Retinopathy of prematurity classification updates: possible implications for treatment

It seems indisputable that the treatment for a condition should be based on the current description and classification of the condition. But here, we ask: is this always so for treatment of severe retinopathy of prematurity (ROP)? The current indications for ROP treatment were defined in 2003 by the Early Treatment for ROP (ETROP) trial, as type 1 ROP, which includes stages 1 or 2 with plus disease in zone I, stage 3 with or without plus in zone I, and stages 2 or 3 with plus in zone II.¹ ETROP also defined type 2 ROP that is somewhat less severe but requires increased surveillance. Since then, the International Classification of ROP (ICROP) has been updated twice (in 2005² and 2021³), and critically new treatment modalities, such as antivascular endothelial growth factor agents, have been developed.^4–6 Yet clinical practice and research are still based on earlier iterations of the ICROP,^2,7 and there is evidence that some clinicians are treating severe disease more aggressively than recommended by ETROP.^8–11 Our purpose here is to open a discussion on how the ROP classification updates might affect the indications for ROP treatment, and whether potential changes require collaborative research and/or expert consensus opinion.

ICROP3 refined several of ROP’s ophthalmoscopic features.³ In the absence of recent prospective treatment studies, we believe that ICROP3 presents a timely opportunity for discussion along two lines. First, do revisions in ICROP3 provide an opportunity to revisit and refine treatment indications? Second, how should changes provided in ICROP3 be incorporated into current clinical practice?

Zone: Notch and Subdivision of Zone II

Zone is a major predictor of ROP incidence and severity,¹² but in the absence of retinal landmarks, it has always been one of the most worrisome aspects of ROP categorization.¹³ One purpose of ICROP3 is to reduce ambiguities that may affect diagnostic precision. Zones describe the anteroposterior location of vascularization and ROP by 3 circular zones, the most posterior location of the ROP defining the zone for the eye.

ICROP3 raises two issues. First, zone II should be subdivided into posterior and anterior sections on the basis that the former has greater likelihood to require intervention compared to ROP in anterior zone II. Second, does a notchlike incursion of ROP into the more posterior zone define the overall zone for the eye? Prior to ICROP3, zone was interpreted clinically as the most posterior location of disease and this was used as such in ETROP—the notch was neither considered nor recorded. One topic for discussion is whether an eye with ROP stage 3 in “zone I secondary to a notch” in the absence of plus disease should be treated, as it fulfills the criteria of type 1 ROP based on findings from the ETROP study.

Stage and Aggressive ROP (A-ROP)

The description and inclusion of aggressive posterior ROP (AP-ROP) in the classification of 2005 was a major advance.² However, since that time it has become apparent that AP-ROP is not, as originally thought, exclusively posteriorly located and a specific isolated disease; rather, it can coexist with stage 3 ROP (“classic ROP”), that is, stage 3 and A-ROP lie along a spectrum of proliferation from “flat neovascularization” to classic stage 3. This broader description adds a degree of diagnostic complexity but is highly relevant, because it reflects the condition as it presents across the world.

Evolution of Plus Disease

Plus disease was included in the first international classification of ROP (1984)⁴ but subsequently, the dichotomous categorization of normal vs. plus disease was deemed inadequate and led in 2005 to the introduction of an intermediate level of posterior vascular abnormality (pre-plus disease).² This abnormality was not retrospectively incorporated in the 2003 ETROP indications for treatment. In addition, grading between plus, pre-plus, and normal appearing posterior vessels was reported to vary between observers, albeit with closer agreement at the normal and severe ends.^13–19 It is an inevitable consequence of increasing the diagnostic granularity of plus disease diagnosis that the categorization of certain clinical features and images might change. For example, Figure 1, in 2005 was designated as preplus, but in 2021 the same image was judged by a panel of experts to represent plus. Compounding this change is the reported recent trend to diagnose preplus and plus at an earlier stage of the disease.¹⁹ But what is the effect of these evolutionary and ICROP3-induced changes on the decision to treat?

FIG 1.

This image was designated in 2005² as preplus, but in 2021,³ this same image was judged by a panel of experts to represent plus disease. Permission to reproduce previously published image from Arch Ophthalmol 2005;123:991–9.

The use of artificial intelligence–based or computer-assisted objective methods that can be easily applied at the bedside to grade the vascular changes associated with ROP is eagerly awaited. In the meantime, ICROP3 acknowledges that the changes do not progress in a stepwise fashion, and the newly introduced traffic light schema demonstrates visually that vascular severity forms a continuous spectrum that has increased granularity in clinical diagnosis.³

Possible Implications for Practice

After the ETROP trial, treatment guidelines (ie, type 1 ROP) place the diagnostic weight on three features: zone, stage, and plus, with the relevance of disease extent in sectors considered less important. However, all parameters of disease may be inextricably linked with regard to pathophysiology. We therefore question the logic of not including one ROP component when assessing disease severity. In our opinion, plus disease, however carefully clinically graded, is an imperfect surrogate as a complete reflection of disease extent and severity. As we strive to improve the outcome of ROP, we open for debate how the decision to treat acute phase ROP and post-treatment reactivation may be fine-tuned, by creating an algorithm/formula that incorporates all acute-phase ROP components. The debate should include management of the less affected fellow eye and the danger of lowering the threshold for treatment inappropriately low. Whether type 1 as currently defined remains or is differently labeled will follow from this debate. A recent exchange in J AAPOS exemplifies the current situation with a plea for formal evaluation.^20,21

The purpose of ICROP3 is to improve diagnostic precision, which, we hope, will permit more detailed outcome analysis. Now, with certain evolutionary changes in practice and most of all the introduction of ICROP3, is a natural time to rethink whether any of the changes contained therein could influence the decision to treat. The ETROP treatment recommendations have had enormous impact on the care of premature infants worldwide. We hope that ICROP3 will provide the framework for clinicians and scientists to study how advances since that time may continue to improve the quality of care.

