Browse Articles
Discover research articles across all indexed journals
Real-world efficacy of immune checkpoint inhibitor rechallenge in patients with advanced esophageal squamous cell carcinoma: A retrospective study.
e16046 Background: Immunotherapy combined with chemotherapy is the standard first-line (1L) treatment for advanced esophageal squamous cell carcinoma (ESCC). However, data regarding the efficacy of sequential immune checkpoint inhibitor (ICI) therapy following disease progression are limited. This study evaluates the real-world clinical efficacy of immune rechallenge in advanced ESCC patients to identify optimal strategies for overcoming resistance. Methods: This single-center, retrospective study analyzed patients with advanced ESCC treated at the First Affiliated Hospital of Zhengzhou University between Jan 1, 2020, and Dec 31, 2024. Eligible patients received ≥2 cycles of ICI-based therapy in both 1L and 2L settings. Primary endpoints included progression-free survival (PFS), objective response rate (ORR), and disease control rate (DCR). Subgroup analyses were performed based on different 2L treatment regimens. Results: A total of 135 patients were enrolled (median age: 64 years; 71% male). The median PFS for second-line treatment (mPFS2) was 7.23 months (95% CI: 5.75–9.30). The ORR and DCR were 14.81% and 69.63%, respectively. Subgroup analysis by 2L regimen revealed that the triple combination (ICI + anti-angiogenic targeted therapy + chemotherapy, n = 39) achieved the longest mPFS2 of 9.03 months (95% CI: 5.95–25.00), followed by ICI + anti-angiogenic therapy (n = 39) with 6.77 months (95% CI: 4.96–11.66), ICI + chemotherapy (n = 48) with 5.75 months (95% CI: 4.07–9.66), and ICI monotherapy (n = 9) with 4.67 months (95% CI: 2.73–NR). Multivariate COX regression identified the triple combination as an independent prognostic factor for improved PFS2. Regarding target consistency, patients maintaining the same immune checkpoint target (n = 123, 91.11%) achieved an mPFS2 of 7.23 months (95% CI: 5.75–9.66), compared to 6.37 months (95% CI: 2.60–NR) in those who switched targets (n = 12). Notably, efficacy was independent of the duration of response to prior 1L treatment (PFS1 < 6 months vs. ≥6 months). Conclusions: Immune rechallenge represents a viable therapeutic strategy for advanced ESCC. The triple combination of immunotherapy, anti-angiogenic targeted therapy, and chemotherapy yielded the optimal progression-free survival benefit. Furthermore, maintaining the original immune checkpoint target demonstrated a trend toward superior efficacy compared to switching targets, and efficacy was observed regardless of the duration of benefit from prior first-line treatment. Clinical trial information: ChiCTR2500113885. Subgroup analysis of efficacy. n mPFS2（m） ORR（%） DCR（%） ITT 135 7.23 14.8 69.6 2L treatment I 9 4.67 0 66.7 I+C 48 5.75 12.5 56.3 I+A 39 6.77 20.5 87.2 I+C+A 39 9.03 15.4 69.2 Target Consistency Same Target (Yes) 123 7.23 15.5 68.3 Switch Target (No) 12 6.37 8.3 83.3
Who really benefits from chemotherapy in high-risk localized soft tissue sarcoma? A virtual twin analysis of ISG-STS 1001.
11573 Background: Randomized trials of perioperative chemotherapy in high-risk localized extremity/truncal soft tissue sarcoma (etSTS) show modest average survival benefits, leaving uncertainty about which patients derive clinically meaningful benefit. Conventional hazard ratio-based reporting reflects population-level effects but offers limited guidance for individual decisions. Counterfactual approaches, such as the virtual twin framework, enable estimation of patient-specific outcomes under alternative treatment strategies, translating randomized evidence into clinically interpretable absolute benefit measures. Methods: Individual patient data from the ISG-STS 1001 randomized trial in high-risk localized etSTS with updated follow-up were analyzed using a virtual-twin framework with counterfactual predictions to estimate each patient’s outcomes under alternative treatment scenarios: standard epirubicin-ifosfamide (EI) vs histotype-tailored (HT) chemotherapy (proxy for no effective chemotherapy). Patients with myxoid liposarcoma were excluded. Cox proportional hazards models for overall survival (OS) and disease-free survival (DFS) were adjusted for pretreatment covariates. For each patient, counterfactual 5-year OS and DFS probabilities under EI and HT were estimated and individualized absolute treatment effects were defined as the difference between counterfactual predictions. Bootstrap resampling was used to quantify uncertainty and treatment benefit explored across levels of baseline risk and tumor characteristics. Results: The cohort included 218 pts (median follow-up 116 months). In covariate-adjusted models, mean absolute improvement in 5-year OS with EI was 7.5% (95%CI: 4.5%-10.2%). Absolute OS benefit was inversely correlated with Sarculator-predicted 10-year survival, indicating greater benefit among patients with poorer baseline prognoses. Pts in the highest Sarculator-predicted survival quartile (pOS >69%) had a mean 5-year OS benefit of 5.9% (95%CI: 3.5%-9.0%), with 63% achieving ≥5% benefit. In contrast, pts in the lowest predicted survival quartile (pOS <46%) had a mean benefit of 8.7% (95%CI: 5.3%-10.4%), with 98% achieving ≥5% benefit. Benefits varied by histology (undifferentiated pleomorphic sarcoma = 6.6%, leiomyosarcoma = 7.2%, synovial sarcoma = 8.3%, malignant peripheral nerve sheath tumor = 8.9%) and, within each subtype, was inversely correlated with Sarculator-predicted survival. Conclusions: Individualized counterfactual prediction reveals substantial heterogeneity in the absolute benefit of chemotherapy in high-risk localized etSTS, driven by baseline prognostic risk and histology. This framework provides clinically relevant, patient-centered estimates that support risk-adaptive decision-making and clarify chemotherapy benefits across risk groups beyond average trial effects.
Impact of lung cancer screening in patients diagnosed with small cell lung cancer.
