Publications

BACKGROUND: This pilot study evaluated the feasibility of integrating an immunosuppressant area under the concentration-time curve (AUC) monitoring for tacrolimus and mycophenolic acid (MPA) into pediatric kidney transplantation care.

METHODS: Dedicated test codes, an AUC requisition form, and a coordinated laboratory sampling process were established for tacrolimus and MPA AUC. AUC was calculated using the ISBA 3.0 Bayesian pharmacokinetic platform. AUC results were correlated with doses, trough concentrations (C0), 3 h post-dose concentrations (C3h), and clinical outcomes.

RESULTS: The AUC protocol was successfully integrated into the routine clinical workflow. Tacrolimus AUC showed correlations with dose (r = 0.85) and C0 (r = 0.82); similarly, MPA AUC showed correlation with dose (r = 0.61) and C3h (r = 0.65). Of the 21 Tacrolimus AUC measurements, 76% were within the target range, and 24% were below the range. For MPA AUC measurements, 65% (13/20) were within the target range, 5% (1/20) were below the range, and 30% (6/20) were above the range. Following individual AUC measurements, the tacrolimus dose was adjusted after 43% (9/21) of measurements, and the mycophenolate mofetil (MMF) dose was adjusted after 50% (10/20) of measurements.

CONCLUSION: This AUC pilot study demonstrated the feasibility of integrating AUC-guided monitoring into the routine management of pediatric kidney transplant recipients.

INTRODUCTION: Dairy consumption has been linked to the development of autoimmune diseases. We aimed to examine the association between dairy intake and risk of incident inflammatory bowel disease.

METHODS: We conducted a prospective cohort study of 197,763 participants without a baseline diagnosis of inflammatory bowel disease in 1986 in Nurses' Health Study, 1991 in Nurses' Health Study II (NHSII), and 1986 in Health Professionals Follow-up Study. Data on dairy intake were collected every 2-4 years using a validated semi-quantitative food frequency questionnaire and modeled according to quintiles for total intake and quartiles for components of dairy. We used Cox proportional hazard modeling to estimate adjusted hazard ratios and 95% CIs.

RESULTS: Through the end of follow-up in 2016 in Nurses' Health Study and Health Professionals Follow-up Study, and 2017 in Nurses' Health Study II, we documented 347 Crohn's disease cases and 428 cases of ulcerative colitis (UC). In our primary analysis, we observed an inverse association between baseline dairy intake (P trend = 0.04) and risk of UC (adjusted hazard ratio of 0.72 [95% CI 0.52-1.00 comparing extremes of quintiles]). Among dairy components, baseline yogurt consumption (hazard ratio = 0.70; 95%CI 0.5-0.99; P trend = 0.05) was most strongly associated with decreased risk of UC. There was no consistent association between dairy intake and risk of Crohn's disease.

DISCUSSION: In 3 large prospective cohort studies, we observed a suggestive inverse association between baseline dairy intake, particularly from yogurt, and risk of UC. Future studies are needed to confirm these results.

Developing cells undergo extensive metabolic adaptations to support growth and differentiation. Here, using spatially resolved mass spectrometry imaging and stable isotope tracing, we systematically investigate metabolic remodeling in mouse brains at postnatal day 14 and day 28, a period coinciding with the transition from a maternal milk diet to solid food. Untargeted metabolomics reveals global shifts in lipid composition, and region-specific remodeling of central energy metabolism, including increased glycolytic intermediates in grey matter-enriched regions and a global decrease in tricarboxylic acid (TCA) cycle metabolites after weaning. Despite these marked changes in metabolite levels, the glucose incorporation rate remains constant across these developmental stages. Notably, weaning mice onto a milk-replacement diet demonstrates that the observed metabolic adaptations are largely diet-independent. Together, our data suggest that postnatal brain metabolic remodeling is an intrinsically programmed feature of maturation providing region-specific metabolic reorganization to support developmental demands.

