Molecular targets for tumour progression in gastrointestinal stromal tumours.

BACKGROUND AND AIMS: The distinction between benign and malignant gastrointestinal stromal tumours (GISTs) is often unclear at the clinical and histopathology levels. GISTs are believed to arise from the stem cells of Cajal. In order to define genetic biomarkers and identify target genes related to GIST progression, we analysed and compared benign and malignant GISTs with verified follow up data using cDNA expression arrays. METHODS: Eight genes were frequently overexpressed in malignant GISTs and their overexpression was confirmed using quantitative real time reverse transcription-polymerase chain reaction. These genes included ezrin (villin 2 (VIL2)), collagen 8 alpha 1 subunit (COL8A1), G2/mitotic specific cyclin B1 (CCNB1), high mobility group protein (HMG2), TSG101 tumour susceptibility protein, CENP-F kinetochore protein, protein tyrosine kinase 2 (FAK), and protein kinase DYRK2. To test these genes in a clinical setting, we obtained diagnostic samples of 16 additional GISTs that were classified at diagnosis as benign, malignant, and uncertain malignant potential (UMP). RESULTS: There was remarkable gene overexpression in all malignant GISTs. Statistical analyses revealed significant correlations between overexpression of several gene pairs in malignant GISTs. We found the strongest correlations (rho>0.70) among the significant correlations (p<0.01) between CCNB1-CENP-F (rho = 0.87) and CCNB1-FAK (rho = 0.73). Gene expression of the UMP GISTs suggested two different groups. Three UMP GISTs had gene expression consistent with malignant tumours and their follow up data revealed that indeed these patients had recurrences later on. On the other hand, UMP GISTs that had low gene expression levels continued free of disease for several years. CONCLUSIONS: These results provide insight into the oncogenesis of GISTs and suggest that testing the expression profile of a number of genes may segregate GISTs into groups of different tumour behaviour.

Frequent loss of SMAD4/DPC4 protein in colorectal cancers.

BACKGROUND AND AIMS: Loss of DNA sequences from chromosome 18q21 is a major genetic change in colorectal tumorigenesis. Multiple genes have been identified in this area. One of these, DPC4 (deleted in pancreatic cancer 4, also known as SMAD4), is mutated in a minor subset of colorectal carcinomas as well as in germlines of humans predisposed to colon tumours. PATIENTS AND METHODS: The involvement of SMAD4 in sporadic colorectal neoplasia was evaluated by immunohistochemistry in 53 unselected cases and 27 cases displaying microsatellite instability. RESULTS: SMAD4 expression was absent in 20 of 53 (38%) unselected colorectal carcinomas, and reduced in another 15 (28%) cases. However, 26 of 27 cancers displaying microsatellite instability and TGF-betaIIR mutations were positive for SMAD4 immunostaining. CONCLUSIONS: Loss of SMAD4 expression may play a more prominent role in colon cancer than anticipated based on genetic evidence, but not in mutator phenotype tumours.

Gastrointestinal stromal tumors, intramural leiomyomas, and leiomyosarcomas in the rectum and anus: a clinicopathologic, immunohistochemical, and molecular genetic study of 144 cases.

Gastrointestinal stromal tumors (GISTs), the specific KIT-positive mesenchymal tumors of the gastrointestinal tract, have been sporadically reported in the rectum, but there are few clinicopathologic series. In this study we analyzed the clinicopathologic features of 133 anorectal GISTs, 3 intramural leiomyomas (LMs), and 8 leiomyosarcomas (LMSs) from the files of the Armed Forces Institute of Pathology and the Haartman Institute of the University of Helsinki. Ninety-six GISTs were documented as KIT-positive and three additional ones as CD34-positive. Thirty-four tumors were included by their histologic similarity to KIT- or CD34-positive cases. GIST-specific c-kit gene mutations, mostly in exon 11, were documented in 18 of 29 cases (62%). The GISTs occurred in adults with the age range of 17-90 years (median 60 years) with a significant male predominance (71%). The tumors ranged from small asymptomatic intramural nodules to large masses that bulged into pelvis causing pain, rectal bleeding, or obstruction. They were mostly highly cellular spindle cell tumors; four tumors had an epithelioid morphology. The tumors coexpressed CD34 and KIT and were rarely positive for smooth muscle actin or desmin and never for S-100 protein. Seventy percent of patients with tumors >5 cm with more than 5 mitoses/50 high power fields (HPF) (n = 31) died of disease, whereas only one tumor <2 cm with <5 mitoses/50 HPF (n = 21) recurred and none caused death. Long latency was common between primary operation and recurrences and metastases; either one occurred in 60 of 111 patients with follow-up (54%). Distant metastases were in the liver, bones, and lungs. Three benign actin- and desmin-positive and KIT-negative intramural LMs, similar to those seen in the esophagus, were identified. There were eight LMSs, six of which formed a polypoid intraluminal mass and were actin-positive and KIT-negative. Despite high mitotic counts, only one LMS patient died of disease. A great majority of rectal smooth muscle and stromal tumors are GISTs, which have a spectrum from minimal indolent tumors to overt sarcomas. Intramural LMs are exceptional, and true LMSs are rare, and similar to colonic ones, often present as intraluminal polypoid masses that appear to have a better prognosis than GISTs with similar mitotic rates.