It is now over 30 years since treatment was shown to be effective for sight-threatening ROP. Now, with advances in retinal imaging and new treatment modalities, the purpose of treatment is not only to prevent blindness but also to promote the development of normal vision. To this end, we propose some discussion points for ophthalmologists to consider:

• What are the treatment implications of defining posterior zone II?

• What are the treatment implications of a notch incursion of disease into zone I?

• What are the treatment implications of neovascularization as seen in A-ROP and classic ROP, and how might this be described and recorded?

• Should circumferential extent of disease be included in the decision to treat?

• Are there treatment implications for pre-plus disease, particularly when accompanied by other findings and/ or risk factors?

• How should reactivation be managed?

• What should be the management of the peripheral avascular retina?

• Do the different stages of stage 5 ROP require individual surgical approaches?

• Do the current ROP treatment guidelines need to be revised?

Some of the issues raised here can only be resolved by future research requiring clinical trials and large datasets, which inevitably will take time. Other topics, including several listed above are pressing, and may need to rely on expert consensus opinion. This will form the basis on which to offer, as soon as possible, practical guidance to ophthalmologists. We hope this paper will stimulate discussion within the community regarding future directions in ROP management.

ICROP 3 Classification Committee

Audina Berrocal, MD, Bascom Palmer Eye Institute, Miami, FL; Gil Binenbaum, MD, MSCE, Retina Consultants Ltd, Des Plaines, IL; Michael Blair, MD, Retina Consultants Ltd, Des Plaines, IL, and Department of Ophthalmology, University of Chicago, Chicago, IL; J. Peter Campbell, MD, MPH, Casey Eye Institute, Oregon Health & Science University, Portland, OR; Antonio Capone Jr, MD, Associated Retinal Consultants, Oakland University, Rochester, MI; Yi Chen, MD, China-Japan Friendship Hospital, Beijing, China; Shuan Dai, MD, Queensland Children’s Hospital, Brisbane, Australia; Anna Ells, MD, Calgary Retina Consultants, Calgary, Alberta, Canada; Brian Fleck, MD, University of Edinburgh, Edinburgh, Scotland; William V. Good, MD, Smith-Kettlewell Eye Institute, San Francisco, CA; M. Elizabeth Hartnett, MD, Department of Ophthalmology and Visual Sciences, University of Utah, Salt Lake City, UT; Shunji Kusaka, MD, PhD, Kindai University, Osakasayama, Japan; Andrés Kychenthal, MD, KYDOFT Foundation, Santiago, Chile; Domenico Lepore, MD, A. Gemelli Foundation IRCSS, Department of Ageing and Neuroscience, Catholic University of the Sacred Heart, Rome, Italy; Birgit Lorenz, MD, PhD, Justus-Liebig-University Giessen, Germany, and Department of Ophthalmology, Universitaetsklinikum Bonn, Bonn, Germany; Maria Ana Martinez-Castellanos, MD, APEC, Mexico City, Mexico; Susan R. Ostmo, MS, Casey Eye Institute, Oregon Health & Science University, Portland, OR; Şengül Özdek, MD, Department of Ophthalmology, Gazi University, Ankara, Turkey; Dupe Ademola-Popoola, MD, Department of Ophthalmology, University of Ilorin, Ilorin, Nigeria; James D. Reynolds, MD, Ross Eye Institute, Department of Ophthalmology, University at Buffalo, Buffalo, NY; Parag K. Shah, MD, Aravind Eye Hospital, Coimbatore, Tamil Nadu, India; Michael Shapiro, MD, Retina Consultants Ltd, Des Plaines, IL; Andreas Stahl, MD, Department of Ophthalmology, University Medicine Greifswald, Germany; Cynthia Toth, MD, Department of Ophthalmology, Duke University, Durham, NC; Anand Vinekar, MD, PhD, Narayana Nethralaya Eye Institute, Bangalore, Karnataka, India; Linda Visser, MD, Department of Ophthalmology, University of KwaZulu-Natal, Durban, South Africa; David K. Wallace, MD, MPH, Department of Ophthalmology, Indiana University School of Medicine, Indianapolis, Indiana; Wei-Chi Wu, MD, PhD, Chang Gung Memorial Hospital, Linkou, Taoyuan, Taiwan and Chang Gung University, College of Medicine, Taoyuan, Taiwan; Peiquan Zhao, MD, Department of Ophthalmology, Xinhua Hospital affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai, China; Andrea Zin, MD, PhD, Fernandes Figueira Institute, FIOCRUZ and Brazilian Institute of Ophthalmology, Rio de Janeiro, Brazil.