Ziegler Dan, Winter Karsten, Strom Alexander, Zhivov Andrey, Allgeier Stephan, Papanas Nikolaos, Ziegler Iris, Brüggemann Jutta, Ringel Bernd, Peschel Sabine, Köhler Bernd, Stachs Oliver, Guthoff Rudolf F, Roden Michael
Institute for Clinical Diabetology, German Diabetes Center at Heinrich Heine University, Leibniz Center for Diabetes Research, Düsseldorf, Germany.
Department of Endocrinology and Diabetology, Medical Faculty, Heinrich Heine University, Düsseldorf, Germany.
PLoS One. 2017 Mar 15;12(3):e0173832. doi: 10.1371/journal.pone.0173832. eCollection 2017.
Corneal confocal microscopy (CCM) has revealed reduced corneal nerve fiber (CNF) length and density (CNFL, CNFD) in patients with diabetes, but the spatial pattern of CNF loss has not been studied. We aimed to determine whether spatial analysis of the distribution of corneal nerve branching points (CNBPs) may contribute to improving the detection of early CNF loss. We hypothesized that early CNF decline follows a clustered rather than random distribution pattern of CNBPs. CCM, nerve conduction studies (NCS), and quantitative sensory testing (QST) were performed in a cross-sectional study including 86 patients recently diagnosed with type 2 diabetes and 47 control subjects. In addition to CNFL, CNFD, and branch density (CNBD), CNBPs were analyzed using spatial point pattern analysis (SPPA) including 10 indices and functional statistics. Compared to controls, patients with diabetes showed lower CNBP density and higher nearest neighbor distances, and all SPPA parameters indicated increased clustering of CNBPs (all P<0.05). SPPA parameters were abnormally increased >97.5th percentile of controls in up to 23.5% of patients. When combining an individual SPPA parameter with CNFL, ≥1 of 2 indices were >99th or <1st percentile of controls in 28.6% of patients compared to 2.1% of controls, while for the conventional CNFL/CNFD/CNBD combination the corresponding rates were 16.3% vs 2.1%. SPPA parameters correlated with CNFL and several NCS and QST indices in the controls (all P<0.001), whereas in patients with diabetes these correlations were markedly weaker or lost. In conclusion, SPPA reveals increased clustering of early CNF loss and substantially improves its detection when combined with a conventional CCM measure in patients with recently diagnosed type 2 diabetes.
角膜共焦显微镜检查（CCM）显示，糖尿病患者的角膜神经纤维（CNF）长度和密度（CNFL、CNFD）降低，但CNF丢失的空间模式尚未得到研究。我们旨在确定对角膜神经分支点（CNBP）分布进行空间分析是否有助于提高对早期CNF丢失的检测。我们假设早期CNF下降遵循CNBP的聚集而非随机分布模式。在一项横断面研究中，对86例新诊断的2型糖尿病患者和47例对照者进行了CCM、神经传导研究（NCS）和定量感觉测试（QST）。除了CNFL、CNFD和分支密度（CNBD）外，还使用包括10个指标和功能统计的空间点模式分析（SPPA）对CNBP进行了分析。与对照组相比，糖尿病患者的CNBP密度较低，最近邻距离较高，所有SPPA参数均表明CNBP的聚集增加（所有P<0.05）。高达23.5%的患者中，SPPA参数异常增加超过对照组的第97.5百分位数。当将单个SPPA参数与CNFL相结合时，28.6%的患者中有≥1个指标超过对照组的第99百分位数或低于第1百分位数，而对照组为2.1%，而对于传统的CNFL/CNFD/CNBD组合，相应的比率分别为16.3%和2.1%。在对照组中，SPPA参数与CNFL以及几个NCS和QST指标相关（所有P<0.001），而在糖尿病患者中，这些相关性明显较弱或消失。总之，SPPA揭示了早期CNF丢失的聚集增加，并且在与新诊断的2型糖尿病患者的传统CCM测量相结合时，显著提高了对其的检测。
