Browse Articles
Discover research articles across all indexed journals
Clinical trial in progress: Phase II trial of neoadjuvant tebentafusp in patients with locally advanced primary uveal melanoma.
TPS9607 Background: Uveal melanoma (UM) is the most common intraocular malignant tumor in adults. Primary UM are effectively treated by plaque radiotherapy or enucleation; however, up to 50% of UM patients ultimately succumb to advanced disease. Tebentafusp-tebn is a bispecific gp100 peptide human leukocyte antigen (HLA)-directed cluster of differentiation 3 (CD3) T-cell engager that became the first FDA approved treatment for unresectable or metastatic UM. Among the metastatic UM patients who received tebentafusp on clinical trial (IMCgp100-01 (n=19), IMCgp100-102 (n=146), and IMCgp100-202 (n=245)), 12 patients had recurrent orbital tumors. All orbital lesions were stable or had achieved shrinkage with tebentafusp. The best percentage changes from baseline were reported as -3 to -40% reduction in intra-ocular target lesions. All intra-ocular non-target lesions were stable or achieved a complete response after tebentafusp treatments. These preliminary data indicate that tebentafusp may have potential for treating surgically unresectable primary UM. Methods: This is an investigator-initiated, prospective, single arm phase II trial evaluating neoadjuvant tebentafusp in patients with locally advanced primary UM [NCT06414590]. Patients must be 18 years or older, HLA-A*02:01 with a primary UM with T3 or T4 category tumor size. Patients cannot have a symptomatic UM that requires immediate ophthalmological intervention such as enucleation. Patients will be treated with up to 8 weeks of neoadjuvant tebentafusp, followed by radioactive plaque or enucleation. Tebentafusp will be administered as follows: 20mcg on Day 1, 30mcg on Day 8, 68 mcg on Day 15, and weekly doses of 68 mcg thereafter. The first three treatments will require inpatient hospitalization for overnight observation. The primary endpoint of the study is regression (defined as ≥ 20% reduction in tumor volume) of primary UM after tebentafusp treatment in 20% of treated patients. Secondary endpoints include toxicity assessment and ctDNA analysis in both blood and aqueous humor. Exploratory endpoints include changes to visual acuity, radiation dose to the fovea, and treatment (enucleation to plaque brachytherapy). The efficacy of this combination treatment will be assessed using the Simon’s two stage design. In stage I, a total number of 8 patients are accrued and if there are 0 responses among these 8 patients, further enrollment of patients may be stopped. Otherwise, an additional 11 patients will be accrued in stage II, resulting in a total sample size of 19 patients. Enrollment began in September 2025 at Thomas Jefferson University Hospital, in conjunction with Wills Eye Hospital. Clinical trial information: NCT06414590 .
Longitudinal trends in pembrolizumab uptake for triple-negative breast cancer (TNBC): A comparative analysis of Appalachian and non-Appalachian populations at a tertiary care center.
e23412 Background: Appalachian populations face unique barriers leading to cancer care disparities. As pembrolizumab becomes standard for triple-negative breast cancer (TNBC), equitable access for rural patients is critical. This study evaluates immunotherapy uptake between Appalachian and non-Appalachian patients at the University of Kentucky. Methods: A retrospective analysis using Epic SlicerDicer identified TNBC patients (July 2021- December 2025) at the University of Kentucky, segmented by Appalachian and non-Appalachian counties. Data on demographics, comorbidities, staging, and insurance were extracted. Chi-square tests compared groups; annual pembrolizumab uptake was calculated with confidence intervals. Results: 405 patients with TNBC were identified (220 Appalachian, 185 non-Appalachian). Groups were matched for age < 49 (p = 0.8) and gender (p = 0.91); median ages were 57 and 60 years, respectively. Appalachian patients demonstrated a higher prevalence of obesity (59.5% vs 50.3%; p = 0.06) and Type 2 Diabetes (20.0% vs 13.0%; p = 0.053) compared to the non-Appalachian cohort. Other comorbidities, including GERD (p = 0.19), depression (p = 0.11), hypertension (p = 0.21), and anxiety (p = 0.84), showed no significant regional differences. Non-Appalachian patients were significantly more likely to hold private insurance (40.5% vs 28.6%; p = 0.011), whereas Appalachian patients had significantly higher rates of Medicare enrollment (39.5% vs 29.2%; p = 0.029). Medicaid (p = 0.52) and self-pay (p = 0.2) rates were comparable between regions. Pembrolizumab uptake increased over time in both Appalachia and other counties, but consistently lagged in Appalachia. With an average of (29.7%) in the non-Appalachian cohort compared to the Appalachian cohort (21.4%), a difference that trended towards but did not reach statistical significance (p = 0.053). Despite this disparity in immunotherapy utilization, mortality rates remained similar between the two groups (p = 0.83), with 35 deaths in the Appalachian cohort and 28 in the non-Appalachian cohort. Conclusions: A persistent disparity in pembrolizumab uptake, with almost 15% gap in peak usage, suggests that systemic barriers to immunotherapy persist in rural Kentucky. Addressing insurance disparities and metabolic burden is critical. Further investigation into referral patterns and coverage is needed to ensure equitable access as standards of care evolve. Patient characteristics and insurance distribution. Variable Appalachian (n=220) Non-Appalachian (n=185) P-value Age < 49 63 55 0.8 Obesity 131 93 0.06 Type 2 DM 44 24 0.053 Pembro Use 47 55 0.053 Private Insurance 63 75 0.011 Medicare 87 54 0.029 Medicaid 36 26 0.52 Self Pay 13 17 0.2 Death 35 28 0.83 (P-values calculated via Chi-square analysis).
HRD-positive breast cancer genomics: Cross-application of ovarian cancer HRD scores.
