Search

Breast cancer claims 11,400 lives on average every year in the UK, making it one of the deadliest diseases. Mammography is the gold standard for detecting early signs of breast cancer, which can help cure the disease during its early stages. However, incorrect mammography diagnoses are common and may harm patients through unnecessary treatments and operations (or a lack of treatment). Therefore, systems that can learn to detect breast cancer on their own could help reduce the number of incorrect interpretations and missed cases. Various deep learning techniques, which can be used to implement a system that learns how to detect instances of breast cancer in mammograms, are explored throughout this paper. Convolution Neural Networks (CNNs) are used as part of a pipeline based on deep learning techniques. A divide and conquer approach is followed to analyse the effects on performance and efficiency when utilising diverse deep learning techniques such as varying network architectures (VGG19, ResNet50, InceptionV3, DenseNet121, MobileNetV2), class weights, input sizes, image ratios, pre-processing techniques, transfer learning, dropout rates, and types of mammogram projections. This approach serves as a starting point for model development of mammography classification tasks. Practitioners can benefit from this work by using the divide and conquer results to select the most suitable deep learning techniques for their case out-of-the-box, thus reducing the need for extensive exploratory experimentation. Multiple techniques are found to provide accuracy gains relative to a general baseline (VGG19 model using uncropped 512 × 512 pixels input images with a dropout rate of 0.2 and a learning rate of 1 × 10-3) on the Curated Breast Imaging Subset of DDSM (CBIS-DDSM) dataset. These techniques involve transfer learning pre-trained ImagetNet weights to a MobileNetV2 architecture, with pre-trained weights from a binarised version of the mini Mammography Image Analysis Society (mini-MIAS) dataset applied to the fully connected layers of the model, coupled with using weights to alleviate class imbalance, and splitting CBIS-DDSM samples between images of masses and calcifications. Using these techniques, a 5.6% gain in accuracy over the baseline model was accomplished. Other deep learning techniques from the divide and conquer approach, such as larger image sizes, do not yield increased accuracies without the use of image pre-processing techniques such as Gaussian filtering, histogram equalisation and input cropping.

In this study we use artificial intelligence (AI) to categorise endometrial biopsy whole slide images (WSI) from digital pathology as either "malignant", "other or benign" or "insufficient". An endometrial biopsy is a key step in diagnosis of endometrial cancer, biopsies are viewed and diagnosed by pathologists. Pathology is increasingly digitised, with slides viewed as images on screens rather than through the lens of a microscope. The availability of these images is driving automation via the application of AI. A model that classifies slides in the manner proposed would allow prioritisation of these slides for pathologist review and hence reduce time to diagnosis for patients with cancer. Previous studies using AI on endometrial biopsies have examined slightly different tasks, for example using images alongside genomic data to differentiate between cancer subtypes. We took 2909 slides with "malignant" and "other or benign" areas annotated by pathologists. A fully supervised convolutional neural network (CNN) model was trained to calculate the probability of a patch from the slide being "malignant" or "other or benign". Heatmaps of all the patches on each slide were then produced to show malignant areas. These heatmaps were used to train a slide classification model to give the final slide categorisation as either "malignant", "other or benign" or "insufficient". The final model was able to accurately classify 90% of all slides correctly and 97% of slides in the malignant class; this accuracy is good enough to allow prioritisation of pathologists' workload.

Obesity has reached epidemic proportions and is often accompanied by elevated levels of pro-inflammatory cytokines that promote many chronic diseases, including cancer. However, not all obese people develop these diseases and it would be very helpful to identify those at high risk early on so that preventative measures can be instituted. We performed an extensive evaluation of the effects of obesity on inflammatory markers, on innate and adaptive immune responses, and on blood cell composition to identify markers that might be useful in distinguishing those at elevated risk of cancer. Plasma samples from 42 volunteers with a BMI>35 had significantly higher CRP, PGE2, IL-1RA, IL-6 and IL-17 levels than 34 volunteers with normal BMIs. Of the cytokines and chemokines tested, only IL-17 was significantly higher in men with a BMI>35 than women with a BMI>35. As well, only IL-17 was significantly higher in those with a BMI>35 that had type 2 diabetes versus those without type 2 diabetes. Whole blood samples from participants with a BMI>35, when challenged with E. coli, produced significantly higher levels of IL-1RA while HSV-1 challenge resulted in significantly elevated IL-1RA and VEGF, and a non-significant increase in G-CSF and IL-8 levels. T cell activation of PBMCs, via anti-CD3 plus anti-CD28, resulted in significantly higher IFNγ production from volunteers with a BMI>35. In terms of blood cells, red blood cell distribution width (RDW), monocytes, granulocytes, CD4+T cells and Tregs were all significantly higher while, natural killer (NK) and CD8+ T cells were all significantly lower in the BMI>35 cohort, suggesting that obesity may reduce the ability to kill nascent tumor cells. Importantly, however, there was considerable person-to-person variation amongst participants with a BMI>35, with some volunteers showing markedly different values from controls and others showing normal levels of many parameters measured. These person-to-person variations may prove useful in identifying those at high risk of developing cancer.

