Primary esophageal succinate dehydrogenase deficient epithelioid gastrointestinal stromal tumor: A diagnostic pitfall of poorly differentiated carcinoma

AIM: This study investigated the immune cells within in the tumor stroma of primary and metastatic gastrointestinal stromal tumors (GISTs) by immunohistochemistry. GISTs appear usually as histological homogeneous tumors with a dense cell structure. However, GISTs have been shown to have a tumor stroma, in which different immune cells of the innate and acquired immune system are embedded. The frequency and function of these cells in GISTs remain widely unknown. In this work the various immune cells are described morphologically and their frequency in primary GISTs as well as peritoneal and liver metastases is being compared. Furthermore, the criteria proliferation activity, morphology, tumor size, organ localization and malignancy are examined and compared with the number of immune cells in the primary and metastatic GISTs. In a small number of metastases real time RT-PCR was performed to gain an overview of the functionality of immune cells using RT-PCR.MATERIAL & METHODS: A total of 188 untreated primary GISTs and 52 untreated GIST metastases were immunohistochemically processed to determine the percentage of tissue-associated fibrohistiocytic cells (Kim-1P) and cells of macrophage lineage (CD68). In addition, the number of several lymphocytes (T-Lymphocytes: CD3, B-Lymphocytes: CD20, NK-cells: CD56) was studied in these samples. For this purpose punches of the paraffin blocks of resected specimen or biopsies were lined up site by site in 22 tissue microarryas (TMAs) using a manual tissue arrayer. The photographs of 3-6 punches have been analysed computer-assisted with subsequent statistical evaluation of the results. This analysis resulted in very precise numbers of the percentage of immune cells in the respective tissue preparations. Furthermore, reverse transcription and real-time RT-PCR were used to detect the expression of the proinflammatory cytokines interleukin 1β, interleukin 6 and tumor necrosis factor α in a small sample size of snap-frozen tissue samples of metastases.RESULTS: GISTs have a tumor stroma, in which many immune cells are embedded. In the primary GISTs Kim-1P+ cells (28,8% ±7,1) dominated, but also lymphocytic cells (CD3+(2,2% ±1,8), CD20+(0,6% ±0,7), CD56+(1,1% ±0,9)) and macrophages (CD68+(3,6% ±2,1)) were present. Interestingly, the number of immune cells differ in the metastases of GISTs. On the one hand metastases show more lymphocytic cells than the primary GISTs (CD3+: 7,3% ±2,3 (p<0,01); CD20+: 1,8% ±0,3 (p<0,05)), on the other hand, the peritoneal metastases differ from the liver metastases. Peritoneal metastases have a significant higher number of Kim-1P+ cells (31,8% ±7,5) than liver metastases (18,2% ±3,8; p<0,01), while the latter in turn contain more CD3+ T-lymphocytes (11,7% ±1,8) than the peritoneal metastases (4,4% ±2,6; p<0,01). At the same time, the two sites differ in proliferation activity. Liver metastases have a lower proliferation activity (12,9% ±8,2) than peritoneal metastases of GISTs (18,3% ±7,3; p<0,05). In the RT-PCR of fresh-frozen tissue from metastases of a small cohort, one patient with a high percentage of Kim 1P+ cells (49,5% ±17) did not only show a comparatively high expression of Il-6 (CT value 26,8) and Il-1β (CT value 26,1) but also the lowest expression of TNF (CT 34,1) as well as a rapid clinical progress.CONCLUSION: The different numbers of the various immune cells in primary GISTs compared with metastatic GISTs as well as within the two metastatic sites, suggest a location specific microenvironment, which may play a part in the tumor growth of primary and secondary GISTs. Further studies will be needed to understand the type of communication between the immune cells of the GIST stroma and the tumor itself. In the future, the individuality of this tumor will come to the fore, elucidating that tumors with a different composition of their stromal cells, for example immune cells, might also need a different and individual treatment. Thus, not only the description of the immune cells is necessary, but also the understanding of GISTs as a family of tumors, which should not be seen as uniform, homogeneous tumors.