1037 Background: Homologous recombination deficiency (HRD) is a well-established biomarker for PARP inhibitor (PARPi) therapy in ovarian cancer and is increasingly being explored in breast cancer. However, current definitions of HRD positivity in breast cancer rely heavily on ovarian cancer–derived thresholds, raising questions about their biological relevance. Here, we characterize the genomic landscape of HRD-positive breast cancer and assess the applicability of ovarian cancer–based HRD scoring criteria. Methods: We performed next-generation sequencing on 1,159 breast cancer samples to evaluate HRD status, somatic alterations, and germline variants. HRD scores were calculated as the sum of loss of heterozygosity (LOH), telomeric allelic imbalance (TAI), and large-scale state transitions (LST), using established ovarian cancer cutoffs. These results were compared with HRD scores from 617 ovarian cancer cases analyzed using the same methodology. Results: Pathogenic or likely pathogenic BRCA mutations were detected in 105 breast cancer patients (9.1%), including 33 somatic and 80 germline events, all of which occurred in HRD-positive tumors. Overall, 26.2% (304/1,159) of breast cancer cases were classified as HRD-positive, with a median HRD score of 21. HRD scores were significantly higher in triple-negative breast cancer (TNBC) than in non-TNBC subtypes (P < 0.001). HRD-positive tumors exhibited elevated genomic instability, marked by significantly higher copy number variation burden (P < 0.001) and tumor mutational burden (TMB) (P < 0.001), along with enriched TP53 alterations. Comparative analysis revealed that ovarian cancers had significantly higher HRD scores than breast cancers (median: 43 vs. 21, P < 0.001), suggesting that applying ovarian cancer–derived thresholds may underestimate HRD positivity in breast cancer. Conclusions: HRD-positive breast cancer displays distinct genomic features, including increased genomic instability and TP53 alterations. However, HRD scores in breast cancer are markedly lower than those in ovarian cancer, indicating that ovarian cancer–based cutoffs may not be optimal for breast cancer. Future studies should refine breast cancer–specific HRD thresholds to improve patient selection for PARPi therapy.
Surgical outcomes and postoperative complications of robotic versus conventional nipple-sparing mastectomy in breast cancer: A meta-analysis.
585 Background: Breast cancer is the most common type of cancer, affecting millions of women worldwide. Nipple-sparing mastectomy has emerged as a technique that leaves most of the healthy breast skin, areola, and nipple. This technique is widely used but there is little evidence comparing conventional and robotic nipple-sparing mastectomy. This meta-analysis aimed to compare the efficacy, safety, and postoperative complications of conventional and robotic nipple-sparing mastectomy. Methods: A systematic search of PubMed, Scopus, Web of Science, and the Cochrane Library was conducted to identify studies comparing robotic and conventional nipple-sparing mastectomy. The primary outcome was nipple and skin necrosis; secondary outcomes included operative time and Clavien-Dindo Grade III complications. Pooled estimates were calculated using a random-effects model, and heterogeneity was assessed using the I2 statistic. Results: A total of nine studies encompassing 1,220 participants (423 in the robotic group and 797 in the conventional group) were included in the final analysis. Robotic nipple-sparing mastectomy (RNSM) was associated with a significantly lower risk of nipple and skin necrosis compared to conventional nipple-sparing mastectomy (CNSM), with a pooled odds ratio of 0.47 (95% CI: 0.26–0.86; P = 0.01) and low heterogeneity (I2 = 20.5%). Conversely, RNSM was associated with a significantly longer operative time, with a weighted mean difference of 52.86 minutes (95% CI: 20.28–85.43; P = 0.001) and substantial heterogeneity (I2 = 96%). Regarding safety, there was no statistically significant difference in Clavien-Dindo Grade III complications between the two techniques (OR = 0.60; 95% CI: 0.35–1.05; P = 0.07; I2 = 0%), although the results showed a trend favoring the robotic approach. Conclusions: Robotic nipple-sparing mastectomy demonstrates a superior safety profile regarding ischemic complications, significantly reducing the risk of nipple and skin necrosis compared to the conventional approach. However, it is associated with longer operative times. The incidence of major complications (Grade III) is comparable between both techniques.
Economic analysis of first-line treatment for advanced melanoma in a public hospital.
e13618 Background: Access to immune checkpoint inhibitors (ICIs) for advanced cutaneous melanoma remains constrained in Brazil’s public health system due to reimbursement limitations. We estimated the direct medical costs of first-line (1L) treatment for advanced cutaneous melanoma at a Brazilian public reference hospital, comparing chemotherapy- and immunotherapy-based strategies. Methods: We conducted a retrospective observational cohort study including patients with advanced cutaneous melanoma treated at the Brazilian National Cancer Institute between January 2020 and December 2023. During this period, chemotherapy with dacarbazine was consistently available, whereas immunotherapy with pembrolizumab became available starting in December 2021. Individual-level resource utilization data were extracted from medical records and institutional databases. Direct medical costs of 1L treatment were estimated from the public healthcare system perspective using a bottom-up microcosting approach. Values were reported in US dollars (USD) with 1 USD = 5.62 Brazilian Reais (reference year 2024). Results: Eighty-seven patients received 1L treatment during the study period: 43 (49%) received dacarbazine and 44 (51%) received pembrolizumab. The total direct medical cost of dacarbazine-based treatment was 311,492.59 USD, corresponding to a mean cost of 7,244.01 USD per patient. In contrast, pembrolizumab-based treatment generated a total cost of 3,364,519.60 USD, with a mean cost of 76,466.35 USD per patient. Cost composition differed markedly between strategies. In the dacarbazine group, hospitalization accounted for 78% of total costs, followed by diagnostic exams and procedures (10%). In the pembrolizumab group, drug acquisition and administration accounted for 90% of total costs, whereas hospitalization represented only 7%. Notably, in absolute terms, hospitalization costs were similar between groups. Costs for exams and supportive care (e.g. transfusion) were approximately doubled in the pembrolizumab group. Conclusions: First-line treatment strategies for advanced cutaneous melanoma exhibit different cost structures within the Brazilian public healthcare system. Chemotherapy-based regimens are associated with lower overall costs, largely driven by hospitalization and supportive care, while immunotherapy-based treatment results in a significantly higher budget impact driven primarily by drug acquisition. Although ICIs are associated with meaningful survival benefits, their affordability remains a major challenge in publicly funded systems. Future analyses incorporating survival, quality-of-life outcomes, and productivity measures are needed to contextualize these costs and inform value-based policy decisions. Strategies such as dosing optimization or interval modification warrant further evaluation to improve sustainability without compromising clinical benefit.
