Perioperative Nivolumab in Resectable Lung Cancer.

BACKGROUND: Standard treatment with neoadjuvant nivolumab plus chemotherapy significantly improves outcomes in patients with resectable non-small-cell lung cancer (NSCLC). Perioperative treatment (i.e., neoadjuvant therapy followed by surgery and adjuvant therapy) with nivolumab may further improve clinical outcomes. METHODS: In this phase 3, randomized, double-blind trial, we assigned adults with resectable stage IIA to IIIB NSCLC to receive neoadjuvant nivolumab plus chemotherapy or neoadjuvant chemotherapy plus placebo every 3 weeks for 4 cycles, followed by surgery and adjuvant nivolumab or placebo every 4 weeks for 1 year. The primary outcome was event-free survival according to blinded independent review. Secondary outcomes were pathological complete response and major pathological response according to blinded independent review, overall survival, and safety. RESULTS: At this prespecified interim analysis (median follow-up, 25.4 months), the percentage of patients with 18-month event-free survival was 70.2% in the nivolumab group and 50.0% in the chemotherapy group (hazard ratio for disease progression or recurrence, abandoned surgery, or death, 0.58; 97.36% confidence interval [CI], 0.42 to 0.81; P<0.001). A pathological complete response occurred in 25.3% of the patients in the nivolumab group and in 4.7% of those in the chemotherapy group (odds ratio, 6.64; 95% CI, 3.40 to 12.97); a major pathological response occurred in 35.4% and 12.1%, respectively (odds ratio, 4.01; 95% CI, 2.48 to 6.49). Grade 3 or 4 treatment-related adverse events occurred in 32.5% of the patients in the nivolumab group and in 25.2% of those in the chemotherapy group. CONCLUSIONS: Perioperative treatment with nivolumab resulted in significantly longer event-free survival than chemotherapy in patients with resectable NSCLC. No new safety signals were observed. (Funded by Bristol Myers Squibb; CheckMate 77T ClinicalTrials.gov number, NCT04025879.).

Effects of immune-related adverse events (irAEs) and their treatment on antitumor immune responses.

Immune checkpoint inhibitors (ICIs) are potentially life-saving cancer therapies that can trigger immune-related adverse events (irAEs). irAEs can impact any organ and range in their presentation from mild side effects to life-threatening complications. The relationship between irAEs and antitumor immune responses is nuanced and may depend on the irAE organ, the tumor histology, and the patient. While some irAEs likely represent an immune response against antigens shared between tumor cells and healthy tissues, other irAEs may be entirely unrelated to antitumor immune responses. Clinical observations suggest that low-grade irAEs have a positive association with responses to ICIs, but the correlation between severe irAEs and clinical benefit is less clear. Currently, severe irAEs are typically treated by interrupting or permanently discontinuing ICI treatment and administering empirically selected systemic immunosuppressive agents. However, these interventions could potentially diminish the antitumor effects of ICIs. Efforts to understand the mechanistic relationship between irAEs and the tumor microenvironment have yielded meaningful insights and nominated therapeutic targets for irAE management that may preserve or even boost ICI efficacy. We explore the clinical and molecular relationship between irAEs and antitumor immunity as well as the role that irAE treatments may play in shaping antitumor immune responses.

Spatially Resolved Niche and Tumor Microenvironmental Alterations in Gastric Cancer Peritoneal Metastases.

