Improved breast cancer histological grading using deep learning.

BACKGROUND: The Nottingham histological grade (NHG) is a well-established prognostic factor for breast cancer that is broadly used in clinical decision making. However, ∼50% of patients are classified as grade 2, an intermediate risk group with low clinical value. To improve risk stratification of NHG 2 breast cancer patients, we developed and validated a novel histological grade model (DeepGrade) based on digital whole-slide histopathology images (WSIs) and deep learning. PATIENTS AND METHODS: In this observational retrospective study, routine WSIs stained with haematoxylin and eosin from 1567 patients were utilised for model optimisation and validation. Model generalisability was further evaluated in an external test set with 1262 patients. NHG 2 cases were stratified into two groups, DG2-high and DG2-low, and the prognostic value was assessed. The main outcome was recurrence-free survival. RESULTS: DeepGrade provides independent prognostic information for stratification of NHG 2 cases in the internal test set, where DG2-high showed an increased risk for recurrence (hazard ratio [HR] 2.94, 95% confidence interval [CI] 1.24-6.97, P = 0.015) compared with the DG2-low group after adjusting for established risk factors (independent test data). DG2-low also shared phenotypic similarities with NHG 1, and DG2-high with NHG 3, suggesting that the model identifies morphological patterns in NHG 2 that are associated with more aggressive tumours. The prognostic value of DeepGrade was further assessed in the external test set, confirming an increased risk for recurrence in DG2-high (HR 1.91, 95% CI 1.11-3.29, P = 0.019). CONCLUSIONS: The proposed model-based stratification of patients with NHG 2 tumours is prognostic and adds clinically relevant information over routine histological grading. The methodology offers a cost-effective alternative to molecular profiling to extract information relevant for clinical decisions.

Artificial intelligence as the next step towards precision pathology.

Pathology is the cornerstone of cancer care. The need for accuracy in histopathologic diagnosis of cancer is increasing as personalized cancer therapy requires accurate biomarker assessment. The appearance of digital image analysis holds promise to improve both the volume and precision of histomorphological evaluation. Recently, machine learning, and particularly deep learning, has enabled rapid advances in computational pathology. The integration of machine learning into routine care will be a milestone for the healthcare sector in the next decade, and histopathology is right at the centre of this revolution. Examples of potential high-value machine learning applications include both model-based assessment of routine diagnostic features in pathology, and the ability to extract and identify novel features that provide insights into a disease. Recent groundbreaking results have demonstrated that applications of machine learning methods in pathology significantly improves metastases detection in lymph nodes, Ki67 scoring in breast cancer, Gleason grading in prostate cancer and tumour-infiltrating lymphocyte (TIL) scoring in melanoma. Furthermore, deep learning models have also been demonstrated to be able to predict status of some molecular markers in lung, prostate, gastric and colorectal cancer based on standard HE slides. Moreover, prognostic (survival outcomes) deep neural network models based on digitized HE slides have been demonstrated in several diseases, including lung cancer, melanoma and glioma. In this review, we aim to present and summarize the latest developments in digital image analysis and in the application of artificial intelligence in diagnostic pathology.

Validation of risk stratification models in acute myeloid leukemia using sequencing-based molecular profiling.

Risk stratification of acute myeloid leukemia (AML) patients needs improvement. Several AML risk classification models based on somatic mutations or gene-expression profiling have been proposed. However, systematic and independent validation of these models is required for future clinical implementation. We performed whole-transcriptome RNA-sequencing and panel-based deep DNA sequencing of 23 genes in 274 intensively treated AML patients (Clinseq-AML). We also utilized the The Cancer Genome Atlas (TCGA)-AML study (N=142) as a second validation cohort. We evaluated six previously proposed molecular-based models for AML risk stratification and two revised risk classification systems combining molecular- and clinical data. Risk groups stratified by five out of six models showed different overall survival in cytogenetic normal-AML patients in the Clinseq-AML cohort (P-value<0.05; concordance index >0.5). Risk classification systems integrating mutational or gene-expression data were found to add prognostic value to the current European Leukemia Net (ELN) risk classification. The prognostic value varied between models and across cohorts, highlighting the importance of independent validation to establish evidence of efficacy and general applicability. All but one model replicated in the Clinseq-AML cohort, indicating the potential for molecular-based AML risk models. Risk classification based on a combination of molecular and clinical data holds promise for improved AML patient stratification in the future.

The comparative metabonomics of age-related changes in the urinary composition of male Wistar-derived and Zucker (fa/fa) obese rats.

