retina-216-0951 pap 1. - sp-rop...ophthalmoscopy or color photography.30,31 overall, wfa seems to...

Review

THE CUTTING EDGE OF RETINOPATHY OFPREMATURITY CARE

Expanding the Boundaries of Diagnosis andTreatment

YOSHIHIRO YONEKAWA, MD,*† BENJAMIN J. THOMAS, MD,‡ ARISTOMENIS THANOS, MD,§¶BOZHO TODORICH, MD, PHD,§¶ KIMBERLY A. DRENSER, MD, PHD,§¶ MICHAEL T. TRESE, MD,§¶ANTONIO CAPONE, JR., MD§¶

Purpose: To discuss the latest advances and controversies in the diagnosis and care ofinfants with retinopathy of prematurity (ROP).

Methods: Literature review.Results: Retinopathy of prematurity remains a major global issue. Industrialized nations

now treat profoundly premature infants with posterior and aggressive disease, and middle-income nations are experiencing ROP epidemics. Remote digital imaging may address thedecreasing ratio of ROP providers to premature infants, in addition to improving patientcare. Widefield angiography, optical coherence tomography, and the Wnt signaling path-way have provided new insights into ROP pathogenesis. Anti–vascular endothelial growthfactor treatment is increasing in popularity, but the dearth of information to guide dosing,unpredictable reactivation, persistent vascular abnormalities, the “crunch” phenomenon,and the presently unknown effects of systemic vascular endothelial growth factor sup-pression remain issues to continue investigating. Neurodevelopmental delay has beenraised as a potential consequence, but the evidence currently is weak. Vitrectomy is thetreatment of choice for Stages 4 and 5. Illumination techniques, ab interno incisions,plasmin-assisted vitrectomy, staged surgery in the interest of corneal clearing for advancedStage 5, and immediate sequential bilateral vitreoretinal surgery, are useful techniques.

Conclusion: We are making progress in ROP management. Our goal as clinicians is tocontinue expanding the boundaries of our abilities to keep this blinding disease in checkglobally.

RETINA 0:1–18, 2017

The history of retinopathy of prematurity (ROP)since its first descriptions in the 1940s1 is rela-

tively short. Seminal translational studies and clinicaltrials in subsequent decades have uncovered the path-ophysiology of ROP and rendered it an often prevent-able disease under ideal circumstances. Retinopathy ofprematurity is arguably the best-understood vasoproli-ferative retinopathy in terms of pathogenesis, timing,and severity of disease progression, and treatmentresponse. Yet, there is still much to do, as ROP con-tinues to be a leading cause of blindness among

prematurely born infants worldwide. This review willdiscuss the latest advances and controversies in theclinical care of infants with ROP.

Global Focus

Retinopathy of prematurity was born as a consequenceof medical progress. Advances in neonatology in the20th century allowed more premature infants to survive.Positive pressure mechanical ventilation was one of thesedevelopments, but oxygen delivery was unregulated. The

1

Copyright ª by Ophthalmic Communications Society, Inc. Unauthorized reproduction of this article is prohibited.

combination of prematurity and high oxygen exposureresulted in ROP as a new disease entity.1 Since then,outcomes have improved in industrialized countries withthe development of sophisticated oxygen delivery/monitoring technologies, diligent programs for screeningat-risk infants, and treatments based on large prospective,multi-center clinical trials.2,3 Although advanced ROP(Stages 4 and 5 ROP) is uncommon in such high-income countries, micropremature infants—the smallest,youngest, and sickest infants who tend to develop moreaggressive disease—remain a challenge even under idealcircumstances.4,5 Conversely, premature infants born inlow-income countries still have poor chances of survivalbecause of inadequate resources. The incidence of ROPis therefore relatively low in these nations.6

The regions currently experiencing epidemics ofROP are the middle-income countries.7 Neonatal med-icine has improved in middle-income countries, but inmany circumstances, neonatal intensive care units(NICUs) still lack sophisticated technology for themonitoring of oxygen delivery. Oxygen blenders andpulse oximeters, for example, which are in widespreaduse in the United States are not readily available inmany NICUs throughout the world. This unfortunatelyrecreates the situation that industrialized nations expe-rienced at the onset of the ROP epidemic when oxygenwas unregulated. As a consequence, infants are at riskof developing ROP despite greater gestational agesand birth weights.6–9 Retinopathy of prematurity pro-viders in middle-income countries are currently inun-dated with ROP patients, often with limited manpowerand access to digital imaging technologies for screen-ings and treatment. The current bedside screeningmodel may not be sustainable, and new paradigmsfor education, screening, and treatment may berequired to thwart this epidemic.

Remote Digital Fundus Imaging

Bedside examination with binocular indirect oph-thalmoscopy has been the gold standard for ROP

screening. However, there is an increasing discordancebetween the number of premature infants requiringscreening and the number of ophthalmologists per-forming it. Communities with limited access to ROPproviders may be burdensome to isolated providerscovering multiple hospitals over wide geographicareas. Lower volume and lower acuity NICUs mayalso not provide physicians tasked with diseasesurveillance with adequate experience in managingatypical or advanced stages. Photographic screeningusing digital fundus imaging addresses these logisticissues. Neonatal intensive care unit staff can be trainedto obtain fundus images that can be forwarded to theROP provider for interpretation. Infants with ROPfindings severe enough to require bedside examinationor treatment can be promptly and efficiently identified.Images of challenging cases could be easily sent forexpert consultants in a timely manner. Inherentlyrelated to this topic is the complex medicolegal climatesurrounding ROP. Photodocumentation is a method ofdemonstrating sound clinical practice and judgment,should legal action occur.In addition to its logistic and legal advantages, the

data quality of photographic screening is more objec-tive than binocular indirect ophthalmoscopy examina-tions. Data logged in the chart after binocular indirectophthalmoscopy examinations are subjective sche-matic renditions of the actual fundus findings. Sequen-tial digital images for side-by-side comparison providemore accurate documentation of clinical featuresfundamental to detection of disease progression.10–16

With more than 15 years of studies validating theaccuracy and sensitivity of “store-and-forward” tele-medicine in ROP screening, clinical trials and livetelemedicine programs have demonstrated that photo-graphic screening using contact widefield digital fun-dus images can effectively detect treatment and/orreferral-warranted ROP.10–14 These results are consis-tent among different camera operators, includingtrained ophthalmologists,14,17,18 trained neonatal per-sonnel,19,20 and ophthalmic photographers.20–22 Ana-lytical software can also enhance remote digital fundusimaging (RDFI) analysis. Weight gain–based riskprediction models23,24 and automated quantification ofvascular tortuosity25,26 are two such examples.In our current RDFI programs at our institutions,

trained NICU nurses obtain weekly digital fundusphotographs of all infants who meet the screeningcriteria. The weekly imaging paradigm eliminates therisk of skipped examinations resulting from misinter-preted disease severity, as it may occur when thetiming of subsequent examinations is predicated on theexaminer’s sense of the stage of ROP and plus diseasestatus. The images are securely uploaded to a server

From the *Massachusetts Eye and Ear Infirmary, Harvard Med-ical School, Boston, Massachusetts; †Boston Children’s Hospital,Harvard Medical School, Boston, Massachusetts; ‡Florida RetinaInstitute, Jacksonville, Florida; §Associated Retinal Consultants,Royal Oak, Michigan; and ¶Oakland University William BeaumontSchool of Medicine, Auburn Hills, Michigan.

Consultant for Novartis (A. Capone), consultant for Synergetics(A. Capone, K. A. Drenser, M. T. Trese), founder and equity ownerof FocusROP and Retinal Solutions (A. Capone, K. A. Drenser,M. T. Trese). The remaining authors have no financial/conflictinginterests to disclose.

Reprint requests: Antonio Capone, Jr., MD, Associated RetinalConsultants William Beaumont Hospital, 3555 W. Thirteen MileRoad, Suite LL-20, Royal Oak, Royal Oak, Michigan MI 48073;e-mail: [email protected]

2 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES � 2017 � VOLUME 0 � NUMBER 0


mailto:[email protected]

and ROP providers receive message alerts that thereare images to be interpreted. The physicians thensecurely interpret the images remotely, and the imagesand recommendations are provided to the NICU andincluded in the chart. Bedside examinations are per-formed for potentially treatment warranted infants or ifthere are any questions of image quality.Remote digital fundus imaging will hopefully

improve access to care for at-risk infants. The currentlive programs have been successful to date, not onlybecause of the benefits of RDFI screening, but alsobecause the inherent limitations are taken into consid-eration. For example, image artifacts, poor dilation,and media opacities may not permit accurate gradingof images. Bedside examinations should be performedin those instances, particularly if there are concerningrisk factors. Another limitation is the relative difficultyof imaging the far periphery. Most infants in the NICUwith Type 1 ROP have disease that is posterior enoughto adequately image with wide-angle cameras. How-ever, milder disease in anterior Zone II and Zone IIImay be more difficult to capture.27 Telescreening pro-grams should therefore always have a clear backupalgorithm for timely bedside examinations. Finally,another consideration is that ROP screening scheduleswere developed for bedside examinations. We circum-vent the follow-up schedule of 1 weeks versus 2 weeksby photographing weekly, but it is still unclear whento stop screening. Fortunately, infants are usually dis-charged from the NICU and are being screened inclinic by this point. If not, a bedside examination isgenerally recommended to make such decisions. Atthe current time, RDFI should be used to enhanceROP screening programs as an adjunct, without com-pletely replacing bedside indirect ophthalmoscopy.27

Innovative Imaging in Retinopathy of Prematurity

Imaging technologies have transformed the manage-ment of ROP. The RetCam system is currently the most

widely used for fundus imaging, with which widefield(up to 130°) digital color images and fluorescein angi-ography can be obtained bedside or under anesthesia.Noncontact Optos imaging (Optos, Marlborough, MA)by holding infants in clinic have been shown (“flying-baby”), but the safety of this technique must be properlyevaluated, because these infants are often the youngest,smallest, and sickest neonates.28,29 Widefield fluores-cein angiography (WFA) and spectral domain opticalcoherence tomography have been the two major advan-ces in ROP imaging for the past several years.

Widefield Fluorescein Angiography

Although the diagnosis of ROP is based on ophthal-moscopy or color photography findings, WFA is a usefulancillary test in several situations—particularly in eyeswith atypical findings (Figure 1). The characteristicStage 3 ridge tissue seen when ROP is located primarilyin Zone 2 is often absent in eyes with aggressive pos-terior ROP, with Stage 3 characterized by flat neovas-cularization that can be difficult to identify, especially ifthe view is limited from poor dilation, vitreous haze ofprematurity, or a dense tunica vasculosa lentis. Wide-field fluorescein angiography helps in identifying theflat neovascularization and more accurately delineatesthe borders between vascular and avascular retina. How-ever, for milder disease, WFA allows early identifica-tion of vascular changes that are not yet detectable byophthalmoscopy or color photography.30,31 Overall,WFA seems to improve the sensitivity of diagnosiscompared with photography alone.32–34 Studies havealso shown that WFA is useful in monitoring diseaseprogression and regression.35–37 Finally, when thetempo of disease is not fully consistent with ROP, orif the severity is disproportionately worse for the gesta-tional age/birth weight, other syndromic and familialvitreoretinopathies should be considered in the differen-tial diagnosis,38–40 and WFA is useful to distinguishthese clinical entities (discussed below in ROP or famil-ial exudative vitreoretinopathy [FEVR]? or Both?).