BACKGROUND AND AIMS: Peritoneal metastasis (PM) in gastric cancer (GC) is associated with poor prognosis and significant morbidity. We sought to understand the genomic, transcriptomic, and tumor microenvironment (TME) features that contribute to peritoneal organotropism in GC. METHODS: We conducted a comprehensive multi-omic analysis of 548 samples from 326 patients, including primary tumors, matched normal tissues; peritoneal metastases, and adjacent-normal peritoneal tissues. We used whole exome sequencing, whole transcriptome sequencing, and digital spatial profiling to investigate molecular alterations, gene expression patterns, and TME characteristics associated with PM. RESULTS: Our analysis identified specific genomic alterations in primary tumors, including mutations in ELF3, CDH1, and PIGR, and TME signatures, such as stromal infiltration and M2 macrophage enrichment, associated with increased risk of PM. We observed distinct transcriptional programs and immune compositions in GCPM compared with liver metastases, highlighting the importance of the TME in transcoelomic metastasis. We found differential expression of therapeutic targets between primary tumors and PM, with lower CLDN18.2 and FGFR2b expression in PM. We unravel the roles of the TME in niche reprogramming within the peritoneum, and provide evidence of pre-metastatic niche conditioning even in early GC without clinical PM. These findings were further validated using a humanized mouse model, which demonstrated niche remodeling in the peritoneum during transcoelomic metastasis. CONCLUSION: Our study provides a comprehensive molecular characterization of GCPM and unveils key biological principles underlying transcoelomic metastasis. The identified predictive markers, therapeutic targets, and TME alterations offer potential avenues for targeted interventions and improved patient outcomes.

The Relationship Between Health-Related Quality of Life and Overall Survival in Patients With Advanced Renal Cell Carcinoma in CheckMate 214.

BACKGROUND: In CheckMate 214 (median follow-up, 25.2 months), nivolumab plus ipilimumab yielded greater overall survival (OS) benefit than sunitinib in patients with intermediate-/poor-risk advanced renal cell carcinoma (aRCC). Health-related quality of life (HRQoL) assessed by the Functional Assessment of Cancer Therapy-Kidney Symptom Index-19 (FKSI-19) was also more favorable for the nivolumab plus ipilimumab group than the sunitinib group. We investigated whether HRQoL scores can predict OS of patients with 5 years follow-up in CheckMate 214. PATIENTS AND METHODS: CheckMate 214 was an open-label, phase III trial in previously untreated aRCC (N = 1096). Patients with intermediate-/poor-risk disease (International mRCC Database Consortium prognostic score ≥ 1; n = 847) were randomized to either nivolumab plus ipilimumab or sunitinib monotherapy. Pooled data for OS and FKSI-19 total and subscales (disease-related symptoms [DRS], DRS-physical [DRS-P], and function/well-being [FWB]) were analyzed. Relationships between HRQoL and OS were assessed using Cox proportional hazard models with baseline and longitudinal scores. Associations between HRQoL changes and OS were assessed by landmark analyses. RESULTS: Patients with higher FKSI-19 total and subscale scores at baseline had longer OS than patients with lower scores (HR ≤ 0.834; P < .0001). Longitudinal models indicated stronger associations between HRQoL and OS (HR ≤ 0.69; P < .001 for each). At 3 months after randomization, patients with stable/improved HRQoL versus baseline had longer median OS than patients with worsened/unobserved HRQoL versus baseline (55.9 and 26.0 months, respectively; HR = 0.56; 95% CI, 0.46-0.67; P < .0001). Results at 6-, 9-, and 12-month landmarks were consistent with these findings. CONCLUSION: In aRCC, patient-reported outcomes are important for HRQoL and prognostic evaluation. CLINICALTRIALS.GOV IDENTIFIER: NCT02231749; https://clinicaltrials.gov/ct2/show/NCT02231749.

Patient-reported outcomes with first-line nivolumab plus cabozantinib versus sunitinib in patients with advanced renal cell carcinoma treated in CheckMate 9ER: an open-label, randomised, phase 3 trial.

