Real-world efficacy and safety of Brolucizumab injection in vitrectomised eyes (BIVE) - a prospective multicentric study

Background

The main objective was to evaluate the efficacy and safety of anti-VEGF Brolucizumab injection in vitrectomised eyes for post-vitrectomy recalcitrant macular edema in patients with nAMD/ PCV and recalcitrant diabetic macular edemas as per pro-re-nata protocol.

Patients and methods

A prospective, multi-centric non-randomised interventional study was performed where intravitreal injection of Brolucizumab was administered for post-vitrectomy recalcitrant macular edema as per pro-re-nata protocol. Vitrectomy was performed for sub-macular haemorrhage (Group A) and recalcitrant diabetic macular oedema (Group B).

Main outcomes measured

Pre- and post-injection BCVA and Central Foveal Thickness (CFT) were noted at 4, 12, 24, 36, 52, and 76 weeks.

Results

Eighty-one intravitreal injections of Brolucizumab were given in 51 eyes. The BCVA improved from pre-injection LogMAR 0.80 ± 0.32 to 6-month post-injection 0.65 ± 0.23 (Group A, p-value < 0.001, 95% CI) and LogMAR 0.70 ± 0.37 to 0.60 ± 0.34 (Group B, p = 0.378, 95% CI). Significant reduction in CFT noted from pre-injection 367.75 ± 124.12 microns to 6-month post-injection 291.09 ± 81.42 micron (A, p = 0.001, 95% CI) and from 468.45 ± 145.69 micron to 342.19 ± 117.91 micron (B, p = 0.002, 95% CI).

Conclusion

Brolucizumab may have the potential to be durable, effective and safe in vitrectomised eyes in both post-nAMD and diabetic patients.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12886-026-04621-2.

Summary statement

We prospectively evaluated the efficacy and safety of intravitreal Brolucizumab injection to treat recalcitrant macular edema cases in vitrectomised eyes.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12886-026-04621-2.

Background and rationale

Over the past few years, intravitreal anti-Vascular Endothelial derived Growth Factor (anti-VEGF) injections have been the major treatment modality for age-related neovascular membrane, diabetic macular oedema, macular oedema secondary to retinal vascular occlusion or any recurrent persistent macular oedema because of their anti-angiogenic properties [1]. Vitrectomy is often needed to treat cases of subretinal haematoma formation secondary to age-related macular degeneration (AMD) or polypoidal choroidal vasculopathy (PCV) or cases of recalcitrant diabetic macular oedema [2]. However, postoperative macular oedema is a challenging treatment situation. There are few postoperative situations where anti-VEGF injection may be required in vitrectomised eye, operated to treat submacular haematoma, secondary to neovascular age-related macular degeneration (n-AMD) or polypoidal choroidal vasculopathy (PCV) to treat reactivation of choroidal neovascular growth. Moreover, anti-VEGF injection is often necessary for the recurrence of macular oedema after surgical intervention of refractory diabetic macular oedema [3]. However, vitrectomy changes the intraocular milieu which leads to certain difficulties with anti-VEGF injections in the vitrectomised eye. These include (i) reduced half-life of anti-VEGF molecule [4], (ii) rapid clearance of the drug from the vitreous cavity [4], (iii) insufficient sustained therapeutic level is maintained in vitrectomised eye [4] and (iv) vitrectomy may alter cytokine profile which may reduce effectiveness of anti-VEGF [5]. Because of these problems, more frequent injections of anti-VEGF are required (at 2-week intervals rather than the standard four weeks) [4, 6]. Moreover, anatomical and functional outcomes of intravitreal anti-VEGF injections are sub-optimal in vitrectomised eyes. There are also potential risks of injecting steroids like triamcinolone acetonide (TA) intravitreally like endophthalmitis, lens opacifications and intraocular pressure (IOP) elevation [7]. Brolucizumab (Pagenax, Novartis, India), a newly developed smallest anti-VEGF molecule has demonstrated a longer duration of action and improvement in anatomical and functional outcomes in previous clinical studies in a quarterly 12-week regimen, indicating its potential to reduce treatment burden [8]. This Brolucizumab has recently received FDA approval for managing neovascular AMD as well as diabetic macular oedema [8, 9]. Considering these unique features of Brolucizumab, we decided to use this anti-VEGF agent in vitrectomised eyes.

