Standardized Classification of Lung Adenocarcinoma Subtypes and Improvement of Grading Assessment Through Deep Learning.

The histopathologic distinction of lung adenocarcinoma (LADC) subtypes is subject to high interobserver variability, which can compromise the optimal assessment of patient prognosis. Therefore, this study developed convolutional neural networks capable of distinguishing LADC subtypes and predicting disease-specific survival, according to the recently established LADC tumor grades. Consensus LADC histopathologic images were obtained from 17 expert pulmonary pathologists and one pathologist in training. Two deep learning models (AI-1 and AI-2) were trained to predict eight different LADC classes. Furthermore, the trained models were tested on an independent cohort of 133 patients. The models achieved high precision, recall, and F1 scores exceeding 0.90 for most of the LADC classes. Clear stratification of the three LADC grades was reached in predicting the disease-specific survival by the two models, with both Kaplan-Meier curves showing significance (P = 0.0017 and 0.0003). Moreover, both trained models showed high stability in the segmentation of each pair of predicted grades with low variation in the hazard ratio across 200 bootstrapped samples. These findings indicate that the trained convolutional neural networks improve the diagnostic accuracy of the pathologist and refine LADC grade assessment. Thus, the trained models are promising tools that may assist in the routine evaluation of LADC subtypes and grades in clinical practice.

Artificial Intelligence-Based Tool for Tumor Detection and Quantitative Tissue Analysis in Colorectal Specimens.

Digital pathology adoption allows for applying computational algorithms to routine pathology tasks. Our study aimed to develop a clinical-grade artificial intelligence (AI) tool for precise multiclass tissue segmentation in colorectal specimens (resections and biopsies) and clinically validate the tool for tumor detection in biopsy specimens. The training data set included 241 precisely manually annotated whole-slide images (WSIs) from multiple institutions. The algorithm was trained for semantic segmentation of 11 tissue classes with an additional module for biopsy WSI classification. Six case cohorts from 5 pathology departments (4 countries) were used for formal and clinical validation, digitized by 4 different scanning systems. The developed algorithm showed high precision of segmentation of different tissue classes in colorectal specimens with composite multiclass Dice score of up to 0.895 and pixel-wise tumor detection specificity and sensitivity of up to 0.958 and 0.987, respectively. In the clinical validation study on multiple external cohorts, the AI tool reached sensitivity of 1.0 and specificity of up to 0.969 for tumor detection in biopsy WSI. The AI tool analyzes most biopsy cases in less than 1 minute, allowing effective integration into clinical routine. We developed and extensively validated a highly accurate, clinical-grade tool for assistive diagnostic processing of colorectal specimens. This tool allows for quantitative deciphering of colorectal cancer tissue for development of prognostic and predictive biomarkers and personalization of oncologic care. This study is a foundation for a SemiCOL computational challenge. We open-source multiple manually annotated and weakly labeled test data sets, representing a significant contribution to the colorectal cancer computational pathology field.

Performance of HPV E6/E7 mRNA RT-qPCR for screening and diagnosis of cervical cancer with ThinPrep Pap test samples.

Recent research has shown that oncogenic human papillomavirus (HPV) DNA, which is currently used in the screening and diagnosis of cervical cancer, can be detected not only in high-grade cervical lesions, but also in low-grade cervical lesions and normal tissues. For this reason, HPV tests targeting the E6 and E7 mRNA of five oncogenic HPV strains (HPV genotypes 16, 18, 31, 33, and 45), which are known to be responsible for the oncogenesis of cervical cancer, have been commercialized using a real-time nucleic acid sequence based amplification (NASBA) assay. Previous data has shown that the real-time NASBA assay has higher clinical specificity than HPV DNA testing (97.1% vs. 53.7%). However, the sensitivity of the real-time NASBA assay was lower than that of HPV DNA testing (41.1% vs. 100%). Despite the fact that there are more than 16 oncogenic HPV genotypes known to cause cervical cancer (HPV genotypes 16, 18, 31, 33, 35, 39, 45, 51, 52, 53, 56, 58, 59, 66, 68, and 69), the commercialized real-time NASBA kit was designed to detect only five genotypes (16, 18, 31, 33, and 45). Therefore, in the present study, CervicGen HPV RT-qDX (Optipharm), a commercial diagnostic kit targeting a HPV E6/E7 mRNA based on RT-qPCR assay was evaluated with RNA extracted from ThinPrep Pap samples, and the results were compared to real-time NASBA data. The sensitivity and specificity of the RT-qPCR assay were 91% and 98.6%, respectively, for the detection of cervical intraepithelial neoplasia CIN2(+) high-grade cervical lesions. Therefore, the CervicGen HPV RT-qDX assay showed a significantly higher sensitivity (91.1%) compared to the real-time NASBA assay (41.1%). In normal cytohistology cases, the specificity was 98.6% and 53.7% for HPV mRNA RT-qPCR and HPV DNA testing, respectively. These results demonstrate that HPV mRNA RT-qPCR better reflects clinical diagnosis. In conclusion, it is suggested that HPV mRNA RT-qPCR overcomes the shortcomings of lower specificity seen in the DNA assay and the lower sensitivity of the commercialized HPV mRNA real-time NASBA assay when testing from ThinPrep Pap samples.

