Molecular analysis of phyllodes tumors reveals distinct changes in the epithelial and stromal components

Phyllodes tumors are fibroepithelial mammary lesions that tend to behave in a benign fashion but may undergo sarcomatous transformation. A study of clonality in these tumors has suggested that the epithelial component is polyclonal, but the stroma is monoclonal, and thus forms the neoplastic component of the lesion. In this study microsatellites on chromosome 1q and chromosome 3p were assessed for allelic imbalance (AI) in 47 phyllodes tumors; in all cases stroma and epithelium were analyzed separately. Ten of 42 (24%) phyllodes tumors showed AI at one or more markers on 3p, and 14 of 46 (30%) showed AI on chromosome 1. Five tumors had changes in both the epithelium and stroma. Eight tumors had changes only detectable in the stroma and eight, changes in the epithelium only. Three tumors exhibited low-level microsatellite instability in the epithelium but not in the stroma. The results show that AI on 3p and 1q does occur in phyllodes tumors and that it can occur in both the stroma and epithelium, sometimes as independent genetic events. These unexpected findings throw into doubt the classical view that phyllodes tumors are simply stromal neoplasms and raise questions about the nature of stromal and epithelial interactions in these tumors.     

Array comparative genomic hybridization analysis of chromosomal imbalances and their target genes in gastrointestinal stromal tumors

Gastrointestinal stromal tumor (GIST) is the most common mesenchymal tumor of the gastrointestinal tract. The tumors characteristically harbor KIT or PDGFRA mutations, and mutant tumors respond to imatinib mesylate (Glivectrade mark). Chromosomal imbalances resulting in altered gene dosage are known to have a role in the molecular pathogenesis of these tumors, but the target genes remain to be identified. The present study aimed to identify some of these genes. In total, 35 GIST samples were screened for chromosomal imbalances by array-based comparative genomic hybridization. A cDNA array was used to define the minimal common overlapping areas of DNA copy number change. Eight confirmative, replicate hybridizations were performed using an oligonucleotide array. The most recurrent copy number losses were localized to 14q, 22q, and 1p. Gains were less common with 8q being the most recurrent. Two recurrent deleted regions of 14q were 14q11.2 harboring the PARP2, APEX1, and NDRG2 genes and 14q32.33 harboring SIVA. Additional target candidates were NF2 at chromosome 22, CDKN2A/2B at 9p, and ENO1 at 1p for copy number losses, and MYC at 8q for copy number gains. Array CGH proved to be an effective tool for the identification of chromosome regions involved in the development and progression of GISTs.     

High incidence of microscopic gastrointestinal stromal tumors in the stomach

Gastrointestinal stromal tumors (GISTs) are rare mesenchymal neoplasms with an annual incidence of approximately 10 to 20 per 1 million cases. Although pathologists have often observed incidental small GISTs in the stomach resected from patients with gastric cancer, no report on the real incidence of gastric GISTs is available. In this study, 100 whole stomachs resected from patients with gastric cancer were sectioned at 5-mm intervals and hematoxylin and eosin-stained slides (a mean of 130 slides for each case) were examined for microscopic GISTs. KIT (CD117), CD34, and desmin expression of the incidental tumors was evaluated by immunohistochemistry, and genomic DNA extracted from formalin-fixed and paraffin-embedded tumor tissues was analyzed for c-kit gene mutations in exon 11. In 35 of the 100 whole stomachs, we found 50 microscopic GISTs, all of which were positive for KIT and/or CD34 and negative for desmin. Most microscopic GISTs (45/50, 90%) were located in the upper stomach. Two of the 25 (8%) microscopic GISTs had c-kit gene mutations. Fifty-one leiomyomas with positive expression for desmin were observed in 28 of the 100 stomachs. Both leiomyomas and GISTs were found in 12 stomachs. These results indicate that microscopic GISTs are common in the upper portion of the stomach. Considering the annual incidence of clinical GISTs, only few microscopic GISTs may grow into a clinical size with malignant potential. Further studies are required to clarify the genetic events responsible for the transformation of microscopic GISTs to clinical GISTs.     