AIM: This study investigated the immune cells within in the tumor stroma of primary and metastatic gastrointestinal stromal tumors (GISTs) by immunohistochemistry. GISTs appear usually as histological homogeneous tumors with a dense cell structure. However, GISTs have been shown to have a tumor stroma, in which different immune cells of the innate and acquired immune system are embedded. The frequency and function of these cells in GISTs remain widely unknown. In this work the various immune cells are described morphologically and their frequency in primary GISTs as well as peritoneal and liver metastases is being compared. Furthermore, the criteria proliferation activity, morphology, tumor size, organ localization and malignancy are examined and compared with the number of immune cells in the primary and metastatic GISTs. In a small number of metastases real time RT-PCR was performed to gain an overview of the functionality of immune cells using RT-PCR.MATERIAL & METHODS: A total of 188 untreated primary GISTs and 52 untreated GIST metastases were immunohistochemically processed to determine the percentage of tissue-associated fibrohistiocytic cells (Kim-1P) and cells of macrophage lineage (CD68). In addition, the number of several lymphocytes (T-Lymphocytes: CD3, B-Lymphocytes: CD20, NK-cells: CD56) was studied in these samples. For this purpose punches of the paraffin blocks of resected specimen or biopsies were lined up site by site in 22 tissue microarryas (TMAs) using a manual tissue arrayer. The photographs of 3-6 punches have been analysed computer-assisted with subsequent statistical evaluation of the results. This analysis resulted in very precise numbers of the percentage of immune cells in the respective tissue preparations. Furthermore, reverse transcription and real-time RT-PCR were used to detect the expression of the proinflammatory cytokines interleukin 1β, interleukin 6 and tumor necrosis factor α in a small sample size of snap-frozen tissue samples of metastases.RESULTS: GISTs have a tumor stroma, in which many immune cells are embedded. In the primary GISTs Kim-1P+ cells (28,8% ±7,1) dominated, but also lymphocytic cells (CD3+(2,2% ±1,8), CD20+(0,6% ±0,7), CD56+(1,1% ±0,9)) and macrophages (CD68+(3,6% ±2,1)) were present. Interestingly, the number of immune cells differ in the metastases of GISTs. On the one hand metastases show more lymphocytic cells than the primary GISTs (CD3+: 7,3% ±2,3 (p<0,01); CD20+: 1,8% ±0,3 (p<0,05)), on the other hand, the peritoneal metastases differ from the liver metastases. Peritoneal metastases have a significant higher number of Kim-1P+ cells (31,8% ±7,5) than liver metastases (18,2% ±3,8; p<0,01), while the latter in turn contain more CD3+ T-lymphocytes (11,7% ±1,8) than the peritoneal metastases (4,4% ±2,6; p<0,01). At the same time, the two sites differ in proliferation activity. Liver metastases have a lower proliferation activity (12,9% ±8,2) than peritoneal metastases of GISTs (18,3% ±7,3; p<0,05). In the RT-PCR of fresh-frozen tissue from metastases of a small cohort, one patient with a high percentage of Kim 1P+ cells (49,5% ±17) did not only show a comparatively high expression of Il-6 (CT value 26,8) and Il-1β (CT value 26,1) but also the lowest expression of TNF (CT 34,1) as well as a rapid clinical progress.CONCLUSION: The different numbers of the various immune cells in primary GISTs compared with metastatic GISTs as well as within the two metastatic sites, suggest a location specific microenvironment, which may play a part in the tumor growth of primary and secondary GISTs. Further studies will be needed to understand the type of communication between the immune cells of the GIST stroma and the tumor itself. In the future, the individuality of this tumor will come to the fore, elucidating that tumors with a different composition of their stromal cells, for example immune cells, might also need a different and individual treatment. Thus, not only the description of the immune cells is necessary, but also the understanding of GISTs as a family of tumors, which should not be seen as uniform, homogeneous tumors.

