Brolucizumab-associated intraocular inflammation in Indian patients by VRSI study group

Abstract

Context:

Concerns about brolucizumab’s (Pagenax^®) association with intraocular inflammation (IOI) limit its use despite its cost-effectiveness and efficacy. This multicentric study analyzes IOI incidence across 21 tertiary eyecare centers in India since its introduction in October 2020.

Purpose:

To determine the real-world incidence rate of IOI in Indian patients secondary to intravitreal brolucizumab across 21 tertiary eye care centers in India.

Settings and Design:

Retrospective multicentric, survey-based study.

Methods:

Data including number of patients treated, clinical indications, side effects encountered, and IOI case details was collected via Google Forms in 21 Indian tertiary eye care centers since October 2020. Mean, median, frequency, and standard deviation were calculated for statistical analysis.

Results:

All centers used pro re nata protocol for brolucizumab injections with a minimum injection interval of 8 weeks. The incidence of IOI was 0.79% (21 events out of 2655 eyes). Treatment indications included idiopathic polypoidal choroidal vasculopathy, neovascular age-related macular degeneration, diabetic macular edema, and off-label uses. IOI was experienced after the first injection (57%) in majority of cases with a median onset of 14 days (range: 1–65 days). IOI was mild in 28.5%, moderate in 33%, and severe in 38% of cases. Eighteen out of 21 IOI eyes recovered preinjection best corrected visual acuity or better.

Conclusions:

Our study found a lower IOI incidence (0.79%) with brolucizumab (Pagenax) in Indian patients compared to previously reported literature. IOI events were mostly mild to moderate, and post-treatment, most patients improved or maintained BCVA. Larger prospective multicentric studies with PRN dosing protocol are needed to confirm these findings.

Intravitreal anti-vascular endothelial growth factor (anti-VEGF) drugs are the mainstay of treatment for various retinal pathologies such as neovascular age-related macular degeneration (nAMD) and diabetic macular edema (DME). Phase III trials have demonstrated their efficacy with either monthly or two-monthly dosing schedules.[1,2] However, in real-world clinical practice, it is difficult to have good patient compliance with monthly follow-up or injections. Thus, an effort is being made to come up with drugs which have more durable action, thereby reducing the frequency of visits of the patients to the clinic without compromising the efficacy or the final visual outcome.

To address the challenges associated with frequent dosing, a novel humanized single-chain antibody fragment named RTH258, later known as brolucizumab, has been introduced. In the first human trial with this drug, it was found to be non-inferior to ranibizumab and it also offered the advantage of an extended duration of action, reducing the need for frequent injections.[3] Brolucizumab’s small molecular size (26 kDa), high stability, and solubility allow the administration of 6 mg in a single 50 µl intravitreal injection, enabling the delivery of a higher molar dose within the same volume as current vascular endothelial growth factor (VEGF) inhibitors in clinical use, thereby ensuring a longer duration of action.[4,5] Marketed as Beovu^® by Novartis (Basel, Switzerland), brolucizumab received Food and Drug Administration (FDA) approval for nAMD in October 2019 and for DME in June 2022. The Drug Controller General of India granted approval for the drug, which has been available as Pagenax^® (Novartis India Ltd, Mumbai, India) since October 2020.

Despite the favorable outcomes demonstrated in the HAWK and HARRIER Phase 3 trials, where brolucizumab showed superior anatomic results and noninferior visual outcomes compared to aflibercept in nAMD patients, there was a concern regarding its association with intraocular inflammation (IOI).[6] A post hoc analysis of these trials reported an IOI rate of 4.5%, with 2.1% of eyes experiencing concomitant vascular occlusions.[7] Consequently, FDA approved a label update for brolucizumab on June 11, 2020, including additional safety information.[8]

