Secondary Logo

Journal Logo

Review Article

Screening for Retinoblastoma: A Systematic Review of Current Strategies

Vempuluru, Vijitha S. MD; Kaliki, Swathi MD

Author Information
Asia-Pacific Journal of Ophthalmology: March-April 2021 - Volume 10 - Issue 2 - p 192-199
doi: 10.1097/APO.0000000000000378
  • Open

Abstract

Retinoblastoma (RB) is the most common intraocular cancer of childhood with an incidence ranging from 3.4 to 42.6 cases per million live births across the world.1 Survival rates in developed countries in the present era have crossed 95% but on the other end of the spectrum, children continue to succumb to the disease in low- and middle-income countries, that is attributed to late presentation.2–8 This gap can be bridged by the implementation of effective screening strategies.7–9 Tumors in younger children tend to involve the posterior pole and move towards the periphery with advancing age, although not without exceptions.10 It is also known that high-risk features such as optic nerve invasion, massive choroidal invasion, and anterior segment invasion are more common with increasing age which translates to increased risk of systemic metastasis and death.11,12 Therefore, if diagnosed and treated early, the disease can be managed with less radical forms of therapy, allowing salvage of a child's life and the affected eye with the benefit of vision salvage in select cases.9,12

Screening protocols for population at risk13,14 for retinoblastoma are well established, but without universal screening of neonates and infants, this grave disease may not be identified until it reaches an advanced stage. This review attempts to summarize the various screening programs, devices, and strategies employed across the world for early detection of RB and enable the formulation of an effective screening protocol which is universally acceptable.

METHODS

A PubMed® search was performed to identify articles published in MEDLINE® database with specific reference to screening of neonates and infants for retinoblastoma. Search with “retinoblastoma ‘AND’ early detection”, and “retinoblastoma ‘AND’ screening” produced 4,357 articles of which 59 articles were related to early diagnosis, detection and screening of RB. After checking for duplication and excluding articles on screening of ‘at-risk children’, screening of ‘family members’, pre-natal diagnosis of RB, case reports, and commentaries, 15 studies were included, which were on: 1) novel screening devices or tools utilized for early detection of RB; 2) hospital- or community-based prospective studies aimed at detection of RB in neonates and infants; 3) eye screening programs in children as a part of which RB was diagnosed early

Screening devices, applications, and software were evaluated based on the principle, sensitivity, specificity, limitations, and cost. Prospective and retrospective studies were summarized. Details of target population, screening devices, methodology, results, and conclusions were assessed.

RESULTS

Novel Devices with Scope of Utility in Screening for RB

Novel tools (alternative to direct or indirect ophthalmoscopy) utilized in screening for RB are based on either detection of leukocoria or imaging the retina. The devices described in literature include ArcLightTM, Portable Eye Examination Kit (PEEK), iCam (Optovue), and RetinaScope. Smartphone-based applications include CRADLE, MDEyeDetector, and soft fusion classifier leukocoria detector (Table 1).15–27 The ArcLightTM is a hand-held diagnostic tool which can be used on ophthalmoscope (Fig. 1), anterior segment loupe, or an otoscope with attachments. It is low-cost equipment with a solar panel and ergonomic design intended for use in low resource settings.15,16 PEEK is an adaptor clip which, when combined with a mobile phone, can be used to capture retinal images with minimal training. It has been used widely in Kenya to study disorders of the posterior segment.18,19 iCAM Optovue, Inc. (Fremont, CA) is a portable fundus camera which utilizes infrared light emitting diode (LED) to capture retinal images which can be viewed on a computer screen.20 Mndeme et al demonstrated the utility of ArcLightTM, PEEK and iCAM in detecting media opacities by trained nurses in ophthalmic as well as pediatric clinics with a sensitivity of 93%, 90%, and 98% respectively for ArclightTM, PEEK and iCAM.15 Trained ophthalmic nurses were capable of using all these devices effectively, hence they can be potential screening tools for screening in children.15 Patel et al described a widefield smartphone-based retinal imaging system (RetinaScope) for pediatric fundus photography. The device was used to image 6 eyes with retinoblastoma among 43 children with various ocular pathology.21

