Mitigating disparity?: Treatment patterns, survival, and recurrence rates by race, ethnicity, and hospital site across a large urban health system.

Objective: National data have shown worse endometrial cancer (EC) outcomes among racial and ethnic minorities. We aimed to analyze EC patient outcomes within a large urban academic health system, with a focus on patterns of care and recurrence rates. Methods: This was a retrospective chart review of EC patients at three system hospitals from 1/1/07-12/31/17. Demographic and clinical factors, including time from EMB to surgery, rate of chemotherapy completion, persistent or recurrent disease, and palliative care referrals were extracted. Descriptive statistics and survival curves were generated. Analysis was done using SAS version 9.4. Results: Black patients had lower overall survival compared to all others on univariate analysis only (p < 0.0001). Hospital site was associated with OS, with the academic anchor and satellite 1 having higher rates of all-cause mortality compared to satellite 2 (HR 4.68 academic anchor, 95 % CI 1.72-12.76, HR 5.36 satellite 1, 95 % CI 1.85-15.52). Time from EMB to surgery and rates of persistent disease following primary treatment were higher in Black patients. After adjusting for stage and grade, chemotherapy completion rate was significantly associated with race. Palliative care was utilized more for Black than White patients after adjusting for stage and grade (p = 0.005). Conclusions: Racial disparities in EC are caused by a complex web of interconnected factors that ultimately lead to worse outcomes in Black women. While precision medicine has helped to close the gap, social determinants of health should be addressed, and models focusing on the complex interactions between biologic, genetic, and social factors should be utilized.

Deciduous tooth biomarkers reveal atypical fetal inflammatory regulation in autism spectrum disorder.

Atypical regulation of inflammation has been proposed in the etiology of autism spectrum disorder (ASD); however, measuring the temporal profile of fetal inflammation associated with future ASD diagnosis has not been possible. Here, we present a method to generate approximately daily profiles of prenatal and early childhood inflammation as measured by developmentally archived C-reactive protein (CRP) in incremental layers of deciduous tooth dentin. In our discovery population, a group of Swedish twins, we found heightened inflammation in the third trimester in children with future ASD diagnosis relative to controls (n = 66; 14 ASD cases; critical window: -90 to -50 days before birth). In our replication study, in the US, we observed a similar increase in CRP in ASD cases during the third trimester (n = 47; 23 ASD cases; -128 to -21 days before birth). Our results indicate that the third trimester is a critical period of atypical fetal inflammatory regulation in ASD.

Artificial intelligence in neuropathology: deep learning-based assessment of tauopathy.

Accumulation of abnormal tau in neurofibrillary tangles (NFT) occurs in Alzheimer disease (AD) and a spectrum of tauopathies. These tauopathies have diverse and overlapping morphological phenotypes that obscure classification and quantitative assessments. Recently, powerful machine learning-based approaches have emerged, allowing the recognition and quantification of pathological changes from digital images. Here, we applied deep learning to the neuropathological assessment of NFT in postmortem human brain tissue to develop a classifier capable of recognizing and quantifying tau burden. The histopathological material was derived from 22 autopsy brains from patients with tauopathies. We used a custom web-based informatics platform integrated with an in-house information management system to manage whole slide images (WSI) and human expert annotations as ground truth. We utilized fully annotated regions to train a deep learning fully convolutional neural network (FCN) implemented in PyTorch against the human expert annotations. We found that the deep learning framework is capable of identifying and quantifying NFT with a range of staining intensities and diverse morphologies. With our FCN model, we achieved high precision and recall in naive WSI semantic segmentation, correctly identifying tangle objects using a SegNet model trained for 200 epochs. Our FCN is efficient and well suited for the practical application of WSIs with average processing times of 45 min per WSI per GPU, enabling reliable and reproducible large-scale detection of tangles. We measured performance on test data of 50 pre-annotated regions on eight naive WSI across various tauopathies, resulting in the recall, precision, and an F1 score of 0.92, 0.72, and 0.81, respectively. Machine learning is a useful tool for complex pathological assessment of AD and other tauopathies. Using deep learning classifiers, we have the potential to integrate cell- and region-specific annotations with clinical, genetic, and molecular data, providing unbiased data for clinicopathological correlations that will enhance our knowledge of the neurodegeneration.