Mesenchymal tumors of muscularis mucosae of colon and rectum are benign leiomyomas that should be separated from gastrointestinal stromal tumors--a clinicopathologic and immunohistochemical study of eighty-eight cases.

Most mesenchymal tumors of the gastrointestinal tract are currently classified as specific gastrointestinal stromal tumors. However, true leiomyomas are more common in the esophagus, and they have been occasionally noted in the colon and rectum, but the small number of reported cases does not allow for clinicopathologic profiling. This study was undertaken to characterize 88 tumors of the muscularis mucosae of the colon and rectum. Seventy tumors were obtained form the files of AFIP and 18 cases from the Department of Pathology of the Haartman Institute of the University of Helsinki. The lesions, except one, were removed by snare polypectomy as incidental lesions at cancer or polyp surveillance; one small tumor was an incidental finding in the rectal resection specimen. The tumors had a significant male predominance in both institutions (overall 2.4:1). They occurred in age range of 38 to 85 years (median 62 years). The lesions were typically small (range 1 to 22 mM, median 4 mM) and located predominantly in the rectum and sigmoid (72%). All tumors were composed of well-differentiated, eosinophilic smooth muscle cells that were seen immediately beneath the mucosa obliterating the muscularis mucosae layer and merging with it. Two tumors had significant atypia ("symplastic leiomyoma"); mitotic activity was seen in one of these tumors, but not in others. The lesional cells were uniformly positive for smooth muscle actin and desmin and negative for CD34, CD117 and S100-protein, based on immunohistochemical studies on 20 to 24 cases with each marker. No gastrointestinal stromal tumors were identified among the tumors of muscularis mucosae, and no CD117-positive cells, except mast cells, were seen in the muscularis mucosae layer. None of the patients had morbidity related to the tumor. Based on follow-up data on 29 patients, leiomyomas of muscularis mucosae are benign. They should be separated from gastrointestinal stromal tumors that have a clinicopathologic spectrum including frequent disease-related mortality. Snare polypectomy is an adequate treatment, but ensuring the complete removal and follow-up are necessary precautions for tumors with any atypia or mitotic activity.

Chromosomal aberrations in malignant gastrointestinal stromal tumors: correlation with c-KIT gene mutation.

Gastrointestinal stromal tumors (GISTs) are distinctive, KIT positive mesenchymal neoplasms. The genetic alterations leading to the malignant behavior of these tumors are not well known. In this study, we looked for recurrent numerical chromosomal changes, which may be associated with malignant GISTs, using interphase fluorescence in situ hybridization (FISH). Fourteen malignant primary tumors and two intra-abdominal recurrences were analyzed. Nine benign tumors were studied for comparison. In all cases, the presence of mutations in exons 9, 11 and 13 of the KIT gene were evaluated. Sixteen centromeric enumeration probes (CEP) for chromosomes 1, 2, 3, 4, 6, 7, 8, 9, 10, 11, 12, 15, 16, 17, 18, and X and three locus specific probes (LSI) for 22q11.2 (BCR-locus), 13q14 (RB1-locus) and 14q32 (IgH-locus) were used. The most common changes seen in malignant GISTs were losses of 14q32 and 22q11. However, these changes were commonly detected in benign tumors and represent early changes related to the pathogenesis of GISTs. Losses of chromosomes 1 and 9 were the only recurrent numerical changes seen exclusively in malignant GISTs. Other recurrent numerical changes seen predominantly in malignant tumors were gain of chromosome 8 and losses of chromosomes 7 and 15. The concurrent loss of chromosome 7 and gain of chromosome 8 (in 4 cases) was never seen together with loss of chromosomes 9 or 15 and only once with loss of chromosome 1. Mutations in KIT were found in the majority of malignant GISTs (64%) confirming a previously shown correlation between presence of such mutations and malignancy. KIT mutations were seen in four of five malignant GISTs with loss of chromosome 9, but only in one of four malignant tumors with loss of chromosome 1. These observations may reflect the different pathways leading to malignant transformation of GISTs.