Predictors of survival of patients with cancer-associated thrombosis: Experience from a tertiary hospital in Sub-Saharan Africa.
e24127 Background: Cancer significantly increases the risk of thrombosis. Recent advances in cancer treatment may have led to more cases of cancer-associated thrombosis (CAT). CAT is linked with substantial morbidity and mortality. Methods: This study aimed to assess the clinical characteristics, treatment patterns, and survival outcomes of patients with CAT at Tikur Anbessa Specialized Hospital (TASH), a tertiary hospital in Addis Ababa, Ethiopia. A single-center institution-based retrospective study was done on 274 patients with CAT. A structured, original data abstraction tool was used to extract data from patients' medical records. Bivariate and multivariate binary logistic regression were used to determine factors associated with 1-year mortality. Results: The median (interquartile range (IQR)) age of the patients was 48 (18-87) years, and 183 (66.8%) of them were female. 106 (38.7%) of the patients had at least one comorbidity. The median (IQR) values for the Charlson Comorbidity Index (CCI) and Khorana Risk Score were 5.5 (2.0-7.0) and 1 (0-2), respectively. Gynecological malignancies were the most common cancers associated with thrombosis (29.1%), followed by hematological malignancies (24.4%) and intra-abdominal malignancies (18.2%). The three most common types of thrombosis identified were deep vein thrombosis (DVT) (72.6%), pulmonary embolism (PE) (10.9%), and portal vein thrombosis (PVT) (10.9%). The most frequently utilized anticoagulant was rivaroxaban (40.1%), followed by warfarin (38.0%). The one-year overall survival rate was 81%, and higher Khorana Risk Score (adjusted odds ratio (AOR), 1.55; 95% confidence interval (CI), 1.17-2.07; p = 0.003) and having solid organ malignancies (AOR, 2.98; 95% CI, 1.10-8.07; P = 0.032) were significantly associated with increased 1-year mortality. Conclusions: In this cohort, the one-year survival rate for patients with CAT was higher than previously reported in regional studies. The Khorana Risk Score and the type of malignancy were significant predictors of survival.
Impact of adverse social determinants of health on long-term mortality and comorbidities in patients with cancers of the lip, oral cavity, and pharynx: A propensity score–matched cohort study using federated electronic health records.
e18124 Background: Social determinants of health (SDOH), encompassing socioeconomic and environmental factors, are pivotal in modulating cancer outcomes, yet their long-term implications in head and neck cancers, particularly cancers of the lip, oral cavity, and pharynx (CLOP), remain inadequately characterized amid rising disparities in oncology care. This study investigates the association between adverse SDOH and clinical outcomes in patients with CLOP. Methods: This retrospective, propensity score-matched cohort study utilized data from the TriNetX network, comprising electronic medical records from 88 healthcare organizations. Adults (aged ≥18 years) with CLOP (ICD-10-CM: C00-C14) were stratified into cohorts without (n = 265,544) and with (n = 6,235) adverse SDOH (ICD-10-CM Z59/Z60 codes for housing instability, food insecurity, low income, transportation barriers, and social environment issues). Propensity score matching (1:1) balanced cohorts (n=5430 each) on demographics, BMI, comorbidities, and Eastern Cooperative Oncology Group performance status. The index event was initial CLOP diagnosis and outcomes were assessed from 1-day post-index onward. Analyses included measures of association, Kaplan-Meier survival with log-rank tests and hazard ratios (HRs; Cohort 2 [SDOH] vs Cohort 1 [non-SDOH]), with statistical significance at p<0.05. Results: Matched cohorts were balanced (mean [SD] age, 64 [14] years; 66% male; 63% White). Adverse SDOH was associated with increased mortality (risk, 30.3% vs 25.4%; risk difference, 0.050 [95% CI, 0.032-0.067]; P < .001; HR, 1.76 [95% CI, 1.63-1.89]; P = .038; median survival, 2057 vs 5111 days). Severe sepsis risk was higher with SDOH (6.7% vs 5.7%; risk difference, 0.010 [95% CI, 0.000-0.019]; P = .04; HR, 1.69 [95% CI, 1.44-1.98]; P = .035). Multiple conditions demonstrated comparable associations with worse survival: respiratory disease (HR, 1.23 [95% CI, 1.10-1.39]; P = .38), malnutrition (HR, 1.30 [95% CI, 1.17-1.45]; P = .62), acute kidney failure/CKD (HR, 1.39 [95% CI, 1.24-1.56]; P = .006), infectious diseases (HR, 1.47 [95% CI, 1.33-1.62]; P = .66), and metabolic disorders (HR, 1.27 [95% CI, 1.13-1.42]; P = .11). Dysphagia risk was lower with SDOH (25.6% vs 32.0%; risk difference, -0.064 [95% CI, -0.086 to -0.043]; P < .001; HR, 0.88 [95% CI, 0.81-0.97]; P = .073). Mental/behavioral disorder risk was lower with SDOH (25.4% vs 28.8%; risk difference, -0.034 [95% CI, -0.065 to -0.003]; P = .04; HR, 1.22 [95% CI, 1.06-1.40]; P = .03). Conclusions: Adverse SDOH significantly worsen long-term survival and select morbidity risks in CLOP patients, highlighting the need for multifaceted interventions integrating social support into oncologic frameworks to mitigate disparities and improve equitable oncology care.
Chemotherapy, endocrine therapy, and cognitive impairment in long-term breast cancer survivors.
523 Background: Chemotherapy and endocrine therapy have known neurobiological effects, yet evidence on their associations with long-term cognitive impairment remains limited. This study examined the relationship between prior chemotherapy, endocrine therapy, and cognitive impairment among long-term breast cancer survivors. Methods: Between July and September 2025, 1,338 participants from the Chicago Multiethnic Epidemiologic Breast Cancer Cohort were surveyed and completed the Functional Assessment of Cancer Therapy-Cognition (FACT-Cog), a validated self-report instrument assessing 4 specific domains: perceived cognitive impairment (PCI; scores 0–72), cognitive-related quality of life (QoL; scores 0–16), perceived cognitive abilities (PCA; scores 0–27), and comments-from-others (scores 0–16). Lower scores reflect worse cognitive function. Cognitive impairment was defined as a PCI score <54. Multivariable logistic regression was used to evaluate associations between chemotherapy and endocrine therapy with cognitive impairment. Multiple linear regression was used to assess associations with continuous FACT-Cog domain scores. All models were adjusted for age, race/ethnicity, radiotherapy, surgery history, and educational attainment. Results: The mean age at survey of the 1,338 study participants was 64.3 years (SD 11.7), and the median time from diagnosis to survey was 9.4 years (IQR 6.1–14.1). Overall, 299 (22.3%) met criteria for cognitive impairment. Additionally, 594 (44.4%) and 851 (63.6%) of the participants had prior chemotherapy and endocrine therapy, respectively. In adjusted analyses, chemotherapy (adjusted odds ratio [aOR] 1.39; 95% CI 1.03 to 1.89; p =0.033) and endocrine therapy (aOR 1.86; 95% CI 1.35 to 2.57; p <0.001) were significantly associated with cognitive impairment. Chemotherapy was associated with lower PCI ( β −2.11; 95% CI −3.61 to −0.61; p =0.006), and lower QoL ( β −0.60; 95% CI −1.01 to −0.19; p =0.04). Endocrine therapy was associated with lower scores in PCI ( β −2.52; 95% CI −4.02 to −1.02; p =0.001), QoL ( β −0.55; 95% CI −0.96 to −0.14; p =0.009), PCA ( β −1.54; 95% CI −2.29 to −0.79; p <0.001), and comments-from-others ( β −0.28; 95% CI −0.49 to −0.07; p =0.008). In joint models, those who received both chemotherapy and endocrine therapy (aOR 2.46; 95% CI 1.46 to 4.18; p =0.001) and those who received endocrine therapy alone (aOR 1.71; 95% CI 1.04 to 2.81; p =0.035) had significantly higher odds of cognitive impairment than participants who received neither therapy. Conclusions: In this multiethnic cohort of long-term breast cancer survivors, prior chemotherapy and endocrine therapy were independently associated with worse perceived cognitive functioning across multiple domains. These findings underscore the need for future studies to identify neuroprotective strategies and interventions to mitigate treatment-related cognitive sequelae.
