Use of tissue microarrays and immunohistochemistry to standardize the diagnosis of gastrointestinal stromal tumors

We assessed the concordance among seven general pathologists with respect to histologic diagnosis and interpretation of c-kit proto-oncogene (KIT) and platelet-derived growth factor receptor alpha (PDGFRA) immunostaining of 36 cases of primary spindle-cell tumor, predominantly of the gastrointestinal tract, mesentery, and retroperitoneum, based on review of a tissue microarray (TMA) subjected to immunohistochemistry with antibodies to KIT/CD117, PDGFRA, vimentin, desmin, smooth muscle action, CD34, and S-100 protein. Tumors included 20 molecularly analyzed gastrointestinal stromal tumors (GISTs), 4 leiomyosarcomas, 4 schwannomas, 4 desmoid-type fibromatoses, and 4 solitary fibrous tumors. The mean overall concordance with original diagnosis for each histologic type was 91.1%, with a mean kappa value of 0.91. With respect to PDGFRA immunostaining, the four GISTs with PDGFRA mutation were interpreted as cytoplasm positive, but the 16 GISTs with c-kit mutation were interpreted as weak or positive. These results indicate that the overall concordance with original diagnosis in mesenchymal tumors with the use of immunohistochemical panels is high, despite the use of TMAs. To some extent, PDGFRA immunophenotyping may be useful in GISTs with PDGFRA mutation, but it was not highly reproducible or specific. Therefore, in KIT-negative or weakly positive GISTs, mutation analysis will be required.     

The role of KIT in the management of patients with gastrointestinal stromal tumors

In recent years, immunohistochemical staining for KIT (CD117) has become integral to the diagnosis of gastrointestinal stromal tumors (GISTs), nearly 90% of which harbor activating mutations in the KIT receptor tyrosine kinase gene. Approximately 80% of patients with metastatic GIST show at least some clinical response to the targeted small molecule KIT inhibitor imatinib. The response to imatinib is closely correlated with the presence and type of KIT mutation. GISTs with the most common KIT exon 11 mutations have the highest response rate by far, whereas GISTs lacking mutations in KIT or the alternative receptor tyrosine kinase PDGFRA show much lower rates of response to imatinib. Less than 5% of GISTs are KIT-immunonegative; and many of these tumors have activating mutations of PDGFRA, some of which are also inhibited by imatinib. Most patients who initially respond to imatinib become resistant and eventually progress, which coincides with the selection of imatinib-resistant secondary KIT mutations in the kinase domain. Sunitinib has recently been approved for patients with GIST, principally those who fail imatinib therapy; and additional small molecule inhibitors are in the pipeline. It is becoming evident that alternative approaches to direct KIT inhibition will be required for long-term survival of patients with advanced GISTs. This review examines the role of KIT in the diagnosis and management of patients with GIST.     

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.     

Expression of stem cell factor (SCF), a KIT ligand, in gastrointestinal stromal tumors (GISTs): a potential marker for tumor proliferation

Gastrointestinal stromal tumors (GISTs) show a high incidence of gain-of-function mutations of the c-kit gene, which encodes a receptor tyrosine kinase KIT. This mutation is seen independently of metastasis and/or recurrence of tumors; thus, the factors involved in tumor proliferation rate and malignancy are still not known. Some mesenchymal and epithelial tumors have been reported to co-express KIT and its ligand, stem cell factor (SCF), for autonomous proliferation by the autocrine mechanism. The purpose of this study is to clarify whether GIST cells produce SCF, despite mutated KIT with constitutive activation. Immunohistochemically, we examined the co-expression of KIT and SCF in 36 GIST cases. All cases were found to be KIT-positive, and of these, 21 cases, including four recurrent or metastatic GISTs, showed co-expression of SCF. MIB-1 labeling index was significantly higher, and the average tumor size was larger in SCF-positive cases. By confocal microscopy, KIT was expressed on the cellular membrane, around which SCF was distributed less densely. Western blot analysis revealed that the membrane-bound SCF of 31 kDa was found to be approximately 10 times more abundant than the soluble SCF of 18.5 kDa, suggesting continuous KIT activation. These results indicate that proliferation of GIST cells can be caused not only by the gain-of-function mutation of c-kit, but also by the autocrine mechanism of the SCF/KIT system. Thus, SCF expression would be a useful marker for tumor proliferation.     

