: Pro Re Nata brolucizumab for early onset and treatment-naïve diabetic macular edema: a prospective study

OBJECTIVES To determine the e�cacy and safety of brolucizumab therapy administered on a pro re nata (PRN) basis without loading dose in treatment naïve patients with diabetic macular edema (DME) for 1 year follow-up. METHODS Patients with recent DME (< 6 months) received a mandatory brolucizumab injection at inclusion and other injections could be given on a PRN basis with an 8-week interval (between injections) at minimum. Rescue therapy with other anti-VEGF was possible in case of incomplete DME resolution after the second brolucizumab and with a minimum of 1-month treatment free interval between 2 injections. The primary outcome measure was the change in best-corrected visual acuity (BCVA) at 12 months. Secondary outcome measures included the change in central sub�eld thickness (CST), the change in hard exudate surface area and microaneurysms at 1 year.


INTRODUCTION
Brolucizumab therapy has been recently approved for patients with treatment-naïve diabetic macular edema (DME).The KITE and KESTREL studies have demonstrated non-inferiority of the molecule given in q8/q12 intervals based upon disease activity, when compared to q8 injections of a ibercept. 1Moreover, brolucizumab showed higher reduction in intraretinal and subretinal uid during the study period, explaining the reduction in number of injections for the group of patients treated with this molecule. 1,2his better anatomical outcome could be due to the low molecular weight of brolucizumab which in turn allows a higher dose of active molecule to be concentrated in the standard 0.05 ml volume of drug typically injected into the vitreous cavity.The longer duration of action of this molecule is of interest for the treatment of chronic disease such as age-related macular degeneration (AMD) 3 or DME.However, concerns about potential adverse events induced by brolucizumab have quickly replaced the initial enthusiasm for the molecule, 4,5 explaining a timid start of its prescription in routine practice.Yet, the randomized controlled trials KITE and KESTREL demonstrated a comparable incidence of adverse events between a ibercept and brolucizumab in treatment naïve DME patients. 3These immunological complications following injections are poorly understood but its rather unusually high occurrence questions the strictness of a mandatory dosing regimen and the concept of a loading dose for a drug that is injected in such high concentrations as brolucizumab.We have recently shown that the number of injections of brolucizumab required to treat neovascular AMD under a strict PRN regimen without loading dose was less than what is usually given with conventional molecules (i.e bevacizumab, ranibizumab and a ibercept), without reduced visual gain. 6e aim of the present prospective study is to evaluate functional and anatomical outcomes at 12 months for treatment naïve DME patients treated with brolucizumab in a strict PRN regimen, without resorting to a loading dose.

MATERIALS AND METHODS
This prospective study enrolled patients with treatment naïve DME between April 2021 and November 2021.Patients included were treated with intravitreal brolucizumab, received in PRN regimen for 12 months from baseline but rst injection was mandatory.Informed consent was obtained from the patients at inclusion after complete discussion of the disease process, the treatment alternatives, and potential adverse events.The study adhered to the tenets of the declaration of Helsinki and was approved by the local ethics committee (Sudhalkar Eye Hospital) under the number 2021/02/n°01.Inclusion criteria was an age ≥ 18 years old, central macular sub eld thickness (CST) ≥ 300 µm, occurrence of symptoms < 6 months, no other disease associated to macular edema except diabetes mellitus, best-corrected visual acuity (BCVA) comprised between 0.3 logMAR (20/40 Snellen equivalent) and 1.0 logMAR (20/200 Snellen equivalent) evaluated on the Early Treatment for Diabetic Retinopathy Study (ETDRS) chart.One eye of each patient was included in the study.If both eyes were eligible, one eye was randomly selected using a random number table.Exclusion criteria included past episodes of ocular in ammation/uveitis, other ocular conditions that can affect the retina, history of pars plana vitrectomy and presence of high-grade cataract preventing correct imaging of the macula.Patients with uncontrolled diabetes mellitus (HbA1c ≥ 10%) or any uncontrolled systemic disease were excluded.
Patients with incomplete follow-up were also excluded.

Examination
All patients underwent a complete ophthalmological examination including intraocular pressure (IOP), standard 7-eld fundus photography and fundus uorescein angiography (FA) (Zeiss Visu 500, Carl Zeiss Meditec, Jena, Germany) and spectral domain (SD)-OCT (Opko/OTI Inc., Miami, FL, USA) at baseline to determine eligibility.This ophthalmological examination was repeated at each visit except for fundus photography which was performed only at 6 and 12 months, and FA which was performed at baseline and at 12 months.When required, patients with cataract underwent phacoemulsi cation and intraocular lens (IOL) implantation accompanied with intravitreal brolucizumab injection on the same day.A single surgeon (AS) performed all surgeries.Diabetes mellitus (and all systemic comorbidities) were managed in strict consult with a physician.