8102 Background: The advent of immunotherapy significantly improved overall survival (OS) for limited-stage SCLC (LS-SCLC), highlighting how early detection may drive maximal therapeutic benefit. Lung cancer screening (LCS) with low-dose CT scans (LDCT) in high-risk individuals reduced the risk of lung cancer mortality and identified more Stage I lung cancer (National Lung Screening Trial (NLST), NELSON). Within the NLST, SCLC distribution by stage (I-IV) was similar between LDCT or chest x-ray, thereby limiting the perceived absolute impact of LCS for SCLC detection. Here, we report on the utility of LCS in SCLC in the era of immunotherapy. Methods: We retrospectively reviewed the charts of patients diagnosed with SCLC in the Indiana University (IU) Health System from 2018 to 2024. Patients were stratified by whether SCLC was identified from LDCT or non-screen detected. Kaplan-Meier, Wilcoxon rank sum, Pearson’s Chi-squared, and Fisher’s exact test were applied. Results: Among the 301 LCS-eligible individuals with SCLC, 67 patients (22.3%) received LDCT preceding their diagnosis. With LDCT, a significant shift in stage was observed ( P <0.001). In the non-LDCT cohort, 28% were diagnosed with LS-SCLC and 72% with ES-SCLC. In the LDCT cohort, 60% were diagnosed with LS-SCLC and 40% with ES-SCLC. After robust adjustment for confounders in a multivariable logistic regression model, LCS was independently associated with higher odds of LS-SCLC at diagnosis (odds ratio 4.22; 95% CI, 2.20-8.28, P <0.001). Significant delays post-LDCT were noted including time from scan to biopsy (52 vs 10 days, P <0.001) and biopsy to treatment (20 vs 13 days, P <0.001). LCS did not improve response rates, treatment completion rates, or progression-free survival, regardless of stage. At diagnosis, patients with ES-SCLC in the LDCT cohort tended to have less CNS involvement (11% vs 30%, P = 0.06) but had more bone metastases (74% vs 45%, P = 0.006), compared to non-LDCT. Survival between LDCT and non-LDCT was similar for LS-SCLC (22.9 vs 40.4 months, P = 0.15) and ES-SCLC (12.2 vs 10.8 months, P = 0.89). To account for stage migration and unequal follow-up between groups (41.4 vs 25.3 months), a pre-specified, stage-agnostic 1-year OS was compared. The LDCT group had a higher 1-year OS compared to the non-LDCT group (69.2% vs 54.6%, P = 0.03). Conclusions: Use of LCS in eligible patients resulted in a significant “stage-shift” to LS-SCLC for individuals diagnosed with SCLC. At a pre-specified cut-off, LCS significantly improved OS for patients diagnosed with SCLC. Characteristic LDCT (No) (n = 234) LDCT (Yes) (n = 67) Overall (n = 301) P value Age (Q1, Q3) 66 (61, 72) 68 (64, 73) 67 (62, 72) 0.1 Sex (female) 145 (62%) 41 (61%) 186 (62%) >0.9 ECOG PS (0-1) 150 (70%) 53 (84%) 203 (73%) 0.024 Current tobacco 171 (73%) 50 (75%) 221 (73%) 0.9 COPD 119 (51%) 48 (72%) 167 (55%) 0.003 +FH cancer 128 (59%) 46 (73%) 174 (62%) 0.047 +FH lung cancer 57 (26%) 22 (35%) 79 (28%) 0.2
Impact of pre-existing cancer on short-term outcomes and mortality following left ventricular assist device implantation at urban teaching centers in the United States.
e23056 Background: Cancer patients are at higher odds of various procedural complications. A recent study suggested a rise in the use of left ventricular assist device (LVAD) implantation in recent years, yet the demographics and mortality remain poorly described. Methods: We conducted a retrospective study of adults undergoing LVAD implantation using the National Inpatient Sample, excluding COVID-19 admissions. Patients were stratified by pre-existing cancer. Demographics, comorbidities, and hospitalization characteristics were compared. The primary outcome was in-hospital mortality. Results: Among 25,935 left ventricular assist device recipients, 745 (2.87%) had cancer. Cancer patients were older (59.3±1.0 vs 56.9±0.2 years; p=0.017) and more often female (28.9% vs 22.6%; p=0.083). Most were White (66.9% vs 58.8%), followed by Black (23.0% vs 28.2%) and Hispanic (3.6% vs 6.9%). Elective admissions were similar (28.2% vs 26.8%). Cardiovascular and systemic comorbidities were comparable, including acute myocardial infarction (6.0% vs 4.3%), prior myocardial infarction (13.4% vs 14.7%), dyslipidemia (33.6% vs 37.0%), diabetes mellitus (36.2% vs 38.1%), hypertension (73.8% vs 77.8%), chronic kidney disease (49.7% vs 47.1%), obesity (17.5% vs 21.3%), chronic obstructive pulmonary disease (16.1% vs 15.1%), obstructive sleep apnea (15.4% vs 19.1%), peripheral vascular disease (36.2% vs 37.4%), coagulopathy (40.9% vs 45.3%), history of coronary artery bypass grafting (6.0% vs 7.3%), history of percutaneous coronary intervention (10.7% vs 9.6%), prior stroke (5.4% both), smoking (21.5% vs 26.8%), alcohol abuse (2.0% vs 3.6%), and drug abuse (5.4% vs 4.4%) (p>0.05). Primary expected payer differed (p=0.007), with cancer patients more often covered by private insurance (45.6% vs 33.1%) and less often by Medicaid (6.0% vs 14.5%) or Medicare (46.3% vs 48.3%). Unadjusted in-hospital mortality was higher in cancer patients (14.1% vs 10.6%), but doubly robust adjustment showed no significant association (adjusted odds ratio 1.54, 95% confidence interval 0.83–2.84; p=0.172). Length of hospital stay and inflation-adjusted mean hospital charges were similar. Conclusions: Cancer patients undergoing LVAD implantation were older but presented with a comorbidity profile comparable to non-cancer patients. While short-term in-hospital mortality did not show a statistically significant difference after adjustment, these findings may be influenced by survivor bias and highly selective candidacy criteria for advanced heart failure therapies. Further prospective studies are needed to evaluate long-term post-discharge outcomes and to determine if these results remain consistent across various cancer stages and types.
Integrated genomic and transcriptomic profiling of driver–pathway relationships in colorectal cancer.
e15723 Background: Relationships between DNA driver alterations and downstream transcriptional pathway activation in colorectal cancer (CRC) are not well defined in routine specimens. We assessed driver–pathway concordance using paired DNA and transcriptome profiling. Methods: Tumors from 192 CRC patients (male = 85; female = 107) were profiled by a 517-gene DNA panel and AmpliSeq tissue transcriptome. Pathogenic/likely pathogenic SNV/indels defined driver status. A priori driver–pathway associations based on established CRC biology ( APC →WNT/β-catenin; RAS/RAF→MAPK; PIK3CA →PI3K/AKT/mTOR; SMAD4 →TGF-β/EMT) were evaluated. Pathway activity scores were computed as the average log 2 fold-change of curated pathway marker genes (scores calculated when ≥4 marker genes were available). Mutant vs wild-type groups were compared using Mann–Whitney; Δmedian with bootstrap 95% CI; BH-FDR across a priori tests. Multivariable linear regression adjusted for age, sex, and AJCC stage. Results: Median age was 55 years (IQR 44–63). AJCC stage was IV in 122/192 (63.5%) and unknown in 58/192 (30.2%); adenocarcinoma comprised 166/192 (86.5%). Frequent drivers were TP53 152/192 (79.2%), APC 109/192 (56.8%), KRAS 102/192 (53.1%), BRAF 20/192 (10.4%), PIK3CA 28/192 (14.6%), and SMAD4 30/192 (15.6%); RAS/RAF alterations occurred in 128/192 (66.7%). APC -mutant tumors showed higher WNT/β-catenin activation (evaluable n = 139; Δmedian +0.66, 95% CI +0.15 to +1.40; p = 0.0059; FDR = 0.029), remaining significant after adjustment (β = +0.74, 95% CI +0.04 to +1.44; p = 0.039). SMAD4 -mutant tumors showed lower TGF-β signaling (evaluable n = 159; Δmedian −0.87, 95% CI −1.94 to −0.30; p = 0.0116; FDR = 0.029), also significant after adjustment (β = −1.10, 95% CI −2.12 to −0.08; p = 0.036). RAS/RAF→MAPK and PIK3CA →PI3K/AKT/mTOR were not significant (FDR > 0.05). Conclusions: In routine CRC specimens, paired genomic and transcriptomic profiling demonstrated biologically coherent driver–pathway concordance for APC /WNT activation and SMAD4 /TGF-β signaling. Integrated multi-omics provides pathway-contextual interpretation alongside DNA comprehensive genomic profiling and may support molecular stratification in translational CRC studies.
