Description and Management of Retinopathy of Prematurity Reactivation after Intravitreal Anti-VEGF Therapy

Abstract

Purpose of review:

To review the literature regarding reactivation of retinopathy of prematurity (ROP) after treatment with anti-vascular endothelial growth factor (anti-VEGF) agents.

Recent findings:

Reactivation can occur after anti-VEGF or laser. Risk factors for reactivation includes patient and disease related factors. Various studies are evaluating the use of different anti-VEGF agents and reactivation rates. However, the definition of reactivation varies between studies.

Summary:

The literature has varied definitions of reactivation, which is often used interchangeably with recurrence. It is important to recognize features of reactivation of ROP to appropriately manage patients and conduct clinical trials. The International Classification of Retinopathy of Prematurity (ICROP) 3^rd edition has established a consensus guideline regarding terminology describing reactivation.

Introduction

Retinopathy of prematurity (ROP) is the leading cause of preventable childhood blindness globally. [1] It is characterized by the arrest of vascularization in the retina leading to a proliferative vitreoretinopathy. [2] In utero, a physiologic state of hypoxia and a vascular endothelial growth factor (VEGF) - mediated environment drives normal vascularization. In preterm infants, the peripheral retinal vasculature is not fully developed, and at birth, a hyperoxic state leads to vasoconstriction and vaso-obliteration of blood vessels characterized as Phase 1 of ROP. In Phase 2, a hypoxic state induces the release of VEGF leading to pathologic angiogenesis. [3] Major risk factors for ROP include low birth weight, gestational age, and supplemental oxygen. [2] Screening guidelines in the United States include infants with birth weight of ≤ 1500 grams or gestational age ≤ 30 weeks. [4] ROP is classified according to the International Classification of ROP (ICROP) guidelines primarily based on stage, zone, and plus disease. [5] The purpose of this paper is to summarize the literature regarding reactivation after anti-vascular endothelial growth factor (anti-VEGF) treatment for ROP.

Treatment of ROP

The Cryotherapy for Retinopathy of Prematurity (CRYO-ROP) and the Early Treatment for Retinopathy of Prematurity (ETROP) trials established treatment guidelines for ROP. [6, 7] Options included cryotherapy or laser photocoagulation, with the latter becoming the preferred treatment method for Type 1 (treatment-requiring) ROP. [3] However, laser photocoagulation may prove to be challenging in cases of poor pupillary dilation and poor visualization secondary to media opacities. Furthermore, many resource-limited areas lack the capacity to support laser photocoagulation for ROP. Intravitreal anti-VEGF has also been used for treatment of ROP and can bypass many of the barriers to treatment with laser. [8] Other advantages include shorter procedure times and lower risk for myopia or astigmatism. In addition, anti-VEGF treatment had advantages for zone I or aggressive posterior ROP compared with laser. [9]

Several anti-VEGF agents, including bevacizumab, [9–12] ranibizumab, [13] aflibercept, pegaptanib, [14] and conbercept, [15,16] have been evaluated in the treatment of ROP as monotherapy or in comparison to laser therapy. Other studies have investigated its use in combination with laser [17, 18] or rescue therapy. The Bevacizumab Eliminates the Angiogenic Threat for Retinopathy of Prematurity (BEAT-ROP) evaluated 0.625 mg of intravitreal bevacizumab (IVB) vs laser and found a treatment effect for zone I disease but not zone II in patients treatment with IVB. [19] Wallace et al. compared various doses of bevacizumab and showed that 0.004 mg may be the lowest effective dose for ROP. [20] The Ranibizumab versus laser therapy for the treatment of very low birthweight infants with retinopathy of prematurity (RAINBOW) trial studied 0.2 mg vs 0.1 mg of intravitreal ranibizumab (IVR) vs laser. This trial found that the 0.2 mg dose was potentially superior to laser. [21] The 2 year results also demonstrated an acceptable safety profile. The RAINBOW extension study will investigate the long-term safety and efficacy of IVR. Intravitreal aflibercept compared to laser therapy for ROP are being evaluated by the BUTTERFLEYE and FIREFLEYE clinical trials. [22, 23]

Although anti-VEGF is used for treatment of ROP, the optimal drug and dose, ocular efficacy, need for retreatment, and long-term systemic effects are still under investigation. [24–28] Ocular complications of intravitreal anti-VEGF in retinal diseases include endophthalmitis, intraocular inflammation, rhegmatogenous retinal detachment, intraocular elevation, and ocular hemorrhage. Systemic adverse events include myocardial infarction, thromboembolic events, stroke, and hypertension as well as poor neurodevelopmental outcomes. [29, 30] For ROP, recurrence or reactivation of disease after anti-VEGF treatment is a concern.

