Machine learning to predict curative multidisciplinary team treatment decisions in oesophageal cancer.

BACKGROUND: Rising workflow pressures within the oesophageal cancer (OC) multidisciplinary team (MDT) can lead to variability in decision-making, and health inequality. Machine learning (ML) offers a potential automated data-driven approach to address inconsistency and standardize care. The aim of this experimental pilot study was to develop ML models able to predict curative OC MDT treatment decisions and determine the relative importance of underlying decision-critical variables. METHODS: Retrospective complete-case analysis of oesophagectomy patients ± neoadjuvant chemotherapy (NACT) or chemoradiotherapy (NACRT) between 2010 and 2020. Established ML algorithms (Multinomial Logistic regression (MLR), Random Forests (RF), Extreme Gradient Boosting (XGB)) and Decision Tree (DT) were used to train models predicting OC MDT treatment decisions: surgery (S), NACT + S or NACRT + S. Performance metrics included Area Under the Curve (AUC), Accuracy, Kappa, LogLoss, F1 and Precision -Recall AUC. Variable importance was calculated for each model. RESULTS: We identified 399 cases with a male-to-female ratio of 3.6:1 and median age of 66.1yrs (range 32-83). MLR outperformed RF, XGB and DT across performance metrics (mean AUC of 0.793 [±0.045] vs 0.757 [±0.068], 0.740 [±0.042], and 0.709 [±0.021] respectively). Variable importance analysis identified age as a major factor in the decision to offer surgery alone or NACT + S across models (p < 0.05). CONCLUSIONS: ML techniques can use limited feature-sets to predict curative UGI MDT treatment decisions. Explainable Artificial Intelligence methods provide insight into decision-critical variables, highlighting underlying subconscious biases in cancer care decision-making. Such models may allow prioritization of caseload, improve efficiency, and offer data-driven decision-assistance to MDTs in the future.

The Oesophageal Cancer Multidisciplinary Team: Can Machine Learning Assist Decision-Making?

BACKGROUND: The complexity of the upper gastrointestinal (UGI) multidisciplinary team (MDT) is continually growing, leading to rising clinician workload, time pressures, and demands. This increases heterogeneity or 'noise' within decision-making for patients with oesophageal cancer (OC) and may lead to inconsistent treatment decisions. In recent decades, the application of artificial intelligence (AI) and more specifically the branch of machine learning (ML) has led to a paradigm shift in the perceived utility of statistical modelling within healthcare. Within oesophageal cancer (OC) care, ML techniques have already been applied with early success to the analyses of histological samples and radiology imaging; however, it has not yet been applied to the MDT itself where such models are likely to benefit from incorporating information-rich, diverse datasets to increase predictive model accuracy. METHODS: This review discusses the current role the MDT plays in modern UGI cancer care as well as the utilisation of ML techniques to date using histological and radiological data to predict treatment response, prognostication, nodal disease evaluation, and even resectability within OC. RESULTS: The review finds that an emerging body of evidence is growing in support of ML tools within multiple domains relevant to decision-making within OC including automated histological analysis and radiomics. However, to date, no specific application has been directed to the MDT itself which routinely assimilates this information. CONCLUSIONS: The authors feel the UGI MDT offers an information-rich, diverse array of data from which ML offers the potential to standardise, automate, and produce more consistent, data-driven MDT decisions.

Analgesia in the Initial Management of Acute Pancreatitis: A Systematic Review and Meta-Analysis of Randomised Controlled Trials.

BACKGROUND: The optimal analgesic strategy for patients with acute pancreatitis (AP) remains unknown. OBJECTIVE: The present systematic review and meta-analysis aims to compare the efficacy of different analgesic modalities trialled in AP. METHODS: A systematic search of PubMed, MEDLINE, EMBASE, CENTRAL, SCOPUS and Web of Science conducted up until June 2021, identified all randomised control trials (RCTs) comparing analgesic modalities in AP. A pooled analysis was undertaken of the improvement in pain scores as reported on visual analogue scale (VAS) on day 0, day 1 and day 2. RESULTS: Twelve RCTs were identified including 542 patients. Seven trial drugs were compared: opiates, non-steroidal anti-inflammatories (NSAIDs), metamizole, local anaesthetic, epidural, paracetamol, and placebo. Across all modalities, the pooled VAS scores showed global improvement from baseline to day 2. Epidural analgesia appears to provide the greatest improvement in VAS within the first 24 h but is equivalent to opiates by 48 h. Within 24 h, NSAIDs offered similar pain-relief to opiates, while placebo also showed equivalence to other modalities but then plateaued. Local anaesthetics demonstrated least overall efficacy. VAS scores for opiate and non-opiate analgesics were comparable at baseline and day 1. The identified RCTs demonstrated significant statistical and methodological heterogeneity in pain-relief reporting. CONCLUSIONS: There is remarkable paucity of level 1 evidence to guide pain management in AP with small datasets per study. Epidural administration appears effective within the first 24 h of AP although infrequently used and featured in only a single RCT. NSAIDs are an effective opiate sparing alternative during the first 24 h.

Chimeric human opsins as optogenetic light sensitisers.