Fig. 1. Widefield fluoresceinangiography of Stage 3 ROP.Digital color fundus imagingshows a temporal notch withneovascular ridge tissue (A).Fluorescein angiography moreclearly delineates the vascular–avascular junction and extentof Stage 3 (B). Superiorly, thereare also neovascular tufts pos-terior to the leading edge.

ADVANCES IN ROP MANAGEMENT � YONEKAWA ET AL 3


Spectral Domain Optical Coherence Tomography

The micrometer-level axial resolution in vivo imag-ing of spectral domain optical coherence tomographyhas provided new insights into the microstructuralfeatures of ROP. The development of custom OCTplatforms41,42 and hand-held (Envisu; Leica, BuffaloGrove, IL)43–45 or mountable (iVue/iStand; Optovue,Fremont, CA)46,47 devices have allowed imaging ofinfants and young children.Spectral domain optical coherence tomography

evaluation of the posterior pole of premature infantshas revealed heretofore unrecognized acute features ofthe disease, including foveal hypoplasia with persis-tent inner retinal layers,42,46–48 cystoid macularchanges,43,47,49–51 retinoschisis,42,45,52 posterior hya-loidal organization, and vitreoretinal traction.45 Ofinterest, many of these features remain or becomemore pronounced in adolescent and adult patients withhistories of ROP (Figure 2).53 Regarding the fovealhypoplasia, it has been demonstrated that the structuralchanges seen on spectral domain optical coherencetomography do not consistently correlate with visualacuity,54 similar to findings in FEVR.55 However, thechoroid has been shown to be thinner in eyes withadvanced ROP compared with spontaneously re-gressed ROP, and choroidal thinning appears to beindependently associated with worse vision in thesepatients.56

Our understanding of ROP has been enhancedthrough the use of spectral domain optical coherencetomography technologies. We hope that intraoperativeOCT will also have meaningful impact on the surgicalapproach and clinical decision making of thesepatients in the near future.

Wnt Signaling Pathways, a New Angle

Premature birth disrupts the well-orchestratedsequence of normal retinal vascular development.Wnt signaling is an evolutionarily conserved signal

transduction pathway that modulates cellular and tis-sue differentiation.57 Regarding ocular development,the norrin-FZD4 segment has been identified as play-ing a pivotal role in retinal angiogenesis and vascularmaintenance.58–60 Mutations affecting genes of thispathway can result in several pediatric vitreoretinopa-thies, such as Norrie disease, FEVR, and pseudoglio-ma (term ascribed to describe the appearance of thedetached retina in affected patients) and osteoporosissyndrome. Of interest, studies have identified FZD4,LRP5, and TSPAN 12 mutations in patients withadvanced ROP.61–63 Two avenues for further investi-gation are introduced below.

Placental Homeostasis

Wnt signaling plays an important role in placentalhomeostasis.64,65 Markers for angiogenesis and vascu-lar formation are reduced in the corpora lutea ofFZD4-null mice, and these mice are infertile.66 Like-wise, norrin has been localized to the uterine bloodvessels and decidual cells of rats,67 and NDP knockoutmice have defects in vascular development and de-cidualization in pregnancy that leads to embryonicloss.68 In humans, the expression of NDP has beenestablished in the placenta, and FZD4 expression hasbeen localized to placental villous mesenchymal cells.Our laboratory recently showed that a double missensemutation in FZD4 (p.[P33S(;)P168S]) was associatedwith lower than normal birth weights for gestationalage in infants with ROP compared with other prema-ture infants. This finding suggests that wnt signalingdefects may contribute to ROP pathogenesis, indi-rectly through intrauterine growth retardation, anddirectly through retinal vascular developmentalalternations.64,69

Retinopathy of Prematurity or Familial ExudativeVitreoretinopathy? or Both?

One occasionally encounters infants who havea discrepancy between their birth history and fundus

Fig. 2. Spectral domain opticalcoherence tomography of adultROP. Spectral domain opticalcoherence tomography of thesuperior macula in an adultpatient with a history ofuntreated ROP demonstrates ananomalous dense posterior hya-loid with vitreoretinal tractionand underlying cystoid macularchanges.



appearance. These infants are premature but notsignificantly so, yet present with very severe retinop-athy. The possibility for a wnt signaling mutationshould be entertained in these atypical cases, as thesemutations may be disease modifying with potential forprognostic implications. Widefield fluorescein angiog-raphy can provide valuable information by identifyingangiographic characteristics of ROP versus FEVR-likechanges. For example, ROP will have ridge tissue,arteriovenous shunting, and tuft formations. On thecontrary, FEVR will have more arborization thatmay not lead to a ridge, venous–venous shunting,peripheral vessel pruning that may not leak as much asexpected for ROP, and relatively more exudation.Circumferential hyaloidal contraction is also muchmore characteristic of ROP. When patients withangiographic findings suggestive of FEVR are bornprematurely, or if premature infants show angio-graphic signs of FEVR, the distinction between thetwo can be difficult, and the angiograms may containfeatures of both. We have identified that these patientsmay harbor wnt signaling pathway mutations, and termthese infants fROP.38 They have also been termedROPER.39,40 In general, the tempo of disease pro-gression is slower, but unlike classic ROP, recurrencesof vasoactivity may occur in fROP/ROPER despitetimely and appropriate therapy. Severe detachmentscan also be seen, characterized by extensive avascu-larity and a rapidly contractile posterior hyaloid. Werecommend angiography and genetic testing when indoubt, because the correct diagnosis will alter treat-ment, follow-up, and family counseling. Further studyof this phenomenon is required.

Laser Photocoagulation

The International Classification of Retinopathy ofPrematurity standardized the nomenclature of ROPand laid the groundwork for the landmark treatmenttrials, Cryotherapy for Retinopathy of Prematurity(CRYO-ROP), and Early Treatment for Retinopathyof Prematurity (ETROP).2,3,70,71 CRYO-ROP estab-lished the safety and efficacy of peripheral retinal abla-tion in threshold ROP.72 There was a 50% reduction inunfavorable structural outcomes compared withuntreated eyes.72,73 In planning for ETROP, post hocanalysis of the CRYO-ROP natural history cohortstratified eyes into high-risk (Type 1) or low-risk(Type 2) prethreshold disease.74 Laser was preferredover cryotherapy due to lower risks of postoperativeinflammation, hyaloidal contraction, myopia, and easeof posterior treatment.75,76 ETROP randomized infantsto treatment at Type 1 ROP or Type 2 (threshold)

ROP, and demonstrated a further reduction of unfavor-able structural outcomes with earlier intervention.3

Laser Techniques

Laser peripheral retinal ablation for ROP presentsa variety of challenges, both in terms of treatmenttechnique as well as coordinating multidisciplinaryefforts involving anesthesiology, neonatology, and theNICU staff. A near-confluent pattern of the laser (halfa spot diameter apart) is superior to less confluenttreatment.77,78 The endpoint should be a grayish burn,not chalky white. The laser power should be titratedappropriately because the posterior pole will requirehigher power, whereas the thinner anterior retina re-quires less power. Anterior segment ischemia is a rare,but potentially blinding, complication that may occurif the laser is too confluent and hot.79 The laser spansthe avascular retina anterior to the ridge to the oraserrata. The most common locations for missed laserare anteriorly (technically challenging) or in the tem-poral notch (poor uptake). Neovascular fronds and theridge should not be directly lasered, as it may causebleeding. Flat neovascularization that characterizesaggressive posterior ROP may pose a challengebecause the fronds can overlie the underlying avascu-lar retina. Staged laser has been shown to be effectivein such cases: the flat neovascular fronds will regresswith the first laser treatment, and the underlying avas-cular retina will then be exposed, and can be treatedduring a subsequent session.80 Finally, we also cautionagainst severely anemic/thrombocytopenic infantswho are at higher risk for bleeding.Retinopathy of prematurity laser treatment results in

well-described and predictable regression patterns. Inthe rare instances that laser fails, it fails in a predictablemanner. Almost all eyes treated with laser will succeedor fail within the first 9 weeks posttreatment.81 InETROP, 9% of eyes progressed to retinal detach-ment,82 but laser spots were allowed to be up to onespot diameter apart in the study.83 As discussed above,near-confluent laser results in better outcomes,77,78 andthe failure rates at specialized centers are muchlower.84 When lasered eyes do progress to retinaldetachment, they do so in a fashion similar tountreated ROP. However, the peripheral retina treatedwith laser almost never detaches, which helps maintainsurgical planes during vitrectomy.

Role of Anti–Vascular Endothelial GrowthFactor Treatment

Vascular endothelial growth factor is one of the keycytokines that modulates the pathogenesis of ROP.



The VEGF-driven vasoproliferative phase of advancedROP develops in response to the avascular peripheralretina, and therefore it is logical that anti-VEGFtreatment would be considered for ROP. The benefitscompared with laser include the technical ease ofadministration, general accessibility and lower cost ofbevacizumab (Avastin; Genentech, South SanFrancisco, CA), ergonomic benefits for the treater,the ability to perform the procedure without sedation/intubation, less induced myopia, relative preservationof visual fields, and faster regression of neovasculari-zation. Off-label use of anti-VEGF treatment istherefore increasing in popularity, and many studieshave demonstrated the efficacy in providing rapidROP regression.85–90

However, most of the aforementioned benefits ofanti-VEGF treatment are logistic in nature, physician-centered issues, or ophthalmic issues that would beinsignificant unless the retina is attached. Potentialdrawbacks include, but are not limited to, the follow-ing: “crunching” of the fibrovascular proliferation withprogressive retinal detachment, unpredictable earlyand late reactivation, and systemic exposure of anti-VEGF medication. The incidence and predisposingfactors for these potential issues are currentlyunknown at the time of writing.

The BEAT-ROP Controversy

Many neonatologists and ophthalmologists cite theBEAT-ROP study as justification for anti-VEGFmonotherapy for ROP. The study randomized infantswith Stage 3 ROP to bevacizumab or laser therapy andconcluded that bevacizumab was superior to laser forZone I ROP, but not for Zone II. The trial withoutdoubt provided new information in a landscape thatwas only composed of case series, but the data shouldbe interpreted with caution because of several imper-fections in study design and execution:

1. The laser failure rate for Zone I in BEAT-ROP was42%. This is a high laser failure rate that is morethan double the rates reported previously.5,91–93

Had the laser treatment success rates been compa-rable with previous reports, there would not havebeen a statistically significant difference betweenbevacizumab and laser treatment in Zone I.94

2. The definition of laser failure is different than otherreports. Most would consider progression to retinaldetachment a laser failure. However, BEAT-ROP’sdefinition was recurrence of neovascularization.This does not take into account that infants withaggressive posterior ROP may require more thanone laser session.80 If progressive retinal detach-ment was used as the definition of laser failure,

there were 2 Zone I eyes treated with laser thatprogressed, compared with 0 for bevacizumab. Thiswould not be a significant difference. Also note thatfor Zone II eyes, there were 2 treated with bevaci-zumab that progressed to retinal detachment, com-pared with 0 for laser.

