Automated histological classification of whole-slide images of gastric biopsy specimens

BACKGROUND

Automated image analysis has been developed currently in the field of surgical pathology. The aim of the present study was to evaluate the classification accuracy of the e-Pathologist image analysis software.

METHODS

A total of 3062 gastric biopsy specimens were consecutively obtained and stained. The specimen slides were anonymized and digitized. At least two experienced gastrointestinal pathologists evaluated each slide for pathological diagnosis. We compared the three-tier (positive for carcinoma or suspicion of carcinoma; caution for adenoma or suspicion of a neoplastic lesion; or negative for a neoplastic lesion) or two-tier (negative or non-negative) classification results of human pathologists and of the e-Pathologist.

RESULTS

Of 3062 cases, 33.4% showed an abnormal finding. For the three-tier classification, the overall concordance rate was 55.6% (1702/3062). The kappa coefficient was 0.28 (95% CI, 0.26-0.30; fair agreement). For the negative biopsy specimens, the concordance rate was 90.6% (1033/1140), but for the positive biopsy specimens, the concordance rate was less than 50%. For the two-tier classification, the sensitivity, specificity, positive predictive value, and negative predictive value were 89.5% (95% CI, 87.5-91.4%), 50.7% (95% CI, 48.5-52.9%), 47.7% (95% CI, 45.4-49.9%), and 90.6% (95% CI, 88.8-92.2%), respectively.

CONCLUSIONS

Although there are limitations and requirements for applying automated histopathological classification of gastric biopsy specimens in the clinical setting, the results of the present study are promising.

Deep learning-based six-type classifier for lung cancer and mimics from histopathological whole slide images: a retrospective study

BACKGROUND

Targeted therapy and immunotherapy put forward higher demands for accurate lung cancer classification, as well as benign versus malignant disease discrimination. Digital whole slide images (WSIs) witnessed the transition from traditional histopathology to computational approaches, arousing a hype of deep learning methods for histopathological analysis. We aimed at exploring the potential of deep learning models in the identification of lung cancer subtypes and cancer mimics from WSIs.

METHODS

We initially obtained 741 WSIs from the First Affiliated Hospital of Sun Yat-sen University (SYSUFH) for the deep learning model development, optimization, and verification. Additional 318 WSIs from SYSUFH, 212 from Shenzhen People's Hospital, and 422 from The Cancer Genome Atlas were further collected for multi-centre verification. EfficientNet-B5- and ResNet-50-based deep learning methods were developed and compared using the metrics of recall, precision, F1-score, and areas under the curve (AUCs). A threshold-based tumour-first aggregation approach was proposed and implemented for the label inferencing of WSIs with complex tissue components. Four pathologists of different levels from SYSUFH reviewed all the testing slides blindly, and the diagnosing results were used for quantitative comparisons with the best performing deep learning model.

RESULTS

We developed the first deep learning-based six-type classifier for histopathological WSI classification of lung adenocarcinoma, lung squamous cell carcinoma, small cell lung carcinoma, pulmonary tuberculosis, organizing pneumonia, and normal lung. The EfficientNet-B5-based model outperformed ResNet-50 and was selected as the backbone in the classifier. Tested on 1067 slides from four cohorts of different medical centres, AUCs of 0.970, 0.918, 0.963, and 0.978 were achieved, respectively. The classifier achieved high consistence to the ground truth and attending pathologists with high intraclass correlation coefficients over 0.873.

CONCLUSIONS

Multi-cohort testing demonstrated our six-type classifier achieved consistent and comparable performance to experienced pathologists and gained advantages over other existing computational methods. The visualization of prediction heatmap improved the model interpretability intuitively. The classifier with the threshold-based tumour-first label inferencing method exhibited excellent accuracy and feasibility in classifying lung cancers and confused nonneoplastic tissues, indicating that deep learning can resolve complex multi-class tissue classification that conforms to real-world histopathological scenarios.

