ROR2 is a novel prognostic biomarker and a potential therapeutic target in leiomyosarcoma and gastrointestinal stromal tumour

Soft-tissue sarcomas are a group of malignant tumours whose clinical management is complicated by morphological heterogeneity, inadequate molecular markers and limited therapeutic options. Receptor tyrosine kinases (RTKs) have been shown to play important roles in cancer, both as therapeutic targets and as prognostic biomarkers. An initial screen of gene expression data for 48 RTKs in 148 sarcomas showed that ROR2 was expressed in a subset of leiomyosarcoma (LMS), gastrointestinal stromal tumour (GIST) and desmoid-type fibromatosis (DTF). This was further confirmed by immunohistochemistry (IHC) on 573 tissue samples from 59 sarcoma tumour types. Here we provide evidence that ROR2 expression plays a role in the invasive abilities of LMS and GIST cells in vitro. We also show that knockdown of ROR2 significantly reduces tumour mass in vivo using a xenotransplantation model of LMS. Lastly, we show that ROR2 expression, as measured by IHC, predicts poor clinical outcome in patients with LMS and GIST, although it was not independent of other clinico-pathological features in a multivariate analysis, and that ROR2 expression is maintained between primary tumours and their metastases. Together, these results show that ROR2 is a useful prognostic indicator in the clinical management of these soft-tissue sarcomas and may represent a novel therapeutic target.     

Discovered on gastrointestinal stromal tumours 1 (DOG1) expression in non-gastrointestinal stromal tumour (GIST) neoplasms

Hemminger J & Iwenofu O H
(2012) Histopathology  61, 170–177 Discovered on gastrointestinal stromal tumours 1 (DOG1) expression in non‐gastrointestinal stromal tumour (GIST) neoplasms Aims: To further characterize discovered on GIST1 (DOG1) antibody clone K9 expression in a broad range of mesenchymal and epithelial tumours. Methods and results: Formalin‐fixed paraffin‐embedded sections of various tumours were stained with the anti‐DOG1 monoclonal antibody clone K9. The tumours (n = 187) included: gastrointestinal stromal tumours (GISTs) (n = 20); malignant melanoma (n = 19); schwannoma (n = 10); neurofibroma (n = 10); leiomyosarcoma (n = 10); low‐grade fibromyxoid sarcoma (n = 5); angiosarcoma, (n = 10); epithelioid sarcoma (n = 5); clear cell sarcoma (n = 3); synovial sarcoma (n = 10); malignant peripheral nerve sheath tumour (MPNST) (n = 12); alveolar soft part sarcoma (n = 3); chordoma (n = 5); pleomorphic undifferentiated sarcoma (n = 5); perineurioma (n = 4); granular cell tumour (n = 6); acinic cell carcinoma (n = 5); adenocarcinoma, lung (n = 5), colon (n = 10), endometrioid (n = 10), prostate (n = 10) and renal cell (n = 10). Nineteen of 20 GISTs expressed DOG‐1 and 12 of 20 were diffusely positive (≥95%) with moderate to strong intensity. There was focal, predominantly luminal staining of colorectal (three of 10), endometrioid (four of 10) and acinic cell carcinomas (four of five). One case each of spindle cell/desmoplastic melanoma (2+), schwannoma (trace) and MPNST (2+) showed DOG‐1 expression. Conclusions: Our study supports that DOG‐1 is a highly sensitive and specific marker for GISTs and also highlights hitherto unrecognized and unusual patterns of expression in non‐mesenchymal neoplasms.     

Site-independent prognostic value of chromosome 9q loss in primary gastrointestinal stromal tumours

Although the significance of tumour site for estimating malignant potential in gastrointestinal stromal tumours (GISTs) has recently been recognized, site-specific genetic patterns have not to date been defined. This study examined 52 c-kit-positive primary GISTs (with a mean follow-up of 42.3 months in 51 cases) from three different locations (35 gastric, 12 small intestinal, and five colorectal) using comparative genomic hybridization (CGH). In general, tumour site correlated with key prognostic factors, including tumour size, mitotic rate, proliferative activity, and probable malignant potential. Furthermore, several DNA copy number changes showed a site-dependent pattern. These included losses at 14q (gastric 83%, intestinal 35%; p = 0.001), losses at 22q (gastric 46%, intestinal 82%; p = 0.02), losses at 1p (gastric 23%, intestinal 88%; p = 1 x 10(-5)), losses at 15q (gastric 14%, intestinal 59%; p = 0.002), losses at 9q (gastric 14%, intestinal 53%; p = 0.006), and gains at 5p (gastric 11%, intestinal 53%; p = 0.002). These data demonstrate strong site-dependent genetic heterogeneity in GISTs that may form a basis for subclassification. Prognostic evaluation of DNA copy number changes identified losses at 9q as a site-independent prognostic marker associated with shorter disease-free survival (p = 0.03) and overall survival (p = 0.002). Furthermore, 9q loss also appeared to carry prognostic value in predicting overall survival for patients with advanced or progressive GISTs (p = 0.003).     

