不同程度糖尿病视网膜病变患者黄斑区血流的变化及意义:基于OCTA的评价

目的　光学相干断层扫描血管成像（OCTA）评价不同程度糖尿病视网膜病变（DR）患者黄斑区血流的变化及意义。方法　收集2018年8月至2020年8月在本院就诊的93例120眼2型糖尿病患者资料，其中无DR(NDR组)患者30例40眼、轻度非增生型DR（轻度NPDR组）患者32例40眼、中度NPDR（中度NPDR组）患者31例40眼；另选取同期在本院接受健康体检的正常体检者40名40眼为对照组。比较4组受试者浅层毛细血管丛（SCP）血流密度、深层毛细血管丛（DCP）血流密度、黄斑中心凹无血管区（FAZ）旁300 μm区（FD300）血流密度、FAZ面积、FAZ周长、非圆度指数（acircularity index,AI）；多元回归分析SCP血流密度、DCP血流密度、FD300血流密度、FAZ面积、FAZ周长、AI与DR病情的关系；绘制受试者工作特征（ROC）曲线分析OCTA检查指标对早期DR的预测效能。结果　与对照组比较，NDR组、DR组患者SCP血流密度、DCP血流密度均明显下降，DR组患者FD300血流密度也明显下降，FAZ周长、AI均明显增加（均为P<0.05）；且DR组患者SCP血流密度、DCP血流密度、FD300血流密度均低于NDR组，FAZ周长、AI均大于NDR组（均为P<0.05）；DR患者黄斑区SCP血流密度、DCP血流密度、FD300血流密度随DR分期增加而下降，FAZ周长、AI随DR分期增加而上升（均为P<0.05）。多元回归分析结果显示，SCP血流密度、DCP血流密度、FD300血流密度与DR病情严重程度均呈负相关，与FAZ周长、AI均呈显著正相关（均为P<0.05）；且各变量间不存在共线性。FD300血流密度预测早期DR的曲线下面积（AUC）最高（0.794），敏感度、特异度分别为83.75%、66.25%；以AI变量的cut-off值为1.12，其预测早期DR的特异度（86.25%）最佳，敏感度为72.50%。结论　早期DR患者存在黄斑区FD300血流密度下降及FAZ周长、AI增加，上述OCTA参数与DR分期密切相关，可作为早期DR预测指标。

Changes and significance of macular blood flow in patients with diabetic retinopathy of different severity:based on optical coherence tomography angiography

Objective To evaluate the changes and significance of macular blood flow in patients with diabetic retinopathy(DR)of different severity based on optical coherence tomography angiography(OCTA).Methods From August 2018 to August 2020,clinical data of 93 patients(120 eyes)were collected,30 patients(40 eyes)with non-diabetic retinopathy(NDR)(NDR group),32 patients(40 eyes)with mild non-proliferative dibetic retinopthy(NPDR)(mild-NPDR group),and 31 patients(40 eyes)with moderate NPDR(moderate NPDR group)who were treated in the hospital were enrolled in this study.Meanwhile,40 normal people(40 eyes)were recruited as the control group.All subjects received RTVue-XR OCTA.The blood flow density of superficial capillary plexus(SCP),deep capillary plexus(DCP),and foveal avascular area(FAZ)within a 300μm width(FD300),FAZ area,perimeter,and acircularity index(AI)were compared among the 4 groups.Multivariate regression analysis was applied to detect the correlation of SCP blood flow density,DCP blood flow density,FD300 blood flow density,FAZ area,perimeter,AI with the severity of DR.Receiver operating characteristic(ROC)curves were used to analyze the predictive power of OCTA indicators for early DR.Results Compared with the control group,SCP blood flow density and DCP blood flow density were significantly reduced in NDR group,while SCP blood flow density,DCP blood flow density and FD300 blood flow density were significantly reduced,while FAZ perimeter and AI were significantly increased in DR group(all P<0.05).SCP blood flow density,DCP blood flow density and FD300 blood flow density of DR group were significantly lower than those in NDR group,FAZ perimeter and AI were significantly larger than those in NDR group(all P<0.05).With the increase of DR stage,SCP blood flow density,DCP blood flow density,FD300 blood flow density in the macular area of DR patients decreased,while FAZ perimeter and AI increased(all P<0.05).Multiple regression analysis showed that SCP blood flow density,DCP blood flow density,and FD300 blood flow density were negatively correlated with the severity of DR,but significantly positively correlated with FAZ perimeter and AI(all P<0.05),without collinearity among the variables.The area under the curve(AUC)value of FD300 blood flow density was the highest(0.794)for predicting early DR,with a sensitivity and a specificity of 83.75%and 66.25%,respectively.With 1.12 as the cut-off value of AI,the specificity for predicting early DR was the best(86.25%),with a sensitivity of 72.50%.Conclusion Patients with early DR have decreased macular FD300 blood flow density as well as changes in FAZ perimeter and AI.The above OCTA parameters not only are closely related to DR staging but also can be used as predictors of early DR.