KIT 1530ins6 mutation defines a subset of predominantly malignant gastrointestinal stromal tumors of intestinal origin

Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal tumors of the gastrointestinal tract. GISTs express KIT and show gain-of-function KIT mutations. Most of these mutations affect the KIT juxtamembrane domain, but other KIT domains are mutated at a lower frequency. In this study, frequency of GCC TAT insertion mutation (1530ins6) in KIT exon 9 (extracellular domain) and its possible clinicopathologic significance was investigated. Screening of 520 GISTs identified 26 cases with 1530ins6 KIT mutation and confirmed the previously reported low frequency of this type of KIT mutation among GISTs of different locations. Of the 26 tumors with 1530ins6 KIT mutation studied, 21 originated from the small intestine, 1 from the colon, and 3 from the rectum. In 1 case, primary small intestinal versus colonic localization could not be clearly established because of intra-abdominal dissemination. No distinctive morphological features were identified for the cohort of tumors defined by 1530ins6 KIT mutations. Most of the tumors showed predominant spindle cell morphology, and a few cases had epithelioid or pleomorphic histological features. Following previously published criteria based on tumor size and mitotic rate, 22 of 26 (85%) tumors were classified as malignant or potentially malignant, and 4 (15%) were classified as probably benign. A malignant clinical course was documented in 18 of 19 tumors from the malignant category. The survival times of 11 patients who died of disseminated GISTs ranged from 1 month to 105 months (median survival time, 26 months). In contrast, 2 of 4 GISTs assigned as probably benign tumors with follow-up information had long disease-free survival. GISTs carrying 1530ins6 occur exclusively in the intestinal location, and a great majority of these tumors follow a malignant course.     

Nuclear and mitochondrial DNA microsatellite instability in gastrointestinal stromal tumors.

OBJECTIVE: Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal tumors of the digestive tract. Nuclear (nMSI) and mitochondrial microsatellite instability (mtMSI) play important roles in tumorigenesis in various organs. The aim of this study was to evaluate the role of nMSI and mtMSI in GISTs. METHODS: Samples from 74 mesenchymal tumors were collected. nMSI and mtMSI were examined by microsatellite assay at BAT26 and D310 mononucleotide repeats in mtDNA, respectively. We compared nMSI, mtMSI and clinicopathologic features, including patient age and sex, tumor location, tumor size, presence of tumor ulceration and presence of distant metastasis, for 51 GISTs for which these data were available. RESULTS: nMSI and mtMSI were detected in 3 (5%) and 10 (16%) of the 62 GISTs, respectively. There was no significant relationship between nMSI, mtMSI and clinicopathologic features. CONCLUSION: These results suggest that mtMSI may play a role, but that nMSI may play little role in the development of GISTs.     

Pathology of gastrointestinal stromal tumors

Gastrointestinal stromal tumors (GIST) are the most common mesenchymal tumors in the gastrointestinal tract. It was found that most GIST expressed KIT, a receptor tyrosine kinase encoded by protooncogene c-kit. In normal gastrointestinal wall, KIT is expressed by interstitial cells of Cajal (ICC), which are a pacemaker for autonomous gastrointestinal movement. Because both GIST and ICC are double-positive for KIT and CD34, and because familial and multiple GIST appear to develop from diffuse hyperplasia of ICC, GIST are considered to originate from ICC or their precursor cells. It was also found that approximately 90% of the sporadic GIST have somatic gain-of-function mutations of the c-kit gene, and that the patients with familial and multiple GIST have germline gain-of-function mutations of the c-kit gene. These facts strongly suggest that the c-kit gene mutations are a cause of GIST. Approximately half of the sporadic GIST without c-kit gene mutations were demonstrated to have gain-of-function mutations in platelet-derived growth factor receptor-alpha (PDGFRA) gene that encodes another receptor tyrosine kinase. Because KIT is immunohistochemically negative in a minority of GIST, especially in PDGFRA gene mutation-harboring GIST, mutational analyses of c-kit and PDGFRA genes may be required to diagnose such GIST definitely. Imatinib mesylate was developed as a selective tyrosine kinase inhibitor. It inhibits constitutive activation of mutated KIT and PDGFRA, and is now being used for KIT-positive metastatic or unresectable GIST as a molecular target drug. Confirmation of KIT expression by immunohistochemistry is necessary for application of the drug. The effect of imatinib mesylate is different in various types of c-kit and PDGFRA gene mutations, and the secondary resistance against imatinib mesylate is often acquired by the second mutation of the identical genes. Mutational analyses of c-kit and PDGFRA genes are also significant for prediction of effectiveness of drugs including newly developed agents.     