TABLE 1 - Novel Devices and Softwares Employed for Early Detection of Leukocoria and Retinoblastoma
Device/Software Principle User Input Output Validation in Literature (Leukocoria) Advantages Limitations
ArcLight Miniature direct ophthalmoscope Ophthalmologist/ trained pediatrician/ technician/ nurse Direct ophthalmoscopic examination Interpretation of red reflex by the examiner Sensitivity 93% 10 CheapPortableSolar and battery powered Requires hardware and trained personnel with knowledge of interpretation
PEEK Clip-on smartphone attachment for image acquisition Ophthalmologist/ trained / technician/ nurse Analogous to capturing fundus photograph Captured image, to be interpreted by the user Sensitivity 90%10 Portable Requires hardware and trained personnel. Captured images need interpretationMay not be suitable for young children
iCam (Optovue) Portable fundus camera Ophthalmologist/ trained technician/ nurse Analogous to capturing fundus photograph Captured image Sensitivity: 98%10 Portable ExpensiveRequires hardware, trained personnel for image capture and interpretation
CRADLE Application for leukocoria detection Anyone acquainted with smartphone usage Photograph captured through camera or saved on device Automated:“Normal” or “White” Munson et al18:Sensitivity 90% Specificity 20% Accuracy 55%Khadekar et al20:Percentage of detection (%): 0,0,0,0,100% in Group A to Group E RB respectivelyVagge et al19Sensitivity 15.4% Specificity 100%Negative likelihood ratio 1% Free software available for download on iOS and Google PlayUtility established in non- standard-setting, hence can be extrapolated to screening in communityInterpretation not user-dependent Requires access to and knowledge of usage of smart phoneLow detection rates for early disease (Group A and B)
MDEye Detector Application for leukocoria detection Anyone (requires knowledge of smartphone usage) Photograph captured through camera or saved on device Photograph needs to be interpreted by the user Percentage of detection (%)20:0,0,83,100,100 in Group A to Group E RB respectively Free software available for download on iOS and Google Play Interpretation of image is subjective
Soft fusion of classifiers for leukocoria detection Detection of leukocoria from recreational photographs by soft fusion of classifiers - - - Riva-Perea P et al22:Accuracy 92%True positive 89%False positive 11% Interpretation not user dependentSoft fusion is better than other methods of combining classifiers used in image processing Not commercially available
RetinaScope Clip-on smartphone attachment for fundus imaging Ophthalmologist/ trained technician/ nurse Analogous to capturing dilated fundus photograph Captured image Patel et al16:Detection rate of pathology (93–100% with interobserver variation) Portable Requires dilatation and subject cooperationInterpretation of image is subjective
RB indicates Retinoblastoma.

FIGURE 1
FIGURE 1:
Retinoblastoma screening with ArcLightTM. A, Screening device ArcLightTM. B, Usage of ArcLight to examine red reflex in a child.

ComputeR-Assisted Detector of LEukocoria (CRADLE), also popular as “White Eye Detector,” is a freeware application available for download on a smartphone. It is designed to detect leukocoria by analyzing the digital image (Fig. 2). The application is user-friendly and can be used by parents and caretakers to screen children.22,23 However, Vagge et al and Khadekar et al have found low sensitivity and detection rates for this application.24,25 Khadekar et al compared CRADLE with a similar smartphone application MDEyeCare26 and concluded that with modification in the photography parameters, the latter had superior detection rates for retinoblastoma than CRADLE.25 MDEyeCare, however, needs subjective interpretation of the image after capture of the pupillary reflex.25,26 Rivas-Parea designed a software on similar lines for better accuracy using technology of ‘soft fusion of classifiers’ for improved detection of leukocoria.27

FIGURE 2
FIGURE 2:
Retinoblastoma screening in children. A, Photograph showing usage of the CRADLE application on a smartphone by the parent of the toddler. B, Normal reflex in both eyes as seen on the CRADLE application. C, White reflex detected by the CRADLE application in a child with right eye retinoblastoma. CRADLE, ComputeR-Assisted Detector of LEukocoria.

