Diagnosis Of Cancer Research Articles

Impact of the COVID-19 pandemic on cancer diagnoses, oncological care and cancer patients in Germany: a report from the “COVID & Cancer” workshop 2023 of the German Society for Epidemiology (DGEpi)

PurposeThe COVID-19 pandemic was associated with severe disruptions in healthcare worldwide. Cancer patients are at particular risk of adverse consequences from delays in diagnosis and treatment. To evaluate the available data on the impact of the pandemic on cancer diagnoses, oncological care and patient well-being in Germany, the German Society for Epidemiology (DGEpi) in collaboration with the Epidemiological Cancer Registry of Lower Saxony invited to a workshop on “COVID & Cancer” (held on 26–27 October 2023 in Hanover, Germany). This report provides a summary of the scientific presentations, highlights methodological challenges, and recognises essential evidence gaps.MethodsTwelve studies addressing various aspects in relation to cancer diagnoses, oncological care and patient well-being during the COVID-19 pandemic in Germany and two talks sharing experiences from the UK and the Netherlands were presented at the workshop.Results and conclusionsResults from German cancer registries consistently showed lower number of incident cancer diagnoses among adults during the first months of the pandemic compared to the respective months of the years before the pandemic. Data from the cancer registries of Baden-Württemberg and Lower Saxony found especially for breast cancer a notable drop (by approximately one third) in the numbers of diagnoses during the first restriction period (April-May 2020), during which the nationwide mammography screening programme in Germany was temporarily suspended. Overall, the extent and ways, in which the pandemic had adversely affected cancer diagnoses, oncological care and created service backlogs, is still not adequately understood. The long-term consequences are yet to be determined.

Establishment of potential lncRNA-related hub genes involved competitive endogenous RNA in lung adenocarcinoma

Long non-coding RNAs (lncRNAs) have a notable role in the diagnosis and prognosis of cancer. However, the associations between lncRNA-related hub genes (LRHGs) expression and the corresponding outcomes have not been fully understood in lung adenocarcinoma (LUAD). Here, a total of 71 patients diagnosed with LUAD and 60 healthy volunteers at The First Affiliated Hospital of Huzhou University from April, 2023 to December, 2023 were enrolled in the present study. A LRHGs model was established using least absolute shrinkage and selection operator analyses of The Cancer Genome Atlas-LUAD datasets. The underlying mechanisms of the LRHGs were investigated via Gene Set Enrichment Analysis and Gene Set Variation Analysis. Additionally, the diagnostic role of serum HOXD cluster antisense RNA 2 (HOXD-AS2) was assessed by receiver operating characteristic (ROC) curve analysis. Lastly, TCGA-LUAD samples were divided into high- and low-HOXD-AS2 expression groups based on the median expression. The associations between HOXD-AS2 expression and miR-4538 as well as Calmodulin-Dependent Protein Kinase Type II subunit Beta (CAMK2B) levels were conducted through Pearson correlation analysis. A comprehensive analysis identified 141 differentially expressed lncRNAs between 539 LUAD tissues and 59 normal samples. A prognostic marker for overall survival was established by constructing a predictive signature consisting of 9 LRHGs. Subsequently, 474 LUAD samples were categorized into a high or low-risk group based on the median of the risk score. An independent prognostic model was constructed to confirm the validity of this categorization. Further comparisons of the clinicopathological features and LRHG-related pathways were performed between the two groups. Examinations of LRHG expression in two LUAD clusters and of the association between LRHG expression and immune infiltration were also conducted. HOXD-AS2 expression was shown to be elevated in LUAD tissues compared with matched normal tissues, and the serum HOXD-AS2 level was also notably increased in LUAD samples compared with healthy controls. The results of the ROC analysis indicated that the sensitivity and specificity of HOXD-AS2 were higher than that of cytokeratin-19 fragment (CYFRA21-1), which is a serum marker for LUAD. Pearson analyses indicated that miR-4538 level was negatively associated with HOXD-AS2 expression, but CAMK2B level showed positive correlation in LUAD. The results of the present study therefore indicated that the constructed LRHG model, particularly HOXD-AS2, could independently diagnose and predict the prognosis of LUAD, which suggested the underlying mechanism of the HOXD-AS2/miR-4538/CAMK2B, and might offer efficient strategies for LUAD treatment.

