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.
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.
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).
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.
•A novel deep learning model was developed and validated for improved breast cancer histological grading.•The model uses routine histopathology images and provides independent prognostic value for stratification of the NHG 2 group.•Model-based histological grading offers a cost-effective alternative to molecular profiling for improved risk stratification.
Intestinal microfold (M) cells are specialized epithelial cells that act as gatekeepers of luminal antigens in the intestinal tract. They play a critical role in the intestinal mucosal immune ...response through transport of viruses, bacteria and other particles and antigens across the epithelium to immune cells within Peyer's patch regions and other mucosal sites. Recent studies in mice have demonstrated that M cells are generated from Lgr5+ intestinal stem cells (ISCs), and that infection with Salmonella enterica serovar Typhimurium increases M cell formation. However, it is not known whether and how these findings apply to primary human small intestinal epithelium propagated in an in vitro setting.
Human intestinal crypts were grown as monolayers with growth factors and treated with recombinant RANKL, and assessed for mRNA transcripts, immunofluorescence and uptake of microparticles and S. Typhimurium.
Functional M cells were generated by short-term culture of freshly isolated human intestinal crypts in a dose- and time-dependent fashion. RANKL stimulation of the monolayer cultures caused dramatic induction of the M cell-specific markers, SPIB, and Glycoprotein-2 (GP2) in a process primed by canonical WNT signaling. Confocal microscopy demonstrated a pseudopod phenotype of GP2-positive M cells that preferentially take up microparticles. Furthermore, infection of the M cell-enriched cultures with the M cell-tropic enteric pathogen, S. Typhimurium, led to preferential association of the bacteria with M cells, particularly at lower inoculum sizes. Larger inocula caused rapid induction of M cells.
Human intestinal crypts containing ISCs can be cultured and differentiate into an epithelial layer with functional M cells with characteristic morphological and functional properties. This study is the first to demonstrate that M cells can be induced to form from primary human intestinal epithelium, and that S. Typhimurium preferentially infect these cells in an in vitro setting. We anticipate that this model can be used to generate large numbers of M cells for further functional studies of these key cells of intestinal immune induction and their impact on controlling enteric pathogens and the intestinal microbiome.
Celotno besedilo
Dostopno za:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
An increasing volume of prostate biopsies and a worldwide shortage of urological pathologists puts a strain on pathology departments. Additionally, the high intra-observer and inter-observer ...variability in grading can result in overtreatment and undertreatment of prostate cancer. To alleviate these problems, we aimed to develop an artificial intelligence (AI) system with clinically acceptable accuracy for prostate cancer detection, localisation, and Gleason grading.
We digitised 6682 slides from needle core biopsies from 976 randomly selected participants aged 50–69 in the Swedish prospective and population-based STHLM3 diagnostic study done between May 28, 2012, and Dec 30, 2014 (ISRCTN84445406), and another 271 from 93 men from outside the study. The resulting images were used to train deep neural networks for assessment of prostate biopsies. The networks were evaluated by predicting the presence, extent, and Gleason grade of malignant tissue for an independent test dataset comprising 1631 biopsies from 246 men from STHLM3 and an external validation dataset of 330 biopsies from 73 men. We also evaluated grading performance on 87 biopsies individually graded by 23 experienced urological pathologists from the International Society of Urological Pathology. We assessed discriminatory performance by receiver operating characteristics and tumour extent predictions by correlating predicted cancer length against measurements by the reporting pathologist. We quantified the concordance between grades assigned by the AI system and the expert urological pathologists using Cohen's kappa.
The AI achieved an area under the receiver operating characteristics curve of 0·997 (95% CI 0·994–0·999) for distinguishing between benign (n=910) and malignant (n=721) biopsy cores on the independent test dataset and 0·986 (0·972–0·996) on the external validation dataset (benign n=108, malignant n=222). The correlation between cancer length predicted by the AI and assigned by the reporting pathologist was 0·96 (95% CI 0·95–0·97) for the independent test dataset and 0·87 (0·84–0·90) for the external validation dataset. For assigning Gleason grades, the AI achieved a mean pairwise kappa of 0·62, which was within the range of the corresponding values for the expert pathologists (0·60–0·73).
An AI system can be trained to detect and grade cancer in prostate needle biopsy samples at a ranking comparable to that of international experts in prostate pathology. Clinical application could reduce pathology workload by reducing the assessment of benign biopsies and by automating the task of measuring cancer length in positive biopsy cores. An AI system with expert-level grading performance might contribute a second opinion, aid in standardising grading, and provide pathology expertise in parts of the world where it does not exist.
Swedish Research Council, Swedish Cancer Society, Swedish eScience Research Center, EIT Health.