Specificity of DOG1 (K9 clone) and protein kinase C theta (clone 27) as immunohistochemical markers of gastrointestinal stromal tumour

Wong N A C S & Shelley‐Fraser G
(2010) Histopathology 57 , 250–258 Specificity of DOG1 (K9 clone) and protein kinase C theta (clone 27) as immunohistochemical markers of gastrointestinal stromal tumour Aims: DOG1 and protein kinase C (PKC) theta are both sensitive immunohistochemical markers of gastrointestinal stromal tumour (GIST). However, there are conflicting data regarding the specificity of the most commonly used PKC theta antibody (clone 27), and there are no existing data regarding the specificity of the only known commercially available DOG1 antibody (K9 clone) at the time of writing. This study’s aim was to characterize the immunoreactivity patterns of both monoclonal antibodies amongst a wide range of neoplasm types including, in particular, histological mimics of GIST. Methods and results: Immunohistochemistry for DOG1 and PKC theta was performed on whole tissue sections from 23 different neoplasm types (total of 125 cases). Ten of these neoplasm types showed CD117 immunopositivity. Only three (Ewing’s sarcoma, glomus tumour and synovial sarcoma) of the 23 neoplasm types showed DOG1 immunopositivity, and such positivity was often focal and weak in intensity. In contrast, all but four (ganglioneuromas, leiomyomas, desmoplastic small round cell tumours and PEComa/angiomyolipomas) of the 23 neoplasm types showed PKC theta immunopositivity. Conclusions: Compared with CD117, DOG1 (using the K9 antibody) is a more specific marker, whereas PKC theta (using the clone 27 antibody) is a considerably less specific immunohistochemical marker for GIST.     

Risk assessment and pathological reporting of gastrointestinal stromal tumour

Gastrointestinal stromal tumour (GIST) is the most common mesenchymal neoplasm of the GI tract. GISTs form a biological continuum ranging from benign incidentally detected minute lesions (stromal tumourlets) of no clinical significance to large and highly malignant sarcomatous neoplasms. The disease is characterized by indistinct borders between biologically different subsets within its spectrum leading to unpredictable course of the disease in the majority of cases. Availability of effective tyrosine kinase inhibitors underlined the urgent need for a reliable risk assessment system to reliably identify those patients who are at a significant risk for disease relapse and would thus profit from currently available effective targeted molecular therapy. While several risk systems have been established for GIST over the last decade, evidence is growing that no single system is optimal for all patients. Accordingly, an individualized risk system integrating classical parameters (site, size, mitotic index), status of serosal mechanical barrier covering the tumour (presence of tumour rupture or serosal penetration) as well as several other still debatable histological (coagulative necrosis, venous invasion, mucosal infiltration), biomarkers (proliferative index, p16 expression, etc.), and molecular (specific mutation types, adverse chromosomal aberrations, etc.) prognosticators would be of superb value for selecting the most appropriate patient treatment.     

Expression of HSP90 in gastrointestinal stromal tumours and mesenchymal tumours

Kang G H, Lee E J, Jang K T, Kim K‐M, Park C K, Lee C‐S, Kang D Y, Lee S H, Sohn T S & Kim S
(2010) Histopathology 56, 694–701
Expression of HSP90 in gastrointestinal stromal tumours and mesenchymal tumours Aims: Heat shock proteins (HSP) are up‐regulated under conditions of increased stress, including cancer. Recently, HSP90 has been shown to be crucial to the expression and activation of the KIT oncoprotein. The aim was to explore the role of HSP90 expression as a prognostic marker and therapeutic target in gastrointestinal stromal tumours (GISTs) and other mesenchymal tumours. Methods and results: The expression of HSP90 was evaluated by immunohistochemistry in 92 GISTs, 47 mesenteric fibromatoses, six schwannomas, leiomyomas, melanomas, malignant peripheral nerve sheath tumours and leiomyosarcomas. Western blotting was performed in 22 selected cases. HSP90 overexpression was found in 33.7% of GISTs and was correlated with non‐gastric location, mixed histological subtype, high mitotic index, high risk grades, and specific mutation genotypes. In mesenchymal tumours, HSP90 overexpression was found in 66.7% of malignant peripheral nerve sheath tumours, 83.3% of leiomyosarcomas, and 100% of melanomas. HSP90 expression by Western blotting correlated with the results of immunohistochemistry. The Cox proportional hazards model showed that HSP90 expression is an independent predictor of recurrence in GISTs (P = 0.003). Conclusions: Overexpression of HSP90 is predictive of adverse behaviour in GISTs and may provide a therapeutic solution to the challenge of imatinib‐resistant GISTs and other mesenchymal sarcomas.     