Cell markers in gastrointestinal stromal tumors

Stromal tumors of the gastrointestinal (GI) tract have generated considerable controversy about their direction and level of differentiation, particularly about whether the tumor cells are smooth muscle or Schwann cells. In an attempt to characterize these tumors, the immunohistochemical staining patterns of desmin, vimentin, actin, and S-100 protein were studied in 41 GI stromal tumors, using the avidin-biotin method, and compared with normal host smooth muscle and nerve and with esophageal and uterine leiomyomas. Twenty gastric and one rectal tumor stained diffusely with vimentin and actin, but not with desmin, and had scattered strongly S-100-positive cells that might either be trapped Schwann cells or tumor cells. Twenty small bowel tumors stained similarly to the gastric tumors with regard to vimentin, actin, and desmin, but most (17/20) had a unique, strongly positive geographic staining pattern with S-100. No differences in staining were noted between benign and malignant tumors in either gastric or small bowel sites, and most histologic patterns in tumors from similar locations stained similarly. These results suggest that GI stromal tumors are not truly "leiomyomas and leiomyosarcomas," but relatively undifferentiated tumors, with the expression of various antigens depending on their location in the gut.     

High expression of the RNA-binding protein RBPMS2 in gastrointestinal stromal tumors

Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal neoplasms of the gastrointestinal tract and are often associated with KIT or PDGFRA gene mutations. GIST cells might arise from the interstitial cells of Cajal (ICCs) or from a mesenchymal precursor that is common to ICCs and smooth muscle cells (SMCs). Here, we analyzed the mRNA and protein expression of RNA-Binding Protein with Multiple Splicing-2 (RBPMS2), an early marker of gastrointestinal SMC precursors, in human GISTs (n = 23) by in situ hybridization, quantitative RT-PCR analysis and immunohistochemistry. The mean RBPMS2 mRNA level in GISTs was 42-fold higher than in control gastrointestinal samples (p < 0.001). RBPMS2 expression was not correlated with KIT and PDGFRA expression levels, but was higher in GISTs harboring KIT mutations than in tumors with wild type KIT and PDGFRA or in GISTs with PDGFRA mutations that were characterized by the lowest RBPMS2 levels. Moreover, RBPMS2 levels were 64-fold higher in GIST samples with high risk of aggressive behavior than in adult control gastrointestinal samples and 6.2-fold higher in high risk than in low risk GIST specimens. RBPMS2 protein level was high in 87% of the studied GISTs independently of their histological classification. Finally, by inhibiting the KIT signaling pathway in GIST882 cells, we show that RBPMS2 expression is independent of KIT activation. In conclusion, RBPMS2 is up-regulated in GISTs compared to normal adult gastrointestinal tissues, indicating that RBPMS2 might represent a new diagnostic marker for GISTs and a potential target for cancer therapy.     

Diagnostic morphological features of PDGFRA-mutated gastrointestinal stromal tumors: molecular genetic and histologic analysis of 60 cases of gastric gastrointestinal stromal tumors

In this study, 60 gastrointestinal stromal tumors of the stomach were analyzed to elucidate the possible relation of their morphology to the mutation status of KIT and PDGFRA genes. The patients included 27 men and 33 women with a mean age of 63.8 years (range, 12-92 years). Only 1 tumor occurred before the age of 21 years. KIT mutations were detected in 31 cases (51.7%), PDGFRA mutations in 22 cases (36.7%), and 7 cases (11.7%) were KIT and PDGFRA wild type. When the mutation status was correlated with histologic features of the tumors, epithelioid or mixed epithelioid/spindle cell pattern and mast cell infiltration were found as the most reliable signs of PDGFRA mutation. Neoplastic rhabdoid cells and multinucleated giant cells, also previously reported as features of PDGFRA-mutated gastrointestinal stromal tumors, seemed to be less specific but still helpful markers in our study. Finally, tumor-infiltrating lymphocytes and myxoid stroma do not seem to be valuable histologic signs.     