Guidelines, Programmes, and Policies for Screening of RB

According to the World Health Organization (WHO) Guide for Effective Programmes in Cancer Control, programs are recommended for ‘early diagnosis’ (target population being children with white reflex and convergent strabismus, as these are the most common symptoms) but not for ‘screening’.28 However, ocular examination of neonates, infants and children is recommended as a part of various policies and governing bodies in several countries have formulated guidelines for the same. Ocular examination in neonatal period and infancy is crucial for detection of various ocular pathologies including RB. Guidelines for neonate and infant eye examinations could be retrieved only for select countries and notably most of them emphasize the importance of red reflex examination by trained personnel as a mandate. (Table 2).29–37 However, maintenance of a database for RB including details of cases referred from screening programs is not practised worldwide.31

TABLE 2 - Policies and Guidelines on Screening for Retinoblastoma in Different Countries
Geographic Location Organization Screening Guidelines for Newborns, Infants and Children Relevant to Detection of Retinoblastoma
- WHO Guide for Effective Programmes: Cancer Control Programmes recommended for early diagnosis (target population being children with white reflex an convergent strabismus) but not for screening.
Canada Canadian Task Force on Preventive Health Care, Community Pediatric Society, National Retinoblastoma Strategy Canadian guidelines for Cancer Care Red reflex examination from birth to three months of age; failure to visualize a normal red reflex warrants immediate referral to ophthalmologistOphthalmic examinations from 6–12 months, 3–5 years and 6–18 years
USA American Academy of Pediatrics: Policy Statement Mandatory red reflex examination of all infants within first 2 months of life by pediatrician or by a trained primary care ophthalmic clinician
Mexico Official Gazette of the Federation, General Health Law Neonatal eye examination 4 weeks after birth (no technical guidelines)
Latin and South America AHOPCARetMexGALOP Treatment guidelines defined, but no screening protocol/ guidelines
United Kingdom NIPE screening programmeUK National Retinoblastoma Service Red reflex examination within 72 hours of birth and at 6–8 weeks of ageDim or absent red reflex referred to ophthalmology service
Kenya Kenya National Screening Guidelines Ocular examination and genetic testing for children “at risk” for RB
India Rashtriya Bal Swasthya Karyakram Identification of at-risk newborns (family history of retinoblastoma)Ophthalmic examination including red reflex testing using an ophthalmoscopeReferral of infants at risk or with abnormal red reflex to ophthalmologists.Responsibility delegated to: Pediatricians/medical officers of special newborn care unit, staff nurse, optometrist of the district hospital and the ophthalmologist of district hospital/ private hospital
Australia National Children's Vision Screening Project Proposed screening programs for universal red reflex examination in newborn by trained personnel and staged screening approach all pre-school children (<4 years) starting with assessment of visual acuity
New Zealand Ministry of Health Red reflex examination within first week and at 6 weeks of birth by lead maternity care, general practitioner or pediatrician
AHOPCA indicates Asociación Hemato- Oncológica Pediátrica de Centro America; GALOP, Grupo de America Latina de Oncologia Pediatrica; NIPE, Newborn and infant physical examination; WHO, World Health Organization.

Studies Aimed at Detection of RB or Other Ocular Pathologies in Neonates, Infants, and Children

Vagge et al and Khadekar et al explored the utility of smartphone-based applications for detection of leukocoria to determine sensitivity and specificity. Both these studies showed low detection rates in early disease, both in small as well as a larger cohort of patients.24,25 Khadekar et al suggested a modification in MDEyeCare application to improve detection rates.25

Several studies across the world have shown the benefit of ocular examination in neonates by testing for red reflex and retinal imaging. Various ocular pathologies were detected in asymptomatic children including RB.5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,38,39,40,41,42,43 Prevalence of any form of ocular anomaly ranged from 5% to 24% in several studies.21,38–45 The details of the same are summarized in Table 3.