The largest real-world analysis, which included 10,654 and 11,161 eyes from the Intelligent Research in Sight (IRIS) Registry and Komodo Health database, respectively, found an overall incidence of IOI and/or retinal occlusion to be 2.4%, with certain risk factors such as prior ocular inflammation, occlusion, and female sex associated with higher risk.[9] This lower incidence of IOI, compared to the controlled clinical trials like HAWK and HARRIER, highlighted the significance of real-world data in assessing the safety of brolucizumab. A recent multicentric study conducted in centers across eastern India by Chakraborty et al.,[10] involving 758 injections of brolucizumab for nAMD, reported the incidence of IOI as 1.7% (13 IOI events out of 758 injections), with vascular occlusion in one eye. The above study predominantly reported data from eastern India; in contrast, our study had centers across various parts of the country. Secondly, they calculated the IOI incidence out of the total number of brolucizumab injections used rather than the total number of eyes treated by brolucizumab as reported by the global clinical trials such as HAWK and HARRIER, and KESTREL and KITE.[7,11]

This is a retrospective multicentric survey-based study conducted to determine the real-world incidence of IOI secondary to intravitreal brolucizumab used in Indian patients across 21 tertiary eye care centers located at different geographic locations of the country.

Methods

The study was approved by the ethics committee of the respective participating centers, and it adhered to the tenets of the Declaration of Helsinki. This was a multicentric survey-based study performed by the VRSI (Vitreo-Retinal society-India) study group amidst its members/member institutions. The survey was circulated across 21 tertiary eye care centers located in various states and union territories of India (New Delhi [6], Rajasthan [1], Uttar Pradesh [2], Karnataka [3], Kerala [2], Maharashtra [2], Telangana [1], and West Bengal [3]).

To collect the data of the patients who received intravitreal brolucizumab (Pagenax, Novartis India Ltd) from October 2020 to January 2023, a structured questionnaire was developed using Google Forms and an online survey platform was used and sent via email to all the participating centers. The questionnaire included a section on the number of eyes injected and the clinical indications for which Pagenax was given. Another section of the form was dedicated to the observed side effects. The last section included the clinical details of the eyes with IOI, which included the clinical presentation with visual acuity, relevant past ocular and systemic history, the time interval between the injection and the onset of IOI, laterality of inflammation, grade of inflammation, management done, and the final visual outcome and if the treatment plan with anti-VEGF drugs was modified after IOI was encountered.

The grade of inflammation was classified as per the criteria for adverse events (AEs) used in the HAWK and HARRIER study as follows:[6,7]

1. mild: when the patient is aware of the AE, but can easily tolerate the sign or symptom;

2. moderate: if the sign or symptom results in discomfort significant enough to cause interference with the patient’s usual activities; and

3. severe: if the sign or symptom is incapacitating and results in inability of the patient to work or engage in their usual activities.

Statistical analysis was performed using the Microsoft^® Excel for Mac, version 16.66.1. The mean, median, frequency, and standard deviation were calculated.

Results

A total of 21 centers across India participated in this survey-based study, comprising 13 private centers, six nongovernmental organizations, and two government centers. A total of 5046 brolucizumab injections were administered in 2655 eyes from October 2020 to January 2023. The indications for treatment were idiopathic polypoidal choroidal vasculopathy (PCV), nAMD, and DME. Other off-label indications were idiopathic macular neovascularization (MNV), retinal venous occlusions, pachychoroid neovasculopathy, post-vitrectomy macular edema, Coats disease, vitelliform degeneration with cystoid macular edema, and chronic central serous chorioretinopathy (CSCR). In 14 out of the 21 centers, brolucizumab was administered for both treatment-naïve eyes and the eyes unresponsive to other anti-VEGFs, while in the remaining seven centers, it was administered to resistant cases.

The average follow-up duration was 6 months (range: 3–9 months). All participating centers followed a pro re nata protocol from the initiation of injection, with an interval of at least 8 weeks between two consecutive injections irrespective of the indication.

The percentage prevalence of IOI during the study period was 0.79% (21 events out of 2655 eyes). None of the eyes had vascular occlusion.

Among the eyes with IOI, one had a history of prior cataract surgery 1 month back. Two had a history of prior IOI with other anti-VEGFs, one with ranibizumab and the other with bevacizumab. Intravitreal brolucizumab was given unilaterally in 17 eyes (81%) and bilaterally in four eyes. In all patients, IOI occurred unilaterally. There was history of diabetes mellitus, hypertension, and coronary artery disease in one patient each.

