Cancer Network Research Articles

Cytoreductive Surgery Plus Hyperthermic Intraperitoneal Chemotherapy (HIPEC) in Patients with Advanced Ovarian Cancer: A 2-year Survival Analysis Study

Background: During the last few years, Cytoreductive Surgery plus Hyperthermic Intraperitoneal Chemotherapy (HIPEC) has entered the national comprehensive cancer network guidelines as a new protocol for improving patients’ outcomes. However, there is no consensus on its long-term efficiency, and it still is under debate. Objectives: This study aims to evaluate the effectiveness of Cytoreductive Surgery Plus hyperthermic intraperitoneal chemotherapy in patients with advanced ovarian cancer in Iran. Method: Thirty patients with Stage IIIc and IV advanced ovarian cancer underwent cytoreductive surgery plus hyperthermic intraperitoneal chemotherapy at Jam Hospital with a fixed surgical team in Tehran, Iran, from 2019 to 2021. Fourteen patients were new cases, and sixteen of them were recurrent cases. At the end of cytoreductive surgery, by using a hyperthermic intraperitoneal chemotherapy device, Cisplatin was circulated in the peritoneal cavity for 90 minutes at a dose of 80-100 mg/ m2 at 43°C. Results: Among 30 patients with 54.97±10.74 years of mean age, the mean overall survival was 564.967 days, and 2-year survival rates were 66.7%. According to Fisher's exact test, there was a statistically significant relationship between disease-free after surgery and mortality rate (p=0.00). However, there was no statistically significant relationship between recurrence after surgery and mortality rate (p=0.093). Conclusion: Based on these findings, cytoreductive surgery plus hyperthermic intraperitoneal chemotherapy had a survival rate of 66.7% within two years in advanced ovarian cancer patients. However, to achieve better results, careful selection of patients and complete cytoreductive surgery should be performed.

Factors Associated with Head and Neck Cancer Postoperative Radiotherapy Delays: A Systematic Review and Meta-analysis.

Initiating postoperative radiotherapy (PORT) within 6 weeks of surgery for head and neck squamous cell carcinoma (HNSCC) is included in the National Comprehensive Cancer Network Clincal Practice Guidelines and is a Commission on Cancer quality metric. Factors associated with delays in starting PORT have not been systematically described nor synthesized. PubMed, Scopus, and CINAHL. We included studies describing demographic characteristics, clinical factors, or social determinants of health associated with PORT delay (>6 weeks) in patients with HNSCC treated in the United States after 2003. Meta-analysis of odds ratios (ORs) was performed on nonoverlapping datasets. Of 716 unique abstracts reviewed, 21 studies were included in the systematic review and 15 in the meta-analysis. Study sample size ranged from 19 to 60,776 patients. In the meta-analysis, factors associated with PORT delay included black race (OR, 1.46, 95% confidence interval [CI]: 1.28-1.67), Hispanic ethnicity (OR, 1.37, 95% CI, 1.17-1.60), Medicaid or no health insurance (OR, 2.01, 95% CI, 1.90-2.13), lower income (OR, 1.38, 95% CI, 1.20-1.59), postoperative admission >7 days (OR, 2.92, 95% CI, 2.31-3.67), and 30-day hospital readmission (OR, 1.37, 95% CI, 1.29-1.47). Patients at greatest risk for a delay in initiating guideline-adherent PORT include those who are from minoritized communities, of lower socioeconomic status, and experience postoperative challenges. These findings provide the foundational evidence needed to deliver targeted interventions to enhance equity and quality in HNSCC care delivery.

Real-world use of a CSF circulating tumor cell assay in the diagnosis and management of leptomeningeal metastasis.

2029 Background: The diagnosis of leptomeningeal metastasis (LM) can be challenging due to variability in clinical symptoms, radiographic, and cerebrospinal fluid (CSF) findings. The National Comprehensive Cancer Network (NCCN) guidelines recommend assessment of circulating tumor DNA (ctDNA) to increase sensitivity of tumor cell detection and assess the response to treatment. We hypothesize that the real-world use of an assay for the direct detection and enumeration of circulating tumor cells (CTCs) in the CSF is clinically useful for accurate diagnosis of LM, particularly in an early stage of disease. Methods: We report a series of 55 patients treated at our institution with a cancer diagnosis and either neurological symptoms or radiographic findings that suggested LM. All patients underwent a lumbar puncture, assessment for CSF CTCs, and conventional hospital-based cytology testing between 6/1/2020 and 8/1/2023. Survival data cutoff was 12/31/2023. Patient outcomes are reported. Results: 55 patients were tested, and 30 patients were found to be positive for CTCs and diagnosed with LM. Only 10 patients were concurrently positive for CTCs and cytology at our institution, whereas 20 patients with positive CTCs had a negative or ambiguous cytology. No patients with a negative CTC result (25 patients) were subsequently diagnosed with LM, whereas 4 patients with a positive CTC result remained without progressive neurological symptoms. 5 patients diagnosed with LM were alive at the time of data cutoff. The histologic breakdown for LM was: breast (11), non-small cell lung cancer (NSCLC) (16), gastrointestinal (GI) (3). Median overall survival (OS) was longer in patients with a positive CTC and negative cytology result compared to patients with concurrent positive CTC and cytology (172 vs. 63 days, p = 0.06). Median OS was also longer in breast cancer LM compared to NSCLC LM (235 vs. 63 days, p = 0.008). Most patients diagnosed with LM received therapies including Ommaya reservoir placement, intrathecal chemotherapy, and/or radiation. Conclusions: The use of a CSF circulating tumor cell assay aided the diagnosis of LM and led to timely initiation of treatment. A negative result of the CTC assay was associated with ruling out LM. Diagnosis of LM with the CTC assay and negative cytology was associated with longer survival, possibly due to earlier treatment initiation, and not solely a lead time bias. Breast cancer LM patients had improved survival compared to NSCLC and GI cancer patients, possibly due to more effective treatment options.

