE tumor samples and sequenced on the HiSeq 2500 (Illumina). Overall survival (OS) and progression-free survival (PFS) were compared between groups of pts categorized by tertiles of gene expression (high [H], medium [M] and low [L]). Logrank P -values provide a non-parametric unadjusted assessment of differences. Likelihood ratio tests (LRT) were computed from multivariate Cox proportional hazards models, adjusting for age, sex, ECOG, tumor side, number of metastatic sites, KRAS , CMS, and treatment. Results: PKM gene expression was associated with cet treatment outcomes with PKM- low tumors showing significantly longer PFS and OS (median PFS: 14.3 vs 9.8 vs 8.1 months, L vs M vs H, P &lt; 0.0001, LRT P = 0.011; median OS: 45.9 vs 31.2 vs 20.9 months, P &lt; 0.0001, LRT P = 0.12). Low SLC2A1 was also associated with longer OS in cet-treated pts (32.4 vs 35.8 vs 25.2 months, P = 0.02, LRT P &lt; 0.0001). Similar results were observed in bev-treated pts where low SLC2A1 expressing tumors had longer OS (37.4 vs 26.1 vs 29 months, P = 0.037, LRT P = 0.069) and a non-significant trend for longer PFS (13.1 vs 11 vs 9.5 months, P = 0.076). CAV1 low gene expression was associated with longer OS in cet-treated pts (37.4 vs 34 vs 21.5 months, P = 0.0048, LRT P = 0.013), while no significant associations were found in bev-treated (interaction LRT P = 0.0009). No significant results were observed for SLC16A1 . Conclusions: The intratumoral expression of metabolic genes was prognostic and predictive in mCRC pts treated with first-line therapy. GLUT1 (encoded by SLC2A1 ) and PKM2 (encoded by PKM ) promote aerobic glycolysis of cancer cells, known as Warburg effect, supporting rapid cell proliferation and survival. CAV1 has been shown to be involved in mitochondrial bioenergetics and fatty acid metabolism and to play either a tumor suppressor or oncogenic role depending on the cancer type and stage. Lower expression of SLC2A1 and PKM was associated with improved survival and increased benefit from cet treatment in our cohort. Low CAV1 expression was also associated with increase survival suggesting a cancer-promoting role in mCRC. Our results suggest that targeting cancer cell metabolism through novel inhibitors of the aforementioned pathways may be a promising therapeutic strategy.

Challenges in patient selection for surgical options in V600E BRAF–mutated metastatic colorectal cancer.

280 Background: V600E BRAF mutated metastatic colorectal cancer (mCRC), which represents almost 10% of mCRC, is associated with a decreased disease-free interval and increased risk of recurrence compared to other mCRC molecular subtypes; however, data show wide outcome heterogeneity within this subgroup. Surgery is known to be associated with improved outcomes in patients with mCRC, particularly in those with liver-only disease. While many patients with V600E BRAF have very aggressive disease and never have the chance to undergo surgery, it is not clear how to navigate the clinical heterogeneity and choose patients that might benefit the most from loco-regional treatment. Methods: We retrospectively evaluated V600E BRAF mutated mCRC patients from two independent high-volume institutions (MDACC and IOV IRCCS) who underwent metastectomy from Oct 2009 to Oct 2023. Patients with MSI-H disease and those who had R2 surgery were excluded. Clinicopathological features, such as sex, age, histology, radicality and site of surgery, were collected from institutional datasets. Relapse free survival (RFS) was defined as the time from surgery to first relapse or progression. Overall survival (OS) was defined as the time from surgery to the date of death or last follow up. Survival was estimated with the Kaplan Meier method, and the Cox proportional Hazard model was adopted for survival comparison. Results: We collected data from 71 V600E BRAF mutated mCRC patients (N=37 MDACC, N=34 IOV IRCCS); 49% were female, 61% had right-sided primary tumor and 62% had synchronous metastatic disease. Median age at metastatic disease diagnosis was 62 years old (IQR: 61-28). For 52% of patients, surgery was performed on liver-only disease; 35% involved the peritoneum. There were no emergency interventions. Median RFS was 7.9 months (95% CI: 5.4 – 10.5) and median OS was 39 months (95% CI: 28.8 – 51.7). No clinicopathological features, including sex, sidedness, histology, differentiation, presentation (synchronous, oligometastatic), or site of intervention, were associated with RFS. However, radicality (R1) was associated with worse RFS (HR=3.5 [95% CI: 1.8 – 6.5], p=0.0003) and OS (HR=2.7 [95%CI: 1.3 – 5.6], p=0.0122). Conclusions: In this retrospective evaluation of patients with V600E BRAF mutated mCRC who underwent metastectomy, only surgical radicality, observed in the post-operative setting, was associated with recurrence and survival outcomes. Therefore, while surgery should be evaluated whenever radical resection is feasible, the identification of patients with V600E BRAF mutated mCRC who might benefit the most from metastasectomy during pre-operative prognosis evaluation remains challenging.