BACKGROUND: In the CheckMate 9ER trial, patients with advanced renal cell carcinoma who received first-line nivolumab plus cabozantinib had significantly better progression-free survival compared with those given sunitinib. In this study, we aimed to describe the patient-reported outcome (PRO) results from CheckMate 9ER. METHODS: In this open-label, randomised, phase 3 trial done in 125 cancer centres, urology centres, and hospitals across 18 countries, patients aged 18 years or older with previously untreated advanced renal cell carcinoma with a clear-cell component, a Karnofsky performance status of 70% or more, and available tumour tissue were randomly assigned (1:1) via interactive response technology to nivolumab 240 mg intravenously every 2 weeks plus oral cabozantinib 40 mg per day, or oral sunitinib 50 mg per day monotherapy for 4 weeks in 6-week cycles. The primary endpoint of progression-free survival was reported previously. PROs were analysed as prespecified exploratory endpoints at common timepoints (at baseline and every 6 weeks) until week 115. Disease-related symptoms were evaluated using the 19-item Functional Assessment of Cancer Therapy-Kidney Symptom Index (FKSI-19), and global health status was assessed with the three-level EQ-5D (EQ-5D-3L) visual analogue scale (VAS) and UK utility index. PRO analyses were done in the intention-to-treat population. Change from baseline was assessed using mixed-model repeated measures. A time-to-deterioration analysis was done for first and confirmed deterioration events. This study is registered with ClinicalTrials.gov, NCT03141177, and is closed to recruitment. FINDINGS: Between Sept 11, 2017, and May 14, 2019, 323 patients were randomly assigned to nivolumab plus cabozantinib and 328 to sunitinib. Median follow-up was 23·5 months (IQR 21·0-26·5). At baseline, patients in both groups reported low symptom burden (FKSI-19 disease-related symptoms version 1 mean scores at baseline were 30·24 [SD 5·19] for the nivolumab plus cabozantinib group and 30·06 [5·03] for the sunitinib group). Change from baseline in PRO scores indicated that nivolumab plus cabozantinib was associated with more favourable outcomes versus sunitinib (treatment difference 2·38 [95% CI 1·20-3·56], nominal p<0·0001, effect size 0·33 [95% CI 0·17-0·50] for FKSI-19 total score; 1·33 [0·84-1·83], nominal p<0·0001, 0·45 [0·28-0·61] for FKSI-19 disease-related symptoms version 1; 3·48 [1·58-5·39], nominal p=0·0004, 0·30 [0·14-0·47] for EQ-5D-3L VAS; and 0·04 [0·01-0·07], nominal p=0·0036, 0·25 [0·08-0·41] for EQ-5D-3L UK utility index), reaching significance at most timepoints. Nivolumab plus cabozantinib was associated with decreased risk of clinically meaningful deterioration for FKSI-19 total score compared with sunitinib (first deterioration event hazard ratio 0·70 [95% CI 0·56-0·86], nominal p=0·0007; confirmed deterioration event 0·63 [0·50-0·80], nominal p=0·0001). INTERPRETATION: PROs were maintained or improved with nivolumab plus cabozantinib versus sunitinib. Compared with sunitinib, nivolumab plus cabozantinib significantly delayed time to deterioration of patient-reported outcome scores. These results suggest a benefit for nivolumab plus cabozantinib compared with sunitinib in the treatment of patients with advanced renal cell carcinoma. FUNDING: Bristol Myers Squibb.

Comparing patient global impression of severity and patient global impression of change to evaluate test-retest reliability of depression, non-small cell lung cancer, and asthma measures.

PURPOSE: Score reproducibility is an important measurement property of fit-for-purpose patient-reported outcome (PRO) measures. It is commonly assessed via test-retest reliability, and best evaluated with a stable participant sample, which can be challenging to identify in diseases with highly variable symptoms. To provide empirical evidence comparing the retrospective (patient global impression of change [PGIC]) and current state (patient global impression of severity [PGIS]) approaches to identifying a stable subgroup for test-retest analyses, 3 PRO Consortium working groups collected data using both items as anchor measures. METHODS: The PGIS was completed on Day 1 and Day 8 + 3 for the depression and non-small cell lung cancer (NSCLC) studies, and daily for the asthma study and compared between Day 3 and 10. The PGIC was completed on the final day in each study. Scores were compared using an intraclass correlation coefficient (ICC) for participants who reported "no change" between timepoints for each anchor. RESULTS: ICCs using the PGIS "no change" group were higher for depression (0.84 vs. 0.74), nighttime asthma (0.95 vs. 0.53) and daytime asthma (0.86 vs. 0.68) compared to the PGIC "no change" group. ICCs were similar for NSCLC (PGIS: 0.87; PGIC: 0.85). CONCLUSION: When considering anchor measures to identify a stable subgroup for test-retest reliability analyses, current state anchors perform better than retrospective anchors. Researchers should carefully consider the type of anchor selected, the time period covered, and should ensure anchor content is consistent with the target measure concept, as well as inclusion of both current and retrospective anchor measures.