Expression of CD133 in neuroendocrine neoplasms of the digestive tract: a detailed immunohistochemical analysis.

Gastroenteropancreatic neuroendocrine tumors (GEP-NETs) are potentially malignant with variable biologic behavior that originates from neuroendocrine cells of digestive tract. Recently, the existence of cancer stem cells (CSC) was demonstrated in tumors of gastrointestinal tract. CD133 is a transmembrane glycoprotein that serves as a CSC marker in various malignancies. However, the expression of CD133 in neuroendocrine neoplasms (NEN) of digestive tract has not been studied. We evaluated tissue expression of CD133 by immunohistochemistry in 90 NENs of digestive tract with their matched non-neoplastic mucosa including stomach (n=15), small intestine (n=7), appendix (n=3), colon (n=8), rectum (n=41), pancreas (n=2), gallbladder (n=4) and liver (n=10). Tumors were divided according to 2010 WHO classification. CD133 was expressed in 30.3% (17/56) of well-differentiated neuroendocrine tumors (NET), 26.1% (6/23) of poorly-differentiated neuroendocrine carcinomas (NEC) and 63.6% (7/11) of mixed adenoneuroendocrine carcinoma (MANECs). MANEC refers to existence of both adenocarcinoma and NEC together, each one comprising at least 30% of the tumor. CD133 was expressed in cytoplasm, luminal-side of cell membrane, or both and the staining pattern correlated with tumor growth pattern. CD133 expression was not significantly correlated with tumor grade, site, expression of neuroendocrine markers (chromogranin-A and synaptophysin) and patients' survival. Thus, CD133 expression may lack prognostic significance in GEP-NETs. Importantly, CD133 was not detectable in non-neoplastic neuroendocrine cells of digestive system including pancreatic islets. In conclusion, CD133 is expressed in poorly-differentiated NECs and well-differentiated NETs of the digestive tract.

Integrating cytology into routine digital pathology workflow: a 5-year journey.

Despite recent advances in digital imaging, the adoption of digital cytology is challenging due to technical limitations. This study describes our 5-year institutional experience with the implementation of digital cytology. The routine cytology workflow included conventional two-step screening by cytotechnologists, followed by sign out by pathologists. We introduced sign out of cytologic cases using a microscopic digital imaging platform operated by cytotechnologists, which allowed for remote review of slides by cytopathologists via video streaming. We also provided cytologic correlation to support the virtual slide-based sign out of histopathological specimens and for a weekly pathology-radiology conference. In addition, positive cytology cases were archived for integration into the laboratory information system and for prospective computational pathology studies. We also summarized lessons learned over the years and outlined our vision for future developments. This unique experience may serve as a role model for other institutions.

Substantial improvement of histopathological diagnosis by whole-slide image-based remote consultation.

Consultation by subspecialty experts is the most common mode of rendering diagnosis in challenging cases in pathological practice. Our study aimed to highlight the diagnostic benefits of whole-slide image (WSI)-based remote consultation. We obtained diagnostically challenging cases from two institutions from the years 2010 and 2013, with histological diagnoses that contained keywords "probable," "suggestive," "suspicious," "inconclusive," and "uncertain." A total of 270 cases were selected for remote consultation using WSIs scanned at 40 × . The consultation process consisted of three rounds: the first and second rounds each with 12 subspecialty experts and the third round with six multi-expertise senior pathologists. The first consultation yielded 44% concordance, and a change in diagnosis occurred in 56% of cases. The most frequent change was from inconclusive to definite diagnosis (30%), followed by minor discordance (14%), and major discordance (12%). Out of the 70 cases which reached the second round, 31 cases showed discrepancy between the two consultants. For these 31 cases, a consensus diagnosis was provided by six multi-expertise senior pathologists. Combining all WSI-based consultation rounds, the original inconclusive diagnosis was changed in 140 (52%) out of 266 cases. Among these cases, 80 cases (30%) upgraded the inconclusive diagnosis to a definite diagnosis, and 60 cases (22%) changed the diagnosis with major or minor discordance, accounting for 28 cases (10%) and 32 cases (12%), respectively. We observed significant improvement in the pathological diagnosis of difficult cases by remote consultation using WSIs, which can further assist in patient healthcare. A post-study survey highlighted various benefits of WSI-based consults.