Cardiovascular MRI-derived Right Atrial Strain for Improved Risk Stratification in Patients with Severe Aortic Stenosis.

Purpose To assess the prognostic implications of cardiac MRI-derived imaging markers in individuals with severe aortic stenosis (AS). Materials and Methods This prospective study (German Clinical Trials Register, DRKS00024479) enrolled individuals with severe AS who underwent cardiac MRI before transcatheter aortic valve replacement (TAVR) from January 2017 to March 2022. Image analyses included myocardial volumes, cardiac MRI feature tracking-derived left atrial (LA) and right atrial (RA) as well as left ventricular (LV) and right ventricular (RV) strain, myocardial T1 mapping, and late gadolinium enhancement analyses. Cardiovascular (CV) mortality was defined as primary clinical end point. Cox proportional hazards models were used to determine the association between cardiac MRI-derived parameters and CV mortality. Results The final analysis included 145 participants (median age, 80 years [IQR, 75-83 years]; 91 male). Participants who experienced CV death during follow-up had significantly enlarged RV end-diastolic volumes (median, 82.9 [IQR, 70.8-96.0] mL/m2 vs 62.8 [54.7-76.0] mL/m2; P < .001) and impaired strain values of all cardiac chambers compared with those who survived (LV global longitudinal strain [GLS], -18.1% [-13.1% to -20.4%] vs -22.5% [-16.1% to -27.3%], P = .02; RV GLS, -22.9% [-18.6% to -25.4%] vs -27.9% [-22.9% to -32.0%], P = .002; LA atrial reservoir strain [Es], 9.5% [7.2%-15.4%] vs 14.3% [9.0%-18.1%], P = .04; RA Es, 12.4% [6.8%-14.4%] vs 16.2% [11.2%-22.1%], P < .001). RA reservoir strain independently helped predict CV mortality after adjustment for other cardiac MRI markers and clinical parameters of heart failure (hazard ratio, 0.82 [95% CI: 0.71, 0.95]; P = .008). Within the subgroup of participants with high extracellular volume values, RA strain further identified participants with AS at high risk for CV mortality (P = .001 on log-rank testing). Conclusion In individuals with AS undergoing TAVR, several cardiac MRI parameters were significantly associated with CV mortality. RA strain was an independent predictor of CV mortality and may provide more optimized patient management. Keywords: Cardiac MRI, Aortic Stenosis, TAVR, Risk Assessment, Strain Analyses German Clinical Trials Register no. DRKS00024479 Supplemental material is available for this article. © RSNA, 2025.

A comprehensive study of clinical and molecular features in patients with metastatic colorectal cancer with and without liver mets.