Gold nanorods (GNRs) are increasingly being investigated for cancer theranostics as they possess features which lend themselves in equal measures as contrast agents and catalysts for photothermal therapy. Their optical absorption spectral peak wavelength is determined by their size and shape. Photothermal therapy using GNRs is typically established using near infrared light as this allows sufficient penetration into the tumour matrix. Continuous wave (CW) lasers are the most commonly applied source of near infrared irradiation on GNRs for tumour photothermal therapy. It is perceived that large tumours may require fractionated or prolonged irradiation. However the true efficacy of repeated or protracted CW irradiation on tumour sites using the original sample of GNRs remains unclear. In this study spectroscopy and transmission electron microscopy are used to demonstrate that GNRs reshape both in vitro and in vivo after CW irradiation, which reduces their absorption efficiency. These changes were sustained throughout and beyond the initial period of irradiation, resulting from a spectral blue-shift and a considerable diminution in the absorption peak of GNRs. Solid subcutaneous tumours in immunodeficient BALB/c mice were subjected to GNRs and analysed with electron microscopy pre- and post-CW laser irradiation. This phenomenon of thermally induced GNR reshaping can occur at relatively low bulk temperatures, well below the bulk melting point of gold. Photoacoustic monitoring of GNR reshaping is also evaluated as a potential clinical aid to determine GNR absorption and reshaping during photothermal therapy. Aggregation of particles was coincidentally observed following CW irradiation, which would further diminish the subsequent optical absorption capacity of irradiated GNRs. It is thus established that sequential or prolonged applications of CW laser will not confer any additional photothermal effect on tumours due to significant attenuations in the peak optical absorption properties of GNRs following primary laser irradiation.

BACKGROUND:Understanding the emergence and spread of multidrug-resistant tuberculosis (MDR-TB) is crucial for its control. MDR-TB in previously treated patients is generally attributed to the selection of drug resistant mutants during inadequate therapy rather than transmission of a resistant strain. Traditional genotyping methods are not sufficient to distinguish strains in populations with a high burden of tuberculosis and it has previously been difficult to assess the degree of transmission in these settings. We have used whole genome analysis to investigate M. tuberculosis strains isolated from treatment experienced patients with MDR-TB in Uganda over a period of four years. METHODS AND FINDINGS:We used high throughput genome sequencing technology to investigate small polymorphisms and large deletions in 51 Mycobacterium tuberculosis samples from 41 treatment-experienced TB patients attending a TB referral and treatment clinic in Kampala. This was a convenience sample representing 69% of MDR-TB cases identified over the four year period. Low polymorphism was observed in longitudinal samples from individual patients (2-15 SNPs). Clusters of samples with less than 50 SNPs variation were examined. Three clusters comprising a total of 8 patients were found with almost identical genetic profiles, including mutations predictive for resistance to rifampicin and isoniazid, suggesting transmission of MDR-TB. Two patients with previous drug susceptible disease were found to have acquired MDR strains, one of which shared its genotype with an isolate from another patient in the cohort. CONCLUSIONS:Whole genome sequence analysis identified MDR-TB strains that were shared by more than one patient. The transmission of multidrug-resistant disease in this cohort of retreatment patients emphasises the importance of early detection and need for infection control. Consideration should be given to rapid testing for drug resistance in patients undergoing treatment to monitor the emergence of resistance and permit early intervention to avoid onward transmission.

MicroRNA (miR) abnormalities play a key role in the pathogenesis of chronic lymphocytic leukemia (CLL). High levels of miR-155 have been detected in human neoplasms, and overexpression of miR-155 has been found to induce lymphoma in mice. High levels of miR-155 were detected in CLL cells and STAT3, which is known to induce miR-21 and miR-181b-1 expression, is constitutively activated in CLL. Given these findings, we hypothesized that STAT3 induces miR-155. Sequence analysis revealed that the miR-155 promoter harbors two putative STAT3 binding sites. Therefore, truncated miR-155 promoter constructs and STAT3 small interfering RNA (siRNA) were co-transfected into MM1 cells. Of the two putative binding sites, STAT3-siRNA reduced the luciferase activity of the construct containing the 700-709 bp STAT3 binding site, suggesting that this site is involved in STAT3-induced transcription. Electrophoretic mobility shift assay confirmed that STAT3 bound to the miR-155 promoter in CLL cells, and chromatin immunoprecipitation and luciferase assay confirmed that STAT3 bound to the 700-709 bp but not the 615-624 bp putative STAT3 binding site in CLL cells. Finally, STAT3-small hairpin RNA downregulated miR-155 gene expression, suggesting that constitutively activated STAT3 binds to the miR-155 gene promoter. Together, these results suggest that STAT3 activates miR-155 in CLL cells.