Microphthalmia transcription factor in the immunohistochemical diagnosis of metastatic melanoma: comparison with four other melanoma markers.

The diagnosis of metastatic malignant melanoma (MMM) may be difficult in surgical pathology, often complicated by the unpredictable spread of this tumor and its great variability on histologic evaluation. Traditionally used immunohistochemical markers on melanomas are insufficient because of either a relative lack of specificity (S100 protein) or variably reported sensitivity (HMB45). Information about some newer markers, such as tyrosinase (TYR) and Melan A, is more limited. Recently, based on the study of a small number of tumors, it was suggested that microphthalmia transcription factor (MITF) is 100% sensitive in the identification of metastatic melanoma. In the current study, we compared the diagnostic usefulness of MITF with that of four other markers in 266 cases of conventional metastatic melanomas from different sites, 33 cases of desmoplastic melanomas, and 1 case of melanoma with rhabdoid features. The specificity of MITF was evaluated by using a representative sample of control tumors. Microphthalmia transcription factor with nuclear positivity was seen in 235 of 266 cases of conventional MMM (88%), usually in more than 30% of tumor cells. However, some melanomas had only foci of MITF- and TYR-positive cells, whereas the majority of cells were generally S100 protein-positive. Only 1 of 30 desmoplastic melanomas (3%) had MITF-positive cells, representing epithelioid foci resembling conventional melanoma. Two cases had TYR in a similar pattern; all were HMB45-negative. One metastatic melanoma with rhabdoid features was negative for MITF and other markers except the S100 protein. Half of the S100 protein negative conventional melanomas (6 of 12) were MITF-positive, whereas 4 of 20 (20%) TYR-negative tumors had reactivity for MITF. The percentages of positive cases of MMM (10% or more tumor cells positive) diagnosed with the four other markers in descending order were 90% (S100 protein and TYR), 78% (melan-A), and 66% (HMB45). Microphthalmia transcription factor appeared to be specific, because significant reactivity was not found in 112 carcinomas, 20 lymphomas, 20 angiosarcomas, 20 fibrous histiocytomas, and 20 malignant peripheral nerve sheath tumors. However, positive nuclei were found focally among reactive histiocytes, especially in osteoclasts, epithelioid histiocytes, and sporadic other histiocytes. Microphthalmia transcription factor may be a valuable addition to the marker panel used in diagnosing melanoma, in combination with S100, TYR, and the other markers, but it is not present in cases of desmoplastic melanomas.

Immunohistochemical spectrum of GISTs at different sites and their differential diagnosis with a reference to CD117 (KIT).