242 Background: Colorectal cancer (CRC) most frequently metastasizes to the liver, and patients (pts) with liver metastasis (mets) at the time of metastatic diagnosis tend to have poorer outcomes compared to pts with non-liver mets. However, the reason for this discrepancy is not well-defined. We thus aimed to identify any distinct molecular or clinical characteristics that could contribute to the poor outcomes associated with liver mets of CRC. Methods: Pts diagnosed with metastatic CRC from 2014-2022 were identified using the institutional molecular database of CRC. Demographic, clinical, and molecular data were collected by reviewing electronic medical records. Time to next treatment was defined as the time between 1st and 2nd line treatment. We utilized Fisher’s exact tests for categorical variables and both Kaplan-Meier analysis and multivariate Cox proportional hazards models to analyze survival data, with a p-value threshold of &lt;0.05 for statistical significance. Results: We identified a total of 217 pts, including pts with liver mets (N=146) and non-liver mets (N=71) at time of diagnosis. Pts with non-liver mets were more likely to have rectal primary than those pts with liver mets (40% vs. 15%, p=0.0001), indicating a distinct pattern of recurrence. Common driver mutations such as KRAS, BRAF, and TP53 were observed at equal frequencies in the liver and non-liver cohorts. In pts without liver mets, age had a significant effect on OS (p &lt; 0.05, HR 1.1). Pts without liver mets with colon primary had worse OS compared to rectum primary (HR= 0.48, 95% CI: 0.24–0.96, p = 0.038). For pts without liver mets, KRAS was not a predictor of outcomes, while BRAF was associated with poorer OS (p&lt;0.001, HR 7.09, 95%CI: 2.31-21.8). Contrarily, for pts with liver mets, KRAS mutation was significantly associated with worse survival (mOS 40.4 vs 61.7 mos, HR 2.04, 95% CI: 1.26-3.30, p=0.004). BRAF mutations were also associated with worse OS for this cohort (p=0.039 and HR 2.86, 95%CI: 1.06-7.77). Median time to the next treatment was significantly shorter in pts with liver mets (13.6 mos) than in pts with non-liver mets (22.6 mos) (HR 1.82, 95% CI 1.28-2.59; p&lt;0.001). OS was longer for pts with non-liver mets compared to those with liver mets (mOS 52.9 mos vs 46.7 mos) though it did not reach statistical significance, likely due to the small sample size. Conclusions: Liver metastasis of CRC is associated with systemic therapy resistance with a shorter time on treatment, indicating treatment resistance associated with liver mets is not limited to immunotherapy. Given similar molecular characteristics in pts with and without liver mets, treatment resistance may be related to the inherent tumor microenvironment of the liver. In this study, pts with distinct molecular features had poorer outcomes, highlighting the importance of developing targeted therapeutics in this population.

A multicenter, prospective phase II trial of second-line aflibercept plus FOLFIRI in patients with metastatic colorectal cancer refractory to anti-EGFR agents (HGCSG1801): Updated analyses.

147 Background: The HGCSG1801 trial evaluated the treatment outcomes of aflibercept (AFL) plus FOLFIRI for patients(pts) with metastatic colorectal cancer (mCRC) refractory to an oxaliplatin-based regimen combined with an anti-EGFR agent. Here we report results of the updated efficacy, safety, and a biomarker analysis. Methods: This was a prospective open-label phase II trial. AFL (4 mg/kg iv) followed by FOLFIRI (irinotecan 180 mg/m 2 , leucovorin 200 mg/m 2 iv, bolus 5-fluorouracil [5-FU] 400 mg/m 2 , and infusional 5-FU 2400 mg/m 2 /46 h) was given every 2 weeks until progression or unacceptable toxicities. The primary endpoint was 6-month progression-free survival (PFS) rate, and the secondary endpoints included overall survival (OS), PFS, overall response rate (ORR), disease control rate (DCR), and adverse events. Angiogenic factors were analyzed in plasma sample using a Luminex multiplex assay using the Cox proportional hazards model. This study was sponsored by the Non-Profit Organization Hokkaido Gastrointestinal Cancer Study Group and supported by a grant from Sanofi. Results: Forty-three pts were enrolled between November 2019 and October 2022. The data cut-off date was April 30, 2024. The 6-month PFS rate was 59.0% (90% confidence interval [CI], 45.8–72.1%). Median PFS and OS were 7.3 months (95% CI, 5.5–11.0 months) and 18.8 months (95% CI, 12.9–26.6 months), respectively. The ORR was 20.9% (95% CI, 10.0–36.0%) and DCR was 88.4% (95% CI, 74.9–96.1%). No deaths and no new safety signals with a causal relation to the study treatment were observed. A biomarker analysis using pretreatment plasma samples showed a trend toward better PFS in pts with low VEGF-A (hazard ratio [HR] 0.40 [95% CI, 0.18-0.91]), TSP-2 (HR 0.40 [95% CI, 0.18-0.88]) or IL-8 levels (HR 0.43 [95% CI, 0.20-0.93]). There was also a trend toward better OS in pts with low levels of TSP-2 (HR 0.30 [95% CI, 0.11-0.81]) or TIMP-1 (HR 0.20 [95% CI, 0.06-0.69]). Conclusions: These updated data further support the activity and manageable safety profile of AFL plus FOLFIRI for pts with mCRC who failed an oxaliplatin-based regimen combined with an anti-EGFR agent. It was suggested the association between some angiogenic factors in plasma samples and the activity of AFL plus FOLFIRI in mCRC. Clinical trial information: jRCTs011190006 .

