Sir,

Thank you for your interest in our article published in Eye.1

The aim of our work1 was to report real-world outcomes using Aflibercept in treatment-naive eyes with nAMD per the available evidence (VIEW 1&2).

In your response, you did not determine whether your cohort was composed of only treatment-naive eyes or it was a mixed group of switched and treatment-naive ones.

We believe that visual outcomes of our work are satisfactory; in the integrated analysis of the VIEW 1&2 RCTs,2 the proportion maintaining vision (losing <15 ETDRS L) was 95% (94% in our cohort), whereas the proportion gaining >15 ETDRS L was 31% (24% in our cohort of the published work).

In our research work, classifying lesions as active/inactive (wet/dry) post loading has been of great benefit; at end of year 1, 136 eyes (53%) were inactive, whereas 119 eyes (47%) remained active. Both groups showed VA gains irrespective of the presence/absence of fluid. In both groups, the proportion maintaining vision (losing <15 ETDRS letters) was 94%. The macular status post loading was a reliable indicator of the macular status at end of year 1. Both groups received the same treatment post loading (Aflibefcept Q8W).

We would be interested in knowing the proportion of eyes in your cohort that did not require further injections post loading and what re-injection criteria were used in your centre.

In the modified IVAN protocol you implemented, it is not clear whether post-loading Aflibercept was used monthly (Q4W) or bi-monthly (Q8W). In case you used it monthly (Q4W), what would be the benefit of using Aflibercept every month if the current evidence suggested that its use bi-monthly (Q8W) was not inferior to monthly ranibizumab?2

Once a patient commences anti-VEG therapy, maintaining post-loading gain is a clinical challenge. In your protocol, a REACTIVE regimen was adopted in inactive eyes post loading. We carried on with a PROACTIVE protocol; bi-monthly (Q8W). It might have over-treated a proportion of eyes, but on the other hand, it did not leave the macula under the mercy of the CNV activity as you adopted post loading in inactive eyes.

When monthly treatment was discontinued and a more reactive clinician-guided approach using ranibizumab as required was used, over the subsequent 2 years (HORIZON study) there was a gradual reduction in mean vision gain. This trend continued in the further long-term follow-up (SEVEN-UP) study.3

The landmark (VIEW) studies4 also showed good results adopting a proactive regimen in year 1 as well as year 2 maintaining the gains achieved in year 1; Q8W in year 1 and a PRN approach being used but capped in year 2, so that treatment was given Q12W even if VA and OCT were stable and not strictly needing a treatment according to the PRN approach.

The CATT study5 showed that it was possible to achieve good visual acuity results implementing a PRN protocol, but this was achieved through very strict monthly monitoring, a very low retreatment threshold, and a relatively high number of treatments in the PRN arms; features not easily adopted in a real-world clinical setting as they generate more cost and capacity problems.

The PRN approach does not maintain post-loading gain. The UK EMR study6 of 12 951 eyes receiving 92 976 injections showed a gradual reduction (post loading) of visual acuity to below baseline over the course of three years (PRN implemented).

On the other hand, a proactive approach using Aflibercept as per VIEW showed good results in routine clinical practice across the UK7 in addition to results comparable to VIEW as achieved in our cohort.1

This trend of losing vision post loading when PRN regimen is applied can be observed in Figure 1 of your response; your VA results at M12 (3–4 ETDRS L gain) were inferior to both MARINA and ANCHOR, and inferior to what had already been achieved in the major VIEW 1&2 RCTs using Eylea2 and inferior to what’s been achieved in our cohort.1

In our research, a total number of seven injections were given; a total of only three follow-up visits were booked per patient per year. In your cohort, Figure 2 shows an average number of five injections/eye/year 1, and an average number of 12 follow-up visits (four times the number of follow-ups in our cohort)/eye/ year were implemented (Figure 3). It is obvious that, in terms of capacity, our model does solve many of the issues (capacity, booking, and so on) from which most of the MR departments in the UK suffer.

In Table 1, a comparison between Southampton Protocol and the modified IVAN (Your Centre) is made. Our protocol shows better VA outcomes, it is evidence-based, it sorts out capacity issues in addition to the easy implementation. There is definitely a difference in the cost of treatment/eye/year but at the expense of having to fit patients into office visits at an average of 12 follow-ups in your modified IVAN protocol. The IVAN-modified protocol might comply with being slightly less costly but it is not more cost-effective than Southampton protocol.

Table 1 Southampton vs modified IVAN protocol
Full size table

A treatment protocol in which the maximum VA gain is achieved with the minimum of clinic visits would not only alleviate capacity issues but also offer a cost-effective option. Our results demonstrate that for treatment-naive patients with nAMD, treated in a real-world clinical setting, implementing a PROACTIVE fixed-dosing bi-monthly regimen of Aflibercept does offer not only efficacy, safety, and quality, but also cost-effectiveness.