A Case of Microsatellite Instability-High Colon Cancer in a Young Woman With Familial Adenomatous Polyposis.

Two major molecular pathways of colorectal carcinogenesis, chromosomal instability (CIN) and microsatellite instability (MSI), are considered to be mutually exclusive. Distinguishing CIN from MSI-high tumors has considerable therapeutic implications, because patients with MSI-high tumors can derive considerable benefit from immune checkpoint inhibitors, and tumors that evolved through the CIN pathway do not respond to these agents. Familial adenomatous polyposis (FAP) is a genetic syndrome that is defined by a mutation in the APC gene and is thought to lead to carcinogenesis through the CIN pathway. Here, we report a case of a young woman with FAP who was treated for medulloblastoma as a child and developed advanced MSI-high colon cancer as a young adult. Her response to second-line immunotherapy enabled resection of her colon cancer, and she is free of disease >10 months after surgery. This case highlights the potential for overlap between the CIN and MSI carcinogenic pathways and associated therapeutic implications.

Vitamin D intake is associated with decreased risk of immune checkpoint inhibitor-induced colitis.

BACKGROUND: There is a lack of predictive markers informing on the risk of colitis in patients treated with immune checkpoint inhibitors (ICIs). The aim of this study was to identify potential factors associated with development of ICI colitis. METHODS: We performed a retrospective analysis of melanoma patients at Dana-Farber Cancer Institute who received PD-1, CTLA-4, or combination ICIs between May 2011 to October 2017. Clinical and laboratory characteristics associated with pathologically confirmed ICI colitis were evaluated using multivariable logistic regression analyses. External confirmation was performed on an independent cohort from Massachusetts General Hospital. RESULTS: The discovery cohort included 213 patients of whom 37 developed ICI colitis (17%). Vitamin D use was recorded in 66/213 patients (31%) before starting ICIs. In multivariable regression analysis, vitamin D use conferred significantly reduced odds of developing ICI colitis (OR 0.35, 95% CI 0.1-0.9). These results were also demonstrated in the confirmatory cohort (OR 0.46, 95% CI 0.2-0.9) of 169 patients of whom 49 developed ICI colitis (29%). Pre-treatment neutrophil-to-lymphocyte ratio (NLR) ≥5 predicted reduced odds of colitis (OR 0.34, 95% CI 0.1-0.9) only in the discovery cohort. CONCLUSIONS: This is the first study to report that among patients treated with ICIs, vitamin D intake is associated with reduced risk for ICI colitis. This finding is consistent with prior reports of prophylactic use of vitamin D in ulcerative colitis and graft-versus-host-disease. This observation should be validated prospectively in future studies.

Historical Evolution and Provider Awareness of Inactive Ingredients in Oral Medications.

PURPOSE: A multitude of different versions of the same medication with different inactive ingredients are currently available. It has not been quantified how this has evolved historically. Furthermore, it is unknown whether healthcare professionals consider the inactive ingredient portion when prescribing medications to patients. METHODS: We used data mining to track the number of available formulations for the same medication over time and correlate the number of available versions in 2019 to the number of manufacturers, the years since first approval, and the number of prescriptions. A focused survey among healthcare professionals was conducted to query their consideration of the inactive ingredient portion of a medication when writing prescriptions. RESULTS: The number of available versions of a single medication have dramatically increased in the last 40 years. The number of available, different versions of medications are largely determined by the number of manufacturers producing this medication. Healthcare providers commonly do not consider the inactive ingredient portion when prescribing a medication. CONCLUSIONS: A multitude of available versions of the same medications provides a potentially under-recognized opportunity to prescribe the most suitable formulation to a patient as a step towards personalized medicine and mitigate potential adverse events from inactive ingredients.

Activation of Notch1 synergizes with multiple pathways in promoting castration-resistant prostate cancer.