The clinical indications for using brolucizumab in the eyes which had IOI were nAMD (12 eyes), PCV (five eyes), DME (one eye), Coats disease (one eye), idiopathic choroidal neovascular membrane (CNVM) (one eye), and branch vein occlusion-related macular edema (one eye). Thirteen eyes (62%) had a history of previous anti-VEGFs, whereas eight eyes (38%) were treatment naïve. The median number of anti-VEGF injections received before brolucizumab was 10 (range: 3–30 injections).

IOI related AE occurred after the first injection in 12 eyes (57%), after the second injection in six (28%) eyes, after the third injection in two eyes (9%) and after the fourth injection in one eye.

The median number of days between IOI and brolucizumab injection was 14 days (range: 1–65 days). The time interval between brolucizumab injection and the onset of mild inflammation was 32 days (range: 4–60 days) and for moderate inflammation, it was 30 days (range: 2–65 days). For eyes with severe grade of inflammation, the median number of days between injection and the onset of IOI was 14 days (range: 1–21 days).

The presenting complaints were blurring or diminution of vision in 16 eyes (76%), floaters in five eyes (24%), redness in three eyes (14%), and ocular pain in four eyes (19%). IOI was noted to be mild in six eyes (28.5%), moderate in seven eyes (33%), and severe in eight eyes (38%). Figs. 1 and 2 highlight the various clinical presentations of IOI and their response to treatment.

Figure 1.

(a) slit lamp photograph showing AC cells at presentation (b) wide field fundus photograph showing dense central vitritis with vascultis in peripheral vessels (black arrow) at presentation (c) wide field fundus photograph showing resolving vitritis after receiving oral and topical steroid for 3 weeks (d) Wide-field angiography (mid-phase) images showing mid disc hyperemia but no signs of vasculitis at 3 weeks. [Reproduced with permission from: Chakraborty D, Mondal S, Sengupta S, Abbas Z, Chandra K, Boral S, et al. Incidence, clinical features, risk factors, and outcomes of intraocular inflammation following brolucizumab in Indian eyes –A multicentric study. Indian J Ophthalmol 2023;71:1979-85]

Figure 2.

(a) Magnified slit lamp photograph showing keratic precipitates (KPs) (b) Slit lamp photograph after 2 weeks showing reduction of intra-ocular inflammation. [Reproduced with permission from: Chakraborty D, Mondal S, Sengupta S, Abbas Z, Chandra K, Boral S, et al. Incidence, clinical features, risk factors, and outcomes of intraocular inflammation following brolucizumab in Indian eyes –A multicentric study. Indian J Ophthalmol 2023;71:1979-85]

The median best corrected visual acuity (BCVA) at baseline before brolucizumab injection was 0.6 log of minimum angle of resolution (logMAR; range: 0.2–1 logMAR), that is, Snellen’s equivalent of 6/24.

Among the six eyes with mild inflammation, the median preinjection BCVA was 0.6 logMAR (6/24) (range: 0.5–1) and the median post-IOI BCVA at 1 month was 0.7 logMAR (6/30) (range: 0.5–1.8). At the end of 3 months, BCVA in all mild IOI eyes was equal to or better than pre-IOI BCVA.

In the seven eyes with moderate inflammation, the median preinjection BCVA was 0.6 logMAR (6/24) (range: 0.2–0.8) and the median post-IOI BCVA at 1 month was 0.3 logMAR (6/12) (range: 0.2–0.8). At the end of 3 months, BCVA for six out of seven moderate IOI eyes was equal to or better than pre-IOI BCVA. The eye which experienced a one line drop in vision from a preinjection BCVA of 0.2 logMAR (6/9) to a post-IOI BCVA at 3 months of 0.3 logMAR (6/12) was a case of treatment-naïve branch retinal vein occlusion associated macular edema, which was treated with topical steroids. One eye with moderate inflammation experienced intraocular pressure (IOP) rise, which was controlled with topical IOP-lowering drops.