3. The primary endpoint of the study was altered dur-ing the course of the study, and it is unclear why.

4. A reading center was not established until 20months into the study.

5. The study recommends injecting 2.5 mm posteriorto the limbus, which poses a high risk for retinalperforation in the infant eye. We recommend inject-ing through the pars plana, 0.5 mm to 1.0 mm pos-terior from the limbus.

Beyond these limitations in design and execution,there are several noteworthy incorrect statements in theDiscussion section

1. that anti-VEGF injections are a one-time treatmentand that there will not be recurrences (see Early andLate Reactivation discussion below)

2. that anti-VEGF injections do not leave the eye intothe systemic circulation (see Systemic Exposurediscussion below)

3. that anti-VEGF treatment will allow full retinal vas-cularization (see Persistent Vascular Abnormalitiesdiscussion below).

Although the BEAT-ROP study was influential, wecaution practitioners to interpret the data critically.95

Incorporating Anti–Vascular Endothelial GrowthFactor Treatment Into Practice

In general, laser monotherapy is our preferredtreatment paradigm for Type 1 ROP, and vitrectomyfor Stages 4 and 5. Nevertheless, there are threeclinical scenarios for which we consider anti-VEGFtreatment for Type 1 ROP:

1. When the fovea has not fully vascularized and laserwould affect a large portion of the macula. Interest-ingly though, the macula does vascularize afterlaser treatment to a surprising degree, so the lasercan be very close to the fovea, or even partiallyinvolving it, and the macula can broaden anatomi-cally with further ocular development over time.96

However, we do not know yet if normal fovealdevelopment takes place.

2. When poor fundus visualization precludes lasertreatment. This can occur as a consequence ofa prominent tunica vasculosa lentis, cornel stromalhaze, vitreous haze of prematurity, or vitreous hem-orrhage. If the latter, the view should be good



enough to confirm that there is no retinal detach-ment or fibrotic components may crunch. Other-wise a vitrectomy would be indicated.

3. When the anesthesia risk is too high for lasertreatment.

At the time of writing, there are no hard indicationsor contraindications for anti-VEGF treatment, andmany groups use combinations of laser and anti-VEGF treatments successfully. We provide our per-sonal preferences based on the currently availableliterature, our patient population, and ROP manage-ment experiences in the paragraph above. Robust datato allow standardization of treatment and follow-up ofeyes managed with ant-VEGF agents are currentlylacking.97

RAINBOW Study

To that end, the RAINBOW study (RAnibizumabCompared With Laser Therapy for the Treatment ofINfants BOrn Prematurely With Retinopathy of Pre-maturity; NCT02375971) was designed to providea better assessment of anti-VEGF treatment for ROP.RAINBOW is an ongoing, Phase 3, randomized,multicenter, prospective clinical trial that investigatesthe efficacy and safety of ranibizumab (Lucentis;Genentech) in infants with Type 1 ROP comparedwith laser therapy. The primary outcome is visualfunction at the patients’ fifth birthdays. Secondary out-come measures were anatomical outcomes, ocular andsystemic adverse events, absence of active ROPbeyond 52 weeks after treatment and recurrence ofROP. In addition, systemic measures such as headcircumference, leg length, weight, hearing, and respi-ratory function will be assessed. It is hoped that thisstudy will provide better information on the dosingregimen (0.2 and 0.1 mg ranibizumab) as well. Ofnote, the RAINBOW trial is not investigating bevaci-zumab, but the Pediatric Eye Disease InvestigatorGroup is currently recruiting for a smaller prospectivestudy to examine lower dosing of bevacizumab(NCT02390531).

Anti–Vascular Endothelial Growth FactorInjection Technique

When we use anti-VEGF agents in our practices, weprefer to use 0.17 mg (a third of 0.5 mg) to 0.25 mg(half dose) of ranibizumab when possible. The con-centration of systemic exposure and the degree andlength of systemic VEGF suppression are favorablecompared with bevacizumab and aflibercept,98

although it is unknown if these pharmacokinetic differ-ences are clinically significant.99 The two problems are

that ranibizumab is not reimbursed for ROP, and mosthospitals do not have it on formulary. Samples orextensive research protocols are thus usually required.The majority of practices therefore use a quarter to halfdose of bevacizumab, and there is absolutely no faultin doing so until there are advancements in regulatoryprocesses.After topical and/or subconjunctival anesthesia, the

intravitreal injection is performed 0.5 mm to 1.0 mmposterior to the limbus after povidone iodine prep ofthe conjunctiva and a pediatric speculum has beenplaced. Because the injected volume is small, careshould be taken that the medication is actuallydelivered. The smallest possible syringes are thereforerecommended. The intraocular pressure can risedramatically. After injection, optic nerve perfusionshould therefore be confirmed.In our practices, most patients also receive sub-

sequent laser treatment if possible to prevent early andlate reactivation. The timing of this subsequent laservaries and is tailored to the individual infant’s eye,health, and logistic issues: earlier if there are signsof reactivation,100 and later if there are systemic issuesthat should be well controlled first to minimize theanesthesia risk,3 and as needed if the patient may bedischarged home to where ROP care may be limited.

Considerations in Anti–Vascular EndothelialGrowth Factor Treatment

Early and Late Reactivation

Pharmacologic suppression of VEGF is currentlyprovided as a bolus dose through intravitreal injec-tions. When the medication is cleared from the eye,there is a resurgence of VEGF unless the underlyingcause of the retinopathy has been addressed. Althoughthe VEGF surge is intense yet short-lived in ROPcompared with most adult retinopathies such asdiabetic retinopathy, some infants will require morethan one treatment session. Anti-VEGF treatment ofROP is far from a “one and done” paradigm. The ratesof recurrence or incomplete response requiring multi-ple injections or laser supplementation are not clearlyknown. Rates after bevacizumab have varied from0%,101 2%,102 7%,103 14%,104 to 46%,105 and afterranibizumab, as low as 0%102 and as high as83%.106 The variability is likely due to relatively smallsample sizes, different patient populations, variablefollow-up, and because the timing of anti-VEGF treat-ment is not standardized. Nevertheless, the recurrencerate is not negligible and necessitates a meticulousfollow-up schedule, as recurrent disease activationand progressive detachment can occur as late as 2.5



years after treatment.107 In general though, reactiva-tions tend to occur within 1 month to 3 months afterinitial treatment.100,103,108 Anti-VEGF treatmentshould be avoided if the infant’s family may not beable to keep strict follow-up appointments.

Persistent Vascular Abnormalities

After laser treatment, follow-up examinations occurevery 1 week to 2 weeks based on examinationfindings, with clear endpoints for cessation of acuteexaminations.109 However, there are no long-term pro-spectively acquired data to guide follow-up of infantstreated with anti-VEGF monotherapy. Many infantstreated with anti-VEGF agents have vascular abnor-malities and/or persistently avascular peripheral retinaeven after years of follow-up (Figure 3).36,110,111

These peripheral retinal abnormalities are theoreticallyrisk factors for early or late recurrences, or otherpathologies. Peripheral avascular retina is prone tolattice-like changes, retinal breaks, and retinal detach-ment in teenage years.112,113 Consequently, we prefertreating persistent avascular retina with laser to helpcontrol the acute disease,104,108 to avoid the uncer-tainty of reactivations, and to minimize late ROP-associated rhegmatogenous retinal detachment.114,115

In our practices, patients after anti-VEGF treatmentare clinically followed, but fluorescein is very helpfulif there is any suspicion for reactivation, and/or if thedecision is made for laser.

Crunch Phenomenon

The dynamic cytokine milieu of the developingretina is complex and not fully understood yet.Vascular endothelial growth factor and many othercytokines are involved in ROP pathogenesis.116 Anti-VEGF injections alter multiple cytokines, includingelevation of TGF-b, a potent profibrotic agent.117–119

In addition to this iatrogenic rise in TGF-b, prematureinfants experience an endogenous rise of TGF-b asthey approach term.120,121 This unopposed TGF-bcan cause rapid contraction of the fibrovascular mem-branes to cause progressive retinal detachment. Such“crunching” has been most commonly discussed inproliferative diabetic retinopathy117 but has been re-ported in ROP as well.100,122–126 This phenomenon islikely not common, and incidence data are not avail-able to date. When detachment occurs on the heels ofanti-VEGF pharmacotherapy, they are more likely tobe atypically configured.In an effort to better characterize eyes with ROP that

“crunch” after anti-VEGF treatment, we organized aninternational multicenter study to examine suchcases.126 We found that progression to retinal detach-ment was noted a mean of 70 days after the anti-VEGFinjection; 11% within 1 week, and 49% within 4weeks. The time to detachment negatively correlatedwith the postmenstrual age at injection, i.e., youngerinfants had a longer latency period before detachment,and older infants detached more quickly. Three“crunch” configurations were noted: conventional pro-gression, very posterior with prepapillary contraction,and relatively peripheral but with very tight circumfer-ential tractional vectors (Figure 4). We hypothesizethat anti-VEGF agents may induce fibrosis and con-traction of the immature prepapillary vascular precur-sor cells for the prepapillary configuration, and of theflat neovascularization for the circumferential config-uration. All eyes with conventional detachments wererepaired successfully, but the anatomical success ratefor the prepapillary and circumferential configurationswere approximately two-thirds each. Because thesedetachments are difficult to repair, we recommend try-ing to avoid anti-VEGF agents in eyes with evidenceof fibrotic change along the ridge, as well as in eyeswith existing tractional retinal detachment.

Fig. 3. Persistent vascular abnormalities after anti-VEGF for aggressive posterior retinopathy or prematurity (AP-ROP). One month after anti-VEGFinjection for AP-ROP, there are large areas of avascular peripheral retina and persistent neovascularization (A). Nasally, there is hyaloidal contractionwith an evolving traction retinal detachment (B). There is diffuse vascular leakage throughout the posterior pole, not just from the neovascular tissues(A–C). There are vessels that have grown anteriorly at the previous vascular–avascular junction, but only a meager amount, and these are incompetentvessels that leak fluorescein (C).



