Manochakian, Rami

9002 Background: The field of AI is rapidly evolving. With recent development of more user-friendly AI software, its integration into healthcare across different medical specialties, including HemOnc has emerged. Data detailing the perception and concerns of the HemOnc workforce about AI’s roles is lacking and is needed if AI is to be integrated in HemOnc CliPr and MedEd. Methods: Questionnaires about the perception and concerns regarding AI in CliPr and MedEd were created on REDCap and approved by IRB. We surveyed the entire HemOnc workforce at the 3 Mayo Clinic major sites between 11/07/24 and 01/20/25, including physicians, both faculty (FAC) and fellows (FEL), advanced practice providers (APP), and nurses (RN). Participation was voluntary. Simple statistical analyses were employed for the results. Results: 344 participants (PTP) responded to the survey, 118 physicians (41 FEL and 77 FAC), 49 APP, and 177 RN. 64% of PTP report having used AI but only 31% used it in HemOnc MedEd, and 28% in HemOnc CliPr, with 67% considering themselves to have little to no knowledge about AI. 94% of PTP believe AI will be integrated in HemOnc MedEd, with such integration being seen as beneficial by 90%. Among physicians, 85% of FEL and 92% of FAC report that fellowship programs should incorporate AI training into curricula. 95% of PTP believe AI will be incorporated into HemOnc CliPr. Meanwhile, 62% of PTP (50.8% of physicians, and 68% of APP and RN) are concerned about risks it may pose to CliPr. Nevertheless, 90% of PTP would embrace AI’s use in HemOnc. Table 1 details the perceptions and concerns about AI in HemOnc MedEd and CliPr. The main perceived risks are decreased time spent with patients (pts), and inaccuracies or worse pt care. Meanwhile, 33% believe AI would increase efficiency and quality of care, while 30% believe it would increase time spent with pts. 14% of PTP worry their role could be replaced by AI. Conclusions: To the best of our knowledge, this is the largest assessment of HemOnc workforce’s perception and concerns about AI. Based on our survey, the HemOnc workforce, in general, endorses AI use, but with some concerns raised. While incorporation in MedEd is perceived with excitement, many envision the use in CliPr involves risks and challenges. Proper systematic education of the workforce about AI, with well-designed CliPr integration methods are needed to mitigate the legitimate existing concerns. Perceptions & concerns about AI in HemOnc MedEd & CliPr. FEL % APP % RN % FAC % AI will assist documentation in CliPr 98 82 79 92 AI will be used in pt communication 73 80 78 87 AI will screen pts for clinical trial enrollment 85 86 76 96 AI will increase the efficiency and quality of CliPr 43 37 24 46 AI will increase time spent with pts 39 27 25 38 AI will decrease time spent with pts 29 41 48 23 AI will lead to inaccuracies or worse CliPr 12 20 25 20 AI may replace the HemOnc workforce 11 16 19 5

e16610 Background: Adrenocortical carcinoma (ACC) is a rare malignancy with an incidence of 1-2 per million per year. The optimal treatment remains unclear as does the benefit of chemotherapy (chemo) in patients (pts) with resectable ACC. Methods: The SEER database (2000–2021) was used to identify ACC pts > 18 years. Variables of interest included age, race, household income, rural-urban classification, stage, chemo, radiation therapy (RT), and surgery. Statistical analyses were conducted using ANOVA, Kruskal-Wallis, and Kaplan-Meier methods to compare survival outcomes according to variables of interest. Results: Of the 2,598 pts identified, 1,564 (60%) were female. The majority of pts (85%) were white, followed by black (9%). Most pts (87%) resided in metropolitan areas. Pts with localized, regional, or distant metastatic disease comprised 25%, 17% and 36% of the group, respectively. 40% of pts underwent radical or total surgical removal of the primary site disease while 9%, 2%, and 1% underwent partial surgical removal, excision, or debulking surgery, respectively, and 30% did not undergo surgery. 34% of pts received chemo and 8% received RT. The median overall survival (mOS) for the entire cohort was 15 months (m) (95% CI: 14–17 m). No significant differences in OS were observed among pts of different races or household incomes. Distant metastatic disease was significantly associated with worse mortality when compared with regional or localized diseases (mOS: 5 m vs. 20 m vs 60.6 m, respectively, p < 0.001). Surgical intervention of any kind was associated with longer survival when compared to those who did not undergo surgery (18 m vs. 3 m, HR 0.33, p < 0.0001). Among the 971 pts with distant metastatic disease, pts who received chemo had a superior mOS compared to those who did not receive chemo or with unknown chemo status (8 m vs. 2 m, p < 0.0001). This OS advantage remained significant even after excluding pts who underwent surgery (5 m vs. 2 m, p < 0.0001). While chemo was also associated with longer mOS (29 m vs. 27 m, p < 0.0001) in localized disease, this survival advantage was not present in pts with localized disease managed with surgery (27 m vs. 27 m, p = 0.061). Although chemo was associated with numerically longer mOS in pts with regional disease, the difference was not significant. Conclusions: ACC with distant metastatic disease carries a poor prognosis yet demonstrates superior survival in pts who receive chemo. Surgery is associated with better survival outcomes, but there is no additive benefit from chemo in those with localized disease managed with surgical resection.