Impact of American Society of Breast Surgeons (ASBS) guidelines on genetic testing (GT) rates among patients with breast cancer (BC).

e22537 Background: Identifying patients with pathogenic germline genetic variants can inform clinical decision-making and risk assessment. Previous studies have suggested low rates of GT in BC patients overall and disproportionately lower rates of GT among black patients. This may be explained in part by clinicians failing to understand the current National Comprehensive Cancer Network (NCCN) GT guidelines, which are complex. In February 2019, the ASBS released updated guidelines recommending GT for all patients with a personal history of BC. The objective of this study was to evaluate the impact of the ASBS guideline change on GT utilization for all BC patients, and for black BC patients in particular. Methods: This retrospective cohort study included 4386 BC patients who were diagnosed and/or treated at the Siteman Cancer Center (SCC) in St. Louis, MO from January 1, 2017 to December 31, 2020. The pre-guideline period included those patients diagnosed between January 1, 2017 and July 25th, 2019 and the post-guideline period included those diagnosed between July 26, 2019 and December 31, 2020. Within the study period, Myriad Genetics was used almost exclusively for GT by breast oncologists and surgeons at SCC. Myriad data was queried to identify all BC patients at SCC with GT. Data including self-reported race/ethnicity, BC subtype (triple negative breast cancer (TNBC) vs other), insurance type (public vs private), and family history of BC (yes vs no) was collected from the electronic medical record (EMR) independently by authors. Multivariate logistic regression analyses and difference in difference (DiD) analyses were performed to compare GT rates pre- and post-guideline change. Results: There was no significant difference in the rate of GT among BC patients pre- vs post-guideline change (17% vs 14.8%, p = 0.05). Similarly, there was no significant difference in GT rates pre- vs post-guideline change for non-Hispanic black (NHB) (14.6% vs 13.1%) and non-Hispanic white (NHW) subgroups (17.8% vs 15.2%) using DiD analysis (p = 0.96). Patients with a family history of BC (odds ratio (OR) 1.41, 95% confidence interval (CI) 1.17-1.69, p < 0.01) and TNBC (OR 1.47, 95% CI 1.15-1.86, p < 0.01) were more likely to receive GT. Patients aged greater than 50 years (OR 0.44, 95% CI 0.37-0.54, p < 0.001) and those with public insurance were less likely to receive GT (OR 0.82, 95% CI 0.67-1.02, p = 0.07). DiD analyses identified no significant differences pre- and post-guideline change for these subgroups. Conclusions: Despite the recent broadening of GT guidelines by ASBS, GT rates remain low among patients with BC. We found no significant impact of the guideline change on the uptake of GT among our population of BC patients overall or by race/ethnicity, family history, BC subtype, age, or insurance status. Multidisciplinary efforts are needed to improve guideline implementation and uptake of GT in BC patients.

Reporting of immune-related adverse events (irAEs) in US Food and Drug Administration (FDA) approvals of immune checkpoint inhibitors.

e14693 Background: Evaluating immune checkpoint inhibitor (ICI) toxicity is complex given the interplay between the host immune system and tumor microenvironment leading to unpredictable timing and severity of irAEs. Monitoring of irAEs is critical as they can cause significant impact on morbidity and quality of life. Thus, we evaluated the consistency and pattern of reporting of irAEs in trials leading to FDA drug approvals. Methods: Using the FDA website, we identified 73 primary articles for ICI FDA approvals over a 10-year period (2011-2021). Data were collected independently by all authors as categorical variables for reporting of any grade and grade 3 or higher toxicity in the trial arm that led to FDA approval for 29 irAEs classified by the National Comprehensive Cancer Network Version 1.2024 guidelines. Our analyses were based on a binary classification of reported versus not reported, including the manuscript text, tables, and supplement, and reporting trends were analyzed based on the study year, phase, primary tumor, and monotherapy versus combination ICI therapy. Results: The majority of approvals were associated with Phase II (N=22) or Phase III (N=39) studies. Approvals included 14 primary tumors and 2 tumor agnostic approvals with the most common tumors including lung (N=22) and skin (N=14). irAE reporting ranged between 0% and 100% with a mean of 40% and a standard deviation of 35. Across the 29 irAEs, 45% were categorized as severely underreported with less than 25% reporting in trials, 17% were classified as underreported in 26-50% of trials, 17% were moderately reported in 51-75% of trials, and 21% were well reported in at least 75% of trials. For example, fatigue, diarrhea, and hypothyroidism were well reported at 100%, 99%, and 93%, respectively, while myocarditis, aseptic meningitis, and giant cell arteritis were severely underreported at 18%, 4%, and 1%, respectively. Additionally, certain systems were more reported than others, including gastrointestinal (100%) and endocrine (97%), while other systems were inconsistently reported such as cardiology (15%) and neurology (48%) (see Table). Conclusions: Our study identified significant inconsistencies in irAE reporting in clinical trials that led to ICI FDA approvals. In particular, rarer irAEs were reported infrequently while more common irAEs were well reported. Unified efforts are needed to standardize the irAE reporting in the primary literature to ensure reliable dissemination of information for timely recognition and treatment in clinical practice. [Table: see text]

Characterization of social determinants of health in patients considered for participation in treatment clinical trials at a community cancer center serving the Great Plains.

