MIRS: an AI scoring system for predicting the prognosis and therapy of breast cancer

Current scoring systems for prognosis of breast cancer are available but usually consider only one prognostic feature. We aim to develop a novel prognostic scoring system based on both immune-infiltration and metastatic features to not only assess the patient prognoses more accurately but also guide therapy for patients with breast cancer. Computational immune-infiltration and gene profiling analysis identified a 12-gene panel firstly characterizing immune-infiltrating and metastatic features. Neural network model yielded a precise prognostic scoring system called metastatic and immunogenomic risk score (MIRS). The influence of MIRS on the prognosis and therapy of breast cancer was then comprehensively investigated. MIRS significantly stratifies patients into high risk-group (MIRShigh) and low risk-group (MIRSlow) in both training and test cohorts. The MIRSlow patients exhibit significantly improved survival rate compared with MIRShigh patients. A series of analyses demonstrates that MIRS can well characterize the metastatic and immune landscape of breast cancer. Further analysis on the usage of MIRS in chemotherapy suggests that MIRShigh patients may benefit from three chemotherapeutic drugs (Cisplatin, Tamoxifen and Vincristine). Higher immune infiltration and significantly prolonged survival are observed in MIRSlow patients, indicating a better response in immune checkpoint inhibitor therapy. Our analysis demonstrates that MIRS could effectively improve the accuracy of prognosis for patients with breast cancer. Also, MIRS is a useful webtool, which is deposited at https://lva85.github.io/MIRS/, to help clinicians in designing personalized therapies for patients with breast cancer.

Management of Fetal Supraventricular Tachycardia: Case Series from a Tertiary Perinatal Cardiac Center

<b><i>Background:</i></b> Fetal supraventricular tachycardia is a relatively uncommon cardiac rhythm abnormality which is often associated with adverse perinatal outcomes if untreated. Although there are several treatment modalities and protocols in use globally, there is no consensus as to the most effective antiarrhythmic to manage this condition. <b><i>Aim:</i></b> This study aimed to evaluate perinatal outcomes following prenatal maternal therapy for fetal supraventricular tachycardia. <b><i>Materials and Methods:</i></b> This was a 20-year retrospective cohort study. Institutional records were reviewed for antenatal therapy choice and maternal and fetal outcomes. <b><i>Results:</i></b> Sixty-nine cases met diagnostic criteria for fetal SVT, of which 56 (81%) received maternal antiarrhythmic therapy. Digoxin was the most common, but least effective, first-line therapy in 28 patients, achieving successful rate reversion in 35.7%. Thirty-one patients (55%) required second-line therapy, and this was most successful with digoxin and flecainide polytherapy achieving rate reversion in 17 of 18 cases (94.5%) at a median of 3 days (1.5–7). Hydrops was present in 23 (33%) cases at initial presentation, 16 of which achieved rate reversion. There was minimal difference in treatment efficacy comparing single- or multiple-agent treatment in the setting of hydrops (50% vs. 42.8%). Side effects occurred in 14/56 treated patients (25%) but were severe in only 8 (14.3%) women, most commonly with digoxin and flecainide polytherapy (6 of 8 cases). There were 3 (4%) fetal deaths amongst the study cohort. <b><i>Conclusions:</i></b> Digoxin and flecainide polytherapy were well tolerated and successfully achieved rhythm and rate control in fetuses with prenatally diagnosed supraventricular tachycardia. The presence of hydrops was a poor prognostic feature.

Bearing Prognostics: An Instance-Based Learning Approach with Feature Engineering, Data Augmentation, and Similarity Evaluation

We propose an instance-based learning approach with data augmentation and similarity evaluation to estimate the remaining useful life (RUL) of a mechanical component for health management. The publicly available PRONOSTIA datasets, which provide accelerated degradation test data for bearings, are used in our study. The challenges with the datasets include a very limited number of run-to-failure examples, no failure mode information, and a wide range of bearing life spans. Without a large number of training samples, feature engineering is necessary. Principal component analysis is applied to the spectrogram of vibration signals to obtain prognostic feature sequences. A data augmentation strategy is developed to generate synthetic prognostic feature sequences using learning instances. Subsequently, similarities between the test and learning instances can be assessed using a root mean squared (RMS) difference measure. Finally, an ensemble method is developed to aggregate the RUL estimates based on multiple similar prognostic feature sequences. The proposed approach demonstrates comparable performance with published solutions in the literature. It serves as an alternative method for solving the RUL estimation problem.