Comparative genomic hybridization analysis of N-methyl-N'-nitrosoguanidine-induced rat gastrointestinal tumors discloses a cytogenetic fingerprint

Exposure to N-nitroso compounds is thought to play a key role in the development of gastric cancer in humans. The alkylating agent N-methyl-N'-nitrosoguanidine (MNNG) is carcinogenic in a number of animal models and its preferential target tissue is the gastrointestinal (GI) tract. The genetic synteny among rats and humans makes the rat a useful model for induced tumorigenesis. However, because of the limited availability of genetic information, cytogenetic and molecular studies are rarely performed in the rat. We report an investigation of eight MNNG-induced rat gastric tumors by comparative genomic hybridization (CGH). The tumors were from forestomach (induced by a single dose of MNNG) and from pylorus (induced by chronic exposure). CGH identified a genetic fingerprint of chromosomal imbalances common to the two types of the tumors. Frequent gains were observed at 9q11-q12, 15q22-25, and Xq11-q12. Forestomach carcinomas were also characterized by gains in 7q11-q12, 20q13, and Yq12. Homology studies between the rat and human genomes indicate the presence of genes within these regions with potential relevance to tumorigenesis in the GI tract. Our findings provide new insights into the location of genes involved in MNNG-induced gastric cancer initiation and/or progression in the rat.     

Ultrastructural studies of gastrointestinal stromal tumors

Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal tumors in the gastrointestinal tract (GIT). Although interstitial cells of Cajal has been suggested as origin of this tumor, the cytological and ultrastructural features of GISTs are heterogeneous and unclear. A total 10 cases of normal gastrointestinal tissue (control), 13 GISTs of the stomach (8), small intestine (3), mesocolon (1) and liver (1), and 2 gastrointestinal autonomic nervous tumor (GANT) of small intestine were ultrastructurally studied. Normal interstitial cells of Cajal (ICC) were abundantly present around the myenteric plexuses or individually scattered through the wall of GIT. ICC was characterized by slender cytoplasmic processes, well-developed endoplasmic reticulum (ER), mitochondria, Golgi apparatus, caveolae and intermediate filaments. The GISTs and GANTs had overlapping ultrastructures. The most common and important ultrastructural features of GISTs were rich villous cytoplasmic processes, dispersed intermediate filaments and abundant SER, and those of GANTs were neurosecretory granules and skenoid fibers. Compared with ICC, the GISTs and GANTs had remarkably reduced caveolae and gap junctions. Our study suggested that ultrastructural analysis gives much information to investigate lineage differentiation of neoplastic cells and make a differential diagnosis of these tumors from other mesenchymal tumors and between GISTs and GANTs.     

Morphological characteristics of stromal gastrointestinal tumors

Stromal tumors are singled out from smooth muscle and neurogenic neoplasms into a special group due to differences in CD117 expression caused by mutation of c-kit gene. Out of 57 stromal tumors, 37 (64,9%) located in the stomach, 17 (29,8%) in the small intestine and 3 (5,3%) in the colon. Immunohistochemically, all the tumors expressed CD117 and vimentine. Smooth muscle actin was found in 82% tumors, S-100 protein in 75%, neuron-specific enolase in 66% cases. Malignant tumors were in 93% cases, and in 7% benign. Metastases were observed in 47.7% cases, recurrences in 14%. The liver was most frequent site of metastases (88.9%), peritoneum (51.9%). 21% patients died of progression of the underlying disease during the follow-up of 6-60 months.     

Large-scale methylation analysis of human genomic DNA reveals tissue-specific differences between the methylation profiles of genes and pseudogenes

Cytosine in CpG dinucleotides is frequently found to be methylated in the DNA of higher eukaryotes and differential methylation has been proposed to be a key element in the organization of gene expression in man. To address this question systematically, we used bisulfite genomic sequencing to study the methylation patterns of three X-linked genes and one autosomal pseudogene in two adult individuals and across nine different tissues. Two of the genes, SLC6A8 and MSSK1, are tissue-specifically expressed. CDM is expressed ubiquitously. The pseudogene, psi SLC6A8, is exclusively expressed in the testis. The promoter regions of the SLC6A8, MSSK1 and CDM genes were found to be essentially unmethylated in all tissues, regardless of their relative expression level. In contrast, the pseudogene psi SLC6A8 shows high methylation of the CpG islands in all somatic tissues but complete demethylation in testis. Methylation profiles in different tissues are similar in shape but not identical. The data for the two investigated individuals suggest that methylation profiles of individual genes are tissue specific. Taken together, our findings support a model in which the bodies of the genes are predominantly methylated and thus insulated from the interaction with DNA-binding proteins. Only unmethylated promoter regions are accessible for binding and interaction. Based on this model we propose to use DNA methylation studies in conjunction with large-scale sequencing approaches as a tool for the prediction of cis-acting genomic regions, for the identification of cryptic and potentially active CpG islands and for the preliminary distinction of genes and pseudogenes.     