Absence of c-kit gene mutations in gastrointestinal stromal tumours from neurofibromatosis type 1 patients

Most sporadic gastrointestinal stromal tumours (GISTs) have somatic c-kit gene mutations that are considered to be causal. Neurofibromatosis type 1 (NF1) is caused by mutations of the NF1 gene and NF1 patients have an increased risk of developing GISTs. Since most neoplasms are considered to develop as a result of the combination of several gene mutations, these findings suggest that GISTs from NF1 patients might have somatic c-kit gene mutations and that sporadic GISTs from non-NF1 patients might have somatic NF1 gene mutations. The present study analysed 29 GISTs from seven NF1 patients for c-kit gene mutations and ten sporadic GISTs from ten non-NF1 patients for NF1 mutations. Exons 9, 11, 13, and 17 of the c-kit gene were amplified and directly sequenced after the extraction of genomic DNA from wax-embedded tissues from 26 GISTs from five NF1 patients. The whole coding region of the c-kit cDNA and the whole coding region of the NF1 cDNA were amplified and directly sequenced after RNA extraction and cDNA synthesis in three fresh GIST tissues from two NF1 patients and ten fresh GIST tissues from ten non-NF1 patients. Of the ten sporadic GISTs, eight had heterozygous mutations at exon 11, and one at exon 9, of c-kit. Heterozygous NF1 gene mutations were detected in GISTs from the two NF1 patients from whom fresh tissues were available. None of the 29 GISTs derived from NF1 patients had detectable c-kit gene mutations and none of the ten GISTs derived from non-NF1 patients had detectable NF1 mutations. These results suggest that the pathogenesis of GISTs in NF1 patients is different from that in non-NF1 patients.     

Follicular dendritic cell sarcoma of the colon mimicking stromal tumour

AIMS: Follicular dendritic cell tumours are very rare neoplasms that often occur in lymph nodes. We report here a case in the colon, a hitherto unreported site, in a 37-year-old female. The differentiation from gastrointestinal stromal tumour is emphasized. METHODS and RESULTS: The tumour was tan, elastic and solid with surface ulceration. Microscopically, it was composed of oval to spindle tumour cells with syncytial cytoplasm arranged in fascicular and whorled patterns. There were many infiltrating lymphocytes. The histological appearance resembled gastrointestinal stromal tumour, thymoma or meningioma. Distinct from the stromal tumour, the lymph node was also involved by the tumour. Immunohistochemically, the tumour cells were positive for CD21, CD35 and CD68, but negative for cytokeratin, CD34, smooth muscle actin, desmin, S100 protein, epithelial membrane antigen, leukocyte common antigen, HMB-45 and c-kit. In-situ hybridization study was negative for Epstein-Barr virus RNA sequences. Ultrastructurally, the tumour cells possessed cytoplasmic processes joined by desmosomes. CONCLUSIONS: This entity should be considered in the list of differential diagnoses for gastrointestinal stromal tumour. The lymph node metastasis and immunohistochemical features are of value for identification of this rare neoplasm.     

Familial multiple gastrointestinal stromal tumours with associated abnormalities of the myenteric plexus layer and skeinoid fibres

AIMS: Multiple familial gastrointestinal stromal tumours are rare. We report the third family with two cases of multiple gastrointestinal stromal tumours showing skeinoid fibres. Associated abnormalities of the myenteric plexus layer are described and new hypotheses for the histogenesis of gastrointestinal stromal tumours are formulated. METHODS AND RESULTS: Multiple gastrointestinal stromal tumours developed in the duodenum and proximal jejunum were removed from mother and son. No history of a specific syndrome or of mastocytosis was known. Light microscopy revealed typical gastrointestinal stromal tumours with skeinoid fibres. An unusual abnormality of the myenteric plexus layer, showing a diffuse spindle cell hyperplasia, was noted in the macroscopically normal digestive wall. No abnormalities of the ganglion cells were associated. Tumours and the spindle cell hyperplasia showed similar morphological and immunohistochemical features with expression of CD34 and CD117 antigens. Follow-up revealed recurrences in the mother. CONCLUSION: The morphological characteristics of these two cases of familial gastrointestinal stromal tumours and of the associated abnormalities of the myenteric plexus layer, help to better explain the histogenesis of multiple familial gastrointestinal stromal tumours. The hyperplasia of the myenteric plexus could be considered a risk factor for recurrent tumours.     