e13704 Background: Avera Cancer Institute, a community-based cancer network comprised of six rural cancer centers in SD and MN, provides care to a five-state region in the upper mid-west covering 72,000 square miles and offers several clinical trials to patients to offer access to advanced treatment for malignancies. Malignances are known to disproportionally affect patients that are older, rural, minorities, and have lower socioeconomic status, which are all less likely to enroll into clinical trials. Methods: Patients evaluated for participation in clinical trials from 2018-2022 were identified in an institutional database. Trial type was designated as either a treatment or nontreatment trial. Status of trial involvement was designated between enrolled or screened. Their demographic information was obtained from the electronic health record (EHR). Demographic information obtained from the EHR includes race, ethnicity, language spoken, gender, employment status, insurance type, and if they live in an urban or rural area as defined by the US Census Bureau. Clinical trial portfolio during this time includes trials for breast, gynecological, gastrointestinal, lung, and hematologic malignancies. Results: 4,397 patients were considered for participation in a clinical trial at our institution in the studied timeframe. Of these patients, 1887 were screened for treatment trials and of that group, 773 were enrolled. 35% of participants considered for clinical trials were from urban areas while 65% considered were rural. 48% of the urban population ultimately were enrolled on a clinical trial while 44% of rural population participated. Participants who lived in an urban area 25-50 miles from our study site had an enrollment rate of 49% compared to rural enrollment of 33%, from 50-100 miles rates of 53% and 37%, from 100-250 miles rates of 29% and 41%, and from greater than 250 miles rates of 44% and 33%, respectively. 99.5% of the population was non-Hispanic, 97% were White, and 99% were English speaking. 74% of the total population were female and 26% were male. Females had an enrollment rate of 51% while males had an enrollment of 29%. Patients with no insurance enrolled in trial 62% compared to 36% of those with commercial insurance The average national Area Deprivation Index percentile for screened patients was 57 vs 59 for enrolled patients. Conclusions: This retrospective study demonstrates a greater likelihood for females, urban patients, and white, English-speaking patients to enroll into clinical trials. Our data also demonstrate a preference for uninsured patients to enroll into clinical trials, possibly due to the financial benefits they may provide. Further directions include comparing our patients considered for clinical trials to all patients treated at our institution to identify potential disparities in care.

Evaluating the utility of routine survivorship exams in breast cancer: Modality of detection of recurrence in a prospective cohort study.

1629 Background: National Comprehensive Cancer Network surveillance guidelines for invasive breast cancer recommend a history and physical exam every 3-12 months for the first 5 years, followed by annually, and mammography every 12 months. However, frequent in-office visits create time toxicity for breast cancer survivors, may have low utility in recurrence detection, and can create false reassurance. This study evaluates the modality of breast cancer recurrence in an integrated health system with access to primary and specialty care. Methods: We investigated mode of detection of recurrent breast cancer in the Pathways Study, a prospective cohort study of adult women diagnosed with and treated for invasive breast cancer in Kaiser Permanente Northern California, with enrollment from 2006 to 2013, and follow-up through December 2021. In this analysis, we included 4,431 women diagnosed at Stages I-III. Cohort members were followed for recurrences, deaths, and other outcomes. Of primary interest was modality of detection of recurrence, categorized as patient-driven symptoms, surveillance mammogram, diagnostic testing (i.e., other surveillance imaging for recurrence such as PET CT or MRI), routine clinical breast exam, or incidental finding. Results: A total of 617 (13.9%) women were identified as having breast cancer recurrence. Patients with recurrence shared similar demographic characteristics to those without recurrence. The recurrence population had median age of 59.2 years, BMI of 28.2 kg/m2, and racial/ethnic composition of 63% white,13% Hispanic,12% Asian, and 10% black. The median time from diagnosis to recurrence was 4 years. Seventy percent of recurrences were found to have distant metastatic disease, of which only 6.7% had concurrent local nodal and/or breast involvement. For those without metastases, 17.7% had recurrence localized in the ipsilateral breast, 2.8% in both the breast and locoregional nodes, and 9.3% only in locoregional nodes. Most recurrences were detected by patient driven symptoms (69.4%). Among the total cases, 8.1% of patients were diagnosed by routine mammogram, 10.9% with diagnostic testing and 4.5% by incidental finding. Recurrences in 6.8% of patients were detected by routine physician exam. Overall survival for those with recurrence was 31.1% and most deaths were attributed to breast cancer (89%). Conclusions: Our study, the largest investigation of breast cancer recurrence modality of detection in a US community oncology setting with comprehensive care, revealed the majority of recurrences were identified through patient self-detection with only 6.8% detected by physician exam. Most recurrences were distant, underscoring the need for alternative screening modalities to detect metastatic recurrences earlier, especially in the context of emerging treatments and improved survival rates for oligometastatic disease.

Changes in non-small cell lung cancer (NSCLC) next-generation sequencing (NGS) rates after electronic health record (EHR) integration using large-scale, multi-institutional, real-world data.

e13619 Background: NSCLC treatment guidelines recommend comprehensive NGS testing to detect genomic alterations and gene expression in key biomarkers that are relevant in diagnosis, therapy selection, prognosis, and clinical trials. While 70% of patients with NSCLC will have biomarker alterations related to therapeutic options, a large minority of patients still do not receive biomarker testing. Clinical workflow challenges outside the EHR ecosystem, like paper requisitions and faxed results, have been cited as barriers to adoption. To reduce sequencing barriers, Tempus deploys NGS through direct result EHR connection. To evaluate whether integrating comprehensive NGS ordering and resulting into the EHR would improve testing rates, we retrospectively analyzed NSCLC NGS orders from 29 clinical networks which represent over 1500 medical oncologists, before and after EHR integration. Methods: Clinical networks that partnered with Tempus to perform an NGS EHR integration were identified. Only networks with at least 6 months of pre- and post-integration unique patient NGS orders were included. De-identified patient information from these sites were obtained from the Tempus laboratory information management system. NSCLC diagnosis was determined through ICD9/10 codes. NGS tests ordered were limited to xT/xR (solid tumor DNA and RNA), xE (solid tumor whole exome), or xF (circulating tumor DNA). Aggregate unique patients with orders were summated on a per site basis. Descriptive statistics were used to characterize the findings. Results: From May 2019 to July 2023, 29 clinical cancer networks (48% academic medical center, 45% regional health system) that underwent a Tempus NGS EHR integration were identified. Of the 29 networks considered, 3 were excluded due to insufficient data. Of the evaluable 26 networks, Epic EHR was used in the majority (24 of 26). In aggregate, a total of 1825 NSCLC patients were sequenced pre-integration as compared to 2796 patients sequenced post-integration, representing a 53% increase. Academic medical centers experienced a higher increase versus regional health systems (65% v 48%). When examined on a per network basis, the median and mean increase in NSCLC patients sequenced were 49% and 118%, respectively. Conclusions: In this retrospective analysis, Tempus EHR integrations increased ordering of comprehensive NSCLC NGS by 53%. Despite study limitations such as the incomplete resolution into internal alternate sequencing workflows, this large-scale analysis provides further evidence that EHR-based workflow improvements reduce the barrier to appropriate sequencing.