Low-dose aspirin to reduce recurrence rate in colorectal cancer patients with PI3K pathway alterations: 3-year results from a randomized placebo-controlled trial.

LBA125 Background: Colorectal cancer (CRC) affects 1.9 million individuals globally each year. Among patients with stage II-III CRC, 20-40% develop metastatic disease. Aspirin lowers the incidence of adenomas and CRC in high-risk patients. In addition, observational studies suggest that post-diagnosis aspirin treatment improves disease-free survival (DFS) in unselected populations. Furthermore, retrospective findings indicate that somatic PIK3CA mutations predict treatment response, but requires validation in randomized trials. Methods: The ALASCCA trial was a randomized, double-blind, multicenter, placebo-controlled trial with two parallel arms, across 33 hospitals in Sweden, Denmark, Finland, and Norway. Patients with stage I-III rectal cancer or stage II-III colon cancer exhibiting somatic alterations in the PI3K signaling pathway were included. Patients were randomized to receive either 160 mg of aspirin daily or placebo, initiated within three months post-surgery and continued for three years. To detect a hazard ratio (HR) of 0.36 for the primary outcome of time to recurrence (TTR) assessed at 3 years, with 80% power and a two-sided alpha of 0.05, 150 patients with PIK3CA mutations in exon 9 and/or 20 (“Group A”) were required per arm. An additional 300 patients with other somatic PI3K pathway driver alterations (PIK3CA outside exon 9/20, PIK3R1, or PTEN; "Group B") were required for secondary analyses. A stratified Cox proportional hazards model was fitted for the primary efficacy analysis. Results: A total of 3508 patients were screened for somatic alterations in the PI3K pathway. Of the 2980 patients with conclusive genomic analyses, 1103 patients (37%) had an alteration in the PI3K pathway: 515 patients (17.3%) in Group A and 588 patients (19.7%) in Group B. In total, 626 patients were randomized. After three years of follow-up, the HRs for TTR comparing aspirin to placebo were 0.49 (95% CI; 0.24-0.98; p=0.044) in Group A and 0.42 (95% CI; 0.21-0.83; p=0.013) in Group B. For DFS, the HRs were 0.61 (95% CI; 0.34-1.08; p=0.091) in Group A, and 0.51 (95% CI; 0.29-0.88; p=0.017) in Group B. Three patients experienced aspirin-related severe adverse events (one GI-bleeding, one hematoma, one allergic reaction). Conclusions: Primary endpoint was met. Adjuvant treatment with 160 mg aspirin daily for three years reduced recurrence rate in CRC patients with somatic alterations in the PI3K signaling pathway. These findings could lead to immediate changes in clinical praxis for about a third of CRC patients. Clinical trial information: NCT02647099 .

Clinical outcome of radiofrequency ablation in elderly hepatocellular carcinoma patients over 80 years old.

