Faculty Bibliography

INTRODUCTION/UNASSIGNED:The International Association for the Study of Lung Cancer grading system for resected invasive lung adenocarcinoma provides robust prognostic stratification for recurrence risk and overall survival. However, most patients present at advanced stages, where only small biopsy samples are available, leaving no established prognostic tool for such cases. We propose a practical biopsy-based grading method for pulmonary adenocarcinoma. METHODS/UNASSIGNED:We analyzed 267 paired biopsies and resections. The following two biopsy grading (bxG) models were tested:•Model 1: Based on the predominant pattern-grade 1 (lepidic), grade 2 (acinar/papillary), grade 3 (solid/micropapillary/complex glandular).•Model 2: Any percentage of high-grade patterns classified as grade 3, regardless of predominance. BxG 1 and 2 are defined as predominant patterns as described.For analysis, grades 1 and 2 were combined (G1-2), and grade 3 was considered a separate category. Diagnostic performance was assessed using resections as the reference standard. Interobserver reliability was evaluated, and discrepant cases were reviewed. RESULTS/UNASSIGNED:= 21) were primarily due to sampling limitations and artifacts, such as tangential sectioning. Adjusting for these factors improved interobserver agreement from moderate to almost perfect. CONCLUSION/UNASSIGNED:Biopsy grading using the presence of any high-grade pattern correlates well with International Association for the Study of Lung Cancer-based resection grading. Although bxG1 to 2 biopsies may underestimate tumor grade, bxG3 biopsies demonstrate more than 90% specificity and PPV for predicting G3 resections, supporting their potential utility as a clinical prognostic marker.

BACKGROUND:Prediction of treatment outcomes is essential for improving clinical management, particularly in patients with soft tissue sarcoma, where treatment options remain suboptimal. Given the limitations of current therapies, there is increasing interest in combining neoadjuvant radiotherapy with immunotherapy, referred to as neoadjuvant radioimmunotherapy (NRIT). We aim to develop a predictive model for assessing pathologic treatment response in patients undergoing NRIT or chemoradiotherapy, integrating radiomic features with radiologist assessments, clinical data, and pathology findings. MATERIALS AND METHODS/METHODS:Radiomic and semantic features were extracted from pre- and post-treatment MRI scans. The XGBoost algorithm was used for feature selection and model development. Models included a model based on clinical variables and semantic features, a model based on radiomic features and clinical features and a model using all available features. RESULTS:Study cohort included 213 patients (mean age of 54 years, male/female of 1.6). There were 17 patients in the prospective arm. The best model used all radiomic, clinical, and semantics features. It achieved an area under the receiver operating characteristic curve (AUC) of 0.72 (95% CI = 0.51-0.89) on the hold-out testing set. CONCLUSION/CONCLUSIONS:Multi-modal radiomic-based models are effective in identifying patients at higher risk of non-response to neoadjuvant therapy. Furthermore, the performance of multi-modal radiomics-based models exceeded those based solely on radiologist evaluations. Our findings underscore the potential of radiomics in enhancing precision medicine by enabling identification of treatment response in STS patients undergoing NRIT before surgical excision of the tumor.

BACKGROUND:The widespread use of next-generation sequencing has allowed refinement of the classification and diagnosis of salivary gland neoplasms, leading to identification of recurrent gene fusions in a majority of salivary gland carcinoma types, and characterization of several novel entities. A small proportion of salivary gland carcinomas do not meet the diagnostic criteria for any known tumor type and are therefore classified as "salivary gland carcinoma, not otherwise specified". Given the ever-growing arsenal of tools to classify these lesions, the number of cases diagnosed as such is expected to continue decreasing. CASE PRESENTATION/METHODS:In this article we describe a novel DLG1::BRAF fusion in a high grade salivary gland carcinoma arising in the tongue of a 78 year-old woman. This tumor had solid and cribriform architectural features and was composed of a dual population of abluminal and mucinous cells. Immunohistochemically, it had variable SOX10 expression, variable and strong MUC4 reactivity and expression of p63 and p40 (delta Np63) in the abluminal cell population. The gene fusion retained the kinase domain of BRAF, without the self-inhibitory CR1 domain, which is expected to lead to upregulation of BRAF protein. The patient had a complete resection of her tumor, without evidence of local recurrence or metastasis 10 months after diagnosis. CONCLUSION/CONCLUSIONS:These findings may represent a previously undescribed type of salivary gland tumor. However, additional reports of similar lesions are necessary for definitive characterization.