Gastrointestinal (GI) stromal tumor (GIST) is the designation for the major subset of GI mesenchymal tumors and encompasses most tumors previously classified as GI smooth muscle tumors. Although GISTs typically express CD117 (KIT), often express CD34, and sometimes express alpha-smooth muscle actin (SMA), the relative frequency of these markers has not been characterized in large series of GISTs of different sites, and the CD117 expression has not been fully characterized in intra-abdominal tumors. In this study, we immunohistochemically analyzed 292 GISTs throughout the GI tract, including omentum and mesentery, and compared the immunoreactivities with 211 other tumors that may enter in the differential diagnosis. GISTs were defined in this study as CD117-positive primary spindied or epithelioid mesenchymal tumors of the GI tract, omentum, or mesentery. The CD34 positivity of GISTs varied from 47% in small bowel to 96 to 100% in rectum and esophagus, whereas SMA expression showed the opposite patterns and was most frequent in the GISTs of small bowel (47%) and rarest in the GISTs of rectum and esophagus (10-13%). Desmin was seen only occasionally. S100 positivity was rare but was seen most frequently in small intestinal GISTs (15%). True leiomyomas from esophagus, muscularis mucosae of colorectum, and pericolic leiomyomas similar to uterine leiomyomas were negative for CD117 and CD34 and positive for SMA and desmin (46 of 46). Inflammatory fibroid polyps of stomach and small intestine were negative for CD117 but were often positive for CD34 (6 of 8) and variable for SMA (3 of 8). Inflammatory myofibroblastic tumors involving gastric or colonic wall were negative for CD117 but some showed CD117-positive endothelia. GI schwannomas were all negative for CD117 and positive for S100 protein (11 of 11). Extremely focal CD117 positivity was seen in the neoplastic cells of some retroperitoneal leiomyosarcomas and liposarcomas. Among other CD117-positive tumors were intestinal metastatic melanomas (8 of 11) and extraskeletal Ewing's sarcomas (5 of 11), two of which were abdominal. In conclusion, strong CD117 expression defines most primary GI mesenchymal tumors as GISTs, which show different patterns for CD34 and SMA in various parts of the GI tract. Some unrelated CD117-positive tumors (melanomas, Ewing's sarcomas) should not be confused with GISTs.

Mutations in exons 9 and 13 of KIT gene are rare events in gastrointestinal stromal tumors. A study of 200 cases.

Gastrointestinal stromal tumors (GISTs), the most common mesenchymal tumors of the gastrointestinal tract, typically express the KIT protein. Activating mutations in the juxtamembrane domain (exon 11) of the c-kit gene have been shown in a subset of GISTs. These mutations lead into ligand-independent activation of the tyrosine kinase of c-kit, and have a transforming effect in vitro. Several groups have studied the clinical implication of the c-kit mutation status of exon 11 in GISTs and a possible relationship between c-kit mutations and malignant behavior has been established. Recently, a 1530ins6 mutation in exon 9 and missense mutations, 1945A>G in exon 13 of the c-kit gene were reported. The frequency and clinical importance of these findings are unknown. In this study we evaluated 200 GISTs for the presence of mutations in exons 9 and 13 of c-kit. Six cases revealed 1530ins6 mutation in exon 9 and two cases 1945A>G mutation in exon 13. All tumors with mutations in exon 9 and 13 lacked mutations in exon 11 of c-kit. None of the analyzed tumors had more than one type of c-kit mutation. All but one of the eight tumors with mutations in exon 9 or 13 of the c-kit gene were histologically and clinically malignant. All four of six cases with exon 9 mutation of which location of primary tumor was known, were small intestinal, suggesting that this type of mutation could preferentially occur in small intestinal tumors. Exon 9 and 13 mutations seem to be rare, and they cover only a small portion (8%) of the balance of GISTs that do not have mutations in exon 11 of c-kit. This finding indicates that other genetic alterations may activate c-kit in GISTs, or that KIT is not activated by mutations in all cases.

Gastrointestinal stromal tumors and leiomyosarcomas in the colon: a clinicopathologic, immunohistochemical, and molecular genetic study of 44 cases.

Gastrointestinal stromal tumors (GISTs), mesenchymal tumors largely specific for the gastrointestinal tract, have been well defined in the stomach and small intestine, but have not been extensively documented or contrasted with true smooth muscle tumors in the colon. This study was undertaken to determine the clinicopathologic features of GISTs of the colon, excluding the rectum, and to compare them with leiomyosarcomas (LMSs) of the same location. A total of 37 colonic GISTs and seven LMSs from the files of the Armed Forces Institute of Pathology and the Haartman Institute of the University of Helsinki were analyzed. The GISTs occurred predominantly in adults older than 50 years of age (median, 67 yrs), and most were histologically malignant; four small benign tumors (< or = 1 cm) were incidentally detected, and 10 others had minimal mitotic activity (five or fewer mitoses per 50 high-power fields). The colonic GISTs were typically transmural tumors with frequent intraluminal and outward bulging components. Histologically, they usually showed a spindle cell pattern (92%), whereas 8% were epithelioid. Most tumors (19 of 25) were positive for CD117 (KIT) and for CD34 (16 of 27); six tumors coexpressed alpha-smooth muscle actin and CD117; none showed desmin or S-100 protein. C-kit mutations in exon 11 were seen in 5 (36%) of 14 colonic GISTs. None of the patients with incidental small tumors had a recurrence, whereas 2 of 10 patients with tumors larger than 1 cm but minimal mitotic activity died of the disease with liver metastasis. Nearly all patients whose tumor was larger than 1 cm and showed more than five mitoses per 50 high-power fields died of disease; half had evidence of metastasis. LMSs were typically intraluminally bulging, polypoid masses that showed a histologic likeness to differentiated smooth muscle cells. They occurred in five men and two women with a median age of 61 years. Most LMSs were high-grade histologically and showed smooth muscle actin, desmin, or both. All were negative for CD34 and CD117 and lacked c-kit mutations. Five of the seven patients died of disease, and two had a long-term survival, despite high mitotic activity. These results show that KIT-positive GISTs are more common than LMSs of the colon, and these tumor groups have clinicopathologic differences that warrant their separation.

