Search

The pathogenesis of neuroendocrine carcinomas (NECs) and mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) in the gastrointestinal tract remains poorly understood. This study aims to characterize the clinicopathologic and molecular features of NEC/MiNEN in patients with inflammatory bowel disease (IBD). Eighteen surgically resected IBD-associated intestinal carcinomas with a minimum of 30% neuroendocrine component were collected from 6 academic centers and compared with a control group of 12 IBD-associated carcinomas lacking neuroendocrine differentiation. Both groups exhibited a male predominance and similar age distribution. The NEC/MiNEN group was more likely to have a higher percentage of Crohn disease (9/18 vs 1/12; P = .024), occur in the rectum (9/18 vs 3/12) and small intestine (4/18 vs 0/12) (P < .01), be diagnosed on resection without a preceding biopsy (6/18 vs 0/12; P = .057), and have unidentifiable precursor lesions (10/18 vs 1/12; P = .018) than the control group. Synchronous carcinoma, advanced tumor stage (pT3 and pT4), and lymph node metastasis occurred at similar rates; however, the NEC/MiNEN group had a higher incidence of angiovascular invasion (14/18 vs 4/12; P = .024), distant metastasis (8/18 vs 1/12; P = .049), mortality (8/18 vs 2/12; P = .058), and worse survival (Kaplan-Meier; P = .023) than the control group. All tested cases were mismatch repair proficient. A Ki-67 proliferation index ranged from 25% to 100%. Next-generation sequencing in 11 NEC/MiNEN cases revealed low tumor mutational burdens but complex genetic abnormalities commonly involving TP53 (9/11; 82%), FBXW7 (4/11; 36%), and APC (3/11; 27%) genes, with the other genetic alterations randomly occurring in 1 or 2 cases. The neuroendocrine component, which shared similar molecular alterations as the nonneuroendocrine component, was subcategorized into intermediate (G3a) and high grade (G3b); the higher grade correlated with more genetic alterations. In conclusion, IBD-associated NEC/MiNEN shows diverse histologic features, variable precursor lesions, intricate genetic abnormalities, and aggressive biologic behavior. The classification and grading of gastrointestinal NEC/MiNEN may be refined for better clinical management., (Copyright © 2024 United States & Canadian Academy of Pathology. Published by Elsevier Inc. All rights reserved.)

Gastric mixed adenoneuroendocrine carcinoma (MANEC) is a clinically aggressive subtype of mixed neuroendocrine-nonneuroendocrine neoplasm (MiNEN) with unclear clonal origin. In this study, we analyzed high-resolution copy number (CN) profiling data using the OncoScan CNV Assay in the neuroendocrine carcinoma (NEC) and adenocarcinoma components of eight MANECs. Some common CNVs, including the gain of CCNE1 (19q12) and the loss of FAT1 (4q35.2), were frequently detected in both components; these CNVs were verified by FISH, qPCR and immunohistochemistry staining assays in samples with sufficient material. The identification of common CNVs in both components supports the likelihood of single clonal origin of morphologically heterogeneous tumor cells and suggests several novel genetic events potentially involved in the development of gastric MANEC. We also detected and validated some CNVs and alterations specific for the NEC component, such as MAPK1 loss and MAPK signaling pathway alterations, which could contribute to the neuroendocrine differentiation of gastric MANEC. In addition, we found that the NEC component presented more CNVs and greater CN loss than the adenocarcinoma component (P = 0.007 and P = 0.004, respectively); the NEC components from different cases were not clustered in the hierarchical clustering analysis, indicating the marked genetic heterogenicity of the NEC component in gastric MANEC. In summary, this study describes the cytogenetic characteristics of each component of gastric MANEC, providing some clues for further studies on the development and progression of gastric MANEC as well as providing some potential therapeutic targets.

Gastroblastoma is a rare distinctive biphasic tumor of the stomach. The molecular biology of gastroblastoma has not been studied, and no affirmative diagnostic markers have been developed. We retrieved two gastroblastomas from the consultation practices of the authors and performed transcriptome sequencing on formalin-fixed paraffin-embedded tissue. Recurrent predicted fusion genes were validated at genomic and RNA levels. The presence of the fusion gene was confirmed on two additional paraffin-embedded cases of gastroblastoma. Control cases of histologic mimics (biphasic synovial sarcoma, leiomyoma, leiomyosarcoma, desmoid-type fibromatosis, EWSR1-FLI1-positive Ewing sarcoma, Wilms' tumor, gastrointestinal stromal tumor, plexiform fibromyxoma, Sonic hedgehog-type medulloblastomas, and normal gastric mucosa and muscularis propria were also analyzed. The gastroblastomas affected two males and two females aged 9-56 years. Transcriptome sequencing identified recurrent somatic MALAT1-GLI1 fusion genes, which were predicted to retain the key domains of GLI1. The MALAT1-GLI1 fusion gene was validated by break-apart and dual-fusion FISH and RT-PCR. The additional two gastroblastomas were also positive for the MALAT1-GLI1 fusion gene. None of the other control cases harbored MALAT1-GLI1. Overexpression of GLI1 in the cases of gastroblastomas was confirmed at RNA and protein levels. Pathway analysis revealed activation of the Sonic hedgehog pathway in gastroblastoma and gene expression profiling showed that gastroblastomas grouped together and were most similar to Sonic hedgehog-type medulloblastomas. In summary, we have identified an oncogenic MALAT1-GLI1 fusion gene in all cases of gastroblastoma that may serve as a diagnostic biomarker. The fusion gene is predicted to encode a protein that includes the zinc finger domains of GLI1 and results in overexpression of GLI1 protein and activation of the Sonic hedgehog pathway.