Epstein-Barr Virus-Associated Langerhans Cell Histiocytosis of the Thyroid Gland.

Thyroid gland involvement of Langerhans cell histiocytosis (LCH) is extremely rare in both systemic and isolated disease. The role of viral infection in LCH development is not yet fully understood. Although several viruses are proposed as etiologic factors, such as Epstein-Barr virus (EBV) and human herpesvirus 6 (HHV-6), they seem to play a bystander role in LCH. A 29-year old female patient with a prior history of multisystemic LCH (pituitary gland and skull bone), presented with a thyroid nodule. The patient underwent a total thyroidectomy and the histological examination revealed nodular lesions composed of sheets and clusters of histiocytes in the inflammatory background. The histiocytes stained positive for S-100 and CD1a and were negative for HHV-8, cytomegalovirus, and VE1 (anti-BRAFV600E) on immunohistochemistry. The EBER in situ hybridization for EBV showed frequent positive-stained cells. The conventional PCR analysis for EBV was positive and qPCR analysis confirmed a significant DNA copy number difference (p = 0.02) between the tumor and adjacent non-neoplastic thyroid tissue. PCR analysis for HHV-6, HPV, HSV was negative in both tumor and benign samples. In conclusion, the presented case showed a rare thyroid involvement by LCH associated with EBV infection, which has not been reported before. Further studies are required to investigate a possible etiologic link between EBV infection and LCH.

Basaloid Squamous Cell Carcinoma of the Uterine Cervix: Report of a Case With Molecular Analysis.

There is a lack of knowledge about molecular alterations in basaloid squamous cell carcinoma (BSCC) of the uterine cervix. A 72-year-old woman with a history of previous subtotal hysterectomy and current vaginal bleeding was referred to our hospital. Initially, adenoid cystic carcinoma (ACC) was diagnosed upon cervical cytology and biopsy. Chest imaging showed multiple metastatic lesions in both lungs. The surgical specimen showed BSCC with diffuse p16 immunoreactivity and negativity for S-100, c-kit, and neuroendocrine markers. There was a focal minor ACC component, which could have explained the previous cytology and biopsy diagnosis. Next-generation sequencing with two different panels showed coexisting PIK3CA mutation and NTRK2 fusion with 10 additional variants of unknown significance (ATR, DAXX, FAM123B, JAK1, KEL, MLL2, NOTCH2, PALB2, POLD1, POLE). The MYB gene fusions were not identified. The patient received chemotherapy with TRK inhibitor larotrectinib and carboplatin, which caused shrinkage of metastatic lung nodules. This is the first report of cervical BSCC with extensive molecular workup, which detected multiple genetic events, including targetable ones, which are potentially implicated in the development of a tumor. The accumulation of data and further studies on this tumor are necessary to define its diagnostic criteria and its clinical and biological behavior.

Prevalence of BRAFV600E mutation in Asian series of papillary thyroid carcinoma-a contemporary systematic review.

Papillary thyroid carcinoma (PTC), the most common malignancy of the endocrine system, is frequently driven by BRAFV600E mutation, which was reported in 35-60% cases in Western series. Numerous studies have recently emerged from Asian countries and regions; however sufficient summary is lacking to date. BRAF mutation serves as a diagnostic and prognostic tool in thyroid cancer, therefore establishing a rate of BRAF on the national scale could be of practical significance. We performed systematic reviews of available literature to investigate the prevalence of BRAF mutation in series of PTC from various Asian countries and regions. Out of the total 3,966 reports identified via initial screening, 138 studies encompassing over 40,000 PTCs were included for the final analysis. A vast majority (90.2%) of PTCs with known BRAF status were from East Asia, including China, South Korea, and Japan, with BRAF mutation rates of 71.2%, 75.5%, and 70.6%, respectively. Less abundant Indian and Saudi Arabian series found 45.6% and 46.3% prevalence of BRAFV600E in PTC, respectively. Much limited evidence was available from Thailand, Iran, Kazakhstan, Taiwan, Singapore, Indonesia, Hong Kong, Philippines, Vietnam, Iraq, and Myanmar. No relevant publications were found from other highly populated countries, such as Pakistan, Bangladesh, and Malaysia. After grouping by geographic region, we found that the highest rate of BRAFV600E was reported in the PTC series from East Asia (76.4%). Much lower rate (45-48%) was seen in PTC cohorts from South Asia, Central Asia, and the Middle East while the Southeast Asian series were in between (57%). Further subgroup analysis revealed that studies employing fresh frozen tissue and fine-needle aspirates showed higher rates of BRAF compared to those used formalin-fixed paraffin-embedded tissues. We found that the PTC series enrolled patients' cohorts after 2010 demonstrated a higher rate of BRAF compared to the earlier series. Finally, pediatric PTCs had lower BRAF prevalence compared to the baseline rate for the country. In conclusion, despite considerable among and within countries heterogeneity, the Asian PTC series showed a higher prevalence of BRAFV600E mutation than that in Western series. Causes of geographic heterogeneity, whether genuine (etiology, genetics) or methodology-related should be further investigated.