A head-to-head comparison of [68Ga]Ga-DOTA-FGFR1 and [18F]FDG PET/CT in the diagnosis of lung cancer

A head-to-head comparison of [68Ga]Ga-DOTA-FGFR1 and [18F]FDG PET/CT in the diagnosis of lung cancer

Characteristics of Breast Cancer Patients Who Refused Conventional Treatment in a Teaching Hospital in Southwestern Nigeria

Background Patients’ adherence to treatment recommendations is in the form of a spectrum: complete adherence or total rejection of the offered standard treatments. The aim of this study was to describe the characteristics of breast cancer patients who refused standard treatments in order to engender a focused intervention. Methods This was a retrospective study of breast cancer patients from 2017-2022. Data generated were analyzed using SPSS version 23.0. The results were presented using descriptive statistics. Inferential statistics for subgroup analysis was by chi-square test or risk difference with confidence intervals Results Among 343 patients, 86 (25% (95% CI 21-30) did not initiate conventional oncologic treatment. All were females and their age range was 33–83 years (mean, 49.9±12.0 years). The majority of the women were married (73%), educated (79%), low-income earners (69%), rural dwellers (63%) with advanced disease (81%). None of the patients had government insurance. The mean duration of symptoms was 11.9±9.7 months. Among 71patients with vital information, 50 (70% (95%CI 58-71) were confirmed dead in the last 3 years. The probability of death within 3 years was 38% (95%CI 8.0-68%) and was higher among those diagnosed as late-stage disease. Conclusions A large number of patients still fail to accept conventional treatments after breast cancer diagnosis. Understanding the patients’ characteristics may be necessary to recognize situations where active education and support can help patients accept standard treatment thereby prolonging survival.

An open codebase for enhancing transparency in deep learning-based breast cancer diagnosis utilizing CBIS-DDSM data

Accessible mammography datasets and innovative machine learning techniques are at the forefront of computer-aided breast cancer diagnosis. However, the opacity surrounding private datasets and the unclear methodology behind the selection of subset images from publicly available databases for model training and testing, coupled with the arbitrary incompleteness or inaccessibility of code, markedly intensifies the obstacles in replicating and validating the model's efficacy. These challenges, in turn, erect barriers for subsequent researchers striving to learn and advance this field. To address these limitations, we provide a pilot codebase covering the entire process from image preprocessing to model development and evaluation pipeline, utilizing the publicly available Curated Breast Imaging Subset of Digital Database for Screening Mammography (CBIS-DDSM) mass subset, including both full images and regions of interests (ROIs). We have identified that increasing the input size could improve the detection accuracy of malignant cases within each set of models. Collectively, our efforts hold promise in accelerating global software development for breast cancer diagnosis by leveraging our codebase and structure, while also integrating other advancements in the field.

Nanoparticles and their application in the diagnosis of hepatocellular carcinoma

Abstract Most patients are at advanced stages when they are diagnosed with hepatocellular carcinoma, leading to poor prognosis and a low 5-year survival rate. Serological markers, ultrasound, computed tomography, magnetic resonance imaging, positron emission tomography, and liver biopsy are the common clinical diagnostic techniques for liver cancer. Effective interventions in the early stage will be beneficial to improve the prognosis of liver cancer patients and reduce the global burden. Therefore, it is urgent to develop new diagnostic methods to improve the diagnosis and management of liver cancer. Nanotechnology has become a new frontier subject in medical detection along with the application of nanomaterials in the manufacture of drug carriers, diagnostic tools, and therapeutic devices. Many studies have shown that nanoparticles (NPs) can be applied to the clinical diagnosis of liver cancer in combination with existing technologies, providing a new method for the early diagnosis of liver cancer. In this review, we elaborate on the theoretical basis and characteristics of NPs in the diagnosis of liver cancer, and the research progress and prospects of NPs in the diagnosis of liver cancer are summarized.