TABLE 3 - Comprehensive Review of Neonatal and Childhood Eye Screening Programs for Early Detection of Childhood Eye Disorders Including Retinoblastoma
Author, year Population Intervention (Methods) Comparison Outcome Study Type
Vagge et al19, 2019 122 children (244 eyes) Screened with CRADLE for detection of leukocoria Cycloplegic fundus examination Sensitivity-15.4%Specificity-100%Negative likelihood ratio-0.85% Qualitative
Khedekar et al20, 2019 34 eyes of 23 RB patients and 4 normal children Screened with MDEyeCare (modified) and CRADLE for detection of leucocoria Clinical (fundus) examination Percentage of detection (%)MDEyeCare: 0,0,83,100,100 in Group A to Group E RB respectivelyCRADLE: 0,0,0,0,100 in Group A to Group E RB respectively Qualitative
Hussain et al33, 2019 Staged screening of 33,549 children (<16 years) for ocular anomalies Phase 1: Screening by trained community health workersPhase 2: Comprehensive ophthalmologistPhase 3: Pediatric ophthalmologist Referred children evaluated by pediatric ophthalmologists 3 cases (0.008%) of retinoblastoma identified, prevalence of 0.09 per 1000 screened was noted Quantitative
Simkin et al34, 2019 All infants born between (350) June 2015 and December 2016 Dilated WFDRI using RetCam Shuttle (Clarity MSI USA) at a community birth centre, New Zealand Images reviewed by ophthalmologist through telemedicine Various ocular abnormalities (15.9%) detected including retinal hemorrhages, congenital cataract and optic nerve hypoplasia. No casess of retinoblastoma were diagnosed Quantitative
Mndeme et al10, 2018 a) 1152 children (<5 years)b) 41 cases and 60 controlsc) 2728 children Torchlight exam, red reflex assessment in Reproductive and Child Health ClinicsI-Cam, ArcLight, Portable Eye Examination Kit, torchlight use in Hospital settingReproductive and Child Health Clinics Indirect ophthalmoscopyIndirect ophthalmoscopyIndirect ophthalmoscopy for referred children Red reflex testing more sensitive than torchlight in detecting ocular media disorders, 94.7% versus 42% (p=0.0005)Sensitivity: I- Cam 97.56%, Arclight 92.68%, Portable Eye Examination Kit 90.2%, torchlight 7.3%Arclight was easy to use initial learning curve was easier Qualitative
Goyal et al35, 2018 1152 neonates from March 2014 to Oct 2015 Dilated fundus examination of apparently healthy neonates a civil hospital in Eastern India None Significant ocular findings in 172 babies (14.9%) including retinal hemorrhages, congenital glaucoma, cystic fovea, uveal coloboma, albinotic fundus, fundal jaundice Quantitative
Cagini et al36, 2017 22,272 children born in Umbria, Italy between January 2012 to December 2013 Red reflex was tested and labelled as positive or equivocal; later referred to ophthalmologist Ophthalmoscopy for referred children 3/ 461 (0.013% of total) neonates with equivocal reflex had ocular pathology, 1 (0.004%) of which was retinoblastoma Quantitative
Jayadev et al37, 2015 1450 preterm infants from January 2011 to December 2011 WFDRI captured over multiple sessions using RetCam Shuttle (Clarity MSI USA) All images analyzed by pediatric retinal specialists 2 eyes (0.14%) with retinoblastoma diagnosed early Quantitative
Vinekar et al38, 2015 1021 term infants WFDRI over multiple sessions using RetCam Shuttle (Clarity MSI USA) Children with abnormal images examined clinically 4.7% had ocular anomalies1 child with retinoblastoma detected Quantitative
Luo et al39, 2014 779 preterm and term infants WFDRI Images analyzed 1 retinoblastoma detected amongst 69 children with abnormal fundus findings Quantitative
Li LH et al 40, 2013 3573 healthy neonates within 42 days of birth Flash light, Retinoscope, WFDRI None 871 (24.1%) abnormal cases2 cases (0.06) of retinoblastoma Quantitative
CRADLE indicates ComputeR-Assisted Detector of LEukocoria; RB, retinoblastoma; WFDRI, Widefield digital fundus imaging.

DISCUSSION

American Academy of Ophthalmic Oncologists and Pathologists has formulated clear guidelines for detection of RB in children ‘at risk’ for retinoblastoma and these are practised widely.13 However, about 90% of patients diagnosed with RB, who do not have a positive family history, are sporadic, thus may not be diagnosed early, often leading to worse outcomes. Proven risk factors for poor survival rates with RB in developing countries include delay in diagnosis and treatment abandonment.7–9 Although the latter revolves around the socio-economic status, cultural, and literary background, the former can be tackled by increasing awareness and implementation of effective screening programs.7,9