In the eight eyes with severe inflammation, the median preinjection BCVA was 0.8 logMAR (6/36) (range: 0.5–1.2) and the median post-IOI BCVA at 1 month was 1.3 logMAR (6/120) (range: 0.8–1.3). At the end of 3 months, BCVA for six out of eight eyes was equal to or better than pre-IOI BCVA. In patients with vision drop, one eye was a treatment-naïve PCV with preinjection BCVA of 1 logMAR (6/60), which dropped to a BCVA of 1.8 logMAR (Finger counting at 1 metre [FC-1 m]) immediately post-IOI. Prompt vitrectomy and oral steroids were given, but the patient was lost to follow-up. The second eye was a treatment-naïve nAMD with preinjection BCVA of 0.6 logMAR (6/24), and the patient presented with intense inflammation with BCVA 1.8 logMAR (FC-1 m). The patient was treated with oral and topical steroids and regained a BCVA of 0.8 logMAR (6/36).

Five out of six eyes with mild inflammation were treated with topical steroids, and one patient received additional oral steroids.

Five out of seven eyes with moderate inflammation were treated with oral as well as topical steroids, with one eye requiring vitrectomy and one eye treated with topical steroids only.

One eye experienced IOP rise, requiring IOP-lowering drugs.

Five out of eight eyes were treated with oral and topical steroids, with two eyes requiring additional vitrectomy and one eye treated with posterior sub-Tenon triamcinolone injection.

Most eyes (15 out of 21 eyes) were shifted to another anti-VEGF after the occurrence of IOI with brolucizumab. Of the remaining eyes (six eyes), one eye with treatment-naïve PCV did not require further injections post-IOI episode, whereas one patient was lost to follow-up. Interestingly, four eyes, one with mild IOI and three with severe IOI, were continued with brolucizumab without any modification of treatment such as combining brolucizumab with oral or topical or intravitreal steroids. One out of the three eyes with severe inflammation was being treated for resistant nAMD with a history of cataract surgery 1 month before the first brolucizumab injection and presented with an endophthalmitis-like appearance, which was treated with prompt vitrectomy and switched to other anti-VEGF post-IOI. However, this eye had to be switched back to brolucizumab because of poor response to the other anti-VEGF given.

On further follow-up, none of these four eyes experienced a recurrence of IOI on repetition of brolucizumab during subsequent visits.

Discussion

Brolucizumab is the new anti-VEGF drug for the management of various retinal pathologies. This retrospective, multicentric, survey-based study reports the real-world incidence of IOI following intravitreal brolucizumab injection given for both approved indications such as nAMD, DME and off-label indications such as CSCR, retinal venous occlusions, Coats disease, post-vitrectomy macular edema, idiopathic MNV, vitelliform degeneration with macular edema in patients of Indian origin.

This study shows an incidence of IOI as 0.79%, which is much lesser than that reported previously.[7,9,10,12,13] Large clinical trials such as HAWK and HARRIER, and KESTREL and KITE reported an IOI incidence of 4.5% and 5.3%, respectively.[7,11] The OCTOPUS and SWIFT trials for nAMD reported an even higher incidence rate of 10.5%, which they attributed to proactive monitoring for IOI-related signs and symptoms.[14] Real-world studies from Caucasian populations report IOI rates of 2.4% (IRIS registry and Komodo database) to 4.6%.[9,13] Higher incidence rates of IOI ranging from 10% to 22% were reported from Asian studies (Japan and Korea).[15,16,17,18] A brief summary of major Asian studies is presented in Table 1.

Table 1.

