Objective:To evaluate the clinical value of preoperative Naples prognostic scores (NPS) in patients with resectable Siewert type II-III esophagogastric junction adenocarcinoma (AEG).Methods:In this retrospective observational study we collected and analyzed relevant data of patients with Siewert Type II-III AEG treated in the Department of Gastric Cancer, Tianjin Medical University Cancer Institute and Hospital from January 2014 to December 2018. NPS were calculated using preoperative albumin concentration, total cholesterol concentration, neutrophil/lymphocyte ratio, and lymphocyte/monocyte ratio and used to allocate patients into three groups: NTS-0 (0 points), NTS-1 (1-2 points) and NTS-2 (3-4 points). Kaplan–Meier was used to calculate disease-free survival (DFS) and overall survival (OS) in each NPS group and the log-rank test to compare these groups. Univariate and multivariate survival analyes were performed using the Cox regression model. Time-dependent receiver operating characteristic curves were constructed to compare the relationships between four commonly used tools for evaluating inflammatory responses and nutritional status:NPS, systemic inflammatory response scores, nutrient control status (CONUT), and prognostic nutrition index (PNI).Results:The study cohort comprised 221 patients with AEG of median age 63.0 (36.0–87.0) years. There were 190 men (86.0%) and 31 women (14.0%). As to pTNM stage, 47 patients (21.3%) had Stage I disease, 68 (30.8%) Stage II, and 106 (48.0%) Stage III. One hundred and forty-seven patients (66.5%) had Siewert Type II disease and 74 (33.5%) Siewert type III. There were 45 patients (20.4%) in the NPS-0, 142 (64.2%) in the NPS-1 and 34 (15.4%) in the NPS-2 groups. Higher NPS scores were significantly associated with older patients (χ2=5.056, P=0.027) and higher TNM stages ( H=5.204, P<0.001). The median follow-up was 39 (6-105) months; 16 patients (7.2%) were lost to follow-up. The median OS in the NPS-0, NPS-1, and NPS-2 groups were 78.4, 63.1, and 37.0 months, respectively; these differences are statistically significant ( P=0.021). Univariate and multivariate Cox regression analysis identified the following as independently and significantly associated with OS in patients with Siewert Type II-III: TNM stage (Stage II: HR=2.182, 95%CI: 1.227-3.878, P=0.008; Stage III: HR=3.534, 95%CI: 1.380-6.654, P<0.001), tumor differentiation (G3: HR=1.995, 95%CI: 1.141-3.488, P=0.015), vascular invasion (HR=2.172, 95%CI: 1.403-3.363, P<0.001), adjuvant chemotherapy (HR=0.326, 95%CI: 0.200-0.531, P<0.001), NPS (NPS-1: HR=2.331, 95%CI: 1.371-3.964, P=0.002; NPS-2: HR=2.494, 95%CI: 1.165-5.341, P=0.019), SIS group (NPS-1: HR=2.170, 95%CI: 1.244-3.784, P=0.006; NPS-2: HR=2.291, 95%CI: 1.052–4.986, P=0.037), and CONUT (HR=1.597, 95% CI: 1.187-2.149, P=0.038). The median DFS in the NPS-0, NPS-1, and NPS-2 groups was 68.6, 52.5, and 28.3 months, respectively; these differences are statistically significant ( P=0.009). Univariate and multivariate Cox regression analysis identified the following as independently and significantly associated with DFS in patients with Siewert Type II-III AEG: TNM stage (StageⅡ: HR=2.789, 95%CI:1.210-6.428, P=0.016; Stage III: HR=10.721, 95%CI:4.709-24.411, P<0.001), adjuvant chemotherapy (HR=0.640, 95% CI: 0.432-0.946, P=0.025), and NPS (NPS-1: HR=1.703, 95%CI: 1.043-2.782, P=0.033; NPS-2: HR=3.124, 95%CI:1.722-5.666, P<0.001). Time-dependent receiver operating characteristic curves showed that NPS was more accurate in predicting OS and DFS in patients with Siewert Type II-III AEG than were systemic inflammatory response scores, CONUT, or PNI scores. Conclusion:NPS is associated with age and TNM stage, is an independent prognostic factor in patients who have undergone resection of Siewert type II-III AEG, and is better than SIS, CONUT, or PNI in predicting survival.