e20141 Background: TET are rare and challenging entities in clinical practice. Prospective trials in thymoma and thymic carcinoma (TC) are scarce. Limited number of small trials investigated the role of ICI in TET, and showed some clinical efficacy, but also with immune related adverse events (irAE). Guidelines list ICI as a treatment option in the 2 nd line, although no FDA approval was granted. Aside from the small trials, there is limited real world data about the efficacy and toxicity of ICI in TET. Methods: We performed a retrospective chart review of the medical records of all patients (pts) with TET who were treated with ICI between 3/25/2011 and 11/30/2024 at 5 Mayo Clinic sites. We collected data about baseline characteristics, diagnosis, irAE, and outcomes. Kaplan-Meier method was used to calculate median progression free survival (mPFS), and hazard ratio (HR) with Cox proportional hazards regression. Results: 30 pts were included, 28 diagnosed with TC, and 2 with malignant thymoma. 70% of pts were male. Median age at time of diagnosis was 61 years (range: 15 to 82), and median ECOG at the time of diagnosis was 0. 29 pts received Pembrolizumab, and one received Durvalumab. Prior to ICI, 86.67% of patients received chemotherapy. 70% of pts discontinued ICI use, 60% due to progression, and 10% due to irAE. Overall response rate (ORR) was 23.33%, with 3.33% complete response, and 20% partial response. Disease control rate (DCR) was 63.33%. The mPFS was 10.13 months. At the time of analysis 60% of pts were still alive. 40% of patients had irAE. While the two patients with thymoma experienced irAE, only 33.33% of those with TC had these events. 26.67% of pts had thyroiditis. Pancreatitis, pneumonitis, dermatitis, and encephalitis each occurred in 3.33% of pts. The median time to irAE onset was 75 days. 25% of irAE were grade 1, 66.67% grade 2, 0% grade 3, and 8.33% grade 4, represented by the case of encephalitis. Resolution of irAE occurred in 33.33% of cases. When stratified according to the occurrence of irAE, the mPFS was the same in both groups, with or without irAE, being 10.13 months, HR=0.99 (95% CI 0.39 – 2.54). Conclusions: In this real-world cohort of pts with TET, efficacy of ICI matched what has been previously reported, with similar ORR and DCR, but with longer mPFS. However, irAE incidence was higher, but with less grade 3/4 events. Differently than what has been suggested in other diseases, irAE occurrence did not correlate with better outcomes. These results validate the role of ICI in the treatment of TET, especially TC. Nonetheless, improving patient selection, and developing other therapeutic strategies, including combinations of ICI and other drugs, for these malignancies remain essential. Importantly, the high incidence of irAE indicates that future investigations must also focus on safety of interventions and supportive care in TET.

