Faculty Bibliography

Transformed follicular lymphoma (tFL) is typically associated with chemotherapy resistance and a poor prognosis. There are limited data regarding outcomes after CD19-directed chimeric antigen receptor T-cell (CAR T) therapy in relapsed/refractory (R/R) tFL. A total of 923 adult patients with R/R tFL who received commercial CD19 CAR T therapy between 2017 and 2023 were identified in the Center for International Blood and Marrow Transplant Research registry. Median age was 64 years (range: 30-86) and median prior lines of therapy was 4 (range: 1-18). Most patients (78%) received axicabtagene ciloleucel, with 67% of patients having resistant disease at the time of CAR T infusion. At a median follow-up of 25 months (range: 1-72) from CAR T infusion, the 2-year overall survival (OS) was 57% (95% CI: 53-60) and progression-free survival (PFS) was 43% (95% CI: 40-47). The 2-year cumulative incidences of relapse or progression (rel/prog) and non-relapse mortality (NRM) were 47% (95% CI: 44-51) and 9% (95% CI: 7-11), respectively. The overall response rate to CAR T was 76%, with a complete response rate of 63%. Grade ≥ 3 cytokine release syndrome (CRS) was observed in 7.1% and grade ≥ 3 immune effector cell-associated neurologic syndrome (ICANS) in 21.6% of patients. Multivariable analysis suggested that resistant disease status at the time of CAR T, use of bridging therapy, and high comorbidity index ≥ 3 were associated with inferior PFS and OS. Older age ≥ 60 significantly increased the risk of NRM. Our study suggests that CD19 CAR T is effective and safe for tFL.

There is a growing amount of literature on the benefits of using open-ended questions (OEQs) to assess knowledge in medical education. However, it is unknown how many US medical schools include OEQs in their assessment toolkits and how they are being used. The purpose of this study was to determine if OEQ assessments are an emerging trend in US medical education. We distributed an online survey to assessment leadership at all 156 US accredited allopathic medical schools between September 2022 and April 2024. Questions focused on the use or future interest of OEQs to assess medical knowledge in the pre-clerkship and clerkship curriculum. We calculated descriptive statistics for prevalence and use rates, and completed a conventional content analysis for open-ended comments. Seventy-eight US medical schools completed the survey (50% response rate). Forty schools (51%) reported using OEQs for medical knowledge assessment. OEQs were used during the pre-clerkship (28 schools), clerkship (two schools) or both parts of the curriculum (10 schools). On average, OEQs accounted for 20% of the pre-clerkship and 11% of the clerkship assessments at each school. Schools used OEQs to assess students' understanding, assess certain types of knowledge, and develop students' deeper learning. Representatives at schools not currently using OEQs reported considering using them in the future but expressed concerns about the amount of time needed to implement them. Numerous schools are using OEQs to assess medical knowledge, suggesting that this assessment format is feasible. Institutions can be innovative in their assessments by extending beyond multiple-choice questions and incorporating other question formats, such as OEQs, to fit their educational needs. This study provides a foundation for future research to explore the utility of OEQs and how to overcome the challenges of implementing OEQ assessments.

OBJECTIVE:To examine the associations between patient out-of-pocket (OOP) costs and nonadherence to glucagon-like peptide 1 receptor agonists (GLP-1a), and the consequent impact on adverse outcomes, including hospitalizations and emergency department (ED) visits. RESEARCH DESIGN AND METHODS/METHODS:This retrospective cohort study used MarketScan Commercial data (2016-2021). The cohort included nonpregnant adults aged 18-64 years with type 2 diabetes who initiated GLP-1a therapy. Participants were continuously enrolled in the same private insurance plan for 6 months before the prescription date and 1 year thereafter. Exposures included average first 30-day OOP costs for GLP-1a, categorized into quartiles (lowest [Q1] to highest [Q4]). Primary outcomes were the annual proportion of days covered (PDC) for GLP-1a and nonadherence, defined as PDC <0.8. Secondary outcomes included diabetes-related and all-cause hospitalizations and ED visits 1 year post-GLP-1a initiation. RESULTS:Among 61,907 adults who initiated GLP-1a, higher 30-day OOP costs were associated with decreased adherence. Patients in the highest OOP cost quartile (Q4: $80-$3,375) had significantly higher odds of nonadherence (odds ratio [OR]1.25; 95% CI 1.19-1.31) compared with those in Q1 ($0-$21). Nonadherence was linked to increased incidence rates of diabetes-related hospitalizations or ED visits (incidence rate ratio [IRR] 1.86; 95% CI 1.43-2.42), cumulative length of hospitalization (IRR 1.56; 95% CI 1.41-1.72), all-cause ED visits (IRR 1.38; 95% CI 1.32-1.45), and increased ED-related costs ($69.81, 95% CI $53.54-$86.08). CONCLUSIONS:Higher OOP costs for GLP-1a were associated with reduced adherence and increased rates of adverse outcomes among type 2 diabetes patients.