Objective:To investigate the clinicopathological characteristics and prognostic factors of gastrointestinal stromal tumor (GIST) with initial surgical resection.Methods:The retro-spective cohort study was conducted. The clinicopathological data of 847 GIST patients who under-went initial surgical resection in Tianjin Medical University Cancer Institute & Hospital from January 2011 to December 2020 were collected. There were 405 males and 442 females, aged (60±10)years. Measurement data with normal distribution were represented as Mean± SD, and comparison between groups was conducted using the t test. Measurement data with skewed distribution were represented as M(range). Count data were described as absolute numbers or percentages, and comparison between groups was conducted using the chi-square test. Comparison of ordinal data was conducted using the nonparameter rank sum test. The Kaplan-Meier method was used to calculate survival rates. Univariate analysis was conducted using the Log-rank test. Multivariate analysis was conducted using the COX regression model. Results:(1) Clinicopatholo-gical characteristics. Of 847 patients, the tumor primary location was stomach in 585 cases, jejunum and ileum in 142 cases, duodenum in 76 cases, colorectum in 10 cases, esophagus in 3 cases, and extra-gastrointestinal in 31 cases. There were 13 cases with liver metastasis and 22 cases with abdominal metastasis. The tumor maximum diameter was (7±5)cm, and the number of nuclear divisions was 4(range, 0-60) cells/50 high-power field or 5 mm 2. According to risk classification of National Institutes of Health (NIH), 31 cases were of extremely low risk, 238 cases were of low risk, 213 cases were of moderate risk, 365 cases were of high risk. There were 839 of 847 patients positive for CD117, 788 cases positive for Dog-1, 710 cases positive for CD34, respectively. There were 272 cases with Ki-67 <5%, 214 cases with Ki-67 of 5%- 9%, 198 cases with Ki-67 ≥10%, 163 cases with missing data. R 0 resection was in 814 cases and non-R 0 resection was in 33 cases. (2) Gene testing and postoperative adjuvant therapy of GIST patients. ① Gene testing. Of 847 patients, 424 underwent genetic testing. The proportion of genetic testing was 1.89%(1/53) in 2011, 9.76%(8/82) in 2012, 8.45%(6/71) in 2013, 15.66%(13/83) in 2014, 50.00%(40/80) in 2015, 55.26%(42/76) in 2016, 73.86%(65/88) in 2017, 68.27%(71/104) in 2018, 80.65%(75/93) in 2019, 88.03%(103/117) in 2020, respectively. Of 424 with genetic testing, 338 cases had KIT mutation, 31 cases had PDGFRA mutation, 55 cases were wild type. ② Adjuvant therapy. Of 847 patients, 253 patients underwent postoperative adjuvant therapy. The proportions of postoperative adjuvant therapy were 8.82%(21/238), 41.78%(89/213), 39.18%(143/365) in patients of low risk, moderate risk, high risk. Of 578 patients with moderate to high risk, the proportion of postoperative adjuvant therapy was 15.15%(5/33) in 2011, 14.71%(10/68)in 2012, 22.45%(11/49) in 2013, 29.09%(16/55) in 2014, 41.38%(24/58) in 2015, 46.15%(24/52) in 2016, 32.81%(21/64)in 2017, 60.00%(45/75) in 2018, 60.42%(29/48) in 2019, 61.84%(47/76) in 2020, respectively. Of 253 patients underwent postoperative adjuvant therapy, 247 cases received imatinib had 6 cases received sunitinib. (3) Comparison of clinicopathological characteristics of GIST with non-gastric origin and gastric origin. Of 847 patients, 262 cases had non-gastric origin and 585 cases had gastric origin. There were significant differences in gender, the number of tumor, tumor maximum diameter, Ki-67 index, risk classification of NIH, and R 0 resection between the two groups ( χ2=8.62, 8.40, 12.97, 6.57, Z=-6.15, χ2=17.19, P<0.05). (4) Analysis of influencing factors for recurrence-free survival rate in GIST patients. Results of multivariate analysis showed that the year of initial diagnosis, primary site, tumor maximum diameter, mitotic image, risk classification of NIH, R 0 resection, genetic testing and postoperative adjuvant therapy were independent factors influencing recurrence-free survival rate in GIST patients with initial surgical resection ( hazard ratio=0.58, 0.61, 2.00, 1.71, 5.81, 2.56, 0.65, 0.38, 95% confidence interval as 0.39-0.85, 0.45-0.83, 1.46-2.74, 1.24-2.35, 3.16-10.69, 1.63-4.02, 0.46-0.94, 0.25-0.56, P<0.05). Conclusions:GIST with initial surgical resection is common located in stomach, with high positive rate in CD117 and Dog-1. The number of people undergoing genetic testing and targeted therapy for GIST is increasing year by year. There are significant differ-ences in clinicopathological characteristics between GIST with non-gastric origin and gastric origin. The year of initial diagnosis, primary site, tumor maximum diameter, mitotic image, risk classifica-tion of NIH, R 0 resection, genetic testing and postoperative adjuvant therapy are independent factors influencing recurrence-free survival rate in GIST patients with initial surgical resection.

