Faculty Bibliography

Amiodarone-induced liver injury (AILI) is a known risk of amiodarone therapy, with presentations ranging from asymptomatic aminotransferase elevations to severe or fatal hepatitis and cirrhosis. Due to limited understanding of its histopathologic features, we conducted a retrospective cross-sectional re-analysis of liver biopsy samples from patients on amiodarone from two centers. Of the 48 liver biopsy samples, 42 (87%) exhibited histologic evidence of AILI. All patients showed minimal or mild macrovesicular steatosis. Ballooned hepatocytes were observed in 36 cases (86%), with 25 (69%) displaying a periportal distribution, 8 (22%) centrilobular, and 3 (8%) panacinar in distribution. Mallory-Denk bodies were found in 36 samples (76%)-18 (50%) were numerous and 18 (50%) multiple. Cholestasis was present in 10 patients, 7 (70%) of whom died. In contrast, 10 (31%) of the 32 patients without cholestasis died. This represents a significantly increased mortality risk for patients with AILI and cholestasis (p = 0.03). While AILI shares features with the more generally known metabolic dysfunction-associated steatotic liver disease, our findings indicate that a prominence of periportal distribution of ballooned hepatocytes and Mallory-Denk bodies despite a minimum of macrovesicular steatosis are characteristic of AILI. Furthermore, cholestasis in biopsy samples may suggest a poorer prognosis in patients on amiodarone.

BACKGROUND:Mutations in DICER1 are uncommon, poorly understood, and infrequently found in thyroid nodules. METHODS:The objective of this study was to investigate category III/IV thyroid nodules according to The Bethesda System for Reporting Thyroid Cytopathology with DICER1 gene mutations detected in fine-needle aspiration cytology samples using ThyroSeq v3 molecular testing, with a focus on an exploration of the clinical and histopathologic outcomes of these nodules. In this multicenter study spanning more than 6 years, nodules were retrospectively analyzed for patient demographics, clinical course, cytologic features, and histopathology, where available. RESULTS:In total, 88 patients with somatic DICER1 mutations were included, with a mean age of 39.6 years and a female predominance. All mutations were in the somatic hotspot region, most commonly at the codon 5437 site. Most excised nodules showed benign histologic features (65.9%). Interestingly, the rate of malignancy was higher in this cohort compared with that in the national average. CONCLUSIONS:DICER1 mutations appear to confer a higher risk of malignancy, but are not associated with any specific cytological or histopathological distinguishing features.

INTRODUCTION/BACKGROUND:Papillary thyroid cancer (PTC) often follows an indolent course with a favorable prognosis. This has led to evolving guideline-based, low-intensity treatment options for low-risk patients. Recently, the purported benefit of total thyroidectomy (TT) over unilateral lobectomy for PTC with clinical lateral neck nodal metastasis (cN1b) has come into question. MATERIALS AND METHODS/METHODS:A retrospective analysis of the National Cancer Institute's Surveillance, Epidemiology, and End Results database was performed to study patients with PTC with ipsilateral cN1b disease from 1975 to 2020. Kaplan-Meier curves and log-rank tests were used to compare disease-specific survival (DSS) difference between lobectomy and TT at 10 y. Multivariable Cox proportional hazards analysis was performed to determine the independent association of lobectomy versus TT with DSS, correcting for age and lymph node ratio, defined as the ratio of pathologically positive lymph nodes to total number examined. RESULTS:Among 2943 patients (median [interquartile range] age, 45 [26] y), 42 underwent lobectomy and 2901 underwent TT. Unadjusted DSS at 10 y in the lobectomy and TT groups were 51.0% (95% confidence interval, 31.4%-82.8%) and 86.8% (95% confidence interval, 84.8%-88.9%), respectively. On multivariable analysis of all patients, older age (hazard ratio [HR], 1.08; P < 0.001) and male gender (HR, 1.74; P < 0.001) were associated with lower adjusted DSS, whereas treatment with TT (HR, 0.387; P = 0.005) and receipt of radioactive iodine (RAI) (HR, 0.604; P < 0.001) were associated with improved adjusted DSS. In addition, we observed that the magnitude of survival benefit conferred by RAI and TT were reduced with decreasing age (P < 0.001). CONCLUSIONS:This longitudinal cohort study suggests that, while TT is associated with a DSS benefit in most patients with PTC and ipsilateral cN1b disease, this association may not exist in a smaller cohort of younger patients. These findings raise the possibility that unilateral surgical clearance without RAI could offer adequate oncologic outcomes in selected younger individuals; however, further investigation is warranted to confirm these observations and inform clinical decision-making.

