Figure 1. Study selection process flow diagram .

3.1. Public Assistance

Hill et al. in 201628addresses the development of a multidisciplinary and multisectoral strategy embracing health professionals, students, survivors of RB, their families, and members of the government of Kenya. It was structured from annual meetings and resulted in the development of National Retinoblastoma Strategy Guidelines, that counted with facilitated access to government-subsidized health insurance, the creation of a nursing home, centralization of the pathological diagnosis of the disease, and the acquisition of artificial eyes for the children.

3.2. Surveillance strategies

New software technologies like CRADLE® and MDEyeCare® have been developed to detect leukocoria, the main and most early signal of RB20. Using this principle, Bernard et al.3 with their EyeScreen® mobile application, Vargas-Cuentas21 with RetinoApp® and Khedekar et al20 with MDEyeCare® demonstrated that mobile apps are promising screening strategies. EyeScreen® application achieves a sensitivity of 87% and specificity of 73% 8. RetinoApp®, associated with an algorithm that detected leukocoria, correctly identified 93.33% of cases and 95% of controls21. MDEyeCare® had good results for advanced intraocular disease, but low sensitivity for initial stages.
Ademola-Popoola, 201717evaluated the use of a cell phone’s basic camera’s photo and video functions along with a 20D lens after eye dilation. Although images captured through this approach were not as clear as those obtained through fundoscopy, new camera technologies can overcome this limitation.
Another strategy to enhance screening effectiveness relied on a clinical assessment of the risk of harboring germline RB1 mutations. Kaliki et al. (2019)19successfully executed this strategy by conducting ophthalmological examinations on parents of children diagnosed with RB. If at least one parent had a spontaneously regressed RB, there is a presumed risk of a sibling having a germline RB1 mutation. Similarly, Yousef et al.22devised a calculated risk model based on factors such as family history, the presence of affected family members, and the classification of unilateral or bilateral disease. The study’s findings revealed that the calculated risk model resulted in enhanced detection of the disease at earlier stages, thereby providing increased opportunities for ocular salvage.

3.3. International partnership

In the study conducted by Qaddoumi et al.29, it was observed that the twinning strategy facilitated the provision of high-quality care and early identification of issues that could potentially harm the patients. Furthermore, the training of skilled professionals in developing countries was found to reduce the need for sending professionals abroad for extended periods to enhance their expertise, thereby reducing associated costs. Additionally, it is crucial to support the development of a well-established infrastructure to ensure effective multidisciplinary interventions.
In another study conducted by Wilimas, et al.16, twinning strategies were implemented in Central America countries in collaboration with St. Jude Children’s Research Hospital and the ocular oncology team at the Hamilton Eye Institute (UTHEI). The primary focus of the program was on early diagnosis and specialist referrals, and as the program expanded, there was an increase in ocular salvage16. However, the study also highlighted the need for additional resources and inputs to support the program.
Jordan was another country that benefited from the twinning program with the St. Jude Children’s Research Hospital. The study conducted by Yousef et al. demonstrated improvements in mortality rates, ocular salvage, and increased autonomy throughout the program, as well as successful decision-making with the implementation of the telemedicine strategy30.
Another example of an international partnership was between the Children’s Cancer Institute and St. Jude Children’s Research Hospital, as described in the article by Al Haddad et al8. This partnership aimed to address the needs of children with RB in Lebanon. The study observed an annual increase in admissions of patients with RB; however, there was a higher number of diagnoses of patients with locally advanced tumors (stages D and E) and lower rates of ocular salvage for those with early-stage disease (A, B, and C). Most cases of early-stage disease were diagnosed in bilateral RB, with the second eye being more advanced8.
The My Child Matters program, which supports projects in LMICs, including the African Retinoblastoma Network also demonstrated promising results. The program covers 55 projects in 42 countries and has initiatives related to awareness of early diagnosis and training of professionals in disease management. The program resulted in a reduction of extraocular disease, increased survival, and ocular preservation15. The Mali Retinoblastoma Program pioneered the strategies developed, had substantial results, and was extended to the Democratic Republic of Congo, Senegal, Côte d’Ivoire, and Madagascar15,31.
Finally, it is crucial to highlight a significant 20-year cooperation between Garrahan Hospital in Argentina with Memorial Sloan Kettering Cancer Center in New York, which serves as an exemplary model to be followed. Throughout the stages of this process, it became evident that the primary focus was on empowering the country to develop local capabilities to deliver high-quality patient care. The program, initiated in 1995, initially concentrated on medical education through oncologist visits to New York, alongside the development of a protocol aimed at enhancing ocular preservation and reducing late effects. Subsequently, with financial support, the program facilitated the research and treatment protocol for RB in Latin America, leading to the incorporation of the first center in the region capable of performing intra-arterial chemotherapy. One notable impact of the program has been the significant improvement in the survival and ocular preservation of patients afflicted with the disease, resulting in a three-year probability of disease-free survival, which rose from 0.84 before program implementation to 0.9714,32,33.