Objective:To evaluate the clinical value of preoperative Naples prognostic scores (NPS) in patients with resectable Siewert type II-III esophagogastric junction adenocarcinoma (AEG).Methods:In this retrospective observational study we collected and analyzed relevant data of patients with Siewert Type II-III AEG treated in the Department of Gastric Cancer, Tianjin Medical University Cancer Institute and Hospital from January 2014 to December 2018. NPS were calculated using preoperative albumin concentration, total cholesterol concentration, neutrophil/lymphocyte ratio, and lymphocyte/monocyte ratio and used to allocate patients into three groups: NTS-0 (0 points), NTS-1 (1-2 points) and NTS-2 (3-4 points). Kaplan–Meier was used to calculate disease-free survival (DFS) and overall survival (OS) in each NPS group and the log-rank test to compare these groups. Univariate and multivariate survival analyes were performed using the Cox regression model. Time-dependent receiver operating characteristic curves were constructed to compare the relationships between four commonly used tools for evaluating inflammatory responses and nutritional status:NPS, systemic inflammatory response scores, nutrient control status (CONUT), and prognostic nutrition index (PNI).Results:The study cohort comprised 221 patients with AEG of median age 63.0 (36.0–87.0) years. There were 190 men (86.0%) and 31 women (14.0%). As to pTNM stage, 47 patients (21.3%) had Stage I disease, 68 (30.8%) Stage II, and 106 (48.0%) Stage III. One hundred and forty-seven patients (66.5%) had Siewert Type II disease and 74 (33.5%) Siewert type III. There were 45 patients (20.4%) in the NPS-0, 142 (64.2%) in the NPS-1 and 34 (15.4%) in the NPS-2 groups. Higher NPS scores were significantly associated with older patients (χ2=5.056, P=0.027) and higher TNM stages ( H=5.204, P<0.001). The median follow-up was 39 (6-105) months; 16 patients (7.2%) were lost to follow-up. The median OS in the NPS-0, NPS-1, and NPS-2 groups were 78.4, 63.1, and 37.0 months, respectively; these differences are statistically significant ( P=0.021). Univariate and multivariate Cox regression analysis identified the following as independently and significantly associated with OS in patients with Siewert Type II-III: TNM stage (Stage II: HR=2.182, 95%CI: 1.227-3.878, P=0.008; Stage III: HR=3.534, 95%CI: 1.380-6.654, P<0.001), tumor differentiation (G3: HR=1.995, 95%CI: 1.141-3.488, P=0.015), vascular invasion (HR=2.172, 95%CI: 1.403-3.363, P<0.001), adjuvant chemotherapy (HR=0.326, 95%CI: 0.200-0.531, P<0.001), NPS (NPS-1: HR=2.331, 95%CI: 1.371-3.964, P=0.002; NPS-2: HR=2.494, 95%CI: 1.165-5.341, P=0.019), SIS group (NPS-1: HR=2.170, 95%CI: 1.244-3.784, P=0.006; NPS-2: HR=2.291, 95%CI: 1.052–4.986, P=0.037), and CONUT (HR=1.597, 95% CI: 1.187-2.149, P=0.038). The median DFS in the NPS-0, NPS-1, and NPS-2 groups was 68.6, 52.5, and 28.3 months, respectively; these differences are statistically significant ( P=0.009). Univariate and multivariate Cox regression analysis identified the following as independently and significantly associated with DFS in patients with Siewert Type II-III AEG: TNM stage (StageⅡ: HR=2.789, 95%CI:1.210-6.428, P=0.016; Stage III: HR=10.721, 95%CI:4.709-24.411, P<0.001), adjuvant chemotherapy (HR=0.640, 95% CI: 0.432-0.946, P=0.025), and NPS (NPS-1: HR=1.703, 95%CI: 1.043-2.782, P=0.033; NPS-2: HR=3.124, 95%CI:1.722-5.666, P<0.001). Time-dependent receiver operating characteristic curves showed that NPS was more accurate in predicting OS and DFS in patients with Siewert Type II-III AEG than were systemic inflammatory response scores, CONUT, or PNI scores. Conclusion:NPS is associated with age and TNM stage, is an independent prognostic factor in patients who have undergone resection of Siewert type II-III AEG, and is better than SIS, CONUT, or PNI in predicting survival.