PDGFR expression in differential diagnosis between KIT-negative gastrointestinal stromal tumours and other primary soft-tissue tumours of the gastrointestinal tract

AIMS: To investigate the value of platelet-derived growth factor receptors (PDGFRs) by immunohistochemistry in discriminating KIT-negative gastrointestinal stromal tumours (GISTs) from other soft-tissue neoplasms of the digestive tract. METHODS AND RESULTS: One-hundred and sixty-seven primary gastrointestinal mesenchymal tumours (125 GISTs, 15 intra-abdominal desmoids, 12 leiomyomas, eight leiomyosarcomas, three schwannomas, two solitary fibrous tumours, and one case each of inflammatory pseudotumour and fibroid polyp) were reclassified based on morphology and on the immunohistochemical panel recommended by the National Institutes of Health consensus on GIST. All cases were then tested with antibodies specific for PDGFR alpha and beta. Of 125 GISTs, 117 were KIT-positive (93.6%) and eight KIT-negative (6.4%). All the KIT-positive GISTs were negative for both PDGFRs, while all the eight KIT-negative GISTs expressed PDGFR-alpha, with two of them also coexpressing PDGFR-beta. Among the 42 non-GIST tumours, only a small percentage (26.6%) of desmoids immunostained for PDGFR-alpha, two of them coexpressing PDGFR-beta. CONCLUSIONS: Immunostaining with PDGFR-alpha is a helpful marker in discriminating between KIT-negative GISTs and other gastrointestinal mesenchymal lesions: all KIT-negative GISTs were positive for PDFGR-alpha, while none of the other gastrointestinal mesenchymal tumours analysed, except a small subset of desmoids, was reactive with anti-PDGFRs. These preliminary data demonstrate the suitability of commercially available antibodies to detect immunohistochemically the mutually exclusive expression of KIT and PDGFR-alpha previously reported in GISTs by molecular biological techniques. Since PDGFR exists in the form of a homodimer (alphaalpha, betabeta) or heterodimer (alphabeta) and two of the KIT-negative GISTs coexpressed both PDGFR isoforms, further investigations are required to elucidate the role of PDGFR-beta in GISTs.     

Cytologic diagnosis of gastrointestinal stromal tumors of the stomach by endoscopic ultrasound-guided fine-needle aspiration biopsy: cytomorphologic and immunohistochemical study of 12 cases.

Gastrointestinal stromal tumor (GIST) is an uncommon tumor, which was usually diagnosed by endoscopic biopsy or surgical resection. This study evaluated the efficacy and accuracy of endoscopic ultrasound (EUS) -guided fine-needle aspiration (FNA) biopsy in the diagnosis of GIST and reported its cytomorphologic features. Twelve patients with gastric GIST were diagnosed through EUS-guided FNA. Immediate on-site evaluation and cytologic diagnoses were given in nine cases (75.0%) with an average of three passes. Cell blocks provided diagnostic material in three cases (25.0%). Spindle cells were present in the cytologic material in all cases. Two patients had subsequent surgical resections. Immunohistochemical (IHC) studies performed in cell blocks and two surgical specimens all supported the original diagnoses. In the two cases with surgical resections, IHC results in cell blocks were similar to that in the resected specimens. This study demonstrated that when combining smears and cell blocks, EUS-guided FNA is accurate and efficient in the diagnosis of GIST. IHC reactivity in cell blocks correlated with that of the main tumors.     

Comparative ultrastructural analysis and KIT/PDGFRA genotype in 125 gastrointestinal stromal tumors

GISTs are the most common mesenchymal neoplasms of the digestive tract and are thought to originate from or differentiate toward the interstitial cell of Cajal lineage. Almost all GISTs express KIT protein and the majority show activating mutations in either KIT or PDGFRA proto-oncogenes. Ultrastructurally, these tumors have been shown to have either a smooth muscle, neuronal, dual, or null phenotype. The objective of this study was to investigate the relationship between ultrastructural features and genotype in a large series of 125 histologically confirmed and CD117 positive GISTs. PCR analysis for the presence of KIT exon 9, 11, 13, and 17 and PDGFRA exon 12 and 18 mutations was performed. There were 62 (50%) tumors located in the stomach and 45 (36%) in the small bowel. Overall, KIT mutations were detected in 93 (75%) patients: 86 (69%) in exon 11, and 7 (6%) in exon 9. A PDGFRA mutation was detected in 7 (6%) cases and 25 (19%) cases had no mutation. Ultrastructurally, skeinoid fibers were seen in 55 (44%) cases and were more common in small bowel than stomach GISTs, and occurred in only in 1 of 16 patients with an ITD (KIT) exon 11 or PDGFRA mutation. Focal actin microfilaments were identified in 82 (65%) cases and did not correlate with location or mutation type. Rare neurosecretory-type granules (NS-G) were seen in 34 (27%) of cases, but were seen in most of the cells in only 5 (4%) cases. GISTs showing both NS-G and microtubules were associated with KIT exon 11 genotype and spindle cell morphology. PDGFRA mutated cases were associated with gastric location, predominantly epithelioid morphology and lacked NS-G.     