Social health, activity behaviors, and quality of life among young adult cancer survivors: Protocol for a prospective observational study

Approximately 85,000 adolescent and young adults (AYAs; age 15–39) are diagnosed with cancer in the United States annually. Experiencing a cancer diagnosis as an AYA can substantially impact social connections and social health. This paper describes the design and protocol of an observational study to prospectively assess social health and its association with physical activity and quality of life among AYAs after a cancer diagnosis. The study uses a longitudinal observational design to prospectively explore the relationships between social health and activity behaviors (physical activity and sedentary time) at four clinically significant timepoints over the course of 12 months among AYAs newly diagnosed with cancer. Patients are recruited at three hospitals and surveyed at each time period. Multiple dimensions of social health are assessed including social support, social roles, loneliness, social anxiety, and social networks. A wrist accelerometer is worn for one week at each assessment period. Change in social network structures will be analyzed using egocentric social network analysis. Structural equation models will be fitted to analyze the relationship between social constructs and physical activity. Findings from this study will address gaps in our understanding of the impact of a cancer diagnosis on multiple dimensions of social health for AYAs and the potential role social factors play in physical activity and quality of life. Understanding these processes will inform age-tailored interventions to improve health and quality of life outcomes for this at-risk population.

The role of a Macmillan senior pharmacy technician in a regional cancer centre in Northern Ireland: an evaluation

Abstract Introduction In the UK, up to two-thirds of all cancer patients have at least one comorbidity, with around half having multiple long-term conditions (LTCs).1 Older people with cancer often require multiple medications including cancer-specific treatments, supportive care medications, and medications for pre-existing LTCs. They are therefore at higher risk of the adverse effects of polypharmacy, inappropriate prescribing, drug-drug interactions, and drug-disease interactions. Complexity of medication regimens and treatment burden can cause confusion and medicines mismanagement, thereby influencing a patient's ability to adhere to treatment with the outcome being an increased chance of treatment failure. It is thus essential that medicines be optimised in these patients.2 The Oncology Pharmacy Team based in the North West Cancer Centre, located on the Altnagelvin acute hospital site in the Western Health and Social Care Trust, Northern Ireland, secured Macmillan funding to provide a specialist medicines optimisation service to older patients about to start cancer treatment (the Medicines Optimisation in Older People with Cancer [MOOC] Service). A pharmacy technician and pharmacist were recruited and the service implemented from October 2023. The roles are new and innovative and require extensive evaluation. Aim To evaluate the introduction of the Macmillan senior pharmacy technician into the oncology pharmacy team, analysing impact on medicines optimisation and patient safety. Methods A census approach was taken to include all patients with a new cancer diagnosis and medicines reconciliation (MedRec) completed by the pharmacy technician during the study period. The technician identified and rectified any issues prior to the patient’s first scheduled treatment. Patients with significant medicines-related concerns, or aged ≥65 years and taking ≥8 drugs, were referred to the Macmillan pharmacist for holistic medication review. Data collected included number of MedRecs completed by the pharmacy technician (relative to total completed by the team weekly), cancer diagnosis, medicines-related problems identified, interventions made, other activities completed, and time directed away from the pharmacist. Analysis of patient age and number of medications taken informed whether inclusion criteria for MOOC pharmacist referral were appropriate. Results The specialist medicines optimisation service started in October 2023, however, data collection for the purpose of this study commenced on 1st January and took place over a two-month period.101 patients (aged 69.0 ± 5.7 years, range 57 -74 years)) taking an average of 12.6 medicines (range = 8-18), and recently diagnosed with cancer were reviewed by the pharmacy technician. On average, the technician completed 76.4% of medicines reconciliations each week. 120 medicines-related problems were identified, and 178 interventions made. The service yielded net cost savings of £826.34 (pharmacist time saved). Nine patients were referred to the MOOC pharmacist service; a further five patients did not consent to referral. Discussion / Conclusion Although the evaluation of the MOOC service is ongoing, this interim analysis demonstrated that integration of the pharmacy technician into the oncology team yielded cost savings and reduced and improved provision of patient-centred healthcare and safety. Inclusion criteria for the MOOC pharmacist review were confirmed via analysis of patient age and number of medicines taken.