Objective:To investigate the clinical efficacy of antiperistaltic uncut Roux-en-Y anastomosis (Uncut RY) in laparoscopic distal gastrectomy (LDG) for gastric cancer (GC).Methods:The clinicpathological data of 309 patients who underwent LDG for GC between Jan 2020 to Dec 2021 in Tianjin Medical University Cancer Institute & Hospital were collected. Among them, 102 cases received antiperistaltic Uncut RY anastomosis, 96 cases did RY anastomosis, and 111 cases did B Ⅱ+Braun anastomosis. The general data and clinical efficacy were compared.Results:The operative time in Uncut RY group and B Ⅱ+Braun group was shorter than that in RY group ( F=3.356, P=0.036). The incidence of postoperative short-term complications was not statistically different among three groups ( P>0.05). Compared to Uncut RY group and RY group, the B Ⅱ+Braun group had a significantly higher incidences of reflux gastritis (19.8% vs. 6.9% vs.5.2%, F=13.889, P=0.001). The incidence of RY stasis syndrome in RY group (15.6%) was significantly higher than that in Uncut RY group (2.0%) and B Ⅱ+Braun group (2.7%) ( F=19.320 ,P=0.000). There were no significant differences in postoperative 1-year nutritional indexes among three group ( P>0.05). Conclusions:The antiperistaltic Uncut RY anastomosis in LDG for GC is safe and feasible, which can reduce the incidences of reflux gastritis and RY stasis syndrome.

Abstract Allergic diseases can occur in all systems of the body, covering the whole life cycle, from children to adults and to old age, can be lifelong onset and even fatal in severe cases. Children account for the largest proportion of the victims of allergic disease, Children s allergies start from scratch, ranging from mild to severe, from less to more, from single to multiple systems and systemic performance, so the prevention and treatment of allergic diseases in children is of great importance, which can not only prevent high risk allergic conditions from developing into allergic diseases, but also further block the process of allergy. At present, there is no consensus on the management system of allergic children in kindergartens and primary schools. The "Consensus on Allergy Management and Prevention in Kindergartens and Primary Schools", which includes the organizational structure, system construction and management of allergic children, provides evidence informed recommendations for the long term comprehensive management of allergic children in kindergartens and primary schools, and provides a basis for the establishment of the prevention system for allergic children.