Objective

The main objective was to evaluate the efficacy and safety of anti-VEGF Brolucizumab injection in vitrectomised eyes for post-vitrectomy recalcitrant macular edema in patients with nAMD/ PCV and recalcitrant diabetic macular edema as per pro-re-nata protocol.

Method: participants, interventions

Study setting

It was a prospective, non-randomised interventional study taking consecutive patients of post-vitrectomy recurrent macular oedema cases from November 2020 to October 2023. This study was approved by the Institutional Review Board (IRB) [Disha Eye Hospitals Pvt Ltd Ethics Committee. Registration No. ECR/846/Inst/WB/2016/RR-19], and consent to participate was obtained from each patient. This study involved human subjects, and informed consent was obtained from every patient after being informed about the procedure and its possible complications. Our study was strictly adhered to the tenets of the Declaration of Helsinki.

Eligibility criteria

The criteria for inclusion of candidates were as follows: (i) All treated eyes were vitrectomised eyes, where vitrectomy was performed to treat post n-AMD or PCV sub-macular haemorrhage (Group A) and recalcitrant diabetic macular oedema (either in severe non-proliferative to early proliferative diabetic retinopathy (Group B), (ii) All study eyes were treated after vitrectomy with either loading doses of anti-VEGF injections (Ranibizumab in both groups, at one-month intervals) or intravitreal steroids [ either intravitreal triamcinolone acetonide (IVTA) or dexamethasone implant in Group B ] as a standard first-line of treatment, (iii) Recurrent macular oedema in the form of intra-retinal fluid (IRF), and sub-retinal fluid (SRF) with or without pigment epithelial detachment (PED) on EDI-OCT were considered for an indication of treatment, (iv) Repetition of treatment was considered on the basis of recurrence of macular oedema with or without vision loss, (v) Pro-re-nata (PRN) protocol, means that injections were administered only when the disease shows signs of activity, such as intraretinal or subretinal fluid, was followed as the treatment.

Exclusion criteria

We have excluded (i) sub-macular haemorrhage cases due to other reasons like macro aneurysm or trauma (in Group A), (ii) cases with associated vascular occlusion, (iii) compromised optic discs (iv) absent perception of light and (v) associated severe dysfunction of heart, kidney, liver, lungs or any other organ.

Interventions

Baseline diagnosis was performed on the basis of FFA, ICGA and OCT findings. In Group A, a recurrence of choroidal neovascular membrane activity was noted on OCT after vitrectomy, done to treat submacular haemorrhage (SMH), secondary to neovascular AMD and PCV. 25-gauge Vitrectomy was done in this group along with sub-retinal injection of a cocktail combination of r-tPA (recombinant tissue Plasminogen Activator), Ranibizumab and filtered air (freshly manually prepared aseptically during surgery) with finally gas tamponade. Prior approval from the EC was also taken beforehand for this subretinal injection of the drugs. In Group B, recurrence of diabetic macular oedema (DME) after 25-gauge vitrectomy with Brilliant Blue G dye assisted Internal Limiting Membrane (ILM) peeling (at least 2-disc diameter in size), performed to treat refractory DME (with or without proliferative changes). The injection Brolucizumab (6 mg in 0.05 ml) was administered aseptically by a 30G needle via pars plana route after adequate topical anaesthesia inside a sterile operative room. During post-injection follow-up, we followed the injection schedule as per pro-re-nata protocol, which means that injections are administered only when the macular EDI-OCT showed signs of activity, such as intraretinal or subretinal fluid. Post-injection follow-up was done at 4, 8, 12, 24, 52 and 76 weeks with detailed slit-lamp anterior segment examination, fundus photography (for any inflammatory reaction) and EDI-OCT macula. On EDI-OCT, Central Foveal Thickness (CFT) and pattern of fluid in the form of intraretinal fluid (IRF), sub-retinal fluid (SRF) or pigment epithelial detachment (PED) were noted. Both functional (BCVA) and anatomical (EDI-OCT changes) parameters were evaluated at each follow-up. We stopped recruiting any patients, 6 months before the stipulated end-point of our study. Ocular and systemic adverse effects were also assessed.