Indirect ophthalmoscopic examination with scleral indentation under general anesthesia is the gold standard for establishing or ruling out the diagnosis RB in children.46 However, it cannot be employed as a screening tool for detection of RB due to various constraints, such as need for pupillary dilatation, general anesthesia or sedation, and technical expertise. Research for better screening tools has paved way to the development of numerous devices for retinal imaging and detection of media opacities.20 Some of these have shown potential use in screening for RB as well.18–20 Screening for retinoblastoma can be performed by 2 basic techniques: 1) red reflex assessment and 2) wide-field digital retinal imaging (WFDRI).16–27 Red reflex is a simple test, which is easy to perform with adequate training.47 However, it is not fool proof and has a sensitivity of about 85% and specificity of 39%, which further varies with pupillary dilatation.48 For RB, the location & size of intraocular tumors can affect the degree of distortion of red reflex. WFDRI overcomes these limitations by providing a wider field of view to detect peripheral lesions as well.20 However, lack of portability and high cost limit the use of WFDRI, giving rise to development of various compact devices such as PEEK and iCAM.18–20 One inherent disadvantage with any of these methods includes the subjectivity and need for interpretation by expert personnel. Teleconsultation by sharing the images captured is a feasible option to ensure accurate interpretation.19 Smartphone-based applications for detection of leukocoria came into existence when parents of a child with RB noticed a white reflex in his eyes in serial photographs and they offered these images to be analyzed which heralded the development of CRADLE and MDEyeCare.22,23,26 Being freewares, these applications can be used widely on a larger scale. CRADLE also displays the result as a “normal” or “white” eye, thereby eliminating the need for interpretation. CRADLE has been shown to have a lead time of 284 ± 547 days and 50 ± 103 days from the diagnosis of unilateral and bilateral retinoblastoma respectively.23 Further, artificial intelligence and deep learning have enormous potential in this field with scope for interpretation of pupillary reflex as well as retinal imaging for diagnosis.49

For the formulation of policies on establishment of screening programs, the WHO defines criteria for ‘early diagnosis’ and ‘screening’ of a disease and the choice between the two is determined by weighing the proven benefit of screening, on disease outcomes, and availability of resources. Establishing a ‘screening program’ is more complex than an ‘early detection program’ (which targets at-risk populations) and WHO at present does not recommend a screening programme for RB even in high-resource setting. However, such programmes can be taken up as research projects with clear objectives in terms of effectiveness. Accountability and documentation after implementation play a major role in both early detection and screening programmes to propose changes in program methodology or introduce new programs.28 Guidelines of several nations reflect significant heterogeneity in terms of timing of examination, number of visits and the personnel involved in screening (Table 2). Notably, maintaining a national registry for RB is not uniformly practised and the importance of the same needs to emphasize.31 Further, although there are specific guidelines for examination of children with RB and screening of at-risk children, technical guidelines on mandatory infant eye examination including red reflex testing are detailed only by few organizations.28–37 There is a need to compose uniform guidelines across the world and ensure compliance to the same. The methodology for screening can be customized according to the resources available. Red reflex examination is current standard of care for screening. Policies do not recommend WFDRI as yet for universal ocular examination in children, and researchers have demonstrated its utility in neonatal eye examinations.28–37

Early detection of RB has been reported from universal eye screening programs in the pediatric age group.15,39,41–45 With a reported prevalence of ocular pathology of nearly 25% in some studies, universal ocular examination for neonates certainly needs consideration.45 Although WFDRI is ideal, cost constraints limit its use. Goyal et al noted ocular pathology in 15% in a study of 1152 apparently healthy neonates, but analyzing the cost involved, and they concluded that the ‘inexpensive’ red reflex examination is a more viable alternative for universal screening and recommended that WFDRI should be limited to children at risk of development of ocular disease.40 Vinekar et al, on the other hand, extrapolated their results of WFDRI in 1021 term infants to national scenario (India), and estimated that about 226,950 infants requiring treatment may go undiagnosed if routine WFDRI is not performed.43 For translation of screening strategies to programs, it is important to take into account the ‘number needed to screen’.50 Defined as ‘the number of people that need to be screened to prevent one death or adverse event’, the number needed to screen can be calculated from clinical trials on screening or estimated from studies on prevalence of unrecognized disease and clinical trials on treatment.50 For RB, this would involve estimating the population with early disease (ie, earlier than the most common group at presentation for a given population), adverse events being as vision loss, eye removal, metastasis, or death from the disease. The cost of resources for its implementation needs to be weighed against a benefit of reduction in the adverse events, which may vary in developed countries and developing countries, to determine the feasibility.