It has been shown that employing multiple atlas images improves segmentation accuracy in atlas-based medical image segmentation. Each atlas image is registered to the target image independently and ...the calculated transformation is applied to the segmentation of the atlas image to obtain a segmented version of the target image. Several independent candidate segmentations result from the process, which must be somehow combined into a single final segmentation. Majority voting is the generally used rule to fuse the segmentations, but more sophisticated methods have also been proposed. In this paper, we show that the use of global weights to ponderate candidate segmentations has a major limitation. As a means to improve segmentation accuracy, we propose the generalized local weighting voting method. Namely, the fusion weights adapt voxel-by-voxel according to a local estimation of segmentation performance. Using digital phantoms and MR images of the human brain, we demonstrate that the performance of each combination technique depends on the gray level contrast characteristics of the segmented region, and that no fusion method yields better results than the others for all the regions. In particular, we show that local combination strategies outperform global methods in segmenting high-contrast structures, while global techniques are less sensitive to noise when contrast between neighboring structures is low. We conclude that, in order to achieve the highest overall segmentation accuracy, the best combination method for each particular structure must be selected.
BACKGROUND AND PURPOSE—Chronic impairment of the arm and hand is a common consequence of stroke. Animal and human studies indicate that brief bursts of vagus nerve stimulation (VNS) in conjunction ...with rehabilitative training improve recovery of motor function after stroke. In this study, we tested whether VNS could promote generalization, long-lasting recovery, and structural plasticity in motor networks.
METHODS—Rats were trained on a fully automated, quantitative task that measures forelimb supination. On task proficiency, unilateral cortical and subcortical ischemic lesions were administered. One week after ischemic lesion, rats were randomly assigned to receive 6 weeks of rehabilitative training on the supination task with or without VNS. Rats then underwent 4 weeks of testing on a task assessing forelimb strength to test generalization of recovery. Finally, the durability of VNS benefits was tested on the supination task 2 months after the cessation of VNS. After the conclusion of behavioral testing, viral tracing was performed to assess synaptic connectivity in motor networks.
RESULTS—VNS enhances plasticity in corticospinal motor networks to increase synaptic connectivity to musculature of the rehabilitated forelimb. Adding VNS more than doubled the benefit of rehabilitative training, and the improvements lasted months after the end of VNS. Pairing VNS with supination training also significantly improved performance on a similar, but untrained task that emphasized volitional forelimb strength, suggesting generalization of forelimb recovery.
CONCLUSIONS—This study provides the first evidence that VNS paired with rehabilitative training after stroke (1) doubles long-lasting recovery on a complex task involving forelimb supination, (2) doubles recovery on a simple motor task that was not paired with VNS, and (3) enhances structural plasticity in motor networks.
Summary
Immune checkpoint therapy to reverse natural killer (NK) and T cell exhaustion has emerged as a promising treatment in various cancers. While anti‐programmed cell death 1 (PD‐1) pembrolizumab ...has recently gained Food and Drug Administration (FDA) approval for use in recurrent or metastatic cervical cancer, other checkpoint molecules, such as T cell immunoreceptor with immunoglobulin (Ig) and immunoreceptor tyrosine‐based inhibition motif (ITIM) domains (TIGIT) and T cell immunoglobulin and mucin‐domain containing‐3 (Tim‐3), have yet to be fully explored in this disease. We report expression of TIGIT, Tim‐3 and PD‐1 on subsets of peripheral blood NK (CD56dim/negCD16bright/dim/neg and CD56brightCD16dim/neg) and T cells. The percentages of these cells were increased in women with cervical cancer and pre‐malignant lesions. PD‐1+ NK and T cells were likely to co‐express TIGIT and/or Tim‐3. These cells, with an apparently ‘exhausted’ phenotype, were augmented in patients. A subset of cells were also natural killer group 2 member D (NKG2D)‐ and DNAX accessory molecule 1 (DNAM‐1)‐positive. PD‐1int and PD‐1high T cells were notably increased in cervical cancer. Soluble programmed cell death ligand 1 (PD‐L1) was higher in cancer patient blood versus healthy donors and we observed a positive correlation between sPD‐L1 and PD‐1+ T cells in women with low‐grade lesions. Within the cancer group, there were no significant correlations between sPD‐L1 levels and cervical cancer stage. However, when comparing cancer versus healthy donors, we observed an inverse association between sPD‐L1 and total T cells and a correlation between sPD‐L1 and CD56dim NK cells. Our results may show an overview of the immune response towards pre‐cancerous lesions and cervical cancer, perhaps giving an early clue as to whom to administer blocking therapies. The increase of multiple checkpoint markers may aid in identifying patients uniquely responsive to combined antibody therapies.
Overview of immune cells in the blood of patients with cervical cancer and precursor lesions. The co‐expression of checkpoint molecules PD‐1, TIGIT and Tim‐3 describe distinct NK and T cell populations that increase in accordance with disease progression. Tumor expression of checkpoint ligands, or soluble PD‐L1, may influence the development of these cells.
Background
Surgery for catecholamine‐producing tumours can be complicated by intraoperative and postoperative haemodynamic instability. Several perioperative management strategies have emerged but ...none has been evaluated in randomized trials. To assess this issue, contemporary perioperative management and outcome data from 21 centres were collected.
