Objective: This study investigated changes of plasma cytokines and aimed to build a dynamic nomogram for diabetic macular edema (DME) in type 2 diabetes mellitus (T2DM).

Methods: In a pilot cohort, plasma samples were selected from 9 T2DM patients and 9 DME patients to screen for cytokine differences. The screening cytokines were then validated by enzyme-linked immunoassay in a cohort, which contained 100 DME (DME group) and 100 T2DM patients without DME (T2DM group). A dynamic nomogram for predicting DME was developed, based on the plasma cytokines.

Results: In the pilot cohort, 11 plasma cytokines were significantly increased in the DME group. In the validation cohort, platelet-derived growth factor (PDGF)-BB, tissue inhibitors of metalloproteinase (TIMP)-1, angiopoietin (ANG-1), and vascular endothelial cell growth factor receptor (VEGFR)-2 were confirmed to be significantly elevated in the DME group. The dynamic nomogram demonstrated good calibration and discrimination, with an area under the receiver operating characteristic curve (AUC) of 0.88. In the test set, sensitivity, specificity, and AUC were 73.3%, 80.0%, and 0.84, respectively.

Conclusion: Plasma cytokines were closely associated with DME. A novel dynamic monogram including ANG-1, PDGF-BB, TIMP-1, and VEGFR2 was a novel tool for predicting DME.