National Comprehensive Cancer Network guideline compliance in an oncology value-based care program serving Medicare members.

e23095 Background: Value-based care (VBC) focuses on improving patient health outcomes while reducing the overall cost of healthcare. National Comprehensive Cancer Network (NCCN) cancer treatment guidelines standardize and improve quality of care. While studies have shown that compliance to NCCN guidelines improves survival, little information exists regarding the impact of compliance in VBC arrangements. This study aimed to evaluate compliance to NCCN regimen guidelines in Medicare members being treated for cancer under a VBC arrangement. Methods: This is a cohort study of individuals receiving care, in a single state, enrolled in a VBC contract within a large national insurer from February 2023 to October 2023. Medicare members with bladder, colon, lung, melanoma, breast, hepatocellular, rectal, pancreatic, multiple myeloma, gastric, ovarian, esophageal, lymphoma, kidney, leukemia, head and neck, biliary tract and uterine cancer with at least one oncology claim within the contract year were included. NCCN compliance was a quality measure of the VBC. NCCN regimen compliance was identified from pharmacy and medical claims and defined as concordant if the entire prescribed treatment regimen matched an NCCN regimen (Level 1 and 2a); members not receiving an NCCN recommended regimen were deemed to be non-concordant. Results: This study included 561 patients, the mean [SD] age was 75.7 [8.2] years and 286 [51.0%] members identified as male. The total population captured through the automated prior authorization platform yielded 100% NCCN compliance. Conclusions: In this study, individuals receiving chemotherapy participating in a shared savings VBC demonstrated 100% compliance to NCCN guidelines.

Examining the relationship between ECOG performance status and immunotherapy outcomes: Insights from real-world data analysis.

11131 Background: The impact of Eastern Cooperative Oncology Group Performance Score (ECOG) on immunotherapy (IO) outcomes is intricate, as historically clinical trials have primarily enrolled patients with ECOG scores of 0 or 1. The National Comprehensive Cancer Network (NCCN) advises against administering IO to patients with ECOG scores of 2 or higher. However, conflicting findings from research studies and variations in real-world clinical practice complicate this relationship. Our study aims to elucidate the association between ECOG status and the outcomes of immunotherapy. Methods: Data collected from SUNY Upstate Medical University treated cancer patients who underwent immunotherapy and chemotherapy was used for analysis. The Kaplan-Meier method and Cox regression were used to analyze survival probability based on therapy, age, and ECOG. Results: Of 813 patients included in the study, 46.99% (n=382) received immunotherapy. 51.85% (n=421) were females. The mean age was 64.1 years in IO group and 67.7 years in non-IO/chemo group. 52.5% (n=187) of patients who were ≤ 64 years received IO compared to only 42.7% (n=195) of patients who were ≥ 65 years (p=0.005). 48.6% (n=118) of patients with ECOG score 0 received IO, while only 40.5% (n=72) patients with ECOG score of ≥3 received IO (p=0.279). Patients who received IO with ECOG score of ≥3 had higher probability of survival compared to other groups (p<0.0001). While, in non-IO group, patients with ECOG 0 had higher probability of survival compared to ECOG 3 (HR 1.773, 1.209 – 2.600). There was no difference in overall survival in patients receiving IO based on age categories (p=0.2627, HR 1.148, 0.901 – 1.462). Conclusions: Our findings suggest that relying solely on ECOG status to determine eligibility for immunotherapy may be overly restrictive. Patients with significant comorbidities could still derive benefits from immunotherapy. Further investigation is warranted to comprehensively assess the influence of ECOG status on immunotherapy outcomes.

Comparative analysis of ChatGPT-4.0’s management of six gastrointestinal cancers according to the NCCN guidelines.

e13654 Background: The integration of Natural Language Processing (NLP) into healthcare holds tremendous promise. ChatGPT-4.0 (OpenAI, San Francisco, CA) is a widely recognized large language model that can comprehend and generate text, answer questions, and perform other language-related tasks. However, pitfalls and errors have been described in its clinical application. We explored the ability of ChatGPT-4.0 (ChatGPT) to guide clinical decision-making in 6 gastrointestinal cancers using the National Comprehensive Cancer Network (NCCN) Clinical Practice Guidelines as a framework. Methods: We reviewed the NCCN Guidelines for Anal Squamous Cell Carcinoma (AN), Small Bowel Adenocarcinoma (SB), Ampullary Adenocarcinoma (AA), Biliary Tract Cancers (BT), Pancreatic Adenocarcinoma (PN), and Gastric Cancer (GA). Up to 2 clinical questions were designed for each decision-making page. Questions were categorized by type ( Wo rkup, Treatment, Surveillance, Diagnostics, or References). ChatGPT was queried in a reproducible fashion. To account for variable prompt engineering of our non-validated assessment tool, up to 3 follow-up questions were allowed. Responses were rated on a Likert scale: 5) Correct; 4) Correct, with missing information requiring clarification; 3) Correct, but unable to complete answer; 2) Partially incorrect; 1) Absolutely incorrect. Subgroup analysis was conducted on Correctness (defined as scores 1-2 vs 3-5) and Accuracy (scores 1-3 vs 4-5). Variance between ChatGPT responses to each cancer was analyzed. Descriptive statistics were used, and significance was tested with binary logistic regression. Results: A total of 270 questions were generated (range per cancer 32-68). The score frequency distribution was: 5) 45.2%; 4) 19.3%; 3) 13.3%; 2) 13.7%; and 1) 8.5%. On subgroup analysis, Correctness was seen in 210 (77.8%) of questions, and Accuracy with 174 (64.4%). The difference in Correctness scores between cancers was not statistically significant, and there was no statistically significant difference in scores by question type. There was a statistically significant difference in the Accuracy of ChatGPT between cancers (Table). Conclusions: ChatGPT was significantly more likely to provide accurate responses to questions regarding GA and PN versus AN or SB. It demonstrates a limited capacity to assist with complex clinical decision-making in 6 gastrointestinal cancers. However, the Accuracy level is below the acceptable threshold for implementation into clinical use. Further analysis of the expanding capabilities of ChatGPT and other NLP-based tools is warranted in this rapidly evolving domain. Future studies would benefit from a validating grading instrument. [Table: see text]