Histopathological prognostic factors in ANCA-associated glomerulonephritis

Anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV) are a group of multisystemic autoimmune diseases characterized by necrotizing inflammation of small vessels. Kidney involvement is frequent in granulomatosis with polyangiitis (GPA) and microscopic polyangiitis (MPA), and accounts for a significant proportion of the morbidity and mortality related to these diseases. Despite improvement in therapeutic management of ANCA-glomerulonephritis (ANCA-GN), end-stage kidney disease (ESKD) still occurs in up to 30% of affected patients within 5 years following diagnosis. Thus, identifying patients for whom aggressive immunosuppressive therapy will be more beneficial than deleterious is of great importance. Several clinical, biological and histological factors have been proposed as predictors of ESKD. The kidney biopsy is essential not only for the diagnosis, but also for evaluating renal prognosis. In this review, we discuss the prognostic value of renal lesions at the diagnosis of ANCA-GN by analyzing each compartment of the nephron. We also review existing ESKD risk classification in ANCA-GN and finally propose an example of a standardized pathology report that could be used in routine practice.

A new entity in the pathological spectrum of vulvar neoplasms: The first report of a primary endometrioid adenocarcinoma

We report the first evidence of primary endometrioid adenocarcinoma arising in the vulva, expanding the morphological spectrum of the glandular neoplasms in this anatomic region.

The predictors of long-term outcomes after targeted therapy for primary Aldosteronism

Unilateral primary aldosteronism is thought to be a surgically curable disease, and unilateral adrenalectomy is the mainstay treatment. The Primary Aldosteronism Surgical Outcome (PASO) consensus was developed to assess clinical and biochemical outcomes to standardize the classification of surgical outcomes. However, fewer than half of patients are cured of hypertension after adrenalectomy; therefore, preoperative patient counseling and evaluation might be necessary. Moreover, current studies show that genetic mutations and histopathology classification are associated with the treatment outcome. The Task Force of Taiwan PA recommends using a specific scoring system, including the PASO score and nomogram-based preoperative score, to predict the clinical outcome before adrenalectomy. Herein, we discuss the associations of current histopathological classification and specific somatic gene mutations with clinical outcomes after surgery.

Clinicopathological Criteria Defining Mucinous Appendiceal Tumors from 2476 Appendectomies: a Single-Center Retrospective Study

OBJECTIVES

Mucinous appendiceal tumors (MATs) constitute 0.2-0.3% of appendectomies. This retrospective chart review study determines the incidence of MATs among appendectomies at King Abdul-Aziz Specialist Hospital, Taif City, Saudi Arabia, from January 2009 to December 2014. The clinicopathological features, histopathological criteria, management, outcomes of patients, and the impact of histopathological classification on the follow-up period and recurrence are evaluated.

METHODS

Demographic and clinicopathological data were collected from medical records. Microscopic slides from 2476 appendectomies were re-examined to diagnose and classify MATs into low-grade mucinous neoplasms (LAMNs) and mucinous adenocarcinomas (MACAs). CK20, CK7, and cdx2 immunohistochemistry was applied for evaluating pseudomyxoma peritonei. Data were expressed as numbers, percentages, and mean ± standard deviation.

RESULTS

Nine MATs were diagnosed with an incidence of 0.36% of appendectomies, a male:female ratio of 1.25:1 and a mean age of 57.2 years. Acute appendicitis was the commonest clinical presentation. About 66.7% were LAMNs and 33.3% MACAs. Beside appendectomy, MACAs were managed with right hemicolectomy and chemotherapy. The median follow-up was 34 months with recurrence and liver metastases in two MACAs. No recurrences for LAMNs.

CONCLUSIONS

MATs constitute 0.36% of all appendectomies. Classifying MATs into LAMNs and MACAs is more applicable for both clinical and pathology practices as compared to the three- or four-tiered classification schemes.

A Review of Surgical Outcomes of Management of Sinonasal Malignancies: A 8-Year of Clinical Experience (2013-2021) at the Tertiary Centre, Sarawak