A summary of major Asian studies on brolucizumab-induced intraocular inflammation

Citation		Country		Treatment naïve/resistanta:		Total no. of eyes		Clinical indication		Treatment regimen		IOI events		Visual outcome
Chakraborty et al. 2023[13]		India		NA		NA		nAMD		Pro re nata		13 (758 injections, 1.7%)		11/13 recovered vision
Kim et al. 2023[15]		Korea		20 eyes (6.8%) treatment naïve and 274 eyes (93.2%) resistant		294		nAMD		Loading dose		41 (13.9%)		39 out of 41 recovered vision
Bae et al. 2022[16]		Korea		All resistant		34		nAMD		Pro re nata		5 (14.7%)		All recovered vision
Yeom et al. 2023[17]		Korea		All resistant		81		nAMD		Treat and extend		8 (9.9%)		All recovered vision
Matsumoto et al. 2022[18]		Japan		All treatment naïve		68		nAMD		Treat and extend		15 (22.1%)		All recovered vision
Bilgic et al. 2021 (PROBE study)[19]		India		All treatment naïve		27		nAMD		Pro re nata		0		NA
Hirano et al. 2023[20]		Japan		All resistant		23		DME		Single injection		0		NA
Kikushima et al. 2023[21]		Japan		All resistant		60		nAMD		Treat and extend		12 (20%)		All recovered vision
Hsu et al. 2022[22]		Taiwan		All resistant		10		nAMD and PCV§		Single injection		0		NA
Nam et al. 2022[23]		Korea		All resistant		26		PCV		Pro re nata		0		NA
Fukuda et al. 2021[24]		Japan		All treatment naïve		14		PCV		Loading dose		2 (14%)		One had persistent field loss due to BRAO
Maruko et al. 2021[25]		Japan		43 eyes treatment-naive and 84 eyes resistant		127		nAMD		Loading dose		12 (9.4%)		All recovered vision
Bilgic et al. 2021 (REBA study)[26]		India		25 were treatment naïve rest resistant		105		nAMD		Loading dose for treatments naive and monthly for resistant group		0		NA
Chakraborty et al. 2021 (BRAILLE study)[27]		India		20 eyes treatment naïve, 74 eyes resistant		94		nAMD		Pro re nata		0		NA

DME=Diabetic macular edema, IOI=Intraocular inflammation, nAMD=Neovascular age-related macular degeneration, PCV=Polypoidal choroidal vasculopathy. ^aResistant refers to cases that were switched to brolucizumab after poor response to other anti-VEGFs, BRAO=Branch retinal artery occlusion

Most of the clinical trials followed a regimen of initial monthly or six-weekly loading dose of brolucizumab injections, followed by 12-weekly or Pro re nata (PRN) basis injections.[6,11,14] This has been questioned by certain real-world studies from India, such as the PROBE study, which followed a PRN regimen from the start in 27 treatment-naïve neovascular AMD eyes and found good visual and anatomic outcomes with no IOI occurrence.[19] A recent series of 758 injections by Chakraborty et al.,[10] which followed a PRN protocol from the start, found an IOI-related incidence of 1.7% (per injection).

As recommended by Bilgic et al. in the PROBE study and Chakraborty et al., all the participating centers in our study used a PRN protocol from the time of initiation of treatment and there was a minimum interval of 8 weeks between two subsequent injections of brolucizumab, irrespective of the retinal pathology.[10,19] The lower dosing used may possibly cause lesser immunogenicity, and hence is possibly safer.

Racial variation in ocular inflammation has been previously reported. For instance, the Pacific Ocular Inflammation Study found a significantly higher prevalence of uveitis in white populations.[28] Asian-Pacific countries, including Japan, China, and Taiwan, show a high prevalence of noninfectious and idiopathic uveitis conditions.[29] These differences may explain higher incidence of IOI observed in Caucasian populations and Asian studies, particularly in Japan, in comparison to our study.