Reactivation of ROP

Historically, there has been no consensus on the definition of ROP reactivation after anti-VEGF therapy, and studies that do define reactivation vary in their descriptions. Also, terms such as “recurrence” are frequently used interchangeably with “reactivation.” Reactivation has generally been defined by worsening signs, including appearance of a ridge, plus disease, neovascularization, or extraretinal fibrovascular proliferation after a period of initial regression, and often requires treatment. [31] Table 1 displays a comprehensive list of definitions from studies that evaluated reactivation.

Table 1:

Definitions of reactivation, recurrence, and persistence of ROP. Merged columns indicate that terms were used interchangeably in the study.

Study	Anti-VEGF agent	Reactivation	Recurrence		Persistence
Hu et al37	Ranibizumab	Any worsening signs, including an aggravated ridge or plus signs after the initial regression
Chen et al50	Bevacizumab	Reoccurrence of stage disease
Hu et al32	Bevacizumab	Arrest of anterior progression of retinal vasculature associated with a new demarcation line, ridge, or extraretinal fibrovascular proliferation (EFP) or leakage on fluorescein angiography, with or without recurrence of plus disease. Recurrence implied reactivation after a period of regression			Lack of adequate regression after treatment
Wong et al24	Ranibizumab vs bevacizumab	Any worsening in stage of ROP after initial regression
Chuluunbat et al51	Ranibizumab		Regression of ROP but subsequent reappearance of plus disease, preretinal and vitreous hemorrhage, worsening of retinal neovascularization or progression to retinal detachment
Jin et al16	Conbercept vs ranibizumab	Recurrence of retinal abnormality such as ridge and plus disease after the regression of them
Ling et al35	Laser vs Ranibizumab vs bevacizumab			Initial positive response but later deterioration in retinal status, such as reappearance or worsening of plus disease, neovascularization, and requiring re-treatment
Lyu et al36	Ranibizumab			Reappearance of EFP in eyes after an initial resolution of acute-phase ROP – treated Plus disease-sign of activity of recurrence line or a ridge- observed
Mintz-Hittner et al34	Bevacizumab			Return of plus disease and neovascularization, extent of retinal vascularization
Bai et al	Conbercept	Recurrence of ridge or plus disease

Similar to plus disease, there is likely inter-expert variability in the diagnosis of reactivation especially since multiple criteria have been described to characterize this entity. Given the lack of a standardized definition and criteria for reactivation, it is not surprising that there are differences in the management and treatment criteria for reactivated ROP. In addition, small sample sizes limit several studies that evaluated recurrence, making it difficult to generalize results. Differences between studies in terms of definition of reactivation, type of anti-VEGF agent, dosing, and follow up duration make it challenging to directly compare studies. Retrospective studies are also limited as to whether ROP was completely regressed or if it was persistent. [32]

The ICROP 3^rd edition provides a clearer definition of reactivation, defined as recurrence of acute phase features including a range of signs from a new demarcation line to reactivated stage 3 with plus disease, vascular changes such as recurrent vascular dilation and/or tortuosity or new extraretinal vessels. [33] See Figure 1 for examples of reactivated ROP. The guidelines also discuss documentation of reactivation, which should include zone and stage with the modifier “reactivated”.

Figure 1:

Digital fundus images and fluorescein angiography of a patient with reactivation after anti-VEGF

Patient at 48 weeks post-menstrual age (PMA) after intravitreal anti-vascular endothelial growth factor (anti-VEGF) injection × 3 with recurrent neovascularization. Digital fundus images of the posterior pole of the right eye (A) and left eye (D) are shown. Peripheral fundus images of the right eye (B) and left eye (C) are also shown. (B). Fluorescein angiography shows evidence of new areas of new neovascularization (white arrows) in the right eye (C) and left eye (F). The white asterisk represents areas of peripheral avascular retina.

