Large Language Models as a Rapid and Objective Tool for Pathology Report Data Extraction.

Medical institutions continuously create a substantial amount of data that is used for scientific research. One of the departments with a great amount of archived data is the pathology department. Pathology archives hold the potential to create a case series of valuable rare entities or large cohorts of common entities. The major problem in creation of these databases is data extraction which is still commonly done manually and is highly laborious and error prone. For these reasons, we offer using large language models to overcome these challenges. Ten pathology reports of selected resection specimens were retrieved from electronic archives of Koç University Hospital for the initial set. These reports were de-identified and uploaded to ChatGPT and Google Bard. Both algorithms were asked to turn the reports in a synoptic report format that is easy to export to a data editor such as Microsoft Excel or Google Sheets. Both programs created tables with Google Bard facilitating the creation of a spreadsheet from the data automatically. In conclusion, we propose the use of AI-assisted data extraction for academic research purposes, as it may enhance efficiency and precision compared to manual data entry.

Novel ALK immunohistochemistry assay (clone OTI1A4, Dako) is a sensitive, reliable marker for identifying ALK rearrangements in lung adenocarcinomas: A validation study.

OBJECTIVES: We present the first study validating the recent Dako ALK assay (clone OTI1A4, in vitro diagnostic) for detecting ALK rearrangements in lung adenocarcinoma. METHODS: Lung adenocarcinoma cases between 2011 and 2023 were retrospectively collected to create a cohort of 203 samples. Cases were stained with Dako ALK OTI1A4 and Ventana ALK D5F3 and reviewed by 3 pathologists independently. Correlation between assays, including their sensitivity and specificity, was evaluated. RESULTS: The cohort (n = 203) consisted of resections, core needle biopsies, and cell blocks. Agreement between Dako ALK OTI1A4 and Ventana ALK D5F3 assays was "almost perfect," with κ = 0.89. The sensitivity and specificity of the Dako ALK OTI1A4 assay were 93.3% and 96%, respectively, in a subgroup of 55 molecularly confirmed cases (n = 30 with and n = 25 without ALK rearrangement). CONCLUSIONS: Immunohistochemistry-based assays provide a valid and reasonably priced alternative, especially in settings where molecular confirmatory tests are neither offered nor accessible. Given high interassay and molecular concordance, we propose that the novel Dako OTI1A4 assay can be reliably used to identify cases with ALK rearrangement.

18F-FDG PET/CT texture analysis of anthracotic lymph nodes detected with EBUS and comparison with cytological findings.

OBJECTIVE: Lymph node metastasis is the most important factor both in the selection of treatment since many alternatives have been created in recent years, and in the evaluation of prognosis in lung cancer. The most unpredictable cause of lymph node false positivity in fluorine-18-fluorodeoxyglucose (18F-FDG) positron emission tomography/computed tomography (PET/CT) is anthracosis. The aim of this study is to compare 18F-FDG PET/CT texture information of anthracotic (ALN) and metastatic (MLN) lymph nodes, after re-evaluation of the cytological samples obtained from anthracotic lymph nodes by EBUS-TBNA. SUBJECTS AND METHODS: Ninety nine patients, 78 of whom had primary lung cancer were included in the study. Two hundred and three lymph nodes from 99 patients sampled by EBUS-TBNA and diagnosed cytologically as ALN or MLN were evaluated retrospectively. All ALN were classified as grades 1, 2 and 3 cytologically. Volume of interest (VOI) of 203 lymph nodes was re-drawn and maximum standardized uptake value (SUVmax), metabolic tumor volume (MTV) and total lesion glycolysis (TLG) values were recorded. RESULTS: There was a statistically significant difference in MTV and TLG values in MLN and all ALN grades. However, only grade 1-2 ALNs could be differentiated from MLNs with SUVmax, and no statistically significant difference was found in grade 3 ALN and MLN. Metabolic tumor volume and TLG values over 4.10cm3 and 26.57 showed 60% and 59% sensitivity and 83% and 94 specificity respectively for the identification of MLN. CONCLUSION: The contribution of MTV and TLG values of 18F-FDG PET/CT to the differential diagnosis of ALN is much more valuable than SUVmax values, especially for grade 3 anthracosis. It was thought that cytological reporting of only grade 3 ALN could make a better contribution to the 18F-FDG PET/CT evaluation analysis.

Warthin-Like Mucoepidermoid Carcinoma: A Morphological Spectrum - A Report of 3 Cases with Histological and Cytological Findings and Review of the Literature.