e20120 Background: In May 2024, the FDA approved tarlatamab, a Delta-like ligand (DLL3)/CD3-targeted bispecific T-cell engager, for patients with extensive-stage small cell lung cancer (SCLC) with disease progression following platinum-based chemotherapy and at least one other prior line of therapy. We aim to evaluate the real-world outcomes of agent. Methods: We conducted a retrospective multi-site single center analysis of adult patients who received tarlatamab at Mayo clinic sites including Arizona, Florida, Iowa, Minnesota, and Wisconsin. Data analysis was performed with Blue Sky statistics version 10.4. Results: Twenty patients with a median age 67 years (range 43-76) were included. Patients were predominantly women (55%) and all had a history of tobacco use. Metastatic disease commonly involved liver (55%) and CNS (60%). Majority (90%) had received prior PDL-1 therapy. 25% (5/20) of patients received prior gamma knife and in addition 10% (2/20) received whole brain radiation treatment in our cohort. Median time from CNS radiation and tarlatamab initiation was 4 months (range 3-28). Other characteristics, outcomes, and toxicity are presented in Table 1. Cytokine release syndrome (CRS) Grade (G) 1 was observed in 45% patients. G 2, 3 and 4 CRS were observed in 5% patients respectively. G 1 CRS was managed with acetaminophen initially followed dexamethasone 10mg or methylprednisone 1g if no improvement in 24 hours. Tocilizumab was used 20% patients for G 1,2,3 and 4 CRS depending on clinical course. Immune effector cell-associated neurotoxicity (ICANS) G 1 occurred in 10% patients, followed by G 2 in 15% and G 3 in 5% patients. ICANS was managed in a multidisciplinary fashion. All patients had resolution of CRS and ICANS. Conclusions: In this study investigating the real-world outcomes and toxicity of tarlatamab, a higher incidence of ICANS was observed than that reported in the DeLLphi-301 trial. The association between presence of CNS metastatic disease, prior CNS radiation and incidence of ICANS needs to be explored further. Mayo Clinic report of tarlatamab real world outcomes and toxicity. Patient Characteristics, Tarlatamab Outcomes and Toxicity Mayo Clinic Report DeLLphi-301 trial- Part 1 & 2 DeLLphi-301 trial- Part 3 ECOG no. (%) 2-4 4 (20) 0 (0) 0 (0) Metastatic Disease no.(%) 19 (95) 98 (98) 32 (94) Median no. of previous lines of therapy (range) 3 (2-4) 2 (1-6) 2 (2-6) Objective Response (%) 30% 40% CRS no (%) 12 (60) 49 (49) 19 (56) ICANS no (%) 6 (30) 7 (7) 4 (12) Neutropenia Grade 3 no (%) 2 (10) 6 (6) 2 (6) Fatal CRS, ICANS or Neutropenia 0 (0) 0 (0) 0 (0)

e20011 Background: The therapeutic landscape of lung cancer is rapidly evolving, presenting oncologists with the challenge of staying updated amidst an overwhelming influx of data. Clinical decision support (CDS) tools, including artificial intelligence (AI) and large language models (LLMs), may help bridge this gap. Evaluating the accuracy of LLMs in complex, real-world oncology scenarios is crucial to understanding their potential. Methods: Twenty-five de-identified lung cancer cases from the fellows’ clinic at Karmanos Cancer Institute, Detroit, MI, were analyzed. Two LLMs, GPT-4 (OpenAI) and Claude Opus (Anthropic), were assessed using advanced prompting techniques like persona-based and chain-of-thought prompting. Five board-certified lung cancer oncologists from NCI-designated centers evaluated LLM-generated responses based on accuracy, treatment recommendation comprehensiveness, and supportive care planning, using a 1–5 scale. Novel insights, the presence of fabricated information, and harmful recommendations were flagged as binary outcomes. Oncologists were blinded to the LLM source and actual treatment decisions. Results: Table 1 presents patient characteristics. GPT-4 achieved an average accuracy score of 4.2 (95% CI, 3.9–4.4), with 3.7 for comprehensiveness of medical/surgical treatment recommendations and 3.7 for supportive care planning. Six responses (32%) were flagged as potentially harmful, and two (8%) contained inaccuracies. Sixteen GPT-4 responses (64%) were rated trustworthy as a CDS tool. Claude Opus had an average accuracy score of 3.6 (95% CI, 3.1–4.1), scoring 3.6 for treatment recommendation comprehensiveness and 3.5 for supportive care planning. Nine responses (36%) were flagged for potential harm, and five (20%) included inaccuracies. Eleven Claude responses (44%) were deemed trustworthy. Significant differences were observed in accuracy (p=0.04) and trustworthiness (p=0.03) between models using McNemar's test. Other factors showed no statistical significance. Conclusions: GPT-4 outperformed Claude Opus in accuracy and trustworthiness, but both models demonstrated limitations, including harmful recommendations and inaccuracies. These findings highlight the need for improved LLM refinement before routine use as CDS tools in lung cancer treatment. Patient demographics and clinical characteristics. Category subcategory Number Median Age (range)- Yr 65 (26-78) Female 7 Male 18 Histology Adenocarcinoma 10 Squamous Cell Carcinoma (SCC) 7 Small Cell Carcinoma 6 Poorly Differentiated 2 Total 25 Stage NSCLC Stage 3 7 NSCLC Stage 4 13 Small Cell limited stage 3 Small Cell Extensive Stage 2

