Analysis Of Clinical Pathological Features And Prognosis In212Cases Of Patients With Gastrointestinal Stromal Tumor

Backgroud:Gastrointestinal stromal tumor (GIST) is a kind of low proportion tumor in all of digestive tract tumor, accounting for only2.2%of the gastrointestinal tumors. However, GIST is the most common mesenchymal neoplasms of gastrointestinal tract. It was often misdiagnosed as leiomyoma, leiomyosarcoma or neurilemmoma. Mazur and Clark firstly proposed the concept of gastrointestinal stromal tumor in1983. Hirota, Kindlom and Heinrich conducted extensive research about origin, pathogenesis, pathological features and diagnosis of GIST, increasing the level of diagnosis of GIST greatly. At present, the diagnosis of GIST primarily base on the histopathology and immunohistochemical examination about CD117protein.Because of insidious onset, nonspecific clinical manifestations of GIST, To make preoperative diagnosis is difficult. There are some of GIST which comply with pathological morphology and express negatively CD117protein, which increase the difficulty in diagnosis of GIST. The advent of targeted drugs including mesylate imatinib and sunitinib greatly improved the prognosis of patients with GIST. However, there are some patients primarily or secondarily resistant to mesylate imatinib and sunitinib.Objective:1. To investigate the clinical manifestations, primary site, metastasis, surgical resection, targeted therapy, clinical diagnosis and other clinical factors of gastrointestinal stromal tumor.2. To explore the histological features, risk classification and DOG1, IGF1R, CD117, CD34, SMA, Desmin, S-100protein in immunohistochemical examination of gastrointestinal stromal tumor, and to analyze the difference of each protein expression among clinicopathological factors. Especially focusing on DOG1and IGF1R protein to provide the basis for DOG1protein as a new diagnostic marker and IGF1R protein as a new therapeutic target.3. To analysis clinicopathological factors affecting the prognosis of patients with gastrointestinal stromal tumor. To find accurate and scientific indicators reflecting malignant behavior and prognosis of gastrointestinal stromal tumor.Materials and Methods:1. The clinicopathological and follow-up data of212patients between April2004to August2011in Nanfang Hospital were analyzed retrospectively. Factors included gender, age, symptoms, tumor location, size, risk grade, histological type, metastasis, treatment, survival time and so on.2.212cases of GIST and40cases of non-GIST mesenchymal tumor tissue were examined by immunohistochemisty to detect DOG1, IGF1R, CD117, CD34, SMA, S-100and Desmin protein. The application of chi-square test analyzed the differences among each protein in different clinicopathological factors.3. To follow-up212cases of patients with GIST by telephone or letter. Using SPSS18.0software for statistical analysis. Adopting univariate and multivariate survival analysis statistical method to investigate clinicopathological factors influencing on the prognosis of patients with GIST.Results:Among212cases of patients with GIST, male slightly more than female, age of onset was mostly more than50years old. The tumors occurred in each parts of digestive tract, mainly in stomach and small intestine.111(52.4%) patients located in stomach,75(35.4%) in small intestine,13(6.1%) in colorectal,6(2.8%) in esophagus and7(3.3%)in other sites. Gastric GIST mostly located in gastric fundus and body.25(11.8%) cases of patients had definite distant metastasis at the time of diagnosis.The most site of metastasis is the liver. The tumor sizes ranged from1cm to30cm, with a median size of6cm.The tumor sizes were<2cm in diameter in19cases (9.0%),2～5cm in104(49.1%),5～10cm in56(26.4%) and>10cm in33(15.6%). The clinical symptoms were nonspecific. The clinical presentations were mainly gastrointestinal bleeding in36.3%of all patients and abdominal discomfort in45.3%. Complete resection was operated in191(90.1%), palliative resection in16(7.5%).5(2.6%) patients hadn’t underwent surgical resection due to diffuse metastasis or mass invasion. The proportion of tumor diameter>10cm, mitotic count>5/50HPF, high risk group, distant metastasis