IMPORTANCE/UNASSIGNED:Hospital course (HC) summarization represents an increasingly onerous discharge summary component for physicians. Literature supports large language models (LLMs) for HC summarization, but whether physicians can effectively partner with electronic health record-embedded LLMs to draft HCs is unknown. OBJECTIVES/UNASSIGNED:To compare the editing effort required by time-constrained resident physicians to improve LLM- vs physician-generated HCs toward a novel 4Cs (complete, concise, cohesive, and confabulation-free) HC. DESIGN, SETTING, AND PARTICIPANTS/UNASSIGNED:Quality improvement study using a convenience sample of 10 internal medicine resident editors, 8 hospitalist evaluators, and randomly selected general medicine admissions in December 2023 lasting 4 to 8 days at New York University Langone Health. EXPOSURES/UNASSIGNED:Residents and hospitalists reviewed randomly assigned patient medical records for 10 minutes. Residents blinded to author type who edited each HC pair (physician and LLM) for quality in 3 minutes, followed by comparative ratings by attending hospitalists. MAIN OUTCOMES AND MEASURES/UNASSIGNED:Editing effort was quantified by analyzing the edits that occurred on the HC pairs after controlling for length (percentage edited) and the degree to which the original HCs' meaning was altered (semantic change). Hospitalists compared edited HC pairs with A/B testing on the 4Cs (5-point Likert scales converted to 10-point bidirectional scales). RESULTS/UNASSIGNED:Among 100 admissions, compared with physician HCs, residents edited a smaller percentage of LLM HCs (LLM mean [SD], 31.5% [16.6%] vs physicians, 44.8% [20.0%]; P < .001). Additionally, LLM HCs required less semantic change (LLM mean [SD], 2.4% [1.6%] vs physicians, 4.9% [3.5%]; P < .001). Attending physicians deemed LLM HCs to be more complete (mean [SD] difference LLM vs physicians on 10-point bidirectional scale, 3.00 [5.28]; P < .001), similarly concise (mean [SD], -1.02 [6.08]; P = .20), and cohesive (mean [SD], 0.70 [6.14]; P = .60), but with more confabulations (mean [SD], -0.98 [3.53]; P = .002). The composite scores were similar (mean [SD] difference LLM vs physician on 40-point bidirectional scale, 1.70 [14.24]; P = .46). CONCLUSIONS AND RELEVANCE/UNASSIGNED:Electronic health record-embedded LLM HCs required less editing than physician-generated HCs to approach a quality standard, resulting in HCs that were comparably or more complete, concise, and cohesive, but contained more confabulations. Despite the potential influence of artificial time constraints, this study supports the feasibility of a physician-LLM partnership for writing HCs and provides a basis for monitoring LLM HCs in clinical practice.

Patients with diabetes are prone to developing cerebrovascular disease (CVD) due to a multitude of factors. Particularly, the hyperglycemic environment is a key contributor to the progression of diabetes-associated complications. However, there is a dearth of knowledge regarding glucose transporter 1 (GLUT1, also known as SLC2A1)-dependent mechanisms responsible for these adverse effects. Here, we revealed the importance of glucose transporter 1 in preserving brain endothelial cell homeostasis beyond regulating glucose uptake. To elucidate the GLUT1-mediated protective mechanism, we used bulk RNA sequencing (RNA-Seq) to analyze the transcriptomic alterations under hyperglycemia and GLUT1-deficient conditions and validated the critical gene changes in cultured human brain endothelial cells and diabetic mouse models. We found that GLUT1 downregulation is linked to increased expression levels of podocalyxin (PODXL) and decreased thioredoxin-interacting protein (TXNIP) within healthy brain endothelial cells incubated with high glucose, demonstrating an antistress response mechanism. Interestingly, brain endothelial cells isolated from diabetic mice no longer showed a similar protection mechanism. Instead, the diabetic endothelial cells are characterized by considerably enriched GLUT1 and TXNIP expression under a hyperglycemic state. GLUT1 overexpression recaptures the diabetic features, such as elevated expression of TXNIP and NOD-like receptor pyrin domain-containing 3 (NLRP3) inflammasome, along with increased IL-1β production and permeability. Our findings of a GLUT1-dependent regulatory mechanism for the endothelium provide a potentially deeper insight into mechanistic shifts that occur due to the diabetic disease state and the pathogenesis of diabetes-associated vascular complications.NEW & NOTEWORTHY Glucose transporter-1 is known for regulating glucose uptake in brain endothelial cells. This study used global transcriptome analysis and diabetic mouse models to reveal the novel role of glucose transporter 1 in regulating brain endothelial cell homeostasis by reducing the inflammation response and increasing the protection mechanism. Importantly, the glucose transporter 1-dependent protection mechanism is compromised in diabetic conditions, which explains why patients with diabetes have a high risk of cerebrovascular diseases.