Sinonasal carcinoma, a rare and challenging malignancy, originating in the nasal cavity and paranasal sinuses, poses diagnostic and management complexities. This 8-year retrospective analysis at Sarawak General Hospital aims to elucidate demographic trends, histopathological entities, and management outcomes, shedding light on this multifaceted malignancy. Emphasizing the significance of accurate histopathological classification, the study explores the impact on prognostication and treatment strategies. Spanning 2013 to 2021, the study involved 54 patients with sinonasal malignancies. Demographic, clinical, and histopathological details were examined, adhered to the AJCC staging criteria. Analysis involved demographic distributions, tumour characteristics, treatment modalities, and instances of treatment failure. Statistical analysis was done using SPSS version 29.02. The cohort, predominantly male (57.4%) and of Iban ethnicity (44.4%), with a mean age of 52.8 years, exhibited diverse histopathologies, with squamous cell carcinoma as the most common (38.9%). Epistaxis and nasal blockage were common clinical presentations. Advanced stages (III and IV) were prevalent, with the nasal cavity as the primary site. Surgical interventions, mainly endoscopic endonasal excision, were complemented by adjuvant therapies. Complications occurred in 24% of cases. The study highlights a male predilection, occupational risk factors, and a significant association between tobacco smoking and sinonasal cancers. Surgical interventions predominantly utilized the endoscopic approach. Despite a mean survival of 46.6 months, treatment failure occurred in 29.6% of cases, with recurrence and metastasis. Histopathological analysis revealed comparable 5-year disease survival rates between squamous and non-squamous histologies. Treatment failure was significantly associated with the mode of surgery, with open surgery showing a lower incidence. However, nodal status, histopathology types, T staging, and overall staging did display positive associations with treatment failure. This 8-year review provides comprehensive insights into sinonasal carcinoma, addressing demographic, clinical, and histopathological dimensions. The study underlines the complexity of managing this challenging malignancy, emphasizing the need for a holistic approach to patient care. The findings contribute to the understanding of sinonasal carcinoma, guiding clinical decision-making and fostering further research.

The correct methodological approach to the diagnosis of celiac disease: the point of view of the pathologist

The diagnosis of celiac disease relies on the assessment of serological data and the presence of histological alterations in the duodenal mucosa. The duodenal biopsy is pivotal in adults, and in some circumstances in children, to confirm the clinical suspicion of celiac disease. The correct interpretation of duodenal biopsies is influenced by numerous variables. The aim of this overview is to describe the correct methodological approach including the procedures of biopsy sampling, orientation, processing, staining and histopathological classification in order to avoid or minimize the errors and the variability in duodenal biopsy interpretation. Multiple biopsies taken from different sites of the duodenum during endoscopy maximize the diagnostic yield of duodenal histological sampling. Proper orientation of the biopsy samples is of the utmost importance to assess histological features of pathological duodenal mucosa and to avoid artifacts that may lead even an experienced pathologist to a wrong histological interpretation with subsequent misdiagnosis of celiac disease. An immunohistochemical stain for CD3 can be invaluable to aid the pathologist in obtaining a more accurate intra-epithelial T lymphocytes count. A simplified histological classification facilitates the clinician's work and improves the communication between pathologist and clinician. An integrated clinical and pathological approach is required for a correct diagnosis of celiac disease since a relatively large number of conditions may cause duodenal damage with a histological appearance similar to that of celiac disease.

Adrenal aldosterone synthase (CYP11B2) histopathology and its association with disease-induced sudden death: a cross-sectional study

Background

Unidentified cardiovascular risk factors may account for approximately half of sudden deaths, a devastating event with limited preventive tools. We investigated whether adrenal histopathology suggestive of primary aldosteronism, pheochromocytoma, or adrenal masses could explain part of the risk for disease-induced sudden death (DSD).

Methods

In this study, autopsies and histopathological analyses, including aldosterone synthase staining of adrenal glands, were performed on 403 consecutive individuals who experienced sudden death. These individuals were classified into 258 cases of DSD and 144 deaths caused by trauma, suicide, or intoxication, i.e., non-disease-induced sudden death (nDSD). This trial was registered at ClinicalTrials.gov (NCT05446779).

Findings

Adrenal histopathology revealed changes in 31 (7.7%) subjects of the cohort. Of these, the most prevalent findings [25 (6.2%)] were aldosterone-producing adenomas (APA) or nodules (APN), which were associated with myocardial infarction and atherosclerosis at autopsy. Individuals in the DSD group and the subgroup with sudden cardiac death (SCD) were more likely to have APA or APN than individuals in the nDSD group [23 (8.9%) vs. 2 (1.4%), p = 0.002; 16 (8.8%) vs. 2 (1.4%), p = 0.003, respectively]. APA or APN were explanatory factors for DSD (odds ratio [OR] 6.47, 95% confidence interval [CI] 1.40-29.88, p = 0.017) and SCD (OR 10.68, 95% CI 2.02-56.43, p = 0.005). Other findings included two pheochromocytomas, one bilateral adrenal metastasis, and two unilateral adrenal metastases.