3.4. Education

In the study conducted by Elfalah et al.25educating medical students about RB resulted in significantly improved knowledge about the disease, particularly in terms of its severity and diagnosis, following the intervention25.
Another example of medical education was developed by Hill et al.26 in Kenya. They designed a comprehensive and interactive workshop on RB genetics, followed by discussions of patient case studies. The study revealed that participants felt more confident in discussing genetics with patients and improving genetic counseling as a result of this training26. This strategy is of great importance in the country, as there is evidence indicating that the second and third children with RB are often diagnosed at an advanced stage of the disease, highlighting a failure in genetic counseling. Therefore, this tool shows promise in expanding the knowledge of healthcare professionals regarding the disease. However, knowledge assessments conducted one year after the workshop showed low scores, suggesting the need for frequent reinforcement of knowledge.
Concerning familiar awareness about the disease, Soliman et al.34 and Nawaiseh et al18demonstrated that increased knowledge of the disease among parents who had a first affected child led to improved care for subsequent children.
In Honduras, the integration of an RB education program into a nationwide vaccination campaign resulted in a reduction in the occurrence of extraocular disease27. Similar findings were observed in the study of Antonelli et al.24 in Brazil, where campaigns involved publishing articles in non-medical magazines, placing billboards at strategic locations in the city, providing guidance to medical professionals through courses offered at medical schools, and publishing findings in a medical journal to raise awareness of the disease, resulted in a reduction of referral time and, consequently, an increase in the number of detected intraocular tumors.

3.5. Genetic counseling

Joseph et al., 2004, compared the costs of genetic testing and a clinical strategy based on familiar risk assessment. Significant cost savings of 3.5 times for the proband and 6.1 times for the family were observed9.
Thirumalairaj et al.23observed in their study that the genetic analysis of RB is hindered in developing countries due to the extensive size of the gene and the presence of multiple dispersed exons. This results in time-consuming and costly procedures, as demonstrated by the comprehensive DNA sequencing model developed by Joseph et al. in 200613, which leads to an average delay of 84 days in obtaining results and subsequent management, particularly in India. Consequently, Thirumalairaj’s study emphasizes the need for alternative approaches that can offer faster and more efficient outcomes. The strategy they developed is based on prioritizing analysis based on the local frequencies of mutations reported in the literature and genetic databases23.

3.6. Other studies and initiatives

Despite not meeting our inclusion criteria (see Appendix II), we came across other relevant studies that warrant mention. One, the ’One World, One Vision’ symposium in January 2007, presented a global perspective on RB assistance disparities, showcasing successful case studies. The symposium catalyzed advancements in the field and aimed to explore country-specific RB programs 35. The ’Grupo Mexicano de Retinoblastoma’ (RtbMex), established in January 2003, educated health professionals, provided national RB guidelines, and conducted early diagnosis campaigns 36. The CureAll framework, supported by the WHO Global Initiative for Childhood Cancer, aims to improve children’s cancer outcomes globally, prioritizing childhood cancer through regional, national, and global action plans. The initial focus is on six main childhood cancer types, including RB, to assess care landscapes and identify areas for improvement, with the goal of achieving a minimum 60% survival rate worldwide and reducing overall suffering 37.

3.7. Assessment of the risk of bias of the included studies.

Qaddoumi, 200829, the only cohort study, was considered as of moderate quality. The most significant issue identified was related to the identification and strategies for addressing confounding factors. Participants were not clearly free from the outcome at the beginning of the study, and it was unclear whether strategies were employed to manage follow-up losses. Among the twelve cross-sectional studies evaluated, most demonstrated moderate to high methodological quality3,9,17,20,21,23,25,27,28,30. The primary issues identified pertained to insufficient detail in the study sample and a failure to identify strategies for managing confounding factors. Four case series studies were assessed, with the majority exhibiting high methodological quality (3 out of 4 studies)18,19,22. The main problem identified was a lack of clarity regarding the consecutive inclusion of participants (See Appendix III).