Molecular cytogenetic definition of three distinct chromosome arm 14q deletion intervals in gastrointestinal stromal tumors

Gastrointestinal stromal tumors (GISTs) are mesenchymal neoplasms characterized by frequent chromosome arm 14q losses. In this study, the 14q changes in a series of 39 histologically and immunohistochemically confirmed GISTs were analyzed in detail by metaphase and/or interphase fluorescence in situ hybridization (FISH) studies using 21 genetically well-characterized, region-specific 14q11-24 YAC clones. By conventional cytogenetic analysis, acquired clonal chromosome aberrations were found in 17 out of 35 tumors. Chromosome 14 was involved in 13 cases; six specimens showed complete chromosome 14 loss, while the remaining seven had structural abnormalities with the breakpoints residing within the intervals 14q11-13 or 14q22-24. Other recurrent chromosome aberrations included frequent deletions of chromosome 1p (11/17), losses of chromosome 22 (7/17), losses or deletions of chromosome arm 13 (6/17) or 15 (4/17), and gains or translocations involving chromosome 17 (4/17). Combining cytogenetic data with double-color FISH analysis, total or partial losses of 14q material were detected in 29 out of 36 tumors (81%). The 14q losses were found in all stages and histological subtypes. Two most frequent common deletion regions flanked by YACs 931B1 and 761D4, and 802E7 and 892C11 at 14q23-24 (25/30 of each; 83%) could be identified. Furthermore, 21 tumors (70%) shared a region of deletion defined by YACs 957H10 and 931E5 at 14q11-12. Our results suggest the presence of at least three distinct critical deletion regions on chromosome 14 in GISTs.     

Diagnosis of gastrointestinal stromal tumors: a consensus approach

As a result of major recent advances in understanding the biology of gastrointestinal stromal tumors (GIST), specifically recognition of the central role of activating KIT mutations and associated KIT protein expression in these lesions, and the development of novel and effective therapy for GISTs using the receptor tyrosine kinase inhibitor STI-571, these tumors have become the focus of considerable attention among pathologists, clinicians, and patients. Stromal/mesenchymal tumors of the gastrointestinal tract have long been a source of confusion and controversy with regard to classification, line(s) of differentiation, and prognostication. Characterization of the KIT pathway and its phenotypic implications has helped to resolve some but not all of these issues. Given the now critical role of accurate and reproducible pathologic diagnosis in ensuring appropriate treatment for patients with GIST, the National Institutes of Health (NIH) convened a GIST workshop in April 2001 with the goal of developing a consensus approach to diagnosis and morphologic prognostication. Key elements of the consensus, as described herein, are the defining role of KIT immunopositivity in diagnosis and a proposed scheme for estimating metastatic risk in these lesions, based on tumor size and mitotic count, recognizing that it is probably unwise to use the definitive term benign for any GIST, at least at the present time.     

Prognostic factors in malignant gastrointestinal stromal tumors

Gastrointestinal stromal tumors are a heterogeneous group of mesenchymal neoplasms of the gastrointestinal tract in which routine histopathological evaluation fails to reveal definitive evidence of differentiation. Given the heterogeneity in clinical presentation and the frequent morphological overlap, the biological behavior of these neoplasms is difficult to predict. We have evaluated, by Cox Proportional Hazards Regression Analysis, the clinicopathological features of 51 malignant gastrointestinal stromal tumors to identify predictors of survival. In the univariate analysis, survival inversely correlated with size, number of mitoses, and patient's age. In the multivariate analysis, only the degree of necrosis and phenotypic differentiation toward smooth muscle were found to be indicators of poor prognosis. Based on these results, a simple classification scheme for gastrointestinal stromal tumors is proposed. This classification appears to have great prognostic value for these tumors, and may be useful in guiding therapeutic management.