Thoracic SMARCA4-deficient undifferentiated tumor (SMARCA4d-UT) is an uncommon, aggressive lung neoplasm associated with smoking and characterized by loss of SMARCA4 (BRG-1) expression. Although originally considered to be a primary sarcoma, there is growing evidence that these lesions may represent transformation of conventional non-small cell carcinoma. In this study, we probe this relationship based on the clinical, histologic and molecular findings of 18 SMARCA4-deficient malignancies of the lung. Cases diagnosed as SMARCA4d-UT and SMARCA4-deficient carcinoma were retrospectively reviewed, including histologic and immunophenotypic features, and next generation sequencing studies. Of the 18 tumors, 5 were considered to represent undifferentiated SMARCA4d-UT, and 13 SMARCA4-deficient carcinomas, including 11 adenocarcinomas, 1 squamous cell carcinoma, and 1 poorly differentiated non-small cell carcinoma. All 13 carcinomas had a morphologically identifiable undifferentiated component. Survival outcomes were similar in both SMARCA4d-UT and carcinomas. Genetic alterations often seen in lung cancer were identified in 8 cases, including mutations in EGFR (in 2 SMARCA4-deficient adenocarcinomas), KRAS (1 SMARCA4d-UT and 1 SMARCA4-deficient adenocarcinoma), MAP2K1 (1 SMARCA4-deficient adenocarcinoma), and a gene fusion involving EML4::ALK (1 SMARCA4d-UT). The patient with EML4::ALK fusion was treated with alectinib with partial response. Fusions involving BRAF::CHCHD3 and FGFR1::FILIP1 were identified in 2 SMARCA4-deficient adenocarcinomas. High expression of PD-L1 (TPS >50 %) was seen in 12 cases (67 %). These finding further suggest that SMARCA4d-UT and carcinomas with SMARCA4 loss may be on the same spectrum of disease, and accurate histologic distinction between these lesions may be challenging. A unified terminology may be beneficial for appropriate diagnosis and treatment.

Dermatofibrosarcoma protuberans (DFSP) is a neoplasm of the dermis with a tendency for aggressive local growth and recurrence. A minority of DFSP cases may transform into higher-grade sarcoma (fibrosarcomatous transformation [FST]) (DFSP-FST), which is associated with more aggressive behavior and risk of metastasis. The histologic diagnosis of DFSP-FST may be challenging, particularly in small biopsies. Currently, there are no specific markers to reliably support the diagnosis of DFSP-FST. We identified 33 DFSP from 32 patients, including 9 patients with FST and 1 distant metastasis. Tissue microarrays (TMAs) were created using representative areas of all cases, including DFSP-FST, conventional regions from DFSP-FST, and pure conventional DFSP. Digital spatial profiling of the entire transcriptome was performed on the TMAs using the Nanostring GeoMx platform. Expression data were queried to identify differentially expressed genes specific to the regions of transformation in DFSP. Immunohistochemistry for PReferentially expressed Antigen in MElanoma (PRAME) was subsequently performed using a clinically validated protocol. Digital spatial profiling demonstrated significant gene expression differences between DFSP-FST and conventional DFSP. Genes that were overexpressed in DFSP-FST include PRAME, CTAG1B, and MMP11. Immunohistochemical analysis for PRAME showed positive expression in 7 of 10 DFSP-FST (70%) and 1 of 23 conventional DFSP (4%) (P < .0001). Gene expression profiling shows significant upregulation of PRAME and CTAG1B in DFSP-FST compared with conventional DFSP, a finding that was validated by immunohistochemistry for PRAME. Therefore, immunohistochemistry for PRAME may be useful to support the diagnosis of FST, especially in small biopsies. Increased expression of CTAG1B (NY-ESO-1) in DFSP-FST may also offer future therapeutic potential.

BACKGROUND AND OBJECTIVE/UNASSIGNED:Accurate diagnosis of mediastinal tumors is of critical importance to establish appropriate therapy. However, these lesions are relatively uncommon and may be challenging to evaluate, particularly in small biopsy specimens. Thymomas and thymic carcinomas are the most common primary malignant tumors of the mediastinum, but the site can be affected by many other neoplasms that can pose significant difficulty in diagnosis. The objective of this article is to bring awareness to these rarer tumors and offer a diagnostic approach using ancillary techniques guided by clinical and morphological features. METHODS/UNASSIGNED:We discuss and review six challenging cases of mediastinal tumors with overlapping morphologic features. We discuss their unique morphologic, immunophenotypic, and relevant molecular characteristics to support their definitive diagnosis, based on current literature. Sources were obtained via PubMed search and include original studies and review articles published in the English language between 1990 and 2025. Search terms include the diagnostic entities discussed in the article. KEY CONTENT AND FINDINGS/UNASSIGNED:Judicious use of immunohistochemistry and molecular studies is necessary to accurately diagnose mediastinal neoplasms with overlapping histologic features, such as those seen in the cases discussed. CONCLUSIONS/UNASSIGNED:Diagnosis of uncommon mediastinal lesions may be challenging, particularly in small biopsies, as morphological features may be shared among different entities. Awareness of these rare entities, their clinical characteristics and presentation, and differential diagnosis can guide in the selection of appropriate immunohistochemical panels, molecular markers, and molecular diagnostics when appropriate to support the diagnoses.

