Abstract:AIM: To appraise the effect of treatment for diabetic macular edema (DME) in proliferative stage with sufficient panrentinal photocoagulation (PRP) therapy and intravitreal injections (IV) Conbercept and posterior subtenon’s triamcinolone acetonide (STTA) sequential therapy. METHODS: This prospective clinical randomized controlled trial of cross-over design was conducted in three phases. The participants included cases of DME in proliferative stage. They were divided into two groups and treated with PRP before enrollment. Group A were treated with IV-Conbercept 0.5 mg for one month in the 1st phase. Group B were treated with STTA 40 mg (twice per two weeks). The interventions were exchanged in the second phase (2mo) between the two groups. In the third phase (3-6mo) no other treatment was given. Best corrected visual acuity (BCVA), central macular thickness (CMT) measured by OCT and complications were compared. RESULTS: After phase I: in Group A, BCVA improved from 0.201±0.17 to 0.37±0.24 (F=5.88, P=0.004). CMT changed from 449±155.10 to 304.1±84.70 μm (F=14.9, P<0.01). In Group B, BCVA changed from 0.195±0.19 to 0.26±0.20 (F=0.76, P=0.41) while CMT changed from 463.82±152.92 to 366.00±115.40 μm (F=3.70, P<0.03). The improvement of BCVA was better in Group A (P<0.05). After phase II: in Group A, BCVA raised to 0.47±0.27 (F=0.26, P<0.01), CMT reduced to 260.67±62.97 μm (F=-188.3, P<0.01); in Group B, BCVA raised to 0.51±0.26 (F=0.31, P<0.01), CMT reduced to 261.93±50.15 μm (F=-201.9, P<0.01). But there were no difference between two groups (P>0.05). After phase III: in Group A, BCVA maintained 0.42±0.25 (F=0.22, P=0.001), CMT maintained 267.8±58.34 μm, (F=-0.27, P<0.01); in Group B, BCVA was 0.47±0.25 (F=-0.27, P<0.01), CMT was 272.71±49.16 μm (F=-191.1, P<0.01). No serious complications happened in all phases. CONCLUSION: PRP+Conbercept is better than PRP+STTA in DME with proliferative stage but PRP+Conbercept+STTA sequential therapy may be a wiser choice for persistent effectiveness on anatomical as well as functional status.