603 Background: The SURF trial showed the efficacy and safety of radiofrequency ablation(RFA). However, patients over 80 years old were excluded from the study. Aims: To evaluate clinical outcomes of RFA for largest hepatocellular carcinoma (HCC) diameter ≤3 cm, and ≤3 HCC nodules in patients over 80 years. Methods: Patients who underwent RFA at our institution for initial treatment of largest HCC diameter ≤3 cm, and ≤3 HCC nodules from January 2011 to December 2023. Treatment outcome and prognosis were examined in the elderly group for cases 80 years or older and in the non-elderly group for cases under 80 years. The Cox proportional hazards model was used to analyze the factors associated with treatment outcome and prognosis. Results: Of the 518 eligible patients, 136 patients were 80 years or older. Median overall Survival (OS) was 80 (95%CI 60-96) vs. 123 (95%CI 101-NA) months (p=0.021) in the elderly vs. nonelderly group. The median recurrence-free survival was 16 (95%CI 14-22) vs. 26 (95%CI 19-30) months (p=0.023), better in the non-elderly group, Interestingly, for liver disease-related deaths, median OS was 97 (95% CI 80-NA) vs. NR (95% CI NA-NA) months (p=0.62) in the elderly vs. non-elderly group. In the multivariate analysis, factors associated with OS were ALBI grade 2 or 3 (HR 1.67, 95%CI 1.07-2.60), DCP ≥ 40mAU/ml (HR 2.08, 95%CI 1.42-3.05), persistent HCV infection (HCV non-SVR) (HR 5.45, 95%CI 3.07-9.67), nonviral liver disease (HR 4.18, 95%CI 2.31-7.55). HCC recurrence was significantly associated with male (HR 1.51, 95%CI 1.18-1.94), elderly group (HR 1.46,95%CI 1.10-1.95), PT ≥80% (HR 0.69, 95%CI 0.53-0.91), DCP ≥40mAU/ml (HR 1.37, 95%CI 1.07-1.76), and HCV non-SVR (HR 1.73, 95%CI 1.35-2.21). The factors associated with liver disease-related death were ALBI grade 2 or 3 (HR 2.16, 95% CI 1.25-3.73), DCP ≥ 40mAU/ml (HR 2.34, 95% CI 1.48-3.70) and HCV non-SVR (HR 2.22, 95% CI 1.39-3.56). Conclusions: In RFA for diameter ≤3 cm, and ≤3 HCC nodules, over 80 years was not a significant factor associated with OS or liver disease-related death. The results support that RFA would be a promising treatment option for HCC patients over 80 years.

Schizophrenia, Off-Label Antipsychotics, and Dementia Risk in People With HIV.

Comorbidities such as schizophrenia and medication such as antipsychotics may influence the risk of dementia among people living with HIV (PLWH). The objective of this article is to assess the associations among HIV patients with schizophrenia, off-label antipsychotics, and dementia risk. US Department of Veterans Affairs health care facilities from 2000 to September 2023. Retrospective cohort study of PLWH treated by the US Department of Veterans Affairs with history of schizophrenia, off-label antipsychotic use, and neither schizophrenia nor antipsychotic use. Propensity score-matched non-HIV controls were included for the respective HIV groups. The hazard of dementia is estimated using Cox proportional hazards models. PLWH and schizophrenia, were found to have a 2.49 higher hazard of dementia than HIV patients with no history of schizophrenia or antipsychotic medication use (hazard ratio [HR] = 2.49, 95% confidence interval [CI]: 1.85 to 3.35). PLWH and off-label antipsychotic use were found to have a 1.77-fold higher hazard of dementia than HIV patients with no history of schizophrenia or antipsychotic medication use (HR = 1.77, 95% CI: 1.37 to 2.28). Propensity score-matched analysis reveals that, among patients with schizophrenia, those with HIV had a 1.65-fold higher hazard of dementia (HR = 1.65, 95% CI: 1.12 to 2.44). Among patients with no schizophrenia or antipsychotic medication, those with HIV had a 1.47-fold higher hazard of dementia (HR = 1.47, 95% CI: 1.33 to 1.63). This study demonstrates that among PLWH, history of schizophrenia or off-label antipsychotic medication use are associated with substantial increases in dementia incidence. Furthermore, propensity score-matched control analysis reveals that HIV infection itself is independently and significantly associated with elevated dementia risk.