Human opsin-based photopigments have great potential as light-sensitisers, but their requirement for phototransduction cascade-specific second messenger proteins may restrict their functionality in non-native cell types. In this study, eight chimeric human opsins were generated consisting of a backbone of either a rhodopsin (RHO) or long-wavelength-sensitive (LWS) opsin and intracellular domains from Gq/11-coupled human melanopsin. Rhodopsin/melanopsin chimeric opsins coupled to both Gi and Gq/11 pathways. Greater substitution of the intracellular surface with corresponding melanopsin domains generally showed greater Gq/11 activity with a decrease in Gi activation. Unlike melanopsin, rhodopsin and rhodopsin/melanopsin chimeras were dependent upon exogenous chromophore to function. By contrast, wild-type LWS opsin and LWS opsin/melanopsin chimeras showed only weak Gi activation in response to light, whilst Gq/11 pathway activation was not detected. Immunocytochemistry (ICC) demonstrated that chimeric opsins with more intracellular domains of melanopsin were less likely to be trafficked to the plasma membrane. This study demonstrates the importance of Gα coupling efficiency to the speed of cellular responses and created human opsins with a unique combination of properties to expand the range of customised optogenetic biotools for basic research and translational therapies.

Quantitative multiparametric MRI allows safe surgical planning in patients undergoing liver resection for colorectal liver metastases: report of two patients.

It is not uncommon for clinicians to encounter varying degrees of hepatic steatosis in patients undergoing resection for colorectal liver metastases (CRLM). Magnetic resonance imaging is currently the preferred investigation for identification and pre-operative planning of these patients. An objective assessment of liver quality and degree of steatosis is paramount for planning a safe resection, which is seldom provided by routine MRI sequences. We studied two patients who underwent an additional pre-operative multiparametric MRI scan (LiverMultiScanTM) as a part of an observational clinical trial (HepaT1ca, NCT03213314) to assess the quality of liver. Outcome was assessed in the form of post-hepatectomy liver failure. Both patients (Patient 1 and 2) had comparable pre-operative characteristics. Both patients were planned for an extended right hepatectomy with an estimated future liver remnant of approximately 30%. Conventional preoperative contrast MRI showed mild liver steatosis in both patients. Patient one developed post-hepatectomy liver failure leading to prolonged hospital stay compared to patient two who had uneventful post-operative course. Retrospective evaluation of multiparametric MRI scan revealed findings consistent with fibro-inflammatory disease and steatosis (cT1 829 ms, PDFF 14%) for patient 1 whereas patient two had normal parameters (cT1 735 ms, PDFF 2.4%). These findings corresponded with the resection specimen histology. Multiparametric MRI can objectively evaluate future liver health and volume which may help refine surgical decision-making and improve patient outcomes.

Clinical factors associated with successful embolization of lower gastrointestinal bleeding.

AIM: To develop a model of clinical factors that may predict: (1) technically and clinically successful embolization of a bleeding vessel at digital subtraction angiography (DSA) for lower gastrointestinal bleed (LGIB); (2) a negative DSA in the presence of positive CT-mesenteric angiography (CTMA) for LGIB. METHODS: A retrospective cohort study of all DSAs conducted with intent for embolization for acute LGIB over a 10-year period was undertaken. Pre-procedural and intra-procedural clinical variables were evaluated using uni- and multi-variate analysis. RESULTS: One hundred and twenty-three DSAs were evaluated. Technical success was 81% and clinical success 78% where DSA was positive. Technical success was associated with super-selective approach, contrast extravasation on CT, haemoglobin drop, anatomical source and time from CT to DSA on univariate analysis. On multivariate analysis, time from CT to DSA was significant with a higher success probability within 120 min with different factors being salient depending on degree of delay. Clinical success was only associated with activated partial thromboplastin time (<27.5 s). A negative DSA was associated with anatomical source, haemodynamic stability, platelet count and time from CT to DSA on univariate analysis. The latter three remained so on multivariate analysis. CONCLUSION: A triaging approach to utilizing emergency DSA may be helpful. If prolonged delay between CT and DSA is anticipated, with haemodynamic stability and a near-normal platelet count, the DSA may not be fruitful. Technical success may be more likely if DSA occurs within 120 min. Clinical success may be more likely if activated partial thromboplastin time is within normal range.

Deep brain stimulation in multiple system atrophy mimicking idiopathic Parkinson's disease.

Deep brain stimulation (DBS) is approved for idiopathic Parkinson's disease (IPD) but has a poor evidence base in Parkinson-plus syndromes such as multiple system atrophy (MSA). We describe the clinical and neuropathological findings in a man who was initially diagnosed with IPD, in whom DBS was unsuccessful, and in whom MSA was unexpectedly diagnosed at a subsequent autopsy. This case report highlights that DBS is often unsuccessful in MSA and also demonstrates that MSA can masquerade as IPD, which may explain treatment failure in a small group of patients apparently suffering from Parkinson's disease. Additionally, it also presents a case with an unusually long duration of disease prior to death, comparable only to a handful of other cases in the literature.

The putative effects of green tea on body fat: an evaluation of the evidence and a review of the potential mechanisms.

The increase in the prevalence of obesity in recent years has prompted research into alternative methods of modulating body weight and body fat. The last decade has reflected this with a surge in studies investigating the potential of green tea as a natural agent of weight loss, with a view to confirming and elucidating the mechanisms underlying its effect on the body. Currently, it is widely believed that the polyphenolic components present in green tea have an anti-obesogenic effect on fat homeostasis, by increasing thermogenesis or reducing fat absorption among other ways. The data published to date, however, are inconsistent, with numerous putative modes of action suggested therein. While several unimodal mechanisms have been postulated, a more plausible explanation of the observed results might involve a multimodal approach. Such a mechanism is suggested here, involving simultaneous inhibition of the enzymes catechol-O-methyltransferase, acetyl-CoA carboxylase, fatty acid synthase and impeding absorption of fat via the gut. An evaluation of the available evidence supports a role of green tea in weight loss; however the extent of the effects obtained is still subject to debate, and requires more objective quantification in future research.