Procedure and follow-up
Brolucizumab treatment was given under aseptic conditions in a dedicated room.Patients were seen days 1 and every week the month after the injection and then monthly thereafter for 12 months.If a second injection was necessary, the patients were also followed up on days 1 and every week the month after the injection and monthly thereafter.Brolucizumab injections could be given, if necessary, every 8 weeks at minimum, but not before 8 weeks.Patients were eligible for rescue therapy with other anti-VEGF agents (ranibizumab or a ibercept) or intravitreal steroids (dexamethasone implant) and/or focal laser if there was incomplete DME resolution between 2 brolucizumab injections, starting after the second brolucizumab injection and with 1 treatment free interval having passed between 2 injections.The nal choice for rescue therapy was at the Ophthalmologist's discretion.During the study, diabetic retinopathy could be treated by pan-retinal photocoagulation or surgery if necessary.

Outcome Measures
The primary outcome measure was the change in BCVA at 12 months.Secondary outcome measures were the proportion of patients who gained ≥ 15 letters from baseline to month 12, CST changes at 12 months, the change in microaneurysms numbers from baseline to 12 months counted on early, and late phases of uorescein angiography as previously described, 7 and the change in hard exudates area on fundus retinography using ImageJ (available in the public domain at https://imagej.nih.gov/ij/download.htmlInstitute of Health, USA) according to the algorithm previously described. 8ular hypertension (OHT) was de ned as IOP > 25 mmHg and/or an increase of 10 mmHg from baseline over the follow-up period.

Statistical analysis
Descriptive statistics was used to analyze categorical variables in terms of size (absolute frequencies) and proportions (relative frequencies).The signi cance of the change in BCVA over time was determined using the repeated measures ANOVA test.The repeated measures ANOVA test was also used to determine change in CST, the hard exudate area, and IOP changes over time.Statistical signi cance was set at p < 0.05.All patients were managed on an intent-to-treat basis.

RESULTS
A total of 53 patients were included in this study amongst the 152-treatment naïve DME patients examined during the same period of inclusion.Patient characteristics are given in Table 1.During the study period, mean (SD) HbA1c decreased signi cantly from 7.8 (0.3) % to 6.6 (0.4) % (p = 0.017), the mean weight of patients decreased non-signi cantly from 66.1 (7.6) Kg to 63.4 (8.2) Kg (p = 0.27), and the blood pressure remained well controlled throughout the follow-up with none of the patients experienced any hypertensive crisis.Anatomical outcomes decreased signi cantly with treatment and this decrease was sustained over the study period with a statistically signi cant drop from baseline to month 12 (p = 0.013, Table 2 and Fig. 2).The mean area of hard exudates showed a statistically signi cant decrease during the study period (p = 0.012, Table 2) and all patients demonstrated a decrease in hard exudate surface area.The mean (SD) number of microaneurysms also signi cantly decrease from baseline to 12 months in the early phase FA (8.4 (2.4) microaneurysms versus 3.2 (0.8) microaneurysms, p = 0.012) and in the late phase FA (4.3 (1.4) microaneurysms versus 2 (0.6) microaneurysms, p = 0.02).

Number of intravitreal injections
The mean (SD) number of intravitreal injections over the 12 months period 2.6 (0.8) [range: 1-4] in addition to the mandatory baseline injection.A total of 3 eyes (5.6%) required 1 additional brolucizumab

Ocular Hypertension
According to OHT de nition, none of the patients experienced OHT during the study period.

Diabetic retinopathy-induced adverse events
Four patients (7.5%) had developed proliferative disease during the study period and received panretinal photocoagulation.These four patients did not show signi cantly worse mean (SD) visual outcomes at 12 months as compared to the rest of the study population [0.43 (SD-0.16)]logMAR in this subset of 4 patients, p = 0.27).