Pediatric Age-Prioritized Waitlisting Policy Potentially Disadvantages Adolescents on Dialysis Not Listed for Transplant Until Adulthood.

Current kidney transplant (KT) policies offer advantages in waiting time and organ allocation priority to pediatric patients waitlisted before 18 years old. This study evaluates the effects of this policy for patients who are on dialysis before, but not waitlisted until after, age 18. Patients aged 11-25 years and waitlisted between 2001 and 2022 for KT were identified in the OPTN STAR data file for analysis. Cohorts were defined by age and dialysis status at time of listing: Peds if < 18 yo, young adult (YA) if ≥ 18 yo; NYOD-not yet on dialysis or OD-on dialysis at time of listing, with YA groups further segregated by age at dialysis initiation. Cumulative incidence of transplant was calculated for waitlisted patients. Graft survival was assessed using Kaplan-Meier analysis and multivariable Cox proportional hazards modeling. p values < 0.01 were significant. Amongst 35 764 KT registrations, candidates who initiated dialysis as pediatric patients but were not waitlisted until after turning 18 years old (YA + OD < 18) have the highest rate of nontransplantation (33.5%) and longest time on dialysis (median: 2103 days) before deceased donor (DD) KT. YA + OD < 18 patients were sixfold less likely than Peds + OD patients to undergo DDKT at 5 years after listing. YA + OD < 18 recipients had the worst post-KT graft survival of all groups at 5 years with adjusted hazard ratio of 1.477 (95% confidence interval: 1.218-1.792) compared to Peds-NYOD (p < 0.001). Current allocation policies significantly disadvantage candidates who initiate dialysis before, but are not listed until after age 18, and should be re-examined to address these inequities.

Optimal preoperative treatment strategy for conversion surgery in unresectable locally advanced pancreatic cancer: A project study by the Japanese Society of Hepato-Biliary-Pancreatic Surgery.

718 Background: Conversion surgery (CS) for unresectable locally advanced pancreatic cancer (UR-LAPC) has been widely accepted when feasible. However, optimal preoperative treatment strategies remain unclear. To determine the most effective preoperative approach for UR-LAPC in the current treatment landscape, we conducted a project study in the Japanese Society of Hepato-Biliary-Pancreatic Surgery. Methods: We analyzed 465 UR-LAPC patients who underwent CS following chemotherapy with FOLFIRINOX (FFX), gemcitabine plus nab-paclitaxel (GnP), or modified regimens between 2015 and 2020 across 84 Japanese board-certified training institutions. Overall survival (OS) was the primary endpoint. We used the Kaplan-Meier method for estimating survival analyses, the Cox proportional hazards model for multivariate analyses, and Maximally Selected Rank Statistics to determine the optimal preoperative treatment duration. Results: Median OS was 43.8 months, with a 5-year survival rate of 37.2%. The optimal preoperative treatment duration was 6.1 months. Patients receiving &gt;6 months of preoperative treatment (n=350) showed significantly better median OS and recurrence-free survival (RFS) compared to ≤6 months (n=115) (50.4 vs 29.7 months and 15.6 vs 9.1 months, respectively; both P&lt;0.001 ). FFX-based regimens (n=116) showed better OS and RFS than GnP-based regimens (n=349) (65.0 vs 36.5 months and 19.3 vs 12.4 months, respectively; both P&lt;0.01 ). Multivariate analysis identified preoperative treatment &gt;6 months (HR 0.53, 95% CI 0.407-0.697, P&lt;0.001 ) and initial FFX-based regimens (HR 0.63, 95% CI 0.464-0.843, P=0.002 ) as independent prognostic factors. Additional favorable factors included normal CA19-9 and CEA levels, PNI&gt;45 before CS, adjuvant chemotherapy, R0 resection, and Evans grade III/IV. Conclusions: This large-scale retrospective study has demonstrated that preoperative treatment duration &gt;6 months and the use of FFX-based regimens as initial treatment were associated with improved survival outcomes in UR-LAPC patients undergoing CS. These findings provide critical information for considering CS in UR-LAPC patients, especially in current clinical practice.