BACKGROUND: Mucoepidermoid carcinoma (MEC) showing Warthin's tumor (WT)-like features is a low-grade malignancy which should be differentiated from WT. Morphological features may be distinctly different in each case, causing diagnostic difficulties. CASE PRESENTATION: Three cases were presented and discussed with their morphologies. All cases that presented with a mass in the parotid gland went to parotidectomy, and all had preoperative fine-needle aspirations (FNAs). Case 1 was a 16-year-old female; FNA was suggestive of WT and initially interpreted as WT histologically. Case 2 was a 27-year-old male; FNA was interpreted as noninformative due to the presence of cyst fluid only. Case 3 was a 53-year-old male and cytologically was found to be suspicious for MEC which contained squamous and goblet cells on a mucoid background. On histopathological examination, case 2 and case 3 were morphologically consistent with low-grade MEC with WT-like features. Prominent lymphoid stroma and the cystic pattern were the characters of these tumors. Case 1 had the classical WT appearance with some mucinous and squamous metaplasia which could only be interpreted as MEC after the detection of MAML2 rearrangement by FISH. The other 2 showed either focal or relatively diffuse usual low-grade MEC findings, and case 3 was also confirmed by MAML2 rearrangement. CONCLUSION: Cytological and histopathological features revealed a spectrum. Differentiating WT-like MECs from ordinary WTs may be challenging. On the one end of the spectrum, they may look very much like WT, and on the other end, even though usual MEC features are present, still, WT-like appearance may pose diagnostic difficulty. Showing MAML2 rearrangement in these cases is very helpful. The presence of mucinous and squamous cells in an otherwise WT-like looking tumor should be alarming for MEC, and if possible, each case should be analyzed for MAML2 rearrangement.

What Would Next Generation Sequencing Bring to the Diagnosis and Treatment of Sarcomas? A Series of 20 Cases, a Single Institution's Experience.

OBJECTIVE: Soft tissue tumors comprise a small proportion of a pathologist's routine practice. Although morphology and immunohistochemistry are quite helpful for diagnosing these tumors, many require molecular tests. Fluorescence in-situ hybridization has been the most commonly used method for the detection of specific genomic alteration, but next generation sequencing (NGS) could be more informative in many ways. Here we present our targeted NGS experience on soft tissue tumors with a series of 20 cases. MATERIAL AND METHOD: The Laboratory Information System (LIS) was screened for soft tissue tumors that had been sequenced by NGS (between January 2018 - February 2021). 20 consecutive cases were included in the study. All cases were sequenced using a commercial targeted sequencing panel designed for soft tissue tumors. RESULTS: We were able to run a reliable sequencing study for 16 (80%) of the cases but 4 (20%) of them failed in quality tests. We have found pathogenic alterations in 12 (60%) of the cases. The most common alterations were EWSR1 fusions, FLI1 being the most common partner. NGS results drastically changed the initial diagnosis, and thus the treatment modalities, in 3 cases (15%): the case with ETV6-NTRK3 fusion, the case with FUS-TFCP2 fusion, and the case of rhabdomyosarcoma (RMS) that was favored to be of the alveolar subtype and turned out to lack FOXO1 fusions. CONCLUSION: A targeted NGS panel is robust and very informative. It not only allows pathologists to further specify and/or confirm their diagnosis but it could also play an important role in predicting the outcome.

Pathologic Evaluation of Endoscopically Resected Non-Ampullary Duodenal Lesions: A Single Center Experience.

OBJECTIVE: Endoscopic resections are increasingly being used for superficial gastrointestinal lesions. However, application of these techniques in the duodenum remains challenging, due to the technical difficulties and high complication rates. This study projects a western tertiary center's experience in the endoscopic treatment and diagnostic workup of 19 cases of non-ampullary duodenal lesions. MATERIAL AND METHOD: Specimens (12 endoscopic mucosal resections, 6 endoscopic submucosal dissections, and one endoscopic full-thickness resection) were processed following a strict protocol (photographed, mapped digitally and submitted totally) for histopathologic examination. Clinicopathologic characteristics, margin status and follow-up information were analyzed. RESULTS: The mean age of the 16 patients was 52 years (range: 22-81). Mean lesion size was 1.4 cm (range: 0.3-3.6 cm) for all cases, 2 cm for endoscopic submucosal dissections and 1.1 cm for endoscopic mucosal resections. Mean number of blocks submitted was 4/case. Seven neuroendocrine tumors, 3 tubulovillous adenomas were diagnosed along with nine benign lesions. For endoscopic submucosal dissections, en-bloc and R0 resection rates were 100% (n=6/6) and 83% (n=5/6); for endoscopic mucosal resections, they were 92% (n=11/12) and 83% (n=10/12), respectively. Only one patient had procedure-related late perforation that was managed endoscopically. No mortality was encountered. CONCLUSION: Duodenal endoscopic resections proved successful, safe and feasible methods in a tertiary center. The pathologist's role is to designate the accurate diagnosis, related histopathologic parameters and margin status. The gross protocol was found to be essential in evaluating specimen margins and orientation, as well as in size measurement. We recommend following a standardized approach including gross photography and digital mapping when handling these specimens, for both diagnostic and data collection purposes.