CYP1A1 immunohistochemistry is highly specific for angiofibroma of soft tissue among morphological mimics.

AIMS: Angiofibroma of soft tissue (AFST) is a benign, morphologically distinctive tumour type that harbours recurrent AHRR::NCOA2 fusions in 60-70% of cases and shows a non-specific immunophenotype, expressing EMA in roughly half of cases. The AHRR::NCOA2 fusion results in increased expression of cytochrome P450 1A1 (CYP1A1); a recent study demonstrated CYP1A1 immunohistochemistry (IHC) to be moderately sensitive and highly specific for AFST. METHODS AND RESULTS: In this study, we sought to validate these findings in a larger independent cohort of 30 AFST, as well as 215 morphological mimics, including 30 solitary fibrous tumours, 29 myxoid liposarcomas, 28 low-to-intermediate grade myxofibrosarcomas (MFS), 20 atypical spindle cell lipomatous tumours (ASCLT), 20 cellular angiofibromas, 10 cases each of spindle cell lipoma, neurofibroma, malignant peripheral nerve sheath tumour, superficial angiomyxoma, cellular myxoma, soft tissue perineurioma and deep fibrous histiocytoma, and nine cases each of low-grade fibromyxoid sarcoma and mammary-type myofibroblastoma. We found CYP1A1 IHC to be 70% sensitive for AFST, with granular cytoplasmic staining in 21 of 30 tumours, and 98% specific, with staining in only five morphological mimics: two deep fibrous histiocytomas, one MFS, one cellular angiofibroma and one ASCLT. CONCLUSIONS: These findings confirm that CYP1A1 is 70% sensitive, consistent with the prevalence of AHRR::NCOA2 fusions that up-regulate this protein, and that it is highly specific among morphological mimics.

Postinfantile Giant Cell Hepatitis in Native and Allograft Livers: A Multi-Institutional Clinicopathologic Study of 70 Cases.

Postinfantile giant cell hepatitis (PIGCH) is a rare hepatitis pattern in adults with variable etiologies and clinical outcomes. We conducted a multi-institutional retrospective study to define the clinicopathologic characteristics of patients with PIGCH. A total of 70 PIGCH cases were identified and reviewed for pathological features, including fibrosis, cholestasis, inflammation, steatosis, necrosis, and apoptosis, as well as the distribution of giant cells and the maximum number of giant cells per high-power field. Demographic and clinical data, including age, sex, laboratory results, etiologies, and follow-up results, were recorded. Among the 70 cases, 40% (28/70) were associated with autoimmune liver diseases, followed by 9 (13%) with unknown etiology, 8 (11%) with viral infection, 5 (7%) with medications, 5 with combined etiologies, and 4 (6%) with malignancies (mostly chronic lymphocytic leukemia). Notably, another 16% were de novo PIGCH in liver allografts, most of which occurred after a rejection event. During follow-up, 26 (37%) patients died of the disease and 44 (63%) were alive. Deceased patients were characterized by older age (mean age, 54.9 vs 45.5 years; P = .02), higher alkaline phosphatase level (mean value, 253.3U/L vs 166.3 U/L; P = .03), higher fibrosis stage (stage 3-4 vs stage 0-2, 57.7% vs 29.6%; P = .03), being more likely to have de novo PIGCH after transplantation (23.1% vs 11.4%; P = .04), and being less likely to have primary autoimmune liver disease etiology (26.9% vs 47.7%; P = .04). These results indicate that PIGCH is a rare pattern of liver injury associated with different etiologies and variable clinical outcomes. Autoimmune liver disease with PIGCH is associated with better survival, whereas de novo PIGCH in allografts is associated with poorer survival. Older age, higher alkaline phosphatase level, and advanced fibrosis are adverse prognostic factors.

Gene fusions in superficial mesenchymal neoplasms: Emerging entities and useful diagnostic adjuncts.