Altered expression of cell cycle regulatory proteins in gastrointestinal stromal tumors: markers with potential prognostic implications

Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal neoplasms of the digestive tract. The prediction of the malignant potential of GISTs is still difficult. Altered cell cycle regulation may underlie the tumorigenesis and/or the progression of human malignancies. Although p53 and Bcl-2 have been extensively investigated in GISTs, little is known about the frequency of expression and possible clinical implications of alterations of other cell cycle regulatory proteins in these neoplasms. We have previously investigated the role of loss of p16(INK4A) by loss of heterozygosity and immunohistochemistry in the progression of GISTs and found that loss of heterozygosity of 9p and loss of p16 expression are confined to malignant GISTs. This has led us to investigate the role of other cell cycle regulatory proteins in these tumors. Twenty-three cases of GIST (9 low malignant potential [LMP], 10 primary malignant, and 4 intra-abdominal recurrences) were examined. All cases were strongly positive for KIT (CD117). Immunohistochemical stains were carried out on tissue microarrays to evaluate the expression of proteins involved in the G(1)-S transition and proteins that regulate apoptosis including Rb, E2F1, cyclin D1, CDK4, CDK6, p27(KIP1), p21(WAF1/CIP1), p53, Mdm2, Bcl-2, and Bax. The positive phenotypes identified were as follows: Rb, 39.1%; E2F1, 69.6%; cyclin D1, 30.4%; CDK4, 100%; CDK6, 30.4%; 39.1%; p27(KIP1), 47.8%; p21(WAF1/CIP1), 39.1%; p53, 43.5%; Mdm2, 17.4%; Bcl-2, 91.3%; and Bax, 100%. Malignant GISTs are more likely to be associated with a positive E2F1 and p53 phenotype and a negative p16 and p27(KIP1) phenotype. It was concluded that aberration of the cell cycle regulators is a frequent finding and may be a contributing factor to the pathogenesis of GISTs. While some alterations are seen in LMP and malignant GISTs and therefore may represent an early event in molecular tumorigenesis of GISTs, other alterations are more common in malignant GISTs than LMP and therefore have potential utility as complementary tools for the prognostication of GISTs.     

An immunohistochemical and clinicopathological study of gastrointestinal stromal tumors

Immunohistochemical and clinicopathological features of 58 gastrointestinal stromal tumors (GIST) were studied. One occurred in the esophagus, 41 in the stomach, nine in the small intestine, and seven in the large intestine. By using indirect immunoperoxidase staining for Cajal cell markers (c-kit protein and CD34), smooth muscle markers (alpha-smooth muscle actin, desmin, heavy caldesmon and calponin) and Schwann cell markers (S-100 protein and Leu 7), GIST were classified into five groups, such as Cajal cell type (n = 9), myogenic type (n = 5), Schwann cell type (n = 2), mixed cell type (n = 40) and undifferentiated type (n = 2). c-kit Protein (42/58; 72%) and CD34 (45/58; 78%) were commonly and diffusely expressed in GIST. Novel smooth muscle markers, caldesmon (29/58; 50%) and calponin (18/58; 31%), were useful in detecting myogenic characters of GIST. S-100 protein was expressed in 16 (28%) tumors, two of which were also reactive with Leu 7 (CD57). Three small bowel tumors with skeinoid fibers expressed the Cajal cell markers, and were categorizable in GIST. Clinicopathological analyses using aggressive (n = 21) and non-aggressive (n = 21) GIST indicated that the malignant potential was correlated with the intestinal location, large tumor size, high cellularity, necrosis, solid (non-interlacing bundled) pattern of growth, negativity of c-kit protein and/or CD34, high mitotic count, and high MIB-1 labeling.     

Significance of D2-40 expression in the diagnosis of gastrointestinal stromal tumor

We incidentally found the expression of D2-40 in Cajal interstitial cells for the immunohistochemical valuation of lymphatic invasion by gastric and colonic cancer cells. Therefore, we suggested that D2-40 might become an available marker for gastrointestinal stromal tumors (GISTs). In this study, we performed immunohistochemical study of 21 GISTs, As a result, we confirmed a positive reaction for D2-40 in sixteen GISTs. Among sixteen tumors, thirteen tumors were focally positive and remaining three tumors were diffusely positive. We could find no relationship between positive findings for D2-40 and size or location of the tumor. Three of four tumors with intermediate risk of aggressive behavior and two of four tumors with recurrence showed a negative reaction for D2-40. In conclusion, we suggest that D2-40 may be an available marker of GISTs, but the expression of D2-40 seems to be not associated with the degree of risk of aggressive behavior.     