Immunohistochemical evaluation of failed vessel anastomoses in clinical microsurgery.

Failed vessel anastomoses collected from 12 patients during elective free flap surgery, and from one patient after failed mid-hand replantation were subjected to immunohistochemical analysis. The anastomotic failure was due to an obvious thrombosis, poor flow, an excessively sharp pulse or some other reason causing a non-functioning anastomosis. A total of 17 samples were obtained, 13 of them arterial, three between the artery and vein graft, and one venous. The majority of samples were resected during primary surgery and four of them at reoperation. Variables of coagulation and fibrinolysis were analysed repeatedly during the operation in 7/13 patients. Total occlusion was seen in 6/17 samples and a non-occlusive thrombus in 4/17; two of these were due to suture error. Immunohistochemistry showed that, overall, the endothelial cells (PECAM-l, CD 31) were absent and that the staining pattern for platelets (CD 42b and CD 31) and fibrin (fibrin II, T2G1) correlated. In the absence of a thrombus, however, adherent platelets were positive only for CD 42b, not for PECAM-1. Vessel inflammation was a prominent feature at reoperations. Analysis of coagulation and fibrinolytic markers (thrombin-antithrombin III complex, prothrombin fragment 1 + 2 and D-dimer) confirmed the occurrence of thrombosis in three patients undergoing breast reconstruction with clinically obvious thrombosis during primary surgery or at reoperation. Moreover, the patients with active cancer (2/7) were clearly hypercoagulable compared with the other patients. In short, the primary anastomotic failure was associated with loss of endothelial cells, and with co-localised platelet recruitment and fibrin formation at these sites. At reoperation, inflammation was a prominent feature at the vessel site of thrombi.

DNA sequence copy number changes in gastrointestinal stromal tumors: tumor progression and prognostic significance.

To identify genetic changes related to tumor progression and find out diagnostic and prognostic genetic markers in gastrointestinal stromal tumors (GISTs), 95 tumor samples (24 benign GISTs, 36 malignant primary GISTs, and 35 GIST-metastases) from 60 patients were studied using comparative genomic hybridization. DNA copy number changes were detected in all samples. Benign GISTs had a mean of 2.6 aberrations/ sample (losses:gains, 5:1) and significantly fewer DNA copy number changes and fewer gains than malignant primary and metastatic GISTs (P < 0.01). High-level amplifications were not seen in benign GISTs. Malignant primary GISTs had a mean of 7.5 aberrations/tumor (losses: gains, 1.6:1), whereas the mean number of aberrations/metastatic GIST was 9 (losses:gains, 1.8:1). Frequent changes observed in all GIST groups included losses in chromosome arms 1p (51%), 14q (74%), and 22q (53%). Gains and high-level amplifications at 8q and 17q were significantly more frequent in metastatic GISTs (57 and 43%) than in benign GISTs (8 and 0%; P < 0.001) and malignant primary GISTs (33 and 25%; P < 0.05). Gains and high-level amplifications at 20q were only seen in malignant primary and metastatic GISTs (P < 0.01), and gains at 5p were not detected in benign GISTs (P < 0.01). Losses in chromosome arm 9p were never seen in benign tumors (P < 0.001), and they were more frequent in metastatic GISTs than in malignant primary GISTs (63 and 36%; P < 0.05). Losses in 13q were less frequent in benign GISTs than in malignant primary (P < 0.05) and metastatic (P < 0.01) GISTs. Our results show that several DNA copy number changes are related to the behavior of GISTs and can be used as prognostic markers for tumor progression.

