Treat and Extend Anti‐VEGF Treatment Regimens for Neovascular Age‐Related Macular Degeneration: To Stop or Not to Stop?

The prognosis for patients with neovascular age-related macular degeneration (nAMD) has been transformed by first-generation and newer anti-VEGF agents. The initial excitement around anti-VEGF treatment for nAMD in patients and clinicians has been tempered somewhat in recent years by the realisation that long-term treatment is needed for the majority of patients, which has provided a challenge to patients, clinicians and payers. When initially brought to market, anti-VEGF agents such as bevacizumab and ranibizumab were deployed in clinical settings using ‘pro re nata’ (PRN) regimens where patients would need to be seen in clinic on a four weekly basis and injections were only given when there were signs of exudative disease activity (such as fluid on OCT imaging or new macular haemorrhage) [1]. Although PRN treatment was initially widely used internationally, the treatment regimen has gradually fallen out of favour with the realisation that long-term use of PRN treatment can lead to a gradual loss of vision in clinical cohorts with nAMD [2]. This ‘efficacy gap’ between treatment outcomes in clinical trials and clinical practice was associated with several factors, including real-world PRN treatment regimens not consistently reviewing patients every 4 weeks due to clinic capacity or patient factors leading to undertreatment. This materially differed from clinical trials where regular, four-weekly visits were mandated and delivered [3]. Suboptimal monitoring of patients leads to episodic and potentially prolonged periods of untreated nAMD disease activity, leading to structural changes associated with vision loss in the medium to long term. Indeed, emerging evidence from clinical trials [4] and real-world cohorts [5] suggests the best visual acuity outcomes may be achieved if fluctuations in retinal thickness and morphological features of nAMD disease activity are minimised and a ‘treat to stability’ approach is adopted. An additional observation from clinical trials from the PRN treatment era is that in the CATT, the two arms receiving continuous, monthly anti-VEGF injections in the first year of treatment ended up with comparable visual acuity outcomes to the arms receiving PRN treatment from the start when we look at 2 year outcomes at the end of the study [3]. This indicates no long-term benefit of fixed dosing or loading injections, which do not have a disease-modifying effect. The realisation of the shortcomings of PRN treatment in real-world settings co-incided with the introduction of a new agent, aflibercept 2 mg, into clinical practice. The phase III aflibercept 2 mg VIEW studies used fixed bimonthly dosing in Year 1 after three loading injections and a capped PRN approach to treatment in Year 2 [6]. We then saw a paradigm shift away from PRN towards treat and extend treatment regimens in clinical practice from 2012 onwards. Treat and extend is a more forgiving treatment paradigm in real-world settings compared to PRN, as patients receive an anti-VEGF injection at every clinic visit with a treat and extend but where the interval between injections, which is shortened, maintained or increased based on the presence or absence of nAMD disease activity with the aim of maintaining an exudation free macula [7]. Treat and extend has several theoretical advantages over PRN treatment with fewer clinic visits and with individualised treatment based on markers of disease activity leading to fewer episodes of disease reactivation and less fluctuation in retinal thickness as anti-VEGF injections are given proactively in treat and extend regimens rather than reactively when disease features recur as in PRN treatment. This, in turn, is associated with better treatment outcomes reported in clinical practice compared to PRN treatment [8, 9]. Returning to the long-term nature of nAMD treatment, the question is how safe is it to stop or suspend anti-VEGF injections once the treatment interval has been extended out to 12 weeks or beyond without encountering reactivation of nAMD disease activity? In the era of PRN treatment, studies showed a high rate of nAMD disease reactivation after stopping or pausing anti-VEGF injections [10, 11] with poor outcomes. In this issue, Artiaga et al. retrospectively analysed data from their stable nAMD clinics where patients with nAMD are followed up every 3 months once they have successfully been weaned off anti-VEGF injections after using a treat and extend treatment regimen [12]. They identified consecutive nAMD patients who attended the stable nAMD clinic over a 5-month period and collected baseline data on previous anti-VEGF injection usage before stopping treatment and demographic data from electronic patient records. Where both eyes from the same patient were eligible for the analysis, they chose one eye based on more recent injection usage, or for nAMD eyes with same-day injections, they chose to analyse data from the eye which had a longer anti-VEGF treatment duration. They went on to see what proportion of these patients developed reactivation of nAMD over time with a mean observation period of 5 years and collected data on the details of the reactivation, including whether there were symptoms and whether the reactivation was detected at a scheduled stable nAMD clinic visit or whether it was through a patient-initiated clinic visit. Their results showed that of the 286 patients included in the analysis, approximately one-third of patients reactivated after stopping anti-VEGF injections and 73.9% of these were within the first 3 years of follow-up. Moreover, the vast majority of these reactivations (86.7%) were detected at clinic visits rather than unexpectedly between clinic visits, showing the importance of regular follow-up after stopping or pausing anti-VEGF injections. With newer agents such as faricimab and aflibercept 8 mg, treat and extend regimens have the potential to evolve further as the TENAYA/LUCERNE [13] and PULSAR [14] clinical trials used disease activity criteria which heavily relied on nAMD disease features in the central 1 mm subfield rather than considering the entire macular OCT scan. It remains to be seen if the clinical community can achieve extended treatment intervals with these newer agents by using current real-world injection interval adjustment criteria where disease features across the whole macular OCT scan are used to assess nAMD disease activity or whether nAMD treatment in clinical settings evolves to refine nAMD disease activity based on disease features on the central 1 mm macular OCT subfield. As we potentially move beyond 16 weekly injections to clinical treatment paradigms with extended injection intervals of 20 weeks or beyond with newer, more durable agents, we will also discover if this further fractionation of treatment intervals to 20 weeks and perhaps beyond is associated with less reactivation of nAMD after stopping anti-VEGF treatment. For now, however, the results of the work presented by Artiaga et al. in this issue are consistent with other studies suggesting that stopping anti-VEGF after treat and extend treatment paradigms requires nAMD patients to be followed up for several years, given the high rate of reactivation to reduce the incidence of vision loss by detecting and treating new nAMD disease activity in a timely manner. Consultant to Bayer, Boehringer Ingelheim, Genentech and Roche.