OBJECTIVE/UNASSIGNED:Early detection and removal of suspicious polyps during routine colonoscopies play an important role in reducing the risk of colorectal cancer. Patient management and follow-up are determined by the type of polyps removed and the degree of dysplasia present on histological evaluation. Whereas discerning between a benign polyp and a dysplastic one is a trivial task, distinguishing between tubular adenomas with low-grade dysplasia (LGD) and high-grade dysplasia (HGD) is a challenging task. In this study, we trained a deep learning model to distinguish between colorectal adenomas with LGD and HGD. DESIGN/UNASSIGNED:We retrieved 259 slides of adenomatous polyps taken between January 2011 and October 2024. Slides with HGD were reviewed by a subspecialty-trained GI pathologist. After excluding discordant and duplicate cases, 200 slides remained: 71 (35.5%) with HGD and 129 (64.5%) with LGD. The slides were divided into training (160 slides, 80%) and test (40 slides, 20%) sets. After patch generation and stain normalization, a ResNet34 model (pre-trained on ImageNet) was trained using 5-fold cross-validation. Slide classification was determined by aggregating patch-level predictions. RESULTS/UNASSIGNED:The model's slide-level prediction accuracy was 95.0%, correctly classifying all but 2 out of 40 slides. The model achieved an area under the receiver operating characteristic curve score of 0.981 and an F1 score of 0.923. CONCLUSIONS/UNASSIGNED:This study demonstrates that deep learning models can accurately distinguish between colonic adenomas with LGD and HGD. Training on a larger dataset could increase the accuracy and generalizability of the model and should be a focus of further studies.

BACKGROUND/UNASSIGNED:While traditional risk stratification for thyroid cancer primarily focused on tumor pathology, molecular profiling is increasingly recognized for its clinical relevance. There are limited data on tumors with three or more mutations, especially when presenting with non-malignant cytology. This study aims to evaluate the clinical behavior of cytologically non-malignant thyroid nodules with three or more potent oncogenic mutations. METHODS/UNASSIGNED:Electronic medical records of patients at our institution with thyroid nodules who underwent molecular testing were reviewed to identify cases with cytologically benign or indeterminate thyroid nodules that harbored three or more oncogenic mutations. Clinical, cytological, and molecular data were analyzed to assess tumor behavior. RESULTS/UNASSIGNED:Four of six cases were histologically malignant at index surgery. One histologically benign case developed distant metastases 7 years later. Retrospective analysis of the two histologically benign cases and the one case with low-risk histology demonstrated significant intratumoral heterogeneity. The benign case that developed distant metastases was found to have an area of intratumoral heterogeneity whose genetic profile matched the metastases. CONCLUSIONS/UNASSIGNED:Despite bland cytological and ultrasonographic features, these tumors exhibited aggressive behavior. The molecular profile of thyroid cancers should be considered when determining treatment, especially in cases with multiple high-risk mutations, as they may behave more aggressively than predicted by cytology or imaging alone.

OBJECTIVES/OBJECTIVE:Fine needle aspiration (FNA) plays a crucial role in their initial assessment of salivary gland neoplasms. In the Milan System for Reporting Salivary Gland Cytopathology (MSRSGC), the category of Salivary Gland Neoplasm of Uncertain Malignant Potential (SUMP) categorizes lesions with ambiguous features. This study aims to investigate the risk of neoplasm (RON) and risk of malignancy (ROM) within different subgroups of SUMP lesions using data from three large academic institutions. METHODS:We analyzed salivary gland (FNA) cases from three academic institutions post-MSRSGC implementation. Salivary gland FNA cases categorized as Milan IVB (SUMP) with subsequent surgical pathology follow-up were analyzed. Cases were divided into basaloid, oncocytic, and clear cell SUMP subtypes, with RON and ROM assessed and compared. RESULTS:Out of 1377 MSRSGC cases, 231 were SUMP (16.8%), with 101 subjected to surgical pathology follow-up. The overall ROM for SUMP was 20.8%, with variations of 10% to 29.5% observed amongst institutions, but no significant difference was observed among three institutions (p = 0.15). Basaloid and oncocytic SUMP displayed 17.1% and 20.5% ROM, respectively, without significant disparity. However, all clear cell SUMP cases were malignant on surgical resection. CONCLUSIONS:This study highlights the variability in ROM for SUMP lesions and the significantly higher ROM in SUMP cases with clear cell features. These findings emphasize the importance of accurately subcategorizing SUMP lesions, particularly those with clear cell features, for appropriate clinical management.

This study presents two cases of lipid-rich pancreatic neuroendocrine tumors (PanNETs), a rare variant posing significant diagnostic challenges in fine needle aspiration (FNA) cytology and small biopsies. The first case involves an elderly male with a pancreatic tumor, displaying distinct cytoplasmic vacuoles, while the second case is a middle-aged male present with a pancreatic tail mass exhibiting foamy cytoplasm and eccentric nuclei, infiltrating in the stroma. Both cases did not exhibit typical morphologic features of PanNET but demonstrated cytomorphologic features and infiltrative growth patterns that mimicked adenocarcinoma. Further work-up demonstrated that both tumors were immunoreactive for synaptophysin and chromogranin, and were interpreted as well-differentiated, PanNET, lipid-rich variant. The study highlights the overlapping morphological features between lipid-rich PanNETs and other pancreatic neoplasms and underscores the importance of comprehensive cytological and immunohistochemical analysis for accurately diagnosing this variant, particularly due to the risk of misinterpreting it as pancreatic adenocarcinoma. Recognizing lipid-rich PanNETs is crucial for appropriate clinical management, as their identification can significantly impact treatment decisions and patient outcomes.