AIM: To study the effect of 23G minimally invasive vitrectomy surgery assisted with intravitreal injection of ranibizumab(IVR)and panretinal photocoagulation(PRP)and combine both of them in the treatment of proliferative diabetic retinopathy(PDR).

METHODS: Retrospective control study. A total of 142 patients(169 eyes)with PDR were recruited to have 23G vitrectomy surgery and were divided into 4 groups as intravitreal injection of ranibizumab(Group A, 35 patients, 41 eyes), panretinal photocoagulation(Group B, 34 patients, 39 eyes), combine IVR and PRP(Group C, 32 patients, 38 eyes)and control group(Group D, 41 patients, 51 eyes). The average operating time, blooding during operation, the use of filler and electric coagulation, iatrogenic breaks, best corrected visual acuity in 1mo and complication were comparatively analyzed among the 4 groups.

RESULTS: The average operating time was 70.03±8.91min in Group A, 71.13±8.89min in Group B, 68.60±6.73min in Group C and 103.23±17.44min in Group D. The average operating time of A, B, C groups were significantly reduced than Group D, the difference was statistically significant(P_AD,P_BD,P_CD<0.05). The difference of iatrogenic breaks in four groups was no statistically significant(χ²=0.531, P>0.05). Electric coagulation were used in 8 eyes of Group A, in 11 eyes of Group B, in 7 eyes of Group C and in 38 eyes of Group D. The use of electric coagulation in Group A, B, C were significantly reduced than Group D, the difference was statistically significant(χ²_AD=27.499, χ²_BD=19.105, χ²_CD=27.405, P_AD,P_BD,P_CD<0.0167). But there was no significant difference in Group A, B, C(χ²=1.305, P>0.05). Filler were needed in 6 eyes of Group A, in 7 eyes of Group B, in 5 eyes of Group C and in 28 eyes of Group D, the use of filler in A, B, C groups were significantly reduced than D group, the difference was statistically significant(χ²_AD=15.818, χ²_BD=12.698, χ²_CD=18.014,P_AD、P_BD、P_CD<0.0167); but there was no difference in A,B,C groups(χ²=0.360, P>0.05). Continuous or severe bleeding during the operation in the four groups were no difference(χ²=2.809, P>0.05).The best corrected visual acuity of 1mo after surgery was 0.274±0.151, 0.175±0.079, 0.277±0.137, 0.177±0.059 respectively in the four groups; Groups A, B, C were significantly increased than Group D, the difference was statistically significant(P_AD,P_BD,P_CD<0.05). There were 5, 5, 4, 20 eyes happened vitreous hemorrhage again respectively in Group A, B, C, D of 1wk; Group A, B, C were significantly reduced than Group D, the difference was statistically significant(χ²_AD=8.385, χ²_BD=7.675, χ²_CD=9.100, P_AD, P_BD, P_CD<0.0167); but there were no difference after 1 and 3mo among four groups(χ²_1mo=2.933, χ²_3mo=2.249, P_1mo, P_3mo>0.05).

CONCLUSION:The therapy of ranibizumab and panretinal photocoagulation before vitrectomy surgery can effectively shorten the operation time, reduce the use of filler and electric coagulation and complication, effectively increase the visual acuity.