In a setting of limited resources with a large disease burden, red reflex testing would be a feasible option with focus on reducing the mortality first and subsequently improvise. Till WFDRI is widely available for screening, smartphone-based cost-effective strategies and smartphone-based screening applications hold promise and are currently under-utilized. The responsibility lies on the shoulders of health care professions in pursuing research on novel screening techniques.

Summarizing all the above facts, though disease burden and challenges vary across the world (Table 4), universal guidelines for retinoblastoma screening are the need of the hour, allowing leeway for modifications subject to availability of resources in the form of screening devices, manpower, and logistics. Maintenance of an interconnected nationwide database is invaluable from an epidemiological perspective and formulation of health policies; hence strongly recommended. Novel devices and applications need large scale trials for validation and research for an ideal screening tool should be pursued upfront. Till wide-field retinal imaging can be made available for all neonates and infants, no child should be deprived of the benefit of a red reflex examination. Every child should undergo red reflex testing at birth and within first few years of life by trained social workers or pediatricians, if not an ophthalmologist.

TABLE 4 - Worldwide Disease Burden, Clinical Profile, Outcomes, Challenges, and Directions for Screening in Various Regions
Region (N)8 Disease Burden(n)8 IRSS Stage at Presentation8 Predominant Socioeconomic Status3 Survival RatesMean (Range) Current Challenges7,8 Considerations for Additional Screening Tools and Strategies
North America (2) 200 0: 59.5%;I: 39.0%II: 0.5%III: 0.5%IV:0.5%NA:0.5% High-income >99% survival rate2–4>90% retain vision in one eye2–4 Leading cause of mortality: second non-ocular cancers Timely diagnosis and of second cancers
Europe (40) 522 0: 53.6%I: 42.9%II: 1.1%III: 0.6%IV: 1.1%NA:0.6%
Oceania (3) 17 0: 29.4%I: 64.7%IV: 5.9%
Latin and Caribbean (23) 312 0: 23.4%I: 50.0%II: 7.7%III: 4.8%IV:8.3%NA:5.8% Upper-middle income 79% (54–93%)5 Delay in diagnosis leading to development of extraocular or metastatic disease Improve life salvage and attempt to preserve vision through:Collaborative screening programs (Malaysia, South Africa): universal screening with fundus imaging may be feasible (iCAM)
Asia (42) 2276 0: 37.5%I: 39.6%II: 4.0%III: 8.2%IV: 5.8%NA:4.9% Lower-middle income 77% (60–92%)5,6 Delay in diagnosis and refusal to treatment Improve life salvage through:community level screening (PEEK/ CRADLE/ ArcLight)Emphasis on socioeconomic factors
Africa (43) 1024 0: 15.1%I: 35.6%II: 9.5%III: 19.1%IV: 15.6%NA:5.2% Low-income 40% (23–70%)5
IRSS indicates International Retinoblastoma Staging System: N, number of countries; n, number of cases of retinoblastoma over an 18-month-period.

