A three-year follow-up of ranibizumab treatment of exudative AMD: impact on the outcome of carrying forward the last acuity observation in drop-outs.

Purpose: To analyse a 3-year clinical patient cohort of ranibizumab treatment of exudative age-related macular degeneration (AMD), to investigate the impact on visual outcome of carrying forward the last acuity observation in drop-outs and to explore possible differences between the early and the late phase of the study.

Methods: A retrospective study of 312 eyes with neovascular AMD. The patients were followed up monthly, received three initial monthly injections of 0.5 mg ranibizumab and were re-treated pro re nata (PRN). Time-domain optical coherence tomography (TD-OCT) was used until spectral-domain (SD)-OCT was introduced during the last year of enrolment. Sixty-five patients were discontinued from the study.

Primary Outcome: change in best corrected visual acuity (BCVA).

Results: Best corrected visual acuity was 58.4 (CI 56.9-59.9) ETDRS (Early Treatment Diabetic Retinopathy Study) letters. At three months, it had increased by 4.1 letters (p=0.0004), at 12 months by 1.8 letters, at 24 months by 1.0 letter and at 36 months by 0.1 letter. However, if the last available acuity of drop-outs was carried forward one step and included, acuity had increased by 3.9 letters at 3 months (p<0.0001) and by 1.0 letter at 12 months but had decreased by 3.8 letters at 24 months (p=0.019) and by 4.1 letters (p=0.003) at 36 months. At 24 months, the result was significantly (p=0.030) less favourable when drop-outs were included. In patients enrolled during the late phase, BCVA was 59.3 (CI 56.7-62.0). It had increased by 5.7 letters (p<0.0001) at three months and by 5.8 letters at 12 months (p=0.0016). In patients enrolled during the early phase, BCVA was 57.9 (CI 55.0-60.8). At three months, it had increased by 3.5 letters (p=0.0008), but at 12 months, it had decreased by 2.3 letters (ns). The result at 12 months was significantly (p=0.0033) better for the late than for the early phase. The number of injections was also significantly (p=0.011) higher in the late phase. Adverse events were similar to those in earlier clinical trials.

Conclusions: The results of this 3-year cohort showed that the initial average acuity could be maintained over 36 months, which was comparable to those of many other clinical cohorts. However, if the last available acuity of drop-outs was carried forward one step and included, the acuity figures would have fallen significantly. The results in patients enrolled during the late phase of the study were fairly similar to those in clinical trials.