Objective To establish a rat model of blood ocular barrier breakdown induced by anterior segment intraocular analogic surgery. Methods One hundred and fifty healthy adult male rats were randomly divided into control group and model group,75 rats in each group.The rats were anesthetized with 1 ml/kg ketamine hydrochloride/xylazine hydrochloride solution.Three way pipes were attached to a phosphate buffer infusion bag and two intravenous catheters. One catheter was inserted 30° obliquelythrough the transparent cornea anterior to the limbus into the rat's anterior chamber.Then the needle was withdrawn and the sheath was indwelling.Another catheter was connected with a manometer.Intraocular pressure was varied from 0 to 12 mm Hg (1 mm Hg=0.133 kPa) 60 times,30 times per min.The catheter was removed.The eyes were treated with ofloxacin ophthalmic solution after surgery.The 1st,2nd,3rd,5th and 7th day after surgery,the integrity of the blood ocular barrier was assessed by immunohistochemical staining for albumin and quantitative measurement using Evan's blue as a tracer. Results Albumin immunohistochemical staining of the control group was confined to the iris and retinal blood vessels.The choroid was stained at each time point after surgery.Albumin immunohistochemical staining of the model group was abundant around the iris and the retinal vasculature on the 1st day after surgery.The albumin diffused throughout the iris and the retina on the 2nd and the 3rd day after surgery.The albumin reached the retinal vessels on the 5th and 7th day after surgery.The aqueous humor Evans blue leakages of the model group were higher than those of the control group on the 1st,2nd,3rd and 5th day after surgery.The differences were statistically significant (t=25.781,37.433,25.150,19.171; P＜0.01).The Evans blue leakage of the model group was close to that of the control group on the 7th day after surgery. The difference was no statistical significant(t=1.303,P=0.209).The retinal Evans blue leakages of the model group were higher than those of the control group on the 1st,the 2nd and the 3rd day after surgery.The differences were statistically significant (t=11.997,14.622,23.014; P＜0.01).The Evans blue leakage of the model group was close to those of the control group on the 5th and 7th day after surgery. The differences were not statistically significant(t=2.027,0.756 ； P=0.058,0.459).Conclusion This study establishes a rat model of blood ocular barrier breakdown induced by imitating the injury to the anterior segment during intraocular surgery.

Objective To explore the value of combined detection with MMP-9 and uPA in the progno-sis of pancreatic carcinoma. Methods By immunohistochemistry PV methods,the expression of MMP-9 and uPA was respectively studied in 63 surgical specimens of primary pancreatic carcinoma and the survival time of patients with pancreatic carcinoma was analysed. Results The expressions of MMP-9 and uPA were positively related(r=0. 573,P=0. 000). The expression of MMP-9 and uPA significantly correlated with differentiation (r= -0. 271,P=0. 032;r= -0. 333,P=0. 008),TNM stages(r= -0. 449,P=0. 000;r= -0. 430,P=0. 000)and lymph node metastasis(r=0. 329,P=0. 009;r=0. 400,P=0. 001),separately. The expression of MMP-9 had also a significant correlation with tumer size(r= -0. 297,P=0. 018)and distant metastasis(r=0. 320,P=0. 011). Univariate analysis identified that tumor size(χ2 =8. 766,P=0. 012),differentiation(χ2 =29. 050,P=0. 000),clinical stage(χ2 =24. 940,P=0. 000),distant metastasis(χ2 =12. 846,P=0. 000), lymph node metastasis(χ2 =15. 457,P=0. 000),MMP-9(χ2 =32. 700,P=0. 000)and uPA(χ2 =41. 495,P=0. 000)were significantly associated with prognosis. Kaplan-Meier survival analysis showed that 1-year survival rate of patients with MMP-9 ( -),uPA ( -)were significantly longer than that of the patients with MMP-9( ﹢),uPA( ﹢),respectively(χ2 =32. 700,P=0. 000;χ2 =41. 495,P=0. 000);1-year survival rate of patients with MMP-9( -)/uPA( -)was significantly longer than the others( Log-rank test,χ2 = 54. 892, P=0. 000). COX regression revealed that differentiation(RR=2. 315,P=0. 004),clinical stage(RR=1. 694, P=0. 002),MMP-9(RR=0. 165,P=0. 000)and uPA(RR=0. 244,P=0. 007)was independent prognostic factors in pancreatic carcinoma. Conclusion They may have a synergistic function in the the process of growth and invasion in pancreatic cancer between MMP-9 and uPA,and the posssible mechanism is that uPA activate degradation of MMP-9,which is not favorable to prognosis. Combined analysis of MMP-9 and uPA may lead to a more reliable prognostic estimation,as the beneficial supplement of the differentiation,and clinical stage to judge the prognosis of pancreatic cancer.