Spherical DNA Nanomotors Enable Ultrasensitive Detection of Active Enzymes in Extracellular Vesicles for Cancer Diagnosis

Enzymes encapsulated in extracellular vesicles (EV)s hold promise as biomarkers for early cancer diagnosis. However, precise measurement of their catalytic activities within EVs remains a notable challenge. Here, we report an enzymatically triggered spherical DNA nanomotor (EDM) that enables one‐pot, cascaded, and highly sensitive analysis of the activity of EV‐associated or free apurinic/apyrimidinic endonuclease 1 (APE1, a key enzyme in base excision repair) across various biological samples. The EDM capitalizes on APE1‐triggered activation of DNAzyme (Dz) and its autonomous cleavage of substrates to achieve nonlinear signal amplification. Using EDM, we reveal a strong correlation between APE1 activity in EVs and that of their parental cancer cells. Additionally, EV APE1 mirrors the fluctuation of cellular APE1 activity in response to chemotherapy‐induced DNA damage. In a pilot clinical study (n = 63), the EDM‐based assay reveals that more than 80% of active APE1 in serum samples is EV‐encapsulated. Notably, EV APE1 can differentiate early prostate cancer (PCa) patients from healthy donors (HDs) with an overall accuracy of 92%, outperforming free APE1 in sera. We anticipate that EDM will become a versatile tool for quantifying EV‐associated enzymes.

Deep learning method for detecting fluorescence spots in cancer diagnostics via fluorescence in situ hybridization

Fluorescence in Situ Hybridization (FISH) is a technique for macromolecule identification that utilizes the complementarity of DNA or DNA/RNA double strands. Probes, crafted from selected DNA strands tagged with fluorophore-coupled nucleotides, hybridize to complementary sequences within the cells and tissues under examination. These are subsequently visualized through fluorescence microscopy or imaging systems. However, the vast number of cells and disorganized nucleic acid sequences in FISH images present significant challenges. The manual processing and analysis of these images are not only time-consuming but also prone to human error due to visual fatigue. To overcome these challenges, we propose the integration of medical imaging with deep learning to develop an automated detection system for FISH images. This system features an algorithm capable of quickly detecting fluorescent spots and capturing their coordinates, which is crucial for evaluating cellular characteristics in cancer diagnosis. Traditional models struggle with the small size, low resolution, and noise prevalent in fluorescent points, leading to significant performance declines. This paper offers a detailed examination of these issues, providing insights into why traditional models falter. Comparative tests between the YOLO series models and our proposed method affirm the superior accuracy of our approach in identifying fluorescent dots in FISH images.

Spherical DNA Nanomotors Enable Ultrasensitive Detection of Active Enzymes in Extracellular Vesicles for Cancer Diagnosis

Enzymes encapsulated in extracellular vesicles (EV)s hold promise as biomarkers for early cancer diagnosis. However, precise measurement of their catalytic activities within EVs remains a notable challenge. Here, we report an enzymatically triggered spherical DNA nanomotor (EDM) that enables one‐pot, cascaded, and highly sensitive analysis of the activity of EV‐associated or free apurinic/apyrimidinic endonuclease 1 (APE1, a key enzyme in base excision repair) across various biological samples. The EDM capitalizes on APE1‐triggered activation of DNAzyme (Dz) and its autonomous cleavage of substrates to achieve nonlinear signal amplification. Using EDM, we reveal a strong correlation between APE1 activity in EVs and that of their parental cancer cells. Additionally, EV APE1 mirrors the fluctuation of cellular APE1 activity in response to chemotherapy‐induced DNA damage. In a pilot clinical study (n = 63), the EDM‐based assay reveals that more than 80% of active APE1 in serum samples is EV‐encapsulated. Notably, EV APE1 can differentiate early prostate cancer (PCa) patients from healthy donors (HDs) with an overall accuracy of 92%, outperforming free APE1 in sera. We anticipate that EDM will become a versatile tool for quantifying EV‐associated enzymes.

