Comment on “Brolucizumab-associated intraocular inflammation in Indian patients by VRSI study group”

Dear Editor, We read with great interest the paper authored by the VRSI study group on the incidence and outcomes of brolucizumab-induced intraocular inflammation (IOI) among Indian patients with retinal vascular diseases.[1] While the study addresses a critical topic, we identified certain aspects that warrant clarification and further discussion to enhance the scientific rigor and transparency of the findings. The study presents representative images of IOI cases borrowed from a previously published study in your esteemed journal.[2] Although proper permissions may have been obtained, it raises the question of potential case overlap between the two studies. Specifically, the previous publication reported 13 cases of IOI in patients with neovascular age-related macular degeneration (nAMD), while the current study includes 12 such cases. Additionally, the study period reported by the VRSI study group (October 2020 to January 2023) overlaps significantly with that of the previous publication (October 2020 to April 2022) and includes data from three centers featured in the earlier study, suggesting the possibility of overlapping patient data. We believe that it is essential for the authors to confirm whether there is any overlap between these cases. Borrowing images from earlier work without presenting original images, despite having 21 cases available, is atypical in scientific literature unless there is a clear continuity of data between studies. In the absence of such overlap, the practice may raise ethical concerns. A further point of inquiry relates to the reported treatment intervals. The VRSI study mentions that all participating centers adhered to a pro re nata (PRN) regimen, with a minimum of eight-week intervals between brolucizumab injections. However, the median interval reported in the earlier study—where most cases appear to be included—was six weeks for eyes with IOI. This discrepancy raises an important question: how could the VRSI study ensure an eight-week injection interval in cases that were reportedly managed with shorter intervals in the earlier publication? Clarifying this inconsistency is crucial, as the previous study indicated a heightened risk of IOI with more frequent injections. Also, the VRSI study group is in an excellent position to confirm or refute this observed trend with appropriate statistical analysis, given the higher number of cases (n = 21) compared to the previous study (n = 13). We request the authors to explore this in the rebuttal with their existing dataset. Additionally, the VRSI study provides a unique opportunity to conduct comparative analyzes between the cases that developed IOI after the first brolucizumab injection (n = 12) and those that developed IOI after subsequent injections (n = 9). Such an analysis could reveal meaningful trends, offering more insight as opposed to the current classification of inflammation into mild, moderate, and severe categories. This inherently subjective grading is difficult to obtain from physician surveys and lacks a direct impact on clinical metrics such as presentation characteristics and outcomes. We suggest presenting these comparative results in tabular format, with appropriate P values, to allow for a clear interpretation of associations. Since the VRSI has an adverse event monitoring committee, it is assumed that they receive and archive data from all cases of IOI occurring in India. Then why were 21 centers explicitly chosen for this study? Also, there were 21 centers with 21 eyes reporting IOI. Does this mean that there was only one case per center? If there was more than one case at any center, then why were others (with no cases of IOI) involved, and how were they chosen to represent the entire region of India? We also note that data collection was performed using Google Forms, a platform typically employed for surveys rather than for detailed patient follow-up data. To ensure transparency, we recommend that the authors share the structure of the Google Form used, particularly the section dedicated to follow-up data collection. It would be helpful to know whether a separate form was filled out for each patient, and how follow-up data were managed, especially if multiple cases were reported from a single center. In terms of statistical reporting, it would be helpful if standard deviations (SDs) were provided alongside the means for continuous variables, including follow-up duration and visual acuity. Given the relatively small sample size, reporting medians with interquartile ranges might offer more robust insights. Also, since there is a good spread of treatment naïve eyes (no previous antiVEGF, n = 8) and IOI after the first injection (n = 12), authors can easily explore the relationship between these, the question to be answered being, do treatment naïve eyes have a lower (or higher) chance of IOI after the first injection of brolucizumab? An odds ratio (with a 95% confidence interval) will be most appropriate to show this association. Another important observation that impacts the incidence of IOI is that the VRSI study group mentioned that strict protocols were implemented to exclude patients with a history of uveitis. However, this is questionable as two eyes with a history of IOI from other anti-VEGF treatments and one eye within one month of cataract surgery received brolucizumab despite the pro-inflammatory state. This casts a shadow on the inclusion criteria used for the analysis. Excluding these three eyes would further reduce the reported incidence of IOI. Finally, the study reports that four eyes (19%) received repeat brolucizumab injections despite experiencing IOI, with no subsequent recurrence of IOI upon re-challenge. This observation is intriguing and underscores the complexity of IOI in real-world settings. While this finding is valuable, we caution against repeating such a practice without further evidence, as it may pose risks in routine clinical care. In summary, the VRSI study provides important insights into the occurrence of IOI with brolucizumab, but several methodological and reporting issues require further clarification. Addressing these concerns in a rebuttal would strengthen the validity of the findings and contribute meaningfully to the understanding of brolucizumab safety in clinical practice. Financial support and sponsorship: Nil. Conflicts of interest: Dr Sabyasachi Sengupta declares that he receives an honorarium as a speaker from Novartis India, Roche India, Allergan (an Abbvie company), and Lupin India.