Cutaneous mesenchymal neoplasms are diagnostically challenging because of their overlapping morphology, and, often, the limited tissue in skin biopsy specimens. Molecular and cytogenetic techniques have identified characteristic gene fusions in many of these tumor types, findings that have expanded our understanding of disease pathogenesis and motivated development of useful ancillary diagnostic tools. Here, we provide an update of new findings in tumor types that can occur in the skin and superficial subcutis, including dermatofibrosarcoma protuberans, benign fibrous histiocytoma, epithelioid fibrous histiocytoma, angiomatoid fibrous histiocytoma, glomus tumor, myopericytoma/myofibroma, non-neural granular cell tumor, CIC-rearranged sarcoma, hybrid schwannoma/perineurioma, and clear cell sarcoma. We also discuss recently described and emerging tumor types that can occur in superficial locations and that harbor gene fusions, including nested glomoid neoplasm with GLI1 alterations, clear cell tumor with melanocytic differentiation and ACTIN::MITF translocation, melanocytic tumor with CRTC1::TRIM11 fusion, EWSR1::SMAD3-rearranged fibroblastic tumor, PLAG1-rearranged fibroblastic tumor, and superficial ALK-rearranged myxoid spindle cell neoplasm. When possible, we discuss how fusion events mediate the pathogenesis of these tumor types, and we also discuss the related diagnostic and therapeutic implications of these events.

Computational augmentation of neoplastic endometrial glands in digital pathology displays.

The pathologic diagnosis of neoplasia requires localization and classification of lesional tissue, a process that depends on the recognition of an abnormal spatial distribution of neoplastic elements relative to admixed normal background tissue. In endometrial intraepithelial neoplasia (EIN), a pre-cancer usually managed by hysterectomy, a clonally mutated proliferation of cytologically altered glands ('neoplastic-EIN') aggregates in clusters that also contain background non-neoplastic glands ('background-NL'). Here, we used image analysis to classify individual glands within endometrial tissue fragments as neoplastic-EIN or background-NL, and we used the distribution of predictions to localize foci diagnostic of EIN. Nuclear coordinates were automatically assigned and were used as vertices to generate Delaunay triangulations for each gland. Graph statistical variables were used to develop random forest algorithms to classify glands as neoplastic-EIN or background-NL. Individual glands in an independent validation set were scored by a 'ground truth' biomarker (PAX2 immunohistochemistry). We found that exclusion of small glands led to improvement in classification accuracy. Using an inclusion threshold of 200 nuclei per gland, our final model classification accuracy was 77.5% in the validation set, with a positive predictive value of 0.81. We leveraged this high positive predictive value in a point cloud overlay display to assist end-user identification of EIN foci. This study demonstrates that graph theory approaches applied to small-scale anatomic elements in the endometrium allow biologic classification by machine learning, and that spatial superimposition over large-scale tissue expanses can have practical diagnostic utility. We expect this augmented diagnostic approach to be generalizable to commonly encountered problems in other organ systems. © 2020 The Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Aneurysmal bone cyst with an unusual clinical presentation and a novel VDR-USP6 fusion.

Aneurysmal bone cyst is a benign bone neoplasm that most commonly arises from the metaphyses of long bones in the first and second decades of life. Here, we describe a case of an aneurysmal bone cyst that occurred in the distal tibial diaphysis of a 72-year-old female that was concerning for malignancy on imaging, demonstrating cortical breakthrough and soft tissue extension. Histologically, the tumor showed the characteristic morphologic features of aneurysmal bone cyst. Fluorescence in situ hybridization was positive for USP6 rearrangement, and RNA sequencing revealed a USP6 gene fusion with VDR, a novel partner that encodes the vitamin D receptor and that has not been implicated previously in human neoplasia. This case highlights the diagnostic challenges presented by aneurysmal bone cyst in elderly adults, and it expands the genetic spectrum of USP6 rearrangements.

Calcifying nested stromal-epithelial tumor: a clinicopathologic and molecular genetic study of eight cases highlighting metastatic potential and recurrent CTNNB1 and TERT promoter alterations.

