CAP TODAY Pathology/Laboratory Medicine/Laboratory Management
Editors: Rouzan Karabakhtsian, MD, PhD, professor of pathology and director of the Women’s Health Pathology Fellowship, Albert Einstein College of Medicine, Montefiore Medical Center, Bronx, NY; Shaomin Hu, MD, PhD, staff pathologist, Cleveland Clinic; S. Emily Bachert, MD, breast pathology fellow, Brigham and Women’s Hospital, Boston; and Amarpreet Bhalla, MD, assistant professor of pathology, Albert Einstein College of Medicine, Montefiore Medical Center.
Use of subtyping to predict behavior of papillary thyroid microcarcinomas
February 2022—The most common malignant neoplasm affecting the thyroid gland is papillary thyroid carcinoma. The disease can be classified based on a number of morphologic variants, including, but not limited to, classic, follicular, and tall cell. Each of the morphologic subtypes has distinct clinical characteristics, some of which cause certain variants, such as tall cell, to behave more aggressively than others. Papillary thyroid carcinomas measuring no more than 1 cm are classified as microcarcinomas. Although these lesions are thought to be clinically indolent, the authors hypothesized that, as with their larger counterparts, certain histologic variants may lead to worse patient outcomes. To test this hypothesis, they searched the pathology archives at the Hospital of the University of Pennsylvania from 2009 to 2020 for cases of papillary thyroid microcarcinomas. Using these cases, the authors assessed whether different morphologic features correlated with more aggressive clinical behavior. Their findings suggested that certain variants exhibit features that portend a more worrisome clinical course. Therefore, papillary thyroid microcarcinomas should be subtyped to help predict patient outcomes.
Gubbiotti MA, Livolsi V, Montone K, et al. Papillary thyroid microcarcinomas: Does subtyping predict aggressive clinical behavior? Hum Pathol. 2021;114:28–35.
Correspondence: Dr. Maria A. Gubbiotti at maria.gubbiotti@pennmedicine.upenn.edu
Gastrointestinal pathology in samples from COVID-19–positive patients
Although primarily considered a respiratory illness, COVID-19 can have gastrointestinal manifestations. The authors conducted a study to evaluate histopathology and in situ hybridization for SARS-CoV-2 in gastrointestinal samples from patients with recent and remote COVID-19. The study comprised patients who had positive SARS-CoV-2 nasopharyngeal tests and provided a gastrointestinal tissue specimen. SARS-CoV-2 in situ hybridization (ISH) was performed on each sample, and SARS-CoV-2 next-generation sequencing (NGS) was performed on a subset of the samples. Twenty-five patients met the criteria for inclusion in the study. Five had positive SARS-CoV-2 nasopharyngeal tests within seven days of their gastrointestinal procedure. Two of the five were ulcerative colitis patients on steroid therapy who lacked typical COVID-19 symptoms. Their colectomies showed severe ulcerative colitis, and one of the two patients demonstrated SARS-CoV-2 by NGS but a negative ISH. Of the three others, one had an ischemic colon resected as a complication of COVID-19; however, both ISH and NGS were negative. Another had a normal-appearing terminal ileum but positive ISH and NGS. And another had ileal ulcers with SARS-CoV-2 negativity by both modalities. The remaining 20 patients had positive nasopharyngeal tests an average of 53 days prior to the gastrointestinal procedure. None of their samples demonstrated SARS-CoV-2 ISH positivity, but one was positive on NGS despite a negative nasopharyngeal test. The gastrointestinal findings from the SARS-CoV-2–infected patients ranged from normal with virus detected by ISH and NGS to bowel ischemia secondary to systemic viral effects without evidence of virus in the tissue. No distinct histologic finding was identified in those with gastrointestinal tissue specimens demonstrating SARS-CoV-2 positivity in this cohort.