KIT expression in angiosarcomas and fetal endothelial cells: lack of mutations of exon 11 and exon 17 of C-kit.

C-kit proto-oncogene product (KIT, CD117) is a tyrosine kinase growth factor receptor for stem cell factor. This receptor is important for the development and maintenance of hematopoietic stem cells, mast cells, germ cells, melanocytes, and interstitial cells of Cajal and is constitutively expressed in them. Among mesenchymal tumors, KIT seems to be specific for the gastrointestinal stromal tumors, which consistently express this protein. Activating mutations in the tyrosine kinase or juxtamembrane domains of c-kit gene have been found in mastocytoma, seminoma, and gastrointestinal stromal tumors. Following up our initial observation of KIT expression in one angiosarcoma, we examined 50 angiosarcomas, 13 Kaposi sarcomas, 10 epithelioid hemangioendotheliomas, and 31 hemangiomas of different types for KIT expression using a polyclonal antiserum specific to KIT. Adult and fetal tissues and neovascular endothelia in 20 carcinomas were studied for comparison. More than half (56%) of the angiosarcomas representing different clinicopathologic and histologic subtypes and 2 of 13 Kaposi sarcoma were KIT positive. All epithelioid hemangioendotheliomas and hemangiomas were negative, with the exception of two infantile hemangiomas that showed KIT reactivity. The fetal capillary endothelia of lungs, placenta, and soft tissues were also KIT positive, although in soft tissues and placenta, KIT positivity was more prominent in the first trimester. However, endothelia of adult vessels and neovascular capillaries of carcinomas were negative. None of the four KIT-positive angiosarcomas and one KIT-positive Kaposi sarcomas that were studied showed mutations in the juxtamembrane or tyrosine kinase domains of the c-kit gene. These results indicate that KIT expression occurs in a subset of angiosarcomas, and the expression probably represents oncofetal expression (i.e., reversion of the tumor cell phenotype to that of fetal endothelial cells that may show KIT expression).

FGF4 and INT2 oncogenes are amplified and expressed in Kaposi's sarcoma.

Kaposi's sarcoma (KS) is a vascular tumor, the pathogenesis of which has been suggested to include human herpesvirus 8 (HHV-8) as well as various cytokines and growth factors. Very little is known about cytogenetic and molecular genetic changes in KS. We studied DNA copy number changes in KS and found a recurrent gain at 11q13. We then analyzed the amplification and expression status of two known oncogenes, FGF4 and INT2, residing at 11q13. Comparative genomic hybridization, interphase fluorescence in situ hybridization with yeast artificial chromosome probes containing FGF4 and INT2, and immunoperoxidase immunostaining with anti-FGF4 and -INT2 antibodies were used on 12 KS samples. All samples tested were shown by polymerase chain reaction to be HHV-8 positive. A recurrent gain at 11q13 was shown by comparative genomic hybridization in 4 of 10 cases studied. Of six cases studied by interphase fluorescence in situ hybridization, four showed a 3- to 4-fold amplification with the probes containing FGF4 and INT2. Expression of FGF4 and INT2 was found in nine and three cases, respectively, of nine studied. Amplification and expression of these genes is particularly interesting in the context of oncovirus involvement, because INT2 is a homolog of mouse int2 which causes mammary carcinoma in mice when activated by integration of retrovirus mouse mammary tumor virus. This raises the question of whether HHV-8 represents an integrating oncovirus that causes amplification and activation of genomic oncogenes in humans.

High-resolution deletion mapping of chromosome 14 in stromal tumors of the gastrointestinal tract suggests two distinct tumor suppressor loci.

DNA copy number losses at chromosome arm 14q are the most frequently occurring aberrations in gastrointestinal stromal tumors (GISTs). To characterize the deletion at 14q, we performed comparative genomic hybridization (CGH) and high-resolution deletion mapping using a panel of 32 polymorphic microsatellite markers in 30 GISTs. The GISTs were classified according to their metastatic potential and mitotic counts into 15 low-risk and 15 high-risk tumors. Losses with a minimal common overlapping region at 14q12-q24 were detected by CGH in 16 tumors (53) (nine low-risk and seven high-risk). Investigation with microsatellite markers was informative in 690 analyses (72%). Loss of heterozygosity (LOH) with at least one marker was detected in 279 analyses in 24 tumors (80%). Deletions were equally frequent in low-risk and high-risk GISTs. Two common deletion regions were identified at 14q11.1-q12 and 14q23-q24.3. The highest frequencies of deletions were seen in regions corresponding to markers D14S283 (20/28, 71%) at 14q11.1-q12 and D14S258 (17/27, 63%) at 14q23-q24, suggesting that these are two tumor suppressor loci.