Another cause for the lower incidence rate of IOI in our study may be the lack of a proactive approach in early diagnosis of IOI as followed by OCTOPUS and SWIFT trials, leading to underdiagnosis of mild IOI following brolucizumab, which may be asymptomatic.[14] All participating centers also followed a strict protocol of excluding patients with a history of uveitis or underlying autoimmune conditions in accordance with the guidelines issued for patient selection for treatment with brolucizumab.[30,31,32] This may also have resulted in lower IOI incidence. A recent post-marketing survey analysis showed decline in overall retinal vasculitis and/or retinal vascular occlusion cases after the safety signal was issued, which further supports this reasoning.[33]

The majority of IOI cases in our data (12 out of 21, or 57%) experienced it after the first injection, with 28% experiencing it after the second injection (six eyes), two eyes experiencing IOI after the third injection, and one eye having it following the fourth injection. Our results were similar to those of Khanani et al.[9] who found 49% (256 out of 523) IOI events after the first injection, 39% (205 out of 523) IOI events after the second injection, and 12% IOI events after the third or more than three injections (combined IRIS registry and Komodo data). Similar findings were reported by Bodaghi et al.[14] in the OCTOPUS and SWIFT trials, which involved proactive screening for IOI and found most IOI-related events (81.1%) occurring during the loading phase after a mean of 2.6 (standard deviation [SD] = 1.3) brolucizumab injections.

Preexisting local serum antibodies to brolucizumab have been noted in 36%–52% of treatment-naïve patients, which is higher than that observed with other anti-VEGF agents such as ranibizumab, where it is 4%–5%.[34] These preexisting antibodies cause a type 3 hypersensitivity reaction causing inflammation in patients receiving the first dose of the injection.[35,36] Since the median number of days between the injection and IOI in our study was 14 days (range: 1–65 days), this supports the hypothesis of a delayed hypersensitivity reaction.[35]

Majority of the cases were mild to moderate IOI (13 out of 21 eyes), with eight eyes reported as having severe IOI. As reported in literature that eyes with mild to moderate IOI resolve without any sequelae, in our series also 12 eyes with mild to moderate IOI and six eyes with severe IOI resolved completely with recovery of vision to pre-injection BCVA.[7,12] One eye with moderate IOI and one eye with severe IOI experienced a one line drop of vision from pre-injection BCVA at 3 months and one eye which presented with severe inflammation was lost to follow-up. In our series, occlusive vasculitis was not reported in any case.

As per the prescribing information, brolucizumab is to be discontinued during active IOI.[34] However, there are no clear guidelines on resuming anti-VEGF treatment after resolution of the IOI episode. Singer et al.,[7] in their post hoc analysis of the HAWK and HARRIER study, reported that after the first IOI-related AE, approximately 74% of eyes in the trial received further brolucizumab injections, and most of these eyes (24/36 eyes [approximately 67%]) completed the study with brolucizumab only and achieved an overall BCVA gain by the end of study. Recurrence, worsening of IOI, and/or occlusive vasculitis have been reported in literature following the switch back to another anti-VEGF.[14,35,37] In our study, the 15 eyes that were switched to another anti-VEGF did not have recurrence or worsening of IOI during the follow-up period. The four eyes that were continued on or switched back to brolucizumab following resolution of the IOI episode also reported no recurrence of IOI during the follow-up period.

To the best of our knowledge, this study has the largest number of eyes of Indian origin treated with brolucizumab for various indications and analyzed for the incidence of IOI. The other strength of the study is that it was a multicentric study including centers from different geographic locations of the country.

The limitations of this study are that it is a survey-based retrospective study. Secondly, the follow-up of our patients ranged between 3 and 9 months, which may not include the incidence of late IOI or recurrence of IOI on subsequent follow-up visits. Thirdly, it included patients in whom brolucizumab was used in off-label indications. Lastly, risk factor analysis could not be carried out due to the small number of IOI cases as well as lack of detailed clinical information of the non-IOI patients.

Conclusion

To conclude, the incidence of IOI in our study was found to be much lesser (0.79%) in Indian eyes in comparison to that reported in Caucasians (2.4%), and we hypothesize that this possibly could be related to a different immune response in Indian patients.[9] The IOI-related AEs were mild to moderate in majority of the patients, and nearly all eyes after treatment of IOI regained or became better than preinjection BCVA at 3 months follow-up visit. Larger multicentric prospective studies maybe required in future to further confirm the incidence of IOI in patients treated with brolucizumab, including a proactive monitoring of the treated patients.

Financial support and sponsorship:

Nil.

Conflicts of interest:

There are no conflicts of interest.