Congenital gastrointestinal stromal tumour is morphologically indistinguishable from the adult form, but does not express CD117 and carries a favourable prognosis

AIMS: The histological and immunohistochemical features of a congenital stromal tumour of the jejunum are compared with those of adult gastrointestinal stromal tumours (GIST). The literature concerning the diagnosis and prognosis of congenital small intestinal stromal tumours is reviewed. METHODS AND RESULTS: A term female infant presented with intestinal obstruction, from birth. Histology of a 15-mm jejunal nodule showed a predominantly spindle-cell tumour with epithelioid areas. There was a low mitotic count and mild nuclear pleomorphism, extensive necrosis and haemorrhage, and focal calcification. Immunohistochemically, tumour cells stained for muscle specific actin and vimentin. Staining for CD117 (c-kit), S100, desmin and CD34 was negative. The features were compared to those of seven adult cases: no morphological feature was specific to the congenital tumour, which was smaller than the adult cases. There were no ultrastructural features specific for a particular line of differentiation. Immunohistochemical staining patterns were similar, except for CD117, which was strongly positive in all adult tumours, but negative in the congenital tumour. CONCLUSIONS: This congenital jejunal stromal tumour morphologically resembled adult GIST, but lack of c-kit expression suggests that it is nosologically distinct. Despite the presence of histological features which would cause the tumour to be categorized as malignant in an adult, it is apparent from previous reports of congenital small intestinal stromal tumours that the prognosis is favourable.     

(GT)N dinucleotide repeat polymorphism of haem oxygenase-1 promotor region is not associated with inflammatory bowel disease risk or disease course

Haem oxygenase-1 (HO-1) up-regulation was suggested to reduce mucosal tissue damage in inflammatory bowel disease (IBD) and an up-regulation of HO-1 expression in patients with Crohn's disease (CD) and ulcerative colitis (UC) was demonstrated. A HO-1 gene promoter microsatellite (GT)(n) dinucleotide repeat polymorphism was associated with regulation of HO-1 in response to inflammatory stimuli. We therefore hypothesized that IBD patients might segregate into phenotypes with high or low HO-1 inducibility. Ethylenediamine tetraacetic acid blood samples were obtained from 179 CD patients, 110 UC patients and 56 control patients without inflammation. Genomic DNA was purified and the 5'-flanking region of the HO-1 gene containing the (GT)(n) dinucleotide repeat was amplified. Polymerase chain reaction (PCR) products were purified and the length of the PCR fragments was analysed. The number of (GT)(n) repeats in the population studied ranged from 13 to 42. The distribution of the allele frequencies was comparable in patients and controls for both the short and the long alleles. The frequencies of short-, middle- and long-sized alleles were not changed among the groups studied. No correlation was found between IBD and microsatellite instability detected in five individals. Our data indicate that (GT)(n) dinucleotide repeats of the HO-1 promotor region have no significance for the pathophysiology and disease course of IBD.     

KIT and PDGFRalpha mutational analyses of mixed cell-type gastrointestinal stromal tumours

AIMS: To determine whether the epithelioid and spindle components of a mixed cell-type gastrointestinal stromal tumour (GIST) show the same receptor tyrosine kinase mutation and, by inference, the same sensitivity to imatinib. METHODS AND RESULTS: Six mixed cell-type GISTs were identified from 108 gastric GISTs. Clinicopathological and immunohistochemical data of the six neoplasms were collated. For each neoplasm, DNA was extracted separately from the laser-microdissected epithelioid and spindle components and non-neoplastic tissue and sequenced for KIT and platelet-derived growth factor receptor (PDGFR)alpha mutations. The epithelioid component often showed less CD117 and/or CD34 immunoreactivity than the spindle component of the same mixed cell-type GIST. Four mixed cell-type GISTs showed somatic KIT mutations (deletions in exon 11 in three tumours and an insertion in exon 9 in one tumour) and one showed a somatic PDGFRalpha mutation (point mutation in exon 18); in each of the five cases, both epithelioid and spindle components showed identical mutations. CONCLUSIONS: The presence of the same receptor tyrosine kinase mutation in both components of a mixed cell-type GIST suggests that both components should be equally responsive to imatinib treatment, and that such mutation is an early key event in the pathogenesis of these neoplasms.     