Westerhoff M, Jones D, Hrycaj SM, et al. Gastrointestinal pathology in samples from coronavirus disease 2019 (COVID-19)–positive patients. Arch Pathol Lab Med. 2021;145(9):1062–1068.
Correspondence: Dr. Maria Westerhoff at mwesterh@med.umich.edu
Assessment of an artificial intelligence system for prostate cancer detection
The diagnosis of prostate carcinoma is usually made via prostate core needle biopsies obtained through a transrectal approach. Making diagnoses using these biopsies may account for a significant portion of pathologist workloads, yet variability in pathologists’ experience and expertise, as well as fatigue, may adversely affect the reliability of cancer detection. Machine-learning algorithms are increasingly being developed as tools to improve diagnostic accuracy in anatomic pathology. The Paige Prostate AI-based system is one such tool. The machine-learning algorithm was trained on the digital slide archive of Memorial Sloan Kettering Cancer Center (MSKCC) and categorizes a prostate biopsy whole slide image as “suspicious” or “not suspicious” for prostatic adenocarcinoma. The authors conducted a study to evaluate the performance of Paige Prostate on prostate biopsies procured, processed, and independently diagnosed at Yale Medicine. The analysis included 1,876 prostate core biopsy whole slide images. Paige Prostate categorizations were compared to the pathology diagnoses originally rendered on the glass slides for each core biopsy. Discrepancies between the rendered diagnosis and categorization by Paige Prostate were manually reviewed by genitourinary pathologists. Paige Prostate showed a sensitivity of 97.7 percent and positive predictive value of 97.9 percent, as well as a specificity of 99.3 percent and negative predictive value of 99.2 percent in identifying core biopsies with cancer using a data set derived from MSKCC. Areas for improvement were identified in Paige Prostate’s handling of poor quality scans. Overall, these results demonstrate the feasibility of porting a machine-learning algorithm to an institution remote from its training set. They also highlight the potential of such algorithms as a powerful workflow tool for evaluating prostate core biopsies in surgical pathology practices.
Perincheri S, Levi AW, Celli R, et al. An independent assessment of an artificial intelligence system for prostate cancer detection shows strong diagnostic accuracy. Mod Pathol. 2021;34:1588–1595.
Correspondence: Dr. Sudhir Perincheri at sudhir.perincheri@yale.edu
Expression patterns for Bcl-2, EMA, β-catenin, E-cadherin, PAX8, and MIB-1 in thymomas
The expression of immunohistochemical markers has been extensively investigated in thymomas to assist in the differential diagnosis of these tumors. The authors studied six markers to determine their utility in evaluating the tumors. A series of 126 thymomas, including 33 type A, 27 type AB, 20 type B1, 22 type B2, and 24 type B3, were examined using a tissue microarray technique with antibodies to E-cadherin, β-catenin, PAX8, Bcl-2, EMA, and MIB-1. Keratin AE1/AE3 and p63 were used for quality control. A significant finding was strong and consistent positivity for Bcl-2 in type A (90 percent) and type AB (88.8 percent) thymoma, while 100 percent of types B1, B2, and B3 were negative for Bcl-2. The distribution of E-cadherin and β-catenin was not useful for differential diagnosis. E-cadherin and β-catenin were expressed in a high proportion of the tumors (92 to 100 percent), except in B2 thymoma, which showed only 45 percent expression of E-cadherin. A significant increase in expression of the MIB-1 proliferation marker (mean, 12.8 percent nuclear positivity) was also observed in B3 thymoma but not the other histologic types. Statistical significance was confirmed using Kruskal’s nonparameterized test for distribution. EMA was generally negative, except for spindle cells in the fibrous septa in types A and AB thymoma. PAX8 showed less consistent nuclear staining than p63 and was only widely expressed in 55.7 percent of cases. Bcl-2 may serve as a useful marker to separate spindle cell thymomas (types A and AB) from the other types, and the MIB-1 proliferation index may help differentiate B2 from B3 thymoma.