Comparative outcomes in chronic lymphocytic leukemia patients with and without myasthenia gravis: A propensity score–matched analysis.
e19019 Background: Chronic lymphocytic leukemia (CLL) is commonly associated with autoimmune complications, most of which are hematological in nature. Rarely, non-hematological autoimmune disorders such as myasthenia gravis (MG) have been reported to occur in patients with CLL, with only a small number of cases described in the literature since the 1960s. Despite this, the clinical impact remains poorly characterized, particularly regarding outcomes such as mortality, hospitalization, infectious risk, and treatment patterns. Methods: A retrospective cohort study was conducted using electronic health record data to identify adults with CLL, with and without MG. Propensity score matching was performed using age, sex, and additional clinical variables to create two balanced cohorts. Outcomes assessed included mortality, hospitalizations, sepsis/bacteremia, and use of immunomodulatory therapies. Statistical analyses included Kaplan–Meier curves, log-rank tests, and Cox proportional hazards modeling. Results: After matching, each cohort contained 171 patients with similar demographics. Mortality risk was 16.4% in CLL patients and 18.1% in CLL + MG patients (risk difference −0.018; 95% CI: −0.098 to 0.063; p = 0.668). Hospitalization occurred in 29.2% of CLL patients versus 39.2% in CLL + MG patients (risk difference −0.099; 95% CI: −0.199 to 0.001; p = 0.053). Rates of sepsis/bacteremia were 17.0% versus 20.5% (risk difference −0.035; 95% CI: −0.118 to 0.048; p = 0.405). Use of acetylcholinesterase inhibitors was significantly higher in the MG group (46.2% vs. 5.8%; risk difference −0.404; 95% CI: −0.486 to −0.321; p < 0.001), as was corticosteroid usage (74.9% vs. 56.1%; risk difference −0.187; 95% CI: −0.286 to −0.088; p < 0.001) and steroid-sparing agents (15.2% vs. 5.8%; risk difference −0.094; 95% CI: −0.158 to −0.029; p = 0.005). Kaplan–Meier analysis showed no significant difference in overall survival between groups (mortality, log-rank p = 0.738), but survival probability declined more markedly with MG in specific therapeutic subgroups. Conclusions: Patients with both CLL and MG exhibit higher rates of immunosuppressive and immunomodulatory therapy use and trends toward increased mortality and hospitalizations, though most results are not statistically significant except for specific medication exposures. These findings highlight the need for tailored management and further research for this clinical population.
Evaluation of intervention strategies to improve follow-up after FIT/FOBT tests: A systematic review and meta-analysis.
10549 Background: Colorectal cancer (CRC) is the third most common cancer worldwide, with the majority of patients diagnosed at advanced stages. Early detection through screenings has been shown to improve morbidity and mortality. However, follow-up after Fecal Immunochemical Test (FIT) / Fecal Occult Blood Test (FOBT) screenings remains low, with 30%-61% compliance to colonoscopy screening and 28-49% compliance to repeat FIT/FOBT surveillance. Our study aims to evaluate 1) interventions targeting colonoscopy completion post-positive FIT/FOBT and improving adherence to follow-up after negative FIT/FOBT, and 2) the rate of colonoscopy completion post-positive FIT/FOBT and the rate of repeat FIT/FOBT adherence after a negative result. Methods: We searched MEDLINE, Embase, and Cochrane from inception to December 20, 2025, to identify cohort studies or randomized controlled trials (RCTs) comparing interventions to improve follow-up colonoscopy after positive FIT/FOBT and repeat FIT/FOBT after negative FIT/FOBT. We computed risk ratios (RR) with accompanying 95% confidence Intervals (CIs) for each study and pooled the results using a random-effects meta-analysis. Results: Twenty-seven studies involving 30,974 patients were included. 19 studies (11,905 patients) evaluated interventions to improve colonoscopy completion after a positive FIT/FOBT. Patient-level interventions (RR 1.30, 95% CI 1.17–1.44, n=13) and system-level interventions (RR 1.45, 95% CI 1.20–1.77, n=6) significantly increased colonoscopy completion. Patient navigation (RR 1.28, 95% CI 1.15–1.41), patient reminders (RR 1.32, 95% CI 1.04–1.67), and provider reminders (RR 1.45, 95% CI 1.16-1.83) were effective. Patient reminders via messaging (RR 1.55, 95% CI 1.31-1.83) were more effective than calling (RR 1.10, 95% CI 0.91-1.33, p=0.0083). Follow-up colonoscopy increased from 45.8% in controls to 65.9% after interventions. 8 studies (19,069 patients) assessed repeat FIT/FOBT adherence after a negative result. Patient navigation (RR 1.89, 95% CI 1.36–2.63) and patient reminders (RR 1.31, 95% CI 1.01–1.71) significantly improved adherence, increasing rates from 46.0% in controls to 66.5% after interventions. Based on our findings, we developed a protocol to improve colonoscopy completion post-positive FIT/FOBT and adherence to follow-up after negative FIT/FOBT. Conclusions: Patient-level and system-level interventions improve adherence to follow-up colonoscopy post FIT/FOBT. Patient navigation and patient reminders increase completion of follow-up colonoscopy and adherence to follow-up FIT/FOBT. Healthcare systems should adopt patient-level and system-level interventions to enhance compliance with follow-up after initial FIT/FOBT testing.