Metastatic castration-resistant prostate cancer (CRPC) is the primary cause of prostate cancer-specific mortality. Defining new mechanisms that can predict recurrence and drive lethal CRPC is critical. Here, we demonstrate that localized high-risk prostate cancer and metastatic CRPC, but not benign prostate tissues or low/intermediate-risk prostate cancer, express high levels of nuclear Notch homolog 1, translocation-associated (Notch1) receptor intracellular domain. Chronic activation of Notch1 synergizes with multiple oncogenic pathways altered in early disease to promote the development of prostate adenocarcinoma. These tumors display features of epithelial-to-mesenchymal transition, a cellular state associated with increased tumor aggressiveness. Consistent with its activation in clinical CRPC, tumors driven by Notch1 intracellular domain in combination with multiple pathways altered in prostate cancer are metastatic and resistant to androgen deprivation. Our study provides functional evidence that the Notch1 signaling axis synergizes with alternative pathways in promoting metastatic CRPC and may represent a new therapeutic target for advanced prostate cancer.

Scale development with small samples: a new application of longitudinal item response theory.

PURPOSE: Measurement development in hard-to-reach populations can pose methodological challenges. Item response theory (IRT) is a useful statistical tool, but often requires large samples. We describe the use of longitudinal IRT models as a pragmatic approach to instrument development when large samples are not feasible. METHODS: The statistical foundations and practical benefits of longitudinal IRT models are briefly described. Results from a simulation study are reported to demonstrate the model's ability to recover the generating measurement structure and parameters using a range of sample sizes, number of items, and number of time points. An example using early-phase clinical trial data in a rare condition demonstrates these methods in practice. RESULTS: Simulation study results demonstrate that the longitudinal IRT model's ability to recover the generating parameters rests largely on the interaction between sample size and the number of time points. Overall, the model performs well even in small samples provided a sufficient number of time points are available. The clinical trial data example demonstrates that by using conditional, longitudinal IRT models researchers can obtain stable estimates of psychometric characteristics from samples typically considered too small for rigorous psychometric modeling. CONCLUSION: Capitalizing on repeated measurements, it is possible to estimate psychometric characteristics for an assessment even when sample size is small. This allows researchers to optimize study designs and have increased confidence in subsequent comparisons using scores obtained from such models. While there are limitations and caveats to consider when using these models, longitudinal IRT modeling may be especially beneficial when developing measures for rare conditions and diseases in difficult-to-reach populations.

Assessment of Patient-Reported Outcome Instruments to Assess Chronic Low Back Pain.

Objective: To identify patient-reported outcome (PRO) instruments that assess chronic low back pain (cLBP) symptoms (specifically pain qualities) and/or impacts for potential use in cLBP clinical trials to demonstrate treatment benefit and support labeling claims. Design: Literature review of existing PRO measures. Methods: Publications detailing existing PRO measures for cLBP were identified, reviewed, and summarized. As recommended by the US Food & Drug Administration (FDA) PRO development guidance, standard measurement characteristics were reviewed, including development history, psychometric properties (validity and reliability), ability to detect change, and interpretation of observed changes. Results: Thirteen instruments were selected and reviewed: Low Back Pain Bothersomeness Scale, Neuropathic Pain Symptom Inventory, PainDETECT, Pain Quality Assessment Scale Revised, Revised Short Form McGill Pain Questionnaire, Low Back Pain Impact Questionnaire, Oswestry Disability Index, Pain Disability Index, Roland-Morris Disability Questionnaire, Brief Pain Inventory and Brief Pain Inventory Short Form, Musculoskeletal Outcomes Data Evaluation and Management System Spine Module, Orebro Musculoskeletal Pain Questionnaire, and the West Haven-Yale Multidimensional Pain Inventory Interference Scale. The instruments varied in the aspects of pain and/or impacts that they assessed, and none of the instruments fulfilled all criteria for use in clinical trials to support labeling claims based on recommendations outlined in the FDA PRO guidance. Conclusions: There is an unmet need for a validated PRO instrument to evaluate cLBP-related symptoms and impacts for use in clinical trials.

Economic Evaluation of Linaclotide for the Treatment of Adult Patients With Chronic Idiopathic Constipation in the United States.