DISCUSSION
We provide here real-world outcomes of brolucizumab in the treatment of newly diagnosed DME patients.We report signi cant improvements in BCVA over the 1-year study period, showing early rise in BCVA that was sustained over the study period, and associated with a dramatic decrease in macular thickness.This allowed that almost three-quarter of eyes had a nal vision ≥ 20/40 at the end of the follow-up.Most importantly, a majority of patients gained ≥ 15-letters, and this BCVA improvement generally occurred during the rst month after the rst injection.These excellent visual outcomes should nevertheless be nuanced.Our patient group presented with lower BCVA at baseline in comparison to the KITE & KESTREL studies, therefore permitting higher potential visual gains. 9,10Moreover, in our study the DME was of recent onset and patients treatment naive, permitting higher potential gain in BCVA. 11ese functional outcomes were backed up by anatomical outcomes.The decrease in macular thickness occurred shortly after the rst brolucizumab injection and was sustained over the study period.3][14][15] The presence of uid in the retinal space (intraretinal or subretinal) is generally the result of increased vascular permeability, which in DME is a probable consequence of leaking intraretinal microaneurysms.Indeed, increased CST was more frequently observed around microaneurysms 16 nd a recent analysis published in literature demonstrated that the use of anti-VEGF decreased the numbers of microaneurysms in patients with DME. 7 We demonstrate in the current analysis that the numbers of microaneurysms was halved by the end of the follow-up period, probably explaining the sustained anatomical response to the molecule.The same conclusions can be drawn with hard exudate quanti cation which showed a signi cant decrease in surface area between baseline and the end of follow-up.
These functional and anatomical results were obtained with a mean of 2.6 injections in addition to the rst mandatory brolucizumab injection.This low number of injections during the year of follow-up was obtained under a strict prospective PRN regimen, permitting repeat injections only when DME recurred, or in case of an incomplete anatomical response.It should however be noted that less than 15% of patients required rescue therapy with at least one other anti-VEGF therapy over the study period.These results contrasts with the randomized controlled trials KITE & KESTREL where a median of 7 injections were realized in the rst year. 1 However, in these pivotal studies, a loading dose of 5 injections was mandatory, followed by a pro-active regimen with systematic reinjections every 8-or 12-weeks.In the literature, there are to date very few reports on the results of brolucizumab outside of randomized-controlled trials.Chakraborty et al. demonstrated an early experience with the molecule on 13 eyes of patients with chronic DME.They showed that all but one eye who completed 16-weeks follow-up after the rst brolucizumab could achieve 16-weeks treatment interval before the 2nd brolucizumab. 17Combined with our study, these results demonstrate that a sustained response can be obtained with brolucizumab, which was injected in average every 4 months for recent-onset DME.The omission of the loading dose proposed in the present study, and the absence of pro-active management of DME, contribute to the lower number of injections and allow a more individualized treatment induction in these patients.This is of importance with respect to the treatment burden and the risk of ocular adverse events and especially the occurrence of intraocular in ammation with rare cases of occlusive vasculitis that can be associated with visual loss. 4,18The exact mechanism of these intraocular in ammations is currently unknown, but they seem appeared more frequently when injections are given in a monthly dosing (i.e. during the loading dose, or systematic q4 treatment). 19,20However, studies have shown that the incidence of adverse events has been probably noted to be marginally lower in DME patients, probably due to the speci city of the disease, combined with a loading dose where patients were injected every 6 weeks in DME, instead of every 4 weeks in AMD.
Interestingly, the proportion of cataract surgery during the follow-up may appear high and could arti cially increase BCVA gains. 21It is therefore questionable on how the molecule could be associated with this high occurrence of lens opaci cation.However, these cataracts should be attributed to the diabetes itself and the age of the population rather than the molecule. 22Currently, there is no evidence that demonstrates an eventual cataractogenic effect of brolucizumab when injected intravitreally.Finally, the surgery did not worsen treatment outcomes for these patients, demonstrated again the safety of the surgery in DME treated patients. 23Another complication associated with the disease is the development of proliferative retinopathy during the study, which required panretinal photocoagulation.Only 4 patients experienced neovascular complications and their functional outcomes did not differ from other patients.The anti-VEGF agents can reduce the progression of diabetic retinopathy to the proliferative stage, as emphasized with other molecules. 24,25 acknowledge some limitations in the present study.The relatively low number of patients included precludes comparative and subgroup analyses.As a molecule which has been recently approved in DME, brolucizumab needs to nd a place in the therapeutic armamentarium used in DME. 26,27Moreover, we included only patients with early onset of DME symptoms, which is not always the case in a chronic disease such as type 2 diabetes mellitus.The recent nature of the DME could explain not only the excellent response to the molecule in terms of functional and anatomical outcomes, but also the lower number of injections required to control the ocular disease.Furthermore, the functional and anatomical outcomes were probably a combined effect of therapy and the control of systemic factors (as shown by the reduction of HbA1c) that could in uence macular edema.Therefore, the absence of a comparative arm prevents the direct comparison with studies treating patients with classical anti-VEGF.Lastly, the follow-up can be de ned as short as the DME can progress during years, depending on systemic factors.
In conclusion, this study demonstrated e cacity and safety of brolucizumab under PRN regimen in patients with recent history of DME.Most importantly, the low number of injections required over 1 year can decrease the treatment burden and potentially minimize the occurrence of ocular adverse events.