Calcifying nested stromal-epithelial tumor (CNSET) is a rare hepatic tumor that occurs in children and young adults. With <40 cases in the literature, the mechanism for tumorigenesis and the biological behavior of CNSET remain uncertain. Here, we studied the clinicopathologic and molecular genetic features of eight CNSETs. Six patients (75%) were female, and the median age at presentation was 22.5 years (range 14-34 years). The median tumor size was 14 cm (range 2.7-18 cm). All tumors had fibrous stroma that contained organoid nests of epithelioid to spindled tumor cells with moderate amounts of palely eosinophilic cytoplasm and ovoid, vesicular nuclei. Five tumors showed calcifications, and one showed lymphovascular invasion. Necrosis was absent in all. Immunohistochemistry demonstrated nuclear ß-catenin expression in five of five tested tumors and focal to diffuse nuclear WT-1 positivity in five of seven. Hepatocellular markers (HepPar-1, arginase-1, and albumin in situ hybridization) and neuroendocrine markers (synaptophysin, chromogranin, and INSM1) were uniformly negative. Next-generation sequencing demonstrated CTNNB1 alterations in all seven sequenced tumors. Sanger sequencing demonstrated TERT promoter mutations in all six sequenced tumors. Clinical follow-up was available for seven patients (median duration 4.4 years; range 1.2-6.2 years): four (57%) developed metastatic disease; all four developed lung metastases; and two also had abdominal metastases. All four patients with metastatic disease also had persistent or recurrent liver tumors. Three patients with metastases were alive with disease at the most recent follow-up and one died of disease. The other three patients with available follow-up did not develop metastasis or recurrence. One tumor treated with neoadjuvant chemotherapy showed no response, and another showed 90% tumor fibrosis; the latter patient remained disease-free at 6.2 years of follow-up. Our series demonstrates the presence of TERT promoter mutations and CTNNB1 alterations in all sequenced tumors and suggests that CNSET might perhaps be more aggressive than previously reported.

Recent developments in gastroesophageal mesenchymal tumours.

The pathologist's approach to gastroesophageal mesenchymal tumours has changed dramatically during the last 25 years. In particular, gastrointestinal stromal tumour (GIST) has evolved from a wastebasket mesenchymal tumour category to a precisely defined entity with an increasingly detailed genetic subclassification. This subclassification has brought gastrointestinal mesenchymal neoplasia into the realm of precision medicine, with specific treatments optimised for particular genetic subtypes. Molecular genetic data have also greatly improved our understanding of oesophageal mesenchymal tumours, including the discovery that so-called 'giant fibrovascular polyps' in fact represent a clinically distinctive presentation of well-differentiated liposarcoma. Here, we will focus on gastroesophageal mesenchymal tumours for which there have been recent developments in classification, molecular genetics or tumour biology: granular cell tumour, 'giant fibrovascular polyp'/well-differentiated liposarcoma, plexiform fibromyxoma, gastroblastoma and, of course, GIST.

Undifferentiated Embryonal Sarcoma of the Liver With Rhabdoid Morphology Mimicking Carcinoma: Expanding the Morphologic Spectrum or a Distinct Variant?

Undifferentiated embryonal sarcoma of the liver (UESL) is a rare aggressive neoplasm that occurs predominantly in children. Like mesenchymal hamartoma of the liver (MHL), UESL harbors recurrent rearrangements involving 19q13.3 and 19q13.4, a region of the genome that contains a primate-specific cluster of micro-RNAs. Here, we present a case of a high-grade neoplasm that arose in the left hepatic lobe of a 5-year-old male and gave rise to widespread lymph node, visceral, and soft tissue metastases. The tumor was composed of sheets, tubules, and papillae of epithelioid cells with rhabdoid morphology. INI1 and BRG1 expression were retained. Tumor cells diffusely expressed epithelial markers, including multiple keratins. While the morphologic and immunophenotypic features were suggestive of poorly differentiated carcinoma with rhabdoid features, the tumor was found to harbor the t(11;19)(q13;q13.3) translocation characteristic of UESL, as well as a TP53 mutation. Given the clinical presentation, imaging, clinical course, the tumor was classified as UESL with unusual, carcinoma-like histopathologic features. In the context of an unclassified high-grade hepatic tumor in a young child, molecular or cytogenetic testing for chromosome 19q13 alterations should be considered.

Clinicopathologic characterization of malignant chondroblastoma: a neoplasm with locally aggressive behavior and metastatic potential that closely mimics chondroblastoma-like osteosarcoma.