Objective: To analyze the clinicopathological features and gene mutations of primary gastrointestinal stromal tumors (GISTs) of the stomach and intestine and the prognosis of intermediate- and high-risk GISTs. Methods: This was a retrospective cohort study. Data of patients with GISTs admitted to Tianjin Medical University Cancer Institute and Hospital from January 2011 to December 2019 were collected retrospectively. Patients with primary gastric or intestinal disease who had undergone endoscopic or surgical resection of the primary lesion and were confirmed pathologically as GIST were included. Patients treated with targeted therapy preoperatively were excluded. The above criteria were met by 1061 patients with primary GISTs, 794 of whom had gastric GISTs and 267 intestinal GISTs. Genetic testing had been performed in 360 of these patients since implementation of Sanger sequencing in our hospital in October 2014. Gene mutations in KIT exons 9, 11, 13, and 17 and PDGFRA exons 12 and 18 were detected by Sanger sequencing. The factors investigated in this study included: (1) clinicopathological data, such as sex, age, primary tumor location, maximum tumor diameter, histological type, mitotic index (/5 mm2), and risk classification; (2) gene mutation; (3) follow-up, survival, and postoperative treatment; and (4) prognostic factors of progression-free survival (PFS) and overall survival (OS) for intermediate- and high-risk GIST. Results: (1) Clinicopathological features: The median ages of patients with primary gastric and intestinal GIST were 61 (8-85) years and 60 (26-80) years, respectively; The median maximum tumor diameters were 4.0 (0.3-32.0) cm and 6.0 (0.3-35.0) cm, respectively; The median mitotic indexes were 3 (0-113)/5 mm² and 3 (0-50)/5 mm², respectively; The median Ki-67 proliferation indexes were 5% (1%-80%) and 5% (1%-50%), respectively. The rates of positivity for CD117, DOG-1, and CD34 were 99.7% (792/794), 99.9% (731/732), 95.6% (753/788), and 100.0% (267/267), 100.0% (238/238), 61.5% (163/265), respectively. There were higher proportions of male patients (χ²=6.390, P=0.011), tumors of maximum diameter > 5.0 cm (χ²=33.593, P<0.001), high-risk (χ²=94.957, P<0.001), and CD34-negativity (χ²=203.138, P<0.001) among patients with intestinal GISTs than among those with gastric GISTs. (2) Gene mutations: Gene mutations were investigated in 286/360 patients (79.4%) with primary gastric GISTs and 74/360 (20.6%) with primary intestinal GISTs. Among the 286 patients with gastric primary GISTs, 79.4% (227/286), 8.4% (24/286), and 12.2% (35/286), had KIT mutations, PDGFRA mutations, and wild-type, respectively. Among the 74 patients with primary intestinal GISTs, 85.1% (63/74) had KIT mutations and 14.9% (11/74) were wild-type. The PDGFRA mutation rate was lower in patients with intestinal GISTs than in those with gastric GISTs[ 0% vs. 8.4%(24/286), χ²=6.770, P=0.034], whereas KIT exon 9 mutations occurred more often in those with intestinal GISTs [22.2% (14/63) vs. 1.8% (4/227), P<0.001]. There were no significant differences between gastric and intestinal GISTs in the rates of KIT exon 11 mutation type and KIT exon 11 deletion mutation type (both P>0.05). (3) Follow-up, survival, and postoperative treatment: After excluding 228 patients with synchronous and metachronous other malignant tumors, the remaining 833 patients were followed up for 6-124 (median 53) months with a follow-up rate of 88.6% (738/833). None of the patients with very low or low-risk gastric (n=239) or intestinal GISTs (n=56) had received targeted therapy postoperatively. Among 179 patients with moderate-risk GISTs, postoperative targeted therapy had been administered to 88/155 with gastric and 11/24 with intestinal GISTs. Among 264 patients with high-risk GISTs, postoperative targeted therapy had been administered to 106/153 with gastric and 62/111 with intestinal GISTs. The 3-, 5-, and 10-year PFS of patients with gastric or intestinal GISTs were 96.5%, 93.8%, and 87.6% and 85.7%, 80.1% and 63.3%, respectively (P<0.001). The 3-, 5-, and 10-year OS were 99.2%, 98.8%, 97.5% and 94.8%, 92.1%, 85.0%, respectively (P<0.001). (4) Analysis of predictors of intermediate- and high-risk GISTs: The 5-year PFS of patients with gastric and intestinal GISTs were 89.5% and 73.2%, respectively (P<0.001); The 5-year OS were 97.9% and 89.3%, respectively (P<0.001). Multivariate analysis showed that high risk (HR=2.918, 95%CI: 1.076-7.911, P=0.035) and Ki-67 proliferation index > 5% (HR=2.778, 95%CI: 1.389-5.558, P=0.004) were independent risk factors for PFS in patients with intermediate- and high-risk GISTs (both P<0.05). Intestinal GISTs (HR=3.485, 95%CI: 1.407-8.634, P=0.007) and high risk (HR=3.753,95%CI:1.079-13.056, P=0.038) were independent risk factors for OS in patients with intermediate- and high-risk GISTs (both P<0.05). Postoperative targeted therapy was independent protective factor for PFS and OS (HR=0.103, 95%CI: 0.049-0.213, P<0.001; HR=0.210, 95%CI:0.078-0.564,P=0.002). Conclusions: Primary intestinal GIST behaves more aggressively than gastric GISTs and more frequently progress after surgery. Moreover, CD34 negativity and KIT exon 9 mutations occur more frequently in patients with intestinal GISTs than in those with gastric GISTs.