Persistent poverty impacts survival and National Comprehensive Cancer Network guideline-concordant care for patients with melanoma.

1598 Background: Based on guidelines from the National Comprehensive Cancer Network (NCCN), sentinel lymph node biopsy (SLNB) should be offered to patients with high-risk stage IB or stage II melanoma and adjuvant systemic therapy is recommended for patients with stage III disease. Persistent poverty census tracts are defined as areas where at least20% of residents were poor as measured by each of the 1980, 1990, 2000 censuses and 2007 American Community Survey 5-year average. The aim of this study was to describe the association between persistent poverty and adherence to NCCN guidelines and cancer-specific survival in patients with cutaneous melanoma. Methods: Patients diagnosed with melanoma from 2006-2018 were identified in Survival, Epidemiology, and End Results (SEER) Program registries. The analytic cohort was restricted to patients with superficial spreading melanoma, nodular melanoma, malignant melanoma with regression, lentigo malignant melanoma, and acral lentiginous melanoma. Patient demographics such as age, race, and ethnicity and tumor characteristics associated with stage at diagnosis were analyzed using adjusted regression analyses for patients of all melanoma stages. Demographics and tumor characteristics associated with receipt of SLNB were analyzed amongst pathologic stage IB and II cases, and demographics and tumor characteristics associated with receipt of adjuvant systemic therapy were analyzed amongst stage III cases. Kaplan-Meier curves and adjusted accelerated failure time models were used to examine disparities in cancer-specific survival. Results: Of 127,308 total patients, 3.13% (n=3994) lived in census tracts with persistent poverty. In adjusted analyses, persistent poverty was an independent predictor of later pathologic stage at diagnosis (OR 1.32 [95% CI 1.21-1.44] p<0.001). Persistent poverty was also associated with lower likelihood of receiving SLNB according to NCCN guidelines (OR 0.717 [SE 0.072] p<0.01) and with worse cancer-specific survival (HR 1.25 [1.06-1.47], p=0.008) when controlling for age, race, ethnicity, pathologic stage at diagnosis, and histology. In univariable models persistent poverty was associated with increased odds of receiving systemic therapy, but this relationship was not significant on multivariable analysis. Conclusions: Patients from areas with persistent poverty had a later pathologic stage at diagnosis, were less likely to receive a SLNB according to NCCN guidelines, and had worse cancer-specific survival. Efforts to better define and resolve disparities in the treatment and survival of patients with melanoma are warranted.

A targeted panel of actionable high-risk hereditary cancer predisposition genes to identify patients with pathogenic/likely pathogenic variants (PVs) irrespective of meeting established NCCN testing criteria.

10596 Background: Current germline cancer testing guidelines are limited to patients who meet specific personal and family cancer history (PH, FH) criteria. Our objective was to determine if differences in PV rates exist based on meeting National Comprehensive Cancer Network (NCCN) testing criteria in patients who underwent testing for known actionable genes. Methods: 29 genes with actionable NCCN management guidelines were evaluated. Consecutive patients (N=195,615) aged 18 to 100 years had hereditary cancer testing at a commercial laboratory at the ordering clinician’s discretion that included at least the 29 genes between June 2020 and August 2023. Per clinician-reported patient features (e.g. FH, PH, age), patients were categorized as meeting or not meeting current NCCN testing criteria for breast/ovarian/pancreatic (BOP v2.2024), and colorectal cancer (v1.2023). PV rates were calculated for all genes associated with breast, endometrial, ovarian, prostate, pancreatic, and colorectal cancer risk. Results: PVs were reported in 6634/105,194 (6.3%) patients that met the testing criteria and 5134/90,421 (5.7%) that did not. Of patients with a PV, 5134/11,768 (43.6%) did not meet NCCN testing criteria. When comparing patients with PVs based on NCCN testing criteria, patients were similar in age (median [Q1, Q3], 43.6 [35.0, 56.1] v 42.5 [34.7, 56.6] years) and sex (female 91.5% v 91.3%). While PH and FH were comparable, a PH of cancer (31.5% v 36.1%, OR=0.81, 95% CI 0.75-0.88, p<0.001) and/or a FH of cancer (83.9% v 97.1%, OR=0.16, 95% CI [0.13-0.18], p<0.001) were slightly lower in patients who did not meet criteria compared to those who met testing criteria. On evaluating results in genes associated with individual cancer types, similar PV rates between the groups were observed (Table). Among all patients with PVs, the most common genes were CHEK2, BRCA1, BRCA2, and ATM. Most patients had only 1 PV reported (met: 96.5%, not met: 97.5%). Conclusions: In this cohort of patients who underwent germline testing, approximately 40% of patients with PVs in this study did not meet existing NCCN testing criteria. PV rates were similar between patients that met criteria and those that did not. Broadening existing testing criteria may reduce the number of patients who would otherwise be missed, thereby allowing for treatment decision-making and cascade testing of family members. [Table: see text]

