The Christie School of Oncology: Christie Research Publications Repository
Not a member yet
16373 research outputs found
Sort by
Development of a machine learning algorithm to predict abnormalities in serum phosphate in a large oncology cohort
No full textPURPOSE: Serum phosphate is commonly measured in oncology patients because of the relationship between oncologic conditions and treatments with abnormal phosphate. All patients attending our institution, a large specialist oncology center, have a standardized order set (SOS) measured. This consists of 15 biochemical tests, including serum phosphate. Our aim was to understand if abnormalities in serum phosphate could be predicted, using a machine learning algorithm (MLA) by other interrelated variables in the SOS. METHODS: We trained an XGBoost MLA implemented in Python to predict occurrence of abnormal phosphate (1.78 mmol/L) from other results in the SOS. To train and test this algorithm, we used 481,150 test results for 45,174 patients on blood tests between January 2019 and December 2021, with 5,897 abnormal results. RESULTS: This model was trained and tested on a 70%/30% split (train/test result cohort), achieving an area under the receiver operator curve on the test set of 0.866 (95% CI, 0.857 to 0.875). Assigning a threshold for predictions so the model achieves a sensitivity of 0.924 and a specificity of 0.530 and only performing a phosphate test for results above this threshold, the number of phosphate tests would be reduced from 142,647 to 67,873 in this test set, capturing 1,586 of the total 1,716 abnormal results with a small risk (<0.1%) of missing an abnormal result. The model was further validated on a separate validation cohort between January 2022 and December 2023, achieving similar levels of performance. CONCLUSION: A MLA to optimize testing of phosphate has been developed with high sensitivity. Its application in routine care might result in cost-savings and health care efficiencies. The methodology used to develop our MLA model can be applied to other settings where interrelated variables are measured in SOS
Experiences and needs of patients with sarcoma: a qualitative meta-synthesis
No full textPURPOSE: To systematically identify, summarize, and synthesize qualitative evidence on the experiences and needs of patients living with sarcoma, providing insights into their multidimensional challenges. METHODS: This qualitative meta-synthesis follows the ENTREQ guidelines. Seven electronic databases (PubMed, Web of Science, Scopus, Embase, Cochrane Library, CINAHL, and PsycINFO) were searched up to November 2024. Two reviewers independently conducted the study screening and data extraction. The quality was evaluated using the Joanna Briggs Institute's Qualitative Research Standard Assessment tool. Thematic analysis was used for data synthesis. RESULTS: Twenty-seven studies were included in this meta-synthesis. Three analytical themes emerged: (1) 'Self-perceived health challenges after diagnosis' highlighting the physical and emotional challenges faced by patients; (2) 'Mixed experiences during social interactions' reflecting the complex dynamics in relationships with friends, family, and healthcare providers; and (3) 'Unfriendliness in the society towards sarcomas' emphasizing societal barriers such as employment limitations, financial burdens, and stigma. CONCLUSIONS: Sarcoma patients face significant challenges that extend beyond the individual health condition, impacting their interpersonal relationships and social wellbeing. A comprehensive understanding of the experiences and needs of sarcoma patients on their disease journey can help provide effective management for patients with this rare disease. Addressing these challenges requires a holistic approach involving healthcare providers management, organizational change, community support, and public policy reform
Distant brain failure after stereotactic radiosurgery for brain metastases in patients receiving novel systemic treatments
No full textBACKGROUND: Novel systemic therapies, such as immunotherapy and targeted therapies, have shown better systemic disease control in the last decennium. However, the effect of these treatments on distant brain failure (DBF) in patients with brain metastases (BM) remains a topic of discussion. Improving time to DBF leads to longer overall survival (OS), as is reflected in the brain metastasis velocity (BMV). This study presents real world data about the combined effects of local and systemic treatments on DBF and survival. METHODS: A retrospective consecutive cohort study was conducted. Patients with newly diagnosed BM were included between June 2018 and May 2020. Factors associated with DBF were analyzed in multivariate models. The association between BMV and overall survival was analyzed with linear regression analysis. RESULTS: Three hundred and three patients were included. Two hundred and sixty-two (86%) patients received stereotactic radiotherapy, 41 (14%) awaited in first instance the intracranial effect of newly started or switched systemic treatment. Median time to DBF after radiotherapy was 21 months (95% CI 15-27), median OS was 20 months (IQR 10-36). Receiving immunotherapy or targeted therapy were associated with a lower hazard of DBF, compared with chemotherapy. The presence of > 5 initial BM and progressive or stable extracranial disease were associated with increased DBF. BMV was significantly associated with overall survival. CONCLUSIONS: In this retrospective cohort, patients who received immunotherapy or targeted therapy experienced a reduced risk of DBF in comparison to those treated with chemotherapy. A higher BMV was associated with a decreased OS
HER2 destiny: visual scoring using ASCO/CAP guidelines or precise quantification using A.I?