Objective: To investigate the clinicopathological characteristics and prognosis of sporadic multiple primary gastrointestinal stromal tumor (GIST). Methods: A retrospective cohort study was conducted. Case inclusion criteria: (1) postoperative pathological diagnosis of GIST; (2) primary GIST with single lesion or sporadic multiple primary GIST (sporadic GIST was defined as primary GIST other than familial and syndrome-related GIST, and multiple primary GIST was defined as the number of primary GISTs in the same patient ≥ 2); (3) patients with complete clinicopathological data. Those with tumor recurrence or distant metastasis, and with other malignancies were excluded. Medical records of patients with primary GIST who underwent surgical resection in the Union Hospital, Tongji Medical College, Huazhong University of Science and Technology from January 2010 to December 2020 were collected. Patients were divided into sporadic multiple primary GIST group and single primary GIST group according to the number of primary GIST lesions. The clinicopathological data and prognosis of the two groups were observed and compared. Results: A total of 1200 patients with primary GIST were enrolled in this study, including 628 males (52.3%) and 572 females (47.7%), with a median onset age of 58 (19-93) years. Among them, 1165 cases (97.1%) were sporadic primary GIST with single lesion; 35 cases (2.9%) were sporadic multiple primary GIST. Among 35 cases of sporadic multiple primary GIST, 3 cases (8.6%) had acid reflux as the first symptom, which was higher than the single primary GIST group (22/1165, 1.9%) (χ(2)=7.437, P=0.006). There were no significant differences in other clinical characteristics between the two groups (all P>0.05). Patients in the sporadic multiple primary GIST group contained a total of 80 primary tumors. Compared with the single primary GIST group, the sporadic multiple primary GIST group had a higher proportion of tumors originating in the stomach [87.5% (70/80) vs. 59.1% (689/1165)], lower proportion of spindle cell in histology [85.0% (68/80) vs. 93.7% (1092/1165)], higher proportion of positive CD34 [97.5% (78/80) vs. 87.6% (1021/1165)], smaller maximum diameter [maximum diameter ≤2.0 cm: 61.2% (49/80) vs. 28.8% (335/1165)], lower mitotic rate [≤5/50 high-power fields (HPF): 93.8% (75/80) vs. 74.5% (868/1165)], lower risk of recurrence [60.0% (48/80) vs. 23.3% (271/1165)], and the differences were all statistically significant (all P<0.05). The 3-year recurrence-free survival rate in the sporadic multiple primary group and the single primary GIST group was 96.6% and 89.3% respectively (P=0.160), and the 3-year overall survival rate was 100.0% and 92.8%, respectively (P=0.088). Conclusions: The most common type of sporadic multiple primary GIST is multiple tumors originating in the stomach at the same time. Compared with primary GIST with single lesion, sporadic multiple primary GIST presents smaller maximum diameter and lower mitotic rate. The prognosis of patients between two groups is not significantly different.

