Factors predicting the treatment frequency of ranibizumab injections during the second year in diabetic macular edema

Purpose To investigate the predictors of annual treatment frequency in the second year of pro re nata (PRN) intravitreal ranibizumab (IVR) injections for diabetic macular edema (DME). Study design A retrospective study. Methods We reviewed 65 eyes of 60 patients with center-involved DME who received PRN IVR injections after 3 monthly loading doses. The central subfield thickness (CST) and qualitative findings were assessed on the spectral domain optical coherence tomography (SD-OCT) images. We then investigated whether the parameters at the baseline or at the 12-month visit were associated with treatment frequency in the second year. Results The number of ranibizumab injections decreased from 6 (4–8) during the first year to 2 (0–3) during the second year ( P < .001). The injection numbers during the first year ( ρ = 0.259, P = .037) but not during the second year ( ρ = 0.152, P = .226) were modestly related to the logarithm of the minimum angle of resolution (logMAR) improvement at 24 months. Multivariate analyses revealed that the CST ( β = 0.336, P = .005) and hyperreflective walls in the foveal cystoid spaces ( β = 0.273, P = .020) at baseline were associated with the number of IVR injections during the second year. The treatment frequency during the second year was also related to the CST ( β = 0.266, P = .012), hyperreflective walls ( β = 0.394, P = .002), and cumulative doses of ranibizumab injections ( β = 0.294, P = .006) at the 12-month visit. Conclusions The cumulative doses of ranibizumab injections, CST, and hyperreflective walls in the foveal cystoid spaces at the 12-month visit are designated predictors of the treatment frequency of ranibizumab injections during the second year in DME.