Statistical analysis

SPSS software version 20 was used for statistical analysis. The normal distribution test of all parameters was done before statistical analysis. Student-paired t-tests were used to analyze BCVA changes in the pre-operative and post-operative periods. Changes in CFT were also evaluated using Student paired t-tests. A p-value less than 0.05 was considered to be statistically significant.

Results and outcomes

51 consecutive postoperative eyes have been included in our study. A total of 81 intravitreal injections of Brolucizumab were given in our study patients as per pro-re-nata protocol in 2-year time period. The Brolucizumab was administered in 27 (10 eyes n-AMD and 17 eyes PCV) post-vitrectomy eyes (44 injections) in Group A and 24 post-vitrectomy eyes (37 injections) in Group B. The mean age of included patients was 62.94 ± 10.50 (range 28–81) years. The male/ female ratio was 29:22. Total of 30 /51 (58.82%) eyes (16 in Group A and 14 in Group B) were pseudophakic in our study. The overall mean interval between the vitrectomy surgery and Brolucizumab injection was 39.37 ± 28.05 weeks (range: 8–78). Demographic and clinical characteristics of the study patients in both groups are highlighted in Table 1. Separate analyses for each disease group were done to assess improvement in BCVA and CFT. The BCVA improved significantly from pre-injection LogMAR 0.80 ± 0.32 to 0.67 ± 0.26 (p-value less than 0.001, 95% CI) at 1 month, 0.66 ± 0.27 (p-value less than 0.001, 95% CI) at 3 months and 0.65 ± 0.23 at 6 months (p-value less than 0.001, 95% CI) in Group A. The BCVA also improved significantly in Group B from pre-injection from LogMAR 0.70 ± 0.37 to 0.59 ± 0.31 (p-value = 0.006) at 1 month and 0.56 ± 0.27 (p-value = 0.011, 95% CI) at 3 months, but at 6 months, improved to LogMAR 0.60 ± 0.34 (p = 0.378, 95% CI) which was statistically not significant (Graph 1). Significant reduction in CFT was simultaneously noted in both groups at 1 month, 3 months and 6 months (Graph 2). CFT was reduced from pre-injection 367.75 ± 124.12 microns to 301.41 ± 79.09 microns (p-value less than 0.001, 95% CI) at 1 month, 296.89 ± 115.33 micron (p-value = 0.001, 95% CI) at 3 months and 291.09 ± 81.42 microns (p = 0.001, 95% CI) at 6 months in Group (A) Similarly, CFT was reduced from 468.45 ± 145.69 microns to 382.26 ± 159.22 microns (p = 0.013, 95% CI) at 1 month, 372.13 ± 121.15 microns (p = 0.036, 95% CI) at 3 months and 341.16 ± 117.91 micron (p = 0.002, 95% CI) at 6 months in Group (B) Pattern of fluid on EDI-OCT was variable in two groups. IRF was present in 70.45% (31/44) eyes in Group A and 78.38% (29/37) eyes in Group B. SRF was present in 34.09% (15/44) eyes in Group A and 24.32% (9/37) eyes in Group B. PED was present in 52.27% (23/44) eyes in Group A and 2.70% (1/37) eyes in Group B. Mean follow up was 39.27 ± 20.29 weeks (range: 12–107 weeks). Mean injection-free interval was 24.21 ± 14.36 (range: 11–76) weeks in Group A (Fig. 1a & b) and 21.64 ± 7.66 (range: 11–38) weeks in Group B (Fig. 2). The list of adverse events in all three groups has been highlighted in Table 2. Among the ocular adverse events, few non-serious events like mild ocular pain (in 33 episodes), ocular burning sensation (in 21 episodes) and subconjunctival haemorrhage (in 25 episodes) happened after the injection. Only two cases of sub-retinal haemorrhage (SRH, 4.55%) were detected in Group A as serious adverse effects. No other serious adverse events like intraocular inflammation (IOI) vasculitis or retinal vessel occlusion had been noticed in post-injection follow-up. None of our cases developed any systemic thromboembolic events after treatment.

Table 1.