Prognostic significance of p53 gene mutations and protein overexpression in localized gastrointestinal stromal tumours

AIMS: Mutation of c-kit is a relatively early event in the tumorigenesis of gastrointestinal stromal tumours (GISTs). The aim was to determine the prognostic significance of p53 alterations as an additional genetic change in GISTs. METHODS AND RESULTS: We reviewed 125 patients with localized GISTs subjected to complete resection between 1990 and 2002. Mutational analyses of c-kit exons 9, 11, 13 and 17, p53 exons 4-8 and immunohistochemistry for p53 protein were conducted using paraffin-embedded tissues. Alterations of p53 were observed in 50 patients (40.0%). Based on the National Institutes of Health's risk category, p53 alterations were noted more frequently in the higher risk categories (P = 0.041). With a median follow-up of 56.5 months (range: 2.3-126.8), 5-year relapse-free survival (RFS) rates were 61.7% without p53 alterations, compared with only 40.2% with p53 alterations (P = 0.009). Multivariate analysis indicated that p53 alterations comprised an independent, poor prognostic factor for RFS, in addition to c-kit mutations, large size, a high mitotic count and non-gastric primary sites. CONCLUSIONS: Alterations in p53 were more commonly observed in localized GISTs at higher risk of relapse. This suggests that they are significant as an independent, poor prognostic factor.     

The expression of p53 and bcl-2 in gastrointestinal stromal tumours is associated with anatomical site, and p53 expression is associated with grade and clinical outcome

AIMS: To compare the expression of p53 and bcl-2 in gastrointestinal stromal tumours (GISTs) with anatomical site, National Institutes of Health (NIH) risk category (grade), pathological features, and clinical outcome. METHODS AND RESULTS: The immunohistochemical expression of p53 and bcl-2 in 105 GISTs (71 gastric, 20 small intestinal, four colonic, 10 rectal) was recorded. When all GISTs were assessed, there was p53 positivity in 28% and bcl-2 positivity in 77%. Gastric tumours had a lower prevalence of p53 positivity (20%) than intestinal (40-50%). Rectal GISTs had the lowest prevalence of bcl-2 positivity (20%) and gastric and small intestinal the highest (80% and 90%, respectively). In GISTs from all sites, p53 positivity was associated with size > 50 mm, epithelioid cell shape, nuclear atypia, mucosal invasion, and mitotic count > 5/50 high-power fields. In gastric GISTs the associations were the same, apart from size and mitotic count. In GISTs from all sites and in gastric GISTs, p53 expression correlated with NIH risk category. When GISTs from all sites were subjected to univariate survival analysis, an adverse outcome was associated with p53 positivity, NIH risk category, and several established prognostic factors. When gastric GISTs were assessed, the associations were similar although size was not prognostic. In multivariate survival analysis, p53 expression was independently prognostic for gastric GISTs in some models, while it was never independently prognostic for GISTs from all sites. Whether all GISTs or gastric GISTs were assessed, bcl-2 showed no association with clinical outcome or risk category. CONCLUSIONS: Anatomical site influences p53 and bcl-2 expression in GISTs. p53 expression is associated with NIH risk category, various pathological features, and clinical outcome, and may be independently prognostic for gastric GISTs. Bcl-2 expression has no prognostic value.     

Common genetic changes in leiomyosarcoma and gastrointestinal stromal tumour: implication for ataxia telangiectasia mutated involvement

Gastrointestinal stromal tumours (GISTs) are the most common mesenchymal tumours of the gastrointestinal tract. Formerly GISTs were commonly classified histologically as leiomyosarcomas; however, they are now known to arise from the interstitial cells of Cajal. Majority of GISTs overexpress KIT and have characteristic mutations within the gene, which are the targets of drug treatment with tyrosine kinase inhibitors. Leiomyosarcoma is a malignant tumour of smooth muscle differentiation and falls into a group of sarcomas that show complex karyotypic changes with no consistent recurrent genetic abnormality. We have used comparative genomic hybridization in combination with fluorescence in situ hybridization to determine genetic differences between the tumour types. We found leiomyosarcomas and GISTs share common regions of chromosomal 13q and 11q imbalance, in addition to more specific 1p and 8p losses in leiomyosarcoma and 15q and 22q losses in GISTs. More importantly, we have shown for the first time a deletion in the ataxia telangiectasia mutated (ATM) gene locus with decreased/absent expression of ATM protein, and amplification in the region 13q21–q32 in both tumour types, suggesting both regions may play a role in leiomyosarcoma and GIST biology.