REFERENCES

1. Broaddus E, Topham A, Singh AD. Incidence of retinoblastoma in the USA: 1975–2004. Br J Ophthalmol 2009; 93:21–23.
2. Broaddus E, Topham A, Singh AD. Survival with retinoblastoma in the USA: 1975–2004. Br J Ophthalmol 2009; 93:24–27.
3. MacCarthy A, Birch JM, Draper GJ, et al. Retinoblastoma: treatment and survival in Great Britain 1963 to 2002. Br J Ophthalmol 2009; 93:38–39.
4. Abramson DH. Retinoblastoma in the 20th century: past success and future challenges the Weisenfeld lecture. Invest Ophthalmol Vis Sci 2005; 46:2683–2691.
5. Canturk S, Qaddoumi I, Khetan V, et al. Survival of retinoblastoma in less-developed countries impact of socioeconomic and health-related indicators. Br J Ophthalmol 2010; 94:1432–1436.
6. Kaliki S, Patel A, Iram S, et al. Retinoblastoma in India: clinical presentation and outcome in 1,457 patients (2,074 Eyes). Retina 2019; 39:379–439.
7. Naseripour M. Retinoblastoma survival disparity: the expanding horizon in developing countries. Saudi J Ophthalmol 2012; 26:157–161.
8. Global Retinoblastoma Study Group. Global retinoblastoma presentation and analysis by national income level. JAMA Oncol 2020; 6:685–695.
9. Epelman S. Preserving vision in retinoblastoma through early detection and intervention. Curr Oncol Rep 2012; 14:213–219.
10. Abramson DH, Gombos DS. The topography of bilateral retinoblastoma lesions. Retina 1996; 16:232–239.
11. Kaliki S, Gupta S, Ramappa G, et al. High-risk retinoblastoma based on age at primary enucleation: a study of 616 eyes. Eye (Lond) 2019; 34:1441–1448.
12. Kaliki S, Maniar A, Patel A, et al. Clinical presentation and outcome of retinoblastoma based on age at presentation: a review of 1450 children. Int Ophthalmol 2020; 40:99–107.
13. Skalet AH, Gombos DS, Gallie BL, et al. Screening children at risk for retinoblastoma: consensus report from the american association of ophthalmic oncologists and pathologists. Ophthalmology 2017; 125:453–458.
14. Kaliki S, Gupta Rathi S, Patel A. Routine fundus screening of families of children with retinoblastoma: a prospective study of 131 consecutive families. Retina 2019; 39:1326–1332.
15. Mndeme FG, Mgaya E, Allen L, et al. Red reflex examination in reproductive and child health clinics for early detection of paediatric cataract/ocular media disorders: evidence from Kilimanjaro, Tanzania. J AAPOS 2018; 22:E4.
16. Kousha O, Blaikie A. The Arclight and how to use it. Community Eye Health 2019; 32:50–51.
17. Blundell R, Roberts D, Fioratou E, et al. Comparative evaluation of a novel solar powered low-cost ophthalmoscope (arclight) by eye healthcare workers in malawi. BMJ Innov 2018; 4:98–102.
18. Livingstone I, Bastawrous A, Giardini ME, et al. Collaboration. PEEK: Portable Eye Examination Kit. The smartphone ophthalmoscope. Invest Ophthalmol Vis Sci 2014; 55:1612.
19. Navitsky C. The portable eye examination kit. Retina Today 2013; 6:24–27.
20. Panwar N, Huang P, Lee J, et al. Fundus photography in the 21st century-a review of recent technological advances and their implications for worldwide healthcare. Telemed J E Health 2016; 22:198–208.
21. Patel TP, Kim TN, Yu G, et al. Smartphone-based, rapid, wide-field fundus photography for diagnosis of pediatric retinal diseases. Transl Vis Sci Technol 2019; 8:29.
22. Abdolvahabi A, Taylor BW, Holden RL, et al. Colorimetric and longitudinal analysis of leukocoria in recreational photographs of children with retinoblastoma. PLOS One 2013; 8:E76677.
23. Munson MC, Plewman DL, Baumer KM, et al. Autonomous early detection of eye disease in childhood photographs. Science Advances 2019; 5:Eaax6363.
24. Vagge A, Wangtiraumnuay N, Pellegrini M, et al. Evaluation of a free public smartphone application to detect leukocoria in high-risk children aged 1 to 6 years. J AAPOS 2019; 56:229–232.
25. Khedekar A, Devarajan B, Ramasamy K, et al. Smartphone-based application improves the detection of retinoblastoma. Eye (Lond) 2019; 33:896–901.
26. MDEyeCare. [Internet]Available From: Www.Mdeyecare.Com
27. Rivas-Perea P, Baker E, Hamerly G, Shaw BF. Detection of leukocoria using a soft fusion of expert classifiers under non-clinical settings. BMC Ophthalmol 2014; 14:110.
