Publications

Major depressive disorder (MDD) is strongly linked to chronic stress, yet its molecular mechanisms remain poorly understood. We hypothesize that chronic stress induces epigenetic alterations in stress-associated genes, contributing to transcriptional dysregulation in the prefrontal cortex (PFC), a key brain region implicated in MDD. To test this, we performed expression levels of stress-associated genes in the PFC of rats subjected to restraint stress, which revealed significant downregulation of CRHR2 and NR3C1, both critical regulators of the hypothalamic-pituitary-adrenal (HPA) axis, as well as a trend to decreased expression of BDNF and TRKB, central mediators of synaptic plasticity. Given the well-established role of miRNAs in gene regulation, we further investigated stress-associated miRNA expression changes and identified miR-20b-3p and miR-425-3p as significantly upregulated in the restraint-stressed group. Bioinformatics analysis predicted miR-20b-3p as a potential regulator of NR3C1, a relationship confirmed through RISC-mediated immunoenrichment, demonstrating increased miR-20b-3p binding to NR3C1 in stress-exposed rats. To confirm this interaction, we performed parallel analyses in cells ectopically expressing miR-20b-3p mimic, inhibitor, and a non-oligo control. Cells mimicking endogenous miR-20b-3p showed upregulated NR3C1 binding enrichment compared to non-transfected controls, while anti-miR-20b-3p treatment repressed its interaction with the NR3C1 3'UTR. These findings suggest that miR-20b-3p actively engages with NR3C1 and plays a regulatory role in its expression. Collectively, these results highlight a potential epigenetic mechanism by which chronic stress alters NR3C1 expression, possibly contributing to HPA axis dysregulation in MDD. The identification of miR-20b-3p as a key regulator of NR3C1 provides new insights into stress-induced molecular changes and suggests potential therapeutic targets for stress-associated psychiatric disorders.

BACKGROUND AND AIM: Gestational diabetes mellitus (GDM), a common pregnancy-related metabolic disorder, often goes undiagnosed until the second trimester, limiting early intervention opportunities. Given the higher prevalence of GDM in India, there is a critical need to investigate metabolomic biomarkers among Asian Indians, who exhibit greater insulin resistance and are predisposed to developing type 2 diabetes at an earlier age. This study aimed to identify early pregnancy metabolomic signatures predictive of GDM.

METHODS: Among 2115 pregnant women from the STratification of Risk of Diabetes in Early pregnancy (STRiDE) study, we performed untargeted metabolomic profiling using UPLC-MS/MS at early pregnancy (< 16 weeks) plasma samples from 100 women-comprising 50 with GDM and 50 normal (without GDM) based on oral glucose tolerance test (OGTT) at 24-28 weeks. Statistical and machine learning approaches, including logistic regression and random forest (RF), were applied to identify GDM-associated metabolites and construct predictive models. Pathway enrichment analysis was conducted using KEGG database annotations.

RESULTS: A total of 49 metabolites were significantly associated with GDM, primarily involving lipid classes such as phosphatidylcholines, sphingomyelins, and triacylglycerols. RF analysis identified a panel of eight metabolites that achieved best predictive performance (AUC 0.880; 95% CI: 0.809-0.951) for GDM. When combined with conventional clinical risk factors, the integrated model showed comparable prediction of GDM with AUC 0.88;: 95% CI: 0.810-0.952). Enrichment analysis highlighted dysregulated pathways including glycerophospholipid and sphingolipid metabolism, autophagy, and insulin resistance.

CONCLUSION: This study demonstrates the utility of early-pregnancy metabolomic profiling for predicting GDM in Indian women. The eight-metabolite panel offers a promising tool for early risk stratification of GDM, warranting validation in diverse populations.

OBJECTIVE: Metabolomic risk factors for dementia are under studied, especially in Latinos. We examined the relationship between plasma metabolomic profiles and a Magnetic-Resonance Imaging (MRI)-based markers of brain aging in a cohort of older adult Puerto Ricans residing in the greater Boston area.

METHODS: We used multiple linear regression, adjusted for covariates, to examine the association between metabolite concentration and MRI-derived brain age deviation. Metabolites were measured at baseline with untargeted metabolomic profiling (Metabolon, Inc). Brain age deviation was calculated at wave 4 (  9 years from Boston Puerto Rican Health Study (BPRHS) baseline) as chronologic age, minus MRI-estimated brain age, representing the rate of biological brain aging relative to chronologic age. We also examined if metabolites associated with brain age deviation were similarly associated with hippocampal volume and global cognitive function.

RESULTS: Several metabolites, including isobutyrylcarnitine, propionylcarnitine, phenylacetylglutamine, phenylacetylcarnitine (acetylated peptides), p-cresol-glucuronide, phenylacetylglutamate, and trimethylamine N-oxide (TMAO) were associated with worse brain aging. Taurocholate sulfate, a bile salt, was marginally associated with better brain aging. Most metabolites with negative associations with brain age deviation also were inversely, although not significantly, associated with hippocampal volume and cognitive function.

