The American Journal of Surgical Pathology最新文献

Before the prophylactic use of anti-D antibodies in pregnancy, hemolytic anemia of the newborn was the most common cause of hyperbilirubinemia. Nowadays, given the rarity of hemolytic anemia of the newborn, hepatobiliary abnormalities, perinatal infections, and metabolic disorders have become the most common conditions in the differential diagnosis of neonatal cholestasis. Here, we report 3 instances of cholestatic giant cell hepatitis in 3 infants who had Coombs' positive hemolysis due to ABO incompatibility in 1, Rh incompatibility in another, and combined ABO and Rh incompatibility in the third. Although rare, cholestatic neonatal giant cell hepatitis associated with hemolysis still needs to be considered in patients with neonatal cholestasis. A marked elevation of aspartate aminotransferase over alanine aminotransferase can be a helpful clue to an early diagnosis.

The separation of benign from malignant mesothelial proliferations is often a difficult pathologic problem. UHRF1 (ubiquitin-like with plant homeodomain and ring finger domains-1) is a regulator of DNA methylation and an epigenetic driver of various human cancers. It has recently been reported that UHRF1 is overexpressed in mesotheliomas. We asked whether UHRF1 immunohistochemistry could be used to separate benign from malignant mesothelial proliferations. Initial studies showed that UHRF1 stained mesothelial cells but also endothelial and other non-neoplastic cells, so that accurate counting of positive mesothelial cells was difficult. Therefore, we ran dual UHRF1-AE1/AE3 stains on 2 tissue microarrays containing 40 reactive mesothelial proliferations and 61 mesotheliomas and only counted UHRF1 staining in keratin-positive cells. On average 10.3±8.6% (mean±SD; range: 0% to 36, median: 6.8%) of epithelioid mesothelioma cells stained compared with 5.3±4.8% (range: 0% to 15%, median: 4.1%) of reactive epithelial mesothelial cells. This difference was statistically significant but there was too much overlap to use diagnostically. In contrast, 37±26% (range: 2.5% to 95%, median: 31%) of cells in sarcomatoid mesotheliomas compared with 1.2±1.2% (range: 0% to 3.0%, median: 1.0%) of cells in reactive spindle cell mesothelial proliferations stained. To confirm this difference we stained whole sections of 21 sarcomatoid mesotheliomas and 19 cases of organizing pleuritis. Staining of mesothelial cells was seen in 2.1±2.4% (range: 0% to 6.8%, median: 1.0%) of organizing pleuritis cases and 44±22% (range: 14% to 90%, median: 41%) of sarcomatoid mesotheliomas. We conclude that dual UHRF1-AE1/AE3 immunohistochemistry is very useful for separating benign spindle cell mesothelial proliferations from sarcomatoid mesotheliomas.

Although there is early support for schemes based on nuclear grade, necrosis and mitotic rate, there is currently no widely implemented grading system for diffuse pleural mesothelioma (DPM). We investigated current systems and propose a novel Mesothelioma Weighted Grading Scheme (MWGS). The MWGS assigns weighted scores from 0 to 10 based on age (≤74, >74 yrs: 0,1); histologic type (epithelioid, biphasic, sarcomatoid: 0,1,2); necrosis (absent, present: 0,2); mitotic count per 2 mm2 (≤1, 2 to 4, ≥5: 0,1,2); nuclear atypia (mild, moderate, severe: 0,1,2); and BRCA1-associated protein 1 (BAP1) expression (lost, retained: 0,1). A score of 0 to 3 is low grade, 4 to 6 intermediate grade, and 7 to 10 high grade. In 369 consecutive DPMs, median survival was 17.1, 10.1, and 4.1 months for low, intermediate, and high grades (P<0.0001). A progressive increase in score correlated with worsening overall survival (P<0.0001). Interobserver concordance was substantial (κ=0.588), with assessment of nuclear grade being the most subjective parameter (κ=0.195). We compared the MWGS to the 2-tiered system discussed in the World Health Organization (WHO) fifth edition. The WHO system predicted median survival in epithelioid (median 18.0 vs. 11.3 mo, P=0.003) and biphasic (16.2 vs. 4.2 mo, P=0.002), but not sarcomatoid DPM (5.4 vs. 4.7 mo, P=0.407). Interestingly, the WHO grading system was prognostic in cases with BAP1 loss (median survival 18.7 vs. 10.4 mo, P<0.0001), but not retained BAP1 expression (8.9 vs. 6.2 mo, P=0.061). In conclusion, the WHO scheme has merit in epithelioid/biphasic and BAP1-deficient DPM, however, the MWGS can be used for risk stratification of all DPMs, regardless of histologic subtype and BAP1 status.

