Diabetic macular edema (DME) is one of serious ocular complications of diabetes mellitus and is often treated by laser photocoagulation, peribulbar injection of triamcinolone acetonide (TA) and intravitreal injection of ranibizumab.However, some adverse responses occur in each approach.To seek a safe, effective and ecnomic therapy for DME is of clinical significance.

This study was to observe the safety and efficacy of post-sclera injection of TA with a self-made innovative device for DME and compare the outcome with peribulbar injection of TA and the intravitreal injection of ranibizumab.

A prospective non-randomized controlled study was performed.This study protocol was approved by Ethic Committee of Southwest Hospital of Third Military Medical University and complied with Helsinki declaration.Written informed consent was obtained from each patient before any medical treatment.Sixty eyes of 60 patients with DME were included in Southwest Hospital of Third Military Medical University from March 2013 to July 2016.The eyes were divided into post-sclera injection group, peribulbar injection group and intravitreal injection group, with 20 eyes for each group.TA at the dose of 20 mg was injected via posterior sclera with a self-made divice in the post-sclera injection group and via periphery of eyeball in the peribulbar injection group, and 0.5 mg ranibizumab was intravitreally injected in the intravitreal injection group.Best corrected visual acuity (BCVA) was examined and retinal thickness at macular area was measured by OCT in 1 month and 3 months after injection respectively.The outcome and complication were grouply compared.

The BCVA was significantly improved 1 month and 3 months after injection in comparison with before injection in the post-sclera injection group and intravitreal injection group, and BCVA in the post-sclera injection group and intravitreal injection group was superior to that in the peribulbar injection group (all at P=0.000) .No significant difference was found in post-injected BCVA between post-sclera injection group and intravitreal injection group (P=0.244, 0.397). Retinal edema at macular area was gradually disappeared in the post-sclera injection group and intravitreal injection group and that in the peribulbar injection group was still visible after injection.The retinal thickness at macula was (321.85±31.98), (382.75±39.28) and (315.75±40.43)μm at 1 month and was (311.95±32.73), (393.65±33.84) and (302.65±38.99)μm at 3 months after injection in the post-sclera injection group, peribulbar injection group and intravitreal injection group respectively, and the retinal thickness values at macula in the post-sclera injection group and intravitreal injection group were significantly lower than those in the peribulbar injection group (all at P=0.000). The decrease rate of retinal thickness was higher in the post-sclera injection group and intravitreal injection group than that in the peribulbar injection group at various time points after injection (all at P<0.01).

The efficacy and safety of post-sclera injection of TA for DME are similar to intravitreal injection of ranibizumab, which are superior to peribulbar injection of TA.

Diabetic retinopathy/complications; Macular edema/therapy; Treatment outcome; Triamcinolone acetonide; Ranibizumab; Post-sclera injection; Comparative study