Racial and elderly disparities in receipt of guideline-concordant care in stage I-III breast cancer.

e13688 Background: Elderly and non-Hispanic Black (NHB) patients have worse breast cancer specific survival outcomes. We evaluated whether guideline concordant care (GCC) and all-cause mortality in stage I-III breast cancer differed by race among patients ≥ 65 years old. Methods: We used the National Cancer Database (NCDB), to identify a cohort of 25174 NHB and 233357 non-Hispanic White patients (NHW), ≥65 years old, diagnosed with stage I-III breast cancer between 2010 and 2019. We used the National Comprehensive Cancer Network guidelines to define GCC, including surgery, radiation and systemic therapy. Specific systemic therapies and cause of death were unavailable. We used logistic regression models to assess the association between race and GCC. We used Cox proportional hazards models to estimate associations between race and all-cause mortality after adjusting for GCC. Both models adjusted for disease, demographics, income and educational attainment. We assessed the odds of not receiving timely GCC within 30, 60 and 90 days from diagnosis for NHB vs. NHW patients using logistic regression. Results: Overall, 15% of patients did not receive GCC. Being NHB compared to NHW was associated with increased odds of not receiving GCC in univariate (OR=1.24 95%CI: 1.20-1.28) and multivariate analysis (aOR=1.13, 95%CI: 1.08-1.17). Being NHB was associated with 26% increased risk of all-cause mortality in the univariate analysis (HR=1.26, p<0.001), which decreased to 4% increased risk, after adjusting for GCC, age, clinical and sociodemographic factors (aHR=1.04, p=0.027). Non-receipt of GCC (aHR=1.75, p<0.001) and lower educational attainment (aHR 1.18, p <0.001) were independently associated with higher risk of all-cause mortality. Higher median income (aHR 0.91, p<0.001) was independently associated with a decreased risk of mortality. Being NHW compared to NHB was associated with increased odds of initiating treatment within 30 (OR=1.65, 95% CI 1.6-1.69), 60 (OR=2.11, 95% CI 2.04-2.18), and 90 days of breast cancer diagnosis (OR=2.39, 95% CI 2.27-2.51). Conclusions: Among elderly patients with stage I-III BC, being NHB was associated with increased odds of not receiving GCC and less timely treatment initiation. Being NHB was associated with increased all-cause mortality, which was attenuated after adjusting for GCC. Optimizing timely receipt of GCC may reduce inferior survival outcomes among elderly NHB breast cancer patients. Identifying area-level determinants may help design targeted interventions. Limitations include lack of granular systemic therapy data, likely overestimating GCC receipt, biasing the analysis towards the null, and lack of cause-specific death among elderly patients with competing risks of death.

Understanding barriers to guideline-concordant treatment in foregut cancer: From data to solutions.

e13521 Background: Receipt of guideline concordant treatment (GCT) is associated with improved prognosis in foregut (gastric, pancreatic, and hepatobiliary) cancers. However, a large proportion of patients with foregut cancers do not receive GCT. This single institution study from the Deep South sought to understand barriers to receipt of GCT using a root cause analysis (RCA) approach. Methods: A retrospective review of 498 patients with foregut adenocarcinoma diagnosed between 2018-2022 was performed. GCT was defined using National Comprehensive Cancer Network guidelines. The Ishikawa cause and effect model was used to establish main contributing factors to non-receipt of GCT. Results: Overall, 170 patients (34.1%) did not receive GCT. Median (Interquartile range [IQR] time from symptoms to start of workup was 8.8 weeks (4.3-16.4); time from diagnosis to first oncology appointment was 4.6 weeks (1.6- 6.6), and time from oncology appointment to start of treatment was 11 weeks (3.7- 18.2). Root causes of non-GCT included Patient, Physician, Institutional Environment and Broader System-related factors. Non-GCT was often a function of multiple intersecting patient, physician, and institutional factors. Barriers to receipt of GCT along the cancer care continuum included: receipt of incomplete therapy due lack of follow-up with surgical oncologist after initial appointment (16.5%), deconditioning on chemotherapy resulting in non-receipt of surgical therapy (15.3%), multiple delays in care due to patient resource constraints resulting in loss to follow-up after receipt of some therapy (11.2%), physician factors including provider knowledge (leading to delayed or non-diagnosis) or specialist non- adherence to GCT (11.2%), lack of documented treatment plan after referral to oncologist (11.2%), loss to follow up before oncology referral without record of health system attempt to re-establish care (10%), non-referral to oncologist (9.4%), non-referral to surgical oncologist in patients with resectable disease (8.8%), and complications during treatment preventing completion of treatment (6.5%). Findings from this study will allow us to propose targeted solutions to improve receipt of GCT, such as development of automated systems to improve patient follow-up and compliance with GCT; institutional prioritization of resources to enhance staffing for navigation, physical therapy, social services, and care coordination; financial counselling and assistance programs; and development and integration of structured prehabilitation programs into cancer treatment pathways. Conclusions: A substantial percentage of patients with foregut cancer do not receive GCT due to multiple barriers along the cancer care continuum. This RCA approach allows us to propose and develop solutions that may improve receipt of GCT.