DISCUSSION

In the current systematic review, we aimed to find and evaluate strategies to improve access to diagnosis and treatment in LMICs. Retinoblastoma is a tumor which diagnosis and treatment require a multidisciplinary team that includes an ophthalmologist, oncologist, pathologist, and specialized geneticist38,39. Nationally coordinated strategies are important to build this integration and a chain of care. Public policies need to be guided by experts’ knowledge and require governmental funds and when possible, it is important to establish cooperation with high-income countries.
International partnerships can bring the quality of care from high-income countries closer to children around the world8,30,36,40,41. This has been achieved through cooperative programs known as twinning, which involve partnerships between a developing country and a mentoring institution from a developed country. This collaboration has become possible due to the advancement of communication technology on a global scale.
Telemedicine serves as an alternative for extending services to regions lacking specialists by integrating virtual education, mentoring, and case discussions. It also provides an opportunity for professionals to enhance their skills and receive training in their home country.
Concerning strategies of diagnosis, typically, an ophthalmologist can diagnose RB using indirect ophthalmoscopy with a dilated pupil38. However, in resource-limited areas, accessing an ophthalmoscope can be challenging, along with other restrictions such as the need for sedation and technical expertise for diagnosis. Beyond screening, smartphones are being tested for fundus examination. Since they are more accessible, portable, and compatible with teleconsultation, they can be a cost-effective strategy compared to traditional ophthalmoscopy. Promising results have been reported in the current literature, as also reviewed elsewhere42. It is essential to emphasize that healthcare professionals using these devices should receive proper instructions for use and interpretation. Furthermore, expanding the use of the applications for screening to family members or individuals without medical training can increase the chances of early diagnosis, but at the same time, may increase testing anxiety and false positive rates. New studies are needed to evaluate this possibility before implementation.
In our review, we did not find studies on the screening benefits in the general population of LMICs, possibly due to the rarity of the disease and the scarcity of ophthalmology services. Additionally, there is a lack of secure access to anesthesia and genetic testing in these regions43. However, it is important to emphasize the relevance of routine ocular health evaluation early in childhood for the diagnosis of other conditions, which can also provide an opportunity for the diagnosis of RB 44.
There are several ways to enhance the diagnosis and access to treatment for RB through education, whether by guiding the general population or individuals at higher risk regarding the signs and symptoms of the disease or by offering guidance to healthcare professionals. Medical education interventions play a fundamental role because, in numerous countries, newly trained doctors play a pivotal role in healthcare facilities, and the general practitioner will likely be the first healthcare professional to evaluate a child with suspected RB45. For this reason, a clinician’s ability to identify RB when first assessing a child and to make a timely referral needs to be trained24. Thus, it is important to improve teaching strategies, adding information about the disease and its warning signs to the medical curriculum. There is evidence that frequent reinforcement of knowledge is necessary, since important concepts may be lost over time. Continued medical education can be achieved by promoting periodic educational workshops26.
Furthermore, increasing family awareness about the disease also plays a fundamental role. By being well-informed about the disease and its potential risks, parents can take proactive measures to ensure the well-being of their children and seek appropriate medical attention at the earliest possible stage18.
Studies that identified education strategies for the general population’s awareness of the disease have yielded positive results in reducing extraocular disease upon diagnosis24,27, indicating that this approach has the potential to be implemented in countries where late diagnosis is prevalent.
In patients at higher risk of the disease due to positive family history, there are two ways to conduct screening: through genetic counseling and by using risk models calculated based on clinical assessment. Genetic counseling is important for the control and management of RB, identifying individuals at increased risk, and guiding screening and education for healthy lifestyles. RB is primarily caused by biallelic inactivation of the RB1 gene in a precursor retinal cell, followed by progressive mutations in other specific genes. Only a small number of relatives will carry the mutation. However, every child born to a parent with a germline RB1 mutation has a 50% chance of inheriting the mutation22.
In most LMICs, widespread access to genetic testing for RB is limited due to the absence of advanced technology capable of detecting RB1 mutations. Consequently, researchers have suggested calculated risk models to identify patients at a high risk of RB1 mutations, who could benefit from targeted screening programs19,22. However, findings from this review indicated that genetic testing incurred lower costs when compared to calculated risk models. This can be explained, because the latter approach often led to unnecessary and expensive repetitive ophthalmological examinations. Those unnecessary exams also cause heightened concerns among family members, and increase the risk of complications related to anesthesia9.
Therefore, implementing a genetic screening model would not only enable more accurate and cost-effective care for patients with a higher susceptibility to cancer but also aid family decision-making regarding future pregnancies.

4.1. Improvement opportunities

Despite facing numerous challenges, the initiatives identified in this review highlight important acts that may improve the time to diagnosis and access to treatment. By implementing these acts collectively, we have the potential to not only preserve the lives of children but also enhance their overall quality of life by avoiding treatment late effects; and preserving functionality and facial aesthetics. Drawing upon the extracted data, we propose a comprehensive chain of initiatives (Fig. 2) that we believe can yield improved clinical outcomes for patients with RB.