Chondroblastoma is currently classified as a benign neoplasm; however, chondroblastoma and chondroblastoma-like osteosarcoma have morphologic overlap, raising the possibility that some tumors diagnosed as chondroblastoma-like osteosarcoma might actually represent malignant chondroblastoma. The H3F3B K36M point mutation, which has not been reported in osteosarcoma, is identified in 95% of chondroblastomas and is reliably detectable by immunohistochemistry (IHC). We reviewed 11 tumors diagnosed as atypical chondroblastoma, malignant chondroblastoma, or chondroblastoma-like osteosarcoma (median follow-up: 8.8 years; range: 4 months-26.4 years). Seven chondroblastomas with cytologic atypia and permeative growth were designated "malignant chondroblastoma"; six were H3K36M-positive by IHC. Relative to conventional chondroblastoma, malignant chondroblastoma occurred in older individuals (median: 52 years; range: 29-57 years) and arose at unusual sites. Three of four tumors with long-term follow-up recurred, and one patient died of widespread metastases. One was found to have chromosomal copy number alter4ations and a SETD2 mutation in addition to H3F3B K36M. The four remaining tumors were classified as chondroblastoma-like osteosarcoma. Chondroblastoma-like osteosarcoma occurred in younger patients (median: 21 years; range: 19-40 years) than malignant chondroblastoma. In contrast to malignant chondroblastoma, all had regions of malignant cells forming bone. Two of three patients with long-term follow-up developed recurrences, and two died of disease, one with widespread metastases. No mutations in H3F3A/H3F3B were detected by Sanger sequencing. While malignant chondroblastoma and chondroblastoma-like osteosarcoma show significant morphologic overlap, they have distinct clinical presentations and genetic findings. When considering this challenging differential diagnosis, IHC using histone H3 mutation-specific antibodies is a critical diagnostic adjunct.

A New Classification of Benign, Premalignant, and Malignant Endometrial Tissues Using Machine Learning Applied to 1413 Candidate Variables.

Benign normal (NL), premalignant (endometrial intraepithelial neoplasia, EIN) and malignant (cancer, EMCA) endometria must be precisely distinguished for optimal management. EIN was objectively defined previously as a regression model incorporating manually traced histologic variables to predict clonal growth and cancer outcomes. Results from this early computational study were used to revise subjective endometrial precancer diagnostic criteria currently in use. We here use automated feature segmentation and updated machine learning algorithms to develop a new classification algorithm. Endometrial tissue from 148 patients was randomly separated into 72-patient training and 76-patient validation cohorts encompassing all 3 diagnostic classes. We applied image analysis software to keratin stained endometrial tissues to automatically segment whole-slide digital images into epithelium, cells, and nuclei and extract corresponding variables. A total of 1413 variables were culled to 75 based on random forest classification performance in a 3-group (NL, EIN, EMCA) model. This algorithm correctly classifies cases with 3-class error rates of 0.04 (training set) and 0.058 (validation set); and 2-class (NL vs. EIN+EMCA) error rate of 0.016 (training set) and 0 (validation set). The 4 most heavily weighted variables are surrogates of those previously identified in manual-segmentation machine learning studies (stromal and epithelial area percentages, and normalized epithelial surface lengths). Lesser weighted predictors include gland and lumen axis lengths and ratios, and individual cell measures. Automated image analysis and random forest classification algorithms can classify normal, premalignant, and malignant endometrial tissues. Highest predictive variables overlap with those discovered independently in early models based on manual segmentation.

Validation of a targeted next-generation sequencing approach to detect mismatch repair deficiency in colorectal adenocarcinoma.

Mismatch repair protein deficiency is a hallmark of cancers associated with Lynch syndrome and is a biomarker for response to immunotherapy. With the increasing adoption of cancer next-generation sequencing, there has been a movement to develop screening approaches that take advantage of the unique mutational signatures of mismatch repair-deficient tumors. Here, we develop a sequencing-based metric that distinguishes mismatch repair-deficient from mismatch repair-proficient colorectal adenocarcinomas with comparison to immunohistochemical staining. We find that a single criterion of three or more single base pair insertion or deletion mutations per megabase sequenced, occurring in mononucleotide repeat regions of four or more nucleotides, is sufficient to detect mismatch repair deficiency with 96% sensitivity and 100% specificity in a training set of 241 cancers and 96% sensitivity and 99% specificity in a validation set of 436 additional cancers. Using data from the same cohort, we also find that sequencing information from only three genes-ARID1A, KMT2D, and SOX9-is sufficient to detect mismatch repair-deficient colorectal adenocarcinomas with 76% sensitivity and 98% specificity in the validation set. These findings support the notion that targeted next-generation sequencing already being performed for clinical or research purposes can also be used to accurately detect mismatch repair deficiency in colorectal adenocarcinomas.

Myxoid Inflammatory Myofibroblastic Sarcoma: Clinicopathologic Analysis of 25 Cases of a Distinctive Sarcoma With Deceptively Bland Morphology and Aggressive Clinical Behavior.