Trends in mortality, incidence, and disability-adjusted life years of laryngeal cancer in Pakistan from 1990 to 2023: A comprehensive analysis of the Global Burden of Disease study.
e18010 Background: Laryngeal cancer remains a major cause of cancer-related mortality worldwide, particularly in low- and middle-income countries such as Pakistan. Tobacco consumption, air pollution, and occupational exposures are key risk factors contributing to this burden. Despite its clinical significance, national and regional trend data for LC in Pakistan remain uncharacterized. Methods: We extracted data from the Global Burden of Disease GBD 2023 study to estimate age-standardized mortality (ASMRs), incidence (ASIR), and disability-adjusted life years (ASDALYs) rates per 100,000 population from 1990 to 2023. Analyses were performed using linear regression with Joinpoint software to estimate the annual percent change (APC) and average annual percent change (AAPC) with 95% confidence intervals (CIs) to quantify temporal trends. Results: From 1990 to 2023, the age-standardized mortality rate (ASMR) of laryngeal cancer in Pakistan increased from 3.72 to 3.95 per 100,000 population (AAPC: +0.12%). Similarly, the age-standardized incidence rate (ASIR) rose from 4.21 to 4.85 (AAPC: +0.37%), while the age-standardized disability-adjusted life years (ASDALYs) showed a fluctuating upward trend from 101.1 to 112.9 per 100,000 (AAPC: +0.28%). Regionally, Islamabad consistently recorded the highest mortality and incidence rates throughout the study period, whereas Gilgit-Baltistan exhibited the most pronounced increase in mortality (AAPC: +1.03%). Both males (APC: +5.08%) and females (APC: +4.07%) experienced a marked surge in mortality between 2018 and 2021, accompanied by parallel increases in incidence and DALY rates. Gender-specific analysis revealed that males maintained consistently higher ASMR and ASDALY values than females across all years. The 20–54-year age group showed the greatest rise in both ASIR (AAPC: +0.90%) and ASDALY (AAPC: +0.69%) during the study period. Conclusions: Over the past three decades, Pakistan's mortality, incidence, and disability-adjusted life years have all gradually but steadily increased due to laryngeal cancer. The increasing patterns, especially among men and those between the ages of 20-54, point to a developing public health issue. The necessity for region-specific cancer control methods is highlighted by regional differences, with Gilgit-Baltistan and Islamabad exhibiting the highest and fastest rising rates, respectively. Strengthening early detection, reducing tobacco use, and improving access to diagnostic and treatment facilities are crucial to curbing the escalating burden of laryngeal cancer in Pakistan.
Racial disparities and nonconcordance with endocrine and HER2 therapy for breast cancer: A sequential mediation analysis of neighborhood deprivation and insurance status.
1647 Background: Racial disparities in breast cancer treatment persist, but the mechanisms driving inequities may differ by therapy type. We evaluated whether neighborhood racialized economic segregation (mediator 1, M1) and insurance status (M2) sequentially mediate racial disparities in receipt of guideline-concordant endocrine and anti-HER2 therapy. Methods: We conducted a retrospective cohort analysis of women diagnosed with breast cancer using the Florida Cancer Data System (2013–2021). Eligible women included those with invasive, nonmetastatic disease with hormone or HER2-receptor-positive tumors. Treatment nonconcordance was defined as failure to receive recommended systemic therapy based on tumor receptor status. M1 was measured by the racial-income version of the Index of Concentration at the Extremes (ICE) at the census tract level and divided into quintiles from the most to least deprived. Insurance status was categorized as insured versus Medicaid/uninsured/unknown (MUU). Simple frequencies were obtained to assess the relationships between race, the mediators, and therapy nonconcordance. Causal mediation analyses using the parametric g-formula quantified natural course risks and the proportion of the racial disparity mediated by ICE and insurance status, while adjusting for confounders (disease stage, molecular subtype [endocrine only], tumor grade, age, region of the state, and marital status). Results: There were 71,077 and 3,021 women available for the analysis of endocrine and HER2 therapy, respectively. For nonconcordance with endocrine therapy, disparities were present by race (37.2% vs. 29.0%, Black vs. White), the ICE (37.2% vs. 26.2%, most vs. least deprived), and insurance status (36.1% vs. 29.4%, MUU vs. other). Smaller disparities were present for nonconcordance with HER2 therapy by race (18.1% vs. 12.9%, Black vs. White), the ICE (16.2% vs. 14.2%, most vs. least deprived), and insurance status (18.5% vs. 13.4%, MUU vs. other). In causal mediation analysis for nonconcordance with endocrine therapy, the natural course counterfactual risks showed a 5% absolute disparity by race (34.3% vs. 29.3%); 59% of this disparity was mediated by the ICE and insurance status. For nonconcordance with HER2 therapy, the natural course counterfactual risks showed a 4% absolute disparity by race (17.1% vs. 13.1%); none of this disparity was mediated by the ICE and insurance status. Conclusions: Racial disparities in receipt of endocrine therapy are largely mediated by neighborhood deprivation and access to care, whereas these factors played no role in explaining disparities in anti-HER2 therapy. These findings suggest that treatment disparities are therapy-specific and that analyses of aggregate treatment outcomes may obscure distinct mechanisms of inequity.
Clinical performance and readiness of artificial intelligence models for predicting immune-related adverse events in patients receiving immune checkpoint inhibitors: A systematic review.