Uptake of screening breast MRI in high-risk patients without a personal history of breast cancer.
e13527 Background: Early detection of breast cancer reduces breast cancer mortality. Women at higher-than-average risk for the development of breast cancer should begin mammographic screening at an earlier age than their average-risk counterparts and should consider supplemental screening. The American College of Radiology (ACR) Commission on Breast Imaging recommends screening MRI as the supplemental imaging modality of choice for most higher-than-average risk women. This includes women with a lifetime risk of breast cancer of 20% or greater, according to breast cancer risk assessment tools. In this study, we evaluated the uptake of supplemental screening MRI in patients ages 25-85 classified as high risk (Tyrer Cuzick (TC) score greater than 20%) without a personal history of breast cancer (PHBC). Methods: Data was extracted from the EMR of an academic medical center for patients with a mammogram between 07/20/2020 to 07/18/2025, a TC ≥ 20%, and who did not have a personal history of breast cancer. We assessed differences in MRI screening uptake based on access to healthcare, patient demographics, and mammographic breast density in the population that was eligible for supplemental MRI. Chi-squared analyses were used to compare the groups that did and did not undergo supplemental screening MRI. Results: There were 5647 patients identified as having a TC ≥ 20%. Of those, 35% (n =1962) underwent supplemental MRI screening, while 65% (n =3685) did not. Patients who saw a high-risk breast cancer (HRBC) provider (n =1722, 30%) were far more likely to undergo MRI (n =1178/1722, 68%) compared to those without a HRBC provider (n =784/3925, 20%) (p-value <0.001). Patients who were successfully contacted by a nurse navigator (NN) (n = 2037) were also more likely to undergo MRI (n =817/2037, 40%) compared to those who were unable to be reached by a NN (n = 1016/3610, 28%) (p-value < 0.001). Conclusions: There are a variety of factors that influence whether women at a higher-than-average risk for developing breast cancer complete a recommended screening MRI. While multiple variables in our study displayed statistical significance in the uptake of screening breast MRI, having a HRBC provider and being contacted by a NN were the factors that most strongly correlated with patients undergoing screening MRI.
Evaluation of homologous recombination deficiency (HRD) in gastric cancer: Real-world prevalence and outcomes with first-line (1L) platinum-based therapy.
4039 Background: Metastatic gastric and gastroesophageal junction cancers (GC/GEJ) have poor outcomes with limited predictive biomarkers to guide first-line (1L) therapy. HRD predicts sensitivity to platinum chemotherapy and PARP inhibition in several malignancies, but its clinical relevance is not well defined in GC/GEJ. We evaluated the association between HRD alterations, genomic context, and clinical outcomes following 1L platinum-based therapy (PBT) in metastatic GC/GEJ. Methods: We queried the de-identified Tempus multimodal database using the Tempus Lens Platform to analyze a cohort of patients with advanced GC/GEJ. All patients had DNA (xT, 648 gene DNA-seq panel) and/or RNA sequencing (xR, whole transcriptome RNA-seq). Patients were classified as HRD-altered (HRD-alt) if they harbored pathogenic or likely pathogenic germline or somatic alterations in at least one of 23 predefined HRD-associated genes; all others were classified as HRD wild-type (HRD-wt). Tumor mutational burden (TMB-H, ≥ 10 mut/Mb) and microsatellite instability (MSI) were assessed. Real-world (rw) overall survival (rwOS) was estimated using Kaplan–Meier methods. Ηazard ratios (HRs) from Cox-proportional hazards models were adjusted for relevant clinical and molecular covariates at a significance level of p < 0.05. Results: Among 1,066 eligible patients, 257 (24%) were HRD-alt and 809 (76%) were HRD-wt. HRD-alt tumors were associated with TMB-H (21% vs 5.9%, p < 0.001) and MSI-high status (12% vs 1.2%, p < 0.001). Median OS was longer in HRD-alt compared with HRD-wt patients (15 vs 12 months, HR:0.79, p = 0.033). Notably, HRD-alt tumors were strongly enriched for ERBB2 alterations compared to HRD-wt tumors (30% vs 8.8%, p < 0.001), predominantly driven by amplification (86%). Further stratification by HRD and ERBB2 status showed reduced HRs for rwOS in HRD-alt only (HR 0.78; 95%, p = 0.034), dual ERBB2/HRD-alt tumors (HR 0.63; 95% CI, 0.45–0.88, p = 0.006), and ERBB2-alt only (HR 0.52; 95% CI, 0.36–0.76, p < 0.001) compared to the wt/wt group. ERBB2-alt tumors demonstrated distinct co-mutation patterns enriched for cell-cycle and proliferative genes ( CCNE1, TOP2A) , whereas ERBB2 –wt tumors more frequently harbored KRAS , ARID1A , and CDH1 alterations. Conclusions: In this large rw-cohort of metastatic GC/GEJ, HRD alterations identified a biologically distinct subgroup with improved OS and greater durability on 1L PBT. HRD status was associated with ERBB2 amplification, and the HRD/ERBB2 status was associated with improved survival. These findings support HRD as a clinically relevant biomarker in GC/GEJ and justify prospective evaluation of HRD-informed treatment strategies.
An open-label, randomized phase 2 trial of ramucirumab and pembrolizumab versus pembrolizumab as first-line therapy for PD-L1–positive, recurrent or metastatic head and neck squamous cell carcinoma (RM-HNSCC).
6026 Background: Vascular endothelial growth factor (VEGF) is commonly overexpressed in head and neck squamous cell carcinoma (HNSCC). Combination therapies inhibiting VEGF and PD-(L)1 have improved efficacy outcomes compared to PD-(L)1 inhibitors alone in pre-clinical models and in other cancer types. The hypothesis of this randomized phase 2 trial in RM-HNSCC was the objective response rate (ORR) with ramucirumab (a VEGF-2 inhibitor) and pembrolizumab would be higher than pembrolizumab alone. Methods: Eligible patients had untreated RM-HNSCC with PD-L1 CPS >1, an ECOG performance status of 0 or 1, and adequate organ function. RM disease within 6 months of curative-intent systemic therapy was permitted. Patients were randomized (stratification factors: HPV status [+ or -] and PD-L1 CPS [1-19 or ≥20]) 2:1 to ramucirumab 10mg/kg and pembrolizumab 200mg (Arm 1) or pembrolizumab (Arm 2) given on Day 1 of each 3-week cycle. The primary endpoint was tumor response as assessed with RECIST v1.1 by BICR. A two-stage group sequential design was used with an O’Brien-Fleming stopping rule to accept the null hypothesis at a one-sided alternative hypothesis. We hypothesized an ORR of ≥50% in Arm 1 and ≤19% in Arm 2. The required sample sizes for stage 1 and stage 2 to obtain 80% power at the type I error of 5% were 36 and 72, respectively. At the end of stage 1, if the standardized Z test statistic was ≥0.453, the study could proceed to stage 2. Otherwise, the study was to be stopped, and the null hypothesis accepted. Secondary endpoints include duration of response (DoR), progression free survival (PFS), overall survival (OS) and adverse events. Here, the results of the interim analysis are reported per protocol. Results: Thirty-seven patients were enrolled and treated in stage 1. Median age was 65 years (IQR 60-70), 76% were male, and tumor characteristics (including status of HPV and PD-L1 CPS and prior systemic therapy within 6 months) were balanced between the two arms. The ORR was 28% (7 of 25 patients, 95% CI: 12.1-49.4%) in Arm 1 and 33.3% (4 of 12 patients, 95% CI: 9.9-65.1%) in Arm 2 [z= -0.327]. The null hypothesis was accepted, and the trial did not proceed to stage 2. Conclusions: Among patients with previously untreated PD-L1 positive, RM-HNSCC, ramucirumab and pembrolizumab did not result in a higher ORR than pembrolizumab. Clinical trial information: NCT05980000 .