Endometrial carcinoma arising in a bicornuate uterus.

Endometrial carcinomas arising in a bicornuate uterus are rare, only five case of which have been previously reported. We present a case of endometrial cancer arising in a bicornuate uterus, occurring in a 65-year-old woman. Unlike previously reported cases, our case showed mixed endometrial adenocarcinoma and undifferentiated carcinoma in one horn and focal adenocarcinoma in the other. Adequate tissue sampling of both horns is necessary for accurate diagnosis of malignancy in patients with a bicornuate uterus. Physicians should be aware of the possibility of this abnormality in cases when endometrial cancer is suspected but histology fails to confirm.

Comparison of the performance of the NucliSENS EasyQ HPV E6/E7 mRNA assay and HPV DNA chip for testing squamous cell lesions of the uterine cervix.

This study aims to evaluate the clinical performance of the NucliSENS EasyQ assay and compare it with HPV DNA genotyping for the detection of high-grade squamous intraepithelial lesions (HSIL) and cancer in a Korean population. In 188 total thin prep samples, the remaining fluid after cytology slide preparation was tested with Goodgene HPV DNA chips and the NucliSENS EasyQ HPV E6/E7 messenger RNA (mRNA) assay. The sensitivity and specificity of each test were calculated with HSIL and squamous cell carcinoma (SCC) as the disease endpoint. Out of the 188 samples, 139 (74%) were positive for DNA of 14 HPV types, while 57 (30%) cases were positive for E6/E7 mRNA. The DNA test was positive in cytology cases of SCC, HSIL, and atypical squamous cell. The mRNA test yielded results of 75%, 74%, 60%, 56%, and 29% positivity in abnormal cytology cases of SCC, HSIL, atypical squamous cells - cannot exclude HSIL, atypical squamous cells of undetermined significance, and low-grade squamous intraepithelial lesion, respectively. In normal cytology cases, the positivity rates were 9% and 53% for the mRNA and DNA tests, respectively. For detection of HSIL and SCC, the sensitivity of the mRNA test was 74.36% and that of the DNA test was 100%, while the specificities of the tests were 85% and 40.83%, respectively. These findings suggest that the HPV E6/E7 mRNA assay can overcome the shortcoming of low specificity of DNA assays for clinical detection of high-grade cervical lesions and malignancies.

DNA methylation status of a distinctively different subset of genes is associated with each histologic Lauren classification subtype in early gastric carcinogenesis.

DNA methylation change is known to play a crucial role in early gastric carcinogenesis. The present study aimed to identify and validate the correlation between differentially methylated regions (DMRs) and the subtypes of early gastric cancers (EGCs). Illumina Infinium methylation assay (IIMA; 450K BeadChip kit) was performed on fresh tumor and non­tumor tissues of 12 EGCs to screen the methylation status of 450,000 CpG sites. To evaluate the significance of DNA methylation in each histologic subtype, pyrosequencing assay (PA) was performed on 38 EGCs (18 intestinal-, 12 mixed- and 8 diffuse-type) using 12 genes selected from the screening. Between tumors of the intestinal-type (n=6), and diffuse- (n=4) plus mixed-types (n=2), 169 regions showed significant differences (intensity>3,000, Δß>0.2) in IIMA. Hierarchical clustering using the 169 DMRs revealed distinct separation between the two groups. In PA using 12 selected genes from the IIMA results, the aberrant methylation statuses of DVL2 (p=0.0186) and ETS1 (p=0.0222) were significantly related to diffuse- and mixed-types rather than the intestinal-type, while C19orf35 (p=0.019) and CNRIP1 (p=0.0473) were related to the diffuse­type rather than intestinal­type, and GAL3ST2 (p=0.0158) and ITGA3 (p=0.0273) were related to the mixed-type rather than the other two types. The methylation of other genes, CLIP4, XKR6, CCDC57, MAML3 and SDC2, was related with age, tumor location, or Helicobacter infection rather than the histologic subtype. Aberrant DNA methylation of certain genes may be independently involved in each histologic subtype of EGC. Furthermore, mixed-type EGCs may be a distinctive histologic subtype based on the different subset of DMRs compared to those of other subtypes.