Cervicovaginal lavages uncover growth factors as key biomarkers for early diagnosis and prognosis of endometrial cancer

Endometrial cancer (EC) rates are continuing to rise and it remains the most common gynecologic cancer in the US. Existing diagnostic methods are invasive and can cause pain and anxiety. Hence, there is a need for less invasive diagnostics for early EC detection. The study objective was to evaluate the utility of growth factors collected through minimally invasive cervicovaginal lavage (CVL) sampling as diagnostic and prognostic biomarkers for EC. CVL samples from 192 individuals undergoing hysterectomy for benign or malignant conditions were collected and used to quantify the concentrations of 19 growth and angiogenic factors using multiplex immunoassays. Patients were categorized based on disease groups: benign conditions (n = 108), endometrial hyperplasia (n = 18), and EC (n = 66). EC group was stratified into grade 1/2 endometrial endometrioid cancer (n = 53) and other EC subtypes (n = 13). Statistical associations were assessed using receiver operating characteristics, Spearman correlations and hierarchical clustering. Growth and angiogenic factors: angiopoietin-2, endoglin, fibroblast activation protein (FAP), melanoma inhibitory activity, and vascular endothelial growth factor-A (VEGF-A) were significantly (p < 0.0001) elevated in EC patients. A multivariate model combining 11 proteins with patient age and body mass index exhibited excellent discriminatory potential (area under curve = 0.918) for EC, with a specificity of 90.7% and a sensitivity of 87.8%. Moreover, angiopoietin-2, FAP and VEGF-A significantly (p < 0.05–0.001) associated with tumor grade, size, myometrial invasion, and mismatch repair status. Our results highlight the innovative use of growth and angiogenic factors collected through CVL sampling for the detecting endometrial cancer, showcasing not only their diagnostic potential but also their prognostic value.

Upper and Lower Endoscopic Findings in Mesenteric Panniculitis Patients: A Case-Control Study

Background: The natural history and prognosis of mesenteric panniculitis (MP) are not well-described. Despite referral for colonoscopy being common for this indication, colonoscopy findings in MP patients have not been reported. Therefore, we aimed to describe upper and lower gastrointestinal (GI) endoscopy findings in patients with mesenteric panniculitis, compared to matched controls, to investigate their clinical outcomes including incidence of malignancy and mortality. Methods: Retrospective case–control study was conducted, and included patients who were diagnosed with mesenteric panniculitis according to Coulier radiologic criteria on abdominal computerized tomography between 1/2005 and 12/2019, and followed to 12/2021. The case group was compared to a matched control group without MP on abdominal CT. Clinical data and the upper and lower endoscopies’ reports were reviewed in both groups. We excluded patients who, beyond diagnosis of MP, were also diagnosed with current malignancy, significant intra-abdominal morbidity or inflammatory bowel disease. Results: The initial set of 376 patients with MP, after exclusion, included 187 patients. A total of 56.1% were male, with a mean age 60 ± 15 years. Of them, 74 (39%) patients underwent follow-up CT scans, which demonstrated, in 66 (89.2%) patients, a stable MP without any aggravation. Colonoscopy was performed in 89 MP patients, and 98/187 controls. No significant difference in the colonoscopies’ findings was found between the two groups. Gastroscopy was performed in 84 MP and 79 controls. No case of gastric cancer was found. No statistically significant difference was found in the rate of gastroscopy findings. By the end of the follow-up period, malignancy was diagnosed in four patients of the MP group. None were colon cancer. The mortality rate in the MP group was 3.2%, without a significant difference compared to the controls. None were MP related. Conclusions: MP identified on abdominal CT is not associated with pathologic endoscopy findings or future diagnosis of colon cancer, and also has no impact on mortality rate. Since repeating abdominal CT did not reveal any disease progression, the necessity of follow-up imaging for MP should be carefully reconsidered.