In the rapidly evolving field of oncology, education and knowledge sharing have become essential aspects of oncology practice and must-needed tools to help patients conquer cancer. Oncology education is an art that needs to be mastered by applying scientific learning methods through structured frameworks into it, in order to have an impact. To effectively educate and share knowledge with oncology professionals, we propose following a “chain of education”. This chain consists of five links that are essential for creating a successful educational activity: “who”, “why”, “where”, “what”, and “how”. By following the “chain of education” through all links leading to the final key of how to educate, and guided by learning science principles, educators can optimize and enhance their teaching and knowledge-sharing skills and educate learners with more impact. Two common scenarios of medical education in oncology are the education of trainees and the education of community oncologists. In this article, we review them through the prism of the “chain of education”. We discuss how using effective learning science principles can guide small changes that lead to significant impacts in trainees' education and can also leverage social media platforms in the education of community oncologists. Alongside this described framework, we briefly review some available oncology education resources, and we discuss incorporating the “chain of education” into future technologies.

Lung cancer remains the leading cause of cancer-related mortality worldwide. Non-small-cell lung cancer (NSCLC) is the most common type of lung cancer, with nearly half of all patients diagnosed at an advanced stage. Immune checkpoint inhibitors (ICIs) harness the host immune system to combat malignant cells. ICIs, which target programmed death-ligand 1 (PD-L1), programmed cell death 1 (PD-1), and cytotoxic T-cell lymphocyte-4 (CTLA-4), have transformed the treatment landscape for advanced NSCLC. While a subset of patients experiences a long-term durable response, most patients will develop disease progression. New drugs targeting novel pathways are being tested in clinical trials to improve the efficacy of immunotherapy and overcome resistance patterns. This review aims to summarize the currently available ICIs for advanced NSCLC and describe emerging immunotherapies with recently published data from phase I/II clinical trials.

The emergence of immunotherapy has ushered in a new era in the management of non-small cell lung cancer (NSCLC). Various immune check point inhibitors have demonstrated significant benefit in the management of locally advanced NSCLC that are treated with either surgery or concurrent chemoradiation. We provide a comprehensive and up-to-date review of data from key studies, discuss the challenging clinical issue regarding the timing and duration of immunotherapy in patients undergoing surgery, and highlight the unmet needs and future directions of immunotherapy in NSCLC.

Introduction: Myeloma therapeutics have tremendously improved in recent times, leading to unprecedented outcomes. Timely initiation of effective treatment is vital to improving survival and quality of life. Patient refusal to receive treatment is infrequent but certainly leads to increased morbidity/mortality. We studied factors associated with treatment refusal. Methods: The National Cancer Database (NCDB) was queried for adult patients diagnosed with MM between 2004-2021. Those who refused treatment for initial management despite physician's recommendation were identified. Patients with multiple primary malignancies or unknown treatment status were excluded. Patients' demographics and socioeconomic factors were collected. Kaplan-Meier and multivariate Cox regression were used to evaluate the likelihood of treatment refusal in the whole cohort and in each racial/ethnic group. Results: Out of 86,350 eligible patients with MM who were recommended to initiate therapy, 2,508 (2.9%) did not receive any treatment due to refusal. Of these, 52.6% were females, 64% were non-Hispanic White (NHW), 21.3% were non-Hispanic African American (NHAA), and 5.2% were Hispanic. On multivariate analysis, patients who were treated at comprehensive and integrated facilities were more likely to refuse treatment as compared to those at academic facilities (OR=1.36; 95% CI: 1.22-1.52, p&lt;0.001 and OR=1.22; 95% CI: 1.07-1.4, p=0.004, respectively). Patients on Medicare (OR=1.42; 95% CI: 1.16-1.73, p&lt;0.001), and presence of higher Charlson Comorbidity Index (CCI) also had a higher likelihood of refusing treatment. Lower refusal rates were seen for male gender (OR=0.83; 95% CI: 0.76-0.91, p&lt;0.001), NHW racial/ethnic group (as against NHAA; OR=0.34; 95% CI: 0.16-0.71, p=0.004), being insured (OR=0.69; 95% CI: 0.52-0.93, p=0.015), having private insurance (compared to Medicaid; OR=0.47; 95% CI: 0.38-0.59, p&lt;0.001) and having a higher income level (OR=0.82; 95% CI: 0.70-0.96, p=0.013). Among NHAA patients, 2.8% refused treatment and were more likely to have Medicare (OR=1.4; 95% CI: 1.03-1.9, p=0.032). Having private insurance (compared to Medicaid; OR=0.61; 95% CI: 0.43-0.88, p=0.008) had a lower likelihood of refusing treatment. Among NHW, 2.9% refused treatment and they were more likely to be treated at community, comprehensive and integrated facilities compared to academic facilities (OR=1.39; 95% CI: 1.12-1.73, p=0.003, OR=1.54; 95% CI: 1.33-1.77, p&lt;0.001 and OR=1.38; 95% CI: 1.16-1.64, p&lt;0.001, respectively), be from rural areas (OR=1.43; 95% CI: 1.05-1.96, p=0.025), and have a higher CCI. On the other hand, male gender (OR=0.79; 95% CI: 0.72-0.89, p&lt;0.001), private insurance (compared to Medicaid; OR=0.42; 95% CI: 0.30-0.59, p&lt;0.001) and higher income level (OR=0.80; 95% CI: 0.66-0.98, p=0.029) had a lower likelihood of refusing treatment. Among Hispanic patients, 2.3% refused treatment and were more likely to have Medicare (OR=1.96; 95% CI: 1.13-3.4, p=0.016), while those with private insurance (compared to Medicaid; OR=0.32; 95% CI: 0.14-0.73, p=0.007) had a lower likelihood of refusing treatment. Conclusions: This large real-world analysis highlights the intricate sociodemographic and economic factors that affect the decision of a MM to not proceed with the recommendation for treatment. (1) NHW are less likely to refuse treatment as compared to NHAA. (2) Insurance status is an important driver of refusal and transcends race/ethnicity. Furthermore, in this study, we highlight characteristics that are specific to different racial/ethnic groups like CCI which affect only certain racial/ethnic groups. This provides an opportunity to develop targeted interventions such that the benefit of therapeutic advancements can be realized by all.