New issues in gastrointestinal stromal tumors of the stomach

Gastrointestinal stromal tumors (GISTs) constitute the majority of mesenchymal tumors of the gastrointestinal tract. The activating mutations in the receptor tyrosine kinases KIT and PDFRA are key molecular changes in the pathogenesis of these tumors and their recognition has led to the development of targeted therapies. Approximately 10% of GISTs, referred to as wild-type, lack such mutations and respond poorly to treatment with tyrosine kinase inhibitors. These GISTs are almost exclusive to the stomach and include primarily tumors that occur in children and tumors that are part of several tumor syndromes. Recent studies have shown that most wild-type GISTs are succinate dehydrogenase deficient. This review summarizes current advances in the molecular biology of GISTs and discusses the clinical and pathologic features associated with different genotypes.     

胃肠道间质瘤恶性指标的研究

目的 探讨一组临床病理指标在预测恶性胃肠道间质瘤(GIST)中的价值.方法 回顾性研究840例GIST,假设肉眼播散[包括肝转移和(或)腹腔播散]、显微镜下播散(包括淋巴结转移、血管浸润、脂肪浸润、神经浸润和黏膜浸润)以及肿瘤复发为恶性组,剩余为生物学行为不明组,根据此初步分组后进行多因素分析.结果 多因素分析发现以下形态学指标与恶性组密切相关:核分裂象≥10个/50 HPF、肌层浸润、肿瘤性坏死、围绕血管呈簇状排列以及细胞明显异型(均P<0.01).由此获得了2项肉眼播散、5项镜下播散和5项组织形态学指标.485例具有其中任意一项指标者归入恶性组,其余355例为非恶性组.随访资料提示非恶性组患者5年无瘤生存率(DFS)及5年总生存率(OS)分别为99.3%和100%,而恶性组患者5年DFS及5年OS分别为43.9%和59.7%,恶性组与非恶性组5年DFS与5年OS差异有统计学意义(均为P<0.01).结论 此组12项临床及形态学指标能有效地将恶性GIST从非恶性GIST中区分开来,可用于区分GIST的性质.     

Gastrointestinal stromal tumors and KIT-positive mesenchymal cells in the omentum

Gastrointestinal stromal tumor (GIST) is currently considered to be derived from the interstitial cells of Cajal (ICC). To test the hypothesis that omental mesenchymal tumor is also a type of GIST, we evaluated the expression of specific molecules in GIST, and c-kit gene mutation in omental mesenchymal tumors, and we identified a possible counterpart of ICC in the omentum. Immunohistochemically, all of the omental mesenchymal tumors (n = 5) were positive for both KIT and CD34, and three of the five tumors were also positive for an embryonic form of smooth-muscle myosin heavy chain (SMemb). Polymerase chain reaction-single-strand conformational polymorphism analysis (PCR-SSCP) and direct sequencing revealed mutations in c-kit gene exon 11 in all five tumors. As for the ICC counterparts in the omentum, there were some KIT-positive mesenchymal cells resembling ICC at the surface of the omentum. Double fluorescence immunostaining, using anti-KIT polyclonal antibodies and monoclonal antibodies against other molecules, demonstrated that KIT-, CD34- and SMemb-positive cells were present just beneath the mesothelial cells of the omentum. These results show that omental mesenchymal tumor corresponds to GIST of the omentum, and that KIT-positive bipolar mesenchymal cells may be a counterpart of ICC in the gastrointestinal tract. Identification of a new type of KIT-positive mesenchymal cell in the omentum may lead to the discovery of a new physiological role for this organ.     