Shows demographic and clinical characteristics of patients in both groups

Groups	Group A	Group B
Mean Age (years)	67.1 ± 7.25	58.88 ± 8.96
Male: Female	19:8	5:7
Mean interval between vitrectomy and Brolucizumab injection (in weeks)	23.29 ± 11.56	48.72 ± 28.17
Pre-inj BCAVA (LogMAR)	0.80 ± 0.32	0.70 ± 0.37
Post-inj 1mth BCAVA (LogMAR)	0.67 ± 0.26	0.59 ± 0.31
Post-inj 3mth BCVA (LogMAR)	0.66 ± 0.27	0.56 ± 0.27
Post-inj 6mth BCVA (LogMAR)	0.65 ± 0.23	0.60 ± 0.34
Pre-inj CFT (micron)	367.75 ± 124.12	468.45 ± 145.69
Post-inj 1 month CFT (micron)	301.41 ± 79.09	382.26 ± 159.22
Post-inj 3mth CFT (micron)	296.89 ± 115.33	372.13 ± 121.15
Post-inj 6mth CFT (micron)	291.09 ± 81.42	341.16 ± 117.91
IRF	70.45% (31/44)	78.38% (29/37)
SRF	34.09% (15/44)	24.32% (9/37)
PED	52.27% (23/44)	2.70% (1/37)
Mean follow-up (weeks)	41.48 ± 26.51	35.5 ± 15.88
Mean Injection-free Interval (weeks)	24.21 ± 14.36	21.64 ± 7.66

Fig. 1.

Shows A) The pre-operative OCT image of the left eye of a patient of Group A with Submacular haemorrhage (both Sub-RPE and sub-neurosensory retinal haemorrhage, a1) The post-operative fundus picture of the left eye of the same patient treated for Submacular haemorrhage with vitrectomy and a2) OCT macula showing resolved SMH with a postoperative regressed choroidal neovascular component with PED, a3) Recurrence of subretinal fluid after treatment with previous anti-VEGF injection, a4) Resolved SRF after 4 weeks of the first dose of intravitreal injection Brolucizumab, a5) Recurrence of SRF and IRF after 12 weeks which again resolved after the second dose of intravitreal injection Brolucizumab at 16 weeks (a6) and this was maintained till 24 weeks (a7), Third dose was required at 32 weeks (a8) due to recurrence of CNVM with SRF and IRF, (a9) Resolved SRF and IRF at 36 weeks with a sub-retinal scar. Another case of Group A: b1) post-operative fundus picture of left eye treated for SMH with vitrectomy and b2) OCT macula showing completely resolved SMH with NS thinning, extensive ellipsoid zone defect and parafoveal IRF, b3) Recurrence of PED and IRF after treatment with previous anti-VEGF injection, b4) PED was flattened with resolving IRF after 4 weeks of single dose of intravitreal injection Brolucizumab, b5) PED was flattened more and IRF was resolved completely at 12 weeks which was maintained till 76 weeks (b6) with minimal PED and no recurrence of IRF

Fig. 2.

Shows A) The pre-operative OCT image of the right eye of a patient of Group B with recalcitrant DME, planned for vitrectomy and ILM peeling, a1) The post-operative fundus picture of right eye of the same patient, treated for recalcitrant DME with vitrectomy and ILM peeling and a2) OCT macula showing completely resolved IRF with extensive EZ defect after post-vitrectomy intravitreal dexamethasone implant injection, a3) Recurrence of IRF after treatment with previous intravitreal dexamethasone implant injection, a4) resolving IRF after 4 weeks of single dose of intravitreal injection Brolucizumab, a5) IRF was resolved completely at 12 weeks which was maintained till 24 weeks (a6) with no recurrence of IRF

Table 2.

Shows a list of ocular adverse events in both groups

Adverse Events (AE)	Group A Frequency (%)	Group B Frequency (%)
Non-serious AE
1.Mild ocular pain	19/44 (43.18%)	16/37 (43.24%)
2. Ocular burning sensation	13/44 (29.55%)	9/37 (24.32%)
3. Subconjuntival Hemorrhage	12/44 (27.27%)	11/37 (29.73%)
Serious AE
Subretinal Hemorrhage	2/44 (4.55%)	0

Confidentiality

All data collected were coded. The patient’s anonymity was maintained by removing all identifiers. All the data were stored both electronically in a secure, password-protected database as well as in the paper format, kept in a locked cabinet to prevent unauthorized access.