Molecular and genomic testing in Native American patients with metastatic lung and colorectal cancers: A retrospective chart review.

e23136 Background: The Native American (NA) population has the poorest cancer survival of all racial groups in the U.S. A long history of research misconduct has caused many tribal communities to hold unfavorable views toward genomic data collection. Routine molecular and genomic testing is standard of care for most metastatic lung and colorectal cancers but the frequency of such testing in NA patients is unknown. Many NA patients have limited financial resources and must rely on coverage for health care from an Indian Health Service, Tribal Health Service, or Urban Indian Health (ITU) program. However, ITU programs do not operate their own cancer treatment services and must refer their patients elsewhere for such care. Methods: A Native American Navigation program (NANP) was established at the Stephenson Cancer Center (SCC) in Oklahoma to promote cancer health equity and provide high-quality cancer care by facilitating access to care for ITU patients from across the state. The NANP interfaces with ITU programs to obtain approvals for care as well as to coordinate clinical and supportive services, such as transportation or food assistance. We conducted a retrospective chart review of patients referred from an ITU program to the SCC NANP in Oklahoma from 1/2014 to 6/2023. Among patients with metastatic colorectal cancer (CRC), we investigated the frequency of testing for KRAS, NRAS, BRAF as well as MSS/MSI. In patients with metastatic non-small cell lung cancer (NSCLC), we identified the frequency of EGFR, ALK, ROS1, and PD-L1 testing. Results: A total of 305 NA patients were referred to SCC for colorectal and lung cancers. Of these, 68 patients had metastatic disease and qualified for molecular and genomic testing per National Comprehensive Cancer Network guidelines. The mean age of our cohort was 63 with a male-to-female ratio of 1.6:1. Most (85.3%) patients eligible for testing had testing for at least one biomarker. In patients with metastatic CRC, 100% received MSS/MSI as well as KRAS testing and the frequency of NRAS and BRAF testing was 69.6% and 84.8% respectively. In patients with metastatic NSCLC, the frequency of EGFR, ALK, and ROS1 testing was 95.8%, 87.5%, and 79.1%, with PD-L1 testing frequency being 83.3%. Conclusions: The overall frequency of molecular and genomic testing in these navigated NA patients (85.3%) was comparable to the range reported in the general population (66-90%). By facilitating access to cancer care, programs such as the NANP may help NA patients receive appropriate testing to guide treatment. Larger studies are needed to further assess the role of navigation programs in reducing cancer disparities by facilitating access to appropriate molecular and genomic testing for NA patients with metastatic cancers.

Trends in NCCN enforcement of financial conflict of interest policies for guidelines panelists.

11070 Background: Financial conflict of interest (FCOI) among authors of clinical practice guidelines is a potential concern for the independence and integrity of guideline recommendations. The National Comprehensive Cancer Network (NCCN) Guidelines policy stipulates limits on physician-industry FCOI among its panelists (<$50,000 overall and <$20,000 from each individual company, annually). Those in violation will be asked to resign. However, COI policies reliant on physician self-report may be difficult to enforce. The increased transparency of FCOI resulting from Open Payments – and the ensuing attention in 2016 on high levels of FCOI among guideline-writing bodies including the NCCN – may have enabled and motivated increased enforcement of existing FCOI policy. We describe trends in FCOI among NCCN Guidelines panelists and in enforcement when violations occur. Methods: We manually extracted NCCN Guidelines panelist names and terms of service for the 20 most prevalent cancers, 2013-22. We manually linked each physician to their Open Payments records, a federal archive of financial transactions from industry to physicians. For each panelist, we included only the categories of industry payments deemed relevant by the NCCN COI policy: ownership payments and some General Payments categories (eg., speaking fees, consulting) but not others (eg., free meals). We measured payments during each full calendar year of service. If a panelist received payments above NCCN limits (“violation”), we assessed whether their term of service continued for at least 1 additional full calendar year (“retained"), inferring that panelists who were retained had not been asked to resign as stipulated. We assessed whether retention post-violation was less likely after 2016. Results: There were 978 eligible physician-guideline pairs. Mean industry payments increased from $6175 in 2014 to $9169 in 2019, then declined to $6492 in 2021 following the COVID-19 pandemic. There were 143 panelist-years in violation of NCCN FCOI limits. Violations occurred among 99 unique panelists; 32 panelists had multiple observed violations. Across the full study period, panelists with a violation were less likely to be retained versus those without a violation (OR 0.21, 95% CI 0.15-0.31). However, likelihood of post-violation retention was lower after 2016 compared to before. 81% were retained during 2014-15 (versus 89% for those without violation, OR 0.55, 95% CI 0.26-1.31) compared to 47% during 2017-20 (versus 89% without violation, OR 0.10, 95% CI 0.06-0.17). Conclusions: Lower retention among panelists receiving excessive industry payments is consistent with NCCN enforcement of its COI policy. Lower post-violation retention after 2016 is consistent with stricter enforcement following that year. However, even in 2017-20, in 47% of violations, the panelist continued to serve, suggesting that enforcement remains incomplete.

Exploring AI-generated content and professional guidelines in cancer symptom management: A comparative analysis between ChatGPT and NCCN guidelines.

e13610 Background: Artificial intelligence (AI)-driven tools, like ChatGPT, have become widely-available sources for online health information. Limited research has explored the congruity between AI-generated content and professional treatment guidelines. This study seeks to compare recommendations for cancer-related symptoms generated from ChatGPT with guidelines from National Comprehensive Cancer Network (NCCN), a provider focused source requiring users to log in or register to access its recommendations. Utilizing a provider-focused source like NCCN serves as a benchmark to assess whether ChatGPT recommendations align with the standards typically endorsed by clinicians. Methods: We extracted treatment recommendations from four NCCN Supportive Care webpages (Cancer Pain, Antiemesis, Cancer-Related Fatigue, and Distress Management) and five subsections of the NCCN Palliative Care webpage (dyspnea, constipation, diarrhea, sleep disturbances, and anorexia/cachexia). We then entered "How can I reduce my cancer-related [symptom]" into ChatGPT 3.5 and extracted its recommendations. We calculated and compared word count and Flesch-Kincaid Grade Level readability for each NCCN and ChatGPT section. We completed a comparative content analysis focusing on recommendations for medications, consultations, and non-pharmacological strategies. Results: Across the nine NCCN Supportive Care and Palliative Care webpages, the mean word count was 2393.8 (SD=2601.4) and the mean Flesch-Kincaid Grade Level was 17.3 (SD=1.4) vs 382.4 (SD=29.6) and 11.6 (SD=0.8) for ChatGPT. The mean percent agreement between NCCN and ChatGPT recommendations was 44.6% (range 14.3%-81.8%). ChatGPT and NCCN guidelines shared fewer than half of their symptom-related recommendations in all but one section, fatigue. NCCN offered specific medication recommendations across all sections. ChatGPT's recommendations lacked the specificity observed in NCCN's guidelines including often not suggesting any medications. ChatGPT recommended specific medications in the shortness of breath and diarrhea sections that were not recommended by NCCN. NCCN's guidelines often did not include recommendations related to spirituality or palliative care consults, areas that ChatGPT addressed. Conclusions: While ChatGPT provides concise, accessible supportive care advice including many non-medical support recommendations, discrepancies with guidelines raise concerns for patient-facing symptom management recommendations. Overall, AI-generated content like ChatGPT can assist in providing preliminary information but should be researched in conjunction with comprehensive, evidence-based guidance provided by healthcare professionals. Healthcare providers should work with AI developers to ensure data sources are high-quality and accurate.