The number of recognized sarcoma types harboring targetable molecular alterations continues to increase. Here we present 25 examples of a distinctive myofibroblastic tumor, provisionally termed "myxoid inflammatory myofibroblastic sarcoma," which might be related to inflammatory myofibroblastic tumor, and which occurred in 13 males (52%) and 12 females at a median age of 37 years (range: 7 to 79 years). Primary tumor sites were peritoneum (18 patients; 72%), paratesticular (2; 8%), chest wall (1), upper extremity (1), esophagus (1), retroperitoneum (1), and uterus (1). Nine peritoneal tumors (50%) were multifocal at presentation; all other tumors were unifocal. Tumors showed bland-to-mildly-atypical neoplastic myofibroblasts in a myxoid stroma, with prominent inflammatory infiltrates in 22 cases (88%). Most tumors showed delicate branching stromal vessels like those of myxoid liposarcoma, and most showed infiltrative growth through non-neoplastic tissue. Immunohistochemistry demonstrated expression of SMA (19/25 tumors; 76%), desmin (13/22; 59%), and CD30 (5/11; 45%), while ALK was expressed in 1 tumor (of 25; 4%) that was negative for ALK rearrangement. Sequencing of 11 tumors showed seven to harbor tyrosine kinase fusions (4 PDGFRB , 2 PML :: JAK1 , 1 SEC31A :: PDGFRA ). Two instead harbored hot spot KRAS mutations (G12V and Q61H), and 2 were negative for known driving alterations. Clinical follow-up was available for 18 patients (72%; median: 2.7 years; range: 4 mo-12.3 years). Nine patients (50%) were alive with no evidence of disease, 5 (28%) died of disease, and 4 (22%) were alive with disease. Seven patients (39%) experienced peritoneal relapse or distant metastasis. Two patients showed disease progression on conventional, nontargeted chemotherapy. The patient whose tumor harbored SEC31A :: PDGFRA was treated after multiple relapses with imatinib and sunitinib therapy, with progression-free periods of 5 and 2 years, respectively. Despite its bland appearance, myxoid inflammatory myofibroblastic sarcoma harbors a significant risk for disseminated disease, particularly when it occurs in the peritoneum. Targeted therapy could be considered for patients with disseminated disease.

Melanotic PEComa: A Rare But Distinctive Subtype Analyzed in a Series of 7 Cases.

Perivascular epithelioid cell neoplasms (PEComas) are tumors of uncertain cell lineage that show a strong female predominance. Their hallmark is the presence of combined smooth muscle and melanocytic differentiation. In most cases, melanocytic differentiation is detectable only by immunohistochemistry, but there are rare reports of PEComa with extensive melanin accumulation (so-called "melanotic PEComa"). Here we report a clinicopathologic series of 7 melanotic PEComas that occurred across a wide patient age range of 21 to 82 years (median: 41 y) and with a wide anatomic distribution, including 2 cases in the pelvis and 1 case each in the gallbladder, cervix, eyelid, epidural space, and femur. All tumors were heavily pigmented and, like conventional PEComas, were composed of variably sized neoplastic cells with voluminous granular, or less commonly clear, cytoplasm with prominent nucleoli. All tumors expressed HMB45 by immunohistochemistry, and 6 of 7 showed nuclear TFE3 expression. Where tested, tumors were uniformly negative for Mart-1/Melan-A, S100, desmin, and smooth muscle actin. Molecular analysis identified TFE3 gene rearrangement in 5 of 7 cases, 4 of which were demonstrated by fluorescence in situ hybridization and one by whole-exome RNA sequencing which revealed a SFPQ::TFE3 fusion. The one tumor negative for TFE3 by immunohistochemistry was found instead to harbor a SFPQ::TFEB fusion, the first reported example to our knowledge of TFEB fusion in a PEComa. Clinical follow-up was available for 6 of 7 patients (median: 2.5 y: range: 0.75 to 7 y). The patient whose tumor harbored SFPQ::TFEB died of metastatic disease 9 months after diagnosis. The other tumors behaved in an indolent fashion: 4 patients were alive without evidence of disease at the most recent follow-up and 1 patient died of an unrelated cancer 4 years after diagnosis of the melanotic PEComa. Our results expand the morphologic and molecular spectrum of melanotic PEComa, and awareness of this rare but distinctive subtype is important to ensure accurate diagnosis within the broader family of heavily pigmented neoplasms.