Discussion

There is a significant change in the intraocular milieu after vitrectomy. In a vitrectomised eye, two factors are of major concern, like rate of retinal penetration of the drug and the rate of drug clearance from the vitreous cavity [10]. In the present study, we included post-vitrectomy macular oedema cases and evaluated the treatment effectiveness and safety profile of intravitreal Brolucizumab injection in these conditions. Although we have discussed the treatment effectiveness of intravitreal Brolucizumab injection in two different diseases n-AMD and DME, the mechanisms of refractory macular oedema after vitrectomy are different. Post-vitrectomy macular oedema cases can be persistence or recurrence of intraretinal, subretinal or sub-RPE fluid on OCT due to incomplete regression of neovascular membrane in post-AMD/PCV cases or recurrent elevation of intravitreal VEGF level in uncontrolled diabetics. VEGF plays a significant role in retinal vessel hyper-permeability after diabetic vitrectomy. Anti-VEGF molecules have also role in controlling these retinal hyper-permeability conditions, which is related to the resolution of retinal fluid, which is inflammatory in origin. These post-vitrectomy macular oedema cases were refractory to the previous two doses of intravitreal anti-VEGF or steroids. However, we have excluded the sub-macular haemorrhage cases that were other than typical post-AMD/PCV in origin. We assessed the functional and anatomical improvement after intravitreal Brolucizumab in these conditions. Brolucizumab, being the smallest available anti-VEGF agent, have far more rapid penetration in retina compared to its predecessors with minimal subsequent systemic toxicity [11]. The study revealed that the low molecular weight of brolucizumab (4× lower than that of aflibercept and 1.8× lower than ranibizumab) permits a molar dosing around 12-fold higher than with aflibercept and 22-fold higher than with ranibizumab, respectively [12]. Hence, more Brolucizumab molecules will be available to treat neovascular components in vitrectomised aqueous filled vitreous cavities, where the chance of drug clearance from the vitreous cavity is very high. Studies also showed that the time to maximum concentration in the retina is 1–6 h for brolucizumab [13], compared with 6 h for ranibizumab in monkeys [14] and 24 h for aflibercept in a rabbit model [15]. As the rate of retinal penetration is fast (within 1–6 h), more drug molecules will be accumulated in sub-retinal and sub-RPE space in a very short time. The slower rate of retinal penetration for the other intravitreal anti-VEGF molecules is probably because of their larger molecular size [16]. Hence, the availability of anti-VEGF molecules in the vitreous cavity should be theoretically very low in cases of Brolucizumab because of the rapid penetration of the drug through the retina than other anti-VEGF molecules. This may be the probable explanation for Brolucizumab molecules to have a longer duration of action in the vitrectomised eye, unlike other drugs which have a slower rate of penetration and thus a high rate of drug clearance. Niwa Y also showed intravitreally injected both ranibizumab and aflibercept have shorter half-lives in aqueous humour along with shorter duration VEGF suppression in cases of vitrectomized eyes in comparison of non-vitrectomised macaque eyes and this difference increases when vitrectomy is done along with lensectomy [17]. Similarly, Kakinoki M also showed a similar result with Bevacizumab in vitrectomized than non vitrectomized eyes [18]. Moreover, intravitreal tamponade (silicone oil vs. gas) significantly influences the intravitreal drug distribution as well as drug clearance. But we used SF6 gas tamponade in both groups of the present study and the intravitreal anti-VEGF drugs were administered after the absorption of the gas bubble from the vitreous cavity during the post-operative follow-op visits on the basis of OCT changes. Here we followed the PRN protocol and noticed significant improvement in BCVA at 1 month and 3 months in both groups. Still, at 6 months, only in Group A. Mun Y et al. showed that monthly injections of Intravitreal anti-VEGF agents induced vision maintenance with favorable anatomical improvement in vitrectomized eyes with nAMD [19]. Although, visual improvement was seen in Group B at 6 months, it was not statistically significant (p = 0.378). There is also anatomical improvement in OCT morphology of macular status. CFT reduction was also statistically significant in both groups both at 1 month, 3 months and at 6 months. We have concluded that though both functional and anatomical improvement was considerable in both groups till 6 months, functional improvement was not corroborating with the anatomical improvement in diabetic eyes (Group B). Saito K et al. also showed no