摘要目的：:利用结构拼图观察糖尿病（DM）患者全视网膜光凝（PRP）前后角膜上皮基底神经丛（SNP）和朗格汉斯细胞（LC）的变化，并分析二者的相关性。方法：:前瞻性临床研究。选取2019年4─11月就诊于山西省眼科医院准备行PRP治疗且双眼糖尿病视网膜病变Ⅳ期的2型DM患者，选择病情较重的眼为治疗眼，对侧眼为对照眼，分别于PRP治疗前、每次光凝后1周和PRP完成后1个月行角膜共焦显微镜检查，观察涡状结构及其周围2～3 mm区域SNP和LC的变化，并测量涡状区神经纤维长度（NFL）值和LC密度。采用重复测量方差分析比较不同观察时间点LC密度和NFL值，并采用SAS软件的MIXED模型分析重复测量的NFL值和LC密度之间的相关性。结果：:共纳入患者49例。治疗眼接受PRP后部分患者出现SNP神经纤维变细，伴有不同程度的涡状区神经结构缺失的神经损伤表现；各观察时间点NFL值总体比较差异有统计学意义（ F=8.039， P=0.004），且PRP治疗前NFL值[（15.5±3.7）mm/mm 2]与第2次光凝后1周[（15.0±3.5）mm/mm 2]、第3次光凝后1周[（13.4±4.3）mm/mm 2]和第4次光凝后1周[（13.5±4.1）mm/mm 2]比较差异均有统计学意义（均 P<0.05）。同时，治疗眼LC密度增加，并以涡状区为中心聚集，成熟LC浸润区伴有SNP神经结构的缺失；各观察时间点LC密度总体比较差异有统计学意义（ F=12.350， P<0.001），且PRP治疗前LC密度[（40±54）cells/mm 2]与第3次光凝后1周[（79±91）cells/mm 2]、第4次光凝后1周[（98±126）cells/mm 2]以及PRP完成后1个月[（87±102）cells/mm 2]比较差异均具有统计学意义（均 P<0.05）；相关分析显示治疗眼第4次激光后1周LC密度与其基线水平呈正相关（ r=0.674， P<0.001）；且重复测量的NFL值与LC密度呈负相关（ =-0.041）。 结论：:PRP多次光凝可以导致LC密度增加；成熟LC可以导致SNP神经结构破坏。

abstractsObjective::To observe changes of corneal subbasal nerve plexus and langerhans cells in diabetic retinopathy patients after panretinal photocoagulation (PRP) based on the wide-field mosaic, analyze the their correlation.Methods::In this prospective clinical study, type 2 diabetic patients with binocular diabetic retinopathy stage Ⅳ and waited for PRP treatment in Shanxi Eye Hospital from April to November in 2019 were included. Their severe eyes were chosen as the treatment eye and the contralateral eyes were chosen as the control eye. Corneal confocal microscopy (CCM) was performed before PRP treatment, 1 week after every photocoagulation for 4 times, and 1 month after the completion of PRP to examine the changes of SNP and Langerhans cell (LC) over an area of 2-3 mm around the whorl-like complex, and measure the of NFL value and LC density. The NFL value and LC density among each observation time point were compared by repeated measurement ANOVA, and the correlation between repeated measurement of NFL value and LC density before and after PRP treatment was analyze by MIXED model of SAS software.Results::A total of 49 patients were included. After PRP treatment, SNP nerve fibers showed decrease in diameter, accompanied by different degrees of nerve architecture loss in the whorl-like region in the treatment eye of some patients; the overall comparison of NFL values among observation time points was statistically significant ( F=8.039, P=0.004), among which the NFL value differences between pre-PRP treatment [(15.5±3.7)mm/mm 2] and 1 week after the second photocoagulation [(15.0±3.5)mm/mm 2], 1 week after the third photocoagulation [(13.4±4.3)mm/mm 2], and 1 week after the fourth photocoagulation [(13.5±4.1)mm/mm 2] were statistically significant (all P<0.05). Meanwhile, LC was increased, clustering around the whorl-like area; the mature LC infiltration area also were accompanied nerve architecture loss. The overall difference in LC density among observation time points was statistically significant ( F=12.350, P<0.001), among which the LC density differences between pre-PRP treatment [(40±54)cells/mm 2] and 1 week after the third photocoagulation [(79±91)cells/mm 2], 1 week after the forth photocoagulation [(98±126)cells/mm 2] and 1 month after PRP [(87±102)cells/mm 2] was statistically significant (all P<0.05); Correlation analysis showed that the LC density 1 week after the fourth photocoagulation were significantly positively correlated with their corresponding baseline levels in the treatment eye ( r=0.674, P<0.001), and there was a negative correlation between repeated measurement of NFL value and LC density ( =-0.041). Conclusions::Multiple photocoagulation of PRP treatment can lead to the increase of LC density; mature LC can lead to the damage of SNP nerve architecture.