Mesenchymal Neoplasms of the Liver.

Mesenchymal neoplasms of the liver can be diagnostically challenging, particularly on core needle biopsies. Here, I discuss recent updates in neoplasms that are specific to the liver (mesenchymal hamartoma, undifferentiated embryonal sarcoma, calcifying nested stromal-epithelial tumor), vascular tumors of the liver (anastomosing hemangioma, hepatic small vessel neoplasm, epithelioid hemangioendothelioma, angiosarcoma), and other tumor types that can occur primarily in the liver (PEComa/angiomyolipoma, inflammatory pseudotumor-like follicular dendritic cell sarcoma, EBV-associated smooth muscle tumor, inflammatory myofibroblastic tumor, malignant rhabdoid tumor). Lastly, I discuss metastatic sarcomas to the liver, as well as pitfalls presented by metastatic melanoma and sarcomatoid carcinoma.

Hepatic Small Vessel Neoplasm Mimics Prostate Cancer on 18 F-DCFPyl PET/CT.

ABSTRACT: 18 F-DCFPyl is a Food and Drug Administration-approved radiotracer that targets prostate-specific membrane antigen and is used in the detection of recurrent or metastatic prostate cancer. As its use has increased, a growing number of nonprostatic disease entities have been identified that express prostate-specific membrane antigen and can mimic prostate cancer. Thus, the interpreting physician must also consider other variables such as serum prostate-specific antigen levels and the distribution of uptake to avoid an inappropriate diagnosis of metastatic prostate cancer. We describe 18 F-DCFPyl uptake associated with a hepatic small vessel neoplasm, an association previously undescribed in the literature.

Distinctive Nested Glomoid Neoplasm: Clinicopathologic Analysis of 20 Cases of a Mesenchymal Neoplasm With Frequent GLI1 Alterations and Indolent Behavior.

Recently, it has been recognized that a subset of primary soft tissue neoplasms with GLI1 gene alterations exhibit nested architecture and can mimic glomus tumors or well-differentiated neuroendocrine tumors. Here, we report a series of 20 such neoplasms, which we have provisionally termed "distinctive nested glomoid neoplasm." Eleven patients (55%) were female and 9 were male. The median age at presentation was 41.5 years (range: congenital to 74 y). The anatomic distribution was wide, with body sites including the trunk (7 tumors), lower extremity (5), tongue (4), upper extremity (3), and neck (1). Excluding tumors of the tongue, 10 tumors (62%) arose in deep soft tissue and 6 (38%) arose primarily in the subcutis. Tumor size ranged from 0.9 to 11.1 cm (median: 3 cm). Distinctive nested glomoid neoplasms are composed of nests of round-to-ovoid cells with scant, palely eosinophilic cytoplasm and monomorphic nuclei with vesicular chromatin and small nucleoli. The nests are invested by prominent capillary networks, and they are situated within large lobules separated by irregular, thick fibrous septa. Among 18 tumors for which adjacent non-neoplastic tissue could be assessed, perivascular proliferation of tumor cells was identified in 16 tumors (89%). Microcystic architecture was present at least focally in 8 tumors (40%), and myxoid stroma was identified at least focally in 5 (25%). Seven tumors (35%) showed clear cell features. By immunohistochemistry, some tumors expressed MDM2 (7/15; 47%), S100 (5 of 19; 26%), STAT6 (2 of 5; 20%), and AE1/AE3 (1/5; 20%). Tumors rarely expressed pan-keratin (1/10; 10%) or CAM5.2 (1/10), and all tumors were negative for ß-catenin (12 tumors), chromogranin (12), synaptophysin (11), epithelial membrane antigen (10), desmin (10), smooth muscle actin (9), INSM1 (7), and CD34 (6). GLI1 break-apart fluorescence in situ hybridization was performed on 7 tumors, and next-generation sequencing was performed on 15 tumors (10 DNA sequencing only, 1 RNA sequencing only, 4 both DNA and RNA sequencing). Sixteen tumors, including all 15 tested by next-generation sequencing and an additional case tested by fluorescence in situ hybridization only, were found to harbor GLI1 gene alterations: 10 harbored GLI1 gene rearrangements (3 ACTB :: GLI1 , 2 PTCH1 :: GLI1 , 1 HNRNPA1 :: GLI1 , 1 NEAT1 :: GLI1 , 1 TXNIP :: GLI1 , 2 undetermined fusion partners), and 6 harbored GLI1 amplification. Clinical follow-up was available for 10 patients (50%; range: 3 mo to 10 y; median: 6.4 y), including 8 with >1 year of follow-up. Three patients (30%) experienced local recurrence (at intervals of 3 mo to 10 y). None developed distant metastases or died of disease as yet. Overall, our findings support the notion that a subset of GLI1 -altered soft tissue neoplasms are indolent, morphologically distinctive nested glomoid neoplasms that should not be classified as sarcomas.