Meningococcal disease (MCD) is the leading infectious cause of death in early childhood in the United Kingdom, making it a public health priority. MCD most commonly presents as meningococcal meningitis (MM), septicaemia (MS), or as a combination of the two syndromes (MM/MS). We describe the changing epidemiology and clinical presentation of MCD, and explore associations with socioeconomic status and other risk factors. A hospital-based study of children admitted to a tertiary children's centre, Alder Hey Children's Foundation Trust, with MCD, was undertaken between 1977 to 2007 (n = 1157). Demographics, clinical presentations, microbiological confirmation and measures of deprivation were described. The majority of cases occurred in the 1-4 year age group and there was a dramatic fall in serogroup C cases observed with the introduction of the meningococcal C conjugate (MCC) vaccine. The proportion of MS cases increased over the study period, from 11% in the first quarter to 35% in the final quarter. Presentation with MS (compared to MM) and serogroup C disease (compared to serogroup B) were demonstrated to be independent risk factors for mortality, with odds ratios of 3.5 (95% CI 1.18 to 10.08) and 2.18 (95% CI 1.26 to 3.80) respectively. Cases admitted to Alder Hey were from a relatively more deprived population (mean Townsend score 1.25, 95% CI 1.09 to 1.41) than the Merseyside reference population. Our findings represent one of the largest single-centre studies of MCD. The presentation of MS is confirmed to be a risk factor of mortality from MCD. Our study supports the association between social deprivation and MCD.

BACKGROUND: The receptor tyrosine kinase like orphan receptor (ROR)-1 gene is overexpressed in chronic lymphocytic leukemia (CLL). Because Stat3 is constitutively activated in CLL and sequence analysis revealed that the ROR1 promoter harbors gamma-interferon activation sequence-like elements typically activated by Stat3, we hypothesized that Stat3 activates ROR1. METHODOLOGY/PRINCIPAL FINDINGS: Because IL-6 induced Stat3 phosphorylation and upregulated Ror1 protein levels in MM1 cells, we used these cells as a model. We transfected MM1 cells with truncated ROR1 promoter luciferase reporter constructs and found that IL-6 induced luciferase activity of ROR1-195 and upstream constructs. Co-transfection with Stat3 siRNA reduced the IL-6-induced luciferase activity, suggesting that IL-6 induced luciferase activity by activating Stat3. EMSA and the ChIP assay confirmed that Stat3 binds ROR1, and EMSA studies identified two Stat3 binding sites. In CLL cells, EMSA and ChIP studies determined that phosphorylated Stat3 bound to the ROR1 promoter at those two ROR1 promoter sites, and ChIP analysis showed that Stat3 co-immunoprecipitated DNA of STAT3, ROR1, and several Stat3-regulated genes. Finally, like STAT3-siRNA in MM1 cells, STAT3-shRNA downregulated STAT3, ROR1, and STAT3-regulated genes and Stat3 and Ror1 protein levels in CLL cells. CONCLUSION/SIGNIFICANCE: Our data suggest that constitutively activated Stat3 binds to the ROR1 promoter and activates ROR1 in CLL cells.

Subtelomeric regions are often under-represented in genome sequences of eukaryotes. One of the best known examples of the use of telomere proximity for adaptive purposes are the bloodstream expression sites (BESs) of the African trypanosome Trypanosoma brucei. To enhance our understanding of BES structure and function in host adaptation and immune evasion, the BES repertoire from the Lister 427 strain of T. brucei were independently tagged and sequenced. BESs are polymorphic in size and structure but reveal a surprisingly conserved architecture in the context of extensive recombination. Very small BESs do exist and many functioning BESs do not contain the full complement of expression site associated genes (ESAGs). The consequences of duplicated or missing ESAGs, including ESAG9, a newly named ESAG12, and additional variant surface glycoprotein genes (VSGs) were evaluated by functional assays after BESs were tagged with a drug-resistance gene. Phylogenetic analysis of constituent ESAG families suggests that BESs are sequence mosaics and that extensive recombination has shaped the evolution of the BES repertoire. This work opens important perspectives in understanding the molecular mechanisms of antigenic variation, a widely used strategy for immune evasion in pathogens, and telomere biology.

Clinical trial registrationclinicaltrials.gov (NCT02709135).