e13683 Background: Immune checkpoint inhibitors (ICIs) have transformed cancer treatment but are frequently complicated by immune-related adverse events (irAEs), which may result in treatment interruption, hospitalization and mortality. Reliable pre-treatment tools to identify patients at higher risk for irAEs, particularly checkpoint inhibitor pneumonitis (CIP) remain limited. Artificial intelligence (AI)-based prediction models have emerged but their clinical performance and readiness for implementation are unclear. Methods: We systematically searched PubMed, Embase, Scopus and Web of Science (inception-January 2025) for studies developing AI-based models to predict irAEs or CIP in patients receiving ICIs. Eligible studies included radiomics, deep learning, natural language processing and clinical machine-learning approaches. Two reviewers independently extracted data on cancer type, toxicity endpoint, modality, algorithm, validation strategy and model performance. Study design and reporting characteristics were assessed. Due to substantial methodological heterogeneity, results were synthesized descriptively. Results: Fifteen studies met inclusion criteria, including mixed solid tumor cohorts and lung-predominant populations. Predictive targets included any-grade irAEs and organ-specific toxicities, most commonly pneumonitis. Model approaches ranged from clinical and electronic health record–based models to CT- and PET/CT-based radiomics and multimodal imaging-clinical frameworks. Reported discriminatory performance varied widely, with AUC values ranging from 0.64 to > 0.90 across model types. The most consistent performance was observed in deep learning and multimodal radiomics-clinical models for pneumonitis, with AUCs frequently between 0.75 and 0.92, whereas several clinical-only models demonstrated more modest performance. Most studies were retrospective and single-center, used heterogeneous irAE definitions and lacked standardized imaging protocols. External validation and calibration assessment were uncommon. Conclusions: AI-based models for predicting irAEs in ICI-treated patients demonstrate promising performance, particularly for pneumonitis using radiomics and multimodal approaches. However, current evidence is limited by heterogeneity, scarce external validation and uncertain generalizability. Prospective, multi-institutional validation and standardized outcome definitions are needed before AI-driven irAE risk prediction can be integrated into routine oncology practice to guide surveillance and patient counseling.
Closing the gap in germline testing in patients with cancer through a clinic-based awareness initiative: From booth to patient's agenda.
1603 Background: Germline genetic testing (GT) is a standard of care for a broad range of patients (pts) diagnosed with cancer due to expanded guidelines and targeted treatments options. The consequent identification of hereditary cancer susceptibility syndromes (HCS) enables optimal surveillance measures, prophylactic interventions and the identification of at-risk family members. Unfortunately, GT rates in oncology are still low worldwide. Barriers are complex but commonly related to deferred appointments and overwhelmed pts. In Argentina, GT and genetic counseling (GC) have not been endorsed as a public health need, insurance coverage relies on a case-by-case basis and mainstreaming has not been fully implemented, therefore, integrating GT into oncology workflows remains a challenge. Methods: Our aim was to describe the impact on GT intake of an information campaign coupled with an appointment help desk in a cancer center. We conducted a high-risk, one-month awareness social media campaign, following a one-week, in-person, information booth (during 06/26/2024 -07/03/2024), strategically placed near the waiting room of a cancer center in Argentina. Both offered informative material and the possibility to attend a free, hybrid- mode, interactive, “meet the expert” session on 07/11/2024. For those who visited the booth, in-person orientation by health providers and scheduling for a personalized GC session (in person or by telemedicine) in the following 3 months was offered. After informed consent, information from social media, landing page interaction and medical records was extracted for analysis. Descriptive statistics with Stata program was performed. Results: On social media: ≥600 likes, ≥20 comments, ≥350 landing page visits, ≥278 interactions and 976 new e-mail contacts were registered in the landing page. A peak of 119 visits was recorded on 07/09/2024. Media engagement rate: 5,14 %. 814 individuals registered to the hybrid session with an attendance rate of 30% (245/814), with no differences related to format option (in-person 21/73: 29% and virtual 224/741: 30%, p-value 0.794954). At the booth: 165 pts requested in-person information, 160 were scheduled for a visit: first assessments (1As): 114/160 (71%). Compliance with appointment was 64% between 1As (73/114), GT was suggested to 92% of 1As (67/73) with GT intake of 49/73 (67%) by multi-gene germline panel. Overall, 10/160 HCS were identified (diagnostic yield in intention-to-treat): 6,5%: 3 gATM, 2 gBRCA1-2, 2 gHOXB13, 1 gCHEK2, 1 heterozygous MUTYH, 1 heterozygous FANCA. Risk-reducing surgeries were performed in 4/10 carriers. Conclusions: Social media engagement combined with lowering appointments barriers can achieve a high compliance to 1As genetic consultations, GT and consequently identification of HCS among high-risk pts with cancer in resource-limited scenarios.
Major adverse cardiovascular events during elective chemotherapy admissions: Incidence, predictors, and in-hospital outcomes.
e24020 Background: Cardiovascular complications represent a major source of morbidity among patients receiving systemic chemotherapy, yet contemporary data describing acute cardiovascular events during elective chemotherapy hospitalizations are limited. While prior studies have focused on long-term cardiotoxicity, the incidence, predictors, and short-term clinical impact of major adverse cardiovascular events (MACE) occurring during planned inpatient chemotherapy admissions remain poorly characterized. Methods: We conducted a retrospective cohort study using the National Inpatient Sample (NIS). Adult elective hospitalizations involving chemotherapy administration were identified using ICD-10-PCS codes. MACE was defined as a composite of acute MI, ischemic stroke, acute heart failure, or malignant arrhythmias. Multivariable survey-weighted logistic regression models were used to identify predictors of MACE and to evaluate associations with in-hospital mortality and resource utilization, adjusting for demographics, comorbidities, metastatic disease, and hospital characteristics. Results: Among 7,874,281 elective chemotherapy hospitalizations, 5.8% experienced at least one MACE. Patients with MACE were older (65.4 vs 38.8 years, p < 0.001) and more frequently male (55.1% vs 44.8%, p < 0.001). Independent predictors of MACE included older age (aOR 1.03), African American race versus White (aOR 1.21), obesity (aOR 1.30), obstructive sleep apnea (aOR 1.42), coronary artery disease (aOR 3.48), hyperlipidemia (aOR 1.47), peripheral vascular disease (aOR 1.35), sepsis (aOR 3.26), and anemia (aOR 1.31) (all p < 0.001). Female sex was independently protective (aOR 0.92). Solid tumors were associated with lower odds of MACE compared with hematologic malignancies (aOR 0.17). Metastatic disease and tumor lysis syndrome were not independently associated with MACE. MACE was strongly associated with increased in-hospital mortality (aOR 4.83), higher total hospitalization costs (+$78,844), and longer length of stay (+2.85 days) (all p < 0.001). Conclusions: MACE occurs in nearly 1 in 17 elective chemotherapy hospitalizations and is associated with substantial increases in mortality, length of stay, and healthcare costs. Patients with hematologic malignancies and pre-existing cardiovascular comorbidities are at particularly high risk. These findings highlight the need for improved cardiovascular risk stratification and inpatient monitoring strategies during elective chemotherapy admissions.
Prophylactic vs reactive PEG tube placement in early-stage head and neck (H&N) cancer treated with curative-intent radiotherapy: Balancing benefit and burden in real-world practice.