Systemic Exposure

Vascular endothelial growth factor overexpressionis implicated in many pathologic states of the retina,but we must remember that the growth factor playsa critical role in organogenesis and maintenance ofmicrovascular environments throughout the body.During development of the embryo/fetus, its impor-tance has been demonstrated in lung maturation,cardiac development, neuronal survival, ocular devel-opment, renal development, pancreatic development,and bone growth.127 Although anti-VEGF agents aredelivered locally in the vitreous, the agents do enterthe circulation and systemic VEGF is suppressed inROP infants.128 Studies have demonstrated that beva-cizumab is detectable in the circulation for 8 weeksafter intravitreal injection for ROP, with correspondingsystemic VEGF suppression.129–131 One case reportthat measured VEGF levels after ranibizumab showedVEGF suppression for 3 weeks.132 There are manyadult studies also showing that the systemic VEGFsuppression is much more pronounced and longer withbevacizumab and aflibercept (Eylea; Regeneron, Tar-rytown, NY), compared with ranibizumab98,133–136

and pegaptanib (Macugen; Bausch + Lomb, Bridge-water, NJ).137 We currently do not know whether thetemporary systemic VEGF suppression will have last-ing effects, but the potential should be carefully dis-cussed during the parental informed consent process.

Neurodevelopmental Controversy

It is well established that VEGF plays vital roles inneurogenesis.138 Previous small studies did not showparticularly concerning results.139,140 However, tworecent studies have drawn attention to potential neuro-developmental consequences of anti-VEGF therapy ininfants with ROP.141

Lien et al142 from Taiwan recently reported in a ret-rospective study that infants treated with a combinationof bevacizumab and laser (n = 16) had poorer neuro-development compared with infants treated with laser

alone (n = 33), or bevacizumab alone (n = 12). Thelaser + bevacizumab group, however, had the highestpercentage of eyes with zone I disease, and overallsmaller and younger infants. This means that thisgroup likely underwent more anesthesia sessions (neu-rotoxicity of general anesthesia is also a debatedissue143,144), were sicker at baseline, and had lessmature organ systems.A recent study by the Canadian Neonatal Network

published 2 months after the Taiwanese study re-viewed infants treated with bevacizumab (n = 27) orlaser (n = 125), who underwent neurodevelopmentaltesting at 18 months. The authors noted that infantstreated with bevacizumab were more likely to havepoorer motor scores and severe neurodevelopmentaldisabilities. The study has many limitations and needsto be interpreted with caution. Most importantly, thiswas a small retrospective chart review, which is proneto incomplete data, lack of randomization, selectionbias, and in this case recall bias also because somedata were obtained from parental interviews.Many issues placed the bevacizumab group at

a disadvantage. More infants had Zone I disease, andthis difference was not accounted for in the regressionanalysis. Important factors such as birth weight, race,and use of mechanical ventilation were also notfactored into the model. Risk factors such as malesex and longer hospitalization trended toward placingthe bevacizumab arm at a disadvantage, but they werenot considered because of lack of statistical signifi-cance. Multiple “trends” that each does not attaina desirable P value can compound and cause a cumu-lative real effect.Subject grouping was also another issue. For

example, three infants had bevacizumab + laser, andwere included in the bevacizumab group. These in-fants likely required combination treatment becauseof aggressive disease, which is more likely to be seenin smaller and sicker infants. In addition, the lasergroup contained 11 infants who were treated for Type2 ROP, which potentially means that those infants

Fig. 4. Crunch phenomenonafter anti-VEGF treatment. A.Circumferential type of anti–VEGF-associated progressiveretinal detachment, which wasnoted 1 week after bevacizumabfor an infant with aggressiveposterior ROP. B. Prepapillarytype of anti–VEGF-associatedprogressive posterior retinaldetachment that was noted sev-eral weeks after bevacizumab forZone I disease.



were systemically healthier also, further placing thebevacizumab group at a disadvantage.The neurodevelopmental assessment was also

imperfect. For example, scores were not calculated ifthe children were too developmentally delayed. Therewere nine such cases in the laser group, and only 1 inthe bevacizumab group. This removed nine verydevelopmentally delayed infants from the laser arm.Also, 20/70 vision or worse was one definition for“severe neurodevelopmental disability.” An otherwisehealthy child with 20/70 vision is different from a childwith severe cognitive impairments from cerebral palsy.Furthermore, visual acuities were partially collectedfrom parental interviews, and not actual examinations.The study was one step toward learning more about

potential systemic issues with using anti-VEGF agentsfor ROP. However, the observations noted abovesuggest that the study design was biased against thebevacizumab group. Better designed studies arerequired before making any conclusions regardingneurodevelopmental issues.

Advanced Surgical Techniques for Stages 4 and 5

The history of surgical management for ROP isa chronicle of expanding limits.145–147 Once consid-ered to be a largely inoperable disease, advanced ret-inal detachments from ROP have been renderedoperable—however challenging—through a combina-tion of improved pathophysiologic understanding,revised surgical expectations, and novel techniques(Figure 5).Learning to view retinal detachments in the setting

of ROP as a progressive tractional disease withpredictable vectors (Figure 6),148 all serving to distorta highly elastic retina, allows treating physicians toperform more targeted surgeries while leaving thecrystalline lens intact.149 Shorter surgeries reduceanesthetic risk for vulnerable patients, and less-extensive dissections reduce the risk of retinal breaks

and the subsequent devastating effects of proliferativevitreoretinopathy.The goal of intervention for ROP-related retinal

detachments varies with the severity of the detach-ment. In contrast to the repair of retinal detachmentsin adults, in which relieving all points of traction(and, if relevant, reattaching the macula) is the sinequa non of successful treatment. Conversely, repairof retinal detachments in ROP is a balancing act:removing critical points of traction when one cansafely do so without making a retinal break. The goalfor extramacular retinal detachment (Stage 4A ROP)is an undistorted/minimally distorted posterior pole,total retinal reattachment and preservation of the lens,and central fixation vision. This goal is achieved in90% of eyes with Stage 4A ROP.139 Surgery fortraction retinal detachments involving the macula(Stage 4B ROP) is performed to minimize retinaldistortion and prevent total detachment (Stage 5).Successful surgical therapy in more advanced detach-ments is often performed in a stepwise fashion, withintervals between surgeries to allow for the progres-sive reapproximation of the retina toward the under-lying pigment epithelium, which highlights furthersurgical planes to address during subsequent opera-tions—in more succinct terms, success by successiveapproximation.Various manifestations of ROP have been brought

into the surgical fold through novel techniques thathave expanded surgical indications. For example, theposterior detachments that we more commonly seenow in micropreemies or eyes treated with anti-VEGFagents have very tight dissection planes, and maybenefit from hybrid-gauge vitrectomy, with 27-gaugecutters introduced through 23- or 25-gauge cannu-las.150 Shorter cannulas and instruments have recentlybeen developed for vitreoretinal surgery in infants.Although surgical approaches and goals are differ-

ent between ROP and adult surgeries, advances inadult vitreoretinal surgery can be appropriately appliedto ROP surgery. Examples include small gauge

Fig. 5. Surgery for Stage 5ROP. Preoperatively, the infantpresented with a total retinaldetachment with a dense anteriorhyaloidal plaque, which was dis-sected after a lensectomy andcapsulectomy (A). Several monthslater, the posterior pole is attachedand the infant has brisk lightperception.



instrumentation151–153 and endoscope-assisted vitrec-tomy.154,155 Several more broad concepts to facilitaterepair of complex ROP-related retinal detachments arediscussed below:

Novel Illumination Techniques

Successful surgery in ROP is rooted in visualization,but the nuances of the pathologic anatomy cansometimes be obscured by the standard diffuse enface lighting of the operating microscope. To over-come this obstacle, simple positional maneuvers withthe standard lightpipe can be used. In brief, themaneuvers involve using oblique illumination emittedfrom a standard endoilluminator, positioned trans-corneally or intracamerally to highlight subtle tissueplanes and distinguish dysplastic vitreous from under-lying retinal tissue. The endoillumination probe ispositioned obliquely about the surgeon’s viewing axisthrough the operating microscope, directed at the sur-gical plane of interest. The probe is positioned eitherintracamerally through one of the three ports or trans-corneally (held by an assistant) when bimanual dissec-tion is required. The surgical field is viewed using anunlit operating microscope. Each of the illuminationtechniques—direct, retroillumination, and transcleral—helps to render safer opportunities for surgical ap-proaches by making dissection planes moreapparent.156

Ab Interno Incisions

With appropriate visualization, one can then beginto address the difficult task of transecting the vectorsof traction to relax the retina while preserving the lensand avoiding retinal breaks.149 True to the originaldescription of “retrolental fibroplasia,” the retina isoften pulled forward to just behind the lens by sheetsof fibrous proliferation extending from the ridgetoward the lens and ciliary body, narrowing the avail-able space for surgical approach. Radial or circumfer-ential retinal folds and/or tractional retinaldetachments that are in close approximation to theposterior lens capsule can extend for many clockhours. This can be difficult to manage without removalof the lens because the surgical entry space is toonarrow for a vitreous cutter. However, lens removalin the pediatric population is fraught with sequelaeincluding inhibition of visual development and devel-opment of aphakic glaucoma.These complications can be avoided by using the

previously described technique of ab interno incisionsusing a microvitreoretinal blade.157 With the ab inter-no technique, once the sclera is entered, the microvi-treoretinal blade is first directed carefully posterior andthen inserted into the space or tissue between the retinaand posterior lens capsule. The transvitreal prolifera-tive sheet extending anteriorly from the ridge is inter-rupted using the blade for sharp dissection. This

Fig. 6. Tractional vectors inadvanced ROP. The primarytractional vectors in ROP are (A)ridge to ridge (we call this thedrum head), ridge to optic disk,(B) ridge to eye wall, (C) ridgeto ciliary body, and (D) ridge tolens. These vectors need to betransected by vitrectomy toallow retinal reattachment.



relieves anterior retinal traction, and posterior relaxa-tion of the retina is immediately apparent with creationof adequate space for the ensuing vitrectomy and dis-section of proliferation along the retinal surface whichspares the crystalline lens.157 This ab interno incisioncan be extended for many clock hours by sweeping inthe surgical space parallel to the lens capsule using thesclerotomy as a pivot point or by sliding the blade likea saw to release any tractional vectors, as described(Figure 7). Care must be taken to avoid violating thelens equator or causing an unintentional retinal break.Favorable long-term outcomes of this technique havebeen reported.158

Plasmin-Assisted Vitrectomy

Safe, successful surgery in ROP requires removal oftractional vectors, mainly in the form of dysplasticvitreoretinal adhesions. In pediatric patients, thehyaloidal attachment to the retina is particularlystrong. This strong attachment can lead to moredifficult vitreous separation during vitrectomy andhigher risks associated with the procedure. Retinopa-thy of prematurity eyes have incomplete regression ofthe primary hyaloid (including the tunica vasculosalentis), incomplete development of anterior segmentstructures (such as trabecular meshwork, ciliary pro-cesses), and areas of atrophic retina, making this stepeven more difficult. In fact, the hyaloid is left in placein many scenarios without forcing vitreous separation.One of the next horizons for surgery in this populationis augmenting the surgeon’s ability to create separationof the posterior vitreous, most prominently throughpharmacologic means.

