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; S. Emily Bachert, MD, associate pathologist, Brigham and Women’s Hospital, Boston; Amarpreet Bhalla, MD, assistant professor of pathology, Albert Einstein College of Medicine, Montefiore Medical Center; Divya Sharma, MD, associate professor, Department of Pathology and Laboratory Medicine, University of Cincinnati Medical Center; and Paula Toro, MD, pathology resident, Cleveland Clinic.
Pathology of pulmonary disease after pediatric hematopoietic stem cell transplantation
February 2025—Pulmonary complications cause significant morbidity and mortality in post-hematopoietic stem cell transplantation. The histopathology of pulmonary diseases in the context of post-hematopoietic stem cell transplantation (post-HSCT) is poorly characterized, especially in the pediatric population. The authors sought to characterize the pathologic spectrum of pulmonary disease post-HSCT in a pediatric cohort. Fifty-six specimens, including 53 biopsy specimens, corresponding to 53 patients were identified. Biopsy slides were reviewed and assigned to diagnostic categories (infectious, graft-versus-host disease, vasculopathy, indeterminate, and others) by consensus among three pediatric pulmonary pathologists, taking into consideration pathologic, clinical, radiologic, and laboratory findings. The most common diagnostic category was infection (n=20). Vasculopathy, mostly in the form of fibromyxoid intimal expansion, was a common finding in the cohort (n=26) and was the sole finding in a small subset of patients (n=5), with particularly poor outcomes. A subset of biopsies remained indeterminate (n=10), and the findings in this cohort were dominated by acute lung injury. The latter group had poor prognoses, with a short biopsy-to-death interval. The overall clinicopathologic concordance was 40 percent, and agreement was most common in the infectious category. Wedge biopsies led to a change in management in 69 percent of cases versus 23 percent for limited procedures, such as core needle biopsies. The authors’ results suggest that while complications from infections continue to be common post-HSCT, other findings, such as vasculopathy and acute lung injury, portend a particularly poor prognosis and should be actively sought and reported.
Cortes-Santiago N, Deutsch G, Patel KR, et al. The pathology of pulmonary disease after pediatric hematopoietic stem cell transplantation. Am J Surg Pathol. 2024;48(10):1201–1214.
Correspondence: Dr. Nahir Cortes-Santiago at nxcortes@texaschildrens.org
Nuclear DUX4 immunohistochemistry as a marker for CIC::DUX4 fusion in CIC-rearranged sarcoma
CIC-rearranged sarcomas are clinically aggressive undifferentiated round cell sarcomas commonly driven by CIC::DUX4. Due to the repetitive nature of DUX4 and the variability of the fusion breakpoints, molecular testing may miss CIC::DUX4 fusion. IHC stains have been studied as surrogates for CIC::DUX4 fusion. The authors conducted a study to assess the performance of DUX4 IHC in the workup of CIC-rearranged sarcomas (CRS) and its expression in non-CRS round cell or epithelioid neoplasms. Cases of molecularly confirmed CRS (n=48) and non-CRS (n=105) were included. CRS cases consisted of 35 females and 13 males (age range, less than one year to 67 years; median, 41 years). Among the molecularly confirmed non-CRS cases, C-terminal DUX4 expression was investigated in Ewing sarcoma (38 cases), alveolar rhabdomyosarcoma (18 cases), desmoplastic small round cell tumors (12 cases), and synovial sarcoma (five cases), as well as in nonmesenchymal neoplasms such as SMARCA4/SMARCB1-deficient tumors (five cases), carcinomas of unknown primary (three cases), and haematolymphoid neoplasms (four cases). DUX4 IHC was considered positive when strong nuclear expression was detected in more than 50 percent of neoplastic cells. When used as a surrogate for the diagnosis of CRS, the sensitivity and specificity of DUX4 IHC were 98 and 100 percent, respectively. Only one CRS case was negative for DUX4 IHC and harbored a CIC::FOXO4 fusion. The authors concluded that DUX4 IHC is a highly sensitive and specific surrogate marker for the presence of CIC::DUX4 fusion, demonstrating its utility in diagnosing CRS.