PURPOSE: To evaluate the effectiveness and costs of linaclotide (Linzess) versus lubiprostone (Amitiza) in the treatment of adult patients with chronic idiopathic constipation (CIC). DESIGN: A decision-tree model using model inputs derived from published literature, linaclotide phase 3 trial data, and a physician survey. METHODOLOGY: Measures of treatment efficacy were selected based on comparability between trial data, with posthoc analyses of linaclotide required to ensure comparability with available lubiprostone data. Response to therapy was defined as (1) having one of the best two satisfaction answers of a 5-point global treatment satisfaction scale at Week 4 or (2) having a weekly spontaneous bowel movement (SBM) frequency 4 at Week 4. Patients who do not respond to therapy are assumed to accrue costs associated with a treatment failure. Model time horizon is aligned with the lubiprostone clinical trial duration of 4 weeks. Model outputs include response rates, quality-adjusted life-years (QALYs) and direct costs. RESULTS: Linaclotide was associated with lower per-patient costs vs lubiprostone for both definitions of response ($946 vs $1,015 for global assessment and $727 vs $737 for SBM frequency). When treatment response was based on a global assessment of treatment satisfaction, linaclotide was associated with higher effectiveness (response: 39.3% vs 35.0%). For SBM frequency, linaclotide was slightly less effective compared to lubiprostone (response: 58.6% vs 59.6%), but also less costly. Base-case results were robust in sensitivity analysis. CONCLUSIONS: Linaclotide is less expensive with similar effectiveness when compared to lubiprostone for the treatment of CIC in adult patients.

Development and validation of the Diabetic Peripheral Neuropathic Pain Impact (DPNPI) measure, a patient-reported outcome measure.

PURPOSE: Diabetic peripheral neuropathy (DPN) occurs in 26-47 % of diabetes patients and may have negative impacts on physical functioning, sleep, well-being, and quality of life. The Diabetic Peripheral Neuropathic Pain Impact measure (DPNPI) was developed to measure disease impacts and treatment effects. Presented are the DPNPI conceptual development and validation findings. METHODS: The DPNPI was developed following the FDA Guidance for Industry on patient-reported outcome (PRO) measures. Concept elicitation (CE) included literature review, clinical expert interviews, and patient interviews/focus groups. Qualitative data were analyzed following grounded theory principles, and draft items were cognitively debriefed. The measure underwent psychometric validation, and an a priori statistical analysis plan assessed the measurement model, reliability, and validity. Simultaneous analyses of item functioning were conducted using Rasch measurement theory (RMT). All tests were performed for the total score and each domain. RESULTS: Twenty-five patients and three clinical experts participated in CE which resulted in a 27-item validation ready measure. In the validation study (N = 124), nine draft items were dropped due to high missing data and/or high correlations between items. Factor analysis revealed three domains: physical functioning/mobility, sleep, and daily activities. RMT confirmed adequate item fit and placement within domains. Internal consistency ranged from 0.91 to 0.96 and test-retest from 0.84 to 0.91. All prespecified hypotheses for convergent and discriminant validity were met. CONCLUSIONS: CE and psychometric results provide evidence that the final, 18-item DPNPI is a reliable and valid PRO measure of disease impacts and treatment for DPNP. Further validation work should include responsiveness assessment.

Dosimetric evaluation of adaptive planning for five-fraction gynecologic template-based interstitial brachytherapy.