No full textRecurrence of thymic epithelial tumours (TET) following curative treatment; which patients recur and the implications for follow up
No full textProspective randomized phase-II trial of ipilimumab/nivolumab versus standard of care in non-clear cell renal cell cancer - results of the SUNNIFORECAST trial
No full textBACKGROUND: Non-clear cell renal cell cancers (nccRCCs) are a heterogeneous group of more than 20 different entities, but are rarely included in large, randomized trials. Tyrosine kinase inhibitors with or without immune checkpoint inhibition are considered as a standard of care (SOC), but optimal treatment is not yet defined. We designed the first prospective randomized trial comparing ipilimumab/nivolumab to SOC. PATIENTS AND METHODS: We randomized adult patients with previously untreated advanced or metastatic nccRCC 1:1 to nivolumab 3 mg/kg plus ipilimumab 1 mg/kg every 3 weeks for 4 doses followed by fixed dose nivolumab of 240 mg every 2 weeks or 480 mg every 4 weeks or to SOC. Patients were stratified by histology and by IMDC risk score. Central pathology review was mandatory. The primary endpoint was the overall survival (OS) rate at 12 months, secondary endpoints included median OS, response rate, progression-free survival (PFS), safety and quality of life. RESULTS: In total, 157 patients were assigned to receive ipilimumab/nivolumab, and 152 to SOC. The 12-month survival rate was 78% with ipilimumab/nivolumab [95% confidence interval (CI) 71-84%] compared to 68% with SOC (95% CI 60-75%, P = 0.026). Median OS was 33.2 months versus 25.2 months, P = 0.163 [HR 0.81 (0.61-1.099)]. PFS was similar in both arms [HR 0.99 (0.77-1.28)]. The ORR was 32.8% versus 19.3%. No major differences between papillary and non-papillary RCC subtypes were observed for any endpoint. Exploratory analysis showed a significant OS advantage [HR 0.56 (95% CI 0.37-0.86)] associated with a PD-L1 CPS score ≥1. Treatment discontinuation due to toxicity occurred in 27 patients (17%) with ipilimumab/nivolumab and 13 patients (9%) with SOC. CONCLUSIONS: Ipilimumab/nivolumab demonstrated a significantly longer OS at the 12-month milestone and an acceptable toxicity profile. Our results therefore underline a relevant clinical benefit of ipilimumab/nivolumab in previously untreated nccRCC entities compared to current SOC
Molecular determinants of sotorasib clinical efficacy in KRAS(G12C)-mutated non-small-cell lung cancer
No full textMolecular determinants of KRAS(G12C)inhibitor efficacy in KRAS(G12C)-mutated non-small-cell lung cancer (NSCLC) remain poorly characterized. Here we report one of the largest integrated analyses to date of sotorasib clinical efficacy biomarkers from the phase 2 CodeBreaK 100 and phase 3 CodeBreaK 200 studies. We reveal differential sotorasib activity and relative benefit compared to docetaxel across KRAS(G12C)-mutated NSCLC co-mutational subsets and transcriptional subtypes. We also identify low expression of TTF1 and KEAP1 co-mutations/NRF2 activation as major determinants of sotorasib anti-tumor efficacy and adverse prognostic features. Exploratory analyses highlight potential tumor cell-extrinsic contributors to sotorasib anti-tumor activity and suggest that early on-treatment clearance of KRAS(G12C)- circulating tumor DNA may refine clinical response prediction algorithms. Our findings advance precision medicine for patients with KRAS(G12C)-mutated NSCLC and establish a framework for patient stratification and selection for treatment intensification with rationally applied therapeutic combinations
Pilot testing an adapted version of the registered nursing forecasting (RN4CAST) nurse survey in systemic anti-cancer therapy day units: a mixed-methods feasibility study
No full textOBJECTIVES: Systemic Anti-Cancer Therapy (SACT) day units are increasingly challenged by workload pressures and workforce shortages. The Registered Nursing Forecasting (RN4CAST) survey, originally developed to predict workforce needs in acute and geriatric inpatient units, was adapted to create the RN4CAST-SACT-D survey specifically for SACT day units. This study aimed to pilot test and evaluate the feasibility of the adapted RN4CAST-SACT-D survey. METHODS: A convergent mixed-methods feasibility study was conducted, incorporating an embedded qualitative component through semi-structured interviews. Five SACT day units were invited to participate, with all nurses working on the designated census day invited to complete the RN4CAST-SACT-D survey, followed by a semi-structured interview. Additionally, patients attending for chemotherapy on the census day were invited to complete the Patient-Reported Chemotherapy Indicators of Symptoms and Experience (PR-CISE) survey. RESULTS: A total of 43 nurses completed the RN4CAST-SACT-D survey, yielding a response rate of 69.3%. Of these, 12 nurses participated in semi-structured interviews. Additionally, 172 patients completed the PR-CISE survey, with a response rate of 50%. The study identified key challenges and provided recommendations to refine the survey methodology, informed by both survey responses and interview insights. CONCLUSION: The findings suggest that the RN4CAST-SACT-D survey is a feasible tool for assessing workforce-related factors in SACT day units. While the survey methodology was generally well-received, certain aspects require further refinement and testing to enhance its applicability and effectiveness. IMPLICATIONS FOR NURSING PRACTICE: This study highlighted that current staffing levels in SACT day units are largely based on historical data, rather than on present-day needs. This underscores the urgent need for research specifically focused on SACT day unit settings
Inferring binding specificities of human transcription factors with the wisdom of crowds
No full textDNA motif discovery and, particularly, computational modeling of transcription factor binding motifs, has been a mecca of algorithmic bioinformatics for several decades. Here, we report the results of the largest open community challenge in Inferring BInding Specificities (IBIS), where participants all over the world were invited to construct binding specificity models from multi-assay experimental data for poorly studied human transcription factors. The submissions were rigorously tested against a rich held-out dataset. Benchmarking demonstrated a consistent advantage of properly designed deep learning models over traditional positional weight matrices and other machine learning methods. Yet, the positional weight matrices displayed a surprisingly strong performance out of the box, being only slightly behind the best deep learning models. A post-challenge assessment of a selection of other deep learning methods further solidified this finding. IBIS highlights the power of benchmarking in finding adequate DNA motif representations, emphasizes the pros and cons of various machine learning methods applied to DNA motif modeling, and establishes a rich dataset, benchmarking protocols, and computational framework for a fair cross-platform evaluation of future models of transcription factor binding motifs in DNA sequences