Gene fusions involving JAZF1 are a recurrent event in low grade endometrial stromal sarcoma, and have been more recently described in few instances of endometrial stromal sarcoma-like tumors in the genitourinary tract of men. In this article, we describe a previously unreported spindle cell sarcoma harboring an in-frame JAZF1::NUDT5 gene fusion, arising in the chest wall of a 51-year-old man. The tumor had unique morphologic features resembling both endometrial stromal sarcoma and endometrial stromal sarcoma-like tumors, consisting of a mixture of cytologically bland and pleomorphic spindle cells with brisk mitotic activity, within an alternating myxoid and fibrous stroma. It had diffuse immunohistochemical expression of CD10, CD34 and CD56, and variable expression of androgen receptor. To our knowledge, neoplasms with these clinico-pathologic characteristics and novel gene fusion have not been previously reported in the English language literature.

PURPOSE/OBJECTIVE:Necrosis quantification in the neoadjuvant setting using pathology slide review is the most important validated prognostic marker in conventional osteosarcoma. Herein, we explored three deep learning strategies on histology samples to predict outcome for OSA in the neoadjuvant setting. EXPERIMENTAL DESIGN/METHODS:Our study relies on a training cohort from New York University (New York, NY) and an external cohort from Charles university (Prague, Czechia). We trained and validated the performance of a supervised approach that integrates neural network predictions of necrosis/tumor content, and compared predicted overall survival (OS) using Kaplan-Meier curves. Furthermore, we explored morphology-based supervised and self-supervised approaches to determine whether intrinsic histomorphological features could serve as a potential marker for OS in the setting of neoadjuvant. RESULTS:Excellent correlation between the trained network and the pathologists was obtained for the quantification of necrosis content (R2=0.899, r=0.949, p < 0.0001). OS prediction cutoffs were consistent between pathologists and the neural network (22% and 30% of necrosis, respectively). Morphology-based supervised approach predicted OS with p-value=0.0028, HR=2.43 [1.10-5.38]. The self-supervised approach corroborated the findings with clusters enriched in necrosis, fibroblastic stroma, and osteoblastic morphology associating with better OS (lg2HR; -2.366; -1.164; -1.175; 95% CI=[-2.996; -0.514]). Viable/partially viable tumor and fat necrosis were associated with worse OS (lg2HR;1.287;0.822;0.828; 95% CI=[0.38-1.974]). CONCLUSIONS:Neural networks can be used to automatically estimate the necrosis to tumor ratio, a quantitative metric predictive of survival. Furthermore, we identified alternate histomorphological biomarkers specific to the necrotic and tumor regions themselves which can be used as predictors.

Liposarcoma is a malignant soft tissue tumor with several subtypes, the most common of which is well-differentiated liposarcoma (WDL) or atypical lipomatous tumor (ALT). WDL/ALTs are further divided into three histological subtypes, including lipoma-like, sclerosing, and inflammatory. While the majority of these tumors are predominantly fatty, the sclerosing variant demonstrates diverse histologic and radiographic characteristics, including variable amounts of fibrosis and fat. Because of this histological variability and relative rarity, the sclerosing WDL/ALT can present diagnostic dilemmas. We present two cases of sclerosing WDL/ALT, both of which demonstrated high degrees of fibrosis and a paucity of fat, mimicking desmoid fibromatosis and other fibrotic soft tissue tumors. Thus, it is important for radiologists to be aware of the subtypes of liposarcoma and their unique characteristics, and to consider sclerosing WDL/ALT in cases of fibrotic soft tissue tumors.

Histologic grading of tumors is associated with prognosis in many organs. In the lung, the most recent grading system proposed by International association for the Study of Lung Cancer (IASLC) and adopted by the World Health Organization (WHO) incorporates the predominant histologic pattern, as well as the presence of high-grade architectural patterns (solid, micropapillary, and complex glandular pattern) in proportions >20% of the tumor surface. This system has shown improved prognostic ability when compared with the prior grading system based on the predominant pattern alone, across different patient populations. Interobserver agreement is moderate to excellent, depending on the study. IASLC/WHO grading system has been shown to correlate with molecular alterations and PD-L1 expression in tumor cells. Recent studies interrogating gene expression has shown correlation with tumor grade and molecular alterations in the tumor microenvironment that can further stratify risk of recurrence. The use of machine learning algorithms to grade nonmucinous adenocarcinoma under this system has shown accuracy comparable to that of expert pulmonary pathologists. Future directions include evaluation of tumor grade in the context of adjuvant and neoadjuvant therapies, as well as the development of better prognostic indicators for mucinous adenocarcinoma.