Faecal microbiome-based machine learning for multi-class disease diagnosis.

Systemic characterisation of the human faecal microbiome provides the opportunity to develop non-invasive approaches in the diagnosis of a major human disease. However, shared microbial signatures across different diseases make accurate diagnosis challenging in single-disease models. Herein, we present a machine-learning multi-class model using faecal metagenomic dataset of 2,320 individuals with nine well-characterised phenotypes, including colorectal cancer, colorectal adenomas, Crohn's disease, ulcerative colitis, irritable bowel syndrome, obesity, cardiovascular disease, post-acute COVID-19 syndrome and healthy individuals. Our processed data covers 325 microbial species derived from 14.3 terabytes of sequence. The trained model achieves an area under the receiver operating characteristic curve (AUROC) of 0.90 to 0.99 (Interquartile range, IQR, 0.91-0.94) in predicting different diseases in the independent test set, with a sensitivity of 0.81 to 0.95 (IQR, 0.87-0.93) at a specificity of 0.76 to 0.98 (IQR 0.83-0.95). Metagenomic analysis from public datasets of 1,597 samples across different populations observes comparable predictions with AUROC of 0.69 to 0.91 (IQR 0.79-0.87). Correlation of the top 50 microbial species with disease phenotypes identifies 363 significant associations (FDR < 0.05). This microbiome-based multi-disease model has potential clinical application in disease diagnostics and treatment response monitoring and warrants further exploration.

Can transcranial Doppler ultrasound be used for screening cerebral small vessel diseases in the community?

OBJECTIVES: Elderly persons harbouring severe white matter hyperintensity (WMH), a radiological manifestation of cerebral small vessel disease (SVD), have an increased risk of dementia, stroke and poor functional outcomes. A simple screening tool will enhance their recruitment into preventive trials for SVD. We explored the clinical utility of the pulsatility index (PI) of the middle cerebral artery (MCA), obtained from transcranial Doppler ultrasound (TCD), in identifying severe WMH among community elderly persons with vascular risk factors. METHODS: Three hundred and thirty-one dementia- and stroke-free community elderly subjects with hypertension and/or diabetes mellitus underwent TCD to obtain the MCA PI. The WMH volume on 3.0 Tesla MRI was quantified and normalized to each subject's brain volume. The normalized WMH volumes were classified as low (<14.5 ml, 1 standard deviation [SD] above the mean, 84th percentile) or high (≥14.5 ml). The severity of WMH was also rated visually with the Fazekas score. Logistic regression and receiver-operator characteristics (ROC) analysis were performed to evaluate the association between the MCA PI and the severity of WMH. RESULTS: The MCA PI was not an independent predictor of severe WMH. An MCA PI ≥1.095 detected high normalized WMH volumes with an area under the curve (AUC) of 0.553 (95% CI 0.473-0.633), sensitivity of 0.556, and specificity of 0.523. ROC analysis of the MCA PI in predicting high Fazekas scores yielded similar findings. CONCLUSION: In stroke- and dementia-free elderly persons with vascular risk factors, the MCA PI was unable to identify severe WMH. (Word count: 260).

Stroke caused by small vessel occlusion in a patient taking etoricoxib and thalidomide.

We report a 33-year-old man with seronegative arthritis who had an acute infarct at the left lentiform nucleus while taking etoricoxib and thalidomide regularly. Extensive investigations did not find any evidence of large artery atherosclerosis, vasculitis, cardioembolic source or anti-phospholipid antibodies. While it is possible that a short smoking history, hyperlipidemia, and the use of thalidomide could have contributed to the thrombosis of a small penetrator vessel, we postulated that the prolonged use of etoricoxib is another possible contributing factor.