GLI1 Immunohistochemistry Distinguishes Mesenchymal Neoplasms With GLI1 Alterations From Morphologic Mimics.

Glioma-associated oncogene 1 ( GLI1 ) alterations have been described in pericytoma with t(7;12), gastroblastoma, plexiform fibromyxoma, and an emerging class of GLI1 -rearranged or amplified mesenchymal neoplasms including "nested glomoid neoplasm". The immunophenotype of these tumor types is nonspecific, making some cases difficult to diagnose without sequencing. The utility of GLI1 immunohistochemistry (IHC) in distinguishing nested glomoid neoplasms and pericytomas with t(7;12) from morphologic mimics is unknown. To investigate the diagnostic value of GLI1 IHC, we determined its sensitivity and specificity in a "test cohort" of 23 mesenchymal neoplasms characterized by GLI1 alterations, including 12 nested glomoid neoplasms (7 GLI1 -rearranged, 4 GLI1 amplified, and 1 unknown GLI1 status), 9 pericytomas with t(7;12), 1 gastroblastoma, and 1 malignant epithelioid neoplasm with PTCH1 :: GLI1 fusion. GLI1 IHC was 91.3% sensitive in this cohort; all tumors except 2 pericytomas with t(7;12) expressed GLI1. GLI1 was also expressed in 1 of 8 (12%) plexiform fibromyxomas. Nineteen of 22 GLI1-positive tumors showed nuclear and cytoplasmic staining, while 3 showed nuclear staining only. GLI1 IHC was 98.0% specific; among morphologic mimics [40 well-differentiated neuroendocrine tumors, 10 atypical lung carcinoids, 20 paragangliomas, 20 glomus tumors, 20 solitary fibrous tumors, 10 Ewing sarcomas, 10 alveolar rhabdomyosarcomas (ARMS), 10 BCOR -altered sarcomas, 10 myoepitheliomas, 9 myopericytomas, 9 epithelioid schwannomas, 9 ossifying fibromyxoid tumors, 10 biphasic synovial sarcomas, 10 PEComas, 31 gastrointestinal stromal tumors, 10 inflammatory fibroid polyps, 11 pseudoendocrine sarcomas], 5 of 249 tumors expressed GLI1 (2 well-differentiated neuroendocrine tumors, 1 ARMS, 1 Ewing sarcoma, 1 BCOR -altered sarcoma). GLI1 IHC was also performed on a separate cohort of 13 molecularly characterized mesenchymal neoplasms in which GLI1 copy number gain was identified as a putatively secondary event by DNA sequencing (5 dedifferentiated liposarcoma [DDLPS], 2 adenosarcomas, 2 unclassified uterine sarcomas, 1 leiomyosarcoma, 1 ARMS, 1 intimal sarcoma, 1 osteosarcoma); 2 DDLPS, 1 ARMS, and 1 unclassified uterine sarcoma expressed GLI1. Lastly, because pleomorphic sarcomas sometimes show GLI1 amplification or copy number gain, GLI1 IHC was performed on a separate "pleomorphic sarcoma" cohort: GLI1 was expressed in 1 of 27 DDLPS, 1 of 9 leiomyosarcomas, and 2 of 10 pleomorphic liposarcomas, and it was negative in 23 well-differentiated liposarcomas and 9 unclassified pleomorphic sarcomas. Overall, GLI1 IHC was 91.3% sensitive and 98.0% specific for mesenchymal tumor types with driver GLI1 alterations among morphologic mimics. GLI1 expression was less frequent in other tumor types with GLI1 copy number gain. Given its specificity, in the appropriate morphologic context, GLI1 IHC may be a useful diagnostic adjunct for mesenchymal neoplasms with GLI1 alterations.