Updated analysis of a phase 2 study of fruquintinib plus trifluridine/tipiracil (TAS-102) as third-line treatment in patients with metastatic colorectal adenocarcinoma.

145 Background: Fruquintinib and TAS-102 are two standard therapies for patients (pts) with previous-treated mCRC worldwide. The study aimed to evaluate the efficacy and safety of fruquintinib plus TAS-102 as third-line treatment for mCRC. We previously reported that fruquintinib plus TAS-102 was well tolerated with encouraging clinical activity in pretreated mCRC. Here, we reported the updated results at the data cutoff of Sep 3, 2024. Methods: In this open-label, single-arm, multi-center, phase 2 trial (NCT05004831), pts with mCRC who had failed at least two prior standard treatment regimens were enrolled. Eligible pts received fruquintinib (4mg, qd, d1-21) and TAS-102 (35mg/m 2 , bid, d1-5, 8-12) orally every 4 weeks until disease progression or unacceptable toxicity. Primary endpoint was PFS (per RECIST v1.1). Secondary endpoints were OS, ORR, DCR and safety (per NCI-CTCAE v5.0). Results: From Mar 2022 to Aug 2023, 50 eligible pts were enrolled. Median age was 60 years (range 39-76) and 58.0% were male. 82% pts were of left-sided colon and rectal cancer, 42.0% pts were RAS mutant, 58.0% had liver metastasis and 18.0% had peritoneal metastasis. Prior treatments and proportion were standard chemotherapy 100%, anti-VEGF 88.0% and anti-EGFR 26.0%. As of Sep 3, 2024, ORR was 10.9% (95%Cl: 3.6-23.6) and DCR was 73.9% (95%Cl: 58.9-85.7). The median PFS was 6.33 months (m) (95% CI: 4.20-8.62). The 6-m, 9-m and 12-m PFS rates were 53.0% (95% CI: 40.2-70.0), 28.3% (95% CI: 17.4-45.9) and 23.1% (95% CI: 13.2-40.5), respectively. At a median follow-up of 17.6m, median OS was 18.4m (95% CI: 12.0-NA). The 6-m, 9-m and 12-m OS rates were 87.0% (95% CI: 77.8-97.3), 66.9% (95% CI: 54.0-82.9) and 64.3% (95% CI: 51.1-80.8), respectively. Median PFS was comparable in liver metastasis (LM) and non-LM pts (6.33m [95%CI: 4.13-8.62] vs. 6.46m [95%CI: 3.74-NA], P =0.54). Similar results were observed in peritoneal metastasis (PM) and non-PM pts (6.07m [95%CI: 3.74-NA] vs. 6.33m [95%CI: 4.20-8.27], P =0.95). None of RAS gene mutation, prior treatment lines ≥ 3, age ≥ 65 years, metastatic sites ≥ 2, peritoneal metastasis and right-sidedness were identified as factors significantly associated with PFS or OS (by univariable Cox proportional-hazards model). The most common treatment-related adverse events (TRAEs) of any grade and grade ≥ 3 were mainly hematological toxicities. TRAEs (any grade, grade ≥ 3) in ≥ 50% of pts including neutrophil count decreased (80.0%, 54.0%), white blood cell count decreased (70.0%, 26.0%), anemia (58.0%, 20.0%) and proteinuria (50.0%, 4%). No treatment-related deaths were observed. Conclusions: The updated analysis demonstrated encouraging survival benefits of fruquintinib plus TAS-102 as third-line treatment in patients with mCRC with acceptable toxicities. This regimen could be an alternative therapeutic approach for these patients. Clinical trial information: NCT05004831 .

Risk of proteinuria with atezolizumab plus bevacizumab versus lenvatinib in first-line systemic treatment for hepatocellular carcinoma.