in palliative surgical or non-surgical group were greater than in complete surgical resection group. Statistical analysis revealed that the differences between two groups in different tumor size, mitotic count, NIH risk classification and metastasis were statistically significant,the value of x2was56.723,22.506,26.495,164.920respectively; the value of P was<0.001,<0.001,<0.001,<0.001seperately.53cases of metastatic, recrudescent, high-risk patients with GIST had accepted mesylatc imatinib targeted therapy, postoperative adjuvant therapy for41(77.4%),first-line palliative treatment for12(22.6%). The application of mesylate imatinib between adjuvant therapy and palliative treatment group was statistically significant(P<0.05).The time of targeted therapy was1～65months, with a median time of12months.40(75.5%) patients had taken this medicine for a year or more than a year. The side effects of imatinib during taking period mainly were edema (35.8%) and gastrointestinal disorder (13.2%).Adverse reactions are mostly mild to moderate.The general forms of GIST were mainly round (58.0%) and oval (35.3%) the change of hemorrhage, necrosis, cystic degeneration could been seen in the lesions. The spindle cell type was the most common morphology that was found in203(95.8%) cases, followed by epithelial cell type in7(3.3%) and mixed type in2(0.9%). The median mitotic count was7/50HPF,58.0%less than5/50HPF,42,0%greater than or equal to5/50HPF. Biological risk grading of GIST was based on National Institutes of Health (NIH) risk criterion which was improved in2008. there were22(10.4%) patients of very low risk,73(34.4%) of low risk,35(16.5%) of intermediate risk and82(38.7%) of high risk.The positive expression rates of DOG1,CD117,CD34,IGF1R,SMA,Desmin and S-100protein in212GIST tissue were91.2%,94.8%,80.2%,6.1%,453%,3.3%and7.1%respectively. DOG1and CD117protein positive signal was mainly localized in cytoplasm, showing brown particles diffuse distribution. The expressions of SMA, Desmin and S-100protein in GIST tissues mostly scattered punctate or focal positive expression in the cytoplasm. The positive expression rates of DOG1,CD117, CD34,IGF1R,SMA,Desmin and S-100protein in40non-GIST mesenchymal tumor tissue were为5.0%、12.5%、10.0%、2.5%、52.5%、45%and50%respectively.92.5%of GIST were positive for the expression of DOG1and CD117protein,4.7%were all negative for the expression of DOG1and CD117protein.77.8%were positive expression of DOG1and CD34protein for the same time.4.7%were negative for DOG1and CD34protein expression. Among11cases of GIST tissues expressing CD117protein negatively,90.0%were positive for the expression of DOG1protein. Among6cases of GIST tissues expressing DOG1protein negatively,83.8%were positive for the expression of CD117protein. Among42cases of GIST tissues expressing CD34protein negatively,97.6%were positive for the expression of DOG1protein. Among6cases of GIST tissues expressing DOG1protein negatively,83.3%were positive for the expression of CD34protein. Among11cases of GIST tissues expressing CD117protein negatively,72.4%were positive for the expression of IGF1R protein. Among42cases of GIST tissues expressing CD34protein negatively, only4.8%were positive for the expression of IGF1R protein. No significant difference between the expression rates of DOG1,CD117and CD34protein (P>0.05). The difference between the expression rates of IGF1R and CD117protein was statistically significant (χ2=89.391, df=1, P<0.001). The expression rates of IGF1R and CD34protein did not differ significantly(x2=0.171, df=1, P=0.505).The expressions of DOG1,IGF1R, CD117, SMA, Desmin and S-100protein were not correlated with gender, age, location, size, NIH risk classification, cell type and other clinicopathological factors(P>0.05).Expressions of CD34protein in different gender, mitotic number and metastasis were different significantly, the value of χ2was4.691,4.943and4.676respectively; the value of P was0.024,0.026and0.031separetely, in other clinicopathological factors were not different significantly (P>0.05)212cases of GIST patients had not been lost to follow-up. The time of follow-up was from2to89months, the median time was20months. The survival rates of1,2,3year