Search

Background: Pleural effusion can be caused by a wide range of benign and malignant conditions. Pleural biopsy and effusion cytology represent two key methods of pathological diagnosis. To compare the performance these two methods, a large cohort of matched pleural biopsy and effusion cytology with clinical follow‐up was reviewed. Methods: Pleural biopsies and effusion cytology specimens over a period of 18 years were retrieved. Cytology specimens collected within 7 days of pleural biopsy were matched. Reports were reviewed, and the cause for pleural effusion was determined by hospital disease coding and clinical data. Results: Totally, 3026 cases were included. The leading cause of benign effusion was tuberculosis (n = 650). Malignant pleural effusion (MPE) was more common in older females (p < 0.001) and mostly due to lung cancer (n = 959), breast cancer (n = 64), and mesothelioma (n = 48). The inadequate/insufficient (B1/C1) rate of biopsy was higher than cytology (15.6% vs. 0.3%) but the rates for other diagnostic categories were similar. Biopsy and cytology showed a correlation coefficient of 0.315, improving to 0.449 when inadequate/insufficient (B1/C1) cases were excluded. The ROM for benign cytology (C2) was lower than biopsy (B2) (p < 0.001). Compared with biopsy, the diagnostic accuracy was higher in cytology overall and for metastatic carcinomas (p < 0.001) but lower for hematolymphoid malignancies (p = 0.014) and mesotheliomas (p = 0.002). Conclusions: These results suggest that effusion cytology may be better for confirming benignity and diagnosing carcinomatous MPE. In these cases, pleural biopsy may be withheld to reduce procedural risks. However, for suspected hematolymphoid malignancies and mesothelioma, biopsy should be considered. [ABSTRACT FROM AUTHOR]

We thank Zhang and colleagues for their readership and letter.[[1]] Zhang et al. have commented that diagnostic odds ratio (DOR), negative-to-positive likelihood ratio (LR), and area under receiver operating characteristic curves (AUC) could serve additional metrics for exploring the diagnostic performance of pleural biopsy and pleural effusion, and that the lack of AUC constitutes methodological insufficiency. All subsequent analyses were based on two-by-two tables of pleural biopsies and pleural effusion cytology specimens. [Extracted from the article]

Histopathological images of colorectal liver metastases (CRLM) contain rich morphometric information that may predict patients' outcomes. However, to our knowledge, no study has reported any practical deep learning framework based on the histology images of CRLM, and their direct association with prognosis remains largely unknown. In this study, we developed a deep learning-based framework for fully automated tissue classification and quantification of clinically relevant spatial organization features (SOFs) in H&E-stained images of CRLM. The SOFs based risk-scoring system demonstrated a strong and robust prognostic value that is independent of the current clinical risk score (CRS) system in independent clinical cohorts. Our framework enables fully automated tissue classification of H&E images of CRLM, which could significantly reduce assessment subjectivity and the workload of pathologists. The risk-scoring system provides a time- and cost-efficient tool to assist clinical decision-making for patients with CRLM, which could potentially be implemented in clinical practice., Competing Interests: The authors declare no competing interests., (© 2023 The Authors.)

Idiopathic granulomatous mastitis (IGM) is a benign mimic of breast carcinomas. It is defined histologically by the presence of granulomas and inflammation. The closely related cystic neutrophilic granulomatous mastitis (CNGM) shows lipogranulomas, with a reported association with corynebacteria. A large cohort of IGM was reviewed to compare clinical, microbiological and histological features between non-CNGM IGM and CNGM. Cases of IGM were reviewed for histological parameters including the presence of lipogranulomas and composition of inflammatory cells. Clinical data were obtained through hospital records. The cohort included 79 cases, including 51 non-CNGM IGM and 28 CNGM. Comparing non-CNGM IGM and CNGM, there were no differences in clinical or demographical data, other than a younger age of presentation (36.2 vs 41.5 years, p=0.012) for CNGM. Most IGM resolved within the follow-up period (n=57/64, 89.1%), with similar outcomes regardless of treatment (p>0.05). In CNGM, there were more infiltrates of neutrophils (p=0.001), histiocytes (p=0.047), and multinucleated giant cells (p=0.006), but less lymphocytes (p=0.008). Corynebacteria was cultured in two (25%) cases of CNGM, and one non-CNGM IGM (14.3%). Gram-positive bacilli were identified in two cases of CNGM. 'Early' lipogranulomas were observed closely associated to inflamed ducts in three cases of CNGM. Apart from age, there was no distinct clinical or microbiological feature for CNGM. These findings do not support CNGM as a distinct entity. Rather, CNGM-pattern may represent a continuum of IGM, possibly age-related and attributable to ductal inflammation and cystic changes in the breast parenchyma., (Copyright © 2022 Royal College of Pathologists of Australasia. Published by Elsevier B.V. All rights reserved.)

Computational cytology is a critical, rapid-developing, yet challenging topic in medical image computing concerned with analyzing digitized cytology images by computer-aided technologies for cancer screening. Recently, an increasing number of deep learning (DL) approaches have made significant achievements in medical image analysis, leading to boosting publications of cytological studies. In this article, we survey more than 120 publications of DL-based cytology image analysis to investigate the advanced methods and comprehensive applications. We first introduce various deep learning schemes, including fully supervised, weakly supervised, unsupervised, and transfer learning. Then, we systematically summarize public datasets, evaluation metrics, versatile cytology image analysis applications including cell classification, slide-level cancer screening, nuclei or cell detection and segmentation. Finally, we discuss current challenges and potential research directions of computational cytology., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2022 Elsevier B.V. All rights reserved.)