Macedo RT, Baranovska-Andrigo V, Pancsa T, et al. Nuclear DUX4 immunohistochemistry is a highly sensitive and specific marker for the presence of CIC::DUX4 fusion in CIC-rearranged sarcomas: a study of 48 molecularly confirmed cases. Histopathology. 2024. doi.org/10.1111/his.15341
Correspondence: Dr. Josephine K. Dermawan at dermawj@ccf.org
Use of an AI-enhanced, Web-based application to review bile duct brushing cytologic specimens
The authors previously developed an artificial intelligence tool to help cytologists evaluate digital whole slide images generated from bile duct brushing specimens. The aim of a pilot trial featured herein was to assess the effectiveness and accuracy of cytologist review using the AI-enhanced application on a new prospective, biliary whole slide image (WSI) data set. Consecutive bile duct brushing WSI from indeterminate strictures were obtained. A multidisciplinary panel reviewed all relevant information and provided a central interpretation for each WSI of positive, negative, or indeterminate. The WSI were then uploaded to the AI application. The AI scored each one as positive or negative for malignancy, referred to as computer-aided diagnosis (CADx). For each WSI, the AI prioritized cytologic tiles by the likelihood that malignant material was present in the tile. Using the AI, blinded cytologists reviewed all WSI and provided interpretations, referred to as computer-aided detection (CADe). The diagnostic accuracies of the WSI evaluation via CADx, CADe, and the original clinical cytologic interpretation, referred to as the official cytologic interpretation (OCI), were compared. The panel determined that 15 of the 84 WSI were positive, 42 were negative, and 27 were indeterminate. The WSI generated, on average, 141,950 tiles each. Cytologists using the AI evaluated 10.5 tiles per WSI before making an interpretation. Additionally, cytologists required an average of 84.1 s of total WSI evaluation. WSI interpretation accuracies were similar for CADx (0.754; 95 percent confidence interval [CI], 0.622–0.859), CADe (0.807; 95 percent CI, 0.750–0.856), and OCI (0.807; 95 percent CI, 0.671–0.900). The authors concluded that this trial suggests that an AI application that prioritizes WSI based on malignancy risk and selects the most worrisome cytologic tiles for review maintains cytologist accuracy and improves sensitivity.
Marya NB, Powers PD, Bois MC, et al. Utilization of an artificial intelligence–enhanced, web-based application to review bile duct brushing cytologic specimens: A pilot study. Cancer Cytopathol. 2024. doi:10.1002/cncy.22898
Correspondence: Dr. Neil B. Marya at neil.marya@umassmed.edu
Establishing criteria to assess tumor necrosis as a prognostic indicator in colorectal cancer
Tumor necrosis has been considered an independent prognostic factor in colorectal cancer, but its evaluation methods have not been described in sufficient detail for it to be used clinically. To study the potential of tumor necrosis as a prognostic indicator in colorectal cancer, the authors compared three methods for evaluating it: average percentage method (tumor necrosis percentage of the whole tumor), hotspot method (tumor necrosis percentage in a single hotspot), and linear method (diameter of the single largest necrotic focus). Cox regression models were used to calculate cancer-specific mortality hazard ratios (HRs) for tumor necrosis categories in two colorectal cancer cohorts representing more than 1,800 cases. To assess reproducibility, nine investigators evaluated 30 cases, and Spearman’s rank correlation coefficients and Cohen’s kappa coefficients were calculated. The authors found that all three methods predicted colorectal cancer-specific survival independently of other prognostic parameters, including disease stage, lymphovascular invasion, and tumor budding. The greatest multivariable HRs were observed for the average percentage method (cohort one: HR for ≥40 percent versus <3 percent, 3.03; 95 percent confidence interval [CI], 1.93–4.78; cohort two: HR for ≥40 percent versus <3 percent, 2.97; 95 percent CI, 1.63–5.40). All three methods had high reproducibility, with the linear method showing the highest mean Spearman’s correlation coefficient (0.91) and Cohen’s kappa (0.70). The authors concluded that all three methods showed good reproducibility and predictive ability. The findings pave the way for the use of tumor necrosis as a prognostic indicator in colorectal cancer.
Kastinen M, Sirniö P, Elomaa H, et al. Establishing criteria for tumor necrosis as prognostic indicator in colorectal cancer. Am J Surg Pathol. 2024;48(10):1284–1292.
Correspondence: Dr. Juha P. Väyrynen at juha.vayrynen@oulu.fi