Both human-derived plasmin enzyme (both autolo-gous and heterologous) and ocriplasmin (Jetrea;Thrombogenics, Leuven, Belgium) have been usedas pharmacologic vitreolysis agents to facilitate theinduction of posterior vitreous detachment duringvitrectomy. Ocriplasmin is a serine protease andtruncated form of plasmin that is active againstsubstrates, such as fibronectin and laminin, andcleaves the vitreoretinal interface to treat vitreomaculartraction in adults. The use of plasmin enzyme to assistsurgery has been reported in the ROP population.159–161

The data regarding plasmin were retrospective innature and lacked control arms, but it did demonstratean acceptable safety profile and appeared to augmentvitreous separation. Further studies are warranted here.Ocriplasmin demonstrated slightly less encouragingresults in a prospective study of pediatric vitreoreti-nopathies.162 Although the data presented in that studydo not show a clear-cut benefit to the use of ocriplas-min as an adjunct in pediatric retinal surgery, the com-plexity of these surgeries and the small sample sizelimited the analysis and left open the door for futurestudies and pharmacologic agents. The investigatorsdid note that vitrectomy appeared to be easier in thetreatment arms after the unmasking. In general, theoverall ability to dissociate the hyaloid from the retinawith less mechanical force and the improved visuali-zation of surgical planes are advantageous. Ongoingstudies will hopefully be better able to delineate therole for these agents in complex ROP surgery.

Corneal Clearing

A final point of expanding surgical limits pertains toROP eyes with advanced pathology that demonstrateshallowing of the anterior chamber with subsequentlens–cornea touch and central corneal opacification.These cases often present late (and are predominatelyseen in areas with poor ROP screening and treatmentinfrastructures), and although often considered inop-erable due to the corneal status, are indeed amenable tosurgical intervention.146

The surgical approach involves entering through theiris root or limbus at the nasal horizontal meridian witha bent 23-gauge butterfly needle that allows tissueentry with simultaneous infusion (the needle isattached to an infusion line of lactated Ringer’s solu-tion). A temporal wound is made at the horizontalmeridian with an microvitreoretinal blade, then a stan-dard vitreous cutter is inserted into the crystalline lensand lens material is endoaspirated to deepen the ante-rior chamber. The iris can then be mechanically sweptfrom the posterior corneal surface with a cyclodialysis

Fig. 7. Ab interno transection of retrolental membrane. A micro-vitreoretinal blade is inserted through the pars plicata into the surgicalspace between the lens (L) and the retinal fold (RF). The membrane isfirst punctured and then swept laterally. This allows a safer entry site forthe vitreous cutter.



spatula in a side-to-side motion or viscodissected, afterwhich the cutter is again used to remove the fibroticpupillary margin (enlarging the pupil beyond the mar-gins of the central corneal opacity) and performingsurgical iridectomy if required, followed by removalof the remaining lens material and capsule. Afterward,the eye is placed on topical cycloplegics, steroids, top-ical ocular hypotensive, and hypertonic sodium chlo-ride drops. In most cases, the cornea will clearsufficiently to allow posterior segment surgery within2 weeks to 4 weeks, thereby allowing therapeutic ben-efit in eyes with an otherwise abysmal prognosis.

Immediate Sequential BilateralVitreoretinal Surgery

Infants with ROP often have multiple life-threatening comorbidities and are the retina surgeon’shighest anesthesia risk population.163,164 Repeated ses-sions may not be desirable or even feasible. However,ROP is a rapidly progressive bilateral vitreoretinop-athy, where both eyes commonly require surgicalintervention. This had traditionally required repeatedvisits to the operating room for each eye, which com-pounds the anesthesia risk each surgery. Furthermore,the Stage 4A eye could progress to a 4B while await-ing surgery, and the visual potential has now beendrastically reduced. Many of these infants have activecomorbidities that may prevent them from returning tosurgery for the second eye in a timely fashion.One way to circumvent these issues of anesthesia

risk (mortality) and progressive retinal detachment(blindness) is to operate on both eyes consecutivelyduring the same anesthesia session.165 We termed thistechnique immediate sequential bilateral pediatric vit-reoretinal surgery (ISBVS).165 Each eye is treated asa completely separate procedure with re-preping, re-draping, re-scrubbing, and new sets of instruments,medications, and intraocular fluids. Based on the re-ported incidence of postvitrectomy endophthalmitis of0.03% to 0.08%, the risk of bilateral endophthalmitiswould be 1 in 500,000 to 10,000,000 simultaneousvitrectomy procedures.166,167 In comparison, the riskof anesthesia-related death in children is as high as 1 in10,000, all-cause perioperative mortality is 1 in 100 to1,000, even higher in neonates, and even higher inpremature infants.168,169 In 2 small studies of infantsundergoing ROP surgery, serious anesthesia complica-tions occurred in 1 in 29, and 1 in 13.170,171 The risk ofbilateral endophthalmitis is miniscule compared withthese risks of morbidity and mortality, and thereforejustifies the employment of ISBVS in appropriatecases. We conducted a study of 344 surgeries from172 ISBVS procedures from 24 centers worldwide,

speaking to its feasibility and safety.165 Of note, wedo not advocate ISBVS for all infants. The majority ofinfants can be treated with staged bilateral surgery, butISBVS can be a powerful option for appropriatelyselected cases.

Future Horizons

Retinopathy of prematurity may seem to be some-thing we have a firm hold of: We know the riskfactors, and how to screen and treat effectively in themajority of at-risk and affected infants. Yet, ROPpersists as a major cause of childhood blindnessworldwide. Retinopathy of prematurity is deeplyinterwoven with socioeconomic dynamics and medicalprogress, and its most prominent clinical manifesta-tions have evolved over the past several decades sinceTerry’s first description in 1942.1 Current challengesinclude aggressive posterior disease seen most com-monly in profoundly premature infants, as well asclarifying and defining the role of anti-VEGF therapy.The epidemics of the middle-income nations need tobe addressed through training programs172–174 andparadigm-shifts toward remote digital image screen-ing. We hope that translational studies will result inthe eradication of ROP one day.175,176 Until then, thegoal should be to widen and strengthen our net toidentify all infants at risk for the development ofROP—as those who enter the treatment pathway havea theoretical success rate of 99% (90% anatomicalsuccess with laser, and 90% anatomical success withStage 4A ROP). Our goal as clinicians is to expand theboundaries of our abilities to keep this blinding diseasein check globally.

Key words: anti-VEGF, fluorescein angiography,laser, optical coherence tomography, pediatric retina,retinal detachment, retinopathy of prematurity, teleme-dicine, vitrectomy, wnt signaling.

References

1. Terry TL. Extreme prematurity and fibroblastic overgrowth ofpersistent vascular sheath behind each crystalline lens. I. pre-liminary report. Am J Ophthalmol 1942;25:203–204.

2. Cryotherapy for Retinopathy of Prematurity CooperativeGroup. Multicenter trial of cryotherapy for retinopathy ofprematurity. Snellen visual acuity and structural outcome at5 1/2 years after randomization. Arch Ophthalmol 1996;114:417–424.

3. Good WV; Early Treatment for Retinopathy of PrematurityCooperative Group. Final results of the Early Treatment forRetinopathy of Prematurity (ETROP) randomized trial. TransAm Ophthalmol Soc 2004;102:233–248.

4. Emsley HC, Wardle SP, Sims DG, et al. Increased survivaland deteriorating developmental outcome in 23 to 25 week



old gestation infants, 1990-4 compared with 1984-9. Arch DisChild Fetal Neonatal Ed 1998;78:F99–F104.

5. Drenser KA, Trese MT, Capone A. Aggressive posterior ret-inopathy of prematurity. Retina 2010;30:S37–S40.

6. Gilbert C. Retinopathy of prematurity: a global perspective ofthe epidemics, population of babies at risk and implicationsfor control. Early Hum Dev 2008;84:77–82.

7. Gilbert C, Rahi J, Eckstein M, et al. Retinopathy of prema-turity in middle-income countries. Lancet 1997;350:12–14.

8. Shah PK, Narendran V, Kalpana N, Gilbert C. Severe reti-nopathy of prematurity in big babies in India: history repeat-ing itself? Indian J Pediatr 2009;76:801–804.

9. Padhi TR, Rath S, Jalali S, et al. Larger and near-term babyretinopathy: a rare case series. Eye (Lond) 2015;29:286–289.

10. Photographic Screening for Retinopathy of Prematurity(Photo-ROP) Cooperative Group. The photographic screen-ing for retinopathy of prematurity study (photo-ROP).Primary outcomes. Retina 2008;28:S47–S54.

11. Fijalkowski N, Zheng LL, Henderson MT, et al. StanfordUniversity Network for Diagnosis of Retinopathy of Prema-turity (SUNDROP): five years of screening with telemedicine.Ophthalmic Surg Lasers Imaging Retina 2014;45:1–8.

12. Vinekar A, Jayadev C, Mangalesh S, et al. Role of tele-medicinein retinopathy of prematurity screening in rural outreach centersin India—a report of 20,214 imaging sessions in the KIDROPprogram. Semin Fetal Neonatal Med 2015;20:335–345.

13. Quinn GE, Ying GS, Daniel E, et al. Validity of a telemedi-cine system for the evaluation of acute-phase retinopathy ofprematurity. JAMA Ophthalmol 2014;132:1178–1184.

14. Ells AL, Holmes JM, Astle WF, et al. Telemedicine approachto screening for severe retinopathy of prematurity: a pilotstudy. Ophthalmology 2003;110:2113–2117.

15. Myung JS, Gelman R, Aaker GD, et al. Evaluation of vascu-lar disease progression in retinopathy of prematurity usingstatic and dynamic retinal images. Am J Ophthalmol 2012;153:544–551.

16. Campbell JP, Kalpathy-Cramer J, Erdogmus D, et al. Plusdisease in retinopathy of prematurity: a continuous spectrumof vascular abnormality as a basis of diagnostic variability.Ophthalmology 2016;123:2338–2344.

17. Wu C, Petersen RA, VanderVeen DK. RetCam imaging forretinopathy of prematurity screening. J AAPOS 2006;10:107–111.

18. Photographic Screening for Retinopathy of Prematurity(Photo-ROP) Cooperative Group, Balasubramanian M, Ca-pone A, Hartnett ME, et al. The Photographic Screening forRetinopathy of Prematurity Study (Photo-ROP): study designand baseline characteristics of enrolled patients. Retina 2006;26:S4–S10.

19. Lajoie A, Koreen S, Wang L, et al. Retinopathy of prematuritymanagement using single-image vs multiple-image telemedi-cine examinations. Am J Ophthalmol 2008;146:298–309.

20. Roth DB, Morales D, Feuer WJ, et al. Screening for retinop-athy of prematurity employing the retcam 120: sensitivity andspecificity. Arch Ophthalmol 2001;119:268–272.

21. Chiang MF, Keenan JD, Starren J, et al. Accuracy and reli-ability of remote retinopathy of prematurity diagnosis. ArchOphthalmol 2006;124:322–327.

22. Chiang MF, Starren J, Du YE, et al. Remote image basedretinopathy of prematurity diagnosis: a receiver operatingcharacteristic analysis of accuracy. Br J Ophthalmol 2006;90:1292–1296.