CONCLUSION: The metabolites associated with brain aging in this study are generally consistent with prior literature and highlight the potential role of TMAO, BCAA and other microbially derived metabolites in dementia.

PURPOSE: To evaluate the histopathologic effects of resin Yttrium-90 Transarterial Radioembolization (90Y-TARE) on intrahepatic cholangiocarcinoma (iCCA), and correlate dose with outcomes.

MATERIALS AND METHODS: This retrospective, single-center study included adult patients with iCCA treated with 90Y-TARE using resin microspheres. Histopathologic and dosimetry results were evaluated. Imaging response was assessed. Complete pathologic necrosis (CPN) was defined as 100% tumor necrosis, and extensive necrosis as > 90% necrosis. Treatment factors affecting the pathologic necrosis rate were evaluated.

RESULTS: For 18 patients, the median age was 69 years [IQR, 67-71 years], with 56% male, and median tumor size of 6.4 cm [IQR,5.1-8.2 cm]. Surgical resection was performed 127 [IQR, 100-182] days after 90Y-TARE. CPN was achieved in 4/18 (22%) patients, with extensive necrosis in 13/18 (72%) patients. Extensive necrosis rate was higher in patients with a prescribed dose ≥ 180 Gy (10/11 [91%]), compared to patients with a prescribed dose < 180 Gy (3/7 [43%], p = 0.03). Patients treated with 90Y-TARE as a first line treatment had higher odds of achieving extensive necrosis compared to patients who received 90Y-TARE after chemotherapy (OR = 11, p = 0.047). All lesions classified as complete response by mRECIST at the 3-month evaluation exhibited extensive necrosis, corresponding to a sensitivity of 58% and a positive predictive value (PPV) of 100%.

CONCLUSION: 90Y-TARE with resin microspheres is associated with a CPN rate of 22% and extensive necrosis rate of 72% in iCCA, with a prescribed dose of ≥ 180 Gy predicting extensive necrosis.

We summarize a new process for developing algorithm-based recommendations for the ACR. This process is currently applied to the ACR's incidental findings recommendations, and other committees providing evidence-based recommendations may elect to adopt these processes in the future. The prior process relied upon informal consensus and was versatile but more limited in scalability and generalizability. Most importantly, the absence of a formal, evidence-driven process prevented incidental findings and other algorithms from receiving designation as clinical guidelines per the National Academy of Medicine's Trustworthy Guidelines criteria and limited both referrer and policymaker adoption. In response, a committee of key stakeholders was formed with approval of the ACR to develop a new process that would overcome these drawbacks, including members from the ACR's Incidental Findings and Reporting and Data Systems committees, the ACR's Commissions on Quality and Safety and Informatics, academic and private practice settings, as well as ACR staff. Here we present the formal, evidence-driven process for algorithm-based imaging recommendations developed by this committee. This process is generalizable to committees across the ACR.

Medical errors are a significant cause of morbidity and mortality. Literature on common mechanisms to reduce radiological error are reviewed including shifting from general to subspecialized care, improving expertise, instituting shift volume and shift length limits, minimizing non-interpretive tasks, participating in focused educational and multidisciplinary conferences, and increasing practice size. Implementation of shift volume limits and full sub-specialization has the potential to significantly reduce radiologist error rates.

Vaginal bleeding is a relatively common occurrence in the first trimester of pregnancy, but can be distressing for both patients and clinicians. Differential considerations include normal intrauterine pregnancy (IUP), a nonviable IUP, an ectopic pregnancy (EP), or much less commonly gestational trophoblastic disease (GTD). Although the diagnosis of EP, nonviable IUP, and GTD is important, it is also crucial to avoid harming potentially normal pregnancies through early medical or surgical treatment. Fortunately, most diagnoses can be made using a combination of ultrasound (US), serum human chorionic gonadotropin levels, and physical examination. When a diagnosis is in question, serial examinations and close clinical evaluation is paramount. Occasionally, MRI of the pelvis without contrast may be helpful in problem-solving for challenging cases, where grayscale US is limited or when there is high concern for nontubal EP. The American College of Radiology Appropriateness Criteria are evidence-based guidelines for specific clinical conditions that are reviewed annually by a multidisciplinary expert panel. The guideline development and revision process support the systematic analysis of the medical literature from peer reviewed journals. Established methodology principles such as Grading of Recommendations Assessment, Development, and Evaluation or GRADE are adapted to evaluate the evidence. The RAND/UCLA Appropriateness Method User Manual provides the methodology to determine the appropriateness of imaging and treatment procedures for specific clinical scenarios. In those instances where peer reviewed literature is lacking or equivocal, experts may be the primary evidentiary source available to formulate a recommendation.