Risk factors for Reactivation

Multiple risk factors for reactivation after anti-VEGF have been identified in various studies. Patient factors related to immaturity include low gestational age, low birth weight, early PMA, and low Apgar scores. [34–35] Disease factors include zone I ROP, [25, 35] extensive retinal neovascularization, [36] preretinal hemorrhage before treatment, [37] and aggressive posterior ROP. [34] Maternal factors include multiple births. [35] Other risk factors that have been implicated in reactivation of ROP include supplemental oxygen requirements after treatment and longer duration of hospitalization. [34, 36]

Comparison of agents in reactivation

Multiple studies have compared incidence of reactivation between various anti-VEGF agents and laser. Anti-VEGF agents that have been used for ROP include intravitreal bevacizumab (IVB), ranibizumab (IVR), aflibercept (IVA), pegaptanib (IVP), and conbercept (IVC). Reports discuss use of anti-VEGF as monotherapy, in combination with laser, or as salvage therapy. IVB was shown to have increased rates of reactivation compared to laser, [10] and IVR monotherapy was shown to have even higher rates of reactivation than laser or IVB. [13, 24] The reactivation rates after IVB have ranged from 4–14%, IVR 4.3–52%, IVA 7.7%, [38] IVP 11.7%, [14] and IVC 0–16.7%.[15, 16] Effects of anti-VEGF levels may be transient given the short half-life of these agents, potentially increasing the risk for reactivation. [24] Serum VEGF levels were shown to be reduced after IVB and IVR.[39, 40]

Reactivation interval

Among the considerations after ROP treatment is the need to surveillance for reactivation. Reactivation has been reported to occur at varying time periods following laser and anti-VEGF treatment. The term reactivation interval is also not very clearly defined in the literature. Studies such as the BEAT-ROP study refer to this interval as the time from initial treatment until recurrence of treatment-requiring ROP. [19] Most evidence indicates that reactivation occurs 37–60 weeks postmenstrual age (PMA). [33] Reactivation interval also varied by study. The original BEAT-ROP study showed that the reactivation interval was approximately 19 weeks after IVB. [19] The follow up study showed an interval of 16 weeks. Late reactivation of ROP has been described 4 to 24 months after treatment with reports of progression to tractional retinal detachment [41, 42] or during adolescence or adulthood. [43] Eyes treated with anti-VEGF require long term follow up and monitoring. Reactivation after laser was thought to occur within the first 9 weeks after treatment. [44] In the Autrata et al study, reactivation interval after IV pegaptanib was 15.1 weeks. [14]

Treatment of reactivation

Currently, there is no clear consensus on how or when reactivation is treated. Laser or repeat anti-VEGF injections alone, [19, 24] and as combination therapy, have been trialed. [45] It is possible that not all cases of reactivation require treatment and instead close observation may be preferred. [36, 46–48] Therefore, the decision on how or when to treat reactivation is based on the physician’s clinical judgment. A treatment algorithm was suggested by Martinez-Castellanos et al. (Figure 2) where treatment modality is selected based on the reactivation feature. For example, the authors recommended repeat intravitreal anti-VEGF for flat vessels and a combination of anti-VEGF and laser for neovascularization. For fibrovascular proliferation or vitreous traction, vitrectomy with laser and intravitreal anti-VEGF was suggested. [45] A well-controlled prospective clinical trial evaluating different degrees of reactivation and treatment response would be ideal to further guide clinical decision making.

Figure 2:

Proposed algorithm of management of reactivation of ROP

Adopted from Martínez-Castellanos MA, González-H León A, Romo-Aguas JC, Gonzalez-Gonzalez LA. A proposal of an algorithm for the diagnosis and treatment of recurrence or treatment failure of retinopathy of prematurity after anti-VEGF therapy based on a large case series. Graefes Arch Clin Exp Ophthalmol. 2020;258(4):767-772. doi:10.1007/s00417-020-04605-y

Permissions obtained.

Future directions

There is a need for a standardization of the guidelines on how to manage reactivation. The use of ultrasound and fluorescein angiography could be useful in identifying various features of reactivation not readily apparent on fundus examination. [45] Optical coherence tomography has also provided valuable insight into the pathophysiology of ROP. [49] Furthermore, telemedicine can be utilized in areas where there is a shortage of trained ophthalmologists or a lack of resources, allowing for timely identification and management of ROP reactivation. Artificial intelligence algorithms could also be trained to identify specific features of reactivation and treatment patterns.

Conclusion

Reactivation of ROP is important to identify but has been defined in various ways in the literature. The ICROP 3^rd edition provides a consensus definition and further details regarding reactivation of ROP. Further studies are needed to characterize reactivation of ROP, risk factors, and reactivation intervals after anti-VEGF. A standardized treatment protocol is needed to ensure proper management of reactivation and to obtain a better understanding of the outcomes of reactivated ROP after treatment.

Key points:

• Multiple definitions for reactivation of ROP have been described in the literature. The ICROP 3^rd edition provides a consensus definition for reactivation of ROP.

• Multiple anti-VEGF agents have been used in ROP, but we are still learning about reactivation risk and interval with these agents.

• There is still a need to develop consensus guidelines for the treatment of reactivation of ROP.