12150 Background: Timing of Percutaneous Endoscopic Gastrostomy (PEG) placement during curative-intent radiotherapy for H&N cancer remains controversial. We compared 6-month and 1-year outcomes after prophylactic (pPEG) versus reactive (rPEG) PEG placement in real-world practice. Methods: We conducted a retrospective study using the TriNetX Network database. Adults (≥18 years) with H&N cancer treated with curative-intent radiotherapy who received PEG were stratified by PEG timing: pPEG (≤30 days before or at radiotherapy initiation) versus rPEG (30–90 days after radiotherapy initiation). The index date was PEG insertion. After 1:1 propensity-matching with demographics, comorbidities, nutrition/weight variables, BMI, stage, chemotherapy/surgery, 2,752 patients remained in each cohort. Outcomes at 6 months and 1 year were analyzed using Kaplan–Meier methods with log-rank tests and Cox proportional-hazards models (HR, 95% CI). Results: At 6 months, compared with rPEG, pPEG was associated with lower all-cause mortality, but higher hospitalization, sepsis, PEG tube dependence, and PEG tube complications. ICU admission, aspiration pneumonitis, pneumonia, and dehydration were similar between cohorts. At 1 year, all-cause mortality and ICU admission remained similar between groups. However, pPEG was associated with higher hospitalization, aspiration pneumonitis, pneumonia, sepsis, and PEG tube complications, while dehydration and PEG tube dependence were similar (see Table). pPEG was associated with higher risk of dysphagia at 6 months (HR 1.3, 95% CI 1.1–1.6; p<0.01) and 1 year (HR 1.2, 95% CI 1.1–1.5; p=0.01), and malnutrition at 6 months (HR 1.3, 95% CI 1.1–1.6; p<0.01) and 1 year (HR 1.3, 95% CI 1.1–1.6; p<0.01). Unintentional weight loss and PEG tube removal did not differ between groups. Conclusions: In this real-world cohort study, pPEG was not associated with improved 1-year survival and was associated with higher acute care utilization, infections, and PEG-related complications compared with rPEG. These findings support selective, risk-stratified pPEG use rather than routine prophylactic placement. Outcome 6 Month HR (95% CI, p-value) 1 Year HR (95% CI, p-value) All-cause mortality 0.8 (0.7-0.9, p=0.04) 0.9 (0.8-1.1, p=0.6) Hospitalization 1.3 (1.01-1.6, p=0.04) 1.2 (1.1-1.5, p=0.04) ICU admission 1.0 (0.8-1.3, p=0.6) 1.1 (0.9-1.3, p=0.3) Aspiration pneumonitis 1.2 (0.9-1.5, p=0.07) 1.2 (1.1-1.5, p=0.03) Pneumonia 1.2 (0.9-1.4, p=0.1) 1.2 (1.1-1.4, p=0.04) Sepsis 1.3 (1.01-1.5, p=0.03) 1.3 (1.1-1.5, p=0.01) Dehydration 1.6 (0.9-2.6, p=0.08) 1.5 (0.9-2.3, p=0.1) PEG tube dependence 1.2 (1.01-1.3, p=0.04) 1.1 (0.9-1.3, p=0.07) PEG tube complication 1.2 (1.1-1.4, p<0.01) 1.2 (1.1-1.3, p<0.01) HR > 1 indicates higher hazard in pPEG vs rPEG; HR < 1 indicates lower hazard.
Comorbidities as independent predictors of mortality in lung cancer: A real-world multivariate analysis from a Latin American tertiary center with integrated molecular profiling.
e20093 Background: Lung cancer patients frequently present with multiple comorbidities that may significantly influence survival beyond tumor-related factors. However, the independent contribution of non-oncologic conditions to mortality remains underexplored, particularly in real-world Latin American populations. Methods: We conducted a retrospective study of adult patients with non–small cell lung cancer treated at a tertiary center (2016–2023). Clinical, tumor, and comorbidity data and causes of death were collected. All-cause mortality was the primary outcome. Multivariate logistic regression adjusted for age, sex, and tumor stage was performed. Results: A total of 496 patients were included (54.6% women; median age 69 years). At diagnosis, 15% were stage I, 6% stage II, 13.8% stage III, and 65.2% stage IV. PD-L1 expression was assessed using tumor proportion score (TPS) in 254 patients: 41.3% were TPS < 1%, 37.8% had TPS 1–49%, and 20.9% had TPS ≥50%. EGFR were assessed in 298 patients, identifying mutations in exon 19 deletions 17.8% (n = 53), L858R 12.8% (n = 38), exon 20 insertions 3.7% (n = 11), ALK rearrangements were assessed in 275 patients identifying positivity 12.0% (n = 33), reflecting a biologically characterized real-world cohort enriched for actionable alterations. The most frequent comorbidities were smoking history (55.6%), hypertension (46.8%), COPD (16.7%), and diabetes (16.3%); pulmonary embolism (PE) occurred in 10.1%. Overall mortality was 60.7% (n = 296). Cause of death was available in 161 patients, of whom 18.8% were non-oncologic and 81.2% cancer-related. Non-oncologic deaths were mainly due to pulmonary infections (pneumonia/COVID-19), often progressing to septic shock (32.0%), followed by cardiovascular disease (heart failure, 10.0%), thromboembolic events, and chronic kidney disease (7.5%). In bivariate analysis, mortality was higher in men (67.9%) and increased with stage (stage IV: 77.3% vs stages I–II, p < 0.001). COPD (70.4% vs 58.8%, p = 0.02), cerebrovascular disease (78.8% vs 59.4%, p = 0.03), and PE (71.4% vs 59.5%, p = 0.05) were associated with mortality. In multivariate analysis, advanced stage was the strongest predictor of death (stage III OR 4.29, 95% CI 1.73–10.10; stage IV OR 15.20, 95% CI 6.70–33.72). COPD (OR 2.13, 95% CI 1.10–4.11), cerebrovascular disease (OR 3.36, 95% CI 1.11–10.25), and PE (OR 1.55, 95% CI 1.05–3.47) remained independently associated with mortality after adjustment for age, sex, and stage. Conclusions: COPD, cerebrovascular disease, and pulmonary embolism remained independent predictors of all-cause mortality after adjustment for key clinical factors. These results support the implementation of integrated care pathways focused on early comorbidity identification and proactive management to improve survival in lung cancer patients.
Education, income, treatment, and survival: A retrospective cohort study of patients with breast cancer in Sao Paulo, Brazil.
e13620 Background: Socioeconomic factors influence breast cancer stage at diagnosis, access to treatment, and survival, yet their impact in middle-income settings remains underexplored. This study examines the effect of education attainment and GDP on treatment utilization and survival outcomes among patients diagnosed with breast cancer in São Paulo, Brazil. Methods: We conducted a retrospective cohort study including incident cases of invasive breast cancer recorded in the São Paulo FOSP registry between 2000–2019. The primary outcomes were overall survival and differences between expected and observed treatment utilization. The primary explanatory variables of interest were educational attainment and GDP. Overall and cancer-specific survival were estimated using Kaplan–Meier methods and multivariable Cox regression. Using previously published health-services modelling frameworks, we compared expected versus observed utilization of chemotherapy, surgery, radiotherapy, and endocrine therapy, stratified by stage. Guideline-concordant care was evaluated among patients with stage III disease. Results: The cohort included 125,005 patients. Educational attainment showed a clear stepwise association with survival: median overall survival ranged from 6.2 years among unlettered patients to 20.6 years among those with a university education. In multivariable analysis, university-educated patients had a 50% lower risk of death compared with unlettered patients (HR 0.50, p < 0.001). GDP was associated with survival but demonstrated weaker and less consistent gradients. Educational attainment strongly influenced treatment patterns: chemotherapy utilization exceeded modelled expectations in stage II but fell below expected levels in stages III–IV, with the largest shortfalls among patients with lower education. Radiotherapy and endocrine therapy utilization were below expected across all groups. Among patients with stage III disease, receipt of all guideline-concordant care was associated with markedly improved survival (median 9.4 vs 3.6 years; HR 0.44, p < 0.001). University educated patients with stage III breast cancer had the highest proportion that received all guideline concordant care (60.5%) while among unlettered patients, only 36.3% received all guideline-concordant care. Conclusions: Educational attainment emerged as a major determinant of treatment utilization and survival. Lower education was associated with later stage at diagnosis, lower receipt of chemotherapy, radiotherapy, and endocrine therapy, reduced guideline-concordant care, and substantially poorer outcomes. These findings highlight the need for policies that address social and structural barriers to timely, high-quality cancer care and demonstrate the utility of modelling approaches for identifying and quantifying inequities in treatment delivery.