Exploring exosomes: novel diagnostic and therapeutic frontiers in thyroid cancer

In recent years, the incidence of thyroid cancer has surged globally, posing significant challenges in its diagnosis, treatment, and prognosis. Exosomes, as a class of extracellular vesicles, are secreted by nearly all cell types and encapsulate a variety of nucleic acids and proteins reflective of their cell of origin, thereby facilitating critical intercellular communication. Recent advancements in understanding these exosomes have catalyzed their application in oncology, particularly through uncovering their roles in the pathogenesis, diagnosis, and therapy of cancers. Notably, the latest literature highlights the integral role of exosomes in refining diagnostic techniques, enhancing targeted therapies, optimizing radiotherapy outcomes, and advancing immunotherapeutic approaches in thyroid cancer management. This review provides a current synthesis of the implications of exosomes in thyroid cancer tumorigenesis and progression, as well as their emerging applications in diagnosis and treatment strategies. Furthermore, we discuss the profound clinical potential of exosome-based interventions in managing thyroid cancer, serving as a foundational reference for future therapeutic developments.

Equity of access to pathological diagnosis and bronchoscopy for lung cancer in Aotearoa New Zealand

background: Māori are less likely to survive their lung cancer once diagnosed, but it remains unclear whether this is partially driven by poorer access to best-practice diagnostic services. methods: We examined all lung cancer registrations in Aotearoa New Zealand between 2007–2019 (n=27,869) linked to national administrative health datasets and further stratified by ethnicity, tumour type and stage of disease. Using descriptive and regression analyses, we compared ethnic groups in terms of the basis of diagnosis (e.g., histology, cytology), receipt of bronchoscopy and travel distance and time to access bronchoscopy. results: We found no differences in access to a pathological diagnosis between ethnic groups regardless of cancer type or stage. We found that Māori within the cohort were marginally more likely to access bronchoscopy than the majority European group; however, we also found that Māori had a lower odds of living close to the location of their bronchoscopy, and a correspondingly higher odds of living 100–200km (adjusted odds ratio [adj. OR] 1.46, 95% confidence interval [CI] 1.26–1.69) or more than 200km away (1.36, 95% CI 1.15–1.61) than Europeans. conclusion: Interventions that aim to further support Māori to overcome the systematic and cumulative disadvantages in access to cancer care should be broadly supported and resourced.

A causal association between chemokines and the risk of lung cancer: a univariate and multivariate mendelian randomization study

BackgroundObservational studies and experimental evidence have shown that chemokines play important roles in lung cancer development, progression, and treatment. However, few studies have examined the causal association between them.MethodsSummary data of chemokines and lung cancer were obtained from genome-wide association studies. Mendelian randomization (MR) analyses were performed by five methods, Inverse variance weighted (IVW), Weighted median estimation, MR-Egger, Simple mode and Weighted, with IVW as the primary analysis method. Sensitivity analysis was used to assess the reliability of the MR results. Multivariate Mendelian randomization studies were used to infer whether causality was influenced by potential mediators. The expression levels of CCL21 were analyzed by quantitative real-time PCR.ResultsWe found that CCL21 was negatively associated with lung adenocarcinoma risk. CCL25 was positively associated with lung squamous cell carcinoma risk. CCL5 was negatively associated with small cell lung cancer risk. CCL21, CCL24, CCL27, and CCL28 was positively associated with small cell lung cancer risk. After multivariate Mendelian randomization adjustment for smoking behavior, it was found that the effect of CCL25 on lung squamous cell cancer disappeared, and the effect of CCL21 on small cell lung cancer was quite opposite to the univariate. The receiver operating characteristic curve indicated that chemokines had high accuracy in the diagnosis of lung cancer. CCL21 expression levels showed large differences in lung adenocarcinoma cells.ConclusionThese results highlighted the causal effects of chemokines on lung cancer and suggested a mediating role of smoking behavior in the association between chemokines and lung cancer.