Development of a semi-automated method for tumor budding assessment in colorectal cancer and comparison with manual methods

Tumor budding is an established prognostic feature in multiple cancers but routine assessment has not yet been incorporated into clinical pathology practice. Recent efforts to standardize and automate assessment have shifted away from haematoxylin and eosin (H&E)-stained images towards cytokeratin (CK) immunohistochemistry. In this study, we compare established manual H&E and cytokeratin budding assessment methods with a new, semi-automated approach built within the QuPath open-source software. We applied our method to tissue cores from the advancing tumor edge in a cohort of stage II/III colon cancers (n=186). The total number of buds detected by each method, over the 186 TMA cores, were as follows; manual H&E (n=503), manual CK (n=2290) and semi-automated (n=5138). More than four times the number of buds were detected using CK compared to H&E. A total of 1734 individual buds were identified both using manual assessment and semi-automated detection on CK images, representing 75.7% of the total buds identified manually (n=2290) and 33.7% of the total buds detected using our proposed semi-automated method (n=5138). Higher bud scores by the semi-automated method were due to any discrete area of CK immunopositivity within an accepted area range being identified as a bud, regardless of shape or crispness of definition, and to inclusion of tumor cell clusters within glandular lumina ('luminal pseudobuds'). Although absolute numbers differed, semi-automated and manual bud counts were strongly correlated across cores (ρ=0.81, p<0.0001). Despite the random, rather than 'hotspot', nature of tumor core sampling, all methods of budding assessment demonstrated poorer survival associated with higher budding scores. In conclusion, we present a new QuPath-based approach to tumor budding assessment, which compares favorably to current established methods and offers a freely-available, rapid and transparent tool that is also applicable to whole slide images.

The role of bronchoscopy in patients with SARS-CoV-2 pneumonia

BackgroundThe role of bronchoscopy in coronavirus disease 2019 (COVID-19) is a matter of debate. Patients and methods: This observational multicenter study aimed to analyse the prognostic impact of bronchoscopic findings in a consecutive cohort of patients with suspected or confirmed COVID-19. Patients were enrolled at 17 hospitals from February to June, 2020. Predictors of in-hospital mortality were assessed by multivariate logistic regression.ResultsA total of 1027 bronchoscopies were performed in 515 patients (age 61.5±11.2; 73% men), stratified into a clinical suspicion cohort (n=30) and a COVID-19 confirmed cohort (n=485). In the clinical suspicion cohort, the diagnostic yield was 36.7%. In the COVID-19 confirmed cohort, bronchoscopies were predominantly performed in the intensive care unit (n=961; 96.4%) and major indications were: difficult mechanical ventilation (43.7%), mucus plugs (39%) and persistence of radiological infiltrates (23.4%). One hundred forty-seven bronchoscopies were performed to rule out superinfection, and diagnostic yield was 42.9%. There were abnormalities in 91.6% of bronchoscopies, the most frequent being mucus secretions (82.4%), haematic secretions (17.7%), mucus plugs (17.6%), and diffuse mucosal hyperemia (11.4%). The independent predictors of in-hospital mortality were: older age (Odds ratio [OR]=1.06; p<0.001), mucus plugs as indication for bronchoscopy (OR=1.60; p=0.041), absence of mucosal hyperemia (OR=0.49; p=0.041) and the presence of haematic secretions (OR=1.79; p=0.032).ConclusionsBronchoscopy may be indicated in carefully selected patients with COVID-19 to rule out superinfection and solve complications related to mechanical ventilation. The presence of haematic secretions in the distal bronchial tract may be considered a poor prognostic feature in COVID-19.

HMGB1, the Next Predictor of Transcatheter Arterial Chemoembolization for Liver Metastasis of Colorectal Cancer?

HMGB1 is an important mediator of inflammation during ischemia–reperfusion injury on organs. The serum expression of HMGB1 was increased significantly on the 1st day after TACE and decreased significantly which was lower on the 30th day after TACE. Tumor markers of post-DEB-TACE decreased significantly. The correlational analysis showed that patients with low HMGB1 expression had lower risks of fever and liver injury compared those with the higher expression, while the ORR is relatively worse. Patients with lower expression of HMGB1 had longer PFS, better efficacy, and higher quality of life. With the high post-expression, the low expression had lower incidence of fever and liver injury too. There was no statistical difference in the one-year survival among the different groups. The quality of life of all patients was improved significantly. The over-expression of HMGB1 in LMCRC is an adverse prognostic feature and a positive predictor of response to TACE.