Using an AI trial screening tool to assess language as a barrier for enrollment of underrepresented minorities in cancer clinical trials.

e13586 Background: A major limitation of clinical trials is limited generalizability due to low enrollment of underrepresented minorities (URM). The percentage of patients who are non-English speaking (NES) is higher in URM and language barriers may discourage providers to offer participation in clinical trials. We have developed an artificial intelligence-based system to identify patients eligible for clinical trials, Clinical Trials Patient Matching (CTPM). This platform offers an ideal opportunity to explore bias in enrolling patients on clinical trials (CT). Among patients identified by CTPM as eligible for CT we assessed the impact of NES on trial enrollment. Methods: Individual algorithms were constructed within CTPM to identify patients potentially eligible for two trials open at Yale Cancer Center, DARE (NCT04567420) and EMBER-4 (NCT05514054). Eligibility was confirmed by Clinical Trials Office staff and notification of eligibility was communicated to the treating provider. Retrospectively, we evaluated the impact of NES status on clinical trial enrollment. This analysis occurred between 9/1/2022 and 11/30/2023 and all study-related activities were approved by the Yale IRB. Results: Out of the 13,466 unique patients presenting to the Yale Cancer Center and Smilow Cancer Network breast program within the period of study, 5.1% were NES and 94.9% were English speaking (ES). Of the patients identified by CTPM as eligible for the 2 clinical trials 7.9% were NES and 92.14% were ES. However, out of patients enrolled, only 1.59% were NES and 98.41% were ES. For the DARE trial, enrollment rate was 38.3% for ES vs 0% for NES (p=0.0081). For the EMBER4 trial, enrollment was 13.4% for ES vs 14.3% for NES (p=1.0). The aggregate data for both trials revealed an enrollment rate of 29.4% for ES vs 5.6% for NES (p=0.0285). Conclusions: We found that patients who identify as NES were less likely to be enrolled in cancer clinical trials. CTPM appears to be an effective tool to prospectively identify underrepresented populations in clinical trials and serve as a platform for developing interventions to address disparities in trial participation, including language. [Table: see text]

Clinical applicability of guideline-recommended molecular targets and genome-targeted cancer therapies.

11049 Background: Genomic testing has expanded the possibilities of precision medicine, particularly for advanced, treatment-resistant cancer cases. However, the clinical relevance of most genetic alterations remains uncertain, which can lead physicians to overestimate the benefits of tailored therapies. The National Comprehensive Cancer Network (NCCN) guidelines are cancer-specific treatment recommendations that often determine insurance coverage. We assessed the evidence for clinical benefit and actionability of molecular targets for genome-targeted cancer drugs recommended by NCCN. Methods: We identified genome-targeted therapies for solid cancers from the most recent NCCN guidelines. Trial design characteristics were obtained from publications supporting the NCCN recommendations. Genome target actionability was assessed with the European Society for Medical Oncology (ESMO) Scale for Clinical Actionability of molecular Targets (ESCAT). This ranges from genome alternation-therapy combinations that lead to improved outcomes (Tier I) to treatments with potential clinical relevance (Tiers II or III) to those with undetermined relevance (Tiers IV to X). Clinical benefit was evaluated using the ESMO-Magnitude of Clinical Benefit Scale (ESMO-MCBS). Molecular targets at ESCAT Tier I in combination with a given treatment associated with studies demonstrating substantial clinical benefit by ESMO-MCBS (Grades 4-5) were designated as high-benefit, while those linked to studies achieving an ESMO-MCBS Grade of 3 were categorized as promising but unproven. Results: We extracted411 recommendations from NCCN supporting 74 genome-targeted drugs targeting 50 driver alterations. Most recommendations (346/411, 84%) referred to clinical trial data, while one-sixth (65/411, 16%) relied on case reports or preclinical studies. Clinical trials were mostly Phase I or Phase II (271/346, 78%), single arm (262/346, 76%), and evaluated overall response rate as the primary endpoint (271/346, 78%). Over half of target recommendations had Tier I target actionability (246/411, 60%), over one-third were Tier II or III (142/411, 35%), and the rest were undetermined (23/411, 6%). Among 267 scorable trials,12% (32/267) demonstrated substantial clinical benefit (ESMO-MCBS Grades 4-5) and 45% (121/267) were Grade 3. When combining both frameworks,12% (32/267) genomic-based cancer treatments were high-benefit and 33% (88/267) were promising-but-unproven. Conclusions: About one-eighth of genome-targeted cancer therapies recommended in NCCN guidelines received a high benefit rating, while one-third were identified as having a promising but unproven substantial benefit according to the ESCAT and ESMO-MCBS frameworks. Ensuring NCCN recommendations are aligned with well-documented clinical benefits is crucial for promoting informed, evidence-based genomic-guided treatment decisions.