Predicting when a patient with advanced cancer is dying is a challenge and currently no prognostic test is available. We hypothesised that a dying process from cancer is associated with metabolic changes and specifically with changes in volatile organic compounds (VOCs). We analysed urine from patients with lung cancer in the last weeks of life by headspace gas chromatography mass spectrometry. Urine was acidified or alkalinised before analysis. VOC changes in the last weeks of life were identified using univariate, multivariate and linear regression analysis; 12 VOCs increased (11 from the acid dataset, 2 from the alkali dataset) and 25 VOCs decreased (23 from the acid dataset and 3 from the alkali dataset). A Cox Lasso prediction model using 8 VOCs predicted dying with an AUC of 0.77, 0.78 and 0.85 at 30, 20 and 10 days and stratified patients into a low (median 10 days), medium (median 50 days) or high risk of survival. Our data supports the hypothesis there are specific metabolic changes associated with the dying. The VOCs identified are potential biomarkers of dying in lung cancer and could be used as a tool to provide additional prognostic information to inform expert clinician judgement and subsequent decision making.

Background: Accurately recognizing that a person may be dying is central to improving their experience of care at the end-of-life. However, predicting dying is frequently inaccurate and often occurs only hours or a few days before death. Methods: We performed urinary metabolomics analysis on patients with lung cancer to create a metabolite model to predict dying over the last 30 days of life. Results: Here we show a model, using only 7 metabolites, has excellent accuracy in the Training cohort n = 112 (AUC = 0·85, 0·85, 0·88 and 0·86 on days 5, 10, 20 and 30) and Validation cohort n = 49 (AUC = 0·86, 0·83, 0·90, 0·86 on days 5, 10, 20 and 30). These results are more accurate than existing validated prognostic tools, and uniquely give accurate predictions over a range of time points in the last 30 days of life. Additionally, we present changes in 125 metabolites during the final four weeks of life, with the majority exhibiting statistically significant changes within the last week before death. Conclusions: These metabolites identified offer insights into previously undocumented pathways involved in or affected by the dying process. They not only imply cancer’s influence on the body but also illustrate the dying process. Given the similar dying trajectory observed in individuals with cancer, our findings likely apply to other cancer types. Prognostic tests, based on the metabolites we identified, could aid clinicians in the early recognition of people who may be dying and thereby influence clinical practice and improve the care of dying patients.

Background: The integration of artificial intelligence (AI) in radiology has the potential to improve diagnostic accuracy and efficiency. Medical students and junior doctors will likely use AI more frequently in the future, making their perceptions essential for identifying educational gaps. Purpose: To explore the perceptions of UK medical students and junior doctors regarding AI in radiology. Material and Methods: A cross-sectional survey was distributed across UK medical schools and foundation programs. A total of 250 responses were analyzed using descriptive statistics and non-parametric tests, focusing on career impact, clinical effectiveness, educational development, and ethical concerns. Results: Most respondents (55.2%) were undeterred by career uncertainties related to AI, with 64% confident that AI would not replace radiologists. Up to 80.6% supported AI's clinical benefits, and 63.2% endorsed its educational integration. However, there were concerns about job displacement and insufficient AI training. Medical students were more worried about job security than junior doctors, while those committed to radiology were less apprehensive and viewed AI as complementary. Conclusion: Educational programs and regulatory frameworks are essential to facilitate AI integration in radiology. Addressing concerns about job displacement and improving AI education will be key to preparing future radiologists for technological advancements.