Esophageal stromal tumors: a clinicopathologic, immunohistochemical, and molecular genetic study of 17 cases and comparison with esophageal leiomyomas and leiomyosarcomas.

Although rare elsewhere in the gastrointestinal tract, leiomyomas (LMs) are the most common esophageal mesenchymal neoplasms. In contrast, gastrointestinal stromal tumors (GISTs) predominate in the stomach and intestines but have not been documented in the esophagus. This study was undertaken to determine the clinicopathologic features and frequency of esophageal GISTs compared with LMs and leiomyosarcomas (LMSs) of the esophagus. A total of 68 stromal/smooth muscle tumors from the Armed Forces Institute of Pathology and the Haartman Institute of University of Helsinki were reclassified by current histologic and immunohistochemical criteria. There were 17 GISTs, 48 LMs, and three LMSs. The esophageal GISTs occurred in 12 men and five women with a median age of 63 years (range, 49-75 years). All tumors were from the lowest third of the esophagus, and the most common complaint was dysphagia, whereas two tumors were detected incidentally. Histologically the tumors had an overall basophilic appearance and showed combinations of solid, myxoid, and perivascular collarlike patterns with a spindle cell histology in 13 patients and epithelioid histology in four patients. All tumors were positive for CD117 and for CD34, whereas two patients were also positive for alpha-smooth muscle actin (SMA) and three patients were positive for desmin. One patient showed a unique immunophenotype with coexpression of CD117, CD34, SMA, and desmin. Nine patients died of disease, including all who had a tumor larger than 10 cm, and also one patient whose tumor showed five mitoses per 50 high-power fields. In comparison, esophageal LMs (n = 48) occurred in a younger population (median age, 35 years) but, similar to the GIST group, men predominated (67%). All LMs were clinically indolent tumors with no tumor-related mortality. The LMs showed eosinophilic cytoplasm, and were positive for desmin and SMA, and negative for CD117 and CD34. All three LMSs were large high-grade tumors that showed muscle cell markers but no CD117. All patients died of disease. Esophageal GISTs showed mutations in exon 11 of c-kit as described previously in gastric and intestinal GISTs. The separation of GISTs from esophageal LMs is important diagnostically because the former group has a high risk of malignant behavior.

Gastrointestinal stromal tumors: recent advances in understanding of their biology.

Gastrointestinal stromal tumor (GIST) is the preferred term for mesenchymal tumors specific for the gastrointestinal tract (60% in stomach, 30% small intestine, 10% elsewhere). GISTs include most tumors previously designated as leiomyoma, cellular leiomyoma, leiomyoblastoma, and leiomyosarcoma. However, in the esophagus, leiomyoma is the most common mesenchymal tumor. GISTs are composed of spindle (70%) or epithelioid (30%) cells, and 10%-30% are malignant showing intra-abdominal spread or liver metastases. They are immunohistochemically positive for c-kit (CD117), CD34, and sometimes for actin but are almost always negative for desmin and S100-protein. The malignant GISTs especially show activating mutations in the c-kit gene. GISTs and gastrointestinal autonomic nerve tumors (GANT) overlap. The cell of origin is not fully understood, but resemblance to the interstitial cells of Cajal, expression of some smooth muscle markers, and occurrence outside of the GI-tract suggest origin from multipotential cells that can differentiate into Cajal and smooth muscle cells.

Gastrointestinal stromal tumors/smooth muscle tumors (GISTs) primary in the omentum and mesentery: clinicopathologic and immunohistochemical study of 26 cases.