23. Wu C, Löfqvist C, Smith LEH, et al; WINROP Consortium.Importance of early postnatal weight gain for normal retinal

angiogenesis in very preterm infants: a multicenter study ana-lyzing weight velocity deviations for the prediction of reti-nopathy of prematurity. Arch Ophthalmol 2012;130:992–999.

24. Binenbaum G, Ying GS, Quinn GE, et al. The CHOP post-natal weight gain, birth weight, and gestational age retinopa-thy of prematurity risk model. Arch Ophthalmol 2012;130:1560–1565.

25. Wallace DK, Zhao Z, Freedman SF. A pilot study using“ROPtool” to quantify plus disease in retinopathy of prema-turity. J AAPOS 2007;11:381–387.

26. Abbey AM, Besirli CG, Musch DC, et al. Evaluation ofscreening for retinopathy of prematurity by ROPtool ora lay reader. Ophthalmology 2016;123:385–390.

27. Fierson WM, Capone A; American Academy of PediatricsSection on Ophthalmology, American Academy of Ophthal-mology, American Association of Certified Orthoptists. Tele-medicine for evaluation of retinopathy of prematurity.Pediatrics 2015;135:e238–e254.

28. Fung THM, Muqit MMK, Mordant DJ, et al. Noncontacthigh-resolution ultra-wide-field oral fluorescein angiographyin premature infants with retinopathy of prematurity. JAMAOphthalmol 2014;132:108–110.

29. Patel CK, Fung THM, Muqit MMK, et al. Non-contact ultra-widefield imaging of retinopathy of prematurity using theOptos dual wavelength scanning laser ophthalmoscope. Eye(Lond) 2013;27:589–596.

30. Zepeda-Romero LC, Oregon-Miranda AA, Lizarraga-Barrón DS,et al. Early retinopathy of prematurity findings identified withfluorescein angiography. Graefes Arch Clin Exp Ophthalmol2013;251:2093–2097.

31. Azad R, Chandra P, Khan MA, Darswal A. Role of intrave-nous fluorescein angiography in early detection and regres-sion of retinopathy of prematurity. J Pediatr OphthalmolStrabismus 2008;45:36–39.

32. Klufas MA, Patel SN, Ryan MC, et al. Influence of fluores-cein angiography on the diagnosis and management of reti-nopathy of prematurity. Ophthalmology 2015;122:1601–1608.

33. Guagliano R, Barillà D, Bertone C, et al. Fluoresceinangiography-based diagnosis for retinopathy of prematurity:expert-non expert comparison. Eur J Ophthalmol 2013;23:881–886.

34. Patel SN, Klufas MA, Ryan MC, et al. Color fundus photog-raphy versus fluorescein angiography in identification of themacular center and zone in retinopathy of prematurity. Am JOphthalmol 2015;159:950–957.

35. Ng EYJ, Lanigan B, O’Keefe M. Fundus fluorescein angiog-raphy in the screening for and management of retinopathy ofprematurity. J Pediatr Ophthalmol Strabismus 2006;43:85–90.

36. Henaine-Berra A, Garcia-Aguirre G, Quiroz-Mercado H,Martinez-Castellanos MA. Retinal fluorescein angiographicchanges following intravitreal anti-VEGF therapy. J AAPOS2014;18:120–123.

37. Park SW, Jung HH, Heo H. Fluorescein angiography ofaggressive posterior retinopathy of prematurity treated withintravitreal anti-VEGF in large preterm babies. Acta Ophthal-mol 2014;92:810–813.

38. Drenser KA, Dailey W, Vinekar A, et al. Clinical presentationand genetic correlation of patients with mutations affectingthe FZD4 gene. Arch Ophthalmol 2009;127:1649–1654.

39. Gologorsky D, Chang JS, Hess DJ, Berrocal AM. Familialexudative vitreoretinopathy in a premature child. OphthalmicSurg Lasers Imaging Retina 2013;44:603–605.



40. John VJ, McClintic JI, Hess DJ, Berrocal AM. Retinopathy ofprematurity versus familial exudative vitreoretinopathy:report on clinical and angiographic findings. Ophthalmic SurgLasers Imaging Retina 2016;47:14–19.

41. Vinekar A, Sivakumar M, Shetty R, et al. A novel techniqueusing spectral-domain optical coherence tomography (Spec-tralis, SD-OCT+HRA) to image supine non-anaesthetized in-fants: utility demonstrated in aggressive posterior retinopathyof prematurity. Eye (Lond) 2010;24:379–382.

42. Joshi MM, Trese MT, Capone A. Optical coherence tomog-raphy findings in stage 4A retinopathy of prematurity: a theoryfor visual variability. Ophthalmology 2006;113:657–660.

43. Dubis AM, Subramaniam CD, Godara P, et al. Subclinicalmacular findings in infants screened for retinopathy of pre-maturity with spectral-domain optical coherence tomography.Ophthalmology 2013;120:1665–1671.

44. Vajzovic L, Hendrickson AE, O’Connell RV, et al. Matura-tion of the human fovea: correlation of spectral-domain opti-cal coherence tomography findings with histology. Am JOphthalmol 2012;154:779–789.

45. Chavala SH, Farsiu S, Maldonado R, et al. Insights intoadvanced retinopathy of prematurity using handheld spectraldomain optical coherence tomography imaging. Ophthalmol-ogy 2009;116:2448–2456.

46. Gursoy H, Bilgec MD, Erol N, et al. The macular findings onspectral-domain optical coherence tomography in prematureinfants with or without retinopathy of prematurity. Int Oph-thalmol 2016;36:591–600.

47. Erol MK, Ozdemir O, Turgut Coban D, et al. Macular findingsobtained by spectral domain optical coherence tomography inretinopathy of prematurity. J Ophthalmol 2014;6:468653–468657.

48. Recchia FM, Recchia CC. Foveal dysplasia evident by opticalcoherence tomography in patients with a history of retinopa-thy of prematurity. Retina 2007;27:1221–1226.

49. Maldonado RS, O’Connell R, Ascher SB, et al. Spectral-domain optical coherence tomographic assessment of severityof cystoid macular edema in retinopathy of prematurity. ArchOphthalmol 2012;130:569–578.

50. Lee AC, Maldonado RS, Sarin N, et al. Macular features fromspectral-domain optical coherence tomography as an adjunctto indirect ophthalmoscopy in retinopathy of prematurity.Retina 2011;31:1470–1482.

51. Vinekar A, Avadhani K, Sivakumar M, et al. Understandingclinically undetected macular changes in early retinopathy ofprematurity on spectral domain optical coherence tomogra-phy. Invest Ophthalmol Vis Sci 2011;52:5183–5188.

52. Muni RH, Kohly RP, Charonis AC, Lee TC. Retinoschisisdetected with handheld spectral-domain optical coherencetomography in neonates with advanced retinopathy of prema-turity. Arch Ophthalmol 2010;128:57–62.

53. Thanos A, Yonekawa Y, Todorich B, et al. Spectral-domainoptical coherence tomography in older patients with history ofretinopathy of prematurity. Ophthalmic Surg Lasers ImagingRetina 2016;47:1086–1094.

54. Villegas VM, Capó H, Cavuoto K, et al. Foveal structure-function correlation in children with history of retinopathyof prematurity. Am J Ophthalmol 2014;158:508–512.

55. Yonekawa Y, Thomas BJ, Drenser KA, et al. Familial exu-dative vitreoretinopathy: spectral-domain optical coherencetomography of the vitreoretinal interface, retina, and choroid.Ophthalmology 2015;122:2270–2277.

56. Wu WC, Shih CP, Wang NK, et al. Choroidal thickness inpatients with a history of retinopathy of prematurity. JAMAOphthalmol 2013;131:1451–1458.

57. Nusse R. Wnt signaling in disease and in development. CellRes 2005;15:28–32.

58. Xu Q, Wang Y, Dabdoub A, et al. Vascular development inthe retina and inner ear: control by Norrin and Frizzled-4,a high-affinity ligand-receptor pair. Cell 2004;116:883–895.

59. Parmalee NL, Kitajewski J. Wnt signaling in angiogenesis.Curr Drug Targets 2008;9:558–564.

60. Dejana E. The role of wnt signaling in physiological andpathological angiogenesis. Circ Res 2010;107:943–952.

61. Kondo H, Kusaka S, Yoshinaga A, et al. Genetic variants ofFZD4 and LRP5 genes in patients with advanced retinopathyof prematurity. Mol Vis 2013;19:476–485.

62. Wu WC, Drenser K, Trese M, et al. Retinal phenotype-genotype correlation of pediatric patients expressing muta-tions in the Norrie disease gene. Arch Ophthalmol 2007;125:225–230.

63. Hiraoka M, Takahashi H, Orimo H, et al. Genetic screeningof Wnt signaling factors in advanced retinopathy of prematu-rity. Mol Vis 2010;16:2572–2577.

64. Knöfler M, Pollheimer J. Human placental trophoblast inva-sion and differentiation: a particular focus on Wnt signaling.Front Genet 2013;4:190.

65. Sonderegger S, Pollheimer J, Knöfler M. Wnt signalling inimplantation, decidualisation and placental differentiation–review. Placenta 2010;31:839–847.

66. Hsieh M, Boerboom D, Shimada M, et al. Mice null for Friz-zled4 (Fzd4-/-) are infertile and exhibit impaired corpora luteaformation and function. Biol Reprod 2005;73:1135–1146.

67. Kaloglu C, Cesur I, Bulut HE. Norrin immunolocalizationand its possible functions in rat endometrium during the estruscycle and early pregnancy. Dev Growth Differ 2011;53:887–896.

68. Luhmann UFO, Meunier D, Shi W, et al. Fetal loss in homo-zygous mutant Norrie disease mice: a new role of Norrin inreproduction. Genesis 2005;42:253–262.

69. Dailey WA, Gryc W, Garg PG, Drenser KA. Frizzled-4 var-iations associated with retinopathy and intrauterine growthretardation: a potential marker for prematurity and retinopa-thy. Ophthalmology 2015;122:1917–1923.

70. An international classification of retinopathy of prematurity.The Committee for the Classification of Retinopathy of Pre-maturity. Arch Ophthalmol 1984;102:1130–1134.

71. International Committee for the Classification of Retinopa-thy of Prematurity. The international classification ofretinopathy of prematurity revisited. Arch Ophthalmol2005;123:991–999.

72. Multicenter trial of cryotherapy for retinopathy of prematu-rity. Preliminary results. Cryotherapy for Retinopathy of Pre-maturity Cooperative Group. Arch Ophthalmol 1988;106:471–479.

73. Tasman W. Ten-year follow-up from the CRYO-ROP study.Arch Ophthalmol 2001;119:1200–1201.

74. Hardy RJ, Palmer EA, Dobson V, et al. Risk analysis ofprethreshold retinopathy of prematurity. Arch Ophthalmol2003;121:1697–1701.