Cyclooxygenase-2 expression in stromal tumors of the gastrointestinal tract

Cyclooxygenase (COX) is a key enzyme in the conversion of arachidonic acid to prostaglandins, prostacyclin, and thromboxane. COX-2 is expressed in many epithelial malignancies, particularly those of the gastrointestinal (GI) tract. COX-2 has been implicated in the pathogenesis of cancers and has a significant negative effect on survival. To date, little is known about the expression of COX-2 in nonepithelial tumors. The objective of this study was to evaluate the expression of COX-2 in GI stromal tumors (GISTs). We evaluated 15 GISTs using tissue microarray. Tissue blocks were retrieved and stained with hematoxylin and eosin to evaluate the histological tumor type. In addition, immunohistochemistry was performed for COX-2, the macrophage marker, CD68 (KP-1), and KIT (CD117). Two pathologists then evaluated the tissues to determine the extent and intensity of COX-2 expression. The location of CD68-positive cells, and whether these cells were COX-2 positive, was also evaluated. The results showed that 80% (12 of 15) of the tumors expressed COX-2. Expression was noted in the cytoplasm of the tumor cells, with variable intensity of staining among the tumors. COX-2 was expressed in both epithelial cell and spindle cell tumors, but appeared stronger in epithelial lesions. In mixed lesions, COX-2 was expressed to a greater extent in epithelial areas. There was a greater extent of COX-2 expression in malignant tumors and tumors located within the stomach. Tumor-infiltrating macrophages (CD68-positive cells) were identified in all of the lesions; in 80% of cases, those macrophages also expressed COX-2. This study is the first to demonstrate COX-2 expression in stromal lesions of the GI tract. The enzyme may play a role in the proliferation of these lesions, suggesting the potential use of nonsteroidal anti-inflammatory drugs in treatment.     

Frequency, phenotype, and genotype of minute gastrointestinal stromal tumors in the stomach: an autopsy study

Gastrointestinal stromal tumors are the most common mesenchymal tumors of the human digestive tract. Up to 85% of these tumors show somatic gain-of-function mutation of the receptor tyrosine kinase c-KIT gene. A recent study has shown a high frequency (22.5%) of minute gastrointestinal stromal tumors in stomachs examined during routine autopsies. The aims of our study were to confirm the previously reported incidence of gastric gastrointestinal stromal tumors in routine autopsies and to investigate their molecular alterations. Gastrointestinal stromal tumors were collected prospectively from 578 autopsies over an 18-month period. After recording the size and location of each lesion, representative tissue samples were processed for hematoxylin and eosin staining and immunohistochemically stained for CD117 and CD34. Microdissected DNA from all identified gastrointestinal stromal tumors was studied for c-KIT and platelet-derived growth factor receptor α mutations. We identified 17 gastrointestinal stromal tumors in 578 consecutive autopsies (2.9%) located in the gastric body (47%) and fundus (47%). One tumor location was not recorded. All tumors were immunohistochemically positive for CD117 and CD34. DNA analysis showed c-KIT mutations in 11 cases. One platelet-derived growth factor receptor α mutation was found. The incidence of gastric minute gastrointestinal stromal tumors (2.9%) is higher than the reported clinical incidence. All are benign tumors, and most, including minute tumors, contain c-KIT mutations. This finding highlights the fact that c-KIT mutations are an early event in the evolution of gastrointestinal stromal tumors but are not sufficient per se for clinically relevant disease.     

Knock-in murine models of familial gastrointestinal stromal tumours

In recent years, gastrointestinal stromal tumour (GIST) has emerged from a poorly understood group of mesenchymal tumours to a distinct pathological entity that has become a leading model for new therapies targeting kinases. GIST pathogenesis is driven by receptor tyrosine kinase-activating mutations most often in KIT or PDGFRA that may be sporadic or result in familial GIST syndromes. In a recent issue of The Journal of Pathology (J Pathol 2008;214:302-311), Nakai and colleagues report that mutation of the previously un-modelled tyrosine kinase II domain of KIT generates a spectrum of features very similar to that in two human GIST families. Knock-in mouse models of GIST may prove extremely useful in pre-clinical evaluation of kinase-targeted compounds and of the mechanism of drug resistance, but intriguingly it now seems clearer that the distribution of GIST in mouse models and humans is different.     

Collision tumor of the esophagus: report of a case with mixed squamous cell carcinoma and gastrointestinal stromal tumor

Esophageal cancer is mainly divided into squamous cell carcinoma and adenocarcinoma. Epidemiologically, the former contributes to 90% of worldwide esophageal cancer cases, while adenocarcinoma contributes to two-thirds of cases in developed countries. Although other rare types and collision with multiple histological types of tumors do occur in the esophagus, it is very rare for a gastrointestinal stromal tumor (GIST) to collide with an epithelial malignant tumor. To date, only three cases have been reported in the literature. The current study reported a 69-year-old male patient with squamous cell carcinoma and GIST in the middle esophagus. There was no merging of tissue components between these tumors. This study together with a literature review indicates that esophageal collision tumors have been increasingly reported in recent years. Histology and immunohistochemistry are needed to make a differential diagnosis. The exact oncogenic mechanism or the interaction of two independent neoplasms still remains to be determined, and further investigation, such as electron microscopy and genetic analysis, may help to elucidate the pathogenesis of the colliding tumors.