28. Cancer Control: Knowledge Into Action-Who Guide For Effective Programmes. [Internet] Available From: Https://Www.Who.Int/Cancer/Publications/Cancer_Control_Diagnosis/En/Canada.
29. Amit M. Vision screening in infants, children and youth. Paediatr Child Health 2009; 14:246–248.
30. American Academy Of Pediatrics, Section on ophthalmology; american association for pediatric ophthalmology; strabismus; american academy of ophthalmology; american association of certified orthoptists. Red reflex examination in neonates, infants, and children [published correction appears in Pediatrics. 2009 Apr;123(4):1254]. Pediatrics 2008; 122:1401–1404.
31. Ramírez-Ortiz MA, Lansingh VC, Eckert KA, et al. Systematic review of the current status of programs and general knowledge of diagnosis and management of retinoblastoma. Bol Med Hosp Infant Mex 2017; 74:41–54.
32. Chantada G, Luna-Fineman S, Sitorus RS, et al. SIOP-PODC recommendations for graduated-intensity treatment of retinoblastoma in developing countries. Pediatr Blood Cancer 2013; 60:719–727.
33. Newborn And Infant Physical Examination (NIPE) Screening: Programme Overview. [Internet]Available From: Https://Www.Gov.Uk/Guidance/Newborn-And-Infant-Physical-Examination-Screening-Programme-Overview.
34. Ministry Of Health, Kenya. Kenya National Cancer Screening Guidelines Nairobi, November 2018. Available From: Http://Www.Health.Go.Ke.
35. Universal Eye Screening In Newborns: Rashtriya Bal Swasthya Karyakram(Rbsk). [Internet] Available From: Https://Nhm.Gov.In/Images/Pdf/Programmes/Rbsk/Resource_Documents/Revised_Rop_Guidelines-Web_Optimized.Pdf.
36. Centre For Community Child Health. National Children's Vision Screening Project Discussion Paper. Melbourne, Vic: Centre For Community Child Health; 2008.
37. Raoof N, Dai S. Red Reflex Screening In New Zealand: a large survey of practices and attitudes in the auckland region. N Z Med J 2016; 129:38–43.
38. Hussain AHME, Roy T, Ferdausi N, et al. Prevalence of childhood ocular morbidity in a peri-urban setting in Bangladesh: a community-based study. Public Health 2019; 170:103–112.
39. Simkin SK, Misra SL, Battin M, et al. Prospective observational study of universal newborn eye screening in a hospital and community setting in New Zealand. BMJ Paediatrics Open 2019; 3:E000376.
40. Goyal P, Padhi TR, Das T, et al. Outcome of universal newborn eye screening with wide-field digital retinal image acquisition system: A Pilot Study. Eye (Lond) 2018; 32:67–73.
41. Cagini C, Tosi G, Stracci F, et al. Red reflex examination in neonates: evaluation of 3 years of screening. Int Ophthalmol 2017; 37:1199–1204.
42. Jayadev C, Vinekar A, Bauer N, et al. Look what else we found-clinically significant abnormalities detected during routine ROP screening. Indian J Ophthalmol 2015; 63:373–377.
43. Vinekar A, Govindaraj I, Jayadev C, et al. Universal ocular screening of 1021 term infants using wide-field digital imaging in a single public hospital in India - a pilot study. Acta Ophthalmol 2015; 93:E372–E376.
44. Luo R, Liu J, Hu P, et al. Rresults of 779 cases of neonatal fundus screening and risk factors for neonatal fundus diseases. Zhongguo Dang Dai Er Ke Za Zhi 2014; 16:1197–1201. Chinese.
45. Li L-H, Li N, Zhao JY, et al. Findings of perinatal ocular examination performed on 3573, healthy full-term newborns. Br J Ophthalmol 2013; 97:588–591.
46. Springer-Verlag Berlin Heidelberg, Kim JW, Singh AD. Singh AD, Murphee AL, Damato BE. Differential Diagnosis Of Leukocoria. Clinical Ophthalmic Oncology: Retinoblastoma 2015. 16.
47. Bowman R, Foster A. Testing the red reflex. Community Eye Health 2018; 31:23.
48. Mussavi M, Asadollahi K, Janbaz F, et al. The evaluation of red reflex sensitivity and specificity test among neonates in different conditions. Iran J Pediatr 2014; 24:697–702.
49. Ting DSW, Pasquale LR, Peng L, et al. Artificial intelligence and deep learning in ophthalmology. Br J Ophthalmol 2019; 103:167–175.
50. Rembold CM. Number needed to screen: development of a statistic for disease screening. BMJ 1998; 317:307–312.
Keywords:

ArcLightTM; cancer; CRADLE; eye; leukocoria; RB; retinoblastoma; screening

Copyright © 2021 Asia-Pacific Academy of Ophthalmology. Published by Wolters Kluwer Health, Inc. on behalf of the Asia-Pacific Academy of Ophthalmology.