75. Nagata M. Therapeutic possibility in retrolental fibroplasia inthe premature infant with light coagulation. Ganka 1968;10:719–727.

76. Ng EYJ, Connolly BP, McNamara JA, et al. A comparison oflaser photocoagulation with cryotherapy for threshold reti-nopathy of prematurity at 10 years: part 1. Visual functionand structural outcome. Ophthalmology 2002;109:928–934.

77. Banach MJ, Ferrone PJ, Trese MT. A comparison of denseversus less dense diode laser photocoagulation patterns for



threshold retinopathy of prematurity. Ophthalmology 2000;107:324–327.

78. Rezai KA, Eliott D, Ferrone PJ, Kim RW. Near confluentlaser photocoagulation for the treatment of threshold retinop-athy of prematurity. Arch Ophthalmol 2005;123:621–626.

79. Salgado CM, Celik Y, VanderVeen DK. Anterior segmentcomplications after diode laser photocoagulation for prethres-hold retinopathy of prematurity. Am J Ophthalmol 2010;150:6–9.

80. Vinekar A, Jayadev C, Mangalesh S, et al. Comparing theoutcome of single versus multiple session laser photoablationof flat neovascularization in zone 1 aggressive posterior reti-nopathy of prematurity: a prospective randomized study. Ret-ina 2015;35:2130–2136.

81. Coats DK, Miller AM, Brady McCreery KM, et al. Involutionof threshold retinopathy of prematurity after diode laser pho-tocoagulation. Ophthalmology 2004;111:1894–1898.

82. Repka MX, Tung B, Good WV, et al. Outcome of eyesdeveloping retinal detachment during the early treatment forretinopathy of prematurity study (ETROP). Arch Ophthalmol2006;124:24–30.

83. Hardy RJ, Good WV, Dobson V, et al. Multicenter trial ofearly treatment for retinopathy of prematurity: study design.Control Clin Trials 2004;25:311–325.

84. DeJonge MH, Ferrone PJ, Trese MT. Diode laser ablation forthreshold retinopathy of prematurity: short-term structuraloutcome. Arch Ophthalmol 2000;118:365–367.

85. Quiroz-Mercado H, Ustariz-González O, Martinez-CastellanosMA, et al. Our experience after 1765 intravitreal injections ofbevacizumab: the importance of being part of a developingstory. Semin Ophthalmol 2007;22:109–125.

86. Mintz-Hittner HA, Kuffel RR. Intravitreal injection ofbevacizumab (avastin) for treatment of stage 3 retinopathyof prematurity in zone I or posterior zone II. Retina 2008;28:831–838.

87. Kusaka S, Shima C, Wada K, et al. Efficacy of intravitrealinjection of bevacizumab for severe retinopathy of prematu-rity: a pilot study. Br J Ophthalmol 2008;92:1450–1455.

88. Lalwani GA, Berrocal AM, Murray TG, et al. Off-label use ofintravitreal bevacizumab (Avastin) for salvage treatment inprogressive threshold retinopathy of prematurity. Retina2008;28:S13–S18.

89. Quiroz-Mercado H, Martinez-Castellanos MA, Hernández-Rojas ML, et al. Antiangiogenic therapy with intravitrealbevacizumab for retinopathy of prematurity. Retina 2008;28:S19–S25.

90. Dorta P, Kychenthal A. Treatment of type 1 retinopathy ofprematurity with intravitreal bevacizumab (Avastin). Retina2010;30:S24–S31.

91. Capone A, Diaz-Rohena R, Sternberg P, et al. Diode-laserphotocoagulation for zone 1 threshold retinopathy of prema-turity. Am J Ophthalmol 1993;116:444–450.

92. Foroozan R, Connolly BP, Tasman WS. Outcomes after lasertherapy for threshold retinopathy of prematurity. Ophthalmol-ogy 2001;108:1644–1646.

93. Jalali S, Kesarwani S, Hussain A. Outcomes of a protocol-based management for zone 1 retinopathy of prematurity: theIndian Twin Cities ROP Screening Program report number 2.Am J Ophthalmol 2011;151:719–724.

94. Robinson J, Capone AJ. Review of the latest treatmentsfor retinopathy of prematurity: laser photo-ablation versusintravitreal bevacizumab in the management of high-riskpre-threshold (type 1) disease. Expert Rev Ophthalmol2015;10:421–430.

95. Moshfeghi DM, Berrocal AM. Retinopathy of prematurity inthe time of bevacizumab: incorporating the BEAT-ROP resultsinto clinical practice. Ophthalmology 2011;118:1227–1228.

96. Pandya HK, Faia LJ, Robinson J, Drenser KA. Maculardevelopment in aggressive posterior retinopathy of prematu-rity. Biomed Res Int 2015;2015:808639.

97. Sankar MJ, Sankar J, Mehta M, et al. Anti-vascular endothe-lial growth factor (VEGF) drugs for treatment of retinopathyof prematurity. Cochrane Database Syst Rev 2016;2:CD009734.

98. Avery RL, Castellarin AA, Steinle NC, et al. Systemic phar-macokinetics following intravitreal injections of ranibizumab,bevacizumab or aflibercept in patients with neovascularAMD. Br J Ophthalmol 2014;98:1636–1641.

99. Moja L, Lucenteforte E, Kwag KH, et al. Systemic safety ofbevacizumab versus ranibizumab for neovascular age-relatedmacular degeneration. Cochrane Database Syst Rev 2014;9:CD011230.

100. Hu J, Blair MP, Shapiro MJ, et al. Reactivation of retinopathyof prematurity after bevacizumab injection. Arch Ophthalmol2012;130:1000–1006.

101. Cernichiaro-Espinosa LA, Olguin-Manriquez FJ, Henaine-Berra A,et al. New insights in diagnosis and treatment for Retinopathyof Prematurity. Int Ophthalmol 2016;36:751–760.

102. Chen SN, Lian I, Hwang YC, et al. Intravitreal anti-vascularendothelial growth factor treatment for retinopathy of prema-turity: comparison between Ranibizumab and Bevacizumab.Retina 2015;35:667–674.

103. Mintz-Hittner HA, Geloneck MM, Chuang AZ. Clinical man-agement of recurrent retinopathy of prematurity after intra-vitreal bevacizumab monotherapy. Ophthalmology 2016;123:1845–1855.

104. Hwang CK, Hubbard GB, Hutchinson AK, Lambert SR. Out-comes after intravitreal bevacizumab versus laser photocoag-ulation for retinopathy of prematurity: a 5-year retrospectiveanalysis. Ophthalmology 2015;122:1008–1015.

105. Chan JJT, Lam CPS, Kwok MKM, et al. Risk of recurrence ofretinopathy of prematurity after initial intravitreal ranibizu-mab therapy. Sci Rep 2016;6:27082.

106. Wong RK, Hubschman S, Tsui I. Reactivation of retinopathyof prematurity after ranibizumab treatment. Retina 2015;35:675–680.

107. Snyder LL, Garcia-Gonzalez JM, Shapiro MJ, Blair MP. Verylate reactivation of retinopathy of prematurity after monother-apy with intravitreal bevacizumab. Ophthalmic Surg LasersImaging Retina 2016;47:280–283.

108. Wu WC, Kuo HK, Yeh PT, et al. An updated study of the useof bevacizumab in the treatment of patients with prethresholdretinopathy of prematurity in taiwan. Am J Ophthalmol 2013;155:150–158.

109. Fierson WM; American Academy of Pediatrics Section onOphthalmology, American Academy of Ophthalmology,American Association for Pediatric Ophthalmology and Stra-bismus, American Association of Certified Orthoptists.Screening examination of premature infants for retinopathyof prematurity. Pediatrics 2013;131:189–195.

110. Tahija SG, Hersetyati R, Lam GC, et al. Fluorescein angio-graphic observations of peripheral retinal vessel growth ininfants after intravitreal injection of bevacizumab as sole ther-apy for zone I and posterior zone II retinopathy of prematu-rity. Br J Ophthalmol 2014;98:507–512.

111. Lepore D, Molle F, Pagliara MM, et al. Atlas of fluoresceinangiographic findings in eyes undergoing laser for retinopathyof prematurity. Ophthalmology 2011;118:168–175.



112. Kaiser RS, Trese MT, Williams GA, Cox MS. Adult retinop-athy of prematurity: outcomes of rhegmatogenous retinaldetachments and retinal tears. Ophthalmology 2001;108:1647–1653.

113. Robinson J. Adult retinopathy of prematurity: a 35-yearreview. American Society of Retina Specialists Annual Meet-ing; August 26, 2013; Toronto, CA.

114. Ho LY, Ho V, Aggarwal H, et al. Management of avascularperipheral retina in older prematurely born infants. Retina2011;31:1248–1254.

115. Gupta MP, Chan RVP, Anzures R, et al. Practice patterns inretinopathy of prematurity treatment for disease milder than rec-ommended by Guidelines. Am J Ophthalmol 2016;163:1–10.

116. Sato T, Kusaka S, Shimojo H, Fujikado T. Simultaneousanalyses of vitreous levels of 27 cytokines in eyes withretinopathy of prematurity. Ophthalmology 2009;116:2165–2169.

117. Arevalo JF, Maia M, Flynn HW, et al. Tractional retinaldetachment following intravitreal bevacizumab (Avastin) inpatients with severe proliferative diabetic retinopathy. Br JOphthalmol 2008;92:213–216.

118. Forooghian F, Kertes PJ, Eng KT, et al. Alterations in theintraocular cytokine milieu after intravitreal bevacizumab.Invest Ophthalmol Vis Sci 2010;51:2388–2392.

119. Connor TB, Roberts AB, Sporn MB, et al. Correlation offibrosis and transforming growth factor-beta type 2 levels inthe eye. J Clin Invest 1989;83:1661–1666.

120. Drenser KA. Anti-angiogenic therapy in the management ofretinopathy of prematurity. Dev Ophthalmol 2009;44:89–97.

121. Sood BG, Madan A, Saha S, et al. Perinatal systemic inflam-matory response syndrome and retinopathy of prematurity.Pediatr Res 2010;67:394–400.

122. Honda S, Hirabayashi H, Tsukahara Y, Negi A. Acute con-traction of the proliferative membrane after an intravitrealinjection of bevacizumab for advanced retinopathy of prema-turity. Graefes Arch Clin Exp Ophthalmol 2008;246:1061–1063.

123. Jang SY, Choi KS, Lee SJ. Delayed-onset retinal detachmentafter an intravitreal injection of ranibizumab for zone 1 plusretinopathy of prematurity. J AAPOS 2010;14:457–459.

124. Suk KK, Berrocal AM, Murray TG, et al. Retinal detachmentdespite aggressive management of aggressive posterior reti-nopathy of prematurity. J Pediatr Ophthalmol Strabismus2010;47:e1–e4.

125. Patel RD, Blair MP, Shapiro MJ, Lichtenstein SJ. Significanttreatment failure with intravitreous bevacizumab for retinop-athy of prematurity. Arch Ophthalmol 2012;130:801–802.