Human leukocyte antigen (HLA)-bound tumor peptides can be routinely isolated from cancer samples and identified using mass spectrometry (MS). However, MS approaches can be stochastic or rely on spectral libraries, which are not customarily available for individual-specific peptides, thus limiting the ability to discover novel peptides. Here, we introduce Pepyrus, which generates user-defined, individual-specific or disease-specific peptide libraries in Escherichia coli to improve the sensitivity and confidence of MS peptide identification, including lowly abundant neoantigens. Using Pepyrus-generated peptide libraries paired with an HLA-specific data-independent acquisition strategy, we recover >75% of the expected sequences per single injection for libraries of >10,000 peptides and identify 0.1 fmol of spiked-in peptides in a complex background. We apply Pepyrus to create personalized libraries, facilitating identification of clinically relevant HLA peptides, including several novel peptides from cell lines derived from persons with melanoma and renal cell carcinoma. Pepyrus enables identification of rare HLA-bound peptides and provides the ability to generate large training datasets to improve spectra, retention time and ion mobility prediction tools.

Neurotropic viruses invade neural tissues, resulting in severe diseases such as poliomyelitis, rabies, herpesviral encephalitis, and viral meningitis. Given this neurotropism, we investigated whether the infection of the host by these viruses is under circadian control. In this study, we found that the expression of most neurotropic virus receptors exhibits rhythmicity across cells, cerebral organoids, and animal models, with host cell susceptibility modulated by the circadian clock. We identified E2F8 as a clock-controlled gene that mediates the indirect regulation of the circadian clock on neurotropic viruses. Notably, E2F8 regulated the expression of core clock components by binding directly to the promoters of REV-ERBα and PER2, suggesting its role as a potential modulator of circadian rhythms. Additionally, we revealed a seldom-recognized viral strategy to accelerate viral replication in the host: rabies virus disrupts the host circadian clock system primarily through its glycoprotein hijacking the E3 ubiquitin ligase HUWE1 to inhibit proteasomal degradation of REV-ERBα. These findings increase our understanding of the interactions between circadian systems and neurotropic viral dynamics and highlight the potential of chronotherapy for improved antiviral treatments.

Background and objective Large language models (LLMs) are increasingly being explored as adjuncts to medical education; however, comparative data on the performance and error patterns of newer models on standardized licensing-style questions remain limited. This study evaluated two advanced large language models (LLMs) - Gemini-Pro and GPT-4 Turbo - on the National Board of Medical Examiners (NBME) Step 1-style multiple-choice questions to assess accuracy, reasoning quality, and common failure modes relevant to exam preparation and clinical reasoning training. Methods A total of 112 NBME Step 1 questions were collected; seven image- or table-dependent items were excluded, yielding 105 text-only questions. Prompts were standardized to include the clinical stem, query, and answer choices and were submitted via Python API to Gemini-Pro and GPT-4 Turbo. Outputs were independently adjudicated by two third-year medical students and one board-certified physician using binary accuracy scoring and structured evaluation of reasoning features (logical reasoning, internal information use, and external knowledge application). Incorrect responses were categorized as logical, informational, or statistical errors. Comparative analyses included raw accuracy calculations and chi-square testing of reasoning-feature distributions. Results GPT-4 Turbo achieved 90.99% accuracy on the January 2024 NBME Step 1 question set, substantially outperforming Gemini-Pro (54.46%). GPT-4 Turbo demonstrated fewer errors overall, with lower logical (16%), informational (4%), and statistical (4%) error rates compared with older baselines reported in the study (e.g., GPT-3.5 logical errors 42%). GPT-4 Turbo incorporated external information in 76% of correct responses versus 25% for Gemini-Pro, and differences in performance metrics between GPT-4 Turbo and Gemini-Pro were statistically significant (p < 0.05). Conclusions GPT-4 Turbo markedly outperformed Gemini-Pro on text-based NBME Step 1 questions, showing higher accuracy, stronger reasoning consistency, and fewer logical/informational failures. These findings support GPT-4 Turbo's potential role as a high-yield supplementary tool for Step 1-style learning and feedback, while underscoring the need for continued refinement and cautious, supervised integration of LLMs into medical education, given persistent (though reduced) error rates.

COVID-19 has affected millions of individuals worldwide, yet the neuro-ophthalmic consequences among survivors remain incompletely characterized. In this case report, we describe a case of unilateral optic neuritis identified in the emergency department using ultrasound in a previously healthy young woman shortly after confirmed SARS-CoV-2 infection. This case highlights the importance of recognizing neuro-ophthalmic sequelae of COVID-19, outlines the diagnostic evaluation, and demonstrates clinical improvement with corticosteroid therapy.