Activation by point mutation of the H-, K- and N-ras genes is found in many tumors. However, no such mutation has yet been found in human esophageal carcinomas from various parts of the world. We have confirmed the absence of mutation at codons 12, 13 and 61 of K- and N-ras and at codons 12 and 61 of H-ras in 25 primary tumors obtained in France. In contrast, among 7 human esophageal carcinoma cell lines (TE1, TE2, TE3, TE8, TE9, TE10, TE13) with different degrees of tumorigenicity in nude mice, 3 highly tumorigenic lines (TE1, TE2 and TE8) exhibited activation of ras oncogenes; 2 showed a G35 to A35 transition of K-ras gene and one a H-ras G35 to T35 transversion. Since these cell lines had been established from tumors of Japanese patients from Sendai, we examined 3 primary esophageal tumors from Tokyo and 19 from Sendai, including the primary tumors from which the TE cell lines had been derived. No ras mutation was detected, which suggests that the ras gene mutations in the TE cell lines are either due to their long-term culture or that only a small portion of the original tumors contained such mutations. In order to directly examine the effect of ras gene mutation, one of the non-tumorigenic cell lines, TE13, was transfected with a plasmid coding for a mutated H-ras gene (G35 to T35). Transfected clones expressing high levels of mutated ras gene were able to induce tumors in nude mice. Thus, although no primary human esophageal tumor contained mutated ras genes, our studies do not exclude a significant role of mutated ras genes in cell proliferation and malignant transformation of human esophageal cells.

Background; Gastrointestinal stromal tumor (GIST) is known to originate from the interstitial cells of Cajal or its precursors with an activating mutation in the c-kit gene. On the other hand, it has been shown that micro GIST with c-kit mutation was commonly found in stomach resected for gastric cancer or in autopsy specimens. It seems that c-kit mutation initiates GIST and other genetic or epigenetic changes occur in the developmental process of GIST. In this study, we compared gene expression profiles among gastric and intestinal GIST as well as their metastatic liver tumors to identify the genes involved in the process of malignant transformation of GIST. Methods; Fresh frozen samples of primary gastric and intestinal GISTs and metastatic liver GISTs were used in the analysis. Total RNA was extracted and gene expression levels were compared by microarray analysis between 3 primary gastric GISTs and 4 metastatic liver GISTs, and among 6 primary gastric GISTs, 3 primary intestinal GISTs and 6 metastatic liver GISTs. Protein levels of versican and CD9 were analyzed by immunohistochemistry in 104 primary gastric GISTs and 13 metastatic liver GISTs. Results; Expression levels were higher for 165 genes and lower for 146 genes in metastatic liver GIST compared with primary gastric GIST. Among them, we further focused on two genes; versican and CD9, mRNA levels of which were higher and lower, respectively, in the metastatic liver GIST. Immunohistochemical analysis revealed that protein levels of versican and CD9 were higher and lower, respectively, in metastatic liver tumors compared with primary gastric GIST. High expression of versican and loss of CD9 expression correlated with shorter liver metastasis-free survival after gastrectomy. Microarray analysis of primary gastric and intestinal GISTs and metastatic liver GIST classified them into two groups according their gene expressions. One group consisted of primary gastric GIST without postoperative recurrence and the other consisted of clinically malignant GIST, intestinal GIST and metastatic liver GIST. These data suggest that intestinal GIST, regardless of presence of metastasis, shows gene expression profile similar to clinically malignant GIST and further to metastatic tumors and may explain mechanisms by which intestinal GISTs are more likely to show clinically malignant phenotype than gastric GISTs. Conclusions: Gene expression profile of primary intestinal GIST was similar to that in metastatic liver GIST but not to that in primary gastric GIST. Versican and CD9 could be potential prognostic markers of gastric GIST with which to predict the risk of postoperative liver metastasis. Citation Format: Hirotoshi Kikuchi, Tomohiko Setoguchi, Shinichiro Miyazaki, Ichirota Iino, Yoshihiro Hiramatsu, Kinji Kamiya, Manabu Ohta, Takanori Sakaguchi, Hiroyuki Konno. Differential gene expressions in primary gastric and intestinal gastrointestinal stromal tumors and metastatic liver tumors. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 405. doi:10.1158/1538-7445.AM2013-405 Note: This abstract was not presented at the AACR Annual Meeting 2013 because the presenter was unable to attend.