Impact of age on immune-related toxicities and survival in non–small cell lung cancer treated with immune checkpoint inhibitors: A multicenter retrospective study.
e20641 Background: Immune checkpoint inhibitors (ICIs) are a cornerstone of therapy for non-small cell lung cancer (NSCLC). They can cause immune-related adverse events (irAEs), which may vary with age due to immunosenescence, inflammaging, and comorbidity burden. Older adults are underrepresented in clinical trials, limiting real-world data on age-related toxicity. We evaluated differences in irAEs and overall survival (OS) between younger and older NSCLC patients treated with ICIs. Methods: We conducted a multicenter retrospective cohort study using the TriNetX federated electronic medical record network (113 healthcare organizations). Adults with NSCLC who received ICIs between January 2010 and January 2026 were included. Cohort 1 comprised patients aged 18–64 years; cohort 2 had patients aged 65–90 years. Patients with systemic lupus erythematosus were excluded. The index event was initiation of immunotherapy. ICIs included pembrolizumab, nivolumab, atezolizumab, durvalumab, and ipilimumab. Cohorts were propensity score-matched 1:1 for over 30 variables, including sex, race, major comorbidities, and immunotherapy type. After matching, 1,422 patients were included in each cohort. Outcomes included irAEs (diarrhea, rash, autoimmune thyroiditis, pneumonitis, and adrenalitis) and OS. Autoimmune hepatitis and hypophysitis were excluded due to fewer than 10 events in both cohorts. Odds ratios were calculated for irAEs, and OS was evaluated using Kaplan-Meier analysis with log-rank testing. Results: After propensity matching, baseline characteristics and immunotherapy exposure were balanced between cohorts. Pembrolizumab was most frequently used (59.4% vs 56.6%), followed by nivolumab (20.6% vs 24.0%), atezolizumab (7.5% vs 5.2%), durvalumab (12.3% vs 14.0%), and ipilimumab (4.1% vs 3.9%) in younger and older cohorts, respectively. Younger patients had a significantly higher incidence of rash compared to older patients (212 vs 147, p < 0.001), while rates of autoimmune thyroiditis, pneumonitis, adrenalitis, and diarrhea were similar (p > 0.05). Kaplan-Meier analysis showed a significant difference in OS between cohorts (891 days vs 731 days, p = 0.011). Conclusions: This real-world analysis shows that ICIs have a similar overall toxicity profile across age groups in NSCLC. Rates of most irAEs were comparable between younger and older adults after propensity matching. The higher incidence of rash in younger patients may reflect stronger immune activation and could serve as a marker of enhanced response to immunotherapy. Our results show that increased age alone does not lead to worse toxicity; future studies are needed to define optimal patient selection.
Caregiver professional employment and its impact on glioblastoma patient survival: Results from two prospective studies.
2050 Background: Caregiver burden is a crucial factor to consider in the diagnosis and care of patients with glioblastoma (GBM). Despite growing interest in caregiver issues, data on caregiver burden and, in particular, professional and working burden are limited, as well as the potential effect of caregiver burden on patients’ survival. Methods: We collected caregiver data from two prospective observational studies (ISES 1 and ISES 2) investigating the impact of socioeconomic variables (SES) in GBM patients. In particular, we collected data on the presence and type of caregiver, the caregiver's professional situation, and any professional modifications after the GBM diagnosis at time of adjuvant treatment start and after six months (in patients enrolled in ISES 2 study). Results: Overall, 511 GBM isocitrate dehydrogenases wild-type (IDH-wt) patients were included. The median follow-up period was 55.2 months (95% CI, 44.3–86.3), and the median overall survival was 15.7 months (95% CI, 14.8–17.2). A caregiver was present for 472 patients (92.4%), consisting mainly of partners or daughters and sons. At the start of adjuvant treatment, 158 caregivers (30.9%) reduced their working hours, and 55 caregivers (10.8%) stopped working. For 232 patients, a longitudinal assessment of caregiver condition was collected, and 71 caregivers (30.6%) reduced their working hours, while 33 patients (14.2%) stopped their work activities. Patients whose caregivers were in stable employment had a longer life expectancy than those whose caregivers were not (multivariate - HR 0.76, 95% CI 0.60–0.94, p = 0.03), regardless of other known prognostic factors, as confirmed by a multivariate analysis (Table 1). Conclusions: Our study revealed that GBM diagnosis often leads to a negative impact on caregiver work activities. A stable caregiver employment translates independently to a long-life expectancy of patients with GBM, regardless of other recognized prognostic factors. Policies supporting the professional and social activities of caregivers could improve not only the quality of care for GBM patients but also their clinical outcomes. Clinical trial information: 0321-3733778. VARIABLES Univariate analysesHazard Ratio (95%CI) p value Multivariate analysesHazard Ratio (95%CI) p value MGMTUnmethylated vs methylated 2.23 (1.79 – 2.77)p < 0.001 0.43 (0.34-0.55)p < 0.001 SURGERYBiopsyPartial resectionComplete resection Ref0.51 (0.38-0.69)0.50 (0.37-0.69)p <0.001 Ref0.62 (0.43-0.88)0.61 (0.42-0.89)p = 0.008 TREATMENTRT (40Gy)/CT ->CT vsRT (60Gy)/CT ->CT 1.31 (1.04-1.73)p = 0.02 1.04 (0.77-1.49)p = 0.8 ECOG0 vs 1 or more 0.67 (0.54-0.82)p <0.001 0.72 (0.56-0.91)p = 0.007 STEROIDSYes vs No 1.94 (1.57-2.40)p < 0.001 1.75 (1.31-2.13)p < 0.001 AGE (As continuous for year) 1.02 (1.01-1.03)p < 0.001 1.01 (1.0009-1.03)p = 0.04 CAREGIVER (professional condition)Employed vs other 0.73 (0.59-0.90)p = 0.005 0.76 (0.60-0.94)p = 0.03
Risk of lung cancer among patients with idiopathic interstitial lung disease: A cohort study using All of Us research program data.