Introduction: Overall outcomes have improved significantly for MM with a 5-year relative survival rate currently of &gt;60%. With effective treatments, patients from all clinical strata and age groups are expected to have meaningful benefit, but despite excellent therapeutic advancements, some patients continue to have early mortality. We conducted this study to investigate factors associated with early mortality. Methods: The National Cancer Database (NCDB) was queried for adult patients with NDMM between 2004 and 2021. We excluded patients with multiple primary malignancies or lost to follow-up (unknown vital status or last contact). Early mortality was defined as death within 6 months from MM diagnosis. Patient demographics and socioeconomic factors available in NCDB were collected. Kaplan-Meier and multivariate Cox regression were used to evaluate the likelihood of early mortality in this large cohort of MM patients using multi-variate analysis. Results: Out of 255,579 eligible patients with MM, 35,302 (13.8%) died within the first 6 months of diagnosis and 108,534 (42.5%) died after the first 6 months of MM diagnosis. The early mortality cohort had significantly higher mean age at MM diagnosis (73.9 vs. 68.5 years, p&lt;0.001) but no significant differences were noted for gender distribution. There was a statistically significant difference for racial/ethnic distribution (p=0.019), but the percentages were very similar among non-Hispanic White (69% vs 68.1%), non-Hispanic African American (18% vs 18.7%) and Hispanic patients (4.7% vs 4.8%). Patients in the early mortality cohort were less likely to be treated at academic institutions (28.4% vs 37.8%, p&lt;0.001) and have significantly different distribution by insurance type (less likely to have private insurance; 17.3% vs. 30%, more likely to have Medicare; 72.1% vs. 58.7%, p&lt;0.001) although overall insurance access was similar between the two groups (95.3% vs 95.2%, p=0.283). Early mortality was associated with higher morbidity scores as per the Charlson-Comorbidity Index (CCI) (p&lt;0.001) and were more likely to have not received any treatment for MM (53.2% in early mortality group vs. 28.4% in late mortality, p&lt;0.001). Early mortality was also associated with a lower literacy rate (high school graduation rate 21.2% vs. 19.7%, p&lt;0.001) and lower household income (&lt;$46,277; 19.1% vs. 17.7%, p&lt;0.001). Conclusion: This large real-world analysis shows that despite therapeutic advances in recent years, there is a significant percentage of patients with NDMM in the U.S. who suffer from early mortality. There are several sociodemographic and economic factors that were found to be associated with early mortality, with some showing clear trends (older age, treatment facility type, insurance type, income, education, CCI), while others like racial differences were significant but somewhat ambiguous. Identifying and addressing these factors at diagnosis will help provide personalized care plans and minimize early mortality.