Objective:To investigate the clinical efficacy of radical proximal gastrectomy with esophagogastrostomy and double-tract anastomosis for upper gastric cancer.Methods:The retrospective cohort study was conducted. The clinicopathological data of 172 patients who underwent radical proximal gastrectomy for upper gastric cancer in Tianjin Medical University Cancer Institute and Hospital from January 2018 to December 2020 were collected. There were 147 males and 25 females, aged from 25 to 81 years, with a median age of 62 years. All the 172 patients underwent digestive reconstruction. Of the 172 patients, 83 cases undergoing esophagogastrostomy were allocated into esophagogastrostomy group, 89 cases undergoing double-tract anastomosis were allocated into double-tract anastomosis group. Patients were performed radical proximal gastrectomy combined with D 1+ lymph node dissection by attending surgeons from department of gastric cancer. The operator decided to adopt esophagogastrostomy or double-tract anastomosis for digestive reconstruction. Observation indicators: (1) surgical situations; (2) follow-up. Follow-up using outpatient examination, telephone interview, and online APP was conducted at postoperative 1 month, once three months within postoperative 2 years, and once six months within postoperative 2-5 years. The questionnaires of reflux esophagitis, gastroscopy and upper gastrointestinal angio-graphy were conducted to evaluate gastroesophageal reflux and anastomotic stenosis up to February 1, 2021. Measurement data with normal distribution were represented as Mean± SD, and comparison between groups was analyzed using the t test. Measurement data with skewed distribution were represented as M (range), and comparison between groups was analyzed using the Mann-Whitney U test. Comparison of ordinal data was analyzed using the non-parameter rank sum test. Count data were represented as absolute numbers, and comparison between groups was analyzed using the chi-square test. Results:(1) Surgical situations: cases with open, laparoscopic or Da Vinci robotic surgery (surgical method), the number of metastatic lymph node, duration of postoperative hospital stay were 74, 9, 0, 2(range, 0-15), (12±4)days for the esophagogastrostomy group, versus 65, 15, 9, 3(range, 0-28), (11±3)days for the double-tract anastomosis group, respectively, showing significant differences in the above indicators between the two groups ( χ2=10.887, Z=-1.058, t=3.284, P<0.05). (2) Follow-up: 172 patients were followed up for 2-38 months, with a median follow-up time of 13 months. Cases with gastroesophageal reflux and anastomotic stenosis were 58 and 10 for the esophagogastrostomy group, versus 14 and 1 for the double-tract anastomosis group, respectively, showing significant differences in the above indicators between the two groups ( χ2=51.743, 7.219, P<0.05). Conclusions:For upper gastric cancer patients undergoing proximal radical gastrectomy, double-tract anastomosis is more suitable for Siewert type Ⅱ adenocarcinoma of esophagogastric junction in large curvature or lower located tumor. Compared with esophago-gastrostomy, double-tract anastomosis has lower incidence of postoperative gastroesophageal reflux and anastomotic stenosis, without increasing complications.

Objective: To evaluate the potential clinical value of standardized procedures of fine lymph node sorting from gastric can-cer samples after curative resection. Methods: Between January 2016 and December 2017, 727 gastric cancer patients who under- went R0 resection in the Tianjin Medical University Cancer Institute and Hospital were retrospectively included and assigned to either the fine lymph node sorting group or regional lymph node sorting group in accordance with the lymph node sorting methods from the tumor samples of all patients. Both the numbers of examined lymph nodes and metastatic lymph nodes were compared between the two groups. Additionally, correlation analyses were performed between the numbers of examined lymph nodes and metastatic lymph nodes in the two groups. Results: There was no significant difference in sex, age, or tumor size between the two groups (P>0.05), indi-cating that there was comparability between the two groups. The number of examined lymph nodes in the fine lymph node sorting group was significantly higher than that in the regional lymph node sorting group (P<0.001). Furthermore, the number of examined lymph nodes in the fine lymph node sorting group was much higher than that in the regional lymph node sorting group with the same pT, pN, or pTNM stage (P<0.001). The number of metastatic lymph nodes in the fine lymph node sorting group was significantly higher than that in the regional lymph node sorting group (P<0.001). There was a significant positive correlation between the numbers of ex-amined lymph nodes and metastatic lymph nodes in both groups (fine lymph node sorting group r=0.181, P=0.023; regional lymph node sorting group r=0.227, P<0.001). Additionally, the correlation coefficient between the numbers of examined lymph nodes and metastatic lymph nodes in the fine lymph node sorting group was weaker than that in the regional lymph node sorting group. Conclu-sions: The standard procedures of fine lymph node sorting from tumor samples of gastric cancer may increase the number of exam-ined lymph nodes, accurately provide the postoperative pN stage, reduce the stage migration, and should be applied in clinical stan-dardization.

Resveratrol is a non-flavonoid polyphenolic compound which presents in plants such as grapes.It has a variety of physiological effects such as antioxidant ,anti-inflammatory ,anti-cancer , cardiovascular and neurological protection. In cardiovascular disease ,through the anti-inflammatory , anti-oxidation and regulating a variety of factors ,resveratrol involves in atherosclerosis ,hypertension , ischemic heart disease ,heart failure and other cardiovascular disease process ,thus plays a protective function for heart. Related clinical researches are also widely carried out ,and resveratrol may be one of the candidate interventions for prevention and treatment of cardiovascular disease.

OBJECTIVETo investigate the lymph node metastasis pattern in pN1 stage gastric cancer patients and to analyze its risk factors.