were85%、50%、and38%respectively. Univariate survival analysis showed that the3-year survival rates and prognosis of the212GIST patients were correlated with tumor size, NIH risk grade, mitotic count, metastasis, surgical resection and imatinib treatment (P<0.05). Multivariate survival analysis confirmed the independent prognostic role of imatinib treatment (P<0.001).The survival rates of3year in tumor diameter<2cm、≧2-5cm、>5-10cm and>10cm groups were80.0%、61.0%、46.0%and18.0%respectively, with a statistically significant difference(χ2=21.968,P<0.001).3-year survival rates in mitotic count<5/50HPF and>5/50HPF groups were69.0%and33.0%respectively, with a statistically significant difference (χ2=13.531,P<0.001).3-year survival rates in GIST patients with or without distant metastasis were31.0%and54.0%respectively, with a statistically significant difference (χ2=4.658,P=0.031).3-year survival rates in GIST patients with complete surgical resection or with palliative surgery/non-surgical were56.0%and20.0%respectively, with a statistically significant difference (χ2=7.868,P0.005).3-year survival rates in patients with mesylate imatinib targeted therapy was66.0%, with the median survival time of84.00months.3-year survival rates in patients without mesylate imatinib targeted therapy was49.0%, with the median survival time of56.94months. The differences between the two groups were significantly different (χ2=6.708,P=0.010).GIST patients with very low, low, intermediate, high risk grade in our study had83.0%,76.0%,59.0%and27.0%in term of3-year survival rates separately, univariate analysis suggested statistically significant difference between each risk groups (χ2=24.387,P<0.001). Among53cases of patients treated with mesylate imatinib, the number of patients treated by imatinib in very low, low, moderate and high risk group was1(1.9%),11(20.8%),8(15.1%),and33(62.3%) cases respectively. High-risk group accounted for the highest proportion of patients undergoing mesylate imatinib. The application of mesylate imatinib in different NIH risk groups was statistically significantly different (χ2=18.998, P<0.001). Biological risk grading of GIST patients treated with mesylate imatinib was in the light of NIH risk criterion. Among these patients,3-year survival rates in very low risk group, low-risk group, moderate risk group and high-risk group were100%,100%,100%and58.0%respectively.3-year survival rates of patients without the treatment of mesylate imatinib were83.0%,73.0%,49.0%and8.0%respectively. The difference of survival rates in the corresponding each risk groups was the highest. Univariate analysis suggested that the survival differences between patients with or without the treatement fo mesylate imatinib in different NIH risk grade were statistically significant (χ2=45.919,P<0.001).Conclusion:The clinicopathological characteristics of Chinese gastrointestinal stromal tumor correspond to the foreign gastrointestinal stromal tumor.1. Immunohistochemical detection of CD117,CD34,SMA,Desmin and S-100protein was beneficial to assistant diagnosis and differential diagnosis.2. DOG1protein has high sensitivity and high specificity in the diagnosis of gastrointestinal stromal tumor. Especially, for CD117protein negative GIST, DOG1protein can play a complementary role and reduce the rate of misdiagnosis. The application of DOG1protein detected by immunohistochemisty can been promoted in diagnosis process of GIST.3. IGF1R protein mostly express in the CD117protein negative GIST and lowly express in the overall GIST. The application of IGF1R protein as a new treatment target need further exploration research.4. GIST patients with small tumor diameter, low mitotic count, low-risk and non-metastasis have a greater chance of undergoing complete surgical resection. To strive for early radical surgery as far as possible can improve the prognosis of patients with GIST.5. Mesylate imatinib significantly improve the prognosis of patients with GIST, especially for GIST patients in the medium and high risk grade.6. It is scientific, reasonable and practicable to adopt NIH risk classification criterion to judge the biological behavior and clinical prognosis of GIST and to guide treatment.