e20089 Background: Interstitial lung disease (ILD), a collection of heterogeneous conditions characterized by lung inflammation and scarring, shares molecular mechanisms with lung cancer. Previous studies in homogeneous European or Asian cohorts demonstrated a 7-20% increased risk of lung cancer among individuals with ILD. This study assessed the association between idiopathic ILD and lung cancer using a case-control study design with cases defined as patients with lung cancer and controls as patients with colon cancer in a large, diverse U.S. population dataset. Methods: The NIH All of Us Controlled Tier Dataset, version 8, includes 393,596 participants with Electronic Health Records information, from which 2,070 individuals had a diagnosis of lung cancer (though the data did not allow for parsing by histologic types of lung cancer), and 3,565 individuals had a diagnosis of colon cancer. Among these 5,635 individuals with cancer, 178 had a documented idiopathic ILD diagnosis. Univariable followed by multivariable logistic regression models were used to evaluate the association between idiopathic ILD and lung cancer (versus colon cancer), adjusting for age, sex, race, smoking status, income, and education level. Results: Among 5,635 individuals with cancer, 68.7% were White, 14.6% Black or Black-American, 10.9% Hispanic, and 5.8% of other races. Within this group, 2,070 had lung cancer, 129 (6.2%) of whom had idiopathic ILD, compared to 49 (1.4%) of 3,565 individuals with colon cancer [OR 4.77; 95% CI: 3.41-6.66]. In the multivariable analysis, idiopathic ILD remained an independent risk factor for lung cancer [aOR 4.37; 95% CI: 2.94-6.51; p < 0.0001]. Within the logistic regression model, other risk factors included smoking [aOR 3.18; 95% CI: 2.75-3.66; p < 0.0001], female sex [aOR 1.27; 95% CI: 1.11-1.45; p = 0.0004], and lower education levels. Among ILD patients, those with and without lung cancer were similar in age, sex, race, smoking history, income, and education (all p > 0.05). Notably, 94.6% (122 out of 129) of lung cancer cases in the group with ILD occurred in patients with the pulmonary fibrosis subtype. Discussion: This study replicates the finding from prior studies that ILD is associated with an increased risk of lung cancer, but within the diverse population of the All of Us dataset, extending the generalizability of ILD as a risk factor in pulmonary carcinogenesis. Notably, even after accounting for smoking and other potential confounders, ILD remained a strong independent predictor of lung cancer, indicating that ILD may independently drive carcinogenesis. Conclusions: ILD increases lung cancer risk, with smoking as an additive factor. These findings support implementing and testing preventive strategies, including education, smoking cessation, and routine low-dose CT screening, for individuals with ILD.
Real-world quality outcomes following implementation of a comprehensive care program in community oncology practice in western India.
e13575 Background: Malnutrition, psychosocial distress & impaired quality of life (QoL) adversely affect treatment tolerance & outcomes in cancer patients. Real-world data on structured comprehensive care models from community oncology settings in India remains limited. This study evaluated quality outcomes following implementation of an integrated nutritional, psychosocial & QoL care program. Methods: This retrospective study assessed 3 supportive care components: (1) nutritional status using Subjective Global Assessment (SGA) (2) psychosocial wellness using NCCN Distress Thermometer & (3) QoL using EORTC QLQ-C30. Pre & post-intervention outcomes were compared using paired t-test, Wilcoxon signed-rank test, linear transformation for percentage calculation & Kaplan Meier/ Binary logistics regression. Results: Nutritional status: Of 1155 patients (median age 57 years; M:F ratio 1:1.8), most common cancers evaluated were breast (30.7%), head & neck (16.4%), gastrointestinal (14.5%) & lung cancer (5.5%). SGA assessment was performed at baseline & 90.9% were assessed at cycle 1; 44.5% were malnourished at baseline. Median pre-SGA score of malnourished cohort was 5 vs 6 post-SGA assessmemt (p = < 0.001); categorical SGA status improved in 57.5% & remained stable in 32.7%, while 31.9% (205/642) of initially well-nourished patients deteriorated. Male Gender (OR 1.7; 95% CI 1.3–2.1), head & neck cancer (OR 4.2; 95% CI 2.9–6.1), lung cancer (OR 4.0; 95% CI 2.3-7.0), & gastrointestinal cancers (OR 3.1; 95% CI 2.1-4.6) were independently associated with malnutrition (p < 0.001) Psychosocial Wellness status: Of 2273 patients evaluated (median age 58 years; M:F ratio 1:1.4), the common cancers were: breast (25%), gastrointestinal (15.8%), gynaecologic (13%) & head & neck cancers ( 9.8%). A total of 57.4% had baseline distress scores ≥5. Mean distress score significantly reduced from 5.2 at baseline to 3.5 at follow-up (p < 0.001) with median scores improving from 5 to 3. Overall 50.7% improved, 42.8% remained stable & 6.5% worsened. Baseline distress was higher in patients < 60 years (p < 0.001) & varied by disease stage (p = 0.040) QoL status: Of 21 patients evaluated (median age 66 years; M:F ratio 1:1.6), 61.9% had stage IV disease. Common cancers included gastrointestinal (28%), gynecologic (23%) & breast (19%). Mean global QoL scores improved from 46.0 ± 18.6 at baseline to 59.1 ± 16.4 at follow up (p = 0.007). Improvements were observed across domains: emotional functioning (68.3 ± 19.7 to 85.7 ± 15.6), role functioning (73.0 ± 23.8 to 84.1 ± 19.3), social functioning (69.8 ± 26.2 to 81.0 ± 19.2) & physical functioning (81.3 ± 19.6 to 87.6 ± 13.9). Conclusions: Implementation of a structured comprehensive care program in community oncology practice was associated with significant improvements across nutritional status, psychosocial distress & patient-reported quality outcomes.
A phase 1, multicenter, open-label, dose-escalation and expansion study to evaluate the safety, tolerability, pharmacokinetics, and preliminary efficacy of SBO-154 in subjects with advanced solid tumors.