METHODSClinicopathological data of 219 patients who underwent radical gastrectomy and were confirmed as pN1 stage gastric cancer between January 2013 and March 2016 were reviewed. All the patients underwent D2 or extended D2(D2+) lymphadenectomy. The overall metastatic rate was calculated. The risk factors associated with lymph node metastasis were analyzed. The pattern of skip lymph node metastasis and clinicopathological factors related to skip metastasis were analyzed.

RESULTSAmong 219 patients, 119 patients had only one metastatic lymph node, and 100 patients had two metastatic lymph nodes. The relatively higher sites of lymph node metastasis were station No.3[29.2%(64/219)], No.6[18.3%(40/219)] and No.4[11.4%(25/219)]. Compared to patients with tumor diameter ≤5 cm, metastatic rates of station No.3[39.4% (39/99) vs. 20.8%(25/120), P=0.003], No.4[16.2%(16/99) vs. 7.5%(9/120), P=0.045] and No.8[16.2%(16/99) vs. 6.7%(8/120), P=0.025] were significantly higher in those with tumor diameter >5 cm. Skip lymph node metastasis was detected in 56 cases(25.6%) and skip lymph node metastatic rate was significantly higher in patients with tumor diameter >5 cm [34.3%(34/99) vs. 18.3%(22/120), P=0.007]. Logistic regression analysis showed that the tumor size was an independent risk factor for the skip lymph node metastasis in pN1 stage gastric cancer (OR=1.982, 95%CI: 0.978 to 3.921, P=0.033).

CONCLUSIONSThe perigastric station No1 lymph node is the main site of early lymph node metastasis of pN1 stage gastric cancer. General pattern of lymph node metastasis is from proximus to distance, while quite a lot of skip lymph node metastases are observed. Tumor size is an important factor affecting the lymph node metastasis and bigger tumor may result in skip lymph node metastasis easily.

OBJECTIVETo evaluate the impact of primary site, NIH risk and imatinib treatment on the prognosis of patients with gastrointestinal stromal tumors(GIST).

METHODSClinicopathological data of 156 adult patients with GIST treated by imatinib in the Cancer Institute and Hospital of Tianjin Medical University from January 2006 to December 2010 were retrospectively analyzed. According to NIH risk classification, 30 patients were at moderate risk and 126 at high risk. Sixty-seven patients had advanced GIST. Prognosis of patients with different primary tumor site, different NIH risk and different treatment was compared respectively.

RESULTSImatinib therapy was well tolerated in all the patients. Eighty-nine cases received radical operation and adjuvant imatinib treatment. Among 67 advanced GIST cases, 26 received radical operation and adjuvant imatinib treatment, 27 received palliative operation and adjuvant imatinib treatment, and 14 received simple adjuvant imatinib treatment without operation. All the patients had routine follow-up, ranging from 9 to 56(median 27) months. The overall survival (OS) rate was 96% in 1-year, 86% in 2-year, and 71% in 3-year. The OS rate was 95% in 1-year, 77% in 2-year, and 65% in 3-year for patients at high risk, and all 100% in 1-, 2-, 3-year for patients at moderate risk, the differences was statistically significant (P=0.001). The OS rate was 97% in 1-year, 90% in 2-year, and 84% in 3-year for patients with gastric GIST, and 95% in 1-year, 69% in 2-year, and 52% in 3-year for patients with non-gastric GIST, the difference was significant(P=0.000). The OS rate was 98% in 1-year, 95% in 2-year, and 90% in 3-year for patients undergoing radical resection and adjuvant imatinib therapy. For 67 advanced GIST patients with imatinib therapy, none had complete remission, 41 had part remission, 15 had stable disease, indicating 56 advanced GIST cases(83.6%) obtaining clinical benefit. The OS rate was 91% in 1-year, 58% in 2-year, and 43% in 3-year.

CONCLUSIONSThe prognosis of high, and non-gastric and advanced GIST patients is poor. Radical resection combined with early imatinib treatment can improve the prognosis of GIST patients.

Antineoplastic Agents ; therapeutic use ; Benzamides ; therapeutic use ; Combined Modality Therapy ; Follow-Up Studies ; Gastrointestinal Neoplasms ; drug therapy ; pathology ; Gastrointestinal Stromal Tumors ; drug therapy ; Humans ; Imatinib Mesylate ; Piperazines ; therapeutic use ; Prognosis ; Pyrimidines ; therapeutic use ; Retrospective Studies ; Survival Rate