Vitreous fibrovascular membranes (FVMs), the hallmark of proliferative diabetic retinopathy (PDR), cause retinal hemorrhage, detachment and eventually blindness. However, little is known about the pathophysiology of FVM. In this study, we employed single-cell RNA sequencing on surgically harvested PDR-FVMs and generated a comprehensive cell atlas of FVM. A total of 8 cellular compositions were identified, with microglia as the major cell population. We identified a GPNMB+ subpopulation of microglia, which presented both profibrotic and fibrogenic properties. Pseudotime analysis further revealed the profibrotic microglia was uniquely differentiated from retina-resident microglia and expanded in PDR setting. Ligand-receptor interactions between the profibrotic microglia and cytokines upregulated in PDR vitreous implicated the involvement of several pathways, including CCR5, IFNGR1 and CD44 signaling, in the microglial activation within PDR microenvironment. Collectively, our description of the novel microglia phenotypes in PDR-FVM may offer new insight into the cellular and molecular mechanism underlying the pathogenesis of DR, as well as potential signaling pathways amenable to disease-specific intervention.

Parathyroid hormone (1-34, PTH) combined β-tricalcium phosphate (β-TCP) achieves stable bone regeneration without cell transplantation in previous studies. Recently, with the development of tissue engineering slow release technology, PTH used locally to promote bone defect healing become possible. This study by virtue of collagen with a combination of drugs and has a slow release properties, and investigated bone regeneration by β-TCP/collagen (β-TCP/COL) with the single local administration of PTH. After the creation of a rodent critical-sized femoral metaphyseal bone defect, β-TCP/COL was prepared by mixing sieved granules of β-TCP and atelocollagen for medical use, then β-TCP/COL with dripped PTH solution (1.0 µg) was implanted into the defect of OVX rats until death at 4 and 8 weeks. The defected area in distal femurs of rats was harvested for evaluation by histology, micro-CT, and biomechanics. The results of our study show that single-dose local administration of PTH combined local usage of β-TCP/COL can increase the healing of defects in OVX rats. Furthermore, treatments with single-dose local administration of PTH and β-TCP/COL showed a stronger effect on accelerating the local bone formation than β-TCP/COL used alone. The results from our study demonstrate that combination of single-dose local administration of PTH and β-TCP/COL had an additive effect on local bone formation in osteoporosis rats.

PurposeTo develop artificial intelligence (AI)-based deep learning (DL) models for automatically detecting the ischemia type and the non-perfusion area (NPA) from color fundus photographs (CFPs) of patients with branch retinal vein occlusion (BRVO).MethodsThis was a retrospective analysis of 274 CFPs from patients diagnosed with BRVO. All DL models were trained using a deep convolutional neural network (CNN) based on 45 degree CFPs covering the fovea and the optic disk. We first trained a DL algorithm to identify BRVO patients with or without the necessity of retinal photocoagulation from 219 CFPs and validated the algorithm on 55 CFPs. Next, we trained another DL algorithm to segment NPA from 104 CFPs and validated it on 29 CFPs, in which the NPA was manually delineated by 3 experienced ophthalmologists according to fundus fluorescein angiography. Both DL models have been cross-validated 5-fold. The recall, precision, accuracy, and area under the curve (AUC) were used to evaluate the DL models in comparison with three types of independent ophthalmologists of different seniority.ResultsIn the first DL model, the recall, precision, accuracy, and area under the curve (AUC) were 0.75 ± 0.08, 0.80 ± 0.07, 0.79 ± 0.02, and 0.82 ± 0.03, respectively, for predicting the necessity of laser photocoagulation for BRVO CFPs. The second DL model was able to segment NPA in CFPs of BRVO with an AUC of 0.96 ± 0.02. The recall, precision, and accuracy for segmenting NPA was 0.74 ± 0.05, 0.87 ± 0.02, and 0.89 ± 0.02, respectively. The performance of the second DL model was nearly comparable with the senior doctors and significantly better than the residents.ConclusionThese results indicate that the DL models can directly identify and segment retinal NPA from the CFPs of patients with BRVO, which can further guide laser photocoagulation. Further research is needed to identify NPA of the peripheral retina in BRVO, or other diseases, such as diabetic retinopathy.