Interpretation

In this exploratory study, APA or APN were more frequently seen in DSD and SCD than nDSD cases. Whether primary aldosteronism constitutes a novel risk factor for sudden death warrants further study.

Funding

Finnish State Research funds and independent research foundations: Aarne Koskelo Foundation, the Finnish Kidney Foundation, and the Finnish Foundation for Cardiovascular Research.

Modified histopathological classification with age-related glomerulosclerosis for predicting kidney survival in ANCA-associated glomerulonephritis

BACKGROUND

The histopathological classification of ANCA-GN divides patients into four groups based on signs of glomerular injury. However, this classification did not consider age-related glomerulosclerosis. In this study, we aimed to compare the prediction of renal survival between Berden's ANCA-GN histopathological classification and ANCA-GN histopathological classification modified with age-related glomerulosclerosis.

METHODS

Between January 2004 and December 2019, 65 patients diagnosed with ANCA-GN were enrolled. Demographic, laboratory, and histopathologic findings were retrospectively analyzed. Renal survival analyses were compared according to classical and modified ANCA-GN histopathological classifications. Multivariate Cox regression analysis for the factors affecting renal survival was performed.

RESULTS

In Berden's ANCA-GN histopathological classification, 15 patients were in the focal group, 21 in the crescentic, 21 in the sclerotic, and 8 in the mixed group. The ANCA-GN histopathological classification model generated statistically significant predictions for renal survival (p = 0.022). When the histopathological classification was modified with age-related glomerulosclerosis, eight of the nine patients previously classified in the sclerotic group were classified in the mixed and one in the crescentic groups. Modification of histopathological classification with age-related glomerulosclerosis increases the statistical significance in renal survival analysis (p = 0.009). The multivariate Cox regression analysis showed that the disease-related global sclerotic glomeruli percentage and serum creatinine level were significant independent factors.

CONCLUSION

Modification of Berden's ANCA-GN histopathological classification model with age-related glomerulosclerosis may increase the statistical significance of the histopathological classification model.

Pathology diagnosis of intraoperative frozen thyroid lesions assisted by deep learning

BACKGROUND

Thyroid cancer is a common thyroid malignancy. The majority of thyroid lesion needs intraoperative frozen pathology diagnosis, which provides important information for precision operation. As digital whole slide images (WSIs) develop, deep learning methods for histopathological classification of the thyroid gland (paraffin sections) have achieved outstanding results. Our current study is to clarify whether deep learning assists pathology diagnosis for intraoperative frozen thyroid lesions or not.

METHODS

We propose an artificial intelligence-assisted diagnostic system for frozen thyroid lesions that applies prior knowledge in tandem with a dichotomous judgment of whether the lesion is cancerous or not and a quadratic judgment of the type of cancerous lesion to categorize the frozen thyroid lesions into five categories: papillary thyroid carcinoma, medullary thyroid carcinoma, anaplastic thyroid carcinoma, follicular thyroid tumor, and non-cancerous lesion. We obtained 4409 frozen digital pathology sections (WSI) of thyroid from the First Affiliated Hospital of Sun Yat-sen University (SYSUFH) to train and test the model, and the performance was validated by a six-fold cross validation, 101 papillary microcarcinoma sections of thyroid were used to validate the system's sensitivity, and 1388 WSIs of thyroid were used for the evaluation of the external dataset. The deep learning models were compared in terms of several metrics such as accuracy, F1 score, recall, precision and AUC (Area Under Curve).

RESULTS

We developed the first deep learning-based frozen thyroid diagnostic classifier for histopathological WSI classification of papillary carcinoma, medullary carcinoma, follicular tumor, anaplastic carcinoma, and non-carcinoma lesion. On test slides, the system had an accuracy of 0.9459, a precision of 0.9475, and an AUC of 0.9955. In the papillary carcinoma test slides, the system was able to accurately predict even lesions as small as 2 mm in diameter. Tested with the acceleration component, the cut processing can be performed in 346.12 s and the visual inference prediction results can be obtained in 98.61 s, thus meeting the time requirements for intraoperative diagnosis. Our study employs a deep learning approach for high-precision classification of intraoperative frozen thyroid lesion distribution in the clinical setting, which has potential clinical implications for assisting pathologists and precision surgery of thyroid lesions.