[Minimally invasive vitrectomy for the treatment of severe proliferative diabetic retinopathy].

To report the efficacy and safety of minimally invasive vitrectomy for the treatment of severe proliferative diabetic retinopathy (PDR) and the effect of preoperative retinal photocoagulation on prognosis. Case-series study and cohort study. This study included 48 severe PDR patients (53 eyes). There are 28 males and 20 females. The average age was 53.5 (range, 40.0 to 59.0) years old. Patients were divided into two groups according to whether preoperative retinal photocoagulation was performed. Under the guidance of the concept of minimally invasive vitrectomy, all patients received intravitreal injection of conbercept 3 days before standard pars plana vitrectomy with a 27G+ vitrectomy system. To relieve traction, the proliferative fibrovascular membrane was divided into islands as small as possible. It was not necessary to pursue a complete removal of the proliferative membrane. The 27G+ vitrector was flexibly applied as a multifunctional tool for membrane removal by reducing frequencies at which the device entered and left the eye. Intraocular retinal photocoagulation was performed in the main area. The primary outcome measures were best corrected visual acuity (BCVA) and retinal reattachment rate, and the secondary outcome measures were intraoperative and postoperative complications. Statistical analysis was performed using test, rank sum test and χ² test. All patients tolerated intravitreal conbercept, with no serious intraoperative or postoperative adverse events. Postoperative BCVA values were improved significantly compared with preoperative values (χ²=125.11, <0.01). The mean logMAR BCVA improved from 1.90 (1.30, 2.30) preoperatively to 1.00 (0.70, 1.90) at 1 week postoperatively, 0.8 (0.5, 1.3) at 1 month postoperatively, 0.7 (0.40, 1.20) at 3 months postoperatively, and 0.70 (0.40, 1.20) at 6 months postoperatively. The visual function increased progressively with time. Postoperatively, the primary and final reattachment rates were 92.5% (50/53) and 96.2% (51/53), respectively. Abnormal intraocular pressure lasted for more than one week occurred in 2 eyes; Vitreous hemorrhage recurred in 5 eyes; Retinal detachment occurred in 4 eyes (7.5%); No postoperative endophthalmitis, choroidal detachment or incision related retinal hole occurred. The intraoperative and postoperative parameters in the preoperative retinal photocoagulation group were better than the preoperative non-photocoagulation group, but the difference was not significant (>0.05). Minimally invasive vitrectomy is fully qualified for the management of severe PDR, with maximized benefits. .