Gastrointestinal stromal tumor or smooth muscle tumor (GIST) is the designation for a major subset of gastrointestinal mesenchymal tumors that histologically, immunohistochemically, and genetically differ from typical leiomyomas, leiomyosarcomas, and schwannomas. Because GISTs, like the interstitial cells of Cajal, the gastrointestinal pacemaker cells, express CD117 (c-kit protein), the origin of GISTs from the interstitial cells of Cajal has been recently proposed. Comparison of GISTs primary in the omentum and mesentery to GISTs primary in the tubular gastrointestinal tract is of particular diagnostic and histogenetic interest in view of the possible similarity of these tumors with the GIST group. In this study, we analyzed 14 omental and 12 mesenteric primary mesenchymal tumors representing smooth muscle tumors or GISTs. These tumors were phenotypically compared with gastric and small intestinal GISTs, leiomyomas of the esophagus, and leiomyosarcomas of the retroperitoneum. Most (13 of 14) omental and mesenteric (10 of 12) tumors showed histologic features similar to GISTs with elongated spindle cells or epithelioid cells with high cellularity; most of these tumors showed low mitotic activity. Omental and mesenteric GISTs were typically positive for CD117 and less consistently for CD34. They often showed alpha-smooth muscle actin reactivity but were virtually negative for desmin and S-100 protein. One omental and two mesenteric tumors showed features of leiomyosarcoma with ovoid, less elongated nuclei, cytoplasmic eosinophilia; all these tumors had significant mitotic activity. These tumors were positive for alpha-smooth muscle actin and two of them for desmin, but all were negative for CD34 and CD117, similar to retroperitoneal leiomyosarcomas. Tumor-related mortality occurred in the group of mesenteric GISTs, but not in the group of omental GISTs. In contrast, all three patients with a true leiomyosarcoma of the omentum or mesentery had documented liver metastases or died of tumor. In summary, we show that tumors phenotypically identical with GISTs occur as primary tumors in the omentum and mesentery. The occurrence of CD117-positive tumors outside the gastrointestinal tract militates against an origin of these tumors exclusively from the interstitial cells of Cajal.

Recurrent DNA copy number losses associated with metastasis of larynx carcinoma.

Squamous cell carcinomas of the head and neck show frequent and complex chromosome aberrations, but little is known about the changes that occur during the metastatic process. To compare the accumulation of changes in primary and metastatic tumors we analyzed 19 pairs of primary larynx cancer tumors and their metastases. The most frequent changes were found at 3p, 3q, 5p, 9, and 13. Losses at 13, 8p, and 9q were more frequent in metastases than in primary tumors.

Calponin and h-caldesmon in soft tissue tumors: consistent h-caldesmon immunoreactivity in gastrointestinal stromal tumors indicates traits of smooth muscle differentiation.

Currently, the immunohistochemical evaluation of smooth muscle differentiation is usually based on desmin, which also reacts with skeletal muscle and is not present in all smooth muscle tumors, and alpha-smooth muscle actin, which reacts with myoepithelial cells. Neither marker typically reacts with gastrointestinal stromal tumors (GISTs), previously classified as smooth muscle tumors or presently often classified as smooth muscle/stromal tumors. Two cytoskeleton-associated actin-binding proteins, calponin (CALP) and h-caldesmon (HCD), are putative smooth muscle markers that also react with myoepithelia. These markers are of particular interest in the immunohistochemical analysis of tumors; neither of them has been extensively documented in soft tissue tumors. In this study, we evaluated selected normal and reactive tissues and more than 250 mesenchymal tumors for CALP and HCD. Both markers were expressed in parenchymal and vascular smooth muscle cells in various organs and in myoepithelial cells. CALP also reacted with myofibroblasts of desmoplastic stroma. All of our 25 benign smooth muscle tumors from various locations were positive for CALP and HCD, as were most of the retroperitoneal and uterine leiomyosarcomas. HCD was more specific, because CALP also reacted with myofibroblastic lesions. The common reactivity of malignant fibrous histiocytomas with CALP and HCD suggests a combination of myofibroblastic and smooth muscle differentiation in these tumors. The GISTs (c-kit positive, usually actin negative) showed nearly consistent HCD reactivity, suggesting traits of smooth muscle differentiation. GISTs were usually CALP negative and showed a CALP expression pattern similar to that of alpha-smooth muscle actin. Although nonmuscle, nonmyofibroblastic tumors were negative for CALP and HCD, synovial sarcomas showed streaks of CALP-positive cells of unknown significance. CALP and HCD should be explored as markers to identify myofibroblastic and smooth muscle cell differentiation in mesenchymal tumors.