A primary gastrointestinal stromal tumor (GIST) arising in the liver is extremely rare. In our case of GIST, CT and MRI showed a well-defined, weakly enhancing mass with a cystic component in the left lateral segment of the liver that showed homogeneous and avid (18)F-fluorodeoxyglucose ((18)F-FDG) accumulation on positron emission tomography/computed tomography (PET/CT). We herein present a rare case of primary malignant GIST of the liver presenting with peritoneal seeding on CT, gadoxetic acid-enhanced MRI and (18)F-FDG PET/CT.

We present the case of a 71 year old man with recurrence of a Gastro Intestinal Stromal Tumour (GIST) at the gastrojejunal anastomosis eight years following partial gastrectomy for a very small primary gastric GIST. He presented acutely on both occasions with haemodynamic shock secondary to massive haematemesis. During his initial presentation in 2001, an emergency laparotomy was performed, demonstrating a pre-pyloric ulcerative lesion. The histopathology was in keeping with a diagnosis of a gastric GIST with a < 2 cm tumour, with <5 mitosis per 50/HPF, no signs of necrosis and invasion limited to the mucosa. Eight years later the same patient presented with a similar clinical picture of haemodynamic instability secondary to haematemesis. Emergency endoscopy showed an irregularly shaped elevated lesion on the gastrojejunostomy line suggestive of recurrence. He subsequently underwent completion gastrectomy and the histology revealed a 0.8 cm GIST tumour composed of spindle cells with <5 mitosis per 50/HPF, tumor invasion into the submucosa and positive expression of c-kit and SMA. The patient remains recurrence free 18 months post surgery. The literature suggests that tumour size, mitotic rate and tumour site are the most important predictive factors of recurrence. Additional features such as the presence of necrosis, local tumour invasion and positive resection margins, can also influence recurrence rates. In this case the lesion was a gastric GIST, very small (<2 cm), had low proliferation rate (<5 mitosis/HPF), lacked necrosis and was limited to the mucosa. Recurrence of such a primary GIST at the anastomotic line, eight years after initial resection has never been demonstrated among review of several thousand primary GISTs. This case highlights how even the most innocent GISTs can never be described as truly benign.

Gastrointestinal Stromal Tumors: Disease and Treatment Update

Data on treatments and specific outcomes of primary gastrointestinal stromal tumors (GISTs) ≥10 cm are limited. We here report the treatments and survival outcomes concerning a subgroup of primary giant GISTs.Data of 83 consecutive patients with primary GISTs ≥10 cm in a single institution were retrospectively collected. Fifty-eight patients underwent surgery before imatinib mesylate (IM) treatment (Group A), 10 underwent surgical resection following IM therapy (Group B), whereas 15 patients took IM as drug therapy alone (Group C).The baseline clinical characteristics were similar among the 3 groups. However, a lower proportion in Group A had metastatic disease at the time of diagnosis or surgery compared with Groups B and C (8.6% vs 40.0% vs 40.0%, P < 0.05). The median follow-up duration was 21.5 months. No statistically significant differences were observed on progression-free survival (PFS) among the groups. However, patients in Group B showed significantly better overall survival (OS) compared with those in Group C (P = 0.044). Multivariate analysis showed that patients treated with adjuvant IM were associated with better PFS (hazard ratio [HR] 3.01; 95% confidence interval [CI] 1.13–7.97; P = 0.027) and OS (HR 29.11; 95% CI 3.32–125.36; P = 0.004). The subgroup with mitotic count >10/50 high-power fields (HPF) showed worse PFS (HR 3.50; 95% CI 1.19–10.25; P = 0.022) and OS (HR 20.04; 95% CI 1.67–143.79; P = 0.018) than that of mitotic count ≤5/50 HPF.Clinical treatment patterns for primary giant GISTs are different, and the outcomes of different interventions vary. The optimal treatments for these subgroup of patients still require further long-term investigation. Moreover, mitotic count and adjuvant IM are closely associated with PFS and OS in giant GISTs.