TPS3161 Background: Mucin 1 (MUC1) is a transmembrane glycoprotein overexpressed in advanced epithelial malignancies, where it promotes tumor progression and correlates with poor clinical outcomes. The membrane-proximal SEA domain of MUC1 is exposed on tumor cells and undergoes rapid internalization, enabling targeted intracellular delivery of cytotoxic payloads. SBO-154 is a MUC1 SEA–directed antibody–drug conjugate (ADC) conjugated to the clinically validated cytotoxic payload monomethyl auristatin E (MMAE) via a protease-cleavable linker. Unlike the shed MUC1-VNTR that creates an antigen sink limiting efficacy of prior ADCs, the negligibly shed MUC1-SEA epitope enables more effective SBO-154 tumor targeting. Nonclinical studies of SBO-154 showed MUC1-dependent cytotoxicity, dose-dependent tumor inhibition in vivo, and a favorable safety for clinical evaluation. This first-in-human study is designed to evaluate the safety, tolerability, pharmacokinetics (PK), immunogenicity, and preliminary antitumor activity of SBO-154 in subjects with locally recurrent or metastatic solid tumors who have exhausted or are intolerant to available therapies. Methods: This phase 1, open-label, multicenter study comprises two parts: dose escalation (Part 1) and dose expansion (Part 2). Eligible adults (≥18 years) must have RECIST v1.1–measurable, locally recurrent or metastatic solid tumors (excluding sarcoma), documented progression after or intolerance to standard therapy, ECOG performance status 0–1, adequate organ function, and tumor tissue for immunohistochemistry (IHC) analysis. SBO-154 will be administered as an intravenous infusion on Day 1 of each 21-day treatment cycle (Q3W) over 30 minutes. Part 1 will use an mTPI-2 design with intra-patient dose escalation from 0.3 mg/kg to 2.5 mg/kg. Approximately 75 subjects will be enrolled in Part 1 (~50 in dose escalation and ~25 in backfill cohorts), and up to 102 subjects in Part 2. After determining the maximum tolerated dose (MTD) and the recommended dose for expansion (RDE), Part 2 will be initiated with a Simon’s two-stage design. Planned expansion cohorts will include NSCLC, ER+ breast cancer, and ovarian cancer, with enrolment to be restricted to tumors with high MUC1-SEA expression. Treatment will continue until disease progression, unacceptable toxicity, or withdrawal from the study. Primary endpoints: include dose-limiting toxicities (DLTs), treatment-related serious adverse events, dose modifications, and clinical/laboratory safety findings. Secondary endpoints: include PK, immunogenicity and efficacy. Enrollment Status: The study initiated enrolment in August 2025. Cohorts 1 and 2 have been completed without any DLTs, and enrolment in Cohort 3 is ongoing as of December 2025. Clinical trial information: NCT07042100 .
Effect of frailty measures on overall survival in prostate cancer: A systematic review and meta-analysis.
5077 Background: Prostate cancer at all stages is associated with a favorable prognosis and treatment is generally well tolerated, including in the elderly. Frailty screening tools such as the Geriatric 8 (G8) have been associated with decreased survival across different cancers. However, the effect of frailty on overall survival (OS) in prostate cancer is poorly understood. Methods: A comprehensive search of PubMed, Embase, and Scopus was performed through December 2025 for studies evaluating the impact of frailty on OS in patients with prostate cancer. Data were analyzed in two distinct cohorts: patients with metastatic castration-resistant prostate cancer (mCRPC) and those with unspecified prostate cancer (PC) (at any stage). Meta-analysis was performed using a random-effects model to calculate pooled hazard ratios (HR) and 95% confidence intervals (CI). Statistical heterogeneity was assessed using the I² statistic and Chi² test. Results: Of the twelve studies that met the inclusion criteria, two were excluded due to non-availability of survival statistics for analysis. Four studies could not be included in the meta-analysis as hazard ratios were not reported. A total of 7120 patients across seven studies were included in the meta-analysis. In the mCRPC cohort (n=4 studies), frailty was a significant predictor of mortality, associated with a pooled HR of 2.16 (95% CI: 1.34–3.49; P = 0.002). Moderate-to-high heterogeneity was observed in this group (I² = 65%). In the unspecified PC cohort (n=2 studies), the association between frailty and OS did not reach statistical significance, yielding a pooled HR of 1.74 (95% CI: 0.69–4.40; P = 0.24). Heterogeneity for the unspecified PC group was I² = 61%. Conclusions: Frailty is associated with increased mortality in elderly patients with metastatic castration-resistant prostate cancer. This effect of frailty was not statistically significant when the population was expanded to include early-stage prostate cancer. These results highlight the importance of geriatric screening for prognostic assessment in elderly patients with prostate cancer, especially in those with metastatic castration-resistant disease. Study characteristics and hazard ratios for overall survival. Study Cancer type Design N Age Assessment tool Frailty % OS Beardo (2019) mCRPC RS 70 ≥75 G8 NR 3.13 (1.13–8.66) Banna (2022) mCRPC PS 234 78 G8 38% 3.58 (1.72–7.49) Wilk (2022) mCRPC PS 49 71 G8 47% 2.24 (1.15–4.36) Deol (2023) mCRPC RS 5822 75.4 VA-FI 39.70% 1.44 (1.35–1.52) Royset (2023) PC PS 73 73.6 EFS 24.60% 1.27 (1.20–1.35) Siwakoti (2024) PC PS 70 70 CARE-FI 21% 3.55 (1.01–12.47) Abbreviations: Retrospective study (RS); Prospective study (PS); Sample size (N); Geriatric 8 (G8); Veterans Affairs Frailty Index (VA-FI); Edmonton Frail Scale (EFS); Cancer and Aging Resilience Evaluation Frailty Index (CARE-FI); Not reported (NR).
Association of chemotherapy with survival in node-positive penile squamous cell carcinoma stratified by nodal burden: A National Cancer Database analysis.
e17033 Background: Node-positive penile squamous cell carcinoma (PSCC) carries a poor prognosis, yet the survival benefit of systemic chemotherapy remains uncertain given the lack of completed randomized trials and reliance on limited phase II and retrospective data. Although NCCN guidelines recommend chemotherapy for patients with ≥3 positive lymph nodes, whether chemotherapy benefit differs by pathologic nodal burden has not been formally evaluated using interaction analysis. We hypothesized that chemotherapy is associated with improved overall survival (OS) among patients with high nodal burden (≥3 nodes) but not among those with low nodal burden (1–2 nodes). Methods: Using the National Cancer Database (2004–2022), we identified adults with pathologically node-positive, nonmetastatic PSCC who underwent regional lymph node dissection with numeric nodal counts available. Heterogeneity of treatment effect by nodal burden was formally assessed using a chemotherapy × nodal burden interaction term. Nodal burden was classified as low (1–2 positive nodes) or high (≥3 positive nodes). OS was estimated using the Kaplan-Meier method and compared using the log-rank test. Multivariable Cox proportional hazards regression adjusted for age, race, ethnicity, Charlson-Deyo comorbidity score, T stage, tumor grade, primary tumor surgery, radiotherapy, facility type, insurance status, and year of diagnosis. A 90-day landmark analysis was performed to mitigate immortal time bias. Results: Among 1,300 patients, 589 (45%) received chemotherapy. A statistically significant chemotherapy × nodal burden interaction was observed (p = 0.009), indicating differential survival associations by nodal burden. In the overall cohort, chemotherapy was not associated with OS (log-rank p = 0.47). Among patients with high nodal burden, chemotherapy was associated with improved OS (adjusted HR 0.76, 95% CI 0.59–0.98, p = 0.03; log-rank p = 0.002), with median OS of 35.9 months compared with 20.6 months for those not receiving chemotherapy. In contrast, among patients with low nodal burden, chemotherapy was not associated with improved OS (adjusted HR 1.19, 95% CI 0.95–1.49, p = 0.13; median OS 59.0 vs 65.7 months; log-rank p = 0.79). Findings were consistent in landmark analyses (interaction p = 0.008). Conclusions: In this large national cohort, the association between chemotherapy and survival in node-positive, nonmetastatic PSCC differed significantly by pathologic nodal burden. Chemotherapy was associated with improved survival among patients with ≥3 positive lymph nodes but not among those with limited nodal involvement. These findings demonstrate clinically meaningful heterogeneity in chemotherapy benefit and may inform more precise patient selection for systemic therapy in node-positive PSCC.