126. Yonekawa Y, Wu WC, Nitulescu CE, et al. Progressive ret-inal detachment in infants with retinopathy of prematuritytreated with intravitreal bevacizumab or ranibizumab. Retina2017; May 3 [Epub ahead of print].

127. Tolentino M. Systemic and ocular safety of intravitreal anti-VEGF therapies for ocular neovascular disease. Surv Oph-thalmol 2011;56:95–113.

128. Sato T, Wada K, Arahori H, et al. Serum concentrations ofbevacizumab (avastin) and vascular endothelial growth factorin infants with retinopathy of prematurity. Am J Ophthalmol2012;153:327–333.

129. Hong YR, Kim YH, Kim SY, et al. Plasma concentrations ofvascular endothelial growth factor in retinopathy of prematu-rity after intravitreal bevacizumab injection. Retina 2015;35:1772–1777.

130. Kong L, Bhatt AR, Demny AB, et al. Pharmacokinetics ofbevacizumab and its effects on serum VEGF and IGF-1 in

infants with retinopathy of prematurity. Invest OphthalmolVis Sci 2015;56:956–961.

131. Wu WC, Lien R, Liao PJ, et al. Serum levels of vascularendothelial growth factor and related factors after intravitre-ous bevacizumab injection for retinopathy of prematurity.JAMA Ophthalmol 2015;133:391–397.

132. Hoerster R, Muether P, Dahlke C, et al. Serum concentrationsof vascular endothelial growth factor in an infant treated withranibizumab for retinopathy of prematurity. Acta Ophthalmol2013;91:e74–e75.

133. Wang X, Sawada T, Sawada O, et al. Serum and plasmavascular endothelial growth factor concentrations before andafter intravitreal injection of aflibercept or ranibizumab forage-related macular degeneration. Am J Ophthalmol 2014;158:738–744.

134. Yoshida I, Shiba T, Taniguchi H, et al. Evaluation of plasmavascular endothelial growth factor levels after intravitrealinjection of ranibizumab and aflibercept for exudative age-related macular degeneration. Graefes Arch Clin Exp Oph-thalmol 2014;252:1483–1489.

135. Zehetner C, Kralinger MT, Modi YS, et al. Systemic levels ofvascular endothelial growth factor before and after intravitrealinjection of aflibercept or ranibizumab in patients with age-related macular degeneration: a randomised, prospective trial.Acta Ophthalmol 2015;93:e154–e159.

136. Carneiro AM, Costa R, Falcão MS, et al. Vascular endothelialgrowth factor plasma levels before and after treatment of neo-vascular age-related macular degeneration with bevacizumabor ranibizumab. Acta Ophthalmol 2012;90:e25–e30.

137. Zehetner C, Kirchmair R, Huber S, et al. Plasma levels ofvascular endothelial growth factor before and after intravitrealinjection of bevacizumab, ranibizumab and pegaptanib in pa-tients with age-related macular degeneration, and in patientswith diabetic macular oedema. Br J Ophthalmol 2013;97:454–459.

138. Mackenzie F, Ruhrberg C. Diverse roles for VEGF-A in thenervous system. Development 2012;139:1371–1380.

139. Martinez-Castellanos MA, Schwartz S, Hernández-Rojas ML,et al. Long-term effect of antiangiogenic therapy for retinop-athy of prematurity up to 5 years of follow-up. Retina 2013;33:329–338.

140. Araz-Ersan B, Kir N, Tuncer S, et al. Preliminary anatomicaland neurodevelopmental outcomes of intravitreal bevacizu-mab as adjunctive treatment for retinopathy of prematurity.Curr Eye Res 2015;40:585–591.

141. Morin J, Luu TM, Superstein R, et al. Neurodevelopmentaloutcomes following bevacizumab injections for retinopathyof prematurity. Pediatrics 2016;137:e20153218.

142. Lien R, Yu MH, Hsu KH, et al. Neurodevelopmental out-comes in infants with retinopathy of prematurity and bevaci-zumab treatment. Hartnett ME, ed. PLoS One 2016;11:e0148019.

143. Sinner B, Becke K, Engelhard K. General anaesthetics andthe developing brain: an overview. Anaesthesia 2014;69:1009–1022.

144. Byrne MW, Casale P, Garzon M, et al. Pediatric surgeons andanesthesiologists expand the dialogue on the neurotoxicityquestion, rationale for early and delayed surgeries, and prac-tice changes while awaiting definitive evidence. J NeurosurgAnesthesiol 2014;26:391–395.

145. Hirose T, Katsumi O, Mehta MC, Schepens CL. Vision instage 5 retinopathy of prematurity after retinal reattachmentby open-sky vitrectomy. Arch Ophthalmol 1993;111:345–349.



146. Trese MT, Droste PJ. Long-term postoperative results ofa consecutive series of stages 4 and 5 retinopathy of prema-turity. Ophthalmology 1998;105:992–997.

147. Greven C, Tasman W. Scleral buckling in stages 4B and 5retinopathy of prematurity. Ophthalmology 1990;97:817–820.

148. Recchia FM, Capone A. Contemporary understanding andmanagement of retinopathy of prematurity. Retina 2004;24:283–292.

149. Capone A, Trese MT. Lens-sparing vitreous surgery for trac-tional stage 4A retinopathy of prematurity retinal detach-ments. Ophthalmology 2001;108:2068–2070.

150. Yonekawa Y, Thanos A, Abbey AM, et al. Hybrid 25- and27-gauge vitrectomy for complex vitreoretinal surgery. Oph-thalmic Surg Lasers Imaging Retina 2016;47:352–355.

151. Gonzales CR, Boshra J, Schwartz SD. 25-Gauge pars plicatavitrectomy for stage 4 and 5 retinopathy of prematurity. Ret-ina 2006;26:S42–S46.

152. Kychenthal A, Dorta P. 25-gauge lens-sparing vitrectomy forstage 4A retinopathy of prematurity. Retina 2008;28:S65–S68.

153. Wu WC, Lai CC, Lin RI, et al. Modified 23-gauge vitrectomysystem for stage 4 retinopathy of prematurity. Arch Ophthal-mol 2011;129:1326–1331.

154. Wong SC, Lee TC, Heier JS, Ho AC. Endoscopic vitrectomy.Curr Opin Ophthalmol 2014;25:195–206.

155. Marra KV, Yonekawa Y, Papakostas TD, Arroyo JG. Indica-tions and techniques of endoscope assisted vitrectomy. J Oph-thalmic Vis Res 2013;8:282–290.

156. Wong SC, Capone A. Illumination techniques for complexpediatric anterior retinal detachment and associated retrolentalplaque. Retina 2015;35:1905–1907.

157. Ho LY, Ranchod TM, Drenser KA, et al. Ab interno incisionfor pediatric vitreoretinal surgery. Retina 2010;30:1542–1543.

158. Thomas BJ, Yonekawa Y, Trese MT. Complete resolution oflarge retinal fold after transection of retrolental membraneduring lens-sparing vitrectomy for retinopathy of prematurity:a 15-year follow-up. Retin Cases Brief Rep 2016;10:93–95.

159. Wu WC, Drenser KA, Lai M, et al. Plasmin enzyme-assistedvitrectomy for primary and reoperated eyes with stage 5 ret-inopathy of prematurity. Retina 2008;28:S75–S80.

160. Wong SC, Capone A. Microplasmin (ocriplasmin) in pediatricvitreoretinal surgery: update and review. Retina 2013;33:339–348.

161. Joshi MM, Ciaccia S, Trese MT, Capone A. Posterior hyaloidcontracture in pediatric vitreoretinopathies. Retina 2006;26:S38–S41.

162. Drenser K, Girach A, Capone A. A randomized, placebo-controlled study of intravitreal ocriplasmin in pediatric pa-tients scheduled for vitrectomy. Retina 2016;36:565–575.

163. Steward DJ. Preterm infants are more prone to complicationsfollowing minor surgery than are term infants. Anesthesiol-ogy 1982;56:304–306.

164. Meza PA. Anesthesia for infants and children. In: Hartnett ME,Trese MT, Capone A Jr, Keats BJB, Caputo G, eds. PediatricRetina. 2nd ed. Philadelphia, PA: Lippincott Williams & Wil-kins; 2014:577–585.

165. Yonekawa Y, Wu WC, Kusaka S, et al. Immediate sequentialbilateral pediatric vitreoretinal surgery: an international mul-ticenter study. Ophthalmology 2016;123:1802–1808.

166. Oshima Y, Kadonosono K, Yamaji H, et al. Multicenter sur-vey with a systematic overview of acute-onset endophthalmi-tis after transconjunctival microincision vitrectomy surgery.Am J Ophthalmol 2010;150:716–725.

167. Govetto A, Virgili G, Menchini F, et al. A systematic reviewof endophthalmitis after microincisional versus 20-gauge vit-rectomy. Ophthalmology 2013;120:2286–2291.

168. Flick RP, Sprung J, Harrison TE, et al. Perioperative cardiacarrests in children between 1988 and 2005 at a tertiary referralcenter: a study of 92,881 patients. Anesthesiology 2007;106:226–237.

169. van der Griend BF, Lister NA, McKenzie IM, et al. Post-operative mortality in children after 101,885 anesthetics ata tertiary pediatric hospital. Anesth Analg 2011;112:1440–1447.

170. Aoyama K, Kondou Y, Suzuki Y, et al. Anesthesia protocolsfor early vitrectomy in former preterm infants diagnosed withaggressive posterior retinopathy of prematurity. J Anesth2010;24:633–638.

171. Sinha R, Talawar P, Ramachandran R, et al. Perioperativemanagement and post-operative course in preterm infantsundergoing vitreo-retinal surgery for retinopathy of prematu-rity: a retrospective study. J Anaesthesiol Clin Pharmacol2014;30:258–262.

172. Chan RVP, Patel SN, Ryan MC, et al. The global educationnetwork for retinopathy of prematurity (Gen-Rop): develop-ment, implementation, and evaluation of a novel tele-education system (an American Ophthalmological SocietyThesis). Trans Am Ophthalmol Soc 2015;113:T21–T226.

173. Paul Chan RV, Williams SL, Yonekawa Y, et al. Accuracy ofretinopathy of prematurity diagnosis by retinal fellows. Retina2010;30:958–965.

174. Wong RK, Ventura CV, Espiritu MJ, et al. Training fellowsfor retinopathy of prematurity care: a web-based survey. JAAPOS 2012;16:177–181.

175. Hoppe G, Yoon S, Gopalan B, et al. Comparative systemspharmacology of HIF stabilization in the prevention of reti-nopathy of prematurity. Proc Natl Acad Sci U S A 2016;113:E2516–E2525.

176. Ley D, Hansen-Pupp I, Niklasson A, et al. Longitudinal infu-sion of a complex of insulin-like growth factor-I and IGF-binding protein-3 in five preterm infants: pharmacokineticsand short-term safety. Pediatr Res 2013;73:68–74.



retina-216-0951 pap 1. - sp-rop...ophthalmoscopy or color photography.30,31 overall, wfa seems to...

Documents