BACKGROUND: Magnesium's (Mg) role in brain aging is under-studied. We examined associations of dietary and serum Mg with MRI and cognitive measures in a cohort with >10 y follow-up.

METHODS: Adults from the Boston Puerto Rican Health Study who underwent MRI (n = 218 with serum Mg, n = 198 with Mg intake) and cognitive testing (n = 1049 with serum Mg, n = 1363 with Mg intake) were included. Hypomagnesemia was defined as serum Mg < 0.75 mmol/L, and adequate Mg intake as ≥ the Estimated Average Requirement (males : 350 mg/d, females : 265 mg/d). Multivariable linear regression and linear mixed-effects models were used to estimate associations of Mg with brain age gap (z-score), volumes (% of total intracranial volume), and cognitive scores (z-score). Effect modification by diabetes, hypertension, dyslipidemia, calcium: Mg intake ratio, or ApoE4 was evaluated.

RESULTS: In the full sample, hypomagnesemia was not associated with brain MRI measures, but with global cognitive function [β (95% CI) = -0.058 (-0.11, -0.0020)], in fully adjusted models. Adequate Mg intake was associated with volumes of total gray matter [β (95% CI) = 2.83 (0.56, 5.11)], white matter [β (95% CI) = 1.54 (0.22, 2.86)], and multiple subregions, but not cognition. In participants with diabetes, hypomagnesemia was associated with lower volumes of multiple brain regions, which were not observed in participants without diabetes (P  interaction<0.05). Limited or no interaction was detected for other potential modifiers, despite some variation in association direction or magnitude across subgroups.

CONCLUSIONS: Diabetes modified Mg's association with brain volumes. Mg merits more attention among people with diabetes for preventing neurodegeneration.

OBJECTIVE: To determine reader preference for image order and thus, by inference, image timing after contrast administration that maximizes cancer visualization on contrast-enhanced mammography (CEM).

METHODS: This IRB-approved reader study includes consecutive CEMs performed for research or clinical care in patients before a diagnosis of unifocal breast cancer, where the cancer was seen on both craniocaudal (CC) and mediolateral oblique (MLO) recombined images. All CEMs started with the side containing cancer and alternated with the nonaffected side of the same projection. From 2016 to 2018, CC projection was performed first (group 1), and from 2019 to 2020, the MLO projection was performed first (group 2). Five readers evaluated cases for background parenchymal enhancement (BPE) and lesion type. Readers assessed cancer visibility, confidence in margins, and cancer conspicuity using a 5-point Likert scale. Contrast-to-noise (CNR) measurements were also taken.

RESULTS: Seventy-eight female patients were included. Group 1 (CC-first) included 40 patients (51%) and group 2 (MLO-first) included 38 patients (49%). Mean age differed between groups by 5 years (P = .031), otherwise there were no differences in group characteristics. There was an overall preference for earlier-obtained images for cancer visibility, confidence in margins, and lesion conspicuity against BPE (P < .001) and preference for CC projection for lesion conspicuity (P = .045). In 35 instances (35/390, 9%), an individual reader reported a different lesion type on images obtained later, with a majority (28/35, 80%) reporting a less discernible lesion on later-obtained imaging (eg, mass changed to nonmass enhancement).

CONCLUSION: Our study shows significant reader preference for cancer characteristic evaluation of CEM performed at earlier time points.

OBJECTIVE: Understand radiologists' opinions regarding remote breast imaging and determine whether having home workstations is associated with greater job satisfaction or less burnout.

METHODS: A 43-question survey on remote breast imaging was distributed to Society of Breast Imaging members (July 6 to August 2, 2023). Questions regarding job satisfaction and burnout were included. Pearson's chi-squared tests compared demographic variables and responses. Multiple-variable logistic regression assessed associations between home workstations and job satisfaction or burnout.

RESULTS: In total, 424 surveys were completed (response rate 13%, 424/3244). Among the third (31%, 132/424) of breast imaging radiologists with home workstations, top motivations included flexibility/work-life balance (67%; 88/132) and decreased commute time (51%, 67/132). Most felt that working from home improved their efficiency (65%, 86/132). Perceived drawbacks among all breast imaging radiologists included the inability to perform US or physical examination (71%, 300/424) and impaired patient contact (47%, 198/424). Most (57%, 240/424) wished for more breast imaging remote reading opportunities, and one-third (32%, 136/424) saw themselves in a 100% remote reading practice in the future. The majority (60%, 228/388) felt that remote reading would majorly or moderately improve radiologist wellness, but no significant association was found between having home workstations and job satisfaction (P = .301) or burnout (P = .140).

CONCLUSION: The majority of breast imaging radiologists want more opportunities to work remotely, perceiving that it improves work-life balance and efficiency, albeit at the expense of patient contact. However, those currently working from home did not have higher job satisfaction or lower burnout.