Despite the impressive efficacy of chimeric antigen receptor (CAR) T-cell therapy (CART) in B-cell non-Hodgkin lymphomas, durable responses are uncommon. The histopathologic and molecular features associated with treatment failure are still largely unknown. Therefore, we have analyzed 19 sequential tumor samples from 9 patients, prior anti-CD19 CART (pre-CART) and at relapse (post-CART), using immunohistochemistry, fluorescence in situ hybridization, array comparative genomic hybridization, next-generation DNA and RNA sequencing, and genome-scale DNA methylation. The initial diagnosis was diffuse large B-cell lymphoma (n=6), double-hit high-grade B-cell lymphoma (n=1), and Burkitt lymphoma (n=2). Histopathologic features were mostly retained at relapse in 7/9 patients, except the frequent loss of 1 or several B-cell markers. The remaining 2 cases (1 diffuse large B-cell lymphoma and 1 Burkitt lymphoma) displayed a dramatic phenotypic shift in post-CART tumors, with the drastic downfall of B-cell markers and emergence of T-cell or histiocytic markers, despite the persistence of identical clonal immunoglobulin gene rearrangements. The post-CART tumor with aberrant T-cell phenotype showed reduced mRNA expression of most B-cell genes with increased methylation of their promoter. Fluorescence in situ hybridization and comparative genomic hybridization showed global stability of chromosomal alterations in all paired samples, including 17p/TP53 deletions. New pathogenic variants acquired in post-CART samples included mutations triggering the PI3K pathway (PIK3R1, PIK3R2, PIK3C2G) or associated with tumor aggressiveness (KRAS, INPP4B, SF3B1, SYNE1, TBL1XR1). These results indicate that CART-resistant B-cell non-Hodgkin lymphomas display genetic remodeling, which may result in profound dysregulation of B-cell differentiation. Acquired mutations in the PI3K and KRAS pathways suggest that some targeted therapies could be useful to overcome CART resistance.

Metastatic tumors interface with liver in multiple patterns, of which, the rare "sinusoidal" growth pattern can be subtle and easily overlooked on biopsy. We sought to characterize the metastasis-to-liver interface patterns of melanoma compared with other tumor types and assess the incidence of metastatic melanoma in histologically normal-appearing targeted liver lesion biopsies. Liver lesion samples from 54 melanoma patients were assessed. Nearly normal-appearing cases, defined as no obvious malignancy on routine hematoxylin and eosin stain (n=24), were stained with SOX10 and confirmed with MelanA. Tumor-to-liver interface patterns were determined in biopsies overtly positive for metastatic melanoma (n=30) versus other hepatic metastases as controls (colon, n=28; breast, n=20; pancreaticobiliary, n=20; and neuroendocrine, n=28). Of the 24 nearly normal-appearing liver biopsies from melanoma patients, 3 had subtle melanoma cells detected in sinusoids, confirmed with immunohistochemistry. Of 30 livers overtly positive for melanoma, 8 showed the sinusoidal pattern, compared with none in other metastases. In total, 11/33 (33%) cases of metastatic melanoma liver biopsies demonstrated the sinusoidal pattern. We describe 11 metastatic melanoma cases in liver with the rare sinusoidal pattern, 3 of which were subtle and easy to miss on routine hematoxylin and eosin stain. Given that sinusoidal metastasis does not elicit a tissue reaction, it is prudent for the pathologists to be aware of this pattern of metastases and have a low threshold to order immunostains for accurate diagnosis and optimal patient care.