558 Background: Proteinuria presents a challenging complication during systemic therapy for hepatocellular carcinoma (HCC). This study aims to identify risk factors for proteinuria in patients with HCC treated with atezolizumab plus bevacizumab (Atezo/Bev) or lenvatinib (LEN) as first-line systemic treatment. Methods: A retrospective analysis was conducted on 622 consecutive patients with unresectable HCC who received Atezo/Bev or LEN as first-line systemic treatment between October 2013 and October 2022. Cumulative incidence of proteinuria was estimated using Kaplan–Meier curves and compared using log-rank tests. Risk factors for proteinuria were identified using Cox proportional-hazard models, along with propensity score-matched and subgroup analyses. Results: Among 367 patients treated with Atezo/Bev and 255 with LEN, the cumulative incidence of proteinuria at 12 months was 27.5%. In the multivariable analysis, Atezo/Bev treatment (HR, 1.57; 95% CI, 1.03–2.42), diabetes (HR, 1.64; 95% CI, 1.03–2.61), Child–Pugh class B (HR, 3.43; 95% CI, 1.34–8.78), macrovascular invasion (MVI; HR, 1.58; 95% CI, 1.04–2.38), and an estimated glomerular filtration rate ≤60 mL/min/1.73 m 2 (HR, 3.21; 95% CI, 1.84–5.62) were identified as risk factors for proteinuria. A higher risk of proteinuria in Atezo/Bev patients compared with LEN was consistently observed in the PS-matched cohort, particularly pronounced in subgroups with MVI (HR, 2.84; 95% CI, 1.23–6.54) compared with those without MVI (HR, 1.31; 95% CI, 0.69–2.47). Conclusions: Patients treated with Atezo/Bev as first-line systemic treatmentfor HCC exhibited a higher risk of proteinuria compared with those with LEN, particularly when accompanied by MVI.

Effects of the delay of surgery in localized rectal cancer.

63 Background: The treatment landscape for localized rectal cancer has evolved significantly in recent years, with long-course neoadjuvant chemoradiation (CRT) and total neoadjuvant therapy (TNT) emerging as the primary options. However, the optimal timing of surgery remains an important clinical question, particularly in resource-limited settings where surgical delays can occur. Data from previous studies have shown conflicting results. This retrospective study aimed to assess the impact of delaying surgery beyond the longer interval used in previous studies, focusing on its effects on Pathologic Complete Response (pCR), DFS, and OS. Methods: We conducted a retrospective cohort study using electronic medical records from three centers in Brazil between the years 2013-2024. The study included 236 patients with localized rectal cancer treated with neoadjuvant CRT and/or chemotherapy. Patients were divided into two pre-specified groups based on the interval between the end of neoadjuvant therapy and surgery: ≤11 weeks and &gt;11 weeks. The primary endpoint was DFS. Secondary endpoints were pCR rates, and OS. Descriptive statistics were used for population analysis, while the Kaplan-Meier method estimated DFS and OS, with comparisons made using the log-rank test. A Cox proportional-hazards model was used to estimate hazard ratios and 95% confidence intervals. Results: The median age of the study population was 62 years, with males comprising 56.4% of the sample. The majority of the sample consisted of stage III patients, making up 72.2%, followed by stage II patients at 25.8%, and stage I patients at 1.9%. Most patients (77.8%) had surgery more than 11 weeks after completing neoadjuvant therapy, with a median time to surgery of 16 weeks. The pCR rates did not significantly differ between the two groups (29% for ≤11 weeks vs. 21.5% for &gt;11 weeks; p = 0.271). After a median follow-up of 30 months, the median DFS was not reached, and there was no significant difference between the groups (HR: 0.86; 95% CI: 0.48-1.5; p = 0.63). Similarly, the median OS was not reached, and no difference was observed between the groups (HR: 0.85; 95% CI: 0.47-1.54; p = 0.88). Conclusions: Delaying surgery beyond 11 weeks did not result in greater tumor downstaging in patients with localized rectal cancer. Furthermore, DFS and OS were not significantly impacted by extended intervals between neoadjuvant therapy and surgery.