significant change in BCVA was noted with intravitreal brolucizumab in vitrectomized eyes with diabetic macular edema over one year despite a marked reduction in retinal thickness [20]. Probably long-term functional benefit of Brolucizumab in vitrectomised eye in patients with DMEs is not satisfactory less which may be due to pre-existing neurosensory retinal damage due to persistent intraretinal cystic changes. Injection-free interval was variable in these two categories in this present study. Mean injection-free interval was more in Group A (24.21 ± 14.36 weeks) and less in Group B (21.64 ± 7.66 weeks). One patient in Group A showed no recurrence of fluid on OCT till 76 weeks after a single injection of Brolucizumab (Fig. 1b). Long injection-free interval after Brolucizumab injection in both groups in the present study indicate the long durable anti-edema effect of the drug in vitrectomised eye. To the best of our knowledge, none of the previous intravitreal anti-VEGF molecule has proved its efficacy and durability in vitrectomised eyes. The main side effect of brolucizumab is intraocular inflammation (IOI). In the present study, no patients showed IOI. In this preliminary observation, it is difficult to predict any association between vitrectomized eyes and no-IOI after intravitreal Brolucizumab. We faced only two cases of sub-retinal haemorrhage (SRH,4.55%) as serious adverse events. This post-injection occurrence of SRH is probably due to the neovascular AMD/PCV disease process, rather than intravitreal brolucizumab injection. We considered this occurrence of SRH as a serious adverse event, though in the true sense, it was not a serious adverse event secondary to the intravitreal brolucizumab injection itself. The other adverse events were mostly non-serious. Registry trial of KESTREL and KITE study has already proved the efficacy of Brolucizumab in Diabetic macular oedema [9]. The current BIVE study finds Brolucizumab to be durable, effective as well as safe in the vitrectomised eyes. However, it is a non-randomised study and our analysis is a preliminary work to strongly conclude its efficacy in vitrectomised eyes. Confounding factors also could have influenced the outcomes. Some challenges we faced during our study design and statistical analysis. First of all, our sample size was small. Secondly, there was lack of comparator arm in the present study. Thirdly, a longer follow-up is necessary for safety, as IOI may occur late. We took the data up to 6 months as loss to follow up may bias our findings. Moreover, we have used the same dosage in our study like in non-vitrectomised eyes. Further studies are required to determine any alteration in the dosage of intravitreal Brolucizumab in vitrectomised eyes. The other factors that need consideration are existing choroidal neovascular membrane activity (in group A), intravitreal VEGF load as well as renal status (in group B) [21] which many times are responsible for recurrences. A further randomised study comparing the efficacy of Brolucizumab with the other available anti-VEGFs in vitrectomised eyes may be warranted to validate our observation, considering the regression analysis of these multiple factors in recalcitrant macular oedema cases. Moreover, before commenting on its universal application in all vitrectomised eyes, irrespective of the previous agent used, the procedure performed and pre-existing structural damages, we have to conduct a large-scale multi-centric RCT comparing its efficacy and safety with the other anti-VEGF agents.

Conclusion

Brolucizumab may have the potential to be durable, effective and safe in vitrectomised eyes in both post-nAMD and diabetic patients.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

SKB (Main Manuscript writing/ Data compilation/Figure/Table/Graph/ Review of the final manuscript). DA (Main Manuscript modification/ Data compilationFigure/Table/ Review of final manuscript). DC (Figure/Table/Graph/ Review of the final manuscript). AD (Graph/ Modification and Review of final manuscript). TKS (Modification and Review of final manuscript).

Funding

Self/ Hospital (Institution) as no funding or grant support was available.

Data availability

The datasets used and analysed during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Institutional Review Board (IRB) [ Disha Eye Hospitals Pvt Ltd Ethics Committee. Registration No. ECR/846/Inst/WB/2016/RR-19], and consent to participate was obtained from each patient. This study involved human subjects and informed consent was obtained from every patient after being informed about the procedure and its possible complications.

Consent for publication

Not applicable.

Authorship

All authors (SKB, DA, DC, AD, TKS) attest that they meet the current ICMJE criteria for authorship.

Competing interests

The authors declare no competing interests.