Cutaneous neuroendocrine (Merkel cell) carcinoma most often arises de novo in the background of a clonally integrated virus, the Merkel cell polyomavirus, and is notable for positive expression of retinoblastoma 1 (RB1) protein and low expression of p53 compared with the rare Merkel cell polyomavirus-negative Merkel cell carcinomas. Combined squamous and Merkel cell tumors are consistently negative for Merkel cell polyomavirus. Little is known about their immunophenotypic or molecular profile. Herein, we studied 10 combined cutaneous squamous cell and neuroendocrine carcinomas for immunohistochemical expression of p53, retinoblastoma 1 protein, neurofilament, p63, and cytokeratin 20 (CK20). We compared mutation profiles of five combined Merkel cell carcinomas and seven 'pure' Merkel cell carcinomas using targeted next-generation sequencing. Combined tumors were from the head, trunk, and leg of Caucasian males and one female aged 52-89. All cases were highly p53- and p63-positive and neurofilament-negative in the squamous component, whereas RB1-negative in both components. Eight out of 10 were p53-positive, 3/10 p63-positive, and 3/10 focally neurofilament-positive in the neuroendocrine component. Six out of 10 were CK20-positive in any part. By next-generation sequencing, combined tumors were highly mutated, with an average of 48 mutations per megabase compared with pure tumors, which showed 1.25 mutations per megabase. RB1 and p53 mutations were identified in all five combined tumors. Combined tumors represent an immunophenotypically and genetically distinct variant of primary cutaneous neuroendocrine carcinomas, notable for a highly mutated genetic profile, significant p53 expression and/or mutation, absent RB1 expression in the context of increased RB1 mutation, and minimal neurofilament expression.

Well-differentiated hepatocellular carcinoma in non-cirrhotic liver can show morphological features similar to hepatocellular adenoma. In rare instances, hepatocellular carcinoma can arise in the setting of hepatocellular adenoma. This study compares the immunohistochemical and cytogenetic features of the hepatocellular adenoma-like and hepatocellular carcinoma portions of these tumors. Immunohistochemistry for β-catenin, glutamine synthetase, serum amyloid A protein, glypican-3, and heat-shock protein 70 was done in 11 cases of hepatocellular carcinoma arising in hepatocellular adenoma in non-cirrhotic liver. Tumors with nuclear β-catenin and/or diffuse glutamine synthetase were considered β-catenin activated. Fluorescence in situ hybridization (FISH) was done in nine cases for gains of chromosomes 1, 8 and MYC. There were seven men (33-75 years) and four women (29-65 years). Focal atypical morphological features were seen in hepatocellular adenoma-like areas in 7 (64%) cases. Hepatocellular adenoma-like areas showed features of inflammatory hepatocellular adenoma in 7 (64%) cases; 4 of these were also serum amyloid A-positive in the hepatocellular carcinoma portion. β-Catenin activation, heat-shock protein 70 positivity, and chromosomal gains on FISH were seen in the hepatocellular adenoma portion in 55%, 40%, and 56% of cases, and 73%, 60%, and 78% of cases in the hepatocellular carcinoma portion, respectively. In conclusion, the hepatocellular adenoma-like portion of most cases of hepatocellular carcinoma arising in hepatocellular adenoma shows features typically seen in hepatocellular carcinoma such as focal morphological abnormalities, β-catenin activation, heat-shock protein 70 expression, and chromosomal gains. Hepatocellular adenoma-like areas in these tumors, especially in men and older women, may represent an extremely well-differentiated variant of hepatocellular carcinoma, whereas the morphologically recognizable hepatocellular carcinoma portion represents a relatively higher grade component of the tumor.

Desmoplastic melanoma is subclassified into pure and mixed variants with a higher rate of lymph node metastasis in the latter. Given that reasons for these biological differences are not currently known, we investigated these subtypes with techniques that included genetic and immunohistochemical analyses of 43 cases of desmoplastic melanoma (24 pure, 19 mixed). Direct DNA sequencing was performed on BRAFV600E, RET gene (coding region on exon 11) and KIT (exons 11, 13 and 17). Immunohistochemical stains were performed with antibodies to markers of significance with respect to biological potential of nevomelanocytic proliferations and/or desmoplastic melanoma (Ki-67, CD117, nestin, clusterin, SOX10 and CD271/p75NTR). Polymorphism at the RET coding region (RETp) was noted in 33% of pure (8/24 cases) versus 24% of mixed (4/17 cases); BRAFV600E was absent in all cases of pure (0/24 cases) versus 6% of mixed (1/17 cases); no mutations were found in any of the cases on analyses of exons 11, 13 and 17 of the c-KIT gene (P=NS for all). For immunohistochemical analyses of pure versus mixed: mean percentage of Ki-67 nuclear positivity was 5% (s.d.=5.6) versus 28% (s.d.=12.6, P<0.001); CD117 stained 26% (6/23 cases) versus 78% (14/18 cases, P<0.01); nestin stained 83% (n=19/23 cases) versus 89% (16/18 cases, P=NS); clusterin stained 4% (1/23 cases) versus 6% (1/18 cases, P=NS); SOX10 87% (20/23 cases) versus 94% (17/18 cases, P=NS) and CD271 stained 61% (14/23 cases) versus 67% (12/18 cases, P=NS). Increased CD117 staining in the mixed variant suggests that alterations in the KIT protein may be involved in tumor progression. In addition, the proliferative index of the mixed variant was higher than that of the pure variant.