Background: Gastrointestinal stromal tumor (GIST) is the most common mesenchymal tumor of the gastrointestinal tract. The majority of GISTs have a gain-of-function mutation of the c-kit or PDGFRA gene in the interstitial cells of Cajal. Although surgery is the most effective treatment for resectable primary GISTs, postoperative recurrence or metastasis has been observed after surgical resection in 40-90% of patients without adjuvant therapy. Metastatic GISTs are found most commonly in the liver, which are difficult to cure with surgical treatment alone. Although tyrosine kinase inhibitor (TKI) imatinib mesylate (Gleevec) is thought to be the most effective agent for treating GISTs, secondary resistance often arises during therapy due to secondary mutations. In this era of TKIs, control of liver metastasis remains to be an important issue in the treatment of GISTs and mechanisms of liver metastasis need to be elucidated. In this study, we compared gene expression and micro RNA (miRNA) expression profiles between gastric GISTs and metastatic liver GISTs to address mechanisms of GIST liver metastasis and to detect novel molecular markers for prediction of the postoperative prognosis of GIST patients. Methods: Frozen tissue specimens of three gastric GISTs and four metastatic liver GISTs were utilized for cDNA microarray analysis. None of the patients had received imatinib therapy before surgery. Formalin-fixed paraffin-embedded (FFPE) tissues of 104 primary GISTs and 13 metastatic GISTs from 107 patients who had undergone surgery were included for immunohistochemistry investigations. GIST cells were isolated from FFPE tissues of five gastric GISTs at low risk, five gastric GISTs at high risk and six metastatic liver GISTs surgically resected, and miRNA expression was analyzed using TaqMan miRNA array. Results: Expression levels were higher for 165 genes and lower for 146 genes in metastatic liver GIST. The upregulation of five oncogenes including versican and downregulation of four tumor suppressor genes, Cadherin8, Protocadherin10, NRCAM and CD9 were confirmed by qRT-PCR. Immunohistochemistry in 117 GISTs revealed that protein levels of versican and CD9 were higher and lower, respectively, in metastatic GIST. High expression of versican and low expression of CD9 in 104 primary gastric GISTs correlated with poor disease-free survival (P= 0.0078 and P= 0.0018). In miRNA array analysis, miR-122 was the most highly expressed in liver metastasis compared with primary gastric GISTs, consistent with our earlier papers that showed the overexpression of miR-122 and concomitant suppression of CAT1 in colorectal liver metastasis (Cancer Sci 104,624-30,2013). Conclusions: In addition to the c-kit gene mutation, genetic or epigenetic changes other than c-kit and alteration of miRNA expression such as miR-122 play important roles in the metastatic process of gastric GIST. Citation Format: Hirotoshi Kikuchi, Ichirota Iino, Shinichiro Miyazaki, Yusuke Ozaki, Yoshihiro Hiramatsu, Manabu Ohta, Kinji Kamiya, Takanori Sakaguchi, Satoshi Baba, Haruhiko Sugimura, Mitsutoshi Setou, Hiroyuki Konno. cDNA and miRNA microarray analysis comparing gastric and metastatic liver gastrointestinal stromal tumors. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 4008. doi:10.1158/1538-7445.AM2014-4008