Inpatient national mortality analysis in hospitalized patients with cancer diagnoses using machine learning.
e13686 Background: In-hospital mortality among hospitalized patients with cancer diagnoses remains a major clinical and health-system burden. National data describing the prevalence of cancer diagnoses among hospitalized patients and associated inpatient mortality patterns are needed to better characterize high-risk populations. Methods: We performed a cross-sectional analysis of adult hospitalizations in the National Inpatient Sample (NIS) from 2016–2021 that included at least one ICD-10 cancer diagnosis code (C00–C96), totaling 3,310,634 admissions. A tree-based gradient boosting machine learning model was developed to predict in-hospital mortality. Model interpretability was assessed using SHapley Additive exPlanations (SHAP). Results: The most prevalent cancer diagnoses in the dataset were secondary bone cancer (11.3%), secondary liver cancer (10.8%), prostate cancer (7.0%), lung cancer (5.5%), and secondary brain cancer (5.4%). Overall in-hospital mortality was 5.7%. The machine learning model demonstrated strong discrimination (ROC-AUC 0.859; accuracy 94.6%). Among frequently observed cancer diagnoses, ICD-10 codes corresponding to acute leukemia were most commonly present among patients who died during hospitalization, with an in-hospital mortality rate of 13.7%. Other cancer diagnoses with high inpatient mortality included disseminated cancer, secondary adrenal cancer, and primary liver cancer. SHAP analysis identified age, acute kidney injury, sepsis, non-elective admission, and payer status as major contributors to mortality prediction. Conclusions: In this national inpatient analysis of hospitalizations with cancer diagnoses, secondary malignancies were highly prevalent, while acute leukemia ICD-10 diagnoses were most commonly observed among patients who experienced in-hospital death. Machine learning with SHAP-based interpretability identifies clinically relevant inpatient mortality patterns that may assist early risk stratification during hospitalization.
MajesTEC-9: A phase 3 randomized study of teclistamab monotherapy vs investigator’s choice of pomalidomide, bortezomib, and dexamethasone or carfilzomib and dexamethasone (PVd/Kd) in patients (pts) with relapsed refractory multiple myeloma (RRMM).
7507 Background: Triplet and quadruplet regimens have greatly improved outcomes in newly diagnosed multiple myeloma, yet a high unmet need remains for more effective, broadly accessible treatments for anti-CD38- and lenalidomide- (Len) exposed/refractory RRMM. Teclistamab (Tec) is the most widely used BCMA×CD3 BsAb in heavily pretreated RRMM, with improved outcomes in earlier lines of therapy (LOTs). In MajesTEC-3, the synergistic Tec-daratumumab combination significantly improved progression-free and overall survival (PFS/OS) as early as first relapse. MajesTEC-9 (NCT05572515) is the first phase 3 study of Tec monotherapy in pts with 1–3 prior LOTs. Methods: Pts with RRMM and 1–3 prior LOTs including anti-CD38 and Len were randomized 1:1 to receive 28-day cycles (C) of Tec (C1–2: 2 step-up doses then 1.5 mg/kg QW; C3–6: 3 mg/kg Q2W or Q4W depending on response; C7+: 3 mg/kg Q4W) or investigator’s choice of PVd/Kd (21/28-day C). Primary endpoint was PFS by IRC; secondary endpoints included OS, complete response or better (≥CR), and safety. Results: 593 pts were randomized (Tec, n=296; PVd/Kd, n=297). Median (range) age: 70 (34–86) yrs; number of prior LOTs: 2 (1–3); Len refractory: 80%; anti-CD38 refractory: 85%; refractory to last LOT: 92%. With 17.3-mo median follow-up, Tec significantly improved PFS vs PVd/Kd (HR, 0.29; 95% CI, 0.23–0.38; P <0.0001); mPFS: NR vs 8.2 mo; 18-mo PFS rate: 69.8% vs 26.9%. OS was significantly improved with Tec vs PVd/Kd (HR, 0.60; 95% CI, 0.43–0.83; P =0.0020). Of 174 PVd/Kd pts who received subsequent treatment, 68.4% received BsAb or CAR-T. PFS favored Tec across all prespecified subgroups, including anti-CD38-refractory pts and Len-refractory pts. ≥CR rate was significantly higher with Tec vs PVd/Kd (65.9% vs 16.8%; OR, 10.42; 95% CI, 6.89–15.76; P <0.0001). At data cutoff, 65.3% of pts remained on Tec vs 24.0% on PVd/Kd (safety set: Tec, n=291; PVd/Kd, n=283). TEAEs were similar for Tec and PVd/Kd (99.7% vs 97.9%). Grade (Gr) 3/4 TEAEs (84.9% vs 76.3%) and Gr 5 TEAEs (6.5% vs 3.5%) were higher with Tec vs PVd/Kd, noting the median treatment duration was nearly doubled with Tec (13.1 vs 7.0 mo). Treatment discontinuation due to TEAEs was lower with Tec vs PVd/Kd (10.7% vs 13.1%). Gr 3/4 infections were higher with Tec vs PVd/Kd (41.6% vs 29.0%). Any-onset Gr ≥3 infections decreased over time. CRS occurred in 66.0% of pts with Tec (Gr 1/2: 48.8%/16.5%) and ICANS in 4.1% (Gr 1/2: 2.4%/1.4%). Conclusions: Tec monotherapy significantly improved PFS/OS vs standard of care (SoC) in this high unmet need population. CRS and infections were managed with established protocols. Results of phase 3 studies support Tec-based regimens as a SoC as early as second line, across all treatment settings and regardless of prior anti-CD38/Len exposure. Clinical trial information: NCT05572515 .