Methods
Twenty‐one centres contributed outcome data from patients who had surgery for phaeochromocytoma and paraganglioma between 2000 and 2017. The data included the number of patients with and without α‐receptor blockade, surgical and anaesthetic techniques, complications and perioperative mortality.
Results
Across all centres, data were reported on 1860 patients with phaeochromocytoma or paraganglioma, of whom 343 underwent surgery without α‐receptor blockade. The majority of operations (78·9 per cent) were performed using minimally invasive techniques, including 16·1 per cent adrenal cortex‐sparing procedures. The cardiovascular complication rate was 5·0 per cent overall: 5·9 per cent (90 of 1517) in patients with preoperative α‐receptor blockade and 0·9 per cent (3 of 343) among patients without α‐receptor blockade. The mortality rate was 0·5 per cent overall (9 of 1860): 0·5 per cent (8 of 517) in pretreated and 0·3 per cent (1 of 343) in non‐pretreated patients.
Conclusion
There is substantial variability in the perioperative management of catecholamine‐producing tumours, yet the overall complication rate is low. Further studies are needed to better define the optimal management approach, and reappraisal of international perioperative guidelines appears desirable.
Antecedentes
La cirugía de los tumores productores de catecolaminas puede complicarse por la inestabilidad hemodinámica intraoperatoria y postoperatoria. Se han propuesto distintas estrategias de manejo perioperatorio, pero ninguna ha sido evaluada en ensayos aleatorizados. Para evaluar este tema, se han recogido los datos de los resultados y del manejo perioperatorio contemporáneo de 21 centros.
Métodos
Veintiún centros aportaron datos de los resultados de los pacientes operados por feocromocitoma y paraganglioma entre 2000‐2017. Los datos incluyeron el número de pacientes con y sin bloqueo del receptor α, las técnicas quirúrgicas y anestésicas, las complicaciones y la mortalidad perioperatoria.
Resultados
Los centros en su conjunto aportaron datos de 1.860 pacientes con feocromocitoma y paraganglioma, de los cuales 343 pacientes fueron intervenidos sin bloqueo del receptor α. La gran mayoría (79%) de las cirugías se realizaron utilizando técnicas mínimamente invasivas, incluido un 17% de procedimientos con preservación de la corteza suprarrenal. La tasa de complicaciones cardiovasculares fue de 5,0% en total; 5,9% (90/1517) en pacientes con bloqueo preoperatorio de los receptores α y 0,9% (3/343) en pacientes no pretratados. La mortalidad global fue del 0,5% (9/1860); 0,5% (8/1517) en pacientes pretratados y 0,3% (1/343) en pacientes no tratados previamente.
Conclusión
Existe una variabilidad sustancial en el manejo perioperatorio de los tumores productores de catecolaminas, aunque la tasa global de complicaciones es baja. Este estudio brinda la oportunidad para efectuar comparaciones sistemáticas entre estrategias de prácticas terapéuticas variables. Se necesitan más estudios para definir mejor el enfoque de manejo óptimo y parece conveniente volver a evaluar las guías internacionales perioperatorias.
Morbidity and mortality rates are difficult to determine in rare diseases like phaeochromocytoma. To date there has been no randomized study of the perioperative management of these patients. Therefore, an international and interdisciplinary effort was made to provide a broad overview of the current management of phaeochromocytoma.
Adverse effects frequent with preoperative blocking
Nerve damage can cause chronic, debilitating problems including loss of motor control and paresthesia, and generates maladaptive neuroplasticity as central networks attempt to compensate for the loss ...of peripheral connectivity. However, it remains unclear if this is a critical feature responsible for the expression of symptoms. Here, we use brief bursts of closed-loop vagus nerve stimulation (CL-VNS) delivered during rehabilitation to reverse the aberrant central plasticity resulting from forelimb nerve transection. CL-VNS therapy drives extensive synaptic reorganization in central networks paralleled by improved sensorimotor recovery without any observable changes in the nerve or muscle. Depleting cortical acetylcholine blocks the plasticity-enhancing effects of CL-VNS and consequently eliminates recovery, indicating a critical role for brain circuits in recovery. These findings demonstrate that manipulations to enhance central plasticity can improve sensorimotor recovery and define CL-VNS as a readily translatable therapy to restore function after nerve damage.
Recovery from serious neurological injury requires substantial rewiring of neural circuits. Precisely-timed electrical stimulation could be used to restore corrective feedback mechanisms and promote ...adaptive plasticity after neurological insult, such as spinal cord injury (SCI) or stroke. This study provides the first evidence that closed-loop vagus nerve stimulation (CLV) based on the synaptic eligibility trace leads to dramatic recovery from the most common forms of SCI. The addition of CLV to rehabilitation promoted substantially more recovery of forelimb function compared to rehabilitation alone following chronic unilateral or bilateral cervical SCI in a rat model. Triggering stimulation on the most successful movements is critical to maximize recovery. CLV enhances recovery by strengthening synaptic connectivity from remaining motor networks to the grasping muscles in the forelimb. The benefits of CLV persist long after the end of stimulation because connectivity in critical neural circuits has been restored.
PDF