PURPOSE: The purpose of this work was to evaluate whether inter-fraction imaging and replanning enhance treatment delivery adherence to clinical planning objectives in the context of a 5-fraction template-based interstitial brachytherapy (TISB) approach for gynecologic cancer treatment. METHODS AND MATERIALS: This retrospective study analyzed nineteen patients who underwent 5 fractions of interstitial brachytherapy over 3 days using the Syed-Neblett template. A verification CT scan was acquired for applicator assessment and reviewed by a radiation oncologist and medical physicist before each fraction. Eleven patients required replanning at least once during the treatment course. Replanning on the verification CT scan consisted of generating new target and organ-at-risk contours, digitizing catheter positions, and optimizing source dwell times to meet planning objectives. Dwell times and positions from the initial treatment plan were evaluated on the new contours to assess the dose that would have been delivered without replanning (nonadapted). Significance of nonadapted versus adapted dose differences were evaluated using a 2-sided Wilcoxon sum rank test. RESULTS: The average (min, max) change in dose (Gy) between the clinically delivered plans and the nonadapted plans were HR-CTV D90%: -6.5 (-0.6, -15.1), HR-CTV D98%: -6.5 (-0.4, -12.6), Bladder D2cc: -0.5 (0.0, -2.8), Bowel D2cc: -0.8 (0.0, -3.2), Rectum D2cc: -1.1 (0.0, -11.5), Sigmoid D2cc: -1.4 (-0.1, -5.4). Dosimetric changes in HR-CTV coverage were significantly improved with replanning while organ-at-risk differences were nonsignificant (p > 0.05). Fraction 3 was the most common fraction indicated for replanning. CONCLUSIONS: Replanning template-based interstitial brachytherapy can improve target coverage and adherence to planning goals.

Enhancing Precision in Detecting Severe Immune-Related Adverse Events: Comparative Analysis of Large Language Models and International Classification of Disease Codes in Patient Records.

PURPOSE: Current approaches to accurately identify immune-related adverse events (irAEs) in large retrospective studies are limited. Large language models (LLMs) offer a potential solution to this challenge, given their high performance in natural language comprehension tasks. Therefore, we investigated the use of an LLM to identify irAEs among hospitalized patients, comparing its performance with manual adjudication and International Classification of Disease (ICD) codes. METHODS: Hospital admissions of patients receiving immune checkpoint inhibitor (ICI) therapy at a single institution from February 5, 2011, to September 5, 2023, were individually reviewed and adjudicated for the presence of irAEs. ICD codes and an LLM with retrieval-augmented generation were applied to detect frequent irAEs (ICI-induced colitis, hepatitis, and pneumonitis) and the most fatal irAE (ICI-myocarditis) from electronic health records. The performance between ICD codes and LLM was compared via sensitivity and specificity with an α = .05, relative to the gold standard of manual adjudication. External validation was performed using a data set of hospital admissions from June 1, 2018, to May 31, 2019, from a second institution. RESULTS: Of the 7,555 admissions for patients on ICI therapy in the initial cohort, 2.0% were adjudicated to be due to ICI-colitis, 1.1% ICI-hepatitis, 0.7% ICI-pneumonitis, and 0.8% ICI-myocarditis. The LLM demonstrated higher sensitivity than ICD codes (94.7% v 68.7%), achieving significance for ICI-hepatitis (P < .001), myocarditis (P < .001), and pneumonitis (P = .003) while yielding similar specificities (93.7% v 92.4%). The LLM spent an average of 9.53 seconds/chart in comparison with an estimated 15 minutes for adjudication. In the validation cohort (N = 1,270), the mean LLM sensitivity and specificity were 98.1% and 95.7%, respectively. CONCLUSION: LLMs are a useful tool for the detection of irAEs, outperforming ICD codes in sensitivity and adjudication in efficiency.

Posttreatment Exposure Rates for 90Y-Microsphere Patients: A Comparison of Products.

There has been a significant increase in the use of 90Y-microspheres in treating liver malignancies. This increase could be seen over the last 30 y, and Food and Drug Administration approval of 2 products-Sirtex SIR-Spheres and Boston Scientific TheraSphere-has helped in the proliferation of these treatments. As the increase in use of both products rose at our institution, there was a need to determine whether there should be special considerations for patients who receive one product compared with patients who receive the other product. This determination was made by measuring exposure rates for several regions of the patient before and after implantation. An independent-samples t test analysis (ɑ = 0.05) was performed for 50 patients (25 TheraSphere and 25 SIR-Spheres) to determine whether the products behaved similarly to the extent that exposure to others was minimized and that as-low-as-reasonably-achievable principles were kept. The results showed that the products exhibited no significant differences in exposure rates, suggesting that no special considerations are needed for the procedure for one product compared with the other.