Local delivery of SN-38 using a biodegradable implantable membrane for treatment of glioblastoma.
e14100 Background: Glioblastoma (GBM) remains a lethal disease with high rates of local recurrence despite maximal surgical resection and standard chemoradiotherapy. Systemic delivery of cytotoxic agents is limited by the blood–brain barrier and dose-limiting toxicity. CEB-01 is a biodegradable nanofiber membrane designed for local delivery of SN-38, a potent topoisomerase I inhibitor, directly to the surgical cavity. This study evaluated the preclinical efficacy, potency, and local tolerability of CEB-01 in GBM models. Methods: The antitumor activity of SN-38 was assessed in glioblastoma cell lines, including temozolomide-resistant and MGMT-positive and -negative models (GBM cell line panel Sanger Institute). In vivo efficacy and local tolerability of CEB-01 were evaluated in rodent intracranial models following surgical implantation in the brain. Tumor growth, histopathology, and neurological tolerance were assessed. External preclinical data were used to contextualize local SN-38 dose ranges and exposure relative to previously reported delivery strategies. Results: SN-38 demonstrated markedly higher in vitro potency compared with temozolomide and carmustine, with IC50 values approximately 1,000-fold lower across GBM cell lines and preserved activity in MGMT-positive and temozolomide-resistant models. In vivo implantation of CEB-01 showed good local tolerability, with no significant behavioral toxicity, weight loss, or fibrosis observed during long-term follow-up. Histological evaluation revealed only localized cortical changes related to surgery and implantation, comparable to incision controls. Dose-scaling analyses indicated that local SN-38 delivery in the low milligram range is pharmacologically plausible while minimizing systemic exposure, supporting further translational development. Conclusions: These preclinical data support CEB-01 as a promising local drug-delivery strategy for glioblastoma, combining high antitumor potency of SN-38 with acceptable local tolerability. Localized sustained delivery of SN-38 may overcome key limitations of systemic chemotherapy in GBM and warrants further translational and clinical investigation.
Efbemalenograstim alfa for prophylaxis in non–small cell lung cancer patients patients with chemotherapy-induced febrile neutropenia: A randomized, multicenter, exploratory clinical trial.
8657 Background: Neutropenia is a common dose-limiting chemotherapy toxicity, with its severity and duration impairing therapeutic efficacy. In NSCLC management, chemotherapy regimens with high risk febrile neutropenia (FN) are relatively uncommon. However, real-world data show suboptimal prophylactic granulocyte colony-stimulating factor (G-CSF) use in intermediate risk FN patients with additional risk factors. This study evaluated efbemalenograstim alfa's efficacy and safety in maintaining absolute neutrophil count (ANC) during chemotherapy cycles in such NSCLC patients. Methods: This was a randomized, multicenter, exploratory clinical trial (NCT06143735) enrolling NSCLC patients receiving intermediate risk FN chemotherapy with additional risk factors. Eligible patients were randomized 2:1 into two groups: primary prophylaxis group (Group A) received subcutaneous efbemalenograstim alfa (20 mg) 48±4 h post-each chemo cycle starting from Cycle 1; secondary prophylaxis group (Group B) initiated treatment only upon grade ≥3 neutropenia (per CTCAE v5.0) in the prior cycle. All patients completed at least four cycles of platinum-based chemotherapy. The primary endpoint was the incidence of grade ≥3 neutropenia in Cycle 1 between the two strategies. Secondary endpoints included FN incidence, grade ≥3 neutropenia in subsequent cycles, and the occurrence of adverse events (AEs) or serious adverse events (SAEs). Results: As of Dec 31, 2025, 99 patients were enrolled; 93 (median age: 69 years; 61.3% with ECOG 0–1; 90.3% male) received ≥1 platinum-based chemo cycle; 83 received ≥1 efbemalenograstim alfa dose. Specifically, 61 patients were assigned to Group A and 32 to Group B (n = 93). In the first treatment cycle, the incidence of grade ≥3 neutropenia was 19.7% (12/61) in Group A vs. 50.0% (16/32) in Group B, representing a relative risk reduction (RRR) of 60.6% (absolute risk reduction: 30.3%; p = 0.002). Across all treatment cycles, the cycle-specific cumulative incidence of grade ≥3 neutropenia was 16.2% (32/198) in Group A vs. 25.6% (32/125) in Group B (p = 0.0382). FN occurred in 2 patients in both groups. Regarding safety, the most common adverse events (AEs) were hematological toxicities in both groups, with a numerically lower incidence of anemia, leukopenia, neutropenia, and thrombocytopenia in Group A than in Group B. Grade ≥3 AEs (52.5% vs. 43.8%, p = 0.514) and efbemalenograstim alfa-related treatment-emergent adverse events (TRAEs, 31.1% vs. 22.7%, p = 0.587) did not differ significantly. Pain-related TRAEs (mainly ankle pain and neuropathic pain) were reported in only 6 patients (7.2%) overall, with most being grade 1–2 per CTCAE v5.0 criteria. Conclusions: Efbemalenograstim alfa effectively reduces the incidence of chemotherapy-induced neutropenia and presents a favorable safety profile. Clinical trial information: NCT06143735 .
Identifying drivers of early readmission in cancer patients to improve care and outcomes.
e23295 Background: Cancer patients are at increased risk for hospital readmission, which adversely affects outcomes and burdens the healthcare system. In a prior quality improvement study, we identified factors linked to potentially avoidable 30-day readmissions among adult cancer patients, including symptomatic presentation at admission, failure to thrive, absence of a post-discharge caregiver, and lack of outpatient follow-up. The current study compares cancer patients with unplanned 30-day readmissions to those without to further identify opportunities for prevention and improved care. Methods: A retrospective chart review was performed at Stony Brook University Hospital between June 30, 2021 and July 31, 2022. Adult patients with active malignancy were included. Patients followed with non-SBUH cancer specialists or discharged after July 1, 2022 were excluded. Demographics, cancer characteristics, hospital course, and post-discharge factors were analyzed. Independent predictors of readmission were evaluated using a multivariable Cox proportional hazards model. Results: Of 468 patients identified, 99 were readmitted (21.2%), 85 were not readmitted (18.2%), and 284 were excluded. Patients with hematologic malignancies were readmitted more often than those with solid tumors (73% vs 23%, p<0.001; HR = 2.26). Patients with designated health care proxies (HCP) had a higher readmission rate (63% vs 31%, p<0.001). There was no significant difference in the percentage of patients reporting pain on initial admission between readmitted (37.4%) and non-readmitted (27%) patients (p=0.183). However, patients prescribed a new pain medication at discharge had a lower readmission rate (47% vs 69%, p=0.003; HR = 0.58). Readmitted patients had a longer length of stay (10.8 days vs. 6.2 days, p=0.0013) with each additional day of hospitalization associated with a 2.4% increase in the HR for readmission. Risk of readmission predicted by the electronic medical record’s (Cerner) proprietary risk model demonstrated a positive predictive value of 77% and negative predictive value of 50%. Among patients flagged by the EMR to be at high risk, only 33% were readmitted (p<0.001). Code status and caregiver availability were not associated with readmission. Conclusions: Identification of factors associated with readmission can identify patients at risk for deterioration and guide targeted interventions to improve outcomes. Improved pain management may be a readily available option to reduce hospital readmission risk for cancer patients. Current EMR readmission risk assessment models are not validated for cancer patients and fail to identify many patients at higher risk for readmission. A risk assessment tool validated in cancer patients and their unique characteristics is needed. Identifying both risk and protective factors for readmission would optimize care for cancer patients and reduce healthcare costs.