The global metabolite profiles of endogenous compounds excreted in urine by male Wistar-derived and Zucker (fa/fa) obese rats were investigated from 4 to 20 weeks of age using both 1H NMR spectroscopy and HPLC-TOF/MS with electrospray ionisation (ESI). Multivariate data analysis was then performed on the resulting data which showed that the composition of the samples changed with age, enabling age-related metabolic trajectories to be constructed. At 4 weeks it was possible to observe differences between the urinary metabolite profiles from the two strains, with the difference becoming more pronounced over time resulting in a marked divergence in their metabolic trajectories at 8-10 weeks. The changes in metabolite profiles detected using 1H NMR spectroscopy included increased protein and glucose combined with reduced taurine concentrations in the urine of the Zucker animals compared to the Wistar-derived strain. In the case of HPLC-MS a number of ions were found to be present at increased levels in the urine of 20 week old Zucker rats compared to Wistar-derived rats including m/z 71.0204, 111.0054, 115.0019, 133.0167 and 149.0454 (negative ion ESI) and m/z 97.0764 and 162.1147 (positive ion ESI). Conversely, ions m/z 101.026 and 173.085 (negative ion ESI) and m/z 187.144 and 215.103 (positive ion ESI) were present in decreased amounts in urine from Zucker compared to Wistar-derived rats. Metabolite identities proposed for these ions include fumarate, maleate, furoic acid, ribose, suberic acid, carnitine and pyrimidine nucleoside. The utility of applying metabonomics to understanding disease processes and the biological relevance of some of the findings are discussed.

The metabonomics of aging and development in the rat: an investigation into the effect of age on the profile of endogenous metabolites in the urine of male rats using 1H NMR and HPLC-TOF MS.

The effect of aging and development in male Wistar-derived rats on the profile of endogenous metabolites excreted in the urine was investigated using both (1)H NMR spectroscopy and HPLC-TOF MS using electrospray ionisation (ESI). The endogenous metabolites were profiled in samples collected from male rats every two weeks from just after weaning at 4 weeks up to 20 weeks of age. Multivariate data analysis enabled clusters to be visualised within the data according to age, with urine collected at 4 and 6 weeks showing the greatest differences by both analytical techniques. Markers detected by (1)H NMR spectroscopy included creatinine, taurine, hippurate and resonances associated with amino acids/fatty acids, which increased with age, whilst citrate and resonances resulting from glucose/myoinositol declined. A number of ions were detected by HPLC-MS that were only present in urine samples at 4 weeks of age in both positive and negative ESI, with a range of ions, including e.g. carnitine, increasing with age. Age predictions by PLS-regression modelling demonstrated an age-related trend within these data, between 4 and 12 weeks for HPLC-MS and 4-16 weeks for NMR. The possible utility of these techniques for metabonomic investigations of age-related changes in the rat is discussed and the importance of employing suitable control animals in pharmacological and toxicological studies is highlighted.

Temperature-time relationship in collembolan response to chemical exposure.

Effects of temperature on chemical toxicity to a collembolan, Folsomia candida, in relation to time were studied in this experiment. Field soil was used as a test substrate. Collembolans were incubated at three different temperatures (+13, +16, and +19 degrees C) and in two different dimethoate concentrations (1 and 3 mg/kg), clean soil serving as the control. Four destructive samplings were done at 2-week intervals. Dimethoate degradation was also analyzed. Dimethoate 1 mg/kg had a slight effect on both adult growth and reproduction, whereas 3 mg/kg was fatal to F. candida in the soil used. Toxic effects tended to last longer at low temperature than at high temperature, but the differences were not extensive. Temperature was negatively correlated with adult growth but positively correlated with reproduction. The dimethoate degradation rate was similar at all temperatures but differed with the concentration.

Late results of paraoesophageal hiatus hernia repair with fundoplication.

BACKGROUND: Most published reports on results of surgical treatment for paraoesophageal hiatus hernia have been based on patient questionnaires, and seldom included endoscopy or barium meal examinations. METHODS: Eight pure and 14 mixed-type paraoesophageal hernias were evaluated a median of 37 (range 2-241) months after surgical repair. An antireflux procedure was done in 19 cases. Before operation all had endoscopy or barium meal (20 and 19 patients respectively); after operation 19 had endoscopy and 12 also had barium meal examination. Oesophageal 24 h pH monitoring was done in five cases before surgery, and in 11 afterwards. RESULTS: Preoperative symptoms of reflux were reported by 18, and were often accompanied by dysphagia, postprandial vomiting or epigastric pain. Symptoms improved after operation, and 21 of the 22 patients were satisfied with the result. At follow-up examination, a recurrent hernia was found in eight of the 19 patients examined. Four of these hernias were sliding, two were mixed type and two purely paraoesophageal. DISCUSSION: Recurrence of symptoms was associated with persistence of reflux rather than hernia recurrence. Concomitant antireflux procedure is recommended in all operations for mixed-type hiatus hernia, but it should also be considered for purely paraoesophageal hernia if reflux cannot be excluded before operation, or if retro-oesophageal dissection is needed.