Gastrointestinal stromal tumors (GISTs) of the esophagus are a rare entity. Diagnosis of GIST is currently based on immunohistochemical staining of c-KIT or CD34. However, some tumors have clinicopathologic features of GIST but do not express c-KIT or CD34. A few GISTs contain mutations within a receptor tyrosine kinase protein, platelet-derived growth factor receptor alpha (PDGFRA). We herein report a case of esophageal GIST that was KIT negative and had a PDGFRA mutation. A 75-year-old male who had a giant submucosal tumor in the lower part of the thoracic esophagus underwent surgical resection. Immunohistochemical staining of the tumor revealed that it was uniformly negative for KIT, but partially positive for CD34, and negative for S-100 protein and smooth muscle actin (SMA), which led to the final diagnosis of GIST. PDGFRA genetic testing revealed a mutation in exon 12. The mitotic index was over 5/50 high-power fields, and necrotic changes were noted. Adjuvant chemotherapy using imatinib mesylate was administered. The patient has been disease free for 2 years. To the best of our knowledge, this is the only reported case of esophageal GIST that was KIT negative and had a PDGFRA mutation. In cases of digestive submucosal tumor that are difficult to diagnose because of c-KIT or CD34 negativity despite being suspicious for GIST, PDGFRA genetic testing may help for diagnosis of this minority type of GIST.

We report a case of esophagectomy after a primary esophageal gastrointestinal stromal tumor (GIST) was preoperatively treated with imatinib mesylate. A 71-year-old woman was diagnosed with an esophageal submucosal tumor by upper gastrointestinal endoscopy at her health checkup. The tumor was located at the lower thoracic esophagus immediately above the esophagogastric junction and measured 4.5 cm in size. It was diagnosed as GIST of the esophagus for reasons of its high susceptibility to imatinib mesylate. Preoperative treatment with imatinib was performed in an attempt to preserve the esophagus. Although the tumor size was decreased by 36% after the 6-month treatment, transhiatal esophagectomy was required for complete resection, and esophageal preservation could not be accomplished.

Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal tumors of the gastrointestinal tract. They can occur anywhere in the gastrointestinal tract, and rarely outside the digestive tract. We herein report a case of primary gastrointestinal stromal tumor that was resected from the liver of a 56-year-old male, which is the sixth description of a primary hepatic gastrointestinal stromal tumor. The tumor was shown to be completely limited within the liver by radiological, intraoperative and pathological examinations. The pathological results demonstrated an intermediate risk gastrointestinal stromal tumor, and immunohistochemical expression of CD117 was positive. Although rare, we suggested that GISTs should be considered in the differential diagnosis of hepatic nodules, and that not all hepatic gastrointestinal stromal tumors should automatically be considered to be metastases from a primary gastrointestinal site.

Primary mesenchymal lesions of the prostate are exceptionally rare. They comprise 1% of all prostatic neoplasms. Despite its rare location, the diagnosis of primary gastrointestinal stromal tumors (GISTs) of the prostate gland should never be missed. Such a diagnosis can be made after the rolling out of direct extension from adjacent organs, especially the rectum. GIST diagnosis has a clinical impact on patient treatment and clinical outcomes. They harbor a certain KIT activation mutation that responds to pharmacologic therapy inhibitors.The objective of the current study was to provide a thorough review of GIST arising primarily in the prostate gland along with a comprehensive study of GIST pathogenesis, histologic morphology, immunohistochemistry, and molecular studies’ findings, and their importance in differentiating GIST from other prostate mesenchymal tumors. This will emphasize the role of careful spindle cell lesion diagnosis in the prostate gland that can influence the prognostic stratification of clinical management, future follow-up, and disease outcome.Thirteen cases were collected after an extensive and detailed review of the English literature through PubMed, Medknow, Google Scholar, as well as personal experience.The anatomic location of this lesion plays a significant role in the differential diagnosis. It is difficult to establish the accurate primary origin of GIST on core needle tissue biopsy. Thus, clinical, and radiological examinations play a crucial role in rolling out the possibility of rectal GIST secondarily invading and involving the prostate gland.To conclude, primary prostatic GIST is a rare diagnosis. Extraintestinal, particularly rectal, GIST can clinically and radiologically mimic the impression of the prostatic lesion. Before diagnosing primary prostatic spindle cell lesions, such as solitary fibrous tumor (SFT), inflammatory myofibroblastic tumor (IMT), leiomyoma, leiomyosarcoma, or prostatic stromal tumors, one should include CD117/c-Kit in the workup of a prostatic spindle cell lesion. GIST has distinct pathogenesis, and its diagnosis can have a clinical impact on the patient's management plan and clinical outcome.