Association of GLP-1 receptor agonist exposure with mortality, infectious, and hematologic outcomes in chronic lymphocytic leukemia: A real-world retrospective TriNetX study.
e18591 Background: Emerging studies show that Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) might have antineoplastic and immune-mediated effects beyond glycemic control by inhibiting PI3K/AKT/mTOR and ERK/MAPK pathways, modulating inflammatory cytokines, and altering the tumor microenvironment, which are key drivers of chronic lymphocytic leukemia (CLL) progression and immune dysfunction. GLP-1 RAs also improve insulin sensitivity and decrease adipose-mediated systemic inflammation, which relate to infectious risk and mortality in CLL, but their impact on clinical outcomes and survival in patients with CLL remains poorly defined. Methods: We conducted a retrospective cohort study using the TriNetX Network to evaluate outcomes associated with GLP-1 RA use in adults with CLL and type 2 diabetes mellitus. GLP-1-exposed and non-exposed cohorts were compared for infectious complications, hematologic events like cytopenias, ICU admissions, adverse events, cardiovascular (CV), and all-cause mortality, while accounting for exposure to contemporary CLL therapies to assess potential confounding and analyzing time-to-event outcomes using Kaplan-Meier and Cox models after excluding pre-existing events. Results: After propensity score matching, follow-up was well balanced between cohorts, with comparable mean short-term (~326 days) and long-term (758-824 days) durations. GLP-1 use was associated with significantly lower risks of sepsis (3.1% vs 6.1%; HR 0.51, 95% CI 0.37-0.69), pneumonia (4.3% vs 8.0%; HR 0.53, 95% CI 0.40-0.70), thrombocytopenia (3.7% vs 5.9%; HR 0.63, 95% CI 0.46-0.86), and all-cause mortality (2.5% vs 8.1%; HR 0.30, 95% CI 0.22-0.41) when compared to the non-GLP1 cohort, with early and sustained separation of Kaplan-Meier curves. CV mortality and ICU admission rates were significantly lower in the GLP-1 cohort with HRs of 0.69 (95% CI 0.57-0.83) and 0.61 (95% CI 0.49–0.77), respectively. Autoimmune Hemolytic Anemia (AIHA) was rare and similar between groups, with no significant difference in time-to-event. Within the GLP-1 CLL cohort, absolute rates of acute pancreatitis (1.2%) and hypoglycemia (3.0%) were low. Exposure to contemporary CLL therapies was uncommon, including BTK inhibitors (3.3%), BCL2 inhibitors (4.4%), anti-CD20 antibodies (rituximab 3.1%, obinutuzumab 2.8%), and ibrutinib specifically (1.9%), with high treatment-free survival rates across CLL therapy exposure (80-95%), indicating that observed outcomes were unlikely due to treatment-related toxicity or differential therapy exposure. Conclusions: GLP-1 receptor agonists have biologically plausible antineoplastic, metabolic, and immunomodulatory effects that may improve inflammation-related infection and survival outcomes in diabetic patients with CLL, warranting further research.
Real-world data on clinical profile of immune-related adverse events (irAEs) associated with immune checkpoint inhibitors (ICIs) in oncology.
11151 Background: Immune checkpoint inhibitors (ICIs) are widely used in cancer, and are associated with immune-related adverse events (irAEs) in many organs. Real-world data on the spectrum, timing, severity, and outcomes of irAEs is limited. In this study, we aimed to gather real world data on the immune toxicity of ICIs. Methods: We conducted a retrospective cohort study on adult oncology patients on ICI therapy between Jan 2018-Dec 2024. IrAEs were categorized by organ system, time to onset (early < 12 weeks, intermediate 12-24 weeks, and late > 24 weeks), and severity (standard toxicity criteria CTCAEv6.0). Management strategies, therapy discontinuations, hospitalizations, and survival were assessed. Results: We screened 652 patients who received ICIs across different cancer types -169 (26%) were identified with irAEs. Sixty-four percent patients had received ICI alone, and the remainder had received ICIs in combination with other agents. The most common organ affected was thyroid (43%), followed by skin (28.4%), and lung (11.8%), more than one organ was affected in 22.5% patients. IrAEs occurred across a broad time frame following treatment- cumulative incidence before 12 weeks was 41.4%, between 12-24 weeks was 44.4%, and beyond 24 weeks was 91.3%. Musculoskeletal, hepatic, pulmonary, gastrointestinal, dermatologic, and pancreatic irAEs generally presented early, with median onset around 10–14 weeks. Renal, ophthalmologic, infusional, and dermatologic irAEs generally occurred across early, and intermediate time frames, while neurological irAEs tended to occur late. Endocrine irAEs, had a median onset of 20.1 weeks, and frequently occurred beyond 24 weeks. Hepatitis, pneumonitis, colitis, CNS and pancreatic toxicity were among the more severe irAEs, with a median CTCAE grade of 3, while renal, endocrine, ophthalmologic, infusional, and dermatologic irAEs were generally of lower severity (grade 1 or 2). IrAE grade > 2 was associated with increased risk of death (HR 2.7, 95% CI 1.1-6.5). Management of irAEs comprised supportive care alone (62%), systemic corticosteroids (31%), or observation alone (8%). Patients with higher grades of irAEs were more likely to receive steroids. Hospitalization was required in 29%, and ICIs were permanently discontinued in 25%. Complete resolution of irAEs was seen in 40%, 5% died attributable to the irAE, and the remainder had partial resolution of irAE with need for continued replacement therapy (as with endocrine irAEs). Conclusions: In our real-world practice, oncology patients receiving ICIs had irAEs across a broad range of organ systems, with heterogeneous timing and severity. Several clinically meaningful toxicities presented late in the treatment course, underscoring the importance of long-term monitoring. Occurrence of grade 2 or higher iRAEs was associated with higher mortality.