A Machine Learning Approach to Early Detection and Malignancy Prediction in Breast Cancer

Breast cancer is the most common cancer among women, making early detection crucial for effective treatment. Traditional diagnostic methods often face limitations, leading to potential errors in diagnosis. This study explores the transformative potential of artificial intelligence (AI) and machine learning (ML) in breast cancer diagnosis, particularly through models like AdaBoost, SVM, Random Forest, and logistic regression. By analyzing key variables—such as age, tumor size, and menopausal status—this research aims to accurately differentiate between malignant and benign lesions. The findings reveal that the AdaBoost model significantly outperforms others, achieving an impressive AUC of 93.60% and a precision rate of 95.65%. This indicates its exceptional ability to accurately classify cases, minimizing false positives and ensuring reliable detection of true positives. With an F1 score of 86.27%, AdaBoost effectively balances precision and recall, positioning it as a valuable tool in clinical settings. Overall, this study underscores the importance of integrating AI-driven approaches in breast cancer diagnosis, enhancing accuracy and improving patient outcomes while reducing unnecessary invasive procedures. The promising results advocate for the adoption of these advanced techniques in healthcare, paving the way for more personalized and effective treatment strategies.

Cancer incidence and mortality among patients with new-onset atrial fibrillation: A population-based matched cohort study

BackgroundUnderstanding the risk of cancer after the diagnosis of another condition can present opportunities for earlier diagnosis. We examined the risk of cancer diagnosis conditional on prior diagnosis of atrial fibrillation (AF). MethodsLinked electronic health records were used to identify patients aged ≥18 with new-onset AF and age-sex-matched controls. Cumulative incidence of and mortality from cancer (overall and cancer-site specific) within three months, three months to five years and beyond five years from diagnosis of AF were examined. Findings were further validated using Mendelian randomisation (MR). ResultsThe cohort included 117,173 patients with new-onset AF and 117,173 matched controls (median age 78). In the first three months, 2.2% of AF patients were diagnosed with cancer vs. 0.47% in controls (relative risk: 4.7 [95%CI 4.2-5.4] in men and 4.4 [95%CI 3.8-5.0] in women). Nearly 80% of cancers related to thoracic or abdominal organs. Differences in cumulative incidence were only evident in women between three months and five years (subdistribution hazard ratio=1.1 [95%CI 1.01-1.12]) and absent in all patients beyond five years. MR analysis did not support the presence of a causal association between AF and major cancer subtypes. ConclusionThere is a large short-term increase in cancer incidence and mortality following new-onset AF. The findings may reflect incidental identification of AF or paraneoplastic manifestation. New-onset AF confers high short-term risk of cancer diagnosis, at levels comparable with symptomatic risk threshold mandating urgent assessment for suspected cancer.

Breaking news of cancer diagnosis

Breaking news of cancer diagnosis

Investigating the Fairness of Deep Learning Models in Breast Cancer Diagnosis Based on Race and Ethnicity

Breast cancer is the leading cancer affecting women globally. Despite deep learning models making significant strides in diagnosing and treating this disease, ensuring fair outcomes across diverse populations presents a challenge, particularly when certain demographic groups are underrepresented in training datasets. Addressing the fairness of AI models across varied demographic backgrounds is crucial. This study analyzes demographic representation within the publicly accessible Emory Breast Imaging Dataset (EMBED), which includes de-identified mammography and clinical data. We spotlight the data disparities among racial and ethnic groups and assess the biases in mammography image classification models trained on this dataset, specifically ResNet-50 and Swin Transformer V2. Our evaluation of classification accuracies across these groups reveals significant variations in model performance, highlighting concerns regarding the fairness of AI diagnostic tools. This paper emphasizes the imperative need for fairness in AI and suggests directions for future research aimed at increasing the inclusiveness and dependability of these technologies in healthcare settings. Code is available at: https://github.com/kuanhuang0624/EMBEDFairModels.