Monitorización ultrasonográfica de la dinámica vítrea durante la administración de medicación antiangiogénicaanálisis de la difusión y efectividad de la medicación en el edema macular diabético

Résumé

Introduction. The antiangiogenic treatment effectiveness has been proven in the most prevalent macular pathology. However, the response to medication shows a high interindividual variability. Regardless of the etiology and severity of lesions, other factors could be conditioning the variety of responses to this therapy, such as vitreous state or vitreoretinal interface. Purpose To analyze the dynamics of the medication injected into the vitreous cavity, using simultaneous ultrasonography. To assess how the previous vitreous state (existence or absence of posterior vitreous detachment) and the presence of drug reflux influences it. In a second phase, we will study how these variables affect the response of diabetic macular edema treated with Ranibizumab. Finally, to obtain a predictive equation that determines retinal response to antiangiogenic medication in diabetic macular edema. Methods Patients with diabetic macular edema, age-related macular degeneration or retinal venous obstruction treated with antiangiogenic agents were selected. During the intravitreal injection, ocular ultrasound was performed to visualize the entry and subsequent diffusion of the administered medication. Patients were followed monthly, registering visual acuity, funduscopy and OCTs in a PRN pattern during, at least, 6 months. Ultrasound scans were analyzed to establish a pattern classification and assessed its relationship with vitreous states, reflux, as well as with medication effectiveness. Finally, responses to antiangiogenics were analyzed to elaborate, using multiple linear regression, a predictive equation of retinal response to intravitreal injections. Results 332 ultrasonographic sequences from 121 eyes of 95 patients were recorded elaborating a classification based of three patterns: A, B and C. Pattern A was pathognomonic in vitrectomized patients, while pattern B appeared more in patients with posterior vitreous detachment. Pattern C appeared almost as often in patients with and without vitreous detachment. Drug reflux occurred exclusively in type C pattern injections, and within these, reflux was 3 times more frequent in those with a prior posterior vitreous detachment. In the non-vitrectomized diabetic macular edema treated with Ranibizumab subsample, significant reduction of 3'181% of edema was achieved in eyes with pattern B, compared to pattern C, and reduction of 2'838% in eyes with posterior vitreous detachment, compared with eyes without it. There was no statistically significant influence of drug reflux on the effectiveness of the medication. The predictive equation of retinal response to diabetic macular edema achieved a 93% correlation and a margin of error of -1'586±31'826 microns. The factor with the greatest weight in the equation was retinal thickness before the injection. Conclusions This is the first study that records intravitreal behaviour of antiangiogenic medication, through simultaneous ultrasonography, establishing a three-well-differentiated-patterns classification. The analyzed vitreous states (vitrectomized eyes, eyes without posterior vitreous detachment and eyes with posterior vitreous detachment), the ultrasonographic patterns (A, B and C) and the presence of drug reflux are interrelated. Finally, in the retinal response analysis, ultrasonographic patterns and vitreous states both influence retinal response to antiangiogenics. However, drug reflux was not determinant in effectiveness. The developed predictive equation is the first one of its kind that takes first steps to anticipate natural evolution of diabetic macular edema.