We present a case highlighting innominate vein reconstruction for resection of Hurthle cell carcinoma with complex vascular invasion. A 69-year-old man presented with a rapidly enlarging neck mass, dysphagia and dysphonia. Workup demonstrated a 11.2 × 7.0 × 6.5 cm Hurthle cell carcinoma invading the oropharynx and superior mediastinum. We proceeded with left thyroid lobectomy and modified left radical neck dissection. Median sternotomy, resection of the left clavicular head, and partial resection of the left manubrium were performed to circumferentially expose the innominate vein. Tumor thrombus was extruded from the innominate vein followed by patch angioplasty, which remains patent 14 months postoperatively.

AIM/OBJECTIVE:This study aimed to explore the efficacy of MRI-based radiomics models, employing various machine learning techniques, in the preoperative prediction of the digital subtraction angiography (DSA) classification of venous malformations (VMs). MATERIALS AND METHODS/METHODS:In this retrospective study, 160 VM lesions from 153 children were categorized into a training set (n=128) and a testing set (n=32). Radiomic features were extracted from preoperative MRI scans. Feature selection was executed using the intraclass correlation coefficient test, z-scores, the K-best method, and the least absolute shrinkage and selection operator. Diverse MRI sequences and machine learning methods underpinned the development of the radiomics models. The models' efficacy was evaluated using receiver operating characteristic curves and the area under the curve (AUC). RESULTS:Out of 4528 radiomic features derived from CET1 and T2 images, 9 features were significantly associated with DSA classification differentiation. The most effective model for predicting VMs' DSA classification incorporated these 9 features and employed a random forest classifier. This model achieved an AUC of 0.917 in the training set and an excellent discrimination AUC of 0.891 in the testing set. CONCLUSION/CONCLUSIONS:The random forest model, utilizing CET1 and T2 sequences, exhibited outstanding predictive performance in the preoperative distinction of VMs' DSA classification.

OBJECTIVES/OBJECTIVE:Pleural effusions often require invasive sampling to establish underlying etiologies. Transthoracic ultrasound (TUS) has shown promise in the diagnostics of pleural effusions; however, there lacks consensus regarding its clinical application. We evaluated the diagnostic utility of specific TUS findings for exudative effusions, specifically complex parapneumonic effusion and empyema. METHODS:Ultrasound-guided pleural effusion drainage cases were retrospectively reviewed at a single university-based medical center from July 2015 to May 2023. Procedure-related images were reviewed for specific ultrasound findings: anechoic, fibrin, septation, loculation, plankton/swirl sign, hematocrit sign, jellyfish sign, and visceral pleura thickening. Exudative or transudative nature, underlying etiology, and other patient data were collected from chart review. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and positive likelihood ratio (LR+) were calculated for these findings, either individually or in combination, to predict nature and specific etiology for pleural effusion. A multivariable logistic regression model was constructed to analyze the association between TUS findings and complicated parapneumonic effusion and empyema. RESULTS:A total of 389 cases included, 252 (64.8%) were exudative and 137 (35.2%) were transudative effusions. Findings of anechoic and jellyfish sign were more common in the transudative group, while septation, loculation, and pleural thickening appeared more commonly in exudative effusion (P < 0.05). Absence of all three signs of fibrin, septation, and loculation had 83% sensitivity (95% CI 77-90) and 78% NPV (95% CI 70-86) for transudative effusion. Septation and loculation had 98% specificity (95% CI 95-100) and 94% PPV (95% CI 88-100) for exudative effusion. Fibrin, loculation, and septation, when found concurrently, had 99% specificity (95% CI 98-100) and 96% PPV (95% CI 88-100) for exudative effusion. Multivariable logistic regression showed the presence of septation (odds ratio [OR] 5.3, 95% CI 1.7-16.3, P = 0.0038) and loculation (OR 3.3, 95% CI 1.1-10.0, P = 0.0327) were each independently associated with the likelihood of complicated parapneumonic and empyema cases. CONCLUSIONS:TUS features of loculation or septation are specific and predictive for exudative pleural effusion. The presence of septation and loculation was each associated with a higher likelihood of complicated parapneumonic effusion or empyema